Chapter 3

Modulation Techniques with

Emphasis on Analog Systems

3.1 Problems
Problem 3.1
The demodulated output, in general, is

yD (t) = Lpfxc (t) 2 cos[! c t + (t)]g

where Lp f g denotes the lowpass portion of the argument. With

xc (t) = Ac m (t) cos [! c t + 0]

the demodulated output becomes

yD (t) = Lp f2Ac m (t) cos [! c t + 0 ] cos [! c t + (t)]g

Performing the indicated multiplication and taking the lowpass portion yields

yD (t) = Ac m (t) cos [ (t) 0]

If (t) = 0 (a constant), the demodulated output becomes

yD (t) = Ac m (t) cos [ 0 0]

Letting Ac = 1 gives the error

" (t) = m (t) [1 cos ( 0 0 )]

1
2CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

The mean-square error is

D E
2
"2 (t) = m2 (t) [1 cos ( 0 0 )]

where h i denotes the time-average value. Since the term [1 cos ( 0 0 )] is a constant,
we have
2
"2 (t) = m2 (t) [1 cos ( 0 0 )]

Note that for 0 = 0 , the demodulation carrier is phase coherent with the original modu-
lation carrier, and the error is zero. For (t) = ! 0 t we have the demodulated output

yD (t) = Ac m (t) cos (! 0 t 0)

Letting Ac = 1, for convenience, gives the error

" (t) = m (t) [1 cos (! 0 t 0 )]

giving the mean-square error

D E
2
"2 (t) = m2 (t) [1 cos (! 0 t 0 )]

In many cases, the average of a product is the product of the averages. (We will say more
about this in Chapters 5). For this case
D E
2
"2 (t) = m2 (t) [1 cos (! 0 t 0 )]

Note that 1 cos (! 0 t 0) is periodic. Taking the average over an integer number of
periods yields
D E
2
[1 cos (! 0 t 0 )] = 1 2 cos (! 0 t 0) + cos2 (! 0 t 0)
1 3
= 1+ =
2 2
Thus
3
"2 (t) = m2 (t)
2

Problem 3.2
A full-wave recti…er takes the form shown in Figure 3.1. The waveforms are shown in
Figure 3.2, with the half-wave recti…er on top and the full-wave recti…er on the bottom.
The message signal is the envelopes. Decreasing exponentials can be drawn from the peaks
of the waveform as depicted in Figure 3.3(b) in the text. It is clear that the full-wave
3.1. PROBLEMS 3

()Cx t

()Dy t

R C

Figure 3.1: Full wave recti…er.

1.5

0.5

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

1.5

0.5

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Figure 3.2: Output waveforms for a half-wave recti…er (top) and a full-wave recti…er (bot-
tom).
4CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

recti…ed xc (t) de…nes the message better than the half-wave recti…ed xc (t) since the carrier
frequency is e¤ectively doubled.

Problem 3.3
Multiplying the AM signal

xc (t) = Ac [1 + amn (t)] cos ! c t

by xc (t) = Ac [1 + amn (t)] cos ! c t and lowpass …ltering to remove the double frequency
(2! c ) term yields
yD (t) = Ac [1 + amn (t)] cos (t)
For negligible demodulation phase error, (t) 0, this becomes

yD (t) = Ac + Ac amn (t)

The dc component can be removed resulting in Ac amn (t), which is a signal proportional
to the message, m (t). This process is not generally used in AM since the reason for using
AM is to avoid the necessity for coherent demodulation.

Problem 3.4
By evaluating
Z T
1
m2n (t) = m2n (t)dt
T 0

we see that m2n (t) is equal to the values given in the following table. (Note: See the …rst
part of Problem 3.5 for the Part a. Part b has the same value of m2n (t) as Part a. For
Part c it is obvious that m2n (t) = 1. The e¢ ciency is de…ned as

a2 m2n (t)
Ef f =
1 + a2 hm2n (t)i

which gives the values given in the following table.

Part m2n (t) a = 0:2 a = 0:4 a = 0:7 a=1

a 1=3 Ef f = 1:3158% Ef f = 5:0633% Ef f = 14:0401% Ef f = 25%
b 1=3 Ef f = 1:3158% Ef f = 5:0633% Ef f = 14:0401% Ef f = 25%
c 1 Ef f = 3:8462% Ef f = 13:7931 Ef f = 32:8859% Ef f = 50%

Problem 3.5
By inspection, the normalized message signal is as shown in Figure 3.3.
3.1. PROBLEMS 5

()nm t

0 t
T
-1

Figure 3.3: Normalized message signal for Problem 3.5.

Thus
2 T
mn (t) = t; 0 t
T 2
and Z T =2 2 2 3
2 2 2 2 1 T 1
m2n (t) = t dt = =
T 0 T T T 3 2 3
Also

Ac [1 + a] = 40
Ac [1 a] = 10

This yields
1+a 40
= =4
1 a 10
or

1+a = 4 4a
5a = 3

Thus
a = 0:6
Since the index is 0:6, we can write

Ac [1 + 0:6] = 40

This gives
40
Ac = = 25
1:6
6CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

This carrier power is

1 1
Pc = A2c = (25)2 = 312:5 Watts
2 2
The e¢ ciency is
(0:6)2 1
3 0:36
Ef f = = = 0:107 = 10:7%
1+ (0:6)2 13 3:36
Thus
Psb
= 0:107
Pc + Psb
where Psb represents the power in the sidebands and Pc represents the power in the carrier.
The above expression can be written

Psb = 0:107 + 0:107Psb

This gives
0:107
Psb = Pc = 97:48 Watts
1:0 0:107

Problem 3.6
For the …rst signal

T
= ; mn (t) = m (t) and Ef f = 81:25%
6
and for the second signal

5T 1
= ; mn (t) = m (t) and Ef f = 14:77%
6 5

Problem 3.7
(a) The using a root …nding algorithm, or by using a simple MATLAB program, we …nd
that the minimum value of m (t) is 12:8985. Thus the normalized message signal is

1
mn (t) = [9 cos(20 t) 8 cos(60 t)]
12:8985
With the given value of c (t) and the index a, we have

xc (t) = 100 [1 + 0:5mn (t)] cos 200 t

3.1. PROBLEMS 7

or
1
xc (t) = 100 1 + [9 cos 20 t 8 cos 60 t] cos 200 t
2(12:8985)
This yields

xc (t) = 15:5057 cos 140 t + 17:4439 cos 180 t

+100 cos 200 t
+17:4439 cos 220 t 15:5057 cos 260 t

(b) The value of m2n (t) is

1 2
1 h i
m2n (t) = (9)2 + (8)2 = 0:4358
12:8985 2

(c) This gives the e¢ ciency

(0:5)2 (0:4358)
E= = 0:0982 or 9:82%
1 + (0:5)2 (0:4358)

(d) The two-sided amplitude spectrum is shown in Figure 3.4.The phase spectrum is every-
where zero except for the components at f = 70 and f = 130 have phase . The sign
is arbitrary but the phase spectrum must be drawn so that it is odd.

Problem 3.8
1
(a) mn (t) = 17 [9 cos(20 ht) + 8 cos(60i t)]
1 2 1
(b) m2n (t) = 17 2 (9)2 + (8)2 = 0:2509
0:25(0:2509)
(c) Ef f = 1+0:25(0:2509) = 0:0590 = 5:9%
(d) The expression for xc (t) is

1 1
xc (t) = 100 1 + (9 cos 20 t + 8 cos 60 t) cos 200 t
2 17
= 11:7647 cos 140 t + 13:2353 cos 180 t
+100 cos 200 t
+13:2353 cos 220 t + 11:7647 cos 260 t

The amplitude spectrum is drawn as in the previous problem. The phase spectrum is
everywhere zero.

Problem 3.9
8CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

50

45

40

35

30
Amplitude

25

20

15

10

0
-150 -100 -50 0 50 100 150
Frequency

Figure 3.4: Amplitude spectrum for Problem 3.7.

3.1. PROBLEMS 9

The modulator output

xc (t) = A cos[2 (200) t] + B cos[2 (180) t] + B cos[2 (220) t]

can be written
2B
xc (t) = A 1 + cos[2 (20)t cos[2 (200) t]
A
The sideband power is
1 1
Psb = B 2 + B 2 = B 2 Watts
2 2
The carrier power is
A2
P0 =
2
Thus, the e¢ ciency is

Psb B2 2B 2
Ef f = = =
P0 + Psb P0 + B 2 A2 + 2B 2
Since P0 = 100 = A2 =2 p
A= 200 = 14:1421
The e¢ ciency is
B2
Ef f = = 0:4
100 + B 2
Thus
B 2 = 40 + 0:4B 2
So that r
40
B= = 8:1650
0:6
Finally, the modulation index is
2B 2(8:1650)
a= = = 1:1547
A 14:1421

Problem 3.10
The modulator output

xc (t) = 30 cos[2 (200) t] + 4 cos[2 (180) t] + 4 cos[2 (220) t]

can be written
xc (t) = [30 + 8 cos 2 (20) t] cos 2 (200) t
10CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

or
8
xc (t) = 30 1 + cos 2 (20) t cos 2 (200) t
30
By inspection, the modulation index is
8
a= = 0:2667
30
Since the component at 200 Hertz represents the carrier, the carrier power is
1
Pc = (30)2 = 450 Watts
2
The components at 180 and 220 Hertz are sideband terms. Thus the sideband power is
1 1
Psb = (4)2 + (4)2 = 16 Watts
2 2
Thus, the e¢ ciency is
Psb 16
Ef f = = = 0:0343 = 3:43%
Pc + Psb 450 + 16

Problem 3.11
(a) By plotting m (t) or by using a root-…nding algorithm we see that the minimum value
of m (t) is M = 3:432. Thus

mn (t) = 0:5828 cos (2 fm t) + 0:2914 cos (4 fm t) + 0:5828 cos (10 fm t)

The AM signal is

xc (t) = Ac [1 + 0:7mn (t)] cos(2 fc t)

= 0:2040Ac cos[2 (fc 5fm ) t]
+0:1020Ac cos[2 (fc 2fm ) t]
+0:2040Ac cos[2 (fc fm ) t]
+Ac cos(2 fc t)
+0:2040Ac cos[2 (fc + fm ) t]
+0:1020Ac cos[2 (fc + 2fm ) t]
+0:2040Ac cos[2 (fc + 5fm ) t]

The spectrum is drawn from the expression for xc (t). It contains 14 discrete components
as shown
3.1. PROBLEMS 11

Comp Freq Amp Comp Freq Amp

1 fc 5fm 0:102Ac 8 fc 5fm 0:102Ac
2 fc 2fm 0:051Ac 9 fc 2fm 0:051Ac
3 fc fm 0:102Ac 10 fc fm 0:102Ac
4 fc 0:5Ac 11 fc 0:5Ac
5 fc + fm 0:102Ac 12 fc + fm 0:102Ac
6 fc + 2fm 0:051Ac 13 fc + 2fm 0:051Ac
7 fc + 5fm 0:102Ac 14 fc + 5fm 0:102Ac

(b) The e¢ ciency is 15.8%.

Problem 3.12
The modulator output

xc (t) = 25 cos(2 (150) t) + 5 cos(2 (160) t) + 5 cos(2 (140) t)

is
10
xc (t) = 25 1 + cos(2 (10) t) cos(2 (150) t)
25
Thus, the modulation index, a, is
10
a= = 0:4
25
The carrier power is
1
Pc = (25)2 = 312:5 Watts
2
and the sideband power is

1 1
Psb = (5)2 + (5)2 = 25 Watts
2 2
Thus, the e¢ ciency is
25
Ef f = = 0:0741
312:5 + 25
Problem 3.13

(a) From Figure 3.72

x (t) = m (t) + cos ! c t
With the given relationship between x (t) and y (t) we can write

y (t) = 4 fm (t) + cos ! c tg + 2 fm (t) + cos ! c tg2

12CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

Filter

}m t
(){
ℑ
{}
()2m t
ℑ Characteristic

fc − W fc fc + W 2 fc f
0 W 2W

Figure 3.5: Spectrum of sqrare-law device output.

which can be written

y (t) = 4m (t) + 4 cos ! c t + 2m2 (t) + 4m (t) cos ! c t + 2 + 2 cos 2! c t

The equation for y (t) is more conveniently expressed

y (t) = 2 + 4m (t) + 2m2 (t) + 4 [1 + m (t)] cos ! c t + 2 cos 2! c t

(b) The spectrum illustrating the terms involved in y (t) is shown in Figure 3.5. The center
frequency of the …lter is fc and the bandwidth must be greater than or equal to 2W . In
addition, fc W > 2W or fc > 3W , and fc + W < 2fc . The last inequality states that
fc > W , which is redundant since we know that fc > 3W .
(c) From the de…nition of m (t) we have

m (t) = M mn (t)

so that
g (t) = 4 [1 + 5M mn (t)] cos(! c t)
It follows that
a = 0:1 = 5M
Thus
0:1
M=
= 0:02
5
(d) This method of forming a DSB signal avoids the need for a multiplier.

Problem 3.14
3.1. PROBLEMS 13

The …lter for USB SSB will be de…ned by

1
HU (f ) = 1 sgn (f + fc )
2
which gives as the output
1 1
xU SB (t) = Ac m (t) cos(2 fc t) b (t) sin(2 fc t)
Ac m
2 2

Problem 3.15
With
m(t) = 2 cos(2 fm t) + cos(4 fm t)
the Hilbert transform of the message signal is

b
m(t) = 2 sin(2 fm t) + sin(4 fm t)

Thus
1
xc (t) = Ac [2 cos(2 fm t) + cos(4 fm t)] cos(2 fc t)
2
1
Ac [2 sin(2 fm t) + sin(4 fm t)] sin(2 fc t)
2
Performing the multiplication and letting Ac = 4 gives

xc (t) = f2 cos[2 (fc + fm )t] + cos[2 (fc + 4fm )t] + 2 cos[2 (fc fm )t] + cos[2 (fc 4fm )t]g
f 2 cos[2 (fc + fm )t] cos[2 (fc + 4fm )t] + 2 cos[2 (fc fm )t] + cos[2 (fc 4fm )t]g

Note that with the + sign, the (fc + fm ) and the (fc + 2fm ) cancel giving LSB SSB. With
the sign the (fc fm ) and (fc 2fm ) terms cancel giving USB SSB.

Problem 3.16
We assume that the VSB waveform is given by
1
xc (t) = A" cos[2 (fc f1 ) t] +
2
1
A (1 ") cos[2 (fc + f1 ) t]
2
1
+ B cos[2 (fc + f2 ) t]
2
We let y (t) be xc (t) plus a carrier. Thus

y (t) = xc (t) + K cos(2 fc t)

14CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

It can be shown that y (t) can be written

y (t) = y1 (t) cos(2 fc t) + y2 (t) sin(2 fc t)
where
A B
y1 (t) = cos(2 f1 t) + cos(2 f2 t) + K
2 2
A B
y2 (t) = A" + sin(2 f1 t) sin(2 f2 t)
2 2
In the preceding expression, the two terms involving A and B represent the message signal.
Also
y (t) = R (t) cos (2 fc t + )
where R (t) is the envelope and is therefore the output of an envelope detector. It follows
that q
R (t) = y12 (t) + y22 (t)
For K is su¢ ciently large so that y12 (t) y22 (t), R (t) = jy1 (t)j, which is 21 m (t) + K, where
K is a dc bias. Thus if the detector is ac coupled, K is removed and the output y (t) is
m (t) scaled by 21 .

Problem 3.17
With an IF frequency of fIF = 455 kHz we have for high-side tuning
fLO = fi + fIF = 1120 + 455 = 1575 kHz
fIM AGE = fi + 2fIF = 1120 + 910 = 2030 kHz
and for low-side tuning we have
fLO = fi fIF = 1120 455 = 665 kHz
fIM AGE = fi 2fIF = 1120 910 = 210 kHz
With an IF frequency of fIF = 2500 kHz we have, for high-side tuning
fLO = fi + fIF = 1120 + 2500 = 3620 kHz
fIM AGE = fi + 2fIF = 1120 + 5000 = 6120 kHz
and for low-side tuning
fLO = fi fIF = 1120 2500 = 1380 kHz
fLO = 1380 kHz
fIM AGE = fi 2fIF = 1120 5000 = 3880 kHz
fLO = 3880 kHz
3.1. PROBLEMS 15

In the preceding development for low-side tuning, recall that the spectra are symmetrical
about f = 0.

Problem 3.18
Since high-side tuning is used, the local oscillator frequency is

fLO = fi + fIF

where fi , the carrier frequency of the input signal, varies between 5 and 25 M Hz. The
ratio is
fIF + 25
R=
fIF + 5
where fIF is the IF frequency expressed in M Hz. We make the following table

fIF ;MHz R
0:4 4:70
0:5 4:63
0:7 4:51
1:0 4:33
1:5 4:08
2:0 3:86

A plot of R as a function of fIF provides the required plot.

Problem 3.19
The required plot follows.

Problem 3.20
By de…nition

xc (t) = Ac cos [! c t + kp m (t)] = Ac cos [! c t + kp u(t t0 )]

The waveforms for the four values of kp are shown in Figure 3.7. The top-left pane is for
kp = , the top-right pane is for kp = =4, the bottom-left pane is for; kp = , and the
bottom-right pane is for kp = =4. The plots illustrate 2 s of data with a frequency 2 Hz.

Problem 3.21
Since sin(x) = cos(x 2) we can write

xc3 (t) = Ac sin(! c t + sin ! m t) = Ac cos ! c t + sin ! m t

2
16CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

Desired
Signal

ω
ω
1

Local
Oscillator
ω
ω
−
1 ω
2

Signal at
Mixer
Output
ω
ω
−
IF
2 ω ω
−
2
1 ω
2

Image
Signal

ω
ω
−
2
1 ω
2

Image Signal
at Mixer Output
ω
ω
−
3
2 2ω
1 ω
2

Figure 3.6: Plot for Problem 3.17.

3.1. PROBLEMS 17

1 1

0.5 0.5
k p=pi/4
k p=pi

0 0

-0.5 -0.5

-1 -1
0 0.5 1 1.5 2 0 0.5 1 1.5 2

1 1

0.5 0.5
k p=-pi/4
k p=-pi

0 0

-0.5 -0.5

-1 -1
0 0.5 1 1.5 2 0 0.5 1 1.5 2

Figure 3.7: Required plots for Problem 3.20.

18CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

which is n o
xc3 (t) = Ac Re ej(!c t =2) j sin ! m t
e
Since
1
X
j sin ! m t
e = Jn ( )ejn!m t
n= 1

we have ( )
1
X
xc3 (t) = Ac Re ej(!c t =2)
Jn ( )ejn!m t
n= 1

Thus
1
X n o
xc3 (t) = Ac Jn ( ) Re ej(!c t+n!m t =2)

n= 1

Thus
1
X h i
xc3 (t) = Ac Jn ( ) cos (! c + n! m ) t
n= 1
2

Note that the amplitude spectrum of xc3 (t) is identical to the amplitude spectrum for both
xc1 (t) and xc2 (t). The phase spectrum of xc3 (t) is formed from the spectrum of xc1 (t) by
adding =2 to each term.
For xc4 (t) we write

xc4 (t) = Ac sin(! c t + cos ! m t) = Ac cos ! c t + cos ! m t

2
Using the result of the preceding problem we write
( 1
)
X
j(! c t =2) j (n! m t+ n2 )
xc4 (t) = Ac Re e Jn ( ) e
n= 1

This gives
1
X n o
+ n2 )
xc4 (t) = Ac Jn ( ) Re ej(!c t+n!m t 2

n= 1

Thus
1
X h i
xc4 (t) = Ac Jn ( ) cos (! c + n! m ) t + (n 1)
n= 1
2

Compared to xc1 (t), xc2 (t), x3 (t) and xc4 (t), we see that the only di¤erence is in the phase
spectrum.

Problem 3.22
3.1. PROBLEMS 19

Let (t) = cos ! m t where ! m = 2 fm . This gives

n o
xc2 (t) = Ac Re ej!c t ej cos ! m t

Expanding a Fourier series gives

1
X
ej cos ! m t
= Cn ejn!m t
n= 1

where Z =! m
!m
Cn = ej cos ! m t
e jn! m t
dt
2 =! m

With x = ! m t, the Fourier coe¢ cients become

Z
1
Cn = ej cos x
e jnx
dx
2

Since cos x = sin x + 2 Z

1
Cn = ej [ sin(x+ 2 ) nx]
dx
2
With x = x + 2 , the preceding becomes
Z 3 =2
1
Cn = ej [ sin y ny+n 2 ]
dy
2 =2

This gives Z
j n2 1
Cn = e ej[ sin y ny]
dy
2
where the limits have been adjusted by recognizing that the integrand is periodic with
period 2 . Thus
n
Cn = ej 2 Jn ( )
and ( )
1
X
j! c t j (n! m t+ n2 )
xc2 (t) = Ac Re e Jn ( ) e
n= 1

Taking the real part yields

1
X h n i
xc2 (t) = Ac Jn ( ) cos (! c + n! m ) t +
n= 1
2
20CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

The amplitude spectrum is therefore the same as in the preceding problem. The phase
spectrum is the same as in the preceding problem except that n2 is added to each term.

Problem 3.23
From the problem statement
xc (t) = Ac cos [2 (50) t + 10 sin (2 (5) t)]
Since fc = 50 and fm = 5 (it is important to note that fc is an integer multiple of fm )
there is considerable overlap of the portion of the spectrum centered about f = 50 with
the portion of the spectrum centered about f = 50 for large (such as 10). This can also
be seen in the time domain. We can see from the time domain plot (top pane) of Figure ??
that the waveform has a non-zero dc value. This is due to the fact that the spectra centered
about f = +50 and f = 50 overlap and the term at f = 0 (dc) is non-zero. Where terms
overlap, they must be summed.
The time average value appears from the spectrum to be approximately 0.63Ac .

Problem 3.24
The amplitude and phase spectra follow directly from the table of Fourier-Bessel coe¢ cients,
or one can write a simple MATLAB program as was done here.. The single-sided magnitude
and phase spectra are shown in Figure ??. The magnitude spectrum is plotted assuming
Ac = 1.

Problem 3.25
We are given J0 (3) = 0:2601 and J1 (3) = 0:3391. Using
2n
Jn+1 ( ) = Jn ( ) Jn 1( )

with = 3 we have
2
Jn+1 (3) = nJn (3) Jn 1 (3)
3
With n = 1,
2
J2 (3) = J1 (3) J0 (3)
3
2
= (0:3391) + 0:2601 = 0:4862
3
With n = 2,
4
J3 (3) = J2 (3) J1 (3)
3
4
= (0:04862) 0:3391 = 0:3092
3
3.1. PROBLEMS 21

With n = 3 we have

J4 (3) = 2J3 (3) J2 (3)

= 2 (0:3091) 0:4862 = 0:1322

Letting n = 4 we can determine J5 (3) as required

8
J5 (3) = J4 (3) J3 (3)
3
8
= (0:1322) 0:3092 = 0:0433
3

Problem 3.26
The modulated signal can be written
h i
xc (t) = Ref 6 + 6ej2 (30)t
+ 6e j2 (30)t
ej2 (100)t
g

We will concentrate on the term in brackets, which is the complex envelope described in
Chapter 2. Denoting the complex envelope by xec (t), we can write

ec (t) = [6 + 6 cos 2 (30t) + 6 cos 2 (30) t]

x
+j [6 sin 2 (30t) 6 sin 2 (30) t]
= 6 + 12 cos 2 (30t)

It follows from the de…nition of xc (t) that

ec (t) = R (t) ej
x (t)

Obviously
R (t) = 6 + 12 cos 2 (30t)
and
(t) = 0

Problem 3.27
For all three message signals illustrated in Figure 3.73, the frequency deviation in Hz is
simply the given message signal with the ordinate values multiplied by 10.
For the …rst message signal (the top signal given in Figure 3.73) the phase deviation in
radians is given Z Z
t t
(t) = 2 fd m ( ) d = 20 m( )d
22CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

For 0 t 1, we have
Z t
(t) = 20 2 d = 20 t2
0
For 1 t 2
Z t
(t) = (1) + 20 (5 ) d = 20 + 100 (t 1) 10 t2 1
1
= 20 + 100 t 100 10 t2 + 10 = 70 + 100 t 10 t2

For 2 t 3
Z t
(t) = (2) + 20 3d = 90 + 60 (t 2) = 60 t 30
2

For 3 t 4
Z t
(t) = (3) + 20 2d = 150 + 40 (t 3) = 30 + 40 t
3

Finally, for t > 4 we recognize that (t) = (4) = 190 . The required …gure results by
plotting these curves.

The phase deviation for the second message signal (bottom right) is given again by
Z t Z t
(t) = 2 fd m ( ) d = 20 m( )d

For 0 t 1, we have
Z t
(t) = 20 d = 10 t2
0
For 1 t 2
Z t
(t) = (1) + 20 ( 2) d = 10 + 10 t2 1 40 (t 1)
1
= 10 t2 4t + 4 = 10 (t 2)2

For 2 t 4
Z t
(t) = (2) + 20 (6 2 )d = 0 + 20 (6) (t 2) 20 t2 4
2
= 20 (t2 6t + 8)
3.1. PROBLEMS 23

Finally, for t > 4 we recognize that (t) = (4) = 0 (Note that the area under the curve
from 0 to 4 is obviously zero). The required …gure follows by plotting these expressions.

Finally, the phase deviation in radians for the third message signal is again given by
Z t Z t
(t) = 2 fd m ( ) d = 20 m( )d

For 0 t 1, we have Z t
(t) = 20 ( 2 )d = 20 t2
0
For 1 t 2
Z t
(t) = (1) + 20 2d = 20 + 40 (t 1) = 60 + 40 t
1

For 2 t 2:5
Z t
(t) = (2) + 20 (10 4 )d = 20 + 20 (10) (t 2) 20 (2) t2 4
2
= 20 ( 2t2 + 10t + 7)

For 2:5 t 3
Z t
(t) = (2:5) 20 2d = 15 40 (t 2:5)
2:5
= ( 250 + 500 + 140 ) 40 (t + 2:5)
= 390 40 t 100 = 290 40 t

For 3 t 4
Z t
(t) = (3) + 20 (2 8) d = 170 + 20 (t2 9) 20 (8)(t 3)
3
= 20 (t2 8t + 23:5)

Finally, for t > 4 we recognize that (t) = (4) = 20 (16 32 + 23:5) = 150 . The required
…gure follows by plotting these expressions.

Problem 3.28
(a) The phase deviation is
Z t
1
(t) = 2 (20) 4 (x 4) dx
8
24CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

which is, for 0 t 8, Z t

(t) = 40 4dx = 160 t
0
Thus

(t) = 0, t<0
(t) = 160 t; 0 t 8
(t) = 160 (8); t>8

(b) The frequency deviation is, by de…nition

1 d 1
= 80 (t 4) = 80 Hz
2 dt 8
for 0 t 8 and is zero for t < 0 and t > 8.
(c) The peak frequency deviation is 80 Hz.
(d) The peak phase deviation is 160 (8) = 1280 radians.
(e) The transmitted power is

(100)2
P = = 5000 W
2

Problem 3.29
(a) The phase deviation is

(t) = 0, t<3
Z t
4 160 t2 9
(t) = 40 (x 3)dx = 3t + 9
3 3 3 2 2
80
= t2
6t + 9 ; 3 t 6
3
Z t Z t
4 4
(t) = 240 + 40 4 (x 6) dx = 240 + 40 12 x dx
6 3 6 3
41 2
= 240 + 40 12(t 6) (t 36)
32
2 2
= 240 + 40 t + 12t 48
3
80
= 240 + t2 + 18t 72 ; 6 t 9
3
(t) = 480 t>9
3.1. PROBLEMS 25

(b) The frequency deviation in Hz is

1
fd m(t) = 80 (t 6)
3
(c) The peak frequency deviation is 4 (20) = 80 Hz.
(d) The peak phase deviation is, since the area of [(t 6)=3] = 3 is 2 fd (3) (4) = 24 (20) =
480
(e) The modulator output power is
1 1
P = A2c = (100)2 = 5000 Watts
2 2

Problem 3.30
(a) Since the carrier frequency is 1000 Hertz, the general form of xc (t) is

xc (t) = Ac cos [2 (1000) t + (t)]

The phase deviation, (t), is therefore given by

(t) = 40t2 rad

The frequency deviation is

d
= 80t rad/sec
dt
or
1 d 40
= t Hz
2 dt
(b) The phase deviation is

(t) = 2 (500) t2 2 (1000) t rad

(Note that we substracted the phase of the unmodulated carrier from the instantaneous
carrier.) The frequency deviation is
d
= 4 (500) t 2 (1000) = 2000 (t 1) rad/sec
dt
or
1 d
= 1000 (t 1) Hz
2 dt
(c) The phase deviation is

(t) = 2 (1200) t 2 (1000)t = 200 t rad

26CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

and the frequency deviation is

d
= 200 rad/sec
dt
or
1 d
= 100 Hz
2 dt
which should be obvious from the expression for xc (t).
(d) The phase deviation is
p p
(t) = 2 (900) + 10 t 2 (1000) = 2 (100)t + 10 t rad

and the frequency deviation is

d 1 1 5
= 2 (100) + (10) t 2 = 2 (100) + p rad/sec
dt 2 t
or
1 d 5
= 100 + p Hz
2 dt 2 t
Problem 3.31

(a) The peak deviation is (14)(5) = 70 and fm = 10. Thus, the modulation index is 70
10 = 7.
(b) The magnitude spectrum, scaled to Ac , is the Fourier-Bessel spectrum of Figure 3.24
with = 7:
(c) Since is not 1, this is not narrowband FM. The bandwidth exceeds 2fm .
(d) For phase modulation, kp (5) = 7 or kp = 1:4.

Problem 3.32
The results are given in the following table. The deviation ratio is found by de…nition and
the bandwidth follows from Carson’s rule.

Part fd D = 6fd =W = fd =2000 B = 2 (D + 1) W

a 20 0.01 24.24 kHz
b 200 0.1 26.4 kHz
c 2000 1 48 kHz
d 20000 10 264 kHz

Problem 3.33
Since Z Z
1 j(nx sin x) 1
Jn ( ) = e dx = ej( sin x nx)
dx
2 2
3.1. PROBLEMS 27

we can write
Z Z
1 1
Jn ( ) = cos ( sin x nx) dx + j sin ( sin x nx) dx
2 2
The imaginary part of Jn ( ) is zero, since the integrand is an odd function of x and the
limits ( ; ) are even. Thus
Z
1
Jn ( ) = cos ( sin x nx) dx
2
Since the integrand is even
Z
1
Jn ( ) = cos ( sin x nx) dx
0

which is the …rst required result. With the change of variables = x, we have
Z
1
Jn ( ) = cos [ sin ( ) n( )] ( 1) d
0
Z
1
= cos [ sin ( ) n + n ]d
0

Since sin ( ) = sin , we can write

Z
1
Jn ( ) = cos [ sin + n n ]d
0

Using the identity

cos (u ) = cos u cos + sin u sin
with
u= sin + n
and
=n
yields
Z
1
Jn ( ) = cos [ sin + n ] cos (n ) d
0
Z
1
+ sin [ sin + n ] sin (n ) d
0

Since sin (n ) = 0 for all n, the second integral in the preceding expression is zero. Also

cos (n ) = ( 1)n
28CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

Thus Z
n 1
Jn ( ) = ( 1) cos [ sin + n ] d
0
However Z
1
J n( ) = cos [ sin + n ] d
0
Thus
Jn ( ) = ( 1)n J n( )
or equivalently
J n( ) = ( 1)n Jn ( )

Problem 3.34
From
1
X
xc (t) = Ac Jn ( ) cos[(! c + ! m ) t]
n= 1

we obtain
1
X
1
x2c (t) = A2c J2 ( )
2 n= 1 n
Also
x2c (t) = A2c cos2 [! c t + (t)]
which, assuming that ! c 1 so that xc (t) has no dc component, is
1
x2c (t) = A2c
2
This gives
1
1 2 1 2 X 2
A = Ac J ( )
2 c 2 n= 1 n
from which
1
X
Jn2 ( ) = 1
n= 1

Problem 3.35

(a) Peak frequency deviation = 10(8) = 80 Hz.

(b) Since Z
160
(t) = 2 (8) 10 cos(20 t)dt = sin(20 t) = 8 sin(20 t)
20
3.1. PROBLEMS 29

the peak phase deviation is 8 rad.

(c) = 8
(d) The power at the …lter input is

A2c (10)2
Pin = = = 50 W
2 2
The power at the …lter output is determined by the number of spectral components passed
by the …lter. The bandwith of the …lter is 70 Hz, which is 35 Hz each sidfe of the carrier.
Since fm = 10 the …lter passes 3 terms each side of the carrier. Thus the power at the …lter
output is
3
" 3
#
A2c X 2 A2c X
Pout = Jn (8) = J02 (8) + 2 Jn2 (8)
2 2
n= 3 n=1
= 50 (0:172) + 2(0:235) + 2(0:113) + 2(0:291)2
2 2

= 16:74 W

(e) The input and output spectra are determined from the corresponding Fourier-Bessel
spectra.

Problem 3.36
We wish to …nd k such that
k
X
Pr = J02 (10) + 2 J02 (10) 0:80
n=1

This gives k = 9, yielding a power ratio of Pr = 0:8747. The bandwidth is therefore

B = 2kfm = 2 (9) (150) = 2700 Hz

For Pr 0:9, we have k = 10 for a power ratio of 0:9603. This gives

B = 2kfm = 2 (10) (150) = 3000 Hz

Problem 3.37
From the given data, we have

fc1 = 110 kHz fd1 = 0:05 fd2 = n (0:05) = 20

This gives
20
n= = 400
0:05
30CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

and
fc1 = n (100) kHz = 44 MHz
The two permissible local oscillator frequencies are

f`0:1 = 100 44 = 56 MHz

f`0:2 = 100 + 44 = 144 MHz

The center frequency of the bandpass …lter must be fc = 100 MHz and the bandwidth is

B = 2 (D + 1) W = 2 (20 + 1) (10) 103

or
B = 420 kHz

Problem 3.38
For the circuit shown
E (f ) R
H (f ) = = 1
X (f ) R + j2 f L + j2 f C
or
1
H (f ) =
1
1+j 2 f L 2 f C

where
L 10 3 6
L = = = 10 ;
R 103
C = RC = 103 10 9
= 10 6

A plot of the amplitude response shows that the linear region extends from approximately
54 kHz to118 kHz. Thus an appropriate carrier frequency is
118 + 54
fc = = 86 kHz
2
The slope of the operating characteristic at the operating point is measured from the am-
plitude response. The result is
KD = 8 10 6

Problem 3.39
3.1. PROBLEMS 31

We can solve this problem by determining the peak of the amplitude response characteristic.
This peak falls at
1
fp = p
2 LC
It is clear that fp > 100 MHz. Let fp = 150 MHz and let C = 0:001 10 12 . This gives
1 3
L= 2 = 1:126 10
(2 ) fp2 C

We …nd the value of R by trial and error using plots of the amplitude response. An
appropriate value for R is found to be 1 M . With these values, the discriminator constant
is approximately
KD 8:5 10 9

Problem 3.40
For Ai = Ac we can write, from (3.184),

xr (t) = Ac cos(2 fc t) + Ai cos[2 (fc + fi ) t] = Ac fcos(2 fc t) + cos[2 (fc + fi ) t]g

which is
xr (t) = Ac f[1 + cos(2 fi t)] cos(2 fc t) sin(2 fi t) sin(2 fc t)g
This yields
xr (t) = R (t) cos [2 fc t + (t)]
where
1 sin(2 fi t)
(t) = tan
1 + cos(2 fi t)
1 2 fi t 2 fi t
= tan tan =
2 2
This gives
1 d 2 fi t 1
yD (t) = = fi
2 dt 2 2
For Ai = Ac , we have a similar situation to the preceding and
1
yD (t) = fi
2
Finally, for Ai Ac we see from the phasor diagram that

(t) (t) = ! i t
32CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

and
KD d
yD (t) = (2 fi t) = fi
2 dt

Problem 3.41
For m (t) = A cos ! m t and
Z t
Akf
(t) = Akf cos ! m d = sin ! m t
!m
and
Akf
(s) =
s2 + ! 2m
The VCO output phase is

KT Akf KT
(s) = (s) =
s + KT (s + KT ) (s2 + ! 2m )

Using partial fraction expansion, (s) can be expressed as

Akf KT 1 s KT
(s) = 2 2
+ 2
KT + ! m s + K T s2 + ! m s + ! 2m
2

This gives, for t 0

Akf KT KT t KT
(t) = e cos ! m t + sin ! m t
KT 2 + ! 2m !m

The …rst term is the transient response. For large KT , only the third term is signi…cant.
Thus,
1 d Akf KT2
e (t) = = cos ! m t
K dt K KT2 + ! 2m
Also, since KT is large, KT2 + ! 2m t KT2 . This gives

Akf
e (t) cos ! m t
K

and we see that e (t) is proportional to m (t). If kf = K , we have

e (t) m (t)

Problem 3.42
3.1. PROBLEMS 33

From (3.229)
s 0 s2 + 2 f s + 2 R
ss = lim
s!0 s3 + Kt s2 + Kt as + Kt b

1. For …rst-order PLL (a = 0, b = 0) this is

s 2 2
0s +2 f s+2 R 0s +2 f s+2 R
ss = lim = lim
s!0 s3 + Kt s2 s!0 s2 + Kt s
Thus
2 + (2 f =s) + (2 R=s2 )
0s +2 f s+2 R 0
ss = lim = lim
s!0 s2 + Kt s s!0 1 + (Kt =s)
0
For 0 6= 0, f = 0, and R = 0: ss = lim =0
s!0 1 + (Kt =s)
0 + (2 f =s) 2 f
For 0 6= 0, f 6= 0, and R = 0: ss = lim =
s!0 1 + (Kt =s) Kt
0 + (2 f =s) + (2 R=s2 ) 2 R
For 0 6= 0, f 6= 0, and R 6= 0: ss = lim = lim =1
s!0 1 + (Kt =s) s!0 Kt s

2. For a second-order PLL (a 6= 0, b = 0) this is

s 2 2
0s + 2 f s + 2 R 0s + 2 f s+2 R
ss = lim = lim
s!0 s3 + Kt s2 + Kt as s!0 s2 + Kt s + Kt a
Thus
2 R
ss =
Kt a
which is zero for R = 0 and nonzero for R 6= 0 for any …nite values of 0 and f .
3. For a third-order PLL (a 6= 0, b 6= 0) this is

s 0 s2 + 2 f s + 2 R 0
ss = lim 3 2
= =0
s!0 s + Kt s + Kt as + Kt b Kt b
for any …nite values 0, f , and R.

Problem 3.43
The Costas PLL is shown in Figure 3.53. The output of the top multiplier is

m (t) cos ! c t [2 cos (! c t + )] = m (t) cos + m (t) cos (2! c t + )

which, after lowpass …ltering, is m (t) cos . The quadrature multiplier output is

m (t) cos ! c t [2 sin (! c t + )] = m (t) sin + m (t) sin (2! c t + )

34CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

dψ / dt

Figure 3.8: Phase-plane plot for a Costas PLL.

which, after lowpass …ltering, is m (t) sin . The multiplication of the lowpass …lter outputs
is
m (t) cos m (t) sin = m2 (t) sin 2
as indicated. Note that with the assumed input m (t) cos ! c t and VCO output 2 cos (! c t + ),
the phase error is . Thus the VCO is de…ned by

d d
= = K e (t)
dt dt
This is shown in Figure 3.8.
Since the ddt intersection is on a portion of the curve with negative slope, the point A
at the origin is a stable operating point. Thus the loop locks with zero phase error and zero
frequency error.

Problem 3.44
The phase plane is de…ned by
= ! Kt sin (t)
at = 0, = ss , the steady-state phase error. Thus

1 ! 1 !
ss = sin = sin
Kt 2 (100)

For ! = 2 (30)
1 30
ss = sin = 17:46 degrees
100
3.1. PROBLEMS 35

For ! = 2 (50)
1 50
ss = sin = 30 degrees
100
For ! = 2 (80)
1 80
ss = sin = 53:13 defrees
100
For != 2 (80)
1 80
ss = sin = 53:13 degrees
100
For ! = 2 (120), there is no stable operating point and the frequency error and the phase
error oscillate (PLL slips cycles continually).

Problem 3.45
With x (t) = A2 f0 t, we desire e0 (t) = A cos 2 37 f0 t. Assume that the VCO output
is a pulse train with frequency 13 fo . The pulse should be narrow so that the seventh
harmonic is relatively large. The spectrum of the VCO output consists of components
separated by 13 f0 with an envelope of sinc( f ), where is the pulse width. The center
frequency of the bandpass …lter is 37 f0 and the bandwidth is on the order of 13 f0 as shown
in Figure.3.9.Problem 3.46

From the de…nition

Kt t
f (t) = Kt e u(t)
it is clear that in the limit as Kt ! 1, f (0) = 1, and f (t) = 0 for t 6= 0. Also
Z 1 Z 1
1
f (t)dt = Kt e Kt t dt = e Kt t e Kt t 0 = 1
1 0

Therefore the given f (t) satis…es all properties of an impulse function in the limit as Kt !
1.

Problem 3.47
Since
(s) Kt (s + a)
=
(s) s (s + a) + Kt (s + a)
we …rst determine
(s) Kt (s + a) s (s + a)
G(s) = =1 =
(s) s (s + a) + Kt (s + a) s (s + a) + Kt (s + a)
so that
0 2 f 2 R s (s + a)
(s) = + +
s s2 s3 s (s + a) + Kt (s + a)
36CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

This component (the fundamental)

tracks the input signal

Bandpass filter
passband

13f
0 73f
0
f
0

Figure 3.9: Filtering scheme for Problem 3.45.

or
2
0s +2 f s+2 R (s + a)
(s) =
s2 s (s + a) + Kt (s + a)
Multiplying by s and taking the limit gives

2 R (s + a)
ss = lim s 0 +2 f +
s!0 s s (s + a) + Kt (s + a)
or
2 R
ss = lim 2 f +
s!0 s Kt
Thus

For 0 6= 0, f = 0, and R = 0: ss =0
2f
For 0 6= 0, f 6= 0, and R = 0: ss = lim
s!0 Kt
2 R
For 0 6= 0, f 6= 0, and R 6= 0: ss = lim =1
s!0 s Kt
3.1. PROBLEMS 37

Problem 3.48
From the de…nition of the transfer function
s+a
(s) Kt F (s) Kt s+ a
= =
(s) s + Kt F (s) s+a
s + Kt s+ a

which is
(s) Kt (s + a) Kt (s + a)
= = 2
(s) s (s + a) + Kt (s + a) s + (Kt + a) s + Kt a
Therefore
s2 + 2 ! n s + ! 2n = s2 + (Kt + a) s + Kt a
This gives p
!n = Kt a
and
Kt + a
= p
2 Kt a

Problem 3.49
With the assumed VCO output represented by e j (t) we have the output of the complex
phase detector de…ned as

Ac m(t)ej[ (t) (t)]

= Ac m(t) cos[ (t) (t)] + jAc m(t) sin[ (t) (t)]

This is Figure 3.53 with the top branch representing the real component Ac m(t) cos[ (t)]
and the bottom branch representing the quadrature, or imaginary, component Ac m(t) sin[ (t)].
Costas PLLs are often simulated using a complex multiplication in this manner.

Problem 3.50
Since the phase error is assumed small, so that cos [ (t)] 1, the input from the top branch
(see Figure 3.53 in the text.) to the multiplier preceding the loop …lter is simply the phase
error, m(t) cos[ (t)] m(t). The other input to the multiplier, for su¢ ciently small phase
error, is m(t) sin[ (t)] m(t) (t) Thus, the input to the loop …lter preceding the VCO is
m2 (t) (t). Since m(t) is a step function

m2 (t) = m(t) = u(t t0 )

so that the input to the loop …lter is u(t t0 ) (t), which is exactly the loop …lter input to
a PLL. We therefore see that this problem is identical to Example 3.13 in the text and the
phase error as a function of time is that given by (3.252) with t replaced by t t0 . The
38CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

parameter a in the loop …lter transfer function is a = fn = as de…ned by (3.248). This

problem illustrates the equivalence of a Costas PLL to a PLL in cases where the phase error
is small.

Problem 3.51
The message signal is
m(t) = 3 sin 2 (10)t + 4 sin 2 (20)t
The derivative of the message signal is

dm (t)
= 60 cos 2 (10) t + 160 cos 2 (20t)
dt
The maximum value of dm (t) =dt is obviously 220 and the maximum occurs at t = 0.
Thus
0
220
Ts
or
220 220
fs = = 4400
0 0:05
Thus, the minimum sampling frequency is 4400 Hz.

Problem 3.52
Let A be the peak-to-peak value of the data signal. The peak error is 0:25% and the
peak-to-peak error is 0:005 A. The required number of quantizating levels is

A
= 200 2n = q
0:005A
so we choose q = 256 and n = 8. The bandwidth is

B = 2W k log2 q = 2W k(8)

The value of k is estimated by assuming that the speech is sampled at the Nyquist rate.
Then the sampling frequency is fs = 2W = 8 kHz. Each sample is encoded into n = 8
pulses. Let each pulse be with corresponding bandwidth 1 . For our case

1 1
= =
nfs 2W n

Thus the bandwidth is

1
= 2W n = 2W log2 q = 2nW = 2knW
3.1. PROBLEMS 39

x1 (BW = W ) x2 (BW = W )
x3 (BW = 2W )
x4 (BW = 4W )
x5 (BW = 4W )

1 2 3 4 5 6 7 8 9 10 11 12

Output

Figure 3.10: Commutator con…guration for Problem 3.53.

and so k = 1. For k = 1
B = 2 (8; 000) (8) = 112 kHz
Problem 3.53

One possible commutator con…guration is illustrated in Figure 3.10. The signal at the
point labeled “output” is the baseband signal. The minimum commutator speed is 2W
revolutions per second. For simplicity the commutator is laid out in a straight line. Thus,
the illustration should be viewed as it would appear wrapped around a cylinder. After
taking the sample at point 12 the commutator moves to point 1. On each revolution, the
commutator collects 4 samples of x4 and x5 , 2 samples of x3 , and one sample of x1 and x2 .
The minimum transmission bandwidth is
X
B= Wi = W + W + 2W + 4W + 4W = 12W
i

Problem 3.54
The single-sided spectrum for x (t) is shown in Figure 3.11.
From the de…nition of y (t) we have

Y (s) = a1 X (f ) + a2 X (f ) X (f )
40CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

W W

f
f
1 f
2

Figure 3.11: Single-sided spectrum for Problem 3.54.

The spectrum for Y (f ) is given in Figure 3.12. Demodulation can be a problem since it may
be di¢ cult to …lter the desired signals from the harmonic and intermodulation distortion
caused by the nonlinearity. As more signals are included in x (t), the problem becomes more
di¢ cult. The di¢ culty with harmonically related carriers is that portions of the spectrum
of Y (f ) are sure to overlap. For example, assume that f2 = 2f1 . For this case, the harmonic
distortion arising from the spectrum centered about f1 falls exactly on top of the spectrum
centered about f2 .

3.2 Computer Exercises

Computer Exercise 3.1
The MATLAB program is

% File: ce3_1.m
t = 0:0.001:1;
fm = 1;
fc =10;
m = 4*cos(2*pi*fm*t-pi/9) + 2*sin(4*pi*fm*t);
[minmessage,index] = min(m);
mncoefs = [4 2]/abs(minmessage)
mn = m/abs(minmessage);
plot(fm*t,mn,’k’),
grid, xlabel(’Normalized Time’), ylabel(’Amplitude’)
mintime = 0.001*(index-1);
dispa = [’The minimum of m(t) is ’, num2str(minmessage,’%15.5f’), ...
3.2. COMPUTER EXERCISES 41

()Y f

a
1
2a W
2

aW
2

f2− 1f f
1 f
2 2f
2 +
f1 f2 2f
2

2−
f 1f

Figure 3.12: Output spectrum for Problem 3.54.

’and falls at ’,num2str(mintime,’%15.5f’), ’ s.’]; disp(dispa)

%
% Now we compute the efficiency.
%
mnsqave = mean(mn.*mn); % mean-square value
a = 0.5; % modulation index
asq = a^2;
efficiency = asq*mnsqave/(1+asq*mnsqave);
dispb = [’The efficiency is ’, num2str(efficency,’%15.5f’),’ .’]; disp(dispb)
%
% Now compute the carrier amplitude.
%
cpower = 50; % carrier power
camplitude = sqrt(2*cpower); % carrier amplitude
%
% Compute the magnitude and phase spectra.
%
xct = camplitude*(1+a*mn).*cos(2*pi*fc*t);
Npts = 1000;
fftxct = fft(xct,Npts)/Npts;
s1 = [conj(fliplr(fftxct(2:25))) fftxct(1:25)];
mags1 = abs(s1);
angles1 = angle(s1);
42CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

for k=1:length(s1)
if mags1(k) <= 0.01
angles1(k)=0;
end
end
%
% Plot the magnitude and phase spectra.
%
figure
subplot(2,1,1), stem(-24:24,mags1,’.’)
xlabel(’frequency’); ylabel(’magnitude’)
subplot(2,1,2), stem(-24:24,angles1,’.’)
xlabel(’frequency’); ylabel(’phase’)
% End of script file.

Executing the program gives:

» ce3_1
mncoefs =
0.9165 0.4583
The minimum of m(t) is -4.36424 and falls at 0.43500 s.
The efficiency is 0.11607 .

The program also generates Figures 3.13 and ??.

Computer Exercise 3.2

The MATLAB code written for Computer Example 3.2 follows.

% File: ce3_2.m
t = 0:0.001:1;
fm = 1;
fc =10;
m = 2*cos(2*pi*fm*t) + cos(4*pi*fm*t);
mhil = 2*sin(2*pi*fm*t) + sin(4*pi*fm*t);
xctusb = 0.5*m.*cos(2*pi*fc*t)-0.5*mhil.*sin(2*pi*fc*t);
xctlsb = 0.5*m.*cos(2*pi*fc*t)+0.5*mhil.*sin(2*pi*fc*t);
%
% Plot time-domain signals.
%
subplot(2,1,1), plot(t,xctusb)
3.2. COMPUTER EXERCISES 43

1.5

0.5
Amplitude

-0.5

-1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized Time

Figure 3.13: Message signal.

44CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

4
magnitude

0
-25 -20 -15 -10 -5 0 5 10 15 20 25
frequency
2

1
phase

-1

-2
-25 -20 -15 -10 -5 0 5 10 15 20 25
frequency

Figure 3.14: Spectra of AM signal.

3.2. COMPUTER EXERCISES 45

xlabel(’time’); ylabel(’USB signal’)

subplot(2,1,2), plot(t,xctlsb)
xlabel(’time’); ylabel(’LSB signal’)
%
% Determine and plot spectra.
%
Npts = 1000;
fftxctu = fft(xctusb,Npts)/Npts;
fftxctl = fft(xctlsb,Npts)/Npts;
sxctusb = [conj(fliplr(fftxctu(2:25))) fftxctu(1:25)];
magsusb = abs(sxctusb);
sxctlsb = [conj(fliplr(fftxctl(2:25))) fftxctl(1:25)];
magslsb = abs(sxctlsb);
figure
subplot(2,1,1), stem(-24:24,magsusb)
xlabel(’frequency’); ylabel(’USB magnitude’)
subplot(2,1,2), stem(-24:24,magslsb)
xlabel(’frequency’); ylabel(’LSB magnitude’)
% End of script file.

Executing the code gives Figures 3.15 and 3.16.

Computer Exercise 3.3

In this computer example we investigate the demodulation of SSB using carrier resertion
for several values of the constant k.for an assumed message signal.

% File: ce3_3.m
t = 0:0.001:1;
fm = 1;
m = 2*cos(2*pi*fm*t) + cos(4*pi*fm*t);
mhil = 2*sin(2*pi*fm*t) + sin(4*pi*fm*t);
k = 1;
ydt = sqrt((k+m).*(k+m)+mhil.*mhil);
ydt1 = ydt-k;
k = 10;
ydt = sqrt((k+m).*(k+m)+mhil.*mhil);
ydt2 = ydt-k;
k = 100;
ydt = sqrt((k+m).*(k+m)+mhil.*mhil);
ydt3 = ydt-k;
46CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

1
USB signal

-1

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time
2

1
LSB signal

-1

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time

Figure 3.15: Time-domain waveforms.

3.2. COMPUTER EXERCISES 47

0.8
USB magnitude

0.6

0.4

0.2

0
-25 -20 -15 -10 -5 0 5 10 15 20 25
frequency
0.8
LSB magnitude

0.6

0.4

0.2

0
-25 -20 -15 -10 -5 0 5 10 15 20 25
frequency

Figure 3.16: Amplitude spectra.

48CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

2
k=1
0

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
4

2
k=10

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
4

2
k=100

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Figure 3.17: Illustration of demodulation resertion for k = 1, 10, and 100.

subplot(3,1,1), plot(t,m,’k’,t,ydt1,’k--’)
ylabel(’k=1’)
subplot(3,1,2), plot(t,m,’k’,t,ydt2,’k--’)
ylabel(’k=10’)
subplot(3,1,3), plot(t,m,’k’,t,ydt3,’k--’)
ylabel(’k=100’)
% End of script file.

Executing the preceding MATLAB program gives the results illustrated in Figure 3.17. In
the three plots the message signal is the solid line and the dashed line represents the output
resulting from carrier reinsertion. It can be seen that the demodulated output improves for
increasing k and that the distortion is negligible for k = 100.

Computer Exercise 3.4

The MATLAB program illustrating recovery of the message signal from a VSB signal follows.

% File: ce3_4.m
t = 0:0.001:1;
3.2. COMPUTER EXERCISES 49

m(t)
-2

-4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
5
xc(t)

-5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

102
xc(t)+c(t)

100

98

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Time

Figure 3.18: Recovery of m(t) by carrier reinsertion in a VSB system.

m = cos(2*pi*t)-0.4*cos(4*pi*t)+0.9*cos(6*pi*t);
e1 = 0.64; e2 = 0.78; e3 = 0.92;
fc = 25; A = 100;
ct = A*cos(2*pi*fc*t);
xct = e1*cos(2*pi*(fc+1)*t)+(1-e1)*cos(2*pi*(fc-1)*t)-...
0.4*e2*cos(2*pi*(fc+2)*t)-0.4*(1-e2)*cos(2*pi*(fc-2)*t)+...
0.9*e3*cos(2*pi*(fc+3)*t)+0.9*(1-e3)*cos(2*pi*(fc-3)*t);
xdem = abs(xct+ct);
subplot(3,1,1), plot(t,m), ylabel(’m(t)’)
subplot(3,1,2), plot(t,xct), ylabel(’xc(t)’)
subplot(3,1,3), plot(t,xdem), ylabel(’xc(t)+c(t)’)
xlabel(’Time’), axis([0 1 97 103])
% End of script file.

The result of executing the preceding program is illustrated in Figure 3.18. The top pane
shows m(t), the second pane shows the transmitted signal xc (t), and the third shows the
transmitted signal plus unmodulated carrier (ampli…ed so that the amplitude is 100). En-
velope detecting this last signal recovers the message signal.

Computer Exercise 3.5

50CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

The MATLAB code follows.

% File: ce3_5.m
fs = 500;
delt = 1/fs;
t = 0:delt:1-delt;
npts = length(t);
fm = 20;
fd = [10 20 100];
for k=1:3
beta = fd(k)/fm;
cxce = exp(i*beta*sin(2*pi*fm*t));
as = (1/npts)*abs(fft(cxce));
evenf = [as(fs/2:fs) as(1:fs/2-1)];
fn = -fs/2:fs/2-1;
subplot(3,1,k); stem(fn,2*evenf,’.’)
ylabel(’Amplitude’)
end
% End of script file.

Computer Exercise 3.6

For a square-wave message signal, the phase deviation will be a triangular-wave. The
triangular-wave message signal is generated using a 6-term Fourier series. The MATLAB
code follows.

% File: ce3_6.m
fs = 1000; % sampling frequency
delt = 1/fs; % sampling increment
t = 0:delt:1-delt; % time vector
npts = length(t); % number of points
fn = (0:(npts/2))*(fs/npts); % frequency vector for plot
% Generate a triangular-wave signal as the phase deviation corresponding
% to a square-wave message signal using Fourier series and plot.
fm = 5;
b = [1 -1/9 1/25 -1/49 1/81 -1/121];
c = 2*pi*fm*[1 3 5 7 9 11];
m = zeros(1,length(t));
for n=1:length(c)
m=m+b(n)*sin(c(n)*t);
3.2. COMPUTER EXERCISES 51

2
Amplitude

0
-250 -200 -150 -100 -50 0 50 100 150 200 250
2
Amplitude

0
-250 -200 -150 -100 -50 0 50 100 150 200 250
1
Amplitude

0.5

0
-250 -200 -150 -100 -50 0 50 100 150 200 250

Figure 3.19: FM spectra with fm held constant and modulation indices of 0.5, 1 and 5.
52CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

end
m = m/max(m);
plot(t,m), xlabel(’Time’)
% Determine spectra.
fd1 = 0.10;
xct1 = sin(2*pi*250*t+fd1*m); % modulated carrier
asxc1 = (2/npts)*abs(fft(xct1)); % amplitude spectrum
ampspec1 = asxc1(1:((npts/2)+1)); % positive frequency portion
fd2 = 1.0;
xct2 = sin(2*pi*250*t+fd2*m); % modulated carrier
asxc2 = (2/npts)*abs(fft(xct2)); % amplitude spectrum
ampspec2 = asxc2(1:((npts/2)+1)); % positive frequency portion
fd3 = 10.0;
xct3 = sin(2*pi*250*t+fd3*m); % modulated carrier
asxc3 = (2/npts)*abs(fft(xct3)); % amplitude spectrum
ampspec3 = asxc3(1:((npts/2)+1)); % positive frequency portion
% Plot spectra.
figure % new figure
subplot(3,1,1)
stem(fn,ampspec1,’.k’);
ylabel(’Magnitude - fd=0.1’)
subplot(3,1,2)
stem(fn,ampspec2,’.k’);
ylabel(’Magnitude - fd=1’)
subplot(3,1,3)
stem(fn,ampspec3,’.k’);
xlabel(’Frequency’), ylabel(’Magnitude - fd=10’)
% End of script file.

Executing the program yields the results illustrated in Figures 3.20 and 3.21 , respectively.
Figure 3.20 shows the triangular-wave message signal. Figure 3.21 shows the spectrum. As
with Figure 3.29 in the text, all sum, and di¤erence frequencies are present. This non-linear
e¤ect is most easily seen for fd = 10 since higher deviations spread the spectrum more than
small deviation constants.

Computer Exercise 3.7

The MATLAB program for generating Figure 3.42 in the text follows.

% File: ce3_7.m
3.2. COMPUTER EXERCISES 53

0.8

0.6

0.4

0.2

-0.2

-0.4

-0.6

-0.8

-1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time

Figure 3.20: Triangular wave representing the phase deviation.

Magnitude - fd=0.1

0.5

0
0 50 100 150 200 250 300 350 400 450 500
1
Magnitude - fd=1

0.5

0
0 50 100 150 200 250 300 350 400 450 500
1
Magnitude - fd=10

0.5

0
0 50 100 150 200 250 300 350 400 450 500
Frequency

Figure 3.21: Spectra for FM signals with a square-wave message signal.

54CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

fs = 1000; % sampling frequency

delt = 1/fs; % sampling increment
t = 0:delt:1-delt; % time vector
npts = length(t); % number of point
fi = 1; Ac = 1;
%
% Let Ai=0.2.
%
Ai = 0.2;
xcr = Ac+Ai*exp(i*2*pi*fi*t);
psi1 = atan(imag(xcr)./real(xcr));
upsi1 = unwrap(angle(xcr));
udifpsi1 = zeros(1,npts);
for k=2:npts
udifpsi1(k) = (upsi1(k)-upsi1(k-1))/delt;
end
%
% Let Ai=0.9.
%
Ai = 0.9;
xcr = Ac+Ai*exp(i*2*pi*fi*t);
psi2 = atan(imag(xcr)./real(xcr));
upsi2 = unwrap(angle(xcr));
udifpsi2 = zeros(1,npts);
for k=2:npts
udifpsi2(k) = (upsi2(k)-upsi2(k-1))/delt;
end
%
% Let Ai=1.1.
%
Ai = 1.1;
xcr = Ac+Ai*exp(i*2*pi*fi*t);
psi3 = atan(imag(xcr)./real(xcr));
upsi3 = unwrap(angle(xcr));
udifpsi3 = zeros(1,npts);
for k=2:npts
udifpsi3(k) = (upsi3(k)-upsi3(k-1))/delt;
end
%
% Plot results.
3.2. COMPUTER EXERCISES 55

0.5 2

Ai=0.2
0 0

-0.5 -2
0 0.5 1 0 0.5 1
2 50

0
Ai=0.9

0
-50

-2 -100
0 0.5 1 0 0.5 1
2 100
Ai=1.1

0 50

-2 0
0 0.5 1 0 500 1000

Figure 3.22: Interference performance of FM demodulator.

%
subplot(3,2,1); plot(t,psi1); ylabel(’Ai=0.2’);
subplot(3,2,2); plot(t,udifpsi1)
subplot(3,2,3); plot(t,psi2): ylabel(’Ai=0.9’);
subplot(3,2,4); plot(t,udifpsi2)
subplot(3,2,5); plot(t,psi3); ylabel(’Ai=1.1’);
subplot(3,2,6); plot(udifpsi3)
% End of script file.

The results are shown in Figure 3.22. We see that these match closely Figure 3.42 in the
text.

Computer Exercise 3.8

This follows directly from the node names in the program.
Computer Exercise 3.9
The MATLAB code for the PLL simulation with a line added for entering the sampling
frequency follows.
56CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

% File: ce3_8.m
% beginning of preprocessor
clear all % be safe
fdel = input(’Enter frequency step size in Hz > ’);
fn = input(’Enter the loop natural frequency in Hz > ’);
zeta = input(’Enter zeta (loop damping factor) > ’);
fs = 2000; % default sampling frequency
fs = input(’Enter the sampling frequency (default = 2000) > ’);
npts = fs ; % default number of simulation point
T = 1/fs;
t = (0:(npts-1))/fs; % time vector
nsettle = fix(npts/10); % set nsettle time as 0.1*npts
Kt = 4*pi*zeta*fn; % loop gain
a = pi*fn/zeta; % loop filter parameter
filt_in_last = 0; filt_out_last=0;
vco_in_last = 0; vco_out = 0; vco_out_last=0;
% end of preprocesso - beginning of simulation loop
for i=1:npts
if i<nsettle
fin(i) = 0;
phin = 0;
else
fin(i) = fdel;
phin = 2*pi*fdel*T*(i-nsettle);
end
s1 = phin - vco_out;
s2 = sin(s1); % sinusoidal phase detector
s3 = Kt*s2;
filt_in = a*s3;
filt_out = filt_out_last + (T/2)*(filt_in + filt_in_last);
filt_in_last = filt_in;
filt_out_last = filt_out;
vco_in = s3 + filt_out;
vco_out = vco_out_last + (T/2)*(vco_in + vco_in_last);
vco_in_last = vco_in;
vco_out_last = vco_out;
phierror(i)=s1;
fvco(i)=vco_in/(2*pi);
freqerror(i) = fin(i)-fvco(i);
end
3.2. COMPUTER EXERCISES 57

% end of simulation loop - beginning of postprocessor

kk = 0;
while kk == 0
k = menu(’Phase Lock Loop Postprocessor’,...
’Input Frequency and VCO Frequency’,...
’Phase Plane Plot’,...
’Exit Program’);
if k == 1
plot(t,fin,t,fvco), grid
title(’Input Frequency and VCO Freqeuncy’)
xlabel(’Time - Seconds’)
ylabel(’Frequency - Hertz’)
pause
elseif k == 2
plot(phierror/2/pi,freqerror), grid
title(’Phase Plane’)
xlabel(’Phase Error / 2*pi’)
ylabel(’Frequency Error - Hz’)
pause
elseif k == 3
kk = 1;
end
end
% end of postprocessor
% End of script file.

The preceding program was executed twice; once for fs = 400 and once for fs = 1000. The
siumulation for fs = 400 is shown in Figure 3.23. We note that the curves, especially for
large frequency error, are not smooth, This is an indication that the samplingt frequency is
too low.

We therefore increase the sampling frequency from 400 to 1000. The results are illus-
trated in Figure 3.24.

We now see that the curve de…ning the phase plane is much smoother. In addition, for
fs = 400 we see that the PLL skips 4 cycles while acquiring lock. For fs = 1000 we see that
the PLL skips 3 cycles while acquiring lock. By executing the simulation a number of times
we see that additional increases in the sampling frequency have no e¤ect on the results.
An e¤ective way to determine an appropriate sampling frequency is to set the sampling
58CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

Phase Plane
50

40

30
Frequency Error - Hz

20

10

-10
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Phase Error / 2*pi

_
Figure 3.23: Phase-plane plot with fs = 400.
3.2. COMPUTER EXERCISES 59

Phase Plane
50

40

30
Frequency Error - Hz

20

10

-10
0 0.5 1 1.5 2 2.5 3 3.5
Phase Error / 2*pi

Figure 3.24: Phase-plane plot with fs = 1000:

60CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

frequency at a high value and then reduce the sampling frequency until the results change
and the waveforms are distorted. At this point the sampling frequency is too low.

Computer Exercise 3.10

The MATLAB program for the PLL with the limiting phase detector follows. We note
that the phase detector has been modi…ed and a line has been added for entering the phase
detector constant A.

% File: ce3_9.m
% beginning of preprocessor
clear all % be safe
fdel = input(’Enter frequency step size in Hz > ’);
fn = input(’Enter the loop natural frequency in Hz > ’);
zeta = input(’Enter zeta (loop damping factor) > ’);
AA = input(’Enter phase detector constant A > ’);
npts = 2000; % default number of simulation points
fs = 2000; % default sampling frequency
T = 1/fs;
t = (0:(npts-1))/fs; % time vector
nsettle = fix(npts/10); % set nsettle time as 0.1*npts
Kt = 4*pi*zeta*fn; % loop gain
a = pi*fn/zeta; % loop filter parameter
filt_in_last = 0; filt_out_last=0;
vco_in_last = 0; vco_out = 0; vco_out_last=0;
% end of preprocessor - beginning of simulation loop
for i=1:npts
if i<nsettle
fin(i) = 0;
phin = 0;
else
fin(i) = fdel;
phin = 2*pi*fdel*T*(i-nsettle);
end
s1 = phin - vco_out;
s2 = sin(s1); % sinusoidal phase detector
if s2<-AA
s2=-AA;
elseif s2>AA
s2=AA;
end
3.2. COMPUTER EXERCISES 61

s3 = Kt*s2;
filt_in = a*s3;
filt_out = filt_out_last + (T/2)*(filt_in + filt_in_last);
filt_in_last = filt_in;
filt_out_last = filt_out;
vco_in = s3 + filt_out;
vco_out = vco_out_last + (T/2)*(vco_in + vco_in_last);
vco_in_last = vco_in;
vco_out_last = vco_out;
phierror(i)=s1;
fvco(i)=vco_in/(2*pi);
freqerror(i) = fin(i)-fvco(i);
end
% end of simulation loop - beginning of postprocessor
kk = 0;
while kk == 0
k = menu(’Phase Lock Loop Postprocessor’,...
’Input Frequency and VCO Frequency’,...
’Phase Plane Plot’,...
’Exit Program’);
if k == 1
plot(t,fin,t,fvco), grid
title(’Input Frequency and VCO Freqeuncy’)
xlabel(’Time - Seconds’)
ylabel(’Frequency - Hertz’)
pause
elseif k == 2
plot(phierror/2/pi,freqerror), grid
title(’Phase Plane’)
xlabel(’Phase Error / 2*pi’)
ylabel(’Frequency Error - Hz’)
pause
elseif k == 3
kk = 1;
end
end
% end of postprocessor
% End of script file.
62CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

Phase Plane
50

40

30
Frequency Error - Hz

20

10

-10

-20
0 2 4 6 8 10 12
Phase Error / 2*pi

Figure 3.25: PLL acquisition with A = 0:5.

3.2. COMPUTER EXERCISES 63

We see that with A = 0:5, the loop slips 5 cycles, whereas with A_ > 1 (limiter absent), the
PLL slips only three cycles with the parameters given. The acquisition time is therefore
increased.

Computer Exercise 3.11

The MATLAB code for producing the PAM, PWM, and PPM signals follow.

% File: ce3_10.m
clear all; % be safe
N = 1000; % vector length
f = 1; % frequency
x = zeros(1,N); % initialize
zpam = zeros(1,N); % initialize
zpwm = zeros(1,N); % initialize
zppm = zeros(1,N); % initialize
width = 40; % nominal width
for n=1:N
x(1,n) = sin(2*pi*f*(n-1)/N+pi/8);
end
for k=1:100:N
zpam(1,k)=x(1,k); % sample value
for kk=1:width
m=k+kk;
zpam(1,m)=x(1,k); % PAM sigmal
end
widthpwm = round(width*(1.2+x(1,k)));
for kk=1:widthpwm
m=k+kk;
zpwm(1,m)=1.0; % PWM signal
end
for kk=1:5
m=k+widthpwm+(kk-1);
zppm(1,m)=1.0; % PPM signal
end
end
subplot(311)
plot(1:N,zpam), axis([0 1000 -1.2 1.2]), grid
ylabel(’PAM Signal’)
subplot(312)
plot(1:N,zpwm), axis([0 1000 -0.2 1.2]), grid
64CHAPTER 3. MODULATION TECHNIQUES WITH EMPHASIS ON ANALOG SYSTEMS

PAM Signal 0

-1
0 100 200 300 400 500 600 700 800 900 1000

1
PWM Signal

0.5

0
0 100 200 300 400 500 600 700 800 900 1000

1
PPM Signal

0.5

0
0 100 200 300 400 500 600 700 800 900 1000
Sample Index

Figure 3.26: PAM, PWM, and PPM signals.

ylabel(’PWM Signal’)
subplot(313)
plot(1:N,zppm), axis([0 1000 -0.2 1.2]), grid
ylabel(’PPM Signal’), xlabel(’Sample Index’)
% End of script file.

The PAM, PWM, and PPM signals are shown in Figure 3.26.