EE3tp4 Exam 2014

ELECTRICAL ENGINEERING 3TP4
Signals and Systems

DAY CLASS
DURATION OF EXAMINATION: 3 hours
McMaster University Final Examination
Profs. T. Todd and T. Kirubarajan

December 2013
This examination paper includes 6 pages and 7 questions. You are responsible for ensuring that your copy
of the paper is complete. Bring any discrepancy to the attention of your invigilator.
SPECIAL INSTRUCTIONS:
Answer all questions.

One 8.5" 11" formula sheet permitted (two sides).
Use of Casio FX-991 calculator only is allowed.
In all questions be sure to explain any assumptions you make.
Answer all questions on the question sheets only.
Print your name and ID number on all question sheets.
Name:
(10)
ID#:
1. Consider the following two signals, q(t) and p(t).
q(t)
p(t)
2 e (t2)
1
t
t
2
(a) Write out the integral that you would solve to find the convolution, y(t) = q(t) p(t), using the
graphical method, when 1 t 3. Simplify your expression as much as possible, i.e., your
final result must not contain any unit step functions. Do not evaluate the integral.
For 1 < t < 3. y(t) = 2t+1 2e( 2) d or y(t) =
integral limit and 1 mark for correct integrand.
R
R t2
1
2e(t 2) d . 2 marks for each correct
(b) If y(t) is the zero-state output of a causal linear time invariant system, which of q(t) or p(t) can
be the input to the system? Explain your answer.
The input has to be p(t) since p(t) is a non-causal signal. The impulse response must be q(t).
5 marks.
(15)
2. The following periodic signal f (t)
f(t)
1
1
t
1
0.5
has complex exponential Fourier series coefficients given by

ck =
sin(k/2)
sinc(k/2)
=
for k = 0, 1, 3, . . .
k
2
(a) Now consider the following periodic signal, d(t).
d(t)
2
1
t
0.5
0.5
Note the relationship between f (t) and d(t). Use this observation to obtain the complex exponential Fourier series coefficients, dk , of d(t).
P
1/2 + 2n) ek . Thus ek = ck ejk/2 /2. Therefore

(k/2)
dk = ek /jk (for k =
6 0), i.e., dk = sinc
ejk/2 /2jk. d0 = 1/4.
j2k
Note that d (t) = f (t)
n= (t
(b) Find a value for the average power in the signal d(t) accurate to two decimal places.
Obtained as an integral in time.
1 R 1/2
2 1/2 (t
+ 1/2)2 dt = 1/6.
(c) Assume that f (t) is input to a linear time-invariant (LTI) system whose impulse response has
the following Fourier transform.
.
H() =
j +
Find the phase shift (in degrees) that the fundamental frequency component in f (t) will experience.
1
. The phase shift is therefore
The fundamental frequency is 0 = . Therefore H() = j+1
1
tan (1) = 45 degrees.
(10)
3. The Fourier transform of a rectangular pulse

p(t) =
1 |t| 1/2
0 otherwise
is given by
P () =
2 sin(/2)
= sinc( )
Now consider the signal x(t).
x(t)
1
2
t
1
1
(a) Use p(t) and P () along with properties of the Fourier transform to obtain the Fourier transform,
X(), of x(t).
x(t) = p(t 1/2) p(t + 1/2) p(t + 3/2). Using linearity and time-shifting this gives
X() = P ()(ej/2 ej/2 ej3/2 ). Instead, some may write x(t) = p(t 1/2) p2 (t + 1)
and use p (t) sinc( /2).
(b) X() can also be obtained by taking the derivative of x(t) and then applying the Fourier transform derivative theorem. Use this method to obtain X(). Note that the form of your answer
will be different than that in Part 3a.
x (t) = (t + 2) + 2(t) (t 1). Using the shifting theorm gives FT: ej2 + 2 ej .
j2
j
Therefore X() = 2e je .
(15)
4. Answer the following questions that deal with the discrete-time Fourier transform.
(a) Find the discrete-time Fourier transform, X(), of x[n] given below. Note that
for |r| < 1.
(
( 21 )n n 0,
x[n] =
0
otherwise.
X() =
jn
n= x[n]e
1
j
1+ e 2
n=0 r
1
1r
(b) Sketch a plot of |X()| for .

2
1.8
1.6
1.4
1.2
0.8
0.6
-4
-3
-2
-1
(c) Discuss the relationships between the Z transform, the discrete-time Fourier transform and the
discrete Fourier transform.
The DTFT can be found by evaluating the ZT on the unit circle in the complex Z plane. The
DFT will give exact samples of the DTFT provided that the signal is finite duration. If not, the
DTFT samples a smeared version of the DTFT.
(20)
5. Consider the following discrete-time signal.
f [n] =
1 0nq
2 q+1nk
0 otherwise.
where q and k are positive integers and k q + 1.

(a) Sketch a stem plot of f [n].
An example for q = 5 and k = 10.

2.5
1.5
0.5
0
-5
10
15
(b) Write f [n] as a function of time shifted unit step functions.

f [n] = u[n] u[n q 1] + 2u[n q 1] 2[n k 1]. Some of these terms can be combined
together, i.e., f [n] = u[n] + u[n q 1] 2[n k 1].
(c) Using your results from Part 5b and properties of the Z transform, find the Z transform, F (z),
of f [n].
Using time shift property.
z
z1 (1
z q1 + 2z q1 2z k1 ) =
z
z1 (1
+ z q1 2z k1 )
(d) Find the Z transform, X(z), of the following discrete-time signal, x[n].
n
x[n] = sin( )u[n 3]
2
As in the example in the notes, we can write sin(n/2)u[n3] = sin(n/2)u[n]+sin(/2)[n
1] + sin(2/2)[n 2] = sin(n/2)u[n] + [n 1]. Therefore X(z) = z 2z+1 z 1 .
(15)
6. Answer the following questions relating to discrete time signals and systems.
(a) Find a closed form expression for the convolution, h[n]f [n], using the graphical method, where
h[n] = 2n u[n] [n 1]
f [n] = 2n u[n]
h[n] f [n] = (n + 1)2n u[n] 2(n1) u[n 1]
(b) Find the first five non-zero values of the impulse response of the system defined by
y[n + 2] y[n + 1] = x[n + 2] + x[n + 1]
where x[n] and y[n] are the system input and output, respectively.
starting at n = 0 we have 1, 2, 2, 2, 2.
(c) Describe two methods that can be used to obtain the complete response of a discrete-time linear
time invariant system. Include a discussion of their usefulness in different situations.
1. recursion (process inputs in real time. no closed form analytic results needed. future inputs
unknown.) 2. Z transform (closed form solution but analytic forms are required). 3. Find ZIR
and use convolution for ZSR, etc.
(15)
7. Answer the following questions involving Matlab and the lab material.
(a) In Lab 3 you investigated aliasing in signal sampling. The inputs included frequency chirp
signals defined by
c(t) = cos(t2 + 2f1 t + )
where = 2000 and f1 = 100 Hz. What is the instantaneous frequency, f (t) in Hz, of c(t) at
time t?
Take derivative of the argument, i.e., 2t + 2f1 rad/sec. Or t + f1 Hz.
(b) In Part 7a, assume that c(t) is sampled at 16 KHz and reconstructed from the samples. Explain
what you would hear, relating it to theory.
This is half the Nyquist rate. So you hear the frequency rise to 8 KHz then drop back to 100
Hz over the 8 second window.
(c) When the Matlab fft function is called with a vector argument whose size is 1 n, the function
returns the n-point DFT of the input. The fft function can also be called with an optional
second argument, m. When this argument is included and m > n, the fft function will return
an m-point DFT.
Explain what the fft function does when m > n. Describe in detail why you would want to
have the fft function return more DFT points than you have input data points.
It is doing zero padding to give you samples of the DTFT at a higher resolution.
THE END

EE3tp4 Exam 2014

Uploaded by

Copyright:

Available Formats

EE3tp4 Exam 2014

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EE3tp4 Exam 2014

Uploaded by

Copyright:

Available Formats

ELECTRICAL ENGINEERING 3TP4

Signals and Systems

Profs. T. Todd and T. Kirubarajan

Answer all questions.

1. Consider the following two signals, q(t) and p(t).

2e(t 2) d . 2 marks for each correct

2. The following periodic signal f (t)

has complex exponential Fourier series coefficients given by

(a) Now consider the following periodic signal, d(t).

1/2 + 2n) ek . Thus ek = ck ejk/2 /2. Therefore

3. The Fourier transform of a rectangular pulse

Now consider the signal x(t).

(b) Sketch a plot of |X()| for .

5. Consider the following discrete-time signal.

where q and k are positive integers and k q + 1.

An example for q = 5 and k = 10.

(b) Write f [n] as a function of time shifted unit step functions.

You might also like