CHAPTER 22 CLOSED-LOOP CONTROLLERS

Controller switch positions

Temperature
Heater supply
Heater
supply Dead band
Time On
On

Heater supply
On Switch Switch
off on
Off Off
Off
Temperature Time Temperature
Controller Controller
switch point switch points
(a) (b) (c)

Figure 22.4 Two-step control.

elapses before the heater begins to have an effect on the room temperature. In the
meantime the temperature has fallen even more. The reverse situation occurs
when the temperature has risen to the required temperature. Since time elapses
before the control system reacts and switches the heater off, and yet more time
while the heater cools and stops heating the room, the room temperature goes
beyond the required value. The result is that the room temperature oscillates
above and below the required temperature (Figure 22.4(b)).
With the simple two-step system described above, there is the problem
that when the room temperature is hovering about the set value the ther-
mostat might be almost continually switching on or off, reacting to very
slight changes in temperature. This can be avoided if, instead of just a single
temperature value at which the controller switches the heater on or off, two
values are used and the heater is switched on at a lower temperature than
the one at which it is switched off (Figure 22.4(c)). The term dead band is
used for the values between the on and off values. A large dead band results
in large fluctuations of the temperature about the set temperature; a small
dead band will result in an increased frequency of switching. The bimetallic
element shown in Figure 2.46 has a permanent magnet for a switch contact;
this has the effect of producing a dead band.
Two-step control action tends to be used where changes are taking place
very slowly, i.e. with a process with a large capacitance. Thus, in the case of
heating a room, the effect on the room temperature of switching the heater
on or off is only a slow change. The result of this is an oscillation with a long
periodic time. Two-step control is thus not very precise, but it does involve
simple devices and is thus fairly cheap. On/off control is not restricted to
mechanical switches such as bimetallic strips or relays; rapid switching can be
achieved with the use of thyristor circuits (see Section 9.3.2); such a circuit
might be used for controlling the speed of a motor, and operational amplifiers.

22.4 Proportional
mode With the two-step method of control, the controller output is either an on or
an off signal, regardless of the magnitude of the error. With the proportional
mode, the size of the controller output is proportional to the size of the error:
the bigger the error, the bigger the output from the controller. This means
 22.4 PROPORTIONAL MODE

that the correction element of the control system, e.g. a valve, will receive a
signal which is proportional to the size of the correction required. Thus
controller output 5 KPe
where e is the error and KP a constant. Thus taking Laplace transforms,
controller output s 5 KPE s
and so KP is the transfer function of the controller.

A summing operational amplifier with an inverter can be used as a

proportional controller (Figure 22.5). For a summing amplifier we have (see
Section 3.2.3)

V0 Ve
Vout 5 2Rf 1
R2 R1

Figure 22.5 Proportional Summing amplifier

controller. R2
Inverter
R2 R
V0
− R
Ve −
R1 +
+
Vout

The input to the summing amplifier through R2 is the zero error voltage
value V0, i.e. the set value, and the input through R1 is the error signal Ve.
But when the feedback resistor Rf 5 R2, then the equation becomes
R2
Vout 5 2 Ve 2 V0
R1
If the output from the summing amplifier is then passed through an inverter,
i.e. an operational amplifier with a feedback resistance equal to the input
resistance, then
R2
Vout 5 V 1 V0
R1 e
Vout 5 KPVe 1 V0

where KP is the proportionality constant. The result is a proportional

controller.
As an illustration, Figure 22.6 shows an example of a proportional control
system for the control of the temperature of a liquid in a container as liquid
is pumped through it.
 CHAPTER 22 CLOSED-LOOP CONTROLLERS

Figure 22.6 Proportional Summing amplifier

controller for temperature R2
Voltage for R2
control. set point
Power amplifier
Temperature −
sensor
R1 +
Heater

With proportional control we have a gain element with transfer function KP

in series with the forward-path element G s (Figure 22.7). The error is thus

KPG s
E s 5 R s
1 1 KPG s

Figure 22.7 System with E(s) X(s)

Kp G(s)
proportional control. R(s)

Unity feedback

and so, for a step input, the steady-state error is

1 1
eSS 5 lim sE s 5 lim s
sS0 s S 0 1 1 1 KPG s s

This will have a finite value and so there is always a steady-state error. Low
values of KP give large steady-state errors but stable responses. High values
of KP give smaller steady-state errors but a greater tendency to instability.

22.5 Derivative
control With the derivative mode of control the controller output is proportional
to the rate of change with time of the error signal. This can be represented
by the equation
de
controller output 5 KD
dt
KD is the constant of proportionality. The transfer function is obtained by
taking Laplace transforms, thus

controller output s 5 KDsE s

Hence the transfer function is KDs.
With derivative control, as soon as the error signal begins to change,
there can be quite a large controller output since it is proportional to the
 22.5 DERIVATIVE CONTROL

rate of change of the error signal and not its value. Rapid initial responses
to error signals thus occur. Figure 22.8 shows the controller output that
Constant rate of results when there is a constant rate of change of error signal with time. The
change of error with time
controller output is constant because the rate of change is constant and occurs
Error

0
Time immediately the deviation occurs. Derivative controllers do not, however,
respond to steady-state error signals, since with a steady error the rate of
change of error with time is zero. Because of this, derivative control is always
combined with proportional control; the proportional part gives a response to
all error signals, including steady signals, while the derivative part responds to
the rate of change. Derivative action can also be a problem if the measurement
of the process variable gives a noisy signal, the rapid fluctuations of the noise
Controller output

resulting in outputs which will be seen by the controller as rapid changes in

error and so give rise to significant outputs from the controller.
Figure 22.9 shows the form of an electronic derivative controller circuit,
the circuit involving an operational amplifier connected as a differentiator
Time circuit followed by another operational amplifier connected as an inverter.
Figure 22.8 Derivative control.
The derivative time KD is R2C.

Figure 22.9 Derivative Differentiator Inverter

R2
controller.
R
R1
Ve − R
C −
+
+
Vout

Derivative control is never used alone because it is not capable of giving an

output when there is a steady error signal and so no correction is possible. It
is thus invariably used in conjunction with proportional control so that this
problem can be resolved.
With proportional plus derivative control the controller output is given by
de
controller output 5 KPe 1 KD
dt
KP is the proportionality constant and KD the derivative constant, de dt is
the rate of change of error. The system has a transfer function given by

controller output s 5 KPE s 1 KDsE s

Hence the transfer function is KP 1 KDs. This is often written as

1
transfer function 5 KD s 1
TD

where TD 5 KD KP and is called the derivative time constant.

 CHAPTER 22 CLOSED-LOOP CONTROLLERS

Figure 22.10 PD control.

Error
0
Time

Controller output
Proportional
element
Derivative
element

Time

Figure 22.10 shows how the controller output can vary when there is
a constantly changing error. There is an initial quick change in controller
output because of the derivative action followed by the gradual change due to
proportional action. This form of control can thus deal with fast process changes.

22.6 Integral control

The integral mode of control is one where the rate of change of the control
output I is proportional to the input error signal e:
dI
5 KIe
dt
KI is the constant of proportionality and has units of 1 s. Integrating the
above equation gives
Iout t

3I0 30
dI 5 KIe dt
t

30
KIe dt
Iout 2 I0 5

I0 is the controller output at zero time, Iout is the output at time t.

The transfer function is obtained by taking the Laplace transform. Thus
1
Iout 2 I0 s 5 KIE s
Error

0 s
Time
and so
1
transfer function 5 KI
s
Figure 22.11 illustrates the action of an integral controller when there is a
Controller output

constant error input to the controller. We can consider the graphs in two ways.
When the controller output is constant, the error is zero; when the controller
output varies at a constant rate, the error has a constant value. The alternative
way of considering the graphs is in terms of the area under the error graph:
0
Time t

30
Figure 22.11 Integral control. area under the error graph between t 5 0 and t 5 e dt
 22.6 INTEGRAL CONTROL

Thus up to the time when the error occurs the value of the integral is zero.
Hence Iout 5 I0. When the error occurs it maintains a constant value. Thus
the area under the graph is increasing as the time increases. Since the area
increases at a constant rate the controller output increases at a constant rate.
Figure 22.12 shows the form of the circuit used for an electronic integral
controller. It consists of an operational amplifier connected as an integrator
and followed by another operational amplifier connected as a summer to
add the integrator output to that of the controller output at zero time. KI is
1 RIC.

Figure 22.12 Integral controller. Integrator

Summing amplifier
R R
Ve − R

+ −
V0
R +
Vout

The integral mode of control is not usually used alone but is frequently used
in conjunction with the proportional mode. When integral action is added to
a proportional control system the controller output is given by

3
controller output 5 KPe 1 KI e dt

where KP is the proportional control constant, KI the integral control

constant and e the error e. The transfer function is thus
KI KP 1
transfer function 5 KP 1 5 s1
s s TI
where TI 5 KP KI and is the integral time constant.
Figure 22.13(a) shows how the system reacts when there is an abrupt
change to a constant error. The error gives rise to a proportional controller
output which remains constant since the error does not change. There is
then superimposed on this a steadily increasing controller output due to the
integral action. Figure 22.13(b) shows the effects of the proportional action
and the integral action if we create an error signal which is increased from the
zero value and then decreased back to it again. With proportional action alone
the controller mirrors the change and ends up back at its original set point
value. With the integral action the controller output increases in proportion
to the way the area under the error–time graph increases and since, even
when the error has reverted back to zero, there is still a value for the area,
there is a change in controller output which continues after the error has
ceased.
 CHAPTER 22 CLOSED-LOOP CONTROLLERS

Figure 22.13 PI control.

+ +

Error
Error
0 0
− Time − Time

Controller output

Controller output
0 Time Time
Effect of just the proportional action Effect of just the proportional action

Controller output
Controller output

0 Time
Time
Effect of just the integral action Effect of just the integral action

Elements
due to
Controller output
Controller output

0
Time Time
Effect of proportional + integral action Effect of proportional + integral action

(a) (b)

22.7 PID controller

Combining all three modes of control (proportional, integral and derivative)
gives a controller known as a three-mode controller or PID controller.
The equation describing its action can be written as

3
de
controller output 5 KPe 1 KI e dt 1 KD
dt

where KP is the proportionality constant, KI the integral constant and KD the

derivative constant. Taking the Laplace transform gives
1
controller output s 5 KPE s 1 KIE s 1 sKD s
s

and so

1 1
transfer function 5 KPe 1 KI 1 sKD 5 KP 1 1 1 TDs
s TIs
 22.8 DIGITAL CONTROLLERS

A three-mode controller can be produced by combining the various circuits

described earlier in this chapter for the separate proportional, derivative and
integral modes. A more practical controller can, however, be produced with
a single operational amplifier. Figure 22.14 shows one such circuit. The
proportional constant KP is RI R 1 RD , the derivative constant KD is
RDCD and the integral constant KI is 1 RICI.

Figure 22.14 PID circuit. RD CI RI

R
−
CD
+
Error
Output

22.8 Digital
controllers Figure 22.22 shows the basis of a direct digital control system that can be used
with a continuous process; the term direct digital control is used when
the digital controller, basically a microprocessor, is in control of the closed-
loop control system. The controller receives inputs from sensors, executes
control programs and provides the output to the correction elements. Such
controllers require inputs which are digital, process the information in digital
form and give an output in digital form. Since many control systems have
analogue measurements an analogue-to-digital converter (ADC) is used for
the inputs. A clock supplies a pulse at regular time intervals and dictates when
samples of the controlled variable are taken by the ADC. These samples are
then converted to digital signals which are compared by the microprocessor
with the set point value to give the error signal. The microprocessor can then
initiate a control mode to process the error signal and give a digital output.
The control mode used by the microprocessor is determined by the program
of instructions used by the microprocessor for processing the digital signals,
i.e. the software. The digital output, generally after processing by a digital-
to-analogue converter (DAC) since correcting elements generally require
analogue signals, can be used to initiate the correcting action.

Analogue Digital Digital Analogue Output

Micro- Correction
ADC DAC Process
processor element

Clock

Measurement

Figure 22.15 Digital closed-loop control system.