ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY

Republic of the Philippines

La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

I. Introduction

Frequency response analysis in Finite Element Analysis (FEA) is used to calculate the
steady-state response due to a sinusoidal load applied to a structure at a single frequency.
It is a specialized type of transient response analysis that is extremely efficient to solve a
very specific type of model[2].

Frequency response analysis is the technique whereby a sinusoidal test signal is used to
measure points on the frequency response of a transfer function or impedance function.
The basic set-up is shown in Figure 1 in which a sine wave u(t) is applied to a system with
transfer function G(s). After transients due to initial conditions have decayed away, the
output y(t) becomes a sine wave but with a different magnitude Y and relative phase Φ.
The magnitude and phase of the output y(t) are in fact related to the transfer function G(s)
at the frequency (ω rad/s) of the input sinusoid[1].

Figure 1: Linear Transfer Function With Sinewave Input

One very important special case of transfer functions occurs when the input voltage is
a sinusoid. Because a sinusoid is the sum of two complex exponentials, each having a
frequency equal to the negative of the other, we can directly predict the output voltage
by examining the transfer function.
Thus
𝑌
= |𝐺(𝑗𝜔)| = gain at 𝜔.......................................
𝑈
And
Φ = ∠G(jω) = phase at ω......................................

Figure 2: Sinewave input and steady-state sinewave response as displayed on an

oscilloscope. The corresponding vector notation is shown on the right.

1|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

An oscilloscope trace can be used to directly measure the relative phase and magnitude
of the input and output waveforms (Figure 2). However, this method is inherently
inaccurate and requires more sophisticated techniques to remove errors caused by
noise, non-linearity, and other factors.

The fact that frequency response analysis is frequency selective is its primary benefit. The
system response at that frequency can be measured with great precision when only one
part of the spectrum is extracted. When the system under consideration has resonant
characteristics, this has significant advantages. In particular, the various resonances can
each be excited and studied independently without affecting other oscillatory modes[2].

Figure 3a: System Frequency Response

By sequentially measuring the gain and phase at various frequencies, a picture of the
system frequency response can be built-up and plotted either as a diagram in the
complex plane shown in Figure 3a.

Figure 3b: Presentation Of Frequency Response Information

2|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

In the figure 3b it was shown here a joint graph of gain against frequency. The latter
are often termed Bode diagrams because of their use in Bode’s control system design
methods.

In addition to its frequency selective nature frequency response analysis has the advantage
that since waves have well-defined amplitude characteristics. In particular the maximum
amplitude of a sine wave is only √2 times greater than the root mean square value. The
ratio

is termed the crest factor (CF) and is an important factor is designing test signals for
systems analysis. Specifically, one usually wishes to inject as much power (as determined
by the mean square signal level) into a system as compatible with a linear system
response. System linearity is usually related to the peak test signal excursion which can
be tolerated before non-linear effects become evident. The “best” excitation in this respect
is the binary waveform with crest factor of infinity. With a crest factor of 1.41 the sine
wave is reasonably close to the theoretical best, while retaining the great merit of
frequency selective system

RELEVANT FORMULA

Resonant Frequency

It is the frequency at which the magnitude of the frequency response has peak value for
the first time. It is denoted by ωr. At ω=ωr, the first derivate of the magnitude of T(jω) is
zero.
Differentiate M with respect to u

Substitute, u=ur and dM/du = 0 in the above equation.

3|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

Substitute, ur=ωr/ωn in the above equation.

Resonant Peak
It is the peak (maximum) value of the magnitude of T(jω). It is denoted by Mr.
At u=ur, the Magnitude of T(jω) is -

Substitute, ur=√1−2δ2 and 1−u2r =2δ2 in the above equation.

Resonant peak in frequency response corresponds to the peak overshoot in the time
domain transient response for certain values of damping ratio δ. So, the resonant peak and
peak overshoot are correlated to each other.

Bandwidth
It is the range of frequencies over which, the magnitude of T(jω) drops to 70.7% from
its zero frequency value.
At ω=0, the value of u will be zero.
Substitute, u=0 in M.

Therefore, the magnitude of T(jω) is one at ω=0.

At 3-dB frequency, the magnitude of T(jω) will be 70.7% of magnitude
of T(jω) at ω=0
i.e., at ω=ωB,M=0.707(1)=1/√2

Let, ub2 = x

4|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

Consider only the positive value of x

Substitute,

Bandwidth ωb in the frequency response is inversely proportional to the rise time tr in the
time domain transient response.

II. Objectives
1. From the given transfer function/model determine the following: resonant frequency
(fo), resonant peak (fp) and the bandwidth (WB).
2. Plot the graph of the response of the given transfer function.
3. Simulate the given circuits using SCILAB, SCILAB XCOS and plot the response.
4. Design and implement a system that using a Frequency Response.

III. Equipment/Tools/Materials/Components Needed

a. PC/Laptop with SCILAB
IV. Procedure
1. From the given transfer function/model determine the following: resonant frequency
(fo), resonant peak (fp) and the bandwidth (WB).

5|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

Resonant Frequency :

Where 𝝎𝒏 = 𝟏𝟎 𝒌𝑯𝒛 converting into rad/s

= 62831.853108075 rad/sec.
𝜻 = 𝟎. 𝟒

𝒓𝒂𝒅
𝝎𝒓 = 𝝎𝒏 √𝟏 − 𝟐𝜻𝟐 = 𝟔𝟐, 𝟖𝟑𝟏. 𝟖𝟓𝟑𝟏𝟎𝟖𝟎𝟕𝟓√𝟏 − 𝟐(𝟎. 𝟒)𝟐 = 𝟓𝟏, 𝟖𝟏𝟐. 𝟒𝟕𝟑𝟒
𝒔
Resonant Frequency :
𝟏 𝟏
𝑴𝒓 = = = 𝟏. 𝟑𝟔𝟑𝟖𝟔𝟏𝟖𝟏𝟒
𝟐𝜻√𝟏 − 𝟐𝜻𝟐 𝟐(𝟎. 𝟒)√𝟏 − 𝟐(𝟎. 𝟒)𝟐
Bandwidth:

𝝎𝒃 = 𝝎𝒏 √𝟏 − 𝟐𝜻𝟐 √𝟐 − 𝟒𝜻𝟐 + 𝟒𝜻𝟒

= 𝟔𝟐, 𝟖𝟑𝟏. 𝟖𝟓𝟑𝟏𝟎𝟖𝟎𝟕𝟓√𝟏 − 𝟐(𝟎. 𝟒)𝟐 √𝟐 − 𝟒(𝟎. 𝟒)𝟐 + 𝟒(𝟎. 𝟒)𝟒

𝒓𝒂𝒅
= 𝟖𝟔, 𝟑𝟔𝟑. 𝟒𝟓𝟕𝟖𝟔
𝒔

2. Plot the graph of the response of the given transfer function when an input voltage is
unit step function u(t) =10 V and when the input voltage is sinusoid v(t)=10 Sin
3770t.

We may remember from linear systems course that, for a continuous-time transfer
function described in terms of Laplace variable s, frequency response can be
achieved by letting s = jw.

By this relationship, a frequency response of a transfer function can be plotted “the

hard way.” Using
62831.853108^2
𝑃(𝑠) = 2
𝑠 + (2 x 0.4 x 62831.853108075)𝑠 + 62831.853108^2

as an example, we solve for P(jw) manually by substituting s = jw, which gives

62831.853108^2
𝑃(𝑗𝜔) =
𝜔2 + (2 x 0.4 x 62831.853108075)𝑗𝜔 + 62831.853108^2

Now we can plot the magnitude and phase versus frequency by using the following
set of commands

6|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

SCILAB COMMAND:

s = %s;
num = (62831.853108^2);
den = ((s^2)+((2*0.4*62831.853108075)*s)+(62831.853108075^2));
sys = syslin('c', num/den);
f_min = 0.0001; f_max = 1000000
clf()
bode(sys,f_min, f_max, "rad")
title('Figure 4: Bode Plot of the Given Transfer Function')

SCILAB RESULT:

The frequency range that a plant is responsive is called bandwidth. When trying to
identify bandwidth from a Bode plot, we can roughly indicate the frequency point where
the magnitude curve touches the 0 dB line. Furthermore, the magnitude decreases as
the frequency drops toward zero. The magnitude starts to rise at a specific frequency,
reaches a maximum value, and then begins to fall as the frequency rises higher. At
higher frequencies, the phase starts to gradually decline until it asymptotically reaches
-180 degrees, whereas at lower frequencies, it is almost zero degrees.

7|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

SCILAB SCRIPT:
s = %s;
wn = 62831.853108 ;
zeta = 0.4
tf = syslin("c", wn^2 / (s^2+ 2*zeta*wn*s+wn^2));
t = 0:0.01:5;
plot2d(t, csim('step', t, tf));
xlabel("Time[s]");
ylabel("y1")
title("Step Response of the Second Order System");
xgrid(1, 1, 10);

Figure 5: Step Response of the Given Transfer Function Above

Looking at the figure above we could see that the step response of the given transfer
function was a quite not pretty as it has no oscillation as it has no overshoot.

Resonant Peak in Bode Plot

Figure 6: Locating the Resonant Peak

8|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

The highest magnitude is 2.695 dB. Its value, 1.364 rad, is obtained by converting it to
radians., as can be seen from the bode plot.

Resonant Frequency in Bode Plot

Figure 6: Locating the Resonant Frequency

The resonance frequency was around 51,640 rad/s, as we could observe.Wherein,

according to the graph, the maximum magnitude is 2.695 dB, which is found at a
frequency of 5.1640E+04 rad/s.

Bandwidth in Bode Plot

Figure 7: Locating the Bandwidth

As the magnitude declines at -3db may be calculated by following the curve from the
bode plot: 8.632E+04 rad/s. As a result, the bandwidth is around 86,320 rad/s.

XCOS PLOTTING:

Figure 8: XCOS Block Diagram with the Given Transfer Function

9|Page
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

The system block diagram in the previous illustration used the controller and unit step
as input signals.

Figure 9: XCOS Result of the Simulated Block Diagram

The system's reaction with the controller at 𝜻 = 0 is seen in the above figure. 4 and
Unit Step as Input Signal with 𝝎𝒏 = 2rad/s the multiplexer that mixes the signal, the
oscilloscope, the simplified controller with a damping factor of 0.4 and a natural
frequency of 2rad/s, the unit step input signal with an input voltage of 10V, and the
summing block for unity gain feedback. Consequently, the system with a controller
generates an output that has a maximum overrun at t around 3 seconds before
settling at a magnitude of 10 seconds at time t, which is around 8 seconds.

Figure 10: Changing the Input Signal of the Previous Block Diagram

Block Diagram of a System Using a Sinusoid and a Controller as Input Signal using
a multiplexer that mixes the signal, an oscilloscope, a simplified controller with a
damping factor of 0.4 and a natural frequency of 2rad/s, a sinusoid input signal with
the function v(t)=10sin(2t), and a summing block for unity gain feedback.

10 | P a g e
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

Figure 11: Response of the System with Controller

We could see above that if the input is a sinusoidal signal, the output is always a
sinusoidal signal of the same frequency. But the amplitude of the output sinusoidal can
be different from the input amplitude. And the ratio of the output amplitude to input
amplitude is a constant for that particular frequency - whatever the input amplitude
and phase are. Similarly, the difference between the input and output sinusoidal phases
is also constant for the particular frequency - irrespective of the amplitude.

V. Observations
The values of the resonant peak, resonant frequency, and bandwidth when problems are
solved manually rely on the natural frequency and damping factor. The bandwidth is
equivalent to 86,363.458 rad/s, the resonant frequency is equal to 51,812.4734 rad/s, and
the resonant peak is equal to 1.364 rad. Additionally, the resonance's frequency, peak, and
bandwidth figured out by merely producing the bode plot and identifying the precise
places on the graph. The resonant peak, calculated using the transfer function provided,
is 2.695 dB or 1.363797845 rad. The bandwidth is 86,320 rad/s, while the resonant
frequency is 51,640 rad/s. With just a little amount of discrepancy, the calculated values
and data gathered from the bode plot are almost comparable. Therefore, it can be deduced
that we may use the formula to solve for the values of the resonant frequency, resonant
peak, and bandwidth or display the bode plot and just identify the values using Scilab.
The system's response overshoots and briefly oscillates when presented with an input
signal with a unit step and a step function of u(t)=10V before settling and achieving a
constant response. When a sinusoidal input signal is present, the magnitude first generates
an output that is equal to the sinusoidal input signal, but it then rises and eventually levels
out at a magnitude that is bigger than the input signal.

11 | P a g e
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

VI. Conclusions
In summary, computing a frequency response of a system is important for control analysis
and design. SCILAB provide different approaches we can use to get the frequency
response of our system. All in all we have completed all the stated objected above.

VII. Recommendations
I can recommend is to apply what we have learned here about frequency response analysis
into an actual application in order to digest and fully understand what was have discussed
above and to visualize accurately the response.When creating a parallel RLC circuit, Since
a parallel RLC circuit provides a second ordcr differential equation, solving for the total
response of either the inductor current or the capacitor voltage will provide a natural (or
transient) response, and if applicable, a forced or steady-state response the inductance is
conversely as opposed to resistance, which increases proportionally with bandwidth. Both
resonant frequency and the bandwidth are also related to the values of R, L, and C.

VIII. Design Applications

This RLC example is used to compare all the LTI representations. The design refers to a
RLC low passive filter, where the input is represented by the voltage drop "V_in" while
the output "V_out" is voltage across the resistor.

The relation between the input and the output of the system is:

Let the following values of the constants:

• A = 10V ; F = 1x104 Hz

12 | P a g e
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

• R = 50 Ohm
• L = 0.000579 H
• C = 6.503186561415094x10-7 F
with initial conditions: Vout (t) = V’out(t) = 0

Figure 12: Bode Plot and Sinusiodal Result of the Given Values Above

As shown at the above figure was the result of the given scheme diagram above with an
amplitude of 10V, a frequency of 1000 Hz, resistance equals to 50 ohms, inductance
equals to 0.000579 H and 6.503186561415094x10-7F.

13 | P a g e
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

RLC low passive filter data

SCILAB command:
// RLC low passive filter data
mR = 50; // Resistor [Ohm]
mL = 0.000579; // Inductor [H]
mC = 6.503186561415094e-7; // Capacitor [F]
mRC = mR*mC;
mf = 1e+4;
// Define system matrix
A = [0 -1/mL; 1/mC -1/mRC];
B = [1/mL; 0];
C = [0 1];
D = [0];
// State space
sl = syslin('c',A,B,C,D)
h = ss2tf(sl)
sl1 = tf2ss(h)
// Transformation
T = [1 0; 1 1];
sl2 = ss2ss(sl,T)
// Canonical form
[Ac,Bc,U,ind]=canon(A,B)
// zero-poles
[hm]=trfmod(h)

Figure 13: Output of the command trfmod

In Scilab it is possible to move from the state-space representation to the transfer

function using the command ss2tf. The vice versa is possible using the command tf2ss.
In the reported code (right), we use the "tf2ss" function to go back to the previous state
but we do not find the original state-space representation. This is due to the fact that
the state-space representation is not unique and depends on the adopted change of

14 | P a g e
 ILOILO SCIENCE AND TECHNOLOGY UNIVERSITY
Republic of the Philippines
La Paz, Iloilo City

COLLEGE OF ENGINEERING AND ARCHITECTURE

variables. The zeros and the poles of the transfer function can be display using the
command trfmod.

Figure 14: Step Respond and Sinusiodal Response of the Given Values

The signal is shown as a function of time shown on the x-axis, while the current in
amperes is shown on the y-axis.

IX. References
1. Frequency Response Analysis - What Is it? - SIMTEQ Engineering
2. Frequency Response Analysis by N.D. Cogger BSc, PhD, MIOA R.V.Webb BTech,
PhD, CEng, MIEE
3. https://www.tutorialspoint.com/control_systems/control_systems_frequency_response_a
nalysis.htm

15 | P a g e