C HAPTE R

Introduction and Single Phase Converter

fed DC Motor Drive
1.1 Introduction
Direct current (DC) motors have variable characteristics and are used
extensively in variable-speed drives, Dc motors can provide a high starting
torque and it is also possible to obtain speed control over a wide range. The
methods of speed control are normally simpler and less expensive than that
of ac drives. Dc motors play a significant role in modern industrial drives.
Both series and separately excited dc motors are normally used in variablespeed drives; however, series motors are traditionally employed for traction
applications. Due to the commutators, dc motors are not suitable for very
high-speed applications and require more maintenance than those of ac motors.
Controlled rectifiers provide a variable dc output voltage from a fixed
ac voltage, whereas choppers can provide a variable dc voltage from a fixed
dc voltage. Due to their ability to supply a continuously variable dc voltage,
controlled rectifiers and dc choppers made a revolution in modern industrial
control equipment and variable-speed drives, with power levels ranging from
fractional horsepower to several megawatts.
DC drives can be classified, in general, into three types
1. Single-phase drives
2. Three-phase drives
3. Chopper drives
1

1.2 Basic Characteristics of DC Motors

The equivalent circuit for a separately excited dc motor is shown in
Fig. 1-1. When a separately excited motor is excited by a field current of if
and an armature current of ia flows in the armature circuit, the motor develops
a back emf and a torque to balance the load torque at a particular speed. The
field current, i f of a separately excited motor is independent of the armature
current, ia, and any change in the armature current has no effect in the field
current. The field current is normally much less than the armature current.
The equations describing the characteristics of a separately excited
motor can be determined from Fig. 1.1. The instantaneous field current, if, is
described as
.... 1.1
The instantaneous armature current can be found from
di
Va = Ra i a + La a + eb
.... 1.2
dt
The torque developed by the motor is
Td = K f i f i a

.... 1.3

di f torque
The developed torque must beVequal
to
the
load
R
i
L
=
+
f
f f
f dt
d
Td = J
+ B + TL
....
1.4
dt
Where,
= motor speed, rad/s
B = viscous friction constant, N-m/rad/s
K v =voltage constant, V/A-rad/s
K a = Kv = torque constant
La = armature circuit inductance, H
L f = field circuit inductance, H
Ra = armature circuit resistance,
Rf = field circuit resistance,
TL = load torque, N-m

Figure 1.1: Equivalent circuit of separately excited dc motor

Under steady-state conditions, the time derivatives in these equations
arc zero and the steady-state average quantities are
Td = K t I f I a
Vf = Rf I f
E b = K v I f

.... 1.5
.... 1.6

Va = Ra I a + Eb = Ra I a + KV l f

.... 1.7
.... 1.8
.... 1.9

Td = B + TL

The developed power is

Pd = Td

.... 1.10

The relationship between the field current, If, and the back emf, Eg, is
nonlinear due to magnetic saturation. The relationship, which is shown in
Fig. 1-2, is known as magnetization characteristic of the motor.
3

Figure 1.2: Magnetization characteristics

From Eq. (1-3), the speed of a separately excited motor can be found

from = Va Ra I a = Va Ra I a
Kv I f

Vf

.... 1.11
We can notice from Eq.1.11 that the motor speed can be varied by
(1) Controlling the armature voltage, Va, known as voltage control;
(2) Controlling the field current, If, known as field control; or
(3) Torque demand, which corresponds to an armature current, la, for a fixed
field current lf the speed, which corre-sponds to the rated armature voltage,
rated field current and rated armature cur-rent, is known as the base speed.
Kv

Rf

In practice, for a motor speed less than the base speed, the armature
current and field currents are maintained constant to meet the torque demand,
and the armature voltage, Va, is varied to control the speed. For motor speed
higher than the base speed, the armature voltage is maintained at the rated
value and the field current is varied to control the speed. The power developed
by the motor (= torque x speed) remains constant. Figure 1.3 shows the
characteristics of torque, power, armature current, and field current against
the speed.
4

Figure 1.3: Characteristics of separately excited dc motors

The field of a dc motor may be connected in series with the armature
circuit as shown in Fig. 1.4, and this type of motor is called a series motor.
The field circuit is designed to carry the armature current. The steadystate average quan-tities are
E g = K vI f
.... 1.12
Va = Ra I a + E g
.... 1.13
Va = Ra I a + K vI a
.... 1.14
Td = K t I f I a
.... 1.15
Td = B + TL

The speed of a series motor can be determined from equation 1-13.

Figure 1.4: Equivalent circuit dc series motor

The speed can be varied by controlling the (1) armature voltage, Va,
or (2) armature current, which is a measure of the torque demand. Equation
(1-10) indicates that a series motor can provide a high torque, especially
at starting; and for this reason, series motors are commonly used in
traction applications.
For a speed up to the base speed, the armature voltage is varied
and the torque is maintained constant. Once the rated armature voltage
is applied, the speed-torque relationship follows the natural characteristic
of the motor and the power (= torque x speed) remains constant. As the
torque demand is reduced, the speed increases. At a very light load, the
speed could be very high and it is not advisable to run a dc series motor
on no-load. Figure 1.5 shows the char-acteristics of dc series motors.

Figure 1.5: Characteristics of dc series motors

1.3 Single Phase Drives
If the armature circuit of a dc motor is connected to the output of a singlephase controlled rectifier, the armature voltage can be varied by varying the delay
angle of the converter,. The forced-commutated ac-dc converters can also be
used to improve the power factor and to reduce the harmonics. The basic circuit
agreement for a single-phase converter-fed separately excited motor is shown in
Fig. 1.6. At a low delay angle, the armature current may be discontinuous, and
this would increase the losses in the motor. A smoothing inductor, Lm, is normally
connected in series with the armature circuit to reduce the ripple current to an
acceptable magnitude. A converter is also applied in the field circuit to control the
field current by varying the delay angle . Depending on the type of single-phase
converters, single-phase drives may be subdivided into;
1.
Single-phase half-wave-converter drives
2.
Single-phase scmiconverter drives
3.
Single-phase full-converter drives
4.
Single-phase dual-converter drives
7

Figure 1.6: Basic circuit arrangement of a single-phase dc drive

1.2.1 Single-Phase Half-Wave-Converter Drives
A single-phase half-wave converter feds a dc motor as shown in Fig.
1.7a. The armature current is always discontinuous unless a very large inductor
is connected in the armature circuit. A freewheeling diode is always required
for a dc motor load and it is a one-quadrant drive, as shown in Fig. 1-7b. The
applications of this drive are limited to the 1-kW power level. Figure 1-7c
shows the waveforms for a highly inductive load. The converter in the field
circuit should be a semi-converter. A half-wave converter in the field circuit
would increase the magnetic losses of the motor due to high ripple content on
the field excitation.
With a single-phase half-wave converter in the armature circuit, the
average armature voltage is given by
.... 1.16
Where Vm is the peak voltage of the ac supply. With a semiconverter in the
field circuit. the average value of field voltage is given by
.... 1.17

Figure 1.7: Single-phase half-wave converter drive.

1.2.2 Single-Phase Semi converter Drives
A single-phase semi converter feeds the armature circuit as shown in
Fig. l-8a. It is a one-quadrant drive as shown in Fig. 1-8b and is limited to
applications up to 15 kW. The converter in the field circuit should also be a
semi converter. The current waveforms for a highly inductive load are shown
in Fig. 1.8c.
With a single-phase semi converter in the armature circuit, the average
armature voltage is expressed as
.... 1.18
9

Where Vm is the peak voltage of the ac supply. With a semiconverter in the

field circuit. the average value of field voltage is given by
....1.19

Figure 1.8: Single-phase semi converter drive

1.2.3 Single-Phase Full-Converter Drives
The armature voltage is varied by a single-phase full-wave converter
as shown in Fig. 1.9a. 1t is a two-quadrant drive as shown in Fig. 1.9b and is
limited to applications up to 15 kW. During regeneration for reversing the
direction of power flow, the back emf of the motor is reversed by reversing
the field excitation.
10

Figure 1.9: Single-phase full-converter drive

The converter in the field circuit should be a full-wave converter to
reverse the polarity of field current. The current waveforms for a highly
inductive load are shown in Fig. l.9c for powering action. With a singlephase full-wave converter in the armature circuit, the average armature voltage
is given by
Va =

2v

cos aa

for 0 a
11

.... 1.20

With a single-phase full-converter in the field circuit, the average value

of field voltage is equal to
for 0 1

.... 1.21

1.2.4 Single-Phase Dual-Converter Drives

Two single-phase full-wave converters are connected as shown in Fig. 1.10.
Either converter 1 operates to supply a positive armature voltage, Va, or
converter 2 operates to supply a negative armature voltage, - Va. It is a fourquadrant drive and permits four modes of operation: forward motoring, forward
braking (regeneration), reverse motoring, and reverse braking (regeneration).
It is limited to applications up to 15 kW. The polarity of field current is reversed
during the forward and reverse regenerations. The field converter should be a
full-wave one to allow reversing the direction of field current.
If converter 1 operating with a delay angle of gives armature voltage as
.... 1.22
If converter 2 operates with a delay angle of gives armature voltage as
.... 1.23
2Vm gives the field voltage as
where With a full converter in the field
V =circuit,
cos
f

.... 1.24

Figure 1.10: Single-phase dual-converter drive

12

SOLVED EXAMPLES
1.1. The full converter is connected to a 120-V, 60-Hz supply. The load
current
is continuous and its ripple content is negligible. The turns
ratio of the transformer is unity. (a) Express the input current in a Fourier series; determine the HF of the input current, DF, and input PF. (b)
If the delay angle is
, calculate
and
Solution:
a. The waveform for input current and the instantaneous input current can be
expressed in a Fourier series as

where
2 / a

a0 =

1
2

an =

2 / a

an =

is ( t ) d (t ) =

2 + a
1 +a
I
d
wt

I a d ( wt ) = 0
(
)
a

a
+
a
2

is ( t ) cos ntd (t )

2 + a
1 /a
I
cos
n

td

I a cos ntd (t )
(
)
a

a
+
a

4Ia
PF
V
I =,.V ,/V3 , HF , DF
sin n for n=1,3,5,.....
isadc( t ) n= arms
0 + ( an cos nt + bn sin nt )
n
n =1,2,....
= 0 for n=2,4,..
=

bn =
=

2 + a

i ( t ) sin ntd (t )

1 +a
I a sin ntd (t ) I2+ a+ a I a sin ntd (t )

4Ia
cos n for n=1,3,5,.....
n
= 0 for n=2,4,.....
=

13

Because

the input current can be written as

n = tan 1

where

an
= n
bn

and
is the displacement angle of the nth harmonic current. The rms value
of the nth harmonic input current is

and the rms value of the fundamental current is

I s1 =

2 2Ia

The rms value of the input current can be calculated as

1/2

I s = I sn2
n=1,3,5,....

I s can also be determined directly from

aDF
0,cos
21/2
1/2
(
) 4 It a+ )2 2 I a
2+1a= cos
I12s1
0 n= =

222I sin
iIPF
t
=
n2
(
(==)=
cos
a
+

=
=
(
I
d
)
t
= I a=n
sssn
n(
(
)
n
n
a

2cos
a
n

n2
=
1,3,5,...
n
2I
s

the HF is found as
1/2

I 2
HF = s 1
I s1

The DF is

= 0.483 or 48.3%

The PF is found as

14

b. = / 3
and Vn = 0.5 pu

I s1 = 2 2 a = 0.90032 I a and I s = I a

and 48.34%

and
PF =

I s1
cos ( ) = 0.45 ( lagging )
Is

1.2. A 15-hp 220-V 2000-rpm separately excited dc motor controls a load

requiring a torque of TL = 45 N-m at a speed of 1200 rpm. The field circuit
resistance is Rf = 147, the armature circuit resistance is Ra = 0.25, and
2
K
Rv a=
0.25

K
t mm=0.7032
V1/2/ A
/ s voltage
2V
The
radfield
the voltage constant of the motor is K
V/A-rad/s.
c1 =0.7032

VDF
=
V
=
120
V V = 0.5
=
=
cos
cos

=
=
54.02

I(s s) cos
rms
dc
is Vf = 220 V. The viscous friction and
negligible. The
HF no-load
= 2 losses
1 are
=30.4834
I

s
1

armature current may he assumed continuous

and ripple
free. Determine

the (a) back emf, Eg; (b) required armature voltage. Va and (c) rated
armature current of the motor
Solution:
R f = 147

Vf = 220 V,
Td = TL = 45 N-m,
15

(a)

From Eq. (1-4),

Substitute respective values in above equation

From Eq. (1-2),

Substitute respective values in above equation

(b)

From Eq. (1-3),

Substitute respective values in above equation

(c)

1Hp is equal to 746 W

Substitute respective values in above equation

16

C HAPTE R

Three Phase AC-DC Converter fed DC

Motor Drive
2.1 Three-Phase Drives
The armature circuit is connected to the output of a three-phase
controlled rectifier or a forced -commutated three-phase ac-dc converter.
Three-phase drives are used for high-power applications up to megawatts
power level. The ripple frequency of the armature voltage is higher than that
of single-phase drives and it requires less inductance in the armature circuit
to reduce the armature ripple current. The armature current is mostly
continuous, and therefore the motor performance is better compared to that
of single-phase drives. Similar to the single-phase drives, three-phase drives
may also be subdivided into:
1. Three-phase half-wave-converter drives
2. Three-phase semi converter drives
3. Three-phase full-converter drives
4. Three-phase dual-converter drives
2.2 Three-Phase Half-Wave-Converter Drives
Three-phase converters provide higher average output voltage, and in
addition the frequency of the ripples on the output voltage is higher compared
to that of single-phase converters. As a result, the filtering requirement for
smoothing out the load current is simple. For these reasons, three-phase
converters are used extensively in high-power variable-speed drives.
Where Vm is the peak voltage of a Y-connected three-phase ac supply.
.... 2.1

17

With a three-phase semi converter in the field circuit, the average

field voltage is given by
.... 2.2
2.3 Three-Phase Semi converter Drives
A three-phase semi converter fed drive is a one-quadrant drive and is
limited to applications up to 115 kW.
The field converter should also be a single-phase or three-phase semi
converter, with a three-phase semi converter in the armature circuit the average
armature voltage is given by
.... 2.3
With a three-phase semi converter in the field circuit, the average field
voltage is given by
..... 2.4
2.4 Three-Phase Full-Converter Drives
A three-phase full-wave converter drive is a two-quadrant drive and is
limited to applications up to 1500 kW. During regeneration for reversing the
direction of power flow, the back emf of the motor is reversed by reversing
the field excitation. The converter in the field circuit should be a single- or
three-phase full converter to reverse the polarity of field current.
With a three-phase full-wave converter in the armature circuit the
average armature voltage is given by
.... 2.5

18

With a three-phase full converter in the field circuit the average field
voltage is given by
..... 2.6
2.5 Three-Phase Dual-Converter Drives
Two three-phase full-wave converters are connected as in Fig. 1.10a.
Either convener 1 operates to supply a positive armature voltage, Va, or
converter 2 operates to supply a negative armature voltage, -Va. It is a fourquadrant drive and is limited to applications up to 1500 kW. The polarity of
field current is reversed during forward and reverse regenerations. Similar to
single-phase drives, the field converter should be full-wave to allow reversing
the direction of field current. If converter 1 operates with a delay angle of
the average armature voltage is given by
.... 2.7
If converter 2 operates with a delay angle of a 2 gives armature voltage is
given by
a1

.... 2.8

With a three-phase full converter in the field circuit, the average field voltage
is given by
.... 2.9

Principle of three-phase half-wave converters:

Three-phase converters provide higher average output voltage, and in
adition the frequency of the ripples on the output voltage is higher compared
with that of single phase converters. As a result, the filtering requirements
19

for smoothing out the load current and load voltage sre simpler. For these
reasons, three-phase converters are used extensively in high-power variablespped drives. Three single phase half-wave converters in figure can be connected to form a three-phase half-wave converter, as shown in figure.

Figure 2.8: Three-phase half-wave converter

20

When thyristor T1 is fixed at

, the phase voltage

pears across the load until thyristor

ristor

is fixed, thyristor

is fixed at

when thy-

is reverse biased, because the line to line volt-

is negative and

age

ap-

is turned off. The phase voltage

appears across the load until thyristor

is fixed at

. When

thyristor is fixed,
is turned off and
appears across the load until
is fixed again at the beginning of next cycle. Figure shows that
.
Characteristics of the load and this is a two-quadrant converter. Figure shows
the input voltages, output voltages and the current through thyristors for a
highly inductive load. For a resistive load and
the load current would
be discontinuous and each thyristor is self-commutated when the polarity of
its phase voltage is reversed. The frequency of output ripple voltage is
.
This converter is not normally used in practical systems, because the supply
currents contain dc components. However, this converter explains the principle of the three-phase thyristor converter.
the average output voltage for a

If the phase voltage is

continous load current is

tf
>===i0
5V
3
/6362V
/6bn

V
T
3d132bn

t+)
+5m+
V
(3/V3m/6ansin
3 3V
an
cn
abs
Vdc
=
=
Vm sin td (....
t ) =2.10 m cos
dm

22 /6+
2
where Vm is the peak phase voltage. The maximum average output
voltage that occurs at delay angle
is
.... 2.11
and the normalized average output voltage is
Vh =

Vdc
= cos
Vdm

21

The rms output voltage is found from

3
Vrms =
2

5 /6 +

/6 +

Vm2 sin 2 td (t )

1
2

1
2
3
= 3Vm +
cos 2
6

For a resistive load and

Vn =

Vdc
1

=
1 + cos +

Vdm
3
6

.... 2.12

Vrms

5/3/66
9
V
T132an+
3V12m 12

3
1 td
Vdc = 5 Vm sin
t) = 1 + cos +

2
2 (
V
td (+t )22
=rms 3=V2m /6
+ Vm
+sin sin
2.5.1 Gating sequence:

6

224 /6+4 8
3
The gating sequence is as follows:
1. Generate a pulse signal at the positive zero crossing of the phase voltage
. Delay the pulse by the desired angle
and apply to the gate and
cathrode terminals of through a gate-isolating circuit.
2. Generate two more pulses of delay angle
and
gating
and
respectively through gate isolating circuits.

for

2.6 Three phase full converters:

Three-phase converters are extensively used in industrial applicatios
upto the 120-kW level, where a two-quadrant operation is required. Figure
22

as shows a full converter circuit with a highly inductive load. This circuit is
known as a three-phase bridge. The thyristors are fired at interval
.
The frequency of output ripple voltage is

and the filtering requirement is

less than that of half-wave converters. At

already conducting and thyristor
, thyristors
line voltage

, thyristor is

is turned on. During interval

and

conduct and the line-to-

appears across the load. At

ristors is fired and thyristor

off
due
to
natural

is reversed biased immediately,

commutation.
During
thyristors

is

and

, thyis turned
Internal

conduct and the line-to-

line voltage.
appears across the load. If the thyristors are numbered as
shown in figure, the firing sequence is 12, 23, 34, 45, 56 and 61. Figure
shows the waveforms for input voltage, output voltage, input current and
currents through thyristors.
If the line to neutral voltages are defined as
.... 2.13

62)/+2
bn+

V
T
sin
t )
V
t ) 2(2.14
5 / 6+ )

(t6gabantfs=/(=t=/=623VVV+(m/ansin

t....
6162ac
==
+
33VVmmsin
sin
tt +
bn
bc
cn
ca
an
bn
m
cn
an Van
cn
bn
3
26

.... 2.15
The corresponding line-to-line voltages are
.... 2.16
.... 2.17
.... 2.18
23

The average output voltage is found from

.... 2.19

Vdc

3 33V /2+c V d t = 3 /2+ 3V sin t + d

V
=dc = mcos
ab ( )
m

(
/6+
6

/6+c

Figure 2.9: Three-phase full converter

24

The maximum average output voltage for delay angle = 0 , is

.... 2.20
and the normalized average output voltage is
Vn =

Vdc
= cos
Vdm

The rms value of the output voltage is found from

3 /2+ 2 2

Vrms =
3Vm sin t + d (t )
6

/6+

1
2

.... 2.21

Vrms

Figure shows the waveforms for

for

the instanta-

neous output voltage

has a negative part. Because the current through
thyristors cannot be negative, the load voltage cannot be negative, and the
full converter behaves as the semi converter.
Gating sequence. The gating sequence

132an
/+
=6 / 63,is as follows 1
>

V
T
0, T4 3 3Vm
1crossing
Vdm =zero
1. Generate a pulse signal at the positive
3 3 of the 2phase voltage
= 3Vm +
cos 2
. Delay the pulse by the desired angle 2 4and apply
it to the gate
and cathodde terminals of through a gate-isolating circuit.
from each other for gating
2. Generate five more pulses of delayed by
and
respectively, through gate-isolating circuits.
2.7 Three-phase dual converters:
In many variable-speed drives, the four-quadrant operation is general
required and three-phase dual converters are extensively used in applications upto the 2000-01 level. Figure shows three-phase dual converters where
two three-phase converters are connected back to back. We have seen in
section 104 that due to the instantaneous voltage differences between the
25

output voltages of converters, a circulating current flows through the converters. The circulating current is normally limited by circulating reactor ,
as shown in figure. The two converters are controlled in such a way that if
is the delay angle of converter 1, the delay angle of converter 2 is
figure shows the wave forms for input voltages, output voltages, and the
voltage across inductor - . The operation of each converter is identical to
that of a three-phase full converter. During the interval
the line-to-line voltage
output of converter 1 and

appears across the

appears across converter2.

(Labbcr12 =/ 6+1 )1 t ( / 2 + 1 )

Figure 2.10: Three-phase dual converter

26

If the line-to-neutral voltage are defined as

.... 2.22
.... 2.23
.... 2.24
The corresponding line-to-line voltage are
.... 2.25
.... 2.26
.... 2.27

and
are the output voltages of converters 1 and 2 respecIf
tively instantaneous voltage across the inductor during interval
is

V
==
6V
+m01sin

11+)

(i=0201rms

02t t =(2ab
/2+bc1 ) 1 5t
an3/V

t
3
V
cos

t
=

t
t
V
t

sin

=
+

3
sin

+
(3tV=3)mVm=Vcos
(
)
sin
sin

t
t
+

sin
sin

r bc
r
m

bn
cn
ca
ab
bn
m
cn
an

an
cn
bn

m
m

L 2662 /6+1
6
Lr m Lr /6+61 663 1

.... 2.28
The circulating current can be found from

i (t )

.... 2.29
27

The circulating current depends on delay angle

and on inductance

This current becomes maximum when

and
. Even without
any external load, the converters would be continuously running due to the
circulating current as a result of ripple voltage across the inductor. This allows smooth reversal of load current during the changeover from one quadrant operation to another and provides fast dynamic responses, especially for
electrical motor drives.
Gating sequence: The gating sequence is as follows:
1. Similar to the single-phase dual converter, gate the positive converter with
a delay angle of
.
2. Gate the negative converter with a delay angle of
isolating circuits.

through gate

2.8 Three-phase PWM rectifier:

There are two circuit topologies for three phase rectifiers to a current
source rectifier where power reversal is done by dc voltage reversal and (2)
a voltage source rectifier where power reversal is by current reversal at the
dc line, figure shows the basic circuits for these two topologies.
A three phase voltage source rectifier

2with
/3 a feedback control loop is
V

IL0rfD1r2t ==
0
shown in figure. The dc link voltage is maintained at a desired reference
value by using a feedback control loop. It is measured and compared with a
. The error signal switches on and off the six switching devices
reference
of the rectifier. The power flow from and to the ac source can be controller
according to the dc-link voltage requirements. The voltage
is measured
at the dc-side capacitor CD. Controlling the dc-link voltage so that the current flow is reversed at the dc-link can control the power reversal.
In the rectifier mode of operation, the current
is positive and the
capacitor CD is discharger through the dc-load, the error signal demands the
control circuit for more power from the ac-supply. The control circuit ctakes
the power from the supply by generating the appropriate PWM signals for
28

the switching devices. More current flows from the ac to dc side, and the
capacitor voltage is recovered. In the inverter mode of operation becomes
negative and the capacitor CD is overcharged. The error signal demands the
control to discharge the capacitor and return power to the ac mains.
The PWM can control both the active power and reactive power. Thus
this type of rectifier can be used for PF correction. The ac current wave forms
can also be maintained almost sinusoidal, reducing harmonic contamination
to the mains supply. The PWM turns on and off the switches in a preestablished form, usually a sinusoidal waveform of voltage or current. An example the modulation of one phase is shown in figure with amplitude of
for the modulating signal.

VI 0mod

Figure 2.11: Basic topologies for force-commuted PWM rectifiers (a)

current source rectifier; (b) voltage source rectifier.
29

Twelve-pulse converters:
A three-phase bridge gives a six-pulse output voltage for high-power
applications such as high-voltage dc transmission and dc motor drives, a 12pulse output is generally required to reduce the output ripples and to increase
the ripple frequencies. Two & pulse bridges can be combined either in series
or in parallel to produce an effective 12-pulse output. Two configurations are
shown in figure. A 300 phase shift between secondary windings can be accomplished by connecting one secondary in y adn the other in delta

()

Configurations for 12-pulse output

30

C HAPTE R

Modeling of AC-DC Converter fed DC

drive components & design of controller
3. Modeling of AC-DC convertor fed DC drive components & design of
controller
3.1 DC Motor and Load
The dc machine contains an inner loop due to the induced emf it is not
physically seen; it is magnetically coupled. The inner current loop will cross
this back-emf loop, creating a complexity in the development of the model.

TI = Blm

Figure 3.1: DC motor and current-control loop

It is shown in Fig. 3.1; the interactions of these loops can be decoupled
by suitably redrawing the block diagram. The development of such a block
diagram for the dc machine is shown in Fig.3.2, step by step. The load is
assumed to be proportional to speed and is given as
3.1

31

To decouple the inner current loop from the machine-inherent inducedemf loop, it is necessary to split the transfer function between the speed and
voltage into two cascade transfer functions, first between speed and armature
current and then between armature current and input voltage, represented as
3.2

m ( s )
Ia ( s )

Kb
Bt (1 + sTm ) 3.3

Where
Ia ( s )
(1 + sTm )
= K1
Va ( s )
(1 + sT1 )(1 + sT2 ) 3.4
Tm =

J
Bt

.... 3.5
.... 3.6

Btm=( sB)1 +
Blm ( s ) I a ( s )

Va ( s )

Step 1

32

Ia ( s )

Va ( s )

2
B
K11= 21 t 1 Bt Ra
1 Bt Ra K b2 + R
, K b + =RaBt +
+
T1 T 2
JLa
2 J La
4 J La

Figure 3.2: Step by step reduction of Block diagram

33

3.2 Converter
The converter after linearization is represented as
Gr ( s ) =

Kr
1 + sTt 3.7

3.3 Current and Speed Controllers

The current and speed controllers are of proportional-integral type. They
are represented as

Gc ( s ) =

K c (1 + sTc )
sTc

Gs ( s ) =

K s (1 + sTs )
.... 3.8
sTs

.... 3.7

Where the subscripts c and s correspond to the current and speed

controllers, respectively. The K and T correspond to the gain and time constants
of the controllers.
3.4 Current Feedback
The gain of the current feedback is Hc. No filtering is required in most
cases. In the case of a filtering requirement, a low-pass filter can be included
in the analysis. Even then, the time constant of the filter might not be greater
than a millisecond.
3.5 Speed Feedback
Gw ( s ) =

Kw
1 + sTw 3.9

Most high performance systems use a dc tacho generator, and the filter
required is low-pass, with a time constant under 10ms. The transfer function
of the speed feedback filter is
Where Kw is the gain and Tw is the time constant.
34

3.6 Design of Controllers

The overall closed-loop system is shown in Figure 3.3, It is seen that
the current loop does not contain the inner induced-emf loop. The design of
control loops starts from the innermost (fastest) loop and proceeds to the
slowest loop, which in this case is the outer speed loop. The reason to proceed
from the inner to the outer loop in the design process is that the gain and time
constants of only one controller at a lime are solved, instead of solving for the
gain and lime constants of all the controllers simultaneously. Not only is that
logical; it also has a practical implication. Note that every motor drive need
not be speed-controlled but may be torque-controlled, such as for a traction
application. In that case, the current loop is essential and exists regardless of
whether the speed loop is going to be closed. Additionally, the performance
of the outer loop is dependent on the inner loop; therefore, the tuning of the
inner loop has to precede the design and tuning of the outer loop. That way,
the dynamics of the inner loop can be simplified and the impact of the outer
loop on its performance could be minimized. The design of the current and
speed controllers is considered in this section.

Figure 3.3: Block diagram of the motor drive

3.6.1 Current Controller
K K K H
(1 + sTc )(1 + sTm )
GH i ( s ) = 1 c r c
3.10
Tc

s (1 + sT1 )(1 + sT2 )(1 + sTr )

35

The current-control loop is shown in Figure 3.4.The loop gain function

Figure 3.4: Current-control loop

This is a fourth-order system, and simplification is necessary to
synthesize a controller without resorting to a computer. Noting that Tm is on
the order of a second and in the vicinity of the gain crossover frequency, we
see that the following approximation is valid:
1 + sTm = sTm 3.11
The converter after linearization is represented as
3.12

This reduces the loop gain function to

K=

K1K c K r H cTm
3.13
Tc

Tc = T2
GH ii ( s ) =

K (1 + sTc )
(1 + sT11 )(1 + sTr2 ))(1 + sTr )

Where
The time constants in the denominator are seen to have the relationship
Tr < T2 < T1 3.14
The equation (3.12) can be reduced to second order, to facilitate a simple
controller synthesis, by judiciously selecting
3.15

Then the loop function is

3.16
36

The characteristic equation or denominator of the transfer function

between the armature current and its command is

(1 + sT1 )(1 + sTr ) + K 3.17

This equation is expressed in standard form as
3.18

From which the natural frequency and damping ratio are obtained as
.... 3.19

Tl + Tr

Tl + Tr

=
K +1
2
TT
l r

.... 3.20

Where
and are the natural frequency and damping ratio,
respectively. For good dynamic performance,
it is an2 accepted practice to
In >>
T
T2K
+lT+1
r T
T
K + 1
Tdamping
2
l +
r r s + ratio to 0.707 in
have a damping ratio of 0.707. HenceTT
the
s
+
1equaling
=

r
T + T
n
TT
T
+
TT
equation (3.20), we get
l Tr r
l r

K +1= l r
2
TlTr
.... 3.21

Realizing that

K >> 1 3.22
3.23

37

Tells us that K is approximated as

3.24

By equaling (3.13) to (3.24), the current-controller gain is evaluated as

Kc =

1 TI Tc
1

3.25
2 Tr K I K r H cTm

3.6.2 First-Order Approximation of Inner Current Loop

To design the speed loop, the second-order model of the current loop is
replaced with an approximate first-order model. This helps to reduce the order
of the overall speed-loop gain function.

TI2
T
K
I
2TI Tr 2Tt

Figure 3.5: Simplified Current control loop

The current loop is approximated by adding the time delay in the
converter block to T; of the motor: because of the cancellation of one motor
pole by a zero of the current controller, the resulting current limp can he
shown in Figure 3.5. The transfer function of the current and its commanded
value is
38

 3.26

where T3 = TI + Tr'
The transfer function can be arranged simply as
3.27

Where

Ti =

T3
(1 + sTi ) .... 3.28

Ki =

K fi

K fi =

K c K r K lTm H c
.... 3.30
Tc

Hc

1
(1 + K fi )

.... 3.29

1
Ia ( s )
KKi c K I TI Tm
=
The resulting model of the current
loop is a first-order system
(1 + sT3suitable
)
IIaa*((ss)) (1 + sT3 T) c
=
for use in the design of a speed loop the *gain and delay
of
the
current
loop
1 can
I a ( s ) 1 + K I K c K r H cTm
also be found experimentally in a motor-drive system. That would be more
Tc
(1 + sT3 )
accurate for the speed-controller design.
3.6.3 Speed Controller
The speed loop with the first-order approximation of the current-control
loop is shown in Figure 3.6.The loop gain function is
K K K H H
(1 + sTs )
GH s ( s ) = s i b c
3.31
BtTs

(1 + sTi )(1 + sTm )(1 + sT )

39

This is a fourth-order system. To reduce the order of the system for

analytical design of the speed controller, approximation serves. In the vicinity
of the gain crossover frequency, the following is valid

(1 + sTm ) = sTm 3.32

Figure 3.6:Representation of the outer speed loop in the dc motor drive

The next approximation is to build the equivalent time delay of the
speed feedback filter and current loop. Their sum is very much less than the
integrator lime constant, Ts, and hence theKequivalent
time delay. T4 can be
(1 + sT )
i K b H K s
KT,
K 2This
2is verys similar to
2 =
( s )BTw.
considered the sum of the two delays,GH
step
sand

s (1 + sT4 )
tTm Ts
the equivalent time delay introduced in the simplification
of the current-loop
transfer function. Hence, the approximate gain function of the speed loop is
.... 3.33

T4 = Ti + T

.... 3.34
.... 3.35

Where
The closed-loop transfer function of the speed to its command is
40

K2Ks
(1 + sTs )
m ( s )
Ts
=
r* ( s ) s3T + s 2 + sK K + K 2 K s
4
2 s
Ts

( a0 + a1s )
1
=

H ( a0 + a1s + a2 s 2 + a3 s 3 )

a0 =

....3.36

K2 Ks
Ts

a1 = K 2 K s

Where

a2 = 1

3.37

a3 = T4

This transfer function is optimized to have a wider bandwidth and a

magnitude of one over a wide frequency range by looking at its frequency
response. Its magnitude is given by

m ( j )
1
a02 + 2 a12
=
r* ( j ) H {a02 + 2 ( a12 2a0 a2 ) + 4 ( a22 a1a3 ) + 6 a32 3.38
This is optimized by making the coefficients of 2 and 4 equal zero,
to yield the following conditions:
a12 = 2a0 a2
a22 = 2a1a3

Substituting these conditions in terms of the motor and controller

parameters given in (3.37) yields
Ts2 =

2Ts
K s K 2 3.39
41

Resulting in
Ts K s =

2
K 2 3.40

Similarly,

Ts2
2Ts2T4
=
3.41
K s2 K 22 K s K 2
This, after simplification, gives the speed-controller gain as
Ks =

1
2 K 2T4 3.42

Substituting equation (3.42) into equation (3.39) gives the lime constant
of the speed controller as
Ts = 4T4 3.43
3.44

Substituting for Ks and Ts into (3.37) gives the closed-loop transfer

function of the speed to its command as
1 m4T( s ) 1

1 + 4T4 s
4 =
function the corner
*
2 2
3 3
It is easy to prove that for that open-loop
gain
points
r ( s ) H 1 + 4T4 s + 8T4 s + 8T4 s
1
are
and T4 , with the gain crossover frequency being 1 2T4 .In the
vicinity of the gain crossover frequency, the slope of the magnitude response
is -20dB/decade, which is the most desirable characteristic for good dynamic
behavior Because of its symmetry at the gain crossover frequency, this transfer
function is known as a symmetric optimum function. Further, this transfer
function has the following features
i.
Approximate time constant of the system is 4T4.
ii

t
t

3t
1
2T4
4T4
r ( t ) =
2e cos
1 + e
3.45
H
4
T
4

42

The step response is given by

With a rise time of 3.1T4 a maximum overshoot of 43.4%, and a
settling time 16.5T4.
iii. Since the overshoot is high, it can be reduced by compensating
for its cause, i.e., the zero of a pole in the speed command path,
as shown in Figure 3.7. The resulting transfer Function of the
speed to its command is
3.46

3.47

Whose step response is

t
1 a 2compensator
3 21 4Tt 4
3t
Figure 3.7: Smoothing of theovershoot
m ( s ) 1 11 via
T4
s r=*(

t ) == ; 1 + e j
e2 2 sin 3 3
r ( s )2T4HH 4T
1 + 4T44 T
s 4+ 83T4 s + 8T4 s4T4

4
With a rise time of 7.6T4, a maximum overshoot of 8.1% and a settling
time of 13.3T4. Even though the rise time has increased, the overshoot has
been reduced to approximately 20% of its previous value, and the settling
time come down by 19%.
iv. The poles of the closed-loop transfer function are

3.48

The real parts of the poles are negative, and there are no repeated poles
at the origin, so the system is asymptotically stable. Hence, in the
symmetric optimum design, the system stability is guaranteed, and there
43

is no need to check for it in the design process. Whether this is true for
the original system without approximation will be explored in the
following example.
v. Symmetric optimum eliminates the effects due to the disturbance very
rapidly compared to other optimum techniques employed in practical
systems, such as linear or modulus optimum. This approach indicates
one of the possible methods to synthesize the speed controller. Thai the
judicious choice of approximation is based on the physical constants
of the motor, on the converter and transducer gains, and on time delays
is to he emphasized hurt.

T4 = Ti + T
=

T3
+ T
1 + K fi

T + Tr
T4 = l
+ T
1 + K fi

3.49

That the speed-loop transfer function is expressed in terms of T4 is

significant in That it clearly links the dynamic performance to the speedfeedback and current-loop time constants, that a faster current loop with a
smaller speed-filler time constant accelerates the speed response is evident
from this. Expressing T4 in terms of the motor, the converter and transducer
gains, and the time delays by using expressions (3.28) and (3.34) yields
Since Kfi >> 1, T4 is found approximately after substituting for Kfi from
equation (3.29) in terms of gains and time delays as

T4

(Tl + Tr ) T2
Tm

1
+ T 3.50
Kl K c K r H c

This clearly shows the influence of the subsystem parameters on the

system dynamics. A clear understanding of this would help the proper selection
of the subsystems to obtain the required dynamic performance of the speedcontrolled motor-drive system. Further, this derivation demonstrates that the
system behavior to a large degree depends on the subsystem parameters rather
44

than only on the current and speed controller parameters or on the

sophistication of their design.
3.7 Two-Quadrant Dc Motor Drive with Field Weakening
Flux weakening is exercised in the machine by varying the voltage
applied to the field winding. The applied voltage is varied via a single-phase
or three-phase half-wave controller rectifier. Because of the large inductance
of the field winding, the field current is smooth and hardly has ripple. The
converter time lag is negligible compared to the field time constant, and the
speed of response is determined primarily by the field time constant. Therefore,
there is very little justification for using full-wave converters in motor drives
of rating less than 100 kW. A schematic of the two-quadrant dc motor drive
with flux weakening is shown in Figure 3.8.

Figure 3.8: A two-quadrant dc motor drive with field weakening

3.51

The command field current is determined by the induced-emf error

function. The induced-emf error is the difference between the reference induced
emf and the estimated machine-induced emf. The machine-induced emf is
found by subtracting the resistive and inductive drop from the applied voltage.
This induced-emf estimator is machine-parameter sensitive and needs to be
adaptive. The induced-emf error is amplified by a PI controller and limited to
yield the field-current reference. The field-current reference is enforced by a
current-control loop feedback very similar to the armature current-control
45

loop. The outer induced-emf feedback loop enforces a constant induced emf
for speeds higher than the base speed. This amounts to
If the induced emf en is kept at rated value, say 1 p.u. then the field flux
is inversely proportional to the rotor speed. This condition also enables
constant-air gap-power operation.
3.8 Four-Quadrant DC Motor Drive
A four-quadrant dc motor drive has a set of dual three-phase converters
for the power stage. Its control is very similar to that of the two-quadrant dc
motor drive. Converters have to be energized depending on the quadrant of
operation. Converters 1 and 2 are for forward and reverse directions of rotation
of the motor, respectively. The changeover from one converter to another is
safely handled by monitoring speed, current-command, and zero-crossing
current signals. These signals form the inputs to the selector block, which
assigns the pulse-control of signals to the appropriate converter. The converters
share the same current and speed loops. A control schematic of the fourquadrant dc motor drive is shown in Figure 3.9.

Figure 3.9: A four-quadrant DC motor drive

The selector block will switch the converters over only when the current
in the outgoing convener has come to zero. Apart from that, some other
conditions have lobe satisfied to transfer the control of converters. Assuming
46

that a positive speed command is required for forward running and a negative
speed command is required for reverse running, the crossover control for
converters is discussed. The machine is running at rated speed and the speed
command is changed lo reverse rated speed. The current command becomes
negative, indicating that the machine has to become regenerative in order to
decelerate the motor in the forward direction. Forward regeneration is possible
only in quadrant IV, and the converter to provide this quadrant of Operation
is converter 2. Before the current is reversed, it has to go through zero. Thai is
achieved by increasing the triggering angle of converter 1. When the current
is zero, a certain dead time is given to enable the thyristors in converter 1 to
recover reverse blocking capability. After this interval, converter 2 is enabled.
At this time, the armature, current has been forced to zero, but the speed is
still positive. The triggering angle of converter 2 is set such that its output
voltage equals and opposes the induced emf. Then, slowly decreasing the
triggering angle increases the armature current in the opposite direction. By a
varying of the triggering angle, the armature current is fully reversed and
then is maintained at the reference level. When the motor reaches zero speed,
the operation of converter 2 is continued by bringing it into rectification mode.
i.e. with less than900, that will accelerate the motor in the reverse direction
until it matches the speed reference.
The function of the selector block is to determine which converter has
1
to be operating. In the previous discussion,
as soon as the current command
goes negative, the selector block will transfer the control from converter 1 to
converter 2 with proper initial triggering angle. If the triggering angle of
converter 1 is then the initial triggering angle 2 of converter 2 is (180 2 )
to match the output voltage of converter 1. If circulating current is not allowed
between the converters, zero crossing of the armature current is required to
transfer the control from one converter to further. The actual rotor speed is
required to determine the quadrant of operation. Based on the rotor speed, the
armature current, and its command, the selector block identifies the converter
for control and operation. The method dis-cussed here with Figure 3.9 does
not allow circulating current between converters 1 and 2. This type of control
has a drawback: slower current transfer from one to the other bridge, due to
the dead time: but it had the advantage that no additional passive component
in the form of an interphase reactor is needed to limit the circulating current.
47

The other type of crossover control uses a circulating current between the
thyristor bridges. It has the advantage of faster current crossover and high
dynamic response, but it has the disadvantages of requiring additional
interphase reactors and losses associated with (hem during the operation.
3.9 Simulation of the One-Quadrant DC Motor Drive
The equations for various subsystems are derived and then assembled
for computer simulation in this section. Key results are discussed. The
simulation for either a two-or four-quadrant dc motor drive is very similar to
the present development. In the present one-quadrant speed-controlled motordrive simulation, it is assumed that the field current is constant in the constant
torque mode and is varied through a three-phase controlled rectifier for the
field-weakening mode of operation to provide constant power over a wide
speed range.
3.9.1 The Motor Equations
Va = Ra ia + La didta + Mi f m 3.52
The motor equations, including that of a simple load modeling, are given
below

Vf

= Rf if + Lf

di f
dt

3.53

di f

dt
d m
Mi f ia Tl = J
+ Bm 3.54
dt

Rf
Lf

if +

Vf
Lf

These equations can be rearranged in the following form lo facilitate

their solution by numerical integration
dia
R
M
V
= a ia i f m + a
dt
La
La
La

.... 3.55

.... 3.56

48

d m M
B
T
= i f ia m l
dt
J
J
J

.... 3.57

.... 3.58
.... 3.59
.... 3.60
Choose ia, ir. and

as state variables and denote them as

From the motor equations and the above set of definitions, the equations
of the motor are written as
.

x1 =
.

Ra
M
V
x1
x2 x3 + a 3.61
La
La
La

x2 =
.

x3 =

Rf
Lf

x2 +

Vf
Lf

3.62

M
B
T
x1 x2 x3 l 3.63
J
J
J

xx3m==i m
12

af

3.9.2 Fitter in the Speed-Feedback Loop

Figure 3.10 shows the speed-feedback filler. The transfer function of
the filter and tacho generator can be represented as

Figure 3.10: Speed-feedback filter

49

 3.64

x4 = mr 3.63
In time domain, this can be rearranged by letting
3.65

Resulting in
The state variable x4 forms one of the inputs to the speed error / controller
block, and that is considered next.
3.9.3 Speed Controller
The speed-controller block diagram is shown in Figure 3.11. The transfer
function of the speed controller considered in the present analysis is a
proportional-plus-integrator controller, given as
Gs ( s ) = K ps +

Kis
3.66
s

.
1 (s)
x
x4 ) H
G4 =( s ) =( Hmr
x3 =
T m ( s ) 1 + sT

Figure 3.11: Speed-controller block diagram

50

Figure 3.12: State diagram of Speed-controller

Tc* = K ps (r* x4 ) + K is x5 3.67

The torque command signal is derived as

0 Te* + Tmax 3.68

In order to maintain the drive system in the safe operating region, the
torque reference is limited to allowable maximum limits determined by the
converter and motor peak capabilities. In this case, let that be +Tmax. This
torque reference limit is integrated into the simulation as
3.9.4 Current-Reference Generator
The current reference is derived from the torque reference by using the
relationship
Te*
i =
Mi f
*
a

K ps r* K ps x4 K is x5 3.69
.
. +
.
=
M x2 M x2 M x2
51

3.9.5 Current Controller

The current controller is of a proportional-plus-integral type with limiter.
The transfer function of the current controller is
Gi ( s ) = K pi +

Kii
3.70
s

Similar to the speed-controller equations, the current-controller

equations are
x 6 = ( ia* H cia )
.

= ( ia* H c x1 )

3.71

And
vc = K pi ( ia* H cia ) + K ii x6

.... 3.72

vcm = H c I max 3.73

This control voltage vc has to be limited to a value Vcm corresponding

to Imax given in the present case as
3.74

Where Imax is the maximum allowable current in the motor and converter.
0 vc V1 cm
vc
Hence the limiter is prescribed as = cos

Vcm
The lower limit for vc is made zero: the current in the phase-controlled
rectifier cannot be reversed. In the case of a four-quadrant dc motor drive,
that limit can be Vcm.
Linearizing controller: The linearization of the output-voltage to input control
signal is accomplished by the following
3.75

52

It could be seen that, when control-signal command goes maximum

for maximum positive current error, the result is the triggering angle, of
zero. That provides the maximum voltage to the armature of the dc motor,
building up the armature current and reducing the current error.
Bridge converter: The bridge converter for continuous current can be modeled
with its delay as
3.76

Note that, for every 60 degrees,

rolls over, and is updated at the
beginning of each 60 electrical degrees. This can be incorporated into the
simulation as

3.77

ss t= 2 f s

Va = 2V sin st + +
3

s t > , s t = 0
3.9.6 Flowchart for Simulation
3
A flowchart for the digital computer
of the one-quadrant dc
when simulation
s t = 0, update and ma int ainthe same
drive is shown in Figure 3.12. The phase-controlled converter modeling for

until
variations of a greater than 60 degrees
has
to be considered; accordingly,
st =
3 For precise prediction of
the algorithm has to be embedded in the software.
SCR and motor-voltage waveforms lo validate the design of the converter.
The use of the right algorithm is required.

53

Figure 3.12: Flowchart for the simulation of a single-quadrant phasecontrolled dc motor drive
54

In the case of discontinuous armature current, note that a nonzero

armature current exists for the conduction duration, which will be less than
60 electrical degrees. For the remaining part of the cycle of 60 electrical degrees
duration, the armature current is zero; accordingly, the algorithm has to be
modified.
3.10 Two/Four-Quadrant Chopper Drives
During power control, a chopper-fed drive operates in the first quadrant,
where the armature voltage and armature current are positive. In a regenerative
braking, the chopper drive operates in the second quadrant, where the armature
voltage is positive and the armature current is negative.

Figure 3.13: Two-quadrant transistorized chopper drive

Two-quadrant operation, as shown in Figure 10.13a, is required to allow
power and regenerative braking control. The circuit arrangement of a
transistorized two-quadrant drive is shown in Figure 10.13b

55

3.10.1 Power control

Transistor Q1 and diode D1 operate. When Q1 is turned on, the supply
voltage Vs is connected to the motor terminals. When Q1 is turned off, the
armature current which flows through the freewheeling diode D1, decays.
3.10.2 Regenerative control
Transistor Q2 and diode D2 operate. When Q2 is turned on, the series
motor acts as a generator and the armature current rises. When Q2 is turned
off, the motor, acting as a generator, returns energy to the supply through the
regenerative diode D2.
In industrial applications, four-quadrant operation, as shown in Figure
10.14a is required. A transistorized four-quadrant drive is shown in Figure
10.14b.

Figure 3.14: Four-quadrant transistorized chopper drive

3.10.3 Forward power control
Transistors Q1 and Q2 operate. Transistors Q3 and Q4 are off. When Ql
and Q2 are turned on together, the supply voltage appears across the motor
terminals and the armature current rises. When Ql is turned off and Q2 is still
turned on, the armature current decays through Q2 and D4.
56

3.10.4 Forward regeneration

Transistors Q1, Q2, and Q3 are turned off. When transistor Q4 is turned
on, the armature current raises flows through Q4 and D2. When Q4 is turned
off, the motor, acting as a generator, returns energy to the supply through D1
and D2.
3.10.5 Reverse power control
The field current of the motor is reversed. Transistors Q3 and Q4 operate.
Transistors Q1 and Q2 are off. When Q2 and Q4 are turned on together, the
armature current rises and flows in the reverse direction. When Q3 is turned
off and Q4 is turned on, the armature current falls through Q4 and D2.
3.10.6 Reverse regeneration
The field current is in the same direction as in reverse power control.
Transistors Q1, Q3, and Q4 are off. When Q2 is turned on, the armature current
rises through Q2 and D4, When Q2 is turned off, the armature current falls and
the motor returns energy to the supply through D3 and D4.

57

C HAPTE R

DC-DC Converter Drive fed DC Motor Drive

4.1 Introduction
Whenever the source is a constant-voltage dc, such as a battery or diodebridge rectified ac supply, a different type of converter is required to convert
the fixed voltage into a variable-voltage/variable-current source for the speed
control of the dc motor drive. The variable dc voltage is controlled by chopping
the input voltage by varying the on- and off-times of a converter, and the type
of converter capable of such a function is known as a chopper.
4.2 Principle of Operation Of The Chopper
A schematic diagram of the chopper is shown in Figure 4.1. The
control voltage to its gate is vc. The chopper is on for a lime ton, and its off
time is toff, Its frequency of operation is
.......4.1
And its duty cycle is defined as
.......4.2
The output voltage across the load during the on-time of the switch is
equal to the difference between the source voltage Vs and the voltage drop
across the power switch. Assuming that the switch is ideal with zero voltage
drop. the average output voltage Vdc is given as
........4.3

58

Where Vs is the source voltage.

Figure 4.1: chopper schematic and its waveforms

Varying the duty cycle changes the output voltage. Note that the output
voltage follows the control voltage as shown in Fig 4.1 signifying that the
chopper is a voltage amplifier.
The duty cycle d can be changed in two ways:
(i)
By keeping the switching/chopping frequency constant and
varying the on- time, to get a changing duty cycle.
(ii) Keeping the on-time constant and varying the chopping frequency
to obtain various values of the duty cycle.
A constant switching frequency has the advantages of predetermined
switching losses of the chopper, enabling optimal design of the cooling for
the power circuit, and predetermined harmonic contents, leading to an optimal
input filter. Both of these advantages are lost by varying the switching
frequency of the chopper hence, this technique for chopper control is not
prevalent in practice.
59

4.3 Four-Quadrant Chopper Circuit

A four-quadrant chopper with transistor switches is shown in Figure
4.2. Each transistor has a freewheeling diode across it and a snubber circuit to
limit the rate of rise of the voltage. The snubber circuit is not shown in the
figure. The load consists of a resistance, an inductance and an induced emf.
The source is dc, and a capacitor is connected across it to maintain a constant
voltage. The base drive circuits of the transistors are isolated and they
reproduce and amplify the control signals at the output. For the sake of
simplicity it is assumed that the switches are ideal and hence, the base drive
signals can be used to draw the load voltage.
4.3.1 First-Quadrant Operation
First-quadrant operation corresponds to a positive output voltage and
current. This is obtained by triggering T1 and T2 together, as is shown in
Figure 6.3: then the load voltage is equal to the source voltage. To obtain zero
load voltage, eitherT1 orT2 can be turned off.
Assume that T1 is turned off; then the current will decrease in the power
switch and inductance. As the current tries to decrease in the inductance it
will have a voltage induced across it in proportion to the rate of fall of current
with a polarity opposite to the load-induced emf, thus forward-biasing diode
D4. D4 provides the path for armature current continuity during this time.
Because of this, the circuit configuration changes as shown in Fig 4.4. The
load is short-circuited reducing its voltage to zero.
The current and voltage waveforms for continuous and discontinuous
current conduction are shown in Figure 4.5. Note that in the discontinuous
current conduction mode the induced emf of the load appears across the load
when the current is zero.
The load voltage, therefore, is a stepped waveform. The operation
discussed here corresponds to motoring in the clockwise direction, or forward
motoring. It can be observed that the average output voltage will vary from o
to Vs the duty cycle can be varied only from 0 to 1.
60

Figure 4.2: A four quadrant chopper circuit

Figure 4.3: First-quadrant operation with positive voltage and current

in the load
The output voltage can also be varied by another switching strategy.
Armature current is assumed continuous. Instead of providing zero voltage
during turn-off time to the load, consider that T1 and T2 are simultaneously
turned off to enable conduction by diodes D3 and D4. The voltage applied
across the load then is equal to the negative source voltage, resulting in a
reduction of the average output voltage.
61

Figure: 4.4 First-quadrant operation with zero voltage across the load

Figure: 4.5 Voltage and current waveforms in first-quadrant operation

62

The disadvantages of this switching strategy are as follows:

(i)
Switching losses double, because two power devices are turned
off instead of one only.
(ii) The rate of change of voltage across the load is twice that of the
other strategy. If the load is a dc machine, then it has the
deleterious effect of causing higher dielectric losses in the
insulation and therefore reduced life. Note that the dielectric is a
capacitor with a resistor in series.
(iii) The rate of change of load current is high, contributing to
vibration of the armature in the case of the dc machine.
(iv) Since a part of the energy is being circulated between the load
and source in every switching cycle, the switching harmonic
current is high, resulting in additional losses in the load and in
the cables connecting the source and converter.
44.3.2Second-Quadrant Operation
Second-quadrant operation corresponds to a positive current with a
negative voltage across the load terminals. Assume that the loads emf is
negative. Consider that T1 or T2 is conducting at a given time. The conducting
transistor is turned off. The current in the inductive load has to continue to
flow until the energy in it is depleted to zero. Hence, the diodes D3 and D4 will
take over maintaining the load current in the same direction, but the load
voltage is negative in the new circuit configuration, as is shown in Figure 4.6.
The voltage and current waveforms are shown in Fig.4.7. When diodes
D3 and D4 are conducting, the source receives power from the load. If the
source cannot absorb this power, provision has to be made to consume the
power. In that case, the over charge on the filter capacitor is periodically
dumped into a resistor connected across the source by controlling the on-time
of a transistor in series with a resistor.
This form of recovering energy from the load is known as regenerative
braking and is common in low-HP motor drives, where the saving in energy
63

might not be considerable or cost-effective .When the current in the load is

decreasing, T2 is turned on. This allows the short-circuiting of the load through
T2 and T4, resulting in an increase in the load current. Turning off T2 results in
a pulse of current flowing into the source via D3 and D4.

Figure 4.6: Second-quadrant operation, with negative load voltage and

positive current

Figure 4.7: Second-quadrant operation of the chopper

This operation allows the priming up of the current and a building up
of the energy in the inductor from the loads emf. thus enabling the transfer of
energy from the load to the source. Note that it is possible to transfer energy
64

from load to source even when E is lower in magnitude than Vs. this is
particular operational feature is sometimes referred to as boost operation in
dc-to-de power supplies. Priming up the load current can also be achieved
alternatively. By using T1 instead of T2.
4.3.3 Third-Quadrant Operation
Third-quadrant operation provides the load with negative current and
voltage. A negative emf source - E. is assumed in the load. Switching on T
and T4 increases the current in the load and turning off one of the transistors
short-circuits the load decreasing the load current. That way, the load current
can be controlled within the externally set limits. The circuit configurations
for the switching instants are shown in Fig. 4.8. The voltage and current
waveforms under continuous and discontinuous current-conduction modes
are shown in Fig. 4.9.

Figure 4.8: Modes of operation in the third quadrant

65

Note the similarity between first and third-quadrant operation.

Figure 4.9: Third quadrant operation

4.3.4 Fourth-Quadrant Operation
Fourth-quadrant operation corresponds to a positive voltage and a
negative current in the load. A positive load-emf source E is assumed. To
send energy to the dc source from the load, note that the armature current has
to be established to flow from the right side to the left side as seen in Fig.4.2
the direction of current is negative.

Figure 4.10: Fourth quadrant operation of the chopper

66

Assume that the machine has been operating in quadrant I with a positive
current in the armature. When a brake command is received, the torque and
armature current command goes negative.
The armature current can be driven negative from its positive value
through zero. Opening T1 and T2 will enable D3 and D4 to allow current via
the source. Reducing the current magnitude rapidly to zero to establish a
negative current T4 is turned on. That will short-circuit the load, making the
emf source build a current through T4and D2 .When the current has reached a
desired peak.T4 is turned off. That forces D1 to become forward-biased and to
carry the load current to the de input source via D2 and the load. When the
current falls below a lower limit, T4 is again turned on, to build up the current
for subsequent transfer to the source.
The voltage and current waveforms are shown in Fig 4.10. The average
voltage across the load is positive, and the average load current is negative
indicating that power is transferred from the load to the source. The source
power is the product of average source current and average source voltage,
and it is negative, as is shown in Fig 4.10.
4.4 Choppers for Inversion
The chopper can be viewed as a single-phase inverter because of its
ability to work in all the four quadrants, handling leading or lagging reactive
loads in series with an emf source. The output fundamental frequency is
determined by the rate at which the forward-to-reverse operation is performed
in the chopper. The chopper is the building block for a multiphase inverter.
4.5 Choppers with Other Power Devices
Choppers are realized with MOSFETs, IGBTs, GTOs, or SCRs.
depending upon the power level required. The MOSFET and transistor
choppers are used at power levels up to 50 kW. Beyond that, IGBTs, GTOs,
and SCRs are used for the power switches. Excepting SCR choppers, all
choppers are self-commutating and hence have a minimum number of power
switches and auxiliary components. In the case of SCR choppers, commutating
circuits have to be incorporated for each main SCR.
67

4.6 Model of the Chopper

The chopper is modeled as a first-order lag with a gain of Kch .. The
time delay corresponds to the statistical average conduction time, which can
vary from zero to T. The transfer function is then

....... (4.4)
Where Kr=Vs/Vcm Vs is the source voltage, and Vcm is the maximum
control voltage. Increasing the chopping frequency decreases the delay time.
and hence the transfer function becomes a simple gain.
4.7 Inputs to the Chopper
The input to the chopper is either a battery or a rectified ac supply line
rectified ac is the prevalent form of input. The ac input is rectified through a
diode bridge, and its output is filtered to keep the dc voltage a constant. The
use of the diode bridge has the advantage of near-unity power factor, thus
overcoming one of the serious disadvantages of the phase-controlled converter.
It has a disadvantage: it cannot transfer power from the de link into ac mains.
In that case .the regenerative energy has to be dumped in the braking
resistor as shown in Fig 4.11 or a phase controlled converter is connected
antiparallel to the diode bridge to handle the regenerative energy as shown in
Fig 4.12. In the latter configuration the phase controlled converter can have a
smaller rating than other power converters, because the rms value of the
regenerative current will be small-the duration of regeneration is only a fraction
of the motoring time for most loads. The regenerating converter has to be
operated at triggering angles greater than 90, to reverse the output de voltage
across the converter to match the polarity of Vs. The phase-controlled converter
is enabled only when Vs is greater than the allowable magnitude V over and
above the nominal de voltage obtained from the ac source.

68

Figure 4.11: Front-end of the chopper circuit

Figure 4.12: Chopper with regeneration capability

-It is usual to set V to be 15 or 20% of V. In such a case, there is a need
for a step-up transformer in the path of the phase-controlled converter, to
match the dc link voltage Vs. The phase converter is disabled when Vs is
slightly greater than 1.35 V, where V is the line-to-line rms voltage ,to prevent
energy flow from source to dc link and from dc link to source via the phasecontrolled converter.
69

4.8 Steady-State Analysis of Chopper-Controlled DC Motor Drive

The steady-state performance of the chopper-controlled dc motor drive
is obtained either with average values, by neglecting harmonics. or including
harmonics. The justification for using average values is that the average torque
is the useful torque that is transmitted to the load. The torque components due
to the current harmonics produce an average torque of zero over one cycle of
switching. They do not contribute to useful power production. Further. they
result in increased armature losses., because the harmonic currents increase
the effective motor current.
From an output point of view, neglecting harmonics and using only
average values gives easier steady-state computation. The analysis by this
method is known as analysis by averaging. When losses, maximum steadystate current. and precise electromagnetic torque are required to fully analyze
and design the drive system for an application, the true current waveforms in
steady state need to be computed.
Therefore, the harmonics cannot be excluded in the steady-state
computation. A computationally efficient, analytical closed-form expression
is obtained by the novel technique of boundary-matching conditions. This
method is referred to as instantaneous stead State computation. It is assumed
that the rotor speed is constant and the field is separately excited. For the
following analys.is, the field flux is maintained at rated value. For any other
value of the field flux, the derivations need to be changed only with regard to
the induced-emf term.
4.8.1 Analysis by averaging
The average armature current is
.........4.5
Where
.........4.6

Vdc =dvs

70

The electromagnetic torque is

Tav =Kb Iav

..........4.7
..........4.8

The torque, written in terms of duty cycle and speed from equations
(4.5), (4.6) and (4.7), is
........... 4.9
The electromagnetic torque is normalized by dividing it by the base
torque, Tb, and by dividing both its numerator and denominator by the base
voltage, Vb The normalized torque is obtained by simplifying the
expressions and substituting for Vb in terms of the base speed, , and emf
constant:
..........4.10
Where
......... 4.11
....... 4.12
R an , V n and mn are normalized resistance Voltage and speed,
respectively. By assigning the product of normalized torque and p.u. resistance
on the y axis, a set of normalized performance curves is drawn for various
values of duty cycles, normalized voltage, and speed, as is shown in Fig.
4.13.From the characteristics and the given duty cycle and voltage, the torque
can be evaluated for a given speed, if the normalized resistance is known.

71

Figure 4.13: Normalized torque -speed characteristics as a function of

duty cycle and source voltage
4.8.2 Instantaneous Steady-State Computation
The instantaneous steady-state armature current and electromagnetic
torques including harmonics is evaluated in this section both for continuous
and discontinuous current conduction, by boundary-matching conditions. The
relevant waveforms are shown in Fig 4.14. For each of the current-conduction
modes, the performance is evaluated separately.
72

Figure 4.14: Applied voltage and armature current in chopper

controlled of DC motor drive
4.8.3 Continuous and discontinuous Current Conduction
The relevant electrical equations of the motor for ON and OFF times are as
follows:
..... 4.13
..... 4.14
The solution of equation (4.13) is
...... 4.15
73

Where
....... 4.16
Similarly, the solution of equation (4.14) is
...... 4.17
And
....... 4.18
From Figure 4.14, it is seen that
....... 4.19
By using this boundary condition,

and I a1 are evaluated as

......... 4.20

........ 4.21
I

0
Having evaluated Ia0 and la1.youacan
use equations (4.15) and (4.17) to
evaluate the instantaneous armature current in steady state.
The limiting or minimum value of duty cycle for continuous current is
evaluated by equating Ia0 to zero. This value is termed the critical duty cycle,
dc and is given by

........ 4.22
Duty cycles lower than dc will produce discontinuous current in the
motor. Note that this critical value is dependent on the ratio between chopping
time period and armature time constant and also on the ratio between induced
emf and source voltage.
74

4.8.4 Discontinuous Current Conduction

The relevant equations for discontinuous current-conduction mode are
obvious from Fig 4.14
...... 4.23
...... 4.24
With
....... 4.25
...... 4.26
Hence
...... 4.27

...... 4.28
This equation is equal to zero, by the constraint given in equation
(4.25), and, from that, tx is evaluated as
...... 4.29
The solution for the armature current in three time segments is
...... 4.30
.... 4.31
...... 4.32
The steady-state performance is calculated by using equations (4.30)
to (4.32).
75

4.9 Rating of the Devices

The chopper shown in Figure 4.2 is considered, for illustration. The
armature cur-rent is assumed to be continuous. with no ripples. If Imax is the
maximum allowable current in the dc machine, the rms value of the power
switch is dependent on this value and its duty cycle. The duty cycle of one
device is slightly more than the duty cycle of the chopper, because one of the
power switches continues to carry current during freewheeling while the other
is turned off. In order to equal the current loading of the switches, the power
switch carrying the freewheeling current is turned of in the next cycle and the
previously inactive switch is allowed to carry the free-wheeling current
alternately.
The rms value of the power switch current and average diode current
are given by,
...... 4.33
..... 4.34
Where the subscripts d and T refer to the diode and power switch,
respectively. The average diode and rms power switch currents vs. duty cycle
for continuous conduction are shown in Fig1.16 for design use.

Figure 4.16: Transistor and diode currents in the chopper for motoring
operation in the continuous conduction mode
76

The minimum voltage rating for both the devices is,

...... 4.35
Whenever regeneration occurs in the motor drive, the current in the
freewheeling diodes would change, and, depending on the frequency and
duration of regeneration, the diode currents have to be recalculated. To optimize
the chopper rating in comparison to the motor and load demands, the operating
conditions have to be known beforehand. The extreme operating conditions
would then prevail on the design and hence on the final rating of the chopper.
4.10 Pulsating Torques
The armature current has ac components. These ac components or
harmonics pro-duce corresponding pulsating torques. The average of the
harmonic torques is zero and they do not contribute to useful torque and
power. Some high-performance applications, such as machine tools and
robots, require the pulsating torque to be a minimum so as not to degrade
the process and the products. In that case, an estimation of the pulsating
torques is in order.

Figure 4.17: Fourth-quadrant operation of DC drive

77

The pulsating currents are evaluated from the harmonic voltages

and the harmonic armature impedances of the dc machine. The applied
voltage shown in Fig 4.17 is resolved into Fourier components as
..... 4.36
Where
.... 4.37
.... 4.38

.... 4.39
.... 4.40

And n is the order of the harmonic. The armature current is expressed

as
...... 4.41
Where
.... 4.42

.... 4.43
.... 4.44

78

And Zan is the harmonic impedance of the armature. The input power
is

....... 4.45
The right-hand side of equation (4.45) can be simplified further,
because the following pulsating terms reduce to an average of zero.
.... 4.46
.... 4.47
The other terms tend to zero, as is shown below

...... 4.48
The average of the double-frequency term is zero, and the power
factor is nearly zero because
...... 4.49
Hence, the expression (4.48) is almost equal to zero. The average
input power becomes a constant quantity and is expressed as
..... 4.50
Power calculation can use the average values, and there is no need
to resort to harmonic analysis, as is shown by the above derivation. The
fundamental harmonic peak pulsating torque then is
79

...... 4.51
And the fundamental-harmonic armature current is given by
..... 4.52
The fundamental armature current can be alternately expressed in
terms of duty cycle as
...... 4.53
When the duty cycle is 50%, the fundamental armature current, and
hence the pulsating torque, is maximum. For a duty cycle of 100%, there
are no pulsating-torque components. The instantaneous electromagnetic
torque is the sum of the dc and harmonic torques, written as
....... 4.54
Where h is the harmonic order. The harmonic torques, Teh, do not
contribute to the load: their averages are zero,
The fundamental pulsating torque is expressed as a fraction of
average torque, to examine the impact of duty cycle on the pulsating torque.
It is facilitated by the following development.

...... 4.55
80

Where 1 is the fundamental power-factor angle, which can be

extracted from
..... 4.56
And
...... 4.57
A set of normalized curves is shown in Fig 4.18 to indicate the
influence of duty cycle and of the ratio between the induced emf and the
source voltage on the magnitude of the pulsating torque.
By noting that the fundamental is the predominant component among
the ac components, the rms value of the armature current is approximated
as
...... 4.58
Where I1 is the fundamental rms current, given by
...... 4.59

Figure 4.18: Normalized fundamental torque pulsation vs.

duty cycle as a function ol the ratio between induced emf and source
voltage
81

The armature resistive losses are

...... 4.60
This indicates that the thermal capability of the motor is degraded
by the additional losses produced by the harmonic currents.
Minimization of the dominant-harmonic torque is of importance in
many applications, mainly in positioning of machine tool drives the key
to mitigation of the harmonic torque is revealed by the expression for
harmonic current given by equation (4.53). Given a fixed voltage source,
there are only two variables that could be utilized to reduce harmonic
current: the chopping frequency, and the machine inductance. Note that d
cannot be used it is a variable dependent on the speed and load. The
chopping frequency is limited by the selection of power device, by its
switching losses, and by other factors, such as electromagnetic
compatibility. The advantage of varying the carrier frequency is that it is
machine-independent, and hence the solution is contained within the
chopper. This may not be feasible, as in the case of large (>100 hp) motor
drives, but it is possible to increase the armature inductance of the machine
during the design of the motor or to include an external inductor to increase
the effective inductance in the armature path. The latter solution is the
only practical approach in retrofit applications.

82

C HAPTE R

Closed Loop Control of Chopper fed DC Motor Drives

Closed loop control of chopper fed DC motor Drives
Design of current controller, design of speed controller, modeling
of current and speed controller, introduction to simulation of speed controlled
dc motor drive.
5.1 Closed-loop Operations
5.1.1 Speed-Controlled Drive System
The speed-controlled dc-motor chopper drive is very similar to the
phase-controlled dc-motor drive in its outer speed-control loop. The inner
current loop and its control are distinctly different from those of the phasecontrolled dc motor drive. This difference is due to the particular characteristics
of the chopper power stage. The current loop and speed loop are examined.
and their characteristics are explained. in this section. The closed-loop speedcontrolled separately-excited dc motor drive is shown in Figure 4.20 for
analysis. but the drive system control strategy is equally applicable to a series
motor drive.
5.1.2 Current Control Loop
With inner current loop alone, the motor drive system is a torque
amplifier. The commanded value of current is compared to the actual armature
current and its error is processed through a current controller. The output of
the current controller.in conjunction with other constraints, determines the
base drive signals of the chopper switches. The current controller can be either
of the following types:
(i) Pulse- Width-Modulation (PWM) controller
(ii) Hysteresis controller

83

Figure: 5.1 speed controlled dc motor chopper drive

The selection of the current controller affects its transient response (and
hence the Overall speed loop bandwidth indirectly). These two controllers
are described in the following sections.
5.1.3 Pulse-Width-Modulated Current Controller
The current error is fed into a controller, which could be proportional
(P), proportional plus integral (PI), or proportional, integral, and differential
(PID). The most commonly used controller among them is the PI controller.
The current error is amplified through this controller and emerges as a control
voltage, ve. It is required to generate a proportional armature voltage from
the fixed source through a chopper operation. Therefore, the control voltage
is equivalent to the duty cycle of the chopper. Its realization is as follows. The
control voltage is compared with a ramp signal to generate the 00- and offtimes, as shown in Figure 5.21. On signal is produced if the control voltageis
greater than the ramp (carrier) signal; olfsignal is generated when the
controlsignal is less than the ramp signal.

84

Figure: 5.2 Generation of base-drive signals from current error for

forward motoring when One is using the chopper
This logic amounts to the fact that the duration for which the control
signal exceeds the ramp signal determines the duty cycle of the chopper. The
on- and off- time signals are combined with other control features, such as
interlock, minimum on- and off-times, and quadrant selection. The interlock
feature prevents the turning on of the transistor (top/bottom) in the same leg
before the other transistor (bottom/top) is turned off completely. This is ensured
by giving a time delay between the turn-off instant of one device and the turnon instant of the other device in the same phase leg. Simultaneous conduction
of the top and bottom devices in the same leg results in a short circuit of the
de source it is known as shoot-through failure in the literature.
85

Figure 5.2 corresponds to the forward motoring operation in the four

quadrant chopper shown in Figure 5.2. When the motor drive is to operate in
the third and fourth quadrants, the armature current reverses. This calls for a
change in the current-control circuitry. A block diagram of the current controller
Is shown in Figure 7.3 ,including all the constraints pertaining to the
operational quadrant .The speed and current polarities, along with that of the
control voltage, determine the quadrant and hence the appropriate gating
signals. The on-time is determined by comparing the ramp signal with the
absolute value of the control voltage. The current-error signal, which
determines the control voltage Vc is rectified to find the intersection point
between the carrier ramp and Vc A unidirectional carrier-ramp signal can be
used when the control voltage is also unidirectional, but the control voltage
will be negative when the current error becomes negative. It happens for
various cases, such as reducing the reference during transient operation and
changing the polarity of the reference to go from quadrant one to three or
four. Taking the polarity of the control voltage and

Figure: 5.3 PWM current controller implementation with ramp carrier

signal combining it with the polarities of the current and speed gives the
operational quadrant. The chopper on and off pulses generated with the
intersection of rectified Vc and carrier ramp will then be combined with the
quadrant-selector signals of speed, current, and control-voltage polarities to
generate the base-drive signals to the chopper switching devices. An illustration
is given in the drive-system simulation section. Instead of a ramp signal for
carrier waveform ,a unidirectional sawtooth waveform could be used. It is
86

advantageous in that it has symmetry between the rising and falling sides of
the waveform, unlike the ramp signal. Its principle of operation is explained
in the following. The voltage applied to the load is varied within one cycle of
the carrier signal. This is illustrated in Figure 7.4 . The switching logic is
summarized as follows: ia* -ia e+ carrier frequency saw tooth wave form
magnitude.
Tp= = 1, Va = = Vs ........ (5.1)

Figure: 5.4 principle of PWM operation

i: - ia < carrier frequency saw tooth waveform magnitude, Tp = 0, Va =
0.....(5.2)
For a fast response ,the current error is amplified so that a small current
error would activate the chopper control. The PWM controller has the
advantage of smaller output ripple current for a given switching frequency,
compared to the hysteresis current controller described later. The pulses
generated from the PWM controller are substituted for Tp in the block diagram
shown in Figure 5.3.Theyare then processed for quadrant selection, interlock.
and safety features, and appropriate base-drive signals are generated for
application to the chopper circuit.
87

5.1.4 : Hysteresis-Current Controller

The PWM current controller acts once a cycle, controlling the duty
cycle of the chopper. The chopper then is a variable voltage source with
average current control. Instantaneous current control is not exercised in the
PWM current controller. In between two consecutive switchings , the current
can exceed the maximum limit if the PWM controller is sampled and held
once a switching cycle, then the current is controlled on an average but not on
an instantaneous basis. The hysteresis controller overcomes such a drawback
by converting a voltage source into a fast-acting current source. The current
is controlled within a narrow band of excursion from its desired value in the
hysteresis controller. The hysteresis window determines the allowable or preset
deviation of current, i Commanded current and actual current are shown in
Figure 7.5 with the hysteresis windows. The voltage applied to the load is
determined by the following logic:
Ia d+ia*-i ,set Va=Vs
Ia e+ ia*+ i ,set Va=0

............. (5.3)
............. (5.4)

Figure: 5.5 hystersis controller operation

Figure: 5.6 realization of hysteresis controller

88

The realization of this logic is shown in Figure 5.6. The windows is,
can either be externally set as a constant or be made a fraction of armature
current, by proper programming. The chopping frequency is a varying quantity,
unlike the constant frequency in the PWM controller .This has the disadvantage
of higher switching losses in the devices with increased switching frequency
.
5.1.5: Modeling of Current Controllers
The current-error amplifier is modeled as a gain and is given by
Gc(s)=kc

........... (5.5)

lhe chopper is modeled as a first-order lag, with a gain given by

....... 5.6
The PWM current controller has a delay of half the time period of the
carrier wave-form, and its gain is that of the chopper. Hence, its transfer
function, including that of the chopper, is
..........5.7
where Kc is the gain of the PWM current controller, Kr is the gain of the
chopper, and the time constant T is given by,
......... 5.8
The gain of the PWM current controller is dependent on the gain of the
current error amplifier. For all practical purposes, the PWM current control
loop can be modeled as a unity-gain block if the delay due to the carrier
frequency is negligible. The hysteresis controller has instantaneous response:
hence, the current loop is approximated as a simple gain of unity.
5.6 Design of Current Controller
The current loop is not easily approximated into a first-order transfer
function. unlike the case of the phase-controlled-rectifier drive system. The
chopping frequency is considered to be high enough that the time constant of
the converter is very much smaller than the electrical time constants of the dc
89

motor. That leads to the converter model given by the product of the converter
and current-controller gains. Then the closed-loop current-transfer function
is written as
5.9

Where and Kc and Kr are the current-controller and chopper gains,

respectively. The chopper gain is derived as
...... 5.10
where Vs is the de link voltage and Vcm is the maximum control voltage.
The gain of the current controller is not chosen on the basis of the damping
ratio. because the poles are 010stlikely to be real ones. Lower the gain of the
current controller; the poles will be far removed from the zeros. The higher
the value of the gain, the closer will one pole move to the zero, leading to the
approximate cancellation of the zero. The other pole will be far away from
the origin and will contribute to the fast response of the current loop.

90