ELECTRICAL & ELECTRONICS ENGINEERING,

FOURAH BAY COLLEGE, UNIVERSITY OF

SIERRA LEONE

SIGNAL ANALYSIS (EENG-411)

Lecture Presentation By:

Dr. ABDULAI SANKOH

 Continuous Time Fourier Series (CTFS) Analysis
❑ Fourier methods are based on using real or complex
sinusoids as basis functions.
❑ For a linear constant-parameter system, the response to a
sinusoid of frequency  is another sinusoid of exactly the
same frequency modified in amplitude and phase by the linear
system’s transfer function.

❑ If the excitation to a linear

Module system
Content can be expressed as a
Continuous Time Fourier Series (CTFS) then the output can be
found as the modified version using the transfer
characteristics of the system.

❑ Furthermore this concept can be extended to the impulse

response of a linear system, i.e. if the impulse response of a
system is known then its output response to an excitation vi(t)
is the convolution of the input and its impulse response,
where the latter is simply the inverse transform of the transfer
function of the linear system.
 Continuous Time Fourier Series (CTFS) Analysis
❑ The concepts met in this topic have a wide range of
applications in signal analysis and form the basis for many
other more advanced topics in Control Theory and
Communication and Information Theory.
❑ To put in perspective the engineering meaning of the CFTS
expansion and its relevance to spectral analysis, consider a
single sinusoid of peak amplitude Em oscillating at . This
can be represented as :- Module Content
𝒗 𝒕 = 𝑬𝒎 𝒔𝒊𝒏(𝝎𝒕) Eq. (1)
❑ The time domain representation of this signal is shown in Fig.
1.
❑ Em is the peak amplitude, and  the angular frequency in
radians per second.
❑ The frequency of oscillation 𝒇 = 𝝎/𝟐𝝅 and the period of the
waveform is given by 𝑻 = 𝟏/𝒇 .
 Continuous Time Fourier Series (CTFS) Analysis
❑ The frequency is a constant and the amplitude of the
signal is a function of time only
❑ The spectrum or frequency-amplitude characteristics of this
signal can be represented by a vertical line of amplitude Em
and located at fm in the frequency axis as shown in Fig 1b, i.e.
a single frequency component or spectral component located
at the fundamental frequency fm.
Module Content

Fig. 1a Time domain Fig. 1b Frequency domain

representation representation
 Continuous Time Fourier Series (CTFS) Analysis
Lets assume that the x[n] is the time domain cosine signal generating a 2
seconds record of frequency fc=10 Hz that is sampled at a frequency fs=
32∗fc for representing it in the computer memory.
Fc = 10; % carrier frequency
OversampleFactor = 32; % OversampleFactor =
fs = OversampleFactor *fc; % sampling frequency with oversampling factor=32
ts = 1/fs; % sampling interval
T = 2;
t = 0:Ts:T-ts; %Module Content
2 seconds signal duration
s = cos(2*pi*fc*t); % time domain signal (real number)

% plot the signal

subplot(3, 1, 1);
plot(t, s);
xlabel(‘ t = nT_s' ); ylabel(' x[ n] ‘ );
title(' x[ n]=cos(2 \pi 10 t) ‘ );

Fig. 2 Time domain representation

 Continuous Time Fourier Series (CTFS) Analysis
Note:
❑ The FFT length should be sufficient to cover the entire length of the
input signal. If N is less than the length of the input signal, the input
signal will be truncated when computing the FFT.
❑ In this example, the number of samples (N) of the cosine wave of 2
seconds duration is
N = T*fs Eq. (2)
N = T*(OverSampFactor*fs) = 2*32*10 = 640 samples
Module Content
❑ Note that the index for the raw FFT are integers from 1 → N. We need
to process it to convert these integers to frequencies.
❑ Each point/bin in the FFT output array is spaced by the frequency
resolution ∆f, that is calculated as
∆f = fs/N Eq. (3)
❑ where, fs is the sampling frequency and N is the FFT size that is
considered. Thus, for our example, each point in the array is spaced
by the frequency resolution.
∆f = fs/N = 320/640 = 0.5 Hz
 Continuous Time Fourier Series (CTFS) Analysis
Signal Spectrum
Lets consider taking a N = length of signal point FFT, which is the 8th
power of 2.
N= size(s, 2); % length of signal % FFT size N = 256;
S = fft(s, N); % N-point complex DFT, output contains DC at index 1
% Nyquist frequency at N/2+1 th index positive frequencies
% from index 2 to N/2 negative frequencies from index N/2+1 to N
% calculate frequency bins with FFT
df = fs/N % frequency resolution
sampleIndex = 0:N-1; % raw index for FFT
Module plot
Content
f = sampleIndex*df; % x-axis index converted to frequencies

% Plotting the signal in the frequency domain

subplot(3,1,2);
stem(sampleIndex, abs(S)); % sample values on x-axis
title(' X[ k] ' );
xlabel(' k' ); ylabel(' | X(k) | ‘ );

subplot(3,1,3);
stem(f, abs(S)); % x-axis represent frequencies
title(' X[ f] ' );
xlabel(' frequencies (f) ' ); ylabel(' | X(f) | ' );
 Continuous Time Fourier Series (CTFS) Analysis
❑ The sample at the Nyquist frequency ( fs/2) mark the boundary between
the positive and negative frequencies.
❑ Note that the complex numbers surrounding the Nyquist index are
complex conjugates and they represent positive and negative
frequencies respectively.

Module Content

Fig. 3 Magnitude response from FFT plotted against - sample index (top)
and computed frequencies (bottom)
 Continuous Time Fourier Series (CTFS) Analysis
❑ It can be seen that the frequency axis starts with DC, followed by
positive frequency terms which is in turn followed by the negative
frequency terms.
DC → positive frequencies → negative frequencies

❑ Introduce proper order in the x-axis, by rearranging the frequencies in

order. The FFTshift function in Matlab can be used to arrange the
frequencies
Module Content
negative frequencies → DC → positive frequencies.

▪ X[1] represents DC frequency component

▪ X[2] to X[N/2] terms are positive frequency components
▪ X[N/2 + 1] is the Nyquist frequency (Fs/2), which is common to both
the positive and negative frequencies.
▪ X[N/2 + 1] to X[N] terms are negative frequency components.
Continuous Time Fourier Series (CTFS) Analysis

Module Content

Fig. 4 Role of FFTShift in ordering the frequencies

 Continuous Time Fourier Series (CTFS) Analysis

❑ Therefore, when N is even, ordered frequency axis is set as

Module Content
❑ When N is odd, the ordered frequency axis should be set as
 Continuous Time Fourier Series (CTFS) Analysis
%for odd N see equation above
S1 = [(S(N/2+1:N)) S(1:N/2)]; % Method-1 order frequencies without using fftShift
S2 = fftshift(S); % Method-2 order frequencies by using fftshift
df = fs/N; % frequency resolution

sampleIndex = -N/2:N/2-1; % raw index for FFT plot

f = sampleIndex*df; % x-axis index converted to frequencies

% plot ordered spectrum using the two methods

figure; Module Content
subplot(2,1,1);
stem(sampleIndex, abs(S1));
hold on;
stem(sampleIndex, abs(S2), ‘b’ ) % sample index on x-axis
title(' Frequency Domain’ );
xlabel(' k' ); ylabel(' | X(k) | ' );

subplot(2,1,2);
stem(f, abs(S1));
stem(f, abs(S2), ' r' )
xlabel(' frequencies (f) ' ); ylabel('|X(f)|' ); % frequencies on x-axis
 Continuous Time Fourier Series (CTFS) Analysis
❑ Magnitude response of FFT result after applying FFTShift : plotted
against sample index (Fig. 5a) and against computed frequencies
(Fig. 5b).

Fig. 5a

Module Content

Fig. 5b
 Continuous Time Fourier Series (CTFS) Analysis
❑ Fig. 1b representation forms the basis of the
representation of periodic waveforms in the frequency
domain or loosely referred to as the spectrum of the
signal or its spectral behaviour.
Periodic Signals Spectra
❑ Assume a periodic function v(t) with a fundamental period To
expressed in general as:Module Content

v(t) = v(t  nTo).

❑ Using the Continuous Time Fourier Series (CTFS) analysis, v(t)
can be decomposed into the following forms:
𝟐𝝅𝒏𝒕 𝟐𝝅𝒏𝒕
𝒗 𝒕 = 𝒂 𝟎 + σ∞
𝒏=𝟏 𝒂𝒏 𝒄𝒐𝒔 + σ∞
𝒏=𝟏 𝒃𝒏 𝒔𝒊𝒏 (4)
𝑻𝟎 𝑻𝟎

❑ The first term ao is obtained through the integral

𝑻𝟎
𝟏 𝟐
𝒂𝟎 = ‫׬‬ 𝒗 𝒕 𝒅𝒕 (5)
𝑻𝟎 −𝑻𝟎
𝟐
 Continous Time Fourier Series (CTFS) Analysis
❑ ao is the average value of the waveform and often
referred to as the ‘dc’ term. This should remind you of
the normalized power.
❑ an and bn are the CTFS coefficients obtained as follows
𝑻𝟎
𝟐 𝟐 𝟐𝝅𝒏𝒕
𝒂𝒏 = ‫׬‬ 𝒗 𝒕 𝒄𝒐𝒔 𝒅𝒕 𝒏 = 𝟏, 𝟐, . . (6)
𝑻𝟎 −𝑻𝟎 𝑻𝟎
𝟐
𝑻𝟎 Module Content
𝟐 𝟐 𝟐𝝅𝒏𝒕
𝒃𝒏 = ‫׬‬ 𝒗 𝒕 𝒔𝒊𝒏 𝒅𝒕 𝒏 = 𝟏, 𝟐, . . (7)
𝑻𝟎 −𝑻𝟎 𝑻𝟎 v
𝟐

❑ The integral of Eq. (5) to (7) are also referred to as the Euler
formulae for the coefficients of v(t).
❑ It is also worth mentioning that the interval of integration above
i.e, 𝑻𝟎 Τ𝟐 may be replaced by any interval of length T, 𝟎 ≤ 𝒕 ≤ 𝑻
❑ Another form of representation of CTFS is given by :-
𝟐𝝅𝒏𝒕
𝒗 𝒕 = 𝑪𝟎 + σ∞
𝒏=𝟏 𝑪𝒏 𝒄𝒐𝒔( − 𝚽) (8)
𝑻𝟎
 Continous Time Fourier Series (CTFS) Analysis
❑ Where Co= ao
𝑪𝒏 = 𝒂𝟐𝒏 + 𝒃𝟐𝒏 (9)
𝒃𝒏
𝚽𝒏 = tan−𝟏 (10)
𝒂𝒏

❑ Cn in Eq. (9) is referred to as the spectral amplitude and the

spectral components are given by 𝑪𝒏 𝒄𝒐𝒔(𝟐𝝅𝒏𝒇𝟎 𝒕 − 𝚽) at a
frequency of nfo. ThisModule
is a Content
more concise or closed form
representation of the CTFS and is used in the plotting of the
spectrum.
❑ Having revised the expressions of the CTFS expansions let us
now establish the relevance of these equations in terms of
spectral analysis, in other words what is the relevance of these
expressions in the engineering applications?

❑ The integral Eq. (5) to (7) are also referred to as the Euler
formulae for the coefficients of v(t).
 Continous Time Fourier Series (CTFS) Analysis
❑ From Eq. (4), we note that v(t) is made up of a d.c. term Co=ao,
and sinusoids at frequencies ranging from the fundamental,
i.e. n = 1, i.e. at ωo, to infinity, i.e. an infinite number of
harmonics of the fundamental frequency. The amplitudes are
determined by Eq. (9) A sketch of these spectral components
is shown below.

Module Content

Fig. 8 Single sided sketch of the amplitude spectrum of a periodic signal

❑ It should be noted that 𝑪𝟏 = 𝒂𝟐𝟏 + 𝒃𝟐𝟏 , 𝑪𝟐 = 𝒂𝟐𝟐 + 𝒃𝟐𝟐 , 𝑪𝟑 = 𝒂𝟐𝟑 + 𝒃𝟐𝟑

 Continous Time Fourier Series (CTFS) Analysis
❑ You may have noticed that as the harmonics increase, the
spectral amplitudes decreases.
❑ As you try several examples you will observe that this is a
general rule. Interestingly in some cases there occur
variations of this general rule for certain specific harmonics
after which the rule is observed. An example of such a signal
will be dealt with later.
Module Content
Practical significance of spectral plots of Fourier series expansion
❑ The question one may ask now is what really is the practical
significance of these plots in engineering applications?

1. The CTFS analysis provides the engineer or analyst with information

on the spectral behaviour of any waveform that can be so represented,
i.e. its frequency or spectral occupancy.

2. In other words it shows how the spectral components (both in terms of

amplitude and phase) behave as a function of frequency.
 Continous Time Fourier Series (CTFS) Analysis
3. This information also provides us with a means of
determining the minimum bandwidth requirements for the
transmission of signals (with tolerable distortion)

4. as well as their power requirements for filters and

amplifiers since the power of the harmonics is a function of
the square of their amplitudes.
Module Content

❑ The subsequent examples will demonstrate two facts.

I. The level of distortion incurred if enough bandwidth is

not provided for the transmission of signals, i.e. early
truncation of the spectrum of a signal and

II. The nature of the computation and variation of the

spectrum of a periodic signal.
 Continous Time Fourier Series (CTFS) Analysis
❑ E.g-1. Determine the CTFS expression considering the square wave
signal s(t), one period of which is illustrated in the Figure. The
amplitude V and period T of the square wave are unity and 2
seconds respectively.
s(t)

T/2 Content
Module T
t

❑ E.g-2. Represent the triangular waveform shown in the figure below

as a CTFS. f(t)
A V

-T/2 0 T/4 T/2

t

B
 Continous Time Fourier Series (CTFS) Analysis
❑ E.g-1 solution
s(t)
V

T/2 T
t

❑ This waveform is an odd function,

Module Content
which means that a0=0, and an=0 .
The period, T is 2 and the waveform over one period is given by

𝑽 𝟎 ≤ 𝒕 ≤ 𝑻Τ𝟐 𝟏 𝟎≤𝒕≤𝟏
𝒔(𝒕) = ൝ =ቊ
−𝑽 𝑻Τ𝟐 ≤ 𝒕 ≤ 𝑻 −𝟏 𝟏≤𝒕≤𝟐

❑ The Fourier coefficient an and bn are calculated using Eqs. (5) and (7)
𝑻 𝟐
𝒂𝒏 = න 𝒔 𝒕 𝒄𝒐𝒔 𝒏𝝎𝒕 𝒅𝒕 = න 𝒔 𝒕 𝐜𝐨𝐬 𝒏𝝅𝒕 𝒅𝒕
𝟎 𝟎
𝟏 𝟐 𝟏 𝟐
𝟏 𝟏
= න 𝐜𝐨𝐬 𝒏𝝅𝒕 𝒅𝒕 − න 𝒄𝒐𝒔 𝒏𝝅𝒕 𝒅𝒕 = . 𝐬𝐢𝐧 𝒏𝝅𝒕 − . 𝐬𝒊𝒏 𝒏𝝅𝒕 =𝟎
𝟎 𝟏 𝒏𝝅 𝟎
𝒏𝝅 𝟏
 Continous Time Fourier Series (CTFS) Analysis
❑ E.g-3. Find the CTFS of the half-wave rectified sinusoid.
s(t) sin(t)
1

E.g-1 solution s(t) Module Content

V

T/2 T
t

❑ This waveform is an odd function, which means that a0=0, and an=0 .
The period, T is 2 and the waveform over one period is given by

𝑽 𝟎 ≤ 𝒕 ≤ 𝑻Τ𝟐 𝟏 𝟎≤𝒕≤𝟏
𝒔(𝒕) = ൝ =ቊ
−𝑽 𝑻Τ𝟐 ≤ 𝒕 ≤ 𝑻 −𝟏 𝟏≤𝒕≤𝟐
 Continous Time Fourier Series (CTFS) Analysis
❑ The Fourier coefficient an and bn are calculated using Eqs. (5) and (7)
𝑻 𝟐
𝒂𝒏 = න 𝒔 𝒕 𝒄𝒐𝒔 𝒏𝝎𝒕 𝒅𝒕 = න 𝒔 𝒕 𝐜𝐨𝐬 𝒏𝝅𝒕 𝒅𝒕
𝟎 𝟎
𝟏 𝟐 𝟏 𝟐
𝟏 𝟏
= න 𝐜𝐨𝐬 𝒏𝝅𝒕 𝒅𝒕 − න 𝒄𝒐𝒔 𝒏𝝅𝒕 𝒅𝒕 = . 𝐬𝐢𝐧 𝒏𝝅𝒕 − . 𝐬𝒊𝒏 𝒏𝝅𝒕 =𝟎
𝟎 𝟏 𝒏𝝅 𝟎
𝒏𝝅 𝟏

❑ For this particular waveform all theContent

Module A coefficient are zero and hence
there will be no cosine terms in the summation of Eq. (4). The Bn
coefficient are calculated in similar way
𝑻 𝟐
𝑩𝒏 = න 𝒔 𝒕 𝐬𝐢𝐧 𝒏𝝎𝒕 𝒅𝒕 = න 𝒔 𝒕 𝐬𝐢𝐧 𝒏𝝅𝒕 𝒅𝒕
𝟎 𝟎
𝟏 𝟐 𝟏 𝟐
𝟏 𝟏
𝑩𝒏 = න 𝐬𝒊𝒏 𝒏𝝅𝒕 𝒅𝒕 − න 𝐬𝐢𝐧 𝒏𝝅𝒕 𝒅𝒕 = − . 𝐜𝐨𝐬 𝒏𝝅𝒕 − − . 𝐜𝐨𝐬 𝒏𝝅𝒕
𝟎 𝟏 𝒏𝝅 𝟎
𝒏𝝅 𝟏

𝟏 𝟏 𝟐
𝑩𝒏 = 𝒏𝝅 𝟏 − 𝐜𝐨𝐬(𝒏𝝅) + 𝒏𝝅 𝒄𝒐𝒔(𝟐𝒏𝝅) − 𝐜𝐨𝐬(𝒏𝝅) = 𝒏𝝅 (𝟏 − 𝒄𝒐𝒔 𝒏𝝅 )
 Continous Time Fourier Series (CTFS) Analysis
❑ For even values of n the B coefficient are all zeros. Therefore the
trigonometric CTFS representation of the square waveform is
∞
𝟐
𝒔(𝒕) = ෍ 𝟏 − 𝒄𝒐𝒔 𝒏𝝅 . 𝐬𝒊𝒏 𝒏𝝅𝒕
𝒏𝝅
𝒏=𝟏

𝟒
𝒔 𝒕 = 𝐬𝐢 𝐧 𝝅𝒕 (Fundamental)
𝝅
+𝟎 (Second harmonic)
𝟒
+ Module Content
𝒔𝒊𝒏(𝟑𝝅𝒕) (third harmonic)
𝟑𝝅
E.g-2 solution
f(t)
A (0,V) V

-T/2 T/4 T/2

0
t

B (T/2,-V)

❑ The function is an even function since it is symmetrical about the

f(t)-axis, therefore bn = 0.
 Continous Time Fourier Series (CTFS) Analysis
❑ Also, the function has half-wave symmetry, therefore a0 = 0.
❑ Thus, the CTFS series expression of Eq. (4) can be rewritten as
𝑻
𝟐𝝅𝒏𝒕 𝟐 𝟐 𝟐𝝅𝒏𝒕
𝒇 𝒕 = σ∞ 𝒂
𝒏=𝟏 𝒏 𝒄𝒐𝒔 and 𝒂𝒏 = ‫𝒇 ׬‬ 𝒕 𝒄𝒐𝒔 𝒅𝒕
𝑻 𝑻 −𝑻 𝑻
𝟐

❑ The periodic interval [t0, t0 + T] can be chosen rather than [0, T]. For
this example, because of the form of f(t), it is convenient to choose
periodic interval to be [- T /2, T /2].
𝟐 𝟐
𝑻 Module Content
𝟐𝝅𝒏𝒕
𝒂𝒏 = ‫𝒇 ׬‬ 𝒕 𝒄𝒐𝒔( ) 𝒅𝒕
𝑻𝟎 −𝑻 𝑻
𝟐

❑ Since both f(t) and cos(𝟐𝝅𝒏𝒕/T) are even functions, the interval must
have the same value between -T/2 to 0 and between 0 and T/2. As
such, an can be rewritten as;
𝑻
𝟒 𝟐 𝟐𝝅𝒏𝒕
𝒂𝒏 = ‫𝒇 ׬‬ 𝒕 𝒄𝒐𝒔( ) 𝒅𝒕
𝑻 𝟎 𝑻 X1,y1 X2,y2
❑ Determine the equation of the line between A (0,V) and B(T/2,-V),
𝟒𝑽 𝟒𝑽 𝑻
𝒇 𝒕 =− 𝒕 +𝑽=− (𝒕 − )
𝑻 𝑻 𝟒
 Continous Time Fourier Series (CTFS) Analysis
❑ -4V/T is the slope of the line for the interval 𝟎 ≤ 𝒕 ≤ 𝑻Τ𝟐 , therefore,
𝑻
𝟒 𝟐 𝟒𝑽 𝑻 𝟐𝝅𝒏𝒕
𝒂𝒏 = ‫ ׬‬−𝑻 𝒕− . 𝒄𝒐𝒔( ) 𝒅𝒕
𝑻 𝟎 𝟒 𝑻

𝑻 𝑻
𝟏𝟔𝑽 𝟐𝝅𝒏𝒕 𝟒𝑽 𝟐𝝅𝒏𝒕
𝒂𝒏 = − 𝟐 ‫𝒕 𝟐𝟎׬‬. 𝒄𝒐𝒔( ) 𝒅𝒕 + ‫(𝒔𝒐𝒄 𝟐𝟎׬‬ ) 𝒅𝒕
𝑻 𝑻 𝑻 𝑻

❑ Take the integration of an by part.

𝑻Τ𝟐
𝟐
𝟏𝟔𝑽 𝑻 𝟐𝝅𝒏𝒕 𝑻 𝟐𝝅𝒏𝒕
𝒂𝒏 = − 𝒕. 𝒔𝒊𝒏 Module
− 𝟏 . Content. −𝒄𝒐𝒔 + ⋯.
𝑻𝟐 𝟐𝒏𝝅 𝑻 𝟒𝒏𝟐 𝝅𝟐 𝑻
𝟎
𝑻Τ𝟐
𝟒𝑽 𝑻 𝟐𝝅𝒏𝒕
𝒔𝒊𝒏
𝑻 𝟐𝒏𝝅 𝑻 𝟎
❑ Substitute the values of the integral limit.
𝟏𝟔𝑽 𝑻𝟐 𝑻𝟐 𝑻𝟐 𝟐𝑽
𝒂𝒏 = − 𝟐 𝒔𝒊𝒏 𝒏𝝅 + 𝟐 𝟐 𝒄𝒐𝒔 𝒏𝝅 − 𝟐 𝟐 + 𝒔𝒊𝒏 𝒏𝝅
𝑻 𝟒𝒏𝝅 𝟒𝒏 𝝅 𝟒𝒏 𝝅 𝝅𝒏

𝟏𝟔𝑽 𝑻𝟐 𝑻𝟐 𝟐𝑽
𝒂𝒏 = − 𝟐 𝒄𝒐𝒔 𝒏𝝅 − 𝟏 + 𝒔𝒊𝒏 𝒏𝝅 + 𝒔𝒊𝒏 𝒏𝝅
𝑻 𝟒𝒏𝟐 𝝅𝟐 𝟒𝒏𝝅 𝝅𝒏
 Continous Time Fourier Series (CTFS) Analysis
❑ Hence, 𝒔𝒊𝒏 𝒏𝝅 = 𝟎 for all n and 𝒄𝒐𝒔 𝒏𝝅 = 𝟏 for n = even integer
number and -1 for n = odd integer number. Thus
∞
𝟖𝑽 𝟐𝝅𝒏𝒕
𝒇 𝒕 = ෍ 𝟐 𝟐 . 𝒄𝒐𝒔
𝒏 𝝅 𝑻
𝒏=𝒐𝒅𝒅

𝟖𝑽 𝟐𝒏𝝅𝒕 𝟏 𝟐𝒏𝝅𝒕 𝟏 𝟐𝒏𝝅𝒕

𝒇 𝒕 = 𝟐 𝒄𝒐𝒔 + 𝒄𝒐𝒔 + 𝒄𝒐𝒔 +⋯
𝝅 𝑻 𝟗 𝑻 𝟐𝟓 𝑻

Module Content
E.g-3 Solution
s(t) sin(t)
1

t
❑ The function s(t) can be represented over one period given as:

𝒔𝒊𝒏(𝒕) 𝟎 ≤ 𝒕 ≤ 𝝅
𝒔(𝒕) = ቊ
𝟎 𝝅 ≤ 𝒕 ≤ 𝟐𝝅
 Continous Time Fourier Series (CTFS) Analysis
❑ Since there is no half-wave symmetry, a DC value exist. Therefore
𝟏 𝑻 𝟏 𝝅 𝟏 𝟐𝝅
𝒂𝟎 = ‫ 𝒕𝒅 𝒕 𝒏𝒊𝒔 𝟎׬ = 𝒕𝒅 𝒕 𝒔 𝟎׬‬+ ‫= 𝒕𝒅 𝟎 𝟎׬‬
𝑻 𝟐𝝅 𝟐𝝅
𝟏 𝝅 𝟏 𝟏
𝒂𝟎 = −𝒄𝒐𝒔 𝒕 𝟎 = 𝟏+𝟏 =
𝟐𝝅 𝟐𝝅 𝝅
❑ It can be seen that the function is neither an odd nor even symmetry.
Therefore, the values of an and bn must be determine. The CTFS
representation of a periodic signal is given by
Module Content
s 𝒕 = 𝒂𝟎 + σ∞
𝒏=𝟏 𝒂𝒏 𝒄𝒐𝒔(𝒏𝝎𝟎 𝒕) + 𝒃𝒏 𝒔𝒊𝒏(𝒏𝝎𝟎 𝒕)
𝟐 𝑻
𝒂𝒏 = ‫ 𝒕 𝒔 𝟎׬‬. 𝒄𝒐𝒔(𝒏𝝎𝟎 𝒕) 𝒅𝒕
𝑻
𝟐 𝑻
𝒃𝒏 = ‫ 𝒕 𝒔 𝟎׬‬. 𝒔𝒊𝒏(𝒏𝝎𝟎 𝒕) 𝒅𝒕
𝑻
𝟐 𝑻
𝒃𝒏 = ‫𝒏𝒊𝒔 ׬‬ 𝒕 . 𝒔𝒊𝒏 𝒏𝒕 𝒅𝒕 𝝎𝟎 = 𝟏
𝟐𝝅 𝟎

❑ By the trigonometric function product relationship

𝟏 𝟏
𝒔𝒊𝒏 𝒕 . 𝒔𝒊𝒏 𝝅𝒕 = 𝒄𝒐𝒔 𝒕 − 𝒏𝒕 − 𝒄𝒐𝒔 𝒕 + 𝒏𝒕
𝟐 𝟐
 Continous Time Fourier Series (CTFS) Analysis
❑ Therefore, bn can be written as:
𝟏 𝑻 𝟏 𝟏
𝒃𝒏 = ‫׬‬ 𝒄𝒐𝒔 𝟏−𝒏 𝒕− 𝒄𝒐𝒔 𝟏 + 𝒏 𝒕 𝒅𝒕
𝝅 𝟎 𝟐 𝟐
𝝅
𝟏 𝟏 𝟏
𝒃𝒏 = 𝒔𝒊𝒏 𝟏 − 𝒏 𝒕 − 𝒔𝒊𝒏 𝟏 + 𝒏 𝒕
𝝅 𝟐(𝟏 − 𝒏) 𝟐(𝟏 + 𝒏) 𝟎
𝟏 𝟏 𝟏
𝒃𝒏 = 𝒔𝒊𝒏 𝝅 − 𝒏𝝅 − 𝒔𝒊𝒏 𝝅 + 𝒏𝝅 − 𝟎 + 𝟎
𝝅 𝟐(𝟏 − 𝒏) 𝟐 𝟏 +
Module Content 𝒏
𝟏 𝟏 𝟏
𝒃𝒏 = 𝒔𝒊𝒏 𝝅 − 𝒏𝝅 − 𝒔𝒊𝒏 𝝅 + 𝒏𝝅
𝟐𝝅 𝟏 − 𝒏 𝟏+𝒏
❑ Substituting any positive value for n into the above expression, bn = 0.
For n = 1, the undetermined form 0/0 is obtained for the first term.
Using L ’Hospital’s rule, 1/(𝟏 − 𝒏).sin 𝝅 − 𝒏𝝅 can be evaluated for n =
1 as follows;
𝒈𝟏 𝒏 𝒔𝒊𝒏 𝝅−𝒏𝝅 where 𝒈𝟏 𝒏 = 𝒔𝒊𝒏 𝝅 − 𝒏𝝅
𝒈 𝒏 = = ,
𝒈𝟐 𝒏 𝟏−𝒏
𝒈𝟐 𝒏 = 𝟏 − 𝒏
 Continous Time Fourier Series (CTFS) Analysis
❑ According to L ’Hospital’s rule, g(n) can be evaluated as:
𝒈′𝟏 𝒏 −𝝅𝒄𝒐𝒔 𝝅−𝒏𝝅
G 𝒏 = =
𝒈′𝟐 𝒏 −𝟏

❑ If G(1) exists then g(1) = G(1), thus

𝒈 𝟏 = 𝝅 𝒄𝒐𝒔 𝝅 − 𝝅 = 𝝅 and
𝟏 𝟏
𝒃𝟏 = 𝝅 =
𝟐𝝅 𝟐Module Content

❑ The only sine component in the CTFS is ½ sin(t)

❑ Evaluating an
𝟏 𝑻
𝒂𝒏 = ‫ 𝒕 𝒏𝒊𝒔 𝟎׬‬. 𝒄𝒐𝒔 𝒏𝒕 𝒅𝒕
𝝅
𝟏 𝑻 𝟏 𝟏
𝒂𝒏 = ‫𝟎׬‬ 𝒔𝒊𝒏 𝟏 + 𝒏 𝒕 + 𝒔𝒊𝒏 𝟏 − 𝒏 𝒕 𝒅𝒕
𝝅 𝟐 𝟐
𝝅
𝟏 𝟏 𝟏
𝒂𝒏 = − 𝒄𝒐𝒔 𝟏 + 𝒏 𝒕 − 𝒄𝒐𝒔 𝟏 − 𝒏 𝒕
𝟐𝝅 𝟏 + 𝒏 𝟏−𝒏 𝟎
 Continous Time Fourier Series (CTFS) Analysis
𝟏 𝟏 𝟏 𝟏 𝟏
𝒂𝒏 = − 𝒄𝒐𝒔 𝝅 + 𝒏𝝅 − 𝒄𝒐𝒔 𝝅 − 𝒏𝝅 − ( + )
𝟐𝝅 𝟏 + 𝒏 𝟏−𝒏 𝟏+𝒏 𝟏−𝒏

❑ n = 1, yield
𝟏 𝟏 𝟏 𝟏 𝟏
𝒂𝟏 = − + −( + )
𝟐𝝅 𝟐 𝟎 𝟐 𝟎
❑ The two indeterminate terms are both equal to zero by L ’Hospital’s
rule, giving 𝒂𝟏 =0 Module Content

❑ n = 2, yield
𝟏 𝟏 𝟏 𝟐
𝒂𝟐 = − − +𝟏− +𝟏 =−
𝟐𝝅 𝟑 𝟑 𝟑𝝅

❑ n = 3, yield

𝟏 𝟏 𝟏 𝟏 𝟏
𝒂𝟑 = − − − + =𝟎
𝟐𝝅 𝟒 𝟑 𝟒 𝟑
 Continous Time Fourier Series (CTFS) Analysis
❑ Continue with integer values of n>0, it can be observed that for all
odd harmonics (n = 1, 3, 5,..) 𝒂𝒏 is zero, and all even harmonics (n = 2,
4, 6,..)
𝟏 𝟏 𝟏 𝟏 𝟏
𝒂𝒏 = − − − − −
𝟐𝝅 𝟏 + 𝒏 𝟏 − 𝒏 𝟏 + 𝒏 𝟏 − 𝒏

𝟏 𝟐 𝟐 𝟏 𝟐 𝟐 𝟐 𝟏
𝒂𝒏 = − − − = Content +
Module
=
𝟐𝝅 𝟏 + 𝒏 𝟏 − 𝒏 𝟐𝝅 𝟏 + 𝒏 𝟏 − 𝒏 𝝅 (𝟏 − 𝒏𝟐 )

❑ Hence, the half-wave rectified sine wave can be expressed as

𝟏 𝟏 𝟐 𝟏
s 𝒕 = + 𝒔𝒊𝒏 𝒕 + σ∞
𝒏=𝟐,𝟒,𝟔,… 𝒄𝒐𝒔 𝒏𝒕
𝝅 𝟐 𝝅 𝟏−𝒏𝟐
 Assignment 1
❑ Determine the CTFS series of the waveforms shown in Fig. 1a
& b and sketch their spectra up to the fifth harmonic in the
range -2π≤t≤2π. Plot their time function on the same graph
using :
a) The first two terms only of the Fourier series
b) The first three terms only of the series
c) The first four terms only of the series.
d) Comment on the Moduleeffect Content
of these truncations on how
accurately the Fourier series expansion approximates the
original time function.

Fig. 1a Fig. 1b
 Assignment Submission Mode and Deadline
Submission Mode

❑ Paper form submission

❑ MATLAB m-File submission via email: [email protected]

Module Content
Submission Deadline
❑ One week from date of discussion.
THANK YOU