COMPENSATOR DESIGN

1. Introduction:

Because of the prolonged use of the system, the parameters of the system can change and
output of the system may start deviating from the desired value. If the performance of a control system
is not upto the expectations as per the desired specifications, then some changes in the system are
required to obtain the desired performance. The changes can be in the form of adjustment of forward
path gain or feedback gain or inserting a compensating device (called compensator) in the control
systems.

The compensators are used to improve the performance (i.e. transient and steady state) characteristics
in the run-time. The nature of the compensation depends upon the given plant. The compensator may
be electrical, mechanical, hydraulic, pneumatic or any other type of device or network. Usually an
electrical network serves as compensator in many control systems. Among the many types of
compensators commonly employed are the lead compensators, lag compensators and lag-lead
compensators which are going to be discussed in later sections.

Control systems are always designed for a specific purpose. A good control system should
operate with as little error as possible. It should exhibit suitable damping i.e, the controlled output
should follow the changes in the reference input without unduly large oscillations or overshoots. It must
be able to mitigate the effect of undesirable disturbances. Its performance should not be affected by
small changes in certain parameters .the ideas presented here may be quite contradictory since increase
in the loop gain reduces the steady-state error but causes a deterioration in the transient performance
.by increasing the loop gain, we can increase the speed of response and reduce the steady-state error to
step a input, but we also tend to increase the increase the overshoot in the response and reduce the
damping ratio of the dominant poles, as described in the root locus method .

1.2. LINEAR SYSTEM DESIGN:

The design of the control system is the most important function which the control engineer
carries out. Every control system which has been designed for a specific application should meet certain
performance specifications.

a. One method of specifying the performance of a control system is by a set of specifications in

time domain or in a frequency domain such that peak overshoot, gain margin, phase margin,
settling time, steady-state error etc.
b. The other method is by optimality of a certain function.

There are always some constraints which are imposed on the Control system design in addition
to the performance specification. The choice of plant (collection of various devices in a system)
 is not only depending upon the performance specification but also on the size, weight, cost etc.
Although the designer of the Control system is free to choose a new plant, it is generally not
advised due to cost and other constraints. Under these circumstances, it is possible to introduce
some kind of corrective subsystems in order to force the chosen plant to meet the given
specifications. We refer to these subsystems as “compensators” whose job is to compensate for
the deficiency in the performance of the plant. For a given palnt, we design suitable
compensators so that the overall system would meet the given specifications.

1.3. INTRODUCTION TO COMPENSATING NETWORKS

A Control system usually requires some adjustment so that the various conflicting and
demanding specifications shall be met. The adjustment is called “compensation”, which can be
accomplished in several ways. For example, an additional component can be inserted in the forward
path, as shown in the fig 1.1.this is referred to as “cascade (or series) compensation”. The transfer
function of the compensator is denoted by 𝐺𝑐 (𝑠),whereas that of a original process or plant is denoted
by G(s). Alternatively, the compensator may be inserted in the feedback path, as shown in fig 1.2.this is
referred to as “feedback compensation”. A combination of these different two schemes is shown in fig
1.3.

C (s)
+- G (s)

_
-

_
Fig 1.1. Cascade compensation

R (s) C (s)
+_ +_

Fig 1.2. Feedback compensation

 +_ +_

Fig 1.3. Combined cascade and feedback compensation

1.4. Lead compensation

One of the simplest compensators has the first order transfer function

(𝑠+𝑧 ) (𝑠+1⁄ )
𝐺𝑐 (𝑠) = (𝑠+𝑝𝑐 ) = (𝑠+1⁄ 𝜏 ) ………………..(1.1)
𝑐 𝛼𝜏

𝑧
Where α=𝑝𝑐 < 1 , τ > 0
𝑐

𝜏𝑠+1
Thus 𝐺𝑐 (𝑠)=α (𝛼𝜏𝑠+1) ………………………….(1.2)

The s-plane representation of the lead compensator as shown in fig.1.4. Which has a zero at
s=−1 𝜏 and a pole at s=−1⁄𝛼𝜏 with a zero closer to the origin than the pole.
⁄

jω

𝑧𝑐 = −1⁄𝜏 σ

𝑝𝑐 = −1⁄𝛼𝜏

Fig.1.4. the s-plane representation of lead compensator

The lead compensator with the transfer function (1.1) can be realized by an electric lead network as
shown in fig 1.5

R2
R1 eo
ei

Fig.1.5. Electric lead network

Assuming that the circuit is not loaded, kcl for the output node yields,

𝑑(𝑒𝑖 −𝑒𝑜 ) 1 1
𝐶 𝑑𝑡
+𝑅 (𝑒𝑖 − 𝑒𝑜 )=𝑅 𝑒𝑜 ………………………………………(1.3)
1 2

Laplace transforming the equation (with zero initial conditions) gives

1 1
Cs[𝐸𝑖 (𝑠) − 𝐸𝑜 (𝑠)]+𝑅 [𝐸𝑖 (𝑠) − 𝐸𝑜 (𝑠)]= 𝑅 𝐸𝑜 (𝑠) …………………………(1.4)
1 2

The transfer function is

𝐸𝑜 (𝑠) 𝑅2 (𝑠+1⁄𝑅1 𝐶 )
𝐸𝑖 (𝑠)
=𝑅 = 1 …………………………………………..(1.5)
𝑅 ⁄𝑠𝐶 𝑠+
2+ 1 [𝑅2 ⁄𝑅1 +𝑅2 ]𝑅1 𝐶
𝑅1+ 1⁄𝑠𝐶

If we define

τ=𝑅1 𝐶

and α=𝑅2 ⁄(𝑅1 + 𝑅2 ) <1 ,

we recognize that the transfer function (1.5) has the same form as that of lead compensator in
equation.(1.1).

the sinusoidal transfer function of the lead network is given by

𝑗𝜔𝜏+1
𝐺𝑐 (𝑗𝜔)=α (𝑗𝜔𝛼𝜏+1) ;α<1……………….(1.6)

At zero frequency the network has a gain of α<1 or an attenuation of 1⁄𝛼 .it is easy to cancel the d.c
attenuation of the network with an amplification of A= 1⁄𝛼 . thus the lead compensator is visualized
as a combination of a network and an amplifier as shown in fig 1.6
 Fig 1.6. phase-lead network with amplifier.

The sinusoidal transfer function of the lead compensator is given by

𝑗𝜔𝜏+1
𝐺𝑐 (𝑗𝜔)=𝑗𝜔𝛼𝜏+1 ; α<1 ………………(1.7)

Since we have α<1 ,the network output leads the sinusoidal input under steady-state and so the name
lead compensator.
 Fig .1.7. Bode plot for lead network with amplifier of gain A= 1/α.

From equation(1.7),it is evident that the compensator provides a phase lead between the output and
the input ,given at any frequency,ω, given by

Ф(ω) =𝑡𝑎𝑛−1 (𝜔𝜏) - 𝑡𝑎𝑛−1 (𝛼𝜔𝜏) ……………………………(1.8)

𝜔𝜏(1−𝛼)
Or tan Ф = ……………………………… (1.9)
1+𝛼𝜔2 𝜏2

By making use of the condition ,

𝑑Ф(𝜔)
𝑑𝜔
=0 …………………………………..(1.10)

We find that maximum phase lead occurs at a frequency of

1
𝜔𝑚 =
𝜏√𝛼

1 1
=√(𝜏 ) (𝛼𝜏) …………………………….. (1.11)

From the equation (1.11) ,we see that 𝜔𝑚 is the geometric mean of the two corner frequencices of the
1 1
compensator viz..𝜔1 =(𝜏 ) and 𝜔2 = (𝛼𝜏).

The maximum phase lead 𝜙𝑚 at ω =𝜔𝑚 , is given by

Tan 𝜙𝑚 = (1 − 𝛼)⁄2√𝛼 …………………………………(1.12)

Sin 𝜙𝑚 = (1 − 𝛼)⁄(1 + 𝛼) ……………………………….. (1.13)

Thus we have

1−𝑠𝑖𝑛 𝜙
α = 1+𝑠𝑖𝑛 𝜙𝑚 ……………………………………….(1.14)
𝑚

1++𝑗𝜔𝑚 𝜏 1
At ω =𝜔𝑚 |𝐺𝑐 (𝑗𝜔𝑚 )| =| |= …………………………………….. (1.15)
1+𝑗𝜔𝑚 𝛼𝜏 √𝛼

Which the magnitude of 𝐺𝑐 (𝑗𝜔) at the frequency of maximum phase lead, useful in computing α from
the required value of maximum phase lead.

Good to know: For lead compensator, zero is dominant over pole. Thus, the effect of putting a lead
compensator is the addition of a zero to the system transfer function.

1.5. LAG COMPENSATION

 The general form of the transfer function of the lag compensator is
(𝑠+𝑧𝑐 ) (𝑠+1⁄𝜏)
𝐺𝑐 (𝑠) = = …………………………….. (1.16)
(𝑠+𝑝𝑐 ) (𝑠+1⁄𝛽𝜏)

𝑧
Where β =𝑝𝑐 > 1 , τ > 0
𝑐

−1
The s-plane representation of lag compensator is shown in fig.1.8,which has a poles at and a
𝛽𝜏
−1
zero at 𝜏
with the zero located to the left of the pole on the negative real axis.

The electric lag network is shown in fig 1.8

Fig 1.8 Electric lag network

Applying KVL for the loop yields the equation

1 𝑡
i𝑅1 +𝐶 ∫0 𝑖 𝑑𝑡 + 𝑖𝑅2 = 𝑒𝑖 …………………………..(1.17)

Laplace transforming (1.17), we have,

1
(𝑅1 + 𝑅2 + 𝐶𝑠)I(s) =𝐸𝑖 (𝑠) …………………………………….(1.18)

The output voltage is given by

1
𝐸0 (𝑠) = (𝑅2 + ) 𝐼(𝑠) ……………………………..(1.19)
𝐶𝑠
The transfer function of the lag network is therefore given by
1
𝐸0 (𝑠) 𝑅2 +
𝐶𝑠
𝐸𝑖 (𝑠)
= 1
𝑅1 +𝑅2 +
𝐶𝑠

1 𝑠+(1⁄𝑅2 𝐶 )
= 𝑅 +𝑅 [ 1 ] ………………….(1.20)
( 1 2) 𝑠+ 𝑅 +𝑅
1 2
𝑅2 ( )𝑅2 𝑐
𝑅2

From equation (1.16) and (1.17), we have

τ=𝑅2 𝐶

And β=𝑅1 +𝑅2 ⁄𝑅2 > 1

The transfer function of the network is

1 𝑠+1⁄𝜏 1 𝑠+𝑧
𝐺𝑐 (𝑠) = 𝛽 (𝑠+1⁄𝛽𝜏) =𝛽(𝑠+𝑝𝑐 ) ……………………..(1.21)
𝑐

𝜏𝑠+1
𝐺𝑐 (𝑠) = ……………….. (1.22)
𝛽𝜏𝑠+1

𝑧𝑐
Where 𝛽= >1
𝑝𝑐

The sinusoidal transfer function of the lag network is given by

(1+𝑗𝜔𝜏)
𝐺𝑐 (𝑗𝜔) = ………………………………… (1.23)
(1+𝑗𝛽𝜔𝜏)

The Bode plot of the lag network is shown in fig 1.9

From fig. 1.9, it is seen that the lag network has a D.C. gain of unity. It offers a high frequency
1
gain of 𝛽. Usually β is chosen to be greater than 10.
 m

1

Fig. 1.9. Bode plot for phase lag network

Good to know: For lag compensator, pole is dominant over zero. Thus, the effect of putting a lag
compensator is the addition of a pole to the system transfer function.

1.6. LAG-LEAD COMPENSATION

A combination of a lag compensator and a lead compensator is called a lag-lead compensator. The
transfer function of a lag-lead compensator is

𝑠+1⁄𝜏 𝑠+1⁄𝜏
𝐺𝑐 (𝑠) = (𝑠+1⁄𝛽𝜏1 )(𝑠+1⁄𝛼𝜏2 ) ……………………….(1.24)
𝑖 2

(lag) (lead)

where β >1 and α< 1.

The electric lag-lead network is shown in Fig 1.10.

 Fig. 1.10. Lag-lead compensator

Equating the currents at node A yields,

1 𝑑
[𝑒 − 𝑒0 ] + 𝐶1 (𝑒 − 𝑒0 ) = 𝑖 ……………………………..(1.25)
𝑅1 𝑖 𝑑𝑡 𝑖

The voltage 𝑒0 and the current I are related by

1 𝑡
∫ 𝑖
𝐶2 0
𝑑𝑡 + 𝑖𝑅2 = 𝑒 ………………………………(1.26)

Taking the laplace transform of these two equations (1.25 ) and (1.26) (with zero initial conditions) and
eliminating I(s) results in the equation

1 𝐸0 (𝑠)
[𝑅 + 𝐶1 𝑠] [𝐸𝑖 (𝑠) − 𝐸0 (𝑠)] = 1 …………..(1.27)
1 +𝑅2
𝑠𝐶2

The transfer function of the network is therefore,

𝐸0 (𝑠) 𝑅2 + 1⁄𝑠𝐶2
=[ ]
𝐸𝑖 (𝑠) 1 𝑅1 ⁄𝑠𝐶1
𝑅2 + 𝑠𝐶 +
2 𝑅1 + 1⁄𝑠𝐶1

1 1
(𝑠+ )(𝑠+ )
𝑅1 𝐶1 𝑅2 𝐶2
=[ 2 1 1 1 1 ] …..(1.28)
𝑠 +( + + )𝑠+( )
𝑅1 𝐶1 𝑅2 𝐶1 𝑅2 𝐶2 𝑅1 𝐶1 𝑅2 𝐶2
On comparing the equations (1.24) and (1.28) , we have

𝑅1 𝐶1 = 𝜏1 …………………………(1.29)

𝑅2 𝐶2 = 𝜏2 ………..(1.30)

𝑅1 𝑅2 𝐶1 𝐶2 = 𝛼𝛽𝜏1 𝜏2 …………………(1.31)
1 1 1 1 1
𝑅1 𝐶1
+𝑅 +𝑅 = 𝛽𝜏 + 𝛽𝜏 ………………………….(1.32)
2 𝐶1 2 𝐶2 1 2

It can be seen from equations (1.29) to ( 1.32) that

α β =1 ………………(1.33)

the s-plane representation of lag-lead compensator is given in fig 1.11

jω

S-plane

−1⁄𝜏2 −1⁄𝜏2

−𝛽⁄𝜏2 −1⁄𝛽𝜏1 σ

Fig .10.11. S-plane representation of the lag-lead compensator.

Bode plot of the lag-lead compensator is shown in fig .1.12

 Fig .1.12. Bode plot of lag-lead network

10.8. COMPENSATION USING BODE PLOTS

The procedure for to design lead compensation is as follows:

Step (1): Determine the value of K to satisfy the specified error constant.

Usually the error constants (𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎 ) and phase margin are the specifications given.

Step (2) : for this value of k,draw the Bode plot and determine the phase margin (Ф) of the system.

Sstep (3): if Ф𝑠 = specified phase margin

Ф = phase margin of uncompensated system (found out from the bode plot drawn)

ε= margin of safety( since cross over frequency may increase due to compensation)

phase lead required Ф𝑙 = Ф𝑠 − Ф + ε …………………….(1.34)

step (4): let Ф𝑚 = Ф𝑙

1−sin Ф
find α = 1+𝑠𝑖𝑛 Ф 𝑚 ……………….(1.35)
𝑚
if Ф𝑚 > 60𝑜 , two identical networks each contributing a maximum lead of Ф𝑙 ⁄2 are used.

Step (5) : find the frequency 𝜔𝑚 at which the uncompensated system will have a gain equal to
1
-10 log(𝛼), from the Bode plot drawn.

Take 𝜔𝑐2 = 𝜔𝑚 = cross-over frequency of compensated system.

Step (6): corner frequencies of the network are calculated as

1
𝜔1 = 𝜏
= 𝜔𝑚 √𝛼

1
𝜔2 = 𝛼𝜏 = 𝜔𝑚 ⁄√𝛼 …………………..(1.36)

Step (7): draw the magnitude and phase plots of compensated system and check the resulting phase
margin. If the margin is still low, raise the values of ε and repeat the procedure.

Effects of phase lead compensator:

1. The lead compensator adds a dominant zero and a pole. This increases the damping of
the closed loop system.
2. The increased damping means less overshoot, less rise time and less settling time. Thus
there is improvement in the transient response.
3. It improves the phase margin of the closed loop system.
4. The slope of the magnitude plot in Bode plot of the forward path transfer function is
reduced at the gain cross over frequency. This improves gain and phase margins
improving the relative stability.
5. It increases bandwidth of the closed loop system. More the bandwidth, faster is the
response.
6. The steady state error does not get affected.

Limitations of Phase-lead compensator:

1. If the system is unstable or has low stability margin, the additional phase lead required
to realize a certain desired phase margin may be excessive. This may require very small
value of  and may result in large bandwidth.
2. More bandwidth is sometimes not desirable due to the fact that the noise entering the
system at the input may become objectionable. It makes the system more sensitive to
the noise signals because of increase in the high frequency gain.
3. The compensated system may have a larger undershoot than overshoot. So tendency to
over compensate system may lead to a conditionally stable system.
 4. The maximum phase lead available from a single lead network is less than 60 o. Thus, if a
phase lead of more than 60o is required, a multistage lead compensators are required.

Example 1.1. The open loop transfer function of a ufb system is

𝑘
G(s) =
𝑠(𝑠+1)

It is desired to have the velocity error constant 𝑘𝑣 = 12 𝑠𝑒𝑐 −1 and phase margins as400 . Design
a lead compensator to meet the above specifications.
𝑘
Solution: Given G(s) =
𝑠(𝑠+1)

𝑘𝑣 = 12

𝜙𝑝𝑚 = 400

Step 1: For 𝑘𝑣 = 12, the value of k is chosen to be 12

K=12
12
Now, we draw the bode plot of the system G(s) = 𝑠(𝑠+1)

(i) For factor 12:

M = 20 log 12 = 21.58 db
𝜙1 = 00
(ii) For (jω)−1 :
M = -20 log ω db
𝜙2 = -900
(iii) For (1 + jω)−1 :
M = -20 log √12 + 𝜔 2

1 1
Break frequency 𝜔𝑏 = 𝑇 = 1 = 1 rad/sec 𝜙3 = -tan−1 𝜔
 ω 𝜙1 𝜙2 𝜙3 Ф=𝜙1 + 𝜙2 +𝜙3
0.1 0 -90 -5.71 -95.71
0.2 0 -90 -11.31 -101.31
0.5 0 -90 -26.56 -116.56
1 0 -90 -45 -135
2 0 -90 -63.43 -153.43
5 0 -90 -78.69 -168.69
10 0 -90 -84.29 -174.29
20 0 -90 -87.13 -177.13
50 0 -90 -88.85 -178.85
100 0 -90 -89.43 -179.43
500 0 -90 -89.89 -179.89
Step 2: From the Bode plot,

Phase margin = 15𝑜

Step 3: Ф𝑠 = specified phase margin =40𝑜

Ф = phase margi of uncompensated system = 15𝑜 (𝑓𝑜𝑢𝑛𝑑 𝑓𝑟𝑜𝑚 𝐵𝑜𝑑𝑒 𝑝𝑙𝑜𝑡)

ε =margin of safety = 5𝑜 (small value of 5𝑜 is chosen since the G(jω) is decreasing at

a slower rate since final slope of the log-magnitude curve is -40 db/dec only)

therefore, required phase lead Ф𝑙 = Ф𝑠 − Ф + 𝜀

= 40 -15 + 5

Ф𝑙 = 30𝑜

Step 4: Let Ф𝑚 = Ф𝑙

Ф𝑚 = 30𝑜
1−𝑠𝑖𝑛 Ф 1−sin 30 1−.0.5
α = 1+𝑠𝑖𝑛 Ф𝑚 =1+sin 30 = 1+0.5
=0.33
𝑚

step 5: to find the frequency𝜔𝑚 at which the uncompensated system will have a gain
1
= -10 log (𝛼)

1
Gain =-10 log ( ) =-4.8 db
0.33

From Bode plot. 𝜔𝑐2 = 4.6𝑟𝑎𝑑/𝑠𝑒𝑐 Corresponding to -4.8 db gain.

Step 6: 𝜔𝑐2 = 𝜔𝑚 =4.6 rad/sec .

1
Corner frequencies are 𝜔1 = 𝜏
=𝜔𝑚 √𝛼 = (4.6) √0.33 = 2.64 rad/sec
 1 𝜔𝑚 4.6
𝜔2 = = = = 8 𝑟𝑎𝑑/𝑠𝑒𝑐
𝛼𝜏 √𝛼 √0.33

1
𝑠+ 𝑠+2.64 2.64(1+𝑠⁄2.64)
𝜏
Step 7 : 𝐺𝑐 (𝑠) = 1 = 𝑠+8
= 8(1+𝑠⁄8)
𝑠+
𝛼𝜏

0.33 (1+0.378 𝑠)
𝐺𝑐 (𝑠) = (1+0.125 𝑠)

The amplification necessary to cancel the lead network attenuation of 0.33 is

1 1
A =𝛼 =0.33 =3

(1+0.378 𝑠)
𝐺𝑐 (𝑠) =(1+0.125 𝑠) (with amplifier A =3)

OLTF of the compensated system is

12 (1+0.378 𝑠)
G(s) =𝐺𝑓 (𝑠). 𝐺𝑐 (𝑠) =𝑠(𝑠+1)(0.125 𝑠+1)
 Fig 1.16 bode plot for 1.1

Design procedure for Lag compensation

Step 1: Determine the value of K to satisfify the specified error constant.

Step 2: For this value of K draw the Bode plot for uncompensated system . Find the phase margin for the
system from Bode plot.

Step 3: If Ф𝑠 = specified phase margin,

ε = margin of safety (5𝑜 𝑡𝑜 15𝑜 )

for suitable ε, find Ф2 =Ф𝑠 + ε

Step 4: find the frequency 𝜔𝑐2 where the uncompensated system makes a phase margin contribution of
Ф2 .

Step 5: measure the gain of uncompensated system at 𝜔𝑐2

Find β from the equation 𝜔𝑐2 = 20 log 𝛽

Step 6: choose the upper corner frequency (𝜔2 1⁄𝜏) of the network one octave to one decade below
𝜔𝑐2 .

1 𝜔 𝑐2 𝜔𝑐2
𝜔2 = = to
𝜏 2 10

Step 7: thus β and τ are determined which can be used to find the transfer function of Lag compensator.
1
1 𝑠+𝜏
𝐺𝑐 (𝑠) = [ ]
𝛽 𝑠+ 1
𝛽𝜏

Step 8: draw the Bode plot of the compensated system and check if the given specifications are met.

Effects and limitations of phase lag compensator:

1. Lag compensator permits the high gain at low frequencies therefore it acts as low pass
filter. Thus, it improves the steady state performance.
2. The gain crossover frequency is decreased and thus the bandwidth of the system is
reduced.
3. The response is slower due to reduced bandwidth. Thus the rise time and settling time
of the system are usually longer.
4. The transient response lasts for longer time.
5. The system is more sensitive to parameter variations.
6. Lag compensator is like PI (i.e. proportional plus integral) controller and tends to make
the system less stable.

𝐾
Example 1.2: For G(s) = . Design a Lag compensator, given P.M.≥ 35𝑜 and 𝐾𝑣 ≤ 20.
𝑠(𝑠+2)(𝑠+20)

𝐾𝑠
Solution: 𝐾𝑣 = lim 𝑠 𝐺(𝑠) = lim 𝑠(𝑠+2)(𝑠+20)
𝑠→0 𝑠→0

𝐾
= 20
40

K=800
800
G(s) = 𝑠(𝑠+2)(𝑠+20)
 800
= 𝑠 𝑠
40𝑠(1+ )(1+ )
2 20

20
G(s) = 𝑠 𝑠
𝑠(1+ )(1+ )
2 20

Table for construction of Bode plot

Factor corner frequency characteristics of curve phase angle

20⁄𝑠 none st.line of slope -20db/dec passing through Ф1 = -90𝑜
20 log 20=-26db/dec at ω = 1
(1 + 𝑠⁄2) 𝜔1 = 2 st.line of slope = 20 db/dec for ω> 𝜔1 Ф2 = −𝑡𝑎𝑛−1 𝜔⁄2

1 + 𝑠⁄2 0 𝜔2 = 20 st.line of slope -20 db/dec for ω> 𝜔2 Ф3 = −𝑡𝑎𝑛−1 𝜔⁄2 0

Phase angle

ω Ф1 Ф2 Ф3 Ф = Ф1 + Ф2 + Ф3
0.1 -90𝑜 –2.86𝑜 –0.286𝑜 –93.146𝑜
0.2 -90𝑜 −5.71𝑜 –0.573𝑜 -96.283𝑜
0.5 -90𝑜 −14.04 𝑜 – 1.43𝑜 -105.47𝑜
1 -90𝑜 −26.56 𝑜 -2.86𝑜 -119.42𝑜
2 -90𝑜 -45𝑜 -5.71𝑜 -140.71𝑜
5 -90𝑜 -68.20𝑜 -14.04𝑜 -172.24𝑜
6 -90𝑜 −71.56𝑜 -16.70𝑜 -178.26𝑜
7 -90𝑜 -74.05 𝑜 -19.29𝑜 -183.34𝑜
8 -90𝑜 -75.96 -21.80𝑜 -187.76𝑜
10 -90𝑜 -78.69𝑜 -26.57𝑜 -195.26𝑜

From the plot PM = 0

Ф2 = Ф𝑠 + 𝜀 = Ф𝑙 = 35𝑜 + 5𝑜 = 40𝑜

Required phase margin of 40𝑜 occurs at ω = 2 rad/sec.

So new gain cross over frequency – 2 rad/sec.

Zero of compensator is at 1⁄10th of gain cross-over frequency.

i.e., at 2⁄10 =0.2 rad/sec.

now , α |𝐺(𝑗𝜔)|𝑎𝑡 𝜔=2 = 1

800
α. =1
2.√22 +22 √22 +202

800
α .2∗2.828∗20.09 = 1

α=0.142
 1
now , = 0.2 ,αT = 5
𝛼𝑇

1
= 0.0284 T=35.21
𝑇

(1+𝛼𝑇𝑠) (1+5𝑠)
𝐺𝑐 (𝑠) = 𝐺𝑐 (𝑠) =
(1+𝑇𝑠) (1+35.21𝑠)

Transfer function of compensated system is

20 (1+5𝑠)
G(s) = 𝑠(1+𝑠⁄2)(1+𝑠⁄20) . (1+35.21𝑠)

20(1+𝑠⁄2)
= 𝑠(1+𝑠⁄0.02)(1+𝑠⁄2)(1+𝑠⁄20)

𝜔 𝜔 𝜔
∟G(s) = −90𝑜 + 𝑡𝑎𝑛−1 𝜔⁄0.2 - [𝑡𝑎𝑛−1 0.02 + 𝑡𝑎𝑛−1 2 + 𝑡𝑎𝑛−1 20]

=−90𝑜 + 84.29 –(89.43 + 45 +5.71)

=-145.85

PM = 180 -145.85 =34.15

Construction of Bode plot for compensated system

Factor corner frequency characteristics of curve phase angle

20⁄𝑠 none 20db/dec st. line passing through -90𝑜
20 log 20= 26db/dec at ω = 1
(1 + 𝑠⁄0.02) 𝜔1 = 0.02 -20 db/dec line for ω> 𝜔1 Ф1 = −𝑡𝑎𝑛−1 𝜔⁄0.02
(1 + 𝑠⁄0.2 ) 𝜔2 = 0.2 +20 db/dec line Ф2 = −𝑡𝑎𝑛−1 𝜔⁄0.2
(1 + 𝑠⁄2) 𝜔1 = 2 -20 db/dec line Ф3 = −𝑡𝑎𝑛−1 𝜔⁄2
(1 + 𝑠⁄20 ) 𝜔2 = 20 -20 db/dec line Ф4 = −𝑡𝑎𝑛−1 𝜔⁄20

Phase angle

ω Ф′ Ф1 Ф2 Ф3 Ф4 Ф= Ф′ + Ф1 + Ф2 + Ф3 + Ф4
0.1 -90𝑜 -78.69𝑜 +26.56𝑜 -2.86𝑜 -0.286𝑜 −145.28𝑜
0.2 −90𝑜 −84.29𝑜 +45𝑜 −5.71𝑜 -0.573𝑜 -135.57𝑜
0.5 −90𝑜 -87.70𝑜 +68.2𝑜 −14.04𝑜 −1.43𝑜 -124.97𝑜
1 -90𝑜 -88.35𝑜 + 78.69𝑜 -26.56𝑜 -2.86𝑜 −129.58𝑜
2 -90𝑜 -89.43𝑜 +84.29𝑜 -45𝑜 -5.71𝑜 -145.85𝑜
5 -90𝑜 -89.77𝑜 +87.71𝑜 -68.2𝑜 −14.04𝑜 −174.80𝑜
6 -90𝑜 -89.80𝑜 +88.09𝑜 −71.56𝑜 −16.70𝑜 −179.97𝑜
7 -90𝑜 -89.84𝑜 +88.36𝑜 −74.05𝑜 -19.29𝑜 -184.82𝑜
10 -90𝑜 -89.88𝑜 +88.85𝑜 -78.69𝑜 -26.57𝑜 -196.29𝑜
Fig 1.13(a) bode plot of example 1.2
 Fig 1.13(b) bode plot of example 1.2

Design procedure for Lag-Lead compensation

Step 1: For the specified error constant, determine the open-loop gain K

Step 2: Draw the magnitude and phase Bode plots of the uncompensated system with this value of K
and determine the phase margin and gain crossover frequency.

Step 3: First design the lag section to provide only partial compensation of phase margin. Choose the
gain crossover frequency such that it is higher than the gain crossover frequency if the system was fully
lag compensated.

Step 4: Determine the value of  required such that high frequency attenuation provided by the lag
network is equal to the magnitude of the uncompensated system at this frequency.

Step 5: Calculate the value of  1 such that the upper cut-off frequency of the lag network is two octaves
below the gain crossover frequency.
Step 6: Calculate the lower cut-off frequency, 1 = 1  1 . So the lag network design is complete. Find
its transfer function. Draw the magnitude and phase plots of the lag compensated system and
determine the gain crossover frequency and the phase margin from them.

Step 7: For the lead section design, independent value of  cannot be chosen. So select  = 1  and
calculate the maximum phase-lead provided by the lead section using the following formula

1 −   −1  1 − 1  
m = sin −1  = sin 1 + 1  
1 +    

Step 8: To fully utilize the lead effect, choose the compensated crossover frequency to coincide with  m
. This is the frequency where the lag section compensated system has a gain of −10 log  dB. Read this
from Bode plot. Then, m = 1  2  , So calculate  2 ,  2 and write the lead compensation transfer
function.

Step 9: Combine the transfer function of the lag and lead sections to get the lag-lead compensator
transfer function. Write the transfer function of the compensated system and draw the Bode plots for
that transfer function and determine the phase margin.

Step 10: If the specifications are not met, redesign the system by modifying the values of  and  .

Effects of Lag-Lead Compensator

Lag-lead compensator is used when both fast response and good accuracy are desired. Use of
lag-lead compensator increases the low frequency gain which improves the steady state. While
at the same time it increases bandwidth of the system, making the system response very fast.

In general, the phase lead portion of this compensator is used to achieve large bandwidth and
hence shorter rise time and settling time. While the phase lag portion provides the major
damping of the system.
𝐾
Example 1.3: For the unity feedback control system forward path transfer function G(s) = 𝑠(𝑠+4)(𝑠+20).
Design a lag- lead compensator so that PM ≥ 40 and St.S.error for unit ramp input ≤ 0.04 rad.
𝐾
Solution: G(s) =𝑠(𝑠+4)(𝑠+20)

𝐾
𝐾𝑣 = lim 𝑠 𝐺(𝑠) = 80
𝑠→0

80
𝐾
= 0.04 k=2000;

2000
G(s) = 𝑠(𝑠+4)(𝑠+20) ;
 2000
= 𝑠 𝑠
4∗20𝑠(1+ )(1+ )
4 20

25
G(s) = 𝑠 𝑠
𝑠(1+ )(1+ )
4 20

Table for construction of Bode plot

Factor corner frequency characteristics of curve phase angle

25⁄𝑠 none st. line of slope -20 db/sec Ф1 = −90𝑜
passing through 20 log 25=28 db at ω=1
𝜔
(1+𝑠⁄4) 𝜔1 =4 st.line of slope -20db/sec for ω>𝜔1 Ф2 = −𝑡𝑎𝑛−1 4
𝜔
(1+ 𝑠⁄20) 𝜔2 =20 st.line of slope -20 db/sec for ω>𝜔2 Ф3 = −𝑡𝑎𝑛−1 20
Phase angle

ω 𝜙1 𝜙1 𝜙1 Ф=𝜙1 + 𝜙2 +𝜙3
0.1 −90𝑜 −1.43𝑜 −0.29𝑜 −91.72𝑜
0.2 −90𝑜 −2.86𝑜 −0.57𝑜 −93.43𝑜
0.5 −90𝑜 −7.13𝑜 −1.48𝑜 −98.5𝑜
1 −90𝑜 −14.04𝑜 −2.86𝑜 −106.90𝑜
2 −90𝑜 −26.57𝑜 −5.71𝑜 −122.28𝑜
5 −90𝑜 −51.34𝑜 −14.04𝑜 −155.38𝑜
10 −90𝑜 −68.20𝑜 −26.57𝑜 −184.77𝑜
20 −90𝑜 −78.69𝑜 −45𝑜 −213.69𝑜
50 −90𝑜 −85.43𝑜 −68.20𝑜 −243.63𝑜
100 −90𝑜 −87.71𝑜 −78.69𝑜 −256.40𝑜

PM= −4.77𝑜

Phase angle to be contributed by the network = 40 -4.77+5 = −40.23𝑜

1+sin 40.23 1.6458

𝛼2 = = =4.65
1−sin 40.23 0.3541

2000
= √4.65 = 2.1563
𝜔0 √𝜔0 2 +4 √𝜔0 2 +400

𝜔0 √𝜔0 2 + 4 √𝜔0 2 + 400 = 927.5

𝜔0 = 6.5 rad/sec
 Fig 1.14 Bode plot of example of 1.3

1 1 14.02
Now,𝑇 = 𝜔0 √𝛼 =6.5 √4.65 = 14.02 𝛼2 𝑇2
= 4.65
= 3.02
2

1 1 1 1 1 0.302
Putting 𝛼 = 0.1 ; 𝛼 = 0.302 = 𝛼 *𝛼 = = 0.065
1 𝑇1 2 𝑇2 𝑇1 2 1 𝑇1 4.65

Thus the lag-lead compensator has the transfer function as

(1+𝛼1 𝑇1 𝑠)(1+𝛼2 𝑇2 𝑠) (1+0.302𝑠)(1+3.02𝑠)

G(s) = (1+𝑇1 𝑠)(1+𝑇2𝑠)
= (1+0.065𝑠)(1+14.02𝑠)
Comparison of Lead and Lag Compensators
Phase Lead Compensator Phase Lag Compensator
1. It improves the transient response 1. It improves the steady state response
2. It reduces the rise time 2. It increases the rise time
3. No significant reduction in steady state 3. It reduces the steady state error
error
4. It makes the system dynamic response 4. It makes the system dynamic response
faster slower
5. It increases the stability margin i.e. the 5. It reduces the stability margin i.e. the
system becomes more stable system becomes less stable
6. It increases the band width 6. It decreases the band width
7. High frequency noise in the output signal 7. High frequency noise in the output signal
increases or SNR decreases decreases or SNR increases
8. It can not be applied when phase angle of 8. It can not be applied if uncompensated
uncompensated system is increasing system’s phase angle in low frequency
rapidly near crossover frequency region is not sufficient to provide requisite
phase margin.

Good to know: A lead compensator acts like a high pass filter. A lag compensator acts like a
low pass filter. A lag-lead compensator acts like a band-pass filter.

Review Questions
1. Explain what is compensation and why is it required in control system.
2. Which are the important electrical networks used practically for the compensation of
the control systems?
3. Explain the steps to design the lead compensator using the Bode plot.
4. What is lead compensator? Obtain its transfer function and sketch its pole-zero plot.
Also sketch its Bode plot.
5. Write the procedure for the design of a lag compensator.
6. What is lag-lead compensator? Obtain its transfer function.
7. Explain the procedure to design the lag-lead compensator in frequency domain.