Sunteți pe pagina 1din 9

Sulaimany Polytechnic University 2019-2020

Technical College of Engineering third stage


Communication Engineering Dept digital communication lab

Experiment no .1
Sampling of continuous-Time Signals

Student:
1-Kaso Baxtiar
2-zhiwar awat
3-shangeen baqi
4-rayan jamal

Date of submission: 10-11-2019

1
Theory
Basically sampling is process of converting continuous time analog signal to
discrete time signal so the input can be any continuous time signal and in result the
output will be series of samples with different amplitudes and uniform space
between them, and this is also called pulse amplitude modulation (PAM) and
practically this can be done by multiplying the continuous time signal with series
of pulse which generated by pulse generator ,and this process is one of process that
happens to the signal during digitalizing process

And to get proper sampling we shall ensue the following points :


1-The signal must be band-limited, meaning it does not have spectral component
above B Hz.
2. Sampling interval TS which is the time between the samples must be less than or
equal to 1/2B, Sampling frequency fs or sampling rate should be greater than or
equal to the twice the highest frequency component of message signal. fs≥2B.
And if this two condition applied to a signal the signal can be reconstructed
properly from sampled signal at the end
And y
Generally in the nature the signals exist in analog form, however these signal are
usually transmitted and used in digital form,and in order to transmit these signal
digitally and reconstruct it back accurately, there must be a method converts the
signal from analog to digital, this happens if some conditions that ensure the
reliability of the received signal are satisfied.
The first step in converting an analog signal to digital signal is called sampling;
sampling is representing a continuous time signal in discrete time domain with

2
uniform spaces between the sampled signals. This continuous time signal that can
be represented in its samples can be recovered back to its analog form.
There are two conditions that ensure a signal can be sampled and reconstructed
back:
1. The signal must be band-limited, meaning it does not have spectral component
above B Hz.
2. Sampling interval TS which is the time between the samples must be less than or
equal to 1/2B, Sampling frequency fs or sampling rate should be greater than or
equal to the twice the highest frequency component of message signal. fs≥2B.
According to the sampling theorem, Nyquist rate is the minimum sampling rate
fs=2B,and Nyquist interval is Ts=1/2B.
The second step and the third in converting an analog signal to digital are
Quantization and Encoding,however in this experiment we focus only on sampling
in time domain and frequency domain.
There are 3 types of sampling:
1-Ideal Sampling

3
2-Natural Sampling

3-Flat-top sampling

Ideal Sampling in time domain consist of multiplication of the message signal g(t)
with impulse train δ(t) ,the resulted signal is called sampled signal ՜g(t) in time
domain,The frequency domain sampling of this signal is the fourier transform of
the message signal G(f) repeating with respective shifting as the equation show:

1
G(f)¿ ∑ G(f −nfs)
Ts −∞

4
Procedure
Test1:sampling

Test2: reconstruction

5
Test3:spectrum of the sampled signal

6
Result
-Sampled signal in time domain;

The above graph shows analog signal, impulse train ,sampled signal respectively.

7
2- reconstructed signal:

The above graph shows the message signal and impulse train and the sampled
signal and filtered signal interpolated signal respectively.
3-Sampling in frequency domain:

8
Discussion
In this experiment, we sampled the analog signal in both the frequency and the
time domain.
In the first case, we sampled the sine wave in time domain through multiplying it
with impulse train δ(t) through product block ,and the resulted signal which we
called the sampled signal was displayed through scope. Since the sine wave had
frequency of 1KHz ,the period we used for the pulse generator was 1/5000 or less
than 1/2B in order to have lesser sampling period for more accurate sampled signal
and to prevent aliasing. The signal we get is the naturally sampled signal because
as we can see from the results ,the sampled signal retains the amplitude and the
shape of the message signal.
To reconstruct the signal back to its original analog form, we a Low-Pass filter
with cut-off frequency 2KHz since it’s not ideal in practical, called analog filter
design in Matlab to remove the spectrum of the sampled signal and to have only
the spectrum of the signal remaining.
We also pass the spectrum of the signal through amplifier with gain to increase the
amplitude to its original value.
We obtain the original sine wave back but with phase shift created due to the low-
pass filter.
In the second case, we sample the message in the frequency domain by using
spectrum analyzer, since the sampling equation in the frequency domain of the
natural sampling shows that the spectrum of the sampled signal consist of shifted
signals, we have to use a much greater bandwidth in order to show the shifted
signal.

՜G(f)=coG ( f )+ ∑ cn[G ( f −nfs ) +G ( f +nfs ) ]
1

S-ar putea să vă placă și