Telecommunication Systems I

Chapter 5
Dr. Saeed Abdallah

1
 Chapter 5: Angle Modulation
Motivation:
• In AM modulation, noise is proportional to the bandwidth.
• There was an effort to consider other types of modulation
that may use less bandwidth and are more robust to noise.
Angle Modulation:
• The idea is to carry the signal by varying the angle of the
carrier.
• Can be divided into phase modulation (PM) and frequency
modulation (FM).
• FM is capable of much more noise immunity than AM.
• FM is currently the most used form of analog communication.

2
 Angle Modulation
Conventional Carrier without message
• A conventional sinusoidal signal not carrying any message is given
by
𝑥 𝑡 = 𝐴𝑐𝑜𝑠(𝜔𝑐 𝑡 + 𝜃𝑜 )
where 𝐴 is constant, 𝜔𝑐 is constant, and 𝜙 is constant.

• In others words, 𝑥 𝑡 = 𝐴𝑐𝑜𝑠(𝜃(𝑡)), where 𝜃 𝑡 = 𝜔𝑐 𝑡 + 𝜃𝑜 .

• Hence, the phase is a linear function of time.

Amplitude Moduation (DSB-SC)

• To send a message 𝑚(𝑡) using DSC-SC, we use a sinusoid whose
amplitude consists of the message signal
𝑥 𝑡 = 𝑚 𝑡 cos 𝜔𝑐 𝑡 + 𝜃𝑜 = 𝑚 𝑡 cos(𝜃(𝑡))
where

𝜃 𝑡 = 𝜔𝑐 𝑡 + 𝜃𝑜 = 2𝜋𝑓𝑐 𝑡 + 𝜃𝑜

3
 Angle Modulation
• In Angle modulation, we send signals using a generalized
sinusoid
𝑥 𝑡 = 𝐴𝑐𝑜𝑠(𝜃 𝑡 )
where 𝜃 𝑡 is the generalized angle which does not have to
be linear.
• The idea of angle modulation is to vary 𝜃 𝑡 using the
message 𝑚(𝑡).

4
 Instantaneous Frequency
• Instantaneous frequency is the derivative of the generalized
angle:
𝑑𝜃
𝜔𝑖 𝑡 = = 𝜃 ′ 𝑡 (rad/s),
𝑑𝑡
1 𝑑𝜃
or 𝑓𝑖 𝑡 = (Hz)
2𝜋 𝑑𝑡
• Hence, the generalized angle is the integral of the
instantaneous frequency
𝑡
𝜃 𝑡 = ‫׬‬−∞ 𝜔𝑖 𝛼 𝑑𝛼

5
 Instantaneous Frequency

• 𝜔𝑖 𝑡 is the slope of the angle 𝜃(𝑡) at instant 𝑡.

6
 Instantaneous Frequency
Example:
Find the instantaneous frequency for the following
sinusoids in Hz.
𝜋
• 𝑥 𝑡 = 10 cos 2000𝜋𝑡 +
4
1 𝑑𝜃 𝑡
𝑓𝑖 𝑡 = = 1000𝐻𝑧 (this is a conventional
2𝜋 𝑑𝑡
sinusoid)
• 𝑥 𝑡 = 10 cos 2000𝜋𝑡 + 100 sin 100𝜋𝑡
1 𝑑𝜃 𝑡
𝑓𝑖 𝑡 = = 1000 + 5000cos(100𝜋𝑡)
2𝜋 𝑑𝑡
• The above sinusoid is a generalized sinusoid like the ones
used to transmit messages in Angle Modulation

7
 Angle Modulation
Two types: Phase modulation and Frequency modulation
Phase modulation (PM),
– The phase is varied linearly with 𝑚(𝑡):
𝜃 𝑡 = 𝜔𝑐 𝑡 + 𝑘𝑝 𝑚(𝑡)
– 𝑘𝑝 is a constant, called the modulation sensitivity, with unit
of Rad/Volt.
– Instantaneous frequency:
𝑑𝜃
𝜔𝑖 𝑡 = = 𝜔𝑐 + 𝑘𝑝 𝑚′(𝑡),
𝑑𝑡
1 𝑑𝜃 𝑘𝑝
𝑓𝑖 𝑡 = = 𝑓𝑐 + 𝑚′(𝑡),
2𝜋 𝑑𝑡 2𝜋
– Hence, instantaneous frequency varies linearly with
the derivative 𝑚′(𝑡) of 𝑚(𝑡) .
– If 𝑚 𝑡 varies rapidly, the frequency deviations will be
larger.
8
 Angle Modulation
Frequency Modulation (FM),
– The instantaneous frequency varies linearly with 𝑚(𝑡):
𝜔𝑖 𝑡 = 𝜔𝑐 + 𝑘𝑓 𝑚(𝑡),
𝑘𝑓
𝑓𝑖 𝑡 = 𝑓𝑐 + 𝑚(𝑡)
2𝜋

– 𝑘𝑓 is called the modulation sensitivity with unit of

Rad/s/volt.
– The phase depends on the integral of the message
𝑡 𝑡
𝜃 𝑡 = ‫׬‬−∞ 𝜔𝑖 𝑢 𝑑𝑢 = 𝜔𝑐 𝑡 + 𝑘𝑓 ‫׬‬−∞ 𝑚 𝑢 𝑑𝑢

𝐴2
• The average power of PM and FM is always regardless of
2
the values of 𝑘𝑓 or 𝑘𝑝 .

9
 Relationship between FM and PM
• Phase modulation of 𝑚(𝑡) = frequency modulation of 𝑚′(𝑡).
• Frequency modulation of 𝑚 𝑡 = phase modulation of
‫𝑢𝑑 𝑢 𝑚 ׬‬.

10
 Frequency Deviation
• An important concept for understanding FM is the concept of
frequency deviation.
• The amount of frequency deviation a signal experiences is a
measure of the change of transmitter output frequency from
the rest frequency of the transmitter.
• The rest frequency is the output frequency when no message
is transmitted (i.e., the carrier frequency 𝑓𝑐 ).
• The frequency deviation is Δ𝑓 = 𝑓𝑚𝑎𝑥 − 𝑓𝑐 where 𝑓𝑚𝑎𝑥 is the
maximum value of the instantaneous frequency.
𝑘𝑓 𝑚𝑝
• In FM, the frequency deviation is given by Δ𝑓 = , or
2𝜋
equivalently, Δ𝜔 = 𝑘𝑓 𝑚𝑝 , where 𝑚𝑝 is the peak amplitude of
the signal 𝑚(𝑡).

11
 Frequency Deviation
In FM,
𝑘𝑓
𝑓𝑖 𝑡 = 𝑓𝑐 + 𝑚(𝑡)
2𝜋
• The maximum value of 𝑓𝑖 (𝑡) is achieved when 𝑚(𝑡) take its
maximum value 𝑚𝑝 .
𝑘𝑓
• Hence, 𝑓𝑚𝑎𝑥 = 𝑓𝑐 + 𝑚𝑝
2𝜋
• Hence, the frequency deviation is
𝑘𝑓 𝑚𝑝
Δ𝑓 = 𝑓𝑚𝑎𝑥 − 𝑓𝑐 =
2𝜋
• We can also measure the frequency deviation in Rad/s
Δ𝜔 = 𝑘𝑓 𝑚𝑝

12
 Frequency Deviation
In PM,
𝑘𝑝
𝑓𝑖 𝑡 = 𝑓𝑐 + 𝑚′(𝑡)
2𝜋
• The maximum value of 𝑓𝑖 (𝑡) is achieved when 𝑚′(𝑡) take its
maximum value 𝑚𝑝 ′.
𝑘𝑝
• Hence, 𝑓𝑚𝑎𝑥 = 𝑓𝑐 + 𝑚𝑝 ′
2𝜋
• Hence, the frequency deviation is
𝑘𝑝 𝑚𝑝 ′
Δ𝑓 = 𝑓𝑚𝑎𝑥 − 𝑓𝑐 =
2𝜋
• We can also measure the frequency deviation in Rad/s
Δ𝜔 = 𝑘𝑝 𝑚𝑝 ′

13
FM Example: 𝑘𝑓 = 2𝜋 × 105 , 𝑓𝑐 = 100 MHz
𝑘𝑓
• 𝑓𝑖 (𝑡) = 𝑓𝑐 + 2𝜋 𝑚 𝑡 = 108 + 105 𝑚(𝑡)
• 𝑓min = 99.9 MHz,
• 𝑓m𝑎𝑥 =100.1 MHz.
• Δ𝑓 = 𝑓𝑚𝑎𝑥 − 𝑓𝑐 = 0.1 MHz

FM instantaneous frequency:
• 𝑚(𝑡) increases and decreases linearly with time, so the
instantaneous frequency increases linearly over half a cycle and
then decreases linearly over the remaining half.

14
PM Example: 𝑘𝑝 = 10𝜋, 𝑓𝑐 = 100 MHz

PM instantaneous frequency:
𝑘𝑝
• 𝑓𝑖 (𝑡) = 𝑓𝑐 + 𝑚′ 𝑡 = 108 + 5𝑚′(𝑡)
2𝜋
• 𝑓min = 99.9 MHz, 𝑓m𝑎𝑥 =100.1 MHz, Δ𝑓 = 0.1 MHz
• 𝑚′(𝑡) switches back and forth from -20,000 to 20,000,
so carrier frequency switches back and forth from 99.9
MHz to 100.1 MHz every half-cycle of 𝑚′(𝑡).
15
Another FM Example,

𝑘𝑓 = 2𝜋 × 105

𝑓𝐶 = 100 𝑀𝐻𝑧

𝑘𝑓
• 𝑓𝑖 (𝑡) = 𝑓𝑐 + 𝑚(𝑡) = 108 + 105 𝑚(𝑡)
2𝜋
• 𝑚(𝑡) switches back and forth from 1 to -1 so FM wave switches back and
forth between 99.9 and 100.1 MHz.
• The above transmission scheme is called frequency shift keying.
• This can be used in binary digital transmission where +1 represents binary
1 and -1 represents binary 0.
• The binary message can be detected using two bandpass filters, one tuned
for 99.9 MHz and the other tuned for 100.1 MHz. 16
We can also use PM for the same binary message

𝜋
𝑘𝑝 = , 𝑓𝑐 = 100 𝑀𝐻𝑧
2

𝑘𝑝 1
• For PM, 𝑓𝑖 (𝑡) = 𝑓𝑐 + 𝑚′ 𝑡 = 108 + 𝑚′(𝑡)
2𝜋 4
• Derivative is zero except at points of discontinuity where
there are impulses of strength 2. Frequency stays the same
except at isolated points in time.
• The impulses result in sudden change in the phase 𝜃(𝑡).
• This scheme is referred to as phase shift keying.
17
Phase-Shift Keying (PSK)
Continuing with the previous slide
• If 𝑘𝑝 = 𝜋/2 and 𝑚(𝑡) takes only the values +1 and −1, then
𝜓𝑃𝑀 𝑡 = 𝐴𝑐𝑜𝑠 𝜔𝑐 𝑡 + 𝑘𝑝 𝑚(𝑡)
𝜋
= 𝐴𝑐𝑜𝑠 𝜔𝑐 𝑡 + 𝑚(𝑡)
2

𝜋
𝐴𝑐𝑜𝑠 𝜔𝑐 𝑡 − = 𝐴𝑠𝑖𝑛 𝜔𝑐 𝑡 𝑤ℎ𝑒𝑛 𝑚 𝑡 = −1
2
=൞ 𝜋
𝐴𝑐𝑜𝑠 𝜔𝑐 𝑡 + = −𝐴𝑠𝑖𝑛 𝜔𝑐 𝑡 𝑤ℎ𝑒𝑛 𝑚 𝑡 = +1
2

• At instances where the impulses are located, the carrier phase

shifts by ±𝜋 instantaneously.
• The sinusoids are 180𝑜 out of phase but synchronized with the
carrier.
• It is possible to design a receiver that detects the change of phase
to distinguish between bit 0 and bit 1. 18
Sudden Changes of Phase
• Sudden changes of phase can occur when using phase
modulation if the message 𝑚(𝑡) changes suddenly.
• Remember, the phase in PM modulation is given by
𝜃 𝑡 = 2𝜋𝑓𝑐 𝑡 + 𝑘𝑝 𝑚(𝑡)
• Suppose 𝑚(𝑡) changes suddenly at time 𝑡𝑜 from +1 to −1
• Then the phase changes suddenly from 𝜃 𝑡𝑜− = 2𝜋𝑓𝑐 𝑡𝑜 + 𝑘𝑝
to 𝜃 𝑡𝑜+ = 2𝜋𝑓𝑐 𝑡𝑜 − 𝑘𝑝
• The change of phase is Δ𝜃 = 𝜃 𝑡𝑜+ − 𝜃 𝑡𝑜− = −2𝑘𝑝
• More generally, the change of phase is given by
Δ𝜃 = 𝜃 𝑡𝑜+ − 𝜃 𝑡𝑜− = 𝑘𝑝 (m t + −
o − m(t o ))
• In other words
𝑘𝑝 (𝑛𝑒𝑤 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑚 𝑡 − 𝑜𝑙𝑑 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑚(𝑡))

19
20
21
22
23
24
25
26
27
 Bandwidth of Angle-Modulated Signals
• Angle modulation is nonlinear and difficult to
analyze.
• No properties of Fourier transform can be applied
directly to analyze bandwidth.
• Early developers thought the bandwidth can be
reduced to zero, they were wrong!
• Two approximations for FM:
– Narrowband approximation.
– Wideband approximation.

28
 Bandwidth of Angle-Modulated Signals
• Let 𝐵𝑚 be the bandwidth of the baseband signal 𝑚 𝑡 , and let
𝑎 𝑡 = ‫𝑢𝑑 𝑢 𝑚 ׬‬.
Narrow-band approximation (NBFM):
• If 𝑘𝑓 𝑎 𝑡 ≪ 1, then 𝐵𝐹𝑀 ≈ 2𝐵𝑚 and 𝜓𝐹𝑀 𝑡 ≈ 𝐴൫𝑐𝑜𝑠𝜔𝑐 𝑡 −

29
 Narrowband FM Bandwidth Approximation
𝑡
• Let 𝑎 𝑡 = ‫׬‬−∞ 𝑚 𝑢 𝑑𝑢 , the FM signal is given by

𝜓𝐹𝑀 𝑡 = 𝐴𝑐𝑜𝑠(𝜔𝑐 𝑡 + 𝑘𝑓 𝑎(𝑡))

𝑥2 𝑥4 𝑥3 𝑥5
Note: cos 𝑥 = 1 − + … , sin 𝑥 = 𝑥 − + …
2! 4! 3! 5!
• Using Taylor Series expansion for cosine and sine, we obtain
𝜓𝐹𝑀 𝑡
𝑘𝑓2
= 𝐴(cos 𝜔𝑐 𝑡 − 𝑘𝑓 𝑎 𝑡 sin 𝜔𝑐 𝑡 − 𝑎 𝑡 2 cos 𝜔𝑐 𝑡
2!
3
𝑘𝑓
+ 𝑎 𝑡 3 sin 𝜔𝑐 𝑡 + ⋯ )
3!

• This expansion shows that the true bandwidth is infinite.

|𝑘𝑓 𝑎 𝑡 |𝑛
• However, since 𝑘𝑓 𝑎 𝑡 ≪ 1, → 0 as 𝑛 increases, hence most
𝑛!
of the power will be concentrated in a finite band
30
31
 Narrowband FM Approximation
Hence, if 𝑘𝑓 is very small such that |𝑘𝑓 𝑎(𝑡)| ≪ 1, then all but the
first two terms of the expansion are negligible.
• Hence, narrowband FM Approximation:
𝜓𝐹𝑀 𝑡 ≈ 𝐴(cos 𝜔𝑐 𝑡 − 𝑘𝑓 𝑎 𝑡 sin(𝜔𝑐 𝑡))
𝑡
• Note that 𝑎 𝑡 = ‫׬‬−∞ 𝑚 𝑢 𝑑𝑢 has the same bandwidth as
𝑚(𝑡).
• If 𝑚(𝑡) has bandwidth 𝐵𝑚 , then Narrowband FM has
bandwidth 2𝐵𝑚 (same as bandwidth of DSB-SC and AM).
• Narrowband approximation for phase modulation:
𝜓𝑃𝑀 (𝑡) ≈ 𝐴(cos 𝜔𝑐 𝑡 − 𝑘𝑝 𝑚 𝑡 sin(𝜔𝑐 𝑡))

32
33
 Wideband FM Bandwidth Analysis
• FM signals for which 𝑘𝑓 𝑎 𝑡 > 1 are called wideband FM
(WBFM). In this case, the NBFM approximation 𝐵𝐹𝑀 ≈ 2𝐵𝑚 does
not work.
• However, by using a stair case approximation of the baseband
signal 𝑚(𝑡), it is possible to obtain the following approximation
𝐵𝐹𝑀 = 2(Δ𝑓 + 𝐵𝑚 )

• This is called Carson’s rule.

34
 WBFM Bandwidth Analysis

• Using Carson’s rule, for narrowband case, Δ𝑓 ≪ 𝐵𝑚 , hence

𝐵𝐹𝑀 ≈ 2𝐵𝑚 .
• While for wideband case, Δ𝑓 ≫ 𝐵𝑚 , hence 𝐵𝐹𝑀 ≈ 2Δ𝑓.
Δ𝑓
• Letting 𝛽 = , 𝐵𝐹𝑀 = 2𝐵𝑚 (𝛽 + 1).
𝐵𝑚
• 𝛽 is called the deviation ratio.
• To determine if a signal is NBFM or WBFM, a threshold 𝜀 = 1
is chosen.
• 𝛽 > 1 ⇒ 𝑊𝐵𝐹𝑀, 𝛽 < 1 ⇒ 𝑁𝐵𝐹𝑀
• As 𝛽 increases the FM signal occupies more bandwidth.

35
 Special Case: Sinusoidal Message
• Consider a sinusoidal message signal 𝑚 𝑡 =
𝐴𝑚 𝑐𝑜𝑠(2𝜋𝑓𝑚 𝑡)
• For FM modulation, the instantaneous frequency will be
𝑘𝑓 𝐴𝑚
𝑓𝑖 𝑡 = 𝑓𝑐 + cos(2𝜋𝑓𝑚 𝑡)
2𝜋
𝑘𝑓 𝐴𝑚
• The frequency deviation is Δ𝑓 =
2𝜋
• The bandwidth of the baseband signal is 𝐵𝑚 = 𝑓𝑚 .
𝐾𝑓 𝐴 𝑚
• The deviation ratio is 𝛽 =
2𝜋𝑓𝑚
• The FM-modulated signal is given by
𝑡
𝜓𝐹𝑀 𝑡 = 𝐴𝑐𝑜𝑠 2𝜋𝑓𝑐 𝑡 + 𝐾𝑓 න 𝐴𝑚 cos(2𝜋𝑓𝑚 𝑢)𝑑𝑢
−∞
= 𝐴𝑐𝑜𝑠(2𝜋𝑓𝑐 𝑡 + 𝛽sin(2𝜋𝑓𝑚 𝑡))
36
 Special Case: Sinusoidal Message
• For this special case, we can express 𝜓𝐹𝑀 (𝑡) as
∞

𝜓𝐹𝑀 𝑡 = 𝐴 ෍ 𝐽𝑛 𝛽 cos 2𝜋(𝑓𝑐 +𝑛𝑓𝑚 𝑡)

𝑛=−∞
where 𝐽𝑛 . is the 𝑛𝑡ℎ order Bessel function.
• Hence, the FM signal consists of sum of sinusoids that are
occur at 𝑓𝑐 ± 𝑛𝑓𝑚
• In the frequency domain
Φ𝐹𝑀 𝑓
∞
𝐴
= ෍ 𝐽𝑛 𝛽 𝛿 𝑓 − 𝑓𝑐 + 𝑛𝑓𝑚 + 𝛿 𝑓 + 𝑓𝑐 + 𝑛𝑓𝑚
2
𝑛=−∞
• The spectrum consists of a carrier component and infinitely
many sideband components at frequencies 𝑓𝑐 ± 𝑛𝑓𝑚 , 𝑛 =
1,2,3 …
37
 Special Case: Sinusoidal Message
• The amplitudes of the spectral lines depend on the
value of 𝐽𝑛 (𝛽), which becomes small as 𝑛 increases.
• The number of significant spectral lines depends on
𝛽.
• For 𝛽 ≪ 1, only 𝐽0 and 𝐽1 are significant, so the
carrier consists of a carrier and two sidebands.
• If 𝐵 ≫ 1, there will be many sidebands and the
overall bandwidth will be large.

38
Special Case: Sinusoidal Message

39
 WBPM Bandwidth Analysis
• Results derived for FM can be applied directly to PM by
replacing 𝑚(𝑡) with 𝑚′(𝑡).
′
𝑘𝑝 𝑚𝑝
• Hence, 𝐵𝑃𝑀 = 2 Δ𝑓 + 𝐵𝑚 = 2 + 𝐵𝑚 .
2𝜋

• We conclude that in FM the bandwidth depends on the peak

value of the signal 𝑚(𝑡), while in PM it depends on the peak
value of the derivative 𝑚′(𝑡).

40
Example 1

a) For the above signal, estimate 𝐵𝐹𝑀 and 𝐵𝑃𝑀 for 𝑘𝑓 = 2𝜋 × 105 and
𝑘𝑝 = 5𝜋. Assume that the essential bandwidth of the periodic signal
𝑚(𝑡) is the frequency of the third harmonic.

b) Repeat the results if the amplitude of 𝑚(𝑡) is doubled.

Reminder:
Periodic signals have Fourier Series expansion
 
2
x(t )  
n  
X [n]e jn0 t
 a[0]   a[n] cos(n 0 t )  b[n]sin(n 0 t ) ;  0  2
n1
f0 
T

41
42
43
44
Example 2:

Is the signal 𝜙𝐹𝑀 (𝑡) = 10cos(3000𝜋𝑡 + 100𝜋 cos 2 300𝜋𝑡) an

NBFM or WBFM signal?

45
46
Example 3:
If the signal 𝑚 𝑡 = 3 cos 600𝜋𝑡 is to be FM modulated using a
carrier of 10 KHz, and frequency sensitivity 𝑘𝑓 = 2000𝜋, find
the following (assume 𝐴 = 10 and sin −∞ = 0 ):
a) The modulated signal 𝜙𝐹𝑀 (𝑡)
b) Frequency deviation Δ𝑓 in Hz
c) Deviation ratio 𝛽
d) Estimated FM signal bandwidth in Hz.
e) FM power.

47
48
 Generation of FM Signals
• In the narrowband scenario, we can use the NBFM
approximation to generate the FM-modulated signal.

• It is important to note that the NBFM signal generated using

the above modulator will have distortion because it uses an
approximation.

49
 Generation of FM Signals (Narrowband)
• In particular, the signal generated using the narrowband
approximation will experience amplitude variations, while an
ideal FM signal should have a fixed amplitude.
• A bandpass limiter is a nonlinear device that can be used to
eliminate these amplitude variations.
• The bandpass limiter is the cascade of a hardlimiter and a
bandpass filter.

50
 Generation of FM Signals (Narrowband)
• The bandpass limiter is the cascade of a hardlimiter and a
bandpass filter.

• The hardlimiter produces an output of 1 whenever the input

signal is positive and an output of -1 whenever it is negative.

51
 Generation of FM Signals (Narrowband)

• The resulting signal is an FM signal with fixed amplitude

52
Generation of WBFM: Indirect Method
Indirect method for WBFM Generation
1. Generate NBFM signal (as before).
2. Convert NBFM into WBFM using additional frequency multipliers.
• Frequency multiplier can be realized by using nonlinear device
followed by bandpass filter.

53
Generation of WBFM: Indirect Method
Indirect method for WBFM Generation

• Example of nonlinear device: 𝑦 𝑡 = 𝑎1 𝑥 𝑡 + 𝑎2 𝑥 2 (𝑡).

• Suppose the input to the NL device is the FM signal
𝑥 𝑡 = cos(𝜔𝑐 𝑡 + 𝑘𝑓 ‫)𝛼𝑑 𝛼 𝑚 ׬‬
• If 𝑥(𝑡) passes through the NL device, output is:
𝑦 𝑡 = 𝑎1 cos(𝜔𝑐 𝑡 + 𝑘𝑓 ‫ )𝛼𝑑 𝛼 𝑚 ׬‬+𝑎2 cos 2 𝜔𝑐 𝑡 + 𝑘𝑓 ‫𝛼𝑑 𝛼 𝑚 ׬‬
= 𝑎1 cos(𝜔𝑐 𝑡 + 𝑘𝑓 ‫ )𝛼𝑑 𝛼 𝑚 ׬‬+ 0.5𝑎2 + 0.5𝑎2 cos[2𝜔𝑐 𝑡 + 2𝑘𝑓 ‫]𝛼𝑑 𝛼 𝑚 ׬‬
• Using a bandpass filter centered at 2𝑓𝑐 , we obtain an FM signal with
twice the carrier frequency, and also twice the frequency deviation.

54
Generation of WBFM: Indirect Method
• More generally, nonlinear devices have the characteristic:
𝑦 𝑡 = 𝑎0 + 𝑎1 𝑥 𝑡 + 𝑎2 𝑥 2 𝑡 + ⋯ + 𝑎𝑛 𝑥 𝑛 (𝑡)
• If 𝑥 t = Acos 𝜔𝑐 𝑡 + 𝑘𝑓 ‫ 𝛼𝑑 𝛼 𝑚 ׬‬, then we can show using
trigonometric identities that 𝑦(𝑡) is of the form:
𝑦 𝑡
= 𝑐0 + c1 cos 𝜔𝑐 𝑡 + 𝑘𝑓 ‫ 𝛼𝑑 𝛼 𝑚 ׬‬+ 𝑐2 cos 2𝜔𝑐 𝑡 + 2𝑘𝑓 ‫ 𝛼𝑑 𝛼 𝑚 ׬‬+ ⋯
+𝑐𝑛 cos[𝑛𝜔𝑐 𝑡 + 𝑛𝑘𝑓 ‫]𝛼𝑑 𝛼 𝑚 ׬‬
• Output is a sum of FM waves whose center frequencies are 𝑓𝑐 , 2𝑓𝑐 , …, 𝑛𝑓𝑐
with frequency deviations Δ𝑓, 2Δ𝑓, …, 𝑛Δ𝑓.
• Bandpass filter centered at 𝑛𝑓𝑐 can recover FM signal whose
instantaneous frequency has been multiplied by 𝑛.
• Hence these devices are called frequency multipliers.
• If we wish to change the carrier frequency, we can use a frequency
converter (mixer).

55
Generation of WBFM: Indirect Method
• Armstrong Indirect FM Transmitter:

56
Example:
A narrowband FM (NBFM) signal with a carrier of 1 KHz and 𝛽 = 0.2 is used
to generate a Wideband FM (WBFM) signal with a carrier of 100 KHz and 𝛽 =
7.2. Sketch the Armstrong FM transmitter that can be used to generate the
required FM signal, and specify all the parameters, knowing that
𝑚 𝑡 = 4cos(200𝜋𝑡).

57
Solution:
7.2
𝑀𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝐹𝑎𝑐𝑡𝑜𝑟 = = 36.
0.2
(0) 1
Before multiplier, 𝑓𝑐 = 1 𝑘𝐻𝑧. After multiplier 𝑓𝑐 = 36 𝑘𝐻𝑧.
2
We need a mixer to convert the carrier frequency to 𝑓𝑐 = 100 𝑘𝐻𝑧.
The oscillator frequency of the mixer should be 100 − 36 = 64 𝑘𝐻𝑧.
We can also use 100 + 36 = 136 𝑘𝐻𝑧.
Finally, to find the filter parameters for the BPF, we need to find the
bandwidth of the FM signal after the frequency multiplier. We have
𝐵𝐹𝑀 = 2(Δ𝑓 + 𝐵𝑚 )
Since 𝑚 𝑡 = 4cos(200𝜋𝑡), 𝐵𝑚 = 100 𝐻𝑧.
Δ𝑓 = 𝛽 × 𝐵𝑚 = 7.2 × 100 = 0.72 𝑘𝐻𝑧
𝐵𝐹𝑀 = 2 0.72 + 0.1 = 1. 64 𝑘𝐻𝑧
Hence, the BPF should have a center frequency of 100 𝑘𝐻𝑧 and a
passband bandwidth of 1.64 𝑘𝐻𝑧.

58
 Solution cont’d:

Frequency Frequency WBFM

Bandpass
𝑚(𝑡) NBFM Multiplier Converter Filter
𝜷 = 𝟎. 𝟐 x 36
𝒇𝒄 = 𝟏 𝒌𝑯𝒛 𝜷 = 𝟕. 𝟐 𝜷 = 𝟕. 𝟐
𝒇𝒄 = 𝟑𝟔 𝒌𝑯𝒛 𝒇𝒄 = 𝟏 𝟎𝟎 𝒌𝑯𝒛 Centered at
100𝑘𝐻𝑧 ,
Crystal Bandwidth
𝒇 = 𝟏𝟎𝟎 − 𝟑𝟔 Oscillator 1.64 𝑘𝐻𝑧.
= 𝟔𝟒 𝒌𝑯𝒛

59
Direct Generation of WBFM Signals

60
Direct Generation of WBFM Signals

𝒇 𝒕 = 𝒇𝟎 + 𝑲𝟎 𝒗𝒊𝒏 (𝒕)

61
 Demodulation of FM Signals
• The information about the message signal 𝑚(𝑡) in an FM
signal resides in the instantaneous frequency
𝑘𝑓
𝑓𝑖 (𝑡) = 𝑓𝑐 + 𝑚 𝑡 .
2𝜋
Demodulation Methods:
1. Differentiator followed by envelope detector.
2. Phase-shift Quadrature FM demodulator.
3. Zero-crossings detector.

62
 Demodulation of FM Signals
Differentiator (slope detection):
• The information about the message signal 𝑚(𝑡) in an FM
signal resides in the instantaneous frequency
𝑘𝑓
𝑓𝑖 (𝑡) = 𝑓𝑐 + 𝑚 𝑡 .
2𝜋
• To obtain 𝑚(𝑡) we can use a differentiator followed by an
envelope detector.

63
 Demodulation of FM Signals
Ideal Differentiator (slope detection):

′
• The envelope of 𝜑𝐹𝑀 (𝑡) is directly related to the signal 𝑚(𝑡).
• We can use envelope detection if 𝑘𝑓 𝑚𝑝 < 𝜔𝑐 , since the
envelope will have the same shape as 𝑚(𝑡).

64
 Demodulation of FM Signals
Phase-Shift Quadrature FM Demodulator

65
66
67
 Zero-Crossings Detector
• An amplitude limiter is used to generate rectangular pulse
output.

• A frequency counter is designed to measure the

instantaneous frequency from the number of zero crossings.

68
 Advantages of FM

Unlike the FM, in the AM case the distortion clearly makes it impossible to
recover the original 𝑚 𝑡 .

69
70
71
72
73
74
75
76
77
78
79
80
81
82