Você está na página 1de 470


Input [Various Switching Converters Output[various

sources] [Efficient and Intelligent] loads]

 DC-DC conversion
 AC-DC rectification
 DC-AC inversion
 AC-AC cyclo conversion
Power processing is controlled
Input [Various Switching Converters Output[various
sources] [Efficient and Intelligent] loads]

Control input

Feed forward Controller Feed back

Few words about Power
 Power electronics refers to electronic circuits whose
primary function is to process energy
 Signal level circuits process information
 Power electronic systems comprise a combination of
these two circuit classes interposed between sources
and sinks of electric energy
 The increasing penetration of power electronics into
commercial, industrial, and military applications are
driven by three forces:
Few words..
 1. Decreasing cost making solid-state power
technology more economically attractive
 2. new semiconductor materials allowing the control
of higher voltages and operation at higher
temperatures and frequencies
 3. the need for more sophisticated and energy efficient
control process
 A most unusual application of power electronics is
gaining interest recently- the extraction of energy
from very low power sources
 These range from mechanical vibrations to
electrochemical potentials within the body
 This is an extreme power electronics challenge,
requiring conversion at very low voltages (10s of
millivolts), with an efficiency that produces usable
energy, and a size that is biologically implantable
Power processing is efficient
 Efficiency (η) = Pout/Pin
Ploss = Pin – Pout = Pout[(1/η) -1]

High efficiency leads to low power loss within converter

Small size and reliable operation is possible

Efficiency is a good measure of performance

Efficient and intelligent controller
 Recent trend is to design converters of small in size,
lighter in weight which can process substantial power
at high efficiency
 For certain class of converters (resonant, soft-
switched) efficiency is as high as 97%
 Advancement in VLSI field and signal processing has
created interest in intelligent control of these
Elements/devices for design
Elements/devices for design
Elements/devices for design
Course coverage
 Text book:
 T1. N. Mohan, T. M. Undeland, and W. P. Robbins,
Power Electronics: Converters, Applications, and
Design, John Wiley & Sons Inc. 2003, third edition.
 References:
 R1. Robert W. Erickson and Dragon Maksimovic,
Fundamentals of Power Electronics, Kluwer Academic
Publishers, second edition, 2001
 R2. Class notes and other references
Course coverage
 Introduction of power electronics
 Introduction to power processing, application of power
electronics, elements of power electronics
 Switch realization
 Switch applications, a brief survey of power
semiconductor switches (diodes, MOSFETs, BJTs,
IGBTs, thyristors), switching loss

Course coverage
 DC-DC converters
 Principles of steady state analysis, inductor volt-
second balance, capacitor charge balance, small-ripple
approximation, steady-state equivalent circuit
modeling, losses and efficiency
 Operation and design issues of buck, boost, and buck-
boost converters in continuous and discontinuous
conduction mode
 Cuk converter, isolated converters
Course coverage
 AC to DC converters.
 Line frequency diode rectifiers: line frequency AC-
uncontrolled DC
 Line frequency phase controlled rectifiers and
inverters: line-frequency ac-controlled dc
 The nature of harmonic corruption and power factor
Course coverage
 DC to AC converters.
 The switch mode voltage source inverter (VSI): design
and operational issues.
 Square wave and quasi-square wave inverter; PWM
control strategies.
 Harmonic corruption and power factor problems.
Course coverage
 Converter dynamics and control
 AC equivalent circuit modeling: Basic AC modeling
approach, state-space averaging, circuit averaging and
averaged switch modeling, canonical circuit model
 Converter transfer function and controller design
Course coverage
 Resonant and soft-switching converters
 Series resonant DC-DC converter, Parallel resonant
DC-DC converter, zero voltage and zero current
switching topologies

Course coverage
 AC-AC converters
 The AC power controller: modes of operation, power
factor and harmonic problems, applications.
 Practical issues
 Gate and base drive circuits, snubber circuits,
magnetics and heat sink design
Power loss in an ideal switch
 Switch closed v(t) = 0
 Switch open i(t) = 0
 Power loss in either case
p(t) = v(t) i(t) =0 v(t)

Ideal switch consumes zero power

Simple DC-DC converter
How to meet the load requirement of 50V, 10A
with an available source of 100V
Dissipative realization
Resistive voltage divider (efficiency = ?)
Dissipative realization
Series pass regulator: transistor operates in active region
(Efficiency = ?)
Switch realization
Single pole double throw switch (SPDT) is used
(efficiency = ?)
Voltage level is changed by duty cycle variation
D = switch duty cycle [ 0 ≤ D ≤ 1 ]
Ts = switching period; fs = switching frequency
DC value of Vs = D* Vg
Efficiency = ?
Filter for switching harmonics
Low pass filter using L and C is introduced
Filter cut-off frequency fo is much smaller than switching frequency fs
This is a buck converter
Controller for regulated output
The boost converter
Voltage level is boosted by changing the position
of switch and circuit elements
A single phase inverter
Duty cycle variation is sinusoidal leading to desired sinusoidal output
[ important advantage of switch mode circuits]
Application of power electronics
 1. At various power levels

 Less than 1 watt in battery operated portable equipments

 Tens, hundreds, and thousands watt power supplies for
computers or office equipments
 kW and MW in variable speed motor drives
 1000MW in rectifiers and inverters for utility DC
transmission line
A laptop computer power supply system
Electric vehicle power and drive system
Component weights SiC and Si
based converter
Power system of an earth orbiting
Elements of power electronics
 Analog circuits
 Electronic devices
 Control systems
 Power systems
 Electric machines
 Magnetics
 Signal processing
Looking ahead
 Converter dynamics and control

Small -signal averaged

Looking ahead
 Converter transfer function
 Controller design
 Input filter design
 Modeling of discontinuous conduction mode
 Current program control
 Latest in the field of converter design
 Various applications
Looking ahead

Transformer and inductor design Transformer size with switching frequency

Power devices with ratings
Switch realization
Semiconductor power devices behave as a SPST switch

Switch realization using semiconductor devices depends on the

(SPST) Polarity of the voltage that device must block in the OFF state,
Switch ON i>0 and on the polarity of the current that devices must conduct
Switch OFF v>0 in the ON state
Switch realization
A single quadrant switch is capable
of conducting currents of single
polarity, and of blocking
Voltage of a single polarity

Depending on the converter operating point

[Quadrant of operation] switches can be
realized with Combination of transistors
and diodes
Switch realization
Single quadrant Current bidirectional
switch Two quadrant switch

Voltage bidirectional
Two quadrant switch Four quadrant
Single-quadrant switch
Active switch: switch state is controlled by a third
control terminal

Passive switch: switch state is controlled by the

applied current and/or Voltage at terminals 1 and 2

SCR: turn-ON transition is active , while turn-OFF transition is passive

Single quadrant switch: ON state i(t) and OFF state v(t) are unipolar
The diode
A passive switch

Single quadrant switch

Can conduct positive ON-state Current

Can block negative OFF- state voltage



An active switch controlled by terminal C

Single quadrant switch which can conduct positive

ON state current and can block positive OFF state voltage

An active switch controlled by terminal C

Normally operated as single quadrant switch; can con also conduct

negative current in some cases
Realization of switch using
transistor and diode
Current bidirectional two quadrant

BJT/ anti parallel

diode realization

Instantaneous i-v characteristics

An active switch controlled by terminal C

Operated as two quadrant switch

Can conduct positive or negative ON-state current

Can block positive OFF state voltage

Two quadrant switches
MOSFET body diode

Power MOSFET Power MOSFET and Use of external diodes

characteristics Integral body diode To prevent conduction
Of body diode
A simple inverter

Current bidirectional two quadrant switches are required

Bidirectional battery

When battery is being charged iL is positive and Q1 and D2

Alternatively Conduct, when battery is being discharged,
iL is negative, and Q2 and D1 Alternatively conduct

Example of current bi-directional two quadrant switch

Switching waveforms
Switching transition
Switching loss
 Turn-on crossover interval
tc(on) = tri + tfv
 Energy dissipated during turn-on
Wc(on) = ½ Vd* Io* tc(on)
 Turn-off interval
Tc(off) = trv + tfi
 Energy dissipated during turn-off
Wc(off) = ½ Vd*Io*tc(off)
 Switching power loss
Ps = ½ *Vd*Io*fs(tc(on)+tc(off))
Switching trajectory
Safe operating area
Linear transition
Switching trajectory
Transition with ideal
Switching trajectory
Actual turn-off and trajectory
Turn-on and trajectory
Type of Commutation and Energy Loss
1. Linear
 VoffIontswitch/6
2.Rectangular (clamped inductive)
 VoffIontswitch/2
3. Inductive
 VoffIontswitch/1.5
Efficiency vs. switching frequency
Switching losses are equal to other
Converter losses at critical frequency

fcritical = (Pcond +Pfixed)/Wtot

For fs > Fcritical efficiency drops rapidly

With frequency
A brief survey of power semiconductor devices
 Power diodes

 Power MOSFETs

 Bipolar junction transistor (BJTs)

 Insulated Gate Bipolar Transistors (IGBTs)

 Thyristors (SCR, GTO, MCT)

Challenges for power
semiconductor design
 To obtain high breakdown voltage keeping on-state
resistance and forward voltage low
 This forces a longer switching time (lower switching
 To achieve high breakdown voltage, doping
concentration is low, leading to high resistivity
(dominant contributor to on-resistance of the device)
Wide band gap semiconductor
Power diodes turn-on transient

Converter current i(t) supplies

(i) charge to increase voltage across depletion region
(ii) Charge needed to support the on-state current
(iii) Charge to reduce on-state resistance of n-region
Turn-off transients
ta is the time required for the diode
Reverse current to increase from zero
to its peak IRRM

tb is the time required for trr to fall from

peak value to zero

tb/ta is the softness factor

Actual diode waveforms
Important parameters from data
Switching loss in transistor due to
diode transients

Diode recovered stored charge Qr flows through

Transistor during transistor turn-on transition,
Including switching loss
Qr depends on diode on-state forward current, and
On the rate-of-change of diode current during diode
turn-off transition
Switching loss = Wd * fs

Wd = Vg * iL * tr + Vg Qr
Principles of steady state analysis
SPDT switch changes
DC component

Switch output voltage

Vs(t) = D*Vg
D = duty cycle
0≤ D ≤1
DC component of switch output

Average value of Vs(t) = (D*Ts*Vg)/Ts = D*Vg

Low pass filter to remove switching
harmonics and pass only DC value

V = D*Vg
Basic DC-DC converter, Buck
Boost Converter
Buck-Boost converter
Inductor volt-second balance, capacitor charge
balance, small ripple (linear ripple) approximation

Buck converter containing

Practical low pass filter

Actual output voltage

v(t) = V + vripple(t)
Small ripple approximation

v(t) = V + vripple(t)
Buck converter analysis

Switch position 1 Switch position 2

Inductor current and voltage at
switch position 1

Inductor voltage vL = Vg – v(t)

Small signal approximation gives, vL = Vg – V

Also, vL = L di(t)/dt ; di(t)/dt = (Vg-V)/L

Inductor current changes with

constant slope
Inductor current and voltage at
switch position 2

Inductor voltage vL = – v(t)

Small signal approximation gives; vL = – V

Also, vL = L di(t)/dt ; di(t)/dt = (-V)/L

Inductor current changes with constant slope

Inductor voltage and current
Inductor current ripple
Inductor current at starting
Principle of inductor volt-second
 The requirement that, in equilibrium, the net change in
inductor current over one switching period be zero gives
us the steady state condition in any switching converter:
the principle of inductor volt-second balance
Principle of inductor volt-second
 Inductor relation

Integration over one switching period

In periodic steady state, net change in inductor current is zero

 The right hand side has the units of volt-seconds ,
which states that the net area or the net volt-seconds,
under the v (t) waveform must be zero

In equilibrium, the applied inductor voltage must have zero

dc component
Inductor volt-second balance: buck
converter example

Integral of voltage waveform is the area of rectangles;

 Average voltage is

Zero average voltage leads to

Principle of capacitor charge
Capacitor defining relation

Integral over one complete switching period

 In periodic steady state, net change in capacitor
voltage is zero

Hence the total area (or charge) under capacitor current waveform
is zero under steady State condition, leading to zero average
capacitor current
Boost converter analysis
Boost converter analysis

Switch position 1 Switch position 2

Subinterval1: switch in position1
 Inductor voltage and capacitor current

Small ripple approximation

Subinterval2:switch in position2
Inductor voltage and capacitor current

Small signal approximation

Inductor voltage and capacitor
current waveforms
Inductor voltage balance
Net volt-second applied to inductor
over one switching period

Equating to zero

Voltage conversion ratio

Solving for V
Conversion ratio M(D) of boost
DC component of inductor current
Capacitor charge balance

Equating to zero

Solving for I
Eliminating V
Inductor current ripple

Inductor current slope during

Subinterval 1:
Inductor current slope during
Subinterval 2:

Change in inductor current during subinterval 1

Peak ripple
Capacitor voltage ripple

Capacitor voltage slope during

Interval 1
Capacitor voltage slope during
interval 2

Change in capacitor voltage during interval 1

Peak ripple
Discontinuous conduction mode
 Minimum diode current
= (I – Δi )L

 DC component of current
= V/R

 Current ripple is

Current depends on R, but ripple does not!

Reduction of load current

Increase R until
I = ΔiL
At this condition inductor
current iL(t) and Diode current
iD(t) are both zero at the end of
Switching period. Load current
is still positive and non zero
Further reduction of load current
I < ΔiL

Diode current cannot be negative,

hence, Diode becomes reverse
biased before Ts

We have three subintervals:

D1Ts: transistor conducts
D2Ts: diode conducts
D3Ts: neither diode nor transistor
Conducts, converter operates in
Discontinuous Conduction mode
Mode boundary
Boundary condition

For a buck converter

Leads to

The expression is of the form

Variation of K and Kcrit with D

 Converter operates in DCM at low duty cycle,

and in CCM at high duty cycle
converter operates in CCM for
all duty cycle K > 1
Critical load resistance
Lab Experiment: Feb 6, 2014
Analysis of conversion ratio
Inductor volt-second balance

Capacitor charge balance

Small signal approximation

Subintervals of operations
Sub-interval 2,3
Subinterval 1

Small ripple approximation for v(t); but not for i(t)

Subinterval 2

Small ripple approximation for v(t); but not for i(t)

Subinterval 3

Small signal approximation

Inductor volt-second balance

Inductor volt-second balance

Solving for V
Capacitor charge balance
Node equation

Capacitor charge balance

Inductor current waveform
Inductor current
Peak current

Average current

Triangle formula
Equating to DC component of
DC load current
Solution for V
Inductor volt-second balance

Capacitor charge balance

Solving for V
Buck converter M(D,K)
Boost converter example(DCM)
Mode boundary
Mode boundary
K and Kcrit(D)
[Kcrit(D) vs. D]
Mode boundary(comparison of K
and Kcrit(D)
Subintervals of operations
Sub-intervals 2,3
Subinterval 1
Subinterval 2
Subinterval 3
Inductor volt-second valance
Capacitor charge balance
Inductor and diode current
Equating diode current to load
Solving for V
Boost converter characteristics
Summary of converters
Characteristics comparison in DCM
Key points
 The discontinuous conduction mode occurs in converters containing
current- or voltage- unidirectional switches, when the inductor
current or capacitor voltage ripple is large enough to cause the
switch current or voltage to reverse polarity

 Conditions for operation in the discontinuous conduction mode can

be found by determining when the inductor current or capacitor
voltage ripples and DC components cause the switch on-state
current or off-state voltage to reverse polarity

 The DC conversion ratio M of converters operating in the

discontinuous conduction mode can be found by application of the
principles of inductor volt-second and capacitor charge balance
Key points
 Extra care is required when applying the small-ripple
approximation. Some waveforms, such as the output
voltage, should have small ripple which can be neglected.
Other waveforms, such as one or more inductor currents,
may have large ripple that cannot be ignored

 The characteristics of a converter changes significantly

when the converter enters DCM. The output voltage
becomes load-dependent, resulting in an increase in the
converter output impedance
Inclusion of inductor copper loss
Dc transformer model can be extended, to include
Converter non idealities like inductor copper loss

Replace inductor with a resistor in series to represent the

copper loss
Analysis of non-ideal boost

Pos 1 Pos 2
Circuit equations, switch in
Circuit equation, switch in
Inductor voltage and capacitor
current waveforms
Solution for output voltage
Efficiency for various values of RL
Inclusion of semiconductor
conduction losses

MOSFET on-state resistance Ron

Diode : constant forward voltage VD and on resistance RD

Boost converter example
Inductor voltage and capacitor
Converter efficiency
AC-DC rectification(line frequency
ac- uncontrolled dc)

Uncontrolled utility interface

Large capacitor connected at dc side

As a filter
High distortion in input current

These rectifiers draw a highly distorted Current from


Harmonic standards and guidelines will limit the amount

of current distortion allowed into utility, and a simple
diode rectifiers may not be allowed
Basic rectifier concepts

With resistive load

With inductive load
Load with a dc back emf
Single phase diode rectifier bridge
With Ls = 0 and resistive load
Dc side as constant current
Waveforms analysis

Simplified circuit
With resistive load
With constant load current

The average value of the output voltage

𝑇/2 Vs is the rms
1 2 Value of input
𝑉𝑑0 = 2 𝑉𝑠 sin 𝑤𝑡 𝑑𝑡 = 2 𝑉𝑠 = 0.9𝑉𝑠
𝑇/2 0 𝜋 voltage
Rectifier input current

Input current with constant dc current load

Harmonic component of input

THD= 48.43%
With ac side inductance

Waveforms without Ls
Basic circuit to understand
Understanding commutation

Circuit during commutation

Waveforms during commutation
Circuit after commutation
Reduction in output voltage due to
During commutation
Current commutation in bridge
Circuit during commutation
Reduction in voltage

𝐴𝑢 = 2𝑤𝐿𝑠 𝐼𝑑

2𝑤𝐿𝑠 𝐼𝑑
𝑐𝑜𝑠𝑢 = 1 −

2𝐿𝑠 𝑤𝐼𝑑
𝑉𝑑 = 0.9𝑉𝑠 −
Rectifier with dc side voltage
Normalized Id versus Vd with
constant dc side voltage
Effect of dc side current on PF THD
and DPF
Power and harmonics in non-
sinusoidal systems

v(t) and i(t) expressed as Fourier series

Energy transmitted to load per
Average power
Net energy is transmitted to load only when Fourier series of
v(t) and i(t) contains same frequency term
RMS value and power factor
Increase in losses due to harmonics
 The presence of harmonics always increases its rms
 In case where the voltage contains only fundamental
while current contains harmonics, then the harmonics
increase the rms value of current with average power
 The result is increased (Irms)^2 R losses in the system
 Series resistances (R) always exists in the source, load,
transmission line etc.
Linear resistive load, non-
sinusoidal voltage
Non-linear dynamic load,
sinusoidal voltage
Distortion factor
Distortion factor vs THD
Diode bridge rectifier waveforms
Input line current distortion
Line voltage distortion
Line voltage distortion
L-C filtered output
Output voltage
Three phase rectifiers
Line frequency ac- controlled dc
Rectification and inversion
Resistive load
Inductive load
Load with dc side voltage
Gate trigger control circuit
Practical thyristor converters
Single phase with Ls = 0 and
constant Id
Waveforms α =0
With finite α
Dc-side voltage
ac-side quantities
Effect of Ls
Practical thyristor converters
Discontinuous conduction
Inverter mode of operation
With a dc voltage source
DC-AC inverters

Switch mode inverter for ac motor drive

DC-AC inverters

Switch mode converter for motoring and regenerative braking

in ac motor drives
DC-AC inverters
 Voltage source inverters (VSI)
 Current source inverters (CSI)
 Three categories of VSI
a) Pulse-width modulated inverters
b) square wave inverters
c) single phase inverters with voltage cancellation
PWM inverters
 DC input voltage is constant in magnitude. Diode
rectifier is used to rectify the line voltage
 Inverter must control the magnitude and the
frequency of the output voltage
 This is achieved by PWM of the inverters switches and
hence such inverters are called PWM inverters
 Various PWM schemes are available
 Sinusoidal PWM scheme is discussed
Square wave inverters
 DC input voltage is controlled in order to control the
magnitude of the output ac voltage, and therefore the
inverter has to control only the frequency of the output
 The output ac voltage has a waveform similar to a
square wave, and hence these inverters are called
square wave inverters
Single phase inverters with voltage
 In case of inverters with single phase output, it is
possible to control the magnitude and the frequency of
the inverter output voltage, even though the input to
the inverter is a constant dc voltage and the inverter
switches are not pulse width modulated.
Single phase inverter
Quadrant of operation
Full bridge topology
Output voltage
 The output voltage VAN, with respect to negative dc bus
N, is dictated by the switch states
 When TA+ is ON the output current will flow through TA+
if io is positive or it will flow through DA+ if io is negative.
In either case, TA+ being ON ensures that the point A is at
the same potential as the positive terminal of dc input
voltage, hence
VAN = Vd (TA+ is ON and TA- is OFF)
Similarly when TA- is ON, a negative io will flow through TA-
and positive io will flow through DA-, leading to VAN = 0 (if
TA- is ON and TA+ is OFF)
Output voltage
 VAN depends on the switch status and is independent of
the direction of io. Therefore , the output voltage of
converter leg A, averaged over one time period depends
on the input voltage Vd and the duty ratio of TA+
VAN = (Vd ton + 0.toff)/Ts
= Vd . Duty ratio of TA+
Similarly for leg B
VBN = Vd. Duty ratio of TB+
Output voltage; Vo = VAN - VBN
PWM with bipolar switching
 Switches (TA+ , TB-) and (TB+, TA-) are treated as switch pairs
 Switching signal is generated by comparing a switching
frequency triangular waveform V tri with the control
voltage V control
 when V control > V tri, TA+ and TB- are ON; Otherwise,
TA- and TB+ are ON
 Output voltage is given by
Vo = (Vd/V tri). V control = k . V control

 Output voltage jumps between +Vd and –Vd; Bipolar switching

PWM with bipolar switching
PWM with unipolar switching
 In full bridge converter, regardless of the direction of io, Vo = 0
if TA+ and TB+ are both ON
Similarly Vo = 0 if TA- and TB- are both ON
 A triangular waveform is compared with control voltage
Vcontrol and -Vcontrol for determining the Switching signals
for leg A and leg B respectively
 Comparison of Vcontrol with Vtri Controls leg A switches, whereas
Leg B switches are controlled by comparing –Vcontrol with Vtri
Unipolar switching
TA+ is ON if Vcontrol > Vtri

TB+ is ON if –Vcontrol> Vtri

Output voltage is given by

Vo = (Vd/Vtri)/Vcontrol

Average output voltage is

Same as in the bipolar
Switching scheme
One leg of DC-AC inverter
Sinusoidal PWM
A sinusoidal control signal at the desired frequency is
compared with a triangular waveform

Triangular waveform Vtri is at a switching frequency

fs (carrier frequency). The control Signal Vcontrol is used to
modulate the switch duty ratio and has a frequency f1,
which is the desired fundamental frequency of the inverter
voltage output
Amplitude and frequency
modulation ratio
 Amplitude modulation ratio ma is defined as
ma = (Vcontrol/Vtri)

 Frequency modulation ratio mf is defined as

mf = (fs/f1)

 The switches TA+ and TA- are controlled based on the

comparison of Vcontrol and Vtri

 When Vcontrol > Vtri ; TA+ is ON ; VAo = ½ Vd

 Vcontrol < Vtri ; TA- is ON ; VAo = - ½ Vd
Output voltage waveform
Output voltage magnitude
Output voltage
When Vcontrol = Vcontrol sin(ω1t) ; where Vcontrol ≤ Vtri

The fundamental frequency component (VAo)1 varies

Sinusoidally and in phase with Vcontrol as function of time

(VAo)1 = (Vcontrol/Vtri) sin(ω1t) Vd/2

= ma sin(ω1t) Vd/2

(VAo)1,max = ma Vd/2 ; ma ≤ 1.0

 Which shows that in a SPWM, the amplitude of
the fundamental frequency component of the
output voltage varies linearly with ma for ma ≤ 1.0,
which is the linear range of ma
Harmonic spectrum of output
Output voltage harmonics
 The harmonics appear as sidebands, centered around
the switching frequency and its multiples, that is,
around mf, 2mf, 3mf, and so on. This is true for all
values of ma in the range of 0-1
 Theoretically the frequencies at which the voltage
harmonics occur can be indicated as
fh = (jmf±k)f1
The harmonic order h corresponds to the kth
sideband of j times the frequency modulation
ration mf
h= j(mf)±k

For odd values of j, the harmonics exist only for even

values of k. for even values of j, the harmonics exist
only for odd values of k
Selection of mf
 The harmonic mf should be an odd integer. It results
in an odd symmetry [f(-t) = -f(t)] as well as half
symmetry [f(t) = -f(t+(1/2)T1)] with the time origin.

 Only odd harmonics are present and the even

harmonics are absent in the output voltage waveform
Frequency modulation ratio
 Filtering higher frequency harmonics is easier, so
switching frequency is kept higher. This leads to
more switching losses

 In most applications switching frequency is

selected to be either less than 6kHz or greater than
20kHz to be above the audible range
Small and large mf
 For small mf(mf ≤ 21), the triangular waveform signal
and the control signal should be synchronized to each

 Asynchronous PWM (mf is not an integer) results in

sub harmonics(of the fundamental frequency)

 Large mf(mf > 21)

Generalized harmonics
Over modulation(ma>1.0)
Voltage control by ma
 Simulate single phase rectifiers (both controlled and
uncontrolled) using SIMULINK and explain the waveforms
discussed in lecture class
 Repeat the same for a single phase DC-AC inverter circuit
Full bridge inverter
Unipolar voltage switching
Vcontrol > Vtri ; TA+ is ON ; VAN = Vd
Vcontrol < Vtri ; TA- is ON ; VAN = 0

(-Vcontrol) > Vtri; TB+ is ON; VBN = Vd

(-Vcontrol) < Vtri; TB- is ON; VBN = 0
Combination of switch ON-state
 Four combinations:
 1. TA+, TB- ON; VAN = Vd, VBN = 0, Vo= Vd
 2. TA-, TB+ ON; VAN = 0, VBN = Vd, Vo = -Vd
 3. TA+, TB+ ON; VAN = Vd, VBN = Vd, Vo = 0
 4. TA-, TB- ON; VAN = 0, VBN= 0, Vo = 0
 When both the upper switches are ON, the output
voltage is zero. The output current circulates in the
loop through TA+ and DB+ or DA+ and TB+, depending
on the direction of i0. During this interval, the input
current id is zero. Similar is the case when both bottom
switches are ON
Output voltage magnitude
 The output voltage changes between zero and +Vd or
between zero and – Vd
 This is known as unipolar switching

 Magnitude of fundamental voltage is given by

Vo1(peak) = ma Vd; (ma ≤1.0)
Vd< Vo1(peak) < (4/π) Vd; (ma>1.0)
Spectrum of output voltage
Output voltage harmonics
 This scheme effectively doubles the switching
frequency from output harmonics point of view
 The advantage of effectively doubling the switching
frequency appears in the harmonic spectrum of the
output voltage waveform, where the lowest harmonics
appears as sidebands of twice the switching frequency
 For even mf, the harmonic component at switching
and the sidebands of the switching frequency
harmonics disappear
Square wave operation
 Switches (TA+ , TB- ) and (TB+ , TA- ) are operated as
two pairs with a duty ratio of 0.5
 Output voltage magnitude is regulated by input dc
V01(peak) = (4/π) Vd
Output control by voltage

Switches in two legs are controlled separately

All the switches have a duty ratio of 0.5
Waveform overlap angle α can be controlled
Output voltage
Output voltage
 During this overlap interval, the output voltage is zero
 With α = 0, the output voltage is similar to the square
wave inverter with maximum possible fundamental
output magnitude
 Fundamental and harmonic frequency components of
the output voltage is given by
(Vo)h, max = (4/πh)Vd sin(hβ)
β = (90⁰ - α/2); h is an odd integer
Fundamental and harmonic voltage
Ripple in single phase inverter
Fundamental frequency

Ripple frequency component

Ripple in square wave switching
Ripple in PWM bipolar switching
Selection of inverter switching
 The PWM inverter results in a substantially smaller
peak ripple current component compared to the
square wave inverter
 Advantageous when switching harmonics are pushed
to as high frequency as possible which is achieved by
using higher switching frequency
 Higher switching frequency causes higher switching
losses, reducing the overall energy efficiency
 So, compromise must be made in selecting the inverter
switching frequency
Three phase inverter
PWM waveform
Line to line voltage and its
Blanking time
 Due to finite turn-off and turn-on times associated
with any type of switch, the turn-on of the other
switch in that inverter leg is delayed by a blanking
time tΔ which is chosen to avoid a shoot through or
cross conduction current through the leg
 The blanking time is chosen in few microseconds or
even lesser for fast devices like MOSFETs
Effect of blanking time
Effect of blanking time
 Both switches are off during the blanking time, VAN
depends on the direction iA
 The difference between the ideal and actual output
voltage is
vε = (VAN)ideal – (VAN)actual
Averaging over one time period of the switching
frequency, change in output voltage due to tΔ
ΔVAN = +(tΔ/Ts) Vd; iA >0
ΔVAN = -(tΔ/Ts) Vd; iA <0
Output voltage change
 Similarly for leg B
ΔVBN = -(tΔ/Ts) Vd; iA >0
ΔVBN = +(tΔ/Ts) Vd; iA <0

Output voltage change is given by

ΔVo = ΔVAN – ΔVBN = +(2tΔ/Ts) Vd; io >0
= -(2tΔ/Ts) Vd; io <0
Change in output voltage when
current reverses direction
Effect on sinusoidal output
Effect of output voltage distortion
 The distortion in output voltage at zero crossings
results in low order harmonics such as third, fifth,
seventh, and so on of the fundamental frequency in
the inverter output
Steady state equivalent circuit
 Pin = Pout ; Vg Ig = V I
 V = M(D) Vg ; Ig = M(D) I
Transformer model representation
At secondary side
Construction of equivalent circuit
Inductor volt-second and capacitor
charge balance
Respective expressions
Corresponding circuits
Representation dc-transformer
Equivalent model
Transformer secondary
Converter dynamics and control
What we look at to design these
 1. Dynamic model of the switching converters
 2.How do the variations in the power input voltage, the
load current, or the duty cycle affect the output
 3. what are the small signal transfer functions?
 4. steady state models will be extended to include the
dynamics introduced by the inductors and capacitors
of the converters [what about diodes?]
 Representation of physical phenomena by
mathematical means
 We model only the important dominant behavior of
the system, ignoring other insignificant phenomena
(complicated phenomena)!
 Approximation plays an important role in modeling
 Once we understand the system behavior with
approximate model, models can be refined for more
complicated observations
Neglecting the switching ripple

Modulation in duty cycle is represented by

d(t) = D + Dm cos ωm t; Where, D and Dm are constant with

|D| >> |Dm| and modulation frequency ωm is much smaller than

Switching frequency ωs
Output voltage
Output voltage spectrum

For low switching ripple, switching frequency harmonics

and its side bands are neglected, and only
modulating frequency (low frequency ) harmonics remain
Objective of ac modeling
 To predict how low frequency variation in duty cycle
induce low frequency variation in converter voltages
and currents
 Ignore the switching ripple
 Ignore complicated switching harmonics and its
 Approach will be to remove switching harmonics,
averaging all waveforms over one switching period
Averaging to remove switching

In steady state
Non-linear averaged equations
 These averaged voltages and currents are non-linear
functions of converter duty cycle, voltages and
 Linearization is done by constructing small signal
model of the converter
Buck-boost converter model
Switch at position 1

With small ripple approximation

Switch at position 2

With small ripple approximation

Averaging the inductor waveforms
Low frequency average is found by

Average inductor voltage is given by

Leads to
Inductor current
Inductor current waveforms
Averaging capacitor waveforms
Waveforms , current
Capacitor voltage
The average input current
Perturbation and linearization
Linearization around quiescent
For steady-state value of

The inductor current, capacitor voltage and input current

reaches quiescent values I, V and Vg, given by
Let small variation in

Results in change given by

Small signal assumptions
If ac variations are much smaller than corresponding quiescent

The converter equation can be linearized

Perturbation of inductor equation
The perturbed inductor equation
Neglecting the second order term
and equating dc and ac term from
both sides

This is the linearized equation which reflects

the small signal ac variations
Capacitor equations
Linearized capacitor equation
Average input current

Linearized small signal equation

Equivalent circuit model equations
Inductor loop equation
Capacitor node equation
Input port node equation
Complete equivalent circuit
Dc transformer model
Converter transfer function
The ac output voltage variation in terms of superposition
of two inputs is given by
Control to output and line to
output transfer function
Line to output transfer function
Setting d sources to zero
After dc transformer
Transfer function
Using voltage division formula

Standard form
Salient feature

Gg0 is dc gain; ω0 is corner frequency

Control to output transfer function
Gvd (s)
After dc transformer
With voltage source
With current source
Normalized form
Standard form
Salient feature
Bode plot; control to output
transfer function
Bode plot; line to output transfer
Salient features
Effect of RHP zero
State space averaging
Circuit equations
Matrix form
Basic state space average model
For a converter operating in CCM, state equations are
formulated during both subintervals as follows
During sub-interval 1

During sub-interval 2
Equilibrium(dc) state space average
Solution for dc model
Small signal ac model
Non-ideal buck-boost converter
Input, state, output
Sub-interval 1
State space
Sub-interval 2
State space
Averaged matrices
Dc state equations
Dc solutions
Small signal ac model
Small signal ac state equations
Small signal ac model
Inductor equation
Input and capacitor
Overall model
Controller design
Functional block diagram

1. Maintain constant output voltage v(t) = V

even if disturbances in vg(t) and iload(t) are there
With negative feedback
Functional block diagram with
Model representing variations in
vg, d and iload
Values for different topologies
Transfer function
Regulator small signal block
Block diagram solution
Simplified form
Line to output transfer function
Line to output transfer function without feedback loop

Line to output transfer function with feedback loop

Reduction in transfer function
Feedback reduces the transfer function by a factor

For large loop gain T(s), the output voltage

variation v resulting from vg is highly attenuated
Closed loop output impedance
Open loop output impedance

Closed loop output impedance

Reduction in output impedance
With feedback output impedance reduces by a factor

For large T(s), the influence of load current variations

on the output voltage is reduced
Transfer function from reference to
Bode plot; T(s)
Bode plot
Approximation fro T/(1+T) and
Plot of T/(T+1)
Example output to reference
Well below cross over frequency , T(s) is large
and output follows the reference with ideal gain 1/H(s)
Above cross over frequency

Which is same as the open loop transfer function from

reference to output; the loop has no effect on the transfer
function from reference to output
Plot of 1/(T+1)
Disturbance rejection

Well below cross over frequency, disturbances are reduced by a

factor 1/T(s);

Above cross over frequency feedback loop has no effect on

Addition of a feedback loop can cause an otherwise stable
system to become unstable

Possibility of right half plane poles in the closed loop transfer

Phase margin test
 Allows determination of closed loop stability directly
from the magnitude and phase of T(s)
 A good design tool: observation shows how T(s)
should be shaped to obtain good performance in
transfer function containing 1/1 + T(s) term
Stability test
The crossover frequency fc is defined as the frequency where

The phase margin φm is determined from the phase test of

T(s) at fc as
For one crossover frequency
 If there is only one cross over frequency and T(s)
contains no RHP poles then,
 The quantities 1/(1+ T(s)) and T(s)/(1+T(s)) contain no
RHP poles whenever the phase margin φm is positive
Example: stable closed loop system
Unstable closed loop system
Phase margin and closed loop
damping factor
 A small positive phase margin leads to a stable closed
loop system having complex poles near the crossover
frequency with high Q. the transient response exhibits
overshoot and ringing
 Increasing the phase margin reduces the Q. obtaining
real poles, with no overshoot and ringing, requires a
large phase margin
Closed loop response
Low Q case
High Q case
Q vs. φm
Transient response
Transient response
Regulator design; specifications
 Effect of load current variations on output voltage
regulation [limit on the maximum allowable output
 Effect of input voltage variations on the output voltage
regulation [limits the maximum line-to-output
transfer function]
 Transient response time [requires a sufficiently high
crossover frequency]
 Overshoot and ringing [adequate phase margin must
be obtained]
Lead PD compensator
Maximum phase lead
Phase lead design
For a phase lead of θ at crossover frequency, the pole and
zero frequency is selected as

For unity gain at crossover frequency

Quiescent operating point
Small signal model
Control to output transfer function
Bode plot
Line to output transfer function
Block diagram
Uncompensated loop gain
 Obtain a crossover frequency at 5kHz with phase
margin of 52 degrees
Plot with compensator
1/[1+T(s)] with lead compensator