Transforms
by
B Kanmani
Sequence of presentation
1. Fourier Series: FS
2. Fourier Transform: FT
B Kanmani, BMSCE
1. Fourier Series
B Kanmani, BMSCE
The
Fourier representation
of
continuous time periodic signals:
Fourier Series
B Kanmani, BMSCE
The Fourier Series representation
n 1
n 1
x(t ) a0 an cos n0t bn sin n0t
cn cos n0t n
n 0
B Kanmani, BMSCE
Example-I
B Kanmani, BMSCE
Continuous time periodic signal
B Kanmani, BMSCE
Time domain representation
without using Fourier series
x(t )
x t nT
ONE
where
xONE
1 0t T /2
1 T / 2 t T
B Kanmani, BMSCE
Without Fourier Series
X-one: Equation for signal in one
period T
x(t):
Sum of Time shifted X-one
Infinite Sum
B Kanmani, BMSCE
Without Fourier Series
SET of TWO equations
Difficult to perform operations like:
multiplication, differentiation, addition
with another signal
B Kanmani, BMSCE
10
With Fourier Series
SINGLE EQUATION
B Kanmani, BMSCE
11
Same Example
B Kanmani, BMSCE
12
B Kanmani, BMSCE
13
Time domain representation
using Fourier series
4
x(t )
sin n0t
n 1,3,5,... n
B Kanmani, BMSCE
14
ONE term
Fourier series
representation
x(t )
B Kanmani, BMSCE
sin 0t
15
TWO term
Fourier series
representation
x(t )
sin 0t
sin 30t
3
B Kanmani, BMSCE
16
THREE term
Fourier series
representation
x(t )
sin 0t
sin 30t
3
4
sin 50t
5
B Kanmani, BMSCE
17
Eleven term
Fourier series
representation
11
4
x(t )
sin n0t
n 1,3,5,... n
B Kanmani, BMSCE
18
Infinite term
Fourier series
representation
4
x(t )
sin n0t
n 1,3,5,... n
B Kanmani, BMSCE
19
Some observations
Exact representation when infinite
terms are considered
Signal and the cosine basis functions are
assumed to extend to infinity
Original signal is the weighted sum of
harmonics of sinusoidal signals
B Kanmani, BMSCE
20
Additional Information
Frequency Information
B Kanmani, BMSCE
21
Frequency Spectrum
B Kanmani, BMSCE
22
Time-domain
Frequency-domain
B Kanmani, BMSCE
23
Time-domain
T=1 sec
fo=1Hz
Frequency-domain
B Kanmani, BMSCE
24
Time-domain
T=0.5 sec
fo=2 Hz
Frequency-domain
B Kanmani, BMSCE
25
Observation
Compression
in time domain
leads to
expansion
in frequency domain
B Kanmani, BMSCE
26
Example-II
B Kanmani, BMSCE
27
Half-wave rectified wave
B Kanmani, BMSCE
28
ONE TERM
B Kanmani, BMSCE
29
TWO TERMS
B Kanmani, BMSCE
30
FIVE TERMS
B Kanmani, BMSCE
31
TWENTY TERMS
B Kanmani, BMSCE
32
Fourier Series
x(t )
0.5sin(0t )
1
cos n0t
2
n 2,4,6,... ( n 1)
B Kanmani, BMSCE
33
Time-domain
Frequency-domain
B Kanmani, BMSCE
34
Example-III
B Kanmani, BMSCE
35
THE PERIODIC SIGNAL
B Kanmani, BMSCE
36
ONE TERM
B Kanmani, BMSCE
37
TWO TERMS
B Kanmani, BMSCE
38
THREE TERMS
B Kanmani, BMSCE
39
FIVE TERMS
B Kanmani, BMSCE
40
TWENTY TERMS
B Kanmani, BMSCE
41
Fourier Series
1
1
1
x(t ) sin(0t ) sin(20t ) sin(30t ) ...
2
3
B Kanmani, BMSCE
42
Time-domain
Frequency-domain
B Kanmani, BMSCE
43
Fourier Series
Exact time-domain representation
Periodic continuous time signals
From FS, we can get its spectrum
The spectrum is always discrete
The spectrum contains harmonics of
the fundamental
Reducing time period increases
frequency
B Kanmani, BMSCE
44
Fourier Series
Continuous time
Periodic signals
Discrete spectrum
B Kanmani, BMSCE
45
Fourier Series: Application
B Kanmani, BMSCE
46
Fourier Series: Application
B Kanmani, BMSCE
47
Fourier Series: Application
B Kanmani, BMSCE
48
In all cases: output is sine-wave
B Kanmani, BMSCE
49
What should be the filter cut-off?
B Kanmani, BMSCE
50
In all cases: cut-off is about 10 fc
B Kanmani, BMSCE
51
In general
more than 98% of energy
is contained in the
first TEN harmonics
B Kanmani, BMSCE
52
Another Example
B Kanmani, BMSCE
53
Time-domain
T=1 sec + 1 sec
Frequency = 0.5Hz
Frequency-domain
B Kanmani, BMSCE
54
Time-domain
T=1 sec + 3 sec
Frequency = 0.25 Hz
Frequency-domain
B Kanmani, BMSCE
55
Time-domain
T=1 sec + 7 sec
Frequency = 0.125 Hz
Frequency-domain
B Kanmani, BMSCE
56
B Kanmani, BMSCE
57
2. Fourier Transform
B Kanmani, BMSCE
58
Fourier Transform
Continuous time
Non-Periodic signals
Continuous spectrum
B Kanmani, BMSCE
59
The Fourier Transform
X ( )
x(t )e
jt
1
x(t )
2
jt
X ( )e d
B Kanmani, BMSCE
60
Example - I
B Kanmani, BMSCE
61
Example - II
B Kanmani, BMSCE
62
Example - III
B Kanmani, BMSCE
63
Example - IV
B Kanmani, BMSCE
64
Matlab Command
>> fm=500;
>> Fs=20*fm;
>>Ts=1/Fs;
% sine wave frequency
%Actual sampling rate
%Sampling interval
>> time=0:Ts:4.0;
>> x_1=0.5*cos(2*pi*fm*time);
>> sound(x_1,Fs);
B Kanmani, BMSCE
65
Example - V
B Kanmani, BMSCE
66
Example - VI
B Kanmani, BMSCE
67
Thank you
B Kanmani, BMSCE
68
