School of Engineering

Department of Electrical and Computer Engineering

332:224 Principles of Electrical Engineering II Laboratory

Experiment 3

Steady State Frequency Response Using Bode Plots

1 Introduction

Objectives • To study the steady state frequency response of stable

transfer functions of certain simple circuits using Bode
plots

Overview

This experiment treats the subject of frequency response by the use of Bode plots. The basic
equation and logic of Bode plots are introduced in section 2 for transfer functions with real
and complex-conjugate poles and zeros. The asymptotes for the magnitude and phase plots
are introduced, as is the method for correcting around the corner frequency to obtain realistic
plots.
Four circuits with transfer characteristics that exhibit illustrative examples of the pole and
zero combinations covered in section 2 are realized in section 4. Their theoretical Bode plots
are derived in section 3 and data for plotting their actual ones are extracted in section 4 by
varying the frequency and measuring output/input (magnitude of gain) and phase.

© Rutgers University
Authored by P. Sannuti
Latest revision: December 16, 2005, 2004
by P. Panayotatos
 PEEII-V-2/16

2 Theory

2.1 Introduction to Bode plots1

Bode plots are commonly used to display the steady state frequency response of a stable
system. Let the transfer function of a stable system be H(s). Also, let M(!) and "(!) be
respectively the magnitude and the phase angle of H(j!). In Bode plots, the magnitude
characteristic M(!) and the phase angle characteristic "(!) of the frequency response are
plotted vs. log10(!). Also, in the magnitude plot, instead of plotting M(!), vs. ! one plots 20
log10M(!) which is called the decibel2 (abbreviated as dB) value of M(!). By utilizing the
logarithm, one accommodates a wide range of data on a compact graph. Moreover, as will be
seen shortly from equations (3) and (4), the logarithmic operation permits the decomposition
of the transfer function into simpler parts in a natural way. By analyzing each of the simpler
parts individually, and then combining them appropriately, one can plot relatively easily the
overall frequency response of a composite transfer function. Note here that the DC frequency
!=0 cannot be represented in the logarithmic scale since log10(0) =-#.

2.2 Transfer function standard form

Consider a rational transfer function H(s) in which all the poles and zeros are shown in a
factored form as
Ks r ( s + z1 ) ( s + z2 ) .....( s + zm )
H (s) =
( s + p1 )( s + p2 ).....( s + pn ) (1)

where m, n, and r are some integers. It is convenient to rewrite H(j!) in the form known as a
standard form for Bode plots,

K ( j! ) (1 + j! / z1 ) (1 + j! / z2 ) .....(1 + j! / zm )
r

H ( j! ) = o
(1 + j! / p1 )(1 + j! / p2 ).....(1 + j! / pn ) (2A)

where
Kz1z2 ....zm
Ko = (2B)
p1 p2 ....pn

If any of the poles or zeros are complex, each of the complex conjugate pairs of such poles or zeros
of H(s) can be combined to get a quadratic form of the type s2 + (2$/! 0)s + ! 02, and then, by
factoring out !02 and letting s =j!, it can be reduced to a standard form as

1
The subject is treated in detail in Appendix E of the text.
2
The decibel is treated in detail in Appendix D of the text.
 PEEII-V-3/16

1 + (2$/!0)j! - !2/ !02

2.3 Real Poles and Zeros

First assume that all the poles and zeros are real. Then, one can easily determine the magnitude
and phase functions 20 log10M(!) and "(!) as

20 log10M(!)= 20 log10(|K0|)+ 20 r log10(j!)

+ 20 log10(|1+j!/z1|) + 20 log10(|1+j!/z2|) + ......
- 20 log10(|1+j!/p1|) - 20 log10(|1+j!/p2|)-.... (3)

"(!) = %K0 + r 90o +tan-1(!/z1) + tan-1(!/z2) + ......

- tan-1(!/p1) - tan-1(!/p2) - ..... (4)

The phase of K0 is zero if K0>0 and 180o if K0<0.

The above equations reveal that the magnitude and phase functions 20log10M(!) and "(!) are
obtained by simply adding the contributions due to several individual factors. In what follows,
we examine individually each of such factors. Once this is done, Bode plots of composite
transfer functions can easily be determined. For clarity, we denote 20 log10M(!) by MdB(!).

j!
= 20 log10 1 + (! / p1 )
2
Note that 20 log10 1 +
p1

j!
= 20 log10 1 + (! / p1 ) " 1
2
If ! << p1 then !/p1<<1 and 20 log10 1 +
p1

If ! >> p1 then !/p1>>1 and

j! #!&
= 20 log10 1 + (! / p1 ) " 20 log10 % (
2
20 log10 1 +
p1 $ p1 '

The latter represents a line with slope of 20dB per decade or 6db per octave3 since:

3
A decade is a ten-fold increase in frequency, while an octave is a two-fold increase in frequency.
 PEEII-V-4/16

"!%
If !=10 p1 then 20 log10 $ ' = 20 log10 10 = 20dB
# p1 &

"!%
If !=2 p1 then 20 log10 $ ' = 20 log10 2 ( 6dB
# p1 &
Exactly the same is true for zeros. The only difference is that the contribution of poles is
subtracted (negative slope –20dB/dec) and that of zeros is added (+20dB/dec).

These asymptotes of MdB(!) meet at != p1. Thus, the frequency p1 is called the corner
frequency. At this corner frequency, the actual value of MdB(!) can be computed easily as 3
dB rather than the 0 dB predicted by either asymptote. At half of the corner frequency, i.e., at
! = 0.5p1, one can compute MdB(!) as 1 dB rather than the 0 dB given by the low frequency
asymptote. On the other hand, at twice the corner frequency, i.e., at ! = 2 p1, one can
compute MdB(!) as 7 dB rather than the 6 dB given by the high frequency asymptote. Using
these actual values as a guide, an exact plot of MdB(!) can easily be drawn4.

In terms of the angle, one can choose three cases;

"!%
If ! << p1 then !/p1<<1 and tan -1 $ ' ( 0
# p1 &
"!%
If ! = p1 then !/p1=1 and tan -1 $ ' = tan -11 = 45 o
# p1 &
"!%
If ! >> p1 then !/p1>>1 and tan -1 $ ' ( tan -1) = 90 o
# p1 &
This means that the maximum contribution of any one zero or pole is 90º (+ for zeros, - for
poles). If one approximated “much smaller than p1” by p1/10 5and as “much larger than p1”
by 10 p1,6 then the angle plot has a slope of -45º/dec7 and is flat elsewhere.

4
Figure E-6 in the text shows both the low and high frequency asymptotes of MdB(!) as well as the actual characteristic.
5
tan-1(0.1) = 5.3o≈ 0
6
tan-1(10)= 84.7o≈ 90º
7
+45º/dec for a zero
 PEEII-V-5/16

2.3.1 Constant gain K0:

If H(s)=K0, then MdB(!) = 20 log10|K0|

which is a constant with respect to log10 !. If K0 is positive, then the phase angle of
H(j!) is zero for all !; if K0 is negative, then the phase angle of H(j!) is 180o for all !.

2.3.2 Pole or zero at the origin:

For an ideal differentiator, the transfer function is s, while for an ideal integrator, the
transfer function is 1/s. Thus, the ideal differentiator has a zero at s=0, and an ideal
integrator has a pole at s=0.
For an ideal differentiator, H(j!)= j!, and MdB(!) = 20 log10 !.
Thus, the plot of MdB(!) as a function of log10 ! is a straight line with a slope of 20 dB
per decade or approximately 6 dB per octave. The straight line passes through 0 dB at
!=1 since 20 log10 ! equals 0 at !=1. The phase angle of H(j!)= j! is 90o for all !.

For an ideal integrator, the transfer function is 1/s which is the reciprocal of that of the
differentiator. The Bode plots of 1/s are the negative replicas of those of s. That is, the
magnitude plot is a straight line with a slope of -20 dB per decade and passes through 0
dB at !=1, while the phase angle is -90o for all !.

2.3.3 Simple Real Pole or Zero:

Then the discussion of section 2.3 above is relevant and the asymptotes are at zero db
and ±20dB/decade for zero and pole respectively.

2.3.4 Complex Conjugate Pairs of Poles or Zeros:

We consider next a transfer function having a pair of complex conjugate poles which
together give rise to a quadratic factor, i.e.,

)1
# 2! s2 &
H (s) = %1 + s+ 2(
$ " o "o '

where $ and !0 are respectively called as the damping coefficient and the natural
frequency of the two complex conjugate poles. For this case,
 PEEII-V-6/16

!2 j!
M dB (! ) = "20 log10 1 " 2 + 2#
!o !o

% " (
' 2$
"o *
! (" ) = # tan #1 ' 2 *
and '1# " *.
'& " o2 *)

Two asymptotes of MdB(!), one for low frequencies and another for high frequencies,
can easily be ascertained. For !/! 0 << 1, we have

MdB(!) ≈ -20 log10 1 = 0.

Similarly, for !/! 0 >> 1, we have MdB(!) ≈ -40 log10 (!/! 0).
That is, for high enough frequencies such that ! >> ! 0, the magnitude characteristic
MdB(!) is a straight line with a slope of -40 dB per decade and intersects the frequency
axis at != ! 0.. Both the low and high frequency asymptotes of MdB(!) meet at the
corner frequency ! = ! 0. The exact magnitude characteristic of MdB(!) depends on the
damping coefficient $. As shown in Figure 1, for small values of $, there is significant
peeking in the neighborhood of the corner (or natural) frequency ! 0. For ! < 1 / 2 and
for certain representative frequencies, one can compute exactly MdB(!) as outlined
below:
1. At the corner frequency !0, MdB(!) has a value of -20 log10[2 $].
2. At ! = 0.5 ! 0, MdB(!) has a value of -10 log10[$2+0.5625].
3. It can be shown that the peak of MdB(!) occurs at ! = ! o 1 " 2# 2 and that the peak
value is -10 log10[4 $2(1-$2)].
4. MdB(!) crosses the 0 dB axis at the frequency ! = ! o 2 1 " 2# 2 ( )
For ! > 1 / 2 the exact characteristic of MdB(!) lies entirely below its asymptotic
approximation.

For the angle, as shown in Figure 1, the exact plot of "(!) depends on the damping
factor $ . Without taking into account the value of $ , one can coarsely8 approximate

8
There exist in the literature certain better mid frequency approximations which take into account the value of $ .
 PEEII-V-7/16

the phase angle plot by three asymptotes, similar to the ones derived for the simple real
pole, only now the total contribution of the pair will be 180º rather than 90º of the
simple pole or zero:
Low frequency approximation: For ! < 0.1 ! 0, "(!) can be approximated by 0.
High frequency approximation: For ! > 10 ! 0, "(!) can be approximated by -180o .
Mid frequency approximation: For 0.1 ! 0 < ! <10 ! 0, "(!) can be approximated by a
straight line having a slope of -90o per decade and passing through -90o at ! = ! 0.

Consider next a transfer function with a pair of complex conjugate zeros,

H(s)= 1 + (2$/!0)s+ s2/ !02 .

For this case, the magnitude and phase plots are negative replicas (inverted versions) of
the corresponding plots derived for the case of complex conjugate poles. In particular,
the high frequency asymptote of the magnitude characteristic has a slope of +40 dB /
decade, and the mid frequency asymptote of the phase characteristic has a slope of
+90o/decade.

20

15
% = 0.1
% = 0.2
10
% = 0.3
5 Fig. 1 Log-Magnitude
% = 0.5
and Phase-Angle curves
dB 0
for a pair of complex
% = 0.7
!5 conjugate poles
% = 1.0
!10

% = 0.1
!15
% = 0.2
0"
% = 0.3
"
!45

!90"
"
#
"
% = 1.0 % = 0.7 % = 0.5
!135

"
!180

0.1 0.2 0.4 0.6 0.8 1 2 4 6 8 10

$
$k
!/!&
 PEEII-V-8/16

3 Prelab Exercises
3.1 Derive the transfer functions H(s) for the five circuits in Section 4. Note that the transfer
function of circuit #2 is not the product of two simple transfer functions of type #1circuits.
This is the result of the so called loading effect. Circuits #4 and #5 differ in the sense that the
transfer function of circuit #4 has a quadratic factor with complex roots in its denominator,
whereas the transfer function of circuit #5 has a quadratic factor with real roots in its
denominator.

3.2 Put each of the transfer functions derived in 3.1 in the appropriate standard form of the
transfer functions in section 2. Derive or state any necessary arguments that are essential or
necessary for plotting the theoretical Bode diagrams for each transfer function. Also, plot the
theoretical Bode diagrams for each transfer function. Make sure that you use a semi-log graph
paper for all your plots. Make sure that both plots, the straight line approximation and the
corrected plot, are shown on the same graph paper for every circuit.
 PEEII-V-9/16

4 Experiments

Suggested Equipment:
Tektronix FG 501A 2MHz Function Generator9
Tektronix PS 503 Power Supply
Tektronix DC 504A Counter-Timer
HP 54600A or Agilent 54622A Oscilloscope
741 Op-Amp
Protoboard

9
NOTE: The oscillator is designed to work for a very wide range of frequencies but may not be stable at very low
frequencies, say in the order of 100 Hz or 200Hz. To start with it is a good idea to have the circuit working at some
mid-range frequency, say in the order of 1K Hz or 2K Hz, and then change the frequency slowly as needed.
 PEEII-V-10/16

4.1 Circuit #1

Construct the circuit shown in fig. 2 using R = 2 KΩ and C = 0.1 µF. Apply a sinusoidal input
of amplitude 1 V to the circuit (recheck that it is 1 V whenever changing frequency). Using
the oscilloscope, measure the ratio of the output voltage to the input voltage, and the phase
angle10 over the frequency range from 20 Hz to 20 KHz. Use frequencies of 1, 2, 4, 6 and 8 in
each decade.

R
Vs V0
C Fig. 2 Circuit # 1

Frequency (Hz) output voltage (V) input voltage (V) Phase angle º

10
The phase angle between two sinusoidal signals of the same frequency can be determined as follows: Trace both
signals on two different channels with the same horizontal sensitivities (the same horizontal scale). To calibrate the
horizontal scale in terms of degrees, one can use the fact that the angular difference between the two successive zero
crossing points of a sinusoidal signal is 180 degrees. Thus, by measuring the distance between the successive zero
crossing points of either sinusoidal signal, one can calibrate the horizontal scale in terms of degrees. To determine
the phase difference between the two sinusoidal signals, determine the distance between the zero crossing point of
one signal to a similar zero crossing point of another signal and convert it into degrees.
 PEEII-V-11/16

4.2 Circuit #2
Repeat Section 4.1 for the circuit shown in fig. 3, except that the frequency range should be
from 20 Hz to 100 KHz. Use R1 = 1 KΩ, R2 = 2 KΩ, C1 = 0.1 µF, and C2 = 0.1 µF. Use
frequencies of 1, 2.7, 4.6, 6, and 8 in each appropriate decade.

R1 R2 Fig. 3
Vs C1 C2 V0 Circuit # 2

Frequency (Hz) output voltage (V) input voltage (V) Phase angle º
 PEEII-V-12/16

4.3 Circuit #3

Repeat Section 4.1 for the circuit shown in fig. 4, except that the frequency range should be
from 10Hz to 20 KHz. Use R1 = 1 KΩ, R2 = 10 KΩ, R3 = 1 KΩ, and C = 0.1 µF. Use
frequencies of 1, 1.6, 4, 6, 8, and 9.5 in each appropriate decade.

R1 R2 Fig. 4
Vs R3 V0 Circuit # 3

Frequency (Hz) output voltage (V) input voltage (V) Phase angle º
 PEEII-V-13/16

4.4 Circuit #4

Repeat Section 4.1 for the circuit shown in fig. 5, except that the frequency range should be
from 100 Hz to 100 KHz. Use an R so that Rg + R + RL is around 65 Ω (Rg is the internal
impedance of the function generator, and RL is the d.c. resistance of the inductor), L = 10
mH, and C = 0.1 µF. Use frequencies of 1, 3, 4, 5, 6, and 8 in each appropriate decade.

Rg R RL L
Fig. 5
Vs V0 Circuit # 4
C

Frequency (Hz) output voltage (V) input voltage (V) Phase angle º
 PEEII-V-14/16

4.5 Circuit #5

Repeat 4.4 for the circuit shown in fig. 5, except that the frequency range should be from 100
Hz to 20 KHz. Use Rg + R + RL = 200 Ω, L = 10 mH, and C = 1 µF. Use frequencies of 1,
1.6, 3, 6, and 8 in each appropriate decade.

Frequency (Hz) output voltage (V) input voltage (V) Phase angle º

4.6 A Band Reject Filter (optional)

This is a Band-Reject Filter Circuit. You can attempt this only after completing all the
experiments in Section 4.

1 KΩ

Fig. 6 A Band
0.1 µF 0.1 µF
Reject Filter

1 KΩ 2 KΩ
+
+ +
Vs + - V0
- -
0.1 µF

Apply a sinusoidal input of rms value of 1 V and display both input and output on the screen
of the oscilloscope.
 PEEII-V-15/16

Vary the frequency of the input, read the output on the DVM, and show your instructor that
the circuit actually behaves like a band reject filter.
Find the following:

(1) The maximum value of the output.

(2) The 3 dB frequencies and the corresponding value of the output.
(3) The stop band.
(4) The phase angle at some particular frequencies.
 PEEII-V-16/16

5 Report

5.1 Tabulate all experimental data obtained in Section 4.

5.2 Plot the Bode diagrams (magnitude and phase) for the five circuits in Section 4 using the
experimental data. Make sure that you use a semi-log graph paper for all your plots.
For each circuit, compare the experimental Bode diagrams (plots) with the theoretical Bode
diagrams (plots) previously completed in pre-lab exercise 3.2.

5.3 Prepare a summary.