Analog Modulation Lecture Notes in “Analog Communications & Noise”

by: Dr. Mohammed AlMahamdy Electrical Engineering | University of Anbar

3.6 FREQUENCY MODULATION

We saw in AM that the information contents of 𝑚(𝑡) is transmitted through changing the
amplitude of the carrier in proportion to the amplitude of 𝑚(𝑡). In the angle modulation, the
information content is transmitted through changing the frequency of the carrier instead. The
instantaneous frequency of the carrier is determined by the value of the baseband signal.

In this modulation type, the carrier frequency varies in a proportion to the amplitude of the
message signal 𝑚(𝑡).

3.6.1 Definitions
Let us consider the carrier function as: 𝑦(𝑡) = 𝐴𝑐 cos 𝜃(𝑡), where 𝜃(𝑡) is the angle.
𝑑𝜃(𝑡)
Since the instantaneous frequency 𝜔𝑖 (𝑡) is the slop of 𝜃(𝑡), or 𝜔𝑖 (𝑡) = , then
𝑑𝑡

𝜃(𝑡) = ∫ 𝜔𝑖 (𝜀)𝑑𝜀
0

In the case of the Phase Modulation (PM), the angle varies linearly with 𝑚(𝑡) as:

𝜃(𝑡) = 𝜔𝑐 𝑡 + 𝑘𝑃 𝑚(𝑡)

-41-
Lecture Notes in “Analog Communications & Noise” Analog Modulation
Electrical Engineering | University of Anbar by: Dr. Mohammed AlMahamdy

where: 𝑘𝑃 = modulation sensitivity = constant, and 𝜔𝑐 = the center frequency at 𝑚(𝑡) = 0.

The resulting PM wave is:

𝑦𝑃𝑀 (𝑡) = 𝐴𝑐 cos[𝜔𝑐 𝑡 + 𝑘𝑃 𝑚(𝑡)]

The instantaneous frequency in this case is:

𝑑𝜃(𝑡)
𝜔𝑖 (𝑡) = = 𝜔𝑐 + 𝑘𝑃 𝑚′ (𝑡)
𝑑𝑡
In words: the instantaneous frequency varies linearly with the derivative of the modulating
signal.

In the case of the Frequency Modulation (FM), the instantaneous frequency 𝜔𝑖 (𝑡) is varied
linearly with the modulating signal, or:

𝜔𝑖 (𝑡) = 𝜔𝑐 + 𝑘𝐹 𝑚(𝑡)

where: 𝑘𝐹 = modulation sensitivity = constant. The angle of the carrier becomes:

𝑡

𝜃(𝑡) = ∫[𝜔𝑐 + 𝑘𝐹 𝑚(𝜀)]𝑑𝜀 + 𝜃0

0

If we assume 𝜃0 = 0, the FM signal becomes:

𝑡

𝑦𝐹𝑀 (𝑡) = 𝐴𝑐 cos (𝜔𝑐 𝑡 + 𝑘𝐹 ∫ 𝑚(𝜀)𝑑𝜀)

0

In words: the instantaneous frequency varies linearly with the integral of the modulating signal.

FM and PM are closely related to each other; if we know the properties of the one, we can
determine those of the other. For this reason, the material on angle modulation hereafter is
devoted to FM.

To simplify the analysis of this equation, we consider 𝑚(𝑡) = 𝐴𝑚 cos(𝜔𝑚 𝑡). So,

𝜔𝑖 (𝑡) = 𝜔𝑐 + 𝑘𝐹 𝐴𝑚 cos 𝜔𝑚 𝑡 = 𝜔𝑐 + ∆𝑓 cos 𝜔𝑚 𝑡

Where ∆𝑓 = 𝐴𝑚 𝑘𝐹 =is the peak frequency deviation or the maximum frequency shift away from
𝑓𝑐 (Note: ∆𝑓 ≪ 𝑓𝐶 ). The instantaneous angle of the carrier is:
𝑡
∆𝑓
𝜃(𝑡) = ∫ 𝜔𝑖 (𝜀)𝑑𝜀 = 𝜔𝑐 𝑡 + sin 𝜔𝑚 𝑡 , at 𝜃0 = 0
𝜔𝑚
0

⇒ 𝜃(𝑡) = 𝜔𝑐 𝑡 + 𝛽 sin 𝜔𝑚 𝑡

-42-
Analog Modulation Lecture Notes in “Analog Communications & Noise”
by: Dr. Mohammed AlMahamdy Electrical Engineering | University of Anbar
∆𝑓
Where 𝛽 = = modulation index of the FM signal
𝑓𝑚

Hence, the formula of the FM waveform will be:

𝑦𝐹𝑀 (𝑡) = 𝐴𝑐 cos[𝜔𝑐 𝑡 + 𝛽 sin(𝜔𝑚 𝑡)]

Since the peak deviation is given by ∆𝑓 , the carrier swing = 2∆𝑓 .

FM

f
0

Note that:

o There is no ‘over-modulation’ situation with FM signal.

o 𝛽 can take on any value from 0 to infinity. Its range is not limited as it is for AM.
o As 𝛽 is increased, the signal becomes more resistant to interfering noise however
occupies more bandwidth.

3.6.2 Spectrum of FM Signals

The spectrum of FM signals is rather ‘messy’ as it has may different sidebands spaced at
multiples of 𝑓𝑚 from the carrier. As a result, the bandwidth needed to accommodate an FM
signal is greater than that for an AM signal having the same modulating frequency, (Except for
NBFM).

According to the value of 𝛽, we have the following FM formats:

NARROW BAND FM
In the FM systems where 𝛽 is small (< 0.2), several approximations lead to the following
method of FM generation:
𝑡

𝑦(𝑡) = 𝐴𝑐 [cos(𝜔𝑐 𝑡) + 𝑘𝐹 sin(𝜔𝑐 𝑡) × ∫ 𝑚(𝜀)𝑑𝜀 ]

0

t +
m(t) 
0
dτ k
+
Ac y(t)=NBFM

-90o
cos(wct)

-43-
Lecture Notes in “Analog Communications & Noise” Analog Modulation
Electrical Engineering | University of Anbar by: Dr. Mohammed AlMahamdy

We can easily notice in the Bessel chart that the spectrum and BW occupied by this NBFM
waveform is 2𝑓𝑚 (just like the AM-DSB).

WIDE BAND FM
To determine the spectrum of FM signal with values of 𝛽 > 0.2 (which is typical), we may
simplify matters by using the complex representation of band-pass signals when 𝑚(𝑡) =
𝐴𝑚 cos 𝜔𝑚 𝑡:

𝑦(𝑡) = 𝐴𝑐 cos[𝜔𝑐 𝑡 + 𝛽 sin 𝜔𝑚 𝑡]

= Re[𝐴𝑐 exp(𝑗𝜔𝑐 𝑡 + 𝑗𝛽 sin(𝜔𝑚 𝑡))]

= Re[𝑦̃(𝑡) exp(𝑗𝜔𝑐 𝑡)]

where 𝑦̃(𝑡) is the complex envelope of the FM signal 𝑦(𝑡), defined by:

𝑦̃(𝑡) = Re[𝐴𝑐 exp(𝑗𝛽 sin(𝜔𝑚 𝑡))]

Thus, unlike the original FM signal 𝑦(𝑡), the complex envelope 𝑦̃(𝑡) is a periodic function of time
with a fundamental frequency equal to the modulation frequency 𝑓𝑚 . We may therefore expand
𝑦̃(𝑡) in the form of a complex Fourier series as follows:
∞

𝑦̃(𝑡) = ∑ 𝐶𝑛 exp(𝑗𝜔𝑚 𝑛𝑡)

𝑛=−∞

where the complex Fourier coefficient 𝐶𝑛 is defined by

𝜋/𝜔𝑚

𝐶𝑛 = 𝑓𝑚 ∫ 𝑦̃(𝑡) exp(−𝑗𝜔𝑚 𝑛𝑡) 𝑑𝑡

−𝜋/𝜔𝑚

𝜋/𝜔𝑚

= 𝑓𝑚 𝐴𝑐 ∫ exp(𝑗𝛽 sin(𝜔𝑚 𝑡) − 𝑗𝜔𝑚 𝑛𝑡) 𝑑𝑡

−𝜋/𝜔𝑚

Define a new variable 𝑥 = 𝜔𝑚 𝑡,

𝜋
𝐴𝑐
𝐶𝑛 = ∫ exp(𝑗𝛽 sin(𝑥) − 𝑗𝑛𝑥) 𝑑𝑥
2𝜋
−𝜋

Here the integral on the right-hand side, is recognized as the nth order Bessel function of the
first kind and argument 𝛽. This function is commonly denoted by the symbol 𝐽𝑛 (𝛽), as shown
by:

-44-
Analog Modulation Lecture Notes in “Analog Communications & Noise”
by: Dr. Mohammed AlMahamdy Electrical Engineering | University of Anbar
𝜋
1
𝐽𝑛 (𝛽) = ∫ 𝑒 𝑗(𝛽 sin(𝑥)−𝑛𝑥) 𝑑𝑥
2𝜋
−𝜋

Accordingly, we may reduce 𝐶𝑛 Equation to 𝐶𝑛 = 𝐴𝑐 𝐽𝑛 (𝛽).

We substitute to get the complex envelope of the FM signal in terms of the 𝐽𝑛 (𝛽)
∞

𝑦̃(𝑡) = 𝐴𝑐 ∑ 𝐽𝑛 (𝛽) exp(𝑗𝜔𝑚 𝑛𝑡)

𝑛=−∞

And
∞

𝑦(𝑡) = 𝐴𝑐 Re [ ∑ 𝐽𝑛 (𝛽) exp[𝑗2𝜋(𝑓𝑐 + 𝑛𝑓𝑚 )𝑡]]

𝑛=−∞
∞

∴ 𝑦(𝑡) = 𝐴𝑐 ∑ 𝐽𝑛 (𝛽) cos[2𝜋(𝑓𝑐 + 𝑛𝑓𝑚 )𝑡] Eq.1

𝑛=−∞

𝐽0 (𝛽) cos[2𝜋𝑓𝑐 𝑡]
+𝐽1 (𝛽) cos[2𝜋(𝑓𝑐 + 𝑓𝑚 )𝑡] + 𝐽1 (𝛽) cos[2𝜋(𝑓𝑐 − 𝑓𝑚 )𝑡]
= 𝐴𝑐 +𝐽2 (𝛽) cos[2𝜋(𝑓𝑐 + 2𝑓𝑚 )𝑡] + 𝐽2 (𝛽) cos[2𝜋(𝑓𝑐 − 2𝑓𝑚 )𝑡]
+𝐽3 (𝛽) cos[2𝜋(𝑓𝑐 + 3𝑓𝑚 )𝑡] + 𝐽3 (𝛽) cos[2𝜋(𝑓𝑐 − 3𝑓𝑚 )𝑡]
{… }

𝐽0 (𝛽) cos[2𝜋𝑓𝑐 𝑡]
+2𝐽1 (𝛽) cos[2𝜋(𝑓𝑐 + 𝑓𝑚 )𝑡]
= 𝐴𝑐 +2𝐽2 (𝛽) cos[2𝜋(𝑓𝑐 + 2𝑓𝑚 )𝑡]
+2𝐽3 (𝛽) cos[2𝜋(𝑓𝑐 + 3𝑓𝑚 )𝑡]
{… }

The discrete spectrum of 𝑦(𝑡) is obtained by taking the Fourier transforms of both sides, yields:
∞
𝐴𝑐
𝑌(𝑓) = ∑ 𝐽𝑛 (𝛽)[𝛿(𝑓 − 𝑓𝑐 − 𝑛𝑓𝑚 ) + 𝛿(𝑓 + 𝑓𝑐 + 𝑛𝑓𝑚 )] Eq.2
2
𝑛=−∞

Knowing that: ∑ 𝐽𝑛2 (𝛽) = 1 Eq.3

𝑛=−∞

In the figure below, we plot the Bessel function 𝐽𝑛 (𝛽) versus the modulation index 𝛽 for
different positive integer values of 𝑛, (see the properties of Bessel function: Str. Page 291)

-45-
Lecture Notes in “Analog Communications & Noise” Analog Modulation
Electrical Engineering | University of Anbar by: Dr. Mohammed AlMahamdy

Thus, using Equations 2 & 3 and the curves of the above Figure, we may observe:

(1) The spectrum of an FM signal contains a carrier component and an infinite set of side
frequencies located symmetrically on either side of the carrier at frequency separations
of 𝑓𝑚 , 2𝑓𝑚 , 3𝑓𝑚 …
(2) For the special case of 𝛽 < 0.2 , only the Bessel coefficients 𝐽0 (𝛽) and 𝐽1 (𝛽) have
significant values, so that the FM signal is effectively composed of a carrier and a single
pair of side frequencies at 𝑓𝑐 ± 𝑓𝑚 . (NBFM)
(3) The amplitude of the carrier component varies with 𝛽 according to 𝐽0 (𝛽). That is, unlike
an AM signal, the amplitude of the carrier component of an FM signal is dependent on 𝛽.
As the envelope of an FM signal is constant, so that the average power of such a signal is:
∞
𝐴2𝑐 𝐴2𝑐
𝑃 = ̅̅̅̅̅̅̅
𝑦 2 (𝑡) = , 𝑃𝑇𝑜𝑡𝑎𝑙 = ∑ 𝐽𝑛2 (𝛽)
2𝑅 2𝑅
𝑛=−∞

To sketch the spectrum of an FM signal at a certain 𝛽, we must get the values of 𝐽𝑛 (𝛽) from a
plot or a table of Bessel function (see Error! Reference source not found.). So, the plot may
look like the following depiction, which is plotted using Eq.2 (Note: we assume 𝐴𝑐 = 2V)

-46-
Analog Modulation Lecture Notes in “Analog Communications & Noise”
by: Dr. Mohammed AlMahamdy Electrical Engineering | University of Anbar

J1 J1
J2 J2
FM(f) J3 J0 J3

FC + 2Fm

FC + 3Fm
FC - 3Fm

FC - 2Fm

FC

FC + Fm
FC - Fm
It is evident that, the frequency modulation of 𝑚(𝑡) = cos 𝜔𝑚 𝑡 has infinite number of
sidebands. However, the magnitude of the spectral components of the higher order sidebands
becomes negligible

1
 = 0.2

f
B
1

 =1
f
B
1

 =2
f
B
1

 =5
f
B
1

 =8
f
B

-47-
Lecture Notes in “Analog Communications & Noise” Analog Modulation
Electrical Engineering | University of Anbar by: Dr. Mohammed AlMahamdy

-48-