<br>

alysis
percent
59sech
Stability Analysis 11
Using Bode Plots

nwhen

Introduction to Bode Plot

frequency response is to plot magnitude M

IVhen
o
is varied
from 0 to o there and angle against
becomes difficult is wide range of input
hence it to accomnodate all variations
neandHence
H.IW. Bode suggested
the method
such variations with in
linear
M

Bec.-98 in whích logarithmic values

omain gnitude
are to be plotted against logarithmic values of of
frequencies. Such plots are
Logarithmic plots which allows us
Bec.-98) to show wide range of
nyitude on a single paper. variations n
May-99)

both general Bode plot consists of two plots.

in
expressed in iogarithmic values
Magnitude
against ogarithmic values of
en
loop nquency called Magnitude Plot.
Phe angle in degrees against logarithmic values
of frequency called Phase
Angle Plot.

14.1 Magnitude Plota

The magnitude can be expressed in its logarithmic values by finding out
the value
May-99) 2hg1o |G(jo |, which has a unit as decibel denoted by dB.
For Bode Plot |GG0) = 20 log 1o |
d3. |G(j)
as the Such
vaiues are to be plotted against log 10(o).
decibels
y-2000) So magnitude
plot can be shown as in the Fig. 11.1.
1.12 The Phase Angle Plot
ases as
C-2000)

C-2001)
LGjo)
in degrees

Log
Log
Fig. 11.1 Fig. 11.2 Phase angle plot
Magnítude plot
2001)
Ihthis,
angle of degrees which is to be plotted against
domain G() is to be expressed in
y-2002)

(533)
<br>

534
FCS Bode
11.2. 535
The phase angle plot
can be shown as in the Fig.
may be drawn on same
Pl FCS

X-axis is Log
o both
paper 2 3 4567891 Bode Plots
As for both plots with 3
Comm 567891
X-axis. 2 3
from the frequency response 456 7891
Note :To predict the closed loop stability
magnitude and phase angle of open loop transfer function of
loop system, the open DECADE 1

is to be plotted against Log

o and not only G(jo). Giya
H
DECADE2
DECADE3
Bode plot, magnitude in dB and phase
angle in degrees are LINEAR
for
So
a the
and phase angles of G(jo) Ho). plotted against Log magrtudes Y-AXIS

11.2 Logarithmic Scales (Semilog Papers)

To sketch the magnitude in dB and phase angle
in degrees
against Log
is available on semilog graph paper.
logarithmic scale is used. This
In such ate
X-axis is divided into
a
logarithmic scale which is non linear one. While paperthe to.1 0.2 0.3
divided into linear scale and hence it is called semilog paper. 23
NON LINEAR X-AXIS 10 20 30
Log

The interesting part about X-axis is the distance between 1 and 2 Fig. 11.3 Semilog paper
distance between 2 and 3 and so on. Similarly, on such scale distance is
greater than
between and
10 is equal to the distance between 10 and 100 or between 100 and 1000 1

This distance is called 1 decade.

and so on
11.3 Standard Form of Open Loop T.F. Gljo) H(jo)

This is because Log 1 = 0 and Log 10 = 1. The distance is 1 decade whi.

divided into 10parts according to logarithmic scale i.e. Log 2, Log 3, ... Consider G(s) H(s) =
Ks (s+ Z1) (s+ Z,)
s (s+ P;) (s+ Pz)
Now Log 10 = 1 and Log 100 = 2. The distance is again (2 - 1) i.e.. 1 decade samy
Mato
as between Log 1 and Log 10, further divided into parts as Log 20, Log 30,
. that either s or s will
be present at a time and not both. But
not useful to sketch the Bode Plot. this form is
So X-axis is available, which is divided into two, three, or four such cycles
i Hence it is necessary to rewrite the G(s)H(s)
in the time constant form.
decades.
So it is not necessary to find logarithmic value of o while plotting on xaxis bht G(s) H(s) s² K(1+ Ts) (1+ T, s) ...
the logarithmic scale available takes care of logarithmic value of a The advantage d s (1+T,s) (1+ T, s)...
the scale is wide range of frequencies can be accomodated on a single paper.
where
K= ......
=-0 it is obvious that X -axis cannotasbe
as
As Log 0 calibrated from 0 but pet
Px P, x
requirement the smallest frequency may be selected starting frequencylike 001, 01 Again
etc. This hardly affects the either s or s
is present and not both.
result of the frequency response. Ihe standard time constant form can
The Y axis is divided into linear scale similar to standard be denoted as,
graph paper.
next
G(s) = K(1+T s) (1+ Tzs)
....
To clear the idea of semilog paper and decade paper is shown On H(s)
page.
the graph s
(1+T, s) (1+T, s)

The main advantage using the logarithmic representation is that the

multiplication
K = Resultant system gain P= Type of the system
respectivey
and division of magnitudes get ,Th... = Time constants of different poles and zeros.
replaced by the addition and subtraction respans Each
of
The experimental determination of the factor involved in
data is presented
the transfer function is easier if frequency both low phase
angle G(s)H(s) above contribute to magnitude and
will
in the form of the logarithmic plot. Such plot shows function
can
variations of G(o) H(jo)
in frequency domain. Frequency domain transfer
frequency and high frequency characteristics in be obtained
the same diagram. by substituting s = jo in the above expression,
<br>

536 537
FCS .... Bote FCS Bode Plots
K(1+T; jo) (1+ jo)
T2 P K is constant 20 log 10 K
is always constant
... As
gain though o
is varied
G(jo) H(o) = (Go) (1+T, jo) (1+ T, jo) from
o. magnitude plot will be straight line parallel
occur in the above 0to So its to X-axis.
which very frequently form can
magnitude plot for K> 1 is a line parallel to X-axis
Now basic factors be
So
line. While for at a
K<1it is at a distance of 20distance
of
separately. dB reference 20 Log K

and studied above 0 line. Log K below 0

factors is, db
List of such basic reference
(When Gio) H(o) expressed
1)
system gain K, constant factor.
Resultant s Mag in dB (G(jo) H(jo)l
constant form).
Derivative factors) i.e. (jo)-P
Poles or zeros at the origin. (Integral and
2

zeros at origin will be present. +40

Either poles or Mag plot
for
poles and zeros also called
first order factors of
the form (1 +joT*l K= 10
3) Simple +20

which cannot be factorised into real factors,

4) Quadratic factors
2
of the form OdB - 20 LOG K
20 LOG K
Mag plot.
for
1K>1

1+s+ 1+2% j +
-20 K=0.1

is clear in frequency domain K<1

the behaviour of such factors
Once then byadding j0.1 Log o
logarithmic plots of such factors, the resultant logarithmic plot for any Gfo) Hyo c
10

be obtained. The process of obtaining logarithmic plots for such factors Can be Fig. 11.4 Contribution by K
simplified by using asymptotic approximations for each factor. But by This means that in the variation of G(o) H(o)| effect of
is constant equal to
corrections to such plot if necessary, accurate plot may be obtained.
K

0 Iog for all frequencies. This means

K dB shifts the magnitude plot of K

|Gý9) Hjo) | by a distance of 20 Log K dB upwards if

11.4 Bode Plots of Standard Factors of G(jo) H(o) and downwards if K>l
K<1.
For each factor, procedure to obtain its Bode Plot can be divided into fologi This fact is useful to design 'K for the required specification. In such case
steps. |G(jo) H(jo) | plot can be plotted with K as unknown and then it can be just shifted
Step 1 : Replace 's' by j to convert it to frequency domain. upwards or downwards so as to meet the required specification. This shift is nothing
but 20 log K dB, from which required 'K' can be determined.
Step 2 : Find its magnitude as a function of o'.
:
Express the magnitude in Phase Angle Plot :
Step 3 `dB by 20 log 10 |G(jao) H(jo) |
As G(jo) H(0) = K + j0
-1 imaginary part =
Step 4 : Find phase angle by using tan in degrees. =
tan -1 imj part = tan -1 0 = 0°
real part corresponding
real part K
Step 5 : With required approximations, plot magnitude in dB and phase ang n
So it does not affect the phase angle plot as its contribution to phase angle plot is
degrees against Log o by varying o from 0 to o.
Let us start with the basic factors one by one.
This means that phase plot specifications remain as it is for any positive value of
11.4.1 Factor 1 :
System Gain K' K.
But if K' is negative, it always contributes - i80° to the phase angle plot
G(s) H(s) = K
independent of
G(jo) frequency.
ie. HGo)= K+ j0
|G(jo) H(jo) | = VK
+0= K
Its `dB value = 20 log 10
K dB
<br>

538
Bode
FCS Plchts 539
FCS Bode Plots
magnitude plot for 1 pole at origin is a
So straight line of slope - 20
Phase G(jo)
dB/ decade.
angle in Now ato= 1, | H(jo)| 0 dB i.e. this line intersects
the reference 0 dB
degrees line
at
o=1. magnitude +20 dB
Ato 0.1 iit
+90 = has while at
=10 it has magnitude of -
For K> 0 be 20 dB.
As o 0 cannot
=
indicated, the starting
90 frequency
may be selected 2
requiremnent. Mag in dB 34567891 2 34567891
For K<0 per the
-180 as contribution is valid
This
from 0 to
Log o for
range of o
a line of
20

Slope-20 dB / decade
0.1
1
10 To sketch such 1 POLE AT ORIGIN
dB/decade, first OdB>
slope -20
mark the
intersection point
Fig. 11.5 and -20
of o = with 0 dB line for
1

11.4.2 Factor 2: Poles or Zeros at the Origin (jo)p go up by 20dB

then -40 O=0.1
each 1/10 reduction in =1
0=10
Let us consider for simplicity single pole at the origin = 1 ie
frequency from o
G(s) H(s) = o = 0.1, + 40
+20 dB for
or go down Fig. 11.6 Contribution by 1 pole at
dB for o = 0.01 origin
1 by 20 dB for each 10
times
G(jo) H{a) = -
jo 0+jo increase in frequency rom o = 1 i.e. -20 dB for o = 10, 40 dB for o= 100 and so on.
Then draw a straight line, as shown in the Fig.11.6.
= 1
For magnitude plot |G(jo) Hýo) | Consider two poles at origin
1
G(s) H(s) =
Magnitude in dB = 20 log dB s2

= 20 log (o) dB
GGo) H(jo)'= 1. 1
jo jo
= - 20 log o = 1 1
|G(jo) H(o) |
Thus this equation is similar to y = mx ie. 1 pole at the origin contributes to the
magnitude plot according to the equation `- 20 log o
ie. according to the straight line
of slope - 20 .G(jo) H(jo)| in dB = 20 Log
1

Let us see the unit of the slope.

= 20 Log (o)
Equation is, |Gfo) Hjo) | =- 20 log o = o
If -40 Log
0= straight line of slope - 40 dB/decade.
1 |G(fo) H(o) |=0 dB So it is
o = 10 |G(jo) H(jo) |=-20 dB n logarithmic plot, multiplication gets replaced by addition. One
point is
0= 100 G(jo)
H(0)|=-40 dB
Important
to note that at o= 1, |Gfo) H(o) = 0 dB ie. this line though has slope
0= 0.1 | G(o) H(o)=+ 20 dB 0 dB/ it intersects 0 dB line at o= 1.
Now 10 times changes ie. 1
in frequency range is called 1 decade described earlier Hence
pole at origin reduces decade.
the G(jo) H(o) | at the rate
20 dB per
slope of magnitude plot for 1 pole at the origin is called of-
- 20 dB/decade.
<br>

540 Bode 541

FCS Pota Bode Plots
FCS
general for number of zeros at
P

In
the origin
+80 G(s) H(s) = sP
Spole G(jo) H(jo)= jo: jo jo...P time
+60
3poles Common intersection = oP
|G(o) H(jo)
|
point of o=1 o
+40 with 0 dB line Magnitude in dB= 20 xP Log

..
slope = + 20 xP dB/decade
+20 For 1/s i.e.
family of lines with slopes as +20, +
-20 dB/ decade
it gives =
40 +
20 xP dB/decade passing
o 1
point of with 0 dB line as shown
OdB
through intersection in the Fig. 11.8.
For 1/s
slope 40 dB/ dec
-20
For 1/s'
40 slope60 dB /dec
+60
For 1/s s, +60 dB / dec
60 slope-80 dB/dec
Log o +40
-80 =1 O=10 s, +40 dB/ dec
o=0.1 +20

Fig. 11.7 Contribution by poles at origin2 OdB

Similarly for P
number of poles at the origin s, +20 dB/ dec
-20
Gis) 1
H(S)
=P 40

G(jo) H(o) 11 P times

-60
= 1 1 1 Log o
|Gjo) Hýo)
|
P times =
(o) o=0.1 =10

|G(ja) Ho) |in dB = 20 Log 1 = 20 Log (o)"P =- 20 x P Log

Fig. 11.8 Contribution by zeros at origin

- A zero at the origin increases the magnitude at a rate of +20 dB/dec.
So this is a straight line of slope 20 x P dB/ decade but again intersecting wir

Phase Angle Plot Consider 1 pole at the origin

:

0 dB line at o = 1.
1 1
Therefore magnitude plot for 'P poles at the origin gives a family of lines passus G(s) H(s) = Gfo) Hí)
through intersection of o = 1 and 0 dB line having slope -20 x P dB/ decade as stu
in the Fig. 11.7. 0

Now if there is zero at the origin i.e.

ZG(jo) H(jo) =
jo oe =-90 to
pole at origin is line parallel
G(s) H(s) = s
ie. G(jo) H(jo) = 0 +jo s independent of `o. So phase angle plot of
is
Magnitude in dB = 20 Log o dB. Xaxis contributing -90° to phase angle.
only
This is the equation of a straight decade. The
line whose slope is +20 dB/ is For 2 poles at origin, G(s) H(S) =
zero it
change is the sign of the slope, for pole it is -20
dB/ decade while for only.
20 dB/ decade but for both, intersection of line with 0 dB occurs at o = 1
+
<br>

542 Bode
Plots 543
FCS FCS Bode Plots
G(fo) Ho)
= 1 Consider G(s) H(s) =! so
jo jo e.g. Gfjo) H(o)
=0
< G(jo) H(o)
=l. jo jo
90°-90°
=-180°
Factors are
Constant
= 10,
its contribution to
K
magnitude plot
jo

to each other.
= is
Angle gets added 20 LogK= 20 Log 10 +20 dB.
contribute -90°x
poles at the origin
P
P number of
angle to overall at the origin whose magnitude plot is straight line
In general
constant irrespective of '. 1 pole of slope -20 dB/decade
angle plot. This contribution
is i) passing through intersection point of o=1and 0 dB
phase line.
Similarly for a zero at the origin, Now at
o = 1, total magnitude will be
addition of magnitudes of
G(jo) H(jo) =
K

G(s) H(s) = s ie. jo = 20 dB and 1/s.

due to K +0 dB due to 1 pole at origin
i.e
H(o) = 0 + jo
= + tan =+ 90o = 20 dB
at o =
2G(jo) 1

+
90°. The contribution is same as that of pole, fter addition
of two lines, intersection point of =
1 and 0 dB
will shift
zero at the origin contributes
1 20dB. So to draw resultant of
P number of zeros at the origin, the total
the upwards by the two, we can generalise the procedure
In general
only change is its sign. can angle
of value of a This be shown as in the Fig. as
contribution is +90xP, irrespective 11.9.
Draw magnitude plot for K.
Draw straight line representing pole at origin ie. of slope -20
i) xP dB/ decade,
Y-axis scale passing through intersection point of o=l and 0 dB
1 cm
=90 2 zeros at origin #H) o = 1 and 0 dB on the line representing 20 Log K line.
Shift intersection point of
+180°

1 zero at origin
iv) Draw parallel line to the line representing pole at origin from the point obtained
+90 in step (ii).
The slope of this line will be same as the slope of line representing poles or zeros
at the origin. In this example slope of resultant line will be -20 dB/ decade. This is
1 pole at origin because slope of 20 Log K line is 0 dB/decade.
-90

2 poles at origin
Note :
When two lines are added together, the resultant line always has a siope
-180° which is algebraic addition of the individual slopes of the two lines which are
o
added.
Log
lo=0.1 O=1 0=10 So magnitude plot for above G(s) H(s) is as shown on next page.
Phase Angle Plot :
Fig. 11.9 Angle contribution Prepare the table of individual angle contributions and add them to get resultant
phase angles.
Before going to next factor, let us see the addition of first two factors on semilog
paper. The magnitude plots for poles or zeros at the origin are straight lines having Contribution by K By 1 pole at origin Resultant g
slope -20 x P dB/decade or + 20 x P dB/decade - 90° 90
respectively passing throug
intersection point of o = 1 and 0 dB line. Now adding magnitude plot of 90°
above means to shift the straight line drawn, upwards or downwards kon 10 0 -90°
-
by 20 LOg 40 90° 90
depending on whether K is greater or less than 1. This shift 0
experienced, will be 1000 -90° 90
by all points on net
the straight line representing poles or zeros at the origin. Hence - 90° 909
addition of K
and pole or zeros at
representing 0
origin will be a line parallel to line as shown with phase
poles or zeros
Pnase angle plot is straight line parallel to
at the origin a from X-axis
KdB upwards or downwards
at distance
the 0 dB line.
of 20 Log
angle - 90°.
<br>

544
Bode
FCS Plo 545
FCS
Bode Plots
= 1
Mag in |G(jo) H(o) |
dB
Resultant mag. plot K
20 log
+40
20 dB
In dB magnitude
+20

OdB 1
=- 20 Log 1 +
6?T² dB
pole at origin
-20 dB/ of sketching magnitude
dec Now instead plot exactly acCcording
-20° to this equation we
approximate this into two regions and can
can draw straight line approximated
magnitude plot and then by applying corrections we can
Phase angle modify it to an accurate one
in dB required.
if approximation is,
The

90
Eor low frequency range o<<
T
ie. o T´<<lhence can be neglected.
Resultant phase angle Plot Log Magnitude in dB = - 20
Log 1 = 0 dB.

Io=0.1 0=10 0=100 So for low frequencies it is straight line of 0 dB only. Thus the contribution by
h factor can
be completely neglected for low frequency range, as it is very small.
Fig. 11.10
i) For high frequency range >> T1 .1 <<oT2
So whatever may be the open loop T.F. G(s) H(s) factors, the first step in sketching
Bode plot should be the line adding poles at the origin or zeros at the origin and Magnitude in dB = -20 Log o T dB
20 Log K line by the procedure discussed above. ie. it is straight line of slope - 20 dB/ decade. As again for every decade (10
From the above discussion we can conclude one important fact that "The starting slope times) change in magnitude still decreases by 20 ie. slope is - 20 dB/decade. Thus
o,
gets decided by number of poles or eos for low frequency, magnitude plot is 0 dB line while for high frequency it is straight
of the Bode Plot for the function G(s) H) line of slope -20 dB/dec. Now it is necessary to find a frequency a which separates
at origin present in G(s) H(s)."
the frequency range into two, low and high. This is the value of frequency at which
straight line of slope - 20 dB/ dec in high frequency range intersects with 0 dB line in
as there is
e.g. Starting slope of Bode Plot for G(s) H(S) is -20 dB / decade
S low frequency zone. So such an intersection takes place when
one pole at the origin. - 20 Log oT = 0 dB
20(s + 1) decade
Starting slope of Bode Plot for G(s) H(s) = will be --40 dB/ ie. oT = 1
s (s+2) (s+ 4) i.e.
0 = 1/T
as there are 2 poles at the origin in G(s) H(a). Let us go
t the next facior.
s
frequency at which change of slope from
called Corner Frequency,
0 dB to - 20 dB/ decade occurs is
11.4.3 Factor 3 :
Simple Poles or Zercs (First Order Factors) dernoted by oe

The factor is represented as (1 + Ts)* 1 i.a. Oc = 1/T

(l+ioT)1 is 0 dB line upto
asymptotic i.e. approximate magnitude plot for such factor
Ce

Let us start with a simple pole Oe =

1/T and line of slope 20 dB/ dec, when o> @, ie. above = 1/1. @

Asymptotic Magnitude Plot.

G(s) H(s) = (1+Ts)
(1+ Ts)
e magnitude plot shown above
Error Application :
is called

Nowlet us see asymptotic plot, if required.

G(io) H(o) =
1+Tjo how to apply error correction for such
The
actual equation of magnitude plot is,
<br>

546 547
FCS Bode
FCS Bode Plots
Plc
Magnitude in dB
= - 20 Log 1+o² T2 Approximately 0
dB as =
is less than o,
And 2T ie. in low frequency range.
1
Correction at o =
=1 dB ie. ldB down
On Y-axis 2T
1 cm =20 dB
40 General error
values are.

20 Frequency 0, 20 O
/2
0 dB Error dB down 1dB down 1dB down
20 dB down
-20 Table 11.1
DECADE CHANGE By
applying these errors, the actual magnitude plot may be obtained, if required,
40 IN o
as shown in the
Fig. 11.12.
Slope-20 dB/dec
Log
=0.1 =1 =10 10 =100
10T T
40
Fig. 11.11 Actual Mag
20 plot
Now by approximation at a=0 =
1/T, magnitude in dB = 0 dB. But acthr-l.
can be calculated as, 0 dB

Actual magnitude in dB = - 20 Log /1 +o T² 1 dB down

-20
3 dB down
=
-20 Log1T, T
substituting a, =
1/T
40
1

dB down

= - 20 Log v2 = -3dB -60L

Log o
O=0.1 O=10
Similarly at o

Actual magnitude in dB = - 20 Log,1+T

4 =- 20 Log V5 =
-7d8
Fig. 11.12
For a simple zero, i.e. first order zero,
But approximate magnitude in dB at o
=is G(s) H(s) = (1 + Ts) i.e. G(jo) H(jo) = (1 + joT)

= - 20 Log |Gfo) Hjo)

| =1+(aT)?
T2xT=-6
20 Log oT dB

-.Correction at o = Magnitude in dB = 20 Log 1+oT -dB

is -1 dB ie 1 dB down
T Hence all the analysis is applicable for a simple zero with change in sign of slope.
1 =
While at o =
o

nagnitude plot for simple zero is a straight line of 0 dB upto 1/T and
2T then
straight line more than corner
of slope +20 dB/decade for all frequencies
frequency.
Actual magnitude in dB =
-20 Log 1 + T =- 20 LOg 5 4
The errors are +3 dB for o
and +1 dB for
e =
20 or Hence
4T2 analysis
of a when analysis of a
simple pole is
understood. simple zero is very much simple
=-1 dB
<br>

548
Bode 549
FCS Plcth FCS Bode Plots
20. = T t 634°
Y-Axis scale 1 cm 20 dB 10
=
100, ±84.30
+40 Accurate plot with
Corrections as below.
+20 This can be shown

0 dB
Approximated
-20 (dotted) +90 For simple
Zero
40 +45
Log o
0=1 )=-10
00.1
45 For simple
Fig. 11.13 pole
-90
It is as shown in the Fig. 11.13
Log o
Note that varying the value constant, it shifts the corner
of T
i.e. time frequency
w=0.1 0=1 0=10
or left, but the shape remains same as above.
o,to the right
Phase Angle Plot Consider a simple pole
:

G(s) H(s) = Fig. 11.14

1+Ts
0°
The shapes will remain same, for various values of T time constants.
G(jo) H(ja) =
1
Z G(jo) H(Go) =
= tan oT
1+ joT tan -l 0T
It is important to note that phase angle is t 45° for a zero or pole at w = @ =
1/T
1 Ex. 11.1 Sketch the Bode Plot for the system having
20
While for a simple zero, G(s) H(s) =
s(1 +0.1 s)
G(s) H(s) = + Ts
1

.. 2Gjo) :
First see that given G(s) H(s) is in the proper time constant form or not. If not
G(jo) H(o) =
1+joT H(jo) =
tan-=+ tan 1
oT Sol.
S0, arrange it in the time constant form. Now identify the factors.
So the shape remains the same. Only sign the angles change from negative t
of
K= 20 . =
Its magnitude = 20 Log 20 +26 dB
constructeu v
positive when factor changes from pole to zero. Such plot is to be M
pole at origin. Its magnitude plot is straight line passing through intersection
a as shown below
actually calculating angles for different frequencies, So make table point of o = 1 and 0 dB with slope 20 dB/ decade. -
t tan oT (+ for zero, - for pole im) Simple pole 1
comparing with
1

= +5.71° 1+0.1 s 1+Ts

0.lo, 10T
T= 0.1
O.50,1 + 2660qd rad/sec
2T
o ==10 T
0.1
of

t 45° ie. Asymptotic magnitude plot is

0 dB up to
o = 0
10 and then straight line
slope
20 dB/decade. Procedure to plot resultant,
Draw 20 Log K line ii) Draw line for l pole at origin
<br>

551
550 Bode Bode Plots
FCS = -
FCS Pla simple pole tan -1
oT =-tan
0
dB on 20 log K line and from due to
0.1

point of =l and
Shift intersection a
o this
iü)
line representing
1
pole at origin. This line will pot 1 pole at the origin is always -
draw parallel to have due to 90. If required, more may be
slope
selected to
draw the smooth curve. Let us combine all the things on'o values
-20dB/ decade. semilog paper to
of and
K poles at origin will continue, till next factor complete the
Bode plot.
iv)
This addition becomes
= 10.
dominant i.e. at o= 0-
= onwards will be (-20 Mag
from o 10 dB/der ade Resultant
Hence resultant slope - as Starting in dB
G=10=0
simple pole i.e. resultant 40 dB/ decade. This -20 dB/ decade Y-Axis
slope) + (-20 dB/decade) due to
for sirmple pole cm =20
1
dB
as there is no other tactor present in G(s) H) will +40
continue up to o0
- 40 : Mark the point of intersection 20 Log K = 26 dB
Procedure to draw dB/decad line of o = 10
+20 Effect
and 1/s. From this we want to draw slope of
of K and
line representing addition magnitude of G(jo) H(jo) corresponding
of K Resultant slope
-40 dB/decade. So whatever is the tothis 0 dB>
40 dB / decade
dB for decade change it o ie. at 0=100.
a

intersection point, will get reduced by 40

mark that point and draw the lne.
So 20
of different slopes as
semilog paper itself, draw the lines
OR on -20, - 40,-
very light as shown and then draw parallel to these 6, 40 1
pole at origin
-80, +20 dB/decade etc. lines of Phase
necessary. angle 1/s, -20 dB / decade
the required slope in magnitude plot, wherever -900
Such lines -105
are shown in the +60
Y-Axis scale Resultant phase
Fig. 11.15. -60 dB/ decade 1

cm =20 dB -120 angle plot

+40
Draw such -40 dB / decade
lines very light -135
+20
and then draw
parallel to these 0 dB
-20 dB/ decade -150
lines from the Log
required point of -20 9=0.1 ()=1 O=10 =100
+20 dB / decade
required slope.
For the -40 Fig. 11.16
phase angle plot Log o Observe :
prepare the table o=0.1 0=10
9=1 1) 20 Log K line.
of angles as
below : Fig. 11.15 2) Line of slope -20 dB/ decade as only pole at origin. 1

o in rad/sec due to pole at

1 due to simple OR Resultant ) Intersection point of o and0 dB shifted on 20 Log K line and line parallel to
= 1

- tan 0.1o - 20 dB/decade is drawn which is resultant of K and 1/s.

origin pole =
0.1 - 90° 90.57° 4) This continued till next factor becomes dominant i.e.
o= = 10. Till
o
-0.57° contributes 0 dB only and there is no change in the
0.5 - 90° 92.86° O= Oc= 10, simple pole
2.86° slope.
1
-95.7°
oO from intersection point - 20 dB/decade line and
-90 -5.7° o= 10 line i.e. point P shown
2 -90° - 11.3° - 101.3° dB/decade and hence directly resultant of slope
S changed by -20
- drawn from point P. This is drawn parallel to
10
-90° 135 20) = 40 dB/ decade is
45°
50 - 90° 1680 40dB/ decade line drawn light on semilog paper shoWn.
78.79°
<br>

552 553
Bode Bode Plots
FCS Plsit%
FCS
: Quadratic Factors and
11.4.4 Factor 4 form, frequency, 0>0n as is very low.
Consider quadratic
pole of the For
high
=
1
expressed in time constant
G(s) H(s) form 2

1+s+ Magnitude in dB =-20 Log

1

2
G(jo) H(o)
jo s=- 20 Log
(On
40 Log
(On
=
o, is constant for that factor. -40 Log o + 40 Log wn
where o is variable and equation of straight line
slope 40 dB/decade. of -
1
as j² = 1 This isthe
magnitude plot for quadratic factor is 0 dB line till corner frequency
Hence general
of slope - 40 dB/ decade.
j-|O
1+25
andthen
a straight line
Qc

1 To find corner frequency

- 40 = 0 dB
2 Log
On
On = 1
1
i.e.
=
|Gjo) Hfo)
|
= n

factor.
+45² So o, is the corner frequency for such

.. 1
Magnitude in dB = 20 Log Asymptotic Mag. plot for quadratic factor
2
2 40
+42
On 20
1 0 dB line
0 dB
= 20 Log
2 2
+ 4¿2 40 dB/ decade
On -20 40 dB

. Magnitude in dB = -20 Log 1

2
+ 482 dB
-40
decade
|®n
Log o
Approximation :

10
0.1
For low frequency, 0<<@n O-10
<<1
On
Mag in dB = Fig. 11.17
-20 Log 1 = 0dB corner as the error for the quadratic
above asymptotic plot is not so- accurate
Thus similar to simple But
pole, quadratic pole also is negligible till its the
frequency ocurs. factor not on the value of damping ratio.
only depends on o
but also
<br>

554
FCS Bode 555
on the magnitude plot Plol FCS
Let us see the effect of variation
of necessary to modify Bode Plots
is
Hence it magnitude
above at its Corner frequency for various plot for 2nd order quadratic
2

- 20 Log + 482 shown values of . pole a5

Actual magnitude in dB = use Table 11.2 to
Students can decide correction
formula, for given or
find the corection
Lusing the
Correction = - 20 Log
Now at o = ®n i.e. -= 1
2, dB at o = 0, of pole
Positive correction upwards and negative correction
- Let us magnitude plot for a quadratic zero can downwards.
Actual magnitude in dB
= 20 Log V48. prepare a table The be obtained
11.2
for various of basic asymptote and then by reversing the sign or
values of E and corresponding error values. See Table the slope for various values of .
by reversing the
signs of the corrections
corner frequency Hence it looks like as at
|Accurate magnitude Approximate 11.19 shown in the
error
in dB magnitude in dB Fig
for quadratic 40
+ 13.97 pol
0.1 + 13.97
0.2 + 7.95 0
dB up
+7.95 dB up 30 11 say

+ 4.43 0 + 4.43 dB
0.3
up 20
0.4 + 1.93 + 1.93 dB up 6=0.7 Asymptote +40 dB/ decade
0.7 - 2.92 0 2.92 dB down 10 (=1
On Y-Axis scale
0.9 - 5.10 5.1 dB dowm 1 cm =
10 dB
0 dB
1 - 6.020 6 dB down
¿-0.3
Table 11.2 -10
And due to the errors calculated above the magnitude plot for quadratic pole gets E=0.1

modified as shown below: -20

Quadratic zero
Mag plot
Log o
0.1 1
10

Fig. 11.19 Quadratic zero

Y-Axis scale Let us see phase angle table :
1 cm =
10 dB
20 G(jo) H(jo) for a quadratic pole
10 E=0.1 1 + j2%

= 0°
Z G(jo) H(o)
E=0.3
-10 25 (o / o,)
E=0. tan o,)
-20
|1-(o/
=1

ZG(o) H(jo) = - tan 1

25 (o/o,)
Asymptote of
slope40 dB / decade
1-(o/o)
Logo The table for E = 0.3 is shown below
0 100
2x 0.3 xo/o,
=
tan
Fig. 11.18 Quadratic
pole 1-(o/o,)
<br>

556 Bode
FCS Plots 557
FCS
intersection pont on Bode Plots
Shift this 20 Log line
K

step 3. This is addition of constant and draw parallel

line to the line
drawn in K and number
0.1 -3.46° origin. of poles or Zeros at
the
of this line
0.5 -21.8° Change the slope at various corner
upon which frequencies by
factor is occurring
1
90 ie. dependingmust appropriate value
at corner frequency.
slope be changed For a simple
-
90° and hence must
pole, Do
by- 20 dB/decade, for a
ie. o = 0
=
0, it contributes dB/decade etc. not draw these
individual lines. Change
simple zero by +20
So at =1 approach
to the slope of line
obtained in step 5 by respective value and
0 draw line with resultant
But according to above formula
when >
1, becomes Continue this line till it intersects next corner slope.
180° as 0 positive, frequency line. Change
On
and continue. Apply
necessary correction for the slope
quadratic factor.
must be considered,
case the angle contribution obtained by subtracting18 Prepare the phase angle table and obtain the table of o and
in such
from the positive . vi) y
actual calculation. Piot these points and draw the resultant phase angle
smooth curve obtaining
=+ the necessary phase angle
plot.
=2 =- tan-04]= -(- 21.8) 21.8°
e.g.
somber that at every corner frequency slope of resultant
line must change.
must be considered by applying correction of
.
the actual angle contribution
But
180° 21.8° -
180° 158.19°. =-
This happens because behaviour of tan function 44.6 Frequency Response Specifications
ie.
for the complex quantities with real part negative or imaginary part negative objective of control system design
The basic
be identified on calculator by using above formula. Hence phase angle table bero is meet the performance specifications. These
to dB

specifications are the constraints or limitations M.

put on the mathematical functions describing

the system characteristics. Such frequency
0.1 3.46°
response specifications are described below. 0 dB
0.5 21.8°
1 90°
Consider a general frequency response of a -3 dB
BANDIWIDTH

2
+ 21.8 - 180 = -158.19 systern.

4 + 10.09- 180 =-
170.9° i) Bandwidth:It is defined as the range of Log o

10 + 3.46 - 180 = -176.53° frequencies over which the system will

Fig. 11.20
respond satisfactorily. It can also be
defined as range of frequencies in which
180 the magnitude response is almost flat in nature.
For quadratic zero, sign of the angle should be made positive. So it is defined as range of frequencies over the magnitude of closed loop response
Cjo)
Note : For quadratic factor make sure that its roots are complex. If roots are real ie not drop by more than 3 dB from its zero frequency value.
Rio) does
tactorise it and consider its two components independently as simple factors rai
than quadratic. The above discussion is applicable only when the roots 0 ) Cutoff frequency :
quadratic factor are complex conjugates of each other. It is denoted by @
requency at which the magnitude of the closed loop response is 3 dB down
11.5 Steps to Sketch the Bode Plot from its zero
frequency value is called cutoff frequerncy.
i) Express given G(s) H(s) into time constant
The
range 0
to o, is nothing but bandwidth of the system whose frequency
form. response is
shown.
Draw a line of 20 Log K dB.
origin,
il) Draw a
line of appropriate slope representing or zeros at the
poles
passing through intersection point of o =
1
and0 dB.
<br>

558
Bode
FCS response. It indicates Plots 559
speed of the the ability FCS
Bandwidth indicates the proportional to the rise time. mnargin P.M. :Similar to the Bode Plots
reproduce the input signal.
It
is inversely response. t Phase gain,
it is possible to introduce phase
means small rise time means fast the
system ie. negative angles without affecting
magnitude plot of lag in
bandwidth magnitude Curve
The slope of the
resultant near annount of additional phase lag which can Gf) H(o).
Cutoff rate : the The be introduced in
ii) cutoff rate. cutot reaches on the verge of instability is called the system till
frequency is called value of magnitude
system phase margin P.M.
: the maximum of the postive phase margin means such negative
Resonant peak (M,) It is value of resonant peak dosed The angle introduction
iv)
response as shown. Larger the more \o possible before system becomes unstable. Such system is in system is
frequency step input. It is a measure stable system. While
overshoot of
peak
system for is the margin means present negative phase lag should negative
value of
of relativNe phase be changed by adding positive
hence phase margin is said to be negative and system
stability of the system. angle is unstable.
The frequency at which resonant peak Mathematically it can be defined
vi) Resonant frequency (o,): resonant frequency. M, Occurs
response is called It is inverselyin P.M. = [ZGijo) H(joat o = Ogr
closed loop frequency J-(-180°)]
proportional to the rise time.
= +
P.M. 180° 2Gfjo)
specifications but the specification which are
These are the general Information About
stability and relative stability analysis are discussed below.
mos
11.6.1 More G.M. and P.M.
important from the
requency at which magnib,d. Absoluteand relative stability jrom frequency response plots :
vi) Gain cross over frequency (0gcThe
Gfio) H{o) is unity ie.
1
is called gain cross over frequency. Consider a system with negative feedback, with forward path
transfer function
H(jo) is expressed in dB. And dB value G(s) and feedback path transfer function H(s). Then the closed loop transfer function
Generally magnitude of Gfo) oflis
Gs)
20 Log 10 1 = 0 dB. M(s) =
H{jo) is 0 dB k 1+G (s) H(s)
It can be defined as the frequency at which magnitude of G(jo)

Ogc
With sinusoidal input we can write the ratio of output to input in the closed loop
The frequency at
: phase angle of

Phase cross over frequency (ope) which aS,

vi) G (j@)
G(jo) H(o) is - 180 is called phase cross over frequency, Opc:
MG o) =
1+GGo) H
(jo)
ix) Gain Margin G.M. As seen earlier in the root locus
: as gain is increased the
K

If 1 + G(j o) H (jo) = 0 then M (j o) = o (or infinite).

system stability reduces and for a certain value of it becomes marginally
K

C
as
(jo)
stable. (Except first and second order systems). So gain margin is defined
the
Since M| o)
= = 00
on

margin in gain allowable

by
which gain can be increased till system reaches Rjo)
the verge of instability. Even if theinput R (G o) = 0 (there is no input), M(Go) = 0x = Indeterminate
The positive gain margin means such increase in is possible before systen
K

Hence even without input there can be output. In fact if this condition is exactly
becomes unstable, hence system is stable and negative gain margin means g
kS Satistied the system will oscillate at a constant frequency.
system
than Kmar and system is unstable so `K' is required to be reduced to make the of the :
Magnitude and Phase Condition for Sinusoidal Oscillations
stable. Mathematically it can be defined as reciprocal of the magnitude
G(jo) H(o) measured at phase crossover Uscillations result if +G (jo) Hjo)
1 or H =-1 = 0 G (G o) G @)

frequency.
1 that
Ocemeans, the right-hand side must be a complex number hence this condition
G.M. = actually
|Gjo) Hjo) |
Gi o) Hj o) =
In decibels G.M. = 20 Log 1
-1 +j0
= 12t180°
|Gio) H(j0)| o= 0pc This gives us two conditions.
G.M. = - 20 Log10 |G(j@) se Condition : There must be some frequency (say o =
0pc) at which the
More positive the G.M., mnore
Hjo)| o= @pc
stable is the system. angle of G(jo) H(o) = -180° (say)
<br>

560 561
Bode
FCS Plo FCS Bode Plots
: At thhiis frequency @pc where phase Critical value
Magnitude Condition is-180 GM =
critical K of |G (jo) H (j0)lo- pe
I) magnitude of Gj o) o) = 1
H he actual K Actual value
of |G (jo) H
Thus for oscillations |G(jo) H(o) lo=on1 (0) lo- ape
: Actual value 1
Instability system is unstable.
of
|G (j0) H j0)lo- opc
If|G (j@) (0)loz
H
ope
>1, then the there some in dB = 20 log
1
X

can only be unstable if is frequency at which CM = 20 log (1) - 20

log ()
Thus a system oper angle
= - 180° (or + 180) and at this frequency of
G(i o) H o)
1 system is marginally
stable =-20 log (x)
If |G (jo) H G0) lo-op:= =-20 log [Actual value of G (jo) H G0)
|
(j@)lo=onr >1 system is unstable = -20 log lo-ane
If|G (j0) H [|G (j0) H Go)lozoe!
system is stable.
If|G G0) H(G0)lo-opc <l Summarizing in simple cases we can use,
P.M. = 180 + (Angle of G
11.6.2 Relative Stability (j o) H o) at @= @ge)
us an idea as to how far the system is away trom the
Critical
These measures give critical K | G (jo) H G0) lo= @pc
These are known as Gain Margin (G.M.) and Phase Margin GM =
condition of instability. actual K Actual|G(Go) H ((0)
as G(s) H(s) =
K lo =@pc
(P.M). Consider that a system such s is unstable if
(1+T; s) (1+ T, s)

K= 100 (knoun |G(j@) H

(@)lo=opc
K>100 i.e. at K = 100, the system is marginally stable then for

critical value of K) G.M in dB =- 20 log L|G G0) H (G0) lo=@pe

|GG0) H j@)lo= ope =1
These expressions of G.M. and P.M. can be directly used for mathematical
However if the actual gain is less, (say 50) then the system be stable
will Cain

determination of G.M. and P.M. for a given system.

Margin is the ratio of the Critical Gain to Actual Gain.
Critical value of gain K 11.7 Calculation of G.M. & P.M. from Bode Plot
Gain Margin =
Actual value of gain K
According to definition of G.M.
Phase margin can be understood by considering a particular frequency at whid G.M. = - 20 Log |G(jo) H (G@)
lo=ope dB
= 1. This frequency is known as Gain Cros
the magnitude of loop gain G (o) H(jo) = 0 dB - 20 Log |G(jo) H(o) lo= 0pe db
Over Frequency (oge). At this frequency if angle of G G0) H(j0) = -
140°, the system i Go) lo=ox in dB can be directly read out
-
Stable. It is marginally stable if this angle is 180°. The difference between - 180 (0
Now 20 Log |G ((o) H from magnitude
+ 180°) and the plot.
| G (jo) H (jo) lo=o] is the phase margin.
say point A
P.M. = 180° + EXtend o =0pe line upwards till it intersects resultant magnitude plot
| G(jo) H(o) lo= ge 11.21. The magnitude corresponding to point A is
dShown inthe Fig.
In the above case
|G(jo) H (jo) corresponding to
lo=opc" The difference between dB and magnitude
0
P.M. = 180° + (- 140°) = 40° POnt A is Gain Margin. If point A is below 0 dB. G.M. is positive and if point A is
The phase margin is so defined that for stable system P.M. is positive. above 0 dB, G.M.
is negative.
Thus oge = Gain Cross Over Frequency GH. is 1 Similarly
that frequency at which gain the Bode's P.M. = [2|G Go) Go) lo=ael-180)
H

or gain GH in =
is o
dB 20 log (1) At the gain cross over frequency can be read out from phase angle plot. Extend
= 0 dB. Now
Z| (jo) H ( o)
G

plot
ofG (o) H (jo) just crosses '0 dB line. line
downwards till it lo= @gc
This point of intersection is point
C

= as shown intersets phase angle plot. point and -180° line i.e. point
C
Opc Phase Cross Over Frequency of
in 11.21. The distance between this
is that frequency at which the angle D
is nothingthe Fig. and
G(jo) H{o) =-180° (or + 180°) margin. If point is above -180° line, P.M. is positive
C

if it is but the phase

below -180°
line, P.M. is negative.
<br>

562 563
Bode FCS Bode Plots
Plot
FCS
P.M. and G.M. are positive while system Mag in dB
stable when
System is said to
be
G.M. are negative. Now when system is sia
both P.M. and nature then G.M. and P.M. both is onthe
to be unstable when stable in |Gjo) Hjo)l
marginally are in dB
verge of instability i.e. gc = Opc: This condition gc pc zero
=

when
is useful to design
This is possible the
systems.
marginally stable <
Ope: While for G.M. G.M. = 0 dB
positive i.e. stable system,
Ogc 0 dB
ForG.M. and P.M. and PM
system, Oge > @pc:
negative ie. unstable may be
stable systems G.M. obtained as +o while
In some absolutely as - o0.
may be obtained fot ZGjo) Hjo)
inherently unstable systems GM.
in degrees

All such G.M. and P.M.

conditions are shown in figures below. -180
PM.=0
:
Stability Conditions

Mag in dB
gcpc Log o
IGGu) Ho)l
in dB Fia. 11.23 gc Opc G.M. and P.M. zero, marginally stable system

0 dB
.4 What Should be Values of GM and PM of a Good System ?
+ve G.M.
It is obvious that a system should have gain which is lower than critical value so
LG(jo) Hjo)
C that the system is far removed from unstable conditions.
in degrees
+ve P.M.
This is necessary because for most of the systems the transfer function of
-180
components and systems changes with variation in temperature and pressures of
Surrounding environment. Moreover the gains are also dependent on supply
ac Log o
frequency, supply voltage, loading conditions, variations in control energy sources
such as pneumatic air pressure (in pneumatic systems).

Fig. 11.21 oc <pe G.M. and P.M. positive, stable system It is matter governed by a great degree of experience as to how much low
a
should be the system gain. If gain is high, the gain margin is low, and system's step
Mag in dB response shows high overshoots and long settling time. On the contrary very low
gains give high GM and P.M. but also causes higher steady state error, higher values
IGjo) Hjo)l
in dB
of rise time and settling
time and in general give sluggish response.
A

-ve G.M. Ihus we should keep the gain as high as possible to reduce steady state error and
0 dB
0bain acceptable response speed and yet maintain adequate GM and P.M. These
B
aSkS are, of course, difficult and require tremendous efforts in design of controllers
compernsating networks. An adequate gain margin of 2 (i.e. 6 dB) and phase
LGjw) HGo) d
aiagin of about 30° is generally considered good enough as a thumb rule.
in degrees
11.9
-180 How to Improve the GM and P.M. ?
-ve P.M.
If the closed are not
C adequate, loop system is stable but the gain and phase margins
Opc Ogc Log o how can the system be improved ?
The
easiest way to reduce the'gain'. However this increases the
Fig. 11.22 Ogc > pe G.M. and P.M. negative, unstable syste
to improve is
<br>

564 Bode 565

FCS Better methods Plots FCS Bode Plots
system sluggish. are available IGio)
H(o)| =-20 log
error and makes the or Compensators. Consider o+
20 log Ky
steady state Compensating Networks These
in dB
are adding is varied (reduced) and the the Bode
methods system. If the gain gain |Gijo)
plots of an
unstable stable. Alternatively the phase curve = 1, H(o)|=20 log Ky
system can be made curve at o
brought down the phase is lifted up. The system plot can shown in the Fig. 11.25 Now this
line after extension,
so that the can be This is is intersecting
modified (near opc) known as Gain Compensation thern be
0 dB
suggested are line
modifications andPhase
stabilised. The at @1

Compensation.
compensation without |G(jo) Hjo)| in dB
components which achieve phase
There are no
modifying
LEAD compensator is closest to
this id -20 dB/dec
the gain curve. However the
one of the principle aims of
the control systemstudy
Summarizing, we state that isty
stability, GM and PM and main aim of design is to find suitable
determine GM &
20 log Ky
or controllers to achieve adequate PM at a given
compensating networks gain.
compensating networks are (i) Lead network
Three principle types of (i) Lag
Lag-lead network. Proper design of compensators using Bode's plots,
network and (i) more courses
- M
plot are covered in advanced in control system.
Log o
polar plots and

11.10 Determination of Kp Ky and K from Bode Plot

=1
-
a) K, Static position
error coefficient:
This is defined for type 0 system and given by Fig. 11.25
Lim G(s) H(s) = Lim G(jo) H(jo) Then 0 dB = -20 log 01 + 20 log Ky
Ky =
S0 )>0 20 log 1 = -20 log o + 20 log Ky

Consider Bode plot shown for Type 0 system. As o- the magnitude

0,
error coefficient Kp.
of
Ky = 1
Gijo) H(jo) is nothing but 20 log K which is positional
= 20 log Kp as Ky =
20 log K o0 @1

So frequency at which initial line of slope -20 dB/dec 0 intersects 0 dB line is

4
20 log K
the value of velocity error coefficient.
|G(jo) Hjo)) in dB
c) K, Acceleration error coefficient :

0 dB Lim
This is defined for Type 2 system as
s0
s G(s) H(S). For Type 2 system, as
S0ie. o0 the magnitude plot has two dominating factors namely K i.e. K (for
O) and poles at the origin. So initial slope of magnitude plot is 40 dB/dec and
2
+ Log o equation of this
line is,
Fig. 11.24 |G(jo) Ho)| = -20 log + 20 log K,
o
b) K, - Static velocity error coefficient in dB
This is defined for Type Lim at o= 1, -20 log o
= 0 dB/decade
1 system as s G(s) H(S).
S0 |G(o) H(o) | = 20 log K, dB
are K and I This is dB line at
o0, the factor contributing to magnitude plot for Type 1 System shown.
0
As
The shown in the Fig. 11.26. When extended, this line wìll intersect
pole at origin. So intial slope of magnitude
plot is -20 dB/decade
as Say
a
equation of this line is
<br>

566
Bode
FCS Plcts 567
FCS Bode Plots
origin. Straight
pole at
40 dB/ dec For 1 line of slope - 20
intersection point of o=1 an 0 dB. dB/decade passing through
intersection point of
Shift o=1 and 0 dB on 20 Log K line and draw parallel
20 log Ka dB /decade line drawn. This will continue as a resultant
- 20 to
occurs o
of K
andtill first
Corner frequency i.e. = 2.
s
0 dB as there is simple pole will
Wc1 = 2, it contribute the rate of - 20
At
iv) hence resultant slope after 1 = 2 dB/ decade
1 1
becomes- 20- 20 =-40 dB/decade. This is
addition of K, and This will continue
0=1 till it intersects next cormer

Fig. 11.26
+Log o
()
= 20.
- 20 log (oa ) + 20 log K,
frequency line i.e. 0
dB =
0
E 20, there is simple
pole contributing -20 dB/decade and hence
resultant
= 20 clone after o2 = 20 becomes - 40 - 20 =-60 dB/decade.
ie. +20 log K
(o,)2
log 1
This is resultant of overall G(s) H9) ie. G(jo) Hfo). The final slope is
= 60dB/decade, as there is no other factor present.
K, (o,)?
Step 4 :
Phase Angle plot
:
Convert G(s) H(S) to Grio) Hol
So frequency which initial line of slope -40 dB/decade intersects
at with
n0 dB line 2
error coefficient i.e. G(o) H(ja)
gives the value of square root of acceleration

Solved Problems ZG(jo) H(o) =

22+j0
80 =
<joZ1+o 21+
Ex. 11.2 A unity feedback control system has G(s) Draw the Bode Plot.
2 20
s(s+ 2) (s+ 20)
Determine G.M. P.M. Oge and One. Comment on the stability. 22+ j0 = 0°,
1
ie. pole at origin
1 is - 90° constant
Ljo
Sol.
:
Step 1: Arrange G(s) H(s) in time constant form. 1
-
80 , =1
=-tanl and
20
G(s) H(s) = H(s)
1+j 1+j
s(s+ 2) (s+ 20) 20
80 2 .:.
Phase Angle Table :
Resultant
so120|10) jo
- tan-1
2
tan
-1 0
20
OR

Step 2 : Identify factors: 0.2 -057° -96.27°

90° -57°
i) K = 2 ii) 1 pole at origin -1407°
1
2 -90° -45° -57°
iii) Simple pole 1 1 - 21.8° 18776°
with 8 -75.96°
-900 -195.29g
10 -78.690 2656°
-90° - 219.28°
20 -84.28° -45°
-90° - 240.58°
Oc2=20t
1
iv) Simple pole 1 1 87.13 -63.43°
with T2 = 40
-90°
1+S 20 90 90° 270°
-90° - 60 and
20 To
draw straight lines decade and dB/ decade fromn o1 =2
Step 3 : Magnitude Plot Analysis O2= 20, draw faint lines of- 40 dB/ - 20, - 40, - 60 dB/decade from intersection point
of of slopes points.
i) For K = 2, 20 Log K= 20 Log 2 o= 1 and 0 parallel to them from respective
=6 dB dB line and just draw
<br>

568
Bode 569
FCS Pleh FCS Bode Plots
solution.
Step 5: Bode plot and For a unily
feedback system
Gs) = 800 (s + 2)
11.3
Ex. s (s+ 10)
(s 40) +
Akthe Bode plot , asymptotic in nature. Comment on stability.
Arrange G(s) H(S) in time constant form.
Step 1:
STABLE
rad/sec
rad/sec
Sol. :
Sale dB 800x 2x +1
=
dB SOLUTIO
o=6.35 +21
SYSTEM G(s) H(s) 2
Y-aks K=6
.1 s2x10
=
G.M. 1+40x1+
10) 40)
On Log

Factors are
. 2:
567891
20 Step
Constant K = 4,
1/s
for 1)
1/s2
34 2 poles at the origin,
decade
)
2

dB/
i)
Simple zero, 1+,
S
T =1/2, o= I -=2 rad/ sec.

=100 1
= 1
60 Simple pole, T, 1/10, @a
=10
7891
iv) rad/sec.
decade
1+
10
1 1
dB/
Simple pole, T3 = 1/40, Q = = 40
=6.35
a-1020
v)
s rad/ sec.
40 1+
@pe 40
- Step 3 :
Magnitude plot analysis
For K = 4, 20 Log K= 20 Log 4 = 12dB.
decade

ii) 2 poles at the origin ie. 1/s. It contributes a straight line of slope 40
dB/decade passing through intersection point of o = 1 and 0 dB. So starting
slope becomes - 40 dB/decade.
dB/

.M.
20
i) Shift intersection point of o = 1 and 0 dB on 20 LogKline and draw parallel line
456791
to - 40 dB/decade. This represents addition of K and
1/s. This resultant will
continue till first corner frequency Q1=2.
1") At o1 = 2, simple zero occurs which contributes + 20 dB/dec individually and
hence resultant slope from '2 onwards becomes - 40 +20 =- 20 dB/ dec. This
continues till o = 10.
At Qg = 10, simple pole occurs which contributes - 20 dB/ dec individually and
V

..+20
hence resultant slope from 10 onwards becomes - 20 - 20 40 dB/dec again. =-
dB/dec dB/dec This continues till oa = 40.
dB/de
-
na3 = 40 ,
simple pole occurs which contributes 20- dB/dec- individually and
hence - 20 60 dB/dec. This =
-60 40 -20 resultant slope from 40 onwards becomes 40
tunues upto
Step 4 : Phase
o
o as there is no other factor present in G(s) H(s).
Angle Plot
Fig.11.27
<br>

570
FCS Bode 571
Plots FCS Bode Plots

=
2
G(jo) HGo) jo
10/40
Z4+j0 214 J0 sec
- ABLE

STAR
= 2 dB
2G Go) HGo) jo dB
SOLUTION Cgcadr
+24
21+ 12 M
SVSTEN

10 40 K=
=
G.M,

= + tan
24+j0 0 Z 1+= Log

20
= -180° constant
poles at origin decade

Z2
1
- tan and =- tan -l 0 dB/

ja 10
1 +
jo 40 -60
1+ 100
10 40 dB/decade

g=40
Phase Angle Table
1
+
tanl 2 - tan tanl 40O R
Go)? 10
40
0.2 -180° +57° -114° - 0.28° -17572°
2 -180° +45° -11.3° -2.86° -149.16°
10 -180° +78.6° - 45° -14.03° -160.430
20 -180° +84.28° -63.43° - 2656o 18571°
dB

50 -180° -20
+877° -78.6° - 222°
-51.3º
100 -180° +88.85° - 24354°
-68.19° -84.28° c12

-180° +90° - 90° 90° - 270° dB/decade

P.M

Step 5: Bode Plot and Solution. (See

the Fig. 11.28 on next page) 567891

Ex. 11.4 For a particular 242 (s + 5)

unity feedback system, = G(s)
s (s+ 1) (s +5s + 121) 4
40
Sketch the Bode Plot. Find oor
Sol. : Step 1 :
and o.
G.M.. P.M. Comment on stability.
Arrarnge G(s) H(s) in time constant
form.
242 x5x
(1+) 10(1+7
G(s) H(s) Mag
0.
dB dB>

5 s dB/dec
sx(1+s)(121)(1+ ) s(1+ s) (1 + 0.041st1n in
dB/dec
dB/dec

S+
121 121 0

As Gls) H(S) incudes a

quadratic pole, comparing 1 decide E and -60 40 -20
it with
O,. These values give us proper
magnitude plot
correction to be
of it. applied at o, while sketching
Fig. 11.28
<br>

572
Bode 573
FCS Plots Bode Plots
FCS + 7.95 dB upward shift at
.Comparing exhibiting s =11 and will
1 continue till o as there is
factor.
no other
+2 o, s+
o
$ +5s+ 121 s2
: Phase Angle Plot
. Step 4
2 = 121 O,= 11 1o1+/2)
5 =
= 5 = 0.22 (jo) HGo)
and 2% n 2x11
G
jo(l+ jo) | 1+0.041 jo + jo
jo(1+jo) 1+0.041 jo-
121 121
discussed earlier in quadratic factor
to correction Table 11.2=
Referring 02 analysis,
on 11 1s +/.95 dB upwards, for
approximate correction required at As
j' =-1
precisely as
Or correction can be calculated Z10+j0 20+i
Correction = 20 Log3 V45 =- 20 Log 24 =- 20 L
Log (2 x 0.22)
2 G(j@) HGo)
Zjo 21 +jo 21+0.041 jo -
=+7.13 dB
121
Step 2: Factors. Z10+j0 =
0, 1+j= + tan -1 @
i Constant K = 10, 5
=
iü)
1
pole at the origin, 1/s -90 as 1 pole at origin.
1
1
= 1
jo
Simple pole T, =1 rad/sec.
i) 1+s T, 1
tan o,
1
27+j0
= 5
iv) Simple zero, 1 + rad/sec.
0.041 o
-1
1 =-tan
v) Quadratic pole, 03 =
0, = 11 rad/sec.
1+0.041l jo
(1+0.04151 121
12

:.Phase Angle Table

Step 3 : Magnitude Plot Analysis
Contribution due to K= 10 is 20 Log K = 20 dB.
1
+ tan-l 0 - tan
i)
5 0.041 o
ii) 1 pole at origin, magnitude plot is straight line of slope
- 20 dB/decade, pasing jo -tan

through intersection point of o=1and0 dB line. 121

line to
iii) Shift this intersection point on 20 Log K line and draw parallel 0.23 -94.7
-20 dB/dec. This line represents addition of K = 10 and 1/s. So starting slope D a---a
0.1
.90 +114 -57
- 20 dB/ decade. This will continue till o = 1. 236 -126.0
1
90 +11.3 -45 - 138
iv) At oc = 1, simple pole occurs so it contributes -20 dB/decade individualy
a
5 + 45 -78.6 -144
onwards -90 -149.8
hence resultant will have slope - 20 - 20 = - 40 dB/decade from '1' -828° 34.8
and will continue till o = 5. 90 +58
-67.0 -177.8
= individually and -90 +634 84.2 -
v) At o 5, simple zero occurs so it contributes + 20 dB/decade and +19.5 - 180 =-160.4 261.63
5 onwards
hence resultant will have slope - 40 + 20
will continue till os = 11
=-
20 dB/decade from 20 -90 +75,9 -87.13
180 -270
dB/dec 90 +90 -90 11.29)
vi) At 3 = 11,
quadratic pole with & = 0.22 occurs so it contributes 40 dB/decade obtain the solution.
(See the Fig.
individually and hence resultant will have - -
40 =-60
Stepp
5: Sketch the Bode Plot and
slope 20
<br>

574 575
Bode FCS Bode Plots
FCS Plots
a certain feedback system Gs) H(s) = 3(s+1)(s + 6)

11.5 For
Ex
2(s2+18s + 400)
Sketch the Bode plot and comment on GM. P.M. and stability
Arrange G(s) H(S)
Step 1 :1 in time constant form.
STABLE
Sol.:
sec

rad
dB G(s) H(s) = 3(s+1)(6) (1+s 6)
/
0045(1+ s1+26
dB SOLUTION
+11
12 gc.
+46
SYSTEM
s (400) 1+18
= S+
K= = 400 400
G.M. PM. 400
Log

Step 2 : Factors : i) Constant K= 0.045,

20
i)2 poles at the origin
, ,
Simple zero (1 +
s) T =1,o =1
ii) = 1
dB/decade
iv)
Simple zero, (1
+ ),T, 62=
6

1
,
100
Quadratic pole, s2 O3 =0n=400= 20
(1+0.045 s+ 400
-60

c3=11 Now compare s + 18 s+ 400 with s + 2& o, s +

o2
decade

G.M
=10.1 ..
o = 400, On = 20, and 2 o, =
18,=0.45
dB/
"pc corner frequency is 20 while as &
lts = 0.45, magnitude plot will
exhibit + 2 dB
deca 10
O
overshoot at os = 20 (Referring to correction table given in discussion of quadratic
dB/
pole ).
)
:
Step 3
i) K= 0.045 .:.its contribution is 20 Log K = 20 Log 0.045 =-27 dB.
straight line of slope - 40 dB/dec
dB/decade

P.M 1
2 poles at the origin so magnitude plot is
67891
Resultant

= 1
passing through intersection point of o and 0dB. This is starting slope of
345 -20 magnitude plot.
Shift intersection point of o = 1 and 0 dB line on 20 Log K line ie.
- 27dB
) -
as that point draw parallel to 40 dB/dec line.
downwards (
K<1)and from
till first factor becomes
Inis will represent addition of K/s.This wil! continue
dominant having least corner frequency i.e Q=l
occurs, contributing + 20 dB/decade individually hence
dB/dec
nt =1, simple zero
Tesultant will have slope - 40 + 20 =- 20 dB/dec. Hence '1' onwards slope
dB/dec dB/dec
dB angie of
0

dB/dec contributing up to next corner requency Oc2 =6.

-60 40 -20 esultant will be-20
V)
zero occurs contributing + 20 dB/dec individually
O2 6, another simple
=
= 6

making the slope of the resultant will become

- 20 20 dB/dec from + 0

=
20.
onwards i.e. line parallel to x-axis till next corner frequency o3
Fig. 11.29
<br>

576
FCS Bode 577
, pole occurs
contributing - 40 dB/dec individuallyPlot FCS Bode Plots
Bode plot and obtain
vi) At o3 20 quadratic
=

will becomes
0 - 40 40 dB/decade
from =- Step 5 :
Sketch the
the solution, see
the slope of resultant
hene the Fig,
no other factor BBut 20onwards as Wpc = the Oge is always 0, 11.30.
upto o as there is at Note that less than ooe for any
and will continue 0g =
20 system G.M. =400 dB and systerm is absolutely value of gain. Hence
+ 2 dB. such
show overshoot of for
stable.
Plot.
Step 4 : Phase Angle
0.045 (1 + 0.045 dB ABSOLUTELY

(1 jo) +j) (1+jo)(l+ K=-27

= rad/sec 000
Gjo) H (o) (jo)2
(Go)' 1+0.045 jo+ (jo) (1+0.045
j0 Log
decade

400
SOLUTION

0.19
40 +14
SYSTEM

20 =
20.045 21+joZ1+j dB/
ge P.M.
6
2 G (Go) H0) 40
ZGo)* 1+ 0.045 jo 400

2 0.045+j0 = 0, 1+jo = + tar. o, 21+j=

+
tan -1 0
6
Qc23=20
1 = as 2 poles at origin
-2x90 =-180
o)?
decade

0.045 o
=-tan l
1

dB/

1
1+.045jo 400
20
400 dB7decade

.:.
Phase Angle Table
1
tan -1
Resultant
+
+ tan -1 0 OR
6 0.045 o
- tan -1 -40

1
400 Pgc0.19

0.1 -180 +57 |+ +0.95 0.25 -1736

1
-180 +45 +9.46 - 258 1281
0.1
6 + 8053 -70.9
-180 +45 -1652 +20
dB/dec
10 -180 +84.28 -67.6 dB/dec
dB/dec
dB
+59 30.96
0
20 +8713 -10957
-180 +73.3 -90 40 -20
+23.19 - 180
50 -164.8,
-180 +88.85 +83.15 =-156.8
100 -180 +89.42 + 8656 = -1734 Fig. 11.30
+10.61- 180 -169.38 Ex. 11.6 3 (s + 1).(s + 700)
-180 +90 -180 A
unity feedback system has G(s) =

+90 -180 (s + 18 s + 400)

s²
As two zeros are always contributing more than a Draw the Bode plot and comment on stability.
cannot cross -180* but at the quadratic pole Pras which was
end will run parallel to it. NOre that in this problem a
zero is shifted at s
=-700 from s =-6
<br>

578
Bode 579
FCS Plots,
less significant hence how
effect becomes FCS Bode Plots
present in Ex 11,5. So its we
have seen adding zero it affects th Angle Plot
shown in this problem. Earlier Step 4 : Phase
stability is zero is made less improves
now here we will see if dominant significant the 525(1 +jo)| 1+ jo
stability, 525(1 + jo) 1+
affects the stability
adversely. how
G(jo) Hý o) 700 jo
constant form. 700
G(s) H(s) in time
Sol. : Step 1: Arrange (o)1+0.045jo + (jo) 1+0.045jo -
3(1+ s) (700)1+S 5.25(1 + s) 1+ 400
400
700
G(s) H(s)
700) 25.25+ j0 4
s 1
18 s
1+0.045 s =
1+joL1+ jo
(400)| 1+ S+ +
G (@) H(Go) 700
(400) 400
400
ZG0) Z 1+0.045jo
Factors are
:
:
Step 2 400
L5.25+j0 =
0
i) Constant K = 5.25 21+jo=+ tan o
21+j = + tan
700 700
Two poles at origin, 1/s² 1
0.045o
1
1 =
-180,2 2 =-tan
-1

ii) Simple zero, 1 +s,T, =1, 1+0.045jo 1

400 400
= 20,=0.45. ..Phase Angle Table
=

iv) Quadratic Pole O2 0n

1+0.045 s+
2
S + tan o 0
+ tan -l R

400
(jo)? 700 0.045 o
-tan
v) Simple zero, 1 +
T3
1
Qc3 = 1-700 2
700 1+
400
Step 3
:
Magnitude Plot
0. -180 +5.71 +0.008 -0.257 -174.5
i) Contribution of K is 20 Log K = 20 Log 5.25 = 14 dB
40 dB/decade passing 1
-180 +45 +0.08 - 258 -137.5
ii) Magnitude plot for 1/s is a straight line of slope
through intersection point of o = 1 and 0 dB. 5 - 180° +78.69 +0.409 1349° -114.3
on 20 Log
Shift this point of intersection of o = 1 and 0 dB upwards
K
line and
iü) + 87.13 +163 90 -181.24
from that point draw a line parallel to - 40 dB/decade. This represents
resultant 20 -180
of K/sThis continues till first corner frequency occurs i.e. =
+ 89.42 +10.61 -180 =-1693 - 251.7
1. o1 100 180 +8.13

iv) At o1 simple zero occurs which contributes according- to

= 1,
+
20 dB/ decade
700 +1.47 - 180 =1785 - 223.5
- + trom -180 +89.91 +45
individually and hence resultant slope becomes 40 20 20 dB/decade
n = - 180 l80 180
O1 =1 onwards till o
=
20. +90 +90
with
v) At o2 = 20, quadratic pole occurs which contributes 40 dB/decade Oep 5: Sketch the Bode plot and obtain the solution,
as shown in the Fig. 11.31
overshoot of + 2 dB at o,
= 20 individually and hence resultant slope becouia
zero system insignificant compared to dominant zero
20 - 40 =- 60 dB/decade from o = 20 onwards till Note: As in this is made
= 700. os
individualy t rn system discussed in Ex 11.5, it is clear that G.M. has
reduced considerably
vi) At @s = ,
700 simple zero occurs which contributes + 20 dB/dec onwards system becomes conditionally stable due
fo dB to + 10 dB. ie. absolutely stable
and hence resultant slope becomes - 60 + 20 40. dB/dec from
700
=- citect ofzero less sienificant i.e. relative stability of system has reduced.
becomes
till o0. of system reduces while when
zero are poles are made dominant relative stability
added or made dominant relative stability systemimproves.
of
<br>

580
Bode 581
FCS Plots FCS Bode Plots
a unity feedback system G(s): K
11.7 For s (s +2) (s +
Ex. 10)
Determine marginal value of 'K for which system
o will be marginally
G(s) H(s)
Log
1: Arrange in time constant stable.

rad/sec
STABLE

: Ste
Sol.
G(s) H(s) = K
form

dB SOLUTION

SYSTEM
s2)(1+a0) 0
14
=5
K= K
Log
Wgc decade:
1100
20 K

20 s(1+)(1+
Resultant
dB/
K = K
-40 where
decade
20

dB/ To avoid the confusion, assume new constant K

and calculate Kmarginal by
100
567891 20
60 Bode plot and then corresponding Kmarginal can be determined by the above relation.
:
Step 2 : Factors
4
Oc220

i) K Unknown

) 1/s, 1
pole at origin
1 1
i) Simple pole,
s'
1+
2
M. 1 =
iv) Simple pole ,T, @g = 10
10
1+
10
dB/decade

Step 3 : Now for such problems sketch the magnitude plot without considering
effect of constant. So

line for 1/s ie. having slope - 20 dB/decade passing through

Resulta

Sketch a
40
intersection point of o = 1 and 0 dB line. Now as effect of K i.e. 20 Log K is
unknown the procedure to shift intersection point of o = 1 and 0 dB on 20Log K
and then drawing parallel is not required in such problems. So continue the line
of slope - 20 dB/decade till og = 2.
0.1 as - 20 dB/decade individually
ht Oc1 = 2, simple pole occurs which contributes
dB/dec
dB/dec
- - = - 40 dB/decade from o- = 2
alla hence resultant slope will becomes 20 20
dB
0 onwards till next corner frequency O2=10.
i) At this occurs which contributes -20 dB/decade
40 -20
02
=
10 another simple pole
individually
and hence resultant slope will become
- 40 -
20 60 dB/ decade =-
Oc2 = till o->o as there is no other factor present.
onwards
wt 10

Fig. 11.31
<br>

582
FCS Bode 583
20 Log K' dB at Ple FCS Bode Plots
effect of K will be additon of all
Note Now:
drawn, upwards or downwards without
frequencies = - tan -1 o
magnitude plot 14 J0
shifting the on whether K' is greater or hange i and 10
Log K depending lessthanl.
inay
slope, by distance of 20 10

contribute to phase angle because Table

Now K' is not going to 2K
+j0 = tan 10 .Phase Angle
Due to this phase angle table 1/j0
for any value
of K
selected. constructed --tan
K selected - tan
information of
K
will be true phase angle table and for phase 10
angle
willremain unaffected. pl 0.1
90 -285
as K 057 -93.43
magnitude plot shifts, ogc will change changes but as
– 90
As
unchanged for any K'.
phase
plot - 2656
unchanged ope Will remain s 1
-57 -12226
we will shift the magnitude -90
So to decide proper value of K plot upwards 2 -45 -11.3 -146.3
ge Will get adjusted in such a manner
downwards such that that it will meet ox - 90°
TThis is because opc is not dependent on K the 5 -6819 2656 -18475
required specification. Whatever
shit
common for all frequencies will be 90
-78.69
required which will be nothing but the -45 - 231.69
contribution of K
i.e. 20 Log
K
dB. So equating shift required in dB to 20
Log K we 20 -90 - 84.28 - 23771
can determine corresponding K which will satisfy required specification. The -63.43
shift must be treated
as
positive while downwards shift required if any must -90 90 -90 -
270
be

treated as negative while equating to 20 Log K.

Sten 5 : Sketch the Bode plot and measure the shift required in magnitude
In this problem, specification given is system should be marginally stable plot
a Hll oor = 0pe: (See the Fig. 11.32 on next page.)
i.e. G.M. = P.M. = 0 i.e. gc = Ope. But wpe is independent of K so we will
Now ope = 4.7. To have @ge = Ope the resultant magnitude plot including K must
magnitude plot without K upwards or downwards till oge = Ope:
intersect 0 dB at o = 4.7 i.e. at point B shown on plot. So by proper selection of K we
Equating this shift in 'dB' to 20 Log K, K can be determined and hence Ken can achieve
upward shift which is common at all frequencies to achieve oge= @pe- This
be determined. shift is effect of K' and shifted plot is resultant plot exactly parallel to plot without K'
Step 4 : Phase Angle Plot As this shift A B = 22 dB upwards is +ve
Let us prepare phase angle table where K
is not going to contribute at
all..
20 Log K = 22
= K K = 12.58
Gjo) Hfo)
Now K
2 K
20
an = 20x K

ZGjo) Hjo) K=
ZjoZ(1+)(1+ba =
20x12.58
= 251.78
=
2K+j0 0,4=-90,1 pole at the
origin.pd e
marginal from Bode Plot= 251.78
jo
1 e
value can be checked by Routh's array. By Routh's array
graphically
it is 240. So
=-tan
K from Bode plot we got Kmarginal which very close is
to the actual
1+ marginal
2

hasat
<br>

584 585
Bode FCS Bode Plots
FCS Plots
K(1+
upwards

lin encies upwards

without
dR freguenc

LOGK
P.M. 430
retermine K to get
on0 Shited

he
plot
B
dB
Factors are
condition
must
all
Mag
2
20 Step 2:
at A = unknown
hence K
to thie conmon
Shift
, pole at origin
A
plot
1
stable
magnitude
point B.
) 1/s
to zero, 1+ S
67891

marginal ii)
Simple 2 I =1/2, @e1 = 2
1
Simple pole T, = 1/4, = 4
Shifted

iv) 1+s'4' O
he
1
Simple pole T3 = 1/10, o = 10.
V) 1+ s/10'
K
Step 3
:
Magnitude Plot. Sketch without
i)
Sketch line of slope - 20 dB/decade for 1 pole at origin will continue
till o1 = 2.
Passing through intersection point of o = 1 & 0 dB line.
At o-1 = 2, simple zero occurs contributing + 20 dB/decade so resultant will
have slope - 20 + 20 = 0 dB from o1 =2 onwards till o = 4.
At o2 = 4, simple pole occurs contributing - 20 dB/decade individually so
,
i)
dB| resultant will have slope 0 - 20 = - 20 dB/decade from o = 4 onwards till
Oc3=10
-20
iv) At o = 10, simple pole occurs, contributing - 20 dB/decade individually so
resultant will have slope - 20 - 20 = 40 dB/decade from oa = 10 onwards till
) 0. Then shift this upwards or downwards to adjust oge such that
P.M = +30. This shift will be 20 Log K

+60 - -120 :
dB/dec dB/dec dB/der Step 4 Phase Angle Table
dB -
Phase angle
0

-60 40 -20 G{jo) Hjo)

jo(1+ j jo
10
Fig. 11.32
K (s + 2) LK+j0 Z1+jo2
Find K to get P.M. =+
Ex. 11.8 A
system has Gs) H(s) = 30 =
ZGjo) Hjo)
Sol. :
Step 1 : Arrange G(s) H(s) in standard time constant
s (s + 4) (s+ 10)
Zjo 21+
4
1+ jo10
form
K2)(1+ K A=K
<K +j0 = 0°, 1+jo/2= + tan, =90l pole at origin.
G(s) H(s) = Let jo
20
20
s(4)
<br>

586
FCS Bode 587
1 Ploa
1 =-tan1 0 FCS Bode Plots
. 10 o
jo 14
1+
frequency

o
4 =A208 angie shift
Log
:
Table. s91
teo
Phase Angle phase
this

1 + tan -l
o -tan -l 0
67 PM
upwards
A,
measure

()
2 4 10
5

4
re point
shif which
on
K' point at
-90 + 2.86 -1.43 -0.,57° /4/E/E/4/4/E/ the of and
0.1 -89.14 DistanceCDis
for dB effect
PM.
naving
D
22 C-D=+29
is
+ 26.56 -14.03 -5.7° measure
to
1 -90 - 83.17
s91
= addino 'A
Point C
100

2
- 90 + 45 - 26.56 -11.3 - 82.86 67
,

achieve
point

+63.43° shift 29
4
90 -45 21.8
-93.37°
4 K A reauired

It

10 -90 +78.69 68.19 - 45°

-124.5

20 -90 +84.28 -78.69 63.43 -147.84 567891

-a
-90 +90 - 90 -90 -180
4

Step 5 : Sketch Bode Plot. Now we want P.M. 430 i.e. point of
intersection 3

with resultant phase angle plot must be 30 above - 180 line. (Refer the Fig, 11.3) 2

P.M. = 180° + G(j@) o

H{o) @ge
aloe 7891 decade

30° = 180° + 6

G(jo) Hj©) o= Ogc 5

4
Resultant
7

dB
.2GGw) HG0) lo = oge = -150° 3
-20
2

So shift magnitude plot such that ogc Will be that for which

LGjo) HGo) = -150 represented by point A in the Fig. 11.33. Miag

dB/dec dB/dec

point A' represent point on phase angle plot for which

dB/dec
This shift is 20 Log K. So
2GGo) H(o) = - 150 and corresponding o = 22 must be oge to get P.M. = +30
-60 40 -20
For achieving gc = 22, point C should be at point D i.e. effect of K
required 5

shift + 29 dB upwards. Fig. 11.33

20 Log K
K= +29 Ex. 11,9 A system has G(s) H(s) =
s (1 +s) (1 +0.1s) (1+0.01s)
Determine the value of K
K = 28.183 as K = K/20
so that i) G.M. =+ 10 dB i) PM. =+ 25
K= 563.676 to get P.M. = +30 Sol :: Step 1 : G(s) H(s) constant form.
is already in the time
Step at origin
2: Factors are: i) 1/s i.e. pole 1

1
ii) 1 = 1 simple pole
1+s
T=1,@g
<br>

588
FCS Bode 589
Plots Bode Plots
iii)
1
simple pole,
,
T, = 0.1 oc
=
1=
T 0.1
10 FCS
value of K, this table rermains unchanged
For any Chetch the Bode and hence the value @pc
of
1+0.1s Plot and adjust the shift
1
simple pole, T3 =0.01 O3 = 100 , 11 Step 5
specifications one by one. (See the Fig. 11.34)
to meet the required
iv) 0.01
1+0.01s
Plot.
Step 3 Magnitude
: upwards
dB/decade

K.
Sketch this without effect of
slope - 20 dB/decade
Log

For 1/s,
1
pole at origin straight line of passing dB
i) 0 dB through dB
intersection point of
o =
1and and will continue till +10
- 20 dB/decade individually.
8
-80
o = 1, simple pole occurs which contributes by
At =
ii)
resultant will have slope
- 20- 20 = - 40 dB/dec from shifted WithoutK

oi =1 onwardsSo
G.M.

HN dB
O = 10. os0.=-10

For
occurs which contributes 20 dB/decade be
ii) At O = 10, simple pole
=- individually dB/decade
-
will have slope 40
-
20 60 dB/dec from o
to H
SO resultant =
10 onwards H
67891
dB is
C

Oc3 = 100. +10

so
iv) At occurs
o3 = 100, simple- pole= 80 which contributes
- 20 dB/decade SO resultant 5
=
H(jo)l
D

will have slope - 60 20

34
dB/dec from o3 = 100 onwards G.M.
Point
80
tilloo. Gjo)
This plot shifted to meet the required given specifications.
is to be Once
to get K 2 For i.e.
= +25
for G. M= +10 dB and secondly to get K for P.M.
Step 4 : Phase Angle Plot
= K
67891
decade
dB
+10 e
GGo) H(jo) =
o.1
jo(1+jo) 0.1jo) (1 + 0.01 jo)
(1 + 5 G.M.
4 dB/ pc
2 G(jo) = ZK+j0 +25 rad/sec
Hfo) 40
2 jo 21 + jo Z1+ 0.1 jo. Z1+0.01 jo for =
P.M.
ge 1.7
L K+j0 = 0
-90° as 1
pole at origin,Z
1
=- tana +25
jo 1+jo 567891
dB/decade

1+0.1 jo
=- tan0.1 o, 2 1

1+0.01 jo
=- tan0.01 o

4
Resultant

:.Phase Angle Table is 23

20
1/jo - tan -1 tan 0.1 o- tan - 0.01 o

0.1 96.32 0.1

-90 -5.7 -0.57 0.057
1
-90 - 45 -5.7° -0.57 -141.27
dB/dec dB/dec
+20
dB/dec
Phase
2
-90 -165,87 agie
-63.43 -11.3 -1.14
5 - 90 -78.69 - 26.56 -2.86
-198.11 -60 40 -20

10 - 90 -84.28°
-45 -5.7
- 224.98 Fig. 11.34
-284.1 at point C
50 -90 -88.85 -78.69 - 26.56 = 3.1 rad/sec and is independent of K. It intersects the plot
ipc
which 18 dB down w.r.t dB line. So under such 0

100 -90 89.42 -308.7 is with

out K. Point 'C' is
-84.28 -45 condition G.M. we want it to be
+
10 dB.
is 18 dB but
+

-90 -360
-90 90 -90
<br>

590 Bode 591

Plots Bode Plots
FCS 10 dB. i.e. = + FCS product of
so G.M. all of
them we can construct
G(jo) H((o)
10 dB bethatshifted upwards by 8 dBpoint Cshould
=o=- Taking the the transfer function of
the
iLe. hence should to point system
w.r.t OdB line =
be 10 dB down = + 10 dB, shift
CD
= + 8 dB
20 Log K. Upward shift positive.
Das given
Advantages of Bode Plots
G.M.
shown, to get dB,K= 2.51. 11.12
G.M. = + 10 both low and high frequency characteristics
o t 180 It shows transfer function in
For = of

P.M. = 2 Gfjo) HGo) lo 1)

single diagram.
Now
+
25 = Gjo) H{0) lo =
e + 180° The plots
can be easily constructed using some valid approximations.
of system can be studied I
2)
Relative stability by calculating G.M. and P.M. from the
P.M. = + 25,
So to get
= 25°- 180° = - 155° 3)
Bode Plot.
2 Gfo) HG0) lo =o The various
other frequency domain specifications ike cutoff frequency,
determined
on phase angle plot having
angle -155 i.e. point shown
A
4 bandwidth etc. can be
So find point and
ge = 1.7 rad/sec. constructing complicated polar and Nyquist plots can be easily obtained
frequency o must be Data for
corresponding
from Bode Plot.
at point and it should be at E
K
B
5)
plot without
Now this intersects magnitude to
function of system can be obtained from Bode plot.
get 1.7 as ape and P.M. + 25
6) .Transfer
indicates how system should be compensated to get the desired response.
dB and hence in this problem for P.M.
=
So shift B
is again
E
+ 8 +25 also 7)
The value of system gain K can be designed for required specifications of G.M.
+8 dB .:K= 2.51 8)
Plot.
20 Log K=
G.M. = + 10 dB and and P.M. from Bode
this problem single value of K 2.51 satisfies both
=
In
9
Without the knowledge of the transfer function the Bode Plot of stable open loop
P.M. = +25. svstem can be obtained experimentally.
rare case. For othor 11.10 Determine the transfer function, of a system whose asymptotic gain plot is given
Note that this may not be the case for all problems. This is
Ey

G.M. and P.M. specifications

problems two separate K values must be calculated for below.

11.11 Calculation of Transfer Function from Magnitude Plot

O=1 =1000
case, magnitude plot will be known to us. To decide the transfer function
In this
observe the following points. Starting slope of magnitude plot represents poles or 0 dB
zeros at the origin.
If starting slope - 20 dB/dec there is 1 pole at origin -20 dB/sec +20 dB/sec

- 40 dB/dec there are 2 poles at origin. -20 dB

dB/dec there is no pole at origin.
0
+ 20
dB/dec there is 1 zero at origin and say on.
Fig. 11.35
Then observe the shift in magnitude plot at o = 1 which represents 20 Log rom K
or zero at the origin. The first change
which we can decide value of K. Then this slope changes at each corner frequency s0
DOl. :
Starting slope is 0 dB. So there is no pole
identify the corner frequerncies. The change in slope indicates the respective factor with
n slope occurs at o= 1

corresponding cormner frequencies. If change in slope is - 20 dB/dec i.e. - 20 to - 40 0r

1.e. = 1
0to -
=
-20 - 0 =-
1.
20 and so on then the factor is Change in slope = 20 dB/dec, there is simple pole with @1
simple pole.
If change in slope is + 20 dB/dec i.e.0 to + 20 or
factor is simple zero.The respective
-40 to - 20 and so on te i.e T
= 11
frequency
time constant is reciprocal of the corner
ie. as o=-then T= and hence we can determine the different factors presenl. 1 1
Factor
1+T; s 1+s
<br>

592
Bode 593
FCS Plots
so 20 Log K = 0 dB K=1 FCS Bode Plots
at o = 1
is 0 dB 0 AB
So no pole
The shift Starting slope or zero at
occurs at @, which is unknown but we can
: origin. At each corner frequency
The further change in slope 0 - 20)
= +
20 dB//dec. i.e. determine Sol
changes. slope
(-
- dB and magnitude there
occurs 1,,
change in slope is is simple zero change in slope is at o =
O1 =
1.
it. The
corresponding to o, is 20 at o =
1 is First
magnitude - second slope - initial slope
Now line
ie. o) 0 dB,S slope
20 dB/decade. is
- 20 dB and slope of the is Change in =
is change of
there is decade zero with o = 1
20-0 = 20 dB/ dec.
simple
away from o= 1 factor is
= 1
10 T =1
i.e. i.e.
T 10.1 is (1 + s) i.e. (1 + s)
Factor T
. +T, s) = (1 + 0.1 s) as simple zero. change in slope is at o1
Factor is (1
+
Now next
Next change in slope is at @3 So 03 is O3. Change is 20 -0 = +
20 AB/dec. os Change in slope 0 - 20 =-20 dB/dec so factor is simple pole. Now
So
there is simple zero with
corner frequency o3 knowVn.
at o =
1000 is 0) dB i.e. change 01 is not line.
Now magnitude at o3
= 20 dB and magnitude Let us write equation for that
is +
20 dB// decade. i.e. of
is there from o3 to 1000 and slope 03 and 1000are = m x
+
20 dB +C i.e. magnitude in dB = m Logo+C
decade away. m = + 20 dB/dec
Oc3
= 100
i.e. = 1, magnitude in dB =0 dB
Now at
i.e.
, 1_0.01 Substituting
= 20 Log 1 +C

+ 0.01 s).
C= 0

:. Factor is (1 +T3 s)
(1
- = -
20 dB/doe :*
.. Equation is mag. in dB = 20 Log o
= 1000 which is 0 dB 20
Next change in slope occurs at 0 = 01
= 1000 Now at
there is simple pole with o4
Magnitude in dB = 15 dB given
10.001 .:.
Substituting 15 = 20 Log o1
1 1 @1 = 5.623 rad/sec =
02:
Factor is 1+0.001s 1 1
1+T¢s = = 0.177
T, 5.623
:. Transfer Function is product of all factors,
= (1+0.1s) (1+
0.01s) .. Factor is 1 1
G(s) H(s)
(1+ s) (1+ 0.001s) 1+T,s 1+0.177s
Ex.
11.11 Determine the tansfer function of system whose
corner plot is shown below Next change in slope is at @2

*. Oc3 = - - 0 = - 20 dB/dec
O. change is 20
:. Factor is
-20 dB/dec simple pole.
+20 dB/dec - 20 dB/dec.
lo find o, write eguation for that line having slope
+15 dB
mag in dB = m Log o + C
o +

mag in dB =-20 Log C.

At o = 1000, mag in dB =0
+
0 dB Substituting C
02 1000 0 =-20 Log 1000
C = + 60 dB
Fig. 11.36 *
Equation is, mag in dB = - 20 Log 0+ o0
<br>

594
Bode 595
FCS Plcta
+ FCS Bode Plots
mag.in dB
15
=-20 Log @2
60e th
o,
At
- 20 Log 02 =-45
K= 10.
corner frequency slope changes
177.82 = at each
3
0 = First change
1-0.0562
Then
0.5 change is
- 60)
-(-40) =-20 so factor is simple in slope is at
o= 0.5,
= pole.
T3

1
W1
T =
12
Factor i 1+0.00562 s 1 1

The last change in slope

is at
. Factor is 1+Ts (1+2s)
= 1000= 04 at o =
6,
zero Next change in slope
Change is 0-(- 20) = + 20 dB/dec so factor is simple
Qo4 = 1000
- 40 - 60 ) =+ 20 dB/dec.
1
= 0.001 Change is So factor is simple zero.
= 1
O

T =1/6
(1+T, s) = (1 + 0.001s)
.Factor
..The total transfer function is product of all of them,
(1+s) (1+ 0.001 s) : Factor is (1+ T, s)
G(s) Hs) =
(1+0.177 s) (1 0.00562 s)
+
1s at o = 20 i.e. o,a = 20.
Next change in slope
Ex. 11.12 Determine the transfer function from the magnitude plot shown
below. snge is - 60- (- 40) =-20 dB/dec, so factor is simple pole
T3 = 11 20
Oc3
Mg in dB
1 1

40 dB/dec Factor
1+T3s 1+
S

20
No further change in slope
20 dB
-60 dB/dec
. Overall transfer function is product of all factors
20 10(1+)
0 dB
0.5 G(s) H(s) =
s²(1+2s)(1+

Ex. 11.13 For the plot shown below determine the transfer function.
-40dB/dec -60 dB/dec
dB A
6 dE/octave -6 dB/Octave
Fig. 11.37 32 dB
dB/Octave
Sol. : The starting slope is - 40 dB/dec
:. T.F. must have 2 poles at origin
ie. term 1/s. 12
Now this line should pass through intersection point of o = 1 and 0 dB but ar
Upward Log w

K.
5.0
o=1 it is lifted upwards by 20 dB shown dotted which is due to effect of 0.5
shift of 20 dB at o= 1 means 20 Log K = 20.
Fig. 11.38
<br>

596
Bode 597
FCS FCS
Ple Bode Plots
is to beconverted to dB/decade. Initial slopeis
Sol. : The slope dB/octave G(s) H(s)
79.8 s?
are two zeros at the origin. This is because (1+2s) (1+s) (1+ 0.2s)
+12 dB/octave, hence there
= 20 dB/decade
6 dB/octave circuit shown draw Bode plot for
12 dB/octave
= 40 dB/decade 11.14 For the voltage ratio V,(s)
Ex.
Now the equation of line after
o 0.5 is say V,(s)

20 log o+ C
M=
= 1, M= + 32 dB shown
at
32 = 200 Log 1+ C
V{(s) 1F=
C= 32 V,s)
o = 0.5,
M= 20 Log 0.5 + 32 = +26 dB
At
Now o = 0.5, M = 26 dB is also on the initial line whose equation is
o
M= +40 log +C1 Fig. 11.39
= 40
log 0.5 +C,
Ato =0.5, +26
Sol. : Let us find the transfer function first.
Cf = 38.0412 dB I = I1 +I2
=
Now
for K =1.
V-=s(V;-)
at o 1
Now this line must have M=0 dB I =
But at o =1, M = 40 log 1+38.0412 b1
i.e. M = 38.0412 dB
This is due to contribution of system gain constant K.
20 Log K = 38.0412

K = 79.8
At o, = 0.5, slope changed by -20, there is simple pole. Vi(s) Vy(s)

Factor =
(1+T, s)
1
where T,
=l:205
1
Fig. 11.39(a)
(1+2s)
At @ = 1, slope charnged by -20, there is simple pole. S
-u-1,)
S
1
1 - y-y,
Factor= 1+ T, s
where T, =
1
1.
= 1 I,
=Y2 1

1
V = 12s
1+5
At o, = 5, slope
further changed by -20, there is simple pole. The corresponding signal flow
graph is
Factor= where T3 = 1.15 = 0.2
Now T, =
1+T3 S
1
@
sxxl1
s 2s 2s
Fig. 11.39(b)
71+0.2 s) L=xK-s = - 1,
Hence the transfer function is
<br>

598
FCS Bode FCS
599
Plots
- table is Bode Piots
L, = phase arngle
L, = 1x-=- 2s 0.1
1 0.5
= -1x 1_ two nontouching loops. o 2 10
LL3 2s 2s' -tan -5.7 264 -45 63.4 84.2
A =
1-[Lj + Lz
+Ll+ [LjL3] 11.15 For a
unity feedback system,
= 1.141=
1+1+-+ 2+4l= 2+2 Ex.
G(s) = 10
s 2s 2s ss s(s+ 1)(s+ 4)
2(s +1) Obtain analytically, the gain margin
and phase margin.
S
Maathematically the gain margin is,
Aj =
1 Sol. :
1
-x1 G.M. = 20 Log 1

V,(s) 1
||Gio) Hjo)lo=opc
2(s +1) 4(s +1)
V,() A
oz02c ZG(ja0) Hjo) =
S -l60 1e. imaginary part of G(jo) Hio) at
o=@ is
V2 ja) 1 025 zero. To find @pcr rationalise G(jo) Hjo) and equate imaginary part
to zero.
10 10(-jo)(1- (4-jo)
jo)
V, Ga) 4(1+ jo) (1+ jo) G(jo) =
+
jo(1 jo)(4 + jo) (io)(-jo)(1+ jo) (1-jo) (4 +ja) (4 - j0)
K= 0.25, 20 Log K= -12.04 dB
-10 jo[4-5jo-o*] -50a 10ja(4-o)
is pole with o = 1
o²(1+o')(16+ o) o(1+o)(16+o') o'(+u?) (16+a)
(1+jo)
At o= @pc the imaginary part is zero,
2 3
456 7891 3 4 5 67891
2 3 4 567891 2

4-0 0
2 = 4
+40 dB
=
O 2 rad/sec
+20 dB This is nothing but Ope. Substituting in real part,
0 dB
-12.04 dB |Gjo) H{jo)
la=@pe
50
5x20
-50-0.5
-12 (1+4)(6+4)
-20 dB -20 dB/ decade
= dB
GM. 20
Log6
40 dB
oge first because P.M. is given by,
lo find phase margin, let us calculate
P.M. = 180°+ZG(jo) lo= oge
-30° 1 absolute
At OgerG(jo) =0 dB i.e.
0=
-60 10 = 1
ljo|| 1+ jo|| 4 + jo|
-90 10
= 1

0.1 1 100 L0g

o-Vl+o16 + o²
10
=1 10 = o+o²-l6+o?
Fig. 11.40
<br>

600 601
Bode
FCS Plots FCS Bode Plots
(1 +o')- (16
+ o?)
this we can prepare the resultant slope table as,
Square, 100 =o' From
= 100
.o[16+170 +o] Range of o Resultant slope
error method,
By trial and LHS. = 133.45 Start < 0< 5 + 40
1.5 dB/ dec
0= LH.S. = 80.45
1.3
0= 5<0< 50 + 20 dB/ dec
= 1.45 LH.S. = 118.08
0 50 < o< 0
= 1.385 LH.S. = 100.3 100 0dB/dec
0
Ogc
= 1.385 rad/sec
G(jo),
2G(jo) = -90°-tano-tan ) Replace s by jo to obtain
G(j)
K(jo?
=-90°-54.17°-19.09° -163.26°
=
(1+j0.20) (1+j0.020)
P.M. = 180°+ZG(j0)l o= Mge
= 180°-163.26° = +16730 2 zerOS at - tan 0.2 o - tan 0.02 o
in nature. origin R
As both G.M. and P.M. are positive, system is stable
for the following transfer function and determine gain 0.5 +180° -5.71° + 173..7
Ex. 11.16 Sketch Bode plot
:
K for -57°
crossover frequency oge to be 5 rad/sec (Dec-98) +180° -45° -5.719 + 129.29o
Ks?
G(s) = 10 +180° -63.43° -11.3° +105.271
(1+0.2s) (1+0.02s)
50 +180° 84.28° -45° +50.72°
Sol. :
The give transfer function is
Ks2 100 +180° 88.85° 63.43° +27.72°
G(s) =
(1+0.2s) (1+0.02s)
The Bode plot is shown in Fig. 11.41. (See Fig. on next page)
Let us obtain Bode Plot without value of K. From the Fig. 11.41 to achieve ogc = 5, the point A must be at B i.e. shift required
The various factors are is A to B downwards. This must be the contribution by K downwards.

i) s ie. two zeros at origin, so magnitude plot is straight line of

slope +
20 log K = - 28 dB
dB/decade, passing through intersection of o = 1 and 0 dB, line. log K = -14
ii) Simple pole, 1/1 + 0.2s K = 0.0398 0.04
The corner frequency of this pole is. bx. 11.17 A unity feedback control system has open loop transfer function as
100
G(s)
0.2 s(1+0.1s)\(+0.29)
So magnitude plot is straight line of slope - 20 dB/decade for o > 5 and U db u Sketch Bode plot and determine from it.
for o< 5.
(i) Gain cross over frequency
ii) Simple pole, 1/1 + 0.2s (ii) Phase cross over frequency
The corner frequency of this pole is. (iii) Gain margin
1
(iv) Phase margin
= 50 (v) Closed loop stability of the systemt. (May-2000)
0.2
0 db
So magnitude plot is straight o> 50 and
line of slope 20 dB/decade for
line for o< 50.
<br>

602 603
Bode
FCS Plot FCS Bode Plots

Sol :. O.L.T.F. is
The
G(s) = 100
s(1+0.2 s)(1+0.1
s)
=
Assume H(s) 1 and G(s) H(s) is
Step 1: in time constant
form.
Factors are,
Step 2: = 100
Constant K
1) dB
20 Log K= +40
One pole
atorigin : Straight line of slope - 20 dB/decade
(0 passing through
2)
intersertion point of
o
=l and dB
:
is a simple pole
3) 1+0.2 s
= 0.2
T;
1

=5 rad/sec
rThe straight line of slope - 20 dB/ decade 0 for o>5.
:
is a simple pole
4
1+0.1s
T, = 0.1
gc
= 10 rad/sec
get
T,
to
The straight line of slope 20 dB/decades for o>10
The resultant slope table is

Range of frequency Resultant slope

Starting slope 20 dB/dec

0< o <5 20 dB/dec

5 <o 10 40 dB/dec
10 <o <0 60 dB/dec

dB/ :
Step 3 Phase angle table
40 100
G
(jo) H (jo) =
ja(1+ 02jo)1+01jo)
+

tan 0.2
1
=- tan 0.1 o
jo
=-90o
1+0.2 jo
01+0.ljo

Fig. 11.41
<br>

604 Bode 605

FCS Plts FCS Bode Plots
()
1
- tan 0.1 o tan 0.2 o
OR

1 -90° - 5.71°
- 26.56°
11. 3o - 107°
5 90°
Logk
- 450 - 161.5°
-90° 38.65°
-58° 186.6°
20 10 -90°
1/s
-45° 63.43° - 198.4°
For
50 -90° - 78.69°
-84.28° - 252.9o
-90° 90°
-90° 270°

Step 4 : Bode plot is shown on semi log paper

in the Fig. 11.42. (See Fig. on
previous page.)
Step 5: From the Bode plot,
1)i Oc =18 rad/sec
Ope = 7.6 rad/sec
ii)

ii) G.M. = - 20 dB
iv) P.M = - 48°
radis
The system is unstable.
v)
Ex. 11.18 For a certain control system the asymptotic magnitude plot is shoun in the
8

Fig. i1.43. Draw the phase angle plot. Also determine the stability of the system.
(Dec.-2000)
rad/sec

7.6 dE 0 dB/dec
-20 dB/dec
60

40 -40 dB/dec

400
8
7
0. 4 40
6
5
-60 dB/dec
4

Fig. 11.43
:
Sol. The
starting slope is zero Hence there is no pole at the origin.
1, the plot is shifted 60 dB up from 0 dB line which indicates contribution
by K.tO=

20 log K = 60
1000
Fig. 11.42 K=
<br>

606 Bode
FCS
- 20 Plts 607
change in slope is dB/dec.
First corner frequency is o1= 4. The So
FCS
1 there Bode Plots
= =0.25. is
simple pole with T,
1
1

Factor 1
is, (1+0.25 s)
(1+T;9)
The change in slope is -
=
0c2 40. 20
aB/dec.
Second corner frequency is Sothere
1
with T;
= =0.025
is simple pole 40
OC2
1

Factor 2 is,
(1+T,s) (1+0.025s)
- 20
Third corner frequency is oc3
= 400. The change in slope is dB/der
1
=
So there is simple pole with T3 = 0.0025
Oc3
1 1

Factor 3 is, (1+0.0025 s)

(1+T35)

Hence open loop transfer function is,

1000
G(s) H(s) =
(1+0.25s) (1 +0.025 s) (1+0.0025 s)
1000
=
G(jo)H(o)
(1+0.25jo) (1 +0.025 ja) (1 +0.0025 jo)
The phase angle table is,

tan 10.25 tan- 10.025 o tan- 10.0025 o OR

0.1 - 1.43 -0.14 - 0.014 - 1.84°

1 - 14.03 -1.43 0.143 - 15.87°
40 - 84.28 - 45 - 5.71 -133.5°
100 - 87.7 68.19 - 14.03 - 169.9°
400 - 89.42 - 84.2 - 45 -218.79

The phase angle plot is shown in the Fig. 11.43(a) (See Fig, on next page.)
From Fig. 11.43(a) Ope = 400 while o,pe = 140 so,
Oge pc

The system is unstable in nature.

Fig. 11.43(a)
<br>

608 Bode 609

FCS Plots FCS
Bode Plots
Quadratic pole
Summary
two plots 2 + j25 0°<6<- 180°
Bode plot consists of o |1
in dB against log 22
Magnitude plot i.e. | jo) H(j0)|expressed
G
Correction -tan
against log a
Phase angle plot ie. Z|Gjo) Hf)|
2
i) =-20 log 2%
1.
Bode Plots are drawn
on semilog papers.
Bode at =
Gfjo) HGo) and corresponding Plot are listed
The various factors in below.
Description Magnitude Plot Phase Angle 40 dB / dec
Factor in G(jo) H() Plot
0 dB
K
Constant
K>1
0
LoaK -180
Log
dB 420 Logk K<1
Log

Log o Quadratic zero

1
+j22
1 Pole at the origin -20 dB/ dec

+180°
0 dB
90 0 dB
40 dB/dec
o Log
)=1 Log Log Log o

j Zero at origin The relative stability of system can be judged from the gain
+20 dB /dec margin and phase
margin.
+90°
0 dB| For ge < Opcr System is stable G.M. and P.M. +ve.
C0=1 Log o Log
For ge > 0per System is unstable, G.M. and PM. -ve.

1
For Ogc =0pcr System is marginally stable, G.M. = PM. = 0.
1
Simple pole o =

-20 dB/dec n oT
1+joT =-tan Review Questions
0 dB
+90
What are Bode Plots?
LOg o
0=1/1 Log o 2) State the advantages of Bode Plots.
3) Explain the nature of Bode Plots for
1+joT Simple zero = T
= +tan
i) Poles at origin
+90
0 dB +20 dB/dec i) Simple pole
Log e
o,=1/T Log o i) Simple zero
4)
Explain the margin. Explain how these values
Concept of gain margin and phase
help in studying relative stability.
5)
Writee a
note on frequency domain specifications.
0) Draw the Bode
diagram for
G(s) = 100(0.02s + 1)
(s+ 1)(0.ls + 1) (0.01s + 1)
<br>

611
610 Bode FCS Bode Plots
FCS Plots Determine the value of K in the transfer function given
Bode diagram,
recording the numerical values. dB below such
on the 10) margin is 20 that
Mark the following a) the gain
crossover frequency 30°
1) gain b) the phase margin is
margin K
2) phase Gjo) H(jo)
crossover frequency joj 01 o +1) (j 0.05 o +1)
3) phase
[Ans. : 0ge = 32;
Ope = 60;
4) gain margin
GM. = 10; P.M. = [Ans. : K= 2.63; K= 10]
plot for the following
15° stablel Draw Bode and determine gain
and phase cross over
11)
frequency. Also determine gain and phase margin.
7) Given
10 10
ii)
R
200 (0.1s + 1)
s(01s +1) s(01s +1)?
+ 1)
s (0.2s + 1)
(0.05s
[Ans. : i) Oe = 10 rad/sec , =
G.M. = + o dB., P.M. = + 45°
c
Fig. 11.44 ge =
) @pe= 10 rad/sec
a) Draw the Bode diagram PM. = 0, G.M. = 0 dB.]
b) Is the system stable?
[Ans. : Ogc University Questions
45; @pe =
G.M. = o; PM = 18° stablel
;
Sketch Bode plot for the following transfer function and. determine gain K for
below. gain crossover frequency ogc to be 5 rad/sec
:

Sketch the asymptotic Bode plot for the transfer function given (Dec.-98)
8)
2(s+ 0.25) Ks?
=
G(s) H(s) = G(s)
s²{s+ 1) (s+ 0.5) (1+0.2s) (1+0.02s)
From the Bode plot determine 2) Define the terms
a) The phase crossover frequency i) Gain margin ü) Phase margin (Dec.-98, May-2001)
3) With necessary diagrams explain the following terms :
b) The gain crossover frequency
Gain cross-over frequency
i) ii) Phase cross-over frequency
c) The gain margin.
iii) Gain margin iv) Phase margin (May-99)
d) The phase margin
Determine the value of gain K for a unity feedback system having
s the system stable? = 1.24 rad/s; =
K
:
0pe = 0.4 rad/s; Oge G(s)
|Ans. s(1+ s) (1 +0.1s)
= unstable|
G.M. = -20 dB: PM 35° so that gain margin is 15 dB
below
given
9) Determine the values of gain K
for the open loop transfer function (May-99)
so that Use Bode plot technique.
is +5dB upwards
[Hint : Refer Ex. 11.9. Shift required
a) the gain margin is 15 dB and
so K = 1.78 for G.M. = + 15 dB]
b) phase margin is 60° 5) open loop transter as
K A nity feedback control system has
Gio) H{jo) =
10
joj0.10 + 1) (jo +1) G(s) + 025)
0.38] s(1+01s)(1
[Ans. : K= 2; K= Sketch Bode
plot and determine from
it
<br>

612 613
Bode Bode Plots
FCS Plots FCS
ii) Phase cross-over frequency
feedback control system has
cross-over frequency A unity
i) Gain 10) K
iv) Phase margin
20)
ii) Gain margin system. Gs) = s(s + 2) (s+
v) Closed loop
stability of the
(Dec-991 Bode plots find gain margin
I Hint Refer Ex. 11.2. The answers are, ®ge 6.8, o,8 and phase margin if :
:

Sketching
G.M. = +3 dB, P.M. = +7°, stable.] = 40
system has open loop transfer function
s i) K
6) A
unity feedback control K= 400
=
100 i) = 4000
G(s) s(1+0.1sX1 +0.29) iii) K
On stability of the system.
Comment (Dec-2001)
Sketch Bode plot and determine
fronm it. [Hint : Refer Ex. 11.2. The answers are
cross over frequency (ii) Phase cross over frequency 0
i) G.M. =+ 26 dB P.M. = +60, stable
()Gain
(iv) Phase margin ii) G.M. = +6 dB P.M. = +12°, stable
(ii) Gain margin
(v) Closed loop stability of the system.
=
(May-2000) ii) G.M. =-16 dB P.M. +32°, unstable]
arethe Bode plots ? What information you get from them of the following system with the help
Find the stability
What of a
7) ? Why Bode Plot t S

logarithmic scale is used while plotting the Bode plots ? (Dec-2000)

11)
200
=
For a certain control system the asymptotic magnitude plot is shown G(s) H(s)
8)
Fig.l1.45. Draw the phase angle plot. Also deternmine the stability of the sVstn
in s(s + 10)(s + 2) (May-2002)
(Dec-200)

dB
0
dB/dec
-20 dB/dec
60

40
40 dB/dec

400
0.1 40
-60dB/dec

Fig. 11.45
9)The feedback control system has
(s+ 160)
G(s) H(s) =
s (s+ 4) (s+ 40)

Sketching Bode plots determine closed loop stability of the system. h

[Hint : Refer Ex. 11.3. The answers are, oge =

1, Opc 10,
G.M. + 30 dB, PM = +669,
Stable]
b o obol(May-2001)