You are on page 1of 2

EE 313 FALL 2013

LINEAR SYSTEMS AND SIGNALS COURSE INSTRUCTOR: DR. DAVID BROWN TEACHING ASSISTANT: MANDAR KULKARNI

CE#3 Solution

A general tip You can insert comments to your MATLAB code by typing your comments after a % symbol. MATLAB CODE: t = [-20:0.001:20]; %Array from -20 to 20 with increments of 0.001 k1=1; %constant k2 = 2; %constant x = sin(k1.*pi.*t)./(k1.*pi.*t); %sinc(kt) = sin(pi*kt)/(kt) y = sin(k2.*pi.*t)./(k2.*pi.*t); %another sinc with different width of main lobe x(20001) = 1; %x(20001) is NaN because of division by zero. But by L Hospitals %rule we know that sinc(0) = 1. Hence we need to redefine x(20001) %here. y(20001) = 1; z = 0.001.*conv(x,y); %In order to do convolution of continuous time signals in %MATLAB, we need to multiply by the sampling time (refer %CE2 for little more on this) Here are the plots of x(t), y(t) and z(t) for k1=1 and k2 =2
1

0.8

0.8

0.6

0.6
y(t) = sinc(2t)
-15 -10 -5 0 t 5 10 15 20

x(t) = sinc(t)

0.4

0.4

0.2

0.2

-0.2

-0.2

-0.4 -20

-0.4 -20

-15

-10

-5

0 t

10

15

20

0.5

0.4

0.3
z(t) = sinc(t)*sinc(2t)

0.2

0.1

-0.1

-0.2 -40

-30

-20

-10

0 t

10

20

30

40

You can view the widths of main lobes accurately by changing the x axis limits appropriately (go to Edit->Axis Properties) as shown

0.5

0.4

0.3

z(t) = sinc(t)*sinc(2t)

0.2

This shows that width of main lobe for z(t) is 2 when k1= 1 and k2 = 2. Using the same approach, you can find the widths of the main lobe for x(t) and y(t) to be 2 and 1 respectively. Here you can see the width of z(t) is equal to the maximum of the widths of x(t) and y(t), respectively.
Width = 2

0.1

-0.1

-0.2 -2

-1.5

-1

-0.5

0 t

0.5

1.5

You will find similar observations, when you plot your sinc functions for different values of k1 and k2 as given in the computer assignment. Claim: Width of z(t) = maximum of widths of x(t) and y(t) respectively. Reason: Fourier transform of a sinc function is a rect function. Convolution of two sincs is nothing but multiplication of two rect functions. We know that

( )

Thus, we see that width of rect in frequency domain is inversely proportional to the width of the sinc in time domain. Also, multiplication of two rects is another rect with width equal to the minimum of widths of the two multiplying rects. Thus, convolution of two sincs results is another sinc with the width of main lobe equal to maximum of widths of convolving sincs.

You might also like