CHAPTER 16

DC Choppers

A dc chopper is a dc-to-dc voltage converter. It is a static switching electrical appliance that in one
electrical conversion, changes an input fixed (normally) dc voltage to an adjustable dc output voltage
without inductive or capacitive intermediate energy storage. The name chopper is connected with the
fact that the output voltage is a ‘chopped up’ quasi-rectangular version of the input dc voltage.
In chapters 13 and 14, thyristor devices were used in conjunction with an ac supply that forces thyristor
turn-off at ac supply current reversal. This form of thyristor natural commutation, which is illustrated in
figure 16.1a, is termed line commutation or source commutation.
When a dc source is used with a switching circuit, energy source facilitated switch commutation is
clearly not possible. If the load is an R-C or L-C circuit as illustrated in figure 16.1b, the load current falls
to zero whence the switch in series with the dc supply turns off. Such a natural turn-off process is
termed load commutation.
If the supply is dc and the load current has no natural zero current periods, such as with the R-L load, dc
chopper circuit shown in figure 16.1c, the load current can only be commutated using a self-
commutating switch, such as a GTO thyristor, GCT, IGBT or MOSFET. An SCR is not suitable since once
the device is latched on in this dc supply application, it remains on.
The dc chopper in figure 16.1c is the simplest of the five dc choppers to be considered in this chapter.
This single-ended, grounded-load, dc chopper will be extensively analysed. See example 16.3.

16.1 DC chopper variations

There are five types of dc choppers, of which four are a subset of the fifth - the flexible but basic, four-
quadrant H-bridge chopper shown in the centre of figure 16.2. Notice that the circuits in figure 16.2 are
highlighted so that the derivation of each dc chopper from the fundamental H-bridge four-quadrant, dc
chopper can be seen. Each chopper can be categorized depending on which output Io-Vo quadrant or
quadrants it can operate in, as shown in figure 16.2. The five different choppers in figure 16.2 are
classified according to their output Io - Vo capabilities as follows:
(a) First quadrant - I +Vo +Io using switch/diode T1 D1
(b) Second quadrant - II +Vo -Io using switch/diode T2 D2
(c) Two quadrant - I and II +Vo ±Io using switches/diodes T1 D1 T2 D2
(d) Two quadrant - I and I V ±Vo +Io using switches/diodes T1 D1 T4 D4
(e) Four quadrant - I, II, III, and IV ±Vo ±Io using switches/diodes T1-4 D1-4

In the five choppers in figure 16.2, the numerical subscript of the active switches and diodes specify in
which quadrants operation is possible. For example, the chopper in figure 16.2d uses switches T1 and
T3, (plus diodes D1 and D3) so can only operate in the first (+Io,+Vo) and third (-Io,-Vo) quadrants.
The first-quadrant chopper, I, (+Io,+Vo), in figure 16.2a (and figure 16.1c) can only produce a positive
voltage across the load since the freewheel diode D 1 prevents a negative output voltage. Also, the
chopper can only deliver current from the dc source to the load through the unidirectional switch T1. It is
therefore a single quadrant chopper and only operates in the first quadrant (+Io,+Vo), where Vo ≤ Vs.
The second-quadrant chopper, II, (-Io,+Vo), in figure 16.2b is a voltage boost circuit and current flows
from the load to the supply, Vs. The switch T2 is turned on to build-up the inductive load current. Then
when the switch is turned off current is forced to flow through diode D2 into the dc supply. The two
current paths (when the switch is on and when it is off) are shown in figure 16.2b.

BWW
Chapter 16 DC Choppers 664

io io

io

io

Vo Q1
io
Io
T on

io

off

Figure 16.1. Three basic types of switch commutation techniques:

(a) source commutation; (b) load commutation; and (c) switch commutation.

In the two-quadrant chopper, quadrants I and II chopper, (±Io,+Vo), figure 16.2c, the load voltage is
clamped to between 0V and Vs, because of the freewheel diodes D1 and D2. Because this chopper is a
combination of the first-quadrant chopper in figure 16.2a and the second-quadrant chopper in figure
16.2b, it combines the characteristics of both. Bidirectional load current is possible but the average
output voltage is always positive. Energy can be regenerated into the dc supply Vs due to the load
inductive stored energy which maintains current flow from the back emf source (<Vs) in the load.
The two-quadrant chopper, quadrants I and IV chopper, (+Io,±Vo), figure 16.2d, can produce a
positive voltage, negative voltage or zero volts across the load, depending on the duty cycle of the
switches and the switching sequence. When both switches are switched simultaneously, an on-state
duty cycle of less than 50% (δ < ½) results in a negative average load voltage Vo, while δ > ½ produces a
positive average load voltage. Since Vo is reversible, the power flow direction is reversible, for the
shown current io. Zero voltage loops are created when one of the two switches is turned off.
The four-quadrant chopper in the centre of figure 16.2 combines all the properties of the four subclass
choppers. It uses four switched and is capable of producing positive or negative voltages across the
load, whilst delivering current to the load in either direction, (±Io,±Vo).
The step-up chopper, or boost converter, presented in Chapter 19.4, may be considered a dc chopper
variation, which has second quadrant characteristics, while the step-down converter presented in
Chapter 19.1 can be considered a first quadrant dc chopper.

16.2 First-quadrant dc chopper, QI

The basic first-quadrant dc chopper circuit reproduced in figure 16.3a can be used to control a dc load
such as a dc motor. As such, the dc load has a back-emf component, E  k   , the magnitude and
polarity of which are dependant on the flux  , (field current if) and its direction, and the speed ω and its
direction. If the R-L load (with time constant τ = L /R) incorporates an opposing back emf, E, then when
the switch T1 is off and the diode D1 is conducting, the load current can be forced towards zero by the
opposing back emf. Therefore two output load current modes (continuous and discontinuous load
current) can occur depending on the relative magnitude of the back emf, load time constant, and the
switch on-state duty cycle. Continuous load current waveforms are shown in figure 16.3b, while
waveforms for discontinuous load current, with periods of zero current, are shown in figure 16.3c.
665 Power Electronics

on

Vs T1 vo Vs T1 vo D4

LOAD LOAD
io

D1 off D1 T4

vo (a) vo (d) vo
I II I I

io III IV io IV io

T1 T3
Vs D2 D4
vo

LOAD

T2 T4
D1 D3
vo vo

II II I
(e)
io io

off
Vs D2 Vs T1 D2 vo
vo
LOAD LOAD

on
T2 T2 D1

(b) (c)

Figure 16.2. Fundamental four-quadrant chopper (centre) showing derivation of four subclass dc
choppers: (a) first-quadrant chopper - I; (b) second-quadrant chopper - II; (c) first and second
quadrants chopper – I and II; (d) first and fourth quadrants chopper – I and IV; and
(e) four-quadrant chopper.

In both conduction cases, the average voltage across the load can be controlled by varying the on-to-off
time duty cycle of the switch, T 1. The on-state duty cycle, δ, is normally controlled by using pulse-width
modulation, frequency modulation, or a combination of both. When the switch is turned off the inductive
load current continues and flows through the load freewheel diode, D 1, shown in figure 16.3a

The analysis to follow for all the dc choppers, assumes:

 no source impedance
 constant switch duty cycle
 steady state conditions have been reached
 ideal semiconductors and
 no load impedance temperature effects.
Chapter 16 DC Choppers 666

vo
on
I
Vs T1
io
LOAD

D1 off

(a)

on
T1
Vs Vs
io io

R L + E R L + E
vo D
D21 vo
off

(b) (a) (c)

(b)

conducting devices

T1 D1 T1 D1 T1 D1 T1 D1 T1 D1
on off on off on on off on
iioℓ    ioiℓ
I I I off  

Io
I I
Io
 
I I
t t

vo tx
vo

Vs Vs
Vo
Vo
E E

t t

tT tT
T T

Figure 16.3. First-quadrant dc chopper and two basic modes of chopper output current operation:
(a) basic circuit and current paths; (b) continuous load current; and
(c) discontinuous load current after t = tx.

16.2.1 Continuous load current

Load waveforms for continuous load current conduction are shown in figure 16.3b.
The output voltage vo, or load voltage is defined by
V for 0  t  tT 
v o t    s  (16.1)
0 for tT  t  T 

The mean load voltage (hence mean load current) is

1 tT 1 tT
V o   v o t  dt   Vs dt
T 0 T 0
(16.2)
t Vo  E
 T Vs   Vs whence Io 
T R
where the switch on-state duty cycle δ = tT /T is defined in figure 16.3b.
667 Power Electronics

The rms load voltage is

½ ½
1 tT  1 tT 
Vrms    v o2 t  dt     V s2 dt 
T 0
 T 0
 (16.3)
t
 T Vs   Vs
T
The output ac ripple voltage is
V r  Vr ms
2
Vo2
(16.4)
 
2
  V s   V s  1   
2
  Vs

The maximum rms ripple voltage in the output occurs when δ = ½ giving an rms ripple voltage of ½Vs.
The output voltage ripple factor is
2
Vr V 
RF    r ms   1  FF 2  1
Vo  Vo 
2
(16.5)
  Vs  1 1
  1  1 
  V   
 s 

Thus as the duty cycle   1 , the ripple factor tends to zero, consistent with the output being dc, that is
Vr = 0.

Steady-state time domain analysis of first-quadrant chopper

- with load back emf and continuous output current
The time domain load current can be derived in a number of ways.
 First, from the Fourier coefficients of the output voltage, the current can be
found by dividing by the load impedance at each harmonic frequency.
 Alternatively, the various circuit currents can be found from the time domain
load current equations.

i. Fourier coefficients: The Fourier coefficients of the load voltage are independent of the circuit and
load parameters and are given by
Vs
an  sin2 n 
n
(16.6)
V
bn  s 1  cos 2 n   for n  1
n
th
Thus the peak magnitude and phase of the n harmonic are given by
c n  an2  bn2

n  tan1 an
bn
Substituting expressions from equation (16.6) yields
2V
c n  n s sin  n 
(16.7)
sin2 n 
n  tan 1
 tan1 cot  n   ½   n 
1  cos 2 n 
where
v n  c n sin n t  n  (16.8)
such that

v o t   Vo   c n sin n t  n 
n 1
(16.9)

2Vs
 V s  
n 1
n  sin  n  cos n t    
The load current is given by

i o t  

Vo 
v V 
 c sin n t  
n
i
n 0
n   n  o  n
R n 1 Z n R n 1 Zn
(16.10)

where the load impedance at each harmonic frequency is given by

Z n  R 2  n  L 
2
Chapter 16 DC Choppers 668

ii. Time domain differential equations: By solving the appropriate time domain differential equations,
the continuous load current shown in figure 16.3b is defined by
During the switch on-period, when vo(t) = Vs
dio
L  R i o  E  Vs
dt
which yields
Vs  E
 t
  t
i o t    1  e 
I e for 0  t  t T (16.11)
R
 
During the switch off-period, when vo(t) = 0
di
L o  R io  E  0
dt
which, after shifting the zero time reference to tT, in figure 16.3a, gives
E  t
  t
i o t    1  e    I e  for 0  t  T  t T (16.12)
R 
tT

V 1e  E
where I  s  (A)
R T
R
1e 
tT
(16.13)

Vs e  1 
E
and I   (A)
R T R
e 1
The output ripple current, for continuous conduction, is independent of the back emf E and is given by
tT T tT

V (1  e
 
) (1  e 
)
I p p  i o  I  I  s (16.14)
R T
1e 

which in terms of the on-state duty cycle, δ=tT /T, becomes

 T  1 T

V (1  e 
) (1  e 
)
I p p  s (16.15)
R T
1e 
harmonic rms as % of dc supply V
s

1 1 T/τ

25
pu 5
¾ dc output ¾
mean

½ 1st 2
Ipp ½
harmonic
Vs / R
1
2nd
¼ harmonic ¼ ½
3rd
harmonic

0 0
0 ¼ ½ ¾ 1
0 ¼ ½ ¾ 1
on-state duty cycle δ δ
on-state duty cycle

Figure 16.4. Harmonics in the output voltage and ripple current as a function of duty cycle δ = tT /T
and ratio of cycle period T (switching frequency, fs=1/T) to load time constant τ=L /R.
Valid only for continuous load current conduction.

The peak-to-peak ripple current can be extracted from figure 16.4, which shows a family of curves for
equation (16.15), normalised with respect to Vs / R. For a given load time constant τ = L /R, switching
frequency fs = 1/T, and switch on-state duty cycle δ, the ripple current can be extracted. This figure
shows a number of important features of the ripple current.
 The ripple current Ipp reduces to zero at the extremes, as δ → 0 and δ → 1.
 Differentiation of equation (16.15) reveals that the maximum ripple current Ip  p occurs at δ = ½.
 The longer the load L /R time constant, τ, the lower the output ripple current Ip-p.
 The higher the switching frequency, 1/T, the lower the output ripple.
669 Power Electronics

If the switch conducts continuously (δ = 1), then substitution of tT = T into equations (16.11) to (16.13)
gives a load voltage Vs and a dc load current is
 
V  E  Vo  E 
io  I  I  s  R  I o  (A) (16.16)
R  
The mean output current with continuous load current is found by integrating the load current over two
consecutive periods, the switch conduction given by equation (16.11) and diode conduction given by
equation (16.12), which yields
1 T
i o t  dt 
V o E 
Io 
T 0  R (16.17)

 V s  E  (A)
R

The input and output powers are related such that

Pin  Pout
  V s  E      
Pin  Vs I i  V S    I  I  
 R T  
1 T (16.18)
Pout   v o t  i o t  dt
T 0

 V  E 
 I o2rms R  E I o  I o2rms R  E  s 
 R 
from which the average input current can be evaluated.
Alternatively, the average input current, which is the average switch current, I switch , can be derived by
integrating the switch current which is given by equation (16.11), that is
1 tT
I i  I switch   i o t  dt
T 0
1 tT V s  E  t
  t 

T    1  e   I e  dt

(16.19)
 R 
0
 
 V s  E   
  I  I 
R T
 
 
The term I  I  I p  p is the peak-to-peak ripple current, which is given by equation (16.15). By
Kirchhoff’s current law, the average diode current I diode is the difference between the average output
current I o and the average input current, I i , that is
I diode  I o  I i
 V s  V s  E      
E 
  I  I 
 (16.20)
R R
T  
   
 E 1   
 I  I  
T   R
Alternatively, the average diode current can be found by integrating the diode current given in equation
(16.12), as follows
1 T tT  E  t
  t 
I diode      1  e    I e   dt
T 0  R   (16.21)
     E 1   
 I  I  
T   R
If E represents motor back emf, then the electromagnetic energy conversion efficiency is given by
E Io E I o
  (16.22)
Pin Vs I i
The chopper effective (dc) input impedance at the dc source is given by
Vs
Z in  (16.23)
Ii

For an R-L load without a back emf, set E = 0 in the foregoing equations. The discontinuous load current
analysis to follow is not valid for an R-L, with E = 0 load, since the load current never reaches zero, but
at best asymptotes towards zero during the off-period of the switch.
Chapter 16 DC Choppers 670

16.2.2 Discontinuous load current

With an opposing emf E in the load, the load current can reach zero during the off-time, at a time tx
shown in figure 16.3c. The time 
tx can be found by
 deriving an expression for I from equation (16.11), setting t = tT,
 this equation is substituted into equation (16.12) which is equated to zero, having substituted for t = tx:
yielding
 V E  tT

t x  t T   ln  1  s  1  e 
  (s) (16.24)
 E  
This equation shows that tx > tT.
Alternatively, for a given frequency, 1/T, discontinuous current will occur as the duty cycle is decreased.

Rearranging equation (16.13), after setting I  0 and extracting the duty cycle   tT /T yields
t   E  T 
  T   n 1   e   1  (16.25)
T T  V s  
Conversely, discontinuous conduction will occur as the frequency decreases for a given switch on-time,
according to
1  V  tT 
T     n 1  s  e   1   (16.26)
f  E  

Figure 16.5 can be used to determine if a particular set of operating conditions involves discontinuous
load current.
1
le
ss o t
ib
po n

E / Vs
us
uo
tin
on

¾
sc

us
di

uo
in
nt
co

½ E / Vs

T/τ 0 1 2 5 10 ∞
?
¼

0
0 ¼ ½ ¾ 1
δ
switch on-state duty

Figure 16.5. Bounds of discontinuous load current with E>0.

The load voltage waveform for discontinuous load current conduction shown in figure 16.3c is defined by
V s for 0  t  t T 
 
v o t   0 for t T  t  t x  (16.27)
E for t x  t  T  

If discontinuous load current exists for a period T - tx, from tx until T, then the mean output voltage is
Vo 
T 
1 tT
 0 t
tx

T t
T
Vs dt   0 dt   E dt
x
 thence I o  Vo  E 

 R 
(16.28)
T tx
V o   Vs  E (V) for t x  t T
T
671 Power Electronics

The rms output voltage with discontinuous load current conduction is given by

 
½
1 tT tx T 
Vrms   V s2dt   02d   E 2dt 
T 0 tT tx
 (16.29)
T tx 2
  V s2  E (V)
T
The ac ripple voltage and ripple factor can be found by substituting equations (16.28) and (16.29) into
Vr  Vrms
2
Vo2 (16.30)
and
2
Vr V 
RF    rms   1  FF 2  1 (16.31)
Vo  Vo 

Steady-state time domain analysis of first-quadrant chopper

- with load back emf and discontinuous output current

i. Fourier coefficients: The load current can be derived indirectly by using the output voltage Fourier
series. The Fourier coefficients of the load voltage are
Vs E t
an  sin2 n   sin2 n x
n n T
(16.32)
V E  t 
bn  s 1  cos 2 n    1  cos 2 n x T  n 1
n n  
which using
c n  an2  bn2

n  tan1 an
bn
give

v o t   V o   c n sin n t  n  (16.33)
n 1

The appropriate division by Z n  R 2  n L  yields the output current.

2


ii. Time domain differential equations: For discontinuous load current, I  0. Substituting this
condition into the time domain equations (16.11) to (16.14) yields equations for discontinuous load
current, specifically:

During the switch on-period, when vo(t) = Vs,

Vs  E
 t

i o t    1  e 
 for 0  t  t T (16.34)
 R 
During the switch off-period, when vo(t) = 0, after shifting the zero time reference to tT,
E  t
  t
i o t    1  e    I e  for 0  t  t x  tT (16.35)
R 
where from equation (16.34), with t = tT,

Vs  E
 tT

I  1  e 

(A) (16.36)
R
 
After tx, vo(t) = E and the load current is zero, that is
i o t   0 for t x  t  T (16.37)
The output ripple current, for discontinuous conduction, is dependent of the back emf E and is given by
equation (16.36), that is

V E  tT

I p p  I  s 1  e 

(16.38)
R  

Since I  0 , the mean output current for discontinuous conduction, is
1 tx 1  tT Vs  E  t
 t x -tT E 
t
  t 
Io   i o t  dt   0  1  e   dt    1  e    I e  dt 
T 0 T  R   0 R   


V o E 
R
Chapter 16 DC Choppers 672

 t   tx 
Vs   1  x  E  Vs  T E 
 T  E
Io      (A) (16.39)
R R R

The input and output powers are related such that

Pin  Vs I i Pout  Io2rms R  E I o Pin  Pout (16.40)
from which the average input current can be evaluated.

Alternatively the average input current, which is the switch average current, is given by
1 tT
I i  I switch   i o t  dt
T 0

1 tT Vs  E  t


T  0 R 
 1  e  dt



(16.41)

Vs  E    tT
  Vs  E 


   1  e 
     I
R  T   R T

The average diode current I diode is the difference between the average output current I o and the
average input current, I i , that is
I diode  I o  I i
t 
E  x   (16.42)
  T 
 I
T R
Alternatively, the average diode current can be found by integrating the diode current given in equation
(16.35), as follows
1 t x tT  E  t
  t 
I diode      1  e    I e   dt
T 0  R  
(16.43)
t 
E  x  
 
T
 I  
T R
If E represents motor back emf, then electromagnetic energy conversion efficiency is given by
E Io E I o
  (16.44)
Pin Vs I i
The chopper effective input impedance is given by
Vs
Z in  (16.45)
Ii

Example 16.1: DC chopper (first quadrant) with load back emf

A first-quadrant dc-to-dc chopper feeds an inductive load of 10Ω resistance, 50mH inductance, and back
emf of 55V dc, from a 340V dc source. If the chopper is operated at 200Hz with a 25% on-state duty
cycle, determine, with and without (rotor standstill, E = 0) the back emf:
i. the load average and rms voltages;
ii. the rms ripple voltage, hence ripple factor;
iii. the maximum and minimum output current, hence the peak-to-peak output ripple in the current;
iv. the current in the time domain;
v. the average load output current, average switch current, and average diode current;
vi. the input power, hence output power and rms output current;
vii. effective input impedance, (and electromagnetic efficiency for E > 0); and
viii. sketch the output current and voltage waveforms.
Solution
The main circuit and operating parameters are
 on-state duty cycle δ = ¼
 period T = 1/fs = 1/200Hz = 5ms
 on-period of the switch tT = 1.25ms
 load time constant τ = L /R = 0.05mH/10Ω = 5ms
673 Power Electronics

T1
δ=¼
T=5ms Figure 16.6. Example 16.1.
Vs 10Ω 50mH Circuit diagram.
340V
R L + E

D1 55V

i. From equations (16.2) and (16.3), assuming continuous load current, the average and rms output
voltages are both independent of the back emf, namely
tT
Vo  V   Vs
T s
= ¼×340V = 85V
tT
Vr  V   Vs
T s
 ¼  240V = 120V rms

ii. The rms ripple voltage hence ripple factor are given by equations (16.4) and (16.5), that is
Vr  Vrms
2
Vo2  V s  1   

= 340V ¼  1 - ¼  147.2V ac
and
Vr 1
RF    1  FF 2  1
Vo 
1
 -1  3  1.732 FF  2
¼

No back emf, E = 0
iii. From equation (16.13), with E = 0, the maximum and minimum currents are
tT -1.25ms

V 1e  340V 1  e 5ms
I  s    11.90A
R T
10 -5ms
1e  1  e 5ms
tT
1

V e   1 340V e 4  1
I  s T    5.62A
R  10 e 1  1
e 1

The peak-to-peak ripple in the output current is therefore

 
I p p  I  I
=11.90A - 5.62A = 6.28A
Alternatively the ripple can be extracted from figure 16.4 using T/τ =1 and δ = ¼.
iv. From equations (16.11) and (16.12), with E = 0, the time domain load current equations are
V  t
  t
i o  s 1  e    I e 
R  
 t
 t
i o t   34   1  e 5ms   5.62  e 5ms
 
t
 34  28.38  e 5ms (A) for 0  t  1.25ms
 t
io  I e 

t
i o t   11.90  e 5ms (A) for 0  t  3.75ms

v. The average load current from equation (16.17), with E = 0, is

I o  V o = 85V = 8.5A
R 10Ω
Chapter 16 DC Choppers 674

The average switch current, which is the average supply current, is

 V s  E      
I i  I switch   I  I 
R T  
¼   340V - 0  5ms
 -  11.90A - 5.62A  = 2.22A
10Ω 5ms
The average diode current is the difference between the average load current and the average
input current, that is
I diode  I o  I i
= 8.50A - 2.22A = 6.28A
vi. The input power is the dc supply voltage multiplied by the average input current, that is
Pin  Vs I i =340V×2.22A = 754.8W
Pout  Pin  754.8W
From equation (16.18) the rms load current is given by
Pout
I o rms 
R
754.8W
= = 8.7A rms
10Ω
vii. The chopper effective input impedance is
Vs
Z in 
Ii
340V
= = 153.2 Ω
2.22A
Load back emf, E = 55V
i. and ii. The average output voltage (85V), rms output voltage (120V rms), ac ripple voltage (147.2V
ac), and ripple factor (1.732) are independent of back emf, provided the load current is continuous.
The earlier answers for E = 0 are applicable.
iii. From equation (16.13), the maximum and minimum load currents are
tT -1.25ms

V 1e  E 340V 1  e 5ms 55V
I  s    - = 6.40A
R T
R 10 -5ms
10Ω
1e  1  e 5ms
tT
V e  1 E 340V e 4  1 55V
1

I  s T    1   0.12A
R  R 10 e  1 10
e 1
The peak-to-peak ripple in the output current is therefore
 
I p p  I  I
= 6.4A - 0.12A = 6.28A
The ripple value is the same as the E = 0 case, which is as expected since ripple current is
independent of back emf with continuous output current.
Alternatively the ripple can be extracted from figure 16.4 using T/τ = 1 and δ = ¼.
iv. The time domain load current is defined by
V E  t
  t
io  s 1  e   I e

R  
 t
 t
i o t   28.5   1  e 5ms   0.12e 5ms
 
t
 28.5  28.38e 5ms
(A) for 0  t  1.25ms
E
    t t
io   1  e   I e

R
 
 t
 t
i o t   5.5   1  e 5ms   6.4e 5ms
 
t
 5.5  11.9e 5ms (A) for 0  t  3.75ms
675 Power Electronics

v. The average load current from equation (16.39) is

I o  Vo  E
R
= 85V-55V = 3A
10Ω
The average switch current is the average supply current,
 V s  E      
I i  I switch   I  I 
R T  
¼   340V - 55V  5ms
 -   6.40A - 0.12A  = 0.845A
10Ω 5ms
The average diode current is the difference between the average load current and the average
input current, that is
I diode  I o  I i
= 3A - 0.845A = 2.155A
vi. The input power is the dc supply voltage multiplied by the average input current, that is
Pin  Vs I i =340V×0.845A = 287.3W
Pout  Pin  287.3W
From equation (16.18) the rms load current is given by
Pout  E I o
I orms 
R
287.3W - 55V×3A
= = 3.5A rms
10Ω
vii. The chopper effective input impedance is
Vs
Z in 
Ii
340V
= = 402.4 Ω
0.845A

The electromagnetic efficiency is given by equation (16.22), that is

E Io

Pin
55V×3A
=  57.4%
287.3W

Conducting device
T1 D1 T1 D1 T1 D1 T1 D1 T1 D1 T1 D1

io  io
11.9A I
Io Δio=6.28A 8.5A 

 I 6.4A
5.62A
I
Io 3A Δio=6.28A
t 0.12A I t
tT T
vo vo

Vs 340V Vs 340V

Vo 85V 85V Vo
t E E=55V t
tT tT
o 1¼ms 5ms o 1¼ms 5ms
E=0 E = 55V

Figure 16.7. Example 16.1. Circuit waveforms.

Chapter 16 DC Choppers 676

viii. The output voltage and current waveforms for the first-quadrant chopper, with and without back emf,
are shown in figure 16.7.
♣

Example 16.2: DC chopper with load back emf - verge of discontinuous conduction

A first-quadrant dc-to-dc chopper feeds an inductive load of 10 Ω resistance, 50mH inductance, and
back emf of 55V dc, from a 340V dc voltage source. If the chopper is operated at 200Hz with a 25% on-
state duty cycle, determine:

i. the maximum back emf before discontinuous load current conduction commences with δ=¼;
ii. with 55V back emf, what is the minimum duty cycle before discontinuous load current
conduction; and
iii. minimum switching frequency at E = 55V and tT = 1.25ms before discontinuous conduction.

Solution

The main circuit and operating parameters are

 on-state duty cycle δ = ¼
 period T = 1/fs = 1/200Hz = 5ms
 on-period of the switch tT = 1.25ms
 load time constant τ = L /R = 0.05mH/10Ω = 5ms

First it is necessary to establish whether the given conditions represent continuous or discontinuous load
current. The current extinction time tx for discontinuous conduction is given by equation (16.24), and
yields
 V E  tT

t x  tT   n  1  s  1  e  


 E  
 340V - 55V  -1.25ms

 1.25ms + 5ms  n  1 +  1 - e 5ms   = 5.07ms
 55V  

Since the cycle period is 5ms, which is less than the necessary time for the current to fall to zero
(5.07ms), the load current is continuous. From example 16.1 part iv, with E = 55V the load current falls
from 6.4A to near zero (0.12A) at the end of the off-time, thus the chopper is operating near the verge of
discontinuous conduction. A small increase in E, decrease in the duty cycle δ, or increase in switching
period T, would be expected to result in discontinuous load current.

i. E
The necessary back emf can be determined graphically or analytically.

Graphically:
The bounds of continuous and discontinuous load current for a given duty cycle, switching period, and
load time constant can be determined from figure 16.5.
Using δ = ¼, T/τ = 1 with τ = 5ms, and T = 5ms, figure 16.5 gives E / Vs = 0.165. That is, E = 0.165×Vs
= 0.165×340V = 56.2V

Analytically:
The chopper is operating too close to the boundary between continuous and discontinuous load
current conduction for accurate readings to be obtained from the graphical approach, using figure
16.5. Examination of the expression for minimum current, equation (16.13), gives
tT

V e  1 E
I  s T  0
R  R
e 1
Rearranging to give the back emf, E, produces
tT
e  1
E  Vs T
e  1
1.25ms
e 5ms
-1
= 340V  5ms
= 56.2V
e 5ms
-1
677 Power Electronics

That is, if the back emf increases from 55V to 56.2V then at and above that voltage, discontinuous
load current commences.

ii.  
Again, if equation (16.13) is solved for I  0 then
tT

V e  1 E
I  s T  0
R  R
e 1
Rearranging to isolate tT gives
 E  T 
t T   n  1   e   1  
 Vs  
 55V  5ms 
= 5ms  n  1 + e
5ms
- 1 
 340V  
= 1.226ms
If the switch on-state period is reduced by 0.024ms, from 1.250ms to 1.226ms (δ = 24.52%), operation
is then on the verge of discontinuous conduction.

iii. T
If the switching frequency is decreased such that T = tx, then the minimum period for discontinuous
load current is given by equation (16.24). That is,
 V E  tT

t x  T  t T   n  1  s  1  e  


 E  

340V - 55V  -1.25ms

T  1.25ms + 5ms  n  1 +   1 - e 5ms   = 5.07ms
 55V  
Discontinuous conduction operation occurs if the period is increased by more than 0.07ms.

In conclusion, for the given load, for continuous conduction to cease, the following operating
conditions can be changed
 increase the back emf E from 55V to 56.2V
 decrease the duty cycle δ from 25% to 24.52% (tT decreased from 1.25ms to 1.226ms)
 increase the switching period T by 0.07ms, from 5ms to 5.07ms (from 200Hz to 197.2Hz), with
the switch on-time, tT, unchanged from 1.25ms.
Appropriate simultaneous smaller changes in more than one parameter would suffice.
♣

Example 16.3: DC chopper with load back emf – discontinuous conduction

A first-quadrant dc-to-dc chopper feeds an inductive load of 10Ω resistance, 50mH inductance, and an
opposing back emf of 100V dc, from a 340V dc source. If the chopper is operated at 200Hz with a 25%
on-state duty cycle, determine:

i. the load average and rms voltages;

ii. the rms ripple voltage, hence ripple and form factors;
iii. the maximum and minimum output current, hence the peak-to-peak output ripple in the current;
iv. the current in the time domain;
v. the load average current, average switch current and average diode current;
vi. the input power, hence output power and rms output current;
vii. effective input impedance, and electromagnetic efficiency; and
viii. sketch the circuit, load, and output voltage and current waveforms.

Solution
The main circuit and operating parameters are
 on-state duty cycle δ = ¼
 period T = 1/fs = 1/200Hz = 5ms
 on-period of the switch tT = 1.25ms
 load time constant τ = L /R = 0.05mH/10Ω = 5ms
Chapter 16 DC Choppers 678

D1 +
T1 100V
δ=¼
T=5ms Vs
Vs R L E
10Ω 50mH
340V
δ=¼ 10Ω 50mH
340V
T=5ms
R L E+ T1

D1 100V 0V

0V (a) (b)

Figure 16.8. Example 16.3. Circuit diagram:

(a) with load connected to ground and (b) load connected so that machine flash-over to ground (0V),
by-passes the switch T1.

Confirmation of discontinuous load current can be obtained by evaluating the minimum current given by
equation (16.13), that is
tT

V e  1 E
I  s T 
R  R
e 1
1.25ms

340V e 5ms -1 100V
I=  5ms - = 5.62A - 10A = - 4.38A
10Ω 10Ω
e -1
5ms

The minimum practical current is zero, so clearly discontinuous current periods exist in the load current.
The equations applicable to discontinuous load current need to be employed.
The current extinction time is given by equation (16.24), that is
 V E  tT

t x  t T   n  1  s  1  e  


 E  
 340V - 100V  -1.25ms

= 1.25ms + 5ms  n 1 +  1 - e 5ms  
 100V  
= 1.25ms + 2.13ms = 3.38ms
i. From equations (16.28) and (16.29) the load average and rms voltages are
T tx
V o   Vs  E
T
5ms - 3.38ms
= ¼×340 V +  100V = 117.4V
5ms
T tx 2
Vrms   V s2  E
T
5ms - 3.38ms
= ¼  3402 +  1002 = 179.3V rms
5ms
ii. From equations (16.30) and (16.31) the rms ripple voltage, hence voltage ripple factor, are
Vr  Vrms
2
Vo2
= 179.32 - 117.42 = 135.5V ac
V 135.5V
RF  r = = 1.15 FF  RF 2  1  1.152 +1  1.52
Vo 117.4V
iii. From equation (16.38), the maximum and minimum output current, hence the peak-to-peak output
ripple in the current, are

V E  tT

I  s 1  e 

R  
340V-100V  -1.25ms

=   1 - e 5ms  = 5.31A
10Ω  
679 Power Electronics

The minimum current is zero so the peak-to-peak ripple current is i o = 5.31A.

iv. From equations (16.34) and (16.35), the current in the time domain is
V E  t

i o t   s 1  e 

R  
340V - 100V  t

   1  e 5ms 
10Ω  
 t

 24   1  e 5ms  (A) for 0  t  1.25ms
 
E t
  t
i o t     1  e 
I e
R 
100V  t
 t
   1  e 5ms   5.31e 5ms
10Ω  
t
 15.31  e 5ms  10 (A) for 0  t  2.13ms
i o t   0 for 3.38ms  t  5ms
Conducting device
T D T D T

vo
Vs=340V
117.4V
Vo E=100V
E=100V

io
5.31A

Io 1.74A

iD 5.31A

ID 1.05A

iT
5.31A

IT 0.69A

vT
Vs=340V
240V

0 1.25 3.37 5 6.25 8.37 10 11.25

time (ms) t

Figure 16.9. Example 16.3. Chopper circuit waveforms.

v. From equations (16.39) to (16.42), the average load current, average switch current, and average
diode current are
I o V o  E
R
=117.4V - 100V = 1.74A
10Ω
Chapter 16 DC Choppers 680

t 
E  x  
   T 
I diode  I
T R
 3.38ms 
100V   - 0.25 
5ms  5ms  = 1.05A
 ×5.31A -
5ms 10Ω
I i  I o  I diode =1.74A - 1.05A = 0.69A

vi. From equation (16.40), the input power, hence output power and rms output current are
Pin  Vs I i  340V×0.69A = 234.6W
Pin  Pout  Io2r ms R  E I o
Rearranging gives
I o rms  Pin  E I o / R
= 234.6W - 100V×0.69A / 10Ω = 1.29A

vii. From equations (16.44) and (16.45), the effective input impedance and electromagnetic efficiency,
for E > 0 are
V 340V
Z in  s  = 493Ω
I i 0.69A
E Io E I o 100V×1.74A
  = = 74.2%
Pin Vs I i 340V×0.69A

viii. The circuit, load, and output voltage and current waveforms are plotted in figure 16.9.
♣

16.3 Second-quadrant dc chopper, QII

The second-quadrant dc-to-dc chopper shown in figure 16.2b transfers energy from the load, back to the
dc energy source Vs, a process called regeneration. Its operating principles are the same as those for
the boost switch mode power supply analysed in chapter 19.4. The two energy transfer stages are
shown in figure 16.10. Controlled energy transfer from the back emf E to the supply Vs, is achieved by
varying the switch T2 on-state duty cycle. Two modes of transfer can occur, as with the first-quadrant
chopper already considered. The current in the load inductor can be either continuous or discontinuous,
depending on the specific circuit parameters and operating conditions.

In this analysis, and all the choppers analysed, it is assumed that:

 No source impedance;
 Constant switch duty cycle;
 Steady-state conditions have been reached;
 Ideal semiconductors; and
 No load impedance temperature effects.

16.3.1 Continuous load inductor current

Load waveforms for continuous load current conduction are shown in figure 16.11a.
The output voltage vo, load voltage, or switch voltage, is defined by
0 for 0  t  tT 
v o t     (16.46)
Vs for tT  t  T 

The mean load voltage is

1 T 1 T
Vo   v o t  dt   V s dt
T 0 T tT (16.47)
T  tT
 V s  1    V s
T
where the switch on-state duty cycle δ = tT /T is defined in figure 16.11a.
681 Power Electronics

Vo

II

Io Vs D2 ioff

io R L io

+ R L +
ion E vo E
T2

(a) (b)

Figure 16.10. Stages of operation for the second-quadrant chopper:

(a) switch-on, boosting current and (b) switch-off, energy into Vs.

Alternatively the voltage across the dc source Vs is

1
Vs  Vo (16.48)
1
Since 0 ≤ δ ≤ 1, the step-up voltage ratio, to regenerate into Vs, is continuously adjustable from unity to
infinity.
The average output current is
E V o E Vs 1   
Io   (16.49)
R R
The average output current can also be found by integration of the time domain output current io. By
solving the appropriate time domain differential equations, the continuous load current io shown in figure
16.11a is defined by

During the switch on-period, when vo = 0

di o
L  R io  E
dt
which yields
E t
  t
i o t    1  e 
I e for 0  t  t T (16.50)
R 
During the switch off-period, when vo = Vs
di
L o  R i o V s  E
dt
which, after shifting the zero time reference to tT, gives
E Vs  t
  t
i o t    1  e 
I e for 0  t  T  t T (16.51)
R  
tT T

E V e 
e 
where I   s (A)
R R T
1e 

T tT
(16.52)

E Vs 1  e 
and I   (A)
R R T
1e 
The output ripple current, for continuous conduction, is independent of the back emf E and is given by
T tT T tT

V (1  e
 
)  (e 
e 
)
I p p  I I  s (16.53)
R T
1e 

which in terms of the on-state duty cycle, δ = tT / T, becomes

 T T
V (1  e 
) (1  e 
)
I p p  s (16.54)
R T
1e 

This is the same expression derived in 16.2.1 for the first-quadrant chopper. The normalised ripple
current design curves in figure 16.3 are valid for the second-quadrant chopper.
Chapter 16 DC Choppers 682

Conducting devices
T2 D2 T2 D2 T2 D2 T2 D2 T2 D2

io    io
I I I  
I I
Io
 
I I Io
t t

tx
vo vo

Vs Vs

Vo Vo E E

t t

tT tT
T T

(a) (b)

Figure 16.11. Second-quadrant chopper output modes of current operation:

(a) continuous inductor current and (b) discontinuous inductor current.

The average switch current, I switch , can be derived by integrating the switch current given by equation
(16.50), that is
1 tT
I switch   i o t  dt
T 0

1 tT E  t
  t 

T    1  e    I e   dt (16.55)
R 
0
 
E     
  I  I 
R T  
 
The term I  I  I p  p is the peak-to-peak ripple current, which is given by equation (16.53). By
Kirchhoff’s current law, the average diode current I diode is the difference between the average output
current I o and the average switch current, I switch , that is
I diode  I o  I switch
E V s 1     E     
   I  I  (16.56)
R R T  
   
 V  E  1   
 I  I   s
T   R

The average diode current can also be found by integrating the diode current given in equation (16.51),
as follows
1 T tT  E V s  t
  t 
I diode     1  e    I e   dt
T 0  R    (16.57)
   
 V s  E  1   
 I  I  
T   R

The power produced (provide) by the back emf source E is

 E V s 1    
PE  E I o  E   (16.58)
 R 

The power delivered to the dc source Vs is

   V  E  1    
PVs  V s I diode  V s   I  I   s  (16.59)
T   R 
683 Power Electronics

The difference between the two powers is the power lost in the load resistor, R, that is
PE  PVs  I o2rms R
(16.60)
E I o V s I diode
I o r ms 
R
The efficiency of energy transfer between the back emf E and the dc source Vs is
PVs V s I diode
  (16.61)
PE E Io

16.3.2 Discontinuous load inductor current

With low duty cycles, δ, low inductance, L, or a relatively high dc source voltage, Vs, the minimum output
current may reach zero at tx, before the period T is complete (tx < T), as shown in figure 16.11b. Equation
(16.52) gives a boundary identity that must be satisfied for zero current,
T tT

E V 1e 
I   s  0 (16.62)
R R T
1e 
That is
T tT

E 1e 
 (16.63)
Vs T
1e  
Alternatively, the time domain equations (16.50) and (16.51) can be used, such that I = 0. An
expression for the

extinction time tx can be found by substituting t = tT into equation (16.50). The resulting
expression for I is then substituted into equation (16.51) which is set to zero. Isolating the time variable,
which becomes tx, yields
E  tT

I   1  e  
R 
E V s  t x
 E  tT
 t x
0  1  e 
   1  e 
e
R   R  
which yields
  t T
E 
t x  t T   n  1   1  e  

(16.64)
 V s  E  
This equation shows that t x  t T . Load waveforms for discontinuous load current conduction are shown
in figure 16.11b.

The output voltage vo, load voltage, or switch voltage, is defined by

0 for 0  t  t T 
 
v o t   V s for t T  t  t x  (16.65)
E for t x  t  T 

The mean load voltage is

T  
1 T 1 tx T
V o   v o t  dt  V s dt   E dt
T 0 tT tx

t x  tT T tx t   t 
 Vs  E   x   V s   1  x E
T T T   T  (16.66)
tx
V o  E  V s  V s  E 
T
where the switch on-state duty cycle δ = tT /T is defined in figure 16.11b.

The average output current is

tx
E V o Vs  T Vs  E 
Io   (16.67)
R R

The average output current can also be found by integration of the time domain output current io. By
solving the appropriate time domain differential equations, the continuous load current io shown in figure
16.11a is defined by
Chapter 16 DC Choppers 684

During the switch on-period, when vo = 0

di o
L  R io  E
dt
which yields
E  t

i o t   1  e 

for 0  t  tT (16.68)
R  

During the switch off-period, when vo = Vs

di o
L  R i o V s  E
dt
which, after shifting the zero time reference to tT, gives
E Vs  t
  t
i o t   1  e   I e

for 0  t  t x  t T (16.69)
R  

E  tT

where I   1  e   (A)
R   (16.70)

and I  0 (A)
After tx, vo(t) = E and the load current is zero, that is
i o t   0 for t x  t  T (16.71)

The output ripple current, for discontinuous conduction, is dependent of the back emf E and is given by
equation (16.70),

E  tT

I p  p  I  1  e   (16.72)
R 
The average switch current, I switch , can be derived by integrating the switch current given by equation
(16.68), that is
1 tT
I switch   i o t  dt
T 0

1 tT E  t


T  0

R


1  e 
  dt

(16.73)

E  
  I
R T

The term I  I p  p is the peak-to-peak ripple current, which is given by equation (16.72). By Kirchhoff’s
current law, the average diode current I diode is the difference between the average output current I o and
the average switch current, I switch , that is
I diode  I o  I switch
tx
V s 
T
V s  E   E  
   I (16.74)
R R T
tx 
   T    V s  E 
 I  
T R
The average diode current can also be found by integrating the diode current given in equation (16.69),
as follows
1 t x tT  E V s  t
  t 
I diode   0   1  e   I e  dt

T  R   
(16.75)
tx 
    V s  E 
 
T
 I 
T R
The power produced by the back emf source E is
PE  E I o (16.76)

The power delivered to the dc source Vs is

PVs  V s I diode (16.77)
685 Power Electronics

Alternatively, the difference between the two powers is the power lost in the load resistor, R, that is
PE  PVs  I o2rms R
(16.78)
E I o V s I diode
I or ms 
R
The efficiency of energy transfer between the back emf and the dc source is
PVs V s I diode
  (16.79)
PE E Io

Example 16.4: Second-quadrant DC chopper – continuous inductor current

A dc-to-dc chopper capable of second-quadrant operation is used in a 200V dc battery electric vehicle.
The machine armature has 1Ω resistance in series with 1mH inductance.

i. The machine is used for regenerative braking. At a constant speed downhill, the back emf is
150V, which results in a 10A braking current. What is the switch on-state duty cycle if the
machine is delivering continuous output current? What is the minimum chopping frequency
for these conditions?
ii. At this speed, (that is, E = 150V), determine the minimum duty cycle for continuous inductor
current, if the switching frequency is 1kHz. What is the average braking current at the critical
duty cycle? What is the regenerating efficiency and the rms machine output current?
iii. If the chopping frequency is increased to 5kHz, at the same speed, (that is, E = 150V), what is
the critical duty cycle and the corresponding average dc machine current?

Solution

The main circuit operating parameters are

 Vs = 200V
 E = 150V
 load time constant τ = L /R = 1mH/1Ω = 1ms
Conducting devices
vo
T2 D2 T2 D2 T2 D2
II
io
io

I

Io
D2  
I =0 I =0 t
Vs = 200V
vo
io R L
T2
1Ω 1mH +150V Vs =200V
E=150V

Vo
t

tT
T
Figure 16.12. Example 16.4. Circuit diagram and waveforms.

i. The relationship between the dc supply Vs and the dc machine back emf E is given by equation
(16.49), that is
E Vo E V s 1   
Io  
R R
150V - 200V  1 - δ 
10A =
1Ω
that is
 = 0.3  30% and V o = 140V
Chapter 16 DC Choppers 686

The expression for the average dc machine output current is based on continuous armature inductance
current. Therefore the switching period must be shorter than the time tx predicted by equation (16.64)
for the current to reach zero, before the next switch on-period. That is, for tx = T and δ = 0.3
 E  tT

t x  t T   n  1   1  e  


 Vs  E  
This simplifies to
1ms   0.3T

1  0.3  n 1  150V 1  e 1ms  
T  200V - 150V  
e 0.7T
 4  3e 0.3T

Iteratively solving this transcendental equation gives T = 0.4945ms. That is the switching frequency
must be greater than fs =1/T = 2.022kHz, else machine output current discontinuities occur, and equation
(16.49) is invalid. The switching frequency can be reduced if the on-state duty cycle is increased as in
the next part of this example.

ii. The operational boundary condition giving by equation (16.63), using T=1/ fs =1/1kHz = 1ms, yields
T tT

E 1e 

Vs T
1e 
 -1×1ms
150V 1 - e 1ms
 -1ms
200V
1 - e 1ms
Solving gives δ = 0.357. That is, the on-state duty cycle must be at least 35.7% for continuous machine
output current at a switching frequency of 1kHz.
For continuous inductor current, the average output current is given by equation (16.49), that is
E V o E V s 1   
Io  
R R
150V - V o 150V - 200V× 1 - 0.357 
= = = 21.4A
1Ω 1Ω
V o =150V - 21.4A×1Ω = 128.6V
The average machine output current of 21.4A is split between the switch and the diode (which is in
series with Vs).
The diode current is given by equation (16.56)
I diode  I o  I switch
    V s  E  1   

I  I  
 T  R
The minimum output current is zero while the maximum is given by equation (16.70).

E  tT
 150V  -0.357×1ms

I  1  e     1 - e 1ms  = 45.0A
R   1Ω  
Substituting into the equation for the average diode current gives

I diode =
1ms
  45.0A - 0A  -
200V - 150V   1 - 0.357  = 12.85A
1ms 1Ω
The power delivered by the dc machine back emf E is
PE  E I o = 150V×21.4A = 3210W
while the power delivered to the 200V battery source Vs is
PVs  Vs I diode  200V×12.85A = 2570W
The regeneration transfer efficiency is
PV 2570W
 s = = 80.1%
PE 3210W
The energy generated deficit, 640W (3210W - 2570W)), is lost in the armature resistance, as I R heat
2

dissipation. The output rms current is

P 640W
I or ms  = = 25.3A rms
R 1Ω

iii. At an increased switching frequency of 5kHz, the duty cycle would be expected to be much lower
than the 35.7% as at 1kHz. The operational boundary between continuous and discontinuous
armature inductor current is given by equation (16.63), that is
687 Power Electronics

T tT

E 1e 

Vs T
1e 
 -1+δ 0.2ms
150V 1 - e 1ms
= -0.2ms
200V
1 - e 1ms
which yields δ = 26.9% .
The machine average output current is given by equation (16.49)
E V o E V s 1   
Io  
R R
150V Vo 150V - 200V× 1 - 0.269 
 =  3.8A
1 1Ω
such that the average output voltageV o is 146.2V.
♣

16.4 Two-quadrant dc chopper - Q I and Q II

Figure 16.13 shows the basic two-quadrant dc chopper, which is a reproduction of the circuit in figure
16.2c. Depending on the load and operating conditions, the chopper can seamlessly change between
and act in two modes

 Devices T1 and D1 form the first-quadrant chopper shown in figure 16.2a, and is analysed in
section 16.2. Energy is delivered from the dc source Vs to the R-L-E load.
 Devices T2 and D2 form the second-quadrant chopper shown in figure 16.2b, which is analysed
in section 16.3. Energy is delivered from the generating load dc source E, to the dc source Vs.

The two independent choppers can be readily combined as shown in figure 16.13a.
The average output voltage V o and the instantaneous output voltage vo are never negative, whilst the
average source current of Vs can be positive (Quadrant I) or negative (Quadrant II). If the two choppers
are controlled to operate independently, with the constraint that T 1 and T2 do not conduct
simultaneously, then the analysis in sections 16.2 and 16.3 are valid. Alternately, it is not uncommon
the unify the operation of the two choppers, as follows.

If the chopper is operated such that the switches T 1 and T2 act in a complementary manner, that is either
T1 or T2 is on, then some of the independent flexibility offered by each chopper is lost. Essentially the
consequence of complementary switch operation is that no extended zero current periods exist in the
output, as shown in figures 16.13a and b. Thus the equations describing the features of the first-
quadrant chopper in section 16.2.1, for continuous load current, are applicable to this chopper, with
slight modification to account for the fact that both the minimum and maximum currents can be negative.
The analysis for continuous inductor current in section 16.2 is valid, but the minimum current is not
restricted to zero. Consequently four possible output modes can occur, depending on the relative
polarity of the maximum and minimum currents shown in figure 16.13b and c.
 
i. I > 0, I > 0 and I o > 0
When the minimum current (hence average output current) is greater than zero, the chopper is active
in the first-quadrant. Typical output voltage and current waveforms are shown in figure 16.3a. The
switch

T
 2
and diode D2 do not conduct during any portion of the operating period.
ii. I < 0, I > 0 and I o > 0
When the minimum current is negative but the maximum positive current is larger in absolute
magnitude, then for a highly inductive load, the average output current is greater than zero, and the
chopper operates in the first-quadrant. If the load is not highly inductive the boundary is determined
by

the average

output current I o > 0. The various circuit waveforms are shown in figure 16.13b.
iii. I < 0, I > 0 and I o < 0
For a highly inductive load, if the magnitude of the negative peak is greater than the positive
maximum, the average is less than zero and the chopper is operating in the regenerative mode,
quadrant II. If the load is not highly inductive the boundary is determined by the average output
current

Io < 0.
iv. I < 0, I < 0 and I o < 0
When the maximum current and the average current are both negative, the chopper is operational in
the second-quadrant. Since the load current never goes positive, switch T 1 and diode D1 never
conduct, as shown in figure 16.13c.
Chapter 16 DC Choppers 688

Vo

II I

T1 D2 Q I io Io
Vs
Q II io

R L + E

T2 D1 vo

(a)

Vo Vo
I II

Io Io

T1
Vs Vs D2 off
on

R L + E R L + E

D1 vo T2 vo
off on

Conducting devices
D2 T1 D1 T2 D2 T1 D2 T2 D2 T2
vo
Vs Vs
Vo
E E
Vo t
o o
tT T t tT T
I  0
I  0

io I
o 
t 
 Io I
I 0
I 0 o Io
txD  t Io  0
Io  0

txT I I

is
 t
I o
Is 

Is I
o

t
I 
I

(b) (c)

Figure 16.13. Two-quadrant (I and II) dc chopper circuit where vo > 0:

(a) basic two-quadrant dc chopper; (b) operation and waveforms for quadrant I; and (c) operation
and waveforms for quadrant II, regeneration into Vs.

In all cases the average output voltage is solely determined by the switch T1 on-time duty cycle, since
when this switch is turned on the supply Vs is impressed across the load, independent of the direction of
the load current. When io > 0, switch T1 conducts while if io < 0, the diode in parallel to switch T 1, namely
D1 conducts, clamping the load to Vs.
The output voltage, which is independent of the load, is described by
V for 0  t  tT 
v o t    s  (16.80)
0 for tT  t  T 
689 Power Electronics

Thus
1 tT tT
Vo 
T  0
V s dt 
T s
V   Vs (16.81)
The rms output voltage is also determined solely by the duty cycle,
½
1 tT 
Vr ms  
T
 0
V s2 dt 
 (16.82)
  Vs
The output ac ripple voltage, hence voltage ripple factor are given by equations (16.3) and (16.5), and
are independent of the load:
Vr  Vrms
2
Vo2  Vs  1    (16.83)
and
Vr 1 1
RF   1  FF  1 (16.84)
Vo   
The Fourier series for the load voltage can be used to determine the load current at each harmonic
frequency as described by equations (16.6) to (16.10).

The time domain differential equations from section 16.2.1 are also valid, where there is no zero
restriction on the minimum load current value.

In a positive voltage loop, when vo(t) = Vs and Vs is impressed across the load, the load circuit
condition is described by
V E  t
  t
i o t   s 1  e   I e

for 0  t  t T (16.85)
R  
During the switch off-period, when vo = 0, forming a zero voltage loop
E  t
  t
i o t    1  e    I e  for 0  t  T  t T (16.86)
R 
where
tT

V 1e  E
where I  s  (A)
R T
R
1e 
tT
(16.87)

V e  1 E
and I  s T  (A)
R  R
e 1

The peak-to-peak ripple current is independent of E,

 T  1 T

V (1  e 
) (1  e 
)
I p p  s (16.88)
R T
1e 

The average output current, I o , may be positive or negative and is given by

1 T
i o t  dt 
V o E 
Io 
T  0 R (16.89)

 V s  E  (A)
R
The direction of the net power flow between E and Vs determines the chopper operating quadrant. If
V o > E then average power flow is to the load, as shown in figure 16.13b, while if V o < E, the average
power flow is back into the source Vs, as shown in figure 16.13c.
Vs I s  Io2r ms R  E I o (16.90)
Thus the sign of I o determines the direction of net power flow, hence quadrant of operation.

Calculation of individual device average currents in the time domain is complicated by the fact that the
energy may flow between the dc source Vs and the load via the switch T 1 (energy to the load) or diode
D2 (energy
 
from the load). It is therefore necessary to ascertain the zero current crossover time,
when I and I have opposite signs, which will then specify the necessary bounds of integration.

Equations (16.85) and (16.86) are equated to zero and solved for the time at zero crossover, txT and txD,
respectively, shown in figure 16.13b.
Chapter 16 DC Choppers 690

 

 IR 
t xT   n 1  with respect to t  0
 Vs  E 
  (16.91)
 IR 
t xD   n  1   with respect to t  t T
 E 


The necessary integration for each device can then be determined with the aid of the device conduction
information in the parts of figure 16.13 and Table 16.1.

Table 16.1: Device average current ratings

     
Device and integration bounds, a to b I  0, I  0 I  0, I  0 I  0, I  0
1 Vs  E
b  t
  t
IT 1 
T  a R


1  e 
  I e dt

0 to t T t xT to t T 0 to 0

1 b Vs  E  t
  t
I D1 
T  0 R
1  e


  I e dt

0 to 0 0 to t xT 0 to t T

1 b E  t
  t
IT 2 
T  a

R
1  e   I e dt



0 to 0 t xD to T - t T 0 to T - t T

1 b E  t
  t
I D2 
T  0

R


1  e 
  I e dt

0 to T - t T 0 to t xD 0 to 0

The electromagnetic energy transfer efficiency is determined from

E Io
 for I o > 0
Vs I i
(16.92)
V I
 s i for I o < 0
E Io

Example 16.5: Two-quadrant DC chopper with load back emf

The two-quadrant dc-to-dc chopper in figure 16.13a feeds an inductive load of 10 Ω resistance, 50mH
inductance, and back emf of 100V dc, from a 340V dc source. If the chopper is operated at 200Hz with
a 25% on-state duty cycle, determine:
i. the load average and rms voltages;
ii. the rms ripple voltage, hence ripple and form factors;
iii. the maximum and minimum output current, hence peak-to-peak output ripple in the current;
iv. the current in the time domain;
v. the current crossover times, if applicable;
vi. the load average current, average switch current and average diode current for all devices;
vii. the input power, hence output power and rms output current;
viii. effective input impedance and electromagnetic efficiency; and
ix. sketch the circuit, load, and output voltage and current waveforms.
Subsequently determine the necessary change in
x. duty cycle δ to result in zero average output current and
xi. back emf E to result in zero average load current.

Solution

The main circuit and operating parameters are

 on-state duty cycle δ = ¼
 period T = 1/fs = 1/200Hz = 5ms
 on-period of the switch tT = 1.25ms
 load time constant τ = L /R = 0.05mH/10Ω = 5ms

i. From equations (16.81) and (16.82) the load average and rms voltages are
t 1.25ms
v o  T Vs  ×340V = ¼ ×340V = 85V
T 5ms
Vrms   Vs = ¼ ×340V = 170V rms
691 Power Electronics

vo
δ=¼
T=5ms II I
T1 D2
R L +E io
Vs=340V
10Ω 50mH +100V

δ=¾ T2 D1 vo
T=5ms

Figure 16.14. Example 16.5. Circuit diagram.

ii. The rms ripple voltage, hence voltage ripple factor, from equations (16.83) and (16.84) are
Vr  Vrms
2
Vo2  V s  1   

= 1702 - 852 = 340V ¼  1 - ¼ = 147.2V

Vr 1 1
RF    1= - 1 = 1.732 FF  1  1 2
Vo  ¼  ¼

iii. From equations (16.87) and (16.88), the maximum and minimum output current, hence the peak-to-
peak output ripple in the load current are given by
tT -1.25ms

V 1e  E 340V 1 - e 5ms 100V
I  s T
 = × -5ms
- = 1.90A
R R 10Ω 10Ω
1e  1- e 5ms
tT 1.25ms

Vs e 1 E 340V e 5ms -1 100V
I  T  = × 5ms
- = - 4.38A
R R 10Ω 10Ω
1e -1 e 5ms
The peak-to-peak ripple current is therefore io  1.90A - - 4.38A = 6.28A p-p.

iv. The current in the time domain is given by equations (16.85) and (16.86)
V E  t
  t
i o t   s 1  e   I e

R  
-t -t
340V-100V  
= ×  1 - e 5ms  - 4.38×e 5ms
10Ω  
-t -t
 
= 24×  1 - e 5ms  - 4.38×e 5ms
 
-t
= 24 - 28.38  e 5ms for 0  t  1.25ms

E  t
  t
i o t    1  e

I e
R  
t t
100V  
   1  e 5ms   1.90  e 5ms
10  
t t
 
 10   1  e 5ms   1.90  e 5ms
 
t
 10  11.90  e 5ms for 0  t  3.75ms

v. Since the maximum current is greater than zero (1.9A) and the minimum is less that zero (- 4.38A),
the current crosses zero during the switch on-time and off-time. The time domain equations for the load
current are solved for zero to give the cross over times txT and txD, as given by equation (16.91), or
solved from the time domain output current equations as follows.
Chapter 16 DC Choppers 692

During the switch on-time

-t
i o t   24 - 28.38  e 5ms  0 where 0  t  t xT  1.25ms
28.38
t xT = 5ms×n = 0.838ms
24
During the switch off-time
t
i o t   10  11.90  e 5ms  0 where 0  t  t xD  3.75ms
11.90
t xD =5ms×n = 0.870ms
10
(1.250ms + 0.870ms = 2.12ms with respect to switch T1 turn - on)

vi. The load average current, average switch current, and average diode current for all devices;

Io 
V o E  
 V s E 
R R
 85V - 100V 
= -1.5A
10Ω
When the output current crosses zero current, the conducting device changes.

Table 16.1 gives the
necessary current equations and integration bounds for the condition I  0, I  0 . Table 16.1 shows
that all four semiconductors are involved in the output current cycle.
t
1 tT V s  E    t
IT1    1  e    I e  dt
T t xT R  
-t
1 1.25ms

5ms  0.838ms
 24 - 28.38  e 5ms dt  0.081A

1 t xT V s  E  t
  t
I D1    1  e    I e  dt
T 0 R  
-t
1 0.84ms

5ms 0 24 - 28.38  e 5ms dt   0.357A

1 T -tT E  t
  t
IT 2 
T  txD

R
 1  e   I e dt



t
1 3.75ms
 
5ms 0870ms
10  11.90  e 5ms dt  1.382A

1 t xD E  t
  t
I D2 
T  0

R
 1  e   I e dt



t
1 0.870ms
 
5ms 0
10  11.90  e 5ms dt  0.160A

Check I o  I T 1  I D 1  I T 2  I D 2  - 1.5A + 0.080A - 0.357A - 1.382A + 0.160A = 0

vii. The input power, hence output power and rms output current;
Pin  PVs  V s I i  V s I T 1  I D 1  
 340V×  0.080A - 0.357A  = -95.2W, (charging V s )
Pout  PE  E I o  100V   -1.5A  = -150W, that is generating 150W
From
V s I s  Io2r ms R  E I o
Pout  Pin 150W - 92.5W
Io r ms    2.34A rms
R 10Ω
viii. Since the average output current is negative, energy is being transferred from the back emf E to the
dc voltage source Vs, the electromagnetic efficiency of conversion is given by
Vs I i
 for I o < 0
E Io
95.2W
= = 63.5%
150W
693 Power Electronics

The effective input impedance is

Vs Vs 340V
Z in    = -1214Ω
I i I T 1  I D 1 0.080A - 0.357A
ix. The circuit, load, and output voltage and current waveforms are sketched in the figure 16.15.

Conducting devices
D2 T1 D1 T2 D2 T1 D1 T2

vo

340V

E 100V

Vo 85V
o

io
 1.9A
I t
o

Io -1.5A
txD
-4.38A =0.87ms 
I

is 
1.9A I t
o
Is -0.28A
2.12ms

txT =0.383ms -4.38A
I
tT =1.25ms T =5ms

Figure 16.15. Example 16.5. Circuit waveforms.

x. Duty cycle δ to result in zero average output current can be determined from the expression for the
average output current, equation (16.89), that is
V  E
Io  s 0
R
that is
E 100V
   = 29.4%
V s 340V

xi. As in part x, the average load current equation can be rearranged to give the back emf E that results
in zero average load current
V  E
Io  s 0
R
that is
E  Vs  ¼×340V = 85V
♣

16.5 Two-quadrant dc chopper - Q I and Q IV

The unidirectional current, two-quadrant dc chopper, or asymmetrical half H-bridge shown in figure
16.16a incorporates two switches T 1 and T4 and two complementary diodes D1 and D4. In using
switches T1 and T4 the chopper operates in the first and fourth quadrants, that is, bi-directional voltage
output vo but unidirectional load current, io.
Chapter 16 DC Choppers 694

The chopper can operate in two quadrants (I and IV), depending on the load and switching sequence.
Net power can be delivered to the load, or received from the load provided the polarity of the back emf E
is reversed. Because of this need to reverse the back emf for regeneration, this chopper is not
commonly used in dc machine control. On the other hand, the chopper circuit configuration is commonly
used to meet the converter requirements of the switched reluctance machine, which only requires
unipolar current to operate. Also see chapter 19.6 for a bidirectional smps variation.
The asymmetrical half H-bridge chopper has three different output voltage states, where one state (the
zero output state) has redundancy (two possibilities). Both the output voltage vo and output current io
are with reference to the first quadrant arrows in figure 16.16a.

Vs T1 D4
vo
io
I LOAD

IV io + vo
D1 T4

(a)

Conducting devices
T1 T1 D1 T1 T1 T1 D1 T1 D1 D1 D1 T1 D1 D1
D4 T4 T4 T4 D4 T4 T4 D4 D4 T4 D4 D4 D4 T4

1 T1 D2 1
δ
T1 D2
½ T1 D2 ½
δ

o o

T1 T1 on T1 off T1 on T1 off T1 on T1 off

T1

T4 T4
T4 on T4 off T4 on T4 off T4 on T4 off

+Vs o tT T 2T
vo vo
Vo

Vo
-Vs
o tT T 2T 
I Io
io io

Io I

is I -is

I Is Is

(c) (b)

Figure 16.16. Two-quadrant (I and IV) dc chopper

(a) circuit where io>0: (b) operation in quadrant IV, regeneration into Vs; and (c) operation in quadrant I.

State #1
When both switches T1 and T4 conduct, the supply Vs is impressed across the load, as shown in
figure 16.17a. Energy is drawn from the dc source Vs.
T1 and T4 conducting: vo = Vs
695 Power Electronics

State #2
If only one switch is conducting, and therefore also one diode, the output voltage is zero, as shown
in figure 16.17b. Either switch (but only one on at any time) can be the on-switch, hence providing
redundancy, that is
T1 and D4 conducting: vo = 0
T4 and D1 conducting: vo = 0
State #3
When both switches are off, the diodes D1 and D4 conduct load energy back into the dc source Vs,
as in figure 16.17c. The output voltage is -Vs, that is
T1 and T4 are not conducting: vo = -Vs

0V
+Vs D3 D3 -Vs
Vs T1 D4 Vs T1 D4 Vs T1 D 34
D

LOAD LOAD LOAD

+ Vs - - Vs +

D
D 12 TT34 D
D 12 0V
TT34 D 21
D TT34

(a) (b) (c)

Figure 16.17. Two-quadrant (I and IV) dc chopper operational current paths: (a) T1 and T4 forming a +Vs
path; (b) T1 and D4 (or T4 and D1) forming a zero voltage loop; and (c) D1 and D4 creating a -Vs path.

The two zero output voltage states can most effectively be used if alternated during any switching
sequence. In this way, the load switching frequency (load ripple current frequency) is twice the
switching frequency of the switches. This reduces the output current ripple for a given switch operating
frequency (which minimises the load inductance necessary for continuous load current conduction).
Also, by alternating the zero voltage loop, the semiconductor losses are evenly distributed. Specifically,
a typical sequence to achieve these features would be
T1 and T4 Vs
T1 and D4 0
T1 and T4 Vs
T4 and D1 0 (not T1 and D4 again)
T1 and T4 Vs
T1 and D4 0, etc.

The sequence can also be interleaved in the regeneration mode, when only one switch is on at any
instant, as follows
D1 and D4 -Vs (that is T1 and T4 off)
T1 and D4 0
D1 and D4 -Vs
T4 and D1 0 (not T1 and D4 again)
D1 and D4 -Vs
T1 and D4 0, etc.
In switched reluctance motor drive application there may be no alternative to using only ±Vs control loops
without the intermediate zero voltage state.

There are two basic modes of chopper switching operation.

 Multilevel switching is when both switches are controlled independently to give all three output
voltage states (three levels), namely ±Vs and 0V.
 Bipolar switching (or two level switching) is when both switches operate in unison, where they
turn on together and off together. Only two voltage output states (hence the term bipolar), are
possible, +Vs and – Vs.
Chapter 16 DC Choppers 696

16.5.1 DC chopper:– Q I and Q IV – multilevel output voltage switching (three level)

The interleaved zero voltage states are readily introduced if the control carrier waveforms for the two
switches are displaced by 180°, as shown in figure 16.16b and c, for continuous load current. This
requirement can be realised if two up-down counters are displaced by 180°, when generating the
necessary triangular carriers. As shown in figures 16.16b and c, the switching frequency 1/Ts is
determined by the triangular wave frequency 1/2T, whilst advantageously the load experiences twice
that frequency, 1/T, hence the output current has reduced ripple, for a given switch operating frequency.

i. 0 ≤ δ ≤ ½
It can be seen in figure 16.16b that when δ ≤ ½ both switches never conduct simultaneously hence the
output voltage is either 0 or -Vs. Operation is in the fourth quadrant. The average output voltage is load
independent and for 0 ≤ δ ≤ ½, using the waveforms in figure 16.16b, is given by
1 T V  t 
V o   Vs dt  s T  tT   Vs 1  T  (16.93)
T Tt T  T 
Examination of figure 16.16b reveals that the relationship between tT and δ must produce
when   0 : tT  T and v o  Vs
when   ½ : tT  0 and v o  0
that is
tT
 ½
T
(the period of the carrier, 2T, is twice the switching period, T) which after substituting for tT /T in equation
(16.93) gives
 t 
V o  V s  1  T 
 T  (16.94)
 V s 1  2   V s  2  1 for 0    ½
Operational analysis in the fourth quadrant, δ ≤ ½, is similar to the analysis for the second-quadrant
chopper in figure 16.2b and analysed in section 16.3. Operation is characterised by first shorting the
output circuit to boost the current, then removing the output short forces current back into the dc supply
Vs, via a freewheel diode. The characteristics of this mode of operation are described by the equations
(16.50) to (16.79) for the second-quadrant chopper analysed in 16.3, where the output current may
again be continuous or discontinuous. The current and voltage references are both reversed in
translating equations applicable in quadrants Q II to Q IV.

ii. ½ ≤ δ ≤ 1
As shown in figure 16.16c, when δ ≥ ½ and operation is in the first quadrant, at least one switch is
conducting hence the output voltage is either +Vs or 0. For continuous load current, the average output
voltage is load independent and for ½ ≤ δ ≤ 1 is given by
1 tT V
V o   V s dt  s t T (16.95)
T 0 T
Examination of figure 16.16c reveals that the relationship between tT and δ must produce
when   ½ : tT  0 and v o  0
when   1 : tT  T and v o  Vs
that is
 tT 
  ½  1
T 
which on substituting for tT /T in equation (16.95) gives
t
V o  V s T  V s  2  1 for ½    1 (16.96)
T
Since the average output voltage is the same in each case, equations (16.94) and (16.96) for (0 ≤ δ ≤ 1),
the output current mean is given by the same expression, namely
V o  E Vs (2  1)  E
Io   (16.97)
R R
Operation in the first quadrant, δ ≥ ½, is characterised by the first-quadrant chopper shown in figure
16.2a and considered in section 16.2 along with the equations within that section. The load current can
be either continuous, in which case equations (16.6) to (16.23) are valid; or discontinuous in which case
equations (16.24) to (16.45) are applicable. Aspects of this mode of switching are extended in 16.5.3.
In applying the equations for the chopper in section 16.2 for the first-quadrant chopper, and the
equations in section 16.3 for the second-quadrant chopper, the duty cycle in each case is replaced by
697 Power Electronics

 2δ -1 in the case of δ ≥ ½ for the first-quadrant chopper and

 2δ in the case of δ ≤ ½ for the fourth-quadrant chopper.
This will account for the scaling and offset produced by the triangular carrier signal decoding.
16.5.2 DC chopper: – Q I and Q IV – bipolar voltage switching (two level)
When both switches operate in the same state, that is, both switches are on simultaneously or both are
off together, operation is termed bipolar or two level switching.
From figure 16.18 the chopper output states are (assuming continuous load current)
 T1 and T4 on vo = Vs
 T1 and T4 off vo = - Vs
From figure 16.18, the average output voltage is
Vo 
1 tT
T 0  T
V s dt   V s dt
tT  (16.98)
Vs
 t T  T  t T    2  1V s
T
The rms output voltage is independent of the duty cycle and is Vs.
The output ac ripple voltage is
Vr  Vrms
2
Vo2
(16.99)
 Vs2   2  1 Vs2  2Vs  1   
2

which is a maxima at δ = ½ and a minima for δ = 0 and δ = 1.

Vs T1 D4
io

LOAD
R L + E
D1 vo T4

(a)

Conducting devices
T1 D1 T1 D1 T1 D1
T4 D4 T4 D4 T4 D4

io

I

Io
 
I I t
o
T
vo

Vs
Vo
o
t
tT
-Vs

(b)
Figure 16.18. Two-quadrant (I and IV) dc chopper operation in the bipolar output mode:
(a) circuit showing load components and (b) chopper output waveforms.

The output voltage ripple factor is

Vr 2Vs  1    2  1   
RF    (16.100)
Vo 2  1Vs 2  1
Chapter 16 DC Choppers 698

Although the average output voltage may reverse, the load current is always positive but can be
discontinuous or continuous. Equations describing bipolar output are presented within the next section,
16.5.3, which considers multilevel (two and three level) output voltage switch states.

16.5.3 Multilevel output voltage states, dc chopper

In switched reluctance machine drives it is not uncommon to operate the asymmetrical half H-bridge
shown in figure 16.18 such that
 both switches operate in the on-state together to form +V voltage loops;
 switches operate independently to give zero voltage loops; and
 both switches are simultaneously off, forming –V voltage output loops.
The control objective is to generate a current output pulse that tracks a reference shape which starts
from zero, rises to maintain a fixed current level, with hysteresis, then the current falls back to zero. The
waveform shown in figure 16.19 fulfils this specification.
The switching strategy to produce the current waveform in figure 16.19 aims at:
 For rising current:- use +V loops (and zero volt loops only if necessary)
 For near constant current:- use zero voltage loops (and ±V loops only if necessary to increase
or decrease the current)
 For falling current:- use – V loops (and zero volts loops only if necessary to reduce the fall rate)
Operation is further characterised by continuous load current during the pulse.
Energy is supplied to the load from the dc voltage source during +V loops, and returned to the dc supply
during –V loop periods.
The chopper output current during each period is described by equations previously derived in this
chapter, but reproduced as follows.
In a positive voltage loop, (T1 and T4 are both on), when vo(t) = Vs and Vs is impressed across the load,
the load circuit condition is described by
di o
L  R i o  E  Vs
dt
which yields
Vs  E
 t
  t
i o t   1  e   I e

for 0  t  t  (16.101)
R   
During the first switching cycle the current starts from zero, so I = 0. Otherwise I is the lower

reference, I , from the end of the previous cycle.
The current at the end of the positive voltage loop period is the reference level I , whilst the time to rise
+

to I is derived by equating equation (16.101) to I and solving for time t at the end of the period.
+ + +

Solving io(t ) = I for t , gives

+ + +


Vs  E  I R 



t   ln  (16.102)
Vs  E  I R 
 

In a zero voltage loop, when vo(t) = 0, such as circuit loops involving T1 and D4 (or T4 and D1), the
circuit equation is given by
di o
L  R io  E  0
dt
which gives
E  t
  t
i o t    1  e   I e

for 0t to (16.103)
R 

where I equals the reference current level, I from the previous switching period.
+
 
The current at the end of the period is the reference level I , whilst the time to fall to I is given by

equating equation (16.103) to I and solving for time, t at the end of the period.
o


E  I R 


t   n
o  (16.104)
 E  I R 
 
In a negative voltage loop, when both switches T1 and T4 are off, the current falls rapidly and the circuit
equation, when vo(t) = -Vs, is
di o
L  R i o  E  V s
dt
699 Power Electronics

which gives
E Vs  t
  t
i o t   1  e   I e

for 0  t  t  (16.105)
R  

where I equals the reference current level, I from the previous switching period.
+

 
The current at the end of the period is I , whilst the time to reach I is given by equating equation

(16.103) to I and solving for time t  at the end of the period.

Vs  E  I R 



t   n  (16.106)
Vs  E  I R 
 
The same equation is used to determine the time for the final current period when the current decays to

zero, whence I = 0.

The characteristics and features of the three output voltage states are illustrated in the following
example, 16.6.

Conducting devices
T1 T1 T1 D1 T1 T1 T1 D1
T4 D4 T4 T4 T4 D4 T4 D4
I
+
I o
io
-
I

to t+ to t+ to t+
o o
I o t

is Io

o o
t

energy
recovered

 Io
Vs

vo

o o
t

-Vs

Figure 16.19. Two-quadrant (I and IV) dc chopper operation in a multilevel output voltage mode.

Example 16.6: Asymmetrical, half H-bridge, dc chopper

The asymmetrical half H-bridge, dc-to-dc chopper in figure 16.18 feeds an inductive load of 10 Ω
resistance, 50mH inductance, and back emf of 55V dc, from a 340V dc voltage source. The chopper
output current is controlled in a hysteresis mode within a current band between limits 5A and 10A.
Determine the period of the current shape shown in the figure 16.20:

i. when only ±Vs loops are used and

ii. when a zero volt loop is used to maintain tracking within the 5A band.

In each case calculate the switching frequency if the current were to be maintained within the hysteresis
band for a prolonged period.
How do the on-state losses compare between the two control approaches?
Chapter 16 DC Choppers 700

Solution

The main circuit and operating parameters are

 E = 55V and Vs = 340V
 load time constant τ = L /R = 0.05mH/10Ω = 5ms

 I = 10A and I = 5A
+

Examination of the figure 16.20 shows that only one period of the cycle differs, namely the second
period, t2, where the current is required to fall to the lower hysteresis band level, -5A. The period of the
other three regions (t1, t3, and t4) are common and independent of the period of the second region, t2.

t1: The first period, the initial rise time, t = t1 is given by equation (16.102), where I =10A and I = 0A.
+ +


Vs  E  I R 


t   n
 
V s  E  I R 
 
 340V - 55V - 0A×10Ω 
that is t 1= 5ms  n   = 2.16ms
 340V - 55V - 10A×10Ω 
t3: In the third period, the current rises from the lower hysteresis band limit of 5A to the upper

band limit

10A. The duration of the current increase is given by equation (16.102) again, but with I = I = 5A.

V s  E  I R 


t   n
 
V s  E  I R 
 
 340V - 55V - 5A×10Ω 
that is t 3 = 5ms  n   = 1.20ms
 340V - 55V - 10A×10Ω 
t4: The fourth and final period is a negative voltage loop where the current 
falls from the upper band
 
limit of 10A to I which equals zero. From equation (16.106) with I =I =10A and I = 0A
+


V s  E  I R 



t   n 
V s  E  I R 
 
 340V + 55V + 10A×10Ω 
that is t 4 = 5ms  n   = 1.13ms
 340V + 55V + 0A×10Ω 
The current pulse period is given by
Tp  t1  t 2  t 3  t 4
 2.16ms + t 2 + 1.20ms + 1.13ms
= 4.49ms + t 2

i. t2: When only -Vs paths 

are used to decrease the current, the time t2 is given by equation (16.106),

with I =5A and I =10A,

R 

V  E  I
t   n 
 s
V s  E  I R 
 
 340V + 55V + 10A×10Ω 
that is t 2 = 5ms  n   = 0.53ms
 340V + 55V + 5A×10Ω 
The total period, Tp, of the chopped current pulse when a 0V loop is not used, is
Tp  t1  t 2  t 3  t 4
= 2.16ms + 0.53ms + 1.20ms + 1.13ms = 5.02ms

ii. t2:When a zero voltage loop is used to maintain the current within the hysteresis band, the current

decays slowly, and the period time t2 is given by equation (16.104), with I = 5A and I =10A,

E  I R 


t   n
o 
 E  I R 
 
 55V + 10A×10Ω 
that is t 2 = 5ms  n   = 1.95ms
 55V + 5A×10Ω 
The total period, Tp, of the chopped current pulse when a 0V loop is used, is
Tp  t1  t 2  t 3  t 4
= 2.16ms + 1.95ms + 1.20ms + 1.13ms = 6.44ms
701 Power Electronics

Conducting devices
T1 D1 / T1 T1 D1
T4 D4 T4 D4
10A
io
5A

t1 t2 t3 t4
o o
10A t

is 5A

1.13ms
o o
t
2.16ms 1.2ms
-V 0.53ms
0V 1.95ms

-10A
Vs +340V

vo 0V
loop E=55V
o o
t1 t2 t3 t4 t

-V
loop
-Vs -340V
Tp

Figure 16.20. Example 16.6. Circuit waveforms.

The current falls significantly faster within the hysteresis band if negative voltage loops are employed
rather that zero voltage loops, 0.53ms versus 1.95ms.
The switching frequency within the current bounds has a period t2 + t3, and each case is summarized
in the following table. For longer current chopping, t2 and t3 dominate the switching frequency.
Using zero voltage current loops (alternated) reduces the switching frequency of the H-bridge
switches by a factor of over three, for a given peak-to-peak ripple current.
If the on-state voltage drop of the switches and the diodes are similar for the same current level, then
the on-state losses are similar, and evenly distributed for both control methods. The on-state losses
are similar because each of the three states always involves the same current variation flowing
through two semiconductors. The principal difference is in the significant increase in switching losses
when only ±V loops are used (1:3.42).

Table 16.2 Example 16.6. Switching losses.

Voltage Current ripple Switch Switch loss

t2 + t3
loops frequency frequency ratio
578
±V 0.53ms+1.20ms =1.73ms 578Hz 578Hz = 3.42
169

+V and zero 1.95ms+1.20ms = 3.15ms 317Hz 169Hz 1

♣

16.6 Four-quadrant dc chopper

The four-quadrant H-bridge dc chopper is shown in figure 16.21 where the load current and voltage are
referenced with respect to T1, so that the quadrant of operation with respect to the switch number is
persevered.
The H-bridge is a flexible basic configuration where its use to produce single-phase ac is considered in
chapter 17.1.1, while its use in smps applications is considered in chapter 19.8.2. It can also be used as
a dc chopper for the four-quadrant control of a dc machine.
Chapter 16 DC Choppers 702

With the flexibility of four switches, a number of different control methods can be used to produce four-
quadrant output voltage and current (bidirectional voltage and current). All practical methods should
employ complementary device switching in each leg (either T1 or T4 on but not both and either T2 or T3
on, but not both) so as to minimise distortion by ensuring current continuity around zero current output.
One control method involves controlling the H-bridge as two virtually independent two-quadrant
choppers, with the over-riding restriction that no two switches in the same leg conduct simultaneously.
One chopper is formed with T 1 and T4 grouped with D1 and D4, which gives positive current io but
bidirectional voltage ±vo (QI and QIV operation). The second chopper is formed by grouping T2 and T3
with D2 and D3, which gives negative output current -io, but bi-direction voltage ±vo (QII and QIII
operation).

vo

T1 T3 II I
Vs D2 D4

io III IV io

LOAD

I
vo
T2 T4
D1 D3

Figure 16.21. Four-quadrant dc chopper circuit, showing first quadrant io and vo references.

The second control method is to unify the operation of all four switches within a generalised control
algorithm.
With both control methods, the chopper output voltage can be either multilevel or bipolar, depending on
whether zero output voltage loops are employed or not. Bipolar output states increase the ripple current
magnitude, but do facilitate faster current reversal, without crossover distortion. Operation is
independent of the direction of the output current io.

Since the output voltage is reversible for each control method, a triangular based modulation control
method, as used with the asymmetrical H-bridge dc chopper in figure 16.16, is applicable in each case.
Two generalised unified H-bridge control approaches are considered – bipolar and three-level output.

16.6.1 Unified four-quadrant dc chopper - bipolar voltage output switching

The simpler output to generate is bipolar output voltages, which use one reference carrier triangle as
shown in figure 16.22 parts (c) and (d). The output voltage switches between + Vs and – Vs and the
relative duration of each state depends on the magnitude of the modulation index δ.

If δ = 0 then T1 and T4 never turn-on since T2 and T3 conduct continuously which impresses – Vs
across the load.
At the other extreme, if δ = 1 then T1 and T4 are on continuously and +Vs is impressed across the
load.
If δ = ½ then T1 and T4 are turned on for half of the period T, while T2 and T3 are on for the remaining
half of the period. The output voltage is – Vs for half of the time and + Vs for the remaining half of any
period. The average output voltage is therefore zero, but disadvantageously, the output current
needlessly ripples about zero (with an average value of zero).

The chopper output voltage is defined in terms of the triangle voltage reference level vΔ by
 vΔ > δ, vo = -Vs
 vΔ < δ, vo = +Vs
From figure 16.22c and d, the average output voltage varies linearly with δ such that
Vo 
1 tT
 T

 V s dt   V s dt
T 0 tT 
(16.107)
1  t 
  2tT  T V s   2 T  1 V s
T  T 
703 Power Electronics

Examination of figures 16.22c and d reveals that the relationship between tT and δ must produce
when   0 : tT  0 and v o  V s
when   ½ : t T  ½T and v o  0
when   1 : tT  T and v o  V s
that is
tT
 
T
which on substituting for tT /T in equation (16.107) gives
 t 
V o   2 T  1  Vs
 T  (16.108)
  2  1 V s for 0    1
The average output voltage can be positive or negative, depending solely on δ. No current discontinuity
occurs since the output voltage is never actually zero. Even when the average voltage is zero, ripple
current flows though the load, with an average value of zero amps.

The rms output voltage is independent of the duty cycle and is Vs.
The output ac ripple voltage is
Vr  Vrms
2
Vo2
(16.109)
 Vs2   2  1 Vs2  2Vs  1   
2

The ac ripple voltage is zero at δ = 0 and δ = 1, when the output voltage is pure dc, namely - Vs or Vs,
respectively. The maximum ripple voltage occurs at δ = ½, when Vr = Vs.

The output voltage ripple factor is

Vr 2 V s  1   
RF  
Vo  2  1Vs
(16.110)
2  1     1 
2

     1  FF 2  1
 2  1  2  1 

Circuit operation is characterized by two time domain equations:

During the on-period for T1 and T4, when vo(t) = Vs

di o
L  R i o  E  Vs
dt
which yields
Vs  E
 t
  t
i o t    1  e 
I e for 0  t  t T (16.111)
R  
During the on-period for T2 and T3, when vo(t) = - Vs
di
L o  R i o  E  V s
dt
which, after shifting the zero time reference to tT, gives
V E  t
  t
i o t    s  1  e 
I e for 0  t  T  t T (16.112)
R  
 
The initial conditions I and I are determined by using the steady-state boundary conditions:
tT T

V 1  2e 
e 
E
where I  s  (A)
R T
R
1e 

tT T
(16.113)

V s 2e 
1 e 
E
and I  (A) 
R T
R
1e 
The peak-to-peak ripple current is independent of load emf, E, and twice that given by equation (16.15).

The mean output current is given by

Io 
V o E  
 1  2 V s E 
(A) (16.114)
R R
which can be positive or negative, as seen in figure 16.22c and d.
Chapter 16 DC Choppers 704

Conducting devices
D2 D2 T1 D1 T2 D2 T1 T1 D2 D2 T1 D1 T1 D1 T2 T2 D1D1 T2 D2 T1 D1 T2 T2
T3 D3 T4 T4 D3 D3 T4 D4 T3 D3 T4 T4 D4 D4 T3 D3 T4 D4 T3 T3 D4 D4 T3 D3
1 1
δ T1 D2
½ ½
T1 D2
T1 D2 δ
o o

T1/2 T1 on T2 on T1 on T2 on T1 on T2 on
T1/2

T3/4
T4 on T3 on T4 on
T3/4 T3 on T4 on T3 on

vo +Vs o tT T 2T
vo
Vo o
Vo

o -Vs
o tT T 2T io

io I 
Io I

I 
Io
I
is  is 

I I

(a) (b)

Conducting devices
T1 D1 T2 D2 T1 D1 T1 D1 T2 D1 T1 D1
T4 D4 T3 D3 T4 D4 T4 D4 T3 D3 T4 D4
1 1
δ
½ ½
δ
o o

T1/2 T1 on T2 on T1 on T1 on T2 on T1 on T2 on
T1/2

T3/4
T4 on T3 on T4 on
T3/4 T on T3 on T4 on T3 on
T3 on 4

o tT o tT T
vo +Vs +Vs
vo
Vo
Vo
-Vs -Vs

T io 
io I I
Io

 
I I Io
 
is I is I


I I


 I I 
I I
(c) (d)

Figure 16.22. Four-quadrant dc chopper circuit waveforms:

multilevel (three-level) output voltage (a) with V o > 0 and I o > 0; (b) with V o < 0 and I o < 0;
bipolar (two-level) output voltage (c) with V o > 0 and I o > 0; (d) with V o < 0 and I o < 0.
705 Power Electronics

Figures 16.22c and d show chopper output voltage and current waveforms for conditions of positive
average voltage and current in part (c) and negative average voltage and current in part (d). Each part
is shown with the current having a positive maximum value and a negative minimum value. Such a load
current condition involves activation of all possible chopper conducting paths (sequences) as shown at
the top of each part in figure 16.22 and transposed to table 16.3A. The table shows how the conducting
device possibilities (states) decrease if the minimum value is positive or the maximum value is negative.

Table 16.3A: Four-quadrant chopper bipolar (two-level) output voltage states

Conducting devices sequences

V 0 V 0
T1 D1  T1 D1

T4 D4 I 0 T4 D4

V 0 V 0

T1 D1 T2 D2 I 0 T1 D1 T2 D2

T4 D4 T3 D3 I 0 T4 D4 T3 D3

V 0 V 0
T2 D2  T2 D2

T3 D3 I 0 T3 D3


If the minimum output current is positive, that is, I is positive, then only components for a first and fourth
quadrant chopper conduct. Specifically T 2, T3, D2, and D3 do not conduct. Examination

of figure 16.22c
shows that the output current conduction states are as shown in table 16.3A for I  0 .

If the output current never goes positive, that is I is negative, then

T1, T4, D1, and D4 do not conduct.
The conducting sequence becomes as shown in table 16.3A for I  0 . Because the output is bipolar
(±Vs), the average chopper output voltage, V o does not affect the three possible steady state
sequences. Table 16.3A shows that the conducting devices are independent of the average output
voltage polarity. That is, the switch states are the same on the left and right sides of table 16.3A.

The transition between these three possible sequences, due to a current level polarity change, is
seamless. The only restriction is that both switches in any leg do not conduct simultaneously. This is
ensured by inserting a brief dead-time between a switch turning off and its leg complement being turned
on. That is, dead-time between the switching of the complementary pair (T 1 -T2), and in the other leg the
complementary pair is (T3 -T4 ).

16.6.2 Unified four-quadrant dc chopper - multilevel voltage output switching

In order to generate three output states, specifically ±Vs and 0V, two triangular references are used
which are displaced by 180° from one another as shown in figure 16.22a and b. One carrier triangle is
used to specify the state of the leg formed by T1 and T2 (the complement of T1), while the other carrier
triangle specifies the state of leg formed by switches T3 and T4, (the complement of T3). The output
voltage level switches between +Vs , 0V, and –Vs depending on the modulation index δ, such that 0 ≤ δ
≤ 1. A characteristic of the output voltage is that, depending on δ, only a maximum of two of the three
states appear in the output, in steady-state. The 0V state is always one of the two alternating states. An
alternative method to generate the same switching waveforms, is to us one triangular carrier and two
references, δ and 1-δ.

If δ = 0 then T1 and T4 never turn-on since T2 and T3 conduct continuously which impresses –Vs across
the load. As δ increases from zero, the 0V state appears as well as the – Vs state, the later of which
decreases in duration as δ increases.

At δ = ½ the output is zero since T2 and T3 (or T1 and T4) are never on simultaneously to provide a path
involving the dc source. The output voltage is formed by alternating 0V loops (T 1 and T3 on, alternating
to T2 and T4 on, etc.). The average output voltage is therefore zero.

At the extreme δ = 1, T1 and T4 are on continuously and Vs is impressed across the load. As δ is reduced
from one, the 0V state is introduced, progressively lengthening to all of the period as δ reduces to ½.
Chapter 16 DC Choppers 706

The voltage output in terms of the triangular level vΔ reference is defined by

For 0 ≤ δ < ½
 vΔ > δ, vo= -Vs
 vΔ < δ, vo= 0
For δ = ½
 vΔ > δ, vo= 0
 vΔ < δ, vo= 0
For ½ > δ ≥ 1
 vΔ > δ, vo= 0
 vΔ < δ, vo= Vs

# From figure 16.22b for δ < ½, the average output voltage varies linearly with δ such that
Vo 
1 tT T

 0 dt   V s dt
T  0 tT 
(16.115)
1t 
tT
 T V s   T  1 V s

T  T 
Examination of figure 16.22b reveals that the relationship between tT and δ must produce
when   0 : tT  0 and v o  Vs
when   ½ : tT  T and v o  0
that is
tT
 ½
T
which on substituting for tT /T in equation (16.115) gives
t 
V o   T  1  Vs
 T  (16.116)
  2  1 V s

# From figure 16.22a for δ > ½, the average output voltage varies linearly with δ such that
Vo 
1 tT T

 Vs dt   0 dtT  0 tT  (16.117)
t
 Vs T
T
Examination of figure 16.22a reveals that the relationship between tT and δ must produce
when   ½ : tT  0 and v o  0
when   1 : tT  T and v o  Vs
that is
  tT
 1  ½
T 
which on substituting for tT /T in equation (16.117) gives
V o   2  1V s (16.118)
Alternately, if one pole produces δ then the other pole produces the complement, 1 - δ, such that the
output is the difference, δ – (1 - δ) = 2δ - 1.
Since the same expression results for δ ≤ ½ with bipolar switching, the average output current is the
same for the range 0 ≤ δ ≤ 1, that is

Io 
V o E  
 2  1V s E 
(A) (16.119)
R R
which can be positive or negative, depending on δ and the load emf, E.

Although the average voltage equations of the multilevel and bipolar controlled dc choppers are the
same, the rms voltage and ripple voltage differ, as does the peak-to-peak output ripple current. Unlike
the bipolar controlled chopper, the rms voltage for the multilevel controlled chopper is not a single
continuous function.

# For δ ≤ ½ the rms load voltage is

½
1 T 
 t Vs 
2
Vrms   dt 
T T  (16.120)
 1  2 V s
707 Power Electronics

The output ac ripple voltage is

Vr  Vrms
2
Vo2

   2  1V 
2 2
 1  2 V s s
(16.121)

 2 V s  1  2 
The output voltage ripple factor is
2
Vr V 
RF    r ms   1  FF 2  1
Vo  Vo 
(16.122)
2
 1  
 2    1
1  2  1  2 
Thus as the duty cycle   0 , the ripple factor tends to zero, consistent with dc output voltage, that is Vr
= 0. The ripple factor is undefined when the average output voltage is zero, at δ = ½.

The minimum rms ripple voltage in the output occurs when δ = ½ or 0 giving an rms ripple voltage of
zero, since the average is a dc value at the extremes (0V and -Vs respectively). The maximum ripple
occurs at δ = ¼, when Vr = ½Vs, which is the same as when δ = ¾, (but half that obtained with the bipolar
output control method, Vs).

# For δ ≥ ½ the rms load voltage is

½
1
 V s  dt 
T 2
Vrms  
T
 tT
 (16.123)
 2  1 V s
The output ac ripple voltage is
Vr  Vr ms
2
Vo2
(16.124)
     2  1V s   2 V s
2
2  1 1   
2
 2  1 V s
The minimum rms ripple voltage in the output occurs when δ = ½ or 1 giving an rms ripple voltage of
zero, since the average is a dc value at the extremes (0V and Vs respectively). The maximum ripple
occurs at δ = ¾, when Vr = ½Vs, which is half that obtained with the bipolar output control method.

The output voltage ripple factor is

2
Vr V 
RF   FF 2  1   r ms   1
Vo  Vo 
(16.125)
2
1  1 
 2    1
2  1  2  1 

Thus as the duty cycle   1 , the ripple factor tends to zero, consistent with the output being dc, that is
Vr = 0. The ripple factor is undefined when the average output voltage is zero, at δ = ½.

Circuit operation is characterized by three time domain equations.

During the on-period for T1 and T4, when vo(t) = Vs

di o
L  R i o  E  Vs
dt
which yields
Vs  E
 t
  t
i o t    1  e 
I e for 0  t  t T and   ½ (16.126)
R
 
During the on-period for T2 and T3, when vo(t) = -Vs
di
L o  R i o  E  V s
dt
which, after shifting the zero time reference to tT, gives
V E  t
  t
i o t    s 1  e   I e

for 0  t  T  t T and   ½ (16.127)
R  
The third equation is for a zero voltage loop.
Chapter 16 DC Choppers 708

During the switch off-period, when vo(t) = 0

di o
L  R io  E  0
dt
which, after shifting the zero time reference, in figure 16.22a or b, gives
E  t
  t
i o t     1  e    I e 
R 
0  t  t T and   ½ (16.128)
0  t  T  t T and   ½
 
The initial conditions I and I are determined by using the usual steady-state boundary condition
method and are dependent on the transition states. For example, for continuous steady-state
transitions between +Vs loops and 0V loops, the boundary conditions are given by
tT

V 1e  E
where I  s  (A)
R T
R
1e 
tT
(16.129)

Vs e  1 
E
and I   (A)
R T R
e 1
Figures 16.22a and b show output voltage and current waveforms for conditions of positive average
voltage and current in part (a) and negative average voltage and current in part (b). Each part is shown
with the current having a positive maximum value and a negative minimum value. Such a load current
condition involves the activation of all possible chopper conducting paths, which are shown at the top of
each part in figure 16.22 and transposed to table 16.3B. The conducting device possibilities decrease if
the minimum value is positive or the maximum value is negative.

Table 16.3B: A Four-quadrant chopper multilevel (three-level) output voltage states

Conducting devices sequences

V 0 V 0
T1 D1 T1 T1  T1 D1 D1 D1

T4 T4 T4 D4
I 0 D4 D4 T4 D4

V 0 V 0

D2 D2 T1 D1 T2 D2 T1 T1 I 0 T1 D1 T2 T2 D1 D1 T2 D2

T3 D3 T4 T4 D3 D3 T4 D4 I 0 D4 D4 T3 D3 T4 D4 T3 T3

V 0 V 0
D2 D2 T2 D2  T2 T2 T2 D2

T3 D3 D3 D3
I 0 T3 D3 T3 T3


If the minimum output current is positive, that is, I is positive, then only components for a first and fourth
quadrant chopper conduct. Specifically T 2, T3, D2, and D3 do not conduct, thus do not appear in the
output sequence. Examination

of figure 16.22a shows that the output current conduction states are as
shown in table 16.3B for I  0 .

If the output current never goes positive, that is I is negative, then T1, T4, D1, and D4 do not conduct,
thus do not appear in the output device sequence. The conducting sequence is as shown in table 16.3B
for I  0 .

Unlike the bipolar control method, the output sequence is affected by the average output voltage level,
as well as the polarity of the output current swing. The transition between the six possible sequences
due to load voltage and current polarity changes, is seamless. The only restriction is that both switching
devices in any leg do not conduct simultaneously. This is ensured by inserting a brief dead-time
between a switch turning off and its leg complement being turned on.
709 Power Electronics

Example 16.7: Four-quadrant dc chopper

The H-bridge, dc-to-dc chopper in figure 16.21 feeds an inductive load of 10Ω resistance, 50mH
inductance, and back emf of 55V dc, from a 340V dc source. If the chopper is operated with a 200Hz
multilevel carrier as in figure 16.22 a and b, with a modulation depth of δ = ¼, determine:

i. the average output voltage and switch T 1 on-time

ii. the rms output voltage and ac ripple voltage, hence voltage ripple and form factors
iii. the average output current, hence quadrant of operation
iv. the electromagnetic power being extracted from the back emf E.

If the mean load current is to be halved, what is

v. the modulation depth, δ, requirement
vi. the average output voltage and the corresponding switch T1 on-time
vii. the electromagnetic power being extracted from the back emf E?

Solution

The main circuit and operating parameters are

 modulation depth δ = ¼
 period Tcarrier = 1/fcarrier = 1/200Hz = 5ms
 E = 55V and Vs = 340V dc
 load time constant τ = L /R = 0.05mH/10Ω = 5ms

i. The average output voltage is given by equation (16.116), and for δ < ½,
t 
V o   T  1 V s   2  1V s
T 
= 340V×  2×¼ - 1 = -170V
where
tT  2 T = 2×¼× ½×5ms  = 1.25ms
Figure 16.22 reveals that the carrier frequency is half the switching frequency, thus the 5ms in the
above equation has been halved. The switches T1 and T4 are turned on for 1.25ms, while T2 and T3
are subsequently turned on for 3.75ms.

ii. The rms load voltage, from equation (16.120), is

Vrms  1  2 V s
= 340V× 1 - 2×¼ = 240V rms
From equation (16.121), the output ac ripple voltage, hence voltage ripple factor, are
Vr  2 V s  1  2 
= 2 ×340V× ¼  1 - 2×¼ = 170V ac
Vr 170V
RF   1 FF  RF 2  1  2  1.41
Vo 170V

iii. The average output current is given by equation (16.119)

V  E  2  1V s  E
Io  o 
R R
340V×  2×¼ - 1 - 55V
= = -22.5A
10Ω
Since both the average output current and voltage are negative (-170V and -22.5A) the chopper
with a modulation depth of δ = ¼, is operating in the third quadrant.

iv. The electromagnetic power developed by the back emf E is given by

PE  E I o  55V×  -22.5A  = -1237.5W

v. The average output current is given by

Io 
V o E  
 2  1V s E 
R R
when the mean current is -11.25A, δ = 0.415, as derived in part vi.
Chapter 16 DC Choppers 710

vi. Then, if the average current is halved to -11.25A

Vo  E  Io R
=55V - 11.25A×10Ω = -57.5V

The average output voltage rearranged in terms of the modulation depth δ gives
 V 
  ½  1  o 
 Vs 
 -57.5V 
= ½×  1 + = 0.415
 340V 
The switch on-time when δ < ½ is given by
tT  2T = 2×0.415× ½×5ms  = 2.07ms

From figure 16.22b both T1 and T4 are turned on for 2.07ms, although, from table 16.3B, for negative
load current, I o  -11.25A, the parallel

connected freewheel diodes D2 and D3 conduct alternately,
rather than the switches (assuming I o  0 ). The switches T1 and T4 are turned on for 1.25ms, while
T2 and T3 are subsequently turned on for 2.93ms.

vii. The electromagnetic power developed by the back emf E is halved and is given by
PE  E I o  55V×  -11.25A  = -618.75W
♣

Reading list

Dewan, S. B. and Straughen, A., Power Semiconductor Circuits,

John Wiley & Sons, New York, 1975.

Dubey, G.K., Power Semiconductor Controlled Drives,

Prentice-Hall International, 1989.

Mohan, N., Undeland, T. M., & Robbins, W.P., Power Electronics: Converters, Applications & Design,
John Wiley & Sons, New York, 2003.

Problems

16.1. The dc GTO thyristor chopper shown in figure 16.1c operates at 1kHz and supplies a series 5Ω
and 10mH load from an 84V dc battery source. Derive general expressions for the mean load
voltage and current, and the load rms voltage at an on-time duty cycle of δ. Evaluate these
parameters for δ = 0.25.
[21 V, 4.2 A; 42 V]

16.2. The dc chopper in figure 16.1c controls a load of R = 10Ω, L = 10mH and 40V battery. The
supply is 340V dc and the chopping frequency is 5kHz. Calculate (a) the peak-to-peak load
ripple current, (b) the average load current, (c) the rms load current, (d) the effective input
resistance, and (e) the rms switch current.