Fourier series
• powerful tools for representing any
periodic function as a sum of sine and
cosines series.
• Means it has infinite Bandwidth
(BW)
• Provide inside information of the real
time signal in frequency domain or
reciprocal domain
Fourier Series:
Orthogonality conditions
If n and m are different, then
Cos n Sin m =0
Cos n Cos m =0
Sin n Sin m =0
Replace sin and cos with complex function
More than 25 harmonic fn
required to rectangular wave
Each circle represent a
sinusoidal function and
together (in fact you require
infinite numbers) makes a
rectangular function
You all can try to verify by writing series of sinusoidal functions on this link
https://www.desmos.com/calculator
1. sin function
1. Rectangular function
Each single line represent a frequency
Dirichlet Conditions
» A periodic signal x(t), has a Fourier series if it satisfies the
following conditions:
1. x(t) is absolutely integrable over any period, namely
a +T
∫ | x(t ) | dt < ∞,
a
∀a ∈
2. x(t) has only a finite number of maxima and minima over
any period
3. x(t) has only a finite number of discontinuities over any
period
Fourier Transform
» We have seen that periodic signals can be represented with the
Fourier series
» Can aperiodic signals be analyzed in terms of frequency
components?
» Yes, and the Fourier transform provides the tool for this analysis
» The major difference w.r.t. the line spectra of periodic signals is
that the spectra of aperiodic signals are defined for all real
values of the frequency variable 𝜔 not just for a discrete set of
values
Properties of the Fourier Transform
» Linearity: x(t ) ↔ X (ω ) y (t ) ↔ Y (ω )
α x(t ) + β y (t ) ↔ α X (ω ) + β Y (ω )
− jω t0
» Left or Right Shift in Time: x(t − t0 ) ↔ X (ω )e
1 ω
» Time Scaling: x(at ) ↔ X
a a
Properties of the Fourier Transform
» Time Reversal: x(−t ) ↔ X (−ω )
» Multiplication by a Power of t:
n
d
t x(t ) ↔ ( j )
n n
X (ω )
dω n
» Multiplication by a Complex Exponential:
jω 0 t
x(t )e ↔ X (ω − ω 0 )
Examples
Fourier transform time and frequency domain
Correction: At t=0, it should cross 1 in the y-axis. You may consider some kind of
normalization
6 terms of the series
10 terms of the series
20 terms of the series
Sawtooth wave
6 terms of the series
10 terms of the series
20 terms of the series
Triangular wave
Correction: At t=0, it should cross 1 in the y-axis
Oscillations 2
7
8 terms of the series
Wave packets and pulses
Wave packet Pulses
Short burst/envelope Non-sinusoidal waveform
Displacement Displacement
x x
Wave packet propagation
Let us synthesize a wave train by superposing
a number of sinusoidal oscillations spanning a
continuous of frequencies.
Spectral density :
Spectral density :
Displacement
Consider a pulse (here, Gaussian and Rectangular)
Time
Rectangle
Gaussian
Where,
t
At the nodes :
Uncertainty product :
For non-dispersive waves
Signal emitted over a time interval
Travels a distance
Uncertainty product :
= 4π
The value of the product of Δω and Δt depends on the choice of ρ (ω)
Physical significance of uncertainty product
1. For wave packet which last long has very little spread in
frequency
2. A short sharp burst with a small spread in time has within it
a large range of frequencies
eg. Radio and TV signal
Wave packet describe the signal received at a certain point in
space as a function of time
Model of dispersive medium y
x
0,0
k0is spring constant per unit ∆x
T
length
The vertical restoring force
∂2 y 2 ∂ y
2
T k0
= c − ω 2
0 y ; where c =
2
&ω =
2
∂t ∂x µ µ
2 s 2 s 0
y = A e i (ω t − k x ) ω
⇒ ω 2 = cs2 k 2 + ω02
Wave with large k almost non-dispersive propagation ω0
Wave with small k curve deviate from linear behavior
0 k
Cut-off frequency ω0 below which no wave can propagate
