School of Electrical and

Communication Engineering

EE417: DIGITAL SIGNAL PROCESSING

LABORATORY – 3
Study of sampling theorem

Name: Immanuel Sisaho

ID #: 22300670
C/Code: BEEC/4
Lecturer: Prof. R. Kumar

https://dsp-iitkgp.vlabs.ac.in/exp/sampling-theorem/

Prepared by:
Raj Kumar
Introduction
In modern digital signal processing, the conversion of real-life analog signals into digital form
is a foundational step. This experiment focuses on the principle of sampling, which is the first
phase in the analog-to-digital conversion process. The main objective is to understand how a
continuous-time analog signal can be accurately represented as a discrete-time signal, such
that it can later be reconstructed without loss of information.

Through this experiment, students will gain practical experience in observing analog signals
using an oscilloscope and configuring the signal source by adjusting its amplitude, frequency,
and phase. More importantly, the experiment emphasizes the importance of selecting an
appropriate sampling frequency. By adjusting this parameter, students will explore the
conditions under which a sampled signal can be exactly reconstructed, in line with Shannon’s
Sampling Theorem.

This hands-on approach is intended to bridge theoretical concepts with real-world

applications, reinforcing the critical role of proper sampling in ensuring signal integrity during
analog-to-digital conversion.
Theory
Most real-life signals encountered in daily life are analog in nature. These signals are
continuous in time and amplitude. However, to be processed by digital systems such as
microprocessors or computers, analog signals must be converted into a digital format. The first
step in this transformation is converting the continuous-time analog signal 𝑥(𝑡) into a discrete-
time signal 𝑥(𝑛𝑇), a process known as digitization.

Digitization involves two main processes: sampling (digitization in time) and quantization
(digitization in amplitude). Sampling involves recording the signal’s amplitude at specific time
intervals (every T second). This generates a sequence of values 𝑥(𝑛𝑇) from the continuous
waveform 𝑥(𝑡) (Oppenheim et al., 1999).

The sampling frequency 𝑓𝑠 is defined as the inverse of the sampling interval 𝑇, i.e.,

1
𝑓𝑠 =
𝑇

To accurately reconstruct the original signal from its samples, the Shannon’s Sampling
Theorem must be satisfied. It states that:

𝑓𝑠 ≥ 2𝑓𝑀

Where 𝑓𝑀 is the maximum frequency component of the original signal. This minimum rate 2𝑓𝑀
is known as the Nyquist Rate (Shannon, 1949).

If the sampling frequency is lower than twice the maximum signal frequency, aliasing occurs.
In this case, higher frequency components "fold back" into the lower frequency range,
resulting in distortion and loss of information, as the original signal cannot be reconstructed
accurately.

This concept is illustrated using two cases:

1. Correct Sampling: A 1 Hz analog signal sampled at 4 Hz satisfies the Nyquist criterion,

allowing perfect reconstruction (as shown in Fig. 3).
2. Under sampling (Aliasing): A 5 Hz signal sampled at the same 4 Hz violates the Nyquist
criterion, leading to incorrect reconstruction where the 5 Hz signal appears as a 1 Hz
signal due to aliasing (as shown in Fig. 4).

The sample-and-hold circuit often performs this ideal sampling in practical systems, holding
each sampled value until the next is taken.

Once digital processing is complete, the discrete signal 𝑥(𝑛) is usually converted back into an
analog signal 𝑥(𝑡) through digital-to-analog conversion (DAC).
Aim/Objective
This experiment helps students:
• Observe real-world analog signals using an oscilloscope.
• Adjust signal parameters such as amplitude, frequency, and phase.
• Set the sampling frequency correctly to accurately reconstruct a signal from its samples.

The main goal is to explore the sampling theorem and understand how continuous-time analog
signals are converted into discrete-time samples.

Method
1. Click on the Experiment tab SIMULATOR will open the workspace
2. Read the instruction to understand how the simulation to be executed.
3. Observe the block diagram and power on the circuit.
4. Select the amplitude (A) of the input sine wave signal(x(t)).
5. Select the frequency of the signal (fm) for the input signal(x(t)).
6. Select the sampling frequency of the signal(fs).
7. Click on the "Channel 1" button to observe the input signal x(t) on graph.
8. Click on the "Channel 2" to button observe the sampled output signal y(n) on
graph.
9. Click on the "Dual" to observe the input signal and output signal on graph.
10. Change the values of A, fm, and fs to observe the variation in the input and output
signals.
11. Hover on the graph to observe the value of the x(t) and y(n) at that instant of time
T.
12. Save the graph if you are done with your experiment.
13. Note:
• Make sure always input signal amplitude (A) >0 v o
• Make sure always input signal frequency (fm and fs) >0 Hz
• To change the values just scroll by hovering on the knob.
• Make sure that you kept equal time/div in the "Channel 1" and
"Channel 2".
 Result

Figure 1. Graphs of signal sampled at 2fs ≥ fm.

Figure 2.. Graph of signal sampled at more than 2fs ≥ fm.

Figure 3. Graphs of signal sampled at 2fs ≤ fm.

Discussion
The experiment investigates the practical implications of the sampling theorem by observing the
effects of different sampling frequencies on the reconstruction of analog signals. According to
the sampling theorem, to accurately reconstruct a continuous-time signal from its samples, the
sampling frequency (𝑓𝑠) must be at least twice the maximum frequency (𝑓𝑚) present in the
signal, known as the Nyquist rate (Shannon, 1949).
Analysis of Images:
• Case 1: Sampling at or above the Nyquist Rate (at or above 2fs)
In the scenario where a 1 Hz analog signal is sampled at 4 Hz (which is four times the
signal frequency), the sampled points accurately represent the original waveform. The
discrete samples align with the peaks and troughs of the original sine wave, and when
reconstructed, the signal closely matches the original analog input. This demonstrates
that sampling above the Nyquist rate preserves all the information necessary for perfect
reconstruction, as predicted by the theorem (Oppenheim et al., 1999).
• Case 2: Sampling below the Nyquist Rate (Aliasing) (less than 2fs)
When a 5 Hz signal is sampled at 4 Hz, the sampling frequency is less than twice the
signal frequency, violating the Nyquist criterion. The resulting sampled signal does not
capture the true nature of the original waveform. Instead, the reconstructed signal
appears as a lower-frequency waveform (in this case, resembling a 1 Hz signal), a
phenomenon known as aliasing. This distortion occurs because higher frequency
components are misrepresented as lower frequencies, leading to irreversible loss of
information (Proakis & Manolakis, 2006).
These observations from the images reinforce the theoretical understanding that proper
sampling is critical for accurate digital representation and reconstruction of analog signals. The
experiment also highlights the importance of using an oscilloscope to visualize both the original
and sampled signals, making it easier to observe the effects of varying the sampling frequency
in real-time.
.

Conclusion
The experiment successfully demonstrates the validity and practical significance of the sampling
theorem in digital signal processing. When the sampling frequency meets or exceeds twice the
maximum frequency component of the input signal (the Nyquist rate), the original analog signal
can be reconstructed accurately from its samples. However, if the sampling frequency falls below
this threshold, aliasing occurs, resulting in distortion and loss of information that cannot be
recovered.
These findings underscore the necessity of choosing an appropriate sampling frequency in the
design of digital systems to ensure signal integrity. The use of visual tools, such as oscilloscopes
and simulation platforms, provides valuable insight into the consequences of under-sampling
and the importance of adhering to the sampling theorem in practical applications

Reference
• Oppenheim, A.V., Schafer, R.W. and Buck, J.R., 1999. Discrete-time signal processing. 2nd
ed. Upper Saddle River, NJ: Prentice Hall.
• Shannon, C.E., 1949. Communication in the presence of noise. Proceedings of the IRE,
37(1), pp.10-21.
• Proakis, J.G. and Manolakis, D.G., 2006. Digital signal processing: principles, algorithms,
and applications. 4th ed. Upper Saddle River, NJ: Pearson Prentice Hall.