SOLUTION MANUAL

to

Introduction to
Modern Power
Electronics

by

Andrzej M. Trzynadlowski

John Wiley & Sons, Inc.

 Chapter 1

P1.3. (a) From Fig. 1.18(a), Vo,dc = 0.32 pu. From Eq. (1.44),
1
V o, dc = [1 + cos ( 90 o )] = 0.318 pu
π

(b) From Fig. 1.18(a), Vo,1,p = 0.5 pu. Thus, Vo,1 = 0.5/%2 = 0.35 pu

P1.4. (a) From Fig. 1.18(b), Vo,1,p = 0.6 pu

1 π 1
V o=1 [ π - + sin ( π )] = 0.5 pu
2π 2 2

(b) Vo,1 = 0.6/%2 = 0.424 pu

V o, h = 0. 52 - 0. 424 2 = 0.265 pu

(c) dh(V) = 0.265/0.424 = 0.625

P1.5. As in the all subsequent problems involving spectral analysis, formulas

provided in Appendix B are used, in the angle-domain version (substitute 2π
for T and ωt for t). The output voltage waveform, vo(ωt) has both the odd and
half-wave symmetries, so that there are no even harmonics. Consequently,
the peak value, Vo,k,p, of the kth harmonic can be found as

2
 π

V o k, ,p = V o k, ,s = ∫ vo ( ω t)s in( kω t) dω t
2
k = 1 ,3 ,5 ,...
π0
which yields
4V i
V o,k, p = k = 1,3,5,...
kπ

It can be seen that the amplitude, Vo,1,p, of the fundamental voltage is as

much as 27% higher than the dc input voltage, Vi.

P1.6. (a)

115 2
V o, dc = [1 + cos ( 30 o )] = 96.6 V
π

(b)

1 π 1
V o = 115 2 [ π - + sin ( 60 o )] = 113.3 V
2π 6 2

Eq. (1.45) for a generic ac voltage controller was used. It is also valid for the
generic rectifier, since the rms value does not depend on the sign of a
function.

V o, ac = 113. 32 - 96. 6 2 = 59.2 V

(c) kr(V) = 59.2/96.6 = 0.613

P1.7. f1 = 120 Hz

3
P1.8. This is a rather tedious problem. The waveform of output voltage,
vo(ωt), given by

has only the half-wave symmetry. Therefore, only odd harmonics are present
in the spectrum, and the amplitudes of odd harmonics must be computed as

2 2
V o, k, p = V o, k, s + V o, k, c

For instance, with αf = 90o, the peak value of fundamental voltage, Vo,1,p, is

2
V o,1, p = 0. 52 + (-0.3183 ) = 0.5927

P1.9. The waveform of output voltage of the generic rectifier has no half-
wave symmetry, so that all harmonics are present, including the dc
component ("zero harmonic"). Analogous waveform in the generic ac voltage
controller has the half-wave symmetry, and only odd harmonics appear in the
spectrum of output voltage. The dc component and even harmonics are
absent.

P1.10. S1 & S2: D1,2 = 70/100 = 0.7

S3 & S4: D3,4 = 0

S5: D5 = 1 - 0.7 = 0.3

S5 is switching complementarily with S1 & S2.

P1.11. S1 & S2: D1,2 = 0

S3 & S4: D3,4 = 35/100 = 0.35

S5: D5 = 1 - 0.35 = 0.65

S5 is switching complementarily with S3 & S4.

P1.12. tON + tOFF = tON/D = 0.2/0.4 = 0.5 ms

4
 fsw = 1/(tON + tOFF) = 1/0.5 = 2 kHz

P1.13. tON + tOFF = 1/1250 = 0.8 ms

tON = D(tON + tOFF) = 0.3 Η 0.8 = 0.24 ms

tOFF = 0.8 - 0.24 = 0.56 ms

P1.14. Since M = D = tON/(tON + tOFF), then

tON/tOFF = M/(1 - M) = 0.8/0.2 = 4

P1.15. tON + tOFF = 1/(50 Η 60) = 1/3000 s = 1/3 ms 6 fsw = 3 kHz

tON = 0.3/3 = 0.1 ms

P1.16. (a) no. of pulses = fsw/f = 1500/60 = 25 pulses per cycle

(b) D1,2 = 0.333, D5 = 0.667

(c) pulse width = tON = D1,2(tON + tOFF) = D1,2/fsw = 0.333/1500 = 0.222

ms

P1.17. (a) S1 & S2 6 ON S3 & S4 6 OFF S5 6 OFF

(b) S1 & S2 6 OFF S3 & S4 6 OFF S5 6 ON

(c) S1 & S2 6 OFF S3 & S4 6 ON S5 6 OFF

P1.18. Δ = 360o/8 = 45o

αn = (n - 0.5) Η 45o

Dn = 0.75sin(αn)

γn = Dn Η 45o

5
 n αn Dn γn
1 22.5o 0.287 12.9o
2 67.5o 0.693 31.2o
3 112.5o 0.693 31.2o
4 157.5o 0.287 12.9o
5 202.5o 0.287 12.9o
6 247.5o 0.693 31.2o
7 292.5o 0.693 31.2o
8 337.5o 0.287 12.9o

P1.21. Eqs. (1.73) and (1.75) can be modified to

M
io(t1)=io(t0 )+ [V pi _, s (ω it0 )_n- Rio(t0 )- E ∆ T ]
L
and
R E
i o ( t 2 ) = i o ( t 1 )[1 - (1 - M) ∆T] - (1 - M) ∆T
L L

where ΔT = 1/720 s. Substituting the data (if you have the first print of the
book, correct the inductance, L, to 12 mH), yields

io(t1) = 0.971io(t0) + 19.65 |sin(377t0)| - 5.21

and

io(t2) = 0.971io(t1) - 5.21

Since 720/60 = 12, there are twelve switching intervals per cycle. The
initial value, io(0), of the current can be assumed to equal the average ouput
current, Io,that is

6
 Vo-E
i o (0) = I o =
R

where
2
Vo= V i, p M = 108 V
π

Thus, io(0) = 36 A. Computed values of the current are tabulated below.

t (s) io (A) t (s) io (A)

0.0000 36.0 0.0090 29.1
0.0007 29.7 0.0097 23.0
0.0014 23.7 0.0104 27.0
0.0021 27.6 0.0111 21.0
0.0028 21.6 0.0118 32.2
0.0035 32.8 0.0125 26.0
0.0042 26.6 0.0132 39.7
0.0049 40.3 0.0139 33.4
0.0056 33.9 0.0146 44.2
0.0063 44.7 0.0153 37.7
0.0069 38.2 0.0160 41.2
0.0083 35.3 0.0167 34.8

Correction of the initial current to 33.8 A yields the final value of the current
equal to the initial value:

t (s) io (A) t (s) io (A)

0.0000 33.8 0.0090 27.6
0.0007 27.6 0.0097 21.6
0.0014 21.6 0.0104 25.6
0.0021 25.6 0.0111 19.6
0.0028 19.6 0.0118 30.8

7
 0.0035 30.9 0.0125 24.7
0.0042 24.8 0.0132 38.5
0.0049 38.5 0.0139 32.1
0.0056 32.2 0.0146 43.0
0.0063 43.0 0.0153 36.6
0.0069 36.6 0.0160 40.1
0.0083 33.8 0.0167 33.8

The substantial ripple of the current results from the low switching
frequency. In practice, it would be an order of magnitude higher. Note that
once the current equations are found, a simple computer program allows easy
repeating of the calculations with various values of io(0).

P1.22. Substituting the data, including ΔT = 1/500 s, to Eqs. (1.73) and

(1.75), yields

io(t1) = 0.4io(t0) + 9.76sin(314t0)

and

io(t2) = 0.8io(t1)

The initial current, io(0), can be found using Eq. (1.70), which gives io(0) =
-5.92 A. Computed values of the current are tabulated below.

t (s) io (A) t (s) io (A)

0.0000 -5.92 0.0115 3.14
0.0015 -2.37 0.0120 2.51
0.0020 -1.89 0.0135 -4.73
0.0035 4.98 0.0140 -3.79
0.0040 3.98 0.0155 -10.80
0.0055 10.88 0.0160 -8.64
0.0060 8.70 0.0175 -12.74
0.0075 12.76 0.0180 -10.19

8
 0.0080 10.21 0.0195 -9.81
0.0095 9.82 0.0200 -7.85
0.0100 7.86

The poor convergence is due to the low load inductance and low switching
frequency. The initial current must be corrected to -7.87 A to yield the final
value equal to the initial value:

t (s) io (A) t (s) io (A)

0.0000 -7.87 0.0115 3.16
0.0015 -3.15 0.0120 2.53
0.0020 -2.52 0.0135 -4.71
0.0035 4.73 0.0140 -3.77
0.0040 3.78 0.0155 -10.78
0.0055 10.79 0.0160 -8.63
0.0060 8.63 0.0175 -12.74
0.0075 12.74 0.0180 -10.19
0.0080 10.19 0.0195 -9.84
0.0095 9.82 0.0200 -7.87
0.0100 7.86

Note that once the current equations are found, a simple computer
program allows easy repeating of the calculations with various values of io(0).

9
 Chapter 4

P4.1.

3 3 230 2
V o,dc = _ = 155.3 V
2π 3

Io,dc = 155.3/10 = 15.53 A

P4.2. The output voltage of the rectifier varies between VLN,p and VLN,p/2 (see
Fig. 4.3), that is, between 187.8 V and 93.9 V. Therefore, with E = 180 V, the
conduction is discontinuous.

P4.3. The output voltage of the rectifier varies between VLL,p and %3VLL,p/2,
that is, between 650.5 V and 563.4 V. Therefore, with E = 480 V, the
conduction is continuous.

10
P4.4. kE = 480/(460%2) = 0.74

ν = tan-1(377 Η 0.05/2.1) = 83.6o

The diagram in Fig. 4.10 indicates continuous conduction. Thus,

Vo = (3/π) Η 460%2 = 621 V

Io = (621 - 480)/2.1 = 67 A

P4.5. kE = 600/(460%2) = 0.922

αc = sin-1(0.922) - π/3 = 7.2o

αe = 60o - 7.2o = 52.8o

β = 52.8o - 7.2o = 45.6o

P4.8. As seen from the diagram in Fig. 4.21, it is the firing angle of 30o that
is always feasible. This is because the SCRs are fired at the instant when the
corresponding line-to-line voltage is peaking.

P4.10. kE = 260/(%2 Η 230) = 0.8

ν = tan-1(377 Η 0.0007/0.15) = 60.4o

According to Fig. 4.21, the firing angle of 25 o is feasible, and according to Fig.
4.22, the conduction is continuous with this angle. Thus,

Vo = (3/π) Η %2 Η 230 Η cos(25o) = 281.5 V

and
Io = (281.5 - 260)/0.15 = 143.4 A

P4.11. kE = 260/(%2 Η 230) = 0.8

ν = tan-1(377 Η 0.0007/0.15) = 60.4o

11
According to Fig. 4.21, the firing angle of 45 o is feasible, and according to Fig.
4.22, the conduction is discontinuous with this angle. Thus, the extinction
angle must be found from the current waveform given by Eq. (4.46).
Substituting the data, yields
ω t -π /4
io ( ω t)= 1069[sin( ω t - 0.4 ) - 1.62+ 0.918e 1.76 ]
o -

Computing the current waveform allows to pinpoint the extinction angle,

αe, at which the current reaches zero. It occurs at ωt of about 82.5o, which
gives the conduction angle, β, as

β = 82.5o - 45o = 37.5o

The average voltage, Vo,dc, can now be found from Eq. (4.47) as Vo,dc = 263 V.
The average current, Io,dc, is

Io,dc = (263 - 260)/0.15 = 20 A

P4.12. VLL,p = %2 Η 230 = 325 V

ν = tan-1(377 Η 0.0007/0.15) = 60.4o

Substituting ν and αf in condition (4.45) gives kE < -0.52 (this value can also be
read from the diagram in Fig. 4.22). Consequently, E < -0.52 Η 325 V, that is,
E < -169 V.

P4.13. From Eq. (4.49),

Fp = (3/π)cos(60o) = 0.477

P4.14. VLL,p = %2 Η 460 = 650.5 V

With an ideal ac source, Vo,dc = (3/π) Η 650.5 Η cos(40o) = 476 V

With a source with inductance, Vo,dc = 0.9 Η 476 = 428.4 V, that is, the dc
voltage is reduced by ΔVo,dc = 47.6 V.

From Eq. (4.64), XsIo,dc = (π/3)ΔVo,dc = (π/3) Η 47.6 = 49.8 V. This, when
substituted in Eq. (4.63), yields

μ = |cos-1[cos(40o) - 2 Η 49.8/650.5] - 40o| = 12.2o

12
P4.15.

n αn Dn γn α1,n α2,n
1 7.5o 0.487 7.30o 3.85o 11.15o
2 22.5o 0.635 9.52o 17.74o 27.26o
3 37.5o 0.739 11.08o 31.96o 43.04o
4 52.5o 0.793 11.90o 46.55o 58.45o

P4.16.

n αn Dn γn α1,n α2,n
1 7.5o 0.705 10.58o 2.21o 12.79o
2 22.5o 0.792 11.88o 16.56o 28.44o
3 37.5o 0.795 11.92o 31.54o 43.46o
4 52.5o 0.774 11.61o 46.70o 58.31o
P4.17. From Eq. (4.93),
2π π
t ON( = 0.05 sin ( ) = 5.4_ 10 -6 s = 5.4 µs
24_2 π_50
min )
24

and from Eq. (4.95),

2π π
t OFF( = [1 - cos ( )] = 7.1_ 10 -6 s = 7.1 µs
24_2 π_50
min )
24

P4.18. αf1 = 60o, αf2 = 120o

3
Vo= 460 2 cos ( 60 o ) = 311 V
π

At ωt = 15o,

vo,1 = vCB = 460%2cos(15o + 30o) = 460 V

vo,2 = vCA = -460%2sin(15o) = -168.4 V

13
 Δvo = 460 - 168.4 = 291.6 V

P4.19. VLL,pk(max) = 1200/1.4 = 857 V

VLL(max) = 857/%2 = 606 V

Vo(max) = (3/π) Η 857 = 818 V

Io(max) = 100/1.2 = 83 A

Rmin = 818/83 = 9.9 Ω

P4.20.

3 o
V LL, pk(rat) cos ( 30 ) = 400
π

400
V LL, pk(rat) = = 483.7 kV
3
cos ( 30 o )
π

Vrat ∃ 1.2 Η 1.1 Η 483.7 = 638.5 kV

n = 638.5/5 = 128

14
 Chapter 5

P5.2.

377_0.005
_ = tan -1 ( ) = 43. 3o
2

Values of the magnitude control ratio, M, for individual firing angles, αf, can
now be estimated from Fig. 5.3 as approximately equidistant from the
characteristics for the purely resistive and purely inductive loads.

αf M Vo
30o 1.0 120 V
90o 0.8 96 V
150o 0.2 24 V

15
P5.3. From Fig. 5.3, for ν = 0 and M = 0.75, αf = 85o

P5.4. The output voltage is the line-to-neutral voltage, hence Vo = M Η

460/%3. The magnitude control ratio can be found from Fig. 5.12.

αf M Vo
20o 0.98 260 V
90o 0.55 146 V
130o 0.12 32 V

P5.5. Vo = %0.65 Η 115 = 93 V

P5.6. tON + tOFF = 1/(60 Η 20) = 8.33 Η 10-4 s = 833 μs

tON = 0.7 Η 833 = 583 μs

tOFF = 833 - 583 = 250 μs

P5.7. Similarly to the input voltage, the output voltage, vo, is the line-to-line
voltage.

Vo = %0.4 Η 230 = 145 V

Vo,1 = 0.4 Η 230 = 92 V

P5.8. From Eq. (5.38),

3
V o, LN ,1 = _0.8_460 = 351 V
π

Hence, Vo,LL,1 = 351 Η %3 = 608 V

16
P5.9. Permitted states of the matrix converter: 73, 74, 76, 81, 82, 84, 97,
98, 100, 137, 138, 140, 145, 146, 148, 161, 162, 164, 265, 266, 268, 273,
274, 276, 289, 290, 292

P5.10. Vrat ∃ %2 Η (1 + 0.4) Η 460 = 911 V

Io(rat) = 10000/(%3 Η 460) = 12.6 A

Irat ∃ %2/π Η (1 + 0.2) Η 12.6 = 6.8 A

P5.11. Based on results of P5.10, Irat ∃ 6.8/%3 = 3.9 A

P5.12. tON(min) = 2π/(30 Η 377) Η 0.04 = 22.2 Η 10-6 s = 22.2 μs

tOFF(min) = 2π/(30 Η 377) Η (1 - 0.96) = 22.2 Η 10-6 s = 22.2 μs

Chapter 6

P6.1. (1) First quadrant (M > 0, see answer to the next Problem)

(2) Vo = 0.7 Η 240 = 168 V

(3) Io = 168/10 = 16.8 A

P6.2. No. There is no EMF in the load, which therefore cannot supply any
power.

P6.3. tON + tOFF = 1/1200 = 8.33 Η 10-4 s = 833 μs

(a) tON = 0.7 Η 833 = 583 μs

(b) tOFF = 833 - 583 = 250 μs

(c) From Eq. (6.20), fsw(pu) = 0.02/10 Η 1200 = 2.4 pu

17
 From Eq. (6.18), Io,ac(pu) = 0.7(1 - 0.7)/(2%3 Η 2.4) = 0.025 pu

From Eq. (6.19), Io,ac = 240/10 Η 0.025 = 0.6 A

(d) Io = 168/10 = 16.8 A

kr(I) = 0.6/16.8 = 0.036

P6.4. (a) fsw(pu) = 0.02/10 Η 1200 = 2.4 pu

(b) Io,ac(pu) = 0.7(1 - 0.7)/(2%3 Η 2.4) = 0.025 pu

P6.5. (a) Fourth quadrant (see Table 6.1 or Fig. 6.21)

(b) Vo = -0.7 Η 150 = -105 V

(c) Io = (-105 + 120)/0.5 = 30 A

P6.6. (a) The chopper operates in the fourth quadrant (see P.6.5). Thus,
D4 = 1 - 0.7 = 0.3

(b) tON + tOFF = 1/900 = 1.11 Η 10-3 s = 1.11 ms

tON = 0.3 Η 1.11 = 0.33 ms

(c) tOFF = 1.11 - 0.33 = 0.78 ms

(d) τ = 1.7 Η 10-3/0.5 = 0.0034 s

fsw(pu) = 0.0034 Η 900 = 3.06 pu

Io,ac(pu) = 0.7 Η (1 - 0.7)/(2%3 Η 3.06) = 0.0198 pu

Io,ac = 0.0198 Η 150/0.5 = 5.94 A

P6.7. (a) First quadrant, since Vo > 0, while Vo > E implies Io > 0

(b) Io = (125 - 100)/0.5 = 50 A

18
 (c) M = 125/150 = 0.833

(d) D1 = 0.833

P6.8. Vo = E + RIo

(a) First quadrant, since E > 0 and Io > 0

D1 = (210 + 0.1 Η 150)/300 = 0.75

(b) Fourth quadrant, since E < 0 and Io > 0

D4 = (-210 + 0.1 Η 150)/300 + 1 = 0.65

P6.9. (a) Vo = 0.5 Η 300 = 150 V

Io = (150 - 100)/0.1 = 500 A

(b) The M = -0.8 value is erroneous. Change it to -0.6.

Vo = -0.6 Η 300 = -180 V

Io = (-180 + 200)/0.1 = 200 A

P6.10. From Eq. (6.18), with M = 0.5 (which results in the highest ripple),

0.5(1 - 0.5)
0.02 >
2 3 f sw(pu)

which is satisfied if fsw(pu) > 3.6

τ = 0.0005/0.1 = 0.005 s

fsw > 3.6/0.005 = 720 Hz

P6.11. First quadrant: E = 240 V E/Vi = 240/400 = 0.6

M = 0.6 ... 1 D1 = 0.6 ... 1

19
 Second quadrant: E = 240 V E/Vi = 0.6

M = 0 ... 0.6 D2 = 0.4 ... 1

Third quadrant: E = -240 V E/Vi = -0.6

M = -1 ... -0.6 D3 = 0.6 ... 1

Fourth quadrant: E = -240 V E/Vi = -0.6

M = -0.6 ... 0 D4 = 0.4 ... 1

P6.12. Vo,p = 12/(1 - 0.9) = 120 V

tON + tOFF = 1/2000 = 5 Η 10-4 s = 500 μs

tOFF = (1 - 0.9) Η 500 = 50 μs

P6.13. Vrat > 12 V

Irat > 150/12 = 12.5 A

P6.14. Vrat > Vi,p = %2 Η 230 = 325 V

Irat > Io,rat = 3000/300 = 10 A

P6.15. Vrat > Vi,p = %2 Η 460 = 651 V

Irat > Io,rat = 10000/600 = 17 A

P6.16. Vrat > Vo,p = 6/(1 - 0.9) = 60 V

ID(rat) > Io,dc(rat) = 240/6 = 40 A (the average output voltage of 6 V

is used to determine the rated average
current of the chopper)

IS(rat) > 0.9 Η 40/(1 - 0.9) = 360 A

20
 Chapter 7

P7.1. Simple square-wave mode: Vo,1 = 0.9 Η 310 = 279 V

Optimal square-wave mode: Vo,1 = 0.828 Η 310 = 257 V (see pp. 276-
277)

P7.2. In both modes of the inverter, the output voltage waveform has both
the odd and half-wave symmetry. Therefore, Eq. (B.23) can be used, which, in
the angle-domain version yields

4 2
4 π
ck = ∫ s i (n kω t ) ωd t = [ c o( skα d ) - c o( sk ) ]
π αd kπ 2
for k = 1, 3, 5, .... In the simple square-wave mode, αd = 0, and in the optimal

21
squre-wave mode, αd = 0.405 rad.

MODE: simple optimal

square- square-
wave wave
k ck ck
1 1.273 1.170
3 0.424 0.148
5 0.255 0.112
7 0.182 0.173

Note that the low-order voltage harmonics in the optimal square-wave mode
are significantly reduced in comparison with those in the simple square-wave
mode.

P7.3. The switching intervals are 360o/30 = 12o wide.

α13 = (13 - 1/2) Η 12o = 150o

F(m,α13) = 0.6sin(150o) = 0.3

γ13 = 0.3 Η 12o = 3.6o

α1,13 = 150o - 3.6o/2 = 148.2o

α2,13 = 150o + 3.6o/2 = 151.8o

P7.4. VLL,1,p = 2/%3 Η 620 = 716 V (peak line-to-line)

VLL,1 = 716/%2 = 506 V (rms line-to-line)

VLN,1 = 506/%3 = 292 V (rms line-to-neutral)

P7.5. 5 - 1 - 3 - 2 - 6 - 4 - ...

22
P7.6. The switching intervals are 15o wide.

α4 = (4 - 0.5) Η 15o = 52.5o

Phase A: 0 # α4 < π/3

F(m, α4) = 2 Η 0.9 Η cos(52.5o - 60o) - 1 = 0.785

a = (1 + 0.785)/2 = 0.8925

γA,4 = 0.8925 Η 15o = 13.3875o

α1A,4 = 52.5o + 13.3875o/2 = 59.19375o

α2A,4 = 52.5o - 13.3875o/2 = 45.80625o

Phase B: 4π/3 # α4 - 120o < 5π/3

F(m, α4 - 120o) = -1

b = (1 - 1)/2 = 0

no switchings in this phase

Phase C: 2π/3 # α4 - 240o < π

F(m, α4- 240o) = 2 Η 0.9 Η sin(52.5o - 240o - 30o) - 1 =

0.096

c = (1 + 0.096)/2 = 0.548

γC,4 = 0.548 Η 15o = 8.22o

α1C,4 = 52.5o + 8.22o/2 = 56.61o

α2C,4 = 52.5o - 8.22o/2 = 48.39o

The first angle indicates the turn-on instant of the upper (common-anode)
switch in a given phase (leg) of the inverter, the second angle indicates the
turn-off instant of this switch. Vice-versa, the other (common-cathode) switch
turns off at the first angle and turns on at the second angle.

23
P7.7. The switching intervals are 10o wide and 292 μs long.

m = 400%2/620 = 0.912

α19 = (19 - 0.5) Η 10o = 185o 6 Sextant 4 6 α = 5o

X = 3 = 0112, Y = 1 = 0012, Z1 = 0002 = 0

State sequence: X - Y - Z1 = 3 - 1 - 0

dX = d3 = 0.912sin(60o - 5o) = 0.747

dY = d1 = 0.912sin(5o) = 0.079

dZ = d0 = 1 - 0.747 - 0.079 = 0.174

t3 = 0.747 Η 292 = 218 μs

t1 = 0.079 Η 292 = 23 μs

t0 = 0.174 Η 292 = 51 μs

P7.8. The switching intervals are 10o wide and 292 μs long.

m = 400%2/620 = 0.912

α19 = (19 - 0.5) Η 10o = 185o 6 Sextant 4 6 α = 5o

X = 3, Y = 1, Z = 0

State sequence: X - Y - Z = 3 - 1 - 0

dX = d3 = 0.912sin(60o - 5o) = 0.747

dY = d1 = 0.912sin(5o) = 0.079

dZ = d0 = 1 - 0.747 - 0.079 = 0.174

t3 = 0.747 Η 292 = 218 μs

t1 = 0.079 Η 292 = 23 μs

24
 t0 = 0.174 Η 292 = 51μs

States and their durations are the same for both versions of the space-vector
PWM technique. In Problems 7.7 and 7.8, the same state sequence, 3 - 1 - 0,
applies to the switching interval in question. However, in the next switching
interval, the high-quality state sequence would be 1 - 3 - 7, and the high-
efficiency sequence 0 - 1 - 3.

P7.9. In the angle domain, the 19th switching interval extends from 180o to
190o. State 3 occupies the first 0.747 of this interval, from 180o to 180o +
0.747 Η 10o = 187.47o. State 1 occupies the next 0.079 of the interval, from
187.47o to 187.47o + 0.079 Η 10o = 188.26o. State 0 occupies the remaining
part of the interval, from 188.26o to 190o. In the next, 20th, switching interval,
the first state is State 1. The last state in the previous, 18 th, interval was State
7. Consequently, prior to the beginning of the switching interval in question,
all switching variables are 1 (State 7). At 180 o, switching variable a changes
from 1 to 0 (State 3), at 187.47 o, variable b changes from 1 to 0 (State 1), and
at 188.26o, variable c follows suit (State 0). At 190o, variable c changes from 0
to 1 (State 1). Thus:

Switch SA: turns off at 180o

Switch SA': turns on at 180o
Switch SB: turns off at 187.47o
Switch SB': turns on at 187.47o
Switch SC: turns off at 188.26o and turns on at 190o
Switch SC': turns on at 188.26o and turns off at 190o

P7.10. In the angle domain, the 19th switching interval extends from 180o to
190o. State 3 occupies the first 0.747 of this interval, from 180o to 180o +
0.747 Η 10o = 187.47o. State 1 occupies the next 0.079 of the interval, from
187.47o to 187.47o + 0.079 Η 10o = 188.26o. State 0 occupies the remaining
part of the interval, from 188.26o to 190o. In the next, 20th, switching interval,
the first state is State 0 again. The last state in the previous, 18th, interval was
State 3. Consequently, prior to the beginning of the switching interval in
question, a = 0, b = 1, and c = 1 (State 3). The switching variables do not
change at 180o since the first state in the 19th switching interval is State 3
again. At 187.47o, variable b changes from 1 to 0 (State 1), and at 188.26o,
variable c follows suit (State 0). At 190o, the 20th switching interval begins,
and the inverter remains in State 0. Thus:

Switch SA: is not switched

Switch SA': is not switched
Switch SB: turns off at 187.47o

25
Switch SB': turns on at 187.47o
Switch SC: turns off at 188.26o
Switch SC': turns on at 188.26o

Comparison of results of this problem with those of Problem 7.9 well illustrates
the reduction in the number of switchings when the high-quality state
sequence is replaced with the high-efficiency sequence.

P7.11. Switching variable a changes from 0 to 1 at 0o, 14.80o, 89.07o, 92.07o,

170.52o, 189.48o, 267.93o, 270.93o, and 345.2o. It changes from 1 to 0 at
9.48o, 87.93o, 90.93o, 165.2o, 180o, 194.8o, 269.07o, 272.07o, and 350.52o.

P7.12. IL,1,p = (2%3/π) Η 200 = 220.5 A

IL,1 = 220.5/%2 = 155.9 A

VLN,1 = 2 Η 155.9 = 311.8 V (the load inductance produces only

voltage spikes during switchings, and
its contribution to the output voltage is
negligible)

VLL,1 = 311.8 Η %3 = 540.1 V

P7.13. Switching pattern for switch SA is: 2.24o (ON), 5.6o (OFF), 21.26o, 30o,
38.74o, 54.4o, 57.76o, 122.24o, 125.6o, 141.26o, 150o, 158.74o, 174.4o (ON), and
177.76o (OFF). Switching pattern for switch SA' is shifted by 180o, that is,
182.24o (ON), 185.6o (OFF), ..., 357.76o (OFF).

P7.14. Vi = (3/π) Η %2 Η 1.2 = 1.62 kV

VLL,p = 1.065 Η 1.62 = 1.73 kV

VLL = 1.73/%2 = 1.22 kV

VLN = 1.22/%3 = 0.7 kV

P7.15. 2010 = 2023 6 a = 2, b = 0, c = 2

26
P7.16. See Fig. 7.55. Shifting the vAB waveform rightwards by 15o produces
the even and half-wave symmetry. The rms value of the line-to-line voltage
can be calculated as

1 5
π π
3 12

V L L= [ ∫ 1 dω t + ∫ 0 5. dω t] = [ π + 0 . 2 5 ( π - π ) ]= 0 .7 6 7 4
2 2 2 2 1 5 1
π 0 1 π 4 12 4
π
4

The peak value of the fundamental of this voltage is

1 5
π π
4 12
4
V LL,1,p = [ ∫ 1 cos( ω t)dω t + ∫ 0.5 cos( ω t)dω t
π 0 1
π
4

4 1 5 1
= [ sin( π ) - sin(0)+ 0.5sin( π ) - 0.5sin( π )] = 1.065
π 4 12 4

and the rms value, VLL,1 = 1.065/%2 = 0.7531. Hence, the harmonic content is
and the total harmonic distortion of the line-to-line voltage is dh,LL=
0.1274/0.7531 = 0.169.V LL, h = 0. 7638 2 - 0. 7531 2 = 0.1274

Similarly, to acquire the even and half-wave symmetries, the vAN waveform
should be shifted leftwards by 15o. Then

and, using the same approach as before, the rms value, VLN, of the line-to-
neutral voltage is calculated as 0.441, i.e., equal to VLL/%3. The rms value,
VLN,1, of the fundamental line-to-neutral voltage is 0.4348. Thus,

2 2
P7.17. Since, as calculated h =Example
V LN ,in 0. 441 - 7.4,
0. 4348
the =
dc0.074
link voltage is given by

27
then the peak value of this voltage is 200 + 243.8 = 443.8 V, and the inverter
switches must be able to withstand at least this voltage (in contrast, if a
vo = 200 - 243.8 cos (64550t + 0.609) V
regular inverter supplied from the same dc voltage of 200 V was considered,
the switches would have to withstand only the 200 V).

As a result of the clamping, the voltage gain of the inverter is reduced by 24%
(see the conclusion of Example 7.4). To compensate for this drop, the dc input
voltage must be raised by the same 24%, to 248 V. Since voltage pulses are
clipped to 1.3 of that voltage, their amplitude is 1.3 Η 248 = 322.4 V. Thus, in
comparison with the unclamped dc link, the required voltage rating of the
inverter switches is reduced by 27%.

P7.18. According to Eqs. (7.12) and (4.89), duty ratio of switch SA in the nth
switching interval is given by

1 2 1
D A ,n = an = _1+ m [sin( α n )+ sin(3α n )]_
2 3 6
where
1 2π
αn = (n - )
2 18

Thus, the on-time of the switch is

D A, n
t ON =
18_120

and the off-time is

1 - D A, n
t OFF =
18_120

The extremal values of DA,n occur when m = 1. Indeed observing the waveform
of third-harmonic modulating function in Fig. 4.41, it can be seen that, at
certain values of ωt, it reaches the extremal values of -1 and +1,
corresponding to DA,n = 0 and DA,n = 1, respectively. The values of tON and tOFF
for m = 1 and individual switching intervals are listed in the following table.

n DA,n tON (μs) tOFF (μs)

1 0.6484 300.2 162.8

28
 2 0.8849 409.7 53.3
3 0.9904 458.5 4.4
4 0.9944 460.4 2.6
5 0.9811 454.2 8.7
6 0.9944 460.4 2.6
7 0.9904 458.5 4.4
8 0.8849 409.7 53.3
9 0.6484 300.2 162.8
10 0.3516 162.8 300.2
11 0.1151 53.3 409.7
12 0.0096 4.4 458.5
13 0.0056 2.6 460.4
14 0.0189 8.7 454.2
15 0.0056 2.6 460.4
16 0.0096 4.4 458.5
17 0.1151 53.3 409.7
18 0.3516 162.8 300.2

It can be seen that the shortest tON and tOFF times are both 2.6 μs.

P7.19. The peak value of the voltage supplying the diode rectifier constitutes
the highest voltage in the system. Thus,

Vrat ∃ 1.4 Η %2 Η 460 = 911 V

The average output voltage of the rectifier is (3/π) Η %2 Η 460 = 621 V, and it is
equal to the maximum available value of the peak fundamental line-to-line
voltage of the inverter. Therefore, the rms value of this voltage is 621/%2 =
439 V, and it is the rated voltage of the inverter. The rated current of the
inverter is 60000/(%3 Η 439) = 78.9 A. Consequently, the rated current, IS(rat), of
the inverter switches must satisfy the condition

29
 IS(rat) ∃ (%2/π) Η 1.25 Η 78.9 = 45 A

while the rated current, ID(rat), of the inverter diodes can be a half of that value,
that is,
ID(rat) ∃ 45/2 = 23 A

P7.20. The peak value of the voltage supplying the diode rectifier constitutes
the highest voltage in the system. Thus, disregarding the possibility of faulty
operation of the inverter,

Vrat ∃ 1.4 Η %2 Η 2400/2 = 2376 V

The average output voltage of the rectifier is (3/π) Η %2 Η 2400 = 3241 V and,
assuming the square-wave operation, the voltage gain of the inverter is 1.065.
Hence, the maximum available value of the peak fundamental line-to-line
voltage of the inverter is 1.065 Η 3241 = 3452 V, and the rms value of this
voltage is 3452/%2 = 2441 V, and it is the rated voltage of the inverter. The
rated current of the inverter is 200000/(%3 Η 2441) = 47.3 A. Consequently,
the rated current, IS(rat), of the inverter switches must satisfy the condition

IS(rat) ∃ (%2/π) Η 1.25 Η 47.3 = 27 A

while the rated current, ID(rat), of the inverter diodes can be a quarter of that
value, that is,
ID(rat) ∃ 27/4 = 7 A

30
 Chapter 8

P8.1. D = 32/(32 + 8) = 0.8

Vi = Vo/D = 12/0.8 = 15 V

fsw = 1/(32 + 8) = 0.025 MHz = 25 kHz

P8.2. Vo = 0.6 Η 50 = 30 V

Io = 30/10 = 3 A

It can reasonably be assumed that the minimum value of inductance L for

continuous conduction is to be determined for the worst case, that is, the duty
ratio, D, equal zero. Then, from Eq. (8.11),

31
 30 1-0
L= 2 _ 3
= 6.67_10-4 H
2_3 15_10

and from Eq. (8.9),

1 - 0.6
∆V o = 30 = 0.2 V
8_6.67_ 10 _50_ 10 -6 _(15_ 10 3 )2
-4

P8.3. D = 32/(32 + 8) = 0.8

Vi = Vo Η (1 - D) = 12 Η (1 - 0.8) = 2.4 V

fsw = 1/(32 + 8) = 0.025 MHz = 25 kHz

P8.4. D = 32/(32 + 8) = 0.8

Vi = Vo Η (1 - 1/D) = 12 Η (1 - 1/0.8) = -3 V

fsw = 1/(32 + 8) = 0.025 MHz = 25 kHz

P8.5. Assume fsw = 25 kHz as in the preceding Problems

Assume L = 1mH

D = 6/12 = 0.5

Assuming that the capacitor is selected for the worst case, that is, D = 0, it
can be specified from Eq. (8.9) as

1- 0
C≥ _100 = 2_ 10 -5 F = 20 µF
8_ 10 _(25_ 10 3 )2
-3

P8.6. Conversion of a 12-V voltage into a 6-V is, of course, impossible in a

boost converter. Here, the solution could end. However, for comparison with
the buck converter in the preceding Problem, let us assume the same duty
ratio, D, of 0.5. Then, Vo = 12/(1 - 0.5) = 24 V.

Assuming fsw = 25 kHz and R = 200 Ω, the required capacitance, for the worst

32
case of D = 1, can be found from Eq. (8.18) as

1
C≥ _100 = 2_ 10 -5 F = 20 µF
200_25_ 10 3

It can be seen that, in contrast with the buck converter, the ripple of the
output voltage depends on the load, here assumed as 1 kΩ.

P8.7. Assuming, as in the preceding example, fsw = 25 kHz and R = 200 Ω,

the required capacitance can be found from Eq. (8.18) as

1
C≥ _100 = 2_ 10 -5 F = 20 µF
200_25_ 10 3

P8.8. Vo = 1 Η 0.4 Η 15 = 6 V

For the worst case of D = 0, the required inductance can be found from Eq.
(8.11) as

6
L≥ (1 - 0) = 6_ 10 -5 = 60 µH
2_25_ 10 3 _2

Then, from Eq. (8.9),

1 - 0.4
∆V o = _6 = 0.075 V = 75 mV
8_6_ 10 _160_ 10 -6 _(25_ 10 3 )2
-5

P8.9. Io = 100/25 = 4 A

From Eq. (8.55),

Vo 100
D= = = 0.769
k N V i + V o 0.5_60 + 100

and from Eq. (8.56),

33
 2
(1 - 0.769 ) _100
f sw > = 20563 Hz ≈ 20.5 kHz
2_0. 5 2 _4_0.13_ 10 -3

Then, from Eq. (8.18),

0.769
∆V o = _100 = 0.75 V
25_0.2_ 10 -3 _20.5_ 10 3

P8.10. From Eq. (8.55),

0.35
Vo=1 100 = 53.8 V
1 - 0.35
53.8
Io= = 2.69 A
20

For the worst case of D = 0, the inductances can be determined from Eqs.
(8.57) and (8.50) as

(1 - 0 )2 _53.8
L1 = 2 = 6.67_ 10 -4 H
2_ 12 _2.69_30_ 10 3

and

(1 - 0 )2 _53.8
L2 = 2 3
= 6.67_ 10 -4 H
2_2.69_30_ 10

Then, from Eq. (8.9),

1 - 0.35
∆V o = _53.8 = 0.29 V
8_6.67_ 10 _25_ 10 -6 _(30_ 10 3 )2
-4

P8.11. From Eq. (8.58),

80
D= = 0.4
2_2_50

and from Eq. (8.59),

34
 1 - 2_0.4
∆V o = _80 = 0.63 V
32_0.1_ 10 _20_ 10 -6 _(20_ 10 3 )2
-3

P8.12. From Eq. (8.60), for D = Dmax= 0.5, Vi = 100(2 Η 0.5) = 100 V

P8.13. From Eq. (8.58), for D = Dmax= 0.5, Vi = 100(2 Η 2 Η 0.5) = 50 V

P8.17. From Eq. (8.63),

from Eq. (8.62),

0.1_ 10 -3
Zo= = 20 Ω
0.25_ 10 -6

from Eq. (8.65),

35000
kf= = 1.1
31831

and from Eq. (8.66),

15
r= = 0.75
20

P8.18. From Eq. (8.63),

1
f o= = 59314 Hz
2π 60_ 10 -6
_120_ 10 -9

from Eq. (8.62),

60_ 10 -6
Zo= = 22.36 Ω
120_ 10 - 9

from Eq. (8.65),

35
 24000
kf= = 0.405
59314

from Eq. (8.66),

20
r= = 0.8944
22.36

and from Eq. (8.75),

1
Kv= = 0.1657
2 1+ [ π 2
(0.405 -
1
) ]2
8_0.8944 0.405

Thus, Vo = 0.1657 Η 120 = 19.88 V and Io = 19.88/20 = 0.994 A

P8.19. R = 502/500 = 5 Ω

From Eq. (8.63),

1
f o= = 25291 Hz
2π 12_ 10 -6 _3.3_ 10 -6

from Eq. (8.62),

12_ 10 -6
Zo= = 1.91 Ω
3.3_ 10 -6

from Eq. (8.65),

30000
kf= = 1.186
25291

from Eq. (8.66),

5
r= = 2.62
1.91

and from Eq. (8.81),

36
 1
Kv= = 0.74
π 2
28_1.186 2
(1 - 1. 186 ) + ( 2 2
)
4 π _2.62

Thus, Vi = 50/0.74 = 67.6 V.

P8.20. For r = 4, Eq. (8.82) gives

1
Kv=
π 2
(2 - k 2f )2 + [
2
(kf -
1
) ]2
4 π 2
kf

and, as seen in Fig. 8.35, the maximum voltage gain, Kv, occurs with the
frequency ratio, kf, somewhere between 1.25 and 1.5. A computer can be
used to search this range. It turns out that Kv reaches the maximum value of
2.844 when kf = 1.41. Since

18000
kf= = 36 π L r
1
2π 10 -6 L r

then Lr = (1.41/36π)2 = 1.55 Η 10-4 H = 0.155 mH.

37