Chapter 4: Frequency Modulation Theory

Contents
Frequency modulation systems: Instantaneous frequency, frequency deviation,
modulation index, Bessel coefficients, significant sideband criteria, bandwidth of a
sinusoidally modulated FM signal, power of an FM signal, narrowband FM, direct and
indirect FM generation. Principle of operation of PLL. Various methods of FM
demodulation, discriminator, phase-lock loop; limiter, pre-emphasis and de-emphasis,
stereophonic FM broadcasting. FM broadcast band and specification.

1. Angle Modulation
2. Spectrum of an FM Signal
3. FM Signal Generation
4. FM Demodulation
5. Noise in FM Signal
6. FM Broadcast Band and Specifications

1
FREQUENCY MODULATION THEORY
Angle Modulation
• In angle modulation, the amplitude of the modulated carrier is
held constant and either the phase or the time derivative of the
phase of the carrier is varied linearly with the message signal
m(t).
• There are two basic types of angle modulation
• Phase modulation (PM)
• Frequency modulation (FM)
Generally, angle-modulated signal is given by
x(t) = Ac cos [2πfct + θ(t) ]
• The quantity 2πfc t + θ(t) = ψi(t) is called the instantaneous
phase of x(t).
• The quantity θ(t) is called the phase deviation of x(t).
2
Angle Modulation
The instantaneous angular frequency ωi(t) of x(t),
defined as the rate of change of the instantaneous phase
and having unit of radians per sec (rad/s), is

dΨ d (2πf ct +θ (t )) d θ (t )
ωi (t ) = i = = 2πf c +
dt dt dt
The instantaneous frequency fi(t), having unit of hertz
(Hz), of x(t) is accordingly given by
ωi (t) 1 dθ (t )
f i (t ) = = fc +
2π 2π dt
The quantity dθ/dt is called the angular frequency
deviation. 3
Angle Modulation
Example
An angle modulated signal is given by
x(t ) = cos(2000t + 0.1 sin(20t ))
a) Find the frequency of the modulated carrier fc
b) Instantaneous frequency fi

Solution
a) 2πfc=2000, fc=318.31Hz.
b) θ (t ) = 0.1sin(20t )

f i (t ) = f c + 1 dθ (t ) = 318.31+ 0.32cos(20t )
2π dt
4
Phase Modulation (PM)
From angle modulated signal given by
x(t) = Ac cos [2πfct + θ(t) ]
Variation of θ(t) produces Phase Modulation
In phase modulation, θ(t) is proportional to the modulating signal.
Thus
θ (t) = 2πk pm(t )
where kp is the phase sensitivity, with unit volt-1.
In PM, the instantaneous phase varies linearly with the message
signal
Ψi = 2π f ct + 2π k p m ( t )
And, the instantaneous frequency is

f i (t ) = f c + k p dm( t )
5
dt
Phase Modulation (PM)
Example 5 volt
m(t)
0 volt
0.01s
A phase modulated signal has a carrier of frequency 2kHz. Draw the
phase modulated signal given the phase sensitivity is 2volt-1 for the
above modulating signal m(t).
Solution
The instantaneous frequency is f i (t ) = f c + k p dm( t )
• fi=fc+2(0)=2kHz, for m(t)=0. dt
• fi=fc+2(500)=3kHz, for m(t)=500t.

6
2kHz 3kHz 2kHz
Frequency Modulation (FM)
Variation of dθ/dt produces Frequency Modulation
In frequency modulation, dθ/dt is proportional to the modulating
signal.
Thus dθ (t ) = 2π k m(t) or phase angle θ (t ) = 2π k t
m(τ )dτ
f f ∫0
dt
where kf , expressed in Hertz per unit of m(t) (Hz/volt), represents
the frequency sensitivity of the FM signal.
In FM, the instantaneous frequency varies linearly with the message
signal.
f i (t ) = f c + k f m(t )
And, the instantaneous phase is

Ψi = 2π f ct +θ (t ) = 2π f ct + 2π k f ∫0t m(τ )dτ

7
Frequency Modulation (FM)
Example 5 volt
m(t)
0 volt
0.01s
A freq. modulated signal has a carrier of frequency 2kHz. Draw the
freq. modulated signal given the freq. sensitivity is 500 Hz/volt for
the above modulating signal m(t).
Solution
The instantaneous frequency is f (t ) = f + k m(t )
• fi=fc+500(0)=2kHz, for m(t)=0. i c f
• fi=fc+500(500t)=2k+250kt, for m(t)=500t.

8
2kHz 2kHz to 4.5kHz 2kHz
PM signal:
xPM ( t ) = Ac cos  2π f c t + 2π k p m ( t ) 

FM signal:
( t ) = Ac cos  2π f ct + 2π k f ∫0 m (τ ) dτ 
 t
xFM

9
FM – Frequency Deviation
Consider a sinusoidal modulating information signal given by
m(t ) = Am cos(2πf mt )
The instantaneous frequency of the resulting FM signal equals

f i (t ) = f c + k f Am cos(2πf mt )
The maximum change in instantaneous frequency fi from the carrier
frequency fc, is known as frequency deviation ∆f.

∆ f = k f Am
The frequency deviation is a useful parameter for determining the
bandwidth of the FM signals.
10
FM – Phase Deviation and Modulation Index
The phase deviation of FM signal is given by

t ∆f
θ (t ) = 2π k f ∫0 m (τ )dτ = sin(2π f m t )
fm
The ratio of the frequency deviation to the modulation frequency fm is
called the modulation index of the FM signal.

We denote it by
β = ∆ f
fm
β is measured in radians. It represents the phase deviation of the
FM signal. In other words β represents the maximum change in
instantaneous phase ψi from the angle “2πfc t” of the carrier.
11
FM – Modulation Index
• Generally, the modulating signal m(t) might compose of
different sinusoids of mixed frequencies and amplitudes.
• Thus, the modulation index
∆f k f |m(t )|max
β= =
fm f m(max)
where |m(t)|max is the peak amplitude deviation of the
modulating signal and fm is the maximum modulating
frequency.
• β < 1 → Narrow-band FM
• β > 1 → Wide-band FM
12
PM and FM signals
• If m(t ) = Am cos(2πf mt ) , the PM signal is given by

xPM ( t ) = Ac cos  2π f c t + 2π k p Am cos ( 2π f mt ).  .

• If m(t ) = Am cos(2πf mt ) , the FM signal is given by

xFM ( t ) = Ac cos  2π f c t + β sin ( 2π f mt ) 

.

13
FM – Modulation Index
Example
The audio signal to be transmitted in a FM radio ranges from 20 to
15,000 Hz. Calculate the maximum frequency deviation for the
modulation index 2 and 5. If the amplitude of the audio signal stays
at ±2V, what is the frequency sensitivity?

Solution
• fm(max)=15,000.
• β=2: ∆f =β fm(max)=2x15,000=30kHz.
• β=5: ∆f =β fm(max)=5x15,000=75kHz.
• β=2: kf = ∆f /Am=30k/2=15kHz/V.
• β=5: kf = ∆f /Am=75k/2=37.5kHz/V.
14
Summary – Angle Modulation
You should be able to define/describe:
• angle modulation, phase modulation, frequency modulation
You should also be able to define other related terms:
• instantaneous phase and phase deviation
• instantaneous frequency and frequency deviation
• phase sensitivity
• frequency sensitivity
• modulation index

15
Overview: Chapter 4

Frequency Modulation Systems

• Angle Modulation
• Spectrum of an FM Signal
• FM Signal Generation
• FM Demodulation
• Noise in FM Signal
• FM Broadcast Band and Specifications

16
Narrow-band Frequency Modulation
The expression for FM signal
x(t ) = Ac cos(2π f ct + β sin(2π f mt))
= Ac cos(2π f ct )cos(β sin(2π f mt)) −
Ac sin(2π f ct)sin(β sin(2π f mt))
since cos( A + B ) = cos A cos B − sin A sin B
For small values of β, β <1:
cos( β sin( 2πf mt )) ≈ 1
sin( β sin( 2πf mt )) ≈ β sin(2πf mt )
17
Narrow-band FM
Thus, the FM signal becomes

x(t ) = Ac cos(2πf ct ) − Ac sin(2πf ct )β sin(2πf mt )

Then
x(t ) = Ac cos(2π f ct ) +

2 c{
1 β A cos[2π ( f + f )t ] −cos[2π ( f − f )t ]
c m c m }
1
since sin A sin B = [cos( A − B ) − cos( A + B )]
2
18
Narrow-band FM
Ac 
X ( f ) = δ ( f − f c ) + δ ( f + f c )  +
2
1 β A δ ( f − f − f ) + δ ( f + f + f )  +
4 c  c m c m 

1 β A δ ( f − f + f ) + δ ( f + f − f ) 
4 c  c m c m 

Amplitude spectrum (double-sided plot)

19
Wide-band Frequency Modulation
The corresponding FM signal is given by
x(t ) = Ac cos(2πf ct + β sin(2πf mt ))
Suitable for β≥0
It may alternatively be written as

x(t ) = Ac Re exp ( j 2π f ct + j β sin ( 2π f mt ) )

 
= Ac Re{exp ( j 2π f ct ) exp( jβ sin ( j 2π f mt )}
Since exp(jx)=cos(x)+jsin(x); thus cos(x)=Re{exp(jx)}.

20
Wide-band FM
Consider the complex function z(t) given by

z(t ) = exp ( j β sin( 2π f mt ) )

It is periodic with frequency fm and can therefore be expressed as
Fourier series as follows:
∞
z(t ) = ∑ Dn exp( j 2π nf mt )
n =−∞
where the Fourier coefficient is given by
1/ 2 f m
Dn = f m ∫ z(t )exp(− j 2π nf mt )dt
−1/ 2 f m
21
Wide-band FM
1/ 2 f m
Dn = f m ∫ exp[ j β sin(2π f mt )− j 2π nf mt )]dt
−1/ 2 f m

Define a new variable: y = 2πf mt

Hence, we can rewrite Dn as
π
Dn = 1 ∫ exp[ j(β sin( y)− ny)]dy
2π −π
The integral on the right hand side is a function of “n”
and β and is known as the Bessel function of the first kind
of order n and argument β. 22
Wide-band FM
It is conventionally denoted by Jn( β ). That is,
π
J n (β ) = 1 ∫ exp[ j(β sin( y)−ny)]dy
2π −π
Thus
Dn = J n (β )
∞
we get z(t ) = ∑ J n (β ) exp( j 2π nf mt )
n =−∞
Since

x(t ) = Ac Re{exp ( j 2π f ct ) exp( jβ sin ( j 2π f mt )}

23
Wide-band FM
x(t) is accordingly given by
 ∞ 
x(t ) = Ac Re exp ( j 2π f ct )
 ∑ J n (β ) exp( j 2π nf mt )
 n =−∞ 
∞
= Ac ∑ J n (β ) Re{exp[ j 2π ( f c + nf m )t ]}
n =−∞
∞
= Ac ∑ J n (β )cos( 2π ( f c + nf m )t )
n =−∞
The Fourier spectrum of x(t) is, therefore, given by

Ac ∞
X( f )= ∑ J n (β )[δ ( f − f c − nf m ) +δ ( f + f c + nf m )]
2 n=−∞
24
Properties of Bessel Functions
n
Property 1: J n ( β ) = (−1) J − n ( β )

Property 2:
For small values of the modulation index β, we have
J0 (β) ≅ 1
J1 (β) ≅ β/2

∞ 2
Property 3: ∑ J n (β ) =1
n =−∞
25
Table of Bessel Functions

• The first column gives the sideband number, while the first row gives the
modulation index.
• The cells indicate the amplitude values. Maximum is 1 and minimum 0.
• Sidebands with relative magnitude of less than 0.001 have been eliminated.
26
Table of Bessel Functions
• Some of the carrier and sideband amplitudes have negative
signs. This means that the signal represented by that amplitude
is simply shifted in phase 180° (phase inversion). As you can
see, the spectrum of a FM signal varies considerably in
bandwidth depending upon the value of the modulation index.
• The higher the modulation index, the wider the bandwidth of the
FM signal.
• To accommodate the use of different carrier amplitudes, it is
common to normalise the spectrum against Ac/2. Thus, the
amplitudes of the spectrum is directly in proportion to that of the
Bessel function.

27
Amplitude/Magnitude Spectrum

For convenient, only the right hand side of the spectrum is displayed 28
Amplitude Spectrum

•With the increase in the modulation index, the carrier amplitude

decreases while the amplitude of the various sidebands increases.

•With some values of modulation index, the carrier can disappear

completely.

•BT represents the bandwidth of the Wideband FM signal. 29

Power in Angle Modulated Signal
The power in an angle-modulated signal is easily computed
1 2 ∞ 2 1 2
P = Ac ∑ J n (β ) = Ac
2 n =−∞ 2
The power ratio Pr is defined as the ratio of the power
contained in the carrier component and the N components at
each side of the carrier to the total power in x(t).
1 2 N 2
Ac ∑ J n (β )
PN 2 n=− N N
P = = = J0 (β ) + 2 ∑ J n2 (β )
2
r P 1 2 n =1
Ac
2
30
Transmission Bandwidth of FM Signals
• Theoretically, a FM signal contains an infinite number
of side frequencies so that the bandwidth required to
transmit such signal is infinite.
• In general, there is no exact formula to calculate
bandwidth for FM signal. There are, however, 2 widely
used approaches:
i) By Bessel function
ii) By Carson’s rule

31
Bandwidth – Bessel Functions
• Since the values of Jn(β) become negligible for sufficiently large n, the
bandwidth of an angle-modulated signal can be defined by considering
only those terms that contain significant power.
• In practice, the bandwidth of a FM signal can be determined by
knowing the modulation index and using the Bessel function table.

32
N represents the number of significant sidebands
Bandwidth – Bessel Functions
• The bandwidth can then be determined with the simple formula
BT ( Bessel ) = 2 Nf max
where N is the number of significant sidebands.

Example
Calculate the bandwidth occupied by a FM signal with a
modulation index of 2 and a highest modulating frequency of
2.5 kHz using Bessel function table.
Solution
• Referring to the table for β=2, we can see that this produces six
significant pairs of sidebands.
BT ( Bessel ) = 2 × 6 × 2.5k = 30kHz
33
Bandwidth – Carson’s Rule
• An alternative way to calculate the bandwidth of a FM signal
is to use Carson's rule.
• This rule takes into consideration only the power in the most
significant sidebands whose amplitudes are greater than 2
percent of the carrier.
• These are the sidebands whose values are 0.02 or more.
• Carson's rule is given by the expression
BT (Carson) = BW ≅ 2∆f + 2 f m
where, ∆f is the maximum frequency deviation and fm is the
maximum modulating frequency
We may thus define an approximate rule for the transmission
bandwidth of an FM signal generated by a single frequency fm as
follows:
BT (Carson) = BW ≅ 2∆f + 2 f m = 2(β f m + f m ) = 2(β +1)34f m
Bandwidth – Carson’s Rule
Example
Calculate the bandwidth occupied by a FM signal with a
modulation index of 2 and a highest modulating frequency
of 2.5 kHz using Carson’s rule.

Solution
BWT (Carson ) = 2( β + 1) f m = 2(2 + 1)2.5k = 15kHz

Comparing the bandwidth with that computed in the preceding

example, you can see that Carson's rule gives a smaller bandwidth
while by Bessel table it gives a larger bandwidth.

35
Summary – Spectrum of an FM Signal
You should be able to:
• Determine the bandwidth of narrowband FM signals.
• Determine the bandwidth of wideband FM signals: use
table of Bessel functions and Carson's rule.
• Know basic properties of Bessel function of the first kind.
• Plot the amplitude spectrum of narrowband and wideband
FM signals.
• Calculate power of each spectral component in FM
signals.

36
Overview: Chapter 4

Frequency Modulation Systems

1. Angle Modulation
2. Spectrum of an FM Signal
3. FM Signal Generation
4. FM Demodulation
5. Noise in FM Signal
6. FM Broadcast Band and Specifications

37
FM SIGNAL GENERATION

• They are two basic methods of generating frequency-modulated

signals:
i) Direct Method
ii) Indirect Method
• The first category of modulator uses a voltage-controlled
oscillator.
• The second uses PM modulator to produce FM modulator, thus
called the indirect method. This is further categorized by
narrow-band or wide-band design.

38
DIRECT FM SIGNAL GENERATION
• In a direct FM system the instantaneous frequency is directly
varied with the information signal.
• To vary the frequency of the carrier, a linear frequency
modulator called Voltage-controlled Oscillator (VCO) is used.
• The VCO has an average frequency fc where the instantaneous
frequency is deviated by the message signal.
• A voltage variable reactance changes the oscillator frequency
due to the message signal amplitude (the voltage). kf is the
proportionality constant that determines the output frequency
based on the input message signal by
f i = f c + k f m(t )
Message VCO FM
signal signal
39
kf
DIRECT FM SIGNAL GENERATION

Message VCO FM
signal signal
kf
Advantage of direct method: Large frequency deviations
(broadband) is possible with the VCO. High frequency
sinusoids (carrier). The VCO is considerably linear throughout.
Disadvantage of direct method: The VCO is built from
components such as resistor, capacitors and inductors (the
reactance). These devices need to be properly calibrated to
produce consistent and stable sinusoids. At times, the frequency
produced by VCO tends to drift and becomes unstable.

40
DIRECT FM SIGNAL GENERATION
Example
A 1MHz VCO with sensitivity 5kHz/V is modulated by a 1V peak,
5kHz sinusoid. Find:
a) ∆f
b) Modulation index β
c) Express FM signal x(t) mathematically for a cosine carrier and
the cosine modulating signal. Let the carrier amplitude be 5V peak.
Solution
fc is 1MHz since it is 1MHz VCO.
a) ∆f=kfAm=5k(1)=5kHz.
b) β= ∆f/fm=5k/5k=1.
c) x(t)=Accos(2πfct+θ); θ=βsin(2πfmt))
x(t)=5cos(2π106t+sin(2π5x103t))
41
INDIRECT FM SIGNAL GENERATION
Recall that general angle-modulated signal is given by

x(t ) = Ac cos(2πf c t + θ (t ))
The phase angle θ(t) of phase modulation is proportional to the
modulating signal
θ (t) = 2πk pm(t )
The phase angle θ(t) of frequency modulation is

t
θ (t ) = 2π k f ∫0 m(τ )dτ
42
INDIRECT FM SIGNAL GENERATION
• Angle modulation includes frequency modulation (FM) and
phase modulation (PM).
• FM and PM are interrelated; one cannot change without the
other changing. The information signal frequency also
deviates the carrier frequency in PM.
• Phase modulation produces frequency modulation. Since the
amount of phase shift is varying, the effect is that, as if the
frequency is changed.
• Since FM is produced by PM, the later is referred to as
indirect FM.

43
INDIRECT FM SIGNAL GENERATION
Narrow-band PM Modulator

x ( t ) = Ac cos ( 2π f c t + θ ( t ) )
≈ Ac {cos ( 2π f ct ) − θ ( t ) sin ( 2π f c t )}
Modulator
θ (t ) = 2πk p m(t ) + narrowband
PM

+ 90
degree Ac cos(2πf ct )
Phase shifter Crystal oscillator

• The information signal is used to phase modulate a 90 degree phase

shifted carrier, which is then used to produce NB-PM signal. 44
INDIRECT FM SIGNAL GENERATION
Narrow-band FM Modulator
Modulator
θ (t ) = 2πk f m(t ) ∫ + narrowband
FM

Integrator + 90
degree Ac cos(2πfct )
Phase shifter Crystal oscillator

• The information signal is first integrated and then used to phase

modulate a 90 degree phase shifted carrier, which is then used to
produce NB-FM signal.
• For NB-PM and NB–FM, the modulation index is kept small in
order to minimize the distortion in the phase modulator.
45
INDIRECT FM SIGNAL GENERATION
Message signal m(t) Wide-band FM Modulator WB-FM signal

NB-FM
Narrow-band signal Frequency Frequency Frequency
FM Multiplier Converter Multiplier
Modulator xM xN
Crystal oscillator, fo
• The NB-FM signal produce by NB-FM modulator is multiplied
in frequency by means of frequency multiplier (to perform
narrow-band to wide-band conversion). The frequency deviation
∆f and carrier frequency fc of the new waveform is “M” times the
original values.
• The frequency converter converts the carrier frequency by
adding fo while maintaining the frequency deviation.
• The second frequency multiplier performs another frequency
multiplication by “N” times.
46
• Also called the Armstrong indirect FM modulator.
INDIRECT FM SIGNAL GENERATION
Example
a) Design an indirect WB-FM signal generator given:
NB-FM WB-FM
∆f = 5Hz ∆f = 7.5kHz
fc = 400kHz fc = 50MHz
For baseband spectrum 60 to 750Hz.
b) Find the max modulation index for both NB- and WB-FM block.
Solution
• Multiplication needed: 7.5k/5=1500.
• But, final fc=400kx1500=600MHz. This is difficult since we need only
50MHz. Thus, 1500 can be broken into 30x50.
• First frequency multiplier: M=30. Thus, ∆f = 150Hz and fc=12MHz.
• Frequency converter needs a crystal oscillator of fo=11MHz. Thus, at
output of the converter, fc=12-11=1MHz while ∆f = 150Hz.
• Second frequency multiplier: N=50. Thus, ∆f = 7.5kHz and
fc=50MHz. 47
• NB: β=5/60=0.083; WB: β=7.5k/60=125.
Summary – FM Signal Generation
You should be able to:
• Understand how to generate FM signals.
• Explain direct FM signal generation method.
• Explain indirect FM signal generation method.

48
Overview: Chapter 4

Frequency Modulation Systems

1. Angle Modulation
2. Spectrum of an FM Signal
3. FM Signal Generation
4. FM Demodulation
5. Noise in FM Signal
6. FM Broadcast Band and Specifications

49
FM DEMODULATION
The process of FM demodulation is the inverse of frequency modulation

• To perform frequency demodulation, we require a device that

produces an output signal with amplitude directly proportional
to the instantaneous frequency of the FM wave.
• They are two methods of designing a frequency demodulator:
i) Indirect Method
ii) Direct Method
• The first category of demodulator uses a phase-locked loop
(PLL) to match a local oscillator to the modulated carrier
frequency.
• The second of these employ discriminators, which are devices,
that discriminate one frequency from another by transforming
frequency changes into amplitude changes.
50
PHASED-LOCKED LOOP
• A Phase-Locked Loop (PLL) is a negative feedback system that
compares its output phase with the phase of a reference signal
and adjusts itself until both are aligned (or locked). Once the
loop is locked (the phase difference between the output and the
input signals is very close to zero), the desired signal may be
obtained.
• PLL is applicable to both analog and digital communication.

Vo (t )
X C (t )

X v (t )

Fig. Phase-locked loop FM detector

51
PHASED-LOCKED LOOP
• The input is assumed to be an FM modulated signal

xc ( t ) = Ac sin [ω c t + θ i (t ) ]
• The VCO is a sine wave generator whose frequency is
determined by a voltage applied to it from an external source.

ω v = ω c + g vV o , gv is the sensitivity of VCO

Vo (t )
X C (t )

X v (t )

Fig. Phase-locked loop FM detector 52

PHASED-LOCKED LOOP
xv ( t ) = Av cos ∫ ωv (τ ) dτ = Av cos ωc t + g v ∫ Vo (τ ) dτ 
  
t t

 0   0 
• The mixer (or phase detector) has output
xm ( t ) = xc (t ) ⋅ xv (t )

=
Ac Av  i (
sin θ (t ) − g t V τ dτ
v ∫0
o ( ) ) 

 

c (
2  + sin 2ω t + θ (t ) + g t V τ dτ
i v ∫0 o ( ) ) 

• At the output of the LPF, the sum term is suppressed.

53
PHASED-LOCKED LOOP
• The loop compares the phase of the input signal to the phase
of the signal at the output of the VCO.
• Feedback circuits are used to reduce the error term towards
zero.
• The error term is the difference between phase of the input
signal and a reference sinusoidal waveform.
• If the difference in phase is anything other than zero, the
output frequency of the VCO changes in a manner that forces
the difference toward zero.
• The phase deviation of the signal is thus
t
θi (t ) = g v ∫ Vo (τ ) dτ
0

54
 PHASED-LOCKED LOOP
• Since the input is an FM signal θi (t ) = 2π k f ∫ m(τ )dτ

• Thus the desired signal from FM signal is obtained

2πk f ∫ m(τ )dτ = g v ∫ Vo (τ )dτ
gv
m(t ) = Vo (t )
2πk f

• Similarly, if the input is an PM signal θ i (t ) = 2πk p m(t )

• Thus the desired signal from PM signal is obtained

2π k p m ( t ) = g v ∫ V o (τ ) d τ
gv
m (t ) =
2π k p ∫ V o (τ ) d τ
55
PHASED-LOCKED LOOP
• The PLL is the best frequency demodulator, because the filtering
action removes noise and interference and its linear output
reproduces the original modulating signal.
• The PLL has numerous applications in communication systems.
• Some of these are FM detection, frequency synthesis, and pulse
transmission synchronization.

56
DISCRIMINATOR
Slope Detector/FM Discriminator
• The ideal FM discriminator consists of a slope circuit
(differentiator) followed by an envelope detector.
• The slope circuit is a tank circuit where its resonant frequency fo is
slightly above (or below) the carrier frequency of the FM signal.
• The FM signal applied to the slope Voltage
circuit varies in amplitude as frequency Resonant
varies from the carrier frequency. The
amplitude increases as frequency
increases from the carrier and
decreases as frequency drops below
carrier. f
fc fo
Slope
circuit
57
DISCRIMINATOR
The output of the slope circuit is both amplitude as well as
frequency variation of FM signal.

Slope
circuit

The waveform is then passed to an envelope detector.

Envelop detector

58
DISCRIMINATOR

Demodulated
FM signal Slope Envelop detector signal
circuit

• The slope discriminator has disadvantage in which the FM signal

must be constant in amplitude.
• The amplitude of FM signal is affected by noise and interference
and will pass through the circuit.
• Another disadvantage is that the slope is linear over a small band
of frequencies, and thus causes distortion at the output.

59
BALANCED SLOPE DETECTOR
• The slope detector is rarely used because of poor linearity.
• The balanced slope detector provides better linearity than the
single slope detector.
• The upper and lower resonant filter sections of this circuit are
tuned above and below the carrier frequency fc signal, f1 and f2.
• Assume that both filters have a high quality factor (Q-factor).
• The Q-factor of a resonant circuit is a measure of goodness of
the whole circuit.

Demodulated
FM signal
signal

60
BALANCED SLOPE DISCRIMINATOR
• The linearity of the useful portion of the total response,
centered at fc, is determined by the separation of the two
resonant frequencies.

f1
f
f2 fc

61
LIMITER
• If noise or other interference perturbs the amplitude of the carrier,
then the carrier amplitude variations will cause distortion at the
output.
Distorted FM signal – amplitude not constant

• This distortion can be removed by passing through a limiter prior

to the slope detector. The output of the limiter is a square wave.

Limiter

62
LIMITER
• Limiter is a circuit that allows signals below a set of value to pass
unaffected while attenuating the peaks of stronger signals that
exceed this set value. It is used to suppress any amplitude
variations on the input signal, since these would distort the desired
output signal.

63
LIMITER
• The limiter output must then be converted to a sinusoid by
passing it through a band-pass filter (BPF) with center frequency
fc and sufficient bandwidth (correspond to the highest frequency
of the modulating signal fm).
• The result is a constant amplitude sinusoid.

BPF

• It can then be connected to the FM discriminator circuit.

Slope Envelope
Limiter BPF
Detector detector

64
Overview: Chapter 4

Frequency Modulation Systems

1. Angle Modulation
2. Spectrum of an FM Signal
3. FM Signal Generation
4. FM Demodulation
5. Noise in FM Signal
6. FM Broadcast Band and Specifications

65
NOISE IN FM SIGNAL

+ Noise
• Noise is essentially amplitude variations.
• An FM signal, on the other hand, has a constant carrier
amplitude.
• Because of this, FM receivers contain limiter circuits that restrict
the amplitude of the received signal.
• Any amplitude variations occurring on the FM signal are
effectively clipped off.
• This does not hurt the information content of the FM signal.
• Because of the clipping action of the limiter circuits, noise is
almost completely eliminated.
66
INTERFERENCE
• A major benefit of FM system is that interfering signals
produced by another modulating signal of almost the same
carrier frequency will be effectively rejected.
• However, interference suppression works well only when
interference is weaker than the desired signal.
• If the signal of one of the two is stronger, the receiver will lock
onto the stronger signal and eliminate the weaker signal.
• This phenomenon is known as the capture effect in FM.
• However, when the strengths of the two FM signals are nearly
the same, the capture effect alternates.
• You may have experienced this effect yourself when listening
to the FM radio in your car while driving on the highway.
67
NOISE IN FM SIGNAL
• Despite the fact that FM has superior noise rejection qualities,
noise still interferes with an FM signal.
• This is particularly true for the high-frequency components in the
modulating signal which generally has low amplitude.
• Since noise is primarily sharp spikes of energy, it contains a
considerable number of harmonics and other high-frequency
components.
PSD PSD
Flat noise power Parabolic noise power

f f
0 0
PSD of noise at output of PM receiver PSD of noise at output of FM receiver68
NOISE IN FM SIGNAL
PSD PSD decreases
• These high frequencies can at appreciably at higher
times be larger in amplitude than frequencies
the high-frequency content of the
modulating signal. f
0
PSD of typical message signal
High frequency
FM signal received at receiver noise spike
perturbed by interference and
high frequency noise

Interference is rejected after the

limiter circuit but not the high
frequency noise which has higher
amplitude that the message signal

The high frequency noise

appears after BPF
69
NOISE IN FM SIGNAL
• To overcome this problem most FM system use a technique
known as Pre-emphasis and De-emphasis.
• Pre-emphasis: higher frequency component of the message signal
are boosted, thus equalizing the PSD. Overall effect is increased
SNR.

PSD 20log|Hp (f)| PSD

f f f
0 0 0
Before: Message signal Pre-emphasis filter After: Message signal
frequency response 70
NOISE IN FM SIGNAL
• De-emphasis: the higher frequency component of the message
signal are attenuated, thus restoring the original message signal.
PSD 20log|Hd (f)| PSD

f f f
0 0 0
Output of FM receiver De-emphasis filter Original message signal
frequency response

71
PRE-EMPHASIS & DE-EMPHASIS

• At the transmitter the modulating signal is passed through a

simple network which amplifies the high frequency component
more than the low-frequency component.
• The simplest form of such circuit is a simple high pass filter.
• The pre-emphasis circuit increases the energy of the higher
content of the higher-frequency signals so that it will tend to
become stronger than the high-frequency noise component.
• This improves the signal-to-noise ratio (SNR). 72
PRE-EMPHASIS & DE-EMPHASIS

• To return the frequency response to its normal level, a de-

emphasis circuit is used at the receiver.
• This is a simple low-pass filter.
• The de-emphasis circuit provides a normal frequency response.
• The combined effect of pre-emphasis and de-emphasis is to
increase the energy of the high-frequency components of the
modulating signal during the transmission so that they will be
stronger and not masked by noise. 73
 and K=f2/f1

FM radio broadcast:
• In North America & Japan τ = 75µs
• In Europe and Australia τ = 50µs

74
PRE-EMPHASIS & DE-EMPHASIS
• Typical output SNR of an FM receiver without pre-emphasis is
~40-50dB.
• In commercial FM broadcasting, typically f1=2.12kHz and
f2=15khz or more. We may assume reasonably the bandwidth of
audio signal is at 15kHz (although audio spectrum: 20-20kHz,
any frequency higher than 15kHz is too faint to be heard).
• This set of values used for pre-emphasis filter yields an
improvement of ~13dB in the output SNR of the receiver over
no pre-emphasis.
• The PM systems do not require emphasis.

75
STEREOPHONIC FM BROADCASTING
• In FM broadcasting, FCC has assigned 88 to 108 MHz range,
with a separation 200kHz between adjacent stations and a peak
frequency deviation 75kHz.
• Until 1961, all commercial FM broadcasts were monophonic. In
1961, FCC authorized stereophonic system.
• In stereophonic system, FM stereo broadcast is designed to carry
two signals using the same carrier.
• This is a form of Frequency Division Multiplexing (FDM):
which involves assigning non-overlapping frequency ranges to
different signals/users of a medium. Multiplexing here is
a process where multiple analog signals are combined into one
signal over a shared medium.
• This is commonly used in FM radio broadcasting especially in
carrying two distinctive entities of a program, for example, in a
broadcast of a 2-piece orchestra, i.e. the string instruments and
the vocalist.
76
STEREOPHONIC FM BROADCASTING
• This provides a spatial dimension to its perception by a listener
at the receiving end. There are two factors which influence the
specification of standards for stereophonic FM broadcasting.
– Firstly, the transmission has to operate within the allocated FM broadcast
channels.
– Secondly, it must be compatible with the existing monophonic system.
• By operating within the allocated FM broadcast channel, this sets
the permissible frequency parameters and by ensuring the
compatibility, this sets the limitation in the configuration of the
transmitted signal.
• The two figures shown in the next page show the block diagram
of the transmitter of FM stereo and receiver of FM stereo,
respectively.
• Referring to the matrix network, let ml(t) and mr(t) be the signals
detected by the left (L) and right (R) side microphones at the
transmitting side of the system, respectively. 77
STEREOPHONIC FM BROADCASTING
ml(t)+mr(t)

Balanced
ml(t)-mr(t) modulator m(t) FM FM
DSB-SC modulator signal
Matrix network

Pilot carrier, fc=19kHz

m(t ) = ml (t ) + mr (t ) + [ml (t ) − mr (t )]cos(2π (2 f c )t ) + K cos(2π f c t )

FM stereo transmitter
78
STEREOPHONIC FM BROADCASTING
K cos(2π (2 f c )t ) × [ml (t ) − mr (t )]cos(2π (2 f c )t )
= K [ml (t ) − mr (t )]cos 2 (2π (2 f c )t )
K
Received = [ml (t ) − mr (t )][1 + cos(4π (2 f c )t ) ]
FM 2
signal

0-15kHz [ml (t ) − mr (t )]cos(2π (2 f c )t )

demodulator
FM

23-53kHz 0-15kHz
K cos(2π (2 f c )t ) Let the gain of
LPF be 2/K

K cos(2π f c t )
79
FM stereo receiver
STEREOPHONIC FM BROADCASTING
• These two signals will be added and subtracted at the matrix
network, which results in madd(t) = ml(t) + mr(t) and msub(t) =
ml(t) - mr(t). The madd(t) and msub(t) are left unprocessed in
baseband form, which is available for monophonic reception.
• The msub(t) and a subcarrier of 38kHz are applied to a product
modulator, hence producing a DSB-SC modulated signal.
• The madd(t), DSB-SC modulated signal, and a 19kHz pilot signal
(which is used to provide a reference for the coherent detection
of the stereo receiver) form the multiplexed signal given by,
m(t ) = [ml (t ) + mr (t )] + [ml (t ) − mr (t )]cos(4πf c t ) + K cos(2πf c t )

where fc = 19kHz and K is the amplitude of the pilot signal.

• The signal then frequency-modulates the main carrier to produce
the transmitted signal.
• At the receiver, m(t) is retrieved by frequency demodulating the
incoming FM signal.
80
STEREOPHONIC FM BROADCASTING
• Then, m(t) is fed into the demultiplexing system where the individual
components of m(t) are singled-out using three suitable filters.
• The recovered pilot signal is frequency doubled to produce wanted
38kHz subcarrier.
• With the subcarrier, the detection of DSB-SC modulated can be easily
detected using coherent detector, hence recovering msub(t) signal.
• The baseband LPF will pass the madd(t) signal.
• Lastly, the signals are reconstructed back at the matrix network; to the
left-side, ml(t) and right-side, mr(t) and applies them to the respective
speakers.
Pilot carrier Subcarrier

L+R L-R L-R

f (Hz)
50 15k 19k 23 38k 53k

FM baseband spectrum of stereophonic transmission 81

Overview: Chapter 4

Frequency Modulation Systems

• Angle Modulation
• Spectrum of an FM Signal
• FM Signal Generation
• FM Demodulation
• Noise in FM Signal
• FM Broadcast Band and Specifications

82
FM BROADCAST BAND AND
SPECIFICATIONS
• FM broadcasting - invented by Edwin Howard Armstrong that
uses frequency modulation (FM) to provide high-fidelity sound
over broadcast radio.
• Throughout the world, the broadcast band is 87.5 to 108.0 MHz
(or some portion thereof: in the U.S. it is 88.0 to 108.0 MHz).
Japan is the only exception, using exclusively its own unique 76
to 90 MHz band with 0.1 MHz channel spacing.
fc range: 88.1-91.9MHz (noncommercial)
92.1-107.9MHz (commercial)
fc spacing: 200kHz between stations
Bandwidth per station: 200kHz
FCC requirements
for FM broadcast in
Transmitter carrier stability: fc±2kHz
North American
fm range: 50Hz to 15kHz with 75 µs pre-emphasis.
Max. deviation: fc(max) = 75kHz (100% modulation)
83
FM BROADCAST BAND AND
SPECIFICATIONS
• In the former Soviet republics, and some Eastern Bloc nations, an
additional older band from 65.8 to 74 MHz is also used.
• Assigned frequencies are multiples of 30 kHz. This band, sometimes
referred to as the International Radio and Television Organisation
(OIRT) band is slowly being phased out in many countries.
• The frequency of an FM broadcast station (more strictly its assigned
nominal center frequency) is usually an exact multiple of 100 kHz. In
most of the Americas and Caribbean only odd multiples are used.
• In some of Europe, Greenland and Africa only even multiples are
used. In Italy, multiples of 50 kHz are used.
• There are other unusual and obsolescent standards in some countries
including 0.001, 0.01, 0.03, 0.074, and 0.3 MHz.
• All the bands are at VHF band of the radio communication band.
84
FM BROADCAST BAND AND
SPECIFICATIONS
Example
Compare the bandwidth by Bessel function table and Carson’s rule
for FM broadcast by radio based on FCC’s requirements for North
America.
Solution
• fm(max)=15kHz
• Thus, modulation index β=∆f(max)/fm(max) =75k/15k=5.
• From Bessel table, N=10: BT=2Nfm=2(10)(15k)=300kHz.
• From Carson’s rule: BT=2(β +1)fm =2(5+1)(15k)=180kHz.
• In practice, bandwidth of 200kHz is allocated to each FM
transmitter (or station), if we treat this as a reference, thus Bessel
function overestimate by 50% while Carson’s rule underestimate
by 10%.
85
FM BROADCAST BAND AND
SPECIFICATIONS
Example
Calculate the number of channels available for FM broadcast by
radio based on FCC’s requirements for North America. How does
one assign the carrier frequency?

Solution
• FM broadcast by radio transmits from 88 MHz to 108 MHz. The
bandwidth allocated for each station is 200 kHz and 200 kHz
spacing.
• There are thus, (108M-88M)/200k=100 channels.
• The carrier frequency is at the middle of the bandwidth of the
channel, or 100 kHz from the lower end of the channel.
• Thus, we have 88.1 MHz, 88.3 MHz , ..., 107.7MHz, 107.9 MHz.
86
Summary – FM Demodulation, and Noise in FM Signal
You should be able to explain:
• how to demodulate FM signals.
• the functions of phased-locked loop.
• the functions of discriminator.
• the functions of limiter.
• why do we need pre-emphasis and de-emphasis.
• what is stereophonic FM broadcasting.

87
• References
– B.P. Lathi, Modern Digital and Analog Communication
System, 3rd Ed., Oxford University Press, 1998.
– Simon Haykin, Communication Systems, 4th Ed., John
Wiley & Sons, 2001.
– W. Tomasi, Electronic Communications Systems:
Fundamentals Through Advanced, 5th Ed. Upper
Saddle River, New Jersey: Prentice-Hall, 2004.
– Paul H. Young, Electroni Communication Techniques,
4th Ed. Upper Saddle River, New Jersey: Prentice-Hall,
1999.
88