Module 3

EET302 Linear Control System Lecture Notes by T.G.
Sanish Kumar, EED, GEC Thrissur
MODULE 3 – ROOT LOCUS & COMPENSATOR DESIGN

Root locus technique: Construction of Root locus- stability analysis- effect of addition of poles
and zeroes- Effect of positive feedback systems on Root locus
Design of Compensators: Design of lag, lead and lag-lead compensators using Root locus
technique.
PID controllers: PID tuning using Ziegler-Nichols methods.
Simulation based analysis: Introduction to simulation tools like MATLAB/ SCILAB or
equivalent for Root locus based analysis (Demo/Assignment only)
Consider a unity feedback system with forward K

K s ( s  2)
path transfer function, G ( s )  .
s ( s  2)
Characteristic equation is 1  G ( s ) H ( s )  0
s 2  2s  K  0
Roots of characteristic equation, s1 , s2  1  1  K
s1  1  1  K and s1  1  1  K
K=0 s1  0 s2   2 Originates from open loop poles

0<K<1 Real and negative Real and negative
K=1 s1   1 s2  1 Break-away point
K>1 Complex with real part -1 Complex conjugate
with real part -1
K=∞ Terminates at infinity
The path taken by the roots of the characteristic equation when open loop gain K is varied
from 0 to ∞ (or any other system parameter is varied) is called root locus.
Here, root locus originates from s1 = 0 and s2 = -2 and moves along negative real axis and reaches at -
1. The two root loci meet at -1 and then proceeds to infinity at angle ±90˚ to the real axis.
Page 71 of 187
EET302 Linear Control System Lecture Notes by T.G. Sanish Kumar, EED, GEC Thrissur
1 Root locus is symmetrical about real axis

2 Number of branches = order of characteristic equation
3 All branches of root locus start at open loop poles (when K = 0) and ends at either open loop
zeros or infinity (when K = ∞)
4 Number of branches terminating at infinity = n – m where n = No. of poles of G(s)H(s) and
m = No. of zeros of G(s)H(s)
5 A point on real axis lies on the root locus if the sum of the poles and zeros on the real axis to
the right of the point is an odd number.
MAGNITUDE AND ANGLE CRITERION
For a general closed loop system,
C (s) G ( s)

R(s) 1  G (s)H (s)
Characteristic equation is given by,
1  G (s) H (s)  0
G ( s ) H ( s )   1  1  180  (angle shall be odd multiple of 180˚)
G(s) H (s)  1 ----- magnitude criterion
G ( s ) H ( s )   (2 q  1)180 ; q = 0, 1, 2, ----- angle criterion
The values of s which fulfill both the angle and magnitude criteria are the roots of the
characteristic equation or the closed-loop poles.
We first apply the angle criterion to determine whether a point in the s-plane lies on the root
locus. If this condition is satisfied, then we can use the magnitude condition to obtain the
value of K for which this will be a root of the characteristic equation.
K
The open loop transfer function is G ( s )  . Check whether s = -1 is on the root locus.
s ( s  4)( s  5)
K K K
G ( s ) H ( s ) s  1     180 
1 3  4 12 12
G ( s ) H ( s )  180 Angle criterion is satisfied; hence s = -1 is on root locus.
Magnitude criterion, G(s) H (s)  1
K K = 12.
1
12
For K = 12, one of the roots is located at s = -1.
K
The open loop transfer function is G ( s )  . Check whether s = -1 + j2 is on the root locus.
s ( s  2)
Page 72 of 187
K K K
G ( s ) H ( s ) s 1 j 2      180
( 1  j 2)(1  j 2) 5180 5
G ( s ) H ( s )  180 Angle criterion is satisfied; hence s = -1 + j2 is on root locus.
Magnitude criterion, G(s) H (s)  1
K K = 5.
1
5
For K = 5, one of the roots is located at s = -1 + j2.
RULES FOR CONSTRUCTION OF ROOT LOCUS

1 Root locus is symmetrical about the real axis and the number of branches is equal to the order
of the characteristic polynomial.
2 Each branch of the root locus originates from an open loop pole corresponding to K = 0 and
ends at either open loop zero or infinity corresponding to K = ∞.
Number of branches terminating at infinity equals to the difference between the number of
poles and number of zeros of G(s)H(s).
3 A point on the real axis lies on the root locus if the sum of the poles and zeros on the real
axis to the right of the point is an odd number.
4 The (n-m) root locus branches that proceed to infinity do so along the asymptotes with angles
 180(2 q  1)
A  q  0,1, 2, ....( n  m  1)
nm
n = number of open loop poles and m = number of open loop zeros
5 The centroid, the point of intersection of the asymptotes with real axis is given by
sum of poles - sum of zero s
A 
nm
6 The breakaway points (points at which multiple roots of the characteristic equation occur) of
the root locus are the solution of dK  0
ds
7 The angle of departure/arrival (if RL leaves from a complex pole/arrives at a complex zero):
The angle of departure at open loop pole is given by
 d  180   where  is net angle contribution to this pole by all other open loop poles and
zeros
    p   z
  = sum of angles subtended by the phasors drawn to this pole from other poles
p
 = sum of angles subtended by the phasors drawn to this pole from all the zeros
z
The angle of arrival at open loop zero is given by

 a  180   where  is net angle contribution to the zero under consideration by all other
open loop poles and zeros.
    p   z
  = sum of angles subtended by the phasors drawn to this zero from other zeros
p
 = sum of angles subtended by the phasors drawn to this zero from all the poles
z
Page 73 of 187
8 The intersection of root locus with imaginary axis can be determined by using the Routh
criterion or by separating real and imaginary parts.
9 The open loop gain K at any point on the root locus is given by
n
 s  pi
Product of phasor lengths of open loop poles to the point
K  i 1
n

Product of phasor lengths of open loop zeros to the point
i 1
s  zi
K
Consider the root-locus of G ( s )  shown in figure.
( s  3)( s  4)
Net angle contribution at P = G ( s )   K

.  180
( 3.5  j1  3)( 3.5  j1  4)
Net angle contribution at P1 = G ( s )   K

.  71.6
(  5  j1  3)( 5  j1  4)
Net angle contribution at P1 = G ( s )   K

.  71.6
( 2  j1  3)( 2  j1  4)
Net angle contribution on root-locus is always 180˚. Net angle contribution towards left of root locus is
positive and towards right is negative.
K
Plot the root locus of a system whose forward path transfer function is G ( s )  .
( s  3)( s  4)
1 Number of branches of root locus = order of the characteristic polynomial = 2

2 Branches of the root locus originates from open loop poles at s = -3 and s = -4 and both ends
at infinity.
3 On the real axis, the root locus is found between section s = -3 and s = -4 because the sum of
number of poles and zeros to the right of this section is odd.
4 2 branches proceed to infinity along the asymptotes with angles
 180(2 q  1)
A  q  0,1
nm
q  0   A   90
q  1   A   270
Page 74 of 187
5 The centroid (intersection of asymptotes on real axis) is given by

sum of poles - sum of zeros 4  3
A     3.5
nm 2
6 Breakaway points
dK
0
ds
s 2  7 s  12  K  0
K   s 2  7 s  12
dK
 2 s  7  0
ds
s   3 .5
7 Since there is no complex poles and zeros, no angle of departure or arrival.
8 Intersection with imaginary axis
s 2  7 s  12  K  0
Substitute s = j,
  2  j 7  12  K  0
Equating real and imaginary parts,
7   0 and  2  12  K  0
  0 and K   12 ; K is negative
No intersection with imaginary axis.
Page 75 of 187
K
s ( s  2)( s  3)

2 Branches of the root locus originates from open loop poles at s = 0 at s = -2 and s = -3 and
all end at infinity.
3 On the real axis, the root locus is found (a) between section s = 0 and s = -2 and (b) between
section s = -3 and s = -∞ because the sum of number of poles and zeros to the right of this
section is odd.
 180 (2 q  1)
A  q  0,1, 2
nm
q  0   A   60 
q  1   A   180 
q  2   A   300  (same as 60˚)
5 The centroid (intersection of asymptotes) is given by
sum o f poles - sum of zeros 023
A     1.67
nm 3
6 Breakaway points
dK
0
ds
s 3  5s 2  6 s  K  0
K   s 3  5s 2  6 s
dK
  3 s 2  10 s  6  0
ds
3s 2  10 s  6  0
s  0.785 or  2.549 (But, -2.549 does not lie on root locus)
s  0.785
s 3  5s 2  6 s  K  0
1 6
s3
5 K
s2
30  K
s1
5
s0
K
K>0
30 - K > 0
K < 30
0 < K < 30
For K = 30, all elements in one row equal zeros
s3  5s 2  6s  30  0
Auxiliary equation, 5s 2  30  0
Page 76 of 187
 30
s   j 2.45
5
OR (alternate method to find K and )
s 3  5s 2  6 s  K  0
 j 3  5 2  j 6   K  0
 3  6  0 and 5 2  K  0
   2 .4 5 and K  3 0
K
s ( s 2  4 s  13)

2 Branches of the root locus originates from open loop poles at s = 0 at s = -2 + j3 and s = -2
– j3 and all end at infinity.
3 The entire negative real axis is part of the root locus because the sum of number of poles and
zeros to the right of this section is odd.
Page 77 of 187

 180 (2 q  1)
A  q  0,1, 2
nm
q  0   A   60 
q  1   A   180 
q  2   A   300  (same as 60˚)
sum o f poles - sum of zero s 0  2  j3  2  j3
A     1.33
nm 3
6 Breakaway points
dK
0
ds
Characteristic equation is s 3  4s 2  13s  K  0
K   s 3  4 s 2  13s
dK
  3 s 2  8 s  13  0
ds
3s 2  8s  13  0
s  1.33  j1.16
Note: Breakaway point may be real or complex. In order to determine the actual breakaway
point, substitute the breakaway point value in the equation of K to get the value of K. If the
value of K is real positive, the breakaway point is valid one.
K   s 3  4 s 2  13s
When s  1.33  j1.16 ,
K   (  1.33  j1.16) 3  4(  1.33  j1.16) 2  13(  1.33  j1.16)  12.6  j 8.17
When s  1.33  j1.16 ,
K   (  1.33  j1.16) 3  4(  1.33  j1.16) 2  13(  1.33  j1.16)  12.6  j 8.17
Hence, s  1.33  j1.16 is not a valid breakaway point.
7 The angle of departure at open loop pole (-2+j3) is given by
 d  180  
Net angle contribution to this pole,
    p   z =  02  j 3   22  jj 33  123.7  90  213.7
 d  180  213.7   33.7 
The angle of departure at open loop pole (-2-j3) is given by
 d  180  
    p   z =  0 2  j 3   22  jj 33   123.7  90   213.7
 d  180  213.7  393.7    33.7 
s 3  4s 2  13s  K  0
 j 3  4 2  j13  K  0
Page 78 of 187
 3  13  0 and 4 2  K  0

   3 .6 and K  5 2
Plot the root locus of a system whose forward path transfer function is
K
G (s)  .
s ( s  4)( s 2  4 s  20)

2 Branches of the root locus originates from open loop poles at s = 0, s = -4, at s = -2 + j4 and
s = -2 – j4 and all end at infinity.
3 On the real axis, the root locus is found between s = 0 and s = -4 because the sum of number
of poles and zeros to the right of this section is odd.
 180 (2 q  1)
A  q  0,1, 2, 3
nm
q  0   A   45 
q  1   A   135 
q  2   A   225  (same as ±135˚)
q  3   A   315  (same as ±45˚)
Page 79 of 187
sum of p oles - su m of zeros 0  4  2  j4  2  j4

A    2
nm 4
6 Breakaway points: -
Characteristic equation is s ( s  4)( s 2  4 s  20)  K  0
s 4  8s 3  36 s 2  80 s  K  0
K   s 4  8 s 3  36 s 2  80 s
dK
  4 s 3  24 s 2  72 s  80  0
ds
4 s 3  24 s 2  72 s  80  0
s 3  6 s 2  18s  20  0
s1   2; s 2   2  j 2.45; s3   2  j 2.45
K   s 4  8 s 3  36 s 2  80 s
For s1   2 ,
K   (  2) 4  8(  2) 3  36(  2) 2  80(  2)  64
For s 2   2  j 2.45 ,
K   (  2  j 2.45) 4  8(  2  j 2.45) 3  36(  2  j 2.45) 2  80(  2  j 2.45)  100
For s3   2  j 2.45 ,
K   (  2  j 2.45) 4  8(  2  j 2.45) 3  36(  2  j 2.45) 2  80(  2  j 2.45)  100
Hence, s1   2; s 2   2  j 2.45; s3   2  j 2.45 are valid breakaway points.
 d  180  
    p   z =  0 2  j 4   24  j 4   22  jj 44  116.57  63.43  90  270
 d  180  270   90 
 d  180  
    p   z =  02  j 4   24  j 4   22  jj 44   116.57  63.43  90   270
 d  180  270  450    90 
s 4  8s 3  36 s 2  80 s  K  0
 4  j 8 3  36 2  j 80  K  0
8 3  80  0 and  4  36 2  K  0
   3 .1 6 and K  2 6 0
Page 80 of 187
Page 81 of 187
K
s ( s  4)
2

2 Branches of the root locus originates from open loop poles at s = 0 at s = +j2 and s = -j2 and
3 The entire negative real axis is part of the root locus because the sum of number of poles and
zeros to the right of this section is odd.
 180 (2 q  1)
A  q  0,1, 2
nm
q  0   A   60 
q  1   A   180 
q  2   A   300  (same as 60˚)
sum of p oles - sum of zeros 0  j3  j3
A   0
nm 3
6 Breakaway points
dK
0
ds
Characteristic equation is s 3  4s  K  0
K   s 3  4s
dK
 3s 2  4  0
ds
3s 2  4  0
s   j1.15
K  s 3  4s
When s   j1.15 ,
K   ( j1.155) 3  4( j1.155)   j 3.08
When s   j1.155 ,
K   j3.08 (complex conjugate)
Hence, s   j1.155 is not a valid breakaway point.
7 The angle of departure at open loop pole (+j2) is given by
 d  180  
    p   z =  0 j 2    jj 22  90  90  180 
 d  180  180  0 
The angle of departure at open loop pole (-j2) is given by
 d  180  
Page 82 of 187
    p   z =  0 j 2    jj 22   90  90   180 
 d  180  180  360   0 
s 3  4s  K  0
 j 3  j 4   K  0
 3  4  0 and K  0
   2 and K  0
When K = 0, root locus crosses the j axis at j  2
K ( s  1)
s ( s  3.6)
2

2 Branches of the root locus originates from open loop poles at s = 0, at s = 0, at s = -3.6 and
one ends at zero s = -1 and two end at infinity.
Page 83 of 187
3 Points between -1 and -3.6 on negative real axis is part of the root locus because the sum of
number of poles and zeros to the right of this section is odd.
 180(2 q  1)
A  q  0,1
nm
q  0   A   90 
q  1   A   270  (same as 90˚)
su m of po les - sum of zero s 0  0  3.6  1
A     1.3
nm 2
6 Breakaway points
dK
0
ds
Characteristic equation is s 3  3.6 s 2  Ks  K  0
 s 3  3.6 s 2 s 3  3.6 s 2
K 
s 1 s 1
dK ( s  1)(3 s 2  7.2 s )  ( s 3  3.6 s 2 )
 0
ds ( s  1) 2
( s  1)(3 s 2  7.2 s )  ( s 3  3.6 s 2 )  0
3s 3  7.2s 2  3s 2  7.2s  s 3  3.6 s 2  0
2 s 3  6.6 s 2  7.2 s  0
s3  3.3s 2  3.6s  0
s  0 ; s  1.65  j 0.936
s 3  3.6 s 2
K 
s 1
When s  0 ,
K  0;
S = 0 is a valid breakaway point
When s  1.65  j 0.936 ,
K  6.8349.785
Hence, s  1.65  j 0.936 are not valid breakaway point.
7 Since there are no complex poles or complex zeros, there is not need of calculating the angle
of departure or arrival.
s 3  3.6 s 2  Ks  K  0
 j 3  3.6 2  jK   K  0
 3  K   0 and 3.6 2  K  0
  0 and K  0
Page 84 of 187
Plot the root locus of a system whose forward path transfer function is
K
G ( s)  . Also, comment on the stability of the system.
s ( s  6)( s 2  4 s  13)

2 Branches of the root locus originates from open loop poles at s = 0, s = -4, at s = -2 + j3 and s
= -2 – j3 and all end at infinity.
3 On the real axis, the root locus is found between s = 0 and s = -6 because the sum of number of
poles and zeros to the right of this section is odd.
 1 80(2 q  1)
A  q  0,1, 2, 3
nm
q  0   A   45 
q  1   A   135 
q  2   A   225  (same as ±135˚)
q  3   A   315  (same as ±45˚)
sum of poles - su m of zeros 0  6  2  j3  2  j3
A     2.5
nm 4
Page 85 of 187
6 Breakaway points: -
Characteristic equation is s ( s  6)( s 2  4 s  13)  K  0
s 4  10 s 3  37 s 2  78 s  K  0
K   s 4  10s 3  37 s 2  78s
dK
  4 s 3  30 s 2  74 s  78  0
ds
4 s 3  30 s 2  74 s  78  0
s1   4.2; s 2   1.65  j1.386; s 3   1.65  j1.386
K   s 4  10s 3  37 s 2  78s
For s1   4 .2 ,
K   (  4.2) 4  10(  4.2) 3  37(  4.2) 2  78(  4.2)  104.6
For s 2   1.65  j1.386 ,
K   (  1.65  j1.386) 4  10(  1.65  j1.386) 3  37(  1.65  j1.386) 2  78(  1.65  j1.386)
 69.15  j18.12
For s 3   1.65  j1.386 ,
K  69.15  j18.12 (complex conjugate)
Hence, only s1   4.2 is valid breakaway point.
 d  180  
    p   z =  02  j 3   26  j 3   22  jj 33  123.7  36.87  90  250.57 
 d  180  250.57   70.57 
 d  180  
    p   z =  02  j 4   24  j 4   22  jj 44   123.7  36.87  90   250.57 
 d  180  250.57  430.57    70.57 
s 4  10 s 3  37 s 2  78 s  K  0
 4  j10 3  37  2  j 78  K  0
10 3  78  0 and  4  37 2  K  0
   2 .8 and K  2 2 8
For 0 < K < 228, the system is stable because roots lie on the left-hand side of s-plane.
For K = 228, the system is marginally stable because roots lie on the imaginary axis.
For K > 228, the system is unstable because roots lie on the right-hand side of s-plane.
Page 86 of 187
K ( s  3)
s ( s  2)

2 Branches of the root locus originates from open loop poles at s = 0 and at s = -2 and end at
open loop zeros at s = -3 and s = -∞.
section s = -3 and s = -∞ because the sum of number of poles and zeros to the right of this
section is odd.
4 1 branch proceed to infinity along the asymptotes with angles
 1 80 (2 q  1)
A  q  0 (here n – m – 1 = 0)
nm
q  0   A   180 
sum o f p o les - su m o f zeros 023
A   1
nm 2 1
6 Breakaway points
Page 87 of 187
dK
0
ds
Characteristic equation is
K ( s  3)
1 . 0
s ( s  2)
s 2  2 s  K ( s  3)  0
s 2  2s
K 
s3
dK
  ( s  3 )( 2 s  2 )  ( s 2  2 s )  0
ds
2 s 2  8s  6  s 2  2 s  0
 s 2  6s  6  0
s2  6s  6  0
s  1.268 or  4.732 (Both lie on root locus; hence they are breakaway points)
(since s = -4.732 lies between two zeros, it is a break-in point)
7 s 2  2 s  K ( s  3)  0
s 2  (2  K ) s  3 K  0
  2  j (2  K )  3 K  0
  0 and K = -2 (not possible)
The root locus does not cross the imaginary axis.
Page 88 of 187
K ( s  2)( s  3)
s ( s  1)

2 Branches of the root locus originates from open loop poles at s = 0 and at s = -1 and end at
open loop zeros at s = -2 and s = -3.
section s = -2 and s = -3 because the sum of number of poles and zeros to the right of this
section is odd.
4 Breakaway points
dK
0
ds
Characteristic equation is
K ( s  2)( s  3)
1 . 0
s ( s  1)
s2  s
K 
s 2  5s  6
dK
  ( s 2  5 s  6 )( 2 s  1)  ( s 2  s )( 2 s  5 )  0
ds
4 s 2  12 s  6  0
2s 2  6s  3  0
Page 89 of 187
s  0.634 or  2.366 (Both lie on root locus; hence they are breakaway points)
(since s = -2.366 lies between two zeros, it is a break-in point)
5 s 2  s  K ( s 2  5 s  6).  0
(1  K ) s 2  (1  5 K ) s  6 K  0
 (1  K ) 2  j (1  5 K )  6 K  0
  0 and K = -1/5 (not possible)
The root locus does not cross the imaginary axis.
EFFECT OF ADDITION OF POLES
1 Shape of root locus changes and it shifts towards the imaginary axis (towards right)
2 Relative stability decreases
3 System becomes more oscillatory in nature
4 Range of K for stability decreases
5 Settling time increases
Page 90 of 187
EFFECT OF ADDITION OF ZEROS
1 Shape of root locus changes and it shifts towards left of s-plane

2 Relative stability increases
3 System becomes less oscillatory in nature
4 Range of K for stability increases
5 Settling time decreases
Use PI Controller
To improve steady-state error
Use Lag compensator
Use PD Controller
To improve dynamic response
Use Lead compensator
To improve dynamic response and steady-state Use PID Controller
error Use Lead-lag compensator
DESIGN OF COMPENSATORS USING ROOT LOCUS

The root locus approach is used for design purpose when the specification is given in time domain. The
two main specifications which are used in the root locus compensation are damping factor  and
undamped natural frequency n. The dominant closed-loop poles near j-axis are decided by these two
specifications. Therefore, the compensation using root locus reshapes the root locus near j-axis and
origin, so that the dominant closed-loop poles at the desired locations are placed. It is possible to achieve
root locus compensation using lead, lag or lag-lead network.
Complex pole, s1  n  jn 1   2

  cos 1  s1

Peak overshoot, M p  e 1 2 n n 1   2

Settling time, ts  4 
n
n
 
Rise time, tr 
n 1   2
What is the best damping ratio to use? Selection of ξ of industrial control applications requires trade-
off between relative stability & speed of response. A smaller ξ decreases normalized rise time (=ω ntr)
but increases %Mp. Practically, ξ is between 0.4 & 0.7 (ie. Mp is between 5 to 25%).
For accurate system design, ξ should be small as possible to reduce steady state error, e ss since ess is
proportional to ξ.
Page 91 of 187
IMPROVING TRANSIENT RESPONSE

Assume that the desired transient response, defined by percent overshoot
and settling time, is represented by point B.
On the current root locus, A is the point at which the percentage
overshoot is same (ξ is same). By a simple gain adjustment, we can
obtain this point. But, settling time is large (ξωn is small) at point A. With
an existing system, we cannot achieve both the requirements (Mp & ts)
by a simple gain adjustment, since point B does not lie on the root locus.
One way to solve our problem is to replace the existing system with a system whose root locus intersects
the desired design point, B. Unfortunately, this replacement is expensive and counterproductive. Most
systems are chosen for characteristics other than transient response. For example, an elevator cage and
motor are chosen for speed and power. Components chosen for their transient response may not
necessarily meet, for example, power requirements.
Rather than change the existing system, we compensate the system with additional poles and zeros, so
that the compensated system has a root locus that goes through the desired pole location for some value
of gain. One of the advantages of compensating a system in this way is that additional poles and zeros
can be added at the low-power end of the system before the plant. Addition of compensating poles and
zeros need not interfere with the power output requirements of the system or present additional load or
design problems. The compensating poles and zeros can be generated with a passive or an active
network.
A possible disadvantage of compensating a system with additional open-loop poles and zeros is that
the system order can increase, with a subsequent effect on the desired response.
Note: Adding only zero is often problematic because such controller amplifies the high frequency noise.
Adding only pole if often problematic because such controller generates a less stable system (by moving
the closed loop poles to the right).
LEAD COMPENSATOR
1 1
s s
For lead compensator, G ( s )  K T where 0    1 and lag compensator, G (s)  K T
c
c
1 1
s s
T T
where   1 .
Both a lead compensator and a lag compensator have the same shape: So they have a zero at s   1
T
1
and a pole at s   1 or s  . For lead compensators the pole lies more to the left in the complex
T T
plane than the zero and vice versa for lag compensators
RULES FOR DESIGN OF LEAD COMPENSATOR USING ROOT LOCUS

1 List the system specifications and translate them into a desired root location for the dominant
roots.
2 Sketch the uncompensated root locus, and determine whether the desired root locations can
be realized with an uncompensated system (with gain adjustment only)
3 If a compensator is necessary, place the zero of the phase-lead network directly below the
desired root location or to the left of the first to real poles.
Page 92 of 187
4 Determine the pole location so that the total angle at the desired root location is 180˚ and
therefore is on the compensated root locus.
5 Evaluate the total system gain at the desired root location and then calculate the error
constant.
6 Repeat the steps if the error constant is not satisfactory.
K
Consider a unity feedback system with open loop transfer function, G ( s )  . Design a lead
s ( s  8)
compensator to meet the following specifications i) % Mp = 9.5% ii) natural frequency of oscillations,
n = 12 rad/sec iii) velocity error constant K v  10

2 Branches of the root locus originates from open loop poles at s = 0 and s = -8 and both ends
at infinity.
3 On the real axis, the root locus is found between section s = 0 and s = -8 .
 180(2 q  1)
A  q  0,1
nm
q  0   A   90
q  1   A   270
sum of poles - su m o f zero s 08
A    4
nm 2
6 Breakaway points
dK
0
ds
s 2  8s  K  0
K   s 2  8s
dK
 2 s  8  0
ds
s  4
7 Since there is no complex poles and zeros, no angle of departure or arrival; no intersection
with imaginary axis.
Page 93 of 187
1 DOMINANT POLE:

1 2
Mp  e

 ln  M p   2.354
1  2
  2.354  1   2
   2.354 2  1   2 
2
2.354
  0.6
2.354 2   2
Dominant pole, sd   n  j n 1   2 =  0.6  12  j12 1  0.6 2  7.2  j 9.6
P does not lie on root locus and hence simple gain adjustment is not possible.
2 ANGLE TO BE CONTRIBUTED BY LEAD NETWORK
   n  180
Net angle contribution by poles and zeros of uncompensated system to this pole,
n   p  z =  07.2  j 9.6   87.2  j 9.6  126.9  85.2  212
  212.1  180  32
Note:  zc   pc  32 
3 POLE & ZERO OF LEAD COMPENSATOR
Page 94 of 187
Draw horizontal line PA

Draw line PO
Draw bisector PB so that  A P B   B P O   A P O
2
Draw PC and PD so that  C P B   B P D   3 2   1 6 
2 2
Pole of compensator Pc = -16
Zero of compensator Zc = -9
1
Zc    9
T
T = 0.11
1
Pc    16
T
 = 0.56
Transfer function, Gc ( s )  s  9
s  16
4 OPEN LOOP TRANSFER FUNCTION OF COMPENSATED SYSTEM
K ( s  9)
GO ( s ) 
s ( s  8)( s  16)
Magnitude criterion, GO ( s ) 7.2  j 9.6  1
K ( 7.2  j 9.6  9)
 0.00649 K 180 
(  7.2  j 9.6)( 7.2  j 9.6  8)( 7.2  j 9.6  16)
0 .0 0 6 4 9 K  1
K  154
154( s  9)
GO ( s ) 
s ( s  8)( s  16)
5 ERROR REQUIREMENT
154  9
K v  L t sG O ( s )   1 0 .8
s 0 8  16
Kv is greater than 10.
NOTE:
Net angle contribution of uncompensated system at desired pole location s  7.2  j 9.6 ,
K
G ( s ) s 7.2  j 9.6   0.009 K 148
(  7.2  j 9.6)(  7.2  j 9.6  8)
Angle contribution of lead network at desired pole location s  7.2  j 9.6 ,
( 7.2  j 9.6  9)
Gc ( s ) s  7.2  j 9.6   0.75 32
( 7.2  j 9.6  16)
Net angle contribution of compensated system at desired pole location s  7.2  j 9.6 ,
K ( 7.2  j 9.6  9)
Go ( s ) s  7.2  j 9.6   0.0065 K 180
( 7.2  j 9.6)( 7.2  j 9.6  8)( 7.2  j 9.6  16)
Page 95 of 187
K
Consider a unity feedback system with open loop transfer function, G ( s )  . Design
s ( s  4)( s  7)
a lead compensator to meet the following specifications i) % Mp = 12.63% ii) natural frequency of
oscillations, n = 8 rad/sec iii) velocity error constant K v  2.5

section s = -7 and s = -∞
 180 (2 q  1)
A  q  0,1, 2
nm
q  0   A   60 
q  1   A   180 
q  2   A   300  (same as 60˚)
sum o f poles - sum of zero s 047
A     3.6 7
nm 3
6 Breakaway points
dK
0
ds
s 3  11s 2  28s  K  0
K   s 3  11s 2  28s
dK
 3s 2  22 s  28  0
ds
3s 2  22s  28  0
s   1 .6 4
7 s 3  11s 2  28s  K  0
 j 3  11 2  j 28  K  0
 3  28  0 and 11 2  K  0
   5 .3 and K  3 0 8
Page 96 of 187
1 DOMINANT POLE:

1 2
Mp  e

 ln  M p   2.07
1 2
  2.07  1   2
   2.07 2  1   2 
2
2.07
  0.55
2.07 2   2
Dominant pole, sd   n  j n 1   2 =  0.55  8  j8 1  0.552  4.4  j 6.7
   n  180
n   p  z =  04.4  j 6.7   4.4
4
 j 6.7
  7
4.4  j 6.7
 123.3  93.4  68.8  286 
  286  180  106
Since the angle contribution is more than 60˚, the lead compensator is realized as cascade of
two compensators with each compensator contributing half of the required angle.
Page 97 of 187
106 
   53 
2
Draw line PO
2

Draw PC and PD so that  C P B   B P D   5 3   26.5 
2 2
Pole of compensator Pc = -13.5
Zero of compensator Zc = -4.6
1
Zc     4 .6
T
T = 0.217
1
Pc     13.5
T
 = 0.34
Transfer function, Gc ( s)  ( s  4.6) 2 (Two compensators in cascade)
2
( s  13.5)
K ( s  4.6) 2
GO ( s ) 
s ( s  4)( s  7)( s  13.5) 2
K ( 4.4  j 6.7  4.6) 2

GO ( s ) 
( 4.4  j 6.7)(  4.4  j 6.7  4)(  4.4  j 6.7  7)(  4.4  j 6.7  13.5) 2
 0.00091K 180 
0 .0 0 0 9 1 K  1
K  1099
1099( s  4.6) 2
GO ( s ) 
s ( s  4)( s  7)( s  13.5) 2
5 ERROR REQUIREMENT
1099  4.6 2
K v  Lt sGO ( s )   4.56
s0 4  7  13.52
Kv is greater than 2.5.
Page 98 of 187
Note: See the zero of the lead compensator is directly below the desired root location in both problems.
Consider a unity feedback system with open loop transfer function, G ( s )  K2 . Design a lead
s
compensator to meet the following specifications i) %M p  35% ii) settling time (with 2% criterion)
< 4 s iii) acceleration error constant K a  2
1 DOMINANT POLE:

1 2
Mp  e

 ln  M p   1.05
1 2
  1.05  1   2
   1.05 2  1   2 
2
1.05
   0.32
1.05 2   2
Page 99 of 187
For %M p  35% ,   0.32 . Select   0.4

4
ts 
 n
4
n   2.5rad/sec
0.4  4
Dominant pole, sd   n  j n 1   2 =  0.4  2.5  j 2.5 1  0.4 2  1  j 2.3
   n  180
n   p  z =  0 1 j 2.3   01 j 2.3  113.5  113.5  227
  227  180  47
Place the zero of the compensator directly below the desired location at s = -1 (point D).
Draw line PC at   47
Point C gives pole of the compensator.
Pole of compensator Pc = -3.8
Zero of compensator Zc = -1
1
Zc    1
T
T=1
1
Pc     3.8
T
 = 0.26
(s  1)
Transfer function, Gc (s) 
(s  3.8)
K (s  1)
Go ( s)  2
s (s  3.8)
Magnitude criterion, GO ( s ) 1 j 2.5  1
K ( 1  j 2.5  1)
Go ( s)   0.0758 K
( 1  j 2.5) 2 ( 1  j 2.5  3.8)
0 .0 7 5 8 K  1
K  1 3 .2
13.2(s  1)
Go ( s) 
s 2 (s  3.8)
5 ERROR REQUIREMENT
1 3.2  1
K a  L t s 2GO ( s )   3 .47
s 0 3.8
Ka is greater than 2.
Page 100 of 187

LAG COMPENSATOR
RULES FOR DESIGN OF LAG COMPENSATOR USING ROOT LOCUS

roots.
3 If a compensator is necessary, calculate the loop gain at the desired root location and thus the
system error constant.
4 Compare the uncompensated error constant with the desired error constant, and calculate the
necessary increase that must result from the pole-zero ratio  of the compensator.
5 With the known ratio of the pole-zero combination of the compensator, determine a suitable
location of the pole and zero of the compensator so that the compensated root locus will still
pass through the desired root location. Locate the pole and zero near the origin of the s-plane
in comparison to n.
Page 101 of 187

K
Design a lag compensator for a system with open loop transfer function, G ( s )  to
s ( s  1)( s  3)
satisfy the following specifications i) damping ratio 0.5 ii) settling time = 10sec iii) velocity error
constant K v  5.5

section s = -3 and s = -∞.
 180 (2 q  1)
A  q  0,1, 2
nm
q  0   A   60 
q  1   A   180 
q  2   A   300  (same as 60˚)
sum of poles - sum of zeros 0 1 3
A     1.33
nm 3
6 Breakaway points
dK
0
ds
s 3  4 s 2  3s  K  0
K  s 3  4s 2  3s
dK
 3s 2  8s  3  0
ds
3s 2  8s  3  0
s   0 .4 5
s 3  4 s 2  3s  K  0
 j 3  4 2  j 3  K  0
 3  3  0 and 4 2  K  0
   j1.73 and K  1 2
Page 102 of 187

1 DOMINANT POLE:
4
ts 
 n
4
n   0.8rad/sec
0.5 10
Dominant pole, sd   n  j n 1   2 =  0.5  0.8  j 0.8 1  0.52  0.4  j 0.693
2 GAIN K AT DOMINANT POLE
K
G (s)  1
s ( s  1)( s  3) s 4  j 0.693
K
 0.5067 K  1
(  0.4  j 0.693)( 0.4  j 0.693  1)( 0.4  j 0.693  3) s 0.4  j 0.693
K = 1.973
3 DETERMINATION OF 
1.973
G (s) 
s ( s  1)( s  3)
1 .97 3
K v  L t sG ( s )   0 .65 7
s 0 1 3
Desired Kv = 5.5
Page 103 of 187

Factor by which Kv is to be increased = A  5.5  8.37

0.657
  1.2  A  1.2  8.37  10
4 TRANSFER FUNCTION OF LAG COMPENSATOR
The pole and zero of the compensator are placed very close to the origin. There are one open
loop pole at origin and second pole at s = -1. Place the zero of the compensator at s = 0.1 x
second open loop pole ie. at s = 0.1 x -1 = -0.1.
Zero of compensator z c  1   0 .1
T
1  0.1
Pole of compensator pc     0.01
T 10
1
s
Transfer function of lag compensator, Gc ( s)  T s  0.1

1 s  0.01
s
T
5 TRANSFER FUNCTION OF COMPENSATED SYSTEM
1.973(s  0.1)
Transfer function of lag compensator, GO (s) 
s(s  1)(s  3)(s  0.01)
6 ERROR REQUIREMENT
1.973  0.1
K v  L t sG O ( s )   6.58
s 0 1  3  0.01
Kv is greater than 5.5.
The design is acceptable.
K
Design a lag compensator for a system with open loop transfer function, G ( s )  to satisfy
s ( s  2)
the following specifications i) damping ratio 0.45 ii) settling time = 4sec iii) velocity error constant
K v  20

2 Branches of the root locus originates from open loop poles at s = 0 and s = -2 and both ends
at infinity.
3 On the real axis, the root locus is found between section s = 0 and s = -2 .
 180(2 q  1)
A  q  0,1
nm
q  0   A   90
q  1   A   270
sum of poles - sum of zero s 02
A    1
nm 2
6 Breakaway points
dK
0
ds
Page 104 of 187

s 2  2s  K  0
K   s 2  2s
dK
 2 s  2  0
ds
s  1
7 Since there is no complex poles and zeros, no angle of departure or arrival; no intersection
with imaginary axis.
1 DOMINANT POLE:
4
ts 
n
4
n   2.22rad/sec
0.45  4
Dominant pole, sd   n  j n 1   2 =  0.45  2.22  j 2.22 1  0.452  1  j 2
2 GAIN K AT DOMINANT POLE
K
G (s)  1
s ( s  2) s 1 j 2
Page 105 of 187

K
 0.2 K  1
(  1  j 2)(  1  j 2  2)
K=5
5
G(s) 
s ( s  2)
5
K v  L t sG ( s )   2 .5
s 0 2
Desired Kv = 20
20
Factor by which Kv is to be increased = A  8
2.5
  1.2  A  1.2  8  10
loop pole at origin and second pole at s = -2. Place the zero of the compensator at s = 0.1 x
second open loop pole ie. at s = 0.1 x -2 = -0.2.
Zero of compensator z c  1   0.2
T
1  0.2
Pole of compensator pc     0.02
T 10
1
s
Transfer function of lag compensator, Gc (s)  T  s  0.2
1 s  0.02
s
T
5(s  0.2)
Transfer function of lag compensator, GO ( s) 
s(s  2)( s  0.02)
6 ERROR REQUIREMENT
5  0 .2
K v  L t sG O ( s )   25
s 0 2  0.02
LEAD-LAG COMPENSATOR
RULES FOR DESIGN OF LEAD-LAG COMPENSATOR USING ROOT LOCUS

roots.
3 If a compensator is necessary, place the zero of the phase-lead network directly below the
desired root location or to the left of the first to real poles.
4 Determine the pole location so that the total angle at the desired root location is 180˚ and
therefore is on the compensated root locus.
Page 106 of 187

5 Evaluate the total system gain at the desired root location and then calculate the error
constant. If error requirement is satisfied, then only lead compensation is sufficient and
otherwise provide lag compensation.
6 Compare the uncompensated error constant with the desired error constant, and calculate the
necessary increase that must result from the pole-zero ratio  of the compensator.
7 With the known ratio of the pole-zero combination of the compensator, determine a suitable
location of the pole and zero of the compensator so that the compensated root locus will still
pass through the desired root location. Locate the pole and zero near the origin of the s-plane
in comparison to n.
8 Determine the transfer function of lead-lag compensator and compensated system. Check the
error constant of compensated system. If it is satisfied, then the design is accepted and
otherwise, repeat the design by modifying the locations of poles and zeros of the
compensator.
K
Design a lead-lag compensator for a system with open loop transfer function, G ( s )  to
s ( s  0.5)
satisfy the following specifications i) damping ratio 0.5 ii) undamped natural frequency of
oscillations n = 5 rad/s iii) velocity error constant K v  80 sec 1
1 DOMINANT POLE:
Dominant pole, sd   n  j n 1   2 =  0.5  5  j 5 1  0.52  2.5  j 4.3
P does not lie on root locus and hence simple gain adjustment is not possible.
   n  180
n   p  z =  0 2.5  j 4.3   2.5
0.5
 j 4.3
 120.2  115  235.2 
  235.2  180  55.2
Draw line PO
2
Draw PC and PD so that  C P B   B P D   55.2 
  2 7.6 
2 2
Pole of compensator Pc = -10
Zero of compensator Zc = -2.6
1
Zc     2.6
T
T = 0.385
1
Pc    10
T
 = 0.26
Transfer function, Gc ( s )  s  2.6
s  10
Page 107 of 187


K ( s  2.6)
GO ( s ) 
s ( s  0.5)( s  10)
K ( 2.5  j 4.3  2.6)
 0.0211K 180 
( 2.5  j 4.3)(  2.5  j 4.3  0.5)(  2.5  j 4.3  10)
0 .0 2 1 1 K  1
K  4 7 .4
47.4( s  2.6)
GO ( s ) 
s ( s  0.5)( s  10)
5 ERROR REQUIREMENT
4 7 .4  2 .6
K v  L t sG O ( s )   2 4 .6
s 0 0 .5  10
Kv is insufficient.
K v  2 4.6
Desired Kv = 80
Factor by which Kv is to be increased = A  80  3.25
24.6
  1.2  A  1.2  3.25  4
loop pole at origin and second pole at s = -0.5. Place the zero of the compensator at s = 0.1
x second open loop pole ie. at s = 0.1 x -0.5 = -0.05.
Zero of compensator z c  1   0.05
T
1 0.05
Pole of compensator pc    0.0125
T 4
1
s
Transfer function of lag compensator, Gc (s)  T  s  0.05
1 s  0.0125
s
T
5(s  2.6)(s  0.05)
Transfer function of lag compensator, GO (s) 
s(s  0.5)(s 10)(s  0.0125)
9 ERROR REQUIREMENT
4 7.4  2.6  0.0 5
K v  L t sG O ( s )   98 .6
s 0 0.5  10  0.0125
Page 108 of 187

TUNING OF PID CONTROLLER BY ZIEGLER-NICHOLS METHOD
If the mathematical model of the plant can be derived, then it is possible to apply various design
techniques for determining the parameters of the controller that will meet the transient and steady-state
specifications of the closed-loop system. However, if the plant is so complicated that its mathematical
model cannot be easily obtained, then an analytical approach to the design of a PID controller is not
possible. Then, we must resort to an experimental approach to the design of PID controllers. The process
of selecting the controller parameters to meet given specifications is known as controller tuning.
 1 
K p 1   Td s  Plant
R(s)  Ti s  C(s)
Ki  1 
The transfer function of a PID controller is Gc ( s)  K p   K d s  K p 1   Td s 
s  Ti s 
ZIEGLER-NICHOLS OPEN LOOP TUNING METHOD
Ziegler and Nichols suggested rules for tuning PID controllers (to determine the values of proportional
gain Kp, integral time Ti and derivative time Td) based on the experimental step response characteristics
of the plant.
First obtain experimentally the response of the plant to a unit step input as shown in figure below.
The first tuning method is applied for plants that show an S-shaped characteristic for their step response.
The S-shaped curve is characterized by two constants—the delay time, L and time constant, T. A tangent
is drawn to the S-shaped curve at the point of inflection (where the slope of the step response has its
maximum value). The delay time, L is determined by the intersection of the tangent line with the time
axis and time constant, T is determined by intersection of tangent line with final value of step response
c(t)=K.
Ziegler and Nichols suggested setting the values of the parameters Kp, Ti, and Td according to the
following Table.
Page 109 of 187

Type of Kp Kp Kd
Ti  Td 
controller Ki Kp
P T ∞ 0
L
PI 0.9T L 0
L 0.3
PID 1.2T 2L 0.5L
L
PID controller tuned by open loop tuning gives

2
 1
0.6T  s  
 1  1.2T  1   L 
Gc ( s)  K p 1   Td s   1   0.5Ls  
 Ti s  L  2 Ls  s
PID controller has a pole and origin and double zeros at s = -1/L.
Features of Ziegler Nichols open loop tuning:
1. applicable to stable object with no oscillations
2. easy to use
3. some processes do not permit step response tests or it gives a little information about the
process, the step input applied should be small enough for the response to stay within the bounds
of linearity
Page 110 of 187

Module 3

Uploaded by

Copyright:

Available Formats

Module 3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Module 3

Uploaded by

Copyright:

Available Formats

EET302 Linear Control System Lecture Notes by T.G.

Sanish Kumar, EED, GEC Thrissur

MODULE 3 – ROOT LOCUS & COMPENSATOR DESIGN

Consider a unity feedback system with forward K

Roots of characteristic equation, s1 , s2  1  1  K

K=0 s1  0 s2   2 Originates from open loop poles

1 Root locus is symmetrical about real axis

MAGNITUDE AND ANGLE CRITERION

For a general closed loop system,

Characteristic equation is given by,

G ( s ) H ( s )   1  1  180  (angle shall be odd multiple of 180˚)

G(s) H (s)  1 ----- magnitude criterion

G ( s ) H ( s )   (2 q  1)180 ; q = 0, 1, 2, ----- angle criterion

G ( s ) H ( s )  180 Angle criterion is satisfied; hence s = -1 is on root locus.

Magnitude criterion, G(s) H (s)  1

For K = 12, one of the roots is located at s = -1.

G ( s ) H ( s )  180 Angle criterion is satisfied; hence s = -1 + j2 is on root locus.

Magnitude criterion, G(s) H (s)  1

For K = 5, one of the roots is located at s = -1 + j2.

RULES FOR CONSTRUCTION OF ROOT LOCUS

The angle of arrival at open loop zero is given by

Net angle contribution at P = G ( s )   K

Net angle contribution at P1 = G ( s )   K

Net angle contribution at P1 = G ( s )   K

1 Number of branches of root locus = order of the characteristic polynomial = 2

5 The centroid (intersection of asymptotes on real axis) is given by

1 Number of branches of root locus = order of the characteristic polynomial = 3

1 Number of branches of root locus = order of the characteristic polynomial = 3

4 3 branches proceed to infinity along the asymptotes with angles

 3  13  0 and 4 2  K  0

1 Number of branches of root locus = order of the characteristic polynomial = 4

sum of p oles - su m of zeros 0  4  2  j4  2  j4

1 Number of branches of root locus = order of the characteristic polynomial = 3

1 Number of branches of root locus = order of the characteristic polynomial = 3

1 Number of branches of root locus = order of the characteristic polynomial = 4

1 Number of branches of root locus = order of the characteristic polynomial = 2

1 Number of branches of root locus = order of the characteristic polynomial = 2

EFFECT OF ADDITION OF POLES

EFFECT OF ADDITION OF ZEROS

1 Shape of root locus changes and it shifts towards left of s-plane

DESIGN OF COMPENSATORS USING ROOT LOCUS

Complex pole, s1  n  jn 1   2

Peak overshoot, M p  e 1 2 n n 1   2

IMPROVING TRANSIENT RESPONSE

RULES FOR DESIGN OF LEAD COMPENSATOR USING ROOT LOCUS

1 Number of branches of root locus = order of the characteristic polynomial = 2

Draw horizontal line PA

1 Number of branches of root locus = order of the characteristic polynomial = 3

K ( 4.4  j 6.7  4.6) 2

For %M p  35% ,   0.32 . Select   0.4

Page 100 of 187

RULES FOR DESIGN OF LAG COMPENSATOR USING ROOT LOCUS

Page 101 of 187

1 Number of branches of root locus = order of the characteristic polynomial = 3

Page 102 of 187

Page 103 of 187

Factor by which Kv is to be increased = A  5.5  8.37

1 Number of branches of root locus = order of the characteristic polynomial = 2

Page 104 of 187