Angle Modulation (Phase & Frequency Modulation)

EE442 Lecture 7
Spring Semester

Angle Modulation

1
Applications for Various Modulation Techniques

We have studied AM, next is FM (and PM).

2
Amplitude, Frequency and Phase Modulation

With few exceptions,

Phase Modulation (PM)
is used predominantly in
digital communication

3
 Illustrating AM, PM and FM Signals

Carrier signal
Carrier Wave

m(t)

Modulating Signal m(t)

AM

AM Modulated Signal

PM
Angle PM Modulated Signal
Modulation
FM

FM Modulated Signal time

4
 Focus Upon an FM Signal Modulated by a Single-Tone
m(t )
Single-tone modulating signal
BASEBAND
SIGNAL
SIGNAL
FM

5
 Some Observations on FM and PM Waveforms

1. Both FM and PM waveforms are identical except for a phase shift.

2. For FM, the maximum frequency deviation occurs when modulating

signal is at its peak values (i.e., at + mp and – mp).

3. For PM, the maximum frequency deviation takes place at the zero
crossings of the modulating signal m(t).

4. It is difficult to know from looking at a waveform whether the

modulation is FM or PM.

6
 Advantages of Angle Modulation

1. Angle modulation is resistant to propagation-induced selective fading

because amplitude variations are unimportant.

2. Angle modulation is very efficient in rejecting interference (i.e., it

minimizes the effect of noise on the signal transmission). Refer to
slide 10 below.

3. Angle modulation allows for the more efficient use of transmitter power.

4. Angle modulation is capable of handling a greater dynamic range of

modulating signal without distortion as occurs in AM modulation.

5. Wideband FM gives significant improvement in the signal-to-noise ratio

at the output and is proportional to the square of the modulation
index.

7
Phase-Frequency Relationship When Frequency is Constant
 (t )  AC cos( (t ))
(t) is generalized angle

 (t )  AC cos(C t   0 )

 (t )

C t   0
 0 is constant

d (t )
Slope: i (t )   C
dt t t
i
0
time t

8
 Concept of Instantaneous Frequency
 (t )  AC cos( (t ))
Angle
Modulation (t) is generalized angle

 (t )  AC cos(C t   0 )

 (t )

 (t ) C t   0
 0 is constant

d (t )
Slope: i (t )   C
dt t t
0 i

ti time t
9
 Angle Modulation Gives PM and FM

d (t )
t
i ( t ) 
dt t t
and  (t )    ( )d

i
i

Angle
Modulation

Phase Frequency
Modulation Modulation

Frequency modulation and phase modulation are closely related!

10
 Comparing Frequency Modulation to Phase Modulation

# Frequency Modulation (FM) Phase Modulation (PM)

1 Frequency deviation is proportional to Phase deviation is proportional

modulating signal m(t)
2 Noise immunity is superior to PM (and of
course AM)
3 Signal-to-noise ratio (SNR) is better than
PM (and of course AM)
4 FM is widely used for commercial
broadcast radio (88 MHz to 108 MHz)
5 Modulation index is proportional to
modulating signal m(t) as well as the
modulating frequency fm
to modulating signal m(t)
Noise immunity better than
AM, but not FM
Signal-to-noise ratio (SNR) is
not quite as good as with FM
PM is primarily used for mobile
radio services
Modulation index is
proportional to modulating
signal m(t)

11
FM has superior noise immunity compared to AM

FM has better noise (RFI)

rejection than AM, as shown in
this dramatic New York
publicity demonstration by
General Electric in 1940. The
radio has both AM and FM
receivers. With a million-volt arc
as a source of interference
behind it, the AM receiver
produced only a roar of static,
while the FM receiver clearly
reproduced a music program
from Armstrong's experimental
FM transmitter W2XMN in New
Jersey.

https://en.wikipedia.org/wiki/Frequency_modulation

Note: RFI stands for radio frequency interference.

12
 Phase Modulation (PM)
 i (t )  C t   0  k p m(t ) ; Generally we let  0  0,

 PM (t )  AC cos(C t  k p m(t )) Agbo & Sadiku

Section 4.2; p. 159

The instantaneous angular frequency (in radians/second) is

d i (t ) dm(t )
i (t )   C  k p  C  k p m '(t )
dt dt
In phase modulation (PM) the instantaneous angular frequency
i varies linearly with the time derivative of the message signal
m(t) [denoted here by m’(t)].

kp is phase-deviation (sensitivity) constant. Units: radians/volt

[Actually it is radians/unit of the parameter m(t).]

13
 Frequency Modulation (FM)
But in frequency modulation the instantaneous angular frequency
i varies linearly with the modulating signal m(t),
i (t )  C  k f m(t )
t t
 i (t )   

C  k f m( ) d  C t  k f  m( ) d


kf is frequency-deviation (sensitivity) constant. Units: radians/volt-sec.

Then
 t

FM (t )  AC cos  C t  k f  m( ) d  Agbo & Sadiku
Section 4.2; p. 159
  

FM and PM are similar to each other.

For PM the angle is directly proportional to m(t).
For FM the angle is directly proportional to the integral  m(t )dt .

14
 Summary
Message signal is m(t)
d i (t )
Definition: Instantaneous frequency is i (t ) 
dt

Phase Modulation Frequency Modulation

t
 i (t )  C t  k p m(t )
Angle  i (t )  C t  k f  m( ) d


dm(t )
Frequency i  C  kp i  C  k f m(t )
dt

In phase modulation m(t) drives the time variation of phase i.

In frequency modulation m(t) drives the time variation of frequency fC.

15
 A Pictorial View of FM and PM Generation

H(j) = 1/j
m(t ) FM (t )

Phase
t
Modulator
 m( ) d

Generation
Agbo & Sadiku
of FM
Frequency Modulator AC cos(C t )
Figure 4.1
p. 160
H(j) = j
m(t ) d Frequency
PM (t )
dt dm(t ) Modulator
dt Generation
of PM
Phase Modulator AC cos(C t )

We require that H(j) be a reversible (or invertible) operation

so that m(t) is recoverable.

16
FM and PM Generation are Nonlinear Processes

Consider a phase modulated signal:

Let s (t )  AC cos C t  k p [m1 (t )  m2 (t )]
If s1 (t )  AC cos C t  k p m1 (t )  , and
s2 (t )  AC cos C t  k p m2 (t ) 
Then obviously s1 (t )  s2 (t )  s (t )
So PM can't be linear (nor can FM).

17
 Modulation Index  for Angle Modulation
Let the peak values of the message signal m(t) and its first derivative m’(t) be
represented by
Peak value of m(t) = mp = ½(mmax – mmin)
Peak value of m’(t) [= dm(t)/dt] = m’p
Frequency Deviation is the maximum deviation of the instantaneous
modulated carrier frequency relative to the unmodulated carrier frequency.
It is symbolically represented by either  or f.
k f mp
FM:   k f m p or f 
2
k p mp
PM:   k p mp or f 
2
The ratio of the frequency deviation f to the message signal’s bandwidth B
Is called the Frequency Deviation Ratio or the Modulation Index, and is
denoted by  (unitless).
f 
 
B 2 B
18
Equations for FM Wave with Single-Tone Modulation
Carrier signal AC cos(C t ) (volts)
Carrier frequency C  2 fC (radians/sec)
Modulating wave m(t ) Am cos(mt ) Single-tone modulation
Modulating frequency m  2 f m (radians/sec)
Deviation sensitivity k f (radians/volt-second)
Frequency deviation   k f Am (radians/sec)
f  k f Am
Modulation Index    (unitless)
f m m m
A
Instantaneous frequency f i  fC  k f m cos(mt )  fC  f cos(mt )
2
  t  
Remember FM (t )  AC  cos  C t  k f   m(  )d    , generally
  
     
  k f Am 
Tone modulated wave FM (t )  AC  cos  C t  sin(mt )  
  m  
or FM (t )  AC  cos C t   sin(mt )  
19
 Summary of Mathematical Equations for FM and PM

Type of Modulating Signal Angle Modulated Wave

Modulation

AC  cos C t  k p m(t ) 

Phase m(t)
modulation
Frequency
modulation
m(t)
 t
AC  cos C t  k f  m( )d 
 
Am cos(mt)
AC  cos C t  k p Am cos(C t ) 
Phase
modulation
Frequency Am cos(mt)  k f Am 
modulation AC  cos  C t  sin(C t ) 
 m 
k f Am

m

20
 Example

FM (t )  10  cos  2 (10 6 )t  8 sin(2 (10 3 )t )  

f
FM (t )

21
 Solution to Example
Start with the basic FM equation:
FM (t )  AC  cos  2 fC t   sin(2 f mt )  
Compare this to
FM (t )  10 cos  2 (106 )t  8sin(2 (103 )t ) 

(a) We see by inspection that fC = 1,000,000 Hz & fm = 1000 Hz.

(b) The modulation index is  = 8.
(c) The peak deviation frequency f is
f    f m  8  1000  8, 000 Hz

Note: f /fC is 0.008 or 0.8 % deviation frequency to carrier

frequency.

22
 Average Power of a FM or PM Wave
The amplitude AC is constant in a phase modulated or a frequency
modulated signal. RF power does not depend upon the frequency
or the phase of the waveform.

FM or PM (t )  AC cos C t  g( kk , m(t ))

AC2
Average Power  (always)
2

This is a result of FM and PM signals being constant amplitude.

Note: kk becomes kf for FM and kp for PM.

23
 Average Power of a FM or PM Wave

Problem:

Consider an angle modulated signal given by

 (t )  6  cos  2 106 t  2  sin(8000 t   volts

What is the average power of this signal?

Solution:
AC2
Average power  PC  where AC  6 volts
2
62 36
Therefore, PC    18 watts (assumes 1 ohm resistance)
2 2
Note that the result does not depend upon it being FM or PM.

24
 Comparison of FM (or PM) to AM
# Frequency Modulation (FM)
1 FM receivers have better noise
immunity
2 Noise immunity can be improved
by increasing the frequency
deviation
3 Bandwidth requirement is greater
and depends upon modulation
index
4 FM (or PM) transmitters and
receivers are more complex than
for AM
5 All transmitted power is useful so
FM is very efficient
Amplitude Modulation (AM)
AM receivers are very susceptible to
noise
The only option in AM is to increase
the transmission power
AM bandwidth is less than FM or PM
and doesn’t depend upon a
modulation index
AM transmitters and receivers are less
complex than for FM (or PM)
Power is wasted in transmitting the
carrier and double sidebands in DSB
(but DSB-SC & SSB addresses this)
25
 AM, FM and PM Waveforms for Single-Tone m(t)

Carrier signal
Carrier Wave

m(t)

Modulating Signal m(t)

Review:
AM

AM Modulated Signal

PM
Angle PM Modulated Signal
Modulation
FM

FM Modulated Signal time

26
 FM and PM Examples
Sketch FM and PM waveforms for the modulating signal m(t). The constants
kf and kp are 2  105 and 10, respectively. Carrier frequency fc = 100 MHz.
m(t)
m(t) m’(t)
m’(t)

FM PM

kf kp
f i  fC  m(t )  1  10 8  1  10 5  m(t ); f i  fC  m '(t )  1  10 8  5  m '(t );
2 2
mmin  1 and mmax  1 m 'min  20, 000 and m 'max  20, 000
 f i min  108  105 ( 1)  99.9 MHz,  f i min  108  5( 20, 000)  99.9 MHz,
 f i max  108  105 ( 1)  100.1 MHz  f i max  108  5( 20, 000)  100.1 MHz
27
 Frequency Shift Keying is Related to FM
Sketch the FM waveform for the modulating signal m(t). The constant kf is
2  105. Carrier frequency fc = 100 MHz.

m(t)

FM

kf
f i  fC  m(t )  1  10 8  1  10 5 m(t )
2

Since m(t) switches from +1 to -1 and vice versa, the FM wave

frequency switches between 99.9 MHz and 100.1 MHz. This is called
Frequency Shift Keying (FSK) and is a digital communication format.
28
 Example – continued
Sketch the PM waveform for the modulating signal m(t) from prior slide.
The constant kp equals /2. Carrier frequency fc = 100 MHz.

m’(t)
m’(t) kp 1
f i  fC  m'(t )  1  10 8  m'(t )
2 4
PM This is carrier PM by a digital signal
– it is Phase Shift Keying (PSK)
because the digital data is
represented by phase of the
carrier wave.

  
 PM (t )  AC cos C t  k p m(t )  AC cos C t 
m(t )
 2 
 PM (t )  AC sin(C t ) when m(t )  1
 PM (t )   AC sin(C t ) when m(t )  1

29
 Generalized Angle Modulation
Agbo & Sadiku; Section 4.3 on pages 166 to 167
Start with equation (4.8) on page 159, which is
t
 A (t )  AC  cos[C t  k   (t )] where  (t )  m(t )  h(t )   m( )h(t   )d 


and h(t )   (t ) for PM; h(t )  u (t ) for FM

Suppose we use an exponential carrier AC e jC t instead of AC cos(C t ),
The generalized angle modulation becomes
 A (t )  AC  e j ( t  k (t ))  AC  e j ( t )  e jk (t )
C C

and k  k p for PM; k  k f for FM

PM (t )  Re  A (t )   Re  AC  e j t  e
jk p  ( t )
C  ; where  (t )  m(t )   (t )  m(t )

and
t
FM (t )  Re  A (t )   Re  AC  e
jk f  ( t )
 ; where  (t ) 
 m(  ) d 
jC t
e 


30
 Generalized Angle Modulation (continued)

Consider first Frequency Modulation,

  j n k n  n (t ) 
FM (t )  Re  A (t )   Re  AC  e 
jC t f

 n 0 n ! 
 jC t
 k 2f  2 (t ) jk 3f  3 (t )  
FM (t )  Re  AC  e 1  jk f  (t )      
  2! 3!  
 
Now take the real part of the expression above,
 k 2f  2 (t ) 
FM (t )  AC cos(C t )  k f  (t ) sin(C t )  cos(C t )   
 2! 
Note: m(t ) has a bandwidth = B Hz and  (t) has a bandwidth = B Hz,
but  n (t ) has a bandwidth = nB Hz; as n  , bandwidth  

The instantaneous frequency deviations are symmetrical about

carrier frequency C, thus FM is double side-banded. The
effective FM bandwidth = 2nB Hz.
31
 Generalized Angle Modulation (continued)

Consider case where kf is small meaning that |kf (t)| << 1.

It is commonly referred to as narrowband FM (NBFM). We take
only the first two terms in the expansion for FM(t).

FM (t )  AC cos(C t )  k f  (t ) sin(C t ) 

t 
FM: FM (t )  AC  cos(C t )  AC k f   m( )d   sin(C t ) Equation (4.15)
  
By analogy, we can apply parallel analysis for Phase Modulation.
For PM if kp is small meaning that |kp (t)| << 1. This is known
as narrowband PM (NBPM).
PM: PM (t )  AC  cos(C t )  AC k p  m(t )   sin(C t ) Equation (4.16)

32
Generation of Narrowband FM and PM

m(t) + NBFM
 kf 
+
-ACsin(ct)
/2

ACcos(ct)

m(t) + NBPM
kp 
+
-ACsin(ct)
/2

ACcos(ct)

Agbo & Sadiku; Figure 4.5 on page 168

33
Modulation Index  for Angle Modulation (continued)
Parameter  is the modulation index for angle modulation.

 is used to differentiate between narrowband angle modulation and

wideband angle modulation.
Narrowband angle modulation requires  << 1 (Typically < 0.3)
Wideband angle modulation requires  >> 1 (Typically > 5.0)
Equivalently,

Narrowband angle modulation requires f << B

Wideband angle modulation requires f >> B
Comments:
1. Narrowband FM has about the same bandwidth as that of AM.
2. Commercial (broadcast) FM is wideband FM (because of its
superior noise performance).
3. Why even consider narrowband FM? Two reasons:
a. NBFM is easier to generate that WBFM.
b. It is used as in initial step in generating WBFM.

34
 NBFM and NBPM with Tone Modulation

k f mp k f Am
Let m(t )  Am cos(mt ), then m p  Am ; m  2 B; and   
m m
t t
k f Am
Then k f  m(  ) d   k  A

f

m cos(m  )d  
m
 sin(mt )

The time-domain NBFM signal is

1
FM (t )  AC cos(C t )   AC sin(mt )  sin(C t )
2
1 1
FM (t )  AC cos(C t )   AC cos  (C  m )t    AC cos  (C  m )t 
2 2
Comparing to AM:
The 2nd term is the upper sideband and the 3rd term is the lower sideband.

35
 Narrowband FM (NBFM)
1 1
FM (t )  AC cos(C t )   AC cos  (C  m )t    AC cos  (C  m )t 
2 2

AC

Tone modulation  cos(mt)

 = 0.2

(C - m) C (C + m) 

NBPM requires  << 1 radian

(typically less than 0.3 radian)

36
Review: Phasor Interpretation of AM DSB with Carrier
C rotates faster than m

us cos(Ct)

C

cos(mt)
ls
m = |us| = |ls|

Spectrum: DSB AM


C - m C C + m
lower upper
sideband sideband
37
 Narrowband FM Example (Example 4.4)
Exercise: The message signal input to a modulator is m(t) = 4cos(2104t)
and the carrier is 10cos(108t). If frequency modulation is performed
with kf = 1000, verify that the modulated signal meets the criteria of
being narrowband FM. Also, obtain an expression for its spectrum and
sketch this spectrum.

Solution:
First we calculate the modulation index 
Am  4 
  kf  1000  4 
 0.2;   0.3  NBFM
m  2  10 
1 1
AC  10 thus,  AC  (0.2)(10)  1
2 2
We use the equation on the previous slide (slide #31):
1 1
FM (t )  AC cos(C t )   AC cos  (C  m )t    AC cos  (C  m )t 
2 2
FM (t )  10  cos(C t )  cos  (C  m )t   cos  (C  m )t 

38
 Narrowband FM Example (Example 4.4 continued)
FM (t )  10  cos(C t )  cos  (C  m )t   cos  (C  m )t 
The corresponding expression for the specturm becomes
 FM ( )  10  (  C )   (  C )     (  C  m )   (  C  m ) 
   (  C  m )   (  C  m ) 
where C  108  radians/sec and m  2 104 radians/sec

 FM ( )
10
Bandwidth = 2m


-C - C 
(-C - m) (-C + m) (C - m) (C + m)

39
 Wideband FM (WBFM)
WBFM requires  >> 1 radian (much more complicated)
For wideband FM we have a nonlinear process, with single tone
modulation:
FM
WB
(t )  Re  AC exp  jC t  j  sin(mt )  
We need to expand the exponential in a Fourier series in order to
analyze FM
WB
(t ). Modulation Index
 f f
WB
(t )  AC  Jn (  )  cos  2 ( f  nf m )t   
FM
n 
C
fm B
where the coefficients J n (  ) are Bessel functions.

Spectral analysis for tone modulation of WBFM: Agbo & Sadiku, pp. 171-180.
We will not cover this section in EE442 but rather focus upon the physical
interpretation of FM spectrum spread.

40
WBFM (or WBPM) Requires Much More bandwidth Than AM
AC A
t f

fC
Am Carrier
CarrierSignal
Signal(frequency fc )fC)
(frequency A
t f

fm
Message Signal (frequency
(frequency ffmm)) A
A
AM
t f

A WBFM
A Amplitude
AmplitudeModulated
ModulatedSignal
Signal
t f

Frequency
FrequencyModulated
Modulated(FM)
(FM)Signal
Signal
41
FM Spectra as Function of Modulation Index 
 Number of Bandwidth
Sidebands¶
NBFM 0.1 2 2 fm
 = 0.2
0.3 4 4 fm
0.5 4 4 fm
1.0 6 6 fm

 = 1.0 2.0 8 8 fm
5.0 16 16 fm
10.0 28 28 fm

=5 Single-tone
Modulation Index
f 
 = 10  
f m m

BT or BW

42
 Spectra of FM Signals

Single-tone A  = 0.2 A
Modulation Index
f f
f 
   = 1.0
f m m

f increasing & f is constant &

f m is constant =5 f m is decreasing

 = 10

From A. Bruce Carlson, Communication Systems, An Introduction to Signals and Noise in

Electrical Communication, 2nd edition, 1975; Chapter 6, Figure 6.5, Page 229.

43
 Selecting an FM Station
Broadcast FM Radio covers from 88 MHz to 108 MHz
100 stations – 200 kHz spacing between FM stations

Note: 0 dBu = 0.775 volt into 600 ohms

(which is equivalent to 1 mW power
delivered into the 600 ohm resistor)

Service Type Frequency Channel Maximum Highest

Band Bandwidth Deviation Audio
Commercial FM Radio 88.0 to 108.0 200 kHz 75 kHz 15 kHz
Broadcast MHz
44
Measured Spectrum of an FM Radio Signal

Voice modulation

200 kHz

noise

45
 Specifications for Some Commercial FM Transmissions

Service Type Frequency Channel Maximum Highest

Band Bandwidth Deviation Audio
Commercial FM Radio 88.0 to 108.0 200 kHz 75 kHz 15 kHz
Broadcast MHz
Television Sound 4.5 MHz 100 kHz 25 kHz 15 kHz
(analog) above the monaural &
picture carrier 50 kHz
Digital TV has replaced frequency stereo
Public safety – Police, 50 MHz and 20 kHz 5 kHz 3 kHz
Fire, Ambulance, Taxi, 122 MHz to
Forestry, Utilities, & 174 MHz
Transportation
Amateur, CE class A & 216 MHz to 15 kHz 3 kHz 3 kHz
Business band Radio 470 MHz

46
Three Important Frequencies for FM and PM

The three important frequencies in FM and PM are

1. Carrier frequency fC (or C)
2. Maximum modulation frequency fm (or m), and
3. Peak frequency deviation f (or )

Two Definitions of importance:

1. Modulation index 
f f  
    (can be a very large number)
Bm f m m 2 Bm
2. Deviation ratio D
f 
D  (always much less than unity)
fC C
Remember: For FM  = kf mp & for PM  = kp m’p

47
 FM Bandwidth and the Modulation Index 
A. Narrowband FM (NBFM) –  << 1 radian
NB
BFM  2 Bm where Bm is the bandwidth of m(t )

B. Wideband FM (WBFM) –  >> 1 radian

f f 
WB
BFM  2(   1)Bm , where    
Bm f m m
f is the peak frequency deviation f  max  k f m(t )
WB
BFM  2( f  Bm )  2(   1)Bm  Carson's Rule

For PM we have analogous equation,

WB
BPM  2(   1)Bm

48
 Example: Bandwidth of FM Signal
The message signal input to a modulator is 10cos(2  104t). If frequency
modulation with frequency deviation constant kf = 104  is performed, find
the bandwidth of the resulting FM signal.

Solution:
1 k f Am 104  10
  5
2 f m 2 10 4

BFM  2    1 f m  2  5  1 10 kHz  120 kHz

Using Carson's rule to calculate bandwidth BFM

49
 Example: Equal Bandwidth for FM & PM Signals
If phase modulation is performed using the message signal 10cos(2  104t)
used in the previous slide, find the phase deviation constant kp giving the
PM signal the same bandwidth, namely, 120 kHz.

Solution:
For both the FM and PM signals to have the same bandwidth,  and f must
be the same. For FM,  = kf Am; but for PM,  = kp m’p .
Expressing the message signal m(t) = Am cos(mt) gives

d
m '(t )   Am cos(mt )   m Am  sin(mt )  m ' p  m Am
dt
Thus,
104 
kf 1
k f Am  k pm Am  kp    
m 2 104 2
k pm 'p k pm Am 1
Check:  =   k p Am  (10)  5
m m 2

50
 Example: Commercial FM Radio Stations
For commercial FM radio, the audio message signal has a spectral range of
30 Hz to 15 kHz, and the FCC allows a frequency deviation of 75 kHz.
Estimate the transmission bandwidth for commercial FM using Carson’s
Rule.

Solution:
We start by calculating 
f 75 kHz
  5
Bm 15 kHz
Using Carson's rule gives
BFM  2    1 Bm  2  5  115 kHz  180 kHz
The allowed bandwidth for commercial FM is 200 kHz.
Note that Carson's rule slightly underestimates the
bandwidth.

51
 Why Does FM and PM Take Much More Bandwidth?

Observation: The bandwidth required for AM and NBFM are the same.

However, WBFM (wideband FM) requires much more bandwidth. Why?

A Fourier spectrum of an FM signal shows that to keep the amplitude
constant of an FM signal that many components are required to represent
the FM waveform. The frequency spectrum of an actual FM signal has
components extending infinitely, although their amplitude decreases for
sufficiently higher frequencies. Sufficiently higher frequencies applies
to frequencies above the Carson bandwidth rule.

WB
BFM  2( f  Bm )  2(   1)Bm  Carson's Rule
Next we examine the Fourier components this using phasors.

52
 FM Tone-Modulated Signal Spectrum
Review
Review:
For  = 2.0

Note: Only magnitudes of spectral lines shown.:

53
NBFM Constructed From Phasors in FM Modulation

NBFM with tone modulation

(t)
 AC  AC
2 2

fC AC
fm -fm
AC

 AC  AC
2 2

0 fC - fm fC fC + f m

54
 Sidebands Constructed From Phasors in FM Modulation
Animation showing how phase modulation works in the phasor picture -- phase
modulation with a sinusoidal modulation waveform and a modulation depth of π/4
radians. The blue line segments represent the phasors at the carrier and the
harmonics of the modulation frequency.

55
 Generating FM Signals

There are two basic methods to generate FM:

1. Direct method (uses voltage-controlled oscillator to vary the

the frequency linearly with the message signal m(t))
Advantage: Can generate large frequency deviation.
Disadvantage: Carrier frequency tends to drift and
must be stabilized.

2. Armstrong’s indirect method (first generate NBFM with the

message signal with a small frequency deviation and then
frequency multiplication is used to increase the frequency
and frequency deviation to desired levels (generates WBFM)

Advantage: More stable carrier frequency.

Disadvantage: More complex hardware and cost.

56
Direct Generation of FM Signal Using a VCO
VCO is “voltage-controlled oscillator”

+VCC

Varactor diode
RFC

VCO
Q

m(t)  FM (t )

CD
1
osc
LC eq LC Resonator

Ceq is capacitance CD plus capacitance of other capacitors.

57
Narrowband FM Generated by Pulling a Crystal Oscillator
+VCC
R3
Varactor diode  FM (t )

Xtal
m(t)
Q1
A crystal is an
R2 CD R4 electro-mechanical
resonator.

A crystal filter is placed in the feedback loop to stabilize the oscillator.

The frequency of oscillation can be pulled slightly from the high-Q
crystal resonator’s frequency. The frequency deviates only slightly and
is typically only up to about 100 ppm. However, the oscillator is very
stable for m(t) = 0.

58
Generation of Narrowband Frequency Modulation (NBFM)

 t

FM (t )  AC cos  C t  k f  m( ) d 
  
NBFM requires  << 1 radian
DSB-SC
modulator
m(t) + NBFM
 kf 
+
-ACsin(ct)
/2 Oscillator

ACcos(ct)
Carrier
Agbo & Sadiku
Figure 4.5; page 168

59
Indirect Generation of FM Using Frequency Multiplication

In this method, a narrowband frequency-modulated

signal is first generated and then a frequency
multiplier is used to increase the modulation index.
The concept is shown below:

m(t ) FM
NB
(t ) FM
WB
(t )
Frequency
NBFM Multiplier

A frequency multiplier is used to increase both the

carrier frequency and the modulation index by integer N.

60
 Frequency Multipliers
A frequency multiplier is a nonlinear component followed by a bandpass
filter at the multiplied frequency desired.

in (t ) y (t ) out (t )
Section 4.4 Nonlinear Bandpass
Page 181 of Device Filter @ nC
Agbo & Sadiku
We select the nth order nonlinear component of y(t) and pass it through
the bandpass filter.

 t

in (t )  AC  cos C t  k f  m( )d   , and
 0  Note: m(t) is
not distorted
 t

out (t )  AC  cos  nC t  nk f  m( )d   by multiplier.
 0 

Conclusion: Carrier frequency is now nfC and frequency deviation is now nf.
Commercial frequency multipliers are generally 2 and 3.

61
 Armstrong Indirect FM Transmitter Example
Crystal stabilized
voltage-controlled fC 1  200 kHz
fC 2  12.8 MHz
oscillator f1  25 Hz
m(t ) f2  1.6 kHz
NBFM  64
generation Multiplier
 (t )
NB
FM

fC 3  1.9 MHz
FM
WB
(t ) f3  1.6 kHz
 48
PA Multiplier BPF
fC 4  91.2 MHz
A mixer
f4  76.8 kHz does not
change f

Crystal Oscillator
These numbers correspond
to an FM broadcast radio station.

62
 Why are Two Multiplication Chains Used?

FM
NB
(t )
Mixer  WB
FM (t )
NBFM Multiplier Multiplier
generator Chain A Chain B

Oscillator

63
 Many Ways to Perform Frequency Multiplication

In electronics, a frequency multiplier is an electronic circuit that

generates an output signal whose output frequency is a
harmonic (multiple) of its input frequency. Frequency multipliers
consist of a nonlinear circuit that distorts the input signal and
consequently generates harmonics of the input signal.

Most multipliers are doublers or triplers

64
Frequency Multiplication Using Comb Generation
From our discussion on Fourier series and pulse trains:

Tp

Amplitude

Comb frequencies shown

1 1 2
T Tp Tp

65
 Simple Comb Generator
A step recovery diode (SRD) is a p-n junction diode having the ability to generate
extremely short pulses. It is also called snap-off diode or charge-storage diode,
and has a variety of uses in microwave electronics (e.g., pulse generator or
parametric amplifier).

Comb
Generator
Circuit

https://www.edn.com/electronics-blogs/the-emc-blog/4402169/DIY-6-GHz-comb-generator
66
Step Recovery Diode Based Comb Generation

Volts

Time (nanoseconds)
The key to generating a wide comb of frequencies is to
generate very narrow pulses which step recovery diodes
are designed to do.

http://www.mwrf.com/analog-semiconductors/designing-
step-recovery-diode-based-comb-generator
67
Generation of Narrowband Phase Modulation (NBPM)

 PM (t )  AC cos(C t  k p m(t ))

m(t) + NBPM
kp 
+
-ACsin(ct)
/2
ACcos(ct)

Agbo & Sadiku

Figure 4.5; page 168

68
Generation of Phase Modulation

Carrier
frequency fC

+VD

m(t)  PM (t )
kp
CD
Varactor diode

69
 Advantages of FM
Advantages of frequency modulation
1. Resilient to noise: The main advantage of frequency modulation is a reduction in
noise. As most noise is amplitude based, this can be removed by running the
received signal through a limiter so that only frequency variations remain.

2. Resilient to signal strength variations: In the same way that amplitude noise can
be removed, so too can signal variations due to channel degradation because it does
not suffer from amplitude variations as the signal level varies. This makes FM ideal
for use in mobile applications where signal levels constantly vary.

3. Does not require linear amplifiers in the transmitter: As only frequency changes
contain the information carried, amplifiers in the transmitter need not be linear.

4. Enables greater efficiency : The use of non-linear amplifiers (e.g., class C and class
D/E amplifiers) means that transmitter efficiency levels can be higher. This results
from linear amplifiers being inherently inefficient.

70
 Disadvantages of FM
Disadvantages of frequency modulation
1. Requires more complicated demodulator: One of the disadvantages is that
the demodulator is a more complicated, and hence more expensive than the very
simple diode detectors used in AM.

2. Sidebands extend to infinity either side: The sidebands for an FM

transmission theoretically extend out to infinity. To limit the bandwidth of the
transmission, filters are used, and these introduce some distortion of the signal.

71
 Ideal FM Differentiator Demodulator
The ideal FM detector converts the FM signal‘s instantaneous frequency i
to an amplitude that is proportional to i.
Differentiation performs FM to AM conversion
 t

Input: FM (t )  AC cos C t   (t )   AC cos  C t  k f  m(  ) d 
  
d   
t
Output:  FM (t )   AC cos  C t  k f  m(  ) d  
'
dt     
 t
 Both AM
 
 'FM (t )   AC C  k f m(t )  sin  C t  k f  m( ) d  and FM
   included
dFM (t )
FM (t ) d
dt
Envelope
 ACC  AC k f m(t ) 
dt Detector After DC removal

Limiter Differentiator AM allows the

envelope detector
to be used
72
 Bandpass Limiter at the Receiver
For an envelope detector to work well the FM signal’s amplitude
should be constant or flat. We can accomplish with a “hard limiter.”
Factors such as channel noise, interference and channel fading result
in amplitude variations in an FM signal’s amplitude at the receiver.

VL
Output of
Limiter

t 
t 

FM (t ) Band-pass
Filter
@ C
Constant
Limiter Amplitude
Output

4
cos C t   k f t  m( )d  
t
FM (t ) x(t )  FM (t ) 
   
Input
73
 Practical FM Differentiator Demodulator

+ +
Differentiator C R x(t) Envelope
y(t)
at low frequencies _ Detector _

jRC j  / 3dB  1
H ( j )   ; where 3dB 
1  jRC 1  j  / 3dB  RC
1
For  << 3dB  ; then H ( j )  j RC
RC
Multiplication by j in the frequency domain is equivalent
to differentiation in time domain! The high-pass filter acts
as a differentiator for an FM signal. Therefore,

y(t) = ACC RC  ACC RCkf m(t )

Envelope detector extracts m(t)

74
Bode Plot of CR High-Pass Filter

75
 Practical Frequency Demodulators

Frequency discriminators can be built in various ways:

• Time-delay demodulator
• FM slope detector

• Balanced discriminator

• Quadrature demodulators

• Phase locked loops (a superior technique)

• Zero crossing detector

76
 Time Delay Demodulator

FM (t ) + y(t) yd(t)

 Amplifier
(Gain = 1/)
Envelope
Detector
Time
Delay
 Time-Delay Differentiator

This is an implementation of discrete time approximation to differentiation.

1
y (t )  FM (t )  FM (t   ) 

dFM (t ) 1 
 lim  y (t )   lim  FM (t )  FM (t   )  
 0 
dt  0
 
It can be shown that an adequate value for  is less than T/4, where T is the
period of the unmodulated carrier for the FM signal. Again, this relies upon
FM to AM conversion after which the envelope detector recovers m(t).

77
 FM Slope Detector Performs FM to AM Conversion

Envelope
Detector Comment: The
differentiation
operation is
performed by any
FM (t ) x(t) y(t) circuit acting as a
frequency-to-
amplitude
converter.

Slope Slope sets frequency to

approximation voltage conversion scale
Operates on the
skirt of the LC
resonance curve

78
 Balanced Discriminator (Foster-Seeley Discriminator)

Tuned Envelope
Circuits Detectors Centered around fc

• • f
FM (t )
• k m(t)

Transfer Characteristic
Another example of the
use of symmetry in design.

79
 Quadrature Demodulator – Block Diagram

FM signal is converted into PM signal

PM signal is used to recover the message signal m(t)

FM (t )

Phase m(t )
Low-Pass
Phase Comparator
Filter
Shifting Circuit
Circuit

Signal delay t0 times

carrier frequency fC Phase Detector
= 90 degrees (or /2).

80
Using a XOR Gate for Phase Frequency Detector

A B Output
0 0 0
0 1 1
1 0 1
1 1 0
XOR

Exclusive
OR gate

81
 Quadrature Demodulator – Implementation
The signal is split into two components. One passes through a network
providing a basic 90° phase shift in addition to the phase shift from the
signal’s frequency deviation. The mixer output is dependent upon the
phase difference between the two signals; that is, it acts as a phase detector
producing a voltage output proportional to the phase difference and thus
the frequency deviation on the FM signal.

Low-
pass
Filter
C2 Mixer
Incoming Demod
FM signal audio
C1 R L

http://www.radio-electronics.com/info/rf-technology-design/fm-reception/fm-quadrature-
detector-demodulator.php

82
 Phase-Locked Loops (Using Feedback)
A PLL consists of three basic components:
 Phase detector
 Loop filter
 Voltage-controlled oscillator (VCO)
PLL Diagram: Output signal
AC  cos C t  i (t )
is phase
H (s)

Phase Low-Pass Bias

Detector Filter Generator

Oscillator
(VCO)
eo (t )
2 B  cos C t   o (t )

83
Zero-Crossing Detectors

84
 Zero-Crossing Detector Illustration

FM (t ) Zero-
m(t)
Hard Multi- Averaging
crossing
Limiter vibrator circuit
circuit

https://www.slideshare.net/avocado1111/angle-modulation-35636989

85
 Noise in Frequency Modulation
In FM systems noise has a greater effect on the higher
modulating frequencies. It is common practice to artificially
boost the signal level of the higher modulating frequencies to
improve the signal-to-noise ratio of the transmitted FM signal.

This artificial boosting at the transmitter is called “pre-emphasis”

and the removal of the artificial boost at the receiver is called
“de-emphasis.”

The result is an improvement in the quality of received FM

signals.
S
 
 N  FM

S NO f 2 BT
   for f 
 N  FM AC2 2 -BT /2 BT /2 f
0

86
 Pre-Emphasis and De-Emphasis in FM

FM Channel noise acts as interference in

Interference

FM and is uniform over the entire BW.

Voice and music have more energy at
PM lower frequencies, so we need to
FM with Pre- and De-emphasis filters “emphasize “their upper frequencies
by filtering. However, the HF emphasis
Frequency f must be removed at the receiver using
a “de-emphasis” filter.

(Used commercially in recording industry)

Channel
m(t) Pre-emphasis FM FM De-emphasis
Filter Transmitter  Receiver Filter

R1

AWG Noise R1
C
R2
C

Filtering improves SNR in FM transmission.

87
 Typical Pre-Emphasis and De-Emphasis Filters
Transmitter Receiver
Pre-emphasis Filter De-emphasis Filter
R1

R1
R2 C
C

Vout 1  j R1C Vout 1

H ( )   H ( )  
Vin 1  j  R1 R2  C Vin 1  j R1C

H ( ) ( dB) H ( ) ( dB)

-6 dB/octave
+6 dB/octave

2.1 kHz 33 kHz 2.1 kHz

1 1 log( ) 1 log( )
R1C  R1 R2  C R1C

88
 Analog and Digital FM Cellular Telephones
1G analog cellular telephone (1983) – AMPS (Advanced Mobile Phone Service)
First use of cellular concept
Used 30 kHz channel spacing (but voice BW was B = 3 KHz)
Peak frequency deviation f = 12 kHz, and
BT = 2(f + B) = 2(12 kHz + 3 kHz) = 30 kHz
Two channels (30 kHz each); one for uplink and one for downlink
Used FM for voice and FSK for data communication
No protection from eavesdroppers

Successor to AMPS was GSM (Global System for Mobile) in early 1990s
GSM is 2G cellular telephone
Still used by nearly 50% of world’s population
GSM was a digital communication system
Modulating signal is a bit stream representing voice signal
Used Gaussian Minimum Shift Keying (GMSK)
Channel bandwidth is 200 kHz (simultaneously shared by 32 users
This is 4.8 times improvement over AMPS

More to come on cellular . . .

89
 Digital Carrier Modulation – ASK, FSK and PSK

Amplitude
Shift Keying
Digital Signals

Frequency
Shift Keying

Phase
Shift Keying

90
 Digital Phase Shift Modulation

Binary Phase Shift Keying (BPSK)

91
 Questions?

https://www.tutorialspoint.com/principles_of_communicati
on/principles_of_communication_modulation.htm

92
Additional slides

93
 Triangular-Wave FM Generation
Inverter

-1

1
-
Integrator
m(t) FM(t)

Switch vout

vout vin
vin
Schmitt Trigger

time

Can generate
FM up to 30 MHz

94
 Switching-Circuit Phase Modulator
Comparator
m(t) PM(t)
+ Flip-
BPF
Flop

Carrier Oscillator

Sawtooth Generator
m(t) Sawtooth wave

Comparator Output

Flip-Flop Output time

And filtering removes harmonics.

95
 FM System Improvement in SNR

The signal-to-noise ratio (SNR) improvement in an FM system is a

function of modulation index  ,

SNRFM  3 3 (   1)  CNR , where CNR is carrier-to-noise ratio

 B 
SNRFM  3  T   CNR
 2 Bm 
Example: For FM transmission bandwidth BT of 200 kHz and a message
bandwidth Bm of 15 kHz ( = 5.67), the improvement in the SNR at the
output of an FM receiver to have an FM gain of 27 dB above the CNR.

This is essentially a tradeoff between message signal quality (SNR) and

FM transmission bandwidth. Thus, greater transmission bandwidth is
The key to FM’s performance.

96