Você está na página 1de 788



By the same author

Electrical Machinery
. ..
Generalized Theory of Electrical Machines



2-B, Natb Market, Nai Sarak


Dr. P. S. Binibhra '


Ph. D., M.E. (Hons.), F.l.E. (India), M.I.S. T.E.

Ex-Dean, Ex-Prof. and Head of Electrical & Electronics Engg. Dept.

Thapar Institute of Engineering and Techn ology



2-B Nath Market, Nai Sal'ak

Phone s : 2391 23 80 ; 2722 41 79 Fa;{



C.4 Conn aught Pl ace, N. Oel hi1


1''1 011.23.116 01 / 02 / 03 . 011863 7232

Website: www.jdln bookdep ot.co m

E.mall: ~Jle5@j a inbookdep ot. c o m

Published by " Romesh Chander Khanna


2-B, Nath Market, Nai Sarak,


I ISBN No. : 81-7409-215-3

All RWht, Reserved

[This book or part thereof cannot be translated or reproduced in any
form (except for review or criticism) without the 'written .permission
of the Authors and the Publishers]


.. '

Fourth Edition: 20Q6

Reprint : 2007. .
'. ~

. ~

. .....

Price,' Rs. 225.00

Text Composition by " ,

Soft Serve Computers
Krishna Nagar, Delhi-51

Printed at "
Moh an Lal Prfntars
Shahdara, Delh i-32



; '


Loving Memory . .
. of M y Late Parents


Power electronics blend s the three m ajor areas of E lectr ical En gin eering-power,
electronics and control. Under controlled power conciitions, loa d performs better. So ther e has
always been a popular demand to have power modula tors. It is power 'electronics th at has
made possible the availability of a wide variety of controlled power converters. Power
electronics has really r evlut;"(oniLed the art of power conversion and its control. The advent
of power semiconduct or device, 't hyristor', in 1957 has been the most exciting breakthrough,
because its launch gave a boost to the art of power conversion and its control, and t ook this
art to its fore-front. As a result of technological evolution, many more semiconductor devices
such as triacs, asymm etrical thyristors, gate turn off thyristors, power MOSFETs, insulated
gate bipolar transistors, SITs, SITHs and MOS-controlled thyristors are n ow available. The
use of these semicon ductor devices has perva ded the industrial applications r elating to the
field of Electrical, Electr onics, Instrumentation and Control Engineering. In other words ,
power-electronic components find their use in low as well as high-power applications .
The purpose of this book is to provide a good understanding of the power-electronic
components and t he behaviour of power-electronic converters by presenting systematically
all important aspects of semiconduct or devices amI the common type of electric-power
controller s. The book begins with the study of salient features of power diodes, power
transistors, MOS-controlled thyristor, silicon controlled rectifier an d other memb ers of
thyr ' stor f am ily. Th en their applications in the differ ent t yp es of power-conv ert er
configurations are presen t ed in a lucid det ail. In other w ords , this book fo llow s the
bottom-down approach (device characteristics first and t hen their applications). Maj or part
of the book is intended to serve as an intr odu ctory course in' pow er~el ectronics t o t he
undergradu at e students of Electrical, Electronics, Instrumen tation and Control disciplines.
It is presumed that the reader is familiar with the basics of elementary electronics and circuit
theory. The material presented here can be covered in one sem est e with the amiss ' on of some
topics . The instructor, after browsing through the b ook for some tim e, can plan t he course
contents and its sequence without loss of continuity.
The book contains thirteen chapt ers . Chapter 1 gives an overview of merits and demerits
of power -electronic controllers and br' efly discusses the t opics covered in this book . This
chapter also touches upon the significance of power electr onics. Chapter 2 des cribes th e
characteristics of power diodes, powe transistors and MOS-con tr olled thyristors. In C apter
3 are pr esented diode characteristics, rectifiers, performance par amet ers and fil ers. Chapter
4 explains the characteri tics of thyristors in detail and of Triacs, GTOs etc. Thyristor
commuta j on t echniques are given in Chapter 5. In C~pt er 6, the prin ciples of conve sian
from ac to de involving singl e-phase as well as three-phase converters are presented. Chapt ers
7 to 10 pertain to the treatment of de chopper s, inverters, ac voltage controllers an d
cycloconverter s res pectively. While Chapter 11 gives study of several applications of power
electronics, Chapter 12 discusses electric drives. P ower facto improvement an th e m eth ods
o reactive power compensation are detailed in Chapte 13. A large number of ilJu trative
diagrams and a wide variety of worked example add t o the clarity of th e subject matter. The
material given in this boo is class-room tested. In the appendices ; Fo rier analysis, Laplace
transform and a large numb er of objective-type qu es tions r elating to C ap te s 2 t o 13 are


The material added in the present edition includes:

(i) performance parameters of uncontrolled and controlled rectifiers and filters,
(ii) structural modifications in thyristors and GTOs to make than more efficient,
(iii) SIT and SITH and (iv) Chapter 13 on power factor improvement.
Some topics have been r~-written to make the presentation more lucid. Many more
illustrative examples to reinforce the understanding of the subject matter are also included.
Objective-type questions are thoroughly updated. It is hoped that the book in its present form
will serve the purpose for the courses on power electronics of all Indian as well as foreign
The author is grateful to all those students who had interacted with the author, in the
c1ass~room or outside, during the teaching of this subject. This interaction has greatly
influenced the author's style of teaching and writing to a large extent and every effort has
,gone into making the subject matter presentation as easily comprehensible as possible .
. Discussion with several instructors has also been of immense help and inspiration. The author
is beholden to all of them. The author, however, regrets he cannot name them all, for it is a
voluminous task.
Finally, the author expresses his gratitude to his wife for her perennial encouragement,
understanding and patience during the preparation of this book. The author would also thank
his two sons a.'ld daughter-in-law for consistently boosting the author's morale, much needed
during the revision of the book.
Suggestions leading to the improvement of the book will be gratefully acknowledged.

Dr. P.S. Bimbhra



....... ..



1.2 .

Concept of Power Elect r onics . . . . ' .' . .

Applications of Power Electronics . . .. . .
Advantages and Disadvantages of Power-electronic Converters.
Power Electronic Systems . . . . . . .
Power Semiconductor Devices . . . . .
1.6. Types of Power Electronic Converters
1. 7. Power Electronic Modules . . . . . . .




2. 5.



2.8 .


2.1.1. Depletion Layer .. .. .

Basic Structure of Power Diodes
Characteristics of Power Diodes
2.3.1. Diode]- V Characteristics .
2.3.2. Diode Reverse Recovery Characteristics
Types of Power Diodes . . . . .
2.4.1. General-purpose Diodes
2.4.2. Fast-recovery Diodes
2.4.3. Schottky Diodes . . . . .
Power Transistors . . . .. .. .
2.5.1. Bipolar J unction Transistors . Steady-state Characteristics. BJT Switching Performance. Safe OperatingArea . .
Power MOSFETS . . . . .. . . .
2.6. 1. Pmosfet Characteristics . . . . .
2.6.2. Pmosfet Applications . . . . . .
2.6.3. Co~parison of PMOSFET With BJT
Insulated Gate Bipolar Transistor (IGBT)
2.7.1. Basic Structure . .
2.7.2. Equivalent Circuit .
2.7.3. Working . . . . . .
2.7.4. Lat.ch-up in IGBT .
2.7.5. IGBT Characteristics
2.7.6. Switching Characteristics
2.7.7. Application oflGBTh . . .
2.7.8. Comparison oflGBT With MOSFET
Static Induct ion Transistor (SIT)
MOS-controlled Thyristor (MCT)
New Semiconducting Materials.



3. 1. Diode Circuits With de Source

3.1.1. B..esi:Jti"ve LoA
3.1.2. RC Load .


2.1. The p-n Junction . . . .. . . . .

2.3 .

-0 '






14 .





.... .







. 31



. 33


. 34




. . 36

. 37

. .39



. 44


3. 3.



3. 10.


3.1.3. RI Load ..

3.1.4. LC Load .

3.1.5. RIC Load.

Freewheeling Diodes

Diode and L Circuit

Recovery of Trapped Energy.

Single-phase Diode Rectifiers

3.5.1. Single-phase' Half~wave Rectifier
Zener Diodes .. . . . . . . . . . . .. .

Performance Parameters . . ~ .. . . . .

3.7.1. Input Performance Parameters . ~

. 3.7 ~2. Output performance Parameters .

Comparis()nof Single-phase Diode Rectifiers

3.8.1. Single-phase Half-wave Rectifier ..

3.8.2. Single-phase Full-wave Mid-point Rectifier.

3.8.3. Single-phase Full-wave Bridge Rectifier .

Three-phase Rectifiers . . . . . . . . . . . . . . . . : .

3.9.1. Three-phase Half-wave Diode Rectifier .. , .

3.9.2. Three-phase Mid-point 6-pulse Diode Rectifier ' ..

3.9.3. Multiphase Diode Rectifier . . . . . . '.. .

3.9.4. Evolution of Three-phase Bridge Rectifier

3.9.5; Three-phase Bridge Rectifier .. ' . .

3.9.6 .. Three-phase Tw,"lve-pulse Rectifier

3.10.1. Capacitor Filter (c-filter)
3.10.2. Inductor Filter O-filter)
3.10.3. I-C Filter . . . . . . . . .

THYRI STORS . . . . .
4.1. Terminal Characteristics of Thyristnrs . . . . .
4.1.1. Static I-V Characteristics of a Thyristor .
", . .
4.2. Thyristor Turn-on Methods .. . . . . . . . . . . .
4. 3. Switching Characteristics of Thyristors . . . . . . . .
4.3 .1. ' Switching Characteristics During Turn-on . .
4.3.2. Switching Characteristics During Turn-off ..
4.4. Thyristor Gate Characteristics . . .
4.5. Two-transistor Model of a Thyristor
4.6. Thyristor Ratings . . . . . . . .
4.6 .1. Anode Voltage Ratings
4.6.2. Current Ratings . . ..
4. 7. Thyristor Protection . . . . . .
4.7.1. Design of Snubber Circuits .
4.7.2. Overvoltag Protection . . . Suppression of Ovet"Voltages.
4.7.3. Overcurrent Protection .. . .. . ,
4.7.4. Gate Protection . . . . . . . . . .
4.8, Improvement of Thyristor Characteristics
4.8.1. Improvemetlt in dil dt Rating . .
4. 8.1.1. Higher-gate Curr~nt .. .
4. 8.1.2. StructuTal 'Modification of The Device..

I '

~ ..

. 46
' .75
. 77
~ 89
. 97


. . ... .
.~ .

























4.8.2. Improvement in dvl dt Rating .

4.9. Heating, Cooling and Mounting of Thyristors
4.9 .1. Thermal Resistance . . . . . . . .
4.9.2. Heat Sink Specifications . . . . . .
4.9.3. Thyristor Mounting Techniques ..
4.10. Series and Parallel Operation of Thyristors
4.10.1. Series Operation . . . . . . . . .
4.10.2. Parallel Operation . . . . . . . . ~ . ' "
4.11. Other Members of the Thyristor Family . . .
4.11.1. PUT'. (Programmable Unijunction Transistor)
4.11.2. SUS (Silicon Unilateral Switch)
4.11.3. SCS (Silicon Controlled Switch) . . . . .
4.11.4 . Light Activated Thyristors . . . . . . . .
4.11.5. . The Diac (Bidirectional Thyristor Diode)
4.11.6. The Triac .. . . . . . . . . . . . . . .
4.11.7. Asymmetrical Thyristor (ASCR) .. .
4.11.8. Reverse Conducting Thyristor (RCT)
. 4.11.9. Other Thyristor Devices
4.12. Gate Turn off Thyristor (GTO) .. .. . .
4.12.1. Basic Structure . . . . . . Turn-on Process . .
4.12. 2. Static J- V.Characteristics .
. 4.12 .3. Switching Performance ' . Gate Turn-on. . .
4.12.3..2. Gate Turn-off.. .
4.12.4. Comparison Between GTO and Thyristor .
4.12.5. Application of GTOs .. ..
4; 13. Static Induction Thyristor (SITH) . .. .
4.13.1. Basic Structure . . . ; . . . . .
4.13.2. Turn-on and Turn-off Processes
4.13.3. Application of SITH and Comparison with GTO .
4.14. Firing Circuits for Thyristors . . . . . . . . . . .' . . . . .
4.14 .1. Main Features of Firing Circuits . . . . . . . . .
. 4.14.2. Resistance and Resistance-capacitance Firing Circuits
4.14.3.. Unijunction Transistor (UJT)
4.15. PUlse Transformer in Firing Circuits. . . . .
4.16. Triac Firing Circuit . . . . . . . . . . . . . . .
4.17. Gating Circuits for Single-Phase Converters
4. 17. 1. Gate Pulse Amplifiers .
4.17.2. Pulse Train Ga~ing
4.18. Cosine Firing Scheme . . . . .



5. 1. Class A Commutation: Load Commutation
; . .. '.

5.2. Class B Commutation : Resonant-pulse Commutation .

5.3. Class C Commutati on : Complemen~ Commutation .

5.4. Class D Commu tation : Impulse COlIllllutatio

5.5. Class E Commutation : External Pulse Commutat;on
5.6. Cl838 F Commutation: Line Commutation . .. . . .














































(x i i)






6.1. Principle of Phase Control. . . . .

. . . . . . . .
. '.

6.1.1. Single-phase Half-wave Circuit with RL Load


6.1.2. Single-phase Half-wave Cir cuit with RL Load and Freewheeling Diode

6.1.3. Single-phase Half-wave Circuit with RLE Load

6.2. Full-wave Controlled Converters . . . . . . . . .' . . . . . . . . . . . . .


6.3. Single-phase Full-wave Converters . . . . . . . . . . . . . . . . . . . . .


6.3.1. Single-phase Full-wave Mid-point Converter (M-2 Connection)


6.3.2. Single-phase Full-wave Bridge Converters . . . . . .. .

265 Single-phase Full Converter (B-2 Connection) .

268 Single-phase Semiconverter. . . . . . . . . . . Analysis of Twc-pulse Bridge Converter with


Continuous Conduction. . . . . . . . . . . . .

6.4. Single-phase '!\vo-pulse Converters with Discontinuous Load .


6.4.1. Single-phase Full Converter with Disc'ontinuous Current .


6.4.2. Single-phase Semi converter with Discontinuous Current


6.5 . Performance Pa rameters of Two-pulse Converters


6.5.1. Single-phase Full Converter . . . . . . . . . . . . . .


6.5.2. Single-phase Semiconverter . . . . . . . . . . . . . .


6.6. Single-phase Symmet rical and Asymmetrical Semiconverters


6.6.1. Single-phase Symmetrical Semiconverter .


6.6.2. Single-phase Asymmetrical Semiconverter . .


6.7. Three-phase Thyristor Converter s . . .. . . . . . . .


6.7.1. Three-phase Half-w ave Cont rolled Converter


6.7.1. 1. Three-phase 1'1-3 Co verter with R Load.

288 Three-phase M-3 Converter with RL Load.


6.7.2. Three-phase Full Converter s


6.7. 3. Three-phase Semiconvarlers . . . . :: -. . .


6.7. 4. Multi-pulse Controlled Converters . . . . . .


6.8. Performance P arameters of 3-phase Full Converters


6.9. Effect of Source Impedance on the Performance of Converters


6.9.1. Single-phase Full Converter . . . .


6.9.2. Three-phase Full Converter Bridge


6.10. Dual Converters . . . . . . . . . .


6.10.1. Ideal Dual Converter .. .


6.1 0.2. Practical Dual Converter


6.11. Some Worked E amples . . . . . .

7.1. P ':ciple of Chopper Operation .
7.2. Control Stra tegies . . . . . . . .
7.2.1. Time Ra tio Control (trc).
7.3. Step-up Chopper s . . . . . . . . .

7.4. Typ~ of Chopper Circuits . . . .

7.4. 1. Firs t.quadrant, 01' TyJls-A, Chopper .
7.4.2. Second-qu adran t, or Type-B, Chopper .
7.4.3 . Two-quadrant Type-A Chopper, 01' 'li.rpe-C Ch oppe
1 .4.4 . Two-quadrant Type-B Chopper, or Type-D Chop er
7 .4.0. Four-quadrant Chopper, or Type-E Chopper . . . .


347-41. 3










7.5. Steady State Time-domain Analysi of Type-A Chopper

7.5.1. Steady State Ripple . . . . . . . . .
7.5.2. Limit of Continu ous Conduction . .
7.5.3. Computation of Extinction Time tx
7.5.4. Fourier Analysis of Output Voltage
7.6. Thyristor Chopper Circuits . .. .. .
7.6.1. Voltage-commutated Chopper
7.6.2. Current-commutated Chopper
7.6 .3. Load-commutated Chopper.
7.7. Multiphase Choppers . . .. .. .. .


INVERTERS . . . . . . . . . . . . . . .

8.1. Single-phase Voltage Source Inverters: Operating Principle

8.1.1. Single-phase Bridge Inverters . . . . . . . . . . .
8.1.2. Steady-state Analysis of Sin gle-phase Inverter '. .
8.2 . Fourier Analysis of Single-phase Inverter Output Voltage
8.3. Force-commutated Thyristor Inverters . . . . . . .
8.3.1. Modified Mcmurray Half-bri dge Inverter .
8.3.2 . Modified Mcmurray F ull-bridge Inverter '.
8.3.3. McMurray-Bedford Half-bridge Inverter
8.3.4. McMurray-Bedford Full-bridge Inverter
8.4. Three Phase Bridge Inverters . . . . . . . .
8.4.1. Three-phase 180 Degree Mode VSI
8.4.2. Three-phase 120 Degree Mode VSI
8.5, Voltage Control i Single-phase Inverter
8.5.1. Ext ernal Control of ac Output Voltage
8.5 .2. External Control of dc Input Voltage
8. 5.3. Internal Control of Inverter
! .
8.6. Pulse-widt h Modulated Inverters .
8.S.1. Single-pulse Modulation ..
8.6.2. Multiple-pulse Modulation .
8.S. Sinusoidal-pulse Modulation (SPWNl) .
8.6.4. Realization of PWM in Single-phase Bridge Inverters.
8.7. Reduction of Harmonics in the Inverter Output Voltage . .
8.7. 1. Harmonic Reduction by PWM . , .. . .. , . , ..
8.7}. Harmonic Reduction by Transformer Connections.
&7 .3. Harmonic Reduction by Stepped-wave Inverters
8.8, Current Source Inverters . . . . . . . .. . , . . . . . . . .
8.8.1. Single-phase CSI with Ideal Switches ... . , .. ,
8.8.2. Single-phase Capacitor-commutated CSI Wit h R Load
8.8.3. Single-phase Auto-sequential Commutated Inverter (I-phas e ASCl) .
8.9. Series Invert;1 . , . . .. .. . . . . . .
8.9. . Basic Series Inverter . . . . . .

8 ,9.2. Analysis of Basic Series lnverte

8. 9.3. Modified Series Inverter . ,

B.9 . . Half-Bridge Sa 'as In 1erter
8.10. Single~Phase Parallal Inverte
8.10.1. Analysi3 of Panillel Inverter.
B.ll . Good Inyerter . . . . . . , . . .





















































9.1. Principle of Phase Con trol . . . .
9.2 . Principle of Integral Cycle Control . . . .
9.3. Single-phase Voltage Controllers . . . . .
9.3.1. Siilgle-Phase Voltage Controller with R Load
9.3.2. Single-Phase Voltage Controller with RL Load.
9.4. Sequence Control of ac Voltage Controllers . . . . .. . .
. 9.4.1. Two-stage Sequence Control of Voltage Controllers .
9.4 .2. Multistage Sequence Control of Voltage Controllers .
9.4.3. Single-phase SinWloidal Vol~age Controller . . . . . .


CYCLOCONVERTERS . . . . . . , . ~ . ,; . . . . . . . ; ..
10.1. Principle of Cycloconverter OperatioI:l' . .. . . . . . . . . . .
10:1.1. Single-phase to Single-phase. Cireuit-'step-up Cycloconverter Mid-point Cycloeonverter.. . . . . . . . . . . . . . . . Bridge-type Cycloconverter. . . . . . . . . . . . . . .

10.1.2. Single-phase to Single-phase Circuit-step-downCycloconvertel' Mid-point Cycloconverter.. Bridge-type Cycloconverter. . . .. .
10.2. Three-phase Half-wave Cycloconverters . . .; . . . . .
10.2.1. Three-phase to Single-phase Cycloconverters .
10.2.2. Three-phase to Three-phase cycloconverters
10.3. Output Voltage Equation for a Cycloconverter
10.4. Load-eommutated Cycloconverter. . . . . . . . . .' .

533 .

SOME APPLICATIONS . . . . .' . . . . . .

11.1. Switched Mode Power Supply (SMP S)
11. 1. 1. Flyback Converter . . .
11.1.2. Push-pull Converter ..
11.1.3. Half-bridge Converter .
11.1.4. Full-bridge Converter.
11.2. Uninterruptible Power SUpplies
11.3. High Voltage dc Transmission.
11.3.1. Types of HYDC Link . .
11.3.2. Bipolar HYDC System .
11.3.3. Control of HVDC Converters.
11.4. Static Switches . . . . . . . . . .
11.4.1. Si ngle~phase ac Switehes .
11.4.2. de Switches .. . . . . . .
11.4.3. Desigri of S ta tic Switchei! .
11.5. Static Circuit Breakers . .. . . .
11.5.1. Static ae Circuit Breaker s
. 11.5.2. Static de Cir cuit Breakers
11 .6. Solid State Relays . . .. . . .
11.6.1. DC Solid Sta te Relays
11.6.2 . AC Solid State Relays
11.7. Reson n t Convertars . . . .. .
11.7. 1. Zero-cu ent Switching Resonant Converters . L-type 2 CS Resonan ~ Converters. .

: 561
588 .

523 .


,' .

. 572


(xv )

581 M-type ZCS Resonant Converter.

11 .7.2. Zero-voltage-switching Resonant Converters ..
11.7.3 . Comparison Between ZCS and ZVS Converters



12.1. Concept of Electric Drive
12.2. DC Drives . . . . . . . . .
12.2. 1. Basic Perfonnance Equations of dc Motors
12.3. Single-phase dc Drives . . . . . . . . . . . . . . . .
12.3.1. Single-phase Half-wave Converter Drives
12.3.2. Single-phase Semiconverter Drives .
12.3.3. Single-phase Full Converter Drives .
12.3.4. Single-phase Dual Converter Drives
12.4. Three-phase dc Drives . . . . . . . . . . . . .
12.4.1. Three-ph~.se Ha1f-wave Converter Drives.
12.4.2. Three-phase Semiconverter Drives .
12.4.3. Three-phase Full-converter Drives .
12.4.4. Three-phase Dual Converter Drives.
12.5. Chopper Drives . . ' . . . . . . . . . . . . . .
12.5 .1. Power Control or Motoring Control
12.5.2. Regenerative-braking Control .
12.5.3. Two-quadrant Chopper Drives .
12.5.4. Four-quadrant Chopper Drives
12 .6. A.C. Drives . . . . . . . . . . . . . . . .
12.7. Inductio .-Motor Drives . . . . . . . . .
12.8. Speed Control of Three-phase Induction Motors
12.8.1. Stator Voltage Control . . . . . . . . .
12.8.2 . Stator Frequency Control .. . . . . . .
12.8.3 . Sta tor Voltage and Frequency Control
12.8.4. Stator Current Control . . . . .
12.8.5. Static Rotorresistance Control
12.8.S. Sl'p-power Recovery Schemes .
12.8.S.1. Static Kramer Drive.. Static Scherbius Drive.
12.9. Synchronous Motor Drives . . ..
12.9.1. Cylindrical Rotor Motors
12.9.2. Sall nt-pole Motors . . .
12.9.3. Reluctance Motors. . . .
12.9.4. Permanent-Magnet Motors .
12.10. Some Wor ed Examples . . . . .




13.1 . Effect of Poor Power-factor .. . . . . . .

13.2. Methods of Reactive Power Compensatio
13 .2.1. Capacitor Banks . . . . . . . . . .
13.2.2. Synchr6nous Condensers . . . . .
13.2.. Thyristor Controlled Reactors ('1'CRS)
13.3 . Static VAx CompensaUlr (SVc)
13.4. Som Worked Examples .. . . . . . . . . .. .

5 83~674




61 9





. .. , . .. 691698

Appendix: A - FOURIER ANALYSIS . . .

. . . . . 699



Power Semiconductor Diodes and Transistors .

Diode Circuits and Rectifiers . . . .

Thyristors . . . . . . . . . . . . . . .

Thyristor Commutation Techniques

Phase Controlled Rectifiers

Choppers . . . . . . . .

Inverters . . . . . . . . .

AC Voltage Controllers.

Cyc1oconverters ..

Some Applications .. .

E lectric Drives . . . . .

Power Factor Improvement















" ,


Chapter 1

Introd ctO
... -....... ..........

... ... . .. -....... .. .... ... .. ... ........


...... ..-... . . .... .. -.


In this Chapter
concept of Power Electronics
Applications of Power Electronics
Advantages and Disadvantages of Power-electronic Con verters
Power Electronic Systems
Power Semiconductor Devices

Types of Power Electro nic Converters

Power Electronic Modules


"' . . . . . . . . .

_ _ .. .

...... .

.. .

. . . . . . ... .

1<' .. . . . . . . . .

... .

. . . . . . .. . . . . . . . . _ _ _ .

-. . . . . . .. .

'. . ..... . . . . . . . .. .

.... .. .


The object of this chapter is to discuss briefly the concept of power electronics, applications
of power electronics and the types of power converter s desc:t ibed in this book.

Power electronics belon gs partly to power engineers and partly to electronics engineers [2] . Power
engineering is mainly concerned wi th gen eration, trans mission, eli tribution and utilization of electric
energy at high efficiency. Electronics engineering, on the oth er h and,. is guided by distortionless
production, transmission and reception of data and signals of very low power level, of th e order of a
few watts, or milliw atts, without m ch consideration to the efficiency.. ill addition, apparatus
associated with power engineering .s based mainly on electromagn etic principles wh ereas that in
electronics en gineering is based upon physical phenomen a in yaCUUID , gases/vapours and
semiconductors. .
Power electronics is a subject th at con cerns the appl" cation of electronic-priIlciples into situations
that are rated at power level rather than signal level. It may also be defined as a subject that deals
with the apparatus and equipmen t working on the principle of electronics but r ated at powe r level
r ather than signal level. F or e::Iample, semiconductor 'power swit ches such as thyri tors, GTOs etc.
work on the principle of electronics (movement of holes and electrons), but h ave the na.m e power
attached to them only as a description of their power ratings. Similarly, diodes, mercury-arc rectifiers
and thyratrons (gas-fill ed triode), high-pow er .level devices, fann a part of the s"Llbject power
electronics ; because their working is based on the physical phenomena in gases and vapours , an
electronic process. As the inclusion of all such power-rated electronic equipments would be a
voluminous task, the present book is devoted to the study of semi-condu ctor-based power-electronic
compon ents and systems only. It shocld be understood that the techniques used in the design of
high-efficiency and high-energy level pow electronic circuits are quite different from those employed
in the design of low-effici
ency electronic circuits at signal levels.

Th e era of m odern power electron ics began with th e inv.en tion of silicon - controlled-r ectifier
(S CR) by B ell Laboratories in 1956 . Its prototype was introduced by GEC in 1957 and
subsequent ly, GEC i n trod Llced SCR-ba ed system s commercially in 1958. Since then , there
have been eme:rgence of many new power semiconductor devices. Power el e ctron~:: sys .e . ~
today incorp ora te p ower semi :::onductor d 9vice ~ well as microelectronic inte.,.:3 ci ell cuit3 .

P owe Electronics

[Art. 1.2J

The term, 'converter system', in gener al, is used to denote a static device that converts ac to dc,
dc to ac, dc to dc or ac to ac. Conventional power controllers based on thyratr ons, mercury-arc
rectifiers, magnetic amplifiers, rheostatic controllers etc. have been replaced by power electronic
controllers using semiconductor devices in almost all applications. The development of new
power-semiconductor devices, new circuit topologies with their improved performance and their fall
in prices have opened up wide field fOT the new applications ofpower electronic converters. Ajudicious
use of power-semiconductor devices in conjunction with microprocessors or microcomputers has
further enhanced the control strategies and synthesizing capabilities of the power electronic
converters. It is said that power semiconductor devices can be regarded as the muscle and the
microelectronics as the intelligent brain in the modern power electronic systems.
For controlling the power flow to load, all power semiconductor devices, used in a power electronic
converter, are either fully-on or fully-off. In other words, all semiconductor devices in power-electronic
converter operate as switches. Wilen the switch is fully-on, power semiconductor device handles large
current (divided by the load impedance) and negligible voltage drop across it. When the switch is off,
the device handles negligible current vrith full- voltage across it. Therefore, a power semiconductor
device, during on and off periods, has very low power loss in it as compared to the power delivered
by the source to load. This results in higher energy efficiency of the power electronic converter system.
At the same time, low energy loss in the semiconductor device can easily be removed by its efficient
cooling. This all has contributed to the widespread use of power electronic converters in the power
conversion and control systems.
Table 1.1 lists various applications of power electronics. Tills list is however not exhaustive.
No boundaries can be earmarked for the applications of power electronics, especially with the
present trend of integrated design of power-semiconductor devices, microprocessors and the
controlled equipment. The power ratings of power-electronic systems Tfu.J.ge from a few watts
in lamps to several hundred megawatts in HVDC transmission systems. It is believed that in
the early twenty-first cent ury, 60 t o 80% of the electric power consumed in utility systems .vill
pass through power-electronics and this figure will event ually reach 100% in the future.
Tab le 1 .1. Some Applica tion3 of Power El ec tronics
1. Ae rosp a ce:

Space shutt le power supplies, satellit e power supplies, aircraft power systems.
2. Commercial :
Advertising, heating, air conditi oning, central refrigerlltion, comput er and office equipment,
uninterruptible power supplies, elevators, light dimmers and fl ashers .
3. Industrial:
Arc and industrial furn aces, blowers and fans, pumps and compr essors , industrial lasers,
transformer-tap changers, r olling mills, textile mills, excavat ors, cement mills, welding.
4. Residential :
Airconditioning, cooking, lighting, spa ce heating, refrigera tors , electric-dooT openers, dryers, fans,
personal co mputer s , oth er entert ainment equipmen t, v cu um cleaners , was hing an d sewi n g
machines, light dimmers, foo d mixers, electric blankets, food-vvarm ar trays.
5. Telecommuni ca t::i.on :
Battery charger s, power supplies ( dc an d UPS ).
6. Tra nsp ortation;
B ttery chargez:s, traction ont 01 of electri c vehicles , electri c locomoti ves) street cars, t rolley buses,
subways, automotive electronics.
7. Utility systems :
High voltage dc transmission (HVDC), exdtation systSID3, VAR compensation, stati c circuH
bTeak ers , fans and boiler-fe d pump.s, Buppl a-mentaJ'l] ene gy sys elD.3 ( solar, -;vind ).

In traduction

l---\ rt.



The a dvantages possessed by power-electronic system.:; are as und er :
(i) High efficiency du e t o low loss in power -sem icond ctor devices.
(ii) High reliability of power- electronic conver ter systems.
(iii) Long life and less maint enance du e to the abse e of any moving parts .
(iv) Fast dyn am ic respon s e of the p ow er-e lectr oni c systems as compared to
electromechanical converter systems.
(v) Small size and less weight r esult in less flo or space and therefore lower installati on
(vi) Mass p oduction of power -s emiconductor devices has resulted in lower cost of the
converter equipment.
Systems based on p ower electronics, however, suffer from the following disadvantages:
(a) Power-electronic convert er circuits have a tendency to generate harmonics in the supply
system as well as in the load cir cuit.
In the load circuit, the perform ance of the load is influenced, for exam.nle, a high harm on ic
content in the load circuit cau ses commutation problem s in de m achines , in creased m otor
heating and more accoustical noise in both dc and ac machines. So steps must be tak en to
filter these out from the output side of a converter.
In the supply system , the harmonics distort the voltage w avetorm and seriously influence
the performance of oth er equipm ent connected to the same supply line. In addition, the
harmonics in the su pply line can also caus e interference in a udio- and video-equipment (called
raclio interference ). It is , th erefore , n ec'essary to insert filter s on the inpl t sid e of a con verter.
(b) Ac to dc and ac to ac con verters operat e at a low in p ut power fa ctor under certain

operat ing conditions. In order to avoid a low pf, some special measures have t o be adopted.
(c) Power-el ectronic con trollers h av e low ov erload capa city, Th ese conv erters must ,
ther efore, be r ated fO T taking m omentary over loads . As such, cost of power electr onic controller
m ay mcrease.

(d) Regeneration of power is difficult in power electronic converter systems.

The advantages posses ed. by power electronic converters far outiiveigh their disadvantages
mentioned above. As a consequence, semiconductor-based converters are being extensively employed
in systems where power flow is to be regulated. As already stated, conventional p ower controllers
used in many inst allations have already been replaced by semiconductor-based power electronic

The m ajor compon en ts of a powe eledronic' syst em are shown in the form of a block
di gram in Fig. 1. 1. M ain power s ou ce may be an a c supply syst e _ or a de supply ys tem.
The outp t fr om the power electronic circui t may be variabb~ dc, 0 ac voltage, or it may be
a var iable voltage and fre quency. In gen er al, the ou tput of a pow er electronic conver tor ci cu 't
depends u pon the r equirements ofth e lo ad. F or example) iftb e load is a de m ot or, the converter
output must be adju stable dir ect voltage. In case the load is 3-vhase induction m otor, the
converter may ha v ~ adju st able voltage and fr equ ency at its outp ut terminals.
The feedb ack component in Fig. 1.1 me asures a param eter f he 10 ' d, say sp eed in case of
a r ota ting machin e, and comp ares it with th command . The diff:: ence of the two, throug the
digital circuit components , controls f e i stant oftu..rn-on of semicondu.:toT d evices forming the

Power Electronks

[Art. 1.5J
Ma in Power

Sou re e






Power Electronic



Feedback Signal

Fig. 1.1. Block diagram of a typical power electronic system.

solid-state power converter system. In this manner, behaviour of the load circuit can be
controlled, as desired, over a wide range with the adjustment of the command.


Silicon controlled rectifier (SCR) was first introduced in 1957 as a power semiconductor
device. Since then, several other power semiconductor devices have been developed. Must of
these semiconductor devices are listed below in Table 1.2 along with their circuit, or device,
symbol and present maximum ratings.
Table 1.2. Maximum ratings o{power sem iconductor devices
S .No.


Circuit symbol


Diod e








(a) SCR

(6) L'SCE









Upper operating
(req. (kHz)

5000 V/5000 A


7000 V/5000 A


6000 V/3000 A


2500 V/400 A


5000 V/3000 A




~ or~~~:

Voltage I current

[Art. 1.5J

In trod uc tion



Circuit symbol

Voltage I curren t


Upper operating
freq. (kHz)

(e) S1TH

2500 Vl500 A

100 .0

1200 V/40 A




1400 V/400 A


1000 Vl50A


1200 Vl300 A

100 .0

1200 Vl500 A


(j) MeT

M~f-2---itr ~1

(g) Triac















In the above table, the various abbreviations are; SeR (silicon controlled rectifier), LASeR
(light-activated SCR), ASCR (asymmetric al SeR), RCT (revers e conducting thyristor), GTO
(gate-tUIIl off thyristo~) , SITH (static induction thyrist or), MeT (MOS controlle thyTis or),
BJT (bipolar junction transisto ), MOSFET (metal-oxide semicon ductor fiel d effect transistor),
81 (static indudi on transistor) and IGBT (insulated gate bipolar tnms stor).

Power Electronics

[Ar t. 1.6]

Based on (i ) t urn-o n and turn-off characteristics, (ii) gate signal requirements and (iii)
degree of controllability, the power semiconductor devices can be classified as under:
(a ) Di od es . Thes e ar e uncontrolled rectifying devices. Their on and off states are controlled
by power supply.

(b) Thyri st ors. These have controlled turn-on by a gate signal. After thyristors are on,
they remain latched-in on-state due to internal regenerative action and gate loses control.
These can be turned-off by the power circuit.

Controllable swit ches. These devices are turned-on and turned-off by the application
of control signals. The devices which behave as controllable switches are BJT, MOSFET, GTO,

Triac and RCT possess bi-directional current capability whereas all other remaining
devices (diode, SCR, GTO, BJT, MOSFET, IGBT, SIT, SITH and MeT) are unidirectional
current devices.

1.6. TYPES



A power electronic system consists 0: one or more power electronic copverters. A power
electronic convert er is made up of some power semiconductor devices controlled by integrated
circuits. The switching characteristics of power semiconductor devices permit a power
electronic converter to shape the input power of one form to output power of some other form .
St atic pow er converters perform these functions of power conversion very efficiently. Broadly
speaking, power electronic converters (or circuits) can be classified into six types as under:
1. Di ode Rectifi ers. A diode rect ifier circuit converts ac input voltage into a fix ed dc
voltage. The input volt age may be single-phase or three phase . Diode rectifiers find wide use
in el ectric t rac tion , batt ery charging, electroplating, electrochemical pr ocessing) power
supplies, 'vveldin g an d uninterruptible power supply (UPS ) systems.

2. Ac to d c convert ers (Phase-controlled r ectifi ers). These convert constant ac voltage

t o vari able dc outp ut voltage . These rectifiers use lin e voltage for their commutation, as such
these are also called line-commutated or naturally-commutated ac to dc converters.
Phase~controlled converters may be fed from I-phase or 3-phase source. These are used in dc
drives, metallurgical and chemical industries, excitation systems for synchronous m achines
3. D C to d-c converter s (DC Ch oppers). A dc chopper converts fixed dc input voltage to
a controllable de outpu t voltage. The chopper circuits r equire forced, or load, commut ation to
tu rn-off the thyristors . F or lower power cir cuits, th yristors are replaced by power transistors.
Classifi at ion of chopper circuits is dependent upon the type of commutation and also on the
direction of power t1ow. Choppers find wide applications in dc driv es, subway cars , troll ey
t rucks, battery-dri ven vehicles etc.
4. DC to a ,c converters (in'Verte!'s). An inver t er converts fixed de voltage to a variable
voltage. The outpu' may be a variable voltag_ and variable freq eney. These converters use
lin v, load or forced commutation fo turning-off the thyristors. Inverters find wide use in
induct ion-m otoT and synchronous-mot or dTives . indu ction heating, UPS, HVD C trans mission
etc. At pr esent, conventional thyristors are als o being r eplaced by GTOs in high-power
applicati ons and by power transistors in w-pO\vei applic ations.

[Axt. 1.7]

Intro uction

5. AC t o ac convert ers. These convert fixed ac input voltage into va_iable ac output
voltage. These are of two types as under:
(a) AC' voltage controllers (AC voltage regulators). These converter circuits convert fL ed
ac volt age dir ectly to a variable ac voltage at the same fre quency. AC volt age controller employ
two thyristors in antiparallel or a triac. Turn-off of both the devices is obtained by ine
commutation . Output voltage is controlled by varying the firing angle delay. AC voltage
controllers are widely us ed for lighting control, speed control of fans , pumps et c.
(b ) Cyclocon verters. These cir cuits convert input power at one frequency to output power
at a differen t frequ ency through one-stage conversion. Line commutation is more common in
these converters, though forced and load commutated cycloconverters are also employed. These
ar e prim arily used for slow-speed large ac drives like rotary kiln etc .
6. Static switch s. The power semiconductor devices can operate as static switches or
contact ors . Static swit ches possess many advantages over mechanical and electromechanical
circuit breaker s. Depending upon the input supply, the static switches are called ac static
switches or dc static switches.




A power electroni c converter m ay require two, four or more semicon ductor devices
depending upon the circuit configuration. For example, a single-phase half-bridge inverter
requires a power module consisting of two power semiconductor devices; a full-converter (or
H -bridge converter) r equires a power module having four semiconductor devices; a three phase
full converter n eeds a power module having six semiconductor devices. Thus, a power electronic
converter can be ass embled from power modules instead of from individual semiconductor
devices. A power module h as better performance characteristics as compared t o conventional
devices so far as thei.r switching characteristics, operating speed and losses are concerned.
Gate drive circuits for individual devices or power modules are also commer cially available.
As a r esult of these developments, now inte ligent m odules h ave come in the mar 'et .
Inte lligent m odule, also caned smart-power; is state-of-the-art power electronics and it
consists of power module and a periph eral circuit. The peripheral circuit compris es of
interfacing of power module with the inputJoutput through pr oper isolation from low-voltage
signal and from high-voltage power circuit, a dri.ve circuit, protection and diagnostic circuitry
against maloperation like excess current, over volt age etc, microcomputer control and
controlled power supply. The user has merely to connect the existing supply and the load
terminals to the smart-power. At pr esent, int elligent modules are being u sed extensively in
power electr on ics. It is r eported that there are more than twenty manufacturers of int elligent

Po wer semiconductOi d evices form the heart o f modern power electronics, A p o wer
electronics engineer must understand the device thoroughly fo r e fficient, reHabl and
cost-effective d esign of power c onverters. For this p urpose, ch a pter 2 Is devoted 'to the
study of pO'-ll/e r semiconductor d iodes, 'transistors a nd MeT. C hapte r 3 deals with diode
c ircuits a nd recti ners . In ch a pter 4, are discussed in detail t he thyristor characteristics and
its control strat gles . Thyristor c ommut ation t echniques are describ ed In c hapter 5. O t' er
power elec tro nic converte rs mentioned in this c hapte are d escribed in d etail in c hapter
6 o nward.
L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ __ __ __ _ _ _ _ __ _ _ __ _ _ __ _ _ _ _ __ _ _ _ _ _ _ _

Power Ele ctronics

[Prob .5]

. .

Wha t is power electronics? Discuss briefly the concept of power electronics.

(6) What. is a converter? Illustrate yo ur a nswer with examples.
(e) Giv e the reasons lE:ading to the wid esp read use of power electronic converters .
(d) Enumerate at least ten applications of power electronics.
1. 2 . (a ) Give the advantages and disadvantages of power electronic converters .
(6) Describe a power electronic system with its general block diagram.
(e) List the following power semiconductor devices along with their circuit symbols and
m aximum ratings:
1. 3. (a ) Discuss the various types of power elect ronic converters .
(0) Compare a diode with a thyristor
(e) List the semiconductor devices which possess the capability of withstanding
(i ) bidirectional current and (ii ) unidirectional current.
(d ) Give the differences between a triac and a thyristor.
1.4 . (a) Giv e the differences between an ac voltage controller and a cycloconverter.
(6) Wha t is the difference between thyristors and controllable switches? Make a list of
unc ontrolled ar.d controllable switches.
(e) How ma ny semiconductor devices are required for (i ) H-bridge converter and ( ii ) three
phase full converter.
(d ) vVhat is power electronic module ? De ~c rib e smart power.
(a )

Chapt er 2

ower Se ic nductor

Diodes and ansistors

..... -.. -...... .. ... -.............. ... .. ... ... ... . . ............. . ... .. .. ... ..
- ~

In this Chapter

The p-n Junction

Basic Structure of Power Diodes
Characteristics of Power Diodes
Types o f Power Diodes
Powe Transistors
Power Mosfets
Insulated Gat e Bipolar Transistor
Static Induction Transistor (SIT)
Mos-controlled Thyristor (MCT)
New Semiconducting Materials

.- ..... ....... -.. ...... .. .. .. ... -....... -..... ... ........ .. . ... ....-... .. ... .... ... ..

-. ~

.. ...

A low-power diode, called signal diode, is a pn-junction device. A high-power diode, called
power diode, is als o a pn-junction device but with constructional fe at ures somewhat different
from a si gn al diode. Lifewise, power t r ansistors also differ in construction fro m signal
transisto s .
The voltag e, current and power ratings of power diodes and transistors ar e much high er
than the corresponding ratings fo r signal devices. In addition, power devices operate at lower
switchin g speeds whereas signal diodes and transist ors operate at higher switching speeds.
Power s emi conductor devices are u s ed ext ensiv ely in power-electronic circuits . Some
applications of power diodes in clude their use as freewheeling diodes, for ac to dc conversion,
for rec overy of t rapp ed en e gy etc. Power t r ansi stors, used as a switch i ng device 'n
power-electron ic cir cuit s, must operate in the saturation region in or der that their on-state
voltage drop is low. Thei r applications as switching elements 'n elude dc choppers and inverters.
Th e obj ect of t hi s chapte r is t o desc r ib e pow er diodes, pow er t r an sis t ors a nd
MO S -con tr oll ed t hy be or (MC T). A th yris to r is mOTp. im port ant compo n ent of pow er
sem" condu ctor devices;it is, therefore, discussed in detail in Chapter 4.
_ ...<vI.IP
' ~~~



~~~ _


.. .:
;I '

- r.:~?




~~;' ..." ,,,,..,.'1jJ::'~'"'"

. ., .,:.;".',


; ~r

A -n junction for ms t he basic buil ding block of all pow er semicondu ctor devLes. It is ,
th er efor e, worth while h er e to r eview this junction at an introductory l ev el.
A p-n junc bon ' s form ed when p -type semi con ductor is br otlgh t in me tallur gi cal , or
physical, comac wi th n -type sem icondu ctor. A p -r egion h as great "f can en tra tion of holes
whereas n-region has more electron-concent r lltion . In p-region, free h oles are called m aj or ity


Power Electronics

[Art. 2.1]

carriers anci. fr ee electrons min ority carriers. In n-region, fre e electrons'are called maj ority
carriers where as free holes are called minority carriers.
Doping densities in p and n type semiconductors may be different. As such, p-type material
m ay be design at ed p+,p or p-; similarly n-type mater ial as n\n- etc. Rough guidelines for
labelling of p as p + , p - etc and n as n- , n + etc are a s under:
(a ) If doping (or acceptor) density in p-type semiconductor = doping (or donor ) density in
n-type semicondu ctor, then it is called pen junction. For example, if doping density in both p
. a bou tl0 16 cm -3 t10
- 3 JunctIon
. .IS terme d
and n 1ayers IS
0 17 cm,
pen'Junc t'IOn.
(b) If doping density in p-region is much greater than that in n-region, it is called p+n
junction. For example, if doping densities are 10 19 cm- 3 in p layer and 1017 cm- 3 in n layer,.
then it is termed p + n junction.
(c) If doping density in n-type is less than that given in part (b ), the junction is called
p"'n- junction. For example, if doping densities are 10 19 cm- 3 and 10 cm- 3 for p and n types
respectively, then p + n- junction is formed .
(d) If both p and n-layers are heavily doped, it is called p'" n+ junction an d if very lightly
doped, a p-n- junction is formed. For example, if drensity is 10 19 cm- 3 in both p and n layers,
p'" n'" junction is formed.
In general, p + indicates highly doped p r egion, ' n- lightly doped n region and so on.
2.1. 1. Depletion Layer
'When physical cont act between p and n regions is made, fr ee electrons in 11, material diffuse
across th e junction into p m aterial, Fig. 2.1 (a ). Diffusion of each electron from n t o p , leaves
a positive ch arge behind in the n-region near the junction . Similarly, diffusion of each hole
from p t o n , leaves a negative charge behind in the p r egion near the jun ction. As a result of
this diffusion , n region near the junction becom es positively charged ~dp region in the vicinity
of junc tion becomes negatively ch arged, Fig. 2. 1 (b). These charges establish an electric field
across the junction . When this field. ,grows str on g enough, it stops furth er diffu sion. Some
electr ons , as these diffuse from n t o jI ie~ine '~h h oles in p-region and disappear. Similar
recombin ation occurs in n-region . .
Immob ile

( diffuse



Ii I\!I

Junct:o n

Widt h of deplet ion layer

(a l

( b)

Oeple: ion- lcy er width increases

Deplet ion- layer wIdth d?':: "eCses



- +
- +


t "I
- - ++
- - ++
- - - ++ ++

( l





F ig. 2. 1. A p en junction showing (0) direction of holes and electrons diffusion

den} tion region (c ) effect offorward biasin g an d (d). effect orr'everse bi asing.

. (b )

[. r t. 2.2J

Power Semiconduc to r Diodes and Transistors

. 11

When electric field s tops fur ther diffusion, charge carriers (h o es and electr ons) don't move .
As a consequence , opposite charges on each side of the junction produce immobile ions , Fig. 2.1
(b). The region extending in to both p and n semiconductor layers called depletion region or
space-charge region . The wid h of deple tion r egion, or depletion layer , is of th e orde r of
5 x 10- 4 mm . In equilibrium, there is a potential differenc of 0.7 V acros the depletion region
in silicon and 0.3 V across the depletion r.egion in germa.!lium.. This potential differen ce across
the depletion layer is called barrier potential.


When positive terminal of a battery is connected to p-type material and negative terminal
to n-type material, Fig. 2.1 (c), th~ p-n junction is forward biased. Positive terminal of the
battery sucks electrons from p material leaving holes there. These holes travel through p
material towards the n egative charge at p-n junction al'.d thus neutralize partly this negative
charge. Similarly, negative t erminal of the battery injects electrons into n layer. These electrons
move through n material , reach the p -n junction thereby neutralizing partly the positive
charge. As a res' It, width of depletion region gets r educed . .
In case p material is connected to negative terminal of the battery and n material to positive
t erminal of battery, then it can be deduced th at width of depl etion layer in.creases, Fig. 2. 1 (d) .
A rise in jun c60n temperature also decre ases width of depletion layer. As the barri er
potential depends on width of the depletion layer, the barrier potential decreases with rise in
junction temperature.
For p('lwer semiconductor devices, it should be kept in mind that (i) a junction with light ly
doped layer on its on e side requires large breakdown voltage and (ii) a juncti0:l with highly
doped layers on its both sides r equires low breakdown voltage.

Power diodes differ in st ructure from signal diod es. A signal diode. constitutes a simple p-n ..
junction as shown in- Fig. 2.1. The in tricacles in constructing powe diodes arise 'from the need
to make them suit able for high-voltage and h igh-current applicat ions.
The practical realizat' on and th resulting structur e of a power diode is shown in Fig. 2.2
(a). It consists of heav'ly dop8d n" substrate". On this substrate, a lightly doped n- layer is
epi texially grown . Now a heavily doped p "'layer
... y Anode
is diffused into n- layer to form th e anode of
powe r diode , Fig. '2.2 (a ). This shows that n
layer i the basic structural feature not found in
signal di odes. Th e fu ch on of n - lay er is to
:Dri1 !
absorb the depl etion layer of th e rev ers e biased
I region
p"'n- j u nc tion J 1 . The br eak-d own volt age
n '"
Sub strut '
needed in a power d' ode governs the thi kness
of n- layer; greater t he br eakdown voltaD'e,
more the n- layer thickness. The drawback of
-Co hod !?
K. cothodi'
n-layer is to add signi lcant ohmic resistance to
( b)
th e dio d e wh en it is cond u cti g a forward
current. This leads to large p ower dissipation in Fig. 2.2. (a) Structura fea tures of power diode
the diode ; so proper cooling arrangements in
and (b) i 3 cir uit symbol.
large diode ra tings are es ential.


Power Electroni s

[Art. 2.3]

The circuit symbol of a power diode, shown in Fig. 2.2 (b), is t he same as that for a signal
The modifications in the context of diode, presented above, makes them appropriate for
high-power applications. As diode, or p-n junction, is the basic building block of all other power
semiconductor devices; same basic modifications should be implemented in all low-power
semiconductor devices in order to raise their power-handling capabilities.

As stated before, power diode is a two-terminal, p-n semiconductor device. The two
terminals of diode are called anode and cathode, Fig. 2.2 (b) and Fig. 2.3 (a). Two important
characteristics of power diodes are now described.

2.3.1. Diode i-v Characteristics

When anode is positive with respect to cathode, diode is said t o be forward biased. With
increase of the source voltage Vs from zero value, initially diode current is zero. From Vs = 0 to
cut-in voltage, the forward-diode current is very small. Cut-in voltage is also known as
threshold voltage or tum-on voltage. Beyond cut-in voltage, the diod e current rises rapidly and
the diode is said to conduct. For silicon diode, the cut-in voltage is around 0.7 V. When diode
conducts, there is a forward voltage drop of the order ofO .S to 1 V.
For low-power diodes, current in the forward direction increases first exponentially with
voltage and then becomes almost linear as shown in Fig. 2.3 (b). For power diodes, the forward
current grows almost linearly with voltage, Fig. 2.3 (c). The high magnitude of current in a
power diode leads to ohmic drops that hide the exponential part of i-v curve. The n- region, or
drift region, forms a considerable drop in the ohmic resistance of power diodes.
Forward voltage









F ig. 2.3.





A forwar d-biased power diode. i-v characteristics of (b) signal diode

(c) power diode and (d) ideal diode.

When cathode is positive with respect to anode, the diode is said to be re verse biased. In the
reverse biase d condition , a small reverse curren t called leakage current, of t he ord er of
microam pel"es or m ill iamperes (for large diodes) flows. The leak age cur-rent is almost
independent of the magnitude of reverse voltage l..m til this voltage reaches breakdown voltage.
At th is r everse breakdown, voltage remains almost constant but rever e current becomes quite
high-limited only by the ext ernal c'rcui t r esist ance, A l arge r everse br eak down voltage ,
as ociated with h igh r evers e current, leads to excessive power loss that may destroy the diode.
This shows that reverse b eakdown of a power diode must be avoided by _ope a ting it below the
specific peak reverse r epetitive voltage VRRM . Fig. 2.3 (c ) illustrates th.e i-u charac teristi cs of
;, Som e

au thors WT1te u-; characteristics

Power Semlcondueto Diodes and

[Art. 2.3]



power diode and V RRm . For an ideal diode, the i-v characteristics are shown in Fig. 2.3 (d). Here,
volt age drop across conducting diode, VD = 0, revers e leakage current = 0, cut-in voltage = 0 and
r everse breakdown voltage V RRM is infinite.
Diode manufacturers also indicate the value of neak inverse voltage (PlV) of a diode . This
is the largest reverse voltage to which a diode m ay be subjected dur ing its working. PlV is the
same as V RRM'
The power diodes are now available wi th forwar d curren t ratings of 1A t o several thousand
amperes and with reverse voltage ratings of 50 V to 5000 V or mo e.

2.3.2. D iode R everse Recovery Chara cteristics

After the forward diode current decays to zero, the diode continues to conduct in the reverS E;
direction because of the pres s re of stored charges in the dep letion region and the
semiconductor layers. The reverse current flows for a time ca lled reverse recovery time t,.,..
The diode regains its blocking capability until reverse recovery current decays to zero. The
reverse recovery time trr is defined as the time between the inst ant forward diode current
becomes zero and the instant reverse recovery current decays to 25% of its reverse peak value
I RM as shown in Fig. 2.4 (a).
The reverse recovery time is composed of two segments of time ta and t b, i.e. t r, = ta + tb'
Time ta is the time between zero crossing of forward current and peak reverse current I RN/
During the time t a , charge stored in depletion layer is removed. Time tb is measured from the
instant of reverse peak value I RM to
the instant when 0.25ljV.;f is reached,
' ..
Fig. 2.4 (a). During t b , charge from
~trr --i
the semiconductor layers is removed. (a )
: ' t a - tb -...j
The shade d area in Fig. 2 .4 (a)
represents the stored ch arg e, or
reverse recovery charge, QR which

must be rem oved during the reverse

recovery time t rr . The r at' 0 tb1ta is (b)

called the softness factor or S-factor.
o ~~~==~~~~-L----------:
T h is factor i s a measure of t h e
voltage transients that occur dt1 ring
the time diode r ecover . Ita usu al
loss in
~ !
value is unity an d this indicates low (c) dio de ssssssss<ss<ssssJ,,~
, ~
oscillatory reverse recovery process.
In case S factor is small, diode as

Fig, 2.4. Reverse recovery characteristics

lar ge os cill at ory ov er voltages. A

(a ) varia tio n of forward current if
diode with S -factor equal to one i
v oltage drop vf and (c) power loss in a dio de.
called soft-reco very diode and a diode
with S-fact or less than one i called
snappy -re co very diode or fast re covery diode. In Fig. 2.4 (6) is shown th e wav form of
forward-v Itage drop vI' across the diode. The pro uct 0 vr and i f gives the powe loss in a clio e.
Its vari.ation is shown in Fig. 2.4 (c ). The av erage value of vI' i/ gives th e total power 10:>':: in a
diode. Fig. 2.4 (c) reve' Is that major power loss in a diode occurs during th period tb
It is no ticed fro Fi g. 2.4 (a) that pe ak inverse current IRM can be exp esse as


: : :


I '


.. (2 .1 )


Power Electro ks

[Art. 2.4]

where ddi is the rate of change of reverse current. The reverse recovery characteristics of Fig.
. .
2.4 (a) can be taken to be triangular. Under this assumption, storage charge QR. from Fig.
2 .4 (a ), is given by

."ut rr:
QR =-I
:2 R

... (2.2)

If tiT == t a , then fr om Eq. (2.1),



... (2.3)

= t rr dt

From Eqs. (2 .2) and (2 .3), we get

di::;: 2 Q R
rr dt

t .

'. . [2

From Eq. (2.1), with ta == trr>



rr - (di / dt )
we get

:[ . 2 QR

IR.Id = trr . dt = (dUdt)


. -_[2Q R .(diJ~1I2





It is seen fr'om E qs. (2.4) arid (2.5) that reverse recovery time trr and peak inverse current

are dependen t on storage charge and rate of change of current dt" The storage charge

depends upon the forward diode current IF. This shows that reverse recovery time and peak
inverse current depend on forward field, or diode, current.
A power-electronics engineer must know peak reverse current IRM stored charge QR'
S-factor, PN etc in order to be able to design the circuitry employing power diodes . These
parameters are u sually specified in the catalogue supplied by the diode manufacturers.


Diodes are cl assifi ed according to their reverse recovery characteristics. The three types of
power diodes are as under:
(0 General purpose diodes
(ii ) Fast recovery diodes
(iii ) Schottky diodes.
These ar e n ow described briefly.
2.4.1. General purpose Diodes

Th es e diode have rel atively high r everse r ecovery time, of the order of about 25 ~lS . Their
current r atings vary from 1 A t o several t hous and amperes and the r ange of voltage rating is
fr om 50 V to ab ou t 5 kV. Applic ation3 0 power diodes of th is type inclu de battery ch arging,
electr ic a action, electroplating, welding and unin terTUp tible power supplies (UPS).

Power Semjcon ductor Diodes and T ansistors

(Art. 2.5]

2.4.2. Fast.recovery Diodes

T h e diodes with low rever s e recovery time, of about 5 ~s or le ss , are cl assifi ed as
fast-recovery diodes. Thee are used in choppers, commutation circuits, switched mode power
supplies, in ducti on heating etc. Th eir current ratings vary from about 1 A to several t housand
a mperes and voltage r atings from 50 V to about 3 kYo
For volt age ra tin gs below about 400 V, the epitaxial process is used for diode fabrication .
These diodes have fast recovery time, as low as 50 ns.
For voltage ratings above 400 V, diffusion technique is used for the fabrica tion of diodes. In
order to shorten the reverse-recovery time, platinum or gold doping is carried out. But this
doping may increase the forward voltage drop in a diode .
2.4.3. Sc hottky Diodes
This class of diodes u se metal-to-semiconductor junction for rectification purposes instead
of p-n junction. The metal is usually aluminium and semiconductor is silicon. Therefore, a
Schottky diode has aluminium-silicon junction. The silicon is n-type.
When Schottky diode is forward biased, free electrons in n material move towards the Al-n
j unct ion and th en travel thr'ough the metal (aluminium) to constitute the flow of for war d
current. Since metal does not have ~ ny holes, this forward current is due to the movement of
electrons only. As the metal has no holes, there is no storage charge and no-reverse recovery
time . It can , ther efore, be said that rectified current flow in a Schottky diode is by the
movement of m aj ority car riers (electrons) only and the turn-off delay caused by recombination
is avoided . As such , Schottky diode can switch off much faster thanp-n junction diode.
As comp ared t o p-n junction diode, a Sch ottky diode has (i) lower cut-in volt age, (i i ) .~l gher
everse leakage current and (iii) higher oper ating fr equency. Their revers e voltage r a tings ar?
limite d to about 100 V and forward current ratings vary from 1 A to 300 A. Applications of
Schottky diode include high-fr equency instrumentation an d switching power supplies ..
Th e electrical and thermal ch aracteri.stics of power diodes are similar t o those of thyrist ors
wh ich are descr ibed in chapter 4.

P ower d iodes ar e uncontrolled devi ces. In oth er words, their t u rn-on an d turn -off
ch aracteristics ar e n ot under con trol. Power t ransistors, however, possess co n t rolle d
charac teristics. These are turned on when a current signal is given to base, or control, terminal.
The tran ist or reonains in the on-state so long as cont r ol signal is present . When this control
si gnal is r em oved, a power transistor is turned off.
Power transist ors are of our types as under :
Bip olar junction transis ors (BJ Ts)
Metal-oxide-semicondu ctoT fi e d-effect transistors (M OSFET s )
(i ii ) Insulated gate bipolar transistors (I GBTs ) an d
(iv ) Static in duction t r ansistors (SITs ).
(i )

(i i)

These four types are now described one after th e other .

2.5.1. B ip ol ur Jun ction Transist ors

A bipolar transistor is a three-layer, t'i 70 j un ction npn or pnp semiconductor device. With
on e p -r egion sand...vichecl by two n -regions, Fib ' 2. 5 (a ), np n transistor is obtained. With two
p-regi ons sandwichin g one n-region , Fig. 2.5 (6), p np transist or is obtain ed . The ter m' bipolar


Power Electronics

[Art. 2.5]

denotes that the current flow in the device is due to the movement of both h oles and electrons .
A BJT h as three terminals named collector (e), emitter (E) and base (B). An emitter is indicated

by an arrowhead indicating the direction of emitter current. No arrow is associated! with base
.or collector . Power transistors of npn type are easy to manufacture and are cheaper also.
Therefore, use of power npn transistors is very wide in high-voltage and high-current
applications. Hereafter, npn transistors would only be considered.




8 ase






Fig. 2.5. Bipolar junction transistors (a) npn type and (b) pnp type. Steady-state Characteristics. Out of the three possible circuit configurations for
a transistor, common-emitter arrangement is more common in switching applications. So, a common
emitter npn circuit for obtaining its characteristics is considered as shown in Fig. 2.6 (a).

Inp ut characteristics. A graph between 'base current IB and base~emitter voltage VBE
gives input ch ar acteristics. As the base-emitter junction of a transistor is like a diode, IB vers us
V BE graph resembles a diode curve. When collector-emitter voltage V CE~ is more th an V CEl'
base curr en t, for the same VEE' decreases as shown in Fig. 2.6 (b).



VCE 2>'1CE !



Ia5 >184> --- >181


Rc ~Ic


- 1+



f lVaE

t )
Fig. 2.6.

(a) non









I al



Leako ge

(b )

transistor circuit characteristics, (b) input charact eristi cs and


Cu tol1


(c )

output characteristics.

Ou tp ut haracteristics . r graph between coll ector current Ie and coll ector-emitter

vol ts.ge VCE gLe3 output character istics of a tr sis tor. F r zero base current, i.e. for IE = 0, as
VeE is increased, a sm all le akage (collector ) curren t exist as shown in Fig. 2.6 (c). As the base
current is increased from.lB = 0 t o 1Bl. I B2 etc, coll ector cun snt als rises as shm-'nl in Fig. 2.6 (c) .

(Art. 2.5]

Power Semiconductor Diodes and Transistors


Fig. 2.7 (a) shows two ofthe output characteristic curves, 1 for 1B = 0 and 2 for 1B ;f. O. The
initial part of curv'e 2 , characterised by low VeE, is called the saturation region. In this region,
the transistor acts like a switch. The flat part of curve 2, indicated by increasing VeE and almost
constant Ie is the active region. In this region, transistor acts like an amplifier. Almost
vertically rising curve is the breakdown region wh,ich must be avoided at all costs.




Saturation pOint


Rc --








p .

Base -.--'"

Load line


, ( 1,"0

I E~

B :=;) eEO
B .--"",

. Vee





Fig. 2 .7. (a) Output characteristics and load line for npn transistoF and
(b) electron flow in an npn transistor.

For load resistor R e , Fig. 2.6 (a), the collector current Ie is given by
V ee - VeE

. ~

This is the equation of load line. It is shown as line AB in Fig. 2.7 (a). A load line is the locus
of al1 possible oper ating points. Ideally, when tTansistor is on, VeE is zero and Ie =V eelRe This
collector current is shown by point A on the vertical axis . When th e transistor is off, or in the '
cut-off region, Vee appears across collector-em itter terminals and there is no collector current.
This value is indicated by point B on the horizontal axis. For the resistive load, the line joining
points A and B is the load line.
Relation bet wee n a and ~. Most of the electrons, pr oportional to IE given out by emitter,
reach t h e collector as shown in Fig. 2 .7 (b). In other words, collector current I e, though less than
emi tter current IE, is almost equal to IE. A symbol a is used to indicate how close in value these
t wo currents are. Her e a, called forward current gain, is defmed as
a= -[


As I e < IE, value of a varies fr om 0.95 to 0.99.

In a transistor, bas e current is effect ively th e input current and collector current is the
outpu t current. The r atio of collector (output) C Trent Ie to base (input) curre nt IB is kno
the curre ne gain ~.


Powe r Electronics

[Art. 2.5J

As IB is much smaller, ~ is much more than unity; its value varies from 50 ~o 300. In
an other system of analysis, called h parameters, hFE is used in place oJ~.
... (2.7)

Use of KCL in Fig. 2.6

(a )

... (2.8)


Remember that emitter current is the largest of the three currents, collector current is
almost equal to, but less than, emitter current. Base current has the least value. Dividing both
sides of Eq. (2 .8) by Ie, we get









Transistor S'\yitch. Transistor operation as a switch means that transistor operates either
in the saturation region or in the cut-off region and nowhere else on the load line. As an ideal
switch, the transistor operates at point A in the saturated state as closed switch with
VCE = 0 and at point B in the cut-off state as an open s"vitch with Ic = 0, Fig. 2.7 (a). In practice,
the large base current will cause the transistor to work in the saturation region at point A'
with sm all saturation voltage VCES ' Here subscript S is used to denote saturated valu e. Voltage
V eEs represents on-state voltage drop of the transistor which is of the order of abou t 1 V. When
the control, or base, signal is reduced to zero, the transistor is turned off and its operation
shifts to B' in the cut-off region, Fig. 2.7 (a) . A sman leakage current ICED flows in th e collector
circuit when the transistor is off.
For Fig. 2.6

(a ),

KVL for the circuit consisting of VB' RB and emitter gives

VB or
Also , from Fig. 2.6 (a),

VBE = 0
I B =--=- RB
Vec = VCE +Ic Rc
VCE =Vec - Ic.Rc =Vee - ~ IB Re


~ Rc
== Vec - RB (VB -


VCE = V eB + V EE


... (2.11)


.. (2. 13)

If V CES is the collector-emitter saturation vol tage, then collector current l cs is given by

Vcc - V eEs

1cs == - ---::::,.---

... (2.14) .

[Art. 2.5]

Power Semiconductor Diodes and Transist ors


and the corresponding value of minimum base cur r ent, that produces saturation, is

... (2. 15-)

Ifbase current is less than I BS , the transistor operates in the active region, i.e. somewhere
between the saturation and cut-off points. Ifbase current is more than I Bs , V eEs is almost zero
and collector current from Eq. (2 .14) is given by Ies = Vee/Re . This shows that collector cunent
at saturation remains substantially constant even if base current is increased.
With base current more than I BS ' hard drive of transistor is obtained. With hard saturation,
on-state losses of transistor increase. Normally, the practical circuit is designed for hard-drive
of transistor and therefore, base current IB is greater than I BS given by Eq. (2.15 ). The ratio of
IB and I Bs is defined as the overdriv~factor (ODF) .
.~ I
.. .(2.16)




ODF may be as high as 4 or 5.

The ratio of Ies to IBis called forced current gain

~r where


~f = I B < natural current gain ~ or hFE

... (2.17)

The total power loss in the two junctions of a transistor is


... (2.18)

= VBE I B + VeE Ie

Under sat urated state, V BES is gre.ater than V CES , t his means BEJ is fo rward biased .
Further Eq. (2.13 ) shows that V eB is ne~ative under sat u.rated conditions , therefore, CBJ is also
forward biased. In oth er wor ds, under saturated conditio ns, both ju nct ions i.n a power
transistor are forwar d biased.
Example 2.1. A bipolar transistor show n in Fig. 2.6 (a) h as current gain ~ =40. The load
resistance Rc = 10 .0, d c supply voltage Vec = 130 V an d i npu t voltage t o base circu it,
VB = 10 V. For V CES = 1.0 Vand V BES = 1.5 V, calculate.
(a) the value of RB for operation in the saturated state,
(b) the value of RB for an over drive factor 5,
(c ) forced- current gain and

(d) power loss in the transistor for both parts (a) a nd (b) .

Solut ion . H ere ~ = 40, Re;" 10 .0, V ce = laO'V, y"'B = 10 V, VeE ~

(a )

= 1.0 V and V BES = l.5 V

From Eq. (2 .14), for operation in the sa turated state,

1 r ~
~ ,:,

= V cc ~ V CES'=

13 0 - L 0

= 1 2 qo A

From Eq. (2. 15), base current that produ ces saturaGion ,


= 1cs
= 0 322 - A

Value of R B for I Bs = 0.3225 A is given by Eq. (2.11) as,

R = V B - VBES = 10 ~ 1.5 = 2 6 3r:.7 !J

0 .3......
. 0 1.
9 ')5

(A r t. 2.5]

Power Electr onics

(b ) Base current with overdrive, fr om Eq. (2.16), is


I B = ODF X I BS = 5 x 0.3225 = 1.6125 A



10 - 1.5

1.6125 = 5.27

Forced current gain, from Eq. (2.17), is

~f= Ies =


= 8, which is less

than the natural current gain

~ = 40.

(d ) Power loss in transistor, from Eq. (2.18), is


= V BES lBE + V CES Ics

For normal base drive, P r = 1.5 x 0.3225 + 1.0 x 12.9 = 13.384 W

. With overdrive, P r = 1.5 x 1.6125 + 1.0 x 12.9 = 15.32 W
It is seen from above that power loss with hard drive of transistor is more. BJT Switching Performance. When base current is applied, a transistor does
not turn on instantly because of the presence of internal
capacitances. Fig . 2 .9 shows the various switching
waveforms of an npn power transistor with resistive load
between collector and emitter, Fig. 2.8 .

When input voltage VB to bas.e circuit is made - V 2 at

to> junction EB or EBJ is reverse biased, v BE = - V 2 , the
transistor is off, ia = Ie = 0 and vCE = Vee , Fig. 2.9. At time
t I , input voltage VB is made + VI and iB rises to IB l as
sh own in Fig. 2.9. After t 1 , base-emitter voltage VBE begins
to rise gradually fro m - V 2 and collector current ic begins

l -

1) B

t ___. .

t B

i" "
" BE


Fig. 2.8. npn transistor with

resistive load.
to rise from zero (actually a small leakage current I CEO
exists as shown in Fig. 2.7 (a)) and collector- emitter voltage VeE starts falling from its initial
valu e Vee. After some time delay t d , called delay time, the collector current rises to 0.1
I es , veE falls from Vee to 0.9 Vee and vBe reaches VBES = 0.7 V. This delay time is required to
charge the base-emitter capacitance to V BES = 0.7 V. Thus, dp.lay time td is defined as the time
during which the collector current rises from zero to 0.1 Ies an d collector-emitter voltage falls
from Vee to 0.9 Vec.
. After delay time td , collector current rises from 0.1 Ies to 0.9 Ics and veE falls from 0.9
Vee to 0.1 Vee in tim e t r . This time t,. is known as r ise time which depends upo n transistor
junction capa citances. Rise time t,. i defined as the time during which collector curr en t ris es
from O.lIes t o 0.9 Vee and collector-emitter voltage falls fro m 0.9 Vee to 0.1 Vce. This shows
that total turn-on time ton = t d + t r . Value of t on is of the order of 30 to 300 nano seconds. The
transistor remains in the on, or saturated, state so long as input voltage stays at Vl' Fig. 2.9

In case transistor is to be turned off, then input voltage vB and inpu t base urrent iB are
r eversed . At time t 2, inpu t voltage v B to bas e circuit is reversed from VI to -- V 2. t th e same
tim e, base cu rren t changes from I Bl to -IB2 as shown in Fig. 2.9 (b). Negative base current 13 '2
rem oves excess carri ers from the base . The time ts r e quir ~ d to r emove these exc ess carri r is

Power S emicondu ctor Di.odes and

[Art. 2 ..5]



- 1/2




T =1-

tS f


o ~------~--~r_IB~------~~~
. -_I-B2~.~!--------~~



+3 :
' i

( c)




o I 'Ice

~ ~ tn ---l-~




F ig. 2.9. Switching waveforms for npn power transistor of Fig, 2.8.

called storage time and only after t s , base curr en t IB2 begins to decrease towards zero .
Transistor comes out of saturation only after ts' Storage time ts is usually defined as th e time
during which collector current falls from Ies to 0.9 Ies and collector-emitter voltage VeE rises
from V eEs to 0.1 Vee, Fig. 2.9 (d) and (e). Negative input voltage enhances the process of
removal of excess carriers from base and hence reduces the storage time and therefore, the
turn-off time.

After t S ) collector current b egins to fall and collector-emitter volt age st rts building c p.
Time tr, called fall time, is defined as th e time during which collector current drops from
0.9 Ies to 0.1 Ies and collector-emitter voltage rises from 0.1 Vee to 0.9 Vee, Fig. 2.9 (d) and (e) ,
Sum of storage time and fall time gives the transistor turn-off time torr i.e . t ofr ;=' s + tr The
vari ous waveforms during transistor swi tching a e shown in Fig. 2 .9. In t h is figu 8; tn =
conduction period of tt-ans 'stor, to = off period, T = 11f is the periodic time andfis t he switc ing
fr equency.
2 .Ll .1.3. Sale Oper ating Are a. The safe operating ar ea (SOA or SOAR) of a power tran
sis tor specifies the sa e operating limits of coll ect or current l r. versus collector-emitte voltage
V Cg For _eli able op er ation of the transistor, the collector current and v oltage m ust always lie
within this a. ea. Actually, two types of safe operating areas are specifie d by the m anufacturers ,


Power El ectronics

[Art. 2.5]

The forward-base safe operating area (FBSOA) pertains to the transistor operation when
base-emitter jun ction is forward biased to turn-on the transistor. For a power transistor, Fig.
2.10 shows typic-al FBSOA for its dc as well as single-pulse operation. The scale for
Ic and VCE are log~thmic. Boundary AB is the maximum limit for dc and continuous current
for Vc~ less than abouf S0 V.For VeE for more than 80 V, collector current has to be reduced
to boundary BC so as to limit the junction temperature to safe values. For still higher V CE,
current should further be reduced so as to avoid secondary breakdown limit. Boundary CD
defines this secondary breakdown limit. Boundary DE gives the maximum voltage capability
for this particular transistor.
For pulsed operation, power transistor can dissipate more peak power so long as average
power loss is within safe limits of junction temperature. In Fig. 2.10 ; 5 ms, 500 ~s etc. indicate
pulse widths for which transistor is on. It is seen that FBSOA increases as pulse-width is
It should be noted that FBSOA curves, as given by the manufacturers, are for a case
temperature of 25C and for dc and single-pulse operation. In order to take into consideration
the actual working temperature and repetitive nature of the pulses, these curves must be
modified with the help of thermal impedance of the device . .

100 I-:A----~B'

80 .

I B =-5A




1~,0----------JO~0~--------~ E


Fig. 2.10. Typical forward biased safe

operating area (FBSOA) for a power

transistor (logarithmic scale)









Fig. 2.11. Typical reverse-block safe operating

area (RBSOA) for a power transistor.

During turn-off, a tr ansistor is subjected to high current and high voltage with base-emitter
junction reverse biased. Safe operating area for transistor during turn-offis specified as reverse
blocking safe operating area (RBSOA). This RBSOA is a plot of collector current versus
collector-emitter voltage as shown in Fig. 2.11. RBSOA specifies the limits of transistor
operation at turn-off when the base current is zero or when the base-emitter junction is reverse
biased (i.e. with base current negative). With increased reverse bias, area RBSOA decreases
in siz e as shown in Fig. 2.11.
Example 2.2. For a power trans istor, typical s witch ing waveforms are shown in Fig.
2. 12(a). The various parameters of the transistor circuit are as under:
\lce =220 v~ V CES ::::: 2 V, lcs = 80 A , td::::: 0.4 1lS, tr::::: 1 ~s, t,.. = 50 ~s,
ts '=.= 3 ~l S, t f ::::: 2 ,us, to = 40 ,us, f = 5 kHz. Collector to emitter le akage curre nt itlaic= 2 mA.
De termine a verage p ower loss due to collector current during t on and t n . Find
also the peak
instantaneo us po wer loss due to collector current du ring turn on time.

[Art. 2.5]

Power Semiconductor Diodes and Transistors


Solution. During delay t ime , the tim e limits are 0 S t Std. Fig. 2.12 (a ) shows that in this
time, ic (t) = ICEO and vCE (t) = VCC.
:. Instantaneous power loss during delay time is
P d (t) = ic VCE = IcEO VCC = 2 x 10- 3 x 220 = 0.44 W
Average power loss during delay time with 0 ~ t S td is given by
Pd = T

ftd .


ftd0 I

0 LcC t )

VCE (t) dt

CEO Vcc dt

=f I CEO . Vcc . td

,,;, 5 x 103 X 2 X 10- 3 x 220 x 0.4 x 10- 6

= 0.88 mW

f = ~ =frequency ofiransistor switching


o ~ t S tn

During rise time,


= tCS.




=[ Vcc


Vcc - VCES ]

:. Average power loss during rise time is

ICS.t[v _ VCC-VCES . t]dt

T Ot

VCC . Vcc - VCES ]


= f . I cs . tr [-2- =5

10 3 X 80 x 1 X 10- 6 [2~0 -

22~ -.2J = 14.93~ W

- - T ='/1


----l:f -----_

Ic s


. 0 (... t on..l--- t n - --

Fig. 2.12.



Switching wave arms for Examples 2.2


2.3 .


Power Electronics

[Art. 2.5]

Instantaneous power loss during rise time is

PI' (t ) = Ics . t {VCC _ Vce - V CES . t

Ics . t I c s . t
== - t - Vcc 2
[Vcc - VCESJ

.. .(i)

d PI'
. '.tm at whi ch mstantaneous
dt (t) == O
gIves bme
power 1oss d
tr wou Id b e maXlmum.

It is seen from Eq. (i) that

VCCt r
220xlxlO- 6
tm = 2 [Vcc - VCES] = 2 [220 _ 2] = 0.5.046 Ils
Peak instantaneous power loss P rm during rise time is obtained by substituting the value of
t == tm in Eq. (i ).

Y1:c ' tr
Ics (Vcc ' tr)2 [Vcc - VcEsl
I'm tr 2 [Vcc - VcEsl - ~t;
4 [Vcc V cEs12
==- .

Ics . Y1:c
== 80 X 220 = 4440 4 W

4 [Vcc - VCES] 4 [220 -.2J


Total average power loss during turIVon

Pon == Pd + PI' = 0.00088 + 14.933 =14.9339 W

During conduction time, 0 ~ t ~ tn

ic (t) ==lcs and VCE (t) == V CES

Instantaneous poy.,rer loss during tn is

P n (t) == ic . VCE == 1cs . VCES == 80 x 2 = 160 W

Average power loss during conduction period is



Lc,vcE dt=flcs,VCEStn

= 5 x 103 X

80 x 2 x 50 x 10- 6 == 40 W.

Example 2.3. Repeat Example 2.2 for obtaining average power loss during turn-off time
and off-period, and also peak instantaneous power loss during fall tim e due to collector current.

Sketch the instantaneous power loss for period T as a fu nction of ime.

Solution. During storage time, 0

ic (t)

~ t ~ t s

=Ics and VCE (t) = V CES

Instantaneous power loss during ts is

Ps (t)

= ic (t) VCE (t)

=Ics' VCES == 80 x 2 == 160 W

Average power loss during ts is

P s = ~ J~' Ics' V CES . d t = f ICE ' V eEs ' ts

== j


10 3 x 80 x '2 x 3 )( 10- 6 == 2:4 VI!

[A rt. 2.5]

Power Semi co nduc tor Diod es an d Transistors


. f a 11 t'lIne, 0 ::; t ::; t , ~c

. (t)::::: [I' cs - Ics -t I CEO 'J
D Urlng

Du ring t f , ICEO is negligibly small in comparison with I cs ,


iclt)=Ies [l-il


-V s t

Average power loss during fall time is


f T

Ics [ l -

i J'[VCC - VCES . t] dt


Ics Ics]
:::::f(Vcc - V CES ) , tf [ 2-3
:::::f' tf ' s

Wcc -


::::: 5

10 3 X 3 x 10- 6 X 80 x ~ x (220 - 2) ::::: 43.6 W

Instantaneous power loss during fall time is

PP) =Ies


~::::: 0

[l-ifee ~fVeES



gives time tm at which instantaneous power loss dur ing tr would be maximum ,

Here tm = tf l 2.
:. Peak instantaneous power dissipation during tf is






= 80 (220 - 2) ::::: 4360 W

Total average power loss during turn-off proces is

Pof{ = P s + P =2 .4 + 43.6 : : : 46 W


Durin g off-period, 0

::;J ::; to,



and uCE(t)::::: Vce

Instantaneou s power loss during to is

Po(t) -= ic . VeE =I CED ' Vec ::::: 2 x 10- 3 x 220 : : : 0.44 W

Ave rage poweT loss during to is

1 flO

Po = T
::::: Q


Po(t)dt -= f I CEO ' Vee ' to

10 3 X 2,x 10- 3 x 220)( 4 0 x 10- 6::::: 0.0 88 W


P ower E.lectronics

[A r t. 2.6]
P (t)

4440.4 W


,----- --\ - - --

0.44 W



' - - - - - " O.44W

- - ------- ------ - --- ---
____- __--- __
____- 4 - ____+ ___ t

- --



--+----- tn + ts

_ - - - - _.....I .-..-.tf

.....41 .... to


. Fig. 2.12.


Sketch of instantaneous power loss in a transistor for Examples 2.2 and 2.3 .

Total average power loss in power-transistor due to collector current over a period T is
PT = Pon + Pn + Poff + Po = 14.9339 + 40 + 46 + 0.088 = 10l.022 W.
From the data obtained in Examples 2.2 and 2.3 , the power loss variation as a function of
time, over a period T, is sketched in Fig. 2.12(b).
Example 2.4 . Apower transistor has its switching waveforms as shown in; Fig. 2.13. If the
a uerage power loss in the transistor is limited to 300 W, find the switching frequency at which
this transistor can be operated.

Energy loss during turn-on

200 V


t Oil






= J:t~~ x 10' t)(Vc~- :coc x 10'

t Oil






= J (2 x 10 6t ) (200 - 5 x 10 6t ) dt

= 0.1067 watt-sec

Fig. 2.13. Switching waveform for Example 2.4

Energy loss during turn-off

=J;"ff (100 ~ 160 x 10' t)( 275 X 10' t)dt

= 0.1603 watt-sec

Total energy loss in one cycle

= 0.1067 + 0.1603 = 0.267 W-sec

Av erage power loss in transistor

= sWitching frequency x energy loss in one cycle

... Allowable switching frequency,


f= 0.267 = 1123.6 Hz
2.6.: PO~1'I~~Ts l

','. '.::' ~:


, .:

A metal-oxide-s emiconductor fi eld-effect tran ~ ist o ( IOSFET) is a recent device dev elop ed.
by com bin ing t h e areas offield-effect con cept an d MOS technolo gy,

[Art. 2.61

Power Semicond ctor Diodes and Transistors


A power MOSFET has three terminals called drain (D), source (8) and gate (G) in place of
the corresponding three terminals collector, emitter and base for BJT. The circuit symbol of
power MOSFET is as shown in Fig. 2.14 (a). Here arrow indicates the direction of electron flow.
A BJT is a current controlled device whereas a power MOSFET is a voltage-controlled devic e.
As its operation depends up on the flow of majority carriers only, M08FET is a unipolar device.
The control signal, or base current in BJT is much larger than the control sign al (or gate
current) required in a M08FET. This is because of the fact that' gate circuit impedance in
MOSFET is extremely high , of the order of 10 9 ohm. This large impedance permits the
MOSFET gate to be driven directly from microelectronic circuits. BJT suffers from second
breakdown voltage whereas MOSFET is free. from this problem. Power MOSFETs are now
finding increasing applications in low-power high frequency converters.










p-subs trate

So urc ?

Fig. 2.14. N-channel enhancement power OSFET

(a) circuit symbol and (b ) its basic str ucture.

Power MOSFETs are of two types; n-ch annel enhancement M08FET and p-channel
enhancement MOSFET. Out of these two types, n-channel enhancement MOSFET is more
common because of higher mobility of electrons. As such, only this type of 'lOSFET is studi ed
in what follows.
A simplified structure of n-channel planar MOSFET of low power rating is shown in Fig.
2.14 (b). On p-suhstr ate (or body), two h eaVIly doped n+ regions are diffused as shown. An
insulating layer of silicon dioxide (Si0 2 ) is grown on the surface. Now this insulating layer is
etched in or der to embed metallic sour ce and drain terminals . Note that n'" regions make
contact with source and drain t erminals as shown. A layer of m etal is also deposited on Si0 2
layer so as t o form the gate ofMOSFET in between source and drai n terminals, Fig. 2.14 (b).
When gate cir cuit is open, junction be t\ve en n'" region b;,!.ow drain and p-substrate i
reverse biased by input voltage V DD' Therefore, no current flows from drain to source an d lo ad .
'When gate is made positive with r espect to source, an electri c fi eld is established as shown in
Fig. 2.14 (b). Eventually, induced negative charges in the p-SLlbstrate below Si0 2 layer are
formed thu causing the p layer below gate to become an induced n layer. These neg ative
charges, called e ectrons, form n-ch ann el between t vo n '" egions and current can floY'; from
drain to sour ce sh own by t he arrow. If V G L made lore pos 'tiv e, induced n-channel becom es
m ore deep an therefore m ore current fl ows from D to 8. This shows that drain curr ent ID is
enhan ced by the gradu al increase of gate voltage , h ence th e nam e enhancemen t NIOS ET.



P ower Electronics

[Ar t. 2.6]

The m ain disadv antage of n-channel planar MOSFET of Fig. 2.14 (b) is that conducting
n-channel in between drain and source gives large on-state resistance. This leads to high power
dissipation in n-channel. This shows that planar MOSFET construction of Fig. 2.14 (b) is
feasible only for low-power MOSFETs.
The constructional details of high power MOSFET are illustrated in Fig. 2.15. In this figure
is shown a planar diffused metal-oxide-semiconductor (DMOS) structure for n-channel which
is quite common for power MOSFETs. On n+ substrate, high resistivity n-layer is epitaxially"
grown. The thickness of n- layer determines the voltage blocking capability of the device. On
the other side of n+ substrate, a metal layer is deposited to form the drain terminal. Now p
regions are diffused in the epitaxially grown n- layer. Further, n+ regions are diffused in p
regions as shown. As before, Si0 2 layer is added, which is then etched so as to fit metallic
source and gate terminals. A power MOSFET actually consists of a parallel connection of
thousands of basic MOSFET cells on the same single chip of silicon.




dioxide ~:--~~--~~~~~~~l----f::

0 rift


n+ substrate
Curren t path

Fig. 2.15. Basic structure of an-channel DNIOS power MOSFET.

When gate circuit voltage is zero, and V DD is present, n- - p- junctions are reverse biased
and no current flo'ws from drain to source. When gate terminal is made positive with respect
to sou rce, an electric field is established and electrons form n-channel in the p- 'regions as
shown. So a current from drain to source is established as indicated by arrows. With gate
voltage increas ed , current I D also increases as expected. Length of n-channel can be controlled
and therefore on-resistance can be made low if short length is used for the channel .
f\n examination of the basic structure of n-channel DMOS power MOSFET (P MOSFET)
reveals th at a parasitic npn bipolar junction transistor exists between the source and dr ain
as :,hown in Fig. 2.16. The p body acts as the base, n+ layer as the emitter (or source) and n
layer as th e co llec tor (or drain) of this BJT. Since source is connected to both base and emitt er
of parasitic BJT) the saUTce short circuit both base and emitter. As a r esult, p ot enti al differ ence
between base and emitter of the par asitic BJT is zero and therefor e) BJT is ahvays in the
cut-off s Late.
"' A mi. . \lie of s ilicon ato ms and pe nta vale nt atoms, deposited on wa fer, forms a layer of n -type se micon ductor
on hea ted s urfa ce . This iayer is called epitaxial layer.

[Art. 2.6]

Power Semicon du ctor Diodes an d Transistors

Also, vertical travel from source to drain

indicates the existence of a parasitic diode as
shown on the right in Fig. 2 .16. The parasitic
diode , with source acting as anode and drain
as cathode may be used in half-bridge or
full-bridge rectifiers. The parasitic diode also
shows that reverse voltage blocking capability
of PMOSFET is almost zero . This in-built
diode is an advantage in inverter circuits.
In Fig. 2.15, source is negative and drain
is positive. Therefore, electrons flow from



So urce

np n
BJT ~1==='=-----~__.-/



source to n+ layer, then through n-channel of Fig. 2.16. PMOSFET showing parasitic BJT and
p layer and further through n- and n+ layers

parasitic diode .

to drain. The current must flow opposite to the

flow of electrons as indicated in Fig. 2.15. Since the conduction of current is due to the
movement of electrons only, PMOSFET is a majority carri er device. Hence, time delays caused
by removal or recombination of minority carriers are eliminated during the turn-off process of
this device. PMOSFET with a turn-off time of 100 ns are avaiiable. Owingto its low turn-off
time, PMOSFET can be operated in a frequency r ange of 1 to 10 MHz.
2.6.1. PMOSFET Characteristics
The static characteristics of power MOSFET are now described briefly. The basic circuit
diagram for n-channel PMOSFET is shown in Fig. 2.17 where voltage and currents are as
indicated. The source terminal S is t aken as common termin al, as usual, betw een the input
and output of a MOSFET.



--=- "D O


8 10 12


(6 )


Fi g. 2.17. N -chan n el po\ver l'vIOS FET


circ uit di agram and (6) its typical transfer char cteristic.

Characteristics. This characteri tic shows the variation of drah'1 current I D as a

function of gate- source voltage Ves. Fig. 2.17 (b) shows typical transfer characteristics for nchannel
PMOSFET. Threshold voltage VeST is an important parameter of MOSFET. T'v'GST is the minilmun
positive voltage between gate and source to induce n-channel. Thus, for threshold voltage below
VCST' device is in the off-state . Mf\gnitude of VCST is of the order of 2 to 3 V
(6 Ou tp u t Chara cterlstics. PMOSFET outpu t characteri: tics, sh own in Fig. 2.18 (a ),
indicate the varia tion of drain current ID as a fu ncti on of ,h ain-sour ce vol tage V DS' wi b
crat e-soUf.e voltage Ves as a p aram e ter. For low values of Vns . t h~ gr aph be tyveen [D - 1/D: 15
a lm ost 'linear ; hi.:) in dica tes a constan t value of on -resi.'5t ance R DS := V DS' /]D For bven \- ,5, if
(a. ) Transfer


P ower Elec tronics

[ r t . 2.6]

dram current is nearly

constant. A load line intersects the output characteristics at A and B. Here A indicates fully- on
condition and B fully-off state. PMOSFET operates as a switch either at A or at B just like a BJT.
When power MOSFET is driven with large gate-source voltage, MOSFET is turned on,
V DSON is small. Here, the MOSFET acting as a closed s\lvitch, is said to be driven into ohmic
region (called saturation region in BJT). When device turns on, PMOSFET traverses
i D - V DS characteristics from cut-off, to active region and then to the ohmic region, Fig. 2.18
(a) . When PMOSFET turns off, it takes backward journey from ohmic region to cut-off state.

V DS is increased, output characteristic is relatively flat, indicating that

~ Active ~



--~'----------~--~ ~





~--~~--------~I ~







tdn~ tr ~






w hen VGS


.,i' ~----,------H...
, ,


Drai -source volta ge _

Cut- off



F ig. 2. 18. (a) Output characteristics of


~ :

, :

Fig. 2.18. (b) Switching waveforms for


. . (c) S wit ching characteristics. The switching characteristics

influ~nce d to a large extent by the- internal capacitance of the

of a power MOSFET are

device and the internal
impedance of the gate drive circuit. At turn-on, there is an ini.tial delay tdn during which input
capacitance charges to gate threshold voltage V GST . Here tdn is called turn-on delay time.
There is further delay t r , called rise time, during which gate voltage rises to V GSP, a voltage
sufficient to drive the MOSFET into on state. During tr,drain current rises from zero to full-on
current I D . Thus , the total turn-on-time is t on = tdn + t r . The turn-on time can be reduced by
using low-impedance gate-drive source .
As MOSFET is a majority carrier device, t rn-off process is initiated soon after removal
of gat e voltage at time t 1 . The turn-off delay time, t df, is the time during which input capacitance
discharges fro m overdrive gate voltage VI to V osp. The fall time, tf' is the time during which
input capa . tan c discharges from V osp to threshold voltage. During t f l drain current falls from
ID to zero. So when V Gs ::; V GST , PMOSFET tum-off is complete. Switching waveforms for a
power MOSFET are shown in Fig. 2.18 (6).

2.6.2. P 10 SFET Applica ti ons

The on -resistance of MOSFET . cre ase wit h volt age rating; this makes the device very
lossy at high-culTent applications. Since the 01 -resi tance has positive t emperature coefficient,
pa aIlel operat' on of PMOSFETs i r latively easy. The positive t emperature co.efficient also
reduces the second breakdo ;vn effect ' PMOSFETs.

P ower Semi on duct or D iodes at;lcl Transistors

[Ar t. 2.7J


PM OSFETs find applica tions in high-fr equency switching applications , varying from a fev{
watts to few kWs . The device is very popular in switched-mode power s upplies an d inverters .
These are , at present available with 500 V, 140 A ratings.

2.6.3. Comparison of P MOSFET with BJ T

The three terminals in a PM OSFET are designated as gate, source and drain. In a BJT,

the conesponding three term inals are base, emitter and collector. A PMOSFET has sever al
features different from those of BJT. These are outlined below :
(i) BJT is a bipolar device whereas P MOSFET is a unipolar device.
(ii) A PMOSFET has high input impedance (mega ohm) whereas input impedance of
BJT is low (a few k ilo-ohm).
(iii ) PMOSFET has low er switching losses but its on-resistance and conduction loss es
are more. A BJT has higher switching losses but lower conduction loss. So, at high
frequency applications, PMOSFET is the obvious choice. But at lower operating
frequencies Cless than about 10 to 20 kHz ), BJT is superior.
(iu) P MOSFET is voltage controlled device whereas BJT is current controlled device,
(v) PMOSFET has positive temperature coefficient for resistance. Th is makes parallel
operation of PMOSFETs easy. If a PMOSFET shares increased Cl1Trent initially, it
heats up faster, it resistance rises and this increased resistance causes this current
to shift to other devic es in parallel. A BJT has negative temperature coefficient, so
current sharing resistor s are necessary during parallel operation of BJTs.
(vi) In PMOSFETs , secon dary breakdown does not occu. , because it has positive
t em perature coefficien t . As BJT has negative temperature coefficient, sec ond ar y
breakdown does occm , I n BJT, with decrease in resistance with ris e in temperature ,
the curren t'increa~ s. This increased current over the s ame a rea results in hot sp ots
and breakdown of the BJT.
(v ii) PMOSFETs in high er voltage ratings have more conduction loss,
(viii) The state of the art PM OSFETs are available with ratings upon 500 V, 140 A
whereas BJTs are avai1 able wi th ratings upto 12 00 V, 800 A.


IGBT has been developed by combining into it the best qualities of both BJT and PMOSFET.
Thus an IGBT possesses high input impedance like a PMOSFET 3."'1d has low on-state power loss
as in a BJT. Further, IGBT is fre e from second breakdowll problem present in BJT. All these m erits
have made IGBT very popular amongst power-electronics engineers. IGBT is also known as me tal
oxide insulated gate transistor (MOSI GT), conductively-modulated field effect transistor (COMFET)
or gain-modulated FET (GEMFET). I " was also initially called insulated gate transistor (IGT ),

2.7.1. Basi Struct ure

Tg, 2.19 illustrates the basic structure of an IGBT. It is constructed vir tu ally in the sam e
manner as a pO 'vver MOSF ET. There is, however, a m ajor differ ence in the substr ate, The n+
1ayer su bstrate at t he drain in a P MOSFET is now sub tituted in th e I GBT by a p+ layer
substr ate called collector C. Like a pow er MOSFET, an IGBT has also thou sands of bas 'c
SITucture cells con ected approp. 'a ely on a single chip of siheon.
In IGBT, p + substrate is called injection layer because it injects holes into n- 1 yer. The n,- layer
is called drift r egion. As in other semiconductor devices, thickness ,)f n- bYE:T clet 'rmine_ the voltage
blocking capability ofIGBT. The p layer is called body ofIGBT The T' - lay~r in bet-.veen p'" and p
regions serves to accommodat e the dE'pletion layer of pn - jlU'lction i.e. j1.llction J 2 ,


Power Electronics

[Art.. 2.7]





Drift layer


Injection laye r


p + substrate


Current path


Fig. 2.19. Basic structure of an insulated gate bipolar transistor (lGBT)

2.7.2. Equivalent Circ~i~

An examination of Fig. 2.19 reveals that if we move vertically up from collector to emitter,
we come acrossp+, n;"',p layers. Thus, IGBT calf'b-e thought of as the combination of MOSFET
and p+n-p transistor Q1 as shown in Fig. 2.20 (b). Here Rd is resistance offered by n- drift
region, Fig, 2,20 (b) gives an approximate equivalent circuit of an 1GBT,

Dr ift region
resistance, RD



p+ substrate

( b)


Mai n current

,..--p_O_o----:!1---'----- -

-p~-- l

~Pa r osi\ic

: thyr ist r


\lh l
j-(p- _Od y

I resI5 .QI)Ce )


,L _____________________ I

(e )


Fig. 2.20. IGBT (a) basic s tructure sho.. .ving par asiti c tr ansistors and tbyristor (b) a pproxima te
equivaler t circuit (c) exact equivalent circui t a .d (d ) ci cui t symbol.

Power S emiconductor Diodes and Tra nsistors

[. rt. l:-7]

Fig. 2.20 (a) also shows the existence of another path from collector to emitter; thi.s path is
collector, p+, n-,p (n-channel), n+ a..'1d emitter. There is, mus, another inherent transistor Q2 as
n-pn+ in the structure ofIGBT as shown in Fig. 2.20(a ). The interconnection between two transi tors
QI and Q2 is shown in Fig. 2.20 (c ). This figure gives the complet e equivalent circuit of all IGBT.
H ere Rby is the resistance offered by p region to the flow of hole current I h
The two transistor equivalent circuit shown in Fig. 2.20 (c) illustrates that an IGBT
st ructure has a parasitic thyristor in it. Parasitic thyristor is also shown dotted in Fig. 2.20
(a). Fig. 2.20 (d ) gives the circuit symbol of an IGBT.
2.7.3. Working
When collector is made positive with respect to emitter, IGBT gets forward biased. With
no voltage between gate and emitter, two junctions between n- region andp r egion (i.e. junction
J 2 ) aTe reverse bias ed; so no current flows from colledor to emitter, Fig. 2.19.
When gate is made positive with respect to emitter by voltage V0, with gate-emitter voltage
more than the threshold voltage V OET of IGBT, an n-ch annel or inversion layer, is formed in
the upper part of p region just beneath the gate, as in PMOSFET, Fig. 2.19. This n-channel
short-circuits the n- region with n+ emitter regions . Electrons from the n+ emitter begin to flow
to n- drift region through n-channel. As IGBT is forward biased with collector positive and
emitter negative, p+' collector region injects holes i1).~o n - drift region. In short, n - drift region
is flooded with electrons from p-body region and holes from p+ collector region . With this, the
injection carrier density in n- drift region increases considerably and as a result, conductivity
of n - region enhances significantly. Ther efore, IG BT gets t urned on and begins to conduct
fo rward current Ie.
Current Ie, or IE , consists of two cur rent compon ents: (i) h ole current Ih due to inj ected
h oles flowin g fr om collect or, p+n-p transistor Q I' p-body region resistance R by and emitt er and
(ii) electr onic current Ie due to injected electrons flowin g fr om collector, injection layer p +, drift
region n-, n-ch annel r esistance R ch n+ and emitter. This me n s that collector, or load, current
Ie = emitter current IE =lh + Ie.
Maj or compon ent of collector CUlTent is electronic current Ie' i. e. main cw'rent path fOT
collector, or load, current is throu gh p+, n-, drift resistan ce Rd an d n-channel resistance R eh as
shown in Fig. 2.20 (c ). Th erefore, t h e voltage drop in IGBT in its on-state is
V eE .on = Ie Rch + Ie R d + Vj I
=Volt age drop [in n- channel + across drift in n- r egion
+ across forward biased p +n - junction J 11
Here Vj l is usually 0.7 to 1 V as in a p-n diode. Th e voltage drop Ie . R eI! is du e to n- channel
resistance, almos t the same as in a PMOSFET. The voltage dr op Vd{= Ie ,Rd in I GBT is much
less than that in PMOSFET. It1.s du e t o substantial increase in t...~e conductivity caused by
inj ec tion of electron ~ and h oles in n- drift r egion. Th e conductivi ty 'ncrease is th e main reason
for low on-stat e voltage drop in I GBT than it is in PMOSFET
2.7.4. Latch-up in IGBT
I t is seen from Fig. 2.20 (a) an d (c ) that I GBT structure has two inheren t transistors Q1
and Q2' which constitute a parasitic thyrist or. When IGBT i.~ on, th e hole-current flows through
transistor p+n-p andp-body esiscance R by ' Ifload curren tic is large, h ole component of ~ urrcnt
I" would also b large . This larg current WOL1ld increa e the voltage drop Ih . R by which IT ay


[Art. 2.7J

forw ard bias the base p -emitter n + junction of transistor Q2' As a consequence, parasitic
transistor Q 2 gets turned on which further facilitates in the turn-on of parasitic transistor
p~n-p labelled

The parasitic thyristor, consisting of Q 1 and Q2' eventually latches on

through regenerative action , wh en sum of their current gains C'J. 1 + C'J.2 reaches unity as in a
conventional thyristor (discussed in chapter 4). With parasitic thyristor on, IGBT latches up
and after this, collector emitter current is no longer under the control of gate terminal. The
only way now to turn-off the latched up IGBT is by forced commutation of current as is done
in a conventional thyristor. If this latch up is not aborted quickly, excessive power dissipation
may destroy the IGBT. The latch up discussed here occurs when the collector current ICE
exceeds a certain critical value. The device manufactures always specify the maximum
permissible value of load current ICE that IGBT can handle without latch up.
At present, several modifications in the fabrication techniques are listed in the literature
which are used to avoid latch-up in IGBTs. As such, latch-up free IGBTs are available.

2.7.5. IGBT Characteristics

The circuit of Fig. 2.21 (a) shows the various parameters pertaining to IGBT characteristics.
Static I-V or output characteristics of an IGBT (n-channel type ) show the plot of collector
current Ie versus conector~emitter voltage V CE for various values of gate-emitter voltages
V CE1 , V CE2 etc. These characteristics are shown in Fig. 2.21 (b). In the forward direction, the
shape of the output characteristics is similar to that ofBJT. But here the controlling parameter
is gate-emitter voltage VCE because IGBT is a voltage-controlled device. When the device is
off, junction J 2 blocks forward voltage and in case reverse voltage appears across collector and
emitter, junction J 1 blocks it. In Fig. 2.21 (b), V RM is the m aximum reverse breakdown voltage.
The transfer characteristic of an IGBT is a plot of collector current Ie versus gate-emitter
voltage VCE as shown in Fig . 2.21 (c) . This characteristic is identical to that of power MOSFET.
When VCE is less than the threshold voltage V OET ' IGBT is in the off-state.
( A)


VCE I. > VCEJ et c




V~E 1

(a )









'1 CE I



F ig. 2.21. 1GBT

( A)







( b)

circuit diagram (b) static I-V ch aracteristics and



transfer characteris tics.

2.7.6. Swit ching Ch a r acteristics

. S witch ing char a cteristics of an I GBT during turn-on an d turn-off are sketched in Fig. 2.22.
The turn-on tim e is defined as the time bet\oveen th e i 3 ~an s of forwa rd blocking to fo nv ard
011- 5 ate (7 ). Turn-o
time is compos ed of delay time tdT'. and ris e t ime tr> i.e. tOT'. = tdn + t ,.. The
delay tim e is d efined as the t ime for the collector- emitter voltage t o fall fr om VCE to 0.9 V CE '
Here V CE is the initial collector-emitter voltage. Time t d n may also be defined as t he t ime for th e

[Art. 2.7]

Power Semicon du tor Djodes and Transistors

~I tdl

veE. ic






















F ig. 2.22. IGBT turn-on and turn-off characteristics.

collector current t o rise fro m its initial leakage current ICE to 0.1 I c , Here I c is the fin al value
of collector current .
The rise time tr is t he tim e during wh ich collector -emitter voltagefalls from 0.9 VCE t o 0.1
VCE oIt is a.ls o defined as th e t ime for the collector curren t to rise from 0.1 Ic t o its fin al value
Ic. After time to"" the collector cur rent is Ic and the collector;emitter voltage falls to small value
called conduction drop = V CES where subscript S denotes saturated value.
The turn-offtim e is somewhat complex . It consists oft hr ee intervals: (i ) delay time, t d{ (i i )
initial fall tim e, tr1 an d (iii) fin al fall time, tj2 ; i. e. t off =td{+ tn + tj2 .The delay time is the time
during which gate voltage falls from VGE to thresh old voltage V GET, As VGE falls to VGET dur ing
t df, the collector current falls fro~ Ie to 0.9 Ie. At the end of tdf, collector-emitter voltage begins
to rise. The first fall tim e tfl is defined as the time during which collect or cu rrent falls from 90
t o 20% of its initial value I c , or the time during which coll ector -emitter voltage rises fr om
V eEs to 0. 1 V CE'
The final fall time tf2 is the tim e during which collect or current falls from 20 t o 10% of c,
or the tim e uring which collector-emitt er voltage rises fr om 0.1 VCE t o fin al value VCE. see Fig.

2.7.7. Appli cation of IGBTs

IGBTs are widely used in medium power applicat ions su ch ,s dc and ac m otor drives, UPS
system s , poweT supplies and drives for s olenoids, rel ays and contactors. Though IGBTs are
somewhat mor e exp en siv e t han BJTs, yet t h ey aTe be com ing p op lar because of low er
gate-drive requirem ents , lower swit ching losses and smal er ;:,nubber circuit r equirem ents.
IGBT converters are m ore effi ci ent with less size as well as cost, as compar ed to converters
based on BJTs . Recen t ly, I GBT inverter induction-motor drives u sing 15-20 kHz s\vitching
frequ en cy are findin g favour whe.e audio-n oise is obj ectionable. In most applic ations, IGBTs


P ower Electronics

rArt. 2.8J

will eventually push out BJTs. At present, the state of the art I GBTs of 1200 V, 500 A ratings,
0.25 to 20 flS turn-off time with operating frequency upto 50 KHz are available.

2.7.8. Comparison of IGBT with MOSFET

The relative merits and demerits of IGBT over PMOSFET are enumerated below:

(i) In PMOSFET, the three terminals are called gate, source, drain whereas the
corresponding terminals for IGBT are gate, emitter and collector.
(ii) Both IGBT and PMOSFET possess high input impedance.
(iii) Both are voltage-controlled devices .
(iv) With rise in temperature, the increase in on-state resistance in PMOSFET is much
pronounced than it is in IGBT. So, on-state voltage drop and losses rise rapidly in
PM OSFET than in IGBT, with rise in temperature.
(v) With rise in voltage rating, the increment in on-state voltage drop is more dominant
in PMOSFET than it is in IGBT. This means IGBTs can be designed for higher
voltage ratings than PMOSFETs.
In view of the above comparison, (a) PMOSFETs are available upto about 500 V, 140 A
ratings whereas state of the art IGBTs have 1200 V, 500 A ratings and eb) operating frequency
in PMOSFETs is upto about 1 MHz whereas its value is upto about 50 kHz in IGBTs .


SIT is a high-power high-frequency semicondu ctor device. It is the solid-state version of
triode vacuum tube. SIT was commercially introdu ced by Tokin Corporation of Japan in 1987.
Basic structure of SIT is shown in Fig. 2.23 (a) and its symbol in Fig. 2.23 (b) .. It is basically
n+n-n device with a buried grid-like p+ gate structure. It h as short n-channel structure and
p'" gate electrodes are buried in n-n epi-layers as shown. The buried gate structure gives lower
gate-source chann el resistance, lower gate source capacit ance and lower thermal resistance.
SIT is normally an on device, i.e. if V GS = 0 with V DS present, electrons (majority carriers)
would flow from source s to n, pass through gate .0+ electrodes and would then continue their
Electron f low ~

--- -----------1
Source (5)

c urren t ftow

( b)

Fig, 2.2 3. (a) B si c structure of SIT

(b )

device symbol

Power Semiconduc tor Diodes and Transistors -

tAr t. 2.91


journey t hrough n-, n+ and reach drain as shown in Fig. 2. 23. (a ). The drain cu rrent 1D would
flow fr om D to S as shown. If Ves is negative, p+n j un ction s get reverse biased. As a result,
depletion layer is form ed around p + electrodes and this r educes the current flow from its value
when Ves = 0, Fig. 2. 24 (a). At some higher value of everse bias voltage V es , the deplet.ion
layer would grow to such an extent as to cut-off th~ channel completely, F ig. 2.24 (b ) and load
current iD would, therefore, be zero.

__....-----c 5


Ip; T~)rl\i;J;r~'



I D\p\etl oiIOyer \


Vos ~------h-------~


( a)


F ig. 2.24. (a) Lower reverse bias, load current iD redu ced due to depletion layer
(b higher reverse bias, expande d depl etion layer stops current flow.

Although, the device conduction dr op is lower than that of equivalent s eri.es-parallel

oper atio of P MOSFETs , the essentially large drop in SIT makes it unsuitable for gener al
power -electronic applic ations . F or example, a 1500 V, 180 A SIT ha a channel resistance of
0.5 n giving 90 V conduction drop at 180 A. An equiv al ent thyristor or GTO drop may be
. ar oun d 2 V [11J. Though conduction drop in SIT is abn or m ally high, the turn-on and turn-off
tim es of the device are very low. For the SIT cited above, typicaf ton and tofr are 'rround
0.35 ).lS . High conduction drop associated with very low turn-on and turn-off times re~ult in
low on-off energy losses. Thus, SIT is being used in high-power, high-frequency applications
such as MlFM transmitters, induct ion heaters , high-voltage low-current power supplies,
ultrasonic generators etc. SITs with 1200 V, 300 A rating with ton and toff around 0.25 to
0. 35 JlS and 100 kHz operating frequency are available.
SIT is a m ajority carrier (electrons only) device, therefore SOA is limited by junction
temperature. As channel r esistance rises with tem per at ure, parallel operation of SITs is easy.
SIT is normally-on d evice, n ormally-off device is under development.

An M CT is a new device in the fi eld of semiconductor-controlled devices. Ie is basically a

thyristor wi h two 10 SFETs built in t o the gate structure. One MOS 'ET is us ed for turning
on the MGT and th e other for turning off the devic e. An MeT is a h igl -frequency, higb-power,
low- conduction drop switching device.
An MCT co bines into i t the fea tures of both con ventio. a1 fo ur-laye thyris tor a ving
reg en rative action and MOS-gate stru cture . However, in MeT, an cde is the r eference \-vith
r espect to which !?.II gate signals are applied. In a conventional SCR, cathod e is t h e re ~erence
t erminal for o-a te signals .


P owe r Electronics

[Art. 2.9]
Ano de

/:// /////////////1

Gate n








"- {On- FET

Chan nel or
p-Ch annel)

(Of1- FET Channel or n-Channel)

pp buffer


n+ substrate


Fig. 2.25. Basic structure of an MeT.

The basi c structure of an MeT cell is shown in Fig. 2.25. A practical MeT consist of
thousands of these basic cells connected in parallel, just like a PMOSFET (7, 8). This is done in
order to achieve a high-current carrying capacity of the device.
The equiva1ent circuit of MeT is shown in Fig. 2.26 (a) . It consists of one on-FET, one
off-FET and two transistors. One on-FET, a p-channel MOSFET and the other off-FET, an n
channel 1-,'IOSFET, represent MOS-gate structure of MeT. The npnp structure of MeT is
represented by two transist ors npn and pnp as shown in Fig. 2.26 (a). An arrow t owards the
gate t etminal in dicates n-channel MOSFET and the arrow away from the gate terminal as
the p-chann p-l IOSFET. The two transistors in the equivalent circuit indicate
that there is
regenerative fe edback in the MeT just as it is in an ordinary thyristor. Fig. 2.26 (b) gives the
circuit sym bol of an MeT.
An MeT is t urned-on by a negative voltage pulse at the gate with respect to the anode
and is turned-off by a positive voltage pulse. Working of MeT can be understood better by
referring to t he equivalent circuit of Fig. 2.26 (a) .

Turn-on P oce ss. As stated above, MeT is turned on by applying a negative voltage pulse
at the gate with r espect to anode. In other words, for turning-on MeT, gate is made negative
with respect t o anode by th e voltage pulse between gate and anode. Obviously, MeT must be
initially forward biased and then only a negative voltage be applied . With the application of this
negative vol tage pulse, on-FET (p-ch ann el) gets turned on whereas off-FET is already off. With
on-FET on , current begins t o flow fro m anode A, through on-PET and then as the base current
and emitter current 0 npn transistor and then to cathode K. This turns on npn transistor.
As a result, collector current begins to flow i n np n transistor. As off-FET is off, this collector
cu rre nt of npn "ransisto r a cts as the bas e curr ent of p np t ransistor. Subsequently, pnp
transistor is also t urne d on . On bo th t he tran s 'stors ar e on, regenera ti ve action of the
connection sch eme takes place and the thyristor or Me T is turned on .
Note that on-FET and p np transistor are in parallel when MeT is in conduction state.
During the time MeT is on, base cur ent 0 np n transistor fl ows m ainly through p np t ransistor
because of i ts better conducting property.

(Ar t. 2.101

Power Semic on duc tor Diodes and T r ansjstor

;---~_ _ _ __

A [nOde



+ 14 '/


, fOIT1



Off - FET


-7 V

(p- channel)

np n



Fig. 2.26. MeT



equivalent circuit and


circuit symbo~

Turnoff process. For turning-off the MCT, off-FET (or n-channel MOSFErl') is energized
by positive voltage pulse at the gate. With the application of positive voltage pu 1 :e, off-FET is
tu rned on an d on-FET is turned off. After off-FET is turneq on, emitter-base terminals of
pnp transistor are shor t circuited by off-FET. So now anode current begins to flow through
off-FET and therefore base current of pnp transistor begins to .decrease. Further, collector
current of pnp transistor that forms the base current of npn transistor also begins to decrease.
As a consequence, base currents of both pnp and npn transistors, now devoid of stored char ge
in their n and p bases respectively, begin to decay. This regenerative action eventually t urns
off the MCT.
An MCT h as the fcllovving merits:
(i) low fo rward conduction drop,
(ii) fast turn-on and turn-off times,
(iii) low switching losses and
(iu) high gate input impedance, which allows simpler design of drive circ uits.
Main disadvantage of MCT is its low r everse voltage blocking capability.
MeT was commercially introduced in 1992. At that time, it was predicted that its use as
a power -semiconductor device would be so vast that it might challenge the existence of most
of the other devices like SCR, BJT, GTO, IGBT etc. This has, however, not :happened because
an MeT has (i) limited reverse-biased SOA and (ii) its switching frequency is much inferior
to IGBT. At present, MCTs are being promoted for' their use in soft switched converter
topologies, where these inferiorities do not inhibit their use.
2.10~ NEW

At pr es ent, silicon enjoys - monopoly as a semicor.ductor material fo r the commercial
pr oduc ti on of p ower-control dev ices. This is becau se silic on is che aply available an d
semic ondu ctor devices of any size can be easily fabricated on a ingle silicon chip. There ar e,
however, new types of materials lik e gallium assenic (C;'al\s), silicon carbide and diam ond which
possess the desir able properties required for swit ching devices. At pTe en t, st ate-of-the-art
techn ology for these materials is primitive ompar ed with silicon, and many more years of
research invest ment are required befor e thes0 materials become commerc ' ally viable for t he
production of power-controlled devices. Diode, power MOSFET and thyristor made from silicon
carbide have be en established in the labora tory a d are expected to be commercially available
very soon. Superconductive materiah may also b<:: used in the manufa cture of such devic es.
but w .rk in this direc:tion has not yo.t been r eported.


Power Electronics

[pro b. 2]

G rm anium is not used li n the fabrication of thyristors because of the following reasons:
(i) Germanium

has much lower thermal conductivity; its thermal resistance is, there
fore, more . As a c~nsequence, germanium thyristors suffer from more losses, more
temperature rise and therefore lower operating life.
(ii) Its breakdown voltage is much less than that of silicon. It means that germanium
thyristor can be built for small voltage ratings only.
(ii i) Germanium is much costlier. than silicon.

2.1. (a) Why are semiconductor materials designated as p+, p-, n-, n+? Explain.
(b) What is p-n junction? Discuss the formation of depletion layer in p-n junction.
(e) What is barrier potential? How are depletion layer and barrier potential effected by
2.2. (al Explain the effect of forward bias and reverse bias on the depletion layer in a p-n junction.
(b) How is the magnitude of breakdown voltage effected if a junction has highly doped (i)
layers on its both sides and (ii) layer on its one side only.
(e) Describe the structural features of power diodes. How do these differ from signal diodes?
. 2.3. (a) What is a diode? Discuss i-v characteristics of power, signal and ideal diodes.
(6) Describe reverse recovery characteristics of diodes. Show that reverse recovery time and
peak in'!erse current are dependent upon storage charge j:md rate of change of current.
;'-2.4. (a) Describe the various types of power diodes indicating clearly the differences amongst

(b ) What is cut-in voltage in a diode? What are other terms used for cut in voltage?
(e) Discuss the following tenus for diodes:

Softne ss fa cto r, PlY, reverse recovery ti me, reverse recovery current.

(d) For a power diode, the reverse recovery time is 3.9 ,us and the rate of diode-current decay
is 50 A /~ s. For a softness factor of 0.3, calculate the peak inverse current and the storage
[Ans. (d) 150 A, 292.5 ~Cl
2.5. (al Discuss th e power loss in a diode during the reverse recovery transients.
(b) The forward characteristic of a power diode can be represented by vf= 0.88 + 0.015 if'
Determine the average power loss and nus current for a constant current of 50 A for
2/ 3 of a cycle.
1 2 T/ 3
Hint. (6) With T as the time of a cycle, average power loss =T 0
vf' lf db '3' ufIf etc

[Ans. (6) 54.33 W, 40.825 A)

2.6 . (a) Enumerate the types of power transis tor s along with their circuit symbols .
(b) Wh at

is a bipola r junction transistor? 'vVhy is it so called?

Describe t he types of BJTs with their circuit symbols.

(c) De me a an d ~ for BJT and develop a relati on between the two . Why is a less than 1 and
~ m OTe t han 1 ?
(d ) Why is it preferrable to use hard drive for BJ T ?

2.7 . (c'z ) vVhat is th e difference bet we en ~ an orced ~f for BJTs ?

(6) Whflt are th e conditio ns un der which a transistor operates as a switch?
Discuss hard -drive and over drive fact or for BJT.
(c ) Show L at colle ctor current at sat urati on remains s ubstan tially con3:ant even if base
curren t is i n ~r ea sed.

(Prob. 2]

Powe Semiconduc tor Diodes and Transistors


2.8. A bipolar tr ansistor, with current gain P= 50, has load eBistance Re = 10 Q , de supply voltage
Vee = 120 V and input voltage to base circuit, VB = 10 V. For VeEs = 1.2 V and VBES = 1.6 V,
(a ) the value of RB for operation in the saturated state
( b ) the value of RB for an or drive factor 6
(e) forced current gain and
(d ) power loss in t he tr ansistor for both parts (a ) and (b) .
[Ans. (a) 35.354 n ( b ) 5.892 n (c ) 8.33 (d ) 14.6362 W, 16.537 W]

2.9. (a) Explain the s.witching performance of BJT with relevant waveforms. Indicate clearly
turn-on and turn-off times and their components.

(b ) Describe

2.10. (a) Describe the input and output charac

teristic for a BJT. Show the region of the
'. t ransistor characteris tic where it acts
like a switch. .
( b) Typical switching waveforms for a power
transistor are shown in Fig. 2 .27. Show
that switch-on energy loss is given by
Vee l es

Also obtain an expression for the

average value of switch-on loss.
(c) Derive expr essions for the switch-off
ener gy loss and also for its average value
fo r the wa veforms shown in Fig. 2.27 ..
[ Ans (b)











tton...! . "


F ig. 2.27. Pertaining to Frob. 2.10

(b ).

Vee ' I es .
Vee ' I es
Vee ' I es f . ]
f to.. (c) . ?".. ..:' t off,

2.11. In case les =80 A, Vee = 220 V; ton =1.5 ~s and t off = 4 ~s for the switching wavefor ms shown
in Fig. 2.27, find the energy loss during switch-on and swi tch-off intervals. Find als o the
aver age power loss in t he power tr ansistor for a switching frequency of 2 kHz.
Derive the expression5 s-ed.
[Ans 4 .4 mWs, 11 .73 mWs, 32 .267 Wl

2.12. ( a ) For the typical switching waveforms shown in Fig. 2.27 for a power transis tor, find
expressions that give peak ins tantaneous power loss during ton and t off intervals respec
(b) In case I es = 80 A, Vee = 220 V, ton = 1.5 !-LS and tofr = 4 !lS, find the peak value of instan
t ane ous power loss during t 0 1'. and torr intervals re~ pectively.
(Ans. (a )

Ies , Vee Ies ' Vee

(b ) 44 00 "V, 44 00 Wj

2.13. A power transistor is used as a switch and typical waveforms ar e shown in Fig . 2.1 2(a ) Th e
par ameters for the tran"Tht or circuit are as under :

Vee = 2 00 V, VeEs =2.5 V, les"" 60 A, t d = 0.5 ).ls, tr = 1 ~s ,

tn = 40 !ls, ts = 4 )l s, tf "" 3 ).lS, to = 30 )lS, f= 10 kHz.

Coll ector to emi tt er lea age current"" 1.5 JmA.

De termine ave ra ge pow er loss du e to collector current durin t on and t n . F in also t e p eeL

instant aneous power 1053 du e to collector curren t eluring turn-on tim e.

Sket ch t h e ins a tan eo u power loss duri g t on an d t n' (Ans. 20 .50 15 W, 60 Ii , 3037.97 Wl


Power Electronks

2.14. Repeat Prob. 2.13 for obtaining average
power loss during turn-off time and offperiod, and also peak instantaneous
power loss during fall time due to col
lector current.
Sketch the instantaneous power loss

during turn-off time and off-period.

2.15. Fig. 2.28 shows the switching charac

teristics for a power semiconductor
device . Derive the expressions for ener
gy loss during turn-on and turn-off
periods, and also f.or the average
Fig. 2.28. Pertaining to Prob. 2.15 .
switching loss. Sk.etch the variation of
power loss during turn-on and turn-off periods.
For Vs = 220 V, Ia = 10 A, tl = 1 ~s, t2 = 2 ~s, t3 = 1.5 ~s and t4 = 3 ~s , find t.he average value of
power-switching loss in the device for a switching frequency of 1 kHz.

. [AnS. ~ Vs . Ia


+ t2],

~ Vs Ia


+ t 4),

Vs Ia {(ton + torr),


2.16. (a ) Explain the constructional details and working of low-power MO~FET and power MOS
FET and bring out the differences between the two.
(b) Discuss the transfer and output characteristics of power MOSFETs .
2.17. (a ) Describe the switching characteristics of power MOSFETs.
(b) Compare power MOSFETs with BJTs .
2.18. (a ) Discuss how conduction takes place in PMOSFET of n-channel type.
(b) Explai n the formatio~ of parasitic BJT and paras itic diode in a PlY~QSFET. Can parasi tic
diode be used in some power electronic applications ?
2.19. (a ) What is an IGBT ? What are its other nam es ? Give it" basic st ructural features. How
does it differ in stnicture from PMOSFET ?
(b ) Derive the approximate and exact equivalent circu its of an IGBT .from its structural
details . Also describe its output and transfer characteristics.
2.20. (a) Describe the working of an IGBT. How does latch-up occur in an IGBT ?
(b ) Give a comparison between an IGBT and a PlvIOSFET.
2.21. (a) Explain switching characteristics of an IGBT.
(b) Discuss why PMOSFET has no reverse blocking voltage whereas an IGBT has.
(c) Why a re IGBTs becoming popular in their applications to controlled converters ?
Enum erat e some applications of IGBTs .
2.22. (a) What is SIT? Give its basic structural details . E plain its working with relevant diagra ms .
(b ) Tho ugh SIT is not suitable for general power -electronic applications , yet it is being used
in som e specific applications. Explain .
2.23. (n ) Descri be t he basic structure of MOS controlled thyristor (M e T). Give its equiv a lent circu it
and e plain t h e turn-on and turn-off processes.
(b) Giv e the m er its and demerits of MCTs . In what type: o. _Appli cations are MCTs being
promoted at present?
(c) Discuss briefly about the new semicond ucting m ateri als .
2.24. Deduce to show t hat the energy loss during turn-on of a power t ransistor is given by
(VI/6 ) T joules, wher e V :::: off-sta t e voltage, 1 = on-s tate current and T = turn-on time. Assume
t he change of V and I to be li near over T.
H ence, ca lculate t he tur n-on loss of a power transistor for which the voltage an d current,
durin '" the process of t urn-on, chang e linearly from 300 V to zeTO V a nd zero A t o 200 A
respectively in 2 j.l.S.
[Ans. 10 m W -3 or 10 x 10- 3 J ]

Power Semiconductor Diodes and Transistors

[Prob. 2J


2.2 5. Read the following statements carefully and indicate the power semiconductor device' eac

statement represents .
(a ) two-terminal three-layer device
( b ) majority carrier devices
(c) bipolar devices
(d) negative pulse turn-on device
(e) on operation in ohmic region
(f) normally on device
(g) on-state in saturation region
(h) two-terminal two-layer device
(i) uncontrolled turn-on and turn-off device
(j) controlled turn-on and turn-off devices.
[Ans. (a) Power diode (b) PMOSFET, SIT (c) Diode, BJT, IGBT, MeT (d) Me T
(e) PMOSFET (f) SIT (g) BJT (h) signal diode (i ) Diode

Chapt r 3


Diode. Circuit s and

........ ..


......... ...... .................. ..... ........... ...............


In this Chapter

Diode Circuits with DC Source

Freewheeling Diodes
Diode and L Circuit
Recovery of Trapped Energy
Single-phase Diode Rectifiers
Zener Diodes
Performance Parameters
Comparison of Single-phase Diode Rectifiers
Three-phase Rectifi8rs

................ . ................................... ....

.. .. . ........................ ........

A rectifier is a circuit that converts ac input voltage to de output voltage. Semiconductor

diodes are used extensively in power electronic circuit s for the conversion of power from ac t o
dc. A rectifier employing diodes is called an uncontrolled rectifier, because its average output
voltage is a fixed dc voltage.

In t his chap ter, first diode circuits involving different combinations of R, L and C are
studied , and then diode rectifiers are describe ror simplicity, the diodes are considered as
ideal swi tches. An ideal diode has no forwar a'volt a ge drop and reverse recovery time is


In this section, the effect of switching a de source to a circuit consisting of diode and
different circuit parameters is exa mined. Th e conclusions arrived at can then be applied to
similar situations encountered l ater in power-electronic circuits.
3 .1.1. Resistive Load


In the circuit of Fig. 3.1 (a ), when switch S is closed, the current rises instantaneously to
V / R as shown in Fig. 3.1 (b). Here Vs is the dc source v oltage andR is t he loa d esist ance . 'When
switch S is opened at t 1, the current at once falls to zero, Fig. 3.1 (b) . Voltage Un across diode is
zero during th e tilne diode conducts and is equal t o + Vs after diode stops conducting.
3.1 .2.

Re Load

A circui t with de SOUTce, diode and RC load is sho'wn in Fig. 3.2 (0). Vlhen switch S is closed
a t t = 0, KV'L gives


[Art. 3.1]

Diode Circuits an d Rectifiers



Fig. 3.1. Diode circuit with



circuit diagram and



(b) waveforms.





1 !:------L_--=::::~_ _






Fig. 3.2. Diode circuit with RC load (a) circuit diagram and (b) waveforms.

. .
. .
1 "[1(S) q(O)] Vs
Its Laplace transform is R 1(s) + C +
'= .

As the initial voltage acros s C is zero, q(o)

.. :(3.1)

= O. With this, Eq. (3.1) becomes

.[ 1]

1(8) R + - =Vs




Its Laplace inver se is

The voltage ac ross capacit or is


(" .

. RC s + RC



_S .

)=-RVs ..

+ RC

e- tlRC

II .

uc(t) = C 0 .clt

... (3.2 )


= RC

=Vs (1- e- U RC )

II e


... (3.3a )


Power Electronic

[Art. 3.1]


where "t = RC is the time constant for RC circuit. From Eq. (S.Sa), initial rate of change of
capacitor voltage is given by


=[V~ . e- tiRe. RIC]



... (3.4)


source voltage, Vs
RC =---;----:----
(d v/dt)t=o
In Fig. 3.2 (b), current through the circuit and voltage variation across C are shown.
3.1 .3. RL Load
When switch S is closed at t = a in the RL and diode circuit of Fig. S.3 (a), KVL gives

Time constant,

R i +L

~~ = Vs

... (3.5)


Flg. 3.S. Diode circuit with RL load

circuit diagram and (b) waveforms.


With initial current in the inductor as zero, the solution of Eq. (3.5) gives
-!! t


= ~ (1 -

... (S.6)

L )

Initial rate of ris~ of current is



= (Vs
-e L

dt t = 0

The vol tage across L is

. For RL circuit, ;



=L -dt = V



_li t

...(3 .8)

='t is the time constant. The waveforms of current through the circuit and

voltage across indu ctance L are sketched in Fig. 3.3 (b ).

It mllilt be noted that the behaviour of circuits in Figs. 3.1 to 3.3 is not affected whether a
diode is used or not. It is because, in these circuits, the current i does not h ave a t endency to .
reverse; i.e . current i remains unid'rectibnal,


Diode Circuits and Rectifiers



3.1.4. LC Load
A diode circuit with dc source voltage V s' switch S and load LC is shown in Fig. 3.4 (a). When
switch S is closed at t = 0, the voltage equation governing its performance is given by

f 'dt = V

L di
dt + C

Its Laplace transform ~s


As the circuit is initially relaxed, i(O)

[ sc1]

= 0 and Vc


(0) = 0 or q(O) = C . vc(O)

L [8 /(s) _ i(O)] + l.l/(S) + q(O)


/(s) sL + - =Vs




/(s) =


Let we = ~L C . ThIS gIves /(s) = -L--'
. We

+ LC

S + We


= Vs . "'V T.L


S + We

. :' ~ -

. 1






t~= 'T{I Wo










(a )

F i g . 3.4. Diode circuit with L C load (a) circui t diagram and (b) wavefor~s .

Its La plac e inverse is



i(t) = Vs .

~ s'n wot

...(3 .9)

= ~lc is called reson ant fre quency of the circut. Capacitor voltage is given by



1 .t


= C 0 i (t) . dt = C J0 Vs' "'V L

=Vs (1 - cos CI), t )


Wo t .

... (3,10a)


Power Electronics

[A rt. .]

Voltage across inductance is given by

VL(t) =L

When Wo to = 1t or when to
. vc(to ) = 2 Vs and vL(t O) = - Vs
Here to


... (3.10b)

=Vs cos Wo t


from Eq . (3.9), i(to)

=1t/wo = conduction t ime of diod'e


and from Eq. (3. 1 0a),

= 1t ";LC

From Eq. (3.9), circuit or diode curr ent at to/2


attains a peak value of

Ip = Vs . ,jCI L as shown in Fig. 3.4(b). Voltage across diode, soon after diode stops conduction
at to is given by

Waveforms of i(t),


VL - Vc

+ Vs = 0 .- 2 Vs + V s =

- Vs


vL and VD are sketched in Fig. 3.4 (b). It is seen that at to/2 = - 2 ,


diode current reaches peak value, Vc = Vs and vL = O. Also at to =1t/wo =1t (";LC), diode current
decays to zero and capacitor is charged t o voltage 2Vs
' Soon after to, voltage across L is zero and
diode voltag,e VD = - VS'

'---' Example 3.1. For the circuit shown in Fig. 3.5 (aJ, the capacitor is initially charged to a
voltage Vo with upper plate positive. Sw itch S is closed at t = O. Derive expressions for the current
in the circuit and voltage across capacitor C. What is the peak value of diode current? Find also
the energy di ssip ated in the circuit.

S olution. When switch S is closed, KVL gives

R i+

~ f idt= 0

Its Laplac e transform , including the initial v()ltage across capacitor, is

R 1(8) + 1..

[!.ill _
CV o] =

1(s) [R + ~] = Vo
. V


Its s olution, as per Art. 3. 1.2, is



ti RC
R e-.






(a )
F ig. 3.5, P ertaining to Example 3.1 (a) circuit diagram and (b ) waveforms.

Diod e Circuits and Rectifiers

[Art. 3.1]


_ Vo

:. Peak diode current



v c(t) = C 0 ldt - Vo

Capacitor voltage,

~]t Vo . e-tiRC. dt _ Vo


= - V0 e

Current i(t) and voltage vc(t) are sketched in Fig. 3.5 (6).

/ n e r gy dissipated in the circuit =

~ CV~ Joules

Example 3.2. In the diode and LC network shown in Fig. 3.6 (a), the capacitor is initially
charged to voltage Vo with upper plate positive. Switch S is closed at t::; 0. Derive expressions
for current through and voltage across C.
Find the conduction time of diode, peak current through the diode and final steady-state
voltage across C in case Vs = 400 V, Vo = 100 V, L = 100 W1.a'ni:1~C = 30~ Determine also the
. ;.
voltage across diode after it stops conduction.
Souti on. When switch S is closed, KVL for Fig. 3.6 (-li}.gives - ,

l]'d t::; V

L di
dt + C

Its Laplace transform gives

L [s I(s) _ i(O)l + 1.. [I(S) + C V o]::; Vs


As 'nitially i(O) = 0, the above equation becomes

I(s) sL +

Ie] = Vs -s Vo

' -7I /2w o-:




vc 1C


Fig. 3.S. Pertaining to Exa mple 3.2


circuit diagram and (6) waveforms .


Power E lectronics

(Art . 3.1 ]

This equ ation in s-domain can be solved as in Art 3.l.4. Its solution is

= (Vs -

Va) .


1 (Vs - Va)

vc(t) = C


. (Dot dt + Va

(Vs - Va) (1- cos CDot) + Vo

vc(t) = Vo
vc(t) = Vs and at (Do t= 11:, vc(t) = 2 (Vs - V o) + Vo



t = 0,



= 11: / 2,

Diode conduction time,

Peak current through diode,



. .

=-(Do =11: ~LC =~30 x 100

=(Vs = 300


= 2Vs -


x 10- 6 = 54.77 ~s

~ 100

= 164.32 A

Steady state voltage across C occurs when (Dot a = n.

Voltage across diode, after it stops conducting, is given by



v L - Vc

+ V~ = 0 ~ (2Vs - V o) + Vs

=- Vs + Vo =- 400 + 100

300 V.


Example 3.3. In the circuit sho wn in Fig. 3.7 (aJ, the capacitor has i nitial voltage Va with
upper plate positive. The circuit i.s switched ed t = O. Derive expressions for Cllrrent and voltage
across capacitor. Find the conduction time for diode' and steady-state capCiCi tor voltage.
Sqlution. The voltage equation for the circuit of Fig. 3.7 (a), after s\vitch S is closed at
t = 0, is
di+ 1 ~'d t=O
Ldt C

~ 71: /






' _--







_ _

Fig. 3.7. Pertaining to Exampl e 3.3


circuit iagl'am and (6) waveform_ .


Diode CircuJts an d 'Re ctifiers


.1 ]


Its Lapl ac e transform, including initial voltage across capacitC?r, is "

1(s) . sL + 1. [1(S) _ CVo] = 0

l (s)


= Va

Here minus sign is put before Va, because for the direction of positive current flow, polarity
of Va is opposite.
Solution of above s-domain equation, from Art. 3.l.4, is

i(t) =


~ sin Wo t

L a sm wot dt -

vc(t) = C . Va

Voltage across C is
Diode conduction time,

to =~

Steady-state capacitor voltage

Voltage across diode,


Va' = - Va cos wot

= 1t ..JLC

=- Va cos 1t =+ Va
vn = - Va .

'. Waveforms for i, Vc arid Vn are sketched in Fig. 3.7 (b ).

. .. Example 3.4. In the circuit shown in Fig. 3.8 (a), capacitor is initially charged to voltage
V o with upper plate pos itive . Sketch waveforms ofi, Vc, VL and in after the switch S is closed.

Solution . When switch S is closed, capacit or C begins to discharge through Land C. For
:' obtaining i , Vc expressions, refer to Example 3.3.

i = Va


~ sin wot , vc(t ) = - Va cos ooot

t . dtd sm. wot = Va cos ooot = - vc'

vL = L dt = L . Va .



~ vocoswo t

+ Vo

. 10


{ o _..... tLlp
J...C V~ .. , 0


(a )
Fig. 3.S. P er tainin g to Exampl e 3.4 (a) circui . diagram and (6) wave for ms


Power Elec tron ics

[Art. 3.1]

The waveforms for i, vc, uL and iD are shown in Fig. 3.8

i CV6

and i = O. At t =T/ 4, current i reaches peak value Vo

stored in C gets transferred to L as

( b) .

At t

=0, energy stored in Cis

~ = ]p'

Vc = vL

and energy

~ L];. Soon after T 14 (CDo T 14 =1t/ 2), as VL tends to reverse,

diode D gets forward biased. Current Ip now begins to flow as iD through D and as i through
L. If there were no resistance in this closed path, current Ip would continue flowing unabated.
In practice, inherent resistance in the closed path would cause this current to decay

Example 3.5. In the circuit of Fig. 3.9 (a), current in inductor L is
the waveforms of i, Vc and vL after switch S is opened at t = O.

Solution. When switch S is opened at



before t = O. Sketch

current 10 begins to flow in the path

D, C and L. KVL for this path is


1 ~ dt =
di + C
L dt







F ig . 3.9. Pertaining to Example 3.5. (a) Circuit diagram and (6) waveforms .

. L [sf
" (s) - i. (0)] + -1 [ -] (s) - -cVo
Its Laplace transform IS
C s

It is given that initially io

=10 and Vo = O. Therefore, we get

I(s) [ sL +

Its La place inverse is i(t)



s~ 1
=. LIo


s +


where CDo = ~LC as before.

=]0 cos CDot


= ~ Jidt = ~ ~ l o cos CDr! . dt = 10 ~ sin CDot



~~ = L


(10 cos CDot)

= -10 ~ sin CDc t

[Art. 3.1J

Diode Circuits and Rectifiers


At wt =~, current i tends to reverse, but diode D blocks ~his current reversal. Also , at

wt =

~, capacitor is changed to 10 ~sin'~ =10 ~ = Vc and voltage across inductance is

V L =-Io


Thus, after wt

=~, capacitor voltage remains constant at 10 ~ whereas voltage across L

becomes zero because current is now zero. Energy stored in inductance as

i L I~ at wt

= 0


transferred to C at wi

~ n/2 as ~ CV; ~ ~ C ( ~ J.

3.1.5. B.Le Load

A diode in series withRLC circuit is shown in Fig. 3.10 (a). KVL for this circuit, when switch
S is closed at t = 0, is given by
di + C
1 f'd
R l + L dt
l t = s

Critically damped






, .< overd cmped







...... .




Fig. 3.10. Diode circuit with RLC load (a ) circuit diagram and (b) waveforms.

With zero initial conditions, the Laplace transform of above equation is


I(s) R +sL+ sC







Here s2 + ~ s + L~ = 0 is t he char acteristic

s + L s + LC

equ~ion in s-domain. The roots of this equation


s =- ~ -1(fLJ
wh er e

.s= -s ~

- 2L

... (3 .11 )
... (3 .12)


Power Electronics

(A rt. 3.1]


~ called the damping factor.

js called resonant frequencyjn fad/sec

= ~r-t--c-_





=-}w6 Also



= ringing frequency in rad Isee.

=-) w; + ~2

Depending upon the values of ~ andwo, the solution for the current can have three possible
Case 1. In case ~ < wo, it is seen from Eq. (3.11) that the roots are complex and the circuit

is said to he underdamped. The two roots are


=- ~ + j

wr and

j wr

s2 = - So -

and the current is given by



wr L



... (3.15)

Case 2 . If S > wo, the two roots are real and the circuit is said to be overdamped. The two

roots are
and the solution for current is
.(t )

Case 3. In case ~

The roots are


= L ol~
'I ~ -

2 .


. h 01(,.
'I <; ~ _Wo . t

-- --

= wo, the roots are equal and the circuit is said to be

... (3.16)

crit'(cally datnped.. .-__

=s2 =- ~ and the solution for the current is



t . e .,

... (3.17)

Waveforms of current for the three different levels of damping are sketched in Fig. 3.10
(b ).

E xample 3.S. For the circuit of Fig. 3.10 (a), the data is as under:

= lOn,

= 1 mH, C = 5~, Vs =230 V

The circuit is initially relaxed. With switch closed at t = 0, determine (a) current i(t) (b)
time of diode (c) rate of change of current at t = O.

c o n~ u c tio n

Sol ution . (a) From Eq. (3 .12),

S= 10 x

Fro m Eq. (3. 13),

1000 = 5000
Wo = 'LC = [
- 6J1I ,) = ~
O = 14142 .136 rad/s
1 x 10 x 5 x 10
- '1 ~U



Fr om E q. (3 .14),


=[ 1~0

- (5000)2

= 13228. 76 r adls

[Art. 3.2J

Diode Circuits and Rectifiers

Here as

~ <

wo, the circuit is underdamped. The current is, therefore, given by Eq. (3.15 ).

x 1000 . -500Ot. in (13228 76)t

13228.76 x 1 e

'(t) = 230

= 17.3864 e- 5000t . sin (13228.76 t)


Diode stops conducting when


t1 = 1t

.. Conduction time of diode,




J::) e-f,t]
= - Vs
. - [-;./
e ~ . <D cos <D t - sm
<D (
t -..,
dt <Dr L

(c) From Eq. (3.15),



= <Dr = 13228.76 = 237.482 ~s


= Vs

- 230 x 1000

L -

= 230000 A/s


In Fig. 3.3 (a), steady state current, after switch Sis clos.ed, is equal to V/R. Energy stored

in inductance L

is~. L (V/R)2 . lfthe switch S

is now opened, current V/R would eventually

decay to zero. As the current V/R tends to decay with the opening of switch S, a high reverse
voltage appears across switch as well as the diode. High voltage acr oss switch leads to spark
across the switch contacts, thus dissipating the stored energy: In the process , the diode,
subjected to high reverse voltage, may get damaged. In order to avoid such an occurrence, a
diode FD, called freewheeling, or flywheel, diode, is connected across load RL as shown in Fig.
3.11. (a). For understanding how FD comes into play, the working of circuit of Fig. 3.11 (a) is
divided into two modes.

Mode I : When switch S is closed in Fig. 3.11 (a) at t = 0, current flows through
V s ' S, D, Rand L as shown in Fig. 3.11 (b). In this circuit, current i is given by



.. .(3.18)

(1 - e L)

Final value of current,





. i.

(a )



Fig. 3.11. Circuit of Fig. 3.3 with freewheeling diode



~ L <!l,


P ower Electronics

[Art. 3.3J

Mode n : When switch S is opened at

= 0, cu rr ent in the circuit tends to decay

and so a voltage L


~~ is induced! in L which

forward biases freewheeling diode. The

current is, therefore, transferred to the
L _ _ _ _ _ _ _ _ _ _L __
circuit consisting of FD, Rand L as shown
I---- Mode I -----.-I._0- Mod e IT--I
in Fig. 3.11 (c). In this circuit, current is
given by
Fig. 3.11. (d) Current variation in the circuit of Fig. 3.11.
V s --/
R .
".(3 .19)
II = R ' e L

The current i1 will eventually decay to zero exponentially in mode II of Fig. 3.11 (c). The
current build up during mode I and current decay during mode II are shown in Fig. 3.11 (d).


Consider the circuit of Fig. 3.12 (a) where dc source feeds L through diode D. A freewheeling
diode FD is connected across L. When switch S is closed at t = 0, KVL gives

V =L di



... (3.20)


This shows that current i rises linearly with time t . In case switch S is opened at t 1 ,Joad
current L tl begins to flow through FD. As there is no resistance in the circuit formed by

. .
L and FD, current continues to flow at its constant value of L tl. Energy stored in the
inductance is

~ [V:2
L tl .


= ~. ~ tl joules. Current waveforms

are shown in Fig. 3.12 (b).





Fi g. 3.12. Diode circuit with FD and L load


circuit ciiagram and (6) waveforms .


[Art. 3,5]

Diode Cir cuits and Rectifiers



In the ideal circuit of Fig. 3.12 (a), the energy stored in the inductor is trapped. This trapped
energy is not dissipated even when FD conducts because circuit does not contain resistance.
The best way of utilization of this trapped energy is to return it to the source . In this manner,
net energy taken from the source is reduced and the system efficiency improves.
One way of returning this trapped energy back to the source is to add a second winding
closely coupled with the inductor winding as shown in Fig. 3.13. A diode D is also placed in
series with the second winding. The inductor now behaves like a transformer. The two windings
are so arranged that their polarity markings are opposite to each other.




F i g. 3.13. Energy-recovery circuit (a) switch S closed and (6) switch S opened.

vVhen switch S is closed, current i begins to flow and energy is stored in the inductance of
primary winding with Nl turns. The polarity of the secondary winding voltage V 2 is as shown.
The diode D is reverse biased by voltage (V2 + Vs).
When switch S is opened, polarities of voltages VI and V 2 get reversed, the diode is now
forward biased by voltage (V2 - V s )' As a result, diode begins to conduct.a current i 1 into the
posi tive terminal of source voltage Vs and so the trapped energy is fed back to the source.
Energy fedback to de source = Vs x current i 1 dependent upon (V2 - Vs).
The energy stored in L of NI turns is transferred to secondary winding of N2 turns from
where it is fed back into th e de source.


Rectification is the process of conversion of alternating input voltage to direct output
voltage. As stated before, a rectifier converts ac power to de power. In diode-based rectifiers, the
output voltage cannot be controlled.
In this section, uncontrolled single-phase rectifiers are studied. The diode is assumed ideal
as before.
A rectifier m ay b h alf-wave type or full-wave type. A half-vllave rectifier is one in which
cur rent in anyone lin e, connected to ac source, is un idirectional. However, a full-wave rectifier
h as bidirectional current in any one line connected to ac surface.


[Art. 3.5]

Power Electronics

A rectifier may be one-pulse, two-pulse, three-pulse or n-pulse type . The

in any rectifier-configuration is obtained as under:
pulse number


of pulses

= number of load current

(or voltage) pulses during one

cycle of ac source voltage.
3.5.1. Single-Phase Half-wave Rectifier
This is the simplest type of uncontrolled rectifier. It is never used in industrial applications
because of its poor performance. Its study is, however, useful in understanding the principle of
rectifier operation.
In a single-phase half-wave rectifier, for one cycle of supply voltage, there is one half-cycle
of output, or load, voltage. As such, it is also called single-phase one-pulse rectifier.
The load on the output side of rectifier may be R, RL or RL with a flywheel diode . These
are now discussed briefly.
R load: The circuit diagram of a single-phase half-wave rectifier is shown in Fig. 3.14
During the positive half cycle, diode is forward biased, it therefore conducts from wt = 00 to
CDt = 11:. During the_positive half cycle, output voltage Vo = source voltage Vs and load current
io = VoiR . At w t = 1t, . Vo = 0 and for R load, io is also zero.As soon as Vs tends to become negative
after wt = 11:, diode D is reverse biased, it is therefore turned off and goes into blocking state . .
Output voltage, as well as output current, are zero from w t =1t to (j) t = 21t. After w t = 21t, diode
is again forward biased and conduction begins.

( a ).










us :





- .~. -




+.-- .











Fig. 3.14. S ingle-phas e half-wave d 'ode rectifier with R load (a) circuit diagram and (b) waveforms.

For a resistive load, output curren t io has the same waveform as that of the output voltage
Va- Diode voltage VD is zero wh en diode conducts . Diod e is reverse biased fr1fm w t = 1t to
CDt = 21t as shown. The waveforms of us' Va, ia and VD are sketched in Fig. 3.14 (b). Here source
voltage is sinusoidal i.e. Us = Vm sin wt. KVL for the circuit of F ig. 3.14 (a) gives Vs = Vo + VD'
Average valu e of output (or load) voltage,

Vo ;

~s: Vrn sin w t d(w III

[Art. 3.5]

Diode Circuits and Rectifiers

Vm .
- 2n
Rms value of output voltage, Vor




= Vm fTC
-::r2n [ 0

,..(3.21 )


fa ~ sin


wt . d (wt )



1- cos 2wt . d (wt )



.. .. (3.22)

- 2
H ere the subscript 'r' is used to denote rms value .
Average value of load current,
Vo Vm
10 = - = -

... (3.23 )

. ' _ Vor _ Vm
lor - R ~ 2R

.. .(3,24)


Rms value of load current,

.Peak value of load, or diode, current

_ Vm
- R

... (3.25)

, Peak inverse voltage, PlY, is an important parameter in the design of rectifier circuits.

P1V is the maximum voltage that appears across the device (here diode ) during its blocking
state. In Fig. 3.14, PIV = V m = -f2 . Vs = -v; (rms value of transformer secondary voltage ). It is
seen from the waveform of source current is (or io) that the transformer has to handle de
component of is' It leads to magnetic saturation of the transformer case, therefore more iron
10ss8s , more transformer heating and reduced


Power delivered to resistive load


=V or

Input power factor


= (rms load voltage)

Vm Vm


V~ V;
= 4R :::: 2R

=1;,. R

... (3.26)

Power delivered to load

Input VA
Vor . lor Vor -f2vs
=v . ] =Y-= 2V =O.707lag.

= 2 . 2R

(rms load current ).


(b) L load: Single-phas e half-wave diode r ectifier with L load is shown in Fig. 3.15 (a ).
When switch S is closed
at wt = 0, diode starts con ducting. KVL for this circuit gives


io =;:


f sin w t . dt


= - - cos
,At wt


v s = Vo =L -dt - = V m sm w t


= 0,

io:::: 0 , ..


.. (3. 27 a )


Power Electronics

[Art. 3.5J

Substitution of the value of kin Eq. (3.27 a) gives

io = wZ (1 - cos wt)

Output voltage,
Source voltage


va = L (it


Vm .
wL [sm CDt]

... (3 .27 b)


= V m sm CDt =Vs

and both output voltage Vo and output current io are plotted in Fig. 3.15

(b ).

Average value of output voltage, Va = 0




Fig. 3.15. Single-phase one-pulse rectifier with L load (al circuit diagram and (b) waveforms.

The output current io consists of de component and fundamental frequency. component of

frequency w.
Peak value of current 1max occurs at wt = 7t

Average value of current,

. Vm
. 2Vm
= wL (1 + 1)

.. .(3.28)

1 J27t V
10 = 27t 10 w'i(1- cos CDt) d(wt)


= wL =

.. .(3.29)

2 Imay:

Rms value of fu n dament al current, I Ir is given by

r1 [V
\2 J2Tt

I lr 127t

= T2 . wL = wI. = ...[2


0 (coswt)2d(CDt)


... (3.30)


Rms value of rectifie d curr nt = [ 102 + 121rJ1I2



=1.225 10

.. .(3. 31)


[Art. 3.5J

Diod e Circuits and Rectifiers

Voltage across diode,

. vD = 0.
(e) C Load: In Fig. 3.16 (a), when switch S is closed at wt== 0, th e equation governing the
behaviour of the .circuit is

Output volta ge,



= C dts == C dt (Vs sm w t)
= wC Vm cos wt


= ~ Jidt =V m sin wt = v s = VC









wcvm ':

+ -:-






vcr _C

.'37[/ 2

3rr/ 2 2rr

3rr 7rr/2 :


v o=uc

(0 )


Fig. 3.16. Single-phase half-wave di ode r ectifier with C load (a ) circuit diagram and (6) waveforms .

Capacitor is charged to voltage V m at wt


and subsequently this voltage remains

. constant at V m' This is shown as Vo = Vc in Fig. 3.16 (b ).

Capacitor current or load current is maximum at wt ::: O. Its value at wt

The diode conducts for 2 w seconds only fr om wt = 0 t o wt

=1t/ 2 , diode voltage

Vo + Vs = - V m + V m sin wi

voltage is , therefore, zero. After wt


= -

= V (sin wt -



w CVm as

= ~ . During this interval,

di ode

is given by
... (3.3 3)


For Eq. (3.33), the time origin is r edefi ned at wt = 1t/ 2 .

After wt

= 1t/ 2,

diode voltage is plotted as shown in Fig. 3. 16 (b ). At wt = 31t,

Average v alu e of voltage

a cro ~ s




=~ J

= V :::

V m (~in wt - 1) d (rot)


G'D = -

2 V", .

.Power E ectronics

Rms value of fundamental component of voltage across diode,




=[. 2~ fo ~ sin 2 OJt d(OJt)

. ==

" .(3.34 b)

,R ms value of voltage across diode

-= -Jr-"v""'-1-+-v-'i"-r = 1.~'25 V m
Example 3.7. Find the time required to deliver a charge of 200 Ah through a single-phase
half-wave diode rectifier with an output current of 100 A rms and with sinusoidal inpu.t voltage.
Assume diode conduction bver a half-cycle.

Solution. For I-phase half-wave diode rectlfier, rms value of output current,

. V

I or = ~
2R = 100 A or V m. = 200

The charge is delivered by direct current 10 which is given by

= 200 R = 200 A

.Io x time in hours = 200 Ah

:. Time required to deliver this charge

200 x 7t .

= 200 hrs = 7t = 3.1416 hrs

I = Vm

Example 3.8. A single-phase 230 V, 1 kW heater is connected across single-phase 230 V, 50

Hz supply through a diode. Calculate the power delivered to the heater element. Find also the
peak diode current and input power factor.

H eater resIstance,

S o 1u t Ion.

230 n
= 1000

Rms value of output voltage, from Eq. (3.22), is


=..J2 x 230

Power absorbed by heater element

- V;r


2 X 230


230 2

= 500 W

Peak value of diode current, from Eq. (3.25), is given by

..J2 x 230

Input power factor

_ Vor _
- V -

x 1000 = 6 .1 478 A

..J2 x


x 23 0 = 0.7 07 lag.

R E Load : Single-phase half-wave diode rectifi er with l oad resistance R and load
counter emf E is shown in Fig. 3.17 (a). If the switch S is closed at OJt = 0 or when Vs = 0, then
diode would not conduct at OJt = 0 because diode is reverse bi ased until SOUTce voltage Vs equals
E. \hen V m sin 8 1 = E, diode D st arts conducting and the turn-on angle 8 1 is given by

.-I (E]

e1 -- SIn

... (3.313)


[Art. 3.5J

Diode Circuits and Re ctifiers


o '


Uo I i

(77 -8 1)

i' 11:


i :



i i







Fig. 3.17. Single-phase half-wave diode rectifier with RE load

(a) circuit diagram and (6) wave forms .



The diode now conducts from wt = 8 1 to wt = (&.= 8 1), i.e, conduction angle for diode is
(n - 28 1) as shown in Fig. 3.1 i ( b). During the conduction period of diode, th~\oJtage equation.
for the circuitjs
V m sin wt = E +--i-o R

V m sin wt - E

to =
Average value of this current is given by

Io = 2rtR


it -

(Vm sin wt - E ) d(wt)

= 2rtR


Vm cos

81 -

(n - 28 1)J

... (3. 38 )

Rms value of the load current of Eq. (3.37) is

[~ fT8t -8

I =
or 2n

[Vm sin wt - E ]2 .

f:- (V~ sin

wt + E2 - 2 Vm E sin wt) d(wt )]112

~ [2~2 i(V; + E') (n- 28





+ V; sin 28 1 - 4 Vm E cos 91


Power delivered to load,

Supply pf

P = E 10 + 1;1' R "vatts
Power deli vered t o 10ad
(S ou rce volt age) (rms value of ;:;o urc e current )

E 10 + l or R

... (3.4 0)


.. (3 .41)


P ow er Elec troni cs

[Ar t. 3.5]

It is seen from Fig. 3.17 (a) that at rot = 0, UD = - E and at rot = 81 , uD = O. During the period
diode conducts, U D = O. When rot = 3rr/2, Us =- V m and UD:;:: - (Vm + E). Thus PIV for diode is
(Vm + E ) .


.., Example 3.9. A dc battery of constant emf E is being charged through a resistor as shown
in Fig. 3.17 (a). For source voltage of 235 V, 50 Hz and for R = SQ., E = 150 V,
(a) find the value of average charging current,
(b) find the power supplied to battery and that dissipated in the resistor,

(c) calculate the supply pf,

(d) find the charging time in case battery capacity is 1000 Wh and
(e) find rectifier efficiency and PIV of the diode.



The diode will start conducting at an angle 81 , where

8 1 -

. -1

x 230 .

Average value of charging current, from Eq. (3.38), is


~ 2n ~ 8 [2 . ,f2 x 230 cos 27.466


150 (1t - 2 x

= 4.9676 A


Power delivered to battery

= E 10

2~.:g6 x n Jl

= 150 x 4.9676 = 745.14 W

Rms value of charging current, from Eq. (3.39), is


~ [ 2n ~ 64{ (230' + 150') (n - 2 x 27 A66 x 1~0 J+ 230' sin 2 x 27.466


. - 4,f2 x 230 x 150 cos 27.466


~ 9.2955 A

Power dissipated in resistor

=l~r R = (9.2955)2 X 8 = 691.25 W


From Eq. (3 .41), th e supply


= 745.14 + 691.25 = a 672 1


230 x 9.2955

(d) (Power delivered to battery) (charging time in hours)

= Battery capacity in Wh.
:. Charging tim e
= 745.14 = 1.342 h


Rectifier efficiency

= Power deliver ed to battery

Total input power


. (f) PIV of dio de


745.~::~~1.25 x 100 = 51.876%

:;:: Vm + E = -{2 x 230 + 150 = 475.22 V .

RL Load: A sin gle-phase one-pulse diode rectifier fe eding RL load is shown in Fig.

3.18 (a). Current i o continu es t o flow-even aft er source volta ge Vs has become negative; this
i s be caus e of the pr es ence of inductance L in the load circuit . After positive half cycle of source


[Art. 3.5)
















Fig. 3.18. Single-phase half-wave diode rectifier with RL load

(a) circuit diagram and (b) waveforms.

voltage, diode remains on, so the negative half cycle of source voltage appears across load until
load current io decays to zero at wt = ~ . Voltage vR =io R has the same waveshape as that of io
Inductor voltage VL = Vs - vR is also shown. The current io flows till the two areas A and Bare
equal. AreaA (where Us > vR) represents the energy stored by L and area B (where Us < uR) the
energy released by L. It must be noted that average value of voltage VL across inductor L is
When io = 0 at wt = ~; uL = 0, l.iR = and voltage Us appears as reverse bias across diode
D as shown. At ~, voltage vD across diode jumps from zero to Vm sin ~ where ~ > 11:. Here
~ = y is also the conduction angle of the diode.
Average value of output voltage,

. Vo = 21t

= 21t


Vm sin wt d(wt)

(1 - cos ~)

.. .(3.42)

Average value of load or output current

Vo Vm
10 = Ii = 21tR (1 - cos ~)
A general expression for outpu t current io for 0 < wt <

... (3.43)

can be obtained as under:

When diode is conducting, KVL for the circuit of Fig. 3.18 (a) gives


dt = Vm sin wt

Rio + L

The load, or output, curren t io consists of two components , one steady s tate component is
and the other transient component it. Here is is given by


is = ~R2

sin (wt -~)


Power Electronics

(Art. 3.5)

where = tan- 1 ~ and X = wL. Here < is the angle by which rms current Is lags source voltage

. Vs ' The transient component it can be obtained from force-free equation

di t
Ri +L-=O


Its solution gives

it = A e'" T, t

Tota.l solution for current io is, therefore,given by

. io is + it
sin (wt - ) +A e-T,t



where .
Z = ..JR2 +X2

Constant A can be obtained from the boundary condition afwt ~ O.

Atwt = 0, or at {=O, io = O. Thus, from Eq. (3.44)

Vm .
0=:- -y .sin ~ +A


. Vm

. . .



. Substitution of A in Eq. (3.44) gives

io'=~m [sin (wt -


<) + si'n

~. e-tJ

wt ~ ~
It is also seen from the waveform of io in Fig. 3.18 (b) that when wt
condition, Eq. (3.45) gives
for 0



.. .(3.45) . '.

(~- <) + sin , exp [

-1 ~] =


io == O. With this

The solution of this transcendental equation can give the value of extinction angle


(f) RL load with freewheeling diode*: Performance of single-phase one-pulse diode

rectifier with RL load can be improved by connecting a freewheeling diode across the load as
in . Fig. 3.19 (a). Output voltage is Uo = Us for 0 ~ wt ~ n. At wt = re, source voltage Us is
zero, but output current io is not zero because of L in the load circuit. Just after wt = re, as Us
tends to reverse, negative polarity of Us reaches cathode of FD through conducting diode D,
whereas positive polarity of Us reaches anode of F D direct. Freewheeling (or flywheel) diode
FD, therefore, gets forward-biased. As a result, load current io is immediately transferred from
D to FD as Us -tends to reverse.' After wt = re, diode, or source, current is = 0 and diode D is

subjected .to reverse voltage with PlV equal to V m at CDt = 3re, 7; etc.
After wt = n, current freewheels through circuit R, Land FD. The energy stored in L is
now dissipated in R. When energy stored in L = energy dissipated in R, current falls to zero
at CDt = ~ < 2re. Depending upon the value of Rand L, the current m ay not fall to zero even
when CDt = 2re, this is called conti uous conducti on. But in Fig. 3.19 (b) , l0 2.d CUTrent decays to
zero before CDt = 2re; load current is ther efore discontinuous.
* Freewheeling diode is aiso called bypass diode or commutating diode .

Diode Circuits and Re ctifiers


(A rt. 3.5]















D --jFO









Fig. 3.19. Single-phase one-purse diode rectifier with RL load and fre!twnee1ing diode

(a) circuit diagram and (b) waveforms.

The effects of using freewheeling diode are as under:

It prevents the output (or load) voltage from becoming negative.
(ii ) As the energy stored in L is transferred to load R through FD, the system efficiency
is improved.
(iii) The load current waveform is more smooth, the load performance, therefore, gets

The w aveforms for

Us ,

Va, io, vD,


and ifd are drawn in Fig. 3.19 (6).

The expression for the load current io can be obtained from Art. 6.1.2 if required. It is seen
from Fig. 3.19 (b) that
1 J~
= V sin wt d(wt)
o 21t 0 m

average output voltage,

and average load current,

I a -....!!!:.- rtR



... (3.46)
... (3 .47)

(g) Single-phase fun -wave diode rectifier : Th ere are two types of full-wave diode

rectifiers, one i centre-tapped (or mid-point) full -wave diode rectifier and the other is full-wave
diode bridge recti,er. These are now described briefly.
(i) Singl e-phase full-wave mid-point diode r ectifier: Fig. 3.20 (a) illustrates a
single-phase full-wave mid-point rectifier using diodes. The turns ratio fTom each sec ondary
to primary 's taken as l.illty fOT simplicity. vVhen 'a' is positive with respect to ' b', or m id-point
0, diode Dl con ducts for 1t r adi ans. In the next half cycl e, 'b' is positive with respect t o 'a' , or .
mid-poin t 0, and ther efore diode D2 condu cts. The output voltage is shown as Vo in F g. 3.20
(b) . The waveform for output current io (not sh own in the figur e) 15 gimilar to va waveform.


Power Electronics

[Ar t. 3.5]

Vm sin wI








Vs (\J







Fig. 3.20. Single-phase full-wave mid-point diode rectifier
(a ) circuit diagram and (b) waveforms .

When 'a' is positive with respect to 'b', diode D2 is subjected to a reverse voltage of 2vs ' In the
next half cycle, diode Dl experiences a reverse voltage of 2vs' This is shown in Fig. 3.20 (b) as
VDl and VD2' Thus, for diodes Dl and D2 , peak inverse voltage is 2V~. Waveforms of Fig. 3.20
( b ) show that for one cycle of source voltage, there are two pulses of output voltage. So
single-phase fun-wave diode rectifier can also be called single-phase two-pulse diode rectifier.
Source current waveform is is also shown in Fig. 3.20 (b). When Dl conducts, current in
secondary flows upward from 0 to a and therefore, primary current is must flow downward to
balance the secondary mmf from 0 to 7t rad . 'When D2 conducts, secondary current flows
downwar d from 0 to b, therefore, prim ary current is must flow upward to balance the secondary
mmf from It to 27t and so on.
Average outpu t voltage,

...(3.48 a )

Average output current,

Rms value of output voltage,
.. .(3.48 b)

[A rt. 3.5]

Diode Circuits and Rectifiers



Rms value of load current,

= Vor . lor =l~r . R

= Va . lor

Power delivered to load

Input volt amperes
:. I nput



Va' lor

= 1

(ii) Single-phase full-wave diode hridgerectifier : A single-phase full-wave bridge

rectifier employing diodes is shown in Fig. 3.21 (a). When 'a' is 'positive with respect to 'b',
diodes Dl, D2 conduct together so that output voltage is Vab' Each of the diodes D3 and D4 is
subjected to a reverse voltage of Vs as shown in Fig. 3.21 (b). When 'b' is positive with respect
to 'a', diodes D3, D4 conduct together and output voltage is vba' Each of the two diodes D1 and
D2 e~perience a reverse voltage of va as shown.

A comparison of Figs. 3.20 (b) and 3.21 (b) reveals that a diode in mid-point full-wave
rectifier is subjected to PIV of 2Vm whereas a diode in full-wave bridge rectifier has PIV of
V m only. However, average and rms values of output voltage are the same for both rectifier




1: 1




"- ' .


'1. 0 2



Fig. 3.21. Single-phase full-wave diode bridge r ectifier


circuit diagram and



For the waveforms of diode current i Dl or iD2 in Fit 3.21 (b) and also for iD3 , iD4 for the circuit
of Fig. 3.20 (a) (not shown in Fig. 3.20 (b) ), the average and rms values for diode current are
obtained as under:

. I J~
Average va1ue of di ode curren t) ID = -2
1m sin rot . d (cut) =

.. (3.49 a)

n o n

Rms value of diode current,

ID = [

Peak r epetitive diode current,


2~ ( I~ sin' wt d (wt)



... (3.49 b)
... (3.49 c)


Power Electronics

[Art. 3.5]

It can similarly be shown that average value of voltage across each diode in Fig. 3.20 (b) is

Vm. The corresponding rms values of voltage across eac hd'10 d e
- and that in Fig. 3.21 (b) is TC


is V m = {2 Vs in Fig. 3.20 (b) and ~ = Vs in Fig. 3.21 (b).
Three-phase rectifiers using diodes are discussed in APt. 3.9. Example 3.10 is formulated to
illustrate the effect of
reverse recovery time on the average output . voltage.
Example 3.10. In a single-phase full-wave diode bridge rectifier, the diodes have a reverse
recovery time of 40 ~s. For an ac input voltage of230 V, determine the effect of reverse recovery
time on the average output voltage for a supply frequency of (a) 50 Hz and (b) 2.5 kHz.
Solution. Single-phase full-wave diode bridge rectifier is shown in Fig. 3.21 (a) and output
voltage Vo is shown in Fig. 3.21 (b). If reverse
recovery time is taken into consideration, the l..io
diodes D 1 and D2 will not be off atwt = TC in
Fig. 3.21 (b), but will continue to conduct until

as .depicted in Fig. 3.22. The

reduction in output voltage is given by the

cross-hatched area. Average value of this
reduction in output voltage is given by
Fig. 3.22. Effect of reverse recoverY time
on output voltage.



= 1.



sin wt d (cot)




(1 - cos wtrr )

With zero reverse recovery ti m e, average output. voltage, fr om Eq. (3.48),


VO = 2..f2 x 230 = 207.04 V


(:3.50), is .

f= 50 Hz and trr = 40

the reduction in the average output voltage, from Eq.



= ~ (1 - cos 27t f trr )


= -!2 x 230 (1 - cos 27t X 50 X



10- 6 X 18



Percen tage reduction in average outpuc voltage
(b) For

8.1~~;.: 0~0-3 x 100=3.948x 10- 3%

f = 2500 Hz, the reduction in t he average output voltage, from Eq. (3 .50), is

. (1 - cos 21t X 2500 X 40

V ,. = ~ x 230

= 19 .77 V

10- 6 X -180


[Art. 3.5]

Diod e Circuits and Rectifiers


Percentage reduction in average outpu t voltage ~ 21;;, ~074 x 100 = 9.594%.


It is seen from above that the effect of reverse recovery time is negligible for diode operation
at 50 Hz, but for high-frequency operation of diodes, the effect is noticeable.

Example 3.11, A single-phase full bridge diode rectifier is supplied from 230 V, 50 Hz
source. The load consists of R= 10 n and a large inductance so as to render the load current
constant. Determine .
(a) average values o{output voltage and output current,
(b) average and rms values of diode currents,

(c) rins values of output and input currents, and supply pf.
Solution. The circuit diagram and relevant waveforms for this uncontrolled rectifier are
shown in Fig. 3.23.



. I

01 D2---L 0304 -.l--01 02+0304-(









. io




w 't.





. (a)

Fig. 3.23. Pertaining to Example 3 .11

(a )


rIo .




circuit diagram and (b ) waveforms .

Avetage value of output voltage,




= ~ = 2 '1 2 x 230 = 207.04 V .

7t .
7t . .

Average value of output current,

10 = Vo == 207.04 = 20 704 A .
(b ) Average value of diode curren t ,
1 . 7t I
')0 '7 04

= ~7t = ; =...


= 10.352 A





Power Electronics

[A r t. 3.5]


Rm s value of diode current, I Dr =


1 1t


= 12=



= 14.642 A

As load, or output, current is ripple free, rms value of output current

= average value of output current =10 = 20.704 A

10 = 20.704 A

Load power
= Vo 10 = 207.04 x 20.704 W

= Vs Is cos <jl

Inp ut power
230 x 20.704 x cos <jl = 207.04 x 20.704

~' . Supply pf
= cos <jl = 2~~.g4 = 0.90 lagging.

Rms value of source current, Is


o v Example 3.12. A diode whose internal resistance is 20 n is to supply power to a 1000 .0 load
from a 230 V (rms) source of supply. Calculate (a) the peak load current (b) the dc load current
(c) the dc diode voltage (d) the percentage regulation from no load to given load.
(I.A.S., 1983 )
Solution . A voltage of 230 V supplying power to 1000.0, through a single diode, is shown
.in Fi g. 3.24 (a ). Waveforms for the source voltage, load current ioaiid diode voltage vD are shown
in Fig. 3 .24 ( b ).



+ - - io


R=1000 .n.


(a )

(b )

F ig. 3.24. Pertaining to Example 3.12 (a ) circuit diagram and (b) waveforms.
(a )

It is seen from th e w aveform of i o that peak load current 10m is given by

-{2 x 230
10m = R + RD = 1020 = 0.3189 A

H ere R

= load resistan ce and

RD = internal resistance of diode.

(b ) f) C load curren t,


= 21t


f lt

l orn sm wt d (wt)

0.101 51 A

[Art. 3.6J

Diode Circuits a n d Rectifie rs


V D =10 RD - 27t

DC diode voltage,

(d )

At no load, load voltage,


V 01

At given load, load voltage,

10 RD - Vm

230--12 sin wt d(wt)

= 0.10151 x 20. -

230.J2 = - 101.5 V

= Vm
= -f2 x7t 230
= 103.521 V .
= 230-f2

1000:: 101 491 V


x 1020

=Vall -:- Val

:. Voltage regulation



x 100 = 103.521- 101.491 = 1.961%.



Zener diodes are specially constructed to have accurate and stable reverse breakdo'pn
. voltage.
Circuit symboi' for Zener diode is shown in Fig. 3.25 (a) . When it is forward biased, it
behaves as a normal diode. When reverse biased, a small leakage current flows. If the reverse
voltage across Zener diode is increased, a 'value of voltage is "reached at which re verse
breakdown occurs. This is indicated by a sudden increase of Zener current, Fig. 3.25 (6). The
voltage after reverse breakdown remains practically constant over a wide range of Zener
current. This makes it suitable for use as a voltage regulator to furnish constant voltage from
a source whose voltage may vary noticeably.







vo ltage






Vo= Vz

(c )

(a) circuit

Fig . 3.25. Zener diode

symbol (b) I-V characteristics

(c ) use as a voltage re~lator.

For the operation of Zener diode as a voltage regulato~, (i) it must be reverse biased with
a voltage great er than its breakdown, or Zener, voltage and (ii) a series res ist or Rs, Fig. 3.25 .
(c) is necessary to limit the r everse cu rrent t hrough the diode below its rated value.
If V z = voltage across Zener diode, th en it is seen ftom Fig. 3.25



that source current 15


[Art. 3.6]


Load, or output, current, 10 = Iiz where R

E lectronic~

= load resistance. Current through Zener diode,

1z = Is - 10
Power rating of a Zener diode is V z . 1z. These are available in a voltage range from few
volts to about 280 V.
Example 3.13. Design a Zener voltage regulator, shown in Fig. 3.26, to meet the following
Load voltage = 6.8 V, Source voltage Vs is
20 V 20% and load current is 30 rnA 50%.

The Zener requires a minimum current of 1

VL :6.BV
mA to breakdown. The diode D has a forward Vs
voltage drop of O. 6Y.

Solution. When source volbige is maximum

~d load current is minimUm, then source resistance
should be maximum.

Fig. 3.26. Pertaining to Example 3.13 .

. Vs _mlu =VL + (1L - mill + 1z ) Rs -max

_ = (20 X 1.2) - 6.8
= 1075 n .
s - ml1l
[3 0x O.5+1]xlO- 3

Vs -min = V L + (IL -max + 1z) Rs -mill

_ = (20 X 1.2) - 6.8 = 200 n
5 - mm
[30 X 1.5 + 1J X 10- 3

Maximum load resistance ,

= _ V_L_=


Minimum load resistance,

RL _mill




L - max

6 .8

0.5 X 10- 3

30 X 1.5 X 10- 3

=151.5 n

The voltage rating of the Zener diode is

6.8 ~


= 6.2 V .

. E x ample 3.14. The complete circuit shown in Fig. 3.27 (a) rep resents a 25 V de voltmeter

where G is a PM MC galvanometer having full-scale deflection current 1{sd = 200 micro-A and

resistance R o = 500 ohms, and D is a 20-V Zener diode. Find R l and R 2. W hat is the function
of the diode D in this circuit?
. (GATE, 1990) .

Solution. Current through galvanometer,


- I _ Zener voltage
2R .... R

(sd -


2 '

R 2 :500 = 200 X 10- 6



X 10
R.2 -- 20200
- t)'- 00 --




[Art. 3.7]

Diode Circuits and Rectitiers





lz 12
RG=5 00n


Fig. 3.27. Pertaining to Example 3.14.

As Zener diode current is not specified, let it be assumed zero. Therefore, from Fig. 3.27
(b ),

II ~12 =lz == 0


II =12 = 200)lA.

11 = 25 - 20 = 200 x 10- 6


5 X 10 6
Rl = 200 = ~5 kn


Function of Zener diode is to provide a constant voltage to the ga1vanometer circuit.

Whenever voltage across this diode exceeds 20 V, it conducts and the excess current is shunted
away frtlm galvanomete r G. So here diode D prevents overloading of the PMMQ galvanometer.



The input voltage to rectifiers is usually sinusoidal. It is desired that the output voltage
from a rectifier should be constant with no ripples in it. This, however, is not the case. This
shows that the rectified output voltage is m ad up of constant de voltage plus harmoniC
components. The waveform of input and output currents depend on the nature ofload and the
rectifi er configuration. In order to evaluate the overall performan ce of rectifierload
combinations, certain performance parameters relatin g tp their input and output must be
known. The object of this article is to defi n e the various pe-rform ance parameters (or indices)
relating to input as well as output voltages and currents .

3.7.1. Input Performance Parameters

The variou s parameters relating t o the source (or input ) side of t he converter-load
combin ation are defined below :
(i ) I nput power-fact or. Input voltage taken fr om power-supply undertaking is generally
sinusoidal. However, ac input current is usually non-sinusoidal. Under such a condition, only
the fund a mental component of inpu t current t akes pa t in extr acti g mean a c inpu t power
fr om the source .

The inpu t power factor PF is defined as the r atio of mean input power (real power) to the
total r ms input volt amperes (apparent power) given t o the conver t er (or rectifier) system .

If Vs = rm s value of supply phase volt age.

I, = r m s value of supply phase current in cluding fundamental and h armonics

I s 1 = r ms value of fundam ent al component of supply cu rren t Is an d
rp 1 =

phase an gle bet ween supply vol age V s an d fu ndam en tal componentIs1 of supply
eu r ent Is ; see Fi g. 3.28 ;

Power Electronics

[Art. 3.7]


then , the input power factor, as per the definition, is given by

PF =
mean ~c input power
= real power, Vs' lsI' cos (h
total rms mput voltamperes
apparent power, Vs . Is


=y .cos CPl

For a given power demand, if input pf is poor, more input volt-amperes and hence more
input current are taken froin the supply. .
(ii) Input d i splacement factor (DF). As stated above, the phase angle between
sinusoidal supply voltage Vs and fundamental component lsI of supply current Is is CPl' This
angle CPl' shownin Fig. 3.28, is usually known as input displacement angle. Its cosine is called
the input displacement factor DF.

DF= cos <h


DF is also called fundamental power factor.

(iii) Input current distortion factor (CDF). It is defined as the ratio of the rms value of
fundamental component lsI of the input current to the rms value of input, or supply, current Is'


... .(3.53)

It is seen from Eqs . (3.51) to (3.53) that PF = (CDF) x (DF)

or input power factor = (input current distortion factor) x (input displacement factor)

(iv) Input current harmonic factor (HF). Non-sinusoidal input, or supply, current is
made up offundamental current plus current .components of higher frequencies . The harm onic
fa ctor (HF) is equal to the rms value of all the harmonics divided by the rms value of
fundamental component of the input current.
If Ih = rms value of all the harmonic-components combined


='i 1 s - sl


then , as per the defmition,
where, Isn = rms value of nth harmonic content.
H armonic factor is a measure of the harmonic content in the input supply current. HF is
also known as total h armonic distortion (THD). Greater the value of HF (or THD), greater is
the harmonic content and hence greater is the distortion of input supply current. .


HF =

- 1 = ~~-1

... (3.55)

Higher va1ue of input distortion factor CDF indicates lower magnitude ofharinonic content
in the source current.
Non-sinusoidal input current can be esolved in to Fourier series as under :
...(3 .56)
i =; +
(an cos n wt + bit sin nwt)
It = 1,2, 3, ...

=~ +
where a o ::: T2

i(t).dt, an = T2



[Art. 3.7J

Diode Circuits and Rectifiers'


en sin (nwt + <1>n)

=1,2,3, ...

i(t). cos nwt. dt and bn = ~

i(t). sin nwt. dt

Cn=[ a!; b! rand $n =tan- 1

( ~: ) ( 3 5 8 1
Crest Factor (CF). Crest factor for input current is defined as the ratio of peak input
currentlsp to its rms value Is.



Fig. 3.28. Waveforms for source voltage US) source current is, fundamental component
.isl of source current and ~l = phase angle between Us and is1 '
.. .(3.59)

CF is used for specifying the current ratings of power semiconductor devices and other

3.7.2. Ou tput performanc e Par ameters


, ,"

The load, or output, voltage and the load (or output) current at the output terminals of ae
to de converters are unidirectional but pulsating in nature. Fouriers series is used to express
these output quantities in terms of its two components, namely (i) average (or de) value and
(ii) ae component superimposed on de value a s under:

In general, average value of output quantity y is , Yo


= Y de = ~ J y. dt

and its rms value is,

Y or = T


TY 2 dt ]112

wh ere y = instantaneous v alue of the function in terms of t

and T = tim e period for one cycle of y variation .
. . Output de power, P dc

= (average outpu t

voltage, Vo) x (average output current, 10 )

::: Vjo
where subscript '0' denotes output de valu es.

Output ae power Pac = Val"" 101'

where subscript "or" denotes Ims value 01 output qU2.lltitiE':: .

... (3 60)


P ower Electronics

[Art. 3.7]

The various output parameters are now defined below.

(i) R e ctification r atio 11. Rectification r atio, also called efficiency of a converter, is defined

as the ratio of de output power Pde to ac output power Pac'




Rectifier ratio is also known as rectifier efficiency or figure of merit. In case Rd = forward
rectifier resistance, then


... (3.62)

Pac + lor Rd

(ii) Effective, or ripple, value of the ac component of output voltage is given by

Vr=~~r- ~


where VI' is called ripple voltage, or effective value of ae component of output voltage.

(iii) Form factor (FF). It is defined as the ratio of rms value Vo/' of output voltage to the
dc value Va ~utput voltage.

.. .(3.64)


FF is a measure of the shape of the output voltage. The closer FF is to unity, the better is
the dc output voltage wavefor m . For constant de output voltage, rms value of output voltage,
Val' = average value of output voltage, Vo'
Voltage ripple factor (VRF). It is defined as the ratio of ripple voltage VI' to the
average output voltage Va.
(i v)



.. .(3.65)

Substituting the value of VI' from Eq. (3.63) in Eq. (3.65) gives

VRF =[


(t J' -

FF ='/VRF2 + 1

=VFj" -

... (3.66 a)
... (3.66 b)

(v) P er-unit average ou tpu t voltage. It is defined as the r atio of the average output
voltage Va for any value of triggering angle to the average output voltage Vom for zero-degree
firing angle.

.. .(3.67)

(vi) Current ripple fa c tor {CRF). It is defined as the ratio of r m s value of all h armonic
components of output current to the de component ] 0 of the output curren t.

... (3.68)

." .

[Art. 3.8]

Diode Circuits and Rectifiers


Here lor = rms value of output current including de and harmonics,

Ir = rms value of all harmonic components of output current

10 == de .component of output current.

Note that l~r == I; + I~.

(vii) Transformer utilization factor (TUF). If V 2 (= V s) and 12 (= Is) are respectively the

rms voltage and rms current ratings of the secondary winding of a transformer, then TUF is
defined as
... (3.69)

:. Transformer VA rating

... (3.70)

Lower the TUF, higher is the transformer VA rating required..

It is desirable that a rectifier produces a perfect de output voltage so that (i) rms value = de
value (ii) FF = 1.0 (iii) ae component of output voltage = 0 (iv) HF = 0 (vi) PF = 1.0 and TUF = 1.

In this article, the performance parameters of single-phase diode rectifiers feeding resistive
loads are evaluated. The rectifier types discussed are I-phase half-wave rectifier and I-phase
full-wave mid-point and bridge types . The performance parameters are then collated in tabular

3.8.1. Single-phase Half-wave Rectifier

This rectifier, when feeding a resistive load, Fig. 3.14 (a), has waveforms for source voltage
v s , output voltage Vo and output current i o in Fig. 3.14 (b). Its various performance parameters
are obtained as under:

F rom Eq. (3.21), de output voltage, Vo =


. and!
d e output current,

where 1m = ;

I = -~ =~



= maxi mum values of de current as shown in Fig. 3 .14 (b) .

Output de power,
From Eq. (3.22), rms output voltage,
and rms output current,
Output power,
Rectifier efficiency,

.. .(i)

V. or=2

_ Vm _ [m
2.R - 2

or -


V mI ,.

= VOl' l or = -4P dr

TJ == p -=- =

FO :"lTI



- 21t

.. .(ii)


. ~l
Y r'; l"l

== ~

FF= -E!.. =_2 y- -~ - .5708
'). y'm, 2

= 0.4053 or 40.53 %


Power Electronics

[Art. 3.8]

Ripple voltage,
Voltage ripple factor,
Since source voltage


is a sine wave, its rms value, Vs


Load current io waveform is the same as that of source current is'

: . Rms value of source current, Is = rms value of output current, lor = ;.
VA rating of transformer

:. Transformer utilization factor, TUF

A TUF = 0.2865 means that VA rating of transformer is i~F times the de power output.
For a load of 100 watt, a transfo~er having a ratIng

ofO.;~~5 = 349.6 VA would be required .

PIV=-v2Vs =Vm
Peak value of source current,
Rms value of source current,
Crest factor,

I s =1or =_
2 .

Isn 1m
CF =-= =- x 2 = 2
Is 1m

3.8..2. Single-phase Full-w ave Mid-point Rectifier

Its circuit diagram and various waveforms are shown in Fig. 3. 20. Its different performance
parameters are obtained as under.
From Eq. (3.48), de output voltage,
and de output current,

Out put de power ,

R ms output voltage,

Rms output curren t,

Output ac power,

[Art. 3.8]

Diode "Circui s and Rectifiers


Rectifier efficiency,



= -Pdc =-2 V m I m ' --=-=08106








2V = iT2= 1.11


Form factor,

1/ 2

V,=~v;,-v!=[(~ J _(2:," J 1 =0

Ripple voltage,
Voltage ripple factor,

VRF=;- = 0.3077 Vm x 2 V



= 0.483

VRF = ~FF2 - 1 = ~1.112 -1 = 0.482

TUF can be obtained as under:

Rms value of voltage for each secondary winding


. Note that current in


secondary winding flows for half cycle only.

:. Rms value of current in each secondary winding

=; "

VA rating of secondary winding

= 2 [voltage rating of each secondary winding]

x [current rating of each secondary winding]
Vm 1m V m 1m
=2 X12x 2=~=0.707 V m 1m

Primary winding current is, how ever, made up of both positive and negative half cycles.
:, Primary rms current
Primary rms voltage
Primary VA rating
:. Average VA rating of transfo rmer = 0.5 +2.


. V m1m = 0.6035 Vm 1m

P dc
--.2 . V m 1m
average VA rating of transformer It-

PlV for each diode =2Vm

Peak value of source current,


Rms value of source current, Is

- 0 672
0.6035 V m 1m - .


:. CF of input cur ent

3.8,3. Single-phas.e fuU-w

Its circuit diagram is given in Fig. 3.21 (a). It is seen fr om Fig. 3.20 (6) and 3.21 (b) th at
waveform of out ut (or load) type Vo and out current io are identic al in bo th ill! - 2 and B- 2
types of diode rectifiers. Therefor e, in aingle-phase B - 2 diode rectifi er also,


Power Electronics

[Art. 3.8]


2 Vm
Va = --11:-' I" = 11: ' Vor = 12 and lor = .J2
This shows that the rectifier efficiency, FF, ripple voltage V,., VRF are the 3ame for both
types of diode rectifiers . However, PIVof diode in single-phase B- 2 rectifier is Vm whereas
it is 2Vm in I -phase M - 2 rectifier.
T UF: Rms value of source voltage Vs = -J2

Rms value of source current,

VA rating of transformer


VA rating of transformer 11: 2 m m V m 1m 11: 2


Source current waveforms for both types are identical, therefore CF = .f2.
A comparison of three types of I-phase diode rectifiers discussed above is given in the table
below where Vm = -Ii" V" H2re Vs = rms value of sinusoidal source voltage and f = source
frequency in Hz.



(o r one-pulse)


DC output voltage,


Vcltage ripple factor,



efficiency, 11

Rms value of output


utilization factor,


Peak inverse
voltage, PIV


Cr est factor, CF


10 ,

Bridge (B- 2)

Centre-t,-zp (N!2)



Ivoltage, V o,.
I Ripple voltage, V,.


- Vm


Full-wav e (or Two pulse)

'2 V

- - in









0.3856 Vm

0.3077 Vm

0.3077 Vm







40 .53%






'11i !


Number of diodes

Ripple frequency




It is seen from the above table that both u ll-wav e diode r ectifi ers

are better th an the half-wave rectifier in so fa r as vol tage ripple factor, nctificaticn
efficie ncy, T UF and cr est factor ar e concerned,

[Art. 3. ]

Di ode Circuits a nd Rectifiers


(ii) have average output voltage double of that of the half-wave rectifier (for the same

input voltage),
(iii) have ripple frequency double of that of half-wave rectifier.

For both the full-wave rectifiers, the following is observed from the table . .

of B-2 rectifier is superior than the M - 2 type. Therefore, transformer

required in M - 2 configuration is bulky and weighty.
(ii ) PlV of diodes in B-2 rectifier is half of that of the diodes used in M ~ 2 rectifier.
(iii) B-2 rectifier requires four diodes whereas M - 2 requires only two diodes
(iv) Overall, a bridge rectifier using four diodes is more economicaL
(i) TUF


Example 3.15. A load of R = 60 n is fed from I-phase, 230 V, 50 Hz supply through a

step-up transformer and then one diode . The transformer turns ratio is two. Find the VA rating
of traTLsformer.
Solution. The half-wave diode rectifier uses a step-up transformer therefore, ac voltage
applied to rectifier = 230 x 2 = 460 V = Vs


Average value ofload voltage, Vo = p

Output dc power,


-J2 x 460

= 207.04 V

= V;
= 207.04
= 714 .43 W

It is seen from the table that TUF for I-phase half-wave diode r ectifier is 0.2865.

Pde = 714.43 = 249365 VA

:. VA rating of transformer

So choose a transformer with 2.5 kVA (next round figure) rating.

Example 3.16. A 230 V; 50 Hz supply is conne cted to a I-phase transformer which feeds a
diode bridge as shown in F ig. 3.23 (a). Primary to secondary turns ratwIor transformer is 0.5
and load RL has a ripple free current 10 = lOA. Determine (i) average value of output voltage (ii)
input current distortion fa ctor (iii) inp ut displacement factor DF (iv) input power factor (v) input
current harmonic factor HF (o r T HD) and (vi) crest factor.
Solution. Waveforms for supply voltage vs' constant load current io = 10 = 10 A and source
current is are shown in Fig. 3.23 (b).
Rms value of input voltage to bridge rectifi er,

~l = 2~0 = 0.5

vs = 230
= 460 V
The sourc e current, or input current, is can be expressed in Fourier series as under :

is::::: Ide

2. (ancos n oot + bn sin noot)



Here Ide = de v alue of source current

.. .(3.56)

Power Elec tronics

[Art. 3.8]


It can also be stated from the waveform of is that as the area of positive and negative half

cycle are equal,~average value of is i.e. Ide = O.

1 .2 TC
an = - J io(t). cos nrot. cosnrot. d(rot)

= -2



10 cos nwt. d(wt) = - 0


bn = 1.



sin nrot d(rot)


' nwt Ilt0 =:: 0


lor a 11 n.


10 , sin nwt. d(wt)

[- cos nrot]oIt = --- 0 [1 - cos n1t]
for n = 1, 3, 5 ....... (for odd values of n)



=:: 0



= 2, 4, 6 ....... (for even values of n)

Substituting the values of Ide' an and bn in Eq. (3.56), we get

L=:: 41oSinnrotand<l> =::tan-1[jL]=::0

, nIt





1 ..
sm rot + 3" sm 3wt + 5 8m 5rot + "7 sm


Average value of outp ut voltage, Vo


2{2 x 460



twt + ... ]i

= 414.08 V

(ii) Since fundame n t al component of input source current --..R sin rot is in phase with source

voltage V m sin wt, the displacement angle

= O. Also, from

:. Input displacement fact or, DF = cos $1

= cos

above, <1>1 = O.

= 1.

(iii) Rms value of fund am ental component of source current, lsI

of source current, I, [ I;: R

1'" ~ ~ 10 A.

Input current dIst ortIOnal factor, CDF = -I = --;;::-2 x -I

(iu) In pu t



(u i)

= ~o X :J2. A. Rms value


r~ [(019 1r~

= 12 x 2 = 0.9

pf= CDF x DF = 0.9 xl = 0.90 (lagging)


[ (

" J' - 1

J' -

0.4843 or 48.43%

Crest fa ctor. H ere Jsp = 10 = 10 A and I s = 10 A.


CF= 10

= 1.00

Examp le 3.17. A single-phase B 2 diode rectifier is required to supply a de output voltage
J of 230 V to a load of R = 10 n. De termine the diode ra ti ngs and transfo rmer rating required

for this configuration.

Solu ti on.

Average, or d e output voltage,

2V.7l 2{2. Vs

V0 = - =


[Art. 3.9]

Diode Circuits and Rectifiers

:. Rms value of input voltage to rectifier

= transformer secondary voltage,

Average load current,

1m sin rot. d(rot)

and rms value of diode current,

2555 V


3613 A
= -f2 x10255.5 =.

It is seen from the waveform of diode current

diode current is
1t o


= Vo = 230 = 23 A

l e current, 1m = Vm
vaIue O

IDAv = 21



from Fig. 3.23 (b) that average value of

=1m = 36.13 = 11.50 A






=[ 2~ I~l sin




=; =36213 = 18.07 A

PN =-f2vs =-f2 x 255.5 = 361.3 V


Transformer secondary current

Transformer rating

= ~ = 3~3 = 25.55 A = Is
=Vs Is = 255.5 X 25.55 = 6528 VA = 6.528 (kVA)
Pdc = Vo 10 = 230 x 23 = 5290 W

:. Transformer rating

P de


= - - =- - = 6.530 VA = 6.53 kVA 1


Thus, diode ratings are :

I DAv = 11.50 A, I Dr = 18.07 A

Peak diode current, 1m = 36.13 A and PIV = 361~3 V and transformer rating = 6.528 kVA.


The highest possible value of average output voltage from a single-phase full wave (ectifier
is 2V ml1t = 0.63662 V m' At the same time, single-phase rectifiers are suitable up to power loads
of about 15 YW. For higher power demands. three-phase rectifiers are preferred due to the
following reasons :
(i) Higher dc voltage
(ii) Better TUF
(iii) Better input pf
(iu) Less ripple contBnt in output current; therefore better load performance and
(v) lower size of filter circuit parameters because of higher ripple frequency.
Three-phase rectifier are classifIed as under:
(a) Three-phase half-wave rectifier
(b) Three-phase mid-point 6-pulse rectifier
(c) Three-phase bridge rectifier and
(d) Three-phase 12-pulse rectifier.

These are now described one after the other.

. 3.9.1. Thre ,e-PhaBe Half-wave Diode Rectifier

C'r cuit diagram of a three-phase h alf-wave rectifier using three diodes is sho'wn in Fig.
3.29. It uses a 3-phase transfonn er with primary in delta and se condary in star. Th e primary
in delta provides a path for the triplen harm onic currents. This st abilizes the voltages on the


Power Electronics

[Art. 3.9]
vD I



La 01







Fig. 3.29. Three-phase half-wave diode rectifier with common cathode arrangement.

secondary star. The three diodes 01, D2 and D3, one in each phase, have their cathodes
connected together to common load R. Neutral is used to complete the path for the return of
load current. As the cathodes of three diodes are connected together, circuit of Fig. 3.29 is also
known as common-cathode circuit for a 3-phase half-wave rectifier. The three-phase supply
voltage is shown as va (= Van' voltage between a and n), vb' Vc in Fig. 3.30 (a).



rt /2


( b)

10 3




5rr /2

. ,;


. . . . --._.-

",,,,'<. .......





. . . . . ----*"---~""



(e )


........ ,




. . . . . . _.




......... ~

~ ...

0.5 Vmp

Vo ltage of termina l P



Voltage of neutral n

( d)


17/ 2




. .151LD1

(e) ~ . ~_~__
iO1 __________~______~~L-~
_________ _


~---------------2rr=36 0 '--------------~'

150 '

( I)

270 '

O r--.----------~~--~------_r~---------,------------~---w--t

Fig. 3. 3'0 (a ) Line t o eutral so r ee voltages (6) dio de conduction (c) loa d volt age
sourc e cur r ent and (f) volta ge across diode D l.

(d )

load current (e )


[Art. 3. 1)]

Diode Circuits an d Rec tifiers

The rectifier element connected to the line at the highest positive instantaneous voltage
can only conduct. In Fig. 3.29, a diode with the highest positive voltage will begin to conduct
at the cross-over points of the three-phase supply. It is seen fro m Fig. 3.30 (a) that diode Dl
will conduct for CDt = 30 to CDt = 1500 as this diode senses the most positive voltage :Jet' as
compared to the other two diodes, during the interval. Diode D2 will conduct from wt = 150
to 270 0 and diode D3 from CDt = 270 to 390. The conduction of diodes in proper sequence is
shown in Fig. 3.30 (b). When a diode is conducting, the common cathode terminal P rises to
the highest positive voltage of that phase and the other two blocking diodes are reverse biased.
The voltage Vo across the load follows the positive supply voltage envelope and ha.s the
waveform as shown in Fig. 3.30 (c). It should be noted that voltage of the neutral point 'n' is
taken as zero and is given by the reference line wt. The voltage of point P in Fig. 3.29 is shown
by va' Vb' Vc etc above the reference line in Fig. 3.30 (c). The de load voltage v(J varies between
V mp ( = maximum phase voltage) and 0.5 V mp' It is observed that for one cycle of supply voltage,
output voltage has three pulses, the circuit of Fig. 3.29 can therefore be called a 3-phase 3-pulse
diode rectifier or 3-phase half-wave diode rectifier.
Voltage variation across diode D1 can be obtained by applying KVL to the loop consis ting of
D1, phase 'a' winding and load R. So VDl - Va + Vo = 0 or VDl = Va - Vo
When diode D1 conducts,
CDt = 30 to CDt = 150 in Fig. 3.30 (j).


= v o'



= Va -


= O.

This is

shown from

wpe.n diode D2 conducts, Vo = Vb.

vnl = Va - Vb
wt = 180, vb= 0.866 V mp and va = 0,:. vDl = 0 - 0.866 V mp = - 0.866 V mp
CDt = 210, Vb = V mp and Va = - 0.5 V mp '
:. vDl = (- 0.5 -1.0) V mp = - l.5 V mp
CDt = 240, Vb = 0.866 V mp and Va = - 0.866 V mp ,:' vDl = (0 - 0.866 - 0.866) V mp = - \f3Vmp
CDt = 270, Vb = 0.5 V mp and Va = - V mp :. vDl = (- 1 - 0.5) = - 1.5 V mp '
When D3 conducts, V Dl = va - Vc and variation of voltage unl from wt = 270 to wt = 390 is


obtained as outlined before. It is seen from Fig. 3.30 (f) that peak inverse voltage across diode
D1 is ,[3' V mp ; in general prv = '1/3 V mp for each of the three diodes D1, D2 and D3.
AB in single-phase rectifiers, the average output voltage in a 3-phase diode rectifier can be
obtained by considering the output voltage over one periodic cycle.

For a 3-phase diode rectifier of Fig. 3.29, the peri.odicity is 120 or 27t/3 radians as per Fig.
3.30 (c). Here the output volt age comprises of phase voltages Va' Vb, Vc and its average (or m ean)
value V o is given by

V = Are a over one periodicity (shown cross-hatched in F ig. 3.30 (c)

P eriodicity

= p ena.~.lCl't y



d (CDt)

In th e above expression, va is zero at CDt = 0; t herefore, V mp sin wt is written for v~ . Further

Va appears from OJt = 30 to 150 in the ou tput voltage waveform; these are, therefore, the limits
of integration and periodicity is 120 0 = 21t/3 r adians.
3 f5it/6
3-16 .
.. (3 71)
Vo = -2
Vmp sm CDt. d (CDt) = -2- V mp = - 2- Vph = -2 . V ml
n rr~
where V mp = maximum val e of phase voltage, ~ = -f2 Vph
V mi = m aximum value of line voltage, VI = . V mp = '1/6. Vph


Power Electronics

[Art. 3.9]

=[ 2 1 3 j/t/6 (Vmp sin wtf d(wt) ]112

Rms value of out put volt age, Vor



=[ 21t
3 V~p I t _
x 2' w


sin 2wt

= 0.84068 V mp

... (3.72)


Ripple voltage,


= -lv~1' - V~ = V mp -lO.84068
VRF = Va


= 0.151



= 0.827 = 0.1826 or 18.26%

FF = VOl'

= 0.84068 = 1.0165

Rms value of output current, lor = ~r

= 0.8~068

V mp

= 0.84068 Imp

V mp
where Imp = R = peak value of load, or output, current.



Pdc = Va 10 = 21t V mp' 21t Imp


=Vor . lor = (0.84068)2 V mp Imp

=Pde = [ 313 )2x

Rectifier effid en cy



= 0.96765
0.84068 2

or 96.765%

During the interval wt = 30 0 to wt = 150, source current is = ia and the periodicity of source
current is 21t radian s as shown in Fig. 3.30 (e).


:. RMS value of source current, Is

Rms value of source voltage, Vs




(Imp siD: wt)2 d (wt)

]1!2 = 0.4854 Imp

= 0.707 V mp

Transformer h as three phases, therefore,

= 3 Vs Is = 3 x 0.707 V mp x 0.4854 Imp = 1.0295 V mp I mp

VA rating of t r ansform er

DC output power,

Pd ,



x V mp Imp

~ 0.684 Vmp Imp

Transformer VA ratmg

As stated before , Pl V for each diode

= 0.684 Vmp . Imp = 0.6 644

1.0295 V mp . Imp

=-{3 v mp

Following observations can be made fr om th e above analysis.

E ach diod e conducts for 1200 only.
(ii) Th ese are three pulses of output voltage, or output current, dur ing one cycle of input
voltage. It is, therefore, calle 3-ph ase t hree-pul e diode r ectifier.
(i i i) Current in th e transfor mer secondary is unidir ectional, t herefore, de exists in the
transform er second ary current. As a result , transfor mer cor e get s saturat ed leading
to more iron losses and redu ced effi ciency.

[Art. 3.9]

Diode Circuits and Rectifiers


3.9.2. Three-phase Mid-p oint 6-pulse Diode Rectifier

This rectifier is also called six-phase half-wave diode rectifier or three-phase M-6 diode
rectifier. It is seen from the previous section that the performance of three-phase three-pulse
diode rectifier is better than the single-phase two-pulse diode rectifier so far as magnitudes of
average output voltage and ripple content are concerned. From this, it appears to be logical
to think that a rectifier with more number of pulses per cycle would give an overall improved

A 0---..,----.......



c u - - - -__





ic 1


Fig. 3.31. Three-phase mid-point 6-pulse diode rectifier . .

Fig. 3.31 shows a three-phase mid-point 6-pulse rectifier using

six diodes. A three-phase transformer with primary in delta and
secondary in double-star. is used. One diode in each phase is
connected as shown. Note that secondary of each phase winding is
in two halves. The mid-points of the three secondary windings are
connected to form the neutral n. Six-phase supply is available from
six terminals aI' C2, b l , a2, CI and b2 Phase voltages Val, Vbl> Vel are
phase-displaced by 120, similarly V aZ , V bZ , Vcz are displaced by
1200 But voltages Val and VaZ are out of phase by 180 0 However;
Val, V cZ , are out of phase by 60 as shown in Fig. 3.32.
- V
60 0
Th erefore, if v a l -- V mp sID
rot , env c2- mpsm(rot - )




= V mp sin

(rot -120) , va2

= V mp sin (rot vb2 = V mp sin (cDt Vcl

240) = -

= Vmp sin (rot V mp

F i g. 3.32. Six-phase
' CI 331
vo It ages fior F lb'

180) = -

sin (rot - 60) = -



300) =- vbl
Here V mp = maximum value of per phase voltage
The waveform of six-phase voltages , Val' VcZ, Vbl .... are sketched in Fig. 3.33 (a). As before,
a diode sensing the highest positive anode potential gets forward biased and conducts.
Therefore, from wt = 0 to wt = 60 0 , voltage Vb2 is the highest positive , therefore, diode D6
conducts; from wt = 60 to wt = 120, diode. Dl conducts and so on, Fig. 3.33 (a) and (b). Each
diode conducts fo r 6 ". It is seen fr om Fig. 3.33 (c) that load voltage Vo is m ade up of Vb2 from
wt = 0 t o 60 ; Val from rot = 6 0 to 120 0 and so on. Also, Va varies between Vmp and 0.86 6
V m p ' Periodicity of output voltage 1,'0 is 60.


Power Electronics

[Art. 3.9]


( b)




vb 2









v a2




Ve 2






i 02

t bl


te I






( e)

42 0 0






211' =360


270 0




-0.866 Vmp

Fig. 3.33. (a) Line to neutral (or phase) source voltages (b) diode conduction
(c) load voltage (d ) load current (e) source current and (f> voltage
across diode Dl for 3-phase M - 6 diode rectifier of Fig. 3.31.

Average output voltage,


frt /3

Vo -

Rms value of output voltage,


V mp sm rot. d(rot) -

3 V mp



V or =[



=[ 3V21t

frt / 3


rt/ 3

(Vmpsmrot) d(rot)

sin 240



sin 120


Ripple voltage,


FF = ~r

= 0. 95~8 x 1t =1. 009 .

095'"8 V
. U


V, =~~, - ~ =V O.95582_( ~ I ]
VRF = V = 0 . 04~8 x =0.043 or 4 .3%
mp [

= 0.0408 Vmp

... (3.74)

(Art. 3.9]


Rms value of output current, lor = If::::



V mp


=0.9558 Imp

~ V, I, ~( ~ Vm~lmp

Pac = Vorior =(0.9558) V mp Imp

Rectifier efficiency

= ( -3 J2 X

1 2 = 0.9982 or 99.82%

The source current is in phase a l has the same waveform as for the current ial , Fig. 3.33 (d)
and (e). It is seen that periodicity of is is 21t radians.

:. Rms value of source current, I,

VA rating of transformer
DC output power,

~[2~ ( : ' (Imp sin wt)'. d (wi)


f' ~

0.39 Imp

= 6 Vs Is = 6 ~ x 0.39 Imp = 1.655 V mp Imp


P dc


=Vola =-1tX -Imp = 0.912 V mp Imp





= vn ratmgdc0 f trans.f = -655

= 0.551 or

::> .1%.

Voltage variation across any diode, say Dl, can be obtained as done in 3-phase 3-pulse diode \
rectifier. Therefore, VDl = Val - Vo
When D1 conducts, va = Va l '
:. VDl = Val - Val = 0 from wt = 60 to 120.
When D2 conducts from wt = 120 0 to 180, va = vc2 ' :. vDl = Val - vc2
At wt = 150, vc2 = mp an d Lla l = 0.5 V mp ;
:. vDl =(0.5 - 1.0) V mp = - 0.5 V mp
At wt =It, vc2 = 0.866 V mp and Val = 0,
; . VDl =- 0.866 V mp
When D3 conducts from wt = 180 to 240, Vo =Vb I,
:. VDl = Val - Vbl
At wt = 240, Vbl = 0.866 V mp and Val =- 0.866 V mp ' :. vDl =(- 0.866 - 0.866) V mp = - -{3 V mp
When D4 conducts from wt = 2400 to 300 0 Vo = va2 :. vDl = Val - Va2
At wt = 270, Val = - V mp and va2 = V mp'
:. vDl = (-1- 1) V mp =- 2 V mp

Similarly, vDl waveform can be obtained when diodes D5, D6 conduct, this is shown in Fig.
3.33 {f;. It is seen from vDl waveform that PIV for each diode = 2 V mp '
The above analysis reveals the following.
Quality of output is sup erior as compared to 3 pulse rectifier, because RF is 4.3%
and FF is close to unity.
(ii) TUF is poor as compared to 3 pulse rectifier ; it is because of lower value of
conduction angle (= 60) for each phase and diode of this rectifier.
(iii) Output frequency is 6f; siz e of filter, if required, is therefore reduced.
It may be observed from Fig. 3.33 that each phas e winding carries un directional current
and there are six pulses during one cycle of sourc e volt age. That is why it is called six-phase
half-wave diod e rectifier.
At any time, only one secondary phase winding, say phase a 1, carries current, this gets reflected
downward in the primary delta for m mf balance. After furth er c.ut = 1800 , secondary-phase



Power Electronics

[Art. 3.9]

win ding a2 carries current, this gets reflected upward in the primary delta. This shows that
both secondary and primary windings handle alternating current during one cycle of source
voltage; there is therefore no magnetic saturation of the transfcrmer core as it is in the
transformer used in 3-phase half-wave rectifier.
~ 3.9.3. Multiphase Diode Rectifier
For three-phase three-pulse rectifier, or three-phase half-wave rectifier, each phase
conducts for 27t/ 3 radians of a cycle of 21t radians. Three-phase M-6 rectifier may be considered
as 6-phase half-wave rectifier and it is seen that each phase of this type conducts for 27t / 6 rad.
In general, in m-phase half-wave diode rectifier, each phase and diode would conduct for
2 1tl m rad and number of output voltage pulses p would be equal to number of phases m.
For m more than three, an m-phase diode rectifier would have delta-connected primary
and the secondary would have mid-tapped m/2 windings . The number of diodes is equal to m.
Fig. 3.34 shows a few pulses of output voltage waveform for m-pulse half-wave diode rectifier.
Each phase conducts for 21tl m or 21tl p radians, because number of pulses p = number of phases
m for half-wave rectifiers. With time origin AA' taken at the peak value of output voltage in
Fig. 3.34, the instantaneous phase voltage is
v == Vmp cos rot = "-f2 Vp h ' cos rot
where Vph = rms valu e of per-phase supply voltage .
Waveform of output voltage Vo in Fig. 3.34 shows that in m-phase half-wave diode rectifier,
conduction occurs from - ~ to ~, or from - ~ to 7tlp with time origin at AA' and periodicity
is 27tl m, or 27tl p, radians.






~271/P--1 '

Fig. 3.34. Output voltage waveform for m-phase half-wave diode rectifier.

:. Average value of output voltage, Yo = -2-

[ IP

1tl p _ re l p

Y mp

cos rot d (rot)

= Ymp . E..
sin ~
Rms value of output voltage,

Y or = [ L2



.. .(3.75 )

(Ymp cos rot) 2. d (rot)


- 'relp

y2 [IP
= P' mp
(1 + cos 2rot ) d (rot)
[ 41t
_ 'itlp

1 2)J

=V mP1rL21t (~p + -2 sin ~


...(3 .76)

= V mp : 1

Maximum value of load current

Aver age value of diode current,

, 112


- rtlp

1~ cos cot. d

(rot )




... (3 .77 )

[Art. 3.9]

Diode Circuits and Rectifiers

Rmsvalue of diode current, IDr =[2. 11t r-iP


cos wt)2

d (wt)



ltl p







[ =1 ( !!. + -1 sin .J!.

21t p 2


... (3.78 )

Example 3.18.A step-down delta-star transformer, wiih per-phase turns ratio 5, is fed from
3-p hase, 1100 V, 50 Hz source. The secondary of this transformer, th rough a rectifier, feeds a load
R = 10 n. Calculate the average value ofoutput voltage, average and rms values of diode current
and power delivered to load in cas'e the rectifier is (a) 3-phase, 3-pulse type and (b) 3-phase M-6
(a) 3-phase three-pulse type.' Per-phase secondary voltage


Vph == - -5- = 220 V and V mp = '12 X 220 V.

From Eq. (3.71), or from Eq. (3.75) withp == 3, average value of output voltage,
Vo;::;: 21t V mp == 21t X ~ x 22 0 = 257.3 V
Maximum value of load current,

1 = .:..!!!i!. == '12 x 220


=..J2 x 22

From Eq. (3.77), average value of diode current, ID = 22: -f2' sin
From Eq. (3.78), rms value of diode current,

~ 22 xv'2[ ;n(~+ sin;20


~ = 8.575 A

= 15.10 A

From Eq. (3 .76), rms value of output voltage,

sin 2120
3 (1t
Vor == 220 x..J2 [ 21t
- "3 +

Power delivered to load




= 2.61.52 V

26~g22 == 6839. 3 watts

(b) 3-phase M- 6 type.' P er-phase secondary voltage,



= lIOV and V mp =..J2 x llO V

From Eq. (3.73), or fr om Eq. (3."';) withp == 6, average outpu t voltage is


x -Y2 x 110 =148.53 V

Maximum value of load current,


= VRmp = ..J2 10
x 110 = 'if2 X

From Eq. (3 .77), average value of diode current, ID

11 A

= -{2 x 11 ::;in 11:/ 6 = 2.4755 A



Power Electronics

[Art. 3.9J

I l6

1 ( 7t sin 60
Fnm Eq. (3.78), rrns value of diode current, In,. == 11 x --f2 L 27t
From Eq. (3.76), rms value of output voltage,

Power delivered to load

~ = 110 x '1'2[ ;n (~ + sin;oo )




= 6.069 A


148.66 V

14~~62 = 2209.98 = 2210 W

3.9.4. Evolution of Three-phase Bridge R ectifier

Before studying the 3-phase bridge rectifier, let us first examine the evolution of this
rectifier type.
Three-phase half-wave diode rectifier, with common cathode configuration, has already
been discussed in Art. 3.9.1.
For the circuit of Fig. 3.29, the conduction of diodes Dl, D2, D3 indicated in Fig. 3.30 (b), is
again shown in Fig. 3.36 (b), just for the sake of convenience in understanding the evolution of
3-phase bridge rectifier.
. ', .
Consider now a three-ph&se h alf-wave diode rectifier with common-anode arrangement as
shown in Fig. 3.35. In this circuit, diode will conduct only during the most negative part ofthe
supply voltage cycle. This rrieans that a diode will conduct when the neutral i,s positive with
respect to terminal a, b or c. Therefore, for the supply waveform of Fig. 3.36 (a), qiode D5 would
conduct from CDt = 0 to CDt == 90 as the voltage Vb is the most negative for this m terval. From
CDt = 90 to CDt = 210, voltage Vc is the most negative, therefore diode D6 must cO,n duct during
this interval as shown in Fig. 3.36 (c). Similarly, diod~ D4 would conduct from CDt = 210 to
CDt = 330 0 and so on. E ach diode conducts for 1200 (as in common-cathode configu'ration). The
load voltage waveform v o' der ived from Fig. 3.36 (a), follows the negative supply voltage
envelope as shown in Fig. 3.36 (d) for the diode configuration of Fit;>. 3.35. Voltage of neutral n
is fixed at zero by th e reference r ne rot in Fig. 3.36 (d). Th e voltage orterminal Q of Fig. 3.35 is
shown by vb, V c ' Va et c below th e reference line in Fig. 3.36 (d) .




c o--



8 O-------------~


Fig. 3.S5. Three-phase half-wave diode rectifie with common anode arrangement.

The three-phase half-wave rectifier circuits of Figs. 3.29 and 3.35 can be connected in series '
as shown in F:ig. 3.37 (a). An examination of this series connected circuit r eveals that load
current can exist even without neutral wire n. Fo e ampl e, when diode Dl is conducting from
wt = 30 to W: = 150 in Fig. 3.36 (b ), the return path for the curren t is through diod e D5 fr om
wt 30 C to 90 and through dio e D6 from cot = 90 C to 1500 I s ee F ig. 3.36 (b) and (c) . Supply point


[A r t. 3.9]

Diode Circuits a nd Rectifiers


( b)

1 03













Vo ltage oi neut ral n


Vol tage of terminal Q






Fig. 3 .36. (a) Line to neutral source voltages (b) diode conduction for

Fig. 3.29, (e) diode conduction for Fig. 3.35, (d) load voltage.

'a' connected to the anode of Dl is the same as that connected to the cathode of diode D4. The

neutral wire can thus be eliminated and cathode terminal of D4 can be connected to anode of
Dl. Thus, the circuit of Fig. 3.37 (a) can be redrawn as shown in Fig. 3.37 (b). This circuit can
further be rearranged to that shown in Fig. 3.37 (c). The only difference between Fig. 3.37 (a)
and Fig. 3.37 (b) and (c) is that load voltage is equal to line to neutral voltage in Fig. 3.37 (a)
and it is line to line volt age in Fig. 3.37 (b) and (c). The circuit configuration shown in Fig. 3.37
(c) is called 3-phase full waue bridge rectifier, or 3-phase six-puLse bridge rectifier. Note that
diodes Dl, D2, D3 of the bridge would conduct when supply voltage is the most positive,
whereas diodes D4, D5, D6 would conduct when supply voltage is the most negative. Diodes Dl,
D2, D3 may therefore be called a positive diode group and D4, D5, D6 a negative diode group.
The voltage across load would always be the direct emf with the polarity of P positive and that
of Q negative as indic a ed.












(a )


(c 1




F ig. 3.3'7. Evol ution of 3-phase six-pulse rectifi er (a ) circuits of F igs. 3.29 and 3.35

connected in series, (b) circuit obtaine d fro m (a ), (e) circuit of

(b) rearranged and (d) diode numbering scheme alt ered.



Power Electronics

IArt. 3.9]

It may be seen from Fig. 3.36 (b) and (c) that conduction sequence of diodes is D1 (from
positive group), D6 (from negative group), D2, D4, D3, D5 etc. In order that sequence of
conducting diodes is D1, D2, D3, D4, D5, D6, D1. .... (easy to remember), circuit of Fig. 3.37
(c) is redrawn in Fig. 3.37 (d) with diodes numbered D1, D3, D5 for positive group and
D4 (1 + 3), D6 (3 + 3), D2 (5 + 3 - 6) for the negative group.
Line to neutral, or phase, voltages of Fig. 3.30 (a) or 3.36 (a) are redrawn in Fig. 3.38 (a)
as va, Vb' vC ' Fig. 3.36 (b) and Fig. 3.36 (c) are combined and drawn in Fig. 3.38 (b) but with
diode- numbering scheme of Fig. 3.37 (d) . It is seen that for wt = 0 to 30, diodes D5, D6 conduct
together, for wt = 30 to 90, diodes D1, D6 conduct together and so on. Each diode conducts
for 120. At the instant marked 1 (when wt = 30), diode D6 is already on, conduction of diode
D5 stops and that of diode D 1 begins. The magnitude ofload voltage VI at instant 1 is, therefore,
given by

= V rnp sin 30 (from .va) + V mp sin 90 (from Vb) = 1.5 Vm~~''"'

At the instant marked 2, the load voltage has a magnitude ofV2 = V mp sin 60 + V rnp sin 60

=f3vmp '
At the instant marked 3,

V mp


= 1.5 V rnp

= maximum value of phase

(or line to neutral) voltage.

The voltage of the load terminals P and Q of Fig. 3.37 (c) or (d) is shown in Fig. 3.38 (a).
This figure also reveals that at the instants marked 2, 4, 6, 8, 10 etc, the load voltage has a
magnitude off3Vmp ' At the instants marked 1, 3,5,7,9, 11 etc, the magnitude ofload voltage
is 1.5 V mp' The load voltage, or the rectified output volt age, Va can therefore, be plotted as
shown in Fig. 3.38 (c). In this figure, voltage ofterminal Q is shown at zero potential by straight
reference line wt, whereas the potential of terminal P is shown by line voltages v cb' vab' Vac etc.
In fact, if voltage waveform of terminal Q in Fig. 3.38 (a) is made a straight li~~, Fig. 3.38 (c)
Vo ltage of terminal P

(a )
Voltage of ter minal Q









+ve group

I 02

-ve grou p


/ '




3C' '4

1.5 Vmp

.JJV m p


1 cycle =277 =360

a 9






Fig. 3.38. (a ) 3-phase input voltage waveforms (b) conduction sequence of positive an negative
group of diodes (c ) output vo ta e wav eform of 3-phase six-pulse diode bridge of Fig. 3.37 (d) .


[Art. 3.9]

Circuits and Rectifiers


is obtained. It should be remembered that in Fig. 3.38 (c); vab, v ae ' v bc et c are line voltages,
whereas in Fig. 3.38 (a) ; Va' Vb, Vc are phase voltages. The dual subscript ab in vab should be
taken to denote that as per the first subscript 'a', diode connected to phase termin al 'a' from
positive group, i.e . D1 conducts. Asper the second subscript 'b', diode connected to phase
terminal 'b' from the negative group, i.e. D6 conducts. For example, for output voltage Vcb, diode
D5 from positive group and diode D6 from negative group conduct. Note that each diode
conducts for 120.
Fig. 3.38 (c) reveals that there are six pulses for one cycle of supply voltage. Thus ,
three~phase bridge rectifier of Fig. 3.37 (d) can be called 3~phase six~phase diode rectifier or
3-phase B-6 diode rectifier. Here B denotes bridge and 6 denotes the n umber of output~voltage
pulses per cycle of source voltage.


It is thus seen from above that when two 3~phase 3~pulse rectifiers are connected in
antiparallel, a 3~phase 6 pulse rectifier (or a 3~phase bridge rectifier) is evolved.


3.9.5. Three-phase Bridge Rectifier

Power circuit diagram for a 3~phase bridge rectifier using six diodes in shown in Fig. 3.39.
The diodes are arranged in three legs. Each leg has two series-connected diodes. Upper diodes
Dl, D3, D5 constitute the positive group of diodes. The lower diodes D2, D4, D6 form the
negative group of diodes. The three~phase transformer feeding,.the bridge is connected in
delta-star. This rectifier is also caned 3~phase 6~pulse diode rectifier, 3-phase full-wave diode
rectifier, or three-phase B-6 diode rectifier.
Positive group of diodes conduct when these have the most positiv e anode . Similarly,
. n egativE) group of diodes would conduct if these have the most n egative anode. In other words,
diodes D1, D3, D5, forming positive group, would conduct when these exper ience the highest
posi tive voltage. Likewise, diodes D2, D4, D6 would conduct when these are subjected to the
m ost negative voltage.


A 0 > - - - -....







Fig. 3.39. Three-phase bridge rectifier using diodes.

It is seen from the source voltage waveform V s of Fig. 3.40 (a ) that from rot = 30 to 1500 ,
voltage va is more positiv e than the voltages v b' vc' Therefore, diode D1 connected to line 'a'
(as p.er subscript 'a' in va) counducts during the int erval rot = 30 to 150. Likewise, from
rot = 150 to 27 0, voltage vb is m or e positive as compared to va) V c; therefo e, diode D3 connect ed
to line 'b ' (as per the subscript 'b' in Vb) conducts during this interval. Similarly, diode D5 from
th e positive group co .ducts fr om cut = 27 00 t o 390 0 and so on. N ote als o that from (ct ,. 0 to
30 , Vc is the most positive, therefor e, diode D5 from t he positive group conducts for this


Power Electronics

[Art. 3.9]

interval. Conduction of positive group diodes is shown in Fig. 3.40 (b) as DS, Dl, D3, DS,Dl
Voltage ve is the most negative from rot =90 0 to 210 0 Therefore, negative group diode D2
connected to line 'e' (as per subscript 'e' in ve ) conducts during this interval.Similarly, diode
D4 conducts from 210 0 to 330 0 and diode D6 from 330 0 to 450 0 and so on. Note also that from
rot = 00 to 90 0 , Vb is the most negative, therefore diode D6 conducts during this interval.
Conduction of negative group diodes is shown as D6, D2, D4, D6, etc in Fig. 3.40 (b).
During the interval rot = 00 to 300 , it is seen from Fig. 3.40 (b) that diode D5 and D6 conduct.
Fig. 3.39 shows that conduction of D5 connects load terminal P to line terminal e; similarly,
conduction of D6 connects load ter minal Q to line terminal b. As a result, load voltage is
Vpq = Vo = line voltage veb (first positiv~ subscript corresponding to e and second negative
subscript corresponding to b) from rot = 00 to rot = 300 Likewise, during rot =30~ to. ,90 0 , diodes
Dl and D6 conduct. Conduction of diode D1 connects P to a and D6 to b. Therefore, load voltage

(al .









+ 'Ie group
102 -'Ie group






or is .



--130 ~30~


1 T - - - -,


. L I~

~06 01 ~ 01





O r-~-----------'--------~~----------~~--------~-w~t



. Fig. 3.40. Three-phase diode-bridge rectifier

3-phase inpu t voltage waveform
(b) conduction sequence of diodes (c) output voltage waveform
(d) input current waveform (e) diode current wave onn through Dl and
(j) vo tage variation across cHode Dl .

[Art. 3.9]

Diode Circuits and Rectifiers


during this interval is Vo = line voltage vab' Similarly, for interval 90 to 1500 , diodes
D1 and D2 conduct and Vo = line voltage Vac; for interval 150 to 210,.diodes D3 and D2 conduct
and Vo = line voltagevbc and so on . Output, or load, voltage waveform is drawn by a thick curve
in Fig. 3.40 (c).
Average value of load voltage, Vo

= peno
. 1d'lClty
. J~ vab . d


It is seen from Fig. 3.40 (c) that the value

is 60 0 or 1t/3 rad.


at rot = 0 is V ml . sin30 and its periodicity

.. .(3.79 )



where V ml = maximum valu e ofline voltage

V, = rms value of line voltage and

= rms value of phase voltage.

Average value of output voltage, VOl can also be obtained as und~r.: .

to 120. It is because the voltage7

pulse area required extends from rot = 60 0 to rot = 120 0 for the sin rot function.
Vo = V ml . sin cut. d (rot) = ~

... (3.79)
1t ~3

(i) Take any sinusoidal wave and integrate it from 60


(ii) For a cosine function cos cut, voltage pulse of 60 duration extends 30 to the left of
its peak and 30 to t he right of its peak.
. ..
_ -Vo=-

[16 Vmlcos(j)t

. 3V .

1t -~6

...(3 .79)


[~ j7t/3~l sin2cut. d (rot) ]112

Rmsvalue of output voltage, V or=

1t lt/3


.Bipple voltage,
Voltage r ipple factor,



;1 I rot - s:n: I .



lt/ 3

~ "V;, - V; ~[ 0.9558 ~ )'


- (

r - 0.0408 - 0 042 7
- -V
497C1.y O
VRF -:
V - 3 /1t - .

FF = Vor = 0. 9558 V mt = 1.0009



Rm value of output current, l or = ;r = 0_ 9~58 V m l = 0.9558 1m l

= 0.9558 Vml


~ 0.0408 V


Power Elec.tronics

[Art. 3.9]


=(~ 1V

P d<= V,I.

Pac = Yo,.


ml : Iml

(6.9558) V ml .Iml

Vor =

=, (3)21
' "1t x .O.9558 =0.9982 or 99.82%

Rectifier efficiency

it = {:3v

mp. It is seen from

For a resistive load, peak current through each diode is Im[=

the waveform of line current iu (or transformer secondary current is) that (0 periodicity of this
current is 1t rad. and (ii) this current has two pulses, each of 60 durati~n, for each periodicity
of 1t rad .
.. Rms


of line current

= rms
Is =

Transformer VA rating,

value of transformer secondary current

r2rr / 3

I - JI

I[ 1t

. 2

Iml sm rot . d (rot)

rr/ 3

]li2 = 0.7804
,, Imt

Vml '
=3Vs I s = 3 x ~
x 0.7804 I ml






... (3.81)

' "

=(,2.) x 3 x fa
, =0.9541

It is seen from diode-current waveform of Fig. 3.40

(d ) that average value
ofdiode current


2 ]27t13 '
I D =-2

Rms value of diode current, I Dr =[ 22


rr/ 3

tJ(/ 3
rr/ 3



I~ll sin


rot . d (rot)

.~ .(3~. 82)


]112= 0.582 lml



A conducting diode has zero voltage drop across it. Let it be required to sketch in Fig. 3.40
If), the variation of voltage across diode D1 belonging to positive group in Fig. 3.39.
For de output voltage Vab' vae ; diode Dl conducts, therefore, ' voltage across diode D1 is "
zero, i.e. vDl = 0 from rot = 30 to 150 and this is shown in Fig. 3.40 (f). After rot :: 150,diode
D3 conducts for 120 0 with output voltage vbc' Vba ' Now cathode of Dl gets conriected to supply
terminal b through conducting diode D3 for a period of 120 0 , whereas its anode is already
connected to supply terminal 'a'. Therefore, voltage across D1 from wt = 150 to 270.0 is ,
v Dl = va - V b as per KVL CVm - Va + Vb = 0). This voltage Vab during interval of rot ,= 150 to
270 reverse bia es diode D1 and therefore, voltage Vab is shown bel'ow the reference line in
Fig. 3,40 (f> . Examin ation of Fig. 3 40 (c) reveals that waveform Vba from rot = 150 6 to '270 ,
when reversed, becom es v ab in Fig. 3.40 (f>.

After wt = 270, diode D5 conducts for 120 0 an d now voltage VDl = Vac from wt = 270 to
390 This voltage is sh own as v ac and as a reverse bias across diode D'1, so vae is sk etched
below the re f~.ren ce line in Fig, 3.40 (f>. As before, voltage wa veform 'vcu from rot = 27 0'0 to
390 0 in Fig, 3.40 (c), when reversed, becomes Vae in Fig. 3.40 (f; ,

Fig. 3.40 (() revesals th a t PIV fo di ode D1, or any other diod e, is {:3 Vmp =
value of lin e vol tage .


= m aximum

10J .

[Art. 3.9]

Diode Cir cuits and Rectifiers

An examination of waveforms in Fig. 3.40 (a) and

reveals that line voltage Vab leads phase voltage va by

300 ~ Similarly, vbc leads vb by 30 0 .and VQl leadsv c by

. ..30 0 A phasordiagramshowing 3-phase phase voltages

. .. V~, Vii, Vc and the cqrresponding 3-phase line voltages

Vah (=Va - Vb);Vbc (= V~ - V c), Vea (= Ve -:- Va) is drawn

in Fig. 3.41. Line vol~ge Vba is 180 0 away from Vab as

shown. Similarly, line voltages Vcb and VClC are shown in

Fig. 3.41, thus resulting in a six-phase system of line

voltages Vcd" Vue' Vbc ' Vba,Vca ' V cb ' Phasor diagram of

. Fig, 3.41 reveals thatline voltage Vab leads phase voltage

Vet by 30 0, similarly V&: leads Vb by 30 0 and Ven leads

V~ by 30 ; this matches with the waveforms in Fig. 3.40

Fig. 3.41. Six-phase line voltage

as . expected.
Vab. Vae:. Vbe: etc for secondary winding of
delta-star transformer of Fig. 3.39.
Note that Va = Vb= Vc ~ phase voltage, Vph and

Vab =V be == V ca = Vba =Vcb = Vae = line voltage, VI'


3.9.6. rhree-phase TWelve-pulse Rectifier

It has been stated before that as the number of pulses per cycle are increased, the output
de waveform gets improved. So, with twelve pulses per cycle, the quality of output voltage
waveform would definitely be improved with low ripp!e content.

Fig. 3.42 show~ th e circuit diagram for a 3-phase twelve-pulse rectifier using a total of
twelve diodes. A 3-phase transformer with two secondaries and on e delta-connected primary
.feeds the diode rectifier circuit. One secondary winding is connected in star and the other is

in delta. Star-connected secondary feeds the upper 3-phase diode bridge rectifier 1, whereas

the deltaconnected secondary is connected to lower 3-phase diode bridge rectifier 2. Each

bridge rectifier uses six diodes as shown. The two bridges are series connected so that net

output, or load, voltage va = output voltage of upper rectifier, VOl + output voltage of lower

r~ctifier V02'


110 I




80-- -


Fig. 3.42. Three-phase twelve-pulse Tectifier.


Power Electronic

[Art. 3.9]

If Val, V bl , Vel are phase for the upper star, then upper line voltage Vabl (= Val - V bl ) would
lead Val by 30 0 as i~ a three-phase bridge rectifier Of Fig 3.39. Similarly, line voltages
V bel' Veal would lead by 30 0 their corresponding phase voltages V bl and Vel respectively. In Fig.
3.43 (b) are shown all the six line voltages Vab l ' V acl ' V bcl ' Vbal, Veal , Vebl for the upper
star-connected secondary.




:--Vertica I

Fig. 3.43. Voltage phasor diagram for (a) primary line voltages (b) line voltages

for secondary star and (c) line voltages secondary delta.

Phase voltage V a l of upper star must be in phase with primary line, or phase, voltage
as per the tran sfor mer principle. Likewise, line, or phase, voltage Vab2 of the
delta-connected secondary must be in phase with VAl' All the secondary line voltages
V ab2 , V ~2' V bc2 ' V baZ , V ea2 , Veb2 for the delta-connected secondary are shown in Fig. 3.43 (e). The
line voltages V abl for upper rectifier and V ab2 for lower rectifier are phase~displaced by 30
with Vab1leading Vab2 by 30 0 In case input line voltages to upper rectifier 1 and lower rectifier
2 are superimposed ; line voltages Vabl' Vab2' V acl ' Vac2 etc would be obtained; these line voltages
would be phase-displaced from each other by 30 0


In Fig. 3.44 (a) are sketched waveforms for voltages available across upper-star-connected

Six-pulse de output voltage


obtained from upper rectifier 1 is shown in Fig. 3.44

(b); this is identical to the output voltage waveform of3-phase B-6 rectifier of Fig. 3.39. Lower

r ectifier 2 also gives six-pulse de output voltage V02 as shown in Fig. 3.44 (e). As V ab2 lags
. V abl by 30 , peak of vab2 of v 02 is also shown lagging p eak of vabl of VOl' Since the two rectifiers
are serie-s-connected, net output, or load, voltage Vo = UOI + v 02 is obtained by adding the
corresponding ordin ates of VOl .a nd v02' Note th at waveforms of VOl and V02 are phase-shifted
from each other by 300 , therefo e, waveform of output voltage Vo consists of twelve pulses per
cycle of supply voltage.
Peak value of output voltage, Uo =V ml cos

V ml


+ Vml cos 15 = 1.932 V ml

= m aximum

value of line voltage availab e from star-conn ected, or delta-con

nected, s econdary.

For the sake 0 convenience, let the p eak val ue of t welve-puls e output voltage Vo be den oted
by Vp (= 1.932 Vmi)' P eriodicity of Vo is 300 = 11:/6 radians .


[Art. 3.9]

Diode Circuits and Rectifiers






________L -_ _ _ _ _ _

_ _ _ _ _ _ _ _~
- ~
-_ __ _ _ _ _ _~





.J /











yp = l ,~32Vmi\
.: :

1.866 Vmi







Fig. 3.4.,1, Waveforms for (a ) voltages across star-eonnected secondary (b) six-pulss output

voltage vO l from upper bridge 1 (c ) six-pulse output voltage v02 from lower bridge 2 and

(d) resultant twelve-pulse output voltage Va obtained from ~01 + V02'

:. Average value of output voltage, Vo =-



' Vp sin rot , c!. (rot) = 0.98861'6 Vp

1t ,75


= 0,,988616 x 1.932 Vml = 1.91 Vml

Rms value of outp

volt age,

VOl' =

... (3. ,8 3)

[~1': J75os ~ sm2ox d (c.ct) ]112 =0.988668 Vp


R ipple voltage,
Voltage ripple factor,

=0.988668 x 1.932 Vm = 1.9101 Vml

.. . (3. ~4)
Vr = ~V~ - V! = [1.9101 - .91 ]1/2 Vml = 0.019545 ml ,
Vr 0. 019545 . V ml
VRF =--V = 191 V
=0.01023 or,l.023%




Power Electronics

[Art. 3.9]


FF =

Form' factor,

~: = \~~~1 = 1.00005

As voltage ripple factor is sufficiently small, the output voltage from 12-pulse rectifier is
almost pure de voltage.

A comparison of various 3-phase diode rectifiers discussed above is given in the table below:

I = maximum
value of line voltage = ..J3 V mp where V mp = maximum value of line to neutral,

n ,
. or phase, voltage].

I S.No.



DC output voltage, Vo


6-pulse rectifier
B-6 type


3 ..f3vntp
3 Vml

re .

1.91 V ml

6-pulse rectifier
M-6 type


3 V mp






1.9101 V ntl

Rms output voltage,


0.4854 V ml

0.55185 V ml

0.9558 V ml

Ripple voltage, VI'

0.0872 Vml

0.02356 Vml

0.0408 Vml


Voltage ripple factor,


0.1826 or

0.043 or 4.3%

0.0427 or 4.27%


Rectifier efficiency, 11











V ml

1.155 V ml

V ntl


Form factor, FF





0.019545 V ntl .
0.01023 or


It is seen fr om this table th at quality of dc output voltage is improved significantly with

twelve-pulse rectifier. Out of 6-M and 6-B configuration s , 3-phase 6 -B rectifier is
f>. su~r lOr so far as average output voltage, TUF and PlV are connected .
.Q., 0 Exam ple 3.19. A 3-phase bridge rectifier, using diodes , delivers power to a load of
R = 10 n at a de voltage of 400 V Determ ine the ratings of the diodes and of the three-phase .
deltastar transformer.



Solution. It is given that de output voltage,

V o = -1t-ml= .400 V
: . Maximum value ofline voltage,


=. 4003x 7t = 418.88 V

:. Rms value of phase voltage for share-secondary,

V =


'l2Vx Ts = 418.88
T6 = 171 .OV

M aXlm
' um va 1u e 0 fl oa d current, 1m = R
V ml = 418.88
10 = 41.89 A

F rom Eq. (3.81), rms value of phase current,

Transformer r atin g

Is = 0.78041m = 0. 7804 x 41.89 =32.691 A

= 3 Vph . lph = 3 Vs ' Is = 3 x 171 )< 32.691 = 16770. 5 VA
2rr/3 .

Rms val ue of dio ecurrent,

I Dr

= [ 22


r:/ 3

l~! sin 2 wt . d (wt )

] 1/2

[Art. 3.9]

Diode Circuits and Rectifiers

= [ I!


-v3) ]112 = 0 55181


= 0.5518 x 41.89 == 23.115 A

1 f21t / 3 .
1m 41.89
ID =1m sm wt d (rot) = - = --~-~ = 13.334 A

Average value of diode current,

1t 1t/ 3



= 1m = 41.89 A, PIV = V ml = 418.88 V

Peak diode current


Pd = Vo 10= 400 x 10 = 16000 W

Check .'


:. Transformer rating


=TUF = 0.9541 = 16769.73 VA

..r Example 3.20. A

3-phase bridge rectifier charges a 240- V battery. Input voltage to rectifier

is 3-phase, 230 V, 50 Hz . Current limiting resistance in series with battery is an and an inductor
makes the load current almost ripple free. Determine (a) power delivered to battery and the
load (b) input displacement factor (c) current distortion factor (d) input p ower fa ctor (e) input
H F or THD (f) transformer rating.




Arrange output voltage,

=..f2 X Vl =..f2 x 230 V

= 3 V ml = 3..f2 x 230 = 310.56 V



Vo =E +laR

Vo - E
:. Average value of battery chargIng current, 10 = R

= 310.568 -


= 8.82 A

=El~ =240 x 8.82 =2116.8 W ' .

P d = Elo + l!r. R

Power delivered .to battery

Power delivered to load,

S n ce.load current is ripple free , 10r= 10 = 8.82

P d = 240 x 8.82 + 8.822 x 8 = 2739.16 W

Fig. 3:40 (d) shows that phase-a current ia , or transformer

secondary current is, wou ld be constant at 10 = 8.82 A from fiJi = 30 0 to 1500 and -: 10 from
210 0 t o 330 0 and so on . As positive and nega ive half cycles are identical, average value of
is;" 0, -i.e. Ide = '0.
(b) For ripple fr ee load current,



a l

=it2 JS1t/6/t/6 10 cos n rot . d (rot )



5 11:/6
1t2 f 1t/6
10 cos wt. d (rot) = 1t [sin 150


hI = -




10 sm rot. d (rot) =-


sin 30] = 0


[- cos 150

+ cos 30]

From Eq. (3.56), fun amentsl component of source current is givE::n by


2..[3 I0 '
= --;sm rot an d !PI =tan-

nput disp cem ent factor, DF =cos !PI = 1



1= 0



Power Electronics

[Art. 3.10]


Rms value of fundamental component of source current, lsI

12 x21t
Rms value of source current, Is = .[ . ~ x 3

1/ 2

Current distortion factor, CDF


=1; =

2{3 X 10


2{3 10

=--;t x T2

3" .-





x {2. 10

= n= 0.955

(d) Input pi= CDF x DF = 0.955 xl = 0.955 (lagging)

(el HF

=THD =[(::1]'

=[ ( 0.0~55 )2



(n Transformer rating ={3Vs .Is :::;: {3 x 230 x ~x 8.82 =2868.4 VA

Also, transformer rating =


20~:;4114 = 2876.92 VA


A rectifier should provide an output voltage that should be as smooth as possible. In

practice, however, output voltage fr om rectifiers consists of de component plus ae component,
or ae ripples. The ae component is made up of several dominant harmonics. It is more so in
single-phase rectifiers with R load . The ae component does rio useful work. For example, in a
de motor, it is the de current that produces the required torque; in a battery, energy is stored
due to de current only. AC ripples in rectifier output current do not contribute to motor torque, .
or to the energy stored in the battery. AC component merely causes more ohmic losses in the
circuit leading to reduced effi ciency of the system. This shows . that it is of paramount
importance to filter out the unwanted ae component present in the rectifier output. For this
purpose, filters are used. When used on the rectifier output side, these are called de filters;
these tend to make the de output voltage and current as level as possible. The more common
de filters are of L, C and LC type as shown in Fig. 3.45 (a), (b) and (e).


Rectifi er




Re ctifier


Filt er



( b)


Rec tifier

Re ctifier


JFi lter


Fi lter

(d )

Fig. 3.45. (a ), (b ) and (e) de filters, (d) ac n Iter .



Cj r~u its

[Art. 3.10]

and Rectifiers


The non~sinusoidal output current in rectifier circuit causes the supply line current to
contain harmonics. For reducing these harmonics in the supply current, ae filters are used at
the output terminals of rectifier circuits. Fig. 3.45 (d) shows an ac filter of LC type.
An inductor L in series with load R, Fig. 3.45 (a), reduces the ae component, or ae ripples,
considerably. It is because L in series with R offers high impedance to ae component but very
low resistance to de. Thus ac component gets attenuated considerably. A capacitor C across load
R, Fig. 3.45 (b), offers direct short circuit toae component, these are therefore not allowed to
reach the load. However, de gets stored in the form of energy in C and this allows the
maintenance of almost constant de output voltage across the load.
In this article, a simple design of L, C and LC type de filters is presented.

3.10.1. Capacitor Filter (C-Filter)

Fill-wave, or two-pulse, rectifier is more often used than a half-wave, or one-pulse, rectifier.
In the present discussion, therefore, single-phase
full-wave diode rectifier is only examined. Its
ripple frequency is 2 f, where f is the supply
frequency. A capacitor C directly connected
across the load, as shown in Fig. 3.46, serves to
smoothen out the de output wave.

Fig. 3.47 shows the steady state waveforms

pertaining to Fig. 3.46. Fig. 3.48 (a) gives the

circuit model of Fig. 3.46. S ource voltage

Vs = Vm sin wt is sketched in Fig. 3.47 (a). Load

voltage Vo is shown in Fig, 3,47 (6). In this figure,



Fig. 3.46. Single-phase full-wave diode

rectifier with capacitor filter.

from wt = 0 to wt = e, source voltage Vs is le3s

than capacitor voltage Vc = Vo; therefore diodes Dl, D2 are reverse biased and cannot conduct,
During this interval, i.e, from rot = 0 to cot = 8, capacitor discharges through load resistance
R. At (Dt = 9, Vo = Vc = V2 as snown in Fig. 3.47 (b), Soon after wt = 9, source voltage Us exceeds
Vo (= v c ), diodes Dl, D2 get forward biased and begin to conduct.
a result, source voltage
charges capacitor from V 2 to Vm atwt=7t/2, Fig. 3.47 (b) and Fig. 3.48 (b). Soon after
wt = 7t/2, source voltage begins to decrease faster than the capacitor voltage; it is because
capacitor discharges gradually through R. Therefore, after wt = 7t/2, diodes Dl, D2 are reverse
bi ased and capacitor dis ch ar ges through R as shown in Fig. 3.48 (e). The capacitor voltage
falls exponen tially, Fig. 3.47 (b). In the next half cycle, Vc = Vo = V2 at wt = (7t + 8)~ Just after
wt = (7t + 8), Vs > vc ' diodes D3, D4 get forw ar d biased and begin to conduct. The capacitor
voltage rises from V2 t o Vm at wt = 37t/2, Fig. 3.47 (b ). It is seen from this figure that voltage
dro - from m aximum to minimum is V m - Y2' or peak to peak ripple voltage, V rpp = V m - V 2


In Fig, 3.47 (e) is drawn the profile of ripple voltage with the help of Fig. 3.47 (b) , A
horizon tal line at a h eight ~ (Vm + V 2 ), from r eference lin e wt in Fig. 3.47 (6) is now taken as the
r eference llne in F ig. 3 .47 (c) fo plotting voltage pr ofil e U r AB st ated before, peak to peak r ipple
vo tage is V rpp

= Vm -


and peak ripple voltage

V rp ::::

Ripple voltage i se en to be almost triangular in shape.

%(Vm -

V 2)

as shoV'rn in Fig. 3.47 (c).


, Power Electr onics

[Art. 3.10]


=Vm sin wI

wI. '










Capacitor disch,arging



: :l!



! v'2





I .

iI I
: :!










~ig. 3.47. Waveforms for (a) source voltage (b) load voltage with arid without filter

(c) ripple voltage (d) capacitor current and (e) load current for the circuit of Fig. 3.46.

~' .

Charging of Capacitor. From rot ,= e to 1t/2, capacitor charges from V 2 to V m . The equivalent
circuit for capacitor charging of Fig. 3.48 (b) gives the charging cunent ic as under :

, ": , ' du,

tc =. C dt

The charging current

stored in C at wt = n/ 2 is ~


C dt (Vm sm oot)

=w C Vm cos IDt

at wt =1t/2 is roCVm cos 90 0





.. . .85

= V m' Therefor e, energy


Discbarging of capacitor. KVL for the circuit model of Fig. 3.48



~ fi

dt+Ri= 0

(c )

for capacitor


to .













[Art. 3.10]

Diode Circu its ancf Rectifiers






Fig. 3.48. Single-phase full-bridge diode rectifier (a) circuit model

(b) charging and (c) discharging.

In this equation, time origin is taken at rot = 1t/2. Laplace transform..of this equation is
l.[I(Sl_ C'Vm]+R.I(S)::O


1(s) = ;

s+ RC

i(t) = -..E!. e- tiRe :: ~ e-

where't' =RC


v o = R . i(t) = V 111e- tit =V m e- 1I RC

Load voltage.,



. .
- '. . : .,


Pe ak to peak value of ripple voltage, V''PP:: Vm - V 2 = Vm - [voat t == t2j'=-V~ - Vm e- 12/t

. It is known that e- x = 1 - x.

. V,pp

Charging time


=Vm -

Vm ( 1 -

~~ ) = VR~'

is usually small, it can therefore be


7' therefore t2 if and peak to peak ripple voltage, V,pp =


Peak value of rIpple voltage, V rp = 2

As a result, t2

=; . But



= 4 (RC

Variation of ripple voltage, Vr shown in Fig. 3.47 (c), is not a sine wave, therefore rms
va ue 0 V,. of r ipple voltage can be approximat ely found from the relation
... (3.86)

It is seen from the waveform ~ output voltage vo ' Fig. 3.47 (b), that variation of ripple
.voltage VI' shown in Fig. 3.47 (c) is almost triangular and average value Vo of output voltage
is us ually taken as
. Vo= maximum valu e of V m-pe ak value of triangu~ ar ripple voltage

. = Vm -

V,.p = Vm - 4

inim um value of load voltage, V 2 = Vm - Vl'pp

~C = V

=V m [

1 2

m [


4kc 1

kc 1

... (3 .87)


[Art. 3. 10]

Ripple factor RF is given by

ripple voltage, Vr

RF'= - - - - - - - - -
average output voltage , Va

4~.fRCx V

[1 __
1_]=:T2l4' fR



Its simplification gives


C '= 4' ifR [ 1 + F. RF


Waveform of capacitor current is as under:

At cot




0 , le '= OP = -R=-R

, At ' wt = 8, ic = 0' K

' O'BV

=R = R2;


both OP and 0' K are shown negative because capacitor

discharges from wt = 0 to wt = 8.
During charging time t 1, ic (= w CVm cos rot) follows the cosine wave with peak value
wCV:m cos 8 at rot = 8 and ic '= 0 at wt = n/2.
Current ( is positive as the capacitor is getting charged. Soon after the maximum value of V m
at wt

=n/2, capac.'tor begins to discharge and i c

'= -

I Lp where I Lp

= ;. At wt


(n + 8),


decays to

V 2 and lc .=- O'K as before.

Waveform of load current io in Fig. 3.47 (e) is identical with the waveform of va in Fig. 3.47
(b). At wt = 0, io

= ~A, at wt =- 8, io =


at .wt

= n/ 2, io =- ;

and so on.

Eq. (3.86) shows that if C is increased, ripple voltage gets reduced,. But high value of C
increases the amplitude of charging current as per Eq. (3.85). A high charging current would
entail hig~er current rating of diodes. This leads to increased cost of the rectifier-filter circuit.
Thus, a compromise between the value ofC and the magnitude of ripple voltage must be made .
E xample 3.21. A 's inglepha se diode B -2 rectifier is fed from 2 30 V, 50 H z source ahd is
connected to a load of R '= 400 n.

(a) Design a


so that the ripple factor of the output voltage is less than


(b ) W ith the va lue of C obtained in part (a ) , determine t/.. ~ average value of output
(c ) Dete rmine the average value of output voltage without C- filter.

S olution . (a ) From E q. (3 .89), the value of capa citor C to limit the ripple factor RF to 5%


C = 4 fR

11 1 [

1 + '12. RF

1] ='

=- 4 x 50 x 400 1 + -.1"2 x 0.05

189.3 )IF.


[Art. 3.1 0J

Diode Circuits an d R ectifiers

(b) From Eq. (3 .87), average value Vo of 'the output voltage, with filterC, is

V o =Vm [1---.L]=2
4 {Re ' 3 0'-'2[
'1L. 1 _ 4 x 50 x 10
400 x 189.3 ]=303745V

Average value of output voltage without C-filter is

. 2V
V = __
m = 2 '12 x 230 = 207.04 V



It is seen from this example that use of C filter has reduced the RF from 48.2%to 5% and
at the same time, average output voltage Vo has increased from 207.04 V to 303.754V.

3.10.2. Indu ctor Filter (L-filter)

An iriductor filter connected in series with the resistive load serves to provide the requisite
filtering. In a resistive load, current waveform is identical with voltage waveform. An
inductance in series with R load does not allow sudden changes in the load current. As a
consequence, load current profiles becomes noticeably smooth. This has the effect of reducing
. the current ripple factor.
. '.

Fig. 3.49 (a ) shows an in ductor filter L connected in series with R-Ioad in a I-phase
two-pulse diode rectifier. Fig. 3.49 (b) shows the rectified voltage Vo and the load current i o ' It
. is seEm that inductor has a smoothing effect on the load current profile.

:-Filter ....,





, to



Vs= Vmsinwt



(a )

F ig. 3.49. Single-phase diode bridge rectifi er (a ) circuit diagram and ( b ) waveforms for
load voltage, load current , diode Dl current and sour :-" current.

From wt = 0 to wt = 1t; diodes Dl, D2 conduct and fr om wt = 1t t o 21t; diodes D3 , D4 conduct.

The load current is continuous.
,The output v oltage
appendix A),


of the 2-puls e r ectifier can be analysed in to Fourier series as (see

2Vm . 4 Vm
, 4 Vm
. 4 Vm
= - - - 3-' cos 2 wt -15
cos 4wt -3'"' cos 6


whereVm = maximum v alue of source voltage.

. 1t

0 1t


.. .(3.90)


Power Electronics

[Art. 3.10]

Average load voltage,

v = 2Vm

Average load current,

10=]i =nR"




.Zn = ~R2 + (n

Load impedance for nth harmonic,


Magnitude of second harmonic load current]

Similarly, fourth harmonic current,
Load current io can, therefore, be written as


= --;if -


4 Vm

3n ~R2 + (2roL)2 cos (2wt - 92) - 15n ~R2 + (4 roL)2 cos (4 wt - 94)

where first, second and third terms of above current expression are dc, second~harmonic and
fourth-harmonic components of i o' Rms value of harmonic-current components, or ripple
current,Ir is

J2 '

4 Vm

Ir= [ ( 3n'l2


= ' ( -:;c;J2

4 Vm




.' -


. ]]

R2+(4WL)2+'~~ '

1 1.

3 2 . R2 + (2 WL)2 + 15 2 ' R2 + (4 roL)2 + .. '

,.. (3 ,91)

Second-harmonic component seems to be the most dominant component of I,,, Therefore, ' .
neglecting higher-order even harmonics, we get rms value of ripple current,

1 -1 _ 4 Vm
r - 2 - 3 n -fi '-iR2 + (2 roL)2
4 Vm
re R
. 1r
: , 'Current-ripple factor, CRF - - x-
- 10 - 3'Jt '\[2"JR2 + (2 roL)2 2 Vm

= 3 x 42 {2' 'IR 2 +R(2 roL)2 = 0.4715.. ~R2 +R(2wL)2

... (3.91 a)


(3 92)

For good filtering, roL R. Thus, neglecting R in the denominator ofEq. (3.92), we get

CRF =0.4715. 2wL

Fo 50 Hz supply,

. R

=0. 236 wL

CRF = 0.236 (2re X 50) L = 7.51


It is seen from above that

(i ) reduced current ripple r equires large value of L,
(i i) as the load is reduced (or R increased), ripple current inc eases,
(i ii) indu ctive filter is preferred where load resistance R is consistently low or load '
current is invariably high ,
(iv ) induc tor in the load circuit introduces time delay of the load CUl'l'ent with respect
to the load voltage,
(v) as the current rofile becomes more smooth r transformel' utilization factor is im

[A rt . 3.10]

Diode Circuits an d Rectifie rs


Comparison between C and L filters. It is worthwhile at t his stage to compar e a C-filter

with an L-filter.
(a) If load

resistance is low, ripple factor for L-filter is low and pigh for C-filter.
(b) In both C-filter and L-filter, time constant should be large for better waveform, i.e.
for low ripple factor, 't should be high.
. '
. . . . .. ' ' .
(c ) For C-filter, if R is increased, 't (= RC) increases, and therefore, ripple factor gets
reduced . .
Cd) For L-filter, if R is lowered, 't (= LIR) increases, therefore ripple factor becomes low.
This shows. that C-filteris suitable for loads having low curr ent (highload resis,:
tance) consistently. Dfilter is suited for loads requiring high load current (low R)
consistently. .
.' .
(e) A high value of C reduces ripple factor but increases the chargingcurrEmt and the
diode-current rating. . '
. .
. .'
(f) Ahigh value of L reduces the ripple 'factor, but .a delay is introduced in the response .


Inductor filter is bulky, weighty, expensive and causes extra ohmic loss as compared to
C-filter. Besides, L-filt er is noisy in nature. .
Example 3.22.A single-phase two-p1..J,lse diode rectifier has input supply of230 V, 50 Hz and
the load resistance R = 300 n.Calculate the value of inductance to bf!.connected in series with R
so as to limit the current ripple factor to 5%. Find the value of L in case R = 30 O. Determine also
without L .
the value of CRF
Solution. From Eq. (3.92), current ripple factor (CRF) with L is .

CRF= 0.4715. ~R2+ (2rol.J

0.05 =0.4715

[3002 + (2
L = 4.4755 H .

For R = 3000,

For R

= 300,

0.05 = 0.471<>


' .


21t x 50

L)2] 112


'130 + (2 x 21t x 50 x L )
.' L = 0.4477 H

CRF without filter L can be obtainedby putting L = 0 in Eq. (3.92). This gives CRF = 0.4715
without L .
' .
3.10.3. L- C Filter
An L~C filter consists of inductor L in series with the load and capacitor C across the load.
This filter possess es the advantages of both L-filter and C-filter. In addition, ripple factor in
L-C filter has lower value than that obt ained by either L-filter or C-filter for the same values
ofL and C.
Fig. 3.50 (a) show the use of L -C filt er for reducing th e ripple from the output voltage of a
single-phase fu ll-wave di ode rectifier. Its equivalent circuitis g1venin Fig. 3.50 (b). The
inductor L blocks th e dominant harmonics. Capacitor C provides ,aneasy p ath to t h e nth
harmonic ripple currents .
In order t h a t capacit or yields an easy path for harmonics, load impedfu'1ce R must be much '
greater than nth h armonic capacitive reactance; i.e. R
that capacit or provides effe ctive filtering if

~ n~~ ' It has been fOlllld in practice

Power Electronics

[Art. 3.10]








rY'j .


J l



Fig. 3.50.


( b)

Single-phas~ diode bridge rectifier with L - C filter

circuit diagram and (b) its equivalent circuit.


...(3.93 a)

n roC

In case load consists of R and LL in series, then Eq. (3.93 a) becomes

v + ( nro L L )2 = .. 10r'l

...(3 .93 b) .

UJ\.., . .

Under the condition ofEq. (3.93), effect of-loadR, or Rand LL' can be ignored during further
analysis. Therefore, nth harmonic current in Fig. 3.50 is

I" =



where V" = rmsvalue of nth harmonic of rectifier output voltage.
Thus, nth harmonic component ofload voltage VOrl across filter C in Fig. 3.50 (b) is

Vo" =

[n~I~T 1v, = [ (nro}2-;C - 1




Total ripple voltage due to all harmonics is



,~" ]112

... (3.95)

n = 2,4,6,

The Fourier series analysis of output voltage of I-phase full-wav e rectifier is given in Eq.
(3.90). The average value of output voltage Vo = 2Vm/1t. It is seen fromEq. (3.90) that second
harm onic is the most dominant component. Therefore, other harmonic components can be

F rom Eq. (3.90), the rms v alue V 2 of secon d harmonic voltage is

4 Vm
. V 2 = 31t . T2
From Eq. (3 .94) for n =2, ripple voltage is

V - V02 -


rL (2ro)2-LC1 - 1 ]1/2

...(3 .96)


[Prob . 3]

Diode Circuits and Rectifiers

The value of filter capacitor C can be obtained from Eq. (3.93) as






- 2m ~R2


... (3 .97b)

+ (2m LL)2

The VRF is defined as



VRF = Vo = Vo . (200)2 L C _ 1 =

= -f2 [

(200) LC -1

4 Vm )
3 1t . "l2



x (200)2 LC - 1

. . (3.98)

Once filter C is obtained from Eq. (3.97), the value of filter inductor L can be calculated
from Eq. (3.98) for a specified value of VRF.
Example 3.23. A single-phase two-pulse diode rectifier has input supply of 230 V, 50 Hz.
and a load resistance R =50 n and load inductance LL = 10 mHo An LC filter is to 'be used on
the output side so as to reduce the output voltage ripple to 10%. Design the LC filter.
Sol u t ion. From Eq. (3.97 b), the value of filter capacitor Cis


C = 2 X 21t 50 ~502 + (200 1t x 10 x 10- 3)2 = 315.83)lF

From Eq. (3.98),



(2001t)2 x 315.83 xL




O~. = {2
3' (2001t)2.Lx315.83x 10- 6 -1

=~ x 0\ + 1

= 0.045822 H or 45.822 mHo


Vo volts and (b) - Vo volts.


3.1. Capacitor in the circuit of Fig. 3.2 (a) is initially charged with (a)
For both these parts, determine the expressions for current in the circuit and voltage across

capacitor. Sketch the waveforms for current as well as capacitor voltage .

What is the final value of voltage across capacih>r~in each case?

[ Ans.


Vs-Vo -tiRe

-tiRe V
. ' Vo+(Vs-Vo)(l-e
), s

(b ) Va ; Vo

e- tiRe, _ Vo + (Vs + V o) (1 - e- tiRe), V s ]

3.2. A diod e is connected in series with LC circuit. If this circuit is switched on to de source of
voltage Vs at t = 0, derive expressions for current thr ough and voltage across capacitor. The
capa citor is initially charged to a voltage of - Vo. Sketch waveforms for i, Ve, vL and uD'
In case this circuit has Vs = 230 V, Vo =50 V, L =.0.2 mR and C = 10 IlF, detennine the diode

~:~:ction time, diode peak [::~:t::::'::o~i:::t~V:H~::t~:';::::o)C::~cic::~~:

- '\10, 140.496 11 3, 62.61 A 510

V, - 280 V]


Power Electronics



the circuit shown in Fig. 3.4 (a), the circuit is initi ally rel axed. If switch S is closed
at t = 0, sketch the variations of i, uL, Ue and uD as a function of time. Derive the expres
sions describing these functions .
(b ) For part (a), Vs = 220 V, L = 4 mH , C = 5 IlF. Find the diode conduction time and peak
diode c urrent. Determine also ue, uLand uD after diode stops conducting.
fAns. (b) 0.444 ms, 7.778 A, 440 V, 0, - 220 V1

(a) For

3.4. For the circuit shown in Fig. 3.5 (a) ; Vo = 230 V, R = 25 nand C = 10 IlF. If switch S is closed
at t = 0, determine expressions for the current in the circuit and voltage across capacitor C.

Find the peak value of diode current and energy lost in the circuit.

Derive the expressions used.

[Ans. 9.2 e-4000t, - 230 e-4000t, 9.2 A, 0.2645 watt- sec]

3.5. In the circuit shown in Fig. 3.51, switch S is open and

a eurrent of 20 A is flowing through the freewheeling

diode, Rand L. If switch S is closed at t = 0, deter

m(fie the expression for the current through the

switch .
[Ans i(t) = 22 - 2 e- lOOOtJ


VS =220V
3.S. (a) Describe how the energy trapped in an inductor
can be recovered and returned to the source .
(b ) A 230 V, 1 kW heater, fed through single-phase

half-wave diode rectifier, has rated voltage at its

Fig. 3.51. Pertaining to Prob. 3.5.
terminals . Find the ac input voltage. Find also
[Ans. (b ) 325.32 V, 460 V, 8.696 AJ
PlV of diode and peak-diode current.

3.7. (a ) In the circuit shown in Fig. 3.52, a PMMC ammeter is placed in series with diode and a
PMMC volt meter across the diode . Take PMMC instruments ideal. Find the readings on
t hese instrument s . Derive the expressions used for obtaining thes e readings.
(b) If PMMC ammeter is replaced by MI ammeter, find its reading.
[Ans. (a ) 10.352 A, 0 V (b ) 12 .6812 A]
,------( V } - - - - - ,


50 Hz



Fig. 3.52. Pertainin g to Prob. 3.7.

r7 'f/

F i g. 3.53. P ertaining to Prob. 3.8.

3.8. (a ) In the circuit of Fig. 3.53, ideal PMMC voltmeters are placed, one across capacitor and
an other across diode as shown. Find the voltmeter readings . Obtain the expresoions used
for det ermining these readings.
(0) In case PMM C voltmeter 2 is repla ced by MI voltmeter, find its reading.
[Ans. (a) 325.22 V, 325 .22 V (0 ) 398.3 94 V]

3.9. A ba ttery is ch arged by a single-phase half-wave diode rectifier. The supply is 30 V, 50 Hz

and th e battery emf is const ant at 6 V. Find the resistance t o be inserted in series with t he
battery to limit the chargin g current to 4 A. Take a voltage drop of 1 V across diod e. Derive
th e expression used . I

Dra\v waveform of vo ltage acr oss diode and find its P lV.
t e
. - 1 ,pi x 30 et c.
I n . 1 = sm




[2Vm cos 8 1 (E + 1) (It - 2 61 )], 2.5467 n, 49.42 V]

Diode Circuits and Rec tifiers

[Prob. 3]


3 .10. (a) A single-phase half-wave uncontrolled rectifier is connected to RL load. Derive an expres
sion for the load current in terms of Vm' Z, ro etc.
(b) For part (a), Vs == 230 V at 50 Hz, R = 10 n, L = 5 mH, extinction angle = 210. Find

average values of output 'Voltage and output current.


(a) A

single-phase half-wave diode rectifier feeds power to (i) RL load and (ii) RL load with
freewheeling diode across it. Describe the working of this rectifier for both these parts
with relevant waveforms and bring out the differences if any. Hence point out the effect
of using a freewheeling diode.
(b) For part (a), Vs = 230 V at 50 Hz, R = 20 n, L = 1 H. Find the average values of the output
voltage and output current with and without the use of a flywheeling diode.
[Ans. (b) With freewheeli.n g diode : Vo = 103.52V and 10 = 5.176 A

Q o'f"" .. 3.12.

[Ans. (b) 193.172 V, 19.3172 AJ

Without freewheeling diode: Extinction angle

p not known, so

Va' 10 cannot be calculated 1

circuit shown in Fig. 3.54, the output current io is considered constant at 10 because
of large L. Sketch the waveforms of vs , io, va, iD' ifd , and is'
For the above circuit, find
(i) average values of output voltage and output current,
(ii) average and rms values of freewheeling diode current,
(iii) supply pf.
[Ans. (b) (i) 103.52 V, 26.76 A (ii) 13.38 A, 18.925 A (iii) 0.6364Iag]

(a) For the






230 V.
50 Hz



i fd





F ig. 3.54. Pertaining to Prob . 3.12.

Fig. 3.55. Pertaining to Prob. 3.13.

3.13. For the circuit shown in Fig. 3.55, Vs = 160 V, V z = 40 V and zener dio.de current varies from
4 to 40 rnA. Find the minimum and maximum values of Rl so as to all~w voltage regulation
for output current 10 = zero to its maximum value 10m , Also calculate 10m ,
[Ans. 3k n, 30k a, 36 mAl

/3.14 . (a) Enumerate the input perfo rmance par ameters of a rectifier. Discuss how the performance
of a rectifier circuit is influenced in case these parameters have low, or high, value.
(b) Define input power factor, displacement factor DF and current dist ortion factor CDF for
/.....a rectifier system an d show that input power factor = CDF x DF.
3.15. (a ) Defin e input current har monic factor (HF) and crest factor . Express (HF) in terms of
_ /"current distortion factor.~f HF is more, what does it i rucate in a rectifier system .
(b) Define t he following terms :
Rectification ratio, T' pple volta ge, form factor, voltage ripple factor, current ripple fa ctor
and tran sformer utilization factor.
3. 6. F or a single-phase half-wave diode rectifier feeding a res 'stive load R, fmd the values of
recti 1e efficiency, form factor, voltage ripple factor, transformer utilization factor and crest
fa ctor.
3. 7. A sin gle-phase half-wave d'ode r ectifier is de igned to supply de outp ut voltage of 23 0 V to a
load of R = 10 n. Calcula te the ratings of di ode and transfonner for this circuit arra gem ent .
[ADs. I DAV =23 A, ID; =36.1 3 A, PN = 722. 6 V, Trans. rating = 18 .462 kV,\ ]


Power Electr onics


3.18. A single-phase full-wave mid-point diode rectifier feeds resistive load R. For this circuit, .
determine rectifier efficiency, form factor, voltage ripple factor, transformer utilization fac tor
and crest factor.
D\,f;,4 , How does this 'rectifier circuit differ from single-phase full-wave bridge rectifier?

V t..3. 1 ~ (a) Why are three-phase rectifiers preferred over single-phase rectifiers?

(b) For a 3-phase half-wave diode rectifier feeding load R, obtain the following:
Average output voltage, rms output voltage, VRF, FF, TUF andPIV
3.20. Describe the evolution of three-phase six-pulse diode rectifier from 3-phase three-pulse diode
rectifiers with appropriate circuits and waveforms. Hence, derive an expression for the
average output voltage of 3-phase six-pulse diode rectifier.
3.21. Describe a 3-phase M-6 diode rectifier with a circuit diagram and relevant waveforms for
resistive load R.
Hence, derive expressions for average and rms values of output voltage and obtain therefrom
. V o f" VRF, FF, rectifier efficiency and TUF.
, 3.22. A3-phase mid-point 6-pulse di.ode rectifier feeds a load of 10 n at a de voltage of 400 V. Find
the ratings of diodes and the three-phase transformer.
[Ans. IDAV=: 6.667 A, IDr =: 16.337 A, PIV =: 837.76 V, Trans. rating=29.038 kVA]



Describe a 3-phase full-wave diode-bridge rectifier with a circuit diagram and relevant

waveforms for load R .

Hence, derive expressions for average and rms values of output voltage and obtain there from

VRF, FF, rectifier efficiency and TUF.

3.24. A 3-phase full-wave diode rectifier feeds a load requiring constant current 10 and is supplied
from a 3-ph ase delta-star transformer.
(a ) Sketch input volta ge waveforms for Vab, v~, vbc etc., taking vab zero and becoming positive
at rot =: O.
(b ) Sketch waveforms for currents for the three diodes of positive group a'nd phase current
of the t ransformer secondary.
(c ) l<"'r om th e wave orm of secondary phase current , determine current distor~ion factor CDF
and THD.
[Hint. Here al

=: -

..f3 10 , hI =.9. 10 , etc.]


[Ans. (e) 0.955,0.3106]


3.25. A 3-phase full-wave diode rectifier delivers power to an inductive load which takes ripple-free
. current of 120 A. The source voltage is 3-phase, 400 V, 50 V, 50 Hz. Determine the ratings of
diodes, power delivered to load and the rms value of source current.
[Ans. 1DAV = 40 A, 1Dr = 69.284 A, PIV = 565.6 V, 64813.2 W, 97.98 AJ
3.26. Describe a three-phase 12"pulse diode rectifier with circuit diagram and appropri ate
waveforms. Hence derive expressions for average and rms values of output voltage. Fro m
0\ these, obtain VRF and FF.
3.27. (a ) 'Wh at ar e the advantages of 3-ph ase bri dge rect ifier over 3-phase M-6 rectifier.
(b ) For a 3-ph ase p-pulse diode rectifier, prove the following:
Av era ge output voltage,


= V mp ' E. sin ~


and rms outpu t voltage, V = V



(~+ 1. sin 21t )11/2
21t p 2 . P

where Vmp = m aximum value of per-pha e su pply vo1t~ge.

3.28. (a) W ha t ar the fu nctions offilt ers in rectifier circuits ? Distinguish between de and ac filters .
( q ) Explain how the in ductance Land capaci tan e C pI y th eir rol e in r educing the harmonic
. contents in r ectifi er circuits.

Diode Circuits and Rectifins

[Pr ob. 3]


3.29. A single-phase full-wa ve diode rectifier feeds R with 3. .capacitor C directly connected across
load. Describe the opera tion of C as a filter with relev'~mt voltage and current waveforms .
Show that peak ripple voltage is

~ (Vm - V 2).

3.30. A single-phase two-pulse diode rectifier feeds Rand C in parallel. Explain charging and
discharging of capacitor C and derive expressions for ripple factor and the value of filter
capacitor C.
3.31. A single-phase diode bridge rectifier is fed at 230 V, 50 Hz. The load is R = 200 n shunted by
a capacitance of 300 IlF . Neglecting all losses, detennine the average value of load voltage,
VRF, maximum and minimum value ofload current, peak capacitor current and average load
[Ans. 298.12 V, 0.0589, Imax = 1.6261A, Imin = 1.3551 A, 30.651 A 1.4911 AJ
3.32. A single-phase two-pulse diode rectifier feeds load R with an. inductor L in series with it.
Describe the working of L as filter with relevant voltage and current wavefonns. Derive
expression for current ripple factor and show that for 50 Hz supply, CRF = 7.51 x 10- 4.


3.33. A single-phase full-wave diode rect ifier has mean output vo.ltage of 200 V and the load
resistance is 400 n. Determine the inductance required to limit the amplitude of second-har
[Ans. L = 3.48 Hl
monic current in the load to 0.06 A.
3.34. A single-phase full-wave diode rectifier with L-C filter feeds load R. Describe its working and
derive expressions from which the parameters of L-C filter can be obtained.
3.35. (a) Compare C-filter with L-filter. In what tyPe of applications are the two types usually
(b) A single-phase two-pulse diode rectifier is fed from 230 V. The load is R = 200 n. Design
an LC filter so as to get voltage ripple factor of5.89%. Find the rms value of ripple voltage.
[Ans. C = 79.58 IlF, L = 0.28654 H, 12.195 V]


----.. _--.- ... ... . .. ......... ... ............ -_ .................. .. ... ....... ....... ... .... ..... -,

In this, ::hapte r
Term ina l C h aracteristics at Thyristors
Thyristor Turn-on Methods

Switc hing Charac teristics of Thyristors

Thyristor Gate CharacteristiCs
Two-TranSistor Model of a Thyristor
Thyristor Ratings
Thyristor Protection
Improvement of Thyristor Characteristics
Heating, Cooling and Mounting of Thyrts,tors
Series and Para ll e l Operation of Thyristors
Other Members of the Thyristor Family
Gote Turn oN (G .T.O.) Thyristor
Static Induction Thyristor
Firing Circuits tor Thyristors
Pulse Transformer in Firing Circuits
Triac Firing C ircuit

...... -.......................... __ ................................................ ................ As stated before, Bell Laboratories were the first to fabricate a silicon-based semiconductor
device called thyristor. Its first prototype was introduced by GEe (USA) in 1957. This company
did a great deal of pioneering work about the utility of thyristors in industrial applications.
Later on, many other devices having characteristics similar to that of a thyristor were
developed .. These semic~mductor devices, with their characteristics identical with that of a
thyristor, are triac, diac, silicon-controlled switch, programmable unijunction transistor (PUT),
GTO , RCT etc. This whole family of semiconductor devices is given the name thyristor. Thus
the term thyristor denotes a family of semiconductor devices used for power control in dc and
ac systems. One oldest member of this thyristor family, called silicon controlled r ectifier (SCR),
is the most widely used device. At present, the use of SCR is so vast that over the years , the
word thyristor h a.s become synonymous with SCR. It appears that th e term thyristor is now
becoming more common than the actual term SCR. In this book, the term SCR and thyristor
have been used at r andom for the same device SCR. Other members of thyristor fam ily are also
discussed in this chapter.
A thyristor bas characteris tics similar to a thyratron tube. But fro m the construction vi:w
poin t, a thyris to r (a pnpn device) belongs to transistor (p np or n.pn device) fam ily. The name
'thyristor ', is der ived by a combination of the capi tal letters fr om THYRa.tr on and trans ISTOR
This mea.TJ.S that thyristor is a solid state device like a transist or and has charFlcteristics s im ilar
to that of a thjTatron tube. The pr esen t-day r eader may not be fa miliar with thyratr on tube as
this is not being taught these days. Actually, the n:lme 'thyri stor ' ca me in to exis~ence afte r a


[Ar t.

~. l l


. fo rmal' (i"ecision taken at a conference held by lEe (International Electrotechnica l Comml.:ision)

in 1963. Prior to that, it was called s ilic on con trolled r ectifie r. or SCR. It appears that
commission must have evolved the name 'thyristor' as discussed above.
At this conference, the definition of thyristor was decided as under :
(0 It constitutes three or more p-n junctions.
It has two stable states. an ON-state and an OFF-state and can change its state


from one to another.

As per this definition, thyristor now includes a lar ge variety of sem iconductor devices
ha ving similar basic characteristics.
The object of this chapter is to discuss the thyristor ch aracteristics and other related topics
useful for their industrial applications.




Thyristor is a four layer, three-junction, p -n-p -n semiconductor switchin g device. It h as

three terminals ; anode, cathode and gate. Fig. 4.1 (a) gives constructional det~ils of a typical
thyristor. Basically, a thyristor consists of four layers of alternate p-type and n-type silicon
semiconductors forming three junctions J l , J 2 and J 3 as shown in Fi g. 4.1 (a). The threaded
portion is for the purpose of tightening the thyristor to the frame or heat s ink with the help of
a nut. Gate terminal is usually kept near the cathode terminal, Fig. 4.1 (a). Schematic diagram
and circui t sym bol for a thyristor are shown respectively in Figs. 4.1 (b ) and (c ). The terminal
connected to outer p region is called anode (A), the terminal connected to outer n region is called
cathode and th at connected to inner p region is called th e gate (G)' For large current
applications, thyristors need better cooling; this is achieved to a gTeat extent by mounting th em
onto heat sinks. SCR rating h as improved considerably since its introduction in 1957. Now
SeRs of v oltage r ating 10 kV and an rm s current rating of 3000 A w.i th cor responding
power-handling capacity of30 l'Irw are available . Such a high power thyristo r can be switched
on by a low voltage supply of about 1 A and 10 W and this gives us an idea of the immense power
amplification capability (= 3 x 105) of this device. As SeRs ar e solid state devices, they a re
compact, possess high reliability and hav e low loss. Because of these useful featu res, SCR is
almost universally employed these days for all high power-controlled devices.




(Aluminium )


1-- ,I----1 J,
Galt terminal
weJdtd to p
r tgion
Ca ll'lodo:

Fig. 4. 1.

Col ho(J~

(0) Con3tn.:ctional d ':ili!.s (b, Sch~matic diagrum



circut t symbol of a

thy :-iSt8r.


Power EJectronics

[Art. 4.1J

An SCR is so called because silicon is used for its construction and its operation as a rectifier
(very low resistance in the forward conduction and very high resistan ce in the r everse direction)
can be controlled. Like the diode, an SCR is an unidirectional device that blocks the current flow
from cathode to anode. Unlike the diode, a thyristor also blocks the current fl ow from anode to
cathode until it is triggered into conduction by a proper gate signal between gate and cathode
term inals.

For engineering applications of thyristors, their terminal characteristics must be known . In

this article, their static [. V characteristics, dynamic characteristics during tumon and turn-off
processes and their gate characteristics are discussed.
4.1.1. Static I-V Characteristics of a Thyristor
An elementary circuit diagram for obtaining static I- V characteristics of a thyristor is
shown in Fi g. 4.2 (a ). The anode and cathode are connected to main sour ce th rough the load.
The gate and cathode are fed from a source E.I which provides positive gate current from gate
to cathode,

Fig. 4.2 (b) shows static I -V characteristics of a thyristor. Here,vo is the anode voltage
across thyris tor terminals A , K and 10 is the anode current. Typical SCR J-V characteristic
shown in Fig. 4.2 (b) reveals that a thyristor has three basic modes of operation; namely,
reverse blocking mode, forward blocking (off-state ) mode and forward conduction (on-state)
mode. These three modes of operation are now discussed below :
Reverse Blocking Mode. When cathode is made positive with r espect to anode with
switch S open, Fig. 4.2 (a ), thyristor is reverse biased as shown in Fig. 4.3 (a). J unctions J 1, J 3
are seen to be reverse biased whereas junction J 2 is forward biased. The device behaves as if
two diodes are connected in series with reverse voltage applied across them. A s mall leakage
current of the order of a few milliampe res (or a few microamperes depending upon the SCR
r ating) flows. This is reverse blocking mode, called the offstate, of the thyristor. In Fig. 4.2 (b),
reverse blocking mode is shown by OP. If the r everse voltage is increased, then at a critical

Forward conduction mode
(on stote )


Reverse !E.'C~oc;e
\:urr ent


. '>

RevE'r;i.' blocking


V30: forwar d bri'okover voltage

- [0


Forward Iec.lIcge

For .....ord
bloc~in9 mode

" 3i1' Reversi' ':lr ... oltc!own voltc!ii.'

C,p GatE' curren t


Fig. 4.2. (e) Elementary circuit for obtrtining thyristor I-Y characteristics
(b ) S~atic I-Y charac:eristics of!l thyristor.

[Art. 4.2]



breakdown level, ~alled r ever se breakdown voltage

V BR an avalanche oc cur s at J 1 and J 3 and th e
rever se current increases rapidly. A large current
associated with V BR gives rise to more losses in the
SCR. This may lead t o thyr istor dam age as the
junction temperature may exceed its permiss ible
temperature ris e. It should , therefore, be ensured
that maximum. working r everse voltage across a o--c"'C
thyristor does not exceed VBB.. In Fig. 4.2 (b ), reverse G
ava lanche region is shown by PQ . When reverse
___ Rqversil
voltage applied across a thyristor is less than VBRI
the device offers a~ high impedance in the reverse
direction . The SCR in the reverse blocking mode may
therefore be tre ated as an open switch.
(a )
Note that [ .V characteristic after avalanch e Fig. 4.3. (a ) J 2 rorward biased and J 1.J3
bl k
d '
revers e biased (b ) J 2 reverse biased and

b rea k d own d uring

r eve r se
oc 109 mo e I S
db' d
applicable only when load resistance is zero, Fig. 4.2
Jl' .~~ orwar lase.
(b) . In case load resistance is present, a large anode eui+ent associated with avalanche
breakdown at V SR would cause substantial voltage drop across load and as a result, ]. V
characteristic in third quadrant would bend to the right of vertical line drawn at V BR


Forward B locking Mod e: When anode is positive with r espect to the cath ode, with gate
circuit open, thyristor is said t o be forward biased as shown in Fig. 4.3 (b). It is seen from this
figure that junctions J h J 3 are forward biased but junction J 2 is reverse biased. In this mode,
a s mall current, called forward leakage current, flows as shown in Figs. 4.2 (b) and 4.3 (b ). In
Fig. 4.2 (b) , OM represents the fonvard blocking mode ofSe R. As the fonvard leakage curren!
is small, SCR otTers a high impedance. Therefore, a thyristor can be. treated as an open switch
even in the forward blocking mode.
F orward Conduction mode. 'When anode to cathode iorward voltage is increased with
gate circuit opep, reverse biased junction J 2 will h ave an avalanche breakdown at a voltage
called fOl'ward breakover voltage VBO ' After this breakdown, thyristor gets turned on with
point }.If at once shifting to N and then to a point anywhere between N and K. Her e NK
represents the forward conduction mode. A thyristor can be brought from forward blocking
mode to fo rw'ard conduction mode by turning it on by applying (i) a positive gate pulse between
gate and cathode or (ii) a forward breakover voltage across anode and cathode.
Forward conduction mode NK shows that voltage drop across thyristor is of the order of
1 to 2 V depending upon the rating of SCR. It may also be seen from N K that voltage drop
across SCR increases slightly with an increase in anode current. In conducti on mode, anode
current is limited by load im pedance alone as voltage drop across SCR is quite small. This
small voltage drop VT across the device is due to ohmic drop in the four layers, In forward
conduction mode, thyristor is treated as a cl osed switch.


., , . .. .


With anode positi',e with r espect to cat.hode, a thyris tor can be turned on by anyone of the
follow ing techniques: (a) Forward voltage trigging (b) gate triggering (c) du / dt triggering (d)
t emper ature triggering and (eo ) light trigg=ring,

[Art. 4.2J

Power Elt!ctronics

These methods of turningon a thyristor are now discussed one after the other.

Forward voltage triggering. When forward voltage is applied between anode and
cathode with gate circuit open, junction J 2 is reverse bianed. As a result, depletion layer is
form ed across junction J 2 . The width ofthis layer decreases with an increase in anode-cathode
volta.ge. If forward voltage across anodecathod e is gradually increased, a stage comes when
the depletion layer acr oss J 2 vanishes. At this moment, reverse biased junction J 2 is said to
have avalanch e breakdown and the voltage at which it occurs is called forward breakolJer
voltage VaG- The name forward breakover voltage is given because at this voltage VBo i-u
characteristic breaks over and shifts to its on-state position wi th breakover current IBO' At
this voltage, thyristor changes from off-state (high voltage with low leakage current) to on-state
characterised by low voltage across thyristor with large forward current. As other junctions
J t J 3 ar e already fOT'Hard biased, breakdown of junction J 2 allows free movement of carriers
across three junctions and as a result, large forward anode-current flows. As stated before,
thi s forward current is limited by the load impedance. In practice, the transition from off-state
to on-state obtained by exceeding V BO is never employed as it may destroy the device.
(a )

The magnitudes offorward breakover and reverse breakdown voltages are nearly the same
and both are temperatur e dependent. In practice, it is found~ that V BR is slightly more than
V BO' Therefore, forward breakover voltage is taken as the fin al voltage rating of the device
durin g the design of SCR applications.
After the avalanche breakdown, junction J 2 loses its reverse blocking capability. Therefore,
if the anode voltage is r educed below V BO SCR will continue conduction of the current. The
SCR can now be turned off only by r educing the anode current below a certain value called
holding current (defined later).
Gate Triggering. Turning on of thyristors by gate triggering is simple, reliable and
efficient, it is therefore the most usual method of firing the forward biased SCRs. A thyristor
with forward breakover voltage (say 800 V) higher than the normal working voltage (say 400
V) is chosen. This means that thyristor will remain in forw ard blocking state with normal
working voltage across anode and cathode and with gate open. However, when turn-on of a
thyri stor is r equired , a positive gate voltage between gate and cathode is applied. With gate
current thus established, charges are injected into the inner p layer and voltage at which
forward breakover occur s is r educed. The forward voltage nt which the device switches to
on-state depends upon the magnitude of gate current. Higher the gate current, lower is the
forward breakover voltage.

When positive gate current is applied, gate p layer is flooded with electrons from the
cathode. This is because cathode n layer is heavily doped as compared to gate p layer. As the
thyristor is forward biased, some of these electrons reach junction J 2 As a result, width of
depletion layer near junction J'l, is reduced. This causes the junction J 2 to breakdown at an
applied voltage 10v-/eT than the forward bre akover voltage VBo.lfmagnitude of gate current is
increased, more ele ctrons would reachjunctionJ'l" as a con sequence thyristor would get turned
on at a much lower forwar d applied voltage .
Fig. 4.4 (a ) snow3 that for gate current 1, = 0, fo rward breakover voltage is 1,/30' For gate
current lll' forward br eakover, or turn-on voltage is VI which is less than VBO ' For 1,2 > 1,1'
forward breakdover voltage i3 further reduced to V2 < VI ' For 1,3 > 1z'2 1 the forward breakuver
voltage is VJ < V2 , Fig. 4.4 (0). Th e effect of gate current on the forward br eakover voltage of a

[M . 4.2J


Forward voltcgi/'


ReV E'r$i/' leakage

curren t

\ 0'

I g]<Igl<lg'

VJ <

Forward leokcge

- I,


101 102 10 3

Gat Currn t ~)

Fig. 4.4. Effect of gate current on forward breakover voltage.

thyristor can also be illustrated by means of a curve as shown in Fig. 4.4 (b) . For I, < oa, forward
breakover voltage remains almost constant at VBO ' For gate currents 1,1,18'1 and 1,3' the
magnitudes of forward breakover voltages are ox = Vi' OJ = V2 and oz =V3 respectively as
sh own in Fig. 4.4 (a ) and (b). In Fig. 4.4 (a), th e curve marked I, =0 is actually for gate ::urrent
less than 00. In practice, the magnitude of gate current is more than the minimum gate current
required to turn on the SCR. Typical gate current magnitudes are of the order of 20 to 200 rnA.
Once the SCR is conducting a fofward current, reverse biased junction J 2 no longer exists.
As such, no gate current is required for the device to remain in on-state. Therefore, if the gate
current is r emoved, the conducti on of current from anode to cathode remains unaffected.
However, if gate current is reduced to zero before the rising anode current attains a value,
called the latching current, the thyristor will turn-off again. The gate pulse width should
therefore be judiciously chosen to ensure that anode current rises above the latching current.
Thus [atckin current may be deemed as the minimum value of anode current which it must
attmn uring turn-on process to mam am con uc Ion w en gate signal is remove .

Once the thyristor is conducting, gate loses control. The thyristor can be turned-off(or the
thyristor can be return ed to forward blocking state) only if the forward current falls below a
low-level current call ed the h olding current. ,!,hus h olding current may be defined as the
minimum value of anode current below which it must fall for turning-off the thyristor. The
Iitching current IS hIgher than the holding current. Note _that latching current is associated .
\V1th turn-on prct:ess and holdin[..c urrent wi ~h turn-off process. It is usual t o take latching
current as two to three tImes the h oiding current tIl. In mdUstria::t applicatioos, holding current
(typically 10 rnA) is alm ost taken as zero.
(c )

~~ Triggering. Wi th forward voltage across th~ anode

and cathode of a thyristor, the

two oute r junction J 1 JJ are fo rw ard biased, but inner junction J 2 is revers e biased. Thi5
revers e bias ed junction J,!, Fig. 4.3 (b), h as th e characteristics of a capacitor due to charges
exis tin g across the jLlOction. In other words, space-charges exist in the depletion region nea r
junction J 2 and th~:-efore junction J 1 ~e ha.ves like a capacitance. If for-vard voltage is su ddenly


Power Electronics

[Art. 4.21

applied, a charging current th rough junction capacitance CJ may turn on the SCR. .AJrnost the
entir e suddenly applied forward voltage Vo appears acrcssjunction J 2 the char ging current ir:
is , ther efore, given by

. _ <!SI _ !


_ dV.
dt - dt (Cj . Va) - CJ dt + V(l o dt

As the junction capacitance is a lmost constant.

... (4.1 a )

is zero and current ie as


... (4.1 b)

'e= CjTt

Therefor e, if rate of rise of forward voltage dVa / dt is high, the ch arging current ic would
be more. This charging current plays the role of gate current and turns on the SCR even th ough
gate signal is zero. Note th at even if Va is small, it is the rate of change of Va that plays the
role of turning-on the device.
(d ) Temperature Triggering (Thermal Trigge ring) . During forward blocking, most of
the applied voltage appears across revers e biased junction J 2. Thi s voltage across , J 2
associated with leakage current, would raise the temperature of this junction. With increase in
temperature, width of depletion layer decreases. This further leads to more leakage curr ent and
t herefore, more junction temperat ure . With the cumulative process, at. some high temperature
(within the safe limits), depl etion layer of r everse biased junction vanishes and the device gets
turned on .
(e) Light Triggering. For light.triggered SeRs, a recess (or niche) is made in the inner
player as shown in Fig. 4 .5 (a ). When this recess is irradiated , free charge carriers (pairs of
h oles and electrons ) are generated just lik e when gate signal is applied between gate and
cathode. The pulse of light of appropriate wavelength is guided by optical fibres for irradiation .

If the intensity of this light thrown on the t ecess exceeds a certain value, forward-bi ased SCR
is turned on . Such a thyristor is known as light-activated SCR (LASCR).
LASCR may be triggered with a light sou r ce OT with a gate signal. Sometimes a
combination of both light source and gate signal is used to trigger an SCR. For this, the gate is
biased with voltage or current slightly less than that required to turn it on, now a beam oflight
directed at the inner player junction turns on the SCR. The ligh~ intensity required to turnon
the SCR depends upon the voltage bias given to the gate. Higher the voltage (or curren t) bias,
lower the li ght intensity required.
Light trigger ed thyristors have now been used in high-voltage dir ect current (HYDC)
transmission systems. In these several SCRs are connected in series parallel combination and
their light-triggering h as the advantage of electrical isolation between power and control
Example 4.1 . Discuss what would happer. if gate is made positiue with respect to cathode
during the reverse blocking of a thyristor.
S olution . Before answering t his question, it is worthwhile to know
in terna l details of a thyristor.

little more about the

Fig. 4.5 (b) shows cross- section of a conventional cen tr e-gate t hyri stor. In this figure ,
a pproxi mate doping densities in num ber per cubic centimeter are also indicated for all t he four
layers. For example, for PI layer, the doping den sity is 10 19 per cm 3 .

[Art. 4.3]



A node



Light -



P, (pot)

pp "

n,(n-) 10" to 5x






10"cm- 1

P1(p ot )




Fig. 4.5. (0) Elementary LASeR (b) Structural details of conventional centre-gate thyristor.
For semiconductor devices, it should be kept in mind that
(i) junction with lightly doped layers (at least on one side of the junction) requires large
breakdown voltage,
(ii) junction with highly doped layers on both sides requires low breakdown voltage.
When thyristor is in forward blocking state, junctions J 1, J a are forward biased whereas
j unction J 2 is r everse biased. As layer n l is lightly doped around j unction )2' depletion r::egion
of junction J 2 extends mainly into n} layer. Therefore, n 1 layer is made to h ave larger width
to withstand the high voltage during forward blocking state.
For reverse voltage on the device; junctions J 1, J 3 are r~verse biased and J 2 is forward
biased. As layers P2, n2 across junction J 3 are heavily doped, J 3 h as low breakdown voltage.
Layer nl being lightly doped as compared to layer Pl,juDction J 1 has large break'down voltage.
As a consequence, during the reverse blocking of a thyristor, junction J 1 suppor ts most of the
r everse voltage. Even during blocking, the depletion r egion extends into tne nl layer. This
shows that width of layer nl absorbs most of the voltage during forward blocking mode and
also during the reverse blocking mode of a thyristor.
If positive gate voltage is applied between gate and cathode during the rever se blocking
of a thyristor, blocking property of junction JJ disappears as J 3 has low breakdown voltage.
As a result, reverse voltage appears across junction J l ' Positive charge carriers are now inj ected
into the n 1 layer of reverse biased junction J 1. This causes an increase in the reverse leakage
current . The flow of large leakage current associated with high reverse voltage results in
increased power loss across junction J 1 and heat thus generated may raise the junction
temperature above the allowable maximum and this may destroy th.~CR. Such an happening
can be avoided if no positive gate voltage is applied between gate and cathode during the
reverse blocking of SeR. Some manufacturers do specify the maximum positive voltage
(usually less than 0.25 V) that can exist between gate and cathode during the reverse blocking
of a thyristor.


Static and switching characteristics of thyristors are ahvays taken into consideration for
ec onomical and r eliable design of converter equipm ent. Static characteristics of a thyristor


Power Elec lronics

fArt. -lJI

have already be en examined. In this part of the section; switching , dynamic or transient ,
characteristics of thyristors are discussed.
During turn-on and turn-off processes, a thyristor is subjected to different voltages across
it and different currents through it. The time variations of the volta ge across a thyristor and
the current through it during turn-on and turn-off processes give the dynamic or switching
characteristics of a thyristor. Here, first switching characteristics during tum-on are described
and then the switching characteri stics during tum-off.

4.3.1. Switching Charact eristi cs during Turn-on

A forwa rd-biased thyristor is usually turned on by applying a positive gate voltage between
gate and cathode. There is, however, a transition t ime from forward off-state to forward on
state. This transition time, called thyristor tum-on time, is defmed as the time during which
it changes from forward blocking state to final on-state. Total tum-on time can be divided into
three intervals; (i) delay time tel, (ii ) rise time tr and (iii) spread time tp ' Fig. 4 .8.
(i) Delay time td : The delay time t d is measured fr om the instant at which gate current
r eaches 0.9 Ig to the instant at which anode current r eaches 0.110' Here 16 and 10 are
r espectively the final values of gate and anode currents. The delay tim e may also be defined
as the time during which anode voltage falls from Vo to 0.9Vo whe re Vo =initial value of anode
voltage. _-'\n other way of defining delay time is the time during which anode current rises from
forward leakage current to 0.110 wher e 10 = final value of anode current. With the thyristor
initially in the forward blocking state, the anode voltage is OA and anode current is small
leakage current as shown in Fig. 4 .8. Initiation of turnon process is indicated by a rise in
anode current from small forward leakage current and a fall in anode-cathode voltage from
forward blocking voltage OA. As gate current begins to flow fr om gate to cathode with the
application of gate signal. the gate current has non-uniform distribution of current density
over the cathode surface due to the p layer. Its value is much higher near the gate but decreases
r apidly as the distance from the gate increases, see Fig. 4.6 (a). This shows that during delay
time t d , anode current flows in a narrow region near the gate where g-ate current density is
the highest.

The delay time can be decreased by applying high gate curren t and more fo rward voltage
between anode and cathode. The delay time is fraction of a microsecond.






(b )

Fig. -1.6. la ) Distribution of gate and anode currents durin g delay time
(b) Conducting area of cathode (i) during td (ii ) afte r tr (iii) after tp'


[Art. Uj


Rise time t,.: The rise time t,. is the tim e taken by the anode cur rent to rise fr om 0.1

Ia to 0'.9 I a' The rise time is also defined as the time required for the forward bloc kin g off-state

voltage to fa ll from 0 .9 to 0.1 of its initial value OA. The ris e tim e is in versely pr oportional to
the magnitude of gate current and its build up rate. 'Thus t,. can be reduced if high and steep
current pulses are applied to the gate. However, the main factor determin ing t,. is the nature
of anode circuit. For example, for series RL circuit, the rate of rise of a node current is slow,
ther efore, t,. is more. For RC series circuit, dU dt is high, t,. is therefore, less.
From the beginning of ris e time t,., anode current starts spr eading fr om the narrow
condu cting r egi on near the gate. The an ode
current spreads at a rate of about 0.1 mm per \ g
microsecond [2]. As the rise time is small, the
anode current is not able to spread over the entire
cross-section of cathode. Fig. 4.6 (b ) illustrates how
anode current expands over cathode surface ~rea
during tum-on process of a thyristor. Here \the
thyri stor is taken to have single gate electr~e
a way fr om the ce ntre of p -layer. It is seen that
anode current conducts over a small conducting
chann el even after t,. - this conducting channel
area is however, greater than that during t d-.
Fig. 4.7. Typical waveform for
gate current.

During ris e time, turn-on losses in the

thyristor are the highest du e to high anode voltage (Va> and large an ode current (I a) occurring
together in the thyristor as shown in Fig. 4 .8 . As these losses occur only over a small conducting
r egion , local hot spots may be formed and the device may be dam aged.
(iii) Spread time tp : The spread time is the time taken by the anode current to rise from
0.9Ia to I f1" It is also defin ed as the time for the forward blocking voltage to fall fr om 0. 1 of its
ini tial value to the on-state voltage drop (1 to 1.5 V ). During this time, condu ction spreads
over the entire cross-section of the cathode of SCR The spr eading interva:l depends on the
ar ea of cathode and on gate structure of the SCR After the spread time, anode current a ttains
steady state value and the voltage drop across SCR is equal to the on-state voltage drop of the
order of 1 to 1.5 V, Fig. 4 .8.
Total turn-on tim e of an SCR is equal to the sum of delay time, rise time and spread tim e.
Thyristor manufacturers usually specify the rise time which is typically of the order of 1 to 4
)l sec. Tota l tum~on time depends upon the anode circuit par a meters an d the gate signal
wavesh ap es.
Durin g turn-on, SCR may be consider ed to be a charge controlled device. A certain amount
of charge must be injected in to the gate r egion for th e thyri stor conduction to begin. This charge
is dir ectly pr oportion al to the value of gate cu rren t. Therefor e, higher the magnitude of gate
curren t, the lesser time it takes to inject this charge. The turnon time can therefor e be reduced
by usin g higher values of gate currents. The magnitude of gate current is usually 3 to 5 times
the minimum gate current required to tri gger an SCR
When ga te current is sever al times high er than the minimum gate current required, a
thyristor is said to be hard-fired or overdriven. Ha rd -firing or ouerdriuing of a thyri3 tor reduces
its tum-on tim e and enhances i ts dildt capability. A typical waveform for gate current, that is
wide ly used, is shown in Fig. 4.7. This wavefor m has high er initial -nIue of gate current wi th a
very fast rise time. The initial high value of gate curren t is then r educed to a lower value where
it stays fa r sever al micrasecan d3 in order to avaid unwanted turn-off of the device.


[Art. 4,3J

Pow e r Electronics

4.3.2. Switching Characteristics during Turn-off

Thyristor tu m-off means that it has changed from on to off state and is capable of blocking
the forwa rd voltage. This dynam ic process of the SCR from conduction state to forward blocking
state is called commutation process or turn -off process.
Once th e thyristor is on, gate loses control. The SCR can be turned off by reducing the
anode current below h olding curren t , If forward voltage is applied to the SCR at the moment
its anode current falls to zero, the device will not be able to block this forward voltage as the
carriers (holes and e lectrons) in the four layers are still favou rable for conduction . The device
wi ll therefore go into conduction imm ediately even though gate signal is not applied. In orde r
to obviate such an occurrence, it is essential th at the thyristor is reverse biased for a flnite
period after the anode current has reached zero.

VOltog., Vg

Gate pulse

Anode voltoge LIe and
gote curr ent i g

On state voltagE'
drop across SCR


, I
current ]


Anode curre-nl
begins 10





. Reverse voltage
due to powt>r circuit


POWt>f Ioss
( Ucla



C:lmmutation I I

" I




,R ecombination

,I '



t l,




I i..- Steady stole

--I. II - operation

0 .1 10


t, ---i
t-- tq -----'1 I ,

" [I i 'I"~,~_I_
I' Q_~_"Q_O_C_~_"_._n~'


. ....

A~--l g

0.9 I9

o, ~

-. . . ". ~ ....
" :.i '~

--0.9Ve OA =Ve= lnitial anode voltage


t tr - l -tg l

; - - tq
I ,

t , - - --'

' 1

I .

Time In Microsec

fig LS Thyristor voltage and cu rrent wavl! fo r rru du:ing turn-on and turn-off proces.s-es .
'Th )" C;1:l be achie\"erl thNl ugh nat ural comm u ta tion or forced com:":"lutation.


[Art. 4.3)


The turn-off time tq of a thyristor is defined as the time between the instant anode current
becomes z.ero and the instant SCR regains forward blocking capability. During time tq , all the
excess carriers from the four layers of SCR must be removed. This removal of excess carriers
consists of sweeping out of holes from outer p-Iayer and electrons from outer n-layer. The
carrier s around junctionJ2 can be r emoved only by recombina tion . The tum-off time is di vided
into two intervals; r everse r ecovery tim e tTT and the gate recovery time t,r; i.e. tq = tTT + t,r The
thyristor characteristics during turn-on and turn-off processes are sh own in one Fig. 4.8 so as
to gain insight into these processes.
At instant t 1 anode current becomes z.ero. After t 1, anode current builds up in the reverse
direction with the same dildt slope as before t 1. Th e reason for the rever sal of anode current
after tl is due to the presence of catTiers stored in the four layer s. The r everse recovery current
r emoves excess carriers fr om the end junctions J 1 and J 3 between the instants tl and t3' In
other words, r everse recovery current flows due to the sweeping out of holes from top p-Iayer
and electrons from bottom n-Iayer. At instant t 2 when about 60% of the stored charges are
r emoved from the outer two layers , carrier density across J 1 and #3 ..begins to decr ease and
with this reV,e rse recovery current also starts decaying. Th~ reverse current decay is fast in
the beginning but gradual thereafter. The fast decay of r ecovery current causes a 'reverse
voltage across the device due to the circuit inductance. This reve r s~ voltage surge appear s
across the thyristor terminals and may therefor e damage it. In practice, this is avoided by
using protective RC elements across SCR. At instant t 3. when reverse r ecovery current has
fallen to ne arly zero value, endjunctionsJ1 and J 3 recover and SCR is able to block the reverse
voltage . For a thyristor, r everse r ecovery phen omenon between tl and t3 is similar to that of a
r ectifier diode.
At the end of r everse recovery period (t3 - t 1), the middle junction J 2 still has trapped


charges, the refore, the thyristor is not able to block the forward voltage at
The trapped
charges around J 2 i.e. in the inner two layers, cannot flow to the external .circuit, therefore,
these trapped charges must decay only by recombination. This recombination is possible if a
reverse voltage is maintained across SCR, though the magnitude of this voltage is not
important. The rate of recombination of charges is independent of the external circuit
pa.ram eters. The time for the recombination of charges between t3 and t4 is called gate recouery
time t,r' At instant t", junction J 2 recovers and the forw ard voltage can be reapplied between
anode and cathode. The thyristor turn-off time tq is in the range of 3 to 100 )lsec. TI:e turn-off
time is influenced by the magnitude of forward current, dildt at the time of commutation and
junction temperature. An increase in the magnitude of these factors increases the thyristor
turn-off time. If the value of forward current before commutation is high, trapped charges
around junction J 2 are more. The time required for their recombin ation is mor e and ther efore
tu::-c..-off time is increased. But turn-off time decreases with an increase in the magnitude of
reverse voltage. particularly in the range of 0 to - 50 V. This is because high reverse voltage
sucks out the carriers out of the junctions J I , J 3 and the adjacent transition region::; at a fa3tcr
r ate. It is evident from above that turn-off time tq is not a constant parameter of a thyristor.
Th e th)..-ristor turn-off time tq is applicable to a.T). indi,,;dual SCR. In actual practice. thyristor
(or thyristors ) fonn a part of the power circuit. The rum-off time provided to the thyristor by th~
practical circuit is called circllit tum offtime t l! ' It is defined a.5 the time bet\",een the instant anode
current becomes zero and LI-J.e iIlstant reverse voltage due to practical circuit reaches zero, sec Fig.
4.8. Time tc must be greater than tq for reliable turn-off, otherwise the device may turn-on at a n
UndE'3U-ed instant, a proc ess called commutation failure .

Power Electronics

[Art. 4.4]


Thyristors with slow turnoff time (50 - 100 ~sec ) are called converter grade SeRs and
those with fast tum-off time (3 - 50 lJ.Sec) are called inverter.grade SeRs. Converter-grade
SCRs are cheaper and are used where slow turn-off is possible as in phase-controlled rectifiers,
ae voltage controllers, cycloconverter s etc. Inverter-grade SeRa are costlier and are used in
inverters, choppers and force-commutated converters.


The forward gate characteristics of a thyristor are shown in Fig. 4.9 in th e form of a graph
between gate voltage and gate current. Here positive gate to cathode voltage V, and positive
gate to cathode current I, represent de values. As gate-cathode circuit of a thyristor is a p-n
junction , gate characteristics of the device are similar to that of a diode. For a particular type
of SCRs, Vg-Ig characteristic h as a spread between two curves 1 and 2 as shown in Fig. 4.9.
This spread, or scatte r, of gate ' characteristics is due to inadvertent difference in the doping
levels of p and n layers . '!'he gate trigger circuitry must be suitably des igned to take care of
this unavoidable scatter of characteristics. In Fig. 4;9, curve 1 represents the lowest voltage
values that must be applied to turn~on the SCR. Curve Z gives the highest possible voltage
." .
values that can be safely applied to gate circuit.
Each thyristor has maximwn limits as Vgm for gate voltage and Ip for gate current. There is
also rated (average) gate power dissipation PIfI1IJ specified for each SCR These limits should not be
exceeded in order to R.void permanent damage of junction J 3 , Fig. 4.3. There are also minimum limits
for V, and I , for reliable tum.-on, these are represented by ~ and ox respectively in Fig. 4.9. As
stated before. if Vp . I,m and Ptau are exceeded, the thyristor can be destroyed. This shows that
preferred gate drive area for an SCR is bcdefghb as shown in Fig. 4.9.


Minimum ga te voltage a nd current to

trigger an SeR.
Vgm , I,m - Maximum permissible gate voltage
and current.
00 Non- triggering gate voltage.
Oy. ox -




Fig. 4.9. Forward gate characteristics of thyristor.

A non-triggering gate voltage is also prescribed by the manufactur ers of SeRs. This is
indicated by oa in Fig. 4.9. If fl.!""ing circuit generates positive gate signal prior to the desired
i. n.stant of triggering the SCR, it should be ensured that this un wanted sipal is less than the
non-triggering gate voltage 00 . At the same time, all spur ious a T noise s ign als sh ould be less
than the voltage oa.
The design of the firing circuit can be carried ou t with the help of Figs . . t I 0 and 4.11. In
Fig. 4.10 (0 ) is shown a trigger circuit feed ing power to gate-cathode circuit. For this circuit,



Trt gg~r
circuit ,

+ G






L-______L-~-J K
1l ______________ ---J,





,r -1


l _____ __ __________

(a )




(b )

Fig. 4.10. Trigger circuit connected to gate-cathode circuit of an SCR.


=Vg +lg Rs

... (4 .2a )


Es= gate source voltage

Vg = gate-cathode voltage
I N = gate current
R, =gate-source resistance
The internal resistance R, of trigger source should be such that current (E/R~ ) is not
harmful to the source as well as to the gate circuit when SCR is turned on. In case R, is low,
an external resistance in series with R s must be connected.
A resistance Rl is also connected across gate-ca thode terminals, Fig. 4.10 (b ) , so as to
provide an easy path to the flow of leakage current between SCI{ terminals. If I ClHn and V Nmn
are the minimum ga te current and gate voltage to turn- on.SCR, then it is seen fr om Fig. 4. 10
( b ) that current through Rl is Vgmn l R 1 and the trigger source voltage E$ is given by

E,+,m" + VIC JR,+ V,m"

For low-power circuit3, it is customary to
obtain the operating point by utilizing the V-I
characteristics of both source and the device. In
view of this , for selecting the operating poin t
for the circuit of Fig. 4. 10, a load line of th e gate
sou rce volt age Es = OA is drawn as AD in Fi g.
4.11. Here OD = trigger circui t short circuit
current = E,I R J . Let us consider a thyristor
who5e V g-Ig char acteristic is given by curve 3.
In tersecti on of load line AD and Vg -Ig curve 3
gives the operating point S. Thus, for this SCR,
gate voltage =PS and gate current = OP. In
ord er to mini mis e turn -on time and j itte r
(u:1reliable tu m-on ), the load lin e and hence the
operating point 5, whi ch may ch an ge fr om
S! to 5 2. mus t be as cl ose to the P gCl(' curv e as
poss ible . .-\t th e same tim e, the oper ating !Joint

.. .(4.2b )





" 0

Sl~" - '



Igm 0

Gate Curnmt-

Fig. 4. 11. Choice of ga te circu it po:-a:neters


P ower Electronics

[Art. 4.4]


S must li e within the limit curves 1 and 2. The gradie nt of the load line AD (= OAI OD) will
give the required gate source resistance Ro!. The minimum value of gate source series resistance
is obtained by drawing a line AC tangent to Pgau curve.
Gate drive requirements in terms of continuous de signal can be obtained from Fig. 4.11 .
However, it is common to use a pulse to trigger a thyristor. For pulse widths beyond
100 I-lsec, the de data apply [1]. F or pulse widths less than 100 ~sec , magnitudes of gate voltage
and gate current can be increased, see Example 4.2.
As stated before, thyristor is considered to be a charge controlled device. Thus, higher the
magnitude of gate current pulse, lesser is the time to inject the required charge for turning-on
the thyristor. Therefore, SCR turn-on time can be reduced by using gate current of higher
magnitude. It sh ould be ensured that pulse width is sufficient to allow the anode current to
exceed the latching current. In practice, gate pulse width is usually taken as equal to , or gt:eater
than , SCR turn-on time. If T is the pulse width as shown in Fig. 2.12 (a), then T ~ ton
With pulse triggering, grea~er amount of gate power dissipation can be allowed ; this
should, however, be less than the peak instantaneous gate power dissipation Pgm as specified
by the manufacturers. Frequency of firing (or pulse width) for trigger pulses can be obtained
by taking pulse of (i) amplitude Pgm (ii) pulse width'T and (iii) periodicity T 1. Therefore,



Pgm " T f>P

- gau



. or

... (4 .30 )

fT $Pgm

wher e

f = ~ =fre quency of firing, or pulse repetition rate, in Hz,


T = pulse width in sec.

or f= ~;-P.!;"~,T Pgm

In the limiting case,

A duty cycle is defined as the ratio of pulse-on period to periodic time of pulse. In Fig.
4.12 (a), pulse-on period is T and periodic time is T 1. Ther efore, duty cycle 5 is given by


'E ig





Fig. 4.12. (a) Pulse gating a nd (b ) high-frequency carrier gating of SCRs ,

(c) Thyristor protection against reverse overvolta ges .


[Art. U J

From Eq. (4.30 ),


-.E 'S



.. .( 4.3 b)

Sometimes the pulses of Fig. 4.12 (a) are modulated to generate a trai n of pulses as shown
in Fig. 4.12 (b). This technique of firing the thyristor is called highfrequency carrier gating.
The ad vantages offered by this method offiring the scas are lower rating, reduced dimensi ons
and therefore an overall economica1 design of the pulse transformer needed for isolating the
low power circuit from the main power circuit.
For an SCR, Vim and I,m are specified separately. If both of these are used for pulse firing,
then Pim may be exceeded and the thyristor would be damaged. For example, GEC35 thyri stor
has Vim = 10 V and I,m = 2 A. If both these limits are pl aced on C35 , the power dissi pation is
20 W. But this is far excess of the specified Pgnt = 5 W. It should be ensu red that (pulse voltage
amplitude) (pulse current. amplitude) < Pgnt .
There is also prescribed a peak reverse voltage (gate negative with respect to cathode ) that
can be applied across gate-cathode terminals. Any voltage signal, given by the trigger circuit
(or by any interference), exceeding this prescribed limit of about 5 to 20 V may damage the
gate circuit. For preventing the occurrence of such hazard s, a diode is connected either in series
with the gate circuit or across the gate-cathode terminals as shown in Fig. 4.12 (c ). Di ode
across the gate-cathode terminals, called clamping diode, prevents the gate-cathode voltage
from becQming more than about 1 V. Diode in series wi th gate circuit prevents the fl ow of
negative gate source current from becoming more than small reverse leakage current.
The magnitude of gate voltage and gate current for triggering an SCR is inversely
proportional to junction temperature. Thus, at very low temperatur es, gate voltage and gate
current must have high values in order to ensur e turn-on. But P,m should not be exceeded in
any csse;The res istor R 1. connected across gate-cathode terminals, Fig. 4.10 (b ), also serves to bypass
a part of the thermally-generated leakage current across j unction J 2 when SCR is in the
forward blocking mode ; this improves the thermal stability of S'CR.
Example 4.2. (a) The average gate power dissipation for an SCR is 0.5 W. The allowable
gate voltage variation is from a minimum of 2 V to a maximu m of 10 V. Taking ar;e rage gate
power dissipation constant, plot allowable gate voltage as a fun ction of gate current.
(b) If SCR of part (a) is triggered with gate pulses of du ty cycle 0.5, find the new valLle or
average gate power dissipation.
Solution . (a) Here
F or

v"Ig = 0.5 W

Vi = 2 Y, Ig = 0.5/2 = 0. 25 A to
Vg = 10 Y, Ig = 0.5/ 10
= 0.05 A
For other values of ga ts voltaot? V, in between 2
and 10 V, gate current 19 is obtained and plotted in Fig. >c::
- 4
4.13 showing the variation of V, 83 a fun cti on of I, for
constant p,cu.
(b) For this example, T I =2T in Fig. 4.12 (a)

. __ ..


I: -

I f = 0 .5 W.

so th at B = 0 .5. For dc values, V

F or puls e firing, Fi g. 4.12 (a ), the average gate

power dissip ation can be ob tain ed fr om the r elation

: :

:; 1,
..-- ..' I..,

P9 C ~ :0' 5 W

.. -l..J ___~_
' :

. ::.-:+t.::-:1::~::.-::::_. , ...
, I

t: :



0.1 0.15 0.1

Ig in A -



Fig. U3 . Perta ining to E:, :'.mpl i:! 4.2.

Power Electronics


where UN' iJl are th e ins tantaneous val ues of gate voltage and gate current. Therefore, for this
e xampl e, av erage gate power dissipation is given by

V, I, ;;. = (o.o)!

=0.25 W.

As thi s is less th a n the all owable Piau, higher values of ti" i, can be used fo r the pu lse firin g
of SeRs.
Example 4.3. For an SeN. the gate-cathode characteristic has a straight-line slope of 130.
For trigger so urce voltage of 15 V and allowable gate power dissipation of 0.5 watts, compute
the gate-source resistance.
Sol uti on . Here
V, I,=0.5W
-:.L = 130


Thi s gives

I, = [0.51130) ln~ Q.052 = 62 rnA

:. Ga te voltage.
F or th e gate circuit,

V, = 130 x 62 x 10- ' =8.06 V

E, =I, R, + V, = 0.062 R , + B.06 = 15
15 - 8.06
R, = 0 .062 = 1l1.94l1.

Exam p le 4.4. The trigger circuit of a thyristor has a source voltage of 15 V and the load
l ine has a slope of - 120 V per ampere. The minim um gate current to tum- on the SCR is 25
mA Compute
(a) source resistance required in the gate circu it,
(b) the trigger voltage and trigger current for an average gate power dissipation of 0.4 watts.
Solution. (a ) The slope of load line gives the required gate source resis tance. From the
load line, se ri es resis tance required in the gate circuit is 1200.
(b) Here
V, I, = 0.4 W
EJ =R)c + V,
F or th e gate circuit,



15 = 120 I + 0.4

120 Ii - 15 Ig + 0.4 = 0
Its so luti on gives
I, =38.56 mAo or 86.44 rnA


= 0.4 X


= 10 37 V

V - 0.4 X 10' _ 4 6? - V
,- 86.44 - . .. 1

Choose Vg = 4.627 V and I,

=86.44 rnA for minimum gate current of

25 rnA.

Example 4.5. Foran SCR, gate-cathode characteristic is gi uer. by Vi = 1 + IO I~. Gate sou.rce
()oltage is a recta ng ula r pu lse of 15 V with 20 )J sec du. retiol!.. For an average g ate power
d issipation 0/ 0.3 \V and a pec k g ate -dri ve pOlL"er ol5 ~v, compute
(a) the resistance to be connected in series with the S CR gate,
(b) the triggering fre quency and

[A rt. 4.4]



(c) the duty cycle of the triggering pulse.

Solution. (0) Here
V, = 1 + 10 I,.
For pulse-triggering of SCRs,
(P eak gate voltage) (peak gate curr ent) during pulse-on period
= peak gate drive power, P,m'
As the gate pulse width is 2.0 ~ sec (less than 100 ~ sec), the dc data does not apply. Had
the gate pulse width been more than 100 j.LSec, the relation (1 + 10 I.) I, = 0.3 W will hold good .
But as the dc data does not apply, we have here

(1 + 10 I,) I, = 5 W




Its solution gives,

:. Amplitude of current pulse
During the pulse-on period,


E,=R, Ig+V,
15 = R, I, + 1 + 10 I,


15 - 1

R, 0.659 -10= 11.2440

(b )


:. Triggering frequency,

= 0.659 A.
=0.659 A

Pr fJ II

= fl"


H ere T= 20 ~sec

0;:ig' =3kHz

B=fT= 3 x 10' x 20

Duty cycle,

10"= 0.06,
0 '

Example 4 .6. Latching current for an SCR, inserted in between a dc uoltage source of200

V and the load, is 100 mA. Compute the minimum width of gatepulse current required to
turn-on this SCR in case the load consists of (a) L =0.2 H, (b) R = 20 n in series with L =0.2
H and (c) R = 20 n in series with L = 2.0 H.
Solution. (a) When load con sists of pure inductance L, the voltage equation is

E= L


0100 =


~og t


d'L = E



0.1 x 0.2

Thus, minimum gate-pulse is 100 ~sec

(b) The voltage equation for RL load is

L -di
E =R L+


. E( '!!.')


t=R 1 -


t = 100.503




~s e c

: . Minimu m ga te-pulse width is 100.503 ).l5eC

. -E t


= 100 ~sec


Power Electronics

[Art. 4A]

. -E


(1 -!!,)

0.1 = 22000 (1 _ .-1o ,)



t =1005.03 ~see .

This example shows that if load resistance is increased from zero to 20 n, the gate-pulse
width remains almost unaffected. But with an increase in indu ctance from 0.2 H to 2 H, the
gate-pulse width becomes 10 times its previous value.
Example 4.7. Th e gate current of a forward biased SCR is gradually increased from zero
until the device is turned on. It is obserlJed that gate CUTTent, just prior to the instant of ttlrn -on,
is 1 mA and soon after SCR goes in.to conduction, gatt. current decays to about 0.3 rnA. Discuss
how it happens.
Solution . When anode of an SCR is made positive with respect to cathode, a small voltage
E'g genera ted internally, appears across the gate- cathode terminals, Fig. 4.14 (a ). The
magnitude of E'g depends upon applied anode voltage and the device geometry. In the
gate-c athode equivalent circuit of Fig. 4.14 (0), R is the static non-linear gate resistance.
If the SCR is turned on by applying a positive gate signal, then the equivalent circuit for
the trigger circuit is as shown in Fig. 4.14 (bl. Here E, is the gate voltage ~e nerated internally
due to the flow of anode current. The magnitude of E'g is mu ch smaller as compared to Eg. For
a typical SCR, E'g = 0.05 V and Eg '" 0.7 V.

~~;~~i~ '{ ----r-----J'-v--\-"-r---~'_.:G9-,_,;,lo!...,






(a )

Fig. 4. 14. Pertaining to Example 4.7.

Before the SCR starts conducting, gate current

I;:: E,I(R + R J ). Alt er SCR goes into conduction, I, =

I,' = ER-E'
+ R: . As E;

is very small,

E - E
+ R:' Voltage Eg is quite large as

compar ed to E'g, therefore , gate current is reduced fr om a higher value of 1,' to a lower value
oi I,.
In ca3e E, is reduced to zer o, gate current becomes n egative with its value equal to


Iz" = R +

A Under this condition of E

= 0, the voltage appearing across the gate-cathode

term in als i5 Eg - Ig" R.

E xa m pl e 4.9. Gate-cathode characteristics of a thyristor ha ue a spread g iuen. by the
following two relations:

Ig = 2.1 x 10- 3 V; and Ig = 2.1 X 10- 3 V;S



Thy ristors



T he gate source voltage is 16 Vand load line has a slope of - 128 V I A. Calculate the trigger
voltage and trigger current for an average gate power dissipation of 0.5 W
Are the val ues of V" I, obtained here justified? Discuss.
Solut ion. Slope of load line gives gate-source resistance, R.
and E, =I, R.+ VI

= 128 n. Here Vg Ig = 0.5 W

16= I ,x 128+ 0/



I; -16 I, + 0.5 =0

= 62.5 rnA and V, =a v

So point 8 in Fig. 4.11 has V, = 8 V and Ig = 62.5 rnA.
For the same V,; I, = 2.1 X 10- 3
gives mor e I" therefore it represents curve 1
6.11 . Point 8 1 on this curve can be obtained from V, I , = 0.5 Wand I, =2.1 x 10- 3 ~
It solution gives I ,



I, = 2.1 x 10- 3 (

Point 8 2 can be obtained from V, l j


I, = (2.1)

10- 3


of Fig.

orI, = 80.67 rnA and V, .= ~:198 V

=0.5 W and I, = 2.1 x 10- 3 V:5



or I, = 56.01 rnA and Vg = 8.93.V.

Since point 5 (8 V, 62.5 rnA) lies in between 5, (6.198 V, 80 ..67 rnA) and 5 , (8.93 V, 56.01
rnA) as desired, the calculated values of V,; a V and I,l =6'2'.5 rnA are j ustified.



The principle of thyristor operation can be explained with the 'use of its two-tr ansistor
model (or two-transistor analogy). Fig. 4.15 (a) shows schematic diagrum"o-f a thyristor. From
this figure, two-transistor model is obtained by bisecting the two middle layers , along the
dotted line, in two separate halves as shown in F ig. 4.15 (b ). In this figure , junctions
J 1 - J 2 and Jz - J 3 can be considered to constitute pnp and Itpn transistors separately. The
circuit repres entati on of the two-transistor model of a thyristor is shown in Fig. 4.15 (c lIn the off-state of a transistor, collector current Ie is rel a ted to emitter current IE as

Ie = CJ. IE + l cBD
wher e a is the com mon base cu. rrent g ain and ICBO is the comm on base leak age current of
collector- base j un ction of a transis tor.
For transis tor QJ in Fig. 4. 15 (c), emitter current Ie = anode current Ia a nd Ic = collector
current le i' Th er efor e, for Q t,
wher e
an d

ICI = a lIa + I eBol

al = common- base current gain of Ql
Ieaol = co mmon- base leak age cu rr en t oi Q!.

...(4 .4 )

Power Electronics

[Art. 4.5]





G 1,




P ,, P


P n


G [,









G 1,



I "~

P a2'PZ
0, n




(b )

Fig. 4.15. Thyristor (a) its schematic diagram, (b) and (c) its twotransistor model.

Similarly. for transistor Q2, the collector current IC2 is given by

I C2 =

... (4.5 )

ct:z lie + I CB02

ct:z =common-base current gain of Q2

I CB02 =common-base leakage current of Q 2
I. = emitter current of Q2The sum of two collector currents given by Eqs. (4.4) and (4.5) is equal to the external circuit

current f a entering at anode terminal A.

:. I", =IC1 +IC2


=Ci! 1(1. + I CBOl + ~ I. + I CB02

.. ,( 4 .6 )
When gate current is applied. then lit =10 + I , _Substituting this value afl ll in Eq. (4.6) gives
10 =Ci l It:! + I CBOI + ct:z (10. + 1,) + 1eB CY.2


... (4 .7 )
1- (a l + <XV
For a silicon transistor, current gain a is very low at low emitter current. With an increase
in emitter current, a builds up rapidly as shown
Current gain, 0.
in Fig. 4.16. With gate current I, = 0 and with
thyristor forward biased, (al + ~ is very low as
per Fig. 4.16. Under these conditions, Eq. (4.7)
shows tha t forward leakage current ' somewhat
mo re than (l CBOI +lcBo2) fl ows . If. by so m e 0. 75
mea ns, the emitter current of two component
transisto r s ca n be incre ased so that at + a..z
app roaches unity, then as per Eq. (4.7), 10 would 0';0
tend to become infinity thereby turning-on the 0. 25 device. Actually, external load limits the anode
current to a safe value after the thyristor begins
conduction. Th e meth ods of tu rn ing-on a
E mitt ~ r curr ent
thyristor, in fact , are the methods of making
4.16. 'Typical v9.riation of current gain
a 1 + Cl.l to a pproach unity. The3e va ri ous Fig.wi
th emitter curr~n~ oi a thyristor.


=' (l2

[Art. 4.5J



mechanisms for turning-on a thyristor are now discussed below :

(i) GATE Triggering : With anode positive with respect to cathode and with gate current
I, = 0, Eq. (4.6) shows that anode current, equal to the forward leakage current, is somewhat
more than I CBOl + [CB02' Under these conditions, the device is in the forward blocking state.
Now a sufficien t gate-drive current between gate and cathode of thyristor, or the transistor
of Fig. 4.15 (e) is applied. This gate-drive current is equal to base current 182 = I , and emitter
current III of tr ansistor Q2' With the establishment of emitter current l it of Q2' current gain
~ ofQ2 in creases and b ase current Is2 causes the existence of collector current
I C2 = ~2 182 = ~2 1" This amplified current I C2 serves as the base current 181 of transistor Ql '
With the flow of l Bl, collector current lCI = ~1 IBI = ~I ~2 I , of Q1 comes into existence. Currents
lSI and l CI lead to the est ablishm ent of emitter current I II ofQ 1 and this causes cu rrent gain
Ct.} to rise as desired. Now current I, + I CI '= (1 + ~t~2) I , acts as the base current of Q2 and
ther efor e its emitter current IJr. =ICt +1, ri se. With the rise in emitter curr ent 111. , ~ ofQ 2
increases and this further causes IC2 = ~2 (1 + ~l~) I, to rise, As amplified collector current I e'!
is equal to the base current of Qt, current gain a l eventually rises further. There is thus
established a regenerative action internal to the device. This regen erative or positive feedback
effect causes al + ~ to grow towards unity. As a consequence, anode current begins to grow
towards a larger va lu e limited only by load impedance exte rn al to th e dev ice. When
regeneration has grown sufficiently, gate current can he withdrawn. Even after II is removed,
regeneration continues . This characteristic of the thyristor makes it suitable for pulse
triggering. Note that thyristor is a latching device.
After thyristor is turned on, all the four layers are filled with carriers and all junctions are
forward bi ased. Under these conditions, thyristor has ver.y low impedance and is in the forward
on -state.
(ii) Forward-uoltage triggering : If the forward anode to cathode voltage is ' increased, the
collector to emitter voltages of both the transistors are also increased. As a result, the leakage
current at the middle junction J 2 of thyristor increases, which is also the collector cu rrent of
Q2 as well as QI ' With increase in coUector currents ICI and lc2 due to avalanche effect, the
emitter currents of the two transistors also increase causing a l + ~ to approach unity. This
leads to switching action of the device due to r egene rative action. The forward-voltage
triggering for turning-on a thyristor may be destructive and should therefore be avoided.
(iii ) d u/dt triggering: The reversed biased junction J 2 behaves like a capacitor because
of the space.charge present there. Let the capacitance of this junction be Cj . For any capacitor,

= C ~~. In


case it is assumed that entire forward voltage ull appears across rev erse biased

junction J 2 then charging current across the j unction is given by

i =C-

) dt
This ch argi ng or displ ac ement current across junction J'). 13 collector currents of
Q'l and Qt Cu rrents I C'l , I Ct will induce emi tter curren t in Q2' Qt. In case rate of ri se of anode
voltage is large, the emitter currents will be large and as a res ult, 0.1 + ~ will approach un ity
!e ~.d ing to eventu al switching action 01 th e thyri stor.
(iu) Te mperatltN! triggerin.g : At high temperature , the forward leakage current aCr0.55
j~lDction J 2 rises. This leakage current se rves as the collector junction current of th e compon ent





Power Electronics

transistors Q 1 an d Q2' Therefore, an increase in leakage current l eI' I~ leads to an increase in

th e emitter currents of QI' Q2' As a result, (exl +~) approaches unity. Consequentiy, switching
acti on of thyristor takes place.
( 1I ) Light triggering: When light is thrown on silicon, the electron-hole pairs incr ease. In
the forward-biased thyristor, leakage current across Jz increases which eventually incr eases
a, + ~ to unity as explained before and switching action of t.hyristor occurs.
As stated before , gate-triggering is the most common meth, d for turning-on a thyristor.
Light-trigger ed thyristors are used in HYDC applications.


Thyristor ratings indicate voltage, current, power and temperature limits within whi ch a
thyristor can be used without damage or malfunction. Ratings and specifications serve as a
link betwee n the designer and the user of SCR systems.
For reli able operation of a thyristor, it should be ensured that its current and voltage
ratings are not exceeded during its working. One of the major di sadvantages of thyristors is
that they have low thermal time constant. If a thyristor handles voltage, current and power
gr eater than its specified ratings, the junction temperature may rise above the safe limit and
as a resu lt, thyristor may get damaged. Therefore, when SCRs are selected, some safety margin
must be kept in the form of choosing device ratings somewhat higher than their normal working
values. The manufacturers of thyristors make a comprehensive list of the voltage, current,
power and temperature r atin gs after carefully testing the device. If SeRs are operated under
these specified conditions, no damage will be done to SCRs. The object of this section is to
discuss the various SCR ratings.
A thyristor has several ratings such as voltage, current, power, dul dt , dil dt, turn-on time,
turn-off time etc. For correct application of the device in thyristor circuits, a kn owledge of these
r ati ngs is desirable.
Some subscripts are associated with voltage and current ratings for convenietlce in
identifying the m. First subscript letter indicates the direction or the sta~e :
D -+fo rwa rd-b lockin g r egio n with gate circuit open; T --ton-state; R ~re verse
F ---tforward.
Except for the gate G, second subscript letter denotes the operating values.
W -}workin g valu e; R -+repetitive value ; S ---tsurge or non-repetitive value; T -+trigger
Third s ubscript letter I\;/ indicates the maximum or peak valu e.
Ratings with less than three subscripts may not follow these rules. Gate ratings involve
the s ubscript G. Subscript A usually stands for anode and subscript AV for average.
4.6.1. Anode Voltage Ratings
A thyri3tor is made up of fou r layers and three junctions as shown in Fig. 4.1 (b). Th e
middle junction J'2 blocks the forwa r.l voltage whereas the two end junctions J ,. JJ block the
reverse voltage. The anod e voltage r atings indicate the values of maximum voltages tha t a
thyristor can withs tand without a breakdown of the junction area with gate circu it open.
For ac systems, the supply voltage may n ot be a smooth sine wave. The voltage transients
may occur regularly or at random as shown in Fig. 4.17 (a). The diffe r ent anode voltage ratings
are as un de r :
(i) VDIr .\{ -Peck woding fo rward-blocking uoltage. It spec ifi es t he maximum
fonvard-blocking volt age that a thyristor can withstand during its working. Fig. 4.1 i (a ) shows
th at '-li,~'\1 i3 equal to the ma.:timum value of the sine voltage wave.

[Mt. 4.6]



(ii ) V DUM -Peak repetitiue forward-blocking uoltage. It refers to the peak transient voltage
that a thyristor can withstand repeatedly or periodically in its forward-blocking mode. The
rating is specified at a maximum allowable junction temperatUre with gate circuit open or with
a specifi ed biasing resistance between gate and cathode.

Voltage V DRM is encountered when a thyristor is com mutated or turned-off. It may be

recalled t h at during turn-off process, an abrupt change in reverse recovery curr ent is
accompanied by a spike voltage L


; this is responsible for the appearance of


acr oss

thyristor terminals.
(iii) V DSM -Peak surge (or non-repetitiue) forw ard-blocking voltage. It refers to the peak
value of the forvtard surge voltage that does not repeat. Its value is about 130% of V DR.I/, but
V DSM is less than forward breakover volta'ge V BO as shown in Fig. 4.17 (b).


(a )

(b )

Fig. 4.17. Anode voltage ratings during the blocking state of a thyristor.
(iv ) VRWM -Peak working reuerse voltage. It is the ma.'Ximum r everse voltage that a
thyristor can withstand repeatedly. Actually, it is equal to the peak negative value of a sine
voltage wave, Fig. 4.17 (a).
(v) V RR .\! - Peak repetitive reverse uoltage. It specifies the peak reverse transient voltage
that may occur repeatedly in the reverse direction at the allowable maximum junction
temp erature. The transient lasts for a fraction of the time of one cycle, Fig. 4.17 (a). The reason
for the periodic appearance of V RJUf is the same as for V DR.\!'
(vi) VRSM - Peak su rge (or non-repetitive) re ve rse voltage . It represe nts the peak value of
the reve rse surge voltage that does not repeat. Its value is about 130% of VRRJf . But VRS.\I is
less than r everse breakover voltagc-"BR as shown in Fig. 4.17 (b ).
Both VDS.\I and VRS.\f ratings can be increased by conn ecting a diode in series with a
thyristor. Th e anod e vo ltage ratings listed above from (0 to (iii) pertain to forw ard ' blocking
voltages whereas from (iu) to (vi) belong to reverse blocking voltages; a thyristor must be J.ble
to suppOrt these voltages safely with gate circuit open .
(vii) V T - On-state voltage d rop. It is the voltage drop between anode and cathode .....ith
specified forward on-state current an d junction tempe ratu re. Its value is of the ord ~r of 1 to
1.5 V.
(v iii ) Foru'ard du /d t rating. If rate of ri se of fo n~ard anode-to-cilthode vol tage !.:i high,
thyr : ~t o r mny turn on eve n when


Power Electronics

[A r t. 4,6J


there is no gate signal and

voltage is less than fonvard breakover voltage.
When a thyristor is in the forward blocking mode, the applied voltage appears across
junction J 2 as junctions J 1 and J 3 are forwa rd-biased. The reverse biased junction behaves like
a capacitor. When forward voltage is sudd~nl y appli ed to the device , a charging current
C;. du l dt begins to fl ow which may turn on SCR as explained in Art. 4.2. A high value of
du/dt , at which a thyristor just gets turned on is called critical rate of rise of anode voltage or
forward du / dt rating of the device. If applied du / dt exceeds this critical value, thyristor gets
turned on . For applied du / dt lower than forward du / dt r ating, thyristor remains in forward
blocking mode.
(0 )

(b) anode.to.cath~de

The forward du / dt ratin g depends on the junction temperature; hi gher the junction
temperature, lower the forward du / dt rating of the device . In pr actice, du /dt triggering is never
employed as it gives random turn-on of a thyristor. This type of triggering also leads to
destruction of the device through high junction temperature.
(ix) Voltage safety factor (V SF ) ' It is defined as the ratio of peak r epetitive reverse
voltage (VRRM ) to the maximum value of input voltage.

= Peakrepetitive r everse voltage (VRRM)


x rms value of input voltage

Voltage safety factor is usually taken between 2 to 3.

(x) Fin ger voltage. It is the minimum value of forward bias voltage between anode and
cathod e for turning- on t he device by gate triggering. The magnitude of finger voltage is
somewhat more than the normal on-state voltage drop in the thyristor.
4.6.2. Current Ratin gs
A thyristor is made up of semiconductor
material, its the rmal capacity is therefore
quite small. Even for short overcurrents, the
junction temperature may exceed the rated
value and the device may be damaged. As the
j u nction tempe r ature is dependent on the
c,-!rrent handled by a thyristor, a correct
cho ice of curren t r atings is essential for a
long workirrg life of the device. In this part of
,,".,/Temp.curve ,", " \ '.
the article , current r atings of SCRs are
discussed fo r both repetitive and c
non-repetitive type of current waveforms.
~!' ---Tj=125 C
B --f- -- - - -- :~"! ~--- - - 1-Average on-stote current (lTAV)' The o ,:-.-I

forward voltage drop across conducting SCR

is low, th ere fo r e power loss in a thyristor
depends pri mar ily on forward average
on- state cu r rent ITAv. For th e purpose of
f- T --...!
ang lll:
illust ra ting the si gnificance of average
~ 2T = 3aO e' ---I
( b)
on -s tate cur rent, consider a continuous dc
current 0.4. fl owing through the SCR, Fig. Fig. 4. 18. Variation of junction temperature with
4.18 (a ). Me: th e application of this current
constant anode current ic and wit h
a: t = 0, junction temperatu re begins to rise
rectnngulo.r wnve of io .

' .,

r- '--


[Art. 4.6]



until finally it reaches its rated value Tj = 125C. As the SCR has low thermal time constant,
final temperature of 125C is reached in a relatively short time. Suppose now that anode
current is of rectangular waveshape with conduction angle 180' (


x 360' ) as shown in Fig.

4.18 (b) . If the rectangular wave bas. average value equal to the constant current OA in Fig. 4.18
(a ), then current amplitude of rectangular wave in Fig. 4.18 (b) is DC = 2 times OA . As the SCR
has short time constant, junction temperature in Fig. 4.18 (b) is likely to exceed the allowable

temperature of 125C and this is not desirable. In order to limit the temperature to 125C for
rectangular wavefonn of anode current; there are two techniques, (i) provide better cooling to
the thyristor or (ii) reduce the pulse amplitude from OC.
As per the second method, pulse amplitude of anode current is r educed from OC to some
lower value on (say), so that junction temperature remains within limits, Fig, 4.18 (b). But a
reduction in the amplitude of rectangular wave would result in a lower value of average anode
current. This means that for the temperature rise to remain within limits, SCR must be rated
at a lower value of average forward current I TAv when it is conducting a pulsed anode current
than when it is carrying a constant dc. This shows that thyristor is derated when it handles
rec tangular or square wave of anode cumnt. The effect of conduction angle on anode current
ITAv is depicted in Fig. 4.19 (a) for rectangul ar waves. The avera.ge onstate power loss Pa ~ in
this figure is apprOximately given by

= (forward on-state voltage across a thyristor) x I TA V

It can be obtained more accurately fr om the relation



J(instantaneous voltage across SCR) (instantaneous current through SCR) dt

where T = periodic time of the anode current waveform.

The rms current for an SCR is constant whatever the conduction angle may be. But average
current is given by (Irrru I FF') where FF is the form factor of the current waveform. The
conduction angle for sine wave is defined in Fig. 4.19 (b). For the same conduction angle , the








ITAyil A _

(a )

li..,W in A _

Fi g. 4.19. Avera ge onstate power dissipa tion Pa ll as e. fiJn ction of IT~\ ..,for
(0 ) rectangular wa've and (b) hal iw:.w e s inuJoid .


Power Electronics

[Art. 4.6]

form factor for sine wave is higher than , for the rectangular wave (see Examples 4.8 and 4.9),
This means that average current for sine wave will be lower than it is for the r ectangular wave
for the same de (or rms ) current. The derating of the SCR is therefore more for sine waves than
for the square or r ectangular waves . The effect of conduction angle on average current is
depicted in Fig. 4.19 (b) for sine wave. For 180" conduction angle, the anode current in Fig. 4.19
( b ) is less than that in Fig. 4.19 (a ). This diagram is applicable for I-phase half-wave circuit (o r
I-phase one-way, one pulse circuit).
Curves shown in Fig. 4.19 are supplied by the manufacturers ofthyristors and are valid for
supply frequency range 0[50 to 400 Hz. The curve marked dc in Fig. 4.19 (0 ) is applicable when
anode current is continuous dc. The current for different conduction angles are terminated at
different valu es of average current in Fig. 4.19. For example, for 30" conduction angle, I TAV
terminates at [Idcl(form factor)] = (I dcI3.46 4) in Fig. 4.19 (0) for rectangular wave !lnd at
(ldc / 3 .979) in Fig. 4.19 (b) for sine wave. At these terminal points, maximum rms current
rati ngs of the device is reached . Table 4.1 gives different values of form-factor for different
conduction angles of the half-wave sine waveforms.

Table 4.1. Form Factor tor Sine Waves

Condu.ction angle








FCJrm factor








Curves of Fig. 4.19 are applicable when load is purely resistive. In case load i5 inductive in
nature, these curves should be modified. With an improvement in the waveform , i.e. with
waveform becoming more smooth, the form factor decreases and as a consequence, higher
average on-state current I TAy can be handled by the device.
RMS onstate current (IRMs)' By definition, for direct current, rms value I RMS or I rmJ =
average or de value, I dc ' Heating of the resistive elements of a thyristor, such as metallic joints,
leads and interfaces depends on the forward rms current l rnu. The rms current rating is used
as an upper limit for constant as well as pulsed anode current ratings of the thyristor. Its value
is equal to Ide of Fig. 4.19 (a). The value of the rms forward current for an SCR remains the
same for different conduction angles . Average current, however, is dependent on conduction
angle as shown in Fig. 4.19. For example, for 180" conduction angle, the form factor for half-sine
wave is n: / 2, therefore average current is 2 I de/ X or 2 I''M /x. This means that for 180"
conduction angle, thyristor circuit should be designed to carry an average current of 2 I dJrt
instead of I de (or Ir'M)' The derating of the SCR below the dc value depends upon the current
waveshape and it is defined as under :
SCR derating below de value

=ld' - :; =li

1- F~ )

...(4.3 )

where FF is the form factor of the waveform. Its value is always more than one.
For rectangular wave, FF is less as compared to its value for sine wave for the same
conduction angle. Eq. (4.3 ) reveals that SCR derating below dc is less for rectangular wave
than for the sine wave. The average current IT_4v fo r other conduction angles can be comput ed
as discussed above.
The significance of ITA'" and l'nl3 can be highlighted with an example. Suppose ma.'<..imum


r m3 curren t for a thyristor is 35 A. F or 120" conduction angle for sine wave, IT.4v = 1.875

Thyris tors

1M!. 4.6J


= 18.637 A.

This means that thyristor can h andle an average curr ent oi 18.637 A for 12 0 ~
conduction angle and its temperature will rem ain within limits. Suppose an ammeter is placed
in s eries with the SCR for measuring the average current. Now decrease the conduction angle
t o 30" but with average current as measured by the ammeter remaining unchanged at 18.637
A. But a n average current of 18.637 A at 30 conduction angle would require an rms current
oi I rml = 18.637 x 3.9812 = 74.1976 A. But such a large value ofrms current would caus e large
ohmic losses and is, therefore, certainly going to destroy the SCR. This shows that as
conduction angle is reduced, I TAV must be lowered acco rdingly so that rms current is not
exc eeded beyond its rated value and the SCR is not damaged .
The current ratings I TAV and I,rru are of repetitive type. They are dependent on maximum
junction temperature. If better cooling is provided to a thyristor body, th es e ratings can be
upgraded .
As stated above, power loss in a thyristor and its heating is dependent upon the rms
current. Manufacturers also provide curves showing the variation of case temperature Te with







ITAV in A _

I TAV in A ----"

(0 )

Fig. 4.20. Maximum allowable case temperature T ern as a function of frAY for
(a ) rectangular wave and ( b) for balf-wave sinusoid .

average on-state currentITAv Fig. 4.20. These curves can be obtained :from Fig. 4.19 provided
9.k (thermal resistance betw~en junction and thyristor case)" in C/ W is known. If Tj is the
junction temperature, then

TJ.- Te = 9"jt: . P0 11
For SCRs, Tj is usually 125C. Taking 9.k = 0.15CI W for dc current of200 A ; PfJ IJ = 300 W,
fr om Fig. 4.19 (a), is obtained for I TAy = 200 A.

125-Te =0.15 x 300 or Te= 80C

This poin t is pl otted in Fig. 4.2 0 (a ) as A. For 80 A dc, Pa u = 100 W.
125 - T, = 0.15 X 100 or T, = 1l 0' C
Thi s point is plotted as B in Fi g. 4. 20 (a ). For 180 con duction angle , for Ir..w = 140 A.


= 225 W fr om Fig. 4.19 (a ).


= 125 -

0.15 X 225

Far und !rs tandi ng the term th C!rmal rC!sistance read Ar t . 4.8.

=91. 25C

Power Electronics

(Art. 4.6]


This point is plotted as C in Fig. 4.20 (0 ), Other points can be plotted accordingly for
rectangular as well as half-wave sinusoids to obtain the curves of Fig. 4.20. These curves
indica te that for junction temperature Tj = 125"C, lower the average on-state current I TAv
greater is the case temperature that can be allowed for the same conduction angle. For
example, fOT sine wave with 180" conduction angle, for I TAv = 120 A the case temperature
Tem =91 DC ; for !TAV = 80 A the case temperature Tem = 104C and so on.
Surge Current Rating. When a thyristor is wolking under its repetitive voltage and
current ratings, its permissible junction temperature is never exceeded. However, a thyristor
may be subjected to abnormal operating conditions due to faults or short cirtuits. In order to
accommodate these unusual working conditions, surge current rating, I TSM (pea,k
non-repetitive on -state current), of thyristors is also specified. A surge current rating indicates
the maximum possible non-repetitive, or surge, current which the device can withstand. Higher
currents caused by non-repetitive faults or short circuits should occur once in a while during
the life span of a thyristor to prevent its degradation.
Surge currents are assumed to be sine waves with frequency of 50, or 60, Hz depending
upon the supply frequency. This rating is specified in terms of the number of surge cycles with
corresponding surge current peak. Surge current rating is inversely proportional to the
duration of the surge. It is usual to measure the surge duration in terms of the number of cycles
of normal power frequency of 50 or 60 Hz. For example, a three-cycle surge current rating for a
period of 60 msec (3 x 20 msec) for 50 Hz supply consists of three conducting half-cycles, each
followed by a n off-period. Thre e different surge current ratings are provided by the
manufacturers ; as for example, ITSM =3000 A for "2 cycle, ITSM =.2109 A for 3 cycles and

IrsM =1800 A for 5 cycles.

One cycle surge current rating is the peak value of allowable non-recurrent half-sine wave
of 10 ms~c duration for 50 Hz. For duration less than half-cycle i.e. 10 msec, a subcycle surge
current rating is also specified. This rating for 50 or 60 Hz supply is the peak value for a part
of the half-sine wave. The subcycle surge current rating I.b can be determined by equating the
energies involved in one cycle surge and one subcycle surge as follows :

l;b ' t =1' . T

l,b =I


... (4.8 a)


T = time for one half--<:ycle ofsupply frequency, sec

I =one-cycle surge current rating, A
l ,b = subcycle surge current rating, A
t = duration of subcycle surge, sec
f.:j r 50 Hz supply,
T = 10 msec
l,b = 10 . Tt

... (4.8 b)

I-t rating. This rating is employed in the choice of a fuse or other protective equipment for
thyristors. The r ating in terms of amp2-sec specifies the energy that the device can absorb for
a short time before the fault is cleared. It is usually specified for overloads lasting for less than ,
or equal to, one-half cycle of 50 or 60 Hz supply. The ft r ating is given by the relation .
(rms value of one-cycl e surge current )2 x time for one cycle



[Art. 4.6]


~ an example, Pt rating for 4 A(rms) SCRis 10 amp2.sec and for 35 ASCRis 100 amp2.sec.
In order that a fuse (or other protective equipment) protects a thyristor reliably, the Pt rating
of fuse must be less than the ft rating of the series-connected thyristor.
dil dt rating. This rating of a thyristor indicates the maximu~ rate of rise of current from
anode to cathode without any harm to the device. When a thyristor is turned on, conduction
starts at a place near the gate. This small area of conduction spreads to the whole area of
junction. If the rate of rise of anode current (dUdt) is large as compared to the spreading
velocity of carriers across the cathode junction, local hot spots will be fonned near the gate
connection on account of high current density. This causes the junction temperature to rise
above the safe limit and as a consequence, SCR may be damaged pennanently. Therefore, a
limit on the value of dUdt at turnon is specified in amperes per microsecond for all SCRs.
Typical values of dUdt are 20 to 500 A / J.l sec.
Other ratings . In addition to the voltage and current ratings ofthyristors discussed above,
there are some other ratings as under :
(a) Latching and holding currents,

(bl Turn-on an~ ;urn-off times, .

(c) Gate circuit voltage, current and power ra~gl':
These ratings have already been discussed in Art. 4.1 to Art. 4.5.
Detail ed ratings of any SCR during on-state and off state can be obtained from the
manufacturers by quoting the specification sheet number.
Example 4.9. The specification sheet for an SCR gives maximum nns on-state current as
35 A. If this SeR is used in a resistive circuit, compute average .on-stqte current rating faT
half-sine walle current for conduction angles of (a) 180 (b) 90 and (c) 30.
Solution. For half-sine wave of current as s'h own in Fig. 4.21 (a ), .


[Ar t. 4.6]

Power Electronics

- I rm $

FF -


35 x 2 _ 22 282 A

(b ) For 90 conduction angle, 8 1 = 90


Form factor

(e) For 30 conduction angle, 8 1 = 1500



= 2;

[1 + (- 0.866)]

= 0.02 13227 1m

[~: { 1"2 + ~ (- 0.866)}f2 =0.0849035100



Fig. 4.21. Pertaining to (a) Example 4 .9 and ( b) Example 4.10.

- 0.0849035 1m - 3 9818363
- 0.02 13227 1m - .

Form factor

I TAv =3.98 184


Example 4 .10. Repeat Example 4.9 in case the current has rectangular waveshape.

Solution. For the rectangular waveform of current shown in Fig. 4.21 (b),

~Lx 360

Conduction angle





I =I x T =l



a ('

_l-I 2 x T]U2 _ I

rms -



[Art. 4.6)

(a) F or

18 0~


conduction angle, n=-=2
Iau =2 and

I 2 .m
' 1=,2

Form factor

I TAV = T2 = 24.7487 A

(b ) For 90 conduction angle,

= 360
90 =4

I ou ="4 and

="2I ' 1=2

Form factor

I TA V =2"=17 .5A
(c) F or 30 can d uctlon
angI e,

n _- 360
12 -_ 12
I",u = 12


I '

I,.17U = ill

... .

12 =
Form factor = ill ' T = , 12


ITAv ~ill

= 10.1036 A.

Example 4.11. An 8CR has half-cycle surge current rating of 3000 A for 50 Hz supply.
Calculate its onecycle surge current rating and I 2t rating.
Solution. Let I and I,b be the one-cycle and sub-cycle su rge current ratings of the S9 R
r~spe ct ively. Then equating the energies involved in them, w e get

PT=I;b ' t
r x 100 = (3000) x 200




3~0 = 2121.32 A

From Eq. (4.9) ,




1 (--:J2
3000 ) x 100
1 = 45000 Amp 2 . sec.
x 2f=

Thus the SCR has onecycle surge curr ent r ating of 2121.32 A
and I 2 t rating of 45000 amp2.sec.
Example 4.12. In the circuit of Fig. 4.22, the thyristor is gated
with a pulse width of 40 microsec. The latching current of thyristor
is 36 rnA.. For a load of 60 nand 2 H, will the thyristor get turned
on ? Check. If the answer is negative, how this difficulty can be
overcome for the given load. Find the m axim um va lue of tht~
remedial parameter shown dotted.
S olution. The current through load and thyris tor is

. V' (l-e_!!.,)

LT ='R

For the circuit sh own,

-: ~



2H ~


,,, l,,,
,, ,'




L _ _ _ _.._____ -',

Fig. 4.22. P:rtaining to

Example 4.12.

. -_ 300(1
- e. 60
2 ' '' '''' ' ) -- o. 996 x 10 ' -- o. 996mA

Power Electronics

(Art. 4.61


This shows that for a pulse width of 40 ~s, the anode current rises to 5.996 rnA which is far
less than the latching current of 36 rnA. So thyristor will not get turned on.
The remedial parameter, shown dotted in Fig. 4.22, should be resistance, say R I because current
can rise in resistance without any time delay. The value ofRl can be obtained as under:
. 36 10 J 300 300 (1
=T+ 60 -e0.00'2,)


R, = 30.004 x 10 = 9998 n = 9.998 n.


Example 4.13. During forward conduction, a thyristor has static J- V characteristic as

shown by a straight line in Fig. 4.23. Find the average
power loss in the thyristor and its rms current rating for
the following load conditions:
(a) A constant current of 80 A for one-half cycle.

100 A

(b) A constant current of 30 A for one-third cycle.

(c) A half-sine wave vf peak value 80 A.
Solution. It is seen from Fig . .4 .23 that for any
current i a the voltage drop across thyristor is . - liT

- 0.8 . 08 0012'
= O.8 + 2.0100
x La = . +.

Constant current of80 A for one-half cycle is shown

ia =80 A, the voltage drop across Fig. 4.23. PertainiIlg to Example 4.13.
thyristor is liT = 0.8 + 0.012 x 80 = 1.76 V. From the
waveforms of ia vT shown in Fig. 4.24 (a), the average on-state power loss in thyristor is

m Fig. 4.24 (a ). For

Pau = T

VT' ta . dt

=~ J0


1.76 X80dt =1.76~:OX


Waveform of ia gives the rms current rating of thyristor as

-J 80~;

T = 56.577 A




















- ---6 ~8\T --


F ig. -\,.24. Curr ent a nd voltage wave forms pertai ning to Example 4.13.


(Art. 4.7)

Thy ristors
(b )




= 0.8 + 0.012 x 30 = 1.16 V

P = 1.16x30 x T = 115W
Rm s current rating

30 x


17.321 A

Halfsine wave of peak: value of 80 A, Fig. 4.24 (b), can be expressed as ta = 80 sin wt .
UT = 0.8 + 0.012 x 80 sin wt = 0.8 + 0.96 sin wt
From the waveforms for ia and UT shown in Fig. 4.24 (b ), the average onstate power loss
is given by


= 2~ J

= 21.

(0.8 + 0.96 sin wi) (80 sin WI) d (wl)

64 sin WI . d(wI) + 21.


76.8 sin' WI . d(wI)

76.81 w1 - siri2W1 1"

. 1
54 1-coswII' +-2n:
0 . 4n:

Rms current rating

=20.372 + 19.2 =39.572 W

= Imu = 80 = 40 A.



Reliable operation of a thyristor demands that its specified ratings are not exceeded. In
practice, a thyristor may be subjected to overvoltages or overcurrents. During SCR turn-on,
dUdl may be prohibitively large. There may be false triggering of SCR b.y high value of
duldt . A spurious signal across gate-cathode terminal s may lead to unwanted turn-on . A
thyristor must be .protected against all such abnormal conditions for satisfactory and reliable
operation of SCR circuit and the equipment. SCRs are very delicate devices, their protection
against abnormal operating conditions is, therefore, essential. The object of this section is to
discuss various techniques adopted for the protection of SeRs.
(a) dUdt protection. When a thyristor is fOIVIard biased and is turned on by a gate pulse,
conduction of anode current begins in the immediate neighbourhood of the gate-cathode
junction, Fig. 4.6 (a ), Thereafter, th e current spreads acro ss the whole area of j Ullc ti on. The
thyristo r design permits the spread of conduction to the whole junction area as rapidly as
possible. However, if the rate of rise of anode current, i.e. dil dt, is large as compared to the
spread velocity of carners, local hot spots will be formed near the gate connection on account
of high current dens ity. This localised heating may destroy the thyristor. The r ~re, the rate
of rise of anode current at the time of turn-on must be k ept below the specified limiting value.
The value of dUdt can be maintained below acceptable limit by using a small inductor, called
dUdt inductor, in series with the anode circuit. Typical dildt lim it values of SeRs are 20-500
A/'tl sec. The method of determining inductance of dildt inductor is illustrated in Example
Local spot heating can also be avoided by ensuring that the conduction spreads to the
whole are a as rapidly as possible. This can be achi eved by applying a gate curren t nearer to
(but never greater than ) the m ~:dmum specified gate current.
(b ) duldt protection. It has a!:-eady been discussed in .- \rt.-t 2 that ir'ra: : of rise of sudden ly
applied voltage across thyristor is hi gh, th e devic e may get turn ed on. S_lch phe no mena of


Power Electronics

[Art. 4.7J

turning-on a thyristor, called du l dt turn-on must be avoided as it leads to false operation of the '
thyristor circuit. For controllable operation of the thyristor, the rate ofrise offol"'Nard anode to I
cathode voltage dVa/ dt must he kept below the specified rated limit. Typical values ofduldt are'
20 - 500 V/ ~sec . False turn-on of a thyristor by large du / dt .;:an be prevented by using a
sn ubber circuit in parallel with the device,

4.7.1. Design of Snubber Circuits

A snubber circuit consists of a series combination of resistance R. and capacitance C, in
parallel with the thyristor as shown in Fig. 4.25 . Strictly
speaking, a capacitor C, in parallel with the device is
sufficient to prevent unwanted dul dt triggering of the
SCR When switch S is closed, a sudden voltage appears
across the circuit. Capacitor C, behaves like a short
circuit, therefore voltage across SCR is zero. With the
passage oftime, voltage across C, builds up at a slow rate
such that duldt across C, and therefore across SCR is les~
than the specified maximum du l dt rating of the device.
Here the question arises that if C, is enough to prevent
Fig. 4,25, Snubber circuit across SCR.
accidental turn-on of the device bydul dt, what is the need
of putting R, in series with C, ? The answer to this is as under.
Before SCR is fired by gate pulse, C. charges to full voltage V,. When the SCR is turned
on, capa ci tor d ischarges through the SCR and sends a current equal to
V, I (resistance ofl ocal path formed by C, and SCR). As this resistance is quite low, the
turn-on dil dt will tend to be excessive and as a result, SCR may be destroyed. In order to
limit the magnitude of discharge current, a resistance R, is inserted in series with C, as shown
in Fig, 4.25, Now when SCR is turned on, initial discharge current V,IR, is relatively small
and turn-on dUdt is reduced.
In actual practice; R" C, and the load circuit parameters should be such thatduldt across
C, during its charging is less than the specified du l dt rating of the SCR and discharge current
at the turn -on of SCR is within re~son~ble limits. Normally, R C, and load circuit parameters
form an underdrunped circuit so that duldt is limited to acceptable
The design of snubber circuit parameters is quite complex, Here only an approximate
method. of their calculation is presented in Example 4.14. In practice, designed snubber
parameters are adjusted up or down in the final assembled power circuit so as to obtain a
satisfactory performance of the power electronics system.
Example 4.14. Fig. 4,26 (a) shows a thyristor controlling the power in a load resistance
R L . The supply uoltage is 240 V dc and the specified limits for dUdt and duldt for the SCR
are 50 A I J.LSec 'and 300 V 1 )JS~c respectiuely. Determine the ualues of the dil dt inductance and
the snubber circuit p arameters R, and C,.
Solu tion. Snubber circuit parameters R, and C, are connected across SCR and dildt
inductor L in series with anode circuit as shown in Fig. 4.26 (b). When switch S is closed, the
rapacitor behaves like a short circuit and SCR in the forward blocking state offers a very high
r esistance. Th erefor e, the equivalent circuit soon after the instant of closing the switch S is
as shown in Fig. 4.26 (c). For this circuit, the voltage equation is

=(R, + RLJ i + L ~:

... (4.1 0


[Art. 4.7]



Cs ,1-'" cir(.lJit

r-- --- ---------- -.




- - -- - - - - - - - -





Fig. 4.26. (a) Thyristor in series with RL (b) Thyristor protection with L and R~> C!
(c) Equivalent circuit of Fig. 4.26 (b) at the instant switch S is closed.

Its solution gives,

In Eq. (4.10 a ), t is the time in seconds measured from th e instant of closing the switch.
From this equation,

=1 0e- II-r:: . .!.=


R, + RL-._tl -r:
R, +RL
L. '

V, _II~

=r e


The value of dil dt is maximum when t =

. ..(4.]0 b )



The voltage across SCR is given by.


= 240 x 10- ' = 4 8 H




. ~

Rs i

dv o _
dt - R, . dt

(d:;L R,.(~:L

... (4.]] )

From Eq. (4. 10 b) and (4.11),


R,' Y,



R = -L - , Y, dt

... (4.12)

= -48
'- x300= 6a


The circuit of Fig. 4.26, consisting of R, L, C, should be fully analysed to determine the
op timum values of sn ubber circuit par ameters R,> C,o The analysis of this circuit shows that
resistance R, can be ob tained from the relation {9J


Power Electronics

[Art. 4.71


where ~ is the damping factor (or damping ratio). In order to limit the peak voltage overshoot
across thyristor to a safe value, damping factor in the range of 0.5 to 1 is usually used. For
optimum solution of the problem, ~ is taken to be about 0.65.

c'= (~:JL=(2 X~.65J x 4.8 x 10-=0.2253~F

It is seen from Fig. 4.26 (b) that when switchS is closed, capacitor C, is charged to dc supply
voltage before the SCR is triggered.. Now when the SCR is turned on, capacitor C, will discharge
a maximum current of V,IR, and total current through thyristor will be (V,IR, + V.r I RLl . It
should be ensured that this current spike is less than the peak repetitive current rating
(1TRM) of the SCR. Thus if R. is small, the current spike contributed by the discharge of C, will
be large. In order to r educe this spike, R. is normally taken greater than what is required to
limit d uldt . At the same time, value of C. is also reduced so that energy stored in C, is small
and the snubber discharge does not harm SCR when it is turned on. Thus, in the present case,
R, may be chosen somewhat higher than SO, say I on and C, somewhat less than 0.2253 ).1F, say
0.15 ).1F. The adoption of the new value of R, demands a new value of L . From Eq. (4,12),


R, . V, = 103xO~40 = 8 ~H
(d val dt}rrwu

This value of inductance is more than that required to limit di l dt to 50 A/j.lsec.

For ac circuits, maximum value of input voltage (V m ) can be used in place of V, in Eq. (4.12)
for computing R,.
Example 4.15. A thyristor operating from a peak supply voltage of 400 V has the following
specifications :
Repetitiue peak current. I, = 200 A,

(dildt)~ = 50 A/~. (~~ l~ = 200 V/~.

Choosing a factor of safety of 2 for 1p '


(~: Land (~~


design a suitable snubber circuit.

m inimum value of load resistance is 10 n.

Solution. For a factor of safety of 2, the permitted values are I p =-=100A.

50 = 20- AI~, (duldt)m~ = 2
200 = 100 V I IlS

In order t o r estrict the rate of rise of current beyond specified valu e, (di/dt) inductor must
be i:Jerted in series with thyristor. From Example 4.14,

L = (dil dt)=u =

400 x 10-'
= 16 ~H

R =L(dul =1 6 X 10-' x 100 =40

, VJ dt u
10- 6

Before thyristor is turned on, C, is charged to 400 V. 'When thyristor is turned on, the peak
current through the thyristor is
400 400 = 140 ~
10 + 4


[Art. 4.7]



As this peak current through SCR is more than the pennissible peak. current of 100 A, the
magnitude of R, must be increased. Taking R, as 8 n, the peak current through the SCR
= 400
10 + 8" = 90 A, less than the allowable peak current. So choose R, = 8 o.

C.=(,J'L=(\j3)' x 16 x lO- =0.4225~


The value of C, may be lowered as discussed in the previous example, so C, may be taken
as 0.30 ~F.
At the instant switch S is closed, Fig. 4.26, thyristor is open circuited and current through
C, is given by
V, C, dt == R, +RL



. x


10- 6 du = 400
dt 10 + 8
du 400
-d =-18 x

6 = 74.07 V/ ~s
0.3 x 10-

Since designed value of(du / dt) is less than the specified maximum value oflOD
of C, chosen is correct. So choose L = 10).lH, R, =8 n and C, = 0.31J,F.

V/ ~s ,


Example 4.16. Thy ristor shown in Fig.

4.27 has ilt rating of20A 2s. 1ftenninal A gets
short-ci rcuited to ground, calculate the fault
clearance time so that SeR is not damaged.
f\) 230 /2 sin 3141
Solution. The worst possible fault current
should he considered for calculating the fault
clearance time. Maximum fault current occurs
when source voltage is at its peak = 230..J2V.
Fig. 4.27. Pertaining to Example 4.16.
When terminal A gets short-circuited to ground, the resistance offered to s our ce
x l . .
230-12 x 11
=1 + 10
10 + 1 =21/ 11 O. Assummg m8Xlmum fault current =
A to remam constant
during the short clearance time


we get

f;',' f;' (230~



x 11)' . dt =20 A' s

t, =20[ 230N x


x 1000 ms = 0.6892 ms.

4.7.2. Overvoltage Protection

Thyristors are very sensitive to overvoltages j ust as other semi-conductor devices are.
Overvoltage transients are perhaps the main cause of thyristor failure. Transient over voltages
caus e either maloperation of the circuit by u nwante d turn-on of a thyristor or permanent
damage to the device du e to reverse breakdown. A thyristor m ay be subjected to inter nal or
extern al overv oltages ; the former is caused by the thyris tor oper ation whereas the latter comes
fro m the supply lines or the load circuit.
() lnte rna! oue ruoltag es. Large volt ages may be gen er a ted intern ally dur in g the
commutation of a thyris tor. After thyristor an ode curren t reduces to zero, anode current




Power Eleclronics

rever.ses due to stor ed charges. This reverse recovery curren t rises to a peak value at which
time t he SCR begins to block. After this peak, reverse recovery current decays abruptly with
large dildt . Bec au se of the s eries inductance L of th e SCR circuit, large transient voltage


is produced. As this internal overvoltage may be several tim es the breakover voltage of

the devi ce, the thyristor may be destroyed permanently.

(ii) External oueruoltages. External overvoltages are caused due to the interruption of
current fl ow in an inductive circuit and also due to lightning strokes an the lines feeding the
thyristor systems. Wben.a thyristor converteT is fed through a transformer, voltage transients
ar e likely to occur when the transformer primary is energised or de-energised. Such
overvoltages may cause random turn on of a thyristor. As a result, the overvoltages may appear
across the load causing the flow of large fault currents. Overvoltages may also damage the
thyristor by an inverse breakdown. F or reliable oper ation, the overvoltages must be suppressed
by adopting suitable techniques . Suppression of overvoltages. In order to keep the protective components to a
minimum, thyristors are chosen with their peak voltage ratings of 2.5 to 3 times their normal
peak working voltage. The effect of overvoltages is usually minimised by using RC circuits and
non-linear resistors called uoltage clamping devices.

The RC circuit, called snubber circuit, is connected across the device to be protected, see
Fig. 4.30. It provid es a local path for internal overvoltages caused by reverse recovery current.
Snubber circuit is also helpful in damping overvoltage transient spikes and for limiting
d u/dt across the thyristor. The capacitor charges at a slow rate and thus the rate of rise of
forward voltage (d u/dt) across SCR is also reduced . The resistance R, damps out the ringing
oscill ations between th e snubber circui t and the stray circuit inductance. Snubber circuits are
also connected acr os s trans fonner secondary terminals to suppress overvoltage transients
caused by switch ing on or switching off of the primary winding. As snubber circuits provide
only parti al protection to SCR against transient overvoltages, thyristor protection against such
over-lOltages must be upgraded. This is done with the help of voltagec1amping devices .






'E Rrsistonet.,







\ ,.


Vo llage




Prok loull
without any
,."'0 ",





o L~let
:inqL !,rein~
I IIml , t,;J"'" tme, t o
~C [ or l nc:;J




an.! yolt-r:!3istance char acteri s tics of voltage-damping de'lice

Fig. L2J. (e ) Vol ~ ~t:l~e re

(b ) Action of currant limiting fu se i n en !I.e circuit.


[A ct. 4.7]


A voltage-clamping (V. C.) device is a non-line ar resistor connected across SCR as shown in
Fig. 4.30. The v.e. device has falling resistance characteristic with incr easing voltage, F ig. 4.28
(a). Under normal working conditions of voltage below the clamping level, the device has a high
resistance and draws only a small leakage current. When a voltage surge appears , the v.c.
device operates in the low resistance region and produces a virtual short circuit across the S CR.
The incr eased current associated with virtual short circuit produces an increased voltage drop
in the source a nd lin e impedances and as a r esult, voltage across SCR is clamped to a safe value.
After the su rge en ergy is dissipated in the non-linear resistor, the operation of the V.C. device
returns t o its high resistance region . Selenium thyrector diodes, metal oxid e varistors or
avalan ch e diode s uppre ssors are commonly employed for protecting the thyristor circuit
against overvoltages. As the voltage clamping ability of a thyrector is inferior to those of metal
oxide varistor and avalanche-diode suppressor, use of thyrector is on the decline.
It has already been stated that RC snubber is not enough for overvoltage protection ofSCR.
In practic, therefore , a combined protection co nsisting of RC snubber and V.C . device is
provided to thyristors as shown in Fig. 4.30.

4.7.3. Overcurrent Protection

Thyristors h ave small thermal time constants. Therefore, if a thyristor is subjected to
overcurrent due to faults, short circuits or surge currents ; it~ junctt'm}\emperature may exceed
the rated value and the device may be damaged. There is thus a need for the over current
protection of SCRs. As in other electrical systems, overcurrent protection in thyristor circuits is
achieved through the use of circuit breakers and fast- acting fuses as shown in Fig. 4.30.
The type of protection used against overcurrent depends upon whether the supply system
is wea.1( or stiff. In a weak supply network, fa ult current is limited by the source impedance
below the multi-cycle surge current rating of the thyristor. In machine tool and excavator
drives, if th e motor stalls due to ov erloads, the curren t is limited by the source and motor
impedances. The filter inductance commonly empl oyed in dc and ac drives may limit the rate
of rise of fault curr ent below the multi cycle su rge current rating of the thyristor. For all such
systems, overcurrent can be interrupted by conventional fuses and circuit breakers. However,
proper co-ordination is essential to guarantee that (i) fau lt current is interrupted befor e the
thyristor is damaged and (ii) only faulty branches of the n etwork are isolated.
Conventional protective methods are, however, inadequate in electrical stiff supply
networks. In such systems, magnitude and rate ofri se of current is not limited because source '
has negligible impedanc e. As such, fault current and therefore junction tempera ture rise wit hin
a few milliseconds. Special fast-acting current-limiting fuses are, therefore , required for the
protection of thyristors in these stiff supply ne tworks .
The operation of fa5 t-acting current-limiting fuse is"illustr ated in Fig. 4.28 (b ). These fuses
and thyristors are found to have similar th ermal properti es, their co-ordination is therefor e
sim pler. Th e current-limiting fuse consists of one or more fine silver ribbons having very short
fusing tim e. In Fig. 4.28 (b ), fadt is shown to occur at zero crossin g of the ac sine wave, i.f .
at t = O. 'W ithout fuse, the fault current would rise upto A and then would follow dotted curve,
reach pea.1( valu e D and then decrease as shown. A prope rly selected current limiting fuse
melts at A. An arc is then struck. F or a brief interval after A, the cu rTent continues to ris e
depending upon th e circuit parameters nnd the fus e design . This current reaches a pea.l.;, value ,
called peck Ie : :n-ough current, which i3 indicated by point B in Fig. 4.28 (0). _ ote that pe ak
let through current is considerably less than th e pe":lk fault current without the fuse , the latter
is indicated by point D . ..\fter the point B, arc resisnnce L'1Cre a3es and faul t current decr eas es .
At point C, arcin g s tops and the fault current is clear ed. Tp e t otal clearing ti me t~ i:i the su m
of melting ti me t", and a rc ing time t;:, i .~. t: = t", - t:: .


[Art. 4.7]

P ower Electronics

Proper coordination between fast-acting current-limiting fuse and thyristor is essential. A

fuse carries the thyristor current as both are placed in series. Therefore, the fuse must be rated
to carry full-load current plus a marginal overload current for an indefinite period. But the peak
let through current of fuse must be less than the subcycle surge current rating of the SeR. The
voltage acr oss the fuse during arcing period is known as arcing, or recovery, voltage. This
voltage is equal to the sum of source voltage and the emf induced in the circuit inductance
during arcing time tao If the fus e current is interrupted abruptly, induced e.m.f. L


may be

high; as a result arcing voltage would be excessive. It should therefore be ensured during fuse
design a~d coordination that arcing voltage is limited to less than twice the peak supply
voltage. In case voltage rating of the fuse is far in excess of circuit voltage, an abrupt current
interruption would lead to dangerous overvoltages.
When both circuit break er and fastacting current-limiting fuse are used for overcurrent
protection of SCR, Fig. 4.30, the faulty circuit must be cleared before any damage is don e to the
device. A circuit breaker has long tripping time, it is therefore generally used for protecting the
semiconductor devi ce against the continuous overloads or against sur ge currents of long
duration. A fastacting C.L. fus e is used for protecting thyristor s against large surge currents
of very short duration. The tripping time of the circuit breaker, the fu singtime of the
fast-acting fuse must be pr o ~e rly coordinated with the rating of a thyristor. In order that fuse
protects the thyristor reliab ly, the [2t rating of the fuse must be less than that"Of the SCR.
Electronic crowbar protection. As thyristor possesses high surge current capability, it
can be used in an electronic crowbar circuit for overcurrent protection of power converters using
SCRs. An electronic crowbar protection provides rapid isolation of the power converter before
any damage occurs.
Main fusr

Fig. 4.29 illustra tes the basic principle

of electronic crowbar protection. A crowbar


thyristor is connected across the input de

t ..

jb- Got!Ci rTri~f!r

terminals. A current sensing resist or
detects the value of converter current. If it r- 0
exceeds preset value, gate circuit provides
the signal to crowbar SCR and turns it on in
rrglSt rr
a few microseconds . The input terminals
Fig. 4.29. Elementary electronic crowbar circuit.
are then short-circuited by crowbar SCR


Over cur rem

,.. . ______
_. .. . ___ \1 . .,


C. 8 .



Snubber cirC'Jit ,
~-.- .. :\.. _...,

~ inductor



L __._ ____ __ .i

Gol e Protection

,.----------- -- .---'> .- - .--,

H 5


L~ :



... . _. _ _ _ _ __ __ - - - _ _ _ - - - ...I

~ - .-

-_. --- -

Fig. 4.30. Circuit component:.; showing the thyristor protection.

C.B.-Circuit breaku ; F.A. C.L.F.-Fa3t acting C'.l."'Tent limiting ftlse ; H.S.-Heat sin.\. ; ZD-Zener diode.


(Art. 4.7J


and it shunts away the converter overcurrent. The crowbar thyristor current depends upon the
source voltage and its impedance. After some time, main fuse interrupts the fault current. The
fus e may be replaced by a circuit breaker if SCR has adequate surge current rating.
4.7.4. Gate Protection
Gate circuit should also be protected against overvoltages and overcurrents. Overvoltages
across th e gate circuit can cause false triggering of the SCR. Overcurrent may raise junction
temperature beyond specified limit leading to its damage. Protection against over-voltages is
achieved by connecting a zener diode ZD across the gate circuit. A resistor R2 connected in
series with the gate circuit provides protection against overcurrents.
A comm on problem in thyristor circuits is that they suffer from spurious, or noise, fi ring.
Turning-on or turning-off of an SCR may induce trigger pulses in a nearby SCR. Sometimes
transients in a power circuit may also cause unwanted signal to appear across the gate of a
neighbouring SCR. These undesirable trigger pulses may turn on the SCR leading to false
operation of the main SCR. Gate protection against such spurious firing is obtained by using
shielderl cables or twisted ~ate leads. A varying flux caused by nearby transients cannot pass
through twisted gate leads or shielded cables. As such n o e.m.f. is induced in these cables and
spurious firing of thyristors is thus minimised. A capacitor and a resistor are also connected
across gate to cathode to bypass the noise signals, Fig. 4.30. The capacitor should be less than
0.1 ~F and must not deteriorate the waveshape of the gate pulse.
Example 4.17. For the circuit shown in Fig. 4.31,
10 A o.15 )JF
(a) calculate the maximum ualues of dil dt and
d ol dt for the SCR,
(b) find the rms and atJerage current ratings of
the SCR for firing angle delays of90"and 150" and


/2. 230 s in 314\

(c) suggest a suitable voltage rating of tit. SCR.

Solution. (a) From Eq. (4.10),

(~;L =[~)


Fig. 4.31. Pertaining to Example 4.17 .

= '1'2 230 = 21.685 A/ ee.

15 x 10- 8
From Eq. (4. 11),

(~~L =R (~;L = lOx 21.685=21~;85V/~ec

(b ) For 1 5~, X L = 314 x 15 x 10- 8 = 0.00471

n. As

this value of X L is much lower than

= 2 n, the current is primarily limited by 20.


= '1'2 . 230 = 115 . '1'2

For firing angle delays of 90 and 150:, the conduction angles are 90 and 30C respectively
and from Example 4.9, the respective values of iorm factors are rt/{2 and 3.98184.
'1'2 . 115 '1'2
:. For firing angle delay of 90, iT,w =
=73.211 A

an d fo r firin g angle del ay of 150,

'1'2 . 115
ITAv= 3.98184 =40.8l!4 A


[Art. 4.7]

Power Electronics

RMS current rating of the thyristor is 115...[2 = 162.634 A for any conduction angle, but
average currents aTe 73.211 A for conduction angle of 90 and 40.844 A for conduction angle
of 30.
(e) Voltage rating of the SCR = (2.5 to 3) times the peak working voltage
= (2.5 to 3) x 12230 =813.173 V to 975.807 V.
So a voltage rating of about 900 V may be chosen for the SeR.
Example 4.18. For the circuit shown in Fig. 4.32 (a), the initial voltage across capacitor
is Uc (0) = - 100 V. Sketch the time variations of i, ULI Vel iD and iL after the thyristor is turned
Solution. When the thyristor is turned on at t =0, the voltage equation for the circuit is
1 f 'd
L di
dt + C t t


Its Laplace transform is



1 [1(5) C ",(0)] V,
=-;1(5) + C +
1(. ) =

V, L

"N .



=T '






.,~ L....".L---'~_-'-_+-~


iLk:: .


Fig. 4.32. Pertaining to Example 4.18.

300 .
Its Laplace inverse is ,'(t) = "',;sm "'ol


"'0 = ~LC


=L dt =300 cos "'ol

Vc = V, - UL = 200 - 300 cos toot


The current and voltage waveforms are as shown in Fig. 4.32 (b). At 1i: / 2, uL tends to reverse
and as a result, diode D gets forward biased and current i L starts flowing through D as iD ' uL
is therefore zer o from 1t/2 to 'It. Voltage u~ remain,:; 200 V and current i zero from 1t/2 to 7t as
shown in Fig. 4.32 (b ).


[Art. 4.8J



It has be en explained in Art. 4.3.1 that rate of growth of anode current during rise time
of t or: is hi gh, but cathode-conduction area is small. High rise of anode current in a thyristor,
associated with high anode voltage, causes more losses. These high losses, occurring over a
sma ll cathode-conduction area during rise time, may result in hot spots leading to the
destru ction of the device. A high value of du / dt may turn on the thyristor at an unwanted
instant which is undesirable. This all prompts us for an improvement in di/dt as well as
d u/dt ratings of thyristors. A boost in these ratings can be made by doing some structural
modificati ons in thyristo.rs; this is explained below.
4,8.1. Improvemen t in dildt Rating
The rate of rise of anode current (di/dt) in a thyristor depends primarily on the initial
area of cathode conduction during rise time. This implies that if initial cathode conduction
area is increased, the di/dt rating also gets improved. There are two methods of doing this,
(i) by us ing a higher-gate current (ii) by intermixing the
gate-cathode regions.


. " Higher-gate current. At the start of turn on, if
higher-gate current is applied, turned-on area of cathode surface
.' . .........
has to be more for handling this higher-gate current. As a consequence, initial cathode-conduction area for allowing anode curr e n~ to pass through it, increases, and this is what is desired. The
widely used gate current profile is shown in Fig. 4.7. However, FJom gc!e..eny ~ elfe!.: .!
big.her gate current should not be obtained from a gate drive
circuit . The usual way of accomplishing this goal of higher-gate Fig. 4.33. l'o'fain thyristor turn
current is through the use of a pilot thyristor shown in Fig. 4.33.
2n by pilot thyristor.
When pilot thyristor is turned on, a high value of gate current
flows from anode A, pilot thyristor and gate-cathode terminals of main thyristor for switching
. '
4 .8.1.2. Structural m odification of the device . As stated above, the di / dt rating of a
thyristor can be improved by having more cathode-conduction area during delay and rise time
of totl ' This can be achieved by higher.gate current (already discussed ) and by modifying the
gate-cathode geometry. This alteration consists of iritermixing, or interdigitating gate and
cathode regions . The effect of this structural change can be r.ealized by examining the initial
conduction process firs t (i) in side-gate thyristor and then (ii) in centre-gate thyristor.


A side-gate thyristor is shown in Fig. 4.34 (a). When gate current is applied , the
gatecu rrent density is higher near the gate terminal. As a r esult, cathode-conduction area is
small during delay and rise time . This :=.llows that initial conduction occurs over a nar row
channel n ear the gate terminal as shown in Fig. 4.34 (a). Refer ence to Fi g. 4.6 and its rele'l;mt
writeup is also helpfuL
A cen tre-gate thyristor is shown in Fig. 4.34 (b ). \-Vhen positive gate pulse is appli ed, the
gate current flows fr om gate to cathode in all possible direc tions covering a ringshaped area
on the cathod e surface as indic a.ted in Fig. 4.34 (b). Exam ination of Fig. 4.34 (0 ) and (b ) shows
that initial ar ea of cathode conduction is very large in centre-gate thyristor as compared to
that in sid egate thyristor. Thi3 illustrates that initial area of cath ode conduction can be
enhanced significantly by intermixing th: gate-cathode reg:ons approp riately.


Pow er Elec tronics

[Art. 4.8]


f----tt-----j J,



r - - t t ----1 J,

~;;:=\I J'
f--+--+---j J,


Initiol cathode



( bJ

Fig. 4.34 . Initial cathode-conduction area in (a) side-gate thyristor and

(b) centre-gate thyristor.



Another configuration indicating the intermixing, or

interdigitating of the gate-cathode regions is illustrated in
Fig. 4.35.


4.8.2. -Improvement in dv / dt rating

It has already been discussed in Art. 4.2 (du/d/

triggering) that when du/dt is large, high charging currents

flow through the reversed biased junction J 2 which may turn ,,""".,,-L--.,-'-..-'......-'-..-'..-'
on the thyristor: The effect of capacitor-charging current, or
du /d t, can be minimised by using cathode-short structure
show n in Fig. 4.36 (a). Cathode-shorts are realized by
overlapping metal on cathode n+ regions with a narrow
Fig. 4.35. Interdigitating of
p-region in between. Fig. 4.36 (a) shows metallization M and gate-cathode regions in a thyristor.
N which form the cathode.
Most of discho
current dvldt















~ ICcthode short ..l..l

(b )

Fig. 4.36. Thyristor cathoda-3horts (a) 2!ementary intermi:ring


advanc::!d intermixin g.


(Art. 4.9J


In normal structure, discharge current d uldt (acting as gate curren t) flows through J 3
junction, Fig. 4.3 (a), and leads to spurious tum of of SCR. In cathode-short structure, most
of the discharge current (or displacement current) passes through narrow p channels in
between cathode n regions as shown in Fig. 4.36 (a ). Junction JJ shares only a negligible
amount of dvl dt current. A little discharge current flowing through J 3 junction (and acting as
gate current) is too small to turn on the device. Thus, higher valuos of dul dt are now
pennissible with cathode-short structure.
The thermally generated leakage current across junction J 2 also does not pass through
gate-cathode junction J 3 . Therefore, current injection across gate-cathode, or J 3 , junction is
drastically reduced, hence the total discharge current, or dul dt , can be larger without'
turning-on the device. It can, therefore, be inferred that cathode-short structure improves
du / dt rating of the thyristor.



Some power loss occurs in a thyristor during its working. The various components of this
power loss in the junction region of a thyristor are as under :
(i) Forward conduction loss
(ii) Loss due to leakage current during forward an~ reverse blocking
(iii) Switching losses at turn-on and turn-off
(iu) Gate triggering loss
At industrial power frequencies between zero and 400 Hz, the forward conduction loss, or
on-state conduction loss, is usually the major component. But switching losses become
dominant at high operating frequencies. These electrical loss es produce thermal heat which
must be removed from the junction region. The thermal losses and hence the temperature rise
of the device increase with the thyristor rating. The cooling of thyristors, therefore, becomes

more difficult as the SCR rating increases.

The heat produced in a thyristor by electrical loss is dissipated to ambient fluid (air or
water) by mounting the device on a heat sink. When heat due to losses is equal to that
dissipated by the heat sink, steady junction temperature is reached. Thyristor heating and
hence its junction temperature rise is dependent primarily on current handled by the device
during its working. As Buch, current rating of thyristors is often based on thermal
4.9.1. Thermal Resistance
Thermal energy, or heat, flows from a region of higher temperature to a region of lower
temperature. This is similar to the flow of current from higher to lower potential in an electric
circuit. There is thus an analogy between thermal-power flow and current flow as given in
tabular form below :
Eledrnal quantit~s

The rmal quantilus


Heat, J or Ws



Temperature difference, "C


Tbennal power, or rate of heat transfer, W

Thermal resistance, "CIW


Charge, C or As
Potential diffeT~Dce , V
Current. or rate of charge transfer, A
Electrical ~sis tancc, VIA or ohms.

It is seen fr om above that thennal resis tan ce, an alogou3 to electrical resis tanc.e, is the
resistance offer~d to thennal. power flow. Thermal r esis tance is denoted by 8. If p o'~er loss,


Powe r Electron ics

[A r t. 4.91

Pall in watts, causes the temperature of two points to be at T l C and T 2 C wher e Tl > T2 then
t hermal resistance is given by
9 12 = Tl - T , ' CIW

... (4. 13 )


The heat gener a ted in a thyristor due to internal loss es is t aken to be developed at a
j unction within the semiconductor material A simple arrangement of thyristor, its case and
h ea t si nk is shown in Fig. 4.37 (a) . Various temperatures and thermal r esistanc e are also
indicated in this figure. The heat flow from thyristor junction to ambient fluid is as under:

from th e junction to thyristor case, thermal resistance 9jc

(ii ) from t he thyristor case to h eat sink, thermal resistance 9" an d
(iii) fr om the h eat sink to the surrounding ambient fluid (air or water), thermal resistance 9SA '

a jc

9 0s



Fig. 4.37. (0 ) Thyristor with its case and heat sink

(b ) thermnl equivalent circuit for a thyristor.

There is thu s therm al resistance 9jc between junction temperature T, and case temperature
Te' Simil arly, ther e is th erm al r esistance eel between Tc and sink te mp erature T! and 901A
between T, and ambient temperature Tk Using the electrical analogy, a thermal equivalent
circuit depicting the flo'.'V or h eat from junction to ambi ent fluid can be drawn as shown in Fig,
4.37 (b). Her e P au is the average r ate of heat generated at a thyristor junction and is analogous
to constant-cu rrent source. Here

= Tj -Tc ~ Tc -T, =T,- TA


_ TJ.-TA

... (4.14)

wh er e
e', A= 9JC + BCol + B_..
is the total thermal r es istance between junction and ambient.
;on .

The junction-to-case thermal resistan.:e 9jc is specified in th e thyristor data sheet. The
cas:-to-sink therma l r esistance BI'S depends on the size of the device case, flatness of the case
surface, the clam ping pressure and the use of conducting greas e beh. .een the in terfaces. Usual
value orge, vari es between O.0 5 ~C /W and O. 5D C/W. In addition to Bjc , thyristor data sheet also

Thyris tors

[Art. 4.91


specify 9u assuming correct installation procedure and use of the interface thermal lubrication.
The sinkto-am bient thermal resistance 8sA. is independent of the thyristor configuration. The
parameters on which e. . . depends are heat sink material, surface area and fmish of the heat
sink, volume occupied by heat sink and the type of cooling (air cooling or water cooling). For
naturally cooled heat sink, aaA may be equal to O.5CIW and this valu e would be lower for better
cooled heat sinks.
The difference in temperature between junction and ambient can be written from Fig. 4.37
Tr T.=P.,(9.. +9,,+ 9..)


Eq. (4.15) shows that for maximum valu e ofTj (= 125C), POll can be increased by reducing
BaA. This means that by providing efficient cooling system to the SeR, the power dissipation
capability of the device can be increased.
4.9.2. Heat Sink Specifications
The thyristor data sheet specifies maximum junction temperature Tj t thennal resistances
Bjc arid eell. The manufacturers of heat sinks provide catalogs in which sufficient data on heat
sink is available. Fig. 4.38 gives typical data in the forI\l. of curves for standard heat sinks of
aluminium extrusions. These curves relate temperature difference (T lI - T A) in C between heat
sink and ambient uersus average power dissipation POll in watts.


Cur vt



a ..... . . ... 3.2 xIOx7.5

b _._. __ .... 3. 2xl0xI2 . 5
~, ------ .. 10xIOxlO


1------ .. 15.5xI5. 5_22.5



Pay in watts ___

Fig. 4.38. Standard heat sink ratings of aluminium extrusions.

In order to illustrate the use of these curves, choose a particular h eat sink and read POll
and (T, - TA ). Then the thermal resistance of th e h eat sink to ambient is' calculated as
T lI - TA


( ~l m

For maximum specified temperature Tj (usually 125C) and a known ambient

temperature TA , the permi3sible value of POll (with e)C and eu already known) is calculated
fro m Eq. (4.15) with e.sA computed from Eq. (4.16). Ii this Pov is different from tha t chosen
earher from Fig. 4.38, another heat sink with other values of Pall and (TJ - T,,) is tried until
Eqs. (4.15) and (4. 16) are satisfied. After deciding the value of POll' use this value oi Pall in Fig.
4.19 to obtain permissibl e value of average current rating for a given conduction angle and
current waveform.


[Art. 4.9)

Power Elec tronics

In the seconu m ethod of heat sink design, first average armature current is determined
from a known current waveform and conduction angle. Corresponding to this average current,
P cz,; is read off from Fig. 4.19. For this P fW thermal resistance aaA is determined from Eq. (4.15)
as temperatures Tj (= 125C), TA and 9ft. eu are already known . This computed value of 8sA is
used to obtain temperature difference (T.. - T A ) between heat sink and the ambient from Eq.
(4.16). Using these values of Pau and (T, - TA ), an appropriate heat sink is selected from Fig.
4.38, the details of which are usually supplied by the manufacturers.
In the third method of h eat sink selection, first compute average armature current as done
in the second method. For this value of average cUrrent, obtain Pou from Fig. 4.19 and case
temperature Tc from Fig. 4.20 for the known current waveform and conduction angle. An

examination of Fig. 4.38 r eveals that the sink temperature T, in terms of case temperature
T( is given by
...(4 .17)

As ambient temperature is known. (T, - T),) can be calculated. Now, with the knowledge

of Fau and (T. - T A ) . a cl:.oice of suitable heat sink can be made from Fig. 4.38.
Heat sinks are made from metal with high thermal conductivity. Aluminium is the most
commonly used metal. Copper, being a costly metal. is seldom used as a heat sink. material.
Heat dissipation from heat sink takes place primarily by convection. As such, thyristor cooling
by convection can be made more effective by enlarging the cooling surface area by providing
the heat sink with peripheral fins. Heat dissipation also takes place by radiation. Heat sinks
are usually provided with black anodized. finish to enhance the heat dissipation by radiation.
Sometimes the size of naturally-cooled finned heat sink may become large. In such a case,
size of the heat sink can be reduced by using forced air cooling which involves a fan ~l owin g
air over the fins . With forced air cooling, heat-removing capability of the finned heat sink
incr eases by a fact or of two t o three. For dissipating large loss es 10 high-power thyristors ,
water-cooling is usually employedto get a compact size of the heat sink.
4.9.3. Thyristor Mounting Techniques
Internal pow er losses in a thyristor cause high thermal stresses which further give rise to
mechanical for ces. A thyristor must be braced to withstand such mechanical forces. In addition,
SCR mountin g must be so designed as to facilitate heat flow from junction to the case.
Depending upon the low or high power ratings of thyristors, there are five major mounting
techniques for SCRs as described below :
(0 ) Lead-mounting, For load-c~ent rating of about one ampere', lead-mounted SeRs are
used, Fig. 4.39 (a ). Such SCRa do not require any additional cooling or heat sink. Their h ousings

dissipate sufficient heat by radiation and convection.

(b) Stud-mounting. This type of construction shown in Fig. 4.39 (b) is very widely us ed due

to its fl exibility and ruggedoess . The threaded stud forms the anode. The SCR is attached to
a h eat sink by means of threaded stud and nut. Thus anod e gets electrically connected to the
heat sink. If electrical connection between anode and h eat sink is undesirable then mica or
PTFE washers are used in betwee"n the joini."'lg surfaces. Both .mica and PTFE conduct heat
easily put act as insulators to electricity.

(c ) Boltdown mountin.g. This is als o called flat-pack m ounting. This type of device
mounting has tabs \vith one or more holes. Sometimes the hole is provided in the middle a3


[Art. 4.9]

Th:.. ristors


H. S.l.




(0 )




Cool ing fins




~ H .S.



Fig. 4.39. Different SCR mountings and heat sink.

shown in Fig. 4.39 (c). Bolts are pu shed through these holes so as to mount the device on to
heat sink with nuts etc. In case the device is to be insulated from the heat sink, a thin insulating
mica or PTFE washer is used between the device and heat sink and the bolt is made up of
nylon. This type of mounting is used for small and medium ratings.

(d) Press-fit mounting. Press-fit (or pressure-fit) package is designed for insertion into an
appropriate sized hole in the heat sink. The insertion may be done by using a vice and pressing
the device into the hole using wooden block etc. For large sizes, the insertion is carried out by
means of a hydraulic r am. This type of mounting is used for large rated thyristors. Fig. 4.39
(d) illustrates press-fit mounting of SCRs.
(e) Press-pak mounting. This type of mounting is also called "disc" or "hockey-puck"
mounting because of its shape. The SCR is clamped between t\l'O heat sinks, Fig. 4.39 (e) .and
external pressure is applied evenly so that there is no deformation of any part. The heat sinks
may be air, wateT or oil coolfi'd. Such type of mounting is used for thyristo rs of very high current
ratin gs.


[Ac!. 4.9J

Power Electronics

Example 4.19. The data sheet for a thyristor gives the following values:
Jm = 12S'C
9j<=O. l SC/ W
Be.. =O.075C/W
(a) For average power dissipation of 120 W: check whether the selection of heat sink g from
Fig. 4.38 is satisfactory. Use first method of heat sink selection with ambient temperature of
40 C.
(b) A sinusoidal voltage source 0(230 V. 50 Hz feeds power to a resistiue load of R = 2 n.
For a firing angle delay of zero degree, choose a suitable heat sink and find the circuit efficumcy.
(e) For the heat sink chosen in part (a), compute case and junction temperatures in case
the firing angle delay is 60.
Solution. (0 ) For the' heat sink g , Fig, 4.38 gives a value of
T .. -TA=54C for Prlu =120W.
From Eq. (4.16),

9. . = :2~ =

O.4, CIW.

125 - 40
p o. = (0.15 + 0.075 + 0.45) - 125.93 W.

From Eq. (4.15),

As this computed value of Pall is different from the previous value of 120 W, another heat
sink, say f, for which T, - TAo = 58e for Pav = 120 W should be tried .

From Eq. (4.16),

9... = 120 = 0.483

From Eq. (4.15),


125 - 40

= (0.225 + 0.483) = 120.06 W .

This shows that selection of heat sink f is satisfactory.

For Pall = 120 Wand for sinusoidal current, Fig. 4.19 (b) gives average current rating for
the thyristor as 80 A for 1800 conduction angle or ex = 00 , 74 A for a = 60 and 68 A for
cr. = 900
(b) F or a = 0 0 , conduction angle is 180 0 Here second method of heatsink. selection is used .

TAo V=

1 f ' Vm .
Vm 'l'2x 230
2it 0 Ii: sm rot dlcot) = rtR = 7t X 2

.. A

=51.77 =. . 2

For this current, P ou from Fig. 4.19 (b) is 90 W .

From Eq. (4.15),
From Eq. (4.16),


= 1259~ 40 -


= 0.7194'


T. -TA =90xO.7194=64 .75 C

For T, - TA = 64.75e and Pr;J.u= 90 W, Fig. 4.38 shows that heat sink c should be selected.
The use of third method of heat. sink selection is also demonstrated. First I~A~ is calculated
as in the second method. F or this current, Pall = 90 W from Fig. 4.19 (b) and T( =112e from
Fi g. 4.20 (b). Now sm.< temperature from Eq. (4.17) is
T. = 112- 90 x 0.075 =105.25'C
For T, - TA = 65.25C and Pcw = 90 W, Fig. 4.3B shows that heat sink c should be chosen. As
expected, this agrees with the choice made by the use of second m ethod.

Thyrist nrs

[Art. 4.9J

Power delivered to load,



Vr =rms value of load voltage

wher e

v,~[;<J:v;.Sin'wtd(wt)r ~ ~m
P= (V2mJ' .lR ~ (230'if)'
1. = 13225 W .
2) 2


= 13225 + 90 ~ 0.993 pu or 99.3%.

:. Circuit efficiency


For ex = 60, conduction angle is 180 - 60 = 120

ITAV =21t

J' ViSinc.ot . d(oot)=21tR

V (1+coscx)

2< . 2 (1 + cos 60 ) =38.82 A.

For I TAv =38.82 Aand conduction angle of 120,

4.38 for h eat sink e.
T, - TA = 46'C
T, ~ 40 + 46


from Fig. 4.19 (b) is 52 W and from Fig.

86' C.

From Eq. (4. 17), case temperature,

T" = T8+ Pall' Su = 86 + 90 X 0.075 = 92.75C

and junction temperatur e,

T) =T" + Pall'


= 92.75 + 90 X 0.15 = 106.25C.

This example demonstrates that selection of heat sink by second and third methods is more
simpler than by the first method.

Example 4.20. For a thyristor, maximum junction temperature is 125C. The thermal
resistances for the thyristor-sink combination are Sjc = 0.16 and SCI = O.OBoe I W. For a heatsink
temperature of 70C, compute the total average power loss in the thyristor-sink combination.
In case the hea t sinh temperature is brought down to 60C by forced cooling, find the
perce ntage increase in the device rating.
Solution . From the equivalent circuit of Fig. 4.37 (b )
T j = T~ + POll (Sjc + 9c~)

125 - 70
p. ol ~ 0.16 + 0.08 = 229.17 W
Thus total average power los5 in the thyristor-sink combination is 229.17 W. With improv ed

P" , ~

125 - 60

=270.83 W.

Thyristor r atin g is proportional to the square root of aver age power loss.
:. P ercentage increase in thyristor rating
= "270.83 - " 229.1 7 100 = 8 71a
. ".


[Art. 4.10J

Power Electronics


seR ratings have improved considerably since its introduction in 1957. Presently. SCRs
with voltage and current ratings o f 10 kV and 3 kA are available. However, for some industrial

applications, the demand for voltage and current ratings is so high that a single SCR cannot
fulfll such requirements . In sucn cases, SeRs are connected in series in order to meet the h.v.
demand and in parallel for fulfilling the high current demand. For series or parallel connected
SeRs, it should be ensured that each SCR rating is fully utilized and the system operation is
satisfactory. String efficiency is a tenn that is used for measuring the degree of utilization of
SeRs in a string. String efficiency of SeRs connected in series/parallel is defined as
string efficiency
Actual voltage/current rating of the whole string
- [Individual voltage/current rating of one SCR J [Number of SCRs in the stringJ
In practice, this ratio is less than one. For obtaining highest possible string efficiency, the
SCRs connected in s eries/parallel string must have identical J- V characteristics. As SCRa of the
same ratings and specifications do not have identical characteristics, unequal voltage/current
sh aring is bound to occur for all SeRs in a string. As a consequence, string efficiency can never
be equal to one. However, unequal voltage/current sharing by the SCRs in a string can be
minimised to a great extent by using external equalizing circuits.
Even in a string provided with external equalizing circuits, the string efficiency is less than
unity. For a given system, if one extra unit is added to the series7paraUel string, the
voltage/current shared by each device would become lower than its normal rating. The use of
this extra unit will certainly improve the reliability ofthe string though at an increased cost. A
measure of the reliability of string is given by a factor called derating factor DRF defined as
under :
DRF = 1 - string efficiency

For example, for a string voltage of 3300 V, let there be six series-connected SeRa, each of
600 V rating.
:. String efficiency


3300 , /
= 600 x 6 = 0.917 or 91.7 %
DRF = 1 - 0.917 = 0.083 or 8.3%

If one extra unit is connected in series with the same system voltage, then string efficiency
= 600 x 7 = 0.786 or 78.6% and DRF = 1 - 0.786 = 0.214 or 21.4%.
With the addition of extra SCR, DRF has increased from 8.3% to 21.4%, indicating higher
reliability of the string, though at an extra cost.
The object of this section is to study the problems concerning the serieslparallel operation .,
of SCRs and to discuss the measures adopted to overcome these proble m ~.
4.10.1. Series Operation
When system voltage is more than the voltage rating of a single thyristor, SCRs are
conne cte d in seri es in a string. As stated b e fore , thes e SCRs should ha ve thei r J - V
characteristics as close as possible. On account of inherent variations in their characteristics,
the voltage shared by each SCR may not be equal. For ins tance, consider two SeRs with their
static 1- V characteristics as shown in Fig. 4.40. For SCRl, leakage resistance (= VI I lo) is high
whereas for SCR2, it is low (V 2/ Jc). F or the same lea.\cage current 10 in the series connected


[Art. 4.10]




Fig. 4.40. Series connected SeRa.

SCRa, SCR1 supports rated voltage VI whereas SCR2 supports voltage V 2 < VI' Each SCR in
Fig. 4.40 is rated for a forward blocking voltage of VI volts which is always less than its forward
breakover voltage. Here VBOI and VB02 are the forward breakover voltages for thyristors 1 and
2 respectively. It is seen from Fig. 4.40 that two SCRs can support a maximum voltage of
VI + V 2 and not the rated blocking voltage 2V1. The string efficiency for two series connected
SCRs of Fig. 4.40 is,

VI +

V, _!( V,)

-2 I +V

This shows that even though SCRs have identical ratings, voltage shared by each is not the
same and string efficiency is therefore less than one.

A uniform voltage distribution in steady state can be achieved by connecting a suitable

resistance across each SCR such that each parallel combination has the same resistance. This
will require different value of resistance for each SCR which is a difficult propositi'on. A more
practical way of obtaining a reasonably uniform voltage distribution during steady state
working of series-connected SCRs is to connect the same value of shunt resistance R across
each SCR as shown in Fig. 4.41. This shunt resistance R is called the static equalizing circuit.
Magnitude of parallel resistan ce R can be obtained as follows U] .
Consider n thyristors connected in series as shown in Fig. 4.41. Let SCR1 has minimum
leakage current 1bmn and each of the remaining (n -1 ) SCRs have the same leakage current
1bmx > hmn. An examination of Fig. 4.40 (b) reveals that an SCR with lower leakage current
blocks more voltage.
As SCRI has lower leakage current, it will block voltage Vom (say) which..Js more than that
shared by each of the other (11. - 1) SCRs. Here Vbm is the maximum permissible blocking
voltage of SCR I. It is seen from Fig. 4.41 that

!t =1-1bmll and 12=1-101T''::

I =total string current
Voltage across SCR1 is Vbm=IIR

[Art. 4.10]
Power Electr onics
--~--~----------------~.. ,~
: -------------
-. ~

Onl: SCR wit h leokOl}



( Ibmn


,i__ (n-1) SC R, with ',akag' cumnt Ib~


..J"' :

St ri~






I .r'"


cu r r. l:nt




;.~---------- St r in 9 voltagl: Vs -------------.~,

Fig. 4.41. Static-voltage equalization for series-connected string.

Voltage across (n -1) SCRs

=(n -

1) 1.JI

For a string voltage V.. the voltage equation for the series circuit of Fig. 4.41 is


V, =llR+ (n -1) RI, =V'm + (n - 1) R(l -1,=)

=Vbm + (n -l)R [1 1 - (Ibnu:-IbmnH
= V'm + (n - 1) RI 1 - (n - 1) R . t> I,
11 Ib =I bmx - I bmn
RI 1 = V bm , V.. = n V bm - (n - 1) R . 11 Ib


= n Vbm -


.. .(4 .18)

(n - 1) . t> I,

The SCR data sheet contains only maximum blocking current I bmx and rarely 11 lb' In such
a case, it is usual to assume 11 Ib =Ibrru: witll Ibmn = O. With this, the value of R calculated from
Eq. (4.18 ) is low . than what is actually required. The value of minimum leakage, or blocking,
current Ibm may be acquired from manufacturers if required, but data sheet does not give its
Once the value of R is alculated, its power rating is given by


PR =1f


=rms voltage across


It is likely that SCRs do not have identical dynamic characteristics. In such a case,
series -connected SCRs will have unequal voltage distribution during the transient conditions
of turn-on, turn-off and high frequency operation. The dynamic characteristics of two SCRs
during turn-on are shown in Fig. 4.42 (a ) where it is assumed that turn-on time ofSCR2 is more
than that of SCR1 by 11 td . Before these two SCRs are gated, string voltage V, is shared as
V, 12 by each thyristor as shown. Now both SCRs are gated at the same time . F,:, SCR1 has less
turn-on tim e, it gets turned-on at instant tlo whereas SCR2 is yet off. Voltage across SCR1 drops
from V,/ 2 to alm ost zero. At the same inst~t t I , voltage across off SCR2 win boost from V, / 2
to almost full V s' Thus , the voltage shared by two SCRs are unequal. Mer instant t I , voltage
V, a cross SCR2 m ay turn it on in .c ase V, is greater than its breakover voltage. S CR2 will ,
however, get turn ed on a t time (t t + 11 td ) as assumed> Fig. 4.42 (a ).
During t urn-off, thyristor characteristics are shown in Fig. 4.42 (b ). SCR1 is as sumed to
h ave less turn-off time t q l than that ofSCR2, i. e. tql < t q2 . At instant t 2 , SCR1 h as recovere d and
is pa.5sin g through zer o vol tage whereas SCR2 is developing revers e recovery volt age xy. At


[Art. 4. 10J



... tt.ld-et



I ,

I f,



:: ~l1
,., .,


An ode


, ,--

~- 2
. ,,Y
, I tl:



: i

:" I

, tq 1"""",,



: I



,. 1/ :
, .,...2 ..





Fig. 4.42. Unequal voltage distribution for two series connected SCRs during

turn-on and (b) turn-off.

instant t l in Fig. 4.42 (b), both SCRs are developing different reverse recovery voltages given
byab for SCRl and ac for SCR2 as shown, so the two SCRs have unequal voltages across them
at tl ' It is thus seen that .SCRs with different characteristics during tum-ofT time suffer from
unequal voltage distribution during their turn-off processes. It may thus be concluded from
above that series-connected SCRs do suffer from unequal voltage distribution acros s them

during their turn-on and turn-off processes and also during their high-frequency operation
which means more frequent turning-on and turning-off of the devices.
A simple resistor as shown in Fig. 4.41 for static voltage equalization cannot maintain
equal voltage distribution under transient condition . Duritig turn-on and turn-off, the
capacitance of the reverse biased junctions determines the voltage distribution across SCRs
in a series connected string. As reverse biased junctions are likely to have different
capacitances, called self-capacitances, the voltage distribution during tum-on and turn-oft:
periods would be unequal. Voltage equalization under these conditions can, however, be
achieved by employing shunt capacitors as shown in Fig. 4.43. This capacitance has the effect
of removing the inequalities in thyristor self-capacitances. In other words , during tum-on and
turn-off periods, the resultant of shunt capacitance and self-capacitance of each SCR tend to
be equal for each of the series connected SeRs. Thus the shunt capacitors playa dominant
role in equalizing the voltage distribution across the series-connected thyris tors during their
turn-on and turn-off processes.
When any SCR is in the forward blocking state, the capacitor connected across it gets
charged to a voltage existing across that SCR. When this SCR is turned on, capacitor discharges
heavy current through this SCR. For limiting this discharge current spike, a damping r es istor
Rc is used in series with capacitor C as shown in Fig. 4.43. Resi stor Rc also damps out the
high frequency oscillations that mily arise du e to the series combination of R e , shunt capacitor
and circuit inductance. Combination of Rc and C is called the dynamic equaliz ing circuit and
is shown in Fig. 4.43 (a ). Note L1.at the function of Rc and C us ed in Fig. 4.43 is to equalize
the voltage during dynamic (or transient) cond itions and to protect the thyris tors against high
du ldt.


Power Electronics

[Act. 4.10}
Dynamic equall zing
r- -- . ---,



' D,




;, C =,




2) ,,,




r- ' - - - ~- ' ~- - .






r ----



',{ ,

Stot ic equoliz ing

circu it


L. ____ -C __



__ J,




,\ Revers e recovery
curr en'

' It.. Re:ve:rst



Fig. 4.43. Dynamic and static equalizing circuits for series-connected SCRs.

A diode D is also placed across Re- When forward voltage appears.,diode byPasses Rc
during charging time of the capacitor C. This makes the capacitor more effective' ui'Yoltage
equalization and for limiting du/dt across SCR. However, during capacitor discharge,Rc comes
into play for limiting the current spike and rate of cbange of current di/dt. During turn-off
period, when all SCRs are developing r everse voltage, the reverse recovery current ir flows
through all series connected SeRs as shown in Fig. 4.43 (a). However, if one SOR recovers
early, it will not allow the passage of ir from ~he other SORs. If SORI is assumed to recover
fully and earlier than other SORs, then reverse recovery current due to other SCRs can pass
through R connected across SCRI as shown in Fig. 4.43 (b ). In this figure, i, may flow through
C, Rc also in case the conditions are favourable. For simplicity, only two SORs are shown in
Fig. 4.43. The existence of reverse recovery current is desirable as it facilitates the turning-off
process of the series-connected SOR string.


Value of capacitance C shown in Fig. 4.43 can be obtained as under :

In series connected SORs, voltage unbalance during turn-off time js more predominant
than it is during turn-on time, Fig. 4.42. Therefore, choice of capacitor C is based on the reverse
recovery characteristics of SORs . In Fig. 4.44 (b) are shown reverse recovery characteristics
for two SORs of Fig. 4.44 (a ). SORI is assumed to have short reverse recovery time as compared
to SCR 2. Shaded area 11 Q, proportional to the product of current and time, is the difference
in the reverse recovery charges of two thyristors 1 and 2. Under this assumption, SORI
recovers firs t:; it, therefore, goes into blocking state and does n ot allow the passage of excess
charge 11 Q left on SCR2. This charge 11 Q can, however, pass through C as shown in Fig. 4.44

The voltage induced by 11 Q in the capacitor C, connected across SCRI, is 11 QI C; whereas

no voltage is in duced by 11 Q (= Q 2 - QI) in C connected across SOR2. There is thus a difference
in voltages, equ al to Q2 ~ Q1 = I1CQ, t o which the two shunt cap acit ors are charged. The
thyristor with the least reverse recovery time will share the hi ghest transient voltage, say
V bm . As stated above, the voltage difference to which the two shunt capacitors are charged

, ..

[Art. 4.10)




r-' .

'---c. -,



i, =c


~ __ _

_ .J







(b )

Fig. 4.44. (0 ) Flow of reverse recovery current if SCRl recovers 6rst

(b ) Variation of reverse recovery characteristics fOr"two SCRs of.Fig. 4.44 (0 ).
during reverse recovery time is 6 QI C, the transient voltage shared .by slow thyristor 2 must

V'm - '"cQ (less than V'm shared by fast thyristor 1). Thus, in Fig. 4.44 (a),

voltage across fast top thyristor 1 , VI = Vbm

and voltage across slow bottom thyristor 2, V2 = Vb~ - 6 C
: . String vol tage,


an d


=VI + V 2 =Vb," + Vbm -

V'm=~ (v. + "'cQ)

V, = V'm - "'! = ~ (V. - '"cQ)

In order to aid the reverse recovery process of the thyristor s in a string, the string voltage
reverses in polarity as shown in Fig. 4.44 (a ).
Now consider that there are n series-connected SeRa in a string as shown in Fig. 4.45 . If
top S CR h as characteristics similar to SCRI of Fig. 4.44 (b) and the r emaining (n - 1) SCRs
h ave ch aracteristics similar to SCR2 of Fig. 4.44 (b), then SCRI would r ecover first and support
a voltage Vbm. The charge (n - 1) 6 Q fr om the remaining (n - 1) thyristors would pass through
C connected across top fast SCRI and as a result, a voltage (n - 1) 11 QI C would be induced in
C. As befon.::excess charge contributed by each on e of the (n -1 ) thyristors is 60 Q, therefore,
the voltage aCrOSS each one of the slow thyristors is


m -

"'cQ ) as shown in

Fig. 4.45. Thus, fo r

a string of n series-connected thyristors, voltage across fast top thyristor I,V } =Vbm
voltage across each one of the slow thyristors . V2 is

v.-, = Vbm - ~
and voltag e across (n - 1) slow thyris tors

=(n -

1) Vo;!


Power Elec tronics

[MI. 4.10]

=(n-l) (Vbm -~!)

String voltage,


=V, + (n -



= Vbm + (n -1 )(Vbm Its simplification gives Vbm



(n -l)~

C = (n -1) ~ Q.
n V bm

... (4.19)

Voltage across each one of the slow thyristors, in terms of

VJ is given by


(Vbm -

-, .......


=V, + (n n



1) ~





V bm

= 1'1


t =,
" ,i C
= .......

C ~

' -c

Fig. 4.45. String having

n-senes connected "thyristors .

(n - 1). ~Q



__ .l

~. ~Q and this must

be supported by all SCRs together, which is equal to n.Vbm - V


;:: (n_~ t. Q

Durin g the turn- off process, the source voltage V, must

reverse to aid the r everse recovery current. The transient voltage
whi ch each SCR must be able to withstand is Vbm - However, total
voltage acting across th e circuit consisting of VJ thyristors n, 4,

n. V bm -

.. .(4.20)

3, 2 and top C, and per KV1, is V, + (n




(n -l)~Ql

C = (n -1). ~Q
n,Vbm V,

... (4. 19)

4.10.2. Parallel Operation

'When current required by the load is mor e than th e rated current of a single thyristor,
SCRs are connected in parallel in a string. For equal sharing of currents, I-V characteristics
of SCRs during forward conduction must be identical as far as possible.
In Fig. 4.46 (a ) are shown two SCRs in parallel and their characteristics during forwa rd
conduction are shown in Fig. 4.46 (b ). For paralle\.connected SeRs, voltage drop Y r across
them must be equal. Fig. 4.46 (b ) shows that for the same voltage drop Vr SCRI shares n
rated current I ! wh ereas SCR2 carries current 12 much less th an th e rat ed curren t I t The
total current ca rried by th e unit is I , ,.. 12 and not the rated current 21, as required. Therefore,
s tring effi cien cy is given by

[Art. 4.10]



~:- i.

I .... b t

. _ ..9

, '
., ,

LL..J....-;'vc-,- - - V




(el ,

Fig. 4.46. (a ) and (b) Parallel operation of two thyristors


Dynamic resistance decreases as junction temperatul'O:: risesr


Now consider n parallel connected SCRs. For satisfactory operation of these SCRs, they
should get turned on at the same moment. The importance of their simultaneous turn on can
be explained with an example. Consider that SCRI has large turnon time whereas the
r emaining (n -1) SCRs have low turnon time. Under this assumption, (n - 1) SeRs will turn
on first but one SCR1 with longer t~on time is likely to remain off. The voltage drop across
(n - 1) SCRs falls to a low value and SCR1 is therefore subjected to this low voltage. For a given
gate drive power, anode to cathode must have Some minimum forward voltage, called finger
voltage, for a thyristor to turnon. If voltage across SCRl drops to a value less than its finger
voltage, then this thyristor will not tum on. As a consequence, the remaining (n - 1) SeRs,
w~ich are already on, will have to share the entire load current. A$ such, the~e SCRs may be
overloaded and damaged because of heating caused by overcurrents.
If one SCRI in a parallel unit carries more current than. other SCRs, then this SCRI will
have greater junction temperature rise. As a result, its dynamic resi~.tance (= dVrldIa ) during
forward conduction, Fig. 4.46 (c) decreases and this further increases the current shared by this
SCR. In Fig. 4.46 (c), dynamic resistance is oalab wd current shirred is 1'. Because of junction
temperature rise, its dynamic resistance decreases to oalac and current shared by SCRl
increases to 1". This process of anode current rise becomes cumulative and subsequently the
junction temperature of SCRI exceeds its rated value; as a r esult SCRl is damaged. This
sequence of events may engulf another SCR and in this manner all SCRs in the string may be
destroyed permanently. Therefore, wh.3 SeRs are to be operated in parallel, it should be
ensured that they operate at the same temperature. This CfUl be achieved by mounting the
parallel unit on one common heat sink..
Unequal current distribution in a parallel unit is .also caused by the inductive effect of
current carrying conductors. Vlhen SeRs are arranged unsymmetrically as shown in Fig. 4.47
(a), the middle conductor will have more inductance because of more flu.'C linkages fro m two
nearby conductors. A3 a consequence, less current flows through the middle SCR as compar::d
to ou ter two SCRs . This unequal current distribution can be avoided by mounting the SC R.s
symmetrically on the heat sin.., ., shown in Fig. 4.47 (b).


Power Elec tronics

(Art. 4.10)

(Heat sink






Fig. 4.47. Parallel operation ofSCRs (al unsymmetrical arrangement and

symmetrical arrangement on heat sinks (c) current equalization by the use of reactor .

In ae circuits, current distribution can be made more uniform by the magnetic coupling of
the parallel paths as shown in Fig. 4.47 (c). The tapped point A is the mid point of the reactor.
H anode currents are such that II = 12 , then flux produced by two halves of the reactor oppose
each other. As A is the mid point, opposing flux linkages cancel and there is therefore no voltage
drop in the reactor. If currents II and 12 Bre unequal, say 11 > 12 then resultant flux linkages
are not zero. These flux linkages induce emfs in Ll and L2 as shown. Emf across reactor L l
opposes the flow of 11 whereas that across L2 aids the flow of 1a. There is thus a tendency to buck
1! and boost 12 so as to minimise the unbalance of currents in the parallel unit .
When three or more SeRs are connected in parallel, reactors can be arz:anged accordingly
so as to minimise the current unbalance.
Example 4.21. A string of four seriesconnected thyristors is provided with static and
dynamic equalizing circuits. This string has to withstand an off-state uoltage of 10 k V. The static
equalizing resistance is 25000 n and the dynamic equalizing circuit has Rc = 40 nand
C =0.08~. The leakage currents for four thyristors are 21 mA, 25 mA, 18 rnA and 16 rnA
respectiuely. Determine voltage across each SCR in the off-state and the discharge current of
each capacitor at the time of turn-on.
Solution. Let 1 be the string cur r ent in the offstate. Then c'urrent through
static' equalizing resistanc e R of25000 n is (I-leakage current), current through each SCR is its
own leakage current and no current flows through series r')mbination of Rc and C.
:. Voltage across
Voltage across SCR1
Voltage .across SCR2
Voltage across SCR3
Voltage across SCR4

=voltage across each SCR

=(1- 0.021) x 25000 =VI
=(I - 0.025) x 25000 =V,
=(I - 0.018) x 25000 =V,
= (I - 0.016) x 25000 = V,

The sum ofV l , V 2 , Va and V" gives

25000 (41 - 0.08) = V1 + V 2 + Va + V" =string voltage, 1000 0 V
1= 0.12 A
From above, voltage across SCRI

= (0.12 -

0.02 1) x 25000 = 2475 V

Similarly V2 = 2375 V, Va = 2550 V an d Vol = 26 00 V.

..' '.


[Art. 4.10J


Discharge current through SeRt at the time of turn on .

= ;~ =

2:~5 = 61875 A

Similarly, discharge currents through thyristors 2, 3 and 4 are respectively 59.375 A, 63.75
A and 65 A.

Example 4.22. SCRs with a rating of 1000 Vand 200 A are available to be used in a string
to handle 6 k V and 1 kA. Calculate the number of series and parallel units required in case
derating factor is (a) 0.1 and (b) 0.2.

Solution. (a) Derating factor, DRF = 1 - string efficiency



6000 =1 - 1000
n, x 1000
np x 20.0

: . Number of series-connected SCRs,

n, = 1000 x 0.9 = 6.6" 7

Number of parallel-connected SCRs,

x 0.9 = 5.5" 6

np = 200

As above, number of series -connected SCRs,

n, = 1000 x 0.8 = 7.5" 8

and number of parallel-connected SeRs.

x 0.8 = 6.25 " 7

np = 200

With higher value of DRF. more SCRs are required and therefore voltage and current
shared by each device are lower than their normal rating. This increases the string reliability
though at an increased investment.
Example 4.23. It is required to operate 2S0-A SCR in parallel with 350-A SCR with their
respective on-state voltage drops of 1.6 Vand 1.2 V. Calculate the valu.e of resistance to be
inserted.in series with each SCR so that they share the total load of 600 A in proportion to their
current ratings.
Dynamic resistance of 250-A SCRt
Dynamic resistance of 350-A SCR2
Let R be the resistance inserted in series with each SCR. With this, current shared by
SCRI = 600
350 .
~ 250
Tota l res13tance
1. 6

and current shared by

SCR2 = 600

? ~O


Total resls tance

~ 350


Power Electronics

[Act. 4.11]


From above,
Its simplification gives


1.6 R
250 +

250 5
=-=350 7'

R = 0.004 n.

Thus the resistance to be inserted in series with each SCR is 0.004 n.

Example 4.24. Discuss the conditions which must be satisfied for turning-on an SCR with
a gate signal.
Solution. Conditions which must be satisfied for turning-on SCR with a gate signal are
as under:
(a) An SCR must be forward-biased. It means that anode must be positive with respect

to cathode.
(b) Gate pul'Se width must be more than the turn-on time of an SCR. This will ensure that
anode current exceeds the latching current before gate signal is removed.

Cc) Anode to cathode voltage must be more than finger voltage. A finger voltage is that
voltage below which an SCR cannot be .turned on with 'a gate signaL
(d) Magnitude of gate current must be more than the minimum gate current required to
turn-on a thyristor, otherwise the thyristor tum-on will not be reli able.
(e) Magnitude of gate current must be less than the maximum gate current allowed,
otherwise gate circuit may. be damaged.

(f) The gate triggering must synchronize with the ac supply.


The term thyristor includes all four-layer p-n-p-n devices used for the control of power in
ac and dc systems. The silicon controlled rectifier is the most popular member of thyristor
family. There are several other members of thyristor fQ.ll1i1y like PUT, SUS, SCS, triac, diac
etc. All these devices, except triac, are low power devices. Several new devices have been
developed and added to the thyristor family. These recently developed thyristor devices are
asymmetric thyristor CASCR), reverse conducting thyristor (RCT), static induction thyristor
(81TH), gate-assisted tum-off thyristor and gate turn-off (GTO) thyris~or. MOS-controlled
thyristor (MCT) has already been described in Chapter 2. The object of this section is to discuss
other members of the thyristor family.
4.11.1. PUT (Programmable Unijunction Transistor)
It is a pnpn device like an SCR. But the major difference is that gate is connected to n-type
material near the ar..ode as shown in Fig. 4.48 (a). PUT is used mainly in time-delay, logic and
8CR trigger circuits. Its largest rating is about 200 V and 1 A. Circuit symbol and I - V
characteristics of a PUT are shown in Fig. 4.48 (b) and (c) respectively.

In a P UT, G is always biased positive with respect t o cathode. "'Vhen anode voltage exceeds
the gate volt age by a bout 0.7 V,junction J 1 gets forward biased and PUT turns on. When anode
voltage beco mes less than gate vo ltage, PUT is turned off.

[Art. 4,IIJ













L_ _ _ _.



Fig. 4.48. (a) Schematic diagram (b) circuit symbol and (e) J- V characteristics of a PUT.

4,11.2, SUS (Silicon Unilater a l Swit ch )

A SUS is similar to a PUT but with an inbuilt low-voltage avalanche diode between gate
and cathode as shown in Fig. 4.49 (a). Because of the presence of diode. SUS turns on for a fixed
anode-to-cathode voltage unlike an SCR whose trigger voltage and/or current vary widely with
changes in ambient temperature. SUS is used mainly in timing, logic and trigger circuits. Its
r atings are about 20 V and 0.5 A. Circuit symbol. equivalent circuit' and I - V characteristic of
an SUS are shown in Fig, 4.49 (b), (c) and (d) respectively, . _'






__ J

- Va




Fig. 4.49. (a) Schematic diagram (b) circuit symbol

(c) equivalent circuit and Cd) 1-V characteristics of an SUS.



4,11,3, SCg (Silicon Con trolled Switch)

SCS is a tetrode, i.e. four electrode thyristor. It has two [~tes, one anode gate (A G) like a
PUT and an other cathode gate (KG) like an SCR. In other words, ses is a four layer, four
terminalpnpn device; with anode A, cathodeK, anode gateAG and cathode gate KG , Fig. 4.50
(a ). SCS can be turned on by either gate. Circuit symbol and J - V characteristic of an 8eS are
shown in Fig. 4.50 (b) and (c ) respectively.
When a negative pulse is applied to gate AG, junction J 1 is forward biased and 8 eS is
turned on. A positive pulse atAG will reverse bias junction J 1 and turns off the SCS .
A positive pulse at gate KG turns on the devi ce (just like an S CR) and a negati ;> e pulse at
KG turns it off Gust like a G,T,Q,),


Power Electronics

[Art. 4.IIJ






- Vo






Fig. 4.50 .









Schematic diagram (b) circuit symbol and

(e) J. V

characteristic of an SCS.

Its ratings are about 100 V and 200 rnA. This can be operated like an OR gate. Its
applications include :
(i) timing, logic and triggering circuits

(iO pulse generators

(iii) v:oltage sensors


oscillators etc.

4.11.4. Light Activated Thyristors

The circuit symbol and I - V characteristics of light. activated thyristor, also called LASeR,
are shown in Fig. 4.51. LA SeRs are turned on by throwing a pulse of light on the silicon wafer
of thyristor. The pulse of appr opriate wavelength is guided by optical fibres to the special
sensitive ar ea of the wafer. If the intensity of light exceeds a certain value, excess electronhole
pairs are generated due to radiation and forward-biased thyristor gets .t urned on.



(0 )
Fig. 4.51. (a) Circuit symbol and (b) I-V characteristic of LASCR

The primary use of light-fIred thyristors i5 in high-voltage hi gh-current applications, static

r eactive-power compensation etc. A light-fired thyristor has complete electri cal isolation
be tween the light-triggering source and the high-voltage an ode-ca thode circuit. Light-activated
thyris to rs are available up to 6 kV and 3.5 kA, wi th on-state voltage drop of about 2 V and
with light-triggering requirements of5 m\V.


[A n . 4.11]


4.11.5. The Diae (Bidirection al Thyristor Diode )

A cross-sectional view of a diac showing all its layers and junctions is depicted in Fig. 4.52
(a ). If voltage V 12 with tenninal 1 positive with respect to terminal 2, exceeds break-over
voltage Vaal then structure pn pn conducts. In case term in al 2 is positive with respect to
terminal 1 ~d when V 21 exceeds breakover voltage V B02 structure pn pn' conducts. The term
'diac' is obtained from capital letters, DIode that can work on AC. Fig. 4.52 (b) gives the circuit
symbol and Fig. 4.52 (c) the I-V characteristics of a diac. It is seen that diac h as symmetr ical
breakdown characteristics. Its leads are interchangeable. Its turn-on voltage is about 30 V.
When conducting, it acts like a low resistance with about 3 V drop across it. When flat
conducting, it acts like an open switch. A diac is sometimes called a gateless triac.









Vaal Va






Fig. 4.52. (0) Cross-sectional view (b ) circuit symbol and (c) ,!- V characteristics of a di ac.

4.11.6. The Triac

An SCR is a unidirectional device as it can conduct from anode to cathode only and not
from cathode to anode. A triac can, h owever, conduct in both the directions. A triac is thus a
bidirectional thyristor with three terminals. It is used extensively for the control of power in
ac circuits . Triac is the word derived by combining the capital letters from the words TRIode
and AC. Wh en in operation, a triac is equivalent to two SCRs connected in antiparallel. The
circuit symbol and its characteristics are shown in Fig. 4.53 (a) and ( b ) respectively. As the
MT2 positive
I g'l>!q l > l,j O



Igo =O


- lgO

-19 1

-1 92

.\\T 2 n egc:i 'l l!


(0 )

-1 0


F'ig. 4.53. (a ) Circuit symbol and ( b) static 1- V ch aracteris tics of a t:ri ~c.


Power Ele ctronics

(Art. 4.111

triac can conduct in bo th the directions, the terms anode and
ca thode are not applicable to triac. Its three terminals are usually
designated as MTl (main terminal 1), MT2 and the gate by G as
in a thyristor. For understanding the operation of the triac, its
cross-sectional view sh owing all the layers and junctions is
sketched in Fig. 4.54. The gate G is near terminal MTl. .The
cfoss-hatched strip shows that G is connected to N3 as well as P2.
Similarly. terminal MTI is connected to P2 and N z i terminal MT2
to Pi and N 4 .
With no signal to gate, the triac will block both half cycles of
the ae appli ed voltage in case peak value of this voltage is less than
the breakov cr voltage orvBD l or V BD2 of the triac, Fig. 4.53 (b). The
triac can, however, be turned on in each half cycle of the applied Fig. 4.54. Cross-sectional
voltage by ap plying a pos itive or negative voltage to the gate with
view of a triac.
respect to terminal MTl. For convenience, terminal MTI is taken
as the poi nt for measuring the voltage and current at the gate and MT2 terminals.
The tum-o n process of a triac can be explained as under :
(i) MT2 is positi ve and gate current is also positive. When MT2 is positive with respect to
MTI, junction PI NI, P2 N2 are forward biased but junction N I P2 is reverse biased. When
gate terminal is positive with respect to MTl , gate current flows mainly through P2 N2
junction like an ordinary SCR, Fig. 4.55 (a ). When gate current has injected sufficient charge
into P 2 layer, r everse biased junction NI P z breaks down just as in a normal SCR. As a result,
triac starts conducting through P I N I P z N 2 layers. This sh ows that wh~n MT2 ' and gate
terminals are positive with respect to MTl, triac turns on like a conventional thyristor. Under
this condition, triac operates in the first quadrant of Fig. 4.53 (b). The device ism ore sensitive
in this mode. It is recommended m ethod of triggering if the conduction is desired in the first
(ii) MT2 is positive but gate current is negative. When gate tenninal is negative with r espect
to MTl , gate current flows through P 2 N 3 junction , Fig, 4.55 (b) and reverse biased junction
Nt Pz is forward biased as in a normal thyristor. As a result, triac starts conducting through
P , NI P 2 N3 layers initially. With the conduction of P l Nl P z N 3, the voltage drop acr oss this

'. 'I 1,

1, 1,1 Ig
MTI ( - )


., I


MT2(r )

MTt i - )







- _.Pz--J

(P,N, P2 NJ)




(P, N, P, N,)

MT 1(+)





(PzN,P, N,)


Io4TI ( ... )







MT 2{-)

MTz( - )

(e) more sen3itive

(d ) more sensitive
Fig. 4.55. Turning-on process in a triac. Final conduction is through PI N ) P,!N,! in
(0) and (b) and through P,! I'll PI N .. in (c) and (d).

[Art. 4.11 )

Thyristor s


path falls but potential of layer between P2 Ns rises towards the anode potential of MT2. As
the right hand portion of P 2 is clamped at the cathode potential of MTl, a potential gradient
exists across laye r P2 , its left hand region being at higher potential than its right hand region.
A current shown dotted is thus established in layer P2 from left to right. This current is similar
to conventional gate current of an SCR. As a consequence, right-hand part of triac consisting
of main structure PI NI P2 N2 begins to conduct. The device structure PI Nl P 2 N J may be
regarded as pilot SCR and the structure P INI P2N2 as the main SCR. It can then be stated
that anode current of pilot SCR serves as the gate current for the main SCR. As compared
with tum-on process discussed in (i) above, the device with MT2 positive but gate current
negative is less sensitive and therefore, more gate current is required.
(iii) j}IT2 is negative but gate current is positiue. The gate current Ig forward biases P2 N2
junction Fig. 4.55 (e). Layer N 2 injects electrons into P 2 1ayer as shown by dotted arrows. As a
r esult, reverse biased junction N I P I breaks down as in a conventional thyristor. Eventually the
structure P 2 Nt PI N4 is completely turned on. As usual, the current after turn-on is limited by
the extemall oad. As the triac is' turned on by remote gate N 2 , the device is less sensitiue in the
third quadrant with positive gate current.
(iv ) Both MT2 and gate current are negatiue. In thi.s mode, Na acts as a remote gate, Fig.
4.55 (d). The gate current 16 flows from P2 to Na as in a normal thyristor. Reverse-biased
junction N t PI is broken and finally, the structure P2 N} P l N4 is turned on completely. Though
the triac is turned on by rem ote gate N3 in third quadrant, yet the device is more sensitive under
this condition compared with turnon action with positive gate current discussed in (iii) above.
It can, therefore, be concluded from above that :

sens itivity of the triac is greates t in the first quadrant when turned on with positive
gate current and also in the third quadrant when turned on with negative gate
(ii) sensitiv ity of the triac is low in the first quadrant when turned on with negative
gate current and also in the third quadrant when turned-on with positive gate

Thus the triac is rarely operated in first quadrant with negative gate current and in the
third qu adrant with positive gate current.
As the two conducting paths fr om MTI to MT2 or from MT2 to MTI interact with each
other in the structure of the triac; their voltage, current and frequency ratings are much lower
as compared with conventional thyristors. At present, triacs with voltage and current ratings
of 1200 V and 300 A (r ms) ar e available.
Triacs are used extensively in residential lamp dimmers, heat control and for the speed
control of small single-phase series and induction motors.
A tri ac may sometimes oper ate in the rectifier mode rather than in the bidirectional mode.
This may happen due to the fan owing r easons:
For a given value of'positive gate current, a triac may turn on with MT2 positi ve
in first quadrant bu t may fail to turn on with MT2 negative.
(b) 'Wi th constant negati ve gate current, the triac m ay t urn on with MT2 negative in
third quadrant but may not turn on with MT2 positive.
The re ctifi er-mode can be overcome by increasing the value of gate current.


[Art. 4.12J

Power Electronics

4.11 .7 . Asymmetrical Thyristor (ASCR)

A conventional thyristor is able to block a large reverse voltage, but this blocking capability
is not required in several industrial applications. For example, in voltage source inverters
converting de to ae and in some chopper circuits, a freewheeling diode is usually connected in
antiparallel across each thyristor. This fr eewheeling diode damps the thyristor voltage to 1 to
2 V under steady state conditions. An asymmetrical thyristor, or ASCR, is specially fabricated
to have limited reverse yoltage capability ; thi s permits a reduction in tum-on time, turn-off
time and on-state voltage drop in ASCR A typical ASCR may bave r everse blocking capability
of 20 to 30 V and forward blocking voltage of 400 to 2000 V. ASeRs with turn-off time half of
that of a similar rated conventional SCRs have been developed. Fast turn-off ASCRs minimize
the size, weight and cost of commutating components and permit high frequency operation (20
KHz or mor e) with improved efficiency.
4.11.8. Reverse Conducting Thyristor (RCT)
A reverse conducting thyristor is a special case asymmetrical thyristor
with a monolithically integrated antiparallel diode on the same silicon chip.
This construction reduces to zero the reverse blocking capability of RCT. A
current pulse through the diode part of the chip turns off ReT. The
arrangement of ASCR and diode in a single device reduces the heat sink size
and leads to compactness of the converter, The undesirable stray loop
inductance between ASCR and diode is also eliminated and unwanted reverse
voltage transients across ASCR are avoided ; this leads to better tum off
behaviour of RCT. RCTs with 2000 V and 500 A ratings are available. For
high-performance inverter and chopper circuits, RCTs can now be
4.11 .9. Other Thyristor Devices

Fig. 4.56.

thyris tor

Gate-assisted turn-off thyristor (GAT) is. a normal four-layer thyristor, but its turn-off is
achieved by applying a negative gate drive across gate-cathode terminals. In order to reduce
the turn -off time appreciably, the gate-cathode junction is highly interdigitated so that stored
charges can be removed more effectively from the base region . GAT thyristors are extensively
employed in TV deflection circuits at fr equencies around 20 kHz with turn -off times as low as
2.5j.l sec for 200-V devices.
Gate-turn-off thyristor (GTO) and static induction thyristor are described in the next two
articles. The latest semiconductor device to enter the family of thyristors is integrated-gate
commutated thyristor (IGCT).
IG CT is basically a hard-switched GTO. IGCT with 4500 V, 3000 A ratings are available.
Its adv:lntages over GTO are (i ) lower conduction drop . (ii) faster switching speed, (iii)
monolithic by-pass diode, (iu) snubberless operation and (u) ease of series operation \121.

Conv entional thyristors (CTs) are nearly ideal switches for their use in power-electronic
applica tions . These can easily be turned on by positive gate current_Once in the on-state, gate
105:5 control. CTs can now be turned off by expensive and buLlty commut ation circuitry. This
shor:coming of thyristors limit their use up to about 1 kHz applic ation:;. These drawbacks in
thyris tors has led to the development of GTOs.
A GTO is a more versatile power-semicon ductor device. It is like a. CT but with added
features in it. A GTO can easily be turned off by a negative gate pulse of appropri ate amplitude.
Thus, a G'T O is a pn pn device th3.t can be turned-on by a positive ga.te cu rrent and turned-off
by n nega tive gat e current at iw gate cathode terminals.


[Art. 4.12]


Self-turn off capability of GTO makes it the .most suitable device for inverter and chopper
4.12.1. Basic Structure
A GTO is pn pn, there terminal device with anode (A): cathode (K) and gage (G), Fig. 4.57
(a ). The four layers are p np+ n as shown. In CT, anode consists of p. layer, but in a GTO,
anode is made up of n type fingers diffused into p . layer.
Fig. 4.57 (e) gives two alternate circuit symbols for GTO. Since GTO is a four layer
pn pn device just like CT, it can also be modelled by two--transistor analogy as shown in Fig.
4 .57 (b ). The four layer s have different doping levels indicated by p +np+n.+ . Transistor Q 1 is
p +np type and transistor Q2 is np+n+ type, with p. emitter of Q1 as anode A and n+ emitter of
Q2 as cathode K. 1'urnon Process. A GTO is turned on by applying a positive gate current Ig in

the reference direction shown in Fig. 4.57 (b). As GT O is forward biased, regeneration process
starts as in a CT. Current gains ai' <Lz begin to r:1se and when a 1 + a2 = 1, saturation lev el is
r eached an d GTO is turned on. The anode current la is then limited by load impedance.
Anode (A)


n' l " 1n'1 "

~ rn'l




P'r 0,



Gate (6)



Ca thode (K)

p+ l.....J



Fig. 4.57. Gate turn-off thyristor (a) basic structur e

(b) two-transistor analogy and (e) circuit symbol. Turnoff process

The turn-off process in GTO is quite different from th at in
a CT. The two-transistor model is analysed for understanding
the tu rn- off process in a GTO. For co nvenien ce , the
two-transistor E'odel of GTO is redrawn in Fig. 4.58.
From Eq. (2.7 ),
I c, = p,. I.,



From Eq. [2.6).


= P, . Ial

let = Cll l EI

I C2 = ~. IE2

As stated before, for initiating the turn-off process in a GTO,

a nega tive gate current Ig' is applied across gate-cathcde
terminals as shown in Fig. 4.58.

Fig. 4.53. Two-transistor model

fo r GTO with negoative gate
current 1/


Power Electronics

[Act. 4.12J

Now KCL at anode M in Fig. 4.58 gives

l ei-I; -182= 0
IB2=Icl-Ig' =al I,, -I;/


... (4.20 )

Fig. 4.58 also r eve als that f a = l eI + IC2

... (4.21 )


When saturation in Q2 has occurred, 182 = ~, For initiating the turn-ofT process, Qz must

be brought out of saturation. This can be accomplished only if 182 is made less than IC2 /~2' So,
when [82 < (Ic2/~2)' Q2 would shift to active region and regener ative action would eventually
turn-olfthe GTO.

:. For turnmg off of Q 2 (or GTO),




Substituting the value of 182 from Eq. (4.20) and IC2 from E q. (4.21) we get

al Ia - I,' < ; 2 (1 - (Xl)



(1 - 0. 1)





Substitution of


I g' > I , [ a, +: : - 1


... (4 .22)

_ In order that gate current I ,' for turning-off GTO is low, CX:! should be made as near to unity
as possible whereas 0. 1 sh ould be made small.
Th e turn -off gain is defIned as the
turn- off the GTO_

~Qii = I;

:. Turn -off gain.

The turn -off action in GTO

rati c~f


anode current I" to gate current Ig' need ed to


= 01 + O:z -

... (4.23)

now be explained as under:

F ig. 4.58 shows that Ia =IJt and Eq. (4.22) gives Ii' more than I!. So when negative gate
current la' flows be t.,... een gate-cathode terminals, net base current (182 - 19') is reversed, exc ess

carriers are drawn from base p ~ region of Q2 and collector current l e1 of Q1 is diverted into :he


[Art. 4.11J


external gate circuit. This r emoves base drive of transistor Q2' This further rem oves ba3e
current I Bl of transistor Q1 and the GTO is eventually turned off.
(ii) As stated above, a low value of negative gate current requires low value of u i and high

value_of U 2 '
Low value of current gain a l of Q 1 can be achieved (a ) by diffusing gold or other heavy
metal n base of Q 1 transistor (6) or by introducing short-circuiting n+ fingers in the anode p.
layer as shown in Fig. 4.57 (a), (c) or by a combination of both the techniques li sted in (a) and
(6) here. Techniques (a) and (6) are described below.
(1) Gold -doped GTO. A go ld-doped GTO retains its reverse blocking capability.
Gold-doping also reduces turn-off time, therefore, these GTOs
are suitable for high-frequency operation. However, gold-doped
p' 1
GTOs suffer from more on-state voltage drop for a given current
than a similar CT.
,,AnOC;? s ho rt
(ID Anod e-sh orted GT O. The short-circuiting fingers, also
called anode-shorts, leads to short-circuit of the emitter P+ (anode
A) with base n of Ql transistor as shown in Fig. 4.59: For anode
or emitter current la' effective emitt!3r current 11 is reduced
because of anode short. This further decreases collector current
Y' 11
I C I ' Therefore, effective current gain a l of QI' now given by c.o---J._--:,c:_-{~ ~_1
l Cllla gets r educed. So by anode-short structure, Ct l is reduced
.. '
but ~ remains unchanged as desired.
Anode-short, howeve r, reduces r everse voltage blocking
Fig. 4.59. Tw o-trrm.~i s tor
capability. With reverse biased GTO,junction that blocks' reverse
model of GTO with
voltage is J 3 only. Junction J 1 blocks no voltage because of n+
nnode-s ho n _


fingers in between p + anode. As J 3 junction has large doped layers

p., n+ on its two sides, J 3 has lower breakdown voltage, of the order of about 20 to 30 V.
The above is summarised below:
Gold-doped GTO

Anode-shorted GTO


More on-state voltage drop


Low on-state voltage drop


High re .... erse-voltage blocking capability


Low reverse-voltage blocking capability.


Suitabl e for high-frequency operation


Suitable for low-frequ ency operation

4.12.2. Static IY Characteri stics

The static I-V characteristics of a OTO is identical with that of a conventional thyristor.
Latching current for GTO is, h owever, several amperes , S'-';": 2A, as compared to 100 -500 rnA
for a conventional thyristor of the same rating. If gate current is not able to turn on the GTO ,
it behaves like a high-voltage, low gain transistor with considerable anode current. This le ads
to a n oticeable power loss under such conditions.
In the reverse mode, revers e-voltage blocking capabil ity of GTO is low, typically 20 to 30
V, because of (i) anod e shorts and (ii ) large doping densities on both sides of rev erse blockingjunction J 3 , Fig. 4 .57 (a).
4.12 .3. Switching Performance
A basic gate drive circui t for a GTO is shmv n in Fig. 4.6 0 (0) . For turn ing on a GT O, fir s~
transistor TRI is turned on, this in turn s . ."i tches on TR2 to apply a positive gat e-curr ent pul.:5e


Power Electronics

lArt. 4.12]

to tum on GTO. For turning off the GTO, the tum-off circuit should be capable of outputting a
high peak current. Usually, a thyristor is used for this purpose. In Fig. 4.60 (a), turn-off process
is initiated by gating thyristor Tl. When Tl is turned on, a large negative gate current pulse
tUrns off the GTO. Gate turn-on. The turn-on process in n GTO is similar to that of a conventional
thyristor. Gate turn-on time for GTO is made up of delay time, rise time and spread time like
a CT. Further, tum-on time in a GTO can be decreased by increasing its forward gate current
as in a thyristor.
In Fig. 4.60 (b), a steep-fronted gate pulse is applied to tum-on GTO. Gate drive can be
removed once anode current exceeds latching current. However, some manufacturers advise
that even after GTO is on, a continuous gate current, called back porch current 19b as shown ,
should be applied during the entire on-period of GTO. The aim of this recommendation is to
avoid any possibility of unwanted turn-off of the GTO.


Spl~~ vo!tag~

Tail current





Fig. 4.60. Gate turn-off thyristor

... t .. P



basic gate drive circuit and


switching characteristics. Gate turn-off. The tum-off characteristics of a GTO are different from those of
an SCR. Before the initiation of turn-off process, a GTO carries a steady current la. Fig. 4.60
(b) . This figure shows .a typical dynamic turn-off characteristic for a GTO. The total turn-off
'time tq is subdivided into three different periods; namely the storage period ,(t,), the fall period
.(t;) and the tail period (t f ). In other words,

tq =t, + t;+ t/
Initiation of turn-off process starts as soon as negative gate current begins to flow after
t = 0 at instant A. The rate of rise of this gate current depends upon the gate circuit inductance
L and th e gate voltage applied. During the storage period, anode current 1a and anode voltar.e
(equal to on-state voltage drop) r emain constant. Termination of the storage period is indicated
by a fall in If! and rise in Va' During t" excess charges, i.e. holes, in p '" base are removed by
negative gate current and the centre junction comes out of saturation. In other words, during
storage time t., the negative gate current rises to a particular value and prepares the GTO for
turning-off (or commutation) by flushing out the stored carriers. After ts' anode current begins
to fall r apidly and anode voltage starts rising. As shown in Fig. 4.60 (b), the an od e current falls
to a certain value and then abruptly changes its rate of fall. This interval during which anode
current fall s rapidly is the fall time t'l Fig. 4.60 (b ) and is of the order of 1 ~sec {4J. The fall

[Ml. 4.13J



period l; is measured from the instant gate current is maximum negative ta the instant anode
current fall s ta its tail current.
At the time t = l .. + l ,. there is a spike in voltage due ta abrupt change in anode current. After
t,-. anode current ig and anode volta.ge Ug keep moving towards their turn-off values for a time
t/ called tail lime. After t,. anode current reaches zer o value and Ug undergoes a transient
overshoot due to the prese nce af RJ , C.. and then stabilizes to its off-state value equal to the
source va ltage applied to the anode circuit. Here R J and C.. are the snubber circuit parameters.
The turn -ofT process is complete when tail current reaches zero. The over shoot voltage and tail
current can be decreased by increasing the size of C but a compromise with snubber loss must
be made . The duration oft, depends upon the device" chara cteristics [4] .

4.12.4. Comparison b etween GTO and thyristo r

A GTO has the following disadvantages as comp ar ed to a conventional thyristor :
(i ) Magnitude oflatching and holding currents is mor e in a GTO.
(ii ) On state voltage drop and the associated loss is more in a GTO.
(iii) Due to th e multicathode structure of GTO, triggering gate current is higher than
that required for a conventional SCR.
(iu ) Gate drive circuit losses are mor e.
(v ) Its r everse-voltage blocking capability is les s than its forward voltage blocking
capability. But this is no disadvantage s o far as in~erier and chopper circuits ar e
concerned .
In spite of all these demerits, GTO has the fon owing advantages over an SCR :

(ii )
(i v )
( v)


GTO has faster switching speed.

Its surge current capability is comparable with an S CR.
It has more dil dt rating at turn-on .
GTO circuit configuration has lower size and weight as campared to thyristor circuit
GTO unit has higher efficiency because an increase in gatedrive power loss and
on-state loss is more than compensated by the elimination of forced -commutation
GTO unit has reduced acoustical and electromagnetic n oise due to elimination of
commutation ch okes.

4 .1 2 .5 . Application of GTOs

In view of the above facts, GTO device are now being used in (a) highperfonn anc e drive
systems, s uch as the field-oriented control sc heme used in rolling mills , robotics and machine
tooh [41. ( b) tracti on purposes because oi their lighter weight and (e ) adju 5 ~3.bl e-fr2:que ::-lo)
inverter dr ives . At present, GTOs with r atings ':1P to 5000 V and 3000 A are available .



A static in ducti on thyristor, or 81TH, is a three terminal sel f controlled devjce just like a
GT O. It wa s co mmercially introdu ced in Japan in 1988. A si mil a r de-.-i: e, kn own as
field controlled thyristor (FCT), or fi eld- controlled diode (FCD ), was de'/ elopea ea rEe r by
Gene ral Elec tric but could not be comm ercially launched . H owever, commercial use 0[5I1 H is
being promoted by Japanese unive rsities and industries .


Power Electronics

[.-\ rt . 13)

4. 13.1. Basic Structure

The basic structure of 81TH is shown in Fig. 4.61 (a) and the device in Fig. 4.61 (b ). It is
primari ly ap Pnn- diode, withp+ grid-like gate electrodes buried in n layer. Comparison of Fig.
2.23 (a ) and Fig. 4.61 (a) reveals that device structure of SITH is ana1ogous to SIT except that
a p. layer is added on the anode side. In addition, n" type ftngers are diffused in p'" anode layer
just as in a GTO.


(b )


Fig. 4.6 1. Static induction thyristor

(a )

ba!ic structure and (b ) device symbol.

4.13.2. Turn-on and Turn-off processes

A sim plified structure shown in Fig. 4.62 is employed for explaining the turn on and the
turn-off processes in a 81TH. When anode is forward biased with gatecathode voltage Vg equal
to zero, the device behaves like a diode. Load current i(J flows from anode to cathode as p"n
junction is forward biased, Fig. 4.62 (a). This shows that 81TH is a normally-on device just like
SIT. When gate is reverse biased with respect to cathode, i,e. when Vg is negative, a depletion
layer ..."ould be formed as shown in Fig. 4.62 (b), This dElpletion layer blocks the flow of anode
current. With varying negative gate bias, the magnitude of anode current can be controlled.


. ~

v, L

L-==---iJ Anoae
(a )


Fig. 4.62. 81TH (a ) on-conditio n when gate voltage V, is zero and


off-condition when


is nega tive.





If 51TH is reverse biased, with cathode positive and anode n egative, electrons can flow from
anod e, intermixed n'" layer, n , through p + grid, n '" and finally to cathod, Fi g. 4.6 1 (a). Th us,
reverse current from cathode to anode can exist when 51TH is r everse biased. Thi s shows that
81TH does not have any reverse blocking capabi lity due to emitter-shorting (p+ lay er
interdigitated with n+ layers at the anode ).

4.13.3. Application of 81TH and comparison with GTO

At present, SITHs with 2500 V/500 A ratings and 100 kHz operating frequency are
available. 8ITHs with higher power ratings and with normally-off characteristics are likely to
be developed in the near future. Their use in induction heating, high fr equency link dc-ac
converters and HYDe converters is being prom oted by Japanese organisations.
When compared with a GTO, a 81TH (i) is normally-on device unlike GTO (ii) has higher
conduction drop (iii ) has lower turn-off current gain, typically 1 to 3 instead of 4 to 5 for GTO
(iu) has higher switching frequency (u) the du l dt and dtldt ratings are higher and (ui) has more
SO A, (s afe operating area).
An SCR can be switched from off-state to on-state in seve ral ways; these are
forward-v oltage triggering, du l dt triggering, temperature triggering, light triggering and gate
triggering, see Art. 4.2. The instant ofturning on the SCR cannot be controlle~ by the first three
meth ods listed above. Light triggering is used in some applications , particularly in a
s er ies-connected string. Gate triggering is, however, the most common method ofturn ing on the
SCRs, because this method lends itself accurately for turning on the 5CR at the desired instant
of time. In addition, gate triggering is an effic ient and reliable method. In th is secti on, firing
ci rcuits for thyristors are studied in detaiL


4.14.1. Main Featu res of Firing Circui ts

As stated above, the most common method for controlling the onset of conduc tion in an SCR
is by means of gate voltage controL The gate control circuit is also called firing, or triggering,
circuit. These gating circuits are usually low-power electronic circuits. A firing circuit should
fulfil th e follOwing two functi ons .
Shi eldll:d ~






amplif i er




Pu ! s~
trons to rm~ r


transtorme r

Pu isil
tr ansfor mer



DC power

Contro l CirC 1Ji!

Dr iV'lf ci rcuit

Fi g. 4.63. A ge::e: al lnyou t of t~ e firi ng circuit scheme fo r SCR.,.


PO'Nllr (: jrcuit

Power Electronics

[An. 4.141


(i) If power

circuit has more than one SCR, the flring circuit should produce gating pulses
for each SCR at the desired instant for proper operation of the power circuit. These pulses must
be periodic in nature and the sequence of firing must correspond with the type of thyristorised
power controller. For example, in a single-phase semiconverter using two SeRs, the triggering
circuit must produce one firing pulse in each balf cycle; in a 3-phase full converter using SLX
SeRs, gating circuit must produce one trigger pulse after every 60 interval.
(ii) The control signal generated by a firing circuit may not be able to turn-on an SCR It is
therefore comm on to feed the voltage pulses to a driver circuit and then to gate-cathode circuit.
A driver circuit consists of a pulse amplifier and a pulse transfonner.

A firing cir cuit scheme, in general, consists of the components shown in Fig. 4.63. A
r egulated dc power supply is obtained from an alternating voltage source. Pulse generator,
supplied from both ac and dc sources, gives out voltage pulses which are then fed to pulse
amplifier for their amplification. Shielded cables transmit the amplified pulses to pulse
transform ers . The function of pulse transformer is to isolate the low-voltage gate-cathode
circuit from the' high-voltage anode-cathode circuit. Some firing circuit schemes are described
in this section .

4.14.2. Resistance and Resistance-Capacitance Firing Circuits

Rand RC firing circuits are not in commercial use these days. These are presented here for
th e sake of highlighting the basic principles of triggering the SeRs. They otTer simple and
economical firing circuits (3J.

Resistance firing circuits. As statad above, resistance bigger circuits are the simplest

and most economical. They however, suffer from a

limited range of firing angle control (O~ to 90), great
de p end en ce on temperature and difference in
performance between individual SeRs.
Fig. 4.64 shows the most basic resistance biggering
circuit. R2 is th e variable r esistance, R is the stabilizing
resistance. In case R2 is zero, gate current may flow from
source, through load, R 1, D and gate to cathode. This
current should not exceed maximum pe"rmissible gate
current 19m. R 1 can therefore, be found from the relation,


It is thus seen that func tion of Rl is to limit the

t c. ~


rv "'s= Vmsinwt






R , S Igm or R , ~ Igm
... (4.24 0 )
V", = maximum val ue of sourc e volta ge

gate curr ent



Fig. 4.64. Resistance firing circuit.

safe val ue as R2 is varied .

Resis ta nce R should have such a value that maximum voltage drop a cross it does not
exceed maxi mum possible gate voltage V,Ir.. This can happen only when R2 is zero. Under this
condi tion,
R S V,flll

R < Vi m' R l

- V It!

VJ m

..'< 4.24 0)


(Art. .U4]


As resistances R t , R2 are large, gate trigger circuit draws a small current. Diode D all ows

the flow of current during positive half cycle only, i.e. gate voltage
amplitude of this dc pulse can be controlled by varying R 2.

u, is half-wave dc pulse. The

The potentiometer setting R2 determines the gate voltage amplitude, When R2 is large,
current,i is small and the voltage across R, i.e. tlg = iR is also small as shown in Fig. 4.65 (0). As
V,p (peak of gate voltage u,) is less than Vgl (gate trigger voltage), SCR will not turn on.
Ther efore, load voltage tlo = 0, io =0 and supply voltage U~ appears as UT across SCR as shown
in Fig. 4.65 (a>" . Note that trigger circuit consists of resistances on ly, II, is therefore in phase
with source voltage u~ . In Fig. 4.65 (b ), R2 is adjusted such that V,p =V,I. This gives the value of




Vgt ;





. - . .;-.--: . - . ......:... j. - ~

V~ i




v" ., v,. w'


: wI'




:, wt

i,t N




: I

: wi

: I







210' ;




: WI

: wi





Fig. 4.65. Resistance firing of an SeR in a half-wave circuit with dc load
(a) No triggering of SCR (b) a = 90 (c) a. < 90~.

firing angle as 90 0 The various current and voltage waveforms are shown in Fig. 4,65 (b ). In
Fig. 4.65 (c ), Vgp > V~. As soon as va becomes equal t o V,! for the first time SCR is turned on,
The resistance triggerin g cannct give firing angle beyond 90 0 Increasing v, above Vg l turns on
the SCR at firing angl es less than 900 \"Vhen Vg r eaches Vgl for the first ti me, SCR fires, gate
loses control and v, is reduced to almos t zero (about 1 V) value as shown. It may also be seen
that firing a1,;{le can never be equal to zero degree however large VzP may be ; it can, of course,
be brought nearer (2 0 _4) to zero degree firing angle. A relationsh.ip between pea1t gate voltage
VzP and gate trigger vol tage VGt may be expressed as follows:
V gp sin

= V,I


=sin- 1 (Vj IIVgp)

Some s~ud'!n !.J argue th3t in every posi:l're cycle of source, died e ci rcui~ will :e act ive l1:ld will therefore dro'l!
curren: [roo. s:lurce. The current will ca u.se '/olt:!g! d!1)p liD acreS! load iL-"~ .::erefore, v" ..nd i .. should be
shown in. F i~. ".5~ (a l. Ac~ually, load resilta."lce (3 few I,lh::!.s) in ccr::::Jar:so:1. w:~h R\ ... R:l " R (km is quite
,:nail. Ther'! ;",,!e, current d uring the posi tive cycle of source is negligibly S;';"I:lll ;:!.::d li!t'!wise ')~ across the load.

Power Elect ron ics

[Art. 4.14]



vgp = 'R;-,-+-"
Vrn R
ROc,-+- R-;<

. _, [VI' .IR,VR
+ R, + R


As V,I,R 1, R and Vm are fixed, a"'" sin- 1 (R 2) or a"", R 2.

This shows that firing angle is pr oportional to R 2. As R 2 is increased from small value
(i.e. small a), firing angle increases. In any case, a can never be more than 90.
As the firing angle control is from 00 (approximately) to 90 0 , the halfwave power out put
can be controlled from 100% (for a = 00 ) down to 50% (for a = 900 ).

Exampl e 4.23 . Discuss what would happen to the circuit of Fig. 4.64 in case load is shifted
between terminals a and b.
Solu tion. In the circuit of Fig. 4.64 , when SCR is on, voltage vT across it is almost zero
(actually about 1 to 1.15 V) and therefor e voltage across R I , R 2, D, R is a lso nearly zero. As a
result, trigger supply voltage v, is reduced to zero after SCR turn-on. There is thus hardly any
gate current and the associated gate power loss is zero during the time SCR is cond ucting in
Fig. 4.64 .
In case load is shifted between terminals a and b, the circuit may still operate. But after
SCR turn-on, the circuit comprising of R 1, R 2, D and gate to cathode would be subjected to
source voltage. This would cause an increased gate current and the associated gate power loss
would be mo re during SCR turn on . Such an happening would certainly burn out the gate
circui t and destroy the SCR. This shows that load should never be connected between t erminals
a and b in Fig. 4.64.
(b) R C firi ng circuits. The limited range of firing angle contr ol by resistance firing circuit
can be overcome by RC firing circuit. There a re several variations ofRC trigger circuits. Here
only two of them are presented.

RC half-wave trigger circuit. Fig. 4.66 illustr ates RC half-wave trigger circuit. By
varying the value of R, firing angle can be controlled fr om 00 to 1800 In the negative half cycle,

capac itor C ch arges through D2 with lower plate positive to the p eak supply voltage V m at
wt = - 90. After wt = - 90 0 , source voltage v, decreases from - Vm at wt = - 90 0 to zero at
wt = 0. During this period, capacitor voltage Uc may fall from - Vm at wt = - 90 to some lower
value - 0 a at wt = 0" as shown in Fig. 4.67. Now, as
the SCR anode voltag e passes through zero and
becomes positive, C begins to charge through
LO A 0
variable resistance R from the initial volt age - oa.
\-Vhen capacitor charges to positive voltage ~ual to
gate trigger voltage VS/I SCR is fired and after this,
capacitor holds to a small positive voltage, F ig. 4.67. rv vs =Vmslnwt
Diode D1 is used to prevent the breakdown of
ca thode to gate junction through D2 during the
negative half cycle . .'\n examination of Fig. 4.67
re veals that firing a ngle can never be zero and
Fig. 4.66. RC half-wave trigger circui t.

VC f C




In the range of power freque ncies , it may be empirically shown [3J that RC for zero output
voltage is given by

1.3 T
RC >
- 2 -w



... (4.25)

= period orae line frequency in seconds.

The SCR will tri gger when u~ =Vgt + ud' where ud is the voltage drop across diode D1. At the
instant of triggering, if Uc is assumed constant, the current I, t must be supplied by voltage
source through R, D1 and gate to cathode circuit. Hence the maximum value of R is given by
u,:::?: RIgt+u c
u~ ~RI,t + Vgt +ud
u.- V,t - ud


... (4.26)


where Us is the source voltage at which thyristor turns on. Approximate values of Rand C
can be obtained from Eqs. (4.25 ) and (4.26).

, ,




.-' 0:





: 37f



", 1



I I,







Fig. 4.67. Waveiorms for RC baliwave trigger circuit of

Fig. 4.66 (a) high value of R (b) low value of R. '
When SCR triggers, voltage drop across it falls to 1 to 1.5 V. This, in turn, lowers the voltage
across R and C to this low value of 1 to 1.5 V. Low voltage ;:,.cross S CR during conduction period
keeps C discharged in positive half cycle until negative voltage cycle across C appears. Tl-!is
charges C to maxi mum negative voltage - Vn: as shown in Fig. 4.67 by dotted line. In Fig. 4.6 7
(a ), R is more, the time taken for C to charge from - 0 0 to (V,t + ud ) == V.ft is more, firing angle is
more and therefore average output voltage is 10..... In Fig. 4.67 (b). R is less, firing an.gi e is low
and therefore ave rage output voltage is more .
(iO RC full wa ve trigger cirCtlit. A simple RC trigger circuit giving fuI1 wave ou:rut voltage
is shown in Fig. 4.68. Diodes DI- D4 fOlm a fu1l "N!.v e diode bridge. In this cireui:, t he initial


[Mt. 4.14J

Power Electron ics

voltage from which the capacitor C charges is almost zero. The ca pacitor C is set to this low
positive voltage (upper plate positive) by the clamping action of SCR gate. When capac itor
char ges to a voltage equal to Vgt, SCR triggers and rectified voltage ud appears across load as
uo. Th e value of RC is ca l culated by the
empirical r elation [31,


~ 50 T2 "




As per Eq. (4.26), the value of R is given by

U -



where UJ is the source voltage at which thyristor turns on. In Fig. 4.69 (a), firing angle a is
more than 90 0 and in Fig. 4.69 (b) , a < 90 0


vs=Vm sin wt


Fig. 4.68. RC fullwave trigger circuit.

4.14.3. Unijunction Tran sistor (UJT)

Resistance and RC triggering cir cuits described above giv e prolonged pulses. As a result ,
power dissipation in the gate circuit is large. At the same time , R and RC triggering circuits
cannot be used for automatic or feedback control systems. These difficulties can be overcome by
the use of UJT triggering circuits.
Pulse triggering is preferred as it offers several merits over R and RC triggering. Gate
characteristics have a wide spread, Fig. 4.9. Pulses can be adjusted easily to suit such a wide
spectrum of gate characteristics. The power level in pulse triggering is low as the gate drive is
discontinuous, pulse triggering is therefor e more effici ent. As pulses with higher gate current
are pennissible, pulse firing is more reliable and faster. In this section, first U lT is described
along .....ith its charact eristics and then its use as a r elaxation oscillato r for t riggering SCRs is


Vm sinwt



.'*- - ~~
'.. _.*.. - . .... ; ,

" '' I "e ', I. v.gt "e :!

, I


' La: : k..cc


: .

Vjvj l
(a )




j " I
I ' i





"1N N N J etr"'

w t~







(b )

Fig. 4.69. Waveforms for RC halfwave trigger ci r C~lit of

Fig. 4.68 (0 ) high valu e of R (b ) low value of R.

I wi


[Art. 4.U ]


An UJT is made up of an n-type silicon bas e to which p-type emitter is embedded , Fig.
4 .70 (a). The n-type base is lightly doped whereas p-type is heavily doped. The two ohmi c
con tacts provided at each end are called base-on e B1 and base-two B 2 So, an UJT has three
terminals, n amely the emitter E, base-one Bl and base-two B 2 . Between bases B1 ano. 8 2 , the
unijunction behaves like an ordinary r esistance. RBI and R B2 are the internal resistances
respectively fr om bases Bl and B2 to eta point A, Fig. 4.70 (a). Its symbolic representation is
given in Fig. 4.70 ( b) and its equivalent circuit in Fig. 4.70 (c).
When a voltage V BS is applied across the two base tenninals B I and B 2, the potzntial of
point A with respect to 8 1 is given by
V AB l



=R 81 + R B2 . RB i = R 81 + R B2 . V 8S =n VBB

where n = R


B~ is called the intrinsic stand-off ratio, Typical values ofn are 0.51 to 0.82.
+ B2

Interbase resistance RBB = RBI +- RB2 is of the or der of 5-10 kn .

Thi s resistance Ras can easily be measured by a multimeter with emitter open, As stated
before, RBB is brok en up into two r esistances, RBI between emitter and base B 1 and RB2 in
between emitter and base B 2 . Since emitter is nearer to 8 2 , resistance RB2 is less than the
resistance RBI '




"to -~ in~



QS 1.













"~ 'IB8


F ig. 4.70 . (a) Basic structure ofUJT (b) symboli c representation and (c) its equi"alent circu it.

The operation ofUJT can be und erstood with its equivalent circuit of Fi g. 4.70 (c). A:, UJT
is usually operated with both B2 and E biased positive with respect to reference base terminal
B!. DC voltage source VB8 between B2 and Bl is constant . DC source v-EE in ser ies with
resistance RE is considered as l?put to tl~e UJT. Both V BB and VEE are shown in Fig. ~71 (a )
where UJT equivalent circuit is sh own inside the dotted rectangle. As before

=VA = R Bl + R B2 . RBl =11 V SB

Th e magnitude of voltage V, can be varied by regulating external r esis tancz R E . _!.5 long a.5
emitte r voltage Vr < ll. V BB , the E - Bl Wlijunction (or p - n junctior!) is reverse biased and
emit ter current 1, is negative as shown by curve PS in Fig. 4.il (b) , The region PS of very low
current is tr eat ed as 'off' state of UJT. The resistance behveen E - B I j unction is therefore \er.y
high. At ;lOint S, 1, = 0, drop across R z is zero, th erefore 'llf = sour ce volta;;e, i. e. OS = V~ = VEE '


[A ,t. 4.14]

Power Electronics

Actually, offsta te ofUJT extends to a point where emitter voltage V, exceeds VA' or n .V!JB' by
diode voltage VD in E - BI junction. So when V, =fl . VBB + VD point B is reached and E - B I
junction gets forward biased to allow forward current through the diode. Here V D is the forward
voltage drop across E - B I junction (usually 0.5 V).
Point B is called the peak point . Voltage Vp and current Ip pertaining to point B are called
peakpoint lIoltage and peakpoint current respectively. By varying R E Vr is increased till V,
approaches Vp . At this peak point, V, =Vp = fl .VBB + YD. thep emitter begins to inject h oles from
the heavily doped emitter E into the lower base region B t . As n type base is lightly doped , the
holes rar ely get any chance to recombine. The lower base region B I is, therefore, filled up v.;th
additional current carriers (holes). As a r esult, resistance RB l of E - B1 junction decreases. The
fall in R BI causes potential of eta point A to drop.





Saturation rf9!On ~


eqlJlvclent ---.:

re sistance

Cut olf

lood Une ;
'~B(peakpoin l )

( nV

.~ S8

+v )

1 ::

VOl' , .:::. T - -~'I



'~ ~

RI load line
''':'' J

.. .. .::.




:.~ ~

:L__ _ . _ ____
BI _.~:

P 0 101


.. .. xl

volley point


Fig. 4.71. UJT




L. V ~::::"C, __

..... __ _ ....1 VI
. :



equivalent circuit with Vss and VEE , and

(b) typical static V - / characteristics.

This drop in VA' in turn, causes Vf (= VA + V D) to fan . As VEE is constant, fall in Ve gives rise
to more emitter current Ie (= (VEE - Ve) / RE). This increased Ie inj ects more holes into region
B 1 the reby further reducing the resistance RBI and so on. This regeneratiue or snow balling
effect continues till RBi has dropped to a small value (from about 4 kn to around 2 to 25 !l). The
emitter current, limited by external r esistance R E is then giv en by




'Nhen RBI has dropped to a very small value, indicated by point C in Fig. 4.71 (b ), the UJT
has r eached 'on' state. At point C, entire base region Bl is saturated and resistance RBI cannot
decr ease any more. This point C is called the valley point ; Vu and Iu are the corresponding
emitter potential and curr ent. After UJT is on, or after valley point is r eached, an increase in
Ve is accompanied by an increas e in I f; this is indicated by curve CQ. At point Q, Ve is a little
more than its valley poin t voltage Vu. Between points B and C, em itter voltage Vr falls as It
incre ases; UJT, therefore. exhibits negative resistance between these two points. The negative



[Art. 4. 141

resistance region between peak and valley points in Fig. 4.71 (b) gives UJT the switch ing
characte ristics for use in SCR triggering circuits.
At the valley point, the current is given by VuIRB l' Valley-point current , aLso called holding
current, keeps UJT on. When emitter current It falls below l u, UJT turns off.
UJT oscillator triggering. The unijunction transistor is a highly efficient switch ; it5
sw itching tim e is in the range of nanosecond s. Since UJT exhibits negati v~ resistance
characteri stics, it can be used as a relaxation oscillator. Fig. 4.72 (a ) shows a cir cuit diagram
with UJT work1ng in the oscillator mode. The external resistances R 1. R 2 ar e smaH in
comparison with the internal resistances RBI' RB2 of UJT bases. The charging resistance R
should be such that its load line intersects the device characteristics only in th e negati ve
res istance region.
In Fig. 4.72 (a), when source voltage VBB is applied. capacitor C begins to charge through
R exponentially towards '\lBB ' During this charging, emitter circuit of UJT is an open circuit.
Th e capacitor voltage Uel equal to emitter voltage u t is given by


The tim e constan t of the charge circuit is


t l =


When this emitter voltage u t (or IJ e) r eaches the peak-point voltage Vp (= 11 V BS + VD)' the
unijunction between E - B I breaks down. As a result, UJT turns on and capacitor C rapidly
discharges through low resistance RI with a time constant t 2 = R IC. Here t a is much smaller
than tl' When the emitter voltage decays to the valley-point voltage Vu, emitte r current
(V1/(R B1 + R 1 falls below I I) and UJT turns off. The time T required for capacitor C to charge
from initial voltage VI) to peak-point voltage Vp through large r esistance R , can be obtained as
under :


... (4.28 )












Fig. 4.72. UJT oscillator

(a )

Connection diagram and


Vol tage wave rorms.


[Art. 4.14)

Power Electronics

In cas e T is taken as the time period of output pulse duration (neglecting small discharge
time ), then the value of firing angle a 1 is given by

a, = roT = wRC In

I -~

... (4.29 )

where w is the angular frequency of UJT oscillator.

The amplitude of pulse voltage is obtained by drawing a load line B b for Rl 'a s shown in
Fig. 4.71 (b). The vertical projection of Bb, equal to xy, gives the voltage pulse amplitude. With
the discharge of capacitor, the operating points Band b move towards C. For pointsp and q, the
pulse amplitude is XlYl' Eventually, point C is reached at which puls e voltage is zero, then the
operating point shifts to a , Fig. 4.71 (b ). The potential of eta point A is 11Vss , but that of the
emitter is Vu which is less than 'lVBS ' As a result, E-B} unijunction is reverse biased and
ceases to conduct, the UJT turn s off and goes into blocking mode. Capacitor C now again
charges from V t = Vu to voltage 11 V Ba + VD, E - B} unijunctioll breaks down and the above cycle
If the output voltage pulses are used for triggering an SCR, resistance R} should be
su ffi ciently s mall so that normal leakage current drop across R IO when UJT is off, is not able to
tri gger the S CR. In other words,

Vas ' R l
R < SCR trigger voltage V,t




R Bs = R st + R'B2
The emitter-diode forward characteristics vary with temperature in such a manner that
VD decre ases and Ras increases with temperature. In order to provide compensation against
this thermal effect, the value of R2 used in Fig. 4.72 should be calculated from the relation


R, =-TjV

... (4.30)


- ..

The width of triggering pulse is sometimes taken equal to R }C.

In case load line forR intersects the UJT characteristics in the region CQ, Fig. 4.71 (b), the
intersecting point will result in stable operating point and the circuit then cannot work as an
oscillator. This fact fixes the maximum and mini~um values of charging resistor R and the
osc illator output frequency.
Th e maximum value of R is determined by the peak-point values Vp and Ip- When voltage
across C r each es V p the voltage acrossR is V BS - Vr
_ V BB

R n fU -

Vp _

V-SB -

(TlVSB + VD)

... (4 .3Ia)

The mimmum value of R, governed by valley-point values VII and I I,; is given by

R o~

V ss - Vu

... (4.31 b )


E xamp l e -1.2 5. A rela.xa tion oscillator using an UJT, Fig. 4. 72 (a), is to be designed fo r
triggering an S CR. The UJT has th e following deta :

=0.72, Ip = 0.6 rnA , Vp = 18.0

C'2rrent !dc.i. emitter open = 4.2 mA.

V, VI: = 1.0 V , I v = 2.5 TT'. A, Ra3 = 5 k n , No r m a l leak age

[ML 4.14)


The firing frequency is 2 kHz. For C = 0. 0-1



compute the ualues of R, Rl and R'!..

Solution. The valu e of charging resistor R, from Eq. (4.28), is


A s V D is not given,

C In -1=-~

(C In 1 = ~


= 9.82 kn

2000 x 0.04 In 0.; 8

Vp =T)Vss

11 ~ 0 .72 -

ss -



From Eq. (4.30),

R, = 0.72 x 25

With emitt er open.


= 555.55 n

Leakage curre nt (R 1 + R 2 +R BS )

25 3
4.2 x 10-

5000 - 555.55

n = 396.83" 397 n

Example 4.26. If the firing frequency of the SCR in Example 4.25 is changed by uarying
charging resistor R, obtain the maximum and minimum values of R and the corresponding
frequencies .
From Eq. (4.31),


R .

= 25.0 -

1.0 = 9 6 kn
2.5 X 10- 3


From Eq. (4.28),



25 (1-0.72) = 11.67kn
0.6 X 10- 3

=_1_ =_~1,---:-
R e i _1_
n 1-~
1 =1682 .8 Hz" 1.683I;Hz
11.67 x 0.04ln 0.28

fm= =

T m:u:

1 = 2045.7 Hz" 2.05 kHz
9.6 x 0.04 1n 0.28

Synchronhed UJT triggering (or Ramp triggering). A synchronized UJT trigger

circuit using an UJT is shown in Fig. 4.73. Diodes Dl - D4 rectify ac to dc. Resistor Rl lowers
Vde to a suit able value for the zener diode and UJT. Zener diode Z functions to clip the rectified
voltage to <-'itandard level Vl" which remains constant except near the Vde zero, Fig. 4.74. This
volt age V: is ap pli ed to the charging circuit RC. Current it charges c a p a~i to r C at a rate
deter mined by R. Voltage across capacitor is marked by ue in Figs. 4.73 and 4.74. "Then voltage
VI; reaches th e unijunction threshold voltage TlVr the E - Bdunction ofUJT breaks down and
the capacitor C discharges through primary of pulse tran sfor mer sending a current i',! as shown
in Fig. 4.73.
As the curren t i2 is in the form of pulse, windings of the pulse transfor mer have pulse
voltages :g thei r secondary terminal s. Pulses at the two second J. :-Y wi:-_:iings feed the same
in-pha se pu13e to two SCRs of a full- wav e circuit. SCR with positi ve anode vol tage would turn


Power Electronics

[Act. U4]




i, ~



04 ~

zr; ~




;: C


Puln Troosf.








Fig. 4.73. Synchronised UJT trigger circuit.

on . As soo n as the capacitor discharges, it starts to recharge as shown. Rate of rise of capacitor
voltage can be controlled by varying R. The firing angle can be controlled up to about 150:1. This
method of co ntrolling the output power by varying charging resistor R is called ramp control,
openloop control or manual con trol.
As the zener diode voltage V z goes to zero at the end of each half cycle, the synchronization
of the trigger cir cuit with the supply voltage across SCRs is achieved. Thus the time t, equal to
a / w, when the pulse is applied to SCR for the first time, will remain constant for the same value
of R. Small variations in the supply voltage and frequen cy are not going to effect the circuit





, :,,










,, i



",, I
I 2 ,'


Fig. 4.74 . GenerBtion of output pulses for the circuit of Fig. 4.73. Here, t = ct/w.

In case R is r educed so tha t Uc reaches UJT threshold voltage twice in each half cycle as
shown in Fig. 4.74 (b), then there will be two pulses in each half cycle. As the first pulse will be
able to turn-on the SCR, second pulse in each cycle is redundant.
Ramp-and-pedestal triggering. Ra mp and pedestal triggering is an improved version of
syn chr onized-UJT-oscillator triggering . Fig. 4.75 shows the circuit for r amp-an d-pedesta!
triggering of two SCRs conn ected in antiparalleJ for controlling power in an ac load. This trigger
ci rcuit can als o be us ed for triggeri ng the thyri3tor s in a single-phase semiconverter or a
5ingle-phase iull converter. The various voltage waveforms are shown in Fig. 4.76.
Zener diode voltage V": is constant a t ita thr es~-h old voltage. R2 acts as a potentia! divider.
Wiper of R2 controls the value of pedestal voltage Vpd. Diode D allows C to be quickly charg ed
to Vpc! through the low r es is tan ce of the upper portion of R 2. The setting of wiper on R,! is such

Thy r istors

[Acl. 4.14J




, 01



' 03




~ V R,

















Fig. 4.75. Ramp and pedestal trigger circuit for ac load.

that this value of Vpd is always less than the UJT firing point voltage 11 V:. When wiper setting
is such that V pd is small, Fig. 4.76 (al, voltage Vz charges C through R. When this ramp voltage
VI; reaches 11Vz, UJT fires and voltage u ' through the pulse transformer, is transmitted to the
gate circuits of both SCRs Tl and T2. The forward biased SCR Tl is turned on. After this, Vc
reduces to Vpd and then to zero at wt = n. As til; is more than Vpd. during 'the charging of capacitor
C through charging resistor R , diode D is r everse biased and turned off. Thus Vpd does not effect
in any way the discharge of C through UJT emitter and primary of pulse transformer. From
o to n, Tl is forward biased and is turned on. From n to 21t, T2 is forward biased and is turned
on. In this manner, load is subjected to alternating voltage tlo as shown in Fig. 4. is.

Y'~m'inw. ~
.I w'








I . Vp '


I .


I . Wi
I 1

(0 )





- ~~iz
I I W.






Y'~~~ .





'I' r -..... 1

_ ,_

,_ ... ~, ___I[ Y, V~'._.L
+ ~'


Vg Q2


I v ' J lIJl




. ,.




-- I---wr(b )

Fig. 4.76. Waveforms for ramp-andpedes tal circuit of Fig. 4.75.

Wit h the setting of wiper on R.2, pedes tal vol tage Vpd on C can be adjusted. With low
ped estal voltage across C, r amp charging ofC to Tl "llz takes longe r tim e, Fig. 4.76 (0 ) and firi ng
angle de by is therefore more and output voltage is low . With high pedestal on C, voltage- ramp


Power Electronics

(Ar!. 4.15]

charging of C through R r eaches nVz faster, firing a.Y1g1e delay is smaller, Fig. 4.76 (b) and ',
output voltage is high. This shows that output voltage is proportional to the pedestal voltage,
The time T required for the capacitor to charge from pedestal voltage Vpd to nV: can be
obtained from the relation

,= Vpd+( V, - Vpd)(l-.-TI Re )


Note th at (V:

Vpd ) is the effective voltage that chnrges C from Vpd to nVz' From above

... (4.32 )

T =RC In V, (1 _~)
and the firing angle delay


is given by
Vz - V Dd

...(4.33 )

ex, = w RC In V, (1 _ ~)

Pulse transformers are used quite often in firing circuits for SCRs and GTOs , This
transformer has usually two secondaries. The turn ratio from primary to two s~condaries is 2 :
1 : 1 or 1 : 1 : L These transformers are designed to have low winding resistance, low leakage
reactance and iaw inter-winding capacitance. The advantages of using pulse transformers in
triggering semiconductor devices are :
the isolation oflow-voltage gate circuit from h igh-voltage anode circuit and
(ii) the triggering of two or more devices from the same trigger source.
(i )

A square pulse at the primary terminals ofa pulse t ransformer may be transmitted at its

secondary terminals faithfully as a square wave or it may be transmitted as a derivative of the

input waveform. The conditions governing the operation of a pulse tranSformer in these two
functional modes are now examined,
A general layout of the trigger circuit using a pulse transformer is shown in Fig. 4 .77 (0).
Here the function of the diode is to allow the flow of current after the pulse period (i .e: when the
transistor is off) so that energy stored in the primary of pulse transformer is dissipated.
In Fig. 4.77 (a). the trans,istor is acting simply as a switch, turning on when the pulse
applied to its base is at its high level, thereby connecting the de bias VB to the transformer
primary. The advantage of this arrangement are two fold:
(0 ) There need not be a variable strength pulse generator since the pulses may be of the
same amplitude and the strength of the generated pulses may be incr eased simply by varying
the de bias voltage.
(b) The operation of the circuit becomes independent of the pulse characteristics sinc e..!he
only rol e the pulse plays is to turn-on or turnoffthe transistor. Therefore, there is no effect of
pulse dis tortion (e.g. pulse edges or any spike superimposed on the puls e) on the working of this

In Fig, 4.77 (a), RL limits t h e current in the prima ry circuit of puhe tra.'1sformer. Its
equivalent circuit is drawn in Fig, 4.77 (b ), where L is the magnetizing inducta:.lce of the pulse
tr ansformer and Rg is the resistanc e of gate-cathode ci~cuit of an SCR. Fig. 4.77 (c ) 5hoW3 the
transfer of R, to pulse transformer prim ary as R, =


Rg. This circuit can be analysed by


[MI. US]


Puls e tra nsf



N, : N2

(b )




t Pulse


R,,(-N,N, )R


trans former








.' . , ....


Fig. 4.77. (a ) Pulse transformer trigger circuit (b), (c) and (d ) its equivalent ci rcuits .

applying Thevenin's theorem at the terminals a

circuit, where


Vo= VB R , + R L
The voltage equation for Fig. 4.77

(d )


Fig. 4.77

(d )

is the Thevenin's equi .... alent



V0= R O l + L dt
R 1RL .
VSR 1+RL = Rl + RL,+Ldt


) ~;

VB= RL i + L (R ';,R


Its solution is given by


~: [ 1 _ e- L (:I :tRt ) t]

The voltage across L appears as the outpu t voltage. The magnitud e of thi s voltage from
pulse trans former is



o W- ~


(a )

...(4.34 )

21 0

[Act. 4.1;]

Power Elec tronics

Depending upon the values of Ro and L , t her e are two functio nal modes of pulse

Jio > 10 T, where T is th e pulse width (Fig.

(a ) If L is so large as compared with R o that

4.78) of the input signal at G, then from Eq. (4.34),

Rl e e - B Rl + RL

and fo r t = T,

(t IlO T)

"0= V. R ,+ R L
"T = V. R


R "

- 0.1

= 0.904 V.


1+ L

= 0.904"0

Thus the fall in the pulse level during the transmission through the pulse tran sform er at

t = T is very small. This shows that when ;

> 10 T , the input pulse is faithfully transmitted

as square pulse a t the output terminals of pulse transformer as shown in Fig. 4.78 (a).

<b ) If Ro is so large as compared with L that


and for t = T,
This shows that for


~o < ~, then from

Eq. (4.34).

'e -( l O/7')t

"o=V' R ,+ R L


eT = V. R, +RL e

;0 ro,

= 0.0000453 VB R, +RL = 0.0000453'0

the input pulse is transmitted in the fo rm of exponentially

decaying pulses as sh own in Fig. 4.78 (b). It is seen that for a step rise in input voltage, the
pulse transformer output is a positive puls e. In other
words , the input signal is trapsmitted as a derivative of
the input waveform for a step rise. Likewise, for a step
fal l in input voltage, a negative pulse appears at the
pulse transformer output. Fig. 4.78 (b). The operation of
th~ pulse transform er in this mode can be achieved by
uS.mg a small value of L , i.e. by us ing an air core for the
pulse transformer.
L --!-_ __ + --!___--'-_ +.t




It can thus be inferred from above that the deciding v

factor in the waveshape of the output pulses from a
pulse transfo r mer is its inductance . If the pulse
tr ansformer has large inductance, the pulses are
fai thfully r eproduced and ifthe inductance is small, the
pulses are exponentially decaying pulses.
The negative going puls es : an be easily removed by
using a clipper.


1----.J1\l.l. . - - - .r-J.:I\>---.,yrr- . ,



Fig. 4.78. Output voltage wa veform

of a puls e transform er for


(0 ) Ro > l OT

and (b) Ro < 10

Thy r is tors

[Act. . 16[


The amplitude of t he trigger voltage at the secondary te rminals of pu lse t ransformer is


Vg=N- V R



The magnitude of VB should be large enough to pr oduce t rigger voltage Vgr at the gate
circuit of SCR fo r its r eliable turn on, i.e.

N, V. R ,
N 1 ' R I + R l. ~Vg,

V. ~V~~: (l + ~~)


R, =


(~: JR,

V. ~V" ~: [1 +(~:J' ~:1


In practice, exponentially decaying t rigge r pu lses of Fig. 4.78 ( b ) are preferred due to the
following reasons:
{O This pulse waveform is suitable for inj ecting a large charge in the gate circu it fo r
reliable turn on .
(i i )

The duration of this pulse is small, therefore no signi fi cent heating of the gate circuit

is observed.
(iii) For the same gatecathode power, it is permissible to raise VB t o a suitable high
value so that a harddrive of SCR is obtained. A device with a ha rd-drive can
withstand high dU d t at the anode circuit which is desirable.
(iu) The size of th e pulse t ran sfor me r is r educed . For an ext ended pulse, large L (with
iron-cor e) is r equired which increases size and cost of the pulse tr ansformer .

A tr igger ing circuit for a triac using a diac is discu ssed in this. sectio n.
Fig. 4.79 shows a t ri ac fi ring circuit employing a diac. In this circuit, resistor R is variable
whereas resistor Rl has constant resistance. When R is zero, R } protects the diac and triac gate
fr om getting exposed to a lmost full supply voltage.
Resistor R2 limits the current in t he diac and
r--'i:i!co;;."' }.!.-r-- - - ----;
triac gate when diac turns on. The value of C and
potentiometer R are so selected as to give a firing
angle range of n early 0:' and 180:'. In practice,
however, a triggering angle range of 10:' to 170=
is only possible by the firing circuit of Fig. 4.79.


Va ri a ble resistor R controls th~ charging

time of the capacitor C and therefor e the firing
angle of the triac . When R is small, the charging
t ime con st a nt, e qu al to (R 1 + R ) C. is small.
The refor e, sou rce voltage charges capacitor C to Fig. 4.79. Firing circuit ror a triac u5ing n due.


Power Electronics

[M t. 4.16J

diac trigger voltage earlier and firin g angle for triac is small. Likewise, when R is high , firing
angle of triac is large.
When capacitor C (with upper plate positive) charges to breakdown voltage Vd: of diac, diac
turns on. As a consequence, capacitor discharges rapidly thereby applying capacitor voltage Vc
in the form of pulse across the triac gate to turn it on. After triac turn-on at firing angle cr.,
source voltage v, appears acr oss the load during the positive half cycle for (n - cr.) radians. When
Vs becomes zerq at rot = n, triac turns off. After OJt = It, uJ becomes negative, the capacitor C now
charges with lower plate positive. When Vc r eaches Vell of disc, diac and triac turn on and v ,
appears across the load during the negative half cycle for (n - cr.) radians. At Wi = 2n, triac turns
off again and the above process repeats.
The waveforms for v" vc. vT and Vo are shown in Fig. 4.80 (a ) for minimum R and in Fig.
4.80 (b) for maximum R . Here VJ is the source voltage, Vc is the voltage across capacitor, vT is
the vo ltage across triac and Vo is the output or load voltage. After triac turn-on, capacitor C
holds to a small positive voltage.

, I

I, I


.I I

. ,,


l - -,,- /1~dr ~- -- -~:-~I _____1:


.. __



----;1----- :l Vdl1 --:- ,' --::r-:1



--' ra---i, ''( 2Ifter), '" i.


( 1f+0) I



. W

:, II
! wt


(a )

Fig. 4.80. Waveforms for triact firing circu it using Q diac with
pot . R adjusted to minimum and (b) pot. R adju ~t(' d to maximum.

The wave forms shown in Fig. 4.80 ar e for

~o... .
ideal circuit components in Fig. 4.79. In fact,
this circuit produces unsymmetrical waveform
f Ol" the positive and negative half cycles of load
vo ltage. This asymmetry is, to some extent,
due to tria:.-characte ristics but it is mainly due
ru us::VmSlnt.lt
t o hystere sis present in the capacitor. This
means that when lI, is zero, Vc is not zero. In
other words, capacitor retains some charge of
the in iti al voltage applied across its p lates
wh en 30u r ce voltage falls to zero. The
Fig. 4.81. Commercia! tri ne firi ng ci rcu it
wav eforms fo r posi tive and negative half cycles
U!~ing a diac.
ca n, h owev er, b e made symmetrical if
e.d dit ion al r es istance R3 and capacitor C1 are employed as shown in Fig. 4.81. This circuit is


[Ac'. 4.16]


commercially used for controlling the power in lamp dimmers, heat convertors, speed control of
fans etc . For inductive loads, snubber circuit must be used across the triac.
Example 4.27 . The firin g circuit for a triac using a diae, Fig. 4.79, ha.s the following data :
R I = 1000 n, R = zero to 25000 n. C = O.SIJF.
V, = 230 Vat 50 Hz, Diae breakdown voltage = 30 V.

Find the magnitude of maximum and minimum firingangle delays for the triac. The effect
of load impedance may be neglected.
Solution. When the diac is not conducting, the current through R I Rand C is given by


Z=[(R +R)'+(~)'


, ]'12


Z= [

When R =0 :

= [1000' + 6366.2'1"' = 6444.3 n

I I leads V, by an angle

R, +R

"=\ -, lIwC =t _16366.2 =81 07'

an . 1000

230 LO'
I, = 6444.3 L _ 81.07'

:. Voltage across capacitor

V, =J, X,
= 230 L81.07' x 6366.2 L _ 90' = 227.2 L - 8.93'
u, =,[2 . (227.2) sin (Oli - 8.93')

When capacitor voltage
a l is given by


reaches the breakdown voltage of the dine, the triac firing angle

u, = ,[2. (227.2) sin (a, - 8.93') = 30 V


When R = 25000 n


. -1


893' -143'
x30227.2 +.

Z = [26000' + 6366.2'1"' = 26768 n

$.' tan-'

[;~~~;] = 13.76'

230 LO'

I, = 267 68 L- 13 .76'



' x 63 66.2 L- 90' = 54.7 L - 76 .24'
Uc =..J2 x 54 .7 sin (wt - 76 .24)



equal 30 V, let the firing angle be ~

. . 2 x 54 .7 sin (~-76 . 2 . P ) = 30 V

ex, = sin-

(12 ;~4.7 ) + 76.24' = 99.06'

Thus the ma'<imum and rTlifIimurn values offiringangle delays are 99.06 ' and 14.3 respectively.


[Art. 4.17[

P ower Ele ct ron ics


A gate trigge r circuit fo r thyristors in phase-controll ed r ectifi ers sh ould possess the
following :
A circuit for the detection of zero crossing of th e input voltage.
Gener ati on of trigger pulses of required waveshape.
(iii ) DC power supply for pulse amplifier.
(iu) Gate tri gger circuit isolation from the line potential by means of pulse transformers
or optocouplers.
A general block diagram for gate trigger circuit for single-phase converter is shown in Fig.
4.82 . The gating circuit consists of synchronizing transformer, diode rectifi er, zero crossing
detector, firing-angle delay block. pulse amplifier, gate-pulse isolation transformer and pow er
circuit for the con verter.


Synchronizing mid-tapped transformer steps down the supply voltage suitab le for zero
crossing detector and for .delivering de supply Vee to gate trigger circuit. The zero crossing
Power circuit




, It


=C vcc

Sync;;tz ing
Transf .


Zero crossing

(ontrol sigl"lol

Firing OI"Igle


D~ l ay

amplifie r


Gale pulse
transforme r

~~ V"

V9 1


~ C'

Fig. 4.82. Blc c:~ dia -rram of':! th}Tisto, g";!.ting circuit.


1M!. 4.17]


Trans!. vol tage


Romp voi,tage from

zero C. O.










b~ --n I


Control voltage Ec



(211'..-u ) 311'





Fig. 4.83. Waveforms for the ci rcuit of Fig. 4.82.

detec tor conve rts ac synchronizing input voltage into ramp voltage and synchronizes th is ram p
voltage with the zero crossing of the ac supply voltage 8S shown in Fig. 4.83. In the;firing- angle
delay block, the constant amplitude r amp voltage is compared with con tr ol voltage Ee. When
rising ramp voltage equals contr ol voltage E e , a pulse signal of con trolled duration is generated
as shown in Fig. 4.83. Th ese signals are indicated as u, for thyri stors 1 and 2 and uJ fo r
thyristors 3 and 4 for the power circuit of Fig. 4.82. If Ec is lower ed, firin g angle dec reases and
in case Ec is rai se d, firing angle increases. This shows that firing-delay an gle is dir ectly
proportional to the control signal voltage. The pulse output from the firin g-delay angle block
are n ext fed to a pulse amplifier circuit. The amplified puls es ar e th en used for triggering
thyristors 1, 2, 3 and 4 through gate-pulse isolation trans formers as shown.
4.17.1. Gate Pulse Amplifiers
Pulse output from integrated circuits (lCs) may be directly fed to gate-cathode circuit of a
low-power thyristor to turn it on. But in high-power thyristors, trigger-current requirement is
high. The refore, pulses derived from rcs must be am plified and then fed to thyristor fo r its
reliabl e turn on. In a thyristor, anode circuit is subjec ted to high voltage whereas gate circuit
works at a low voltage. Th erefore, an isola tion is essenti al betwe en a thyristor and the
gate-pulse gen er ator. As stated befor e, this isolation is provided by an optocoupl er or a pulse
transform er.
A pulse-amplifi er circuit for amplifying the input pulses is sh own in F ig. 4.84. It consists of
a MOSFET (or a tran sistor ), a pulse transfor mer fo r is olation and diodes 01, D2. Wh en a
voltage of appropriate level is applied to the gate of MOSFET, it gets turned on. :\3 a resu lt,
most of the de voltage Vee appeC'.rs acr oss transformer pr imary a.'1d corresponding pulse voltage
is induced in the transformer secondary. This amplified pulse or;. the secon dary side is applied
to gate and cathode of a thyristo:" to turn it on. When pu lse signal appl ied to the gate of
MOSFET goes to zero, MOSF ET turns off. Th e prim nry current due t o Vee ten ds to fall and
li kewise flux in core also tene!.s to decrease. Due to th is tendency, a vo ltage of opp os ite polarity
is indu ced in both prima.ry and seco ndary windings of pu lse transformer. Diode DI an the


Power Electronics

[Art. 4.1 7]

secondary side of pulse transformer prevents the flow of negative gate current due to the
reverse secondary voltage when MOSFET is off. Reverse voltage in primary, however, forward
biases diode D2 when MOSFET is off. Current flow is thus established in the circuit consisting
of primary, Rand D2. As a consequence, energy in the transformer magnetic core gets
dissipated in R and the core flux gets reset. In case pulse width at the secondary terminals is
to be increased. then a capacitor C is connected across R as shown in Fig. 4.84 (b).





N, iN,



.J tG



v g,

P ~15 e
,i gnal









si gnal


Fig. 4.84. Pulse amplifier circuit using a MOSFET for a thyristor trigger circuit
(a) short-pulse output (b ) long-pulse output.

4.17.2. Pulse Train Gating

Pulse gating is not suitable for inductive, i.e, RL loads, because initiation of thyristor
conduction is not well defined for these types ofloads, This difficulty for such situations can be
overcome by triggering the thyristor continuously. Continuous gating, however, suffers from
some disadvantages like increased thyristor losses and distortion of output pu lse due to
saturation of pulse transformer by continuous pulse. In order to overcome these shortcomings
of continuous gate signal, a train offuing pulses is used to tum on a thyristor. A pulse train of
gating signal is also called high-frequency carrier gating. A pulse train can be generated by
modulating tl~e pulse width at a high frequency (10 to 30 kHz) as shown in Fig. 4.85.
A circuit for generating a pulse train is shown in Fig. 4.85 (a) . This circuit consi .;;!;:! of an
AL'ID-Iogic gate, 555 timer, MOSFET, isolation pulse transformer and diodes Dl, D2. The pulse
signal Vi, obtained from the thyristor trigger circuit and shown in the top of Fig. 4.85 (b ), is fed
to AND gate. The output ut of the timer 555 as shown is also fed to the A.!'\fD gate . T he duty cycle
of the timer should be less than 50% in ord er to allow the transform er fl u x to reset . The pulse
signal Vi and timer output V t are processed in the AND gate to get the waveform output Ux as
shown. The output from A-'N'D gate is ~hen appli ed to pulse amplifier circuit to augment the
amplitude of Ux to Ugif.' The amplified output waveform ug .\: is then app lied across gate-cathode
terminals of a thyristor to turn it on.

[Ad. 4.1 81



Pu lse








V'hnn nhnnnhnnnhn
V'h nnn
pnno ;




Fig. 4.85. Pulse train gating


circuit (b) waveforms.


Cosine firing scheme for thyristors in single phase converters is shown in Fig -t86. The
synchronizing transformer steps down the supply voltage to an appropriate level. The input to
this transformer is taken from the same source from which converter circuit is energized. The
output voltage Ul of synchronizing transformer is integrated to get cosine-wave U2. The dc
control voltage Ec varies from maximum positive Ecm to maximum negative Ecm so that firing
angle can be varied from zero to 180. The cosine wave U2 is compared in comparators 1 and 2
with Ec a nd - Ec. When Ec is high as compared to v 2 output voltage u3 is available fr om
comparator 1. Same is true for comparator 2. So the comparators 1 and 2 give output pulses
1J 3 and V.I respectively as shown in Fig. 4.87. It is seen from this figure that firing angle is



tor 2

Fig . 4.86. Cos ine firing sch em~ for trigge r:ng thyristo rs.




Power Electronics


governed by the intersection of u2 and E,. When E, is maximum, firing angle is zero. Thus,
firing angle a in terms of V2m and E, can be expressed as
V2m cos a =E,

where V 2m

=cos- 1 (~)

...14.36 )


= maximum val ue of cosine signal


' I

Vm sinwt

4n wI




!'< 3lT-+ot)


:.... C211'-tQ)




Fig. 4.87. Waveforms for cosine firing scheme of Fig. 4.86 .

The signals 1)3' u 4 obtained from comparators are fed to c1ockpulse generators 1, 2 to get
clock pulses us, Us as shown in Fig. 4.87. These signals 1)5, U6 energise a JK flip flop to generate
output signals I)j and ur The signal ui is amplified through the circuit of Fig. 4.85 (a) and is then
empl oyed to tum on the SCRs in the positive half cycle. Signal Uj' after amplification, is used to
trigger SCRs in the negative half cycle.
For a single.phase full converter, average output voltage is given by '


... (4.37)

Vo = - - cosa

Substituting the value of a from Eq. (4.36) in Eq. (4.3 7), we get


Vo =2V
- - cos [ cos- IE,
- -] = [2Vm
- - . - - . E,



V 2m

Th} ristors


... (4.38)

This shows that cosine firing scheme provides a linear transfer characteristic between the
average output voltage Vo and the control voltage Ec. This scheme, on account of its linear
tr ansfer characteristic, improves the closed-loop response of the converter syst em. This feature
has mad e the cosine firing scheme quite popular in industrial applications.

4.1. (a ) What is a thyristor? How has this term been coined? Name the most popular thyristor.



(b) What is the definition of thyristor as per IEC ?

(c) Give constructional details of side-gate thyristor. Sketch its schematic di af;:Tam and the
circuit symbol.
(d) Sketch static I-V characteristics of a thyristor. Label the various voltage3. currents and
the operating modes on this sketch.

Des cribe the diffe rent modes of operation of a thyristor with the help of its stati c I-V
<--fb ) Enumerate the various mechanisms by which thyristors can be triggered into co nduction.
Discuss the techniques which result in random turn-on of a thyristor.
4.3. (a ) Describe gate-triggering of a thyristor. Does the gate-current has any effect on the
forward-breakover voltage? Discuss.
lb) How does light triggering of a thyristor differ from gate triggering ? Whe re are LASeRs
used ?
:? 4.4 . (a) Define latching and holding currents as applicable to an SCR. Show these currents on
its static I-V characteristics.
(b) What are the necessary conditions for turning-on of an SCR ? Discu~s.
(c) Define turn-on and turn-off times for an SCR.
4.5. Sketch switching (or dynamic) characteristics of a thyristor during its turn-on and turn-off
processes . Show the va riation of voltage across the thyristor nnd current through it during
these two dynamic processes . Indicate clearly the various interva ls into which turn-on and
turn-off times can be subdivided. Discuss briefly the nature of these curves.
&>~4 :6 . (al Can a forward voltage be appl ied to an SCR soon after its anode current has fallen to
zero? Explain.
Ib) A forward voltage is applied to an SCR soon after reverse recovery current drops nearly
to zero value. Discuss what would happen to the SCR.
(ci Discuss the importance of dildt rating during the turn-on process of a thyristor.
4.7. (al Discuss the conditions which must be satisfied for turning on an SCR with a gate signal.
(b) A thyris tor is conducting a forw ard current. Discuss the basic requirements for commutating (turning-off) this thyris tor.
(0) Bring out clearly how the: anode curre nt expands over the cathode surface area during
the turn-on process of a thyristor.
4.8. (0) Are the tum-on and turn-off times of a thyristor constant? On what factors do thes e
de pend?
(0) In an SCR, the anode curren.t rises li nearly fro m zero to II =- 100 A whereas anode voltage
across S CR falls linearly from VI =- 600 V to zero during its turn-on time of i } = 5 !JS. Derive
an e!'Cpression for the average power los3 in SCR for t} = 5 !-IS. Derive an expression for the
average power loss in SCR for a t rigge!"ing frequency f . In case f""- 100 Hz, find the aVer:lge
pow!:!r 1053 in SCR.
IAns. 1/ 6 VI. /I .l I .!;
5 watts]
(a )


Power Electronics

4.9. (a) Justify the statement, "Higher the gate current, lower is the forward breBkover voltage."
(b ) Wh at is hard-driving for a thyristor? What are its advantages? Sketch a typical waveform
for gate current for hard-driving the thyristor.
(c) For an SCR, the gate-cathode characteristic is given by a straight line with a gradient of
20 volts per ampere passing through origin . The maximum turn-on time is 4 ~s and the
minimum gate current required to quick turn-on is 400 rnA. If the gate source voltage is
15 V, calculate the resistance to be connected in series and the gate-power dissipation.
Given that pulse width is equal to the turn-on time and the average power dissipation is
0.2 W, compute the maximum triggering frequency that will be possible when pulse firing
is used.
(Ans: (e ) 17.5 n, 15.625 kHzl
4.10. Ca) Draw thyristor gate characteristics showing the six gate ratings as specified by the
manufacturers. Discuss these ratings. Indicate clearly the preferred gate drive area . Are
there any other gate ratings in addition to the six mentioned above? If yes, describe
this/these briefly.
(b) The gate-cathode characteristic of an SCR is given by Vg:: 0.5 + 8Ig . For a triggering
frequency of 400 Hz and duty cycle of 0.1, compute the val ue of resistance to be connected
in series with the gate cireuit. The rectangular trigger pulse applied to the gate circuit
has an amplitude of 12 V. The thyristor has average gate-power loss of 0.5 watts.
[ Hint : (b) 0 =

or pulse width, T::

~O~:: 250 IJ.S. A1; T is more than 100 !ls, dc data apPlY]

IAns: Ca) Peak gate-power dissipation and peak reverse gate voltage
(b) 44.23 OJ
4.11. (a) Draw the gate input characteristics ofa batch of thyristors indicating the upper and lower
limi t loci and explain why this variation exists.
(bl Draw the circuit model of a triggering circuit connected to the gatecathode terminals of
a thyristor. Explain the purpose of connecting a resistor across the gate circuit of an SCR.
:./' Ccl A thyristor data sheet gives 1.5 V and 100 mA as the minimum value of gate-trigger
voltage and gate-trigger current respectively. A resistor of 20 n is connected across
gate-cathode tenninals. For a trigger supply voltage of8 V, compute the value of resistance
that should be connected in series with gate circuit in order to ensure tum-on of the device.
[Ans, (b ) 37.143 OJ
4.12. (a l Draw thyristor gate V-I characteristics indicating clearly the gate drive limits. Explain,
with the help of these char'acteristics, the selection of an 'operating point and the choice
of gate circuit parameters.
Discuss also how tum-on time and jitter can be minimised.
( b ) In case gating signal for an SCR consists of a train of pulses instead of continuous dc
signal, explain how the frequency of triggering and other factors are decided .
{cl A thyristor is triggered by a train of pulses of frequency 4 kHz and of duty cycle 0.2.
Calculate the pulse width. In case average gate power dissipation is 1 W, find the
maximum allowable gate power drive.
I Ans: (c ) 50 !lS, 5 WI
4.13 . (a ) Di3cus:; the function of connecting a

diode across gate-cathode terminals,

(ii) diode in series with gate circuit,
(iii) a resistor acros.> gatecathode terminals.
(0 ) How e re the magnit udes o[gate-voJtage and gaLe-current influenced by temperat uTO rise
in a thyristor?
(c) During turnoff of a th),!,istor , idealized voltage and current waveforms are shown in Fig.
4.88. For a triggering frequen cy 0(50 Hz. find the mea n power loss due to turnofT loss .
Aho obtain th e rev e!'3 ed recovery charge.

[P,ob .4]

Thyris tors



2jJ~' I '


O '-----~~'C------C;--_C~_C,--300 A

-+'__ ___

OL-________ - : 2.1.15;""

-100 '1

Fig. 4.88. Pertaining to Prob. 4.13 .




[Ans : (c) 1W, 900 IlCl

4.14. The spread in the gate--cathode characteristics of a thyristor is given by the foll owing two
I, =2.0 x 10- 3 V; and Ig =2.0 x 10- 3 V:s
For an average gate power dissipation of 0.5 W, design the trigger circuit voltage and current
fo r hard-drive.
(Hint: Here I, =79.37 rnA and Vg = 6.3 V. Also I, =47.82 rnA and Vg = 10.456 V. For harddrive, choose I, say 75 m...o\ etc.1
[Ans. Vg = 5.56 V, Ig = 75 mAl

4.15 . (a l Discuss the two-transistor model of a thyristor. Derive an expression for the anode current
and discuss therefrom the turn-on mechanisms of a thyristor .
4.16. (a ) Which cU.rTent rating of an SCR is the most important?
( b) What is the difference between repetitive-current and surge-current ratings of a thyristor?
(c) What are V DR.lI and VRR.\f? Are these ratings different from each other for a thyristor?
(d ) Is it possible to exceed rms current rating of an SCR?
(e ) What are VDWM and VDR.\l? Which ruting is low?
(f) An SeR has maximum rms current rating of 78.5 A. Find its maximum average current
(Ans: (a) RmscuITent (cl VDR.\/ = VRRJf (d) No (e) VDlVMisl'ow if) 50A1
4.17. 1o ) Describe the various anode voltage ratings as opplicBble to an SCR. Indicate these voltage
ratings on a relevant voltage waveform .
Ib) Discuss the signi ficance of duldt in case of thyris tors.
(d Explain why an SCR is derated when it handles pulsed anode current 83 compared to its
rating for constant dc current.
ave rage current rating of an SCR decreases as its conduction angle is reduced .
/ ' Id)

4.1 S~ Define th e following terms relating to SCR and discuss their significance : ~
1i) Forward breakover vo ltage (ii) Peak inverse voltage (iii) critical rate of rise of voltage (iu)
voltage safety factor (u) on-state voltage drop ( ui ) finger voltage.
4.19. (01 The derating of an SCR is more for sine waves than for the square (or rectangular) waves .
Sketch the curves showing average power dissipation as a function of average forward
current for different cond oldion angles for both sine and square waves.
( b) \Vho t is the effect on average curr~nt rating oi an SCR in C"52 inductance is in:;;erted in
the anode circuit? Discuss.




Power El ec tronics

The specification sheet for an SCR gives maximum rms on-s;:nte cuin nt us 50 A. If this
SCR is used in a resistive circuit, compute its average on state current ra ting for conduction angles of30~ and 60 a in case current waveform is (i ) half-::;ine wave nnd (ii) rectangular
[Ans: (cl 30 G (i) 12.56 A (ii ) 14.434 A; 60 (i) 18.00 A (ii ) 20.412 AJ

4.20 .

If a forward voltage is applied to an SCR which is below its breukover voltage , it may
well switch on , particularly if the voltage is applied rapidly. Explain why this is so .
Discuss how the effect mentioned above can be min imi zed .
( b ) A thyristor is placed between a constant dc voltage source of 240 V nnd resistive load
R . The specified limits for dildt and du l dt for the SCR are 60 Nmicro second and 300
V/micro second respectively . Determine the values of the di l dt inductor and the snubber
circuit parameters. Take damping ratio as 0.5.
Discuss how these parameters may be modified to suit the working conditions in the
Derive t he various expressions used.
[Ans: (b) Computed values: 4 J..lH, 5 n, 0.16 J..lF
modified values : 6.4 !lH, 8 n, 0.12 !lFI
4 .2 1. (a ) Snubber circuit for an SCR should primarily consist of capacitor only. But, in actual
practice. a resistor is used in series with the capacitor. Discuss.
(b) R, Land C in an SCR circuit meant for protecting aga inst du l dt and di l dt arc
4 n, 6 J..lH and 6lJ.F respectively. If the supply voltage to the circuit is 300 V, calculate
permissible maximum value:;; of duldt and dildl.
[Hint: (b) Rate of change of voltage across the thyristor at t = 0 when the supply is
switched on is given by
di . l~e
V, 300
= R - + - where I = - = = 7";) !\ etcl
[ Ans: (b) 50 A/J..ls. 212 .5 Vl }lsJ
S dt
' C Rl
(n )

4.22. Following are the specifications of a thyristor operating from a peak supply of 500V:
Repetitive peak current, Ip = 250 A


= 60


(d;,' L 200 YI~s


Take a factor of safety of 2 for the three specifications mentioned above. Design a suitable
snubbe r circuit if the minimum load resi3 tance is 20 n. Take; = 0.65.
I An" 17 ~H, 6 n, 0.5 ~F I
4.23. (a) Discuss how a thyristor may be subjected to internal snd external overvoltages. Describe
the methods adopted for suppressing such overvoltages in thyristor systems .
( b ) During the turn-off process in a thyristor, the revers e recovery current of 10 A is interrupted in a time interval of 4 !ls. The thyristor is connected in series with an inductance
of 6 mH with no resistance in the circuit. If the source voltage during turn-off process is
- 300 V, calculate
(il peak voltage across the thyristor when reo:erse current is interrupted and
(i j ) the value of snubber ci rcuit resistance in case snubber capacitance C, '" 0.3).J.F and
damping ratio is 0.65:[ Ans : (b) - 15.3 kV, 183.85 OJ

Explain the methods adopted for the protection of SCRs against overcurrents.
(b ) A thyris tor. having maximum rms on-state current of 45 A, is used in a resistive circuit.
Compu:e its average ons tate current rati ng fo r halfsine wave for co nduction angl es of
,";13 and ',"'( 12 .
[Ans: (b ) 16.198 A, 20 .26 Al

{a )

4,25, (a ) Describe electronic crowbar prot<!ction scheme employed fo r the overcllrren t protec tion of
power con'/ette rs .
1. 0) Dr aw Co circuit diagra m illustrating the protection of both anode and gate ci rcuits of a n
SC? . Describe briefly t h~ [unctio n of .... arious component .. used.

T hyr istor s



4.26. (a ) Enumer ate the various abnormal conditions against which tbyristors must be protected.
(b ) Describe tbe significance of di l dt and duldt in
(c) Describe, with the help of a circuit diagram, the functi on of various co mponents uded for
the protection of gate circuit of a thyristor.
4.27. (a) Describe the methods employed for improving dildt rating in a thyristor.
(b) Large dIJ l dt may turn on a thyristor at random . Describe how cathodeshorts in thyristors
improve their dIJl dt ratings .
4.28. (a) Discuss briefly the different components of power loss that occur in a thyristor during i~s
working . Which of the power loss component/components islare dominant at power frequencies a nd which at high freque ncies?
( b ) Give the concept of thermal resistance. Describe the analogy between thermal and electri
cal quantities.
(e) Draw the thermal equivalent circuit for an SCR and discuss the various paramete rs
involved in it.
Cd) Describe anyone method of designing the heat sinks for thyristors.
4.29. Ca) For thyristors, various mounting techniques are based on their therm al considerations.
Discuss these mounting techniques with releva nt diagrams.
(b) A t hyristor is rated to carry fullload current wi th an allowable case temperature of
100Ge, for maximum allowable junction temperature of 125C and thermal resistance
between case and ambient as 0.5C/W. Find the sink temper.aty re for an ambient
tempe rature of 40C . Take thermal resistance between sink and ambient as 0. 4 ~C/W.
[Ans, (b) 88' C[


4.30. A thyristor is rated to carry an rms current of 100 A Its maximum allowable j unction
t emperature is 125C.
(0) If this thyristor is made to carry direct current continuously, find the maximum allowable
current rating of the SCR.
(6) If th is SCR is used in a single-phase half-wave circuit with resistive load, find the
maximum allowable average current for firing a ngl es of 0.1 = 30" and ~ = 120.
(c) For part (b ), determine the sink remperatures if average powers dissipated are 200 W for
0. 1 and 150 W fo r 0.2' The value of thermal impedances a re:
a;e = 0.15 C/ W, Be", = O.07C/ W for a1
9je=0 .16C/ W,9c, =0.OaoC/ Wfora2'
[An., (a) 100 A (b) 60 .273 A, 25 .118 A (0) 81 ' C and 89' CI
4.31. A thyristor string is made up of a number ofSCRs connected in series and parallel. The string
has voltage and current rati ngs of 11 kV and 4 kA respectively. Th e voltage and cur~ n t
ratings of avail able SCRs are 1800 V a nd 1000 A respectively. For a string efficiency of 90%,
calculate the number of series and parallel connected SCRs.
For these SCRs, maximum offstate blocking current is 12 mA . Determine the value of static
equalizing resistance for the string. Derive the fonnula used for this resistance .
[Ans: Series- 7. Para llel-5, R = 22 .22 kOJ
4.32. For the th}Tistors of ?rob. 4.31, maximum difference in their reverse recovery charge is 25
mi crocoulombs. Compute the value of dynamic equalizing capaci ta nce of this stri ng. Derive
the formula used for the computation of this capacitance.
[Ans: C == 0.094 ~FJ
4.33. Three s eriesconnected thyTistors, provided with static and dynamic equalizing circuitd. have
to withstand an offst ate volta ge of 8 kV. The static equali zing r esistance is 20 kn and the
dynamic equali. zing ci rcuit has Rc = 40 n a nd C == 0.06 ~I F . These th ree th}'1"istors have ieak,lg:!:
currents of25 rnA, 23 rnA and 22 IT'_A. respectively. De:ermi ne volta ge across each SCR in the
off state and the discharge curren t of each ca pacitor at the time of t urn o ~ .
iAns: 2500 V, 254 0 V, 2560 V; 62. 5 A, 63. 5 A. B.IA]



Power Electron ics

4.34. In a power circuit, foul' SCRs are to be connected in series. Permissible "difference in blocking
voltage is 20 V for a maximum difference in their blocking currents of 1 rnA. Difference in
recovery charge is 10 ~C . Design suitable equalizing ci rcuit.
IAns: Static equalizing resistance = 20 k 0; shunt capacitance = 0.5 ~Fl
4.35. (a ) Discuss how 8CRs suffe r from unequal voltage distribution across them during their
turno n and turn-off processes.
(bl A number of8CRs , each with a rating of2000 V and 50 A, are to be used in series-paraUel
combination in a circuit to handle 11 kV and 400 A. For a derating factor of 0.15, cnlculate
the number of SCRs in series and parallel units .
The maximum difference in their reverse recovery charge is 20 microcoulombs. Calculate
(il the value of dynamic equalizing capacitance and (iil the voltage across each of the slow
thyristors in case one series-connected SCR is fast.
[Ans: (b) n,:7, np=10, C=0.04).lF, 1500V)

4.36. Define string efficiency for series I parallel connected SCRs. Show that stri ng efficiency of two
series connected SCRs is usuallv les s than one.
Derive an expression for the resistance used for sta tic voltage equalization for a series
connected strin g.
4.37. Describe how two series connected SCRs are subjected to unequal voltage distribution during
their dynamic conditions. Derive an expression for capacitance C used in the dynamic equalizing circuit for n series connected SCRs.
4.38. Show that string efficiency for two parallel connected 8CRs is usually less than one.
Discuss the problems associated with the parallel operation of 8CRs and how these are
4.39. (a) Describe briefly the following members of thyristor family.
Illustrate your answer with suitable diagrams.


4.40. (a )
(b )

4.41. (a)
(b )

4.42 . (a)
(b )

4.43. (a )
( b)

4.44. (a )
( b)


Draw the cross-sectional view of the diac and explain how it can conduct in both the
Give the cross-sectional view of a triac and explain its turnon process with relevant
diagrams. Hence show that a triac is rarely operated in first quadrant with negative gate
current and in third quadrant with positive gate current.
Describe LASCR. Give its industrial applications.
Discuss how a triac may sometimes operate in the rectifier mode.
Enumerate the advantages of ASCR and RCT over conventional thyristors.
What is a GTO ? Describe its basic structure.
The turn-off process in a GTO can be described with its twotransistor model. Explain
this in detail.
Bring out clearly the difference between gold-doped GTOs and anode-shorted GTOs .
Describe switching performance in a GTO with relevant voltage and current waveforms.
Give the merits and demerits of GTOs as comp ared t o conventional thyristors .
Define the following terms as applicable to GTOs and discuss their significance.
Turn-oft' gain, backporch current.
Give the application of GTOs.
Describe the bas ic structure of a static induction thyris tor (81TH ).
Explain the turnon and turn-off processes in a SITH. Show that 5 1TH is a normally-on
Compare 81TH with a GTO.





/' 4.45. (a) Discuss the features that the firing circuits for thyristors should possess.

Give the general layout of a firing circuit scheme and explain the function of various
components used in it.
(0 ) Describe the ,.,istance firing circuit used for triggering SeRs. Is it possible to get a firing
angle greater than 90 with resistance firing? Illustrate your answer with appropriate
<'.46. (0) For resistance firing circuits show that firing-delay angle is proportional to the variable
(b ) Resistance firing circuit is used for triggering an SCR in a laboratory. This SCR is
destroyed by a batch of students inadvertently.
A new SCR with the same specification number is installed. But it is found that maximum
firing angle attained is 75 a only. Explain how the desired maximum firing angle of 90 ~
can be obtained.
{Ans: (b) Increase Rl or R2 or else decrease R in Fig. 4.64J
4.47. (a) Draw RC half-wave trigger circuit for one SCR and discuss the function of the various

components used.
Describe, with the help of wavefonns, how the output voltage is controlled by varying the
resistance. Draw the voltage waveform across SeR also.
(b) Describe RC full-wave trigger circuit for one SCR when the load is (i) ac type (ii ) dc type.
Relevant diagrams and waveforms should be drawn to illustrate your answer.
4.48. (a) Compare an UJT ti.ring circuit with Rand RC firing circuits.
(b ) A unijunction transistor, used in relaxation oscillator, has the following data:
T1 =0.67, Iv = 10 rnA, Vu = 2.5 V, Ip;: 15 JJ,A
An oscillator, with an oscillation frequency of 1 kHz, is to be designed by using this UJT.
Compu~ the values of charging resistor and external resistors needed in the ba3e circuits.
Take O:::O.4)J.F and forward-voltage drop of E - Bdunction as 0.5 V. Scu..-ce voltage is 24
V dc and triggering pulse width is 50 J,ls.
[Ans: (b) R = 2.772 k 0 , RmCl:r = 495 k n, Rmin::: 2.15 k 0 ,
R, =621:9 n, Rl =125 nJ
4 .49. (a) Explain the working of an oscillator employing an UJT. Derive expressions for the

frequency of triggering and firing angle delay in terms of eta, chargir.g resistance etc.
A relaxation oscillator, using an UJT, is to be designed for triggering nn SCR. The UJT
has the following data :
11=0.7, Ip =0.5mA, Vp =15 .0V, Vu=O .BV, Iu=2mA, R BB =6k!l.
Normal leakage current with emitter open = 3 mAo
The firing frequency is 1.5 kHz. For C = 0.05 )J.F, compute the values of charging resistor
and the external r esistors connected in the base circuits. Take forward-voltage drop of
E - B 1 junction as zero.
(e) If the frequency of firing the SCR in part (b) is changed by varying charging resistor R ,
obtain the maximum and minimum values of R and the corresponding frequencies.
IAn" (b) 11.074 kn, 47 6.66 n , 666.67 n
(c) 12.858 ill, 10.315 kn, 1.292 ffiz, 1.611 kHzl


4.50. (a) The intrinsic stand-off ratio for an UJT is 0.65. Its intel'bl.'!.se resistance is 10 k n. Calculate
the values of the interbase resistances .
(b ) Estimate the minimum and maximum values of chargi ng resistor in the UJT oscillator
circuit for manu al trigger-angle control of a between 2tj" and 160~ fo r 50 Hz supply.
IAns: (a ) 6.5 k 0, 3.5 k 0, (b ) 2.307 k 0, 1B.457 k OJ
Assume C = 0.4 ).IF and 11 = 0.7.
4.51. (a) Draw and explain the working of an UJT oscillator. Discuss how th e ampli tude of output
voltage pulse can be estimated in this oscillatnr.


Powe r El ectronics


Using a 15-V supply to an UJT, design the oscillator circuit for a frequency of5 kHz . Da ta
fo r UJT is as und er :
., = 0 .65 to 0.75, R B8 = 4 .7 .to 9.1 len
'fake C = 0 .04 J..LF. Missing data may be assumed.
[Hjnt! (6) Ass ume leakage current = 1.88 mAl
(An.: (b) R=4.153kO, R ,= 952.40, R I = 126.30, R_ = 6 kO,
Rmin =3 k
{min = 3460.8 Hz, (max r 6921 .6 Hzl


4.52. A r elaxation oscillator using an UJT is' fabricated to generate pulses for triggering SCRs.
When th e circuit is energised, the circuit fails to oscillate. What could be the plausibe causes
of this failure? How can the circuit be made functi onal?
[Ans. More V BS or less VB! than required ; R < R mill or R > R ma..:.:1
4.53. An UJ T of Fig. 4.71 (a ) has the following parameters :

., =0 .67, VD == 0.7 V,lu =3 rnA, Vu =1 V,lp = 12~. VBS =20 V

(a) Find the value of VEE so as to turn-on UJT if ~E =1 kn.

Find the value of

[Hint. (a ) V EE

to which it must be r educed to tum-off the UJT .

V BB + VD+ /p . RE e~c. 1
IAns. (a ) 14.112 V (b) 4 VI



4.54. Dr aw synchror..ized UJT trigger circuit using a zener diode .

Describe it briefly with relevant voltage and' CUlTent waveforms.
Explain how synchronization of the trigger circuit with the supply voliage across SCR is
' achieved . In case charging resistor is small so that the capacitor voltage reaches UJT
thr eshold v ?lta ~e twice in each half cycle, explain how the circuit operation is influenced.
4.55. (a. ) Draw a circuit diagram for the ramp-and-pedestal trigger circuit used for a single-phase
semiconver ter. Describe its operation with appropriat e waveforms .
For this trigger circuit, derive expressions for the frequency oftrigg'!:ring and firing-angl e
delay in tenus of eta, charging resistor etc.
(b) A fi r ing circuit, u~in g ramp-and-pedestal triggering scheme, has the following dat a :
Char ging resistor = 4 k 0, ch arging capacitor = 0.2 ~F. supply
0.75, zener-diodevoltage 15 V.
fre quency =50 Hz,
Com pute the magnitude of firing angle in case pedestal voltage is (i) zero and (ii ) 4 V.
(Ans: (b) (i) 19.963' (ii) 15.496' 1

,, =

4.5 6. (a ) Describe the use of pulse transformer in the triggering ofSCR! and GTOs. With a suitable
circui t, discuss the conditions under which the input pulse is faithfully transmitted or is
tr ansmitted in the form of exponentially decaying pulse. Which of these two functional
modes is preferred and why?
(b ) The primary of a pulse transformer is connected in series with a t ransistor and a current
limiting r esis tor R L . The data for the triggering circuit is .as under :
RL = 500 0 , gat.e to cathode r esis tance = 200 n
P rimary to secondary turns ratio

= ~, voltage requir ed to trigger the SCR = 3 V.

Comput.e the voltage a pplied to the circuit consisting of transformer primary, RL etc.
Derive the expression used.
[Ans: (b) 16.5 V]
4.57. (a) Describe the trigger circuit for a triac using a diac.
( b) A diac with a breakdown voltage of 35 V, Fig. 4.79, is used for t riggering a tri ac. This
circuit h as RI = 1000 n, R = zero to 280 k n and C = 0.1 J,lF. For a supply voltage of 230
V, 50 Hz ; calculate the maximum and minimum values of firing-delay angles for the triac.
The effect of load imped ance may be neglect ed.
lAns: (b ) 156.5", 7.98 j




4.58. (0 ) Describe 9. gate trigger circuit for a single-phase full converter. Discuss how the adjustment 'of control voltage varies the firing-delay angle.
(b) Describe a gate-pulse amplifier using' a MOSFET.
4.59. (a) Why is pulse-train gating preferred over pulse gating? Explain, with relevant circuit and
waveforms, the pulse-train gating of SeRs.
(b ) Why is the cosine-firing scheme so popular? Describe 8 cosine-firing scheme for the
triggering of thyristors.

Chapter 5

Thyristor Commutation Techniques

............ .. ... .... ...... .. ..... ..... ............... .............. ........... ..... ....... In Ihis Chapter


A Commutation: load Commutation

8 Commutation : Resonant-pulse Commutation
C Commutation: Complementary Commutation
D Commutation: Impulse Commutation
E Commutation : External Pulse Commutation
F Commutation: Une Commutation

.. ........ ..... __ ... ... ... ..... .. ...... .... ..... .. ................... . .. -- .. , ..... ..... __ ..... .... . .


A thyristor is turned on by applying a signal to its gate-cathode circuit. For the purpose of
power control or power cO:lditioning, a conducting thyristor must 'be turned-off as desired. As
stated before, the tum-off of a thyristor means bringing the device from forward-conduction
state to forward-bl ocking state. The thyristor turn-off requires that (i) its anode current falls
below the holding current and (ii ) a reverse voltage is applied to thyristor for a sufficient time
to enable it to recover to blocking state. Commutation is defined as the process of turning-off a
thyristor. Once thyristor starts conducting, gate loses control over the device, therefore,
external means may have to be adopted to commutate the thyristor. Several commutation
te ch niqu es have been devel oped with the sale objective of reducing their turn-off (or
commutation) time,

The use ofthyristor circuits in low-power converters has declined relatively. This is because
of recent advances in semiconductor power devices leading to the availability of power
transistors , GTOs and IGBTs. However, for high-voltage and high-current applications above
about 1 kV and 0.5 kA,' thyristor circuits offer popular circuit configurations.
The classification of thyristor commutation techniques, as reported by various authors, is
not the same, Here, an attempt is made to refer to all these classification techniques. Primarily,
the classification of commutation techniques is based on the manner in which anode current
is reduced to zero and on the configuration of the commutating circuits.
J'hyristor commutation techniques use resonantLC, or underdamped RLC circuits, to force
the current and / or voltagf! oi a thyristor to zero to turn off the device. Several power-electronic
converters employ the circuit co nfigurations used for describing the thyristor commutation
techniques. TherefClre, a study of the various commutation techniques serves as an introduction
and leat.ls to '" better und ers tand ing of the trans!ent ph enomena occuring in power-electronic
converters under switching con ditions .
The various commutati on techniques are now described in this chapter.

For achieving load commutation of a thyristor, th e commutating com ponents L and C are
connected as shown in Fig. 5.1. Here R is th e load resis tance. For low v alue of R, Land C are

Thyristo r Commutation Techniqu es


[Art. 5.1J



r-- --,

Io R

11--: , __ ;


L. _..:







Ic_ 0




Fig. 5.1. Class A or load commutation (a) series capacitor (b) shunt capacitor.

connected in series with R , Fig. 5. 1 (a). For high value of R,lond R is connected across C, Fig.
5.1 (b). The essential requirement for both the circuits of Fig. 5.1 is that the overall circuit must

be underdamped. When these circui~ are energized from dc, current waveforms as shown on
the right hand side of Fig. 5.1 are obtained. Itis seen that current i first rises to ma.umum value
and then begins to fall . When current decays to zero and tencfs' to reverse, thyristor Tin Fig. 5.1
is turned-off on its own at instant A.
~ Load, or class-A, commutation is prevalent in thyristor circui ts supplied from a dc source.
The nature of the circuit should be such that when energized from a dc source, curre nt must
have a natural tendency to decay to zero for the load commutation to occur in a thyristor circuit.
Load comm utation is possible in dc circuits and not in ac circuits. ClassA , or load, commutation
is also called resonant commutation qr self-commutation. A practical circuit employing load
commutation is a series inverter which is described in Chapter 8. Asimple example illustrating
the basic principle of load commutation is given below:
' ..

Example 5.1. The circuit shown in Fig. 5.2 (a) is initially relaxed. The thy ristor T is turned
on at t =O. Determine (a) conduction time of thyristor and (b) voltage across thyris tor and
capacitor after SCR is turned off Ca lculate these values for L = 5 mH, C = 20 p.F and

V. =200 V.
Solution. \vh;n thyristor is turned on , it behaves like a diode. Th erefo re, with S CR on, the
devi ce acts like a closed switch , Fig. 5.2 (b). KVL for this circui t gives
If 'd V
L di
dt + C l t = J
lt3 solution, fro m Art. 3.1.4, is given by Eq. (3.9) which is r epeated here.
i(l) = V.

~ sin ""t

... (5.1 )


P ower Electronics

[A r t. 5.2]


v. ~,




(a )




-l_ _cTl~'

".O.T jf-_ _ _ _w_

' _ _-f__.-.



Fig. 5.2. Ca ) ond (b ) Load commutation circuit (e) waveforms.


Wo =




the resonant frequency of the circui t.

Capacitor voltage, from Eq. (3.IOa) is

ii, (I) = V. (1- cos "'0 I)

... (5.2)

It is seen from above equations that at time t = to=n/Wo. i(t ) = O anl tic: {t } = + 2 V;. TruS''''
shows that 1t/000 sec or 1t ..JLC sec after thyristor is clos ed at t = 0, the charging current becomes
zero, Fig. 5.2 (c) and thyristor is, therefore, turned off on its own. Here
10 =conduction time of the thyristor

=. ~LC


... (5.3)

Voltage tiT across thyristor during its conduction time to is zero. When it stops conducting,
Ur =- 2V~ + V, =- V" It implies that SCR is subjected to a revers e voltage afV, which helps in
its recovery.
For the circuit parameters given, the calculations are as under :
t f
. f th .
esonan requency 0
e ClrcUl , COo = {S x 10- ~ x 20 x 10- 61 1/ 2


3162.27 rad/ , .

=~= 7t " 10 = 9.9346x 10- 4 s

o "'0
= 99.346-.1s.
Volt age across thyristor after it is turned off
=- V. =- 200 V.
Conduction time of thyri stor,

5.2. CLASS B



For explaining class- B, or resonant-pulse, commutation, refer to Fig. 5.3 (0 ). In th is figur e,

sou rce voltage VJ char ges capacitor C to voltage V, with left hand plate positive as shown . Main

Thyristor Commu tation Techniques

[A rt. 5.2 ]

io 1




,~wot=l(_ I'
, ,




+ ~(_.

t( ,



.,.i... . ~D



,I ,I




' T1
tc lor
~ I







,, _1..._. __ :


, ,


t2 t) t, ts


Fig. 5.3. Resonant-pulse commutation (a ) circuit diagram

(b )

waveforms .

thyristor Tl as well as auxiliary thyristor TA are off. Positive direction of capacitor voltage Vc
and capacitor current ic are marked. \Vhen Tl is turned on at t = 0, a constant current 10 is
established in the load circuit. Here, for simplicity,load current is assumed constant.
yptill time tv Uc = V r, ic = 0, i o=10 and iT! =10, Fig. 5.3 (b). F or in itiating th e
commutation of main thyristor Tl, auxiliary thyristor TA is gated at t = tl . With TA on, a
resonant current ic begins to flow from C through TA. L and back to C. This resonant current,
with time measured from instant t l , is given by

ic =- V,

~ sin Wo t = -lp sin Wo t

Minus sign before 1p .sin 000 t is due to the fact that this current fl ows opposite to the
r eference positive direction chosen for ic in Fig. 5.3 (a) .
Capacitor voltage
= V, cos Wo t

... (5.4 )

After half a cycle of ic from instant t 1; ic = 0, Uc = - V, and iTl =10 , .<\fier Jt radians from
instant t l , i.e. just after instant t2 as ic tends to reverse, TA is turned off at t 2 . With Vc = - V"
right-hand plate has pos iti ve p olarity. R es onant curr ent ic n ow b u il ds up thr ough
C, L , D and T l. As this curren t ic grows opposite to forward thyristor current of Tl, net for ward
curr ent in = 10 - ic begin s to decre ase. Finally, wh en ic in the reve rsed direction attain s the
value 10 , fo rw ard current in T l (i n =10 - 10 = 0) is r educed to zero and the device Tl is t urn ed
off at t J . For r eliable commutation, peak r esonan t current Ip must be greater th an load current
10 . As thyri stor is commutated by the gradual build up of r 2 ~o na.T}t current in th e reve rs ed


Power Electronics

[Art. 5.2J

direction , thi:i method of commutation is caUed current commutation, class-B commutation or

resonant-pulse commutation.
After Tl is turned off at t3' constant current 10 flows from V. to load through C, Land D .
Capacitor begins charging linearly from - Va b to zero at t;l and then to V, at t5' As a result, at
instant Is. when Uc = V,. load current io =ic =10 reduces to zero as shown.
It is seen fr om the waveform of i( that main thyristor Tl is turned off when


~ sin 000 (t 3 -

"'0 (I, - I,)



t 2) = 1 0

=sin- 1 (~; J

... (5. 5)

=V, ~ =peak resonant current.

Main thyristor Tl is commutated at t3 ' As constant load current 10 charges C linearly from
Vab at t 3 to zer o at t 4 , SCR Tl is reverse biased by voltage Uc for a period (t4 - t 3 ) = te'
:. Circuit turn-off time for main thyristor,

... (5.6)

tl; =t~-t3 =C 10

Eq. (5.6) shows that tc is dependent on the load current. Waveform of capacitor voltage u~
reveals that the magnitude of reverse voltage Vab across main thyristor TI , when it gets
commu tated, is given by
... (5.7)

Example 5.2. Circuit of Fig. 5.3 (a) employing resonant-pulse commutation (or class -B
com m utation) has C =20 ~ and L =5 ~. Initial uoltage across capacitor is V, = 230 V. For a
constant load current of 300 A, ~alculate
(a) conduction time for the auxiliary thyristor,

(b) uoltage across t~ mam thyristor when it gets commutated and

(c) the circuit turn-off time for the main thyristor.
Solution. Peak value of resonant current,
Ip = V ,-\f[
- /c = 230 -'I
- [20=
s 460 A

"'0 = ~LC

Resonant freq uency,

(0 )


= ~ 100 = 0.1 x 10' radl s

Conduction time for auxiliary thyristor

= -',L =

(0 ) From q. (5.5),

"'0 (I, -


, = 31.416 ~s .

0.1 x 10

(~~~) = 40.706 or 0.71045 ra d.

Voltage across main thyristor, when it gets turned- ofT. is given by Eq. (5.7).


V = V, cos "'0 (I, - I,) = 230 cos (40. 706') = 174 .355 V

[Art. 5.3]

Thyristor Commutation Techniques


Circuit tum-oiTtime for main thyristor, from Eq. (5.6), is


I, = I. - I, = C 1, = 20 x 10

_, 174.355
300 = 11 .6 4 ~s .





------ ---_":\....._- -----+



u;t ITI


~, i,
- V,+


-- - -~T-

(a )

iT I =ll"f" l'





v,[~,+mb i




f/"RI ,



Vs (1- 2p -t/RrC)






e _ '/ R1C
R, ,




Fig. 5.4. Cla33'C commutation (a) and (b ) circui t diagrams (c) wa veforms .



Power Electronics

[Mt. 5.3J

In this type of com mutation , a thyristor carrying load current is commutated by

transferring its load current to another in coming thyristor. Fig. 5.4 (a ) illustrates an :.
arrangement empl oying complementary commutation. In this figure, firin g of SCR Tl
commutates T2 and subsequently, firing of SCR T2 would turn offTl.
Positive and negative directions of voltages and currents are marked in Fig. 5.4 (a ). In this
figure, capacitor is supposed to be initially virgin i.e. uncharged. When Tl is turned on at
t = 0, current through Rl is il =
and through R 2 is ie =
so that thyristor Tl current


iT' =i, + i, =V,

(i, i;I

through R2 fr om ue = O.


begins to flow, Figs. 5.4 (b) and (c). Capacitor C l?egins charging

~e charging current through the circuit V,

C and R2 is given by

ic (t ) = R' . e- t/R~ c

and voltage across capacitor C is given by

!Je(t)= V. (l_e-tlR,c)

Voltage across thyristor T2 is un = Uc (t)

After some time , when transients are over, Ue = un = V, and ie d ecalys to zero . Also
iTI = V, I R I The waveforms for these currents and voltages are shown in Fig. 5.4 (e).

When Tl is to be turned off, T2 is triggered. lfT2 is turned on at t l , then capacitor voltage

ap plies a r everse potential V. across SCR Tl and tUrn s it off. In other words, at

t 1> Un = ,uTl =- V I '

'c =- 2V,
R l an d'.T.!. =V

(2R I + 1) In t h. e CirCUit


conSIS t'log a f

V" R I , C and T2, the capacitor voltage changes from V, to - V, as shown in Fig. 5.4 (c).

Jic dt =V,
1 [I,
C s

For this circuit, KVL gives

R t ic + ~

Its Laplace transform is

R 1 (5)+ -

Its solution gives,

ie (t)=


- - - =-

R' e-IIRLc

As. this current ic (t) flows opposite to the positive direction indicated in Fig. 5.4


... (5.8)


Voltage across capacitor is


... (5.9a )

Volt age across SCR Tl is

un = - Uc = V, [1- 2 e- tlRIC)
... (5.9b)
Note that in Eqs. (5.8) and (5.9), time t is measured from the instant t,. The plots of
capacitor current ic (t) from Eq. (5.8), and capacitor voltage u~ (t) and uTI from Eq. (5.9) are
shown in Fig. 5.4 (c). Current in falls from its value V,

R ,C.


+ ~2) to V, I R, with time constant

Thyristor Commutation Techniques

[Art. 5.4J

'When transients are over after t l ,


= V"~ Ue = - V"~ ic = 0,


= 0, i1'2 = V, I R 2 and i Ti = O.

When TI is turned on to commutate T2 at instant 1"i,.,=O, iT1 =


= - V"~


= 0 and



V,(;, ;j}



With the turn on ofT2 at t I , capncit~r voltage V, suddenly appears as reverse bias across
Tl to turn it off. Similarly, at t 3 , capacitor voltage V, applies a sudden reverse bias across T2 to
turn it off. On account of this, classC' commutation is also called complementary impulse
. commu tation.
Waveforms for voltages and currents are drawn in Fig. 5.4 (c) . Waveform for UTl indicates
that a reverse voltage - V, to zero appears across thyristor Tl for a certain period, This period,
called circuit turn-off time tel for Tl is given by


= 0 =V.

(1_2e-'J RtCj

I" =R j Cln (2)

Similarly, circuit turn-off time for T2 is
I" = R, C In (2)

... (5.10 b)

Example 5.3. Circuit of Fig. 5.4 (a ), employing classC commutation, has

V, =200 V, R J = 10nandR2 = lOOn Determine
(a) peak value of current through thyristors Tl and T2

(b) ualue of capacitor C if each thyristor has turn-off time of 40 )ls. Take a factor of safety 2.
Solution. (a) An examination of Fig. 5.4 (c) reveals that

peak value of current through TI

= V,

and. peak value of current through T2


[~j + ;,] = 200 [110 + I~O]= 24 A

[.l. +.1.]


= 200 [ 10 +
(b) From Eq. (5. 10 0),


Fro m Eq. (5.10 b),

I~O ] = 42 A


Rjln (2)
2 X 40 X 10- 6
10 In (2) =11.542~F

2x40xlO- 6
C = 100 In (2) 1.1042 ~F

So choose a capacitor oflarge size of 11.542 )IF.


For explaining class D. or impulse, commutation, r efer to the circuit of Fig. 5.5 (a). In this
figure, Tl and TA are called m.nin and auxiliary thyri3tors respectively.
Initially, main thyristor T l and auxiliary thyristor TA are off and capacitor is assum ed
charged to voltage YJ with up per pl a te positive. When T l is turned on at t = 0, source voltage


Power Electronics

[Art. 5.4]


_ __

"--_ _ _---,_


. v" In





+ VTA-









Fig. 5.5. Class-D commutation (a ) circuit diagram (b) waveforms .

V, is applied across load and load current 10 begins to fl ow which is assumed to remain constant.
With Tl on at t = 0, another oscillatory circuit consisting of C, TI, L and D is formed where the
capacitor current is given by

Ic SLnWot
. = IpSLnCllot

- t
'e V
= .. 'I

When (00 t = Tt, ie = 0. Between 0 < t < (It!wo), iTl = 1 0 + Ip sin Wo t , Capacitor voltage
change s from + V, t o - VI co-sinusoidally and the lower plate becomes positive. At
Wo t = 1t, ic = 0, iTl = 10 and tic =- V". Fig. 5.5 (b).
At t l auxiliary thyristor TA is turned on. Imm ediately after TA is on, capacitor voltage
V, applies a reverse voltage across main thyristor Tl so that uTI = - V. at tl
SCR Tl is
rurned off and in = O. The load cun:ent is now carried by C and TA. Capacitor gets charged from
- V, to V. with constant load current 10 , The change is, ther efore, linear from + V, to - V, as
shOwn. 'W hen Ut =V" it = 0 a t t2 , thyristor TA is turned off. During the time TA is on from
tl to t 2, Uc = uTl,i c = -10 and to =10 - For main thyristor T1, circuit turn-off time is tc as shown in
Fig. 5.5
, (b).


With the firing of thyristor TA , a reverse voltage V, is suddenly applied across Tl ; this
method of commutati on is therefore, also called uoltage commutation. . With sudden appearance
of reverse voltage across Tl, its current is qu ench ed; in fact the current momentarily reverses
to recover the stored charge ofTI. As an auxiliary thyristor TA is used for turning-off the main
thyristor Tl, this type of commutation is also known as auxiliary commutation.

[A rt. 5.5]

Thyristor Commutation Techniques


When thyristor TA is turned on, capacitor gets connected across Tl to tUrn it off, this type
of commutation is, therefore, also called parallel-capacitor commutatLon .
Example 5.4. Circuit of Fig. 5.5 (a) illustrates class-D commutation. For this circuit,
V, =230 V, L = 20)J.H and C =40 I.J,F. For a constant load current of 120 A , calculate.
(a) peak value of current through capacitance and also through main and auxiliary
(b) circuit turn-off times for main and auxiliary thyristors.



When main thyristor Tl is turned on, an oscillatory current in the circuit

C, '1'1, L and D is set up and it is given by

tc (t) = V, . ~ sin 000 t

:. Peak value of current through capacitor

~ =230~=325.22A

Peak value of current through main thyristor

T1 = Ip + 10 = 325.22 + 120 = 445.22 A

Peak value of current through auxiliary thyristor TA

=10 = 120 A

Waveforms for vTl or ve in Fig. 5.5 (b) indicate ~hat circuit turn-off time for mam
thyristor Tl is the time required for un or ve to change linearly from - V J to zero.

.. I ,


.. Circuit turn-off time for main thjTistor

_ 6 2 30
t, = C 10 = 40 x 10 120 = 76.67 ~s
An examination of Fig. 5.5 reveals that when Tl conducts and during the time upper plate
of C is positive, um =- LIe i .e. auxiliary thyristor TA is reverse biased by LIe' This gives circuit
turn-off time tel for TA


= 2



"'0 = 'JLC =


ho x 40 = ~800

Circujt turn -off time for auxiliary thyristor,





= 2000 = 2 X 106 = 44.43 J..!..5.

In this typ e of comm utati on, a p1l1se of current is ob tained from OJ sepr.rate voltage sou rce
tL :urn off the conducting SCE . The pea.~ value of this current puls e must be more than the load


Power Electronics

[Art. 5.6]

current. Fig. 5.6 shows a circuit using external-pulse

commutation. Here V. is the voltage of the main source
and VI is the voltage of the auxiliary supply. Thyristor +
Tl is conducting and load is connected to source V" VS
Load C
When thyristo r T3 is turned on at t = 0;
V l Tal Land C form an oscillatory circuit. Therefore,
C is charged to a voltage + 2Vt with upper plate
Fig. 5.6. External-pulse
commutation cireui.
positive at t = n ..JLC as shown and as oscillatory
current falls to zero, see Art. 3.1.4, thyristor T3 gets
commutated. For turning off the main thyristor TI, thyristor T2 is turned on. With T2 on, Tl
is subjected to a reverse voltage equal to V. - 2V1 and Tl is therefore turned off. After Tl is off,
capacitor disch arges through the load.

Tbis type of commutation is also 'known as natural commuta~ion. This c;an occur only when
the source is ac, When an SCR circuit is energised from ac source, current has to pass through
its natural zero at the end of every positive half cycle. Then ac source applies a reverse bias
across SCR automatically. As a result, SC~ is turned ofT. This is called natural commutation
because no external circuit is employed to turn-off the thyristor. This method of commutation
is applied to phase-controlled converters, iine-commutated inverters, ae voltage controllers and
step-down cyclocon verters.


Vs =Vmsin wt




. to t U ~




, ......

~ ..... ....










Fig. 5.7. Class F commutation (a) circuit diagram (b) waveforms.

A single-phase half-wave (or one-pulse) controlled converter employing line commutation

is shown in Fig. 5.7 (a). In this figure, thyristor T is fired at firing DJlgle equal to zero, i.e.
when oot = 0, u, = O. Load is resistive in nature. With zero degree firing-delay angle, the
thyristor behaves like a diode. During the positive half-cycle, Uo = uJ and waveshape of load
current io is identical with th e waveshape of Uo for a re3istive load. At wt =tt,
v~ = 0, Uo = 0 and
= 0; therefore T gets turned off at this in3tant. From rot =-1! to Wi = 21t , T
is reverse biased for a period t, = tt/ ro sec, longer than the thyristor turn-off time t q . Here t, is
called the circuit turn-off time.
Another metbod of classification of thyristor commutation technique is as under:
(1) Line commutation: cl ass F
(2) Lo a.d commutation : class A


Thyristor Commutation Techniques



(3) Forced commutation : class B, C and D

External-pulse commutation : class E.

In line, or natural, commutation, natural reversal of ac supply voltage commutates the
conducting thyristor. As stated before,line commutation is widely us ed in ac voltage controllers,
phase-controlled rectifiers and step-down cycloconverters.
In load commutation, L and C are connected in series with the load or C in parallel with
the load such that overall load circuit is under damped . Load commutation is commonly
employed in series inverters.
In forced commutation, the commutating components Land C do not carry load current
continuously. So class B, C and D commutation constitute forced commutation technique!:!. As
stated before, in forced commutation, forward current of the thyristor is forced to zero by
external circuitry called commutation circuit. Forced commutation is usually employed in dc
choppers and inverters.
Example 5.5. In the circuit shown in Fig. 5.8, SCR is forced commutated by circuitry not
shown in the fig ure. Compute the minimum ualue orC so that SCR
does not get turned on due to re-applied dv / dt . The SCR has miniR= 50 !l
mum chargi ng current of 5 mA to turn "it on and its junctio n
capacitance is 25 pF.
Solution. Under steady state, SCR conducts a current
V, 200
=R = 50 = 4 A and voltage across ldeal SCR = voltage v, across


\'lhen SCR is force commutateq. , capacitor C begins charging

from source V, through R so that capacitor voltage u, (= uTl is .given by
V, = V& [l_

Fig. 5.8. Pertaining to

Example 5.5.

e- tI RCl


dU, ] = V . - <IRe . ...l..

[ dt
, e




' 0=


The rate or rise of capacitor voltage

current Cj





...( i )

across SCR may be large. In case SCR charging

happens to be equ al to 5 mA, SeR will get turned on. Here Cj is the junction

capacitance ofSeR.

Cj (~':t O =5mA "

Substituting the value

of(~cl _

fro m Eq. (i) abo..e, we get

, -0

V, 10-3


25 x 10- t2 _ 200 = 5 x 10- 3

bO X C

Power Electronics

[Art. 5.6J


C = 25



10. x 200
250 x 10"


In order to obviate turning on of SCR, the value of capacitance C should be less than
0.02 ~F .


Example 5.S. For a voltage or impulse commutated

thyristor circuit shown in Fig. 5.9, capacitor is initially
charged to V, with polarity as shown . Find the circuit
tu rn -off time for the main thyristor in case
C = 10~, R =Snand V.=200 Vdc.

v, =i';c




Solution . When auxiliary thyristor TA is turned on, .

. .
main thyristor Tl is turned off by means of capacitor Fig. 5.9. Pertammg to Example :l.6.
voltage V. appearing as reverse bias. After Tl is off, KVL for the circuit consisting of
V" C, TA and R in series is given by


R . i (I) + ~

Its Laplace transforme is R I (s) + -1

(I) dl = V.

[IM v.] v.
[ 1]

- -C s


I(s ) R+- = -2V.




I (s)



=-s- . (RsC + 1) = "'j/"

s+ RC

Its Laplace inverse is

The voltage across capacitor C is

(t) =


(t) dt

+ initial voltage across capacitor



2V' e-tlRC_V =V [1 _ 2e- tlRCj
During the time auxiliary SCR TA is on, Ve = !.In =V, [1- e- 1IRC ) . The circuit turn-off time
for Tl is the time taken by lie =Un to change from its value - V, to zero.

O=Vs (l -2e- VRC j

I, =RC In (2) = 5 x 10 x 10" In (2) = 34.6574 ~s.

Example 5.7. For the circuit shown in Fig. 5.1 (a), commutatillg elements L = 20 J,J1{ a.nd
C = 40 ~ a.re connected in series with load resistance R =1 n Check whether self-commutation,
or load commutation, would occur or not. Find the conduc tion time of the thyristor.
Solution. It is seen from Art 3.1.5 that ringing frequ ency
given by

w, =

1/ L~ ti )' =
'" (

w,. in radh.ec , from Eq (3. 14), is

damped frequency of oscillat ion, w.

[Art. 5.6]

Thyristor Commutation Techniques


The condition for underdampir..g is that OOd > 0

L1C-( ~ J






- f4L

4L 4 x 20 x 10- '
Here -C =
= 2. Therefore '''I ~C = ..f2 = 1.41 4 and R = 1 n . As R < 'I ~C ,' the
40 x 10
circuit is underdamped.
Fig. 5.1 (a ) shows that thyristor stops conducting when


, ]'12

10 12
1 X 106
= 2 x 40 - [ 2 x 20 )

:. Conduction time of thyristor, I, =

Example 5.B.

00, t 1 = n .



= 25000 radl sec

1t x 106
= 25000 ~ = 125.664 ~s .

For the circuit shown in Fig. 5.10 (a), d ul dt rating of thyristor T is 400
o. Calculate the ualue of

(a )

V I llS and its junction capacitance is 25 pF. Switch S is closed at t =

C, so that thyristor T is not turned on due to du I dt.

(b ) In case maximum current through thyristor of Fig. 5.10 (a) is limited to 40 A, determine
the ualue of R,.

Sol u tian . (a ) When switch S is closed, the equivalent circuit for Fig. 5.10 (a) is as shown in
Fig. 5.10 (b) wh ere C = Cj + C, . The voltage rise across C is given by

_ v,








--- v,


=F 1=F


(a J

Fig. 5. 10 (0) Pertaining to Example 5:8. Equivalent circuit

(b) for (0) at t = 0 and (e) for (a) when SCR turns on.

The voltag e variation across C is the same as that across C, or Cj of thyristor T .


UC= UT=V,[l - e-

tl t

T} = V, 400 = 200
dl =, RC or 10-' 20 x C

. C= 10x lO- = 002-- F

. o~

C, = C - Cj = 0.025 x 10-' - 0.025x 10-".". 0.025 ~

Vi""hen switch S is closed, C, would be charged to voltage VJ 'w ith upper plate positive
as shown in Fig. 5.10 (c). Now when thyristor is turn ed on , current iT a t that moment through
T would be give n by

I T =R+ 1f

R, =10n.


Power Electronics

[Arl. 3.6]


Example 5.9. For the circuit shown in Fig. 5.11 (a), V,

=200V, L =0.2 mH, C =20!1F,

constant load current 10 = 10 A and capacitor C is initially charged to source voltage V~ with lower
plate positiue. The auxiliary thyristor TA is turned all, at t =0 to commutate the main thyristor TI .
Calculate (a) the time at which the commutation ofmoin thyristor Tl gets initiated (b) the circuit
turn-off time for T 1. Comment on the conduction timl! 1)( auxiliary thyristor.




. ~ F- c





-=- 'I,


[0 =lOA



Fig. 5.11. (a) Pertaining to Example 5.9 (b ) Circuit model when Tl is turned on.

Solution. (a ) When auxiliary thyristor TA is turned on at

t =0, C begins to discharge through L. T1. TA and C.
Eventually. upper plate of capacitor would become positive
with Uc = V,. With TA on, current ic begins to rise as



il l

Note that i" flows opposite to in so that iTl ~ 10 - ie for

0< t < t l , Fig. 5.11 (b) . When it; attains a value of 10 A,
iTl = 0 and therefore. Tl


turnedoff at


is obtained from the relation,

. . rr
f sin




000 t}


o'-'''i,,,,----7":-'~'.--;as shown in Fig. iDo'l" : [Vs}t-Io] : i

5.11 (c). Therefore. time tl at which commutation ofT1 begins

V, -\J


.: ,
lo!---., t~ - - l.. i,

'-">c'~-'-..J-~-ccFig. 5.11 (c). Waveforms

= 10 A

pertaining to Example 5.9.

rc .

.1 10'
dV G.OO xlO-
= ~LC = \I 0.2 x 20 = ~40 an ,'I f = 200 \I 0.2 x 10-' = 40 A
40 sin 00 tl = 10 A


1 .
t '=-sm


{.IQ . - ' (02-)

-10 ) =--sm
. <> = 1-<>. 9S1



(b) After t 1 as i" exceeds 10, diode D1 begins to conduct till time t 2 where
time to of h alf cycle of i, is given by



falls to 10 , Here

n , LC =n = 198.693 ~s
o 00

Circuit turnoff time faT Tl = conduction time of diode D1 = t2 - t} = to- 2 tl = te

= 198.692 - 2 x 15.981 = 166 .73


Thyristor Commutation


[Art. 5.6]


Th e auxili ary thyristor TA will continue conducting .till

capacitor charg~s to V, with upper plate positive and ie falls to
Example 5.10. The circuit of Fig. 5.12 can be used to explain -;; !;- v~ i.e ......,...,.... J ~ ici"lo
class-B commutation. With p ositiue directions indicated for tie>
i e La> iT and u']\ describe, with appropriate wa ueforrns of Lo' ie tic.
i T and uT.how currentcommutation is achieved in this circuit.
Derive expressions for i~ reuerse uoUage across SCR when it is
Fig. 5.12. Pertaining to
Example 5. 10.
turned off and the circuit turn off time. State the assumptions
Solution. The assumptions are as under:
(i) Thyristor is initially off.
(ii) Capacitor C is charged to source voltage V, with upper plate positive.
(iii) Load current is constant and (iu ) SCR is an ideal devic e.
Now. when T is turned on at t = 0, capacitor begins to discharge through C, T , L , Fig. 5 .14

(a ).

Ther efor e, an oscillatory current

i.e =


~ sin Ulo t is e~tablished

Also, at 't = 0, load current 10 is set up so that


Vol cos Ulo t.

ue =

th~5tor current iT = 10 + V, ~ sin Wo t as

shown in Fig. 5.13 and Fig. 5.14 (a) .








c' ~,.3

iT :lO.,.i c

, o~o
R ~
(a) O<t< tl

~ \T:1 0-1<

l.i . 1

" "

---t, l , ,..

o ~----------~~~~~




I T''







Fig. 5.1 3. WaVefOTI!lS for the

circ uit of Fig. 5. 12

Fig. 5. 14. (n) When T is tutn;d on at

tDO . i... =l.+i. (b) t. < t < t., a nd ( r.) t,,< t< t".


[Art. 5.6J

Power Electronics

In Fig. 5.13, at 000 t =lt/ 2, ie =

COo t l


Vc =

0, iT =1 0 + V,

~ (peak value),

liT =

0 and at

=Tt, ie =0, lJe =- V i T =10 and vT = O.

After time t l ie r everses, therefore i T = 10 - it: begins to decrease, Fig. 5.13 and Fig. 5.14 ( b ).

At t 2 ic ri se s to 10 and in

= 0,

therefore T is turned off. Als o at tim t 2 it = 10 , i T = 0,

lie = - ' Vab = - V, cos 000 (t 2 - tl)' UT = - Vob =- V, cos 000 (t 2 - t 1). Note that SCR Tis subje.::ted to
reverse voltage V ab - Between (t3 - t 2), current ie =10 charges C linearly fr om Vob at t2 to zero at
I, so that uT = 0 at
Fig. 5.13 and Fig. 5.14 (c ). This gives


I =C Vob
t3 - t2

:. Circuit turn-off time for T

=(t3 -

t 2) = te =C T

Eventually C charges linearly from zero at t3 to V, at t, as shown in Fig. 5.13. Then ie

reduces to zero. io = 0, ue = V.. (as before),(i T = 0, vT =V, after t,,_



Example 5.11.1n. the circuit shown. in Fig. 5.15 (a), switch S closes at t ~ 0and opens after
10 ms. What will thi currerits in R and L, 8 ms after the switch S opens. A ssume 0.7 V drop
across diode whenever i.t conducts.
Solution. When switch S is closed, Ri + L

It solution

~: =V

,=. V
' [l -e-I/'J


=R1 = 501 =.002 sec an d 1' = 200

50 [ 1 -

At t = 0.01 sec, i = 4 [1 - e- l12 ]

io = 1.54 A inR andL .

= 1.574 A.

' ]
e- -0.02

Therefore, when switch S ope~s after 10 ms ,



Fig. 5. 15. (a ) and (b ). Per taining to Example 5.11

Counting tim e from the instant switch S opens, voltage across L forward biases diode, Fig.
5.11 (b) and current i1 begins to flow.

di 1
L dt + 0. 7 = 0


i 1 - i 1 (0)= - 0.7 t


i 1 = i 1 (0) -0.7 t = 1.574-0.71

Thyristor Commutation T echniques

[Prob. 3]


Wh en I = 0. 008 s, current in L would be il = 1.574 - 0.7 x 0.008 = 1.5684 A

Current in R at t = 0.008 s would iR


Example 5.12 .. In the circuit shown in Fig. 5.16, switch S closes at t = 0 and opens afte r 10
ms. What will be current in R, L and voltage across C, 9 ms after switch S opens. Assume diode
to be ideal.
Solution. 'When switch S is closed, Ri +L

~; =V"
i=4[1_e- tI O,02]

At t = 0.0 1 s, i

= 1.574 A as in the previous example.




'"F == C




Fig. 5.16. Pertaining to Example 5.12

When switch S opens, iR = a and voltage

~croBs L , equal to L ~;: forw~d biases the ideal

diode which begins to conduct at once. An oscillatory circuit consisting of L, C an d diode is

form ed in which current i 1 will decay to zero, Fig. 5,16 (b ). at time tl given by

11 =-"-=< ..JLC =< ,11 x 1 x 10-' =3 .141 ms


Since tl is less than 9 ms, energy in L gets transferred to C as the current il in L decays to

v ' l L ,
c ='2

'2 C


Vc =



x Current when S opens, io


'l/1 x l0-

x 1.574 = 1.574 kV.


Expl ain t he need of commutation in thyristor circuits. Wha t are the differen t methods of
commu tation schemes? Dis cuss one of them, involving tw o thyristors . with a neat
schematic and wa veforms.
(b ) A circuit employing parallelresonance turn-off (or c1ass-B commutation) circuit ha!l
C =50 IJ.F, L = 20 IJ.H, VJ = 200 V and inicial voltage across capacitor is 200 V. Determine
th~ circuit turn-off tim e for main thyristor for load R = 1.5 n .
rAns : (b) 68 ~I3J


(0. )


Distinguish clearly be twee" voltage commutation and current co mmuta tio n in thyristor
circui ts.

Power Eleci:ron ics


Discuss how the voltage across the commutating capacitor is r eversed in a commutating
(c) For the circuit in Fig. S.3 (a), supply voltage V, = 230 V dc, load current 10' = 200 A. circuit
turn-off lime for main thyristor::: 25 J.I.S and reversal current is limited to 150% of 10 ,
Determine the values of commutating components C and L .
[An.: (e) C =29.166 ~F, L =17.143 ~Hl


5.3. In the circuit shown in Fig. 5.9, capacitor C is initially charged to V. = 200 V with polarity as
indicated. Find the circuit turn-off time for mai n thyristor Tl after it is voltage commutated
by thyris tor TA Load current is constant at 40 A and C = 10 ).IF .
[Ans. 50).ls]

(a )


Explain the merits and demerits of self-commutation of SeR and its other methods of
For the circuit shown in Fig. 5.10, given that the load current 10 to be com mutated is 10
A, circuit turn-ofTtime required is 40)Js and the supply voltage is 100 V, obtain the proper
values of commutating components. Take peak resonant current equal to twice the foad
lAns : (b) C:::: 4.619 )JF, L = 115.475 )JH]

Di sc uss , with relevant waveforms, class A and class D types of commutations employed
for thyristors .
(b) For the circuit shown in Fig. 5.10, peak thyristor current =2.5 times the constant load
current, L = 18 )JH and C = 4 )JF. Find the time elapsed from the instant thyristor is turned
on to the instant it gets turned off.
rAns: (b) 32.852 }lsl

5.6. (a.) Enumerate the various commutation techniques used for thyristors.
(b) Describe line-commutation and c!ass-E commutation for thyristors. Name the circuit
configuration where line-commutation is employed.
5.7. (a. ) Discuss, with relevant waveforms, class B and class E types of commutations employed
for thyristor circuits.
(b ) A circuit employing resonant-pulse commutation has C:::: 20 )JF and L = 3 flH. The initial
capacitor voltage = source voltage, Vl =230 V dc. Determine conduction time for alL'Ciliary
thyristor and circuit turn-off time for main thyristor in case constant load current is (i) 300 .-\
nnd (ii) 60 A.
rAns: (b) (i) 24.335 )JS, tc = 13.23)Js (ii) 24 .335)iS, tc = 76.273 IlS]

5.8. (a) Describe class-C type of commutation used for thyristors with appropriate current and
voltage waveforms.
(b) An impulsecommutated circuit is shown in Fig. 5.5 (a). In this circuit, capacitor is initially
charged to source voltage VJ =200 V with upper plate negative . When auxil iary thyristor
is turned on main thyristor gets commutated in 50 )lS. Find the value of C in case load
r esistance i5 20 Q.
Ir peak value of current through main thyristor is limited to twice the full-load current,
calc ulate the value of commutating inductance.
fAns: (b) 3.607 )IF, 1.4428 mHl
5.9. What is com plementary impulse commutatio n? Desc ribe this type of commutatio n with a
circuit di agram and appropriate wavefonns .
Derive expressions for current through and voltage across commutating capacitor. Find also
the circuit tunloff times for the complementary th)T1stors.
5.10. (0 ) A ca pacitor C, initially charged to dc voltage VJ is connected to inductance L through a
thyristor. Determine
(0 the peak value of current through thyris tor Bnd
(ij) the maximum value of dildt through SCR.
(b ) F' OT illustrati ng closs C commutation , ci rcuit of F ig. 5.4 (a ) is empl oyed wher e
VJ :: 200 V and R 1 :: 10 n. Find the value of C 50 that thyris tor Tl i3 commu tat.i!d in
50 )l s .

[Prob. 5]

Thyristor Commutation Techniques


It is required that SCR T2 is turned off naturally when current through it falls below the

holding current of 4 mAo Find the value of R 2.

[Hint: (b) When C is fully charged, current through T2

[ Ans: (a) V,

= holding current = R~ etc.J

."". L


Amp/sec (b) 7.2135 ~F. 50 kn

5.11. In the circuit of Fig. 5.4 (a) employing complementary commutation; V., = 200 V, RI = 20 n
and R2 = 100 Q. Detennine the minimum value of C so that thyristors do not get turned on
due to re-applied duldt. Each SCR has a minimum charging current of 4 rnA to turn it on
and its junction capacitance is 20 pF.
[Ans: 0.1 ~FJ

5.12. For current-commutated circuit of Fig. 5.3 (a ); VI = 230 V, L = 16 ~H and C = 5 ~F. Capacitor
is initially charged to voltage VI with left hand plate positive. Auxiliary thyristor TA is turned
on at t = O. Find the total time for which capacitor current i( exists. The peak resonant current
is 1.5 times the full-load current.

[ Hint: In Fig. 5.3

(b), t5 - t J =

V, +V.b

[Ans , 58.047 ~I

Chapter 6

Phase Controlled Rectifiers

... . _- .-_ .. _-_ .... .............. ............. .. ... ..... -.... .. ... ... ... .. ...... --_. __ ......... ..... In this Chapter


Princ iple of Phase Control

Full-wave Controlled Converters
Single-phase Full-wove Converters
Single-phase Two-pulse Converters with Discontinuous load Current
Performance Parameters of Two-pulse Converte~
Single-phose Symmetrical and Asymmetrica l Semiconverters
Three-phase Thyristor Converters
Performance Parameters of 3-phose full converters
Effect of Source Impedance on the Pe rformance of Converters

Dual Converters
Some Worked Examples

Many industrial applications make use of controllable de power. Examples of such

applications are as follows:
(a) Steel-rolling mills, paper mills, printing presses and textile mills employing de
motor drives.
(b) Traction systems working on dc.
(c) Electrochemical and electrometallurgical processes.
(d ) Magnet power supplies.
(e) Portable hand tool drives.
(f) High-voltage dc transmissi on.
Earlier, dc power was obtained from motor-generator (M G) sets or ac power was converted
to de power by means of mercury-arc rectifiers or thyratrons. The advent of thyristors has
changed the art of ac' to dc conversion. Presently, phase-controlled ac to dc converters employing
thyristors are extensively used for changing constant ac input voltage to controlled dc output
voltage. In an industry where there is a provision for modernization, mercury-arc rectifiers
and thyratrons are being replaced by thyristors.
In phase-controlled rectifiers, a thyristor is turned off as ac supply voltage reverse biases it,
provided anode current has fallen to a level below the ""hCJ~ding current. The turning-off, or
commutation, of a thyristor by supply voltage itself is called natural, or line commutation. In
industrial applications, rectifier circuits make use of more than one SCR. In such circuits, when an
incoming SCR is: tm-ned on by triggering, it immediately reverse biases the outgoing SCR and turns
it off. As pha.5e-<:ontrolled rectifiers need no commutation circuitry, these are simple, less expensive
and are therefo re widely used in industries where controlled dc power is required.
In the study of thyris tor systems, SCRs and diodes are assumed ideal switches which
means that (i) th ere is no voltage drop across them, (i i) no r everse current exists under reverse
voltage conditions and (i ii) holding current is zero.

[.,,1. 6.J]

Phase Controlled Rec tifiers


Trigger circuits are not shown in SCR circuit for convenience.

In this chapter, single~phase and three~phase controlled converters are described and the
effect of source inductance on their performance is examined. Basic oper ating features of dual
converters are also presented.


The simplest form of controlled rectifier circuits consist of a single thyristor feeding dc power
to a resistive load R as shown in Fig. 6.1 (a). The source voltage is u, = Vm sin (llt, Fig. 6.1 (b). An

SCR can conduct only when anode voltage is positive and a gating signal is applied. As such, a
thyristor blocks the flow of load current io until it is triggered. At some delay angle a, a positive
gate signal applied between gate and cathode turns on the SCR. Immediately, full supply voltage
is applied to the load as vo, Fig. 6.1 (b). At the instant of delay angle a, Va rises from zero to
Vm sin a as shown. For resistive load, current io is in phase with va' Firing angle of a thyristor is
measured from the instant it would start conducting if it were replaced by a diode. In Fig. 6.1, if
thyristor is replaced by diode, it would begin conduction at wi = 0, 21t, 41t etc. ; firing angle is
therefore measured from these instants . A firing angle may thus be defined as the angle between
the instant thyristor would conduct if it were a diode and the instant it is triggered.

Firing pulses

Vm sin Cl




V Vm



(io11 TO')

l,,(" ,







Fig. 6.1. Single-phase h:lli-wave thyristor ci rcuit with R load

(0 ) circuit diagram and (0) vol tage and current waveforms


Power Electronics

[Art. 6.1]

A firing angle may also be defined as follows: A firing angle is measured from the angle that
gives the largest average output voltage, or the highest load voltage. If thyristor in Fig. 6.1 is
frred at wt = 0, 2n, 4n etc., the average load voltage is the highest ; the firing angle should thus
be measured from these instants. A firing angle may thus be defined as the angle measured
from the instan t that gives the largest average outpu t voltage to the instant it is triggered.
A critical observation of Fig. 6.1 leads to the emergence of another definition of firing angle.
Thus, a ring angle rna be defined as the angle measured fr om the instant SCR gets forward
~l..- ~
biased to the Instant it is triggere .
Once the SCR is on, load current flows, until it is turned-off by r eversal of voltage at
wt = n, 3n etc. At these angles of1[, 3n, 5n etc. load current frula to zero and soon after the supply
voltage reverse biases the SCR, the device is therefore turned off. It is see n fr om Fig. 6.1 (b) that
by varying the firin g angle ex, the phase relationship between the start of the load current and
the supply voltage can be controlled ; hence the term phase control is used for such a method of
controlling the load currents [3J .
A single-ph ase half-wave circuit is one which produces only one pulse of land current durin g
one cycle of source voltage. As the circuit shown in Fig. 6.1 (a) produces only one load current
pulse for one cycle of sinusoidal source voltage, this circuit represents a single-phase half-wave
thyristor circuit .
In Fig. 6.1 (b ), thyristor conducts from wt =ex to n, (2n + ex) to an and so on. Over the firing
angle delay ex, load voltage tlo = 0 but during conduction angle (1t - ex), tlo =tis. As firing angle is
increased fr om zero to n, the average load voltage decreases from the largest value to zero.
Th e variati on of voltage acr oss thyristor is also shown as tiT in Fig. 6.1 (b). Thyristor
remains on from Wi = a to n, (2n + a) to 3n etc., during these intervals ti T =0 (strictly speaking
1 to 1.5 V ). It is off fyom n to (2n + a), 3n to (4/t + ex) etc., during th ese off intervals vT has the
waveshape of s upply voltage tI, . It may be obse rved that tI, =Vo + tiT. As the thyristor is r everse
biased for 1t r adians , the circuit turn-off time is given by

t, =-w
" sec
where w = 2ft' and, is the supply frequen cy in Hz.
The ci rcuit tum-off time tc must be more than the SCR turn-off tim e tq as specified by the
Average voltage Vo across load R in Fig. 6.1 for the single-phase halfwave circuit in terms
of firing angle ex is given by
Vo= 2n

f"Vm sin wt d


2; (1 +cos ex)

... (6. 1)

The maximum value of aver age output voltage Vo occurs at ex = 0".

Vom = -2=
- 1t

Average load current,

VO = T ( l -rcos ex)



IO =Ji = 2nR (1 + cos a )

...( 6.2 )


Phase Controlled Rectifiers


In some types ofloads, one may be interested in rms value of load voltage Vew Exam ples of
such loads are electric heating and incandescent lamps. Rms voltage Vur in such cases is given

v" = [1...
J' V~ sin' w/ . d (WI)]"

= ~~

[(n-a)+~ sin 2a

The value of r ms current lor is

... (6.3)


l or = Jf

Power deliver ed to resistive loa d = (rms load voltage) (rms load current)



= Vo'!or = If = 1~,R

Input volta mperes

... (6.4 )

= (r ms source voltage) (tota l rms line cur;eiit1


J .

Input power fact or

F;m Eq. (6.3), input

..J2 ...In
lor = 2R

[ (:t -

1 sin 2 a ] "
a) + '2

to load V or ' l or Vor

= P ower delivered
-- V l or =VsInputVA
p{ =

'I;n [ (n - a)

~ sin 2 a

... (6.5)

6.1.1. Single.phase Half.wave Circuit with RL Load

Asingle-phase half-wave thyristor circuit with RL load shown in Fig. 6.2 (a ). Line voltage
v, is sketched in the top of Fig. 6.2 (b). At wt =n , thyristor is turned on by gating signa l (riot-




shown). T he load voltage Vo at once becomes equal to source voltage v, as show n . But the
inductance L forces th e load, or output, current io to rise gradua lly. After some time, io reaches
maximum value and then begins to decr ease. At wt = Tt , Uo is zero but io is not zero because of
the load inductance L . After rot = Tt, SCR is subjected to reverse anode voltage but it will not be
turned off as load current io is not less th an the holding current. At some angle P> It, io reduces
to zero and SCR is turned off as it is already reverse biased. Afte r Wi = p, Vu = 0 and io = O. At
wl = 2n + (,(, SCR is triggered again, Vo is applied to the load and load current develops as before.
Angle ~ is called the extinction angle and (P - a) = y is called the conduction angle.
The waveform of voltage LIT across thyristor T in Fig. ~2 ( b) reveals that when wt = a,
vr = Vm sin CJ. ; from wi = a to P, vr = 0 and at Wi = ~, ur =V m sin ~ . As P> n, ur is negative at
wt = p. T hyristor is therefore reverse biased fr om Wi = ~ t o 2n . Thu s, circuit turn- off time

tc = 2n; ~ sec. For satisfactory commutation, tc should be mor e than

Th e voltage equation fo r the circuit of Fig. 6.2 (0 ), whe n T is on, is

V . . .. sin C!.'l :: R io+L dt


the thyristor turn-off


Power Electronics

[Art. 6. 1J





. I.







L_Vm_Si_nw_t_ _L::.J





Fig. 6.2. Single-pbase half-wave circuit with RL load

(a) circuit diagram and (b) voltage and current waveforms .

The load current io consists of two components, one steady-state

transient component it. Here i, is given by

compo~ent i.

and the other

i, == ~R2:)(l sin (Cilt - 41)


q. = tan- 1 ~ and X = 00. Here $ is the angle by which rms current I. lags V"

The transient component it can be obtained from force-free equation

di t
R I,+L Tt= 0
Its solution gives,

it =A e-(R IL)t


io = i.+t/=

sin(wt_l) +A-(RI L)t

wher e
Constant.4 can be obtained from the boundary condition at oot = a.

At this time t =.!!.

OJ io = O. Thus,
. from Eq. (6.6),


= _

sin (a- 9)+ Ae- RalLw

V m. sin (a _ 41) eRa/tal.


... (6 .6)

[Art. 6.1 1

Phase Controlled Rectifiers


Substitution of A in Eq. (6.6) gives .

io =

~m sin (00/ -

sin (11 - 9) expo\ -

$) - ;

(00/ -(1) }

... ( 6.7 )


It is also seen from the waveform of io in Fig. 6.2 (b ) that when rot = p, load current
to = D. Substituting t his in Eq. (6.7) gives


(~ - $) =sin (11- $). exp \ -

(~ - Il) }

This transcendental equation can be solved to obtain the valu e of extinction angle
case ~ is known, average load voltage Vo is given by

1 J~
Vo = 21t 0. Vm sin rot d (rot) = ~ (cos ex - cos ~)

Average Ioad current, 10 = 2TtR (cos ex - cos "')

Rms load vol tage,

2~ J:V~ sin' d(w/ )

V[ (~ - (1) -"21 (sin 2~ - sin 2(1)]'12
= 2'1:

V., = [

00/ .



.. (6.8)

. ..(69)

... (6. 10)

~s load current can be obtained from Eq. (6 .7) if r equir ed.

: ..
/6.1.2. Single-phase Half-wave Circuit with RL Load and Freewheeling I;>iod e

The waveform ofl oad current io in Fig. 6.2 (b) can be improved by conneCting'a freewheeling
(or flywheeling) diode across load as shown in Fig. 6.3 (a). A freewheeling diode is also called
by-pass or commutating diode . At rot = 0, source vo ltage is becomi ng positive. At some delay
angle a , forward biased SCR is triggered and source voltage vJ appears across load as vo' At
wt = It, source voltage vJ is zero and just after this instant, as v, tends to rever se, freewhe eling
diode FD is fonvard biased through the conducting SCR. As a r esult, load current io is
immedi ately tran sferred from SCR to FD as vJ tends to reverse . At the same time, SCR is
subj ected to r everse voltage and zero current, it is therefore turned off at rot = 1t. It is assumed
that l!uring freewheeling period , load current does not decay to zero until the SCR is triggered
again at (21t + a). Voltage dr op across FD is taken as almost zero , the load voltage Vo is,
therefore, ze ro during the freewheeling period. The voltage variation ac ross SCR is shown as
l)T in F ig. 6.3 (b). It is seen fr om this wave-form that SCR is reverse biased from wi = 1t to
wi =21t. Therefore, circuit turn-off time is

tc =-sec

The source current i, and thyristor current iT have the same w ave form as shown.
Operation of the circui t of Fig. 6.3 (n) can be explained in two modes. In the first mode,
called con duction m ode, SCR conducts from ex to Tt, 21t + a. to 3rt and so on and FD is reverse
biased. Th e duration of this mode is for ((Tt - CX)/ wJ sec. Let the load current at the beginning
of mode 1 be 10 , The exp ression for current io in mode I can be obtained as follows:

j'vl ode I : For conduction mode. the voltagt> equation is


V m sm wt= R~o +L dt

Power Electronics

[Art. 6.1]




' [0


T .;.._ FD'

'I, '







.' T i





. :









rvL.. _V___






Fig. 6. 3. Single-phase half-wave circuit with RL load and 8 freewheeling diode,

(a ) circuit diagram and (b) voltage and current waveforms .

Its solution, already obtained in the previous section, is repeated here from Eq. (6.6 ) as



' o=ysm(wt-41)+Ae (


At Wi =0., i o =10' i.e. att = a ,io=Io


A=[ 1





~m sin(a-) }eRa/1ill.

~mSin {"'t-$)+ [Io - ;Sin {a-$)} exp {- f (t - ~}

(S. l!)

Note that for mode I, a s: Cilt s: n

Mode II : This mode, called freewheeling mock, extends from n to 2n + a, 31t to 41t + a and so
on. In this mode, SCR is reverse biased from it to 21t, 3n to 41t ... as shown by vol tage waveform
Ur in Fig. 6.3 (b ). As the load curr ent is assumed continuous, FD conducts from 1t to (21t + a),
31i to (411: + a ) and so on. Let the current at the beginning of mode 11 be 101 as shown . As load
current is passing through FD, the voltage equation for mode II is

d io


[Art. 6.1]

Phase Controlled Rectifiers

Its s olution is

II - (R/ L J t
=. e

At wi = n,

io =101 ,

l t giv es

A = 101 eR::/ wl.


io=IOlexp [

Note that for mode


- ~I-: )l




Tt < wi S (2n + a )

Average load voltage Vo from Fig. 6.3 (b ) is given by

Vo = 21t

J' Vm sin

wt d (wt ) = V
2; (1 + C05 (X)

Average load current, I = - = - - (1 + cos (X)
o R 21tR

.. .(6.13 )
... (6. 14)

Note that load current io is contributed by SCR fr om a to Tt, (2n + 0 ) to 3n and so on and by
FD from 0 to (X, 1t to (2Tt + a) and so on. Thus the waveshape of thyristor current iT is identical
with the waveshape of io for wt = a to Tt, (2n + a) to 3ft and so on . Similarly, th e wave shape of
FD current iid is identical with the waveform of io for wt = 0 to ex, n to (2ft + a) and so on.
In Fig. 6.2, load consumes power Pl from source for a t o 1t (both Vo and io arc positive)
whereas energy stored in inductance L is return ed to the source as power P2 for n to Jl (va is
n egative a nd iois positive). As a result, net power consumed by the load is th e difference of these
two powers PI and Pa. In Fig. 6.3, load absorbs power for ex to n, but for 1t to (21t + a ), en ergy
stored in L is delivered to load resistance R through the FD. As a consequence, power consum ed
by load is more in Fig. 6.3 . It can, therefore, be concluded that power delivered to load, for the
same firing a ngl e, is more when FD is used. As volt-ampere input is almost the sam e in both
Figs. 6.2 and 6.3, the inputpl(= power delivered to load/input volt-am pere) with the use ofFD
is improv ed . .


It is also seen from Figs. 6.2 (b) and 6.3 (b ) that load current waveform is improved with
FD in Fig. 6.3 (b ). Thus the advantages of using fr eewh eeling diode ar e
input pI is improved
(ii) load cu rren t waveform is improved
(iii) as n r esu lt of (ii) , load performance is better and
(i v) as energy stored in L is trans ferred to R during the freewheeling pe ri od, overall
converter efficiency improves.

It may be seen from Fig. 6.3 (b ) that fr eewheeling diod e prevents the load voltage Uo from
becoming negative. \Vhenever load voltage tends to go negative, FD comes into pl ay. As a result,
load current is t r ansferred fr om main thyristor to FD, allowing the thyristor to regain its
forward blocki ng capabil ity.
It is seen from F igs. 6.2 ( b ) and 6.3 ( b ) th nt sup pl y current i , taken fr om the source i~
uni directiona l an d is in the fo rm of de puls e.5 . Si n~ le phase h alf-wave converter thu s intr odu ces
a dc co:nponent into th e supply line. TPi s is u nd ~ :;i r a bl= as it leads to .5atura tion of the supply
transformer and other diffi culti es (harmonic ::; etc.).
Th e3e shortc oming5 can be overcom e to som e ext e nt by the use of sin gle-ph as e iull wave
circu its discu3s ed in Art. 6.2.



[Art. 6.1J

Power Electronics

6.1.3. Singlephase Half-wave Circuit with RLE Load

A sin gle-phase half-wave controlled converler with RLE load is shown in Fig. 6.4 (a). The
counter emf E in the load may be due to a battery or a de motar. The minimum value of flring
angle is obtain ed from the relation Vm sin wt = E. This is shown to occur at an angle 9 1 in Fig.
6.4 (b).

... (6.15)

In case th}Tistor T is fired at an angle a. < 91 then E > V., SCR is reverse biased and
therefore it will not turn on. Similarly, maximum value of firing angle is 9 2 =1t - 9 1 Fig. 6.4
(b ). During the interval load current io is zero, load voltage uo=E and during the time io ' is
not zero , Uo follows v, curve. For the circuit of Fig. 6.4 (n) and with SCR T on, KVL gives the
voltage differential equation as
.. .(6.16)

The so lution of this equation is made up of two compon ents; namely steady-state current
component i J and the transient current component it. For convenience, i. may be thought of as






Firng pulses








~ a :""' y ...........:



(VmSina-E ) ~

J' ,

: --r



!I i.


. I



Vm s in wt


Fig. 6. 4. Single-phase half-wave ci rcuit with RLE load

(0) circui t dingram and (b) voltage and current wave forms.

P h:.ist: Cuntr OIl t:d K.:..:(ii:.:ts

[An. 6.11


the sum of i"l and i $'~' where is1 i3 the 3teaily 3'::<1te current due to liC source voltage acting alone
and i s'1. is that dut: to de counter emf E acting alone. A3 in the p'ccsentation leading to Eq. (6.6),
i~l due to source voltage V. sin Wi is given by

If only E were present, t hen steady state current

i ,;2

would be given by

ij;l = - (EI R)

The t ransien t current i: is given by

i, _A
- e-tlVL}l

Thus t he total current LO is given by

t il



t.~ or '/

= V'II

.... = a, to
. = O,L' .e. at t = a
. = 0 . Thi s glVes
EAt u..-.
.It = R

; f - - R.

V",r SlO
. ((Ill_I"',
.. ',o = --!
-:'In (Ct -

z L



. (
S lO


Zm.Sln (a -

,11 -

~ Wl-v.J ~

we -.;>

E A - iRI L )(

- R

) ] eRwLw

Er l -exf"" f,. -<ill.R- (

' R.

C1l( - ( l



... {ii. !. 7i
~.; .

\o.l.7J ~:s ~~~ 1 ~':::':::' tc I~:'- .:.c.:; u.t ~ ::.. r..~(. ::: c:-..1..:. .:'(. ... :.
tiring ~fl g!.:: .:r. ;ina the 10;;.,J i(...:;.e.:J.~:I:i.:= ::.r.:.gle y.

::.:-. gh; ~ .:lc~ .:r..:ls U$iVO

iV-lid emf E.

Aver 3.ge voltage across inductance is zero. Thus, a.ver&ge value of luad current can be
obtained by integrating \Vm sin wt - E) / R oetween a ar..d ~ . T he average load curren t 10 is
therefore given by

10 =

2~[ J: (V sin WI - E) d(WI) ]


= 2JtR

Her e conduction angle'y =

[Vm (co. a - cos~) - E (~- a)l

- a . Putting

=y + a

... (6. 18)

in Eq. (6.18 ) gives

10 = 2JtR [Vm Icos a - cos (y + a) -E , yl

Using the trigonometric relation,
. ~
. l..::..!
cosx - cosy = 2 Sill
2 SIn

the above e:cpression lor 10 can be written


IO = 2~[ 2v sin ( a+~ JSin~-E ' Y]


u( 619)

Average load voltage Vo is given by

Vo =E+loR=E +~ 2v sm ( a+n 3in ~ - y ' E ]


= ~ l - J..
) + Vr.mSinru +l2 )'inl2

u.( 6. 20)


POWH E)cClr()nk~

[A n . 6.1 )

The above expression for the average lo&.ct voltage Vo can also be obtained as under :
For periodicity 21t , extending from a to (2:t
Vo = :in

a ), we have

[J" VI{, sin


UJl . d(wt)

In case

-!:- i Va (cos a ~ T.:"


E (2rr T a .-

co.;~) T E (211

CI. -




is is made i::qual co (YT c.) in the a bvve expression , Eq. (6.20 ) can be lihrained.

If lead inductance L is :lero in Fig. 6.4 (ul, then extin..:;tion anglo:: 13 would be: .:'l.ual to
tl! = :'t - 1, i.e. now 13 woula be leSS thz..r. n. Average V:iILle oilo~d current can still be oL~:i.ined
from Eq. (6.1S' by substituting ~ =:! - al . Therefore, a V i;r~gE: load current 1", with L = 0, is

... (6. 18)

Its amplification gives



2~21 (V; + E' ) (~ - Cl) - '1 (sin ~

Powe r delive red to load,

- 2 sin 2Cl) - 2 Vm E (cos Cl-

c~s ~) }

P = I;,. R + 19 E.

Supply power factor

J" . .

(6.2 1)

...(6.22 )
... (6.23)

Tht! time variation of voltage across thyristor is ;;hown as uT in Fig. 6.4 (b). At WI = 0,
u$ = 0, and therefore, uT = - E . At wt =ft, u, = E, therefore u = O. At Wi = a, Us = Vm sin a,
therefore uT = Vm sin a-E. During the conduction angle y =(P - a), uT = O. At wt = p, ui h as
reverse polarity. Therefore , just after thyristor is turned off at Wi = p, voltage
UT =-IVm sin (p - TC ) + E I. It is also possible to write uT = V", sin P- Eat wf = Il, because
Vm sin Pis negati ve fo r P> It. The magnitude ofma.ximum r everse voltage is (Vm + E ) as shown

in Fig. 6.4 (b ). Fig. 6.4 (b) also reveals that circuit


turn~off tim e is

21t + 91 OJ

~ sec.

Example S.l. A s inglephase 230

J kW heater is connecte d across Jphase. 230 V. 50
Hz supply through an SeR. For fir ing angie delay s of 45 and 90, calculate th e power absorbed
in the heater element
Solut ion. Heater resi3tance


[Art. 6.1 ~

PhaSe Cuntrulll.! c1 Hc..:t ifi..::r:.



;/ =

, ;!



' 155 .071

;.( :000 =

~ 5 .; .5 -:

\. ;,:::t::.

( 11- \'
) x 1000 = 250 \v::.tt ::..

~ :23~

Example 6.2. A de: I,c.!tery is charged rhro~gh a resisror R a.s .shOI4"n. in Fig. (;.5 fa). Derive
C!xpre.s.sioH /'cJr the average uaLue of chargirtg curr.!!!! in rerms of VIII" E.,.!? etc. on. the
usswflpticJft thut SeR is {ired cCmtiILuDusly

.'~ I

R :::. ~

F", ,. .:1. ~:.: :,:~!.J.rI:.: l.~l!(J,':;': ui ~:;{J V ,

~1 ~/'.!

j il:J. r/: .; L~b(! (jia ;J':f~i;c. d:/.J.I~.; ./~d ~'~lfrt:H' rur

E :::. 150 V.

, ; " J.'! 1.1 !f,,' ~"'''<1 r

( I.')

[;(j fi'.!.

::,up.0 i,d !i.J tX1.!!":ry wId [ f.a[ d i:,;3ip'-l (.:ci

Cl1.h:u!,;:.!.! chi! .suppLy


tit.! resi.stor.


Solution. Fur the circuit of Fig. 6.5 \a ), t ho:

vol t:1~e

cq us.tion is

Vm sin we =E T ioR
V sinwt-E
. C..!.mo....:=;,;::-'--=-

I, =-


'o',; t

vm sinwt


'"'v Vmsin wl



Fig. 6.5 .

(c )

, ,,



to :




\'~S'tnw t - E

i r (" ,


'\ I


Power circuit diagram (0 1 various waveforms fo r Example 6.2 .

It is seen from Fig. 6.5 that SCR is tur ned on when Vm sin 8 1 = E and is turned off when
Ym sin 8:! = E, where 8:! = It - 8 1, T he battery charging r equires only the average cu rrent 10 given



(A n. 6.IJ


= 21tH (2Vrn cos 8, - E(x - 28,)1
(a) Here

27 660
,= sm. - , \12150
. 230 = .

1 [
r;;_ (
2x27.496xx ',1
10 = 2 . 8 2 ,2 230 cos 27.466 - 100 180
] = ,.9076 A.

(b) Power supplied to battery

E10 = 150 x 4 .9676 = 745.14 W.

For finding the power dissipated in R, rms value of charging current must by obtained.
From Eq. (3.39),


=[ 2.164 {(ISO'''' 230') (

- 2 x 27.466 x


) T (230)' sin 2 x 27.466

- 4 -12 230 150 cos 27.466t)1l1/2 = 9.2955 A.

.. Power dissipated in resistor = (9.2955)2 x 8 = 691.25 Watts .


_ 691.25 + 745 .14 _ 0 -70 I .

0 0 0 ,. ')~ __ - .0 ... :i1 o gTIl", .

From Eq. (6.23/, supply pf

_ .....

::;t. _~ oo

~:imp~..:: -(; . ~ .

Ri!puJ."l Example 6.2 in case ihyri.sm ... is triggered al a {iring

eui!.ry p ositivi! hUll cy:le..
Solution. (a ) Her.:: ex = 35".

P= 02 = r. - e1 = 180 -

a (,gL~

oj 3f," if!

27.466" = 152.534

From Eq. (6.18), average charging current is given by


= 2.: 8 [ {2 x 230 (cos 35'

- cos

152.53~O) -

150 (152.534 - 35) x

1~0 1

= 4.9192A
(b ) Power delivered to battery = E1, = 150 x 4.9192 = 737.88 W

From Eq. (6.21), rms value ofload current is

1o, = [ 2. ~ 64 { (230' + 150') (152.534 - 35) x


2~0 (sin 2 x 152.534 - sin 2 x 35)

- 2. {2 .230 x 150 (cos 35 - cos 152.534) ]

= 9.2874 A

Power dissipated in resistor = 9.28742 x B.= 690.05 W

(c) Supply power ractor

= 6;~.g~ ~.~~~.:8 = 0.6685 lagging

Examples 6.2 and 6.3 demonstrate that an increase in the firing angle reduces the value
of average charging current. rms current and the supply power factor.
Example 6.4. A 230 V, 50 Hz, one-pu.lse SCR controlled converter is triggered at a firing
angle of 40 and the load current e."Ctinguishes at an angle of 210t). Find the circui t tllm off
tim e, average output uoltage and th~ average load current for

(0 )

R = 5 n end L

(b ! R

= 5


n, L =2 mE and E = 11 0 V.


Phase Controlled Rectifiers

26 1

[Ac'. 6.1J

Soluti o n. (0) For this part, refer to Fig. 6.2. It is seen from this figure that circuit tUrn
off tim e tc

_ 2"-~ _ (360-210) " -8333

- 180 X 2l't x 50 - .

From Eq. (6.8), average output voltage

,[2 230
Vo =
[cos 40 - cos 210) = 84.47 1 V
Average load current

10 =

~o = 84;77 = 16.8954 A.

Fig. 6 .4(b) shows that circuit tum-off time is





. - 1 110
19 77.
. - 1 V",
e1 = sm
= sm
"'12 x 230 = .
19.77 - 210) "
t.: = (360180
X 2l't x 50

= 9 432

ms .

From Eq. (6.18), the average charging current is

1 r
10 = 2" . sl ,[2 230 (cos 40 - cos 210) -

no (210 -

40) 180

=6.5064 A.
/.::-- :. Average load voltage, Vo =E loR = n o+ 6.5064 x 5 = 142.532 V.
Example 6.5. A single-phase transformer. with secondary uoltage of 230 V, 50 Hz, deliuers
power to load R = 10 n through a half- waue controlled rectifier drcuit. For a fi ring.angle dela.y
of 60, d etermine (a) the rectification efficiency (b) fo rm factor (e) uoltage ripple factor (d)
transformer utilization factor and (e) PIli of thy ristor.
Solution. Here V, = 230 V,r =50 Hz, R = 10 a,a =60
From Eq. (6.1),

'2 x 230

Va = 2; (1 + cos a ) =
(1 + 005'60) = 77 .64 V


= 7~;4 = 7.764 A

From Eq. (6.3),

Output dc power,

P rk = V., I I)

Output ac power.

Pac = Vori.:r = 145.873 x 14.587 = 2127 .85 W

; 7.64 x 7.764 = 602.8 W

~: = 2~~;:~5 = 0.2333 or 28 .3370

(a ) Rectific ation efficiency

(b) Form facto r,

= V" = 145.873 =1 879



Voltage r ippl e factor, VRF


77 .61

= -IFF' - 1 =,1.t.8 79' - 1 =1.5908


Pow er Elec tronics

[A rt. 6.2J

(d )

TUF = V,l,

(e )


V .. I.

= V, l , =
Vr I"r

230 x 14.587


="2. V, = "2 X no = 325 .22V


There is a large variety of SCR controlled converters (or rectifiers). One way of class,ifying these
ac to dc converters is according to the number e! supply phases on the input side. As per this
classification, the ac to de converters discu..c:sed in Figs. 6.1 to 6.4 are single-phase half-wave
converters. Three-phase controlled rectifiers, as the name suggests, have three-phru:.e supply on their
input side. these are discussed later in this chapter. The other way of classification is according to
the number ofload current pulses per c::--c!e of source voltage. It is seen from Art. 6.1 that single-phase
half-\vave controlled rectifie.r s proouce oJ'ly 011~ p'.llse of lead current dt.:.ri.ng Doe cycle of source
voltage, these can L~er~fe"e be termed a" ~in~:.~~hase one-pulse converters. Thus. the controlied
rectifiers d.isc uss~ ~'! :~;;s. 5.1 !o '3 ..4 c'\.r:,,: a!.l ~i_'1gJ'?-pha~e ~ne F'..tlo::o;: C':;';lverters.

. .'

- I LeAl) ~.----


{a )

Fi. 6.'3 .

(a ~ Sin:o::t~ - ~""'<l;:~ ~W~_":'tl~~ '!l\d-90i~t, ct'r . . . ~ rt<.:.'r

t~~ ? -r~::.>;::-: !"j:-:-? . :<:~ ..... ;,.~ -~':'ir..~ c':'"''-",rt!'":" .

: 0 1'1

The di~ad \ ::'.:1t ~-::c~ ~f s\ngle-ph.<..~~ half-w-3.\-O::. ~r. sin~:l ~t"ha ~ ~ ':'n'.:'- t"Jl,,? ,:::, n\- ~Tto:! ... are
minimi sed by t.':1~ U~!? of s!ngk-t'_~a~'? f 'JJl \\a.... ~. (lr s1nzl ~- pha ~e tW0 cu l :;:~. (';o,,\"e"~'."rs. In
practice. th erE" arp t W I) bal':i...: :("\nfi~urations f')r f1JJ1-\\'a \-~ cC'ntrr,i:~J C':'('l\~ rt <;: r s. One
configuration uses an 1nput transformer with tv,;o \1.indin!:.~ for ?ach ino:_lt oh ;\$~ "'1.0:1;:11:. Thi,.
is cailed mid-point. ':()n. I.;ert.er. A sL'lgk-phase tW'J-pulse mid-poi ... t SCR cp';,\-<.:r',-:- r i~ ~ ~ ~wn in
Fig. 6.6 (a' and a three-phase 6 PlJl~e mid-point converter in Fu::. t3 .G ~ 1: ' .
,- C "~st

f !>uoc-' y






'." .)


tf tf


,. . : - C
0. -






' '2)

Fie:. 6. 7.

(e )

SingJepha5'e two-pulse

' h) Tlv(>-phas",

-" ix- pul~!'


converter and

bridJ?" con\"~rter .




Phase Controlled Rectifiers



The second configuration uses SCRs in the fo rm of a bridge circuit. Single-phase full -wave,
or two-pulse, bridge converter using four SCRs is shown in Fig. 6.7 (a ) and a t hree-phase
six -p ulse bridge converter using si x S CRs in Fig. 6 .7 (b ). A bridge conver ter h as som e
advantages over mid-point converter, these will be discussed after both these configurations a re
studied in the n ext article.

In single-phase two-pulse (or full-wave ) converters, voltage at the output terminals can be
controlled by adjusting the firing angle delay of the thyristors. Mid-point or bridge-type circ uits
may be used for ac to dc conversion. In this section, first mid-point and then bridge-type
configurations are discussed with input from single-phase source.

. 6.3.1. Single-phase Full-wave Mid-point Converter (M-2 Connection)

Th e circuft diagram of a single-phase full-wave converter using a centre-tapped
transformer is shown in Fig. 6.8 (a ). When terminal a is positive with respect to n , terminal n
is positive ..\ith respect to b. Therefore , VII" = V" b or Villi = - Ubll as n is the mid- poin t of seconda ry
'.vinding. Equ ival en t circuit of this arrangement is shown in Fig. 6.B(b ). It is assumed here that
load, or out put, current is continuous and turns ratio from primary to ea ch secondary is unity.



10 1







v tlro


'7 v - , -






: ... V"-.


(b )

2IJmsi no


Fi g. 6.8. S ingle-phase f'.ll1 - wav~ m idpoint convert er (t.l) ci rcuit diag-rarn.

(b) equ ival en t ci rcui t and (c ) va riO\lS volta ge and cu,pcn~ w:! ', i!iorr.5'


P ower Electronics

[Art. 6.3]
Thyri 5 tcr~

T l and T2 are forwa:d biased during positive and negative half cycles
r espectively ; these are therefore t:riggered accordingly. Suppose 12 is already conducting. After
Wi = 0, va." is positi\'e, I I is therefore forward biased and when trigger ed at delay an gle a . TI
gets turned on. At this firing angle (x, supply voltage 2V"m sin ex reverse biases T2. thi s SCR is
therefore turned off. Here Tl is called the incoming thyristor and T2 the outgoing thyristor .
.~ the in:: ":':n i:-;g SCR i~ triggered. ac supply voltage applies reverse bias across the outgoing
thyristor and turos it off. Load CUTTed is also transferred fr om outgoing SCR to incoming
SCR. This process of SCR turn off by natu.ral reversal of ac supply voltage is called n a. tural or
lin p. commu.tation.
From the equivalent circuit of Fig'. 6.8 (b ), it i ~ s-een that if
flo" = V m sin (Ot,
ub" =- u ,,~=- Vm ~i n Wi
r' '1'::. = I.: c" . "',.!, = 2V;"I. s;n

When rot = ex, T1 is triggered . SCR T2 13 subj ected to a reverse voltage uob = 2Vm sin a as
stated before ; current is transferred f!"om T2 t o T1 and as a result T2 is turned off. The
magnit ude of reverse voltage across T2 can also be 0bt.ain~ ~ b;.-- applyin&, KVL t C' the loop
/!fgb.p. of the equi-.:alent circuit of Fig. 6.8 (b) at ~:,,~ l'1st.ant Tl is trigge!"ed. Thus.
I)T2 - '1 /:" . ...011. - !..Tl = IJ


With Tl conducting, un = O. Therefc!'e. the voltage across T2. at the instant


a is gi...-en

1) T2

=- v.'" sin 0: -

VIn sin ex = -

2v-... sin a

This sh ows that SCR T2 is r ever.<;e biased by vo!t.age 2V.., sin a and it is therefore turned
off at rot = a . Thyristor Tl conducts from a. to ;-:: a. Mer ,.:)/. = 1':. Tl is reverse piased but it
\...ill continue conducting as the f'Jrward biased SCR T2 is n ot. get gated. At WI. = j t . a. T2 is

triggered, Tl is reverse biased by voltage '.)f magnitude 2\lm sin 0 , current is transferred from
Tl to T2. Tl is t herefore turned C'ff.
At CJ)t = a. T2 is turned off anel it r e main~ r ever~e bias-<;;:d from (ot = a. to r.. thj ~ c ~.n be seen
f!'om Fig. 6.8 (c). The turn-off time prC'. .-;.ded ?,y this circuit to 8CB. T2 1<:: t~'O'r.efo:-' gh-c.:n by
t ,, = - .




Thyristor Tl is turned off at wt = 1't. a md Fig: . 6.8 {c) re .... eals ~h :;>.': T-;, is ~ '.l'Jjec ted to a
reverse voltage from OJ!: = l't -I- a to Wi = 2l't. Therefore. this circuit provides a tu:-n-'Jff time to
thyristor Tl as

te =

2;;; - (r,: a ) ' :-.: - CI.



which is the s ame as provid ed t l) thyrist or T2 : Eq. (6.24,).

It is seen from voltage waveform I.'c . Fig . 6. &l.c) . that average value of -:)1.ltpu t \ol~ a ge i!'=

gi ven by
'I/o =-Tt

. (1 --:

. (1

Vn-. sin ox . d (oot) =~
cos a

.. .(6.2.5 '

The circuit turn -off tim e t el E q. (6. 24), as provi ded by this circu it of F ig. 6.8 {<::.' mu st be
gr ea te r t ha n SCR t urn-o ff ti me t'l a s gi ven in th e sp eci fi cation sh e!;' t. b case t e "-": t 7 .
co mm uta tion fai lu re w ill occur and the whole s econdary win ding \\ill be sh ort circuited. During
com m uta tion fai lur e. if th e .r a te of rl 3e of fa ult curren t is high. t he inc omin g SCR may be

Phase Controlled Rectifiers

[Art. 6.3J


damaged in case protective elements do n ot clear the fault. Fig. 6.8 (c) reveals that each SCR
is subjected to a peak voltage of 2V m'
, The follo..ving observations can be made from the above study.
(i ) \Then commutation of an SCR is desired, it must be r everse biased and the incoming
SCR must be forward biased.
(i i) Wh~n incoming SCR is gated on, current is transferred fr om outgoing SCR t o incoming
(iii ) The circuit turn-off time must be greater than SCR turn-off time.
It is seen from above that thyristor commutation achieved by means of natural r ever~al
of line voltage l called line or natural commutation, is simple ; it '\ s therefore employed in all
phase-controlled rectifi er s. ac \'oltage controllers and cycloconverters.
V 6.3.2. Single.phase Full-wave Bridge Converters
Phase-controlled slr.gle-phase, or three-phase, full-wave converters are primarily of three
types ; namely unc o:':".troll ed converters, half- controlled converters and fully- con trolled
converters. An uncontrolled con uerter or rectifier u~es only diodes and the le.vel cf dc output
voltage cannot be ccntrolled. A half-controlled conUl!rter or semiconverter uses a mixture of
diodes and thyristors and ther e is a limited
control over the level of dc ou tput voltage. A
f'J.llY-':l)ntrolled conli<:::rtl!r or r.J.ll cOnl.:erter
uses thyristors only and the~e is a ""ider
control over the level of dc output voltag~.

A semiconverter is onequadrant converter.

A one-quadrant converter ha::: one polarity of dc
ou~p ut voltage and curren t at its output
tenninals, Fi:?". 5.9 (a). AtwfJ.qu.adrrmt conv~er
iF O>lO? i'1 wbich v('Itage polarity can reverse but
current direction eanD.')t r everse because of the
(. )
(b )
unidirectionO"j nature of thyristors. Fig. 6.9
(a ) One-quadrant cenverter and
(I,}. In this part of the se<:~on, single-phase
(h: two-Quadrant . co~,'ert.er.
bridge type full converte~ and semiconve:rt.ers
&-r:'e studiei in detail. Single-phase full converter (B2 Co nnect ion)
As.ingle-ph;:.~e full converter bridg2 using-fcur SCR~ is shown in Fig. 6.10 (0). The load is
ass'.1med to be of RLE type, where E is the load circuit emf. Vo : ta~e E may be due to a battery
in th e load circu it or may be g-enerated emf of a de motor. Thyristor p air Tl. T 2 is
simultaneously triggered and n: r adians later. pair T3, T4 is gated t egeth er. When a is positive
with respect to b. supply voltage waveform is. sho\. .'l1 as L''ll) in Fig. 6.10 (b). \Vhen b is positive
with respect to 12, supply voltage waveform is shown dotteci ~ L~g . Obvio1,lsly, L'tV! = - t: ,-,~. The
CU':Tent dire c ~ion!' and voltage polarities sho ....'l1 i.n Fig. 6.10 (0) are treated as positive.
Load c'Jrren t or output current it' is assumed continu'Jus over the working range ; thi~
means that l'n d is alw ays connected to the ac voltage source through the thyrist'J!"s. Between
rot = 0 and C!.'l = CL ; T1, T2 are fo:ward biased through alre ady co nducting SCRs T3 and T4 and
bl ock the iorward voltage. For continu ous current. thyristors T3, T4 conduct after oot = 0 ever.
though these are r everse biased. ""nen iOT\l,; ard biased SCR.5 Tl. 1 2 are trig-gered at (x = 0:,
th-:-y get t..trned on. A$ a result. supply "'oltage Vm sin n immediately appears across thyri~ ~o rs
T3. T4 'a3- a "''vers~ bi as . thes~ are ther efore tur!'led off by naturaL or line. commutation. At
the same ti:ne. load current io flowin.;- tnr()llgc. T3. T4 i ~ tran.~ fe r!'E'd to Tl , T2 at rot = CL . :\ote
. ~.


[Art 6.3J

Power Electronics

that when TI, T2 are gated at wt =a , thes e SeRa will get turned on only if Vm sin ex > E .
Thyristors T1, T2 conduct from rot = ex to 1t + ex. In other words, TI . T2 conduct for 7t radians.
Likewise, waveform of current iT} through Tl (or in through T'2 ) is shown to flow 1t radians in
Fig. 6.10 (b). At wt = n + IX, forward biased SeRs T3, T4 are triggered . The supply voltage turns
off Tl, T2 by natural commutation and the load current is transferred from Tl, T2 to T3, T4.
Voltage across thyristors T1, T2 is shown as un = uT2 and that across T3, T4 as un = uT4'
Maximum revers~ voltage across Tl, T2, T3 or T4 is V m and at the instant of triggering with
firing angle a, each 8-CR is subjected to a r everse voltage of V m sin a. Source current i, is
treated as positive in the arrow direction . Under this assumption, source current is shown
positive when Tl. T2 are conducting and n egative when T3, T4 ar e c0:t:lducting. Fig. B. 10 (b).



~ , ....'

,' I


/ I


I 13

I T2





I r'

I .









ocross TI



I T:-'~~__~~__~____


(a )

(0 )

Fig. 6.10. (0) Single-phase full converter bridge with RLE load
(b) voltage and current wavefonn.;: for conti n ~ous load current.

Phase Controlled Rectifiers

[Art. 6.3]


During a to ft, both t:, and i, are positive, power therefore flows fmm ac source to load.
During the interval 7t to ( 7t .. a), L'~ is negative but i~ is positive, thE> load therefore returns some
of its. energy to the supply system. But the net power flow is from ac sou rce to dc load becau~e
( " - (X) > (X in Fig. 6.10 (b) .
The load terminal voltage,
full-converter output voltage. Vo is shown in Fig. 6.10 (b).
The average value of out put voltage Vo is given by


V I)


f!t . . ", V". .sin rot . d (wt ) = 7


Rm s v:ll1..le of output voltage for

be obtained as und er.


J't.(t V",sm


V' [

O "

... (6.26)

/\/ -2. or B-2. controlled converter can also


= 2ft- wt - "21

. 2

co~ a

1- .. 1' v'... V'
1 5102(.)/
J=T = ~
... (6.27)

~ V,

Eq. ('3.25'1 show~ that if a> 9(F . v~ is

v j
rtegati\"'? T.,,;!' i~ illust r ated in Fi~ . ~ . H' (c ' .

where a I$: ~hown gT''? ater than ~(\~ . It" thi!'

figu~ ,:, . ave rage te rmina l volta,c \"-n is:
o {,-- -X -~ )'negativc. If the l oad ci rcuit emf E is
re-vcrseo. t his source E \\-ill feed power back
T4 OY1o u! '1"2
t" ac supply. This t'peration offu ll c('n"-erter
is knl)wn <''$ irl\'ert~T ('perati-:-n ';'if the
conv,:,rtc ... The f'J1I Cl)nv~rter with fir-ine
angle delay gr~ater thar.. 90-: i~ called
li ll'! -t:.omm.lJf. o t '!d int; uter.
'Jpe r atiC' o I~ lJ $ed in t h e r l?g"'n~rative
Ovtout tu ..... ,,~!
i;I{':;lk.if'ls:' "'Nfe of a de mt:'tor in which ca.c:e
~hE.'n E i ... t:."'J n~ E' r '?mf of the dc m"tf'lr.
_ 'C' __
' \\ '.
Duri"s: 0 to 0:). . ac ~('lJrce ,","ll t<tg"':' " , i~
J;'('o:;iti\'O:; bu t ;;Ie <:ou rce CUlTent i, i~ net;ahve.
power t heref('t r e n"\\-$ from de l'0urCE: t(' a c
source Frl)m a t o Tt. both L.~ and i~ are
p0~it iv~ . pCl\': o:: r there for o::. fl ("\\\'~ from ac
$0urC~ t n dc ~ n IJrc. 3ut the net p(lwt;:r t1t'\\'
v~ .1
i~ from dc sour.:'! to ac ~ource. bec~use
( ;t - 0'. ' < Ct in Fi~ . 6.10 (c ).
In c l) nv~rter operat ion. the average
value t' f ':Iutput voltage v., must be greater
th an I" ad circuit emf E . Dt1rin~ in\"erter
opera:JI) n. load cir cmt emf when inver ted
to ae mlJ~ t 1:-.e more than ac supply voltage.
In ot h~r \\C'd ;::. de ~I')urc'? volt'lgp Emu:;:." br
mor? t ',,\ In itl':er t E.'f \'("IUage VI:" ('n ly ( h ~n
powe:- Wt)IJl d. n nw fro m de l'ou ree t Q ac
curr~ n t wave fo rm for
s ur ~ly S:'- ~ "' rr. . But. in both cl)n\'erter and
FiJ;:. 6.1 0 r.:) inr Q.::- 30i n v .~!':~ r rr :e.c:. : h.\"ri~t o rs must be forwa rci




[Art. 6.3]

Power Electronics

biased and current through SCRs must flow in the same direction as these are unidirectional
devices. This is the reason output current io is shown positive in Fig. 6.10 (c). As before, source
current is is positive when T1, T2 are conducting.

The variation of voltage across thyristors T1, T2, T3 or T4 reveals that circuit turn-off time
for both converter and inverter operations is given by

t, =--5ec
As both the types of phase-controlled converters have been studied, the advantages of
single-phase bridge converter over single-phase mid-point converter can now be statecl. :
(i) SCRs are subjected to a peak inverse voltage of 2 VI'll in mid-point converter and V 1'1\ in
full conVE-rter. Thus for the same voltage and current ratings of SeRs, power handled by
mid-point configuration is about half oft..~at handled by bridge configuration, see Example 6.6.
(ii) In

mid-point converter, each secondary should be able to supply the load power. As such,
the transformer rating in mid-point converter is double the load rating. This, however. is not
the case in single-phase bridge converter.
It may thus be inferred from above that bridge configuration is preferred over mid-point
configurati on. Howe.... er, the choice between these two types depends primarily on cost of the
various components, available source voltage and the load voltage required. Mid-point
configuration is used in case the terminals on dc side hav~ to be grounded.
ExamplE' 6.6. SCRs with peak forwcrd uoltage rating of 1000 V and auerage onstate
current rating of 40 A are used in single-phase mid-point conuerler and single-phase bridge
conuerter. Find the power that these two conlJerters can handle. Use a factor of safety of2.5.
Solution. Maximum voltage across SCR in single-phase mid-point converter is 2Vm , Fig.
6.8. Th erefore, this converter can be designed for a maximum voltage of 21~~~5

=200 V.

:. Maximum average power that mid-point. converter can handle

=l-x-cos a

yTAV= 2x200
x40x 1000=,093kW

SCR in a single-phase bridge converter is subjected to a ma.ximum voltage of"'... , Fig. 15.10.
Therefore, maximum voltage for which this converter can be designed is

1000 =400 V
:. Max;mum average power rating of bridge converter
2 x 40C

= l OCO x x x 40 = 10.186 kW Single.phase 'semiconverter, A singlephase semicon\' ~ rt e r ,ridge wi L~ t;.."o
thyristors and three diodes is shown in Fi g. 6.11 (a). The two thyristors are Tl, T2; the ~'O
d.i odes are Dl, D2 ; the third diode connected across load is freewheeling d.iode FD. The load
is of RLE type as for the full converter bridge. Various voltage and currant waveforms for this
converter are sho\\>-n in Fig. 6.11 (b), where load current is assume d continuous over the
working ra."1ge .
After ~t = 0, thyristor T I is ion vard biased only when source voltage Vm sin IJ.)l exceeds E .
Thus . T l is tri gger ed at a fir ing ang!e delay a such that V III sin a > E . With TI on. load ge ts

[.-\rt. 6. 3J

PhiJ.j ( Controlled Rl!ctiti ers











~TI .,.! T2 i .
"f tt la

"""1r '


T ,, ..;;,

Ib FO +

'---+- - 4



, ..."

:\ ~;r""~J



N r -,




:(2,..,. ) I

, :'








Fig. 6.11. Single-phase s emiconverter bridge (a) Power-eircuit diagram with RLE load and
, (b) voltage nnd current wavefonns for continuow load current.

connected to source through Tl and Dl. For the period wt = ex to It, load current io flows through
RLE, Dl, source and TI and the load termina.l voltage 1)0 is of the same waveshap e as the ac
source voltage US' Soon after rot = It, load vol tage Uo tends to reverse as the ac source voltage
changes polarity. Jus t as Uo tends to r everse (a t wt = It T ), FD gets forward biased and starts
conducting. The load , or output, current ill is transferred from TI, DI to FD. As SCR Tl is
reverse biased at wt = 7t T through FD, T l is turned off at rot =It +. The waveform of current
iTt through thyristor TI is sh own in Fig. 6.11 (b). It flows fr om ex to It, 21t + ex to 31t and so on for
an interval ofCn - a) radian s. The load term l11.als are short circuited through FD, ther efore load,
or output, voltage tlo is zero during it < wt < (it ~ a). After WI. = It, during the n egative h alf cycle,
T2 will be for.vard biased only when source voltage is more than E . At CL'i =It + a, source voltage
exceeds E , 1'2 is therefore triggered. Scan after (It + a), FD is revers e biased and is therefore
turned off ; load cu rrent now shifts from FD to T2, D2. At wi =21t, FD is again forward biased
and ou tput current io is transferre d from 1'2, D2 to FD as explained before. The sourc e current
!J is pC3itive fr om a to it whe :1 T I , DI conduct :lnd is negative fr om (It + a ) to 27t when T'2, D2
cond uct, see Fig. 6.11 (b ).


[An. 6.-'1

During tn .:: Interv:.:al c to It, Tl an d. Dl conC1u..:t and ::.~. ".)urce ddiv.::r:: . ..:f.cr~y to the luau
'::lfCuit. T hl::i cn.:r:;y is partl iiUy "tofo::d In indu .::: t:in .:: ~ L, r..:...rtl:iUy ::,t vfi:C il:::. i::":":[L" l ~' 2r1c: rgy in
iUilCi-l.:lfCUlt I:mf E and parri'-llly dissipated ;AS ~..::::J.t

'r I

R . Dunn::: the tTt." ewho.:dmg: pl:nod It t li tit ..... oJ. "

..:n.:!rg"Y st u r ~.J :n ind .u..:tan..:o.: b f .;:..:v .... <:f t::d. :in..:!. IS
p art ially dl:isip;).t..:d 'itl R Hnd partially added tv ttl":



.....T. i,
, , r ... -----.,

cnl!rg)' stored In IV:id t:m t' E . No enerb'Y h Jcd. t.a..:: k

t o the .sou r..:e during freewh eeling period .




i:i~micu n vl?-rt c r ,

the <lvt:rage output voltage

trom Fig. 6.11 (b ), is given by

vo! . .

" -',





L;'", ~ ;t

V.! .r

=-.' I


:. IJ.. _.; .



Fig-. 6.11 . \cj Convt:ctcr uutput \' Cil~gc as a

(unction ot' nnr,!::" ::in~l.:: t~r s.::mi' and

W ( - - - - , :': !

, 1.
, .. !


n "

V ,,," =



V:. = -l J"'.
V~, sm- we d (wt )



and rms valu.: of Output voltage is

:1:1: '




1 "'
VI/ = -n L VII.::'lnwc d(oot)= ;" ~l .,.. t.:v"u)

= _ !.!.: '


,", ' .,

... _ -;; ;:.'T



~ r.

::.,,'n ')'"
_..... -j


: - -



Thl! vdriatiun of voltag!: acro::.s Tl and '1"2 i.::i abo dtlpn:to::d, in F ig'. 6 .11 ,bi. It i::. ",..:.:.:n irum
thl::::.e w;lvli:forms that circuitturn olf tim~ for the !)cmiconverter is

t, = w-


The variation of average value of converter output voltage as a function of firing angle a is
shown in Fig. 6.11 (c) for semiconverter and full converter, Analy~is of two-pulse bridge converter with continuous conduction. In this
part of th~ section, steady state analysis of singlephase two-pulse converter of both the types
is presented.
S ~ miconv e rter . During c{Jnduction period, the voltage equation for the circuit of Fig, 6.11
(a ), b

uo = u~ = R La TLdt +E
fo r



During fre ewheeling period, the voltage equation is

O=R io+LdtT E


".(6.3 1)

< wt S (1t + a)

Eqs. (6.30 ) and (6.31) can be solved in ti me domain if requir ed. Und er ste3dy-state
co nditions, only average valu e of output voltage and current are requ ired, Therefore,
fr om these two 2quation ~, we get steadystate solution as

op ~ r ating

Va = Rio + E




Vo = V m (1 T cos


I" =



3v~ragl! luad

=a\'~rage \'oltage appl il!d to the: load

\;uTI"t' nt ; E = L.>;,J..:1 cir.: uit . .mt

In ca .:5 ~ loo.d is a dc motor, then E =Km w,,,; R

a rm a ture current ; T" = K", I".

wh~ r e


[AI"(. 6.3}


=r (..

arm&tun:~ir~uit rl;:~ is tan'::l! ;

f;. =1."

= electr omagnetic tor que in Nm.

Km = torque const ant in NmlA, or em! constant in V-sedrad

Vo=r,)g TKmwm

V. -rJ

= "




w /J;-


_ (V,,/:-c)( l



_!..:.... T



Fullconverter. Th~ voltage equati.>n, for the circuit of Fi g. (6 .10,


= t

:.:: R~

...\6. 33,

lCL I,


di J
, L ",1:.'

2 V'"
V\I =- - ClIS Q



In case load is a dc motor, Vo = r)g ,Km{J)",

... \6.341


A comparison of the waveforms for iTt and ioJ in Fig. 6.10 (b ), fo r n single-phase full


tha t in =

~ io tor continuous

:. Average value of thyristor current



load current .

= ~ ( aver3g~ v alu ~ of load current )

=-21 I "

... \6 .351

Th is. h owever. is n ot the case in single-phase semiconverter.

Exa mp le 6.7 . A s ingle-phase fu ll conuerter bridge is connected to RLE load. The source
volcage is 230 V, 50 H2 . The auerage load current of 10 A is conscant ouer the working range.
For R = 0..1 nand L = 2m H, comp ute
(a ) fi ring ang/e del"y for E = 120 V,

(0) fi ring angle dela:1 for E = - 120 V.

Indicate which source is delivering pou:er to load in parts (a) and (b ). S ketch the time
oi Ofl~p u t vo ltage and load cu rrer.! fo r both the parts.



h {'cse Otltp :a current is cssu fr'.:! d cO.'1stcn t, fi nd the inpflt pljor bo th parts (a) and (bJ.

[An, 6.3)

Solution. (0) For E = 120 V, tn !: fu ll conv erter is opera ting as a contl'olled rcctiilcr.
" V



.) . {2 . ->30



COS " = 120 T 10 x 0 .4 = 124 V

= 53.208 =- 53.2:1':
For a = 53.:2 1", power tlows from ac 50urce to de lllsci .
(b ) F or E = - 120 V, the full converter is operating as a l ine comm uta t t!C1 mvertcr.

:2 . v'2 . 230


cos a. = - 120 T 10 x 0.4 =- 116 V

or a = 124.075=: U4.1;;'

For a = 1 24 . 1 ~ , the power flows from de source to ae load.

Output volt::...ge and load current waveform:; for ex = 53.21 CI can be drawn by rererring t.o
Fig. 6.10 ib) and for a = 124.P from Fig. 6.10 (c ).
(c ) For constant load current, rms value ofload current I"r is
1,,=1, = lOA

V.. /orcosl=El o+I!,R

. _ 120 x 10 T lO~ x 0.4 _ ~ ~.) ll ~ ..~

"0 I "
-u,., ... ':t a .,
:..;:S X u

F o'r ,.....' , = ~-'3 ,'_; 1' ,



., . . .. _ :..20 X Iii - 4 0 _ 0 _,..

U = 12:4 .1;,

.... v:>

V -

2::s0 X 1u

" .:;

.:lv4 ... b o o

Example 6.8. (a; A :iingle-phase full conuerter d~li vus p&wer co a resistive load R. Por ac
;source voltage V$' show that average output voltage Vo is given by
Vo =- - (1 Tcas a )


Sketch the time variations of source voltage, output voltage, output current and voltage
across one pair of SCRs. Hence find therefrom the circuit tum-off time.
(b) For the converter of part (a), show that rms value of output current is giuen by

1,,=~ ("-a)+~Sin2a}
Solution. Time variations of source voltage, load
voltage and load current are shown in Fig. 6 .12. At
wt = 1t , Vo = VI = 0 and for resistive load R,





= 0 and io = R = 0,

soon after wt = r.:, supply voltage reverse biases TI , T2;

this pair is therefore turned off. When T3 , T4 i3 triggered
at wt =it + a , output voltage /,10 = uJ up to Wi : 21t. Note
th a t n o SCR conducts during 0 t o a , 1t to (1t + a ) and so
on, Fig. 6.12. F or th e ou tput volt a ge wa veform " 0'
a ver age ou tpu t volta ge Va is


' :'1:

=-1t J


V m. , in 0Jt ' d (col )


= - - (1..,.

C0 3


Fig. 6.12. Per taining to

Example 6.8.

Phase Controlled Rectifiers


[Art. 6.3 ]

The other waveforms can be drawn by referring to Fig. 6.10 (b ).

(b ) Rms value of the output current can be obtained from the waveform io shown in Fig.
6. 12.

~ 1: (;
R ~ 1:

I" = [

=V [

sin oX

d(oX )

(1- cos 2oX) d (oX)


=i[ ~ { (x-a)+~ sm 2a}]'"

Example 6.9. (a) A single-phase controlled rectifier bridge consists of one SCR and three
diodes as shown in Fig. 6.13 (a). Sketch output voltage waveform for a firing angle a for the
SCR and hence obtain an expression for the average output voltage under the assumption of
constant current. Show the conduction of various components as well.
(b) Draw waveforms of current through Tl, Dl, D2 an.d D3 osewrW:tK conatq..n t load current.
(c) For an ac s"Ource voltage of230 V, 50 Hz and firing angle of 45, find the average output
current and power delivered to battery in case load consists of R = 5 n, L = 8 mH and
E =lOOV.

Solution. (a) F or the circuit of Fig. 6. 13 (a). output voltage waveform Vo is shown in Fig.


g~~T1D' ""'0203-!~~T1D '+ 0203-1[

:I :'
0 ,

ii I!







, ,

i ,




Fig . 6.13. (c) Ci rcui t di agrn::: for Ex nmple 6.9




. H::)r" - - - - - I~'




and curre m 'l.'o:::onn3.


Power Electronics

(Art, 6,) J

6.13 (b ). The co nduction of various. comPQnents is also indicated. It is seen that awrage value



is given by


Vo = 21. [ ( Vm sin wt , d(wt ) -


Vm sin wt ' d(wt )

1 ~~

(3 + cos al


(b) The conducti on of various elements shown h~lps

~rawing the waveforms for currents
through Tl, Dl. D2 and D3. For example, D3 conducts from wt = 0 to a, from 1t to?1t + a , from
31t to 47t + a and so.on ; this is shown as iUJ in Fig. 6.13 (b).




-12 230
(3 + cos 45)

10 = i91.8~ - 100
Power delivered to battery


=191.88 V

= i8376 A

'='tno = 100 x 18,376;" 1.8376 kW ,


Example S.lU. A singlephase full converter feeds power to RLE load with R = 6 n,
L = 6 mH and E = 60 V . The ac source voltage is 230 V, 50 Hz . For continuous condtiction, find
the average value of load 'current for a {iring angle delay 'of 50_

In case one of the four SCRs gets ~pen circuited due to a fault, find the new value of avei age
load Cllrrent taking the outpllt current as continuous.
Sketch waueform for the ' new ' output 'volta'ge 'and
indicate the conduction of various SCRs.

vsl :....



2 Vm
2-12 , 230
Vo= - -cosa=
cos 50"

= 133,084 V


10 = V "- E = 133,084 - 60
= 12,181A




. ( 311 +0: )

S uppose SCR T3 in Fig. 6.10 (a ) is damaged and

t a)
is open circuited. With this, output v'oltage waveform
T1T2 -I-- TIT 4
TIT, ~
Uo is as shown in Fig, ~ . 14. Initially, suppose Tl, T2
are conducting from 0. to 7t + 0.. At wi = 7t + Ct., when T3
T4 are gated, only T4 is turned on and as a result, load Fig. 6.14. Pertaining to Example 6.10.
current fre ewh eeling through TI, T4 is zero till T l , 1"2 are triggered again at wt = 27t + a. For
this waveform, average output voltage is given by
Vo =- 27t

Vo =

f' " V

-1223 0


sin wt . d(wt) = V; ' cos a .

cos ;)0 0

10 = 66,54: - 60

27 [' (2:J

= 66.542 V

= 1.0903 A

It is scen t hat load current is reduced radically wit h one SCR getting open circui ted . It is
also observed t hat th:y-ri stor Tl remai ns on.


Controlled R ectifiers

[Act. 6.4 1



So far, sin gle-phase two-pulse converters have been studied on the assumption of
continuous load current. In practice, the output current may become discontinuous at high
values of firing angle or at low values of load current. The term discontinuous is applied to
the condition when load current reaches zero during each half cycle before the next SCR in
sequence is fi red. The term continuous means that load current never ceases but contin ues to
flow through SCRJdiode or their combination. The load performance deteriorates if load current
becomes discontinuous. It is therefore preferable to operate dc load in continuous current mode.
This is prom oted by having freewheeling action and using an external inductor in series with
the load. In this article, working of both single-phase full converler and semi converter is
studied with their load current discontinuous.
6.4.1. Single.phase Full Converter with Discontinuous Current
Power circuit diagram for a single-phase full converter is shown in Fig. 6.10 (a ). For this
converter, when SCR pair T1 T2 is triggered at rot = n, load current begins to build up from
zero as shown. At some angle ~, known as extinction angle, load current decays to zero. Here
P> It. As T1 and T2 are reverse biased after wt =Tt, this pair is commutated at rot = P when
io = O. From a: to ~, output voltage Vo follows source voltage V!, From ~ to (rt + a ), no SCR
conducts, the load voltage therefore jumps from Vm sin ~ to E as shown. At wt = 1t "T" Ct , as pair
T3 T4 is triggered , load current starts to build up again as before and load voltage 1J0 follows
u! waveform as shown. At It + p, io falls to zero, Vo changes from Vm sin (7t +~ ) to E as no SCR
conducts. The source cu rrent i, is also shown in Fig. 6.15 (a).


' I E




TJt7I \;--TlT2 --:Y,__ T3T4 ~ !-T ITZ
I No SCR J1 :
II :

II '
II ;


-t :-~

II :

I conduc ts
(a ) rr < (3 < (rr t o ) II ;
i oWe...
I I, ':
:, :
y /? I I, ..... ( 1' + 1:: ):
: ..... (2'1'-t OJ I :


I :-'
"0 ,

-----: ! :




"1 1



I I,



1[ 11'+/3 )





ti l:

1 I

P< r: a nd V m sin P< E






Fig. 6. r5. Voltage and CU ITe:1t wO\vi:::orm.s for di.scon:'inuo'.ls load

currer.t for a s i..!'l.gle-pha~e fl!!1 co nverter.


[A r .. 6A]

Power Electro nics

Und er some conditions, load current may become zero at wt =~ , \vhere ~ is less than It . It
is assumed here that V m sin P< E. At P, Vo jumps from Vm s in ~ to E . The waveforms for load
current io and load voltage Vo are shown in Fig. 6.15 (b). No SCR conducts from ~ to (n + Ct. )
and during this interval, therefore Uo =E.
From above , the following observations can be ma de :
(i ) Conduction period, a < oot < ~, Tl , T2 conduct and ~' o = u~ . Also
(Tt + <X) < wt It + P), T3 T4 conduct and u" = u$ and so on .
(ii ) Idle period, p < OJt < (n + <x), no circuit element conducts and Vo = E.
The output voltage during discontinuous current mode is less than given by Eq. (6.26). As
stated before, load performance during discontinuous conduction is impaired.
6.4.2. Single-phase Semiconverter with Discontinuous Current
For this converte r. power circuit diagram is given in Fig. 6.11 (a) . For this controlled 2-pulst!
converter, when SCR Tl is triggered at rot = a, load current builds up from zero , rises to a
maximum and then decays to zero at ~ > 1t.
From a to Tt, Tl Dl conduct and Vo = vS' At Wi = 1t, as v, tends to become negative, FD is
iorward biased and starts conducting the load curren t. "Vhen FD condud~ fr:2ll\.1t to ~, Vo = O.










: ( a)


./""'-"....i .


< /311" -to)

...-,. ,



Ii :

.~ I


I' ~



:ig. 6.16. Voitage (l ~d C1.::-rent waverorms fer diseon:in :':Jus

condu ctio:! for a sb gleph ::'!.se 5t!miconv er ter.

[An_ 6.J]

Phruc Coni roll ed Rectifie rs


From p to it + a, no circuit componen t condu cts, therefore l'O = F as shown in Fig. 6.16 (a ).
Du:-ing ~ to it + a, as load current is zero, this makes the load cu:-rent discontinuous. When
T2 is tr iggered at 1[ + a, io builds as shown. At 2it, FD is forwaru biased and starts conduc[.ing
till it +~. During the time FD conducts, lIo = O. From 1t + P to (21t + Ct) , no circuit component
conducts, t herefore LID= E. At l2n + a), T1 is triggered again and the above pmcess rcpeat.i .
Source current i J is also shown in Fig. 6.16 (a).
In case load current becomes zero before it , i.e. for P less than It, then the current and
voltage waveforrrs are shown in Fig. 6.16 (b). Here Vm sin P is assumed less than E. Duri ng p
to 7t + a, no cir:uit component conducts, therefore Uo =E .
From the waveforms for single-phase semiconverler, the following observations are made:
(0) When



P< It + a :

(i) Conduction period,

( ii ) Freewheeling peri od ,

(iii) Idle period,

(b ) When

a < rot < It, TIDl conduct and lI a = lI J . Also

for it + a <ox < 21t, T2D2 conduct and lIo = lis and so on.
7t < wt < p, FD cond uc ts , ifd = io and lIa = O. Also
for 2It < wi < 7t + p, FD conducts , if d = io and lIo = 0 and so on.

p < rot < 7t + a, no circuit component conducts, io = 0 and

P< 'It and V rn sin P< E



(i) Conduction period, ex < cd < ~, T1D1 conduct and lIo = Vol' Also



+ ex < rot < 7t + ~, T2 D2 conduct and lIo =v, and so on

(ii) Freewheeling period, absent and ifi! = O.

(iii) Idle period , ~

io =0 and lIo =E.

< wt < It + ex and 1t + ~ < Wi < 21t + a, no circui t elem e nt condu cts,

The output voltage during discontinuous conduction is not given by Eq. (6.28) . Th~ load
performance with discontinuous load current deteriorates as stated before.
Average output volt age and current. For single~phase full COTlverter. for P> 7t or ~ < It ,
the average load voltage VD, obtained from VDwaveform in Fi g. 6.15 (a) or (b ), is given by
V, =



where y =

Vm sin oll d (oll) + E (n + a -


[cosa-COSP) +E ( 1- 1 )

P- a = conduction angle


V, , --. (cos a - cos ~) + R

E ( 1 - it'
Average load current,lD=Ii =

;;i' . sin gl~ph3Se semi corti.lerler

......i th ~ > 7t , the average Quqmt voltage VD, obtained from vD\Va ... ;. ::.;.1 :: Fig. 6.16 (a), is given by

", =.!it



11_ si n wi . d (CJ::) .;- E ( : -'- .:.:


= ': ,1 + 'os c:.) +E (1_11)


A;e rage load cu;--:~:'.t,

-, ,
lJ = R = r_R (l -)- cos u:) + R 1 -+- ~ )


Power Electronics

[Act. 6.5J



parameters of single-phase full converter and single-phase

Here the performance
semi converter are derived from their input and outpu t waveforms, already f.~ ~ained . The load ,
or output, current is assumed continuous during these derivations.
In general, the instantaneous input current to a converter can be expressed in Fourier
series as under :

i(t ) = a" +
/I '"


(an cos nffit

+ bn sin nwt )

1, 2, 3


1 r"



i (I) d(",,)


cos nwt d


and b, =1.

itt) .

sin nOlI. d


The performance parameters are now obtained first for single-phase full-converter and
then for single-phase semiconductor.
6.5.1. Single-phase full converter
In Fig. 6.10 (b ), the variation of input current, or source current i~, from a to (rt + ex). from
(rt + ex) to (2rt + ex) is continuous but not constant. Here i l is assumed to be ripple free with
amplitude 10 during each half cycle. where 10 = constant load current.

i l (t ) =10 +

CTI sin (nwt + 8T1 )

" '" I , 2, 3


C, =

~a; +b; and e, =tan- 1 ( : : )


I . = -2 [


FHa I . d (OlI) - j,aI , . d (0lI)1= 0


11 .. 0:

=-111 [J"a



cos nwt. d (wt ) -

J211: +CI
!t+ CI

10 cos nwt. d (lOt)]

II sin nwt 11(+

Q- I sin nwt 12:t+ Q]


=- - -0




ex .........

1. 3, 0- ...

= 0 ............ for'l = 2, 4, 6,.... .

b" =-1
1 sin ncut. d (wt ) - f' ,-a 10 sin milt. d (cot )




[ 1-cosnOll 1,a a n1l:

l -cDsn WI 1:'"1.... :1

= - cos n ex .................. forn= 1, 3, 0 , ...


r -~-,:-"

=O ................. ............. .. for r. =2 , -1 6, .. .

C, = [ [ - : :. sin na )' + ( : ; cos ncr )'


= tan- 1 [- tan n o l = -


Phase Controlled Rectifiers

[Ar'- 6.51

displacement angle of nth harmonic current .

i. (t ) = ~ _ 0 sin (a wl - no)


na l ,J, S


... (6.36)

The rms value of nth harmonic input current, from Eq. (6.36), is .
2 >/2Jo '.
I - ..",...-cc - In -

"2 .rite - .


...(6.300 )

- - 2>/2 . I.
Rms va ue of fund~ e~tal curren~, 1' 1== _ ' rr
= 0.,900,32 Ip

[I;.n]1I2 = 1

. Rms va ue of total input current, I. = . ~

Also 9 1 =: - a

Negative sign for 9 1 indicates that,.fundamental, current lags the sourc e voltage . The
various parameters are now obtained.

_. "

From Eq. (3.52), displacement' factor ifF = cos 81 = cos (- ~) = cos a

... (6.37)
. , ,
! .
1. , 2
10 1 2.J2
From Eq. (3.53 ), current distortion factor, CDF = -I1 =
Y-I = = 0.90032


- . 2..f2: ,
Power factor, PF = CDF x DF = - - cos' a

Harmonic factor HF



... (6.38)

.'. ' i

o~ THD =[ CDr
1~;' 1 .]'12
=[ (

From 8q ( 366 a) , voltage

r2 ]"2

25 )2-1 =0.48343-0; 48.34p%

riP,~le ~~ct~~_= [(~: J -) . , .

SubstItutIng the values of V~-fr6m' l!Iq , (6.27) apd V. fr0111 J;:q. (6. 26), we get

. [,-.Y.:;n-c.'"

1. '

- it




.. .(P9)
VRF= . -. -X ,
- 1
~ ~ . cos 2 a
8 cos2 a
Active power input. Only ti:l,e.nns fundam ental component of input current contributes
to the active input power to the ci;nverter. -

:. Active power input,

Pi = 'rms

~alue of ~ource voltage x rms fund amental component

of s ource current x displacement lactor


Vs .161 . cos 9 1


Reactive pow er input,

2{2I. } cosa = -2Vm

-]:) cOSet = V0 I0

...(6.4 0)

Qi = V,. 1st . sin et


(2{2I' )Sin e< =2V.. I , in a

cos a



sin a = V I tan


... (6.4 1 a )
.. .(6.41 0)


[Art. 6.5]

Power Electronics

6.5.2. Single-phase semi converter

For this converter, th~ variation of input current i8 in Fig. 6.11 (b), is shown continuous
from a to n, (rt + a.) to 2 1t and so on, but i, is not constant. A!J before, i, is now assumed ripple
free also with its amplitude 10 ,



a. =.!

I, . d (wt) -




I, d (WI)]


I, co. nWl d (WI) -

I, co. nWl. d (WI)J



= n~



sinnwt I~-I sinnCllt 1~+Cl]

2I, .

=- ~ sm n a ....... .. .. .. for n = I, 3, 5, .. .
=0 ............ ..................... for n = 2,4,6, .. .



I, 8in nOlt . d (Olt) -



= mt (1 + cos n a) ........ for n

I, 8in nWl. d (WI)J

= I, 3, 5 ...

= 0 ............. .. ........ .......... for n = 2, 4, 6 .. .

[ :~ ~in n a) + ( : ; (1 + cos n a) )

C. = (-


= - - (1 + cos na)l/2


It is known that 1 + cos n

e =2 cos2 ~e

2 "2 [
,n a
Cn = - - 2C08 n1t

]V2 =-cos410


a.= at n- 1[ -:;----;--",sin",:,!n,",::" ]
B =an
t -1 bn


= tan- 1

1 + cos no.

2 sin cos

. (/) = ;.

na sm
. ( nwt-2"

4 I,


2 co. l,

T ]=_ net


... (6.42)

n _ l . 3.5

Rm s value of nth harmonic input current, from Eq. (6.42), is

no 2..J21"
l ,n = 'VI?_ nit cos - 2 = n7t . cos -2

Rms fundamental circuit, I,} =


. cos -

. .. (6043 )


[Art. 6 ..5]

Phase Controlled Rectifiers

Rms value of total input current, Is =

l' (rc -




10 [

n: - a

l SI


- n- ]

9 1 = -a2


DF= cos 9 =cos ( - ~ )=cos %




_,_I =

2 -12 I



2 -12 cos !!2
cos .!!VTt
2 I , vn - a C1v~n"'(n-_-a;o

. 0

-1 ]112



- 1 ]112

8 cos 2 ex


n (n - a)

... (6.4 5}

4 (1 +cos a) .


=[ n(n-

2 v'2. cos 2"

=CDF x DF = ~'1 7t (7t - a ) x cos- 2


TC ( TC _

VRF=[ ( ~:)'

... (6.44 )




]112 cos22a


Jt -

a +

(n - a) +


2 sin 2 a

TC ( TC _




. (1 + cos a)

x ~ (1 + cos a )2 -

1 ]~.


n _ 1]
.. .(6.47}

Power input,

Pi = V,. Id cos 9 1
= V,.

2 -12. I ,



V, . I , . -12


=Tt (1 + cos a) . 10 = Vo 10
Reactive power,

( 1 + cos a.)

.. .(6AS}

Qi = V, . l' l sin 9 1
ex. a Vm

= V, . 2 "2
COS- sln "? =-10.sma

AJ.s o,

... (6.46)

.. .(5.49


[Reactive power required in I-phase full converter for the same l ol

Vo = - ( 1 + cos a )

or -

I( I
Q: = 1 + ttcosC
a S!n



0. =




= .,----"-l+ cosa

. a
o tan "2


Power Electronics

[Art. 6';J

Example 6.11. A single -phase full converter; connected to 230 V, 50 Hz source, is feeding a
load'R = 10 0 in series with a large inductance that mahes the load current r?pple free. For a
firing angle of 45 ~, calculate the input and output performance parameters of this converter.
Solution. From Eq. (6.26), V, =



cos a =

2 x .[2 x 230

cos 40 = 146.423 V

I = V, = 146 .423 = 14 6423 A




From Eq. (6. 27 ),

V,,=;n= V, = 230V

As load current is ripple free, rms value of load current,IO,. = 10 = 14.6423 A.

From Eq. (3.60),

P"" = .V, I, = 146.423 x 14.6423 = 2143.97 W

Rms value of source current , l, = 10 = l or = 14.6423 A

P~ = V" x I" = 230 x 14.6423 = 3367.73 W

Output ac power,

From Eq. (3.6 1l, rectification efficiency =



From Eq. (3.64),

~: = ~i:;:;~ = 0.6366 ?r 63.66~


Fr = V, = 146.423 = 1.5708

VRF = -iFF' - 1 = A5708' -1 = 1.2114

As load current is ripple free, CRF = 0

From Eq. 13.660 ),


I" = -

From Eq. (6.36 0 ),

x 14.6423 = 13.183 A

8, =-a=-45'
From Eq. (,.37),

DF = cos a. = cos (+ 45) = 0.707

Rms value of total input current, I, = 10 = 14.6423 A
CDF = 0.90032, PF= CDF x DF = 0.90032 x 0.707 = 0.63653 lag
THD or HF = [

1." - 1]112 = [
1 t

- 1] = 0.48342

From Eq. (6 .40 ), active power = V, I, = 146.423 x 14.6423 = 2143.97 W

From Eq. (6.41 a), reactive power


= - - 10 SIn a


x 14.6423 x sm 45 0 = 2143.963 VAr

Al so, from Eq. (6.41b), Q I = VJ, tan ex = 146.423 x 14.6423 x 1 = 2143.97 VAr
Example 6.12. Repeat E:co.mple 6. 11 for a single-phase semiconductor.
"ro x 230
Solution_ From Eq. (6.29), Vo = ---.!!!. (1 + cos a ) = (1 + cos 45) = 176.72 V

I = V, = 176.72 = 17672 A
F rom Eq. (6. 30).


=V, [


ex) + sin 2ex } ]




[MI. 6.6]

Phase Controlled Rectifiers

~ 230 [
l or

H( ~ )+

sin 90 } ]


~ 2193V

=10 (" .. load curr ent is ripple free ) =17.672 A

From Eq. (6.30),

Pdo ~ V, I ,

176.72 x 17.672 ~ 3122.996 W

Output ac power,

P ~ ~ V" I" ~ 219.3 x 17 .672 ~ 3875.47 W

. P
Rectification efficiency

~ --"'- ~ 3122.996 ~ 0 8058 80.-8%

Poe 3875.47

FF~ V,,~ 219.3

'From Eq. (6.43),

V, '176.72
VRFdFF' - 1 _ -)"1.""247:1""---;-1 ~ 0.735, CRF ~ 0
l sI =--;- x 17 .672 x cos 2= 14.697 A


I. ~ 17.672 'JT':4
~ 1'.304 A


DF = cos 9 1 =cos ( _ !! ) =cos a


CDF ~ 2..[2 x cos 22;:'

=cos 45

~ 0.9603

[ +-~)]
PF ~ CDF x DF ~ 0.9603 x 0.9239 ~ 0.88721ag


THD = CDIi' -1
Power input
Reactive power input,

]'12 =


0.9603' - 1

]11' = 0 .2905

V, I, ~ 176.72 x 17 .672 =3122.996 W


Q, = -

10 sm a

=..[2 xn 230 x 17 .672 x cos 45 =1293.79 Vp...r


A single-phase sem i-converter topology employing two SCRs and three diodes is commonly
used in industrial application s and therefore, th e term "single-phase semiconverter" implies
thi s conve rter only. Here single-phase semiconverters , or s ingl e-phase half-controlled
converters, employing two SeRs and two diodes are also studied. Both of these semi converters
hn.ve two legs.
A single-phase semiconverter, employing one SCR and one diode in . each leg, is called
5ingle -p hase symmetrica.l semiconuerter. The other configuration, using two SeRs in one leg
and two diodes in the other leg, is call ed single-phase asymmetrical semiconuerter. These are
now studied briefly.

6.6.1. Single-phase symmetrical semiconv er ter

It is also called sin gle- phase half-controlled symmelrical conv erter, or single-phase
two-pulse symmetrical converter. As stated before , it has two legs, each leg is made up of one
SCR and one diode in .:ieries as shown in Fig. 6. 17 (al. The \oltage and current waveforms are
drawn as shown in Fig. 6.17 (b ).


Powe r El ectronics

[Act. 6.6]








(a l
Fig. 6.1 7. Single-phase symmetrical semiconverter

(a )

circuit diagram and

(b )


From wi = 0 to rot = 0., let constant load current 10 free wh~el T2 Dl. Soon after wi = 0, T1
gets forward biased through T2. At a firing angle n, T1 is turned on. Load current shifts from
T2 to T1; thyristor T2 is therefore turned off. At Wi = 0, T2 is subjected to a reverse voltage of
Vm sin a which aids in the commutation of T2. Load current n oW flows through Tl, RL, Dl , b,
scurce u!' a and Tl and load voltage Uo = uab' At. rot =Tt, U, = O. Soon after wi = It, b becomes
somewhat positive with respect to a, D2 gets forward biased through D1; as a consequence, load
current now freewheels through Tl D2 fr om rot = It to (It + (X). At rot = (It + a), forward biased
T2 is tur n ed on, Tl is therefore, turned off. Current n ow fl ows through T2 D2 from
rot = (n + a ) to 2Tt and Vo = Vb' Soon after rot = 2 Tt, D1 is forward biased, therefore 10 now begins
to fre ewhe el through 12 D1 and so on. Various waveforms for voltages and currents are drawn
in Fig. 6. 17 (b ). During fr eewheeling periods, Vo = 0 be cause devices are considered ideal.
Wave form for un shows that circuit turn-off time tc is given by

n- a
t, = OJ - sec
6.6.2. Single-phase asymmetrical semiconverter
For th is semiconverter, power ci rcuit diagram is drawn in Fig. 6.18 (a ) and th e relevant
wav eforms in Fi g. 6.18 (6). An exa minat ion of Figs. 6. 17 (0 ) and 6.1S (0 ) r ev eals th at in
5 in gl~ - pha .5 e symm etrical semiconver ter of Fig. 6.17 (a) , the two ca thodes of S CRs T l and T2
are cO:l.:.ected togeth er, th es e ca thodes are, therefor e, at the same potential. As s uch. only one
trigge ring ci rcu it is suffirien t for this s emiconverter. Wh en si ngle gate pu l.;e is appli ed co both
the t:-tyri3to:'s. the S CR \vh ich is for ward-biase d at th at instan t will get ~u rn e d on . I n


P h a.~e

Co n troll ed Rectifiers


[Art. 6.7J

asymmetrical semiconver ter of Fig. 6.18 (a ), however, two separate triggering circuits must be
The thought-process, leading to the understanding of previous single-phase symmetrical
semicon verter, can now be extended to this converter configuration also.
For both symmetrical and unsymmetrical topologies, output voltage V" is given by Eq.
(6.28) and its rms value Vor by Eq. (6.29). The other perform ance parameters can be evaluated
as desi red.
0 ~r-1-~~~-----j~,~,-----,7'~-C",~,


: "~ - '~

01 :-T 'OI ~ 01 i--T202~ 01











f.-. -101 """':'



"0'1. I fI,


/ 317









(a )

Fig. 6.18. Singlephase asymmetrical semiconverter

( bl

circuit diagram and (b) waveforms.

A comparison between single- phase two-pulse type semiconverters

now be made as under :

anafull converters can

Single-phase semic on verter requires two SCRs and three (or tw o) diodes but a
single-phase full conver ter needs fr om SeRs. Single-phase semic onve rter circuits
are. therefore, cheaper.
(ii) Single-phase semiconverter offers one-quadrant operation, whereas si ngle-phase
fu ll convertor can furnish two-quadrant operation .
(iii) Freewheeling action in semiconverter circu its render power factor better than its
value in fu ll conve rter ';ucuits.

Performance parameters of semiconverter circuits a re superior than their corresponding

val ues in full conve rter circuits.

Th e advantage3 of using thre eph ri3e cont rolled converters over singl e.pr.... se cont rolled
converters are the san e (13 pn5se'::;5ed by 3-ph a3c diode rec tifiers over I-ph ~1.5e diode r -::c tiE~rs
?!lumer.J.ted 1:1 f'..rt 3.9. It i:i ?olso discu.:;sed there why thr eephase deltust::1: tr anAu rm~r i.5


Power Elcctronks

[Act. 6.7]

empl oyed for delivering power to three-phase converters . .~1 three-phase controll ed converters
use line-commutation for the turning-off of thyristors.
Three-phase thyristor conve rters may be classified as under :
Three-pulse conver ters
Six-pulse converters
(c) Twelve-pulse converters.
Six-pulse converters include 3-phase full converters, 3-phase semiconverters and six-pulse
mid-point converters . These are now described one after the other.
6.7.1 . Three-phase Half-wave Cont rolled Converter


This converter is also called 3-phase 3-pulse converter or 3-phase M-3 converter. This is
now discussed with different types of loads. Three-phase M-3 Converter with R load. Power circuit diagram of this converter is shown in Fig. 6.19 with resistive load R. A reference to the circuit of Fig. 3.29 and
waveforms of Fig. 3.30 is of considerable help. If firing angle is zero degree, SCR T 1 would
begin conducting from wt =30 0 to 1500 , T2 fr om wt =150 0 to 270 0 and T3 from rot =270 0 to
3900 and so 00. In other words, firing angle for this cont rolled converter would be measured
from rot = 30 0 for TI, from
= 150 0 for T2 and from wt = 270 0 for T3 as indicated in Fig. 6.20
(a). For zero degree firing angle delay, thyristor behaves as a diode and the voltage output
waveform Uo is as shown in Fig. 3.30 (e). The operation of this cqnverter is now described for
a < 30 0 a nd for a > 30 0












Fig. 6.19. Three-phase half-wave thyristor converter feeding R load.

Firing angle < 30. The output voltage waveform uo' for firing angle less than 30 (say
around 150) is sketched in Fig. 6.20 (b ), where T1 conducts from wt =30o +Ct to
wt = 150 0 + a =""""6 + Ct., T2 from 1500 + a to 270 0 + Ct. an d so on. Each SCR conducts for 1200
The waveform of load current


(not shown) woul d be identical with voltage waveform




Average value of outpu t voltage, Va = 21t

a-; Vmp sin wt d (wt)

.. .(6.50 )


ma.ximum value of phase (line to neu tral) voltage

V ml = maximum value of line voltage =..[3. Vmp
Ct. =flring-angle delay

Vn:p =

[Art. 6. 71

Phase Co ntr olled Rectifiers



O:" ~

IX: 0 for T3

0::0 lor T2

lor TI

lor il


.-.~ ........ t







"0" 1 I

TJ--I~T l ~/I--- T 2




",'~ "1-


,I---TJ-! 4



Fig. 6.20 . Three phlise 3pulse converter with R load (a) line to neutral source voltages.
Load. voltage waveforms for (b) 0 < a < 30 e and (c) a:> 30 e
3 Vml
Average load current, I" = Ii = 21t. R cos a

... 16.5001

Rms value of output, or load voltage is




+ 5 ""


v" = 3 V;., [ 1 wt








v;,p sin' wt. d (wt) -j'"

+"" _I sin 22 wt

0.+,; / 6


1 3 -.f3 cas2a
V or=Vmp [ 2+~


Rms Ioa d curre nt,

[1 -.f3



]t" =V [1 -.f3

V"' [1S+8rtcos2a
[or = VO
R , =[f'


'6+8ltcas2a ]'"

.. .(6.51)
... (6.5 1 a)

Firin g angle :> 30". When firing angle is more than 30", Tl would conduct from 30 ~ + a to
180",1'2 from 150" + a to 300" and so on a3 show n in Fig. 6.20 (b ). For R load, when phase
voltage v" reaches zero at rot = 180", current io = 0, '1'1 is th: refore turn ed off. Thus, Tl woul d
conduct from 30" + a to 180". Same i3 true for oth er SCR3. This show,; that each SC R, for firin g
angle:> 30", conducts for (150" - a) only. This also implies that for R load, m aximum pos.5ible
value of[iring angle is 150". Waveform ofia agrees with!l(J wav eform, Fig. 6.20 (e ). Average value
of load voltage,


Power Electronics

(Art. 6.7J


Vmp slnwt. d (wt)

n a ..

V' = -2



Rms value of output voltage, Vor = [ 2

:1. [(

~ = ';.


( 1 +cos(a + 30') I

V!p sin2 wt. d (wt)

a +1tI6


[( 5R

a )+


~ sin (2 a n

/ 3) ]'"

... (6.530)


5 R -a
sin (2a+R / 3)],"
... (6.53 b)
6 .7.1.2. Threephase M3 con vert er with RL load. In Fig. 6.19, load R is now replaced
by load RL . The load inductance L is large so that load current is continuous and constant at
10 as shown in Fig. 6.21. For firing angle < 30 0 , Vo and Vor are given respectively by Eqs. (6.50)
and (6.51). For the firing angle range of30 < a < 900 and 90 0 < a < 1800 , this converter behaves
differently as described here.
30 0 < a < 90 0 For firing angle in this range, let the firing angle be, say, 45 0 for which the various
waveforms in Fig. 6.21 are drawn. Note that T1 conducts from 30 + a to 150 + a, 12 from 150 + a
to 270 0 + a, T3 from 270 0 + a to 390 0 T a and so on. Thus, each SCR conducts for 120 0
vsl 0 &0 for T1




1. 225 V..,

l.22 SVm





' _O.t.t.8

-1.67 V,.. ..

Fig. 6.2 t.

Th r c~- pha 3e

M3 converter


fo r 30 < a < 90 for ripple-free load

C ~lrrer: ~

Phase Cont rolled Re(tifi crs


[-"I. 6.71

iTl (o r La> is not zero because of RL load. Therefore.

T1 would continue cond ucting beyond cot = it . As such , va = V~ goes negative beyond Wi = It. When
T z is t u rn e d on at cot = 150 0 + 0, load current shifts fr om Tl to T2 and a voltage
ua - Vb = [Vm si n (150 + 0) - Vm sin (30+ 0.)1 appe a rs as rever se bias ac r oss T1 t o aid its
commutation. SCR 1'2 co nducts from (150 0 + aJ to (270" + a) and 50 on . The waveform for
in or io. i n or ib and i1"3 or ie a re as shown in Fig. 6.2l.
ill Wi = it, ph ase voltage va is zero, but

The waveform Un fo r voltage ac ross Tl, on the assumption of firing angle 45 0 can be d rawn
as und er ;


is on, un = va

= 0 from cot = 75 0 to 195 0 , Fig. 6.2 1.

- u'a

\Vhen T z is on , vTI = Vo - vb from rot = 195 0 to 315 0 and

When T 3 is on, uTI = va - Vc from wt = 315" t o 435 0 and so on.
When T2 is turned on at wt = 195 0 , un = va - Ub = - V mp sin 15 - V mp sin 75" = -1.225 V mp ;
at oot = 210", vTI= -1.5 V mp ; at wt =240 0 , uTI =.J3vmp ; at wt=270",L'n=-1.5Vmp ; at
wt = 300", uTi =- V mp sin 60- 0 = -0 .866Vmp ' At wt = 3 150 , un ~_: V I11P s in45" + V "'P sin 15"

=- 0.448 Vmp '

AJso, at wt

= 315

T2 gets turned-off whereas T3 is turned on.

un = va - ue ~ - V mp sin' 45" + V mp sin 75 0 = -1.673 V mp '

This shows that at oot

in Fig. 6.21.


uTI at once changes from - 0.448



to - 1.673 Vmfl a;;

.~ hown

At cd = 330 0 , un = - V mp sin 30 - V mp = - 1.5 V mp

At Wi = 360e , un = 0 - 0.866 V mp
At rot = 390", Un

= 0.5 V mp -

0.5 V mp = 0

At Wi = 420 ~ , un = 0.866 V mp
At wi


and also


=V mp sin 75

uTI = Ua - VII =

=- 0.866 V mp

a = 0.866 Vmp
+ V mp sin 150 ~ 1.225 V mp

0 a nd so on.

Average and rms values of output voltage are the same as give n by Eqs. (6.50) and (6.51)
re5pecti vely.
90 ~ < a < 180". For firing angle in this r ange, let a be, say, 165. Under the assumption of
ripple free load current l a, the var ious waveforms are shown in Fig. 6.22. As the output voltage
wav eform u a is be low the reference line, average value of Uo must be negative . It is also eviden t
from Vo =(3 V m!/ 21t ) cos a that \yhen firing angle a is more than 900 , V o is nega tive . F o r. ~
a > 90 ~, three-phase 3-pulse converter operates as a line-commutated inverter which i:i possible
only if the load ci rcuit has a de voltage source of reve rse polarity. as in a single-phase full
converte, a lready discussed in Art. 6.3 .2. 1.

It is also seen from Figs. 6.21 and 6.22 that ave rage value of thyristor current. In.
\'201ue of so un: e curr ent, 1.. .4. = (1" x 120)/360 = 10/ 3.

o ~_ .~!.,~

,:".' . ,.; ..

~. , .. ,,\ _. . __ ... ' . n.l ""' ~ :)'

t /

o. :,ourcecurran ,


T.. -

_ [ J; x 120

~r -


J ="\3

= average


Power Electronics

1M!. 6,7 1


_o::lorT 2 - --





i1: \




r --r-Cl rOr i J - j










_ _






T2 - - - " f - - l J -



'' ''L





l: ~ 1------,



I---- 2/rI J ----I


Fig. 6.22. Three-phase 3-pulse converter waverorms ror 90 < Cl < 180 0 for constant load current.
Waveforms of i u , Lb, ic in Figs. 6.21 and 6.22 show that transformer windings have to carry
de current which is harmful to the transformer. Th e problem can, however, be sorved by using
delta-zigzag connection instead of delta-star connection as shown in Fig. 6.23 .

In this figur e, delta-zigzag transformer.feeds RL load through a 3-phase 3-pulse converter.

Load curr e nt 10 enters the neutral n of secondary zigzag, divides equally in the three
ha lfwindings, i.e . each half winding a, b, c shares a load current 10/ 3. Thi s current. fl ows
through other half windings bl , e l , aI, through SCRs TI , T2 T3 and load RL as shown in Fig.
6 .23 . Note th at each secondary winding is separated into two equal ha\y es which are
appropriately connected to result in zigzag seco ndary.
lo {3



~J lQ~


I ,~


, I,



Ie '0-e


I QI 3


F ig. 6.23. A rl elta-zigz.3g tn!:'!i>iorme:-

f~~ di r: g .1


3-pha3e 3-pulse thyristor CO!1','erte r.

[A.t. 6.7J

Phast: ConI roll ed R '! ctificrs


A care fu l observation of zigzag winding in Fig. 6.23 reveals that same phase winding,
dividing into t wo halves as a , a l ; carries current 10/ 3 in both these J:uil.ves , but in opposite
directi on s. S a me is tru e for phase b, c windings . Since each h alf, "of the three secondary
windin gs , carri es direct current in opposite direction , their magnetic effects cancel each other.
As a resul t, t he core flux and therefore core loss and temperat ure rise remain unaffected . This
show s tha t a 3-phase 3-pulse converter can be used for energizing a dc load provid ed a
delta-zigzag trans former is employed on its input side.

V .,.

Examp le 6. 13. A 3-phase M-3 conuerter is operated from 3-phase, 230 Vi 50 Hz supply
with load resistance R =10 n. An auerage output uoltage of 50% of the maximum pos.ible output
uoltage is req uired. Determine (a) the firing angle (b) auerage and rms "Glues of load current
and (c) rectification efficiency.
So lution. (a) For R load, voltage is continuous when as 30 0 and average output voltage
is given by
3 Villi
V = - - cosCt

. um
Jts maXlm

bl e va I ue IS,
.. Vo m = ~
3 VmI = 3-1221C
x 230 = 15-o.3 V


= V;m = 15;.3 =77.65 V

Requ ired average output voltage

3 Vml
V o =~ cos Ct = Y om cos

V o ="2 Vom


Thi:; sh'ows that actual value of Vo

= V"m cos Ct.,

=~ Y om = 77.65 V

:. 0

= 60

cannot be obtained for a < 30 0 , but

for a > 30 . So by using Eq (6.52), we get


3 Vm
Vo = ~ [1 + cos (0 + 300 )J = 21C


[1 + cos (a + 30 )1

[1 + cos (cu 30 )} = V

x-- = =" 3 Vml Yom 2

! = V, = 77.65 = 7 765 A
o R


Rms voltage from Eq. (6.53 bJ is

V. =


:J~ W6" -C< )+ ~ sin (2 a +n/ 3) J

= V2X 230 r (5"_ 67. 7 x n ) 1. 510

. (2 x 677
60. J ] =10476-V
2 :.r;;
. +
. ;)

!. =

1i~65 :

10.477 A

V, !.
77.65 x 7.765
0 -493
(c J R eC '~lIiler e ffi
llcle ncy = VorIlJr = 10. t 765 x 10.477 = .0
or o. 10
Exa mpl e 6.14. Deriue expressions for the auerage and rms outp ut uoltages for a 3-phase
3-plllse con trolled con uerter by using cosine /imctior. for the supply uoltage . A ssume continu ous

Powe r Electronics

l.-\ n. 6.7 J

S ol u tio n . The variation of output voltage fo r

3phase 3pulse converter is shown in Fig. 6.24 . For using
the cosine function for the input voltage, t he origin must
be taken when instantaneous voltage is maximum . So
here origin is taken at ~O' as shown in Fig. 6.24. With
~O ' as the 'o rigin, VC! = Vrn cos rot and integration must be
made from instant 1 where
where WI

=( ~ + a

rot =- ( ~- a ) to



:- -''''/3:-i.-'!p--...L''''':-~-;'a::f~-

instant 2

--Ia!- .

} Thus, t he average output voltage

Vu for a 3-pRase M3 converter is

Fig. 6.24 . Pertaining to Example 6.14

3 f,..
V.=2" J ~ [ i_"(mpCOSWld(WI)= 2;P



- sinWi

l'-j[~-. l

3Vml '
cos a = - - cos a

2 7t

.. .(6.50)


Now, rms value of output vo ltage V"r is


3 rl ~ o.
V"r = 27t J

-(1-0. )v;,p cos- rot. d (rot)

v'2 _ 3 V~p

IIr - ~

sin 2rot
wt +

Vor = ..f3 VIIIP [




- ( ~-o. )



'6 + 81t cos 2(.( J

=Vmt [ S+S1tcos2a



, .. (6.51 )

E x amp le 6.15. A 3-phase half-wave controlled converter is fed from 3phase, 400 V, 50 Hz
sou rce a nd is connected to load taking a constant cu rrent of 36 A. Thyristors hatle a voLtage
drop of 1,4 V. (a) Calculate average vbalue of load lJoltage for a firing angle 0(30 0 and 60=, (b)
Determine average and rms current ratings as well as PN of thyristors. (c) Find the auerage
p ower dissipated in each thy ristor.
Sol ution . Here, average output voltage,
Vo = 3?Vm1 cos a _ !.I T; Vml = \'2 X 400 V and

For a fir ing ar:le of 30,

(b )


= 60 0 ,

VII = 3 ~ x 400 cos 30 - 1.4 = 232.474 V
3 \ 400


Aver age current r ating of SCR, ITA =

1':15 curr ent

r:l t mg

a rSCR .

cos 60 0


= 1.4 V

1.4 = 133.63 V

I II 36
= 3 = 12 A

I Tr -!!.-~-?08
- \'3 - -./3 - - . 1 ;)-



= ,f3 V m, = Vml = V2 x 400 = 565.6 V
tc .--\n~rag'e powe r dissi p<l ted in each SCR = IT.~ I;T = 12 x 1,4 = 16.8 W

[Art. 6.iJ

Phase Controlled Rectifiers

~ Example


6.16. A 3-phase 3-pulse con verter; fed from de lta-star transform er, is con nected
to a load requiring ripple free current. A freewheeling diode is connected across the load. Shetch
waveforms for sou rce voltage, output voltage, load current, line current and free wheeling diode
current. Obtain expressions for average and rms value of output voltage. thyristor cu rrent and
freewheeling diode current.
Solution. A 3-phase 3-pulse converter feeding RL load and with freewheeling di ode across
RL is shown in Fig. 6.25 (a ). For firing angle < 30, free wheeling di ode does not come into play.
So here, fir ing angle is taken , say 60, just to illustrate how freewhe eling diode comes into
playana to examine its effect on the performance of the converte r.
At cot = Jt, as phase voltage Va tends to go negative, freew heeling diode gets forward biased
through Tl. Therefor e, freewheeling diode starts conducting from rot = Jt till 1'2 is turned on
at rot = 150 + cx. Similarly, when Vb and ve tend to go negative, freewheeling diode comes in to
play, as shown in Fig. 6.25 (b) . Note that each SCR conducts for (150 - ex) and .freewheeling
diode for (a - 30).





( 0--'-'


L--_--"'--~ J-

Bo-- -- - - - J





v. ~

O ~ ~v~~ ~
LTJ--l FD L.-

i i ----l FO L--.T 2


ll:~ !

r- 15O - et



- ..! FO ~TJ _~ FO I--n




II I', I



1\ 1

2 - )----...J




Fii 6.25. Tr...:-ee pr.:u<:l M'; con . . erte: Ca ) circuit (b ) wevefo ..m E:ulmple ~ .0.

2 9~

Power El ectro nics

[Art. 6.7)
Average value of output voltage, V" = -2



Rms thyristor current,

6 - a )+"21

V ml

Vor = 2"1Tt


V mp sin wt . d (wt)
... (6.52 )


S in

(20 + 1t/ 3)


... (6.53 b)


Average thyristor current,

:t/ 6

3 V mp
[1 +OOS (CL + nI 6))




I" = 2n

1" 51t
=21t[ 6 - CL]


{s--CL l

I, [ 2n

{s--CL lJ


Average va lu e of FO current,

Rms value of FO current ,

6.7.2. Three-phase Full Converters

If all the diodes of Fig. 3.39 are r eplaced by thyristors , a three-phase full converter bridge 3S
shown in Fig. 6.26 is obt ained. The three-phase input supply is connected to term in als .4., B, C
and the load RLE is connected across the output terminals of conv erter as shown. As in a singlephase full -converter , thyristor power circuit of Fig. 6.26
works as a three-phase ac to dc converter for firing
angle delay 0 < a. :5 90:: and as three-phase line-commutated inverte r for 90 < a < 180, A three-phase full
converter is, th erefore. preferred where regenerati on of A a 1/,
power is required. The numbering of SCRs in F ig. 6.26 B 0-- --1---4

C"---lL-_-_-_-_-~+-_-_-_-_-~:__ L! "

4is ( = I1, + 3),

+ 3), 2(
3 - 6) for group
the negatand
3, 6 (=
5 3for
th =e 5 +positive
group. This numbe ring scheme is adopted here as it
agrees with the seq uence of gating of the six thyrbtors
in a 3-phase full conv erter.
Fig. 6.26. Power ci rcu it for a 3-phase
For a =0; Tl, T 2, ......T 6 behave like diodes. This
is shown in Fig. 6. 27 (a) . The sequ ence of con duction of
full- converter feeding RLE lon d.
SCRs T 1 to T6 is also ind icate d in thi s figu re. Note that for a = O ~ , T1 is tr igge r ed at
wt = lt / 6, T2 at 90, T3 at 150 and so on. The load voltage has, therefore, the waveform as
shown in Fi g. 3.40 (c). For a =60", the conduc tion sequence of bYristors T1 to T 6 is sh own in
Fig. 6.27 ( b ). H ere T1 is triggered at wt =30" + 6 0~ = 90e , T2 at 90 + 60 = 15 0~ and so on . If the
conduction interval of va riolls thyri stors T1, T2, ..... T6 is shown firs t, then it becomes eas ier t q
draw the voltage and current waveforms . N ote that each SCR conducts for 120", when T l is
t rigge r ed , reve rse bias e d thyri stor T 5 is tu rned off a:"l.d Tl is tu rned on. T6 is al ready
condu cting . .-\s Tl is connected to A and T 6 to B, voltage Vall appears across load . It vane.:: from
1. 5 Vm to zero as shown . H ere V~.D is the ma..'"imum value of phase \oltage. When T2 is turned
on, T6 is co m mu tated fr om th e negative grou p. T1 is already conducting. As Tl and T2 are
conn ec ted to .-\ and C respective ly, voltage u:;~ 8}::pea r5 :lcross lO:ld. Its vahl e vaTi e.:: fr J ;r!
1.5 Vmp to ze ro as show n. This sequence of triggering is conti nu ed for other SeRs.

Phase Controll ed Rect ifiers


[Arl.6 .7J


t--- 120 ------i

"t T1




















I T6












Gro up













.,.V(Z GrOllp

I -~e Group

Fig. 6.27. Voltage waveforms and conduction of thyristors for a 3-phase full converter.

Note that positive group ofSCRs are fired at an interval of 120. Similarly, negative gl:uup
of SCRs are fired with an interval of 120 amongst them. But SCRs from both the groups are
fired at an interval of 60. This means that commutation occurs every 60, alternatively in
upper and lower group of SeRs. Each SCR from both groups conducts for 1200 At any time, two
SCRs, one from the positive group and the other from negative group, must conduct together
for the source to en er gise the load. For ABC phase sequence of the three ~phase supply,
thyristors conduct in pairs; TI and T2, T2 and T3, T3 and T4 and so on .
The sequence of events in Fig. 6.27 can also be shown more conveniently ifline voltages, instead
of phase voltages, are considered. In Fig. 6.28 (a) are shown line voltages Uab, uoe Ubc , Uba etc. For
a = 0, SCRs TI, T2, ....T6 behave as diodes and the output voltage waveform is as shown in Fig.
6.28 (a ) by tlab Ucc , Ubc etc. In this figure, for ex = a, Tl is t1..1l1lJa. on at o..'l = 60 0 , T2 at wt = 120, T3
at wt = 180 and so on. In Fig. 6.28 (a), therefore, firing angle is measured from wt = 60 for T1, from
wt = 120 0 for 1'2, from wt =180 0 for T3 and so on.
The question may arise in the minds oi the readers as to why TI, for a = 0, conduct.s from
wt = 60" and not from wt = 00 H ere the use of subscripts 06, ac , DC, ba etc corne to the rescu e of
readers. As observed, the subscripts in sequence appe(,'r twice. When first subscrip t ap pears
twice, the SCR in the positive group pertainin g to that line conducts for 1 20 ~ . Likevris e, when
second subscript comes twice, the SCR in th e negud ve group pert aining to th at line conduct.s
for 120. For example, first subscript 'a' appe ars hvice in tl ab Uae ; th erefor e SCR from positive


Powe r

[Art. 6.7J


group T1 will begin conduction when Uab appears i.e. at cot = 60". In uue ' uk second subscript 'C'
appears twice, the refore SCR fr om negative group T2 will begin conductioQ..w hen va,' appears
i.e. from cot = 120" in Fig. 6.28 (a). Similarly, first subscript 'b' appears twice in uk ub", so SCR
from positive group T3 will begin conduction when Vbc appears i.e. from cot = 180 0 in Fig. 6.28
(0 ).

For a = 60, T1 is turned on at cot = 60 + 60 = 120, T2 at cot = 180'\ T3 at cot =,240 and so
on. When T1 is turned on at rot 120, T5 is turned off. T6 is al ready conducting. As 'P and T6
are connected toA and B respectively, lo ad voltage must be u"b as shown in Fig. 6.28 (b). ';Vhen
T2 is turned on, T6 is commutated. As T1 and T2 are now conducting, the load voltage is u"c'
Fig. 6.28 (b). In this manner, load voltage waveform can be drawn with the turning on or. off of
uther SC Rs in seque nce. For a = 90, the load voltage is symmetrical about the reference line
wt, therefore its average value is zero. In Fi g. 6.28 (el, load current waveform i" is drawn on the
<Il~ OO


a ~O






t ",b I



"'cb ~

"'oe ~







~""J""'l ""'l """l

llokbO~CO ~Cb

l ~~v

r '~L,

+ve grouo
_ve grour


""J""J '<J ""

db doc ;.{b: ~bO ~ta ~(b




= 150C~


>--0,_ ' 50-----<



o. :1 5


~Q: , = 15 0 - - 1

~l t b r.





I-- <I, = 150---'1

l be


6 =15 0 0-----'l-j



c --------____________________________________
Id )





c -,~------,-~~-~~-~~
c:.;::::::::J wi

Fig. 6.23 . Volt3 ge and current waveforms for a 3-phase rull -conwrter for different fir ing i!ngles .
i boJ


Ph os!: Contr olle d Rec tifiers

[Art. 6. 71


assumption that load is pure L . When v cu is peak positive, slope dio /d t is maximum positive so
t ha t L. (dio/dt) equals peak positive ucn ' Similarly, dioldt is maximum negative wh en U ea iii peak
negative. When current io has peak value, dio(dt is zero and likewise Ucu is zero as shown in Fig.
6.28 (c) . For CJ. = 150, T1 is triggered at wi = 210 T2 at 270 and so on . T he output voltage
waveform is shown in Fig. 6.28 (d ). It is seen from this figur e that average voltage
is reversed
in polarity. This means that de source is delivering power to ae so urce; this is called
line-comm utated inverter operation of the 3-phase full converter bridge . It may be seen from
above t h at for a = o ~ to 90, power circuit of Fig. 6.26 works as a 3-phase full converte r
delivering power from ac source to dc load a nd for a = 90 to 180, it works as a lin e-commu tated
inverter delivering powe r from dc source to ac load. It can work in th e inverter mode only if the
load has a direct emf E due to a battery or a dc motor. It should be noted that direction of current
for both converter and inverter ope rati ons r ema in s fixed but the polarity of output voltage
reve rse!';.

Source current iA in phase A is als o draw n in Fig. 6.28 (d) for 0.= 150 e . For the a rrow
directi on indi cated in Fig . 6.26, iA is t reated as positive. Therefor e iA is positiv e wh en T1 is
conducting , i.e . when first subscript for voltages or currents is 'a'. Likewise iA is negative when
T4 is conducting, i.e. when the second subscri pt for voltages or currents is 'a' . Source current
waveforms for other two phases can al so be drawn accordingly. For other fir ing angles , so urce
currents can be drawn similarly.
Expression for the a~e ra ge output voltage Vo
ca n be obtained by referrin g to Fig. 6.29 where
uab' v a,: etc. are sketched from Fi g. 6.28 ta) fo r ii ring
anglo: de lay Ct. < 30. Note that peri odicity of output
vo ltage is it / 3 r ad ian s. Average va lue of output
voltage is ob tained by finding the dashed area

I 11 i
abed ov e r a pe r iodic cycle , Fi g. 6.29. and then l--- -1 ----.- -, ....,0 ct +ttl
dividin g it by the pe riodic tim e. With 00' as the
Fig. 6.29. Output voltage waveform for a
ori gin a t the maximum value of Ul.lb. Vo is given by
3-ph ase full co nverter.

(~ ." J

Vo =~
Vm 1coswtd(wt )
-(~ -.

n) . ( +'6.] =-3Vn-ml cos a


. ( 0.+'6 - sm
-3Vn- [ sm

... (6.54)


H ere V m: is the maximum value of line voltage.

If s ine function is used for the source voltage, then

wt = o.


=V rnl sin wi



Vo = ~


si n (,:)/;. dt wt)

!.'. - a

3\lm l

r (l 3"2. - (( ,)-

~ - - , - _ C,j S

C0 5

'1 - a l~~ =----;:CO.5 a

( n



a at


[ Ac <. ~ . 7J

P ower Electronics

Rm s value of out put voltage VIN" is

VM=[.:!.n: J
Y"' v;.,si n' wid (wi ) ]"2


a v_
~ .2

V!. : - 2 ml



~ - o

J;- .. u

(1 -

cos 2wt ) d (wt )

3 [3"+2cos2a
n ,f3

...(6 .54a )

Vor=Vml -"' fn

It is observed from Fig. 6.28 th at source cu rrent fo r phase A, i.e. iA (or fo r any other phase)
flo ws for 120 0 fo r every 180 0 Th erefore, in case outpu t current is assumed constant at / 0' the
r ms va lue of source current is
I = - II , 2n x l = 1 - ~
\J o a it
0" "3

Each SC R co ndu cts for 120 0 for every 3600 Therefore, the rms value of thyristor curren.t is

ITI, :


2 2.

\J I O T



_ It
=Io-\J a

6.7.3 . Threc Phase S emi converters

In Fig. 3.39, if diodes Dl, D3. D5 are repl aced by thyris tors Tl , T2 , T3 respectively. a
3-phase semi co nverter bri dge of Fig. 6.30 is obtain ed. A fr eewhee ling diode FD, in parallel
with RLE load , is connected acr oss the out put
t e r min a ls of th e semiconverte r as shown .
Three-phase balanced supply is given to the three
" "t2
inpu t termi na ls A, B , C of Fig. 6.30.
The outpu t voltage Vo acr oss the load terminals A
FO ~ L
is con trolled by varying the firing angles of SCRs B
T 1, T2 and T3. Th e diodes D1, D2 and D3 provide C
merely a return path for the current to the most
,. 0 1
negativ e line termina l.
The se mico nve rter bridge operation for diffe ren t
firin g (ln glcs is sh own in Fi g. 6.31 in t he form of
Fig. 6.30. Power circuit for a 3-phase
volt age a nd cu rrent wavefo rms. The conduction
semiconverter feed ing RLE load.
angles for the SeRs, diod es or FD are also shown .
For a firin g nngle delay of a: O ~, thyristors T1, T2, T3 would behave as diodes and the output
voltage I) f semicon verter would be sy mmetr ical six-pu lse per cycl e as shown in Fig. 6.3 1 (a).
The outj:o-iJt voltage consisting of pulses uel.. Ua b' v ac ' v b.: etc. shown in thi.:i figure is similar to
that shown in Fig. 3.40 (c). The output voltage consists of pulses I..'ab' L'.w t;b: ' ueo etc. as in Fig.
3.40 (c). \\:,e n the fi ring angle is delayed to a = 15 (say) as s hown in F ig. 6.3 1 (b), the triggering
of S CRs T1, T2, T3 is delayed but return diodes Dl, D2, D3 remain un affected so that only
alternate pulses are altered. The load current is continuous and h as little ripple . The FD doe5
no t c:>me into play fo r a : 15 ~. Each S CR an d diode conduct f:>r 1 20~.
In Fig. 6.31 (0), L'cb is the load vol~3.ge from wt : 00 to 60 0 As the first subscrip t indic at es
conduc ting eleme nt in th e posi tive grou p, u ~!> shows that T3 is al ready cO:1Ciucting through
di od~ D2 o~ ne gative !oTfO Up . Voltages U:tO ' Vat" indic ate that , nc:ordl:1.g to the first :subsc ript, T 1
conducts fur 1:200 nnd i~ begin.:i to conduct at wt = 6,Y for Ct = J :\3 .ihown in Fig. 6.3 1 (a ).





[Art. 6.7 1

Phase Controlled Rec tifiers




0. : 0




2 ~9

0" : 0

oo{ .:~':)." v:: 61 .. :~: .~b: I "~_: .~" __ t:~~_







' .... '

..../ ' 1'0




'- v '







I b)










~ 120--+I




" 1 1"



. 'H I group

-"(l I;jroup




V~ Q -W ~ ~ ~ ~'




Id )



" I g I:I "






! I b'.


I! I



I:, I


rig . f~31. Vol t3ge and current wa\'eforms for a 3phase semiconverte r for different firing .mg!e:;.

Similarly, t:bc. (..'bu indicate that T2 conducts for 1 20 ~ and it begins to conduct at wt = 18 0~ for
Ct. = O~ . An SCR with zero degree firing angle behaves like a simple diode. Thus, as per the
definit ion of firing angle, it should be measured from wt = 50 ~ for T1 . from wi = I BO for T2.
fro m Wi = 300 ~ for T3 and so on.
For Ct = 60\ Fig. 6.3 1 (c ), the thyristors are fired so that current retu:ns through on e diode
d'J.!'ing each 121.1 conduction period. For voltage l-'" " Tl and D3 c on dll c~ si multaneou.:;iy for
I:.W; as show n. Si milarly, other elements conduct. P'O does not COr:1 ? into play even fo r
n = 60' . Fu:[her note that .... oltage pu i:5es (Ja il. (..'be l",:o clo not appenr in the output voltage
w;1 :eform for Ct. = 60. It will be seen tha t for ex 2: 60;, \'oltag::! pu13es ~'ab ' t':", L'e:: are eli minated.
Tr.;, l o ~ ..1 C1.1!'rent, assumed co ntinuous for a = 50'> , i.:i not :5nown in Fi g ".31 <('1.


[Ar!. 6.7J

Power Electronics

For firin g angle delay 0[90, voltage , and current wave forms ar e shown in Fig. 6.31 (d ). Th e
output voltage Vo 15 discontinuous . As Vo made up of ucb. v ue, ubo' ut/,.. .. , tends to become negative
at c.ct = 12 0 ~ , 240 e , 360 ~ , FD gets forward biased . Therefore, for each periodi c cycle of 120",
output voltage is equal to line voltage for only 90" and for the r emaining 30, when FD conducts,
Vo = O. F or a = 90", conduction angle of SCRs and diodes is seen to be less than 1200 for every
output pulse. In other words , conduction angle for both positive and negative group elements is
90 n and for the remaining 30". current completes its path through FD as shown in Fig. 6.31 (d )
for a = 90. Voltage pulses uab. v b<; , Un are absent from output voltage Vo for this firing angle as
well. Withou t FD, .after load voltage Vo reaches zero, a diode from negative group would begin
to conduct reducing U o to zero till next SCR in sequence is triggered . For example, at wi = 120",
Uo = u cb = 0 and without FO, 03 from negative group would start conducting through T3 from
wt = 120" to 1500 when 'SCR Tl is gated. This means that without FO, T3 would conduct for
120 from wt = 30 to 150, 02 for 90" from cot = 30" to 120 0 and 03 for 3~ '' from wt = 120 to
150 ~ for thi s periodic cycle of 120 extending from cot = 30 to 150.
For firing angle delay of 120, the voltage and current waveforms are shown in Fig. 6.31 (e).
The load current is now assumed discontinuous. For each periodic cycle of .120, Uo is seen to
have three components . When an SCR is gated, thyristor and diode conduct for 60 only. As Vo
re aches zero and t e nds to become negative, FO gets forward bias ed and therefore starts
conducting for some angle and holds the load voltage to zero. When aU ,the energy stored in
ind uc tance is discharged, FO stops conducting and as a result, load voltage rises to load counter
e mf E. When Vo = E, n one of the elements of semiconverter bridge is conducting, this is
indicated by 0, 0 in Fig. 6.31 (e).
It may be s een from above that in a 3-phase semiconverter, SCRs are gated at an in terval
of 120:0 in a proper sequence. In a single phase semiconvertt:r, SCRs are fir ed at an inte rl<~ : of
180. In ord er to obtain full control of the dc output voltage vo, the range of firing angle is:i
A" t o 180". A three-phase semiconverter has the unique feature of working as a ' six-puls~
conuerter for a < 60 ~ and as a three-pulse con uerter for a 2: 60, a careful obser.ation of Fig. 6.31
rev eals thi s.


3phase semiconverter, each periodic cycle of output voltage has a periodicity of

, 2 0 ~ . AVI'rRga nutput voltaee should. therefore. be calculated over 120 only.
For a < 60 , For firing angle l ess than 60 , the output voltage is redrawn in Fig. 6.32 (a )
from Fig. 6.31 (6 ) for some firing angle less than 30 for convenience. In this figure , area
abcefda divid ed by :In / 3 would give the average value ofouLpuL volta ge Vo. F or area abcdc, bkc
00' as the origin a nd for area dcefd , take AA' as the origin. Then


V o - 211: (A rc a abcda + Aroa dc cfdl

11 ".

(~_au ) V m1coswt . d (Wl).1

= 2"l L [~ _ " l VM1cos wt . d(Wl) + Lv
3V .

Vo = ~ ( 1 + cos a )



With 00' as the origin, an gle




a } s meas ured to toe left of 00' . therefore minus sign

- ( -"6" - a Ij ;:'- lm l'1 nr'lj', r.11:1.
. J:i 5 .l gn 15
. pu t b elore

.. .(6.55)

Phase Controlled Rectifiers

Voltage u(.!b


at rot

) 01

[Ar!. 6. 7J

= 0 and Vc..: = 0

at wt


in Fig. 6.32 (a J. Therefore, V o can also be

obtained as
V O= -2




-3' + u

nl 3

- +0.

Vml sin Wl d (wt) +

nl 3

Vmsin wt d (wt )

3 Vrnl

=2n ( 1 + cos 0 )
Rms valu e of output voltage


v'or :

for a < 60, is given by


V. , : -23

... (6.55)

{f'" ,

_ - -a


) V;. , cos2wtd (wt) + J




3 - [ .1 rot + ------:
,.si"n wi 1""

sin 2wI
+ ..u..+

-(~-. )


V;. , cos-0 wt d( wt )}

6 "'0

Vml 2. ..f3
"'" ~ I 3 + 2 (1 + cas 2 a )


nl 6

- :"'.16


... (6.56 )

For a ~ 60. For a ~ 60, the output voltage waveform is drawn in Fig. 6.32 ( b ) for a firing
angle 60 0 < a < 90) for convenienc e. With 0 0 ' as the origin in this figur e, the ave rage output
voltage Vo is given by
Vo : 2. [Area abedaJ.
: -23





at rot

[J'- "[It'
"2 -

V ml cos

=~ and a


3V (1 + cos O J
wt . d(wt ) =-2It

is measured from rot

=r.: / 3

as sho\vn

... (6.5 5)


Fig. 6.3:2 (b ).

Therefore, V o can also be obtained as

VO = 2/t


3 V ml

= ~

0 '

V m/ sin

( 1 + cos 0 )


~ \.."+-r~- '--.j'X ~
"[ 0
' I, "

wt d (wt )
... (6.55)


liJ t

_ _ 21t / 3 _

Fig. 6.32. Ou tput \'altage

(b )
w n'/efor.n ~

for \ 3-pho.:;; e semico nvertcr fo r (e) a


60' .'l.:1d (0 ) c:. > 60'.


[Act. 6.7J

Power Eleclronks

It is se en from above that express ion for average ou tput voltage is the sam e for both
six-pulse and three-pulse operating modes of a 3-phase semi converter.
Rms value of output voltage Von for a > 60 is given by
V" = [

2~' t~" 1V~l cos' wt.

= V;' .

..y ~ [

(n _ a ) +

d (WI )

~ sin 2" ]'12

.. ,(6.57 )

Example 6.17. (0) .4. 3-phase full converter charges a battery from a three-phase supply of
230 V, 50 Hz. The battery emf is 200 V and its internal resistance is 0.5 .n. On account of
inductance connected in series with the battery, charging current is constant at 20 A. Compllte
the firing angle delay and the supply power factor.
(b) In case it is desired that power flows from dc source to ac load in part (a ), find the 'firing
angle delay for the same current.

Solution. (a ) The battery terminal voltage Vo is


Vo = 200 + 20 x 0.5 = 210 V

210 x 1t
ex =cos 3'12 x 230 = . i)

For constant load current of 10 = 20 A , Fig. 6.28 (d ) reveals that supply current i.-\ is of
rectangular (or square) wave of amplitude 20 A. It is also seen from this figure that iA flows
for 120" (or 27t/ 3 radians) over every half cycle of 180e or 1t radian s.
Rms value of the supply current I. over


radians is

2l't]112 = 20 ~
3 = 16.33 A

I , = [ ;; (20) 3

Rms value of output current, l or = 20 A

Power delivered to load
= Elo + I~r' r = 200 x 20 + (20)2 x 0.5 = 4200 W
.J3 V). cos 9 = 4200 W

Input supply pf = 'l3 x 230 x 16.33 = 0.646 lag.

(b ) "Vhen battery is delivering power, then

Vo = 200 - 20 x 0.5 = 190 V

\'!hen power flows fr om de source to ac load, the 3-ph ase full conv erter then works as a
3ph ase line comm u cated inv erter.

3 Vml
- - cos a = - 190 V


a - cos

- ,[ - 190xlt -1-l?7 7?O

3fi x 230 - ... . - .

[Art. (j,7]

Phase Controlled Rectifie rs


Examp le 6.18. For a 3-phase full converter, sketch the time uari~tlo~,s of input uoltage and
the voltage across one thyristor for one com plete cycle for a fir ing angle dela y of (aJ
0" and (b) 30~.
For both the angles, find the magnitude of reverse uoltage across thi::; SCR a nd it s
commutation time for a three-phase supply uoltage of 230 V 50 Hz.

Solution_ (0) The three-phase input vo.ltage waveform is shown as vab , v ac, Ubl.' vb" etc. in
Fig. 6.33 (a ). A conducting SCR has zero voltage across it. Let the variation of voltage across
SCR T l belonging to positive group of Fig. 6.26 be plotted in Fig. 6.33 (a).

"0. "A

Q:;OO "ab


' be


' bo

' ,b


' 0'





"0' " A.








' ob

C b)


Fig. 6.33. Pertaining to EX3mpie 6.18.

For voltages vab . v ClC ' SCR Tl conducts, therefore voltage across this SCR is VA = V,: - Va = 0
for a peri od of 120", i.e . from Wi = 0 to wt ::: 120 as shown. After Wi = 1 2 0 ~ , SCR T3 conduc ts for
120 0 wi th uk vb as the outp ut vo ltages. Now cathode of Tl is connected to supply term ina l B
through T3 for a period of 120 ~ and its anode to supply terminal A. Therefore, voltage across Tl
from wl::: 120" to 240 is VA = Va - Vb' This voltage reverse biases T1 and it is shown as Va b below
th e reference line in Fi g. 6.33 (0) .
..\Iter Wi = 240 ~ . fo r voltages U CQ Vcb; SCR T5 conducts for 120 0 and therefore cathode ofTI
is conn ected to terminal C th rough T5. Thus, voltage across T1 is VA = L':.! - Ve = Va , below th e
r efe rence line for a period of 120 0 from CJJt =240. At Wi = 21t, T1 is aga in gated and voltage
vA = 0 as shown.
For firing angle delay of zero degree, each SCR is revers e biased for 240 0
:. Com mutatio n time available for SCR turn-off = circuit



turn~o fftim e. te = ;: ::;ec

x 1000
x ?_ l't x ~- 0 = .13.33 msec.


= 3

( b ) For a = 30\ the output voltage waveform and \'oltag{' v...I. ac ross SCR Tl a re ;;l1own !n
Fig. 6.33 (b). At wi ::: 150\ when SCR T1 stops conducting, \'ol t::tge ac r053 T1 mu "t foll ow u,,/J
cu r ve fo r 1 20~ a~ d i.::i cussed in part (a) a b ove. The r efo re , VA j ump:; fr om zero [0

[An. 6.7]


I' ow ~r


Fill/si n 30 = 0.5 VItII as shown. A 3 '1'3 r emains on for 120", VA foll ows uu b curve be low the
reference li ne for 120. At cot = 270", when T5 is gated, VA follows IJ lle for 120 0 till '1'1 is gated
again, Fig. 6.33 (b ). This figure reveals that each SCR is reverse biased for (2400 - a) .
Commutation time avail a ble for SCR turn-ofT = Circuit turn*off time ,

t, =

4J't/ 3 - a



! 5~1

x 1000 =1167 msec.

In case a ~ 60, it would be obs e rved that rotc =1t - u . For both the parts , i.e. for
a = 0" and 0. = 30", the peak rever se voltage across the SCR is ~ . 230 = 325.22 V.
Example 6.19. (a) A 3-phase fullconverter feeds power to a resistive load of 10 n. For a
firing angle delay of 30, the load takes 5 k W. Find the magnitude of per phase input sllPply

(b) Repeat part (a) in case a large reactor in series with load renders the load current ripple

Solution. (a ) For a resistive load , output current waveform is of the same shape as that of
the output voltage wave . It is seen from Fig. 6.28 that for a > 60, the output vo ltage and output
current would be discontinuous for resistive load. However, for a $ 60, output voltage and
current are continuous. Rms valu e of output voltage is already obtained, for u < 60, in Eq.
16.540 ).



I f = 5000 watts and Vml = \'2

For this example,

" 2"
3 ["
2 .v;


- ._-

=5000 x 10

V, = 188.08 V


:. Per phase voltage,

(b )

V ph


=13 = 108.591 V

For a constant load current, average loa d current 10 =rms load current,

1or = VRo = ( 3V1tml casu ) ..!.



-1tcos ex



=5000 W

v, = ~50000 x \122 X 3"cos 30 = 19l.22 V


Vph = 110.40 V.

Example 6.20. (aJ A 3-phase semiconverter feeds power to a resisti ve load of 10 n. For a
firin g angle d eLay of 30, the load takes 5 k W. Find the magnitude of per phase i npll t supply
(b) Repeatpart (a) in case load current is m ade ripple free by connecting an ind uctor in series

with the load.

Ph::.sc Controlled Re ctifiers

[Art. 6.7]


Solution. It is seen from Fig. 6.31 that for ex < 30, the output voltage is continuous. For
a resistive load, output current is also continuous. R~s valu e of output voltage, for ex < 30 is
already obtained in Eq. (6.56).

;{"3 +"2 (1 + cos 60)] = 5000 x 10

... F or

a = 30', -4-

V J = 175 .67 V and Vph = 101.43 V

(a )

For constant load cu rrent, l or = ave rage load current; 10 =Ii


3 Vml
xR = - - ( I +cos30')




(1 + cos ex) R

' -=5DOOW


V, = ,50000 x 3 x 1.866 = 177.44 V and


= 102.45 V

Example 6.21. Repeat Example 6.15 in case (iring angle delay is 90

Solution. (a) F ig. 6.31 shows that for ex = 90 0 , the output voltage is discontinuous. For a
resistive load, output current is also discontinuous, for ex > 60, rms value of output voltage is
already obtained in Eq. (6.57).
: . For ex

= 90



4V; n

2 -.

'2 ) + -21 sin 180

1t - -

V, = ~50000 x


V ph

1 10

x - 1 = 5000 W

= 298.14 V

= 172 .14 V

(b) For constant load current, l or

Vo =[3V
=10 = Ii
(1 + cos a) ]

. [3



x R1

Vrnl (1 + cos 90) ] ' 10 = 5000 W



V, = ,50 DO!} x -3- = 331 ~1 53 V and Vp h = 191.2 V

Example 6.22. Repeat part (a) of Example 6.17 in case 3 -phase full conuerter is replaced
by a 3-phase semicoltuerter. .
Solution. The battery terminal voltage is

Vo = 200 + 20 x 0.5 = 210 V



3 Vml

=""'2rt" (1 + cos ex) = 210 V


-{ 210x2,

a-cos l3.'12,230-

1] -6937'

An ex aminati on of Fig. 6.31 rev eals that fo r fir ing angle a> 60 each SCR conducts for
180 - a. So, in this example, each SCR conducts for (180 - 69.37) =110.63 0 For constant load


P ower Electronics

[M" 6.7J

curr ent of 10 = 20 A, supply cu rrent iA is of squ are wave of amplitude 20 A. As iA flows for
110.63 '01 over eve ry h alf cycl e of 180", t he rm s value of supply current I , is given by


(20)' 110 .63

=20 - /110.63
180 ' 1 180
=Vol, = 210 x 20 =4200 W
pf =:r::
= 0.6724 lag.
, 3 x 230 x 15.68



Power delivered t o load

Input supply

=15.68 A

Example 6.23. A 3-phase full conuerter; fed from delta-star transformer, is connected to load
RL requiring ripple free load curren t. For a firing angle delay of around 45, sketch waveforms

{or (a) inp ut voltage u(l b. ua c ubc etc. (b) load uoltage, load current, thyristor Tl and T4 currents
and phase a , b, c cu rrents.
In case ac supply is 3-phase, 400 V, 50 Hz, leuelload current = 15 A and a =45, calculate
rec tification effic iency, TUF and input power factor.
Sol ution. Th e power circu it diagram of 3-phase full converter, fed from three-phase
delta -s t ar trans form er, is shown in Fig. 6.34 (a).
etc. is sketched in Fig. 6.34 (b ) with Ua b zero at
wt = O. The ma.ximum value of uab , Lla c or u bc isVml ' where Vm l = maximum value of line voltage
V I = ..J3 V ,"p ' Note that uab ' u oc ' ubc: etc. are displaced from each other by aD. angle of60.
(a ) The input voltage waveform Llab' Va e'

Ll bc


A 0---"









c o-_--V"

Fig. 6.34 (a ) Pertaining to Example 6. 23 .

For a firing-angle delay of 45 0 , load voltage waveform va is drawn . Load current 10 is

constant throughout, SCR Tl conducts when first subscript in load voltage is 'a'. So Tl conducts
when load voltage pul ses are uab' Uac' Likewise;T4 conducts for load voltage pulses Ube l Uco '
Source current ia' as shown in Fig. 6.34 (b ), is positive when Tl conducts and negative when
T4 ct?!1d ucts . Current ib is shown displaced by 120 0 from ill' Similarly, ic is drawn .
Here 10 = 15 A, l or = 15 A, because load current is ripple fr ee. Average value of SCR
current, fr om in waveform, is
(b )




ITA =1, x 360 ='3= '3 =5 A

Rm s value. of thyris t or current, I n-

= ~ I; x


~~~ =-t =

= B.66 A

Ph:\se Controlled Rec tifiers



fArl . 6.7]




i: v~ :

l 11a~

v'Lv" "-v,, "-v,, "-v,, ",-v" ""v"


~v" ~u, . ~""

~ ._

' Vrn Lcas 15

iO' .~----~---------II,






Fig. 6.34


Waveform for a 3-phase full converter, Example 6. 23.

Fig. 6.35 illustrates th e output voltage waveform

for a p-pulse controlled converter. A review of Art. 3.9.3
a nd Fig. 3.34 is also helpful. The origin is taken at
00' , the peak value of the instantaneous output v('lltage
va' Firing angle must be measured fr om t he
inters ecti on of phase voltage waveforms as shown .
T he r efo r e, the limits of i n t eg rati on a re from
- (


IX )

to (

( ~-a ) 0'


0 ."


PO D :



:. Average output voltage, Vo is



1--,. _11 ......'

~ + IX ) as illust r ated in Fig. 6.35.



Fig. 6.35. Output voltage waveform

p-pulse controll ed converte'F

fr(' 1V mp cos

wi. d (wt )


(1'. ) . (!! )
p . Vm o



s m Cllt


C a

a )

Rrns value of output voltage, V is


..E..... r; V: cos 2 CJ.U d (wi )
- ."- - .1
[ -9 "

Jr-: \ . .





Power Elec trDn ics

, =
p. V;
- [ -+2sin
r 21 t p
p cos2a

V,, =


v.[1 +(1;;

For 3-phase, 6-pulse controlled converter, p

rms va lu e of source current, I .r =


Here ,

From Eq . (6 .54a ),

~ I; x ~~~

3 V ml
= -n


6, therefore,

cos a =

V,,=-12 x 400 x

~, } cos 2ar

}in (


~= 15. ~

3 -12 x 400

12.25 A

cos 40 = 381.972 V

~ ~+ ~cos 90 ]= 400V

P d, = Vo I, = 381.972 x 15 = 5729.58 W

Pa.c = Vor ' lor = 400 X 15 = 6000 W

Rectification efficiency

~d' = 5~~~'g8 = 0.95493 or 95.493%


Input VA = 3 (per phase source voltage) (per phase rm s source

= 3 x T3 x 12.25

cur re:!1~)

067"1 67 - 1%
TUF = input
VA - ,f3 x 400 x 12.25 = . 0 or .0
Input pf= inputVA ='J3 x 400 x 12 .25 =0.707 lag.

6.7.4. l\'lulti-pulse Controlled Converters

In Fig. 3.31, if diodes Dl ......D6 are replaced by th:yTistors Tl, ..... T6 . a 6-pulse controlled
converter, 3-phase M-6 controlled converter or 6-phase half-waue controlled conuerter i s
obtained . Similarly, in Fig. 3.42, if twelve diodes are repl ace d by twelve SCRs, a 12-pulse
controlled converter or 3-phase 12-pulse controlled converter is obtained. For continuous
conduction mode, each SCR conducts for 2 1t radians (or 60 0 ) for 6-puise converter and for ~;
radians (or 300 ) for 12-pulse converte r. In general . for a p-pulse controlled converter, each SCR
would conduct for 21t radians .


(6) (')

V P -1t

V,, = V. [ 1

=V, [

sin -6




cos a = ..:......:...!!
1t cos a

( ~' ) cos 2a

1 +4~coS2
a ]"'-

For a 3-phase 12- pul:se converter, p = 12, therefore,

V, = Vmp '

~2 )sin ( 1~ )cOS" = 0.98816 Vmp cos O'

IAn. 6.S1

Phase Controll ed Rectifiers


1/ 2

v,,~v,[ 1+(~; )sin ( ;; )cos2a]


In a 6-pulse converter, 'Nith resistive load, continuous conduction occurs for 0 < a < 60 and
di scon tinu ous co nducti on for 60 < a s: 120". The maximum possible firin g angle is therefore,
In a 12-pulse converter, with resistive load. continuous conduction occurs for 0 < a < 75
and discontinuous conduction for 75 < a S 105. The maximum possibl e firin g angle is 105.

Here the performance parameters of 3-phase full converter are derived from the va rious
waveforms sketched in Fig. 6.36. In a 3- phase system, power-factor angle is defined a5 the
angle between phase voltage and phase current. This is the reason for showing the phase
voltage Uti, ub, Uc in Fig. 6.36. Line voltage lJab, Uac1 Uk etc. fac ili tate the sketching of ou tp ut
voltage waveform uo so these are also shown in Fig. 6.36. As before, load current is assumed ,.
rippl e fr ee.
Th e source current is is given by
I, (I) ~ I, +
II '"


t . 2. 3


(nwl + 8")

Here, 1a = 0, because positive and negative half cycles of source current is, about the
reference line OJt , a re identi cal in Fig. 6.36. Further, waveform of source cu rr ent i, reveals that

for phase '0', positive pulse is from

( H7nto(a+I~n}

l[rt ,5



Ia [
= n .1t
For n = 1,

al =


(a+~ )to ( a +

Iacos nOJt . d(cot) -


[ - sin CL.


+ 51t1S


and n ogative pulse i5 from

rTl~-.""I: 1a cosncot
0. +

-j sinnwtl


5; )



+ llr-JS ]


n Tt
1 3 I n genera,I all = - -410 . sm
n a . sm -3 ..... ror n. = , ,Q.



0 ........ . for eo = I , 4. 6 ........ ..


rr+ ,

:0:-:' IrJ sin ncot. d ( wi ) -

cr.- -

= :~[

I -cosn cot



1a' sin !H.ot d


la +S1tI6 u .'tIS

I - cosncot

la+~L-::S ]
a.,. n: / o

({~t) -j


Power Elec tronics

[A rt. 6.8 ]




"to '


I - 0. ~ !- Q---J
) -v~~- . 'v~~ '., " v~~- .



,....... . . . . . ,


o ..i'" -- 1:.~ ~.. __-"..',-._-;.':C''. "-_+-""'-_--""-_.;,
-" _-+,,-_ _ ' --;-;,
' wI

!. --i.i


i '0'

1,_-;-_ _ ,"

r" ,

- - ' - ---f------c---:c:-



~ ---'-_----;-:---L-~---L---,--_ """WI

a+ 1f


---1) _' - +-,,----ii---'-.L---;:;c; "iO'~,-_-"I_'_....J."'-_l'



" -

[ u,
Fig. 6.36. Waveforms for 3-phase full converter with


= 0 when

wt '"' O.


= - - cos n a . sm -3 ...... for n = 1,3,5, ..... .


... ... for n = 2, 4, 6, .... .. .


4 f" .
en =[ [ --;;;.
SlIlna. sm

410 mt
= - sm nn

eo = tan" (
is (t) =

. nn )'
nn ' cos nil. sm

and 8n = tan- [-tannal =-na

~: )= tan"

(- tan nIX) = -

mt .
L.. -4nnf a . sm. -3
SlIl (nrot -


... (6.60)


n " 1,3,5,

410 nn
I Jn = T22. nn sm -3
2"2 I ,
Rms value of fundamenta l current, 1d = l't sin 600

Rms value of nth harmoni c,



... (6.61}

Phase Controlled Rectifiers

[Art. 6.8]

Rm~ value of total source current, 1$=~ I;

~ ~I0-\1~




... (6.62)

DF = cos 91 = cos (- ex) = cos a


1 'I';; ~ -3 ~0 . 955
10 2 ft
CDF x DF ~ - cos a~ 0.995 cos a
I, I,f6

CDF~ - ~ -. I,x-.

PF ~

... (6.63)

HF~THD ~[C~F' - 1 1'" ~[( ~)' -1 ]"" ~0.31084 or 31084%

From Eq. (6.55),

V~ ~ V ml


is.cos 2 a ]"'
'\1 2;; "3 + 2"

_1]"' ~ [3. ~, (.!'.

~[6c:s'a (~ +
~ V~

VRF [ (



ActIVe power mput,Pj





l si_COS 9 1


= 3 - T3' ~ , 1

+ is cos 2<>
} 9. V~II . cos2a

cos ex

line voltage, VI


... (6.64)

... (6.65)

~ - . - . cos a 10= V'/o

where V$ =

_ 1]112

per-phase source voltage.

Reactive power input, Qi = 3 Vs 151 . sin 91

= 3.



T3 Tt 1



sm ex = - . - 10 , sm (X

..(6 .66 a)

=--cosaor --~-o



Q i = cos (x ' 10 , sm a. = V o1o tan

Ct .

...(6.66 b)

Example 06.24. A 3-phase full converter delivers a ripp le free load cu rrent of 10 A with a
firing angle delay of 45. TM input voltage is 3-phase, 400 V, 50 Hz.
(a) Express tM source current in Fourier series.
(b) Find tM DF, CDF, THD and PF.
(c) Calculate the active a reactive input powers.
Solution. It is seen from Eq. (60GO) that source current for a 3-phase full converter is given by

l$(t) =

I... -41

n1t ,

sm -3 sm (nwt - n. o)

""" I, 3, 5

Amplitudes of source current for different harmonics are as under:

410 . n 4 x 10 is
Sin - =
X= 11.03 A and a = 40
For n =1,
4 x 10.
For n;= 3,
8m 1t =


Power Electronics

IArt. 6.8J


For n = 5,

~x 10 sin 300" = - 2 .205 A


4 x ~ sin 420" = 1.575 A

:. i, (I) = 11.03 sin (WI - 45) - 2.205 sin 5 (WI - 45) + 1.575 sin 7 (WI - 4