Scribd
Upload
19 views169 pages

Chapter 8 Consys

Uploaded by

Arkadeep Chowdhury
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
19 views169 pages

Chapter 8 Consys

Uploaded by

Arkadeep Chowdhury
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 169

Chapter 8

• Input is sinusoidal
• Steady state output
will be sinusoidal with
same frequency for
linear systems
• amplitude and phase
will vary with
frequency
Example :
C (s) =
C (t) =
In steady state
Examine
Hence we plot
magnitude and phase
angle of as a
function of frequency
- frequency response
- sinusoidal transfer
function.
Bode plot
• magnitude as a
function of frequency
• phase angle as a
function of frequency
Polar plot
P
r
Example
P2 =∞ =0
. .
- 45º P1
.P
3
Polar plot of

=∞ =0
. .
1
circle
It can be proved that
the plot is a semi circle
1
. .

.
Ex:
-1 =

 =0
Ex:
Real part = 0
Imaginary axis crossing
at  = 1 rad / sec.
Polar plot of

1.
Ex:
Real part = 0
Imaginary axis crossing
at  = 1 /3 rad / sec.

Imaginary part = 0
Real axis crossing at 
= 3 rad / sec.
Polar plot of
1.
Ex:
( How’s that ? )
.

-1
always negative
always negative
Polar plot of
-1
. .
circle
Ex:
Always negative
real axis crossing
Polar plot of

. .
Ex:
= 0 imaginary axis
crossing
= 0 real axis
crossing
Real axis crossing at
Polar plot of

. . .
Bode plot

Taking natural log of both sides


Ex:
for frequencies

= 0 db
For high frequencies
Value is zero db
db 0 db line

. .
0
-20 .
The db magnitude
plot for

is a straight line with


-20 per unit change in
This slop is - 20 db/decade
or
-6 db / octave
The straight line
approximation is
called asymptotic plot
Two asymptotes meet
at the corner
frequency 1 / T.
To get actual plot,
= - 3 db
= - 1 db
= - 3 db
= - 1 db
Ex:
Bode plot for
10 20 log 10
= 20 db
constant for all
frequencies
It is a straight line
with - 20 db/decade
slope ; passes
through 0 db line
for
 = 1 rad/sec
db
20
db

1
.
- 20 db/decade
+ 20 db/decade

db

. 0 db 2.
db
0 db .1

- 20 db/decade
40
- 20 db/decade
20 .
- 40 db/decade
db 8 .
- 20 db/decade 0 db line
0.1 1 2 20
-12
Frequency Response of
Second Order Systems

Critically damped system


Corner frequencies (rad/sec): 1, 1

0<<1 0 db/decade
>1 - 40 db/decade
Overdamped system
Corner frequencies (rad/sec) : 1, 0.5

0 <  < 0.5 0 db/decade

0.5 <  < 1 - 20 db/decade

 >1 - 40 db/decade
Ex:

40 - 20 db/decade

db
0 db line
0.1 1 2

- 40 db/decade
Ex:
40
- 20 db/decade

- 40 db/decade
db
0 db line
0.1 0.5
Ex:

- 20 db/decade
26
db - 40 db/decade
0 db line
0.1 0.5
db 0
. .
.
32 db

0.5 1 5

Slopes in db/octave
Corner frequencies( rad/sec): 0.5 , 1, 5

+ 12 db / octave : s2
. 6
db . 6 38
26 32

0.5 1
20 log K = 38
K = 79.8
Gain margin & phase margin

-1
. -a
. .
Gain margin =

=20 log(1/a) db

for stable system a <1


hence Gain Margin is +ve
Phase margin()
=G(j)H(j) + 180
= 1

phase margin is + ve for


stable system
Gain Cross Over Freq.
db

0 PM GM

Phase Cross Over Freq.


Frequency
Response of
Second Order
Under damped
Systems
Put s = j to get
the sinusoidal transfer function
Put normalised driving
Signal frequency
u=0 M=1 =0
u=1 M = 1/2  = -/2
u= M=0 =-
Correlation between time &
frequency response:
Steady state output
for r(t) = Sin t
c(t) = M Sin(t+)
will give resonant
Frequency where
u = ur
M is max.

ur : Normalised resonant frequency


Where
ur : Normalised resonant freq
r : resonant freq
Mr : resonant peak value
M
Mr

1.0

0.707 Bandwidth

0 r c 

Typical magnification curve of a feedback control system


where
r resonant frequency
c cut off frequency

Bandwidth

The range of frequncies over which


M 1/2 (-3 db)
Normalized
bandwidth
ub

0.707


Bandwidth versus damping factor
Time domain

Frequency domain
For
r = 0
Mr = 1
0 <<1/2 r < n
Mr >1
ng1=[0 0 1];
dg1=[0 1 1];

bode(ng1,dg1)
ng1=[0 0 1];
dg1=[0 1 1];
bode(ng1,dg1)

grid % To show the Grid lines


ng1=[0 0 1];
dg1=[0 1 1];
% to generate a vector of 100
w=logspace(-2,3,100); points logarithmically
equally spaced, including
bode(ng1,dg1,w) boundary points.
ng1=[0 0 1];
dg1=[0 1 1];
w=logspace(-2,3,100);
bode(ng1,dg1,w)
title('Bode Plot for G(s) = 1/(1+s)');
ng1=[0 0 1];
dg1=[0 1 1];
sys=tf(ng1,dg1);
bode(sys)
[Gm,pm,wcp,wcg]= margin(sys);
Gmdb = 20*log10(Gm);
[Gmdb pm wcp wcg]
title('Bode Plot for G(s) = 1/(s+1)');
ans =
Inf -180 NaN 0
ng1=[0 0 1];
dg1=[1 2 1];
% generates a vector of 50 points
w=logspace(-2,3);
logarithmically equally spaced,
bode(ng1,dg1,w) including boundary points.

[mag,phase,w]=bode(ng1,dg1,w);
[Mp,k]=max(mag);
resonant_peak=20*log10(Mp)
resonant_frequency=w(k)
ng1=[0 0 1];
dg1=[1 2 1]; % Critically Damped System
bode(ng1,dg1)
ng1=[0 0 1];
dg1=[1 3 2]; % Over Damped System
bode(ng1,dg1)
ng1=[0 2 3]; % adding a zero between two poles
dg1=[1 3 2];
bode(ng1,dg1)
ng1=[0 0 1];
dg1=[1 1 1];
dg2=[1 0.8 1];
dg3=[1 0.6 1];
dg4=[1 0.4 1];
dg5=[1 0.2 1];
dg6=[1 0.001 1];
bode(ng1,dg1,'b')
hold on
bode(ng1,dg2,'r')
hold on
bode(ng1,dg3,'g')
hold on
bode(ng1,dg4,'y')
hold on
bode(ng1,dg5,'*')
hold on
bode(ng1,dg6,'-')
hold on
hold off
title('bode diagram for different under damped systems')

You might also like