Professional Documents
Culture Documents
x(t )e
T 4
j t
dt
.
T 1 + e 2 j t 1 T e j e j dt = ( ) ( + )= T 2 2 2 2 j 4
4
2) Solution:
F ( ) = U ( x) Ae
ax
j x
dx = Ae
0
( a + j ) x
e ( a + j ) x A dx = A = ( a + j ) 0 ( a + j ) .
3) Solution:
The inverse Discrete Fourier Transform is given by
x ( n) = 1 2
X ( )e
jn
d
.
1 x ( n) = 2
1 j n e d = 2
3
e j n jn
3 2 e = 2 (
3
jn 3
e 2 jn
jn 3
)=
sin(
n ) 3
.
4) Solution:
The Laplace transform of the function f (t ) = sin(t ) is given by F ( s ) = Therefore using the identity L[tf (t )] =
L[t sin(t )] =
1 . s +1
2
d F ( s ) , we have ds
d 1 d 2s F (s) = 2 = 2 ds s + 1 ( s + 1) 2 . ds
MEEN 651 Summer 2004 So the Laplace transform of the function f (t ) = 1 + t sin(t ) , is
F (s) =
1 2s + 2 s ( s + 1) 2 .
5) Solution:
The Laplace transform of a function f (t ) is given by F ( s ) = f (t )e st dt
0 .
Therefore substituting the value of the function in the above relation we get F ( s ) = te
0
at
e dt = te =( a + s ) t dt
st
te ( a + s )t e ( a + s ) t 1 1 1 ( a + s )t e dt F (s) = 0 + = + = (a + s ) ((a + s )) 0 ( s + a) 2 (a + s ) 0 (a + s) 0
6) Solution:
a. The frequencies present in the signal are 2F1 = 500 F1 = 250 Hz , 2F2 = 1000 F2 = 500 Hz , 2F3 = 2000 F3 = 1000 Hz. Therefore Fmax=1000 Hz and according to the sampling theorem the minimum sampling frequency is given by
Fs > 2 Fmax
b. If the sampling frequency is 1000Hz, then according to the sampling theorem the maximum frequency that can be recovered from the discrete signal is given by Fs/2=500 Hz
7) Solution:
a. (1) Signal sampled at 200Hz for 1 second.
It can be seen that there are three frequency components at 10 Hz, 70 Hz and 80 Hz respectively. Note that the magnitude of the 70 Hz component is 3 times that of the 10 Hz and 80 Hz components.
It can be seen that as the number of data points increase, the side lobes in the magnitude plot decrease considerably.
b. For case (3) of part(a), signal downsampled by a factor of 2. The plot is shown in below
Since the sampling frequency is 100 Hz, only frequencies upto 50 Hz can be seen without any aliasing. The remaining frequencies appear as aliased frequencies. Therefore the 10 Hz component appears without any aliasing, whereas the 70 Hz and 80 Hz components appear as 20 Hz and 30 Hz components, i.e.,
c. Therefore a low pass filter is needed before downsampling. The resampled result is shown in below
Mablab Codes: For part (a) (1) time = 1; t = 0:1/200:time; x = sin(20*pi*t)+3*sin(140*pi*t)+ cos(160*pi*t); f = fft(x, 8192); freq = -100:200/8192:100-200/8192; plot(freq, abs(fftshift(f))); grid xlabel('Frequency (Hz)'); ylabel('Magnitude'); title('Original signal sampled at 200Hz for 1 second'); (2) time = 10;
MEEN 651 Summer 2004 t = 0:1/200:time; x = sin(20*pi*t)+3*sin(140*pi*t)+ cos(160*pi*t); f = fft(x, 8192); freq = -100:200/8192:100-200/8192; plot(freq, abs(fftshift(f))); grid xlabel('Frequency (Hz)'); ylabel('Magnitude'); title('Original signal sampled at 200Hz for 10 second'); (3) time = 100; t = 0:1/200:time; x = sin(20*pi*t)+3*sin(140*pi*t)+ cos(160*pi*t); f = fft(x, 8192); freq = -100:200/8192:100-200/8192; plot(freq, abs(fftshift(f))); grid xlabel('Frequency (Hz)'); ylabel('Magnitude'); title('Original signal sampled at 200Hz for 10 second');
For Part (b) and Part(c): time = 100; t = 0:1/200:time; x = sin(20*pi*t)+3*sin(140*pi*t)+ cos(160*pi*t); x1 = dyaddown(x); x2 = resample(x,1,2); f1 = fft(x1, 8192); f2 = fft(x2, 8192); freq1 = -50:100/8192:50-100/8192; figure(1) plot(freq1, abs(fftshift(f1)))
MEEN 651 Summer 2004 grid xlabel('Frequency (Hz)'); ylabel('Magnitude'); title('Signal downsampled by a factor of 2'); figure(2) plot(freq1, abs(fftshift(f2))) grid xlabel('Frequency (Hz)'); ylabel('Magnitude'); title('Signal resampled by a factor of 2');