Scribd
Upload
90 views

Homework 10

1) The Fourier transform of a signal x(t) is calculated from its derivative. Modulating a tenor voice signal and applying a filter is proposed to overlap the frequency ranges of tenor and alto voices. 2) A filter is suggested to select frequencies between 140Hz and 220Hz from a modulated tenor voice. However, the overlapping frequencies from modulation will distort the output. 3) The response of an LTI system to e^-αtu(t) is δ(t). The response to e^-αtcos(βt)u(t) is then derived to be δ(t) - βe^-αtsin(βt)u(t).

Uploaded by

Khalil Ullah
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
90 views

Homework 10

1) The Fourier transform of a signal x(t) is calculated from its derivative. Modulating a tenor voice signal and applying a filter is proposed to overlap the frequency ranges of tenor and alto voices. 2) A filter is suggested to select frequencies between 140Hz and 220Hz from a modulated tenor voice. However, the overlapping frequencies from modulation will distort the output. 3) The response of an LTI system to e^-αtu(t) is δ(t). The response to e^-αtcos(βt)u(t) is then derived to be δ(t) - βe^-αtsin(βt)u(t).

Uploaded by

Khalil Ullah
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

ELEC201 Signals and Systems Spring 2020

HOMEWORK 10

1. Find the Fourier transform of x(t) in the following figure.

Solution: Although F{x(t)} does not seem to be easy, we can find the derivative of x(t).
dx
x0 = = u(t − 1) + u(t − 2) − u(t − 4) − u(t − 5)
dt

dx
Time shifting dt by 3 units to the left gives us

x00 = u(t + 2) + u(t + 1) − u(t − 1) − u(t − 2) = [u(t + 2) − u(t − 2)] + [u(t + 1) − u(t − 1)] = x1 + x2
where x1 = u(t + 2) − u(t − 2) and x2 = u(t + 1) − u(t − 1). We know the Fourier Transform of these functions, and
starting from them, we can calculate F{x0 }, F{x00 } and finally F{x}.
2 sin(2ω)
F{x1 } = F{u(t + 2) − u(t − 2)} =
ω
2 sin(ω)
F{x2 } = F{u(t + 1) − u(t − 1)} =
ω

Using linearity:
2 sin(2ω) 2 sin(ω)
F{x00 } = F{x1 + x2 } = +
ω ω

Using the time shift property:


 2 sin(2ω) 2 sin(ω) 
F{x0 } = e−3jω +
ω ω
Finally, using integral property:
e−3jω  2 sin(2ω) 2 sin(ω) 
F{x} = + + 6πδ(ω)
jω ω ω
sin(αω)
Note that F{x0 }(0) = 6, as αω is 1 for ω = 0.

2. The average voice frequency of a tenor singer lies between 120Hz and 200Hz, while average voice frequency of an alto
singer is between 140Hz and 220Hz. An engineer decides that by modulating the tenor voices with p(t) = π1 cos(2π20t)
and then applying a suitable filter, tenor and alto voice would overlap the same interval.

• Find P (ω) = F{p(t)}


• Draw the spectrum of the modulated tenor voice.
• Suggest a suitable filter.
• Would this design work? Explain.

Solution:

• P (ω) = F{p(t)} = δ(ω + 40π) + δ(ω − 40π)


• Tenor voice before modulation

Components of tenor voice after modulation


So in total, after modulation, we obtain

• We want the output in 2π140 to 2π220 range. So the filter should be


• The output of the filter is

Due to the overlapping of the components of the tenor voice, the output would produce a distorted version of
the input.

3. The response of an LTI system to x1 (t) = e−αt u(t), where α > 0, is δ(t). Find the response of the system to
x2 (t) = e−αt cos(βt)u(t).

Solution: Let H(ω) be the Fourier transform of h(t).

1
H(ω)F[e−αt u(t)] = F[δ(t)] = 1 ⇒ H(ω) = 1 ⇒ H(ω) = α + jω
α + jω

The response of the system to x2 (t) is


 
−1 −αt
y(t) = F H(ω)F[e cos(βt)u(t)] .

1 1 1 1
X2 (ω) = F[e−αt cos(βt)u(t)] = +
2 α + j(ω − β) 2 α + j(ω + β)
 
α + jω 1 1
Y (ω) = +
2 α + j(ω − β) α + j(ω + β)

F[δ 0 (t)] = jω

y(t) = αe−αt cos(βt)u(t) + (e−αt cos(βt)u(t)) ∗ δ 0 (t) = e−αt cos(βt)δ(t) − βe−αt sin(βt)u(t) = δ(t) − βe−αt sin(βt)u(t)
4. Consider x(t) whose Fourier transform X(w) is as depicted below, and p(t) is impulse train with period T= 10−3 ,
∞
X
i.e., p(t) = δ(t − kT ).
k=−∞

X(ω)
1

−500π 500π ω

(a) Determine Fourier transform of s(t) = x(t)p(t) in terms of X(w) and sketch its spectrum.
Solution:
X∞
p(t) = an ejnw0 t
n=−∞
2π
Fundamental frequency of p(t) is w0 = = 2000π.
10−3
Fourier series expansion of p(t) is
Z T /2 Z T /2
1 2π 1 2π 1
an = p(t)e−jn T t dt = δ(t)e−jn T t dt =
T −T /2 T −T /2 T
∞
2π X
Therefore, Fourier transform of p(t) is P (w) = δ(w − 2000πk).
T
k=−∞
∞
F 1 1 X
s(t) = x(t)p(t) −→ S(w) = X(w) ∗ P (w) = X(w − 2000πk)
2π T
k=−∞

(b) Consider an LTI system whose frequency response is given below. Express the output in terms of x(t) when the
input to the system is y(t).

H(ω)

−3000π −1000π 1000π 3000π ω

Solution:
Let z(t) be the output and Z(w) be its Fourier transform.
1 1
Z(w) = Y (w)H(w) = (X(w − 2000π) + X(w + 2000π)) = X(w) ∗ (πδ(w − 2000π) + πδ(w + 2000π))
T πT
1 2
z(t) = 2π x(t) cos(2000πt) = x(t) cos(2000πt)
πT T

You might also like