Chapter 9

Design via Root Locus

1
EE360 (Chapter 9) Learning Outcomes：After completing this chapter the student will be able to：

□State the method of Compensation (Sections 9.1)

□State how does PI controller improve steady-state error (Section 9.2)
□State how does PD controller improve transient response (Section 9.3)
□State how does PID controller improve both steady-state error and transient response(Section 9.4)
Scope of ch. 9 in this course: 9.1 - 9.4
9.1 Introduction
 Week 15 (Chapter 9) Learning Outcomes：
Learning Outcomes in Week 15 (Chapter 9):
After completing this chapter the student will be able
to：
□State the method of Compensation (Sections 9.1)
□State how does PI controller improve steady-state
error (Section 9.2)
□State how does PD controller improve transient
response (Section 9.3)
□State how does PID controller improve both
steady-state error and transient response(Section
9.4)
E1C09 11/03/2010 13:29:48 Page 494

494 Chapter 9 Design via Root Locus

TABLE 9.7 Types of cascade compensators

Function Compensator Transfer function Characteristics

Improve steady-state error
Improve steady-state error
s þ zc
PI K 1. Increases system type.
s
2. Error becomes zero.
3. Zero at zc is small and negative.
4. Active circuits are required to implement.
s þ zc
Lag K 1. Error is improved but not driven to zero.
s þ pc
2. Pole at pc is small and negative.
3. Zero at zc is close to, and to the left of, the
pole at pc .
4. Active circuits are not required to implement.

Improve transient response
Improve transient response
Improve steady-state error and
transient response
Improve steady-state error and
transient response
PD Kðs þ zc Þ 1. Zero at zc is selected to put design point on
root locus.
2. Active circuits are required to implement.
3. Can cause noise and saturation; implement
with rate feedback or with a pole (lead).
s þ zc
Lead K 1. Zero at zc and pole at pc are selected to put
s þ pc
design point on root locus.
2. Pole at pc is more negative than zero at zc .
3. Active circuits are not required to implement.
ðs þ z Þðs þ z Þ
Apago
PID K PDF Enhancer
s
lag lead
1. Lag zero at z lag and pole at origin improve
steady-state error.
2. Lead zero at zlead improves transient
response.
3. Lag zero at zlag is close to, and to the left of,
the origin.
4. Lead zero at zlead is selected to put design
point on root locus.
5. Active circuits required to implement.
6. Can cause noise and saturation; implement
with rate feedback or with an additional pole.
ðs þ zlag Þðs þ zlead Þ
Lag-lead K 1. Lag pole at plag and lag zero at zlag are used
ðs þ plag Þðs þ plead Þ
to improve steady-state error.
2. Lead pole at plead and lead zero at zlead are
used to improve transient response.
3. Lag pole at plag is small and negative.
4. Lag zero at zlag is close to, and to the left of,
lag pole at plag .
5. Lead zero at zlead and lead pole at plead are
selected to put design point on root locus.
6. Lead pole at plead is more negative than lead
zero at zlead .
7. Active circuits are not required to implement.
E1C09 11/03/2010 13:29:59 Page 506

506 Chapter 9 Design via Root Locus

Passive-Circuit Realization
Lag, lead, and lag-lead compensators can also be implemented with passive net-
works. Table 9.11 summarizes the networks and their transfer functions. The transfer
functions can be derived with the methods of Chapter 2.
The lag-lead transfer function can be put in the following form:
  
1 1
sþ sþ
T1 T2
Gc ðsÞ ¼    ð9:50Þ
1 a
sþ sþ
aT 1 T2
where a < 1. Thus, the terms with T1 form the lead compensator, and the terms with T2
form the lag compensator. Equation (9.50) shows a restriction inherent in using this
passive realization. We see that the ratio of the lead compensator zero to the lead
compensator pole must be the same as the ratio of the lag compensator pole to the lag
compensator zero. In Chapter 11 we design a lag-lead compensator with this restriction.
A lag-lead compensator without this restriction can be realized with an active
network as previously shown or with passive networks by cascading the lead and lag
networks shown in Table 9.11. Remember, though, that the two networks must be
isolated to ensure that one network does not load the other. If the networks load
each other, the transfer function will not be the product of the individual transfer
functions. A possible realization using the passive networks uses an operational
amplifier to provide isolation. The circuit is shown in Figure 9.63. Example 9.10
demonstrates the design of a passive compensator.

TABLE 9.11 Passive realization of compensators

Function
Apago PDF Enhancer
Network Transfer function,
Vo ðsÞ
Vi ðsÞ

R2

+ +
1
Lag compensation R2 sþ
R2 R2 C
vi(t) vo(t)
R1 þ R2 1
sþ
C ðR1 þ R2 ÞC
– –

R1

1
Lead compensation + + sþ
R1 C
C 1 1
vi(t) R2 vo(t) sþ þ
R1 C R2 C
– –

R1

+ +
Lag-lead compensation   
C1 1 1
R2 sþ sþ
vi(t) vo(t) R1 C 1 R2 C2
 
1 1 1 1
C2 s þ
2 þ þ sþ
R1 C 1 R 2 C 2 R 2 C 1 R1 R2 C 1 C 2
– –
E1C09 11/03/2010 13:29:56 Page 504

504 Chapter 9 Design via Root Locus

Active-Circuit Realization
In Chapter 2, we derived
Z2(s)
V o ðsÞ Z2 ðsÞ
¼ ð9:44Þ
I2(s) V i ðsÞ Z1 ðsÞ
Z1(s)
Vi (s) V1(s)
– Vo(s) as the transfer function of an inverting operational amplifier whose
Ia(s) configuration is repeated here in Figure 9.60. By judicious choice of Z1(s)
I1(s) + and Z2(s), this circuit can be used as a building block to implement the
compensators and controllers, such as PID controllers, discussed in this
chapter. Table 9.10 summarizes the realization of PI, PD, and PID
controllers as well as lag, lead, and lag-lead compensators using opera-
FIGURE 9.60 Operational amplifier
configured for transfer function realization tional amplifiers. You can verify the table by using the methods of
Chapter 2 to find the impedances.

TABLE 9.10 Active realization of controllers and compensators, using an operational amplifier
Z2 ðsÞ
Function Z1 ðsÞ Z2 ðsÞ Gc ðsÞ ¼  Z1 ðsÞ

R1 R2 R2
Gain 
R1
1
R C
Integration  RC
s
C R
Differentiation RCs
 
Apago PDF Enhancer
R1 R C 2 sþ
1
PI controller R2 R2 C

R1 s
C
R2  
PD controller 1
R2 C s þ
R1 R1 C

C1
R2 C2 2 3
  1
PID controller 6 R2 C 1 R1 C 2 7
6
4 R 1 þ C 2 þ R2 C 1 s þ s 5
7
R1

C1 C2  
1
sþ
C1 R1 C1
Lag compensation   
R1 R2 C2 1
sþ
R2 C2
where R2 C2 > R1 C1
 
1
C1 C2 sþ
C1 R1 C1
  
C2 1
Lead compensation sþ
R2 C2
R1 R2
where R1 C1 > R2 C2
Root Locus – Closed Loop Design using RL

Design Fixed!
Target!
C (s) G (s )
Controller Plant

• Place closed-loop poles at desired location

– by tuning the gain C(s) = K. (for time domain specs)
• If root locus does not pass the desired location, then
reshape the root locus
– by adding poles/zeros to C(s). (How?)
Compensation

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 61

Root Locus – Effect of Adding Poles
• Pulling root locus to the RIGHT
– Less stable
– Slow down the settling

Im Im Im

Re Re Re

Add a pole Add a pole

Root Locus – Effect of Adding Zeros
• Pulling root locus to the LEFT
Im
– More stable
– Speed up the settling

Re

Add a zero
Im Im Im

Re Re Re
Root Locus – Adding Poles/Zeros Remarks
• Adding only zero C ( s ) = K ( s + z ), ( z > 0)
– often problematic because such controller amplifies the
high-frequency noise.

• Adding only pole C ( s ) = K /( s + p ), ( p > 0)

– often problematic because such controller generates a
less stable system (by moving the closed-loop poles to
the right).

• These facts can be explained by using frequency response

analysis.

• Add both zero and pole!

Root Locus – Lead and Lag Compensators

C (s) G (s ) s+z
Controller Plant C (s) = K , ( z > 0, p > 0)
s+ p

• Lead compensator • Lag compensator

Im Im

Re Re

Why these are called “lead” and “lag”?

We will see that from frequency response in this class.
9.1 Introduction

Root Locus:
Possible Design Point via
gain adjustment
Desired design point that
cannot be met via simple
gain adjustment

Response from poles at A

and B
Problem :
How can we design a transient responses that are
not on the root locus

One Solution:
Compensate the system with additional poles and
zeros, so that the compensated system has a root
locus that goes through the desired pole location
Compensation techniques (a) Cascade; (b) feedback
9.2 Improving Steady-state error by cascade
PI (Proportional + Integral) compensation
Placing an open-loop pole at the origin:
Increase the system type by one, therefore steady
state error can be improved.

𝐾2
PI Controller transfer function = 𝐾1 +
𝑠
PI controller
 Improving System Performance
Ideal Integral Compensation (PI)
Lag Compensation

1 Improving System Performance

1 Ideal Integral Compensation (PI)

1 Lag Compensation

ENGI 5821 Unit 8: Design via Root Locus

Improving System Performance

Fundamentally, we wish to improve the performance of our systems

in terms of the principal control systems design criteria: transient
response, steady-state error, and stability.
We saw in the Steady-State Error unit that the addition of
integration (1/s trans. function) into the forward path can reduce
finite SS error to zero. Here we discuss in more detail how to
achieve this.
We also discuss how to improve transient performance. With a
given system the only way we have to adjust transient performance
is to find a suitable point on the RL. Assume we have found point
A on the RL below which gives us the desired %OS.
However, while we are happy with the damping ratio ζ and %OS
at A we would prefer the reduced settling time at B.

Since B is not on the RL it is not achievable by adjusting gain.

Instead we must compensate the system with additional poles or
zeros so that the new system’s RL goes through B.
Compensators can be added to a system to improve either
steady-state error or transient respnose. They can either be
cascaded with the original controller and plant or added to the
feedback path:

If a compensator involves pure integration or pure differentiation

we refer to it as an ideal compensator. Ideal compensators must
be built with active amplifiers. Non-ideal compensators can be
constructed using passive components.
Ideal Integral Compensation (PI)

Steady-state error can be improved by placement of an open-loop

pole at the origin. This increases the system type. Ideally we would
like to add such a zero without affecting the transient response.
This allows transient response to be compensated for separately.
Assume we begin with the system below on the left which is
operating at some point A which yields a desirable transient
response....
To improve steady-state error a compensator is added which places
a pole at the origin. However, the point A is no longer on the RL.
We can achieve the desired steady-state error while maintaining
approximately the same transient response if we also add a zero
close to the origin.
If the zero is close enough to the pole then θpc ≈ θzc which means
they effectively cancel out, leaving A still on the RL. The required
gain will be similar to the original system, since the ratio of the
magnitudes of the added pole to the added zero is approx. 1.
e.g. The following system is operating with a damping ratio of
0.174. Compensate the system to achieve estep (∞) = 0 while
maintaining the system’s current transient response.

We first consider the RL of the uncompensated system...

The current gain is obtained by searching along the radial line
corresponding to ζ = 0.174. We find the location of the dominant
poles to be at −0.694 ± j3.926 for a gain of K = 164.4. We find
the location of the third pole by searching to the left of -10 for a
point on the RL with this gain. We find the third pole at -11.61.
1 K
estep (∞) = Kp = lim = 8.23
1 + Kp s→0 (s + 1)(s + 2)(s + 10)

So estep (∞) = 1/(1 + 8.23) = 0.108.

We add an ideal integral compensator with a zero at -0.1.

This should have only a small effect on the system’s transient

response characteristics. Consider the RL...
The gain K corresponding to ζ = 0.174 is now 158.2. This slightly
shifts the locations of the dominant poles and the third pole. Also,
we now have a small section of the RL between the compensator’s
pole and zero. We search for it using the known value of K and
find it at -0.0902.
The system is now of type 1 and we can easily verify that its
steady-state error for step inputs is reduced to zero. The following
shows the time responses for the uncompensated and compensated
systems:
Ideal integral compensation is also known as PI Control. The
proportional and integral components can be shown as separate
transfer functions in parallel.

The equivalent transfer function of the compensator is

K2
K2 K1 (s + K1 )
Gc (s) = K1 + =
s s
The location of the zero can be adjusted by adjusting K2 /K1 .
Lag Compensation

Ideal integral compensation requires an active circuit for

implementation. If this is not desirable then instead of placing a
pole directly at the origin we can try placing it very nearby. Of
course, we also place a zero nearby so as to minimize the impact
on transient response. This is known as lag compensation.
The static error constant for the system is,
Kz1 z2 · · ·
Kv =
p1 p2 · · ·
With the lag compensator the static error constant becomes
Kzc z1 z2 · · ·
Kvc =
pc p1 p2 · · ·
zc
= Kv
pc

Thus, we get an increase in static error constant (decrease in

eramp (∞)) when zc /pc is large. Are we free to choose how we
make this ratio large?
No. If we want to maintain the same transient response we should
position the zero and pole close to each other so that the system’s
transient response is relatively unaffected.

To make zc /pc a large number our only option is to move both

close to the origin.
e.g. Compensate the system below (again) to improve the
steady-state error by a factor of 10 when the system is operating
with ζ = 0.174.

Previously we found e(∞) = 0.108 and Kp = 8.23 for the

uncompensated system. We now require ec (∞) = 0.0108. Thus,

1
e(∞)c = = 0.0108
1 + Kpc

We can solve for the required Kpc = 91.59. The ratio

Kpc /Kp = 11.13.
Since we previously found the following relationship,
zc
Kvc = Kv
pc

(The previously shown system was type 1 so we were discussing Kv

but in this example the system is type 0 so we are discussing Kp .)
zc /pc must equal the ratio Kpc /Kp . Arbitrarily choosing pc = 0.01
we find,
zc = 11.13pc = 0.111
The root locus for the compensated system is as follows:
The following time response shows that the transient response is
approximately the same as the compensated system, while the
steady-state error has been significantly reduced.
 2nd Order approximation
The next question is, How much farth er from the dominant poles does the third
pole have to be for its effect on the second-order response to be negligible? The
answer of course depends on the accuracy for which you are looking. However, this
book assumes that the exponential decay is negligible after five time constants. Thus,
if the real pole is five times farther to the left than the dominant poles, we assume
that the system is represented by its dominant second-order pair of poles.

See section 4.7 for more details

E1C04 11/03/2010 12:25:29 Page 190

190 Chapter 4 Time Response

the pure second-order system response; c3 ðtÞ, with a third pole close to the
dominant poles, yields the most error.
Students who are using MATLAB should now run ch4p2 in Appendix B.
You will learn how to generate a step response for a transfer
function and how to plot the response directly or collect the
points for future use. The example shows how to collect the points
and then use them to create a multiple plot, title the graph, and
labeltheaxesandcurvestoproducethegraphinFigure4.24tosolve
Example 4.8.
System responses can alternately be obtained using Simulink.
Simulink is a software package that is integrated with MATLAB
to provide a graphical user interface (GUI) for defining systems
and generating responses. The reader is encouraged to study
Appendix C, which contains a tutorial on Simulink as well as
some examples. One of the illustrative examples, Example C.1,
solves Example 4.8 using Simulink.
Another method to obtain systems responses is through the use of
MATLAB’s LTI Viewer. An advantage of the LTI Viewer is that it
displays the values of settling time, peak time, rise time, maximum
response,andthefinal valueonthestepresponseplot.Thereaderis
encouraged to study Appendix E at www.wiley.com/college/nise,
whichcontainsatutorialontheLTIVieweraswellassomeexamples.
Example E.1 solves Example 4.8 using the LTI Viewer.

Apago PDF Enhancer

Skill-Assessment Exercise 4.6

PROBLEM: Determine the validity of a second-order approximation for each of
TryIt 4.2 these two transfer functions:
Use the following MATLAB
and Control System Toolbox 700
statements to investigate the a. GðsÞ ¼
ðs þ 15Þðs2 þ 4s þ 100Þ
effect of the additional pole
in Skill-Assessment Exer-
cise 4.6(a). Move the higher- 360
order pole originally at 15 b. GðsÞ ¼
ðs þ 4Þðs2 þ 2s þ 90Þ
to other values by changing
‘‘a’’ in the code.
ANSWERS:
a=15
numga=100*a; a. The second-order approximation is valid.
denga=conv([1 a],...
[1 4 100]); b. The second-order approximation is not valid.
Ta=tf(numga,denga);
The complete solution is located at www.wiley.com/college/nise.
numg=100;
deng=[1 4 100];
T=tf (numg,deng);
step(Ta, ’. ’,T,’- ’)
 Our Goal in PD and Lead
Compensator

Midterm Examination 5 Spring Semester, 2015

Root Locus – Closed Loop Design using RL

Design Fixed!
Target!
C (s) G (s )
Controller Plant

• Place closed-loop poles at desired location

– by tuning the gain C(s) = K. (for time domain specs)
• If root locus does not pass the desired location, then
reshape the root locus
– by adding poles/zeros to C(s). (How?)
Compensation

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 61

Root Locus – Effect of Adding Poles
• Pulling root locus to the RIGHT
– Less stable
– Slow down the settling

Im Im Im

Re Re Re

Add a pole Add a pole

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 62

Root Locus – Effect of Adding Zeros
• Pulling root locus to the LEFT
Im
– More stable
– Speed up the settling

Re

Add a zero
Im Im Im

Re Re Re

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 63

 Ideal Derivative Compensation (PD)
Lead Compensation
PID Controller Design
Feedback Compensation
Physical Realization of Compensation

1 Ideal Derivative Compensation (PD)

1 Lead Compensation

1 PID Controller Design

1 Feedback Compensation

1 Physical Realization of Compensation

ENGI 5821 Unit 8: Design via Root Locus

Ideal Derivative Compensation (PD)

Generally, we want to speed up the transient response (decrease Ts

and Tp ). If we are lucky then a system’s desired transient response
lies on its RL. However, if no point on the RL corresponds to the
desired transient response then we must compensate the system. A
derivative compensator modifies the RL to go through the
desired point.
A derivative compensator adds a zero to the forward path.

Gc (s) = s + zc

Notice that this transfer function is the sum of a differentiator and

a pure gain. Thus, we refer to its use as PD control (proportional
+ derivative).
We consider various settings for zc when compensating the system
with the following RL:
zc = −4
zc = −3
zc = −2
Midterm EE 360 UMT School of Engineering/ Department of Electrical Engineering Spring 2015

Midterm Examination 5 Spring Semester, 2015

As the zero is moved we get changes in Ts and Tp . In this case,
when the zero is moved to −2 we get the fastest response. All the
while, we are maintaining %OS.

We show how to best place the zero by example...

e.g. Design an ideal derivative compensator for the following
system. The ideal transient response has 16% overshoot and a
threefold reduction in Ts .
The RL for the uncompensated system:

4 4
Ts = = = 3.320
ζωn 1.205
We desire Ts = 3.320/3 = 1.107 for the compensated system.
Thus, the real part of the compensated complex pole,
ζωn = 4/Ts = 4/1.107 = 3.613
The angle made with the positive real-axis must be the same as
before (120.26o ) to maintain 16% overshoot. Therefore we can
determine the imaginary part ωd by trigonometry.
ωd
tan(180o − 120.26o ) =
3.613
ωd = 3.613 tan(180o − 120.26o ) = 6.193
We must now solve for the zero that will place the desired point on
the new RL. At the desired point the sum of angles from the
open-loop poles is −275.6o . To achieve a point on the RL we
require a zero positioned so that the sum of angles equals an odd
multiple of 180o .
−275.6o + θzc = −180o
θz c = 94.6o
What is the coordinate of a zero that makes an angle of 95.6o with
the desired complex pole at −3.613 + j6.193?

6.193
tan(180o − 95.6o ) =
3.613 − σ
σd = 3.006
The RL for the compensated system is as follows:
Notice that the 2nd -order approximation is not as good for the
compensated system. We can determine from simulation that the
following quantities differ from their ideal values:
Ideal Simulated
%OS 16 11.8
Ts 1.107 1.2
Tp 0.507 0.5
A PD controller can be implemented in a similar manner to the PI
controller by placing the proportional and derivative compensators
in parallel:

The overall compensator transfer function is as follows:

K1
Gc (s) = K2 s + K1 = K2 (s + )
K2
Root Locus – Adding Poles/Zeros Remarks
• Adding only zero C ( s ) = K ( s + z ), ( z > 0)
– often problematic because such controller amplifies the
high-frequency noise.

• Adding only pole C ( s ) = K /( s + p ), ( p > 0)

– often problematic because such controller generates a
less stable system (by moving the closed-loop poles to
the right).

• These facts can be explained by using frequency response

analysis.

• Add both zero and pole!

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 64

Root Locus – Lead and Lag Compensators

C (s) G (s ) s+z
Controller Plant C (s) = K , ( z > 0, p > 0)
s+ p

• Lead compensator • Lag compensator

Im Im

Re Re

Why these are called “lead” and “lag”?

We will see that from frequency response in this class.

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 65

Root Locus – Lead Compensators
• Positive angle contribution
Test point Im

∠C Lead ( s ) = θ Lead > 0

Re
-p1 -z 1
s + z1
∠C Lead ( s ) = ∠ = ∠( s + z1 ) − ∠( s + p1 )
s + p1
= θ z − θ p = θ Lead > 0
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 66
Root Locus – Lag Compensators
• Negative angle contribution
Test point Im

∠C Lag ( s ) = θ Lag < 0

s

Re
-z 2 -p2
s + z2
∠C Lag ( s) = ∠ = ∠( s + z 2 ) − ∠( s + p2 )
s + p2
= θ z − θ p = θ Lag < 0
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 67
Root Locus – Rules of Lead/Lag Compensators
• Lead compensator
– Improve transient response
– Improve stability s + z1
C Lead ( s ) = K1
s + p1

• Lag compensator
– Reduce steady state error
s + z2
C Lag ( s ) = K 2
s + p2
• Lead-lag compensator
– Take into account all the above issues.
C LL ( s ) = C Lead ( s )C Lag ( s )
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 68
Root Locus – Example: Radar Tracking System

4
s(s + 2)

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 69

Root Locus – RTS: Lead Compensator Design (1)
• Consider a system

4 C (s) G (s )
G(s) = Controller Plant
s ( s + 2)

• Analysis of CL system for C(s) = 1 Desired pole

– Damping ratio ζ = 0.5 Im
– Undamped natural freq. ωn = 2 rad/s
2 3j
• Performance specification
– Damping ratio ζ = 0.5 Re
– Undamped natural freq. ωn = 4 rad/s CL pole with
C(s) = 1
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 70
Root Locus – RTS: Angle and Mag Conditions
• A point s to be on root locus
 it satisfies
– Angle condition

Odd number
o
∠G(s)H(s) =180 ×(2k +1), k = 0,±1,±2,...

• For a point on root locus, gain K is obtained by

– Magnitude condition
|G(s)H(s)|= 1
K

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 71

Root Locus – RTS: Lead Compensator Design (2)
Evaluate G(s) at the desired pole.
4 −1
G ( −2 + 2 3 j ) = =
( −2 + 2 3 j ) 2 3 j 3+ 3j
o If angle condition is satisfied,
compute the corresponding K.
Desired pole
o In this example, Im

2 3j
Angle condition is not satisfied.

∠G (−2 + 2 3 j ) = −210
Re
Angle deficiency φ = 30

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 72

Root Locus – RTS: Lead Compensator Design (3)
To compensate angle deficiency, design a lead compensator
s+z
C ( s) = K
s+ p
satisfying

Desired pole
∠C ( −2 + 2 3 j ) = 30( =: φ ) Im

2 3j
∠GC ( −2 + 2 3 j ) = −180
Re
There are many ways to design such C(s)!

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 73

Root Locus – RTS: Lead Compensator Design (4)
• Positive angle contribution
Test point Im

s
∠C Lead ( s ) = θ Lead > 0

• Triangle relations
Re
-p1 -z 1
θ p + θ Lead + (π − θ z ) = π

θ z − θ p = θ Lead

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 74

Root Locus – RTS: Lead Compensator Design (5)
How to select pole and zero:
• Draw horizontal line PA Desired pole Im
• Draw line PO
A P
• Draw bisector PB 2 3j
1
∠APB = ∠BPO = ∠APO
2

• Draw PC and PD Re
B O
C
φ D
∠CPB = ∠BPD = -p(=-5.4) -z(=-2.9)
2
• Pole and zero of C(s) are shown in the figure.

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 75

Root Locus – RTS: Lead Compensator Design (6)
Compensator realization:
• One example, using operational amplifiers

C1 C2
R4
R2
R1
R3
- -
vi (t ) + + vo (t )

Vo ( s )  R4  R2 R1C1 s + 1 
=  −  − ⋅ 
Vi ( s )  R3  R1 R2 C 2 s + 1 
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 76
Root Locus – RTS: Lead Compensator Design (7)
• Transfer function z
K

Vo ( s ) R2 R1C1 ( s + R 1C ) R4 R4C1 ( s + R 1C )
C ( s) = = ⋅ = ⋅ 1 1 1 1

Vi ( s ) R1 R2 C2 ( s + R C ) R3 R3C2 ( s + R 1C )
1
2 2 2 2

p
• Lead compensator • Lag compensator

1 Im 1 Im
− −
R2 C 2 R1C1
Re Re

1 1
− −
R1C1 R2 C 2

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 77

Root Locus – RTS: Lead Compensator Design (8)
System responses (uncompensated and compensated)
Compensated system Uncompensated system (C(s)=1)
1.4

1.2

0.8

0.6
Lead compensator gives
0.4 • faster transient response
(shorter rise and settling time)
0.2 • improved stability
0
0 1 2 3 4 5
2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 78
Root Locus – RTS: Lead Compensator Design (9)
Error constants (after lead compensation)
4 ( s + 2.9)
G ( s)C Lead ( s ) = ⋅ 4.675
s ( s + 2) ( s + 5.4)
5

• Step-error constant Unit ramp input

4

K p := lim G ( s )C Lead ( s ) = ∞ 3
s →0

• Ramp-error constant 2

Ramp response
K v := lim sG ( s )C Lead ( s ) = 5.02 1
s →0
0
0 1 2 3 4 5
NOT SATISFACTORY!

Lag compensator can reduce steady-state error.

2009 Spring ME451 - GGZ Week 10-11: Root Locus Page 79