Control Engineering
Lecture 12-Frequency Response Techniques
Muhammad Sajjad Sabir
1
Concept of frequency response
In steady-state response, sinusoidal inputs to a linear system generates sinusoidal
responses of the same frequency.
Sinusoids can be represented as complex numbers called phasors.
M1 cos(wt + ϕ1) M1 = Amplitude of the sinusoid ϕ1 = phase angle of the sinusoid .
M1 cos(wt + ϕ1) = M1 < ϕ1 w is frequency
2
Concept of frequency response
FIGURE
FIGURE 10.2 10.2
Si Siidal fidal
e ef ec e ec e e: a. e: a.e e
; b.
Magnitude frequency response
a fe a f fec if ;cc.i i ; c. ia d a d a ef a ef
Inwords,
In other other words, we define
we define the magnitude
the magnitude frequency
frequency response
response to be
to be the theofra
ratio
Phase the the output
output
frequency sinusoid's
sinusoid's
response magnitude
magnitude to thesinusoid's
to the input input sinusoid's magnitude.
magnitude. We
We define
the response
the phase phase response to be
to be the the difference
difference in angle
in phase phasebetween
angle between the outp
the output and
thesinusoids.
the input input sinusoids. Both responses
Both responses are a function
are a function of frequency
of frequency and only
and apply app
3 FIGURE to the 10.2
steady-state
to the steady-state Si sinusoidal
sinusoidal idalresponse
response off the of e
the c
e system. system.e e: a.
as the
present the input as a phasor in three ways: (1) in polar form,
(10.5)
; (2) in rectangular
Analytical
d (3) using Expressions
Euler's formula, . for Frequency Response
(10.6a)
S e ih i idal i
r the forced response portion of C( ), from which we evaluate
sponse. From Figure 10.3,
(10.6b)
(10.4)
forced solution from the transient solution by performing a
he complex
xpansion on conjugate
Eq. (10.4). of K1,
Thus,
(10.7)
(10.5)
(10.8)
sponse4is that portion of the partial-fraction expansion that
), we obtain (10.8)
Analytical Expressions for Frequency Response
e response is that portion of the partial-fraction expansion that
input waveform's poles, or just the first two(10.10)
terms of Eq.
The sinusoidal
he sinusoidal steady
steady-state − state
output, Css(output,
), is Css(s), is
king the inverse Laplace transformation, we obtain
(10.9)
tion, we obtain
s. (10.6) into Eq. (10.9), we obtain
(10.11)
(10.10)
rse Laplace transformation, we obtain
hich can be represented in phasor form as M ∠ϕ = (M1∠ϕ1) (M
rm as M ∠ϕ = (M1∠ϕ1) (MG∠ϕG),
here M
se function. ∠ϕ
G But G from Eqs. (10.7) andMG < ϕG is the function.
is the frequency response(10.11)
But function
frequency response from Eqs
.
0.8), MG∠ϕG =
s, the frequency G(j ).ofIn
response other words, the frequency response
a system
hose transfer
presented function
in phasor form as M ∠ϕ is G(
= (M ) is
1∠ϕ1) (MG∠ϕG),
s the frequency response function. But from Eqs. (10.7) and
(10.12)
= G(j ). In other words, the frequency response of a system
unction5 is G( ) is
G G
whose transfer function is G( ) is
(10.12)
Plotting Frequency Response
Plo ing F e enc Re on e
G(j ) = MG( ) < ϕG( ) can be plotted in several ways; two of them are (1) as a
function of frequency, with separate magnitude and phase plots; and (2) as a
[Link]
As aplot, where
function of the phasor length
frequency, is the magnitude
with separate andand
magnitude the phase
phasorplots.
angle is
the phase. When plotting separate magnitude and phase plots, the magnitude
2. As a polar plot, where the phasor length is the magnitude and the 1 phasor angle is the
curve can be plotted in decibels (dB) vs. log , where dB = 20 log M. The
phase.
phase curve is plotted as phase angle vs. log . The motivation for these plots
is shown in Section 10.2.
Using the concepts covered in Section 8.1, data for the plots also can be
obtained using vectors on the -plane drawn from the poles and zeros of G( )
to the imaginary axis. Here the magnitude response at a particular frequency
In case of 1
The magnitude curve can be plotted in decibels (dB) vs. Log w, where dB=20 log M.
6
PROBLEM:
Find the analytical expression for the magnitude frequency response and
the phase frequency response for a system G( ) = 1/( + 2). Also, plot both
the separate magnitude and phase diagrams and the polar plot.
SOLUTION:
First substitute = j in the system function and obtain G(j ) = 1/(j + 2)
= (2 − j )/( 2 + 4). The magnitude of this complex number,
, is the magnitude frequency response.
The phase angle of G(j ), ϕ( ) = − tan− 1 ( /2), is the phase frequency
response.
G(j ) can be plotted in two ways: (1) in separate magnitude and phase
plots and (2) in a polar plot. Figure 10.4 shows separate magnitude and
phase diagrams, where the magnitude diagram is
vs. log , and the phase diagram is
ϕ( ) = − tan− 1 ( /2) vs. log . The polar plot, shown in Figure 10.5, is a
plot of for different .
7
FIGURE 10.4 F
FIGURE 10.4 F e
e e e cc e e ee ll ff G(s)
G(s) = = 1/(s
1/(s +
+ 2):
2):
( w2 + 4 )
e a
e a aae ag ii 1de
e20 log
ag de a
a dd . logw
ha ee diag
ha diag a
8 vs phi(w) =a− tan −1(w/2) vs . logw
FIGURE 10.5 F e e c e e l f G(s) = 1/(s + 2):
la l
1
Polar plot :: M(w) < ϕ(w) = < tan −1(w/2) for dif ferent w
w2 + 4
9
Bode Plots
The log-magnitude and phase frequency response curves as functions of log w are
called Bode plots or Bode diagrams.
10
Bode Plots for G(s)=s+a
a function, G( ) = ( + a), for which we want to sketch separate
c magnitude and phase response plots. Letting = j , we have
(10.16)
quencies when approaches zero,
(10.17)
itude response in dB is
(10.18)
= G (j ) and is a constant. Equation (10.18) is shown plotted in
6(a) from = 0.01a to a.
11
Bode Plots for G(s)=s+a (cont.)
a function, G( ) = ( + a), for which we want to sketch separate
c magnitude and phase response plots. Letting = j , we have
(10.16)
quencies when approaches zero,
(10.17)
itude response in dB is
(10.18)
= G (j ) and is a constant. Equation (10.18) is shown plotted in
6(a) from = 0.01a to a.
12
13
Bode Plots for G(s)=s+a (cont.)
a function, G( ) = ( + a), for which we want to sketch separate
c magnitude and phase response plots. Letting = j , we have
(10.16)
quencies when approaches zero,
(10.17)
itude response in dB is
(10.18)
= G (j ) and is a constant. Equation (10.18) is shown plotted in
6(a) from = 0.01a to a.
Please read approximately ten lines on page 966/967 about normalising and
GURE 10.6(Table
scaling. B de10.1land associated
f (s + a):two a. ag i de l ; b. ha e
Figures)
14
the Bode plots for other common transfer functions.
fo G(s) = 1/(s + a)
Bode Plots for
e Bode plots for the transfer function
G(s)=1/(s+a)
(10.22)
has a low-frequency asymptote of 20 log (1/a), which is found
frequency, , approach zero. The Bode plot is constant until the
ncy, a rad/s, is reached. The plot is then approximated by the
cy asymptote found by letting approach ∞. Thus, at high
(10.23)
15
Bode Plots for G(s)=1/(s+a) (cont.)
16
Read it in your free time
Table 10.1 and associated statements (Important)
Bode plots of the following systems:
G(s)=s
G(s)=1/s
You will not leave these topics for next-to-next week. These topics are pre-requisite of
next week’s lecture.
17
