Angle Modulation (Phase & Frequency Modulation)

EE442 Lecture 8

Spring 2017

1
Amplitude, Frequency and Phase Modulation

With few exceptions,

Phase Modulation (PM)
is used primarily in
digital communication

2
 Why Use a Carrier Signal?

Carrier signals are used for two reasons:

(1) To reduce the wavelength for efficient transmission and

reception (the optimum antenna size is ¼ of a wavelength). A
typical audio frequency of 3000 Hz has a wavelength of 100 km and
would need an effective antenna length of 25 km! By comparison, a
typical FM carrier is 100 MHz, with a wavelength of 3 meters, and
would have an 80 cm long antenna (that is 31.5 inches long).

(2) To allow simultaneous use of the same channel, called

multiplexing. Each unique message signal has a different assigned
carrier frequency (e.g., radio stations) and share the same channel.
The telephone company invented modulation to allow phone
conversations to be transmitted over common phone lines.
Mandated by the FCC.

3
 Illustrating AM, PM and FM Signals

Carrier signal
Carrier Wave

m(t)

Modulating Signal m(t)

AM [Chapter 4]

Lathi & Ding; AM Modulated Signal

Page 252-292
PM

PM Modulated Signal [Chapter 5]

FM

FM Modulated Signal time

4
Phase-Frequency Relationship When Frequency is Constant
(t )  A cos( (t ))
(t) is generalized angle

(t )  A cos(C t  0 )

 (t )
C t  0

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

5
 Concept of Instantaneous Frequency
(t )  A cos( (t ))
Angle
Modulation (t) is generalized angle

(t )  A cos(C t  0 )

 (t )
C t  0
 (t )

Figure 5.1 from

Lathi & Ding;
Page 253
d (t )
Slope: i (t )   C
dt t t
0 i

ti time t

6
 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!

7
 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
in PM
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
modulating frequency fm
to modulating signal m(t)
Noise immunity better than AM
but not FM
Signal-to-noise ratio (SNR) is
not as good as in FM
PM is primarily for some
mobile radio services
Modulation index is
proportional to modulating
signal m(t)

8
 Phase Modulation (PM)
 (t )  C t  0  kpm(t ) Generally we let 0  0.

Equation (5.3b)
Let 0  0 PM (t )  A cos(C t  kpm(t )) Lathi & Ding;
Page 254

The instantaneous angular frequency (in radians/second) is

d (t ) m(t ) .
i (t )   C  kp  C  kp m(t )
dt dt
In phase modulation (PM) the instantaneous angular frequency
i varies linearly with the derivative of the message signal m(t)
.
(denoted here by m(t)).

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

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

9
 Frequency Modulation (FM)
But in frequency modulation the instantaneous angular frequency
i varies linearly with the modulating signal m(t),
i  C  k f m(t )
t t
 (t )   (

C  k f m( )) d  C t  k f  m( ) d


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

Then
Equation (5.5)
 t

FM (t )  A cos  C t  k f  m( ) d  Lathi & Ding;
   Page 254

FM and PM are very much related to each other.

In PM the angle is directly proportional to m(t).
In FM the angle is directly proportional to the integral of m(t), i.e.,  m(t )dt

10
 Summary
d (t )
Definition: Instantaneous frequency is i (t ) 
dt

Phase Modulation Frequency Modulation

t
 (t )  C t  kpm(t )
Angle  (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 variation of phase .

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

11
A Pictorial Way to View the Generation of FM and PM

t
H(j) = 1/j
 m( ) d FM (t )
m(t ) 


Phase
Modulator

Frequency Modulator

H(j) = j .m(t )
.
PM (t )
m(t ) d Frequency
dt Modulator

Phase Modulator

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

so that m(t) is recoverable.

12
Equations for FM Wave with Single Tone Modulation
Carrier signal AC cos(C t )
Carrier frequency C  2 fC
Modulating wave m(t ) Am cos(mt ) A single tone frequency
Modulating frequency m  2 f m (radians/sec)
Deviation sensitivity kf
 m  mmin 
Frequency deviation f  k f Am  k f  max 
 2  2 
f
Modulation Index 
fm
Instantaneous frequency f i  fC  k f Am cos(mt )  fC  f cos(mt )
  t  
Remember FM (t )  AC  cos  C t  k f   m( )d    , generally
  
     
  k f Am 
Modulated wave FM (t )  AC  cos  C t  sin(mt )  
  f m  
or FM (t )  AC  cos C t   sin(mt )  

Handout 13
 Generalized Angle Modulation
The first block can be any linear time-invariant (LTI) operator – it need only
be invertible so that we can recover m(t). In general, we have

Note: h(t) is the unit impulse response

 t

GAM (t )  A cos C t   m( )h  t    d 
  

Phase Modulation: h(t )  k p (t ),

Frequency Modulation: h(t )  k f u(t )

We shall focus more on Frequency Modulation in this course

and less on Phase Modulation.

14
 Average Power of a FM or PM Wave
The amplitude A 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 )  A cos C t  f ( k , m(t ))

A2
Average Power  (always)
2

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

15
 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
No such option exists in 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 addresses this)
16
 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
17
Phasor Interpretation of AM DSB with Carrier (continued)

18
 Example 5.1 in Lathi and Ding (pp. 256-257)
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)

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
19
 Example 5.2 in Lathi and Ding (pp. 257-259)
Sketch FM and PM waveforms for the modulating signal m(t). The constants
kf and kp are 2  105 and /2, respectively. Carrier frequency fc = 100 MHz.

FM

kf
f i  fC  m(t )  1  108  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 format.
20
Example 5.2 in Lathi and Ding (pp. 257-259) – continued
Sketch FM and PM waveforms for the modulating signal m(t). The constants
kf and kp are 2  105 and /2, respectively. Carrier frequency fc = 100 MHz.

kp . 1.
f i  fC  m(t )  1  108  m(t )
2 4
PM
This is carrier PM by a digital signal
– it is Phase Shift Keying (PSK)
because digital data is represented
by phase of the carrier wave.

  
 PM (t )  A cos C t  kp m(t )  A cos C t 
m(t ) Lathi & Ding;
 2 
Page 258
 PM (t )  A sin(C t ) when m(t )  1
 PM (t )   A sin(C t ) when m(t )  1
21
 Case I – Narrowband FM (NBFM)
There are two approximations for FM:

◮ Narrowband approximation (NBFM) Lathi & Ding;

◮ Wideband approximation (WBFM) Page 260

 t
 f f
NBFM: FM (t )  A cos  C t  k f  m( ) d   
   fm B
t
If k f  m( ) d

 1, we have NBFM.

t 1
Let k f

 m( ) d  k f sin(mt ),
 = 0.2
Then bandwidth BFM  2 f m

NBPM requires  << 1 radian (fC - fm) (fC + fm) f

fC
(typically less than 0.2 radian)

22
 Narrowband FM (NBFM) Equation
f f
Start with design equation for tone frequency fm :  
fm B
FM
NB
(t )  AC cos  2 fC t   sin(2 fmt ) 
FM
NB
(t )  AC cos  2 fC t  cos   sin(2 fmt )   AC sin 2 fC t  sin   sin(2 fmt ) 
Note: cos   sin(2 fmt )   1, and sin   sin(2 fmt)    sin(2 fmt)
FM
NB
(t )  AC cos  2 fC t    sin  2 fC t   sin(2 fmt )
1
FM
NB
(t )  AC cos  2 fC t    AC cos  2 ( fC  fm )t   cos  2 ( fC  fm )t   NBFM
2
Result from AM modulation with tone frequency:
1
 AM (t )  AC cos  2 fC t    AC cos  2 ( fC  fm )t   cos 2 ( fC  fm )t   AM
2
The difference is the sign (i.e., phase) of the difference frequency term.
Conclusion: The NBFM bandwidth is comparable to that of AM.
 Bandwidth: BT  2 fm

23
 Case II – Wideband FM (WBFM)
WBPM requires  >> 1 radian
For wideband FM we have a nonlinear process, with single tone
modulation:
FM
WB
(t )  Re  AC exp  j 2 fC t  j  sin(2 f mt )  
We need to expand the exponential into a Fourier series so that
we can analyze FM
WB
(t ).

f f
 WB
FM (t )  AC 
n 
J n (  )  cos  2 ( fC  nf m )t  
fm

B
where the coefficients J n (  ) are Bessel functions.

Spectral analysis from tone modulation of WBFM: Lathi & Ding; pp. 264-270
We will not cover this section in ES 442 but rather focus upon a physical
Interpretation of the spectrum spread.

24
FM (or PM) Requires Much More bandwidth Than AM
A A
t f

fC
A Carrier Signal (frequency fc ) A
t f

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

A WBFM
A Amplitude Modulated Signal
t f

Frequency Modulated (FM) Signal

25
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

f
 = 10 
fm

BT or BW

26
 Spectra of an FM Signal

Single-tone
A  = 0.2 A
modulation

f
  = 1.0
fm

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.

27
Measured Spectra of an FM Radio Signal

Voice modulation

200 kHz

noise

28
 Selecting an FM Station
Broadcast FM Radio covers from 88 MHz to 108 MHz
100 stations – 200 kHz spacing between FM stations

29
 Specifications for 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
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

Question: For FM broadcast what is the modulation index ?

30
 FM Bandwidth and the Modulation Index 
Lathi & Ding – Chapter 5 – see pages 261 to 263

Narrowband FM (NBFM) –  << 1 radian

NB
BFM  2B where B is the bandwidth of m(t )
Wideband FM (WBFM) –  >> 1 radian
WB
BFM  2( f  B)  2 B(   1) Carson's Rule
or BT  2( f  f m )

Peak frequency deviation is  = kf Am

f f
Modulation index   
fm B

31
 Phasor Construction of an FM Signal

We are constrained by
constant amplitude for
C
both FM and PM signals.

This is NBFM. The next

slide shows an animation
of this in operation.


C

32
 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.

33
 Direct Generation of FM Signal Using a VCO

VDD

m(t)
LC Tank Circuit

Voltage
FM
NB
(t )
Control
m(t)

Varactor
1
diodes osc
L1C eq

34
Indirect Generation of an FM Signal Using 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 )  (t )
NB
FM
 WB
FM (t )
Frequency
NBFM Multiplier

A frequency multiplier is used to increase both the

carrier frequency and the modulation index by integer N.

35
 Armstrong Indirect FM Transmitter Example
Lathi & Ding;
pp. 275-277 fC 2  12.8 MHz
 (t )
NB
FM fC f2  1.6 kHz
NBFM X64
fC 1  200 kHz generation Multiplier
f1  25Hz FM
NB
(t )
fC 3  1.9 MHz
FM
WB
(t ) f3  1.6 kHz
X48
PA Multiplier BPF
fC 4  91.2 MHz The mixer
does not
f4  76.8 kHz
change f
f LO  10.9 MHz

Crystal Oscillator
Note: These numbers are related
to an FM broadcast radio station.

36
Generation of Narrowband Frequency Modulation (NBFM)

 t

FM (t )  A cos  C t  k f  m( ) d 
  
NBFM requires  << 1 radian
DSB-SC
modulator
m (t) NBFM
 kf 

Asin(ct)
-/2 Carrier

Lathi & Ding; Acos(ct)

Figure 5.10
Page 276

37
Generation of Narrowband Phase Modulation (NBPM)

PM (t )  A cos(C t  kpm(t ))

m (t) NBPM
kp 

Asin(ct)
-/2
Acos(ct)

38
 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.

39
 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.

40
 Practical Frequency Demodulators
Frequency discriminators can be built using various ways:

• FM slope detector

• Balanced discriminator

• Quadrature demodulators

• Phase locked loops (superior technique)

• Zero crossing detector

41
FM Slope Detector Performs FM to AM Conversion

Envelope
Detector

FM (t )

Slope sets frequency to

voltage conversion factor

42
 Balanced Discriminator (Foster-Seeley Discriminator)

Tuned Envelope
Circuit Detector Centered around fc

• • f
FM (t )
•

Transfer Characteristics

43
 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).

44
Using XOR Gate for Phase Frequency Detector

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

45
 Quadrature Demodulator – Implementation

FM (t ) m(t )

FM (t )

Phase shifter
(or delay time)

46
 Phase-Locked Loops

A PLL consists of three basic components:

 Phase detector
 Loop filter
 Voltage-controlled oscillator (VCO)
PLL Diagram:
Output signal
A cos C t  i (t )
is phase
H (s )

Phase Low-Pass Bias

Detector Filter Generator

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

47
Zero-Crossing Detectors

48
Zero-Crossing Detector Illustration

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

49
 Example

FM (t )  10 cos  2 (106 )t  8sin(2 (10 3 )t ) 

f
FM (t )

50
 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 (10 3 )t ) 

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

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

BFM  2 f m (   1)  2,000(8  1)  18,000 Hz

51
 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.

52
 Typical Pre-Emphasis and De-Emphasis Filters
Transmitter Receiver
Pre-emphasis Filter De-emphasis Filter
R1 Lathi & Ding;
Chapter 5,
pp. 286-289
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

53
 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) – later in EE 442
Channel bandwidth is 200 kHz (simultaneously shared by 32 users
This is 4.8 times improvement over AMPS

More to come on cellular . . .

54
 Digital Carrier Modulation – ASK, FSK and PSK

Amplitude
Shift Keying

Frequency
Shift Keying

Phase
Shift Keying

Digital Signals

55
Questions?

56