Introduction to Analog And Digital

Communications

Second Edition

Simon Haykin, Michael Moher

Chapter 4 Angle Modulation

4.1 Basic Definitions

4.2 Properties of Angle-Modulated Waves
4.3 Relationship between PM and FM waves
4.4 Narrow-Band Frequency Modulation
4.5 Wide-Band Frequency Modulation
4.6 Transmission Bandwidth of FM waves
4.7 Generation of FM waves
4.8 Demodulation of FM signals
4.9 Theme Example
: FM Stereo Multiplexing
4.10 Summary and Discussion
 Angel modulation
 The angle of the carrier wave is varied according to the information-bearing
signal

 Lesson 1 : Angle modulation is a nonlinear process, which testifies to its

sophisticated nature. In the context of analog communications, this
distinctive property of angle modulation has two implications :
 In analytic terms, the spectral analysis of angle modulation is complicated.
 In practical terms, the implementation of angle modulation is demanding

 Lesson 2 : Whereas the transmission bandwidth of an amplitude-modulated

wave is of limited extent, the transmission bandwidth of an angle-
modulated wave may an infinite extent, at least in theory.

 Lesson 3 : Given that the amplitude of the carrier wave is maintained

constant, we would intuitively expect that additive noise would affect the
performance of angle modulation to a lesser extent than amplitude
modulation.

3
 4.1 Basic Definitions
 Angle-modulated wave
s (t ) = Ac cos[θ i (t )] (4.1)
 the average frequency in hertz
θ t (t + ∆t ) − θ i (t )
f ∆t (t ) =
2π∆t
 The instantaneous frequency of the angle-modulated signal
f i (t ) = lim f ∆t (t )
∆ →0
t

θ (t + ∆t ) − θ i (t ) 
= lim  t 
∆t → 0
 2π∆t
1 dθ i (t )
= (4.2)
2π dt

θ i (t ) = 2πf c t + φc , for m(t ) = 0

4
1. Phase modulation (PM) is that form of angle modulation in which the
instantaneous angle is varied linearly with the message signal
θ i (t ) = 2πf c t + k p m(t ) (4.3)
s (t ) = Ac cos[2πf c t + k p m(t )] (4.4)

2. Frequency modulation (FM) is that form of angle modulation in which

the instantaneous frequency is varied linearly with the message signal
f i (t ) = f c + k f m(t ) (4.5)

θ i (t ) = 2π ∫ f i (τ )dτ
t

∫
t
= 2πf c t + 2πk f m(τ )dτ (4.6)
0


s (t ) = Ac cos 2πf c t + 2πk f m(τ )dτ  (4.7)
∫
t

 0 
Table. 4.1
5
Table.4.1 Back Next

6
 4.2 Properties of Angle-Modulated Waves
 Property 1 : Constancy of transmitted power
 The amplitude of PM and FM waves is maintained at a constant value equal
to the carrier amplitude for all time.
 The average transmitted power of angle-modulated waves is a constant
1 2
Pav = Ac (4.8)
2

 Property 2 : Nonlinearity of the modulation process

 Its nonlinear character
m(t ) = m1 (t ) + m2 (t )
s (t ) = Ac cos[2πf c t + k p (m1 (t ) + m2 (t ))]
s1 (t ) = Ac cos[2πf c t + k p m1 (t )]
s2 (t ) = Ac cos[2πf c t + k p m2 (t )]

s (t ) ≠ s1 (t ) + s2 (t ) Fig. 4.1
7
Fig.4.1 Back Next

8
 Property 3 : Irregularity of zero-crossings
 Zero-crossings are defined as the instants of time at which a waveform
changes its amplitude from a positive to negative value or the other way
around.
 The irregularity of zero-crossings in angle-modulation waves is also
attributed to the nonlinear character of the modulation process.

 The message signal m(t) increases or decreases linearly with time t, in

which case the instantaneous frequency fi(t) of the PM wave changes form
the unmodulated carrier frequency fc to a new constant value dependent on
the slope of m(t)
 The message signal m(t) is maintained at some constant value, positive or
negative, in which case the instantaneous frequency fi(t) of the FM wave
changes from the unmodulated carrier frequency fc to a new constant value
dependent on the constant value of m(t)

9
 Property 4 : Visualization difficulty of message waveform
 The difficulty in visualizing the message waveform in angle-modulated
waves is also attributed to the nonlinear character of angle-modulated
waves.

 Property 5 : Tradeoff of increased transmission bandwidth for

improved noise performance
 The transmission of a message signal by modulating the angle of a
sinusoidal carrier wave is less sensitive to the presence of additive noise

10
Fig. 4.2
11
Fig.4.2 Back Next

12
13
14
15
 4.3 Relationship Between PM and FM waves
 Fig. 4.3(a)
 An FM wave can be generated by first integrating the message signal
m(t) with respect to time t and then using the resulting signal as the
input to a phase modulation
 Fig. 4.3(b)
 A PM wave can be generated by first differentiating m(t) with respect
to time t and then using the resulting signal as the input to a frequency
modulator

 We may deduce the properties of phase modulation from those of

frequency modulation and vice versa

Fig. 4.3
16
Fig.4.3 Back Next

17
 4.4 Narrow-Band Frequency Modulation
 We first consider the simple case of a single-tone modulation that produces a
narrow-band FM wave
 We next consider the more general case also involving a single-tone
modulation, but this time the FM wave is wide-band

 The two-stage spectral analysis described above provides us with enough

insight to propose a useful solution to the problem
 A FM signal is
m(t ) = A cos(2πf t ) (4.9)
m m

f i (t ) = f c + k f Am cos(2πf mt )
= f c + ∆f cos(2πf mt ) (4.10)

∆f = k f A m ∆f
(4.11) θ i (t ) = 2πf c t + sin( 2πf mt ) (4.12)
The frequency deviation
f m Modulation index of the FM wave

∆f
β= (4.13) θ i (t ) = 2πf c t + β sin(2πf mt ) (4.14)
fm
The phase deviation of the FM wave 18
 The FM wave is
s (t ) = Ac cos[2πf c t + β sin( 2πf mt )] (4.15)

cos( A + B ) = cos A cos B − sin A sin B

s (t ) = Ac cos(2πf c t ) cos[β sin( 2πf mt )] − Ac sin( 2πf c t ) sin[β sin( 2πf mt )] (4.16)

 If the modulation index is small compared to one radian, the approximate form
of a narrow-band FM wave is

s (t ) ≈ Ac cos(2πf c t ) − βAc sin( 2πf c t ) sin( 2πf mt ) (4.17)

cos[β sin( 2πf mt )] ≈ 1
sin[β sin( 2πf mt )] ≈ β sin( 2πf mt )
1. The envelope contains a residual amplitude modulation that varies with time
2. The angel θi(t) contains harmonic distortion in the form of third- and higher
order harmonics of the modulation frequency fm
Fig. 4.4
19
Fig.4.4 Back Next

20
 We may expand the modulated wave further into three frequency
components
s (t ) ≈ Ac cos(2πf c t ) + βAc {cos[2π ( f c + f m )t ] − cos[2π ( f c − f m )t ]} (4.18)
1
2
s AM (t ) = Ac cos(2πf c t ) + µAc {cos[2π ( f c + f m )t ] + cos[2π ( f c − f m )t ]} (4.19)
1
2

 The basic difference between and AM wave and a narrow-band FM

wave is that the algebraic sign of the lower side-frequency in the
narrow-band FM is reversed
 A narrow-band FM wave requires essentially the same transmission
bandwidth as the AM wave.

21
 Phasor Interpretation
 A resultant phasor representing the narrow-band FM wave that is
approximately of the same amplitude as the carrier phasor, but out of
phase with respect to it.
 The resultant phasor representing the AM wave has a different
amplitude from that of the carrier phasor, but always in phase with it.

Fig. 4.5
22
Fig.4.5 Back Next

23
 4.5 Wide-Band Frequency Modulation
 Assume that the carrier frequency fc is large enough to justify rewriting Eq.
4.15) in the form
s (t ) = Re[ Ac exp( j 2πf c t + jβ sin( 2πf m t ))]

= Re s (t ) exp( j 2πf c t )
 ~
(4.20)
 

 The complex envelope is

~
s (t ) = Ac exp[ jβ sin( 2πf mt ) ] (4.21)
~
s (t ) = Ac exp[ jβ sin( 2πf m (t + k / f m )) ]
= Ac exp[ jβ sin( 2πf mt + 2kπ ) ]
= Ac exp[ jβ sin( 2πf mt ) ]
∞

∑ c exp( j 2πnf t )
~
s (t ) = n m (4.22)
n = −∞

24
 The complex Fourier coefficient
1 /( 2 f m ) ~
cn = f m ∫
−1 /( 2 f m )
s (t ) exp(− j 2πnf mt )dt

= f m Ac ∫
1 /( 2 f m )
exp[ jβ sin( 2πf mt ) − j 2πnf mt ]dt (4.23)
−1 /( 2 f m ) x = 2πf mt (4.24)

Ac π
cn =
2π ∫ −π
exp[ j ( β sin x − nx)]dx (4.25)
1 π
J n (β ) =
2π ∫−π
exp[ j ( β sin x − nx)]dx (4.26)

cn = Ac J n ( β ) (4.27)
∞

∑ J (β ) exp( j 2πnf t )
~
s (t ) = Ac n m (4.28)
n = −∞

 ∞ 
s (t ) = Re  Ac ∑ J n ( β ) exp[ j 2π ( f c + nf m )t ] (4.29)
 n = −∞ 
25
 In the simplified form of Eq. (4.29)

s (t ) = Ac ∑ J (β ) cos[2π ( f
n = −∞
n c + nf m )t ] (4.30)

Ac ∞
S( f ) =
2 n = −∞
∑
J n ( β )[δ ( f − f c − nf m ) + δ ( f + f c + nf m )] (4.31)

26
 Properties of single-tone FM for arbitrary modulation index β

1. For different integer values of n,

J n ( β ) = J − n ( β ), for n even (4.32)
J n ( β ) = − J − n ( β ), for n odd (4.33)
2. For small values of the modulation index β
J 0 ( β ) ≈ 1, 
β 
J1 ( β ) ≈ ,  (4.34)
2 
J n ( β ) ≈ 0, n > 2

3. The equality holds exactly for arbitrary β

∞

∑J 2
n ( β ) = 1 (4.35)
n = −∞ Fig. 4.6
27
Fig.4.6 Back Next

28
1. The spectrum of an FM wave contains a carrier component and an infinite
set of side frequencies located symmetrically on either side of the carrier at
frequency separations of fm,2fm, 3fm….
2. The FM wave is effectively composed of a carrier and a single pair of side-
frequencies at fc±fm
3. The amplitude of the carrier component of an FM wave is dependent on the
modulation index β
The average power of such a signal developed across a 1-ohm resistor is
also constant.
1 2
Pav = Ac
2

 The average power of an FM wave may also be determined form

∞

∑
1
P = Ac2 J n2 ( β ) (4.36)
2 −∞

29
Fig. 4.7

Fig. 4.8
30
Fig.4.7 Back Next

31
Fig.4.8 Back Next

32
 4.6 Transmission Bandwidth of FM waves
 Carson’s Rule
 The FM wave is effectively limited to a finite number of significant side-
frequencies compatible with a specified amount of distortion
 Two limiting cases
1. For large values of the modulation index β, the bandwidth approaches, and is only
slightly greater than the total frequency excursion 2∆f,
2. For small values of the modulation index β, the spectrum of the FM wave is
effectively limited to the carrier frequency fc and one pair of side-frequencies at
fc±fm, so that the bandwidth approaches 2fm

 An approximate rule for the transmission bandwidth of an FM wave

 1
BT ≈ 2∆f + 2 f m = 2∆f 1 +  (4.37)
 β

33
 Universal Curve for FM Transmission Bandwidth
 A definition based on retaining the maximum number of significant
side frequencies whose amplitudes are all greater than some selected
value.
 A convenient choice for this value is one percent of the unmodulated
carrier amplitude
Table. 4.2
 The transmission bandwidth of an FM waves
 The separation between the two frequencies beyond which none of the side
frequencies is greater than one percent of the carrier amplitude obtained
when the modulation is removed.

 As the modulation index β is increased, the bandwidth occupied by the

significant side-frequencies drops toward that value over which the carrier
frequency actually deviates.

Fig. 4.9
34
Table.4.2 Back Next

35
Fig.4.9 Back Next

36
 Arbitrary Modulating Wave
 The bandwidth required to transmit an FM wave generated by an
arbitrary modulating wave is based on a worst-case tone-modulation
analysis

 The deviation ratio D

∆f
D= (4.38)
W
 The generalized Carson rule is

BT = 2(∆f + W ) (4.39)

Fig. 4.9
37
38
 4.7 Generation of FM Waves
 Direct Method
 A sinusoidal oscillator, with one of the reactive elements in the tank
circuit of the oscillator being directly controllable by the message signal
 The tendency for the carrier frequency to drift, which is usually
unacceptable for commercial radio applications.
 To overcome this limitation, frequency stabilization of the FM
generator is required, which is realized through the use of feed-back
around the oscillator

 Indirect Method : Armstrong Modulator

 The message signal is first used to produce a narrow-band FM, which is
followed by frequency multiplication to increase the frequency
deviation to the desired level.
 Armstrong wide-band frequency modulator
 The carrier-frequency stability problem is alleviated by using a highly
stable oscillator
Fig. 4.10
39
Fig.4.10 Back Next

40
 A Frequency multiplier
 A memoryless nonlinear device
 The input-output relation of such a device is

v(t ) = a1 s (t ) + a1 s 2 (t ) + ... + an s n (t ) (4.40)


s (t ) = Ac cos 2πf c t + 2πk f m(τ )dτ  (4.41)
∫
t

 0 

f i (t ) = f c + k f m(t ) (4.42)

 A new FM wave is


s (t ) = Ac cos 2πf c t + 2πk f m(τ )dτ  (4.43)
∫
t
' ' '
 0 
f i ' (t ) = nf c + nk f m(t ) (4.44)

Fig. 4.11
41
Fig.4.11 Back Next

42
 4.8 Demodulation of FM Signals
 Frequency Discriminator
 The FM signal is

s (t ) = Ac cos 2πf c t + 2πk f m(τ )dτ 


∫
t

 0 
 We can motivate the formulation of a receiver for doing this recovery by
nothing that if we take the derivative of Eq. (4.44) with respect to time

= −2πAc [ f c + k f m(t )] sin  2πf c t + 2πk f m(τ )dτ  (4.45)


ds (t )
∫
t

dt  0 
d
⇔ j 2πf (4.46)
dt
 A typical transfer characteristic that satisfies this requirement is
 j 2π [ f − ( f c − BT / 2)], f c − ( BT / 2) ≤ f ≤ f c + ( BT / 2)
H1 ( f ) =  (4.47)
0, otherwise

43
 The slope circuit
 The circuit is also not required to have zero response outside the
transmission bandwidth
 The complex envelope of the FM signal s(t) is

S (t ) = Ac exp j 2πk f m(τ )dτ  (4.48)


~

∫
t

 0 
~
 j 2π [ f + ( BT / 2)], − BT / 2 ≤ f ≤ BT / 2
H 1( f ) =  (4.49)
0, otherwise
~
1 ~ ~
S1 ( f ) = H 1 ( f ) S ( f )
2
  1 ~ 1 1
 j π f + B T  S ( f ), − B ≤ f ≤ BT
=  2 
T
2 2 (4.50)
0,
 elsewhere

Fig. 4.12
44
Fig.4.12 Back Next

45
1. Multiplication of the Fourier transform by j2πf is equivalent to differentiating
the inverse Fourier transform
d ~ ~
s (t ) ⇔ j 2πf S ( f )
dt
2. Application of the linearity property to the nonzero part of yields
~
1d ~ 1 ~
s1 (t ) = s (t ) + jπBT s (t ) (4.51)
2 dt 2
the actual response of the slope circuit due to the FM wave s(t) is given by
  2k f  
m(t ) exp j 2πk f m(τ )dτ  (4.52)

~
1
∫
t
s1 (t ) = jπAc BT 1 + 
2   BT    0 

s1 (t ) = Re s1 (t ) exp( j 2πf c t )
 ~

 
1   2k    π
∫
t
= πAc BT 1 +  f m(t ) cos 2πf c t + 2πk f m(τ )dτ +  (4.53)
2   BT    0 2

46
 The envelope detector
 2k f 
  m(t ) max < 1, for all t
 BT 
 Under ideal conditions, the output of the envelope detector is

1   2k f  
v1 (t ) = πAc BT 1 +  m(t ) (4.54)
2   BT  
1   2k f  
v2 (t ) = πAc BT 1 −  m(t ) (4.55)
2   BT  
 The overall output that is bias-free

v(t ) = v1 (t ) − v2 (t )
= cm(t ) (4.56)
Fig. 4.13
47
Fig.4.13 Back Next

48
 Phase-Locked Loop
 A feedback system whose operation is closely linked to frequency
modulation
 Three major components
 Voltage-controlled oscillator (VCO)
 Multiplier
 Loop filter of a low-pass kind

 Fig. 4.14, a closed-loop feedback system Fig. 4.14

 VCO has bee adjusted so that when the control signal is zero, two
conditions are satisfied
1. The frequency of the VCO is set precisely at the unmodulated carrier
frequency fc of the incoming FM wave s(t)
2. The VCO output has a 90◦-degree phase-shift with respect to the
unmodulated carrier wave.

49
Fig.4.14 Back Next

50
 Suppose the incoming FM wave is
s (t ) = Ac sin[ 2πf c t + φ1 (t )] (4.57)
φ1 (t ) = 2πk f ∫ m(τ )dτ
t
(4.58)
0

 The FM wave produced by the VCO as

r (t ) = Av cos[2πf c t + φ2 (t )] (4.59)
φ2 (t ) = 2πkv ∫ v(τ )dτ
t
(4.60)
0

 The multiplication of the incoming FM wave by the locally generated

FM wave produces two components
 A high-frequency component
k m Ac Av sin[ 4πf c t + φ1 (t ) + φ2 (t )]
 A low-frequency component
k m Ac Av sin[φ1 (t ) − φ2 (t )]

51
 Discard the double-frequency term, we may reduce the signal applied to the
loop filter to
e(t ) = k m Ac Av sin[φe (t )] (4.61)

 The phase error is

φe (t ) = φ1 (t ) − φ2 (t )

∫
t
= φ1 (t ) − 2πkv v(τ )dτ (4.62)
0

sin[φe (t )] ≈ φe (t )
e(t ) ≈ k m Ac Avφe (t )
K 0 = k m k v Ac Av (4.64)
K0
= φe (t ) (4.63) Loop-gain parameter of the phase lock loop
kv
 Eq. (4.62), (4.63), (4.65), and (4.60)constitute a linearized feedback model of
the phase-locked loop
∞
v(t ) = ∫
−∞
e(τ )h(t − τ )dτ (4.65)

52
 When the open-loop transfer function of a linear feedback system
has a large magnitude compared with unity for all frequencies, the
closed-loop transfer function of the system is effectively determined
by the inverse of the transfer function of the feedback path.

1. The inverse of this feedback path is described in the time domain by the
scaled differentiator
1  dφ2 (t ) 
v(t ) =   (4.66)
2πkv  dt 

2. The closed-loop time-domain behavior of the phase-locked loop is described

by the overall output v(t) produced in response to the angle Φ1(t) in the
incoming FM wave s(t)

3. The magnitude of the open-loop transfer function of the phase-locked loop is

controlled by the loop-gain parameter K0

53
 We may relate the overall output v(t) to the input angle Φ1(t) by
1  dφ1 (t ) 
v(t ) ≈   (4.67)
2πkv  dt 
1 d
⋅  2πk f ∫ m(τ )dτ 
t
v(t ) ≈
2πkv dt  0 
kf
= m(t ) (4.68)
kv

Fig. 4.15
54
Fig.4.15 Back Next

55
 4.9 Theme Example : FM Stereo Multiplexing
 The specification of standards for FM stereo transmission is influenced
by two factors
1. The transmission has to operate within the allocated FM broadcast channels
2. It has to be compatible with monophonic radio receivers

m(t ) = [ml (t ) + mr (t )] + [ml (t ) − mr (t )] cos(4πf c t ) + K cos(4πf c t ) (4.69)

 The multiplied signal is recovered by frequency demodulating the incoming

FM wave

Fig. 4.16
56
Fig.4.16 Back Next

57
 4.10 Summary and Discussion
 Two kinds of angle modulation
 Phase modulation (PM), where the instantaneous phase of the sinusoidal
carrier wave is varied linearly with the message signal
 Frequency modulation (FM), where the instantaneous frequency of the
sinusoidal carrier wave is varied linearly with the message signal

 Frequency modulation is typified by the equation

s (t ) = Ac cos 2πf c t + 2πk f ∫ m(τ )dτ  (4.70)

 t

 0 

 FM is a nonlinear modulation process

 In FM, the carrier amplitude and therefore the transmitted average power
is constant
 Frequency modulation provides a practical method for the tradeoff of
channel bandwidth for improved noise performance.

58
Fig.4.17 Back Next

Fig. 4.17
59
Fig.4.18 Back Next

Fig. 4.18
60
Fig.4.19 Back Next

Fig. 4.19
61
Fig.4.20 Back Next

Fig. 4.20
62
Fig.4.21 Back Next

Fig. 4.21
63