REED'S

INSTRUMENTATION AND
CONTROL SYSTEMS

By
LESLIE JACKSON
EXTRA FIRST CLASS ENGINEERS CERTIFICATE

B.Sc., C.Eng., F.I.Mar.E., F.R.I.N.A.

PUBLISHED BY THOMAS REED PUBLICATIONS LIMITED

36 COCK LANE, LONDON ECIA 9BY
AND SUNDERLAND

iloo.--_=-
 First Edition - 1970
Second Edition - 1975
Third Edition - 1979
Reprinted 1986

PREFACE

This book aims to bridge the gap between the mathematical

treatment often used by the specialist control engineer and the
(necessarily) narrow descriptive literature of a particular manu-
facturer.
It is written primarily for those with a good general engineer-
ing background who have had-little experience in instrumenta-
tion and control.
The work favours marine engineering but students and
engineers in other industries should find it a useful book as the
subject has a common basis. Text and examples cover the
requirements of all Department of Transport - and all
Technician Education Council (and SCOTEC), previously
ISBN 0 900335 53 X
COLI/OND - syllabuses and examinations for marine engineer
officers and cadets.
Full use has been made of simplified diagrams and the work is
presented from basic principles, using analogues where
appropriate.
The introduction is followed by five chapters on variable
measurement in instrumentation. Chapters 6-8 on telemetering,
electronic devices and final controlling elements link instrumen-
tation to control. Chapters 9-14 cover theory, practice and com-
© THOMAS REED PUBLICATIONS LIMITED ponents of process and kinetic control systems. Chapter 15 and
16 are intended to develop a broader knowledge of the subject
and, by necessity, have a more analytic and mathematical
approach.
A selection of test examples are included at the end of each
chapter and specimen examination questions are added at the
end of the book.
Printed by Thomas Reed Printers Limited If the reader wishes to obtain a full detailed description of a
at the Double Century Press, Sunderland particular component reference should be made to manufac-
turers' instruction manuals.
 ACKNOWLEDGEMENTS CONTENTS

The author wishes to extend thanks to the following for per- INTRODUCTION PAGE
mission generously given to use diagrams, information, etc.: Historical. Utilisation. Cybernetics. Economy.
Safety. Terminology. Comparison of pneumatic-
electronic-hydraulic systems. Open and closed loop.
Institute of Marine Engineers. System. Analogue. Digital. Computers. 1-10
Philips.
Drayton Controls.
Foxboro-Yoxall. CHAPTER 1 - TEMPERATURE MEASUREMENT
Bailey Meters and Controls. Mechanical thermometry; liquid in glass. Filled-
Kent Instruments. system for liquid, vapour, gas, mercury in steel. Bi-
Westinghouse Electric Corporation, Hagan Computer Sys- metallic type (Rototherm).
tems Division. Electrical thermometry; resistance thermometer,
British Standards Institution. Thermistor, thermocouple, electronic unit.
VAF Viscotherm.
Radiation pyrometry; optical pyrometer (disappear-
ing filament). 11-22

CHAPTER 2- PRESSURE MEASUREMENT

Water and mercury manometers. Mercury and
aneroid barometers. Bourdon and Schaffer pressure-
vacuum gauges. Differential pressure cell. Piezoelec-
tric crystals, sensor. Strain gauge pressure sensor. 23-32

CHAPTER 3- LEVEL MEASUREMENT

Direct methods; float operated, sight glasses, probe
elements (capacitive).
Inferential methods; pressure elements. Manometer
types, remote level, Electroflo, Igema. Purge systems,
pneumercator 33-40
CHAPTER 4- FLOW MEASUREMENT CHAPTER 7 - ELECTRONIC DEVICES
Basic principles; quantity meters, rate of flow met- Analogue and digital circuits.
ers, integrators, square root extraction. Semi-conductors; atomic theory, electron conduc-
Inferential-rotational; tlowmeters, mechanical, electri- tion, solid state junctions, rectifier and transistor.
cal, Rotameter. Rectifiers; bridge and centre tap transformer,
Inferential-differential pressure; primary elements, applications, carrier, modulation, d~-modulation,
orifice and venturi flow sensors. Secondary elements, smoothing, filter. Semi-conductor diode, zener diode,
square root extractors, mechanical, pneumatic, electri- characteristics, voltage stabilisation, applications.
cal. Thyristor (silicon controlled rectifier).
Flow sensor/transducer units. 41-54 Amplifiers; rotating electrical (metadyne), magnetic
(transductor), classification, junction transistor, circuit
configurations, transfer characteristics, voltage amp-
lifier, parameters, T-circuit, unijunction transistor
CHAPTER 5 - OTHER MEASUREMENTS (oscillator). Linear mode small signal transistor, power
Electrical. Speed-tachogenerator (analogue and amplifier (push-pull), dc amplifier, chopper, field
digital). Torque-power, torductor. Viscometer. Photo- effect transistor, operational amplifier. Feedback,
electric cells (conductive, vacuum, gas, transistor). negative, positive, gain, stability.
Oil in water sensor. Smoke density detector. Oil mist Oscillators; basic theory, transistor oscillator, har-
detector. pH sensor. Heat (fire) detector. Flame detec- monic oscillators, relaxation oscillators, push-pull
tor. Gas explosion-detector meter. Gas analysis, blocking, multivibrator, univibrator.
oxygen and CO 2 analysers. Relative humidity. Water Other devices; digital, switching, logic, analysers,
analysis, dionic water purity meter, dissolved oxygen cathode ray oscilloscope. Radio communication, elec-
meter. Instrument calibration, testing and adjustment. 55-74 tromagnetic radiation spectrum. 89-114

CHAPTER 6 - TELEMETERING
Pneumatic transmitters; position balance and force
,'(
balance transducers, electro-pneumatic transducer
(I/PP).
Electrical transmitters; variable contact resistance CHAPTER 8 - FINAL CONTROLLING ELEMENTS
transducer, other resistance systems, variable induc- Correcting units; diaphragm control valve, motor,
tance transducer - ratio and balance, variable capaci- correcting element (valve) characteristics, valve
tance transducer, electronic force balance system, vol- positioner. Piston operated, torque actuated, rotary
tage current transducer (V/I). cylinder and wax element control valves.
Receivers; receiver integrator, potentiometric pen Servo-motors; electric dc and ac motors, hydraulic
recorder, XY recorder, position motors (dc), position ram servo, variable delivery pump, hydraulic rotary
motors (ac), synchro, magslip, resolver. 75-88 vane servo, pneumatic piston servo, other servos. 115-126

::nm~IZ!UtlJllll;'l!H'::1!n~i::~~;~!~.i;;;';;-'
 Ii
~'
S.S. BEI(ULAN
OFF. No. a a 8 9-.
J
j

CHAPTER 9- PROCESS CONTROL THEORY

CHAPTER 13 - TYPICAL CONTROL CIRCUITS
Terminology; combination of elements. Steam; exhaust range, gland steam, sootblower, cir-
Response of detecting elements; lags. culating water, steam temperature, steam flow/air flow
Types of control action; two step, proportional, rate, bridge remote control (turbine).
bandwidth. Integral and proportional plus integral.
IC Engine; jacket (or piston) coolant (one and two
Derivative and proportional plus derivative. Three
element), overall coolant systems (ring main, series
term controller (level). Graphs of control actions. Cas-
and parallel), boiler oil purification system, waste heat
cade and split range control. Example -level control. 127-156
flash evaporator control, bridge remote control (IC
engine), outline and detail systems.

General; auto-combustion and attemperator, auxili-

ary automatic boiler, feed regulators, viscosity, refrig-
eration, alternator load sharing, Butterworth heating,
CHAPTER 10 - PNEUMATIC CONTROL PRINCIPLES oily-water separator interface level control, controll-
Two step, relay, one term (P), stack type (P), two able propeller pitch, auto steering (block diagram). 203-238
term (P + I, P + D), three term compound controller
(P + I + D). Air supplies. 157-170

CHAPTER 11 - ELECTRONIC CONTROL PRINCIPLES

Operational amplifier, inverter, summer, multiplier,
two step, one term (P, I), two term (P + D), three term
CHAPTER 14 - KINETIC CONTROL SYSTEMS
compound controller (P + I + D). Black box analysis.
Transistor stabilisation control. 171-184 Position systems; position control servo-mechanism
(dc), system response, damping, overshoot. Stabilising
feedback and rate network. Offset, reset feedforward
and network, block diagrams. Position control servo-
mechanism (ac), hydraulic position control servo-
mechanism.

CHAPTER 12 - ACTUAL CONTROLLER TYPES Speed systems; speed control servo-mechanism (dc),
Mechanical (Synchro). Mechanical-Hydraulic, governor. speed control (ac), Ward Leonard, Thyristor.
Electro-Pneumatic (Philips).
Electronic (Kent, Philips, Foxboro).
Pneumatic (Hagan, Foxboro, Drayton, Bailey (2), Governor systems; Mechanical-hydraulic and elec-
Viscotherm). 185-202 trical-hydraulic speed control servo-mechanisms. 239-254
CHAPTER 15 - CONTROL SYSTEM ANALYSIS
The systems approach; system, state, approach.
SPECIMEN EXAMINATION QUESTIONS 313
System order; analogues, generalised impedance.
First order systems, transfer function, closed loop
transfer function. Second order systems, higher order DoT Class Three 313
systems.
System performance; step, ramp and sinusoidal DoT Class Two 315
input response, Argand diagram. Frequency response
analysis (Nichols and Bode). Stability response, DoT Class One 321
Nyquist diagrams and criteria, Nichols chart. Further
analogues, second order, manufacturing system. TEC/SCOTEC (& METC, OND) 328
Component interaction; internal and external
(series). Higher; TEC/SCOTEC (& CGLI Advanced
Component adjustment; adjusting controllers to METC) 332
plant, empirical setting method. 255-280
Higher; TEC/SCOTEC (& OND Supplementary) 341

INDEX 351

CHAPTER 16 - LOGIC AND COMPUTING

Analogue, analogue computer; elements, control,
systems analysis and approach, model, block, pro-
gramme and circuit diagrams.
Logic circuits; digital algorithm, Boolean algebra,
switching and solid state logic, gates, AND, OR, NOT,
NAND, NOR, memory, delay, flip-flop, logic
sequence engine start. Fluidic logic.
Digital computer; scales of notation, binary scale.
Component units, input, output, memory, control;
peripheral units, reference sets, scanner, analogue-
digital converter, digital instrument.
Data processing; data logger, annunciator system,
integrated system, signal selection and processing,
output, display, control room.
Computer control; direct digital control, program-
mes, marine system, hybrid types. Computer simula-
tion. 281-312
 INTRODUCTION

HISTORICAL
Instrumentation has always been an integral part of technol-
ogy. Development from simple level indicators, Bourdon tubes,
moving iron and moving coil meters, etc. has been rapid. Prog-
ress in electrics and electronics has led to centralised recording
and display stations with associated data processing, computing
and control systems. Application to control with the requirement
of accurate measuring (sensors), variable converting (transduc-
ers) and remote signal transmission (telemetering) has involved
a close relation between measuring, processing and control sys-
tems. The advantage of electrical signal transmission is apparent
in the development of instrumentation. Modular designs and
interchangeable plug-in sub-assemblies have improved servicing
of electronic units. Digital operation is increasingly being
favoured over analogue operation.
The development of control elements is inherent in the history
of man as life form itself utilises classic control principles. The
Watt governor (1788) was one of the first practical applications.
Instability was recognised last century in hunting of steam engine
speed, and ship steering gears, and much analysis followed. The
ship steering gear remains as one of the first control systems and
its development alongside modern bridge equipment continues.
By 1939 fairly complex systems, mainly pneumatic and hyd-
raulic, were in use and development in electronics, related to the
feedback amplifier, has accelerated progress and now leads to
computer control. Systems are generally classified by their
field of operation. Process control such as flow, level and pres-
sure; kinetic control such as displacement, velocity and accelera-
tion; etc.
2 REED'S INSTRUMENTATION AND CONTROL SYSTEMS INTRODUCTION 3

UTILISATION
en
0..
Automation is essentially sequence-controlled mechanisation o
o
and is only an aspect of control. The word tends to varied usage UJ ...J
U
_ (fl .
and as such is not considered further in this text. American I-
U
0::
UJ
<i ...J
terminology tends to use the word cybernetics to describe the 0:: ...J
0..0
entire field of communication and control theory. A full study 0::
I-
Z
programme of the subject, which is generally followed in this u
UJ
o
u
...J
text, is shown in the first sketch. UJ

The degree of utilisation in marine practice varies a great deal.

Individual control loops, from simple to fairly complex, have
been in use for many years. Centralised data handling has been a
more recent innovation. Bridge control and unmanned machin-
ery spaces have developed quite rapidly with improved reliability
of control and alarm loops. The next stage of development is to
link the centralised data handling system to an integrated central I
control system. This requires that a computer is involved in the + w
interface between measurement and control. Computer control ~ IL «~
has developed most recently, from small programmed functions fi
I- <~
Z
0 . r,
I I
gj
0

U
to quite sophisticated direct digitally controlled processes. A z"w
.,8:::;u. I'"
wi a:
;i ~ ~
Q.
15L __.. , >
~z
modern computer can be so programmed not only to control ".
machinery under all conditions but also to have start up,
emergency and shut down procedures. Extension to naviga-
~gu ~z>
o~o
O::<i
W Z
","
UJ2:E
I "-I
I:E,
18,
I
ci
W
"
t;;
:::::>

>-"
CD <iz
:>0
I I C)
0
...J
tional, cargo handling, etc. easily follows. ...J
_1
~
U -
.£
0 u
I-UJ
L...J «...J
:::::>
LL
0
:>
o "
.. ..
"- Ii
ECONOMY u "0
Instrumentation and control results in more efficient opera- Oz
tion and reduced manpower in every case. There is an increased <i
u
en
first cost due to specialised equipment provision which leads to
increased insurance requirements and some increase in certain
running costs, e.g. staff training, skilled maintenance, etc. Over-
all running costs are reduced because·of large cost savings in fuel
and general maintenance, due to efficient operation and close
supervision, as well as staff reductions. The annual saving taking
all factors into account is well proven for controlled plant and the
factor increases with increasing size of plant and machinery. C)I~
z_z
-
0:: 0
UJ ~
I-
SAFETY
In most cases safety is improved by monitoring and con- ~I
UJ I-
U
...J
.
trol. Reliability of both measuring-control alarm equipment is , UJ UJ
I- I-
UJ
o
 4 REED'S INSTRUMENTATION AND CONTROL SYSTEMS INTRODUCTION 5

essential as indeed should be the characteristics of the plant 8. Local hand control of essential machinery.
itself. These aspects should be analysed closely at the design
stage. Unmanned machinery spaces (UMS) are now classified by 9. Adequate settling tank storage capacity.
surveying societies.
Essential requirements for unattended machinery spaces 10. Regular testing and maintenance of instrumentation.
could be summarised thus:
TERMINOLOGY
1. Bridge control of propulsion machinery. The detecting element responds directly to the value of the
The bridge watchkeeper must be able to take emergency con- variable. A measuring element responds to the signal from the
trol action. Control and instrumentation must be as simple as detecting element and gives a signal representing the variable
possible. value. For example pressure (variable), Bourdon tube (detecting
element) and linkage pointer, scale, i.e. pressure gauge (measur-
2. Centralised control and instruments are required in machinery ing element). The measuring unit comprises detecting element
space. and measuring element. Such a unit is used as a monitoring
Engineers may be called to the machinery space in emergency element (to convert, when necessary, the actual variable value
and controls must be easily reached and fully comprehensive. into a converted variable value) of a process control system.
Sensor (American) is a term used for the detecting element.
3. Automatic fire detection system. Is, by its very nature, essentially a transducer.
Alarm and detection system must operate very rapidly. Transducer (American) is a device to convert a signal (repres-
Numerous well sited and quick response detectors (sensors) enting a physical quantity) of one form into a corresponding
must be fitted. signal of another form, retaining the amplitude variations of
4. Fire extinguishing system. energy being converted. For example a microphone is a sound
In addition to conventional hand extinguishers a control fire transducer (acoustic to electrical) and a loudspeaker an electrical
station remote from the machinery space is essential. The station transducer (electrical to acoustic). A transducer may be an
must give control of emergency pumps, generators, valves, ven- integral part of the measuring unit, for example pressure to
tilators, extinguishing media, etc. displacement in a Bourdon pressure gauge. It may also be a
separate unit converter especially suitable to change the signal to
.J
5. Alarm system. a better form for remote transmissions, e.g. displacement to
A comprehensive machinery alarm system must be provided electrical in a differential transformer.
for control and accommodation areas. "I Telemetering may be defined as signal transmission over a
considerable distance. In measurement this involves information
6. Automatic bilge high level fluid alarms and pumping units. transfer from detecting element to a central recording-display
Sensing devices in bilges with alarms and hand or automatic station. In control this involves control operating devices and
pump cut in devices must be provided. related signal transfers. In telemetering systems the measuring
unit is often called the transmitter, usually incorporating a trans-
7. Automatic start emergency generator. ducer, and the recording unit some distance away is then referred
Such a generator is best connected to separate emergency bus .J to as the receiver which may have an associated transducer if
bars. The primary function is to give protection from electrical required.
blackout conditions. The terminology involved further to the above and especially
 6 REED'S INSTRUMENTATION AND CONTROL SYSTEMS INTRODUCTION 7

related to control systems is now fairly extensive. Such terminol- ~i able, simple and resistant to environment. A long commissioning
ogy is covered in some detail at the start of Chapter 9. time should be applied and regular skilled. maintenance is
Instrumentation in this book is generally confined to dynamic required. Signal dc transmission is usually preferred although ac
systems related to recording and control. Obviously the range is signals are essential for certain variables and easy amplification
much wider if extended to include static-laboratory type instru- of ac is an inherent advantage.
ment devices.
CONTROL LOOPS
COMPARISON OF SYSTEMS An open loop system has no feedback and controller action is
Systems, telemetering or control, may be either pneumatic, not related to final result. Consider a domestic central heating
hydraulic, or electronic-electric, or a combination. system as an example, in which fuel supply is varied manually or
Hydraulic systems are generally more restricted in application. automatically by external ambient temperature. Room tempera-
Basically the technique is as for pneumatics but fluid cannot be ture will be maintained at a reasonable value related to outside
allowed to escape and a recovery-storage system is required. conditions. However room temperature does not control fuel
General use is in the higher pressure range. supply so that this is open loop. The word loop is really a
A combination of electronic measure-record instrumentation misnomer.
and pneumatic final power control element is very effective. Now to the open loop shown, add a human operator, so closing
Controllers may be either pneumatic or electronic. The former the loop (dotted lines on sketch). This is a manually controlled
have generally been used because of proven reliability and ease (II closed loop system.
of application to final power transmission. Electronic controllers
POWER POWER
are increasingly being used and the all electronic-electric system
has many obvious advantages.
The advantages of a pneumatic system are:
1. Less expensive initially, this is in spite of tubing and air supply
,. 'f
costs. ,,"
I ,
\
2. No heat generation and safe in explosive atmosphere. I' ~----
I ...'"f'...
c" - " .J " INDICATOR RECORDER
3. Less susceptible to power supply variation, but do have I
I
appreciable time lags. : OPEN CLOSED
I ---

4. Direct application, without transducers, to large final power A

1\
actuators. .J
," \
\..

The advantages of an electronic system are: CONTROL LOOPS

1. Small and adaptable with cheap flexible transmission lines. The automatically controlled closed loop system is actuated by
2. No moving parts, can however generate heat. a signal dependent upon deviation (error) between input (set)
3. Stable, generally accurate and very short time lags. and output values. Deviation only exists when monitored output
(negative feedback) differs from input and this signal controls
4. Low power consumption, direct application to computers, but
power supplied to output. For a closed loop system, as sketched,
often need final element transducers.
output power is only controlled by, and not supplied by, the
In every case equipment must be robust, reliable, interchange- actuating signal. Closed loops have a self regulating property.

fJ~r91~T:·'~~~j#~····iffi¥;!~t
 8 REED'S INSTRUMENTATION AND CONTROL SYSTEMS INTRODUCTION 9

SYSTEM
i.e. Rate of change of variable x = -kx
An assembly of linked components within a boundary. The
dx
motor car is a good example; mechanical, electrical, control and - =-kx
dt
suspension sub-systems within a body-chassis boundary. A sys-
tem may have one input and a related output dependent on the A solution, where C is the x value when time is zero, is:
effect of that system (transfer function G). C
kt = In-
x
()o = Gfh
Voltage (or current) can readily be made analogous to physical
The boundary, represented as a 'black box', may include a com- phenomena. The traditional electric clock is analogue i.e. con-
plex system which need not be analysed if G is provided. More tinuous representation.
complex systems have interconnecting links to related systems.
A system must have input, process, output, and in most systems a DIGITAL
source of power and a means of control. A digital device manipulates "bits", i.e. discrete items of
information - illustrated by the digital clock representation.
States are on/off, equal/unequal, etc. and the binary digit system
is utilised.

q CONTROL IFEEDBACKI
,-----+-------1
I BOUNDARY
P
I
COMPUTERS
The electronic analogue computer is essentially a simulator on
which electrical analogues of various systems can be analysed
and illustrated. The digital computer is a machine for routine,
e. I G Ie
I PROCESS 0 repetitive arithmetic. Hybrid types are a combination.
INPUT 'BLACK BOX' OUTPUT

•I

d SYSTEM
I
POWER

o r-
\
\
~

ANALOGUE
Many different physical phenomena behave in a similar way,
i.e. are analogues of each other. Two examples are air escape
from a storage vessel and electrical charge loss from a capacitor.
Rate of change of pressure a pressure
Rate of change of charge a voltage
 CHAPTER 1

TEMPERATURE MEASUREMENT

This chapter is concerned with the practical aspects of ther-

mometry. No consideration is applied to absolute standards or to
the consideration of special techniques related to extreme temp-
eratures, etc.

MECHANICAL THERMOMETRY
LIQUID IN GLASS THERMOMETER
Mercury can be used from - 38°C (its freezing point) to about
600°C. For the higher temperatures an inert gas at high pressure
is introduced as the boiling point of mercury is about 360°C at
atmospheric pressure; special glass is also required.
Alcohol is used in the range -80°C to 70°C (or toluene) and
pentane can be used to -196°C.
Total immersion types are most accurate, especially when the
fluid is coloured and magnification is used. In many cases only
temperature differential is required so that relatively low accu-
racy partial immersion types are often satisfactory.

FILLED-SYSTEM THERMOMETER
Consist of liquid, vapour or gas filled types. All utilise a bulb,
connecting capillary and usually a bourdon tube mechanism,
responding to pressure change from volume variation (liquid),
for pointer or pen operation. Some systems incorporate a com-
pensating capillary and bourdon tube to allow for changes of
ambient temperature. Alternatively a bi-metallic link for com-
pensation can be incorporated into the mechanism.
 12 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TEMPERATURE MEASUREMENT 13

A common type of liquid filled system utilises mercury in steel POINTER

which can be pressurised for high temperature duty to 600°C.

Such devices are most useful for remote sensing and telemetering
back to a central instrumentation panel. Capillary bore is about
0·02 mm, the scale is generally linear but calibration must allow
for heat variation. Power is sufficient for pointer, pen or trans-
ducer operation.
SI-METAL COIL
Vapour pressure thermometers commonly use freon, alcohol
or ether which partly fills the system as liquid, and the remainder
is vapour filled. Measurement of vapour pressure gives an indica- Fig. 1.1 MECHANICAL TYPE THERMOMETER (BI-METALLIC TYPE)
tion of liquid surface temperature and is usually used in the range
-50°C to 260°C, with the upper limit fixed by the critical temp-
erature of the liquid which must have a low boiling point (high ELECTRICAL THERMOMETRY
vapour pressure). The scale is non-linear, ambient variations can
RESISTANCE THERMOMETER
be neglected but there can be appreciable time lags and the
The electrical resistance of a metal varies with temperature
device is not well suited to remote indication.
and this relationship is usually expressed as P8 = Po(1 + a(J)
where p8 is the specific resistance at temperature (J°C
Gas filled thermometers usually employ nitrogen or helium
po is the specific resistance at temperature O°C
under high pressure, and pressure is proportional to absolute
a is a constant which depends upon the metal and is called
temperature at constant volume. The usual temperature range is
the temperature coefficient of resistance.
-50°C to 430°C and the scale is linear. Compensation for
Figure 1.2 shows diagrammatically a resistance type of temp-
ambient temperature variation is difficult. When used as a sensor
erature measuring unit using the well known Wheatstone bridge
linked to a pneumatic transducer it is a very effective device.
principle. '1'2 is a variable resistance used for balance purposes;
at balance we have:

Rl + rl R3+'
BI-METALLIC THERMOMETER
R2 + r2 R4+r
The principle of operation of bi-metallic devices is that of
differential expansion of two different materials rigidly joined r is the resistance of each of the wires and since each wire will be
together, one on the other, as a strip of bi-metallic material. subjected to the same temperature variation along its length
Figure 1.1 illustrates a typical design usually employed between their resistances will always be equal.
-40°C and 320°C. Invar (36% Ni, 64% Fe) has a low coefficient When the temperature detecting element is subjected to a
of expansion and when welded to a Ni-Mo alloy gives a good temperature alteration its resistance alters and the bridge bal-
bi-metallic strip. The helix coils or uncoils with temperature ance is upset. By using the variable resistor '1'2 balance can be
variation and as one end is fixed the movement rotates shaft and restored (i.e. galvanometer reading returned to zero) and whilst
pointer. The range of the instrument is fixed by the materials this is being done another pointer can be moved simultaneously
used (Rototherm). and automatically to give the temperature - this is known as the

i; ;,.!::"~t:-~~~~~"\t."j~:<'
 TEMPERATURE MEASUREMENT 15
14 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

THERMISTOR (THERMALLY SENSITIVE RESISTOR)

These devices are among a second class of resistance ther-
mometer utilising elements made of semi-conducting material all
r, of which have a characteristic of a resistance decrease with
= ~ 4 temperature increase. Included in this category are carbon resis-
/
.1"' R4
r2
. tors and doped germanium units (see Fig. 1.3).

DETECTING
ELEMENT

PORCE~AIN SHEATH / f
:
iii

rrj~=-==-:===~iof-;---!
..J
<
U
(/)

Lf-------------I f C>

PLATINUM WIRES WOUND LOW RESISTANCE ~

ON CERAMIC FORMER COPPER WIRES UJ
U
Z
Fig.1.2 RESISTANCE THERMOMETER ~
(/)
in
UJ
0:::
U
null balance method. Alternatively, the galvanometer can give u::
the temperature reading directly, in this case no variable resis- ~I
(/)
~ETALLIC CONDUCTORS

tance '1'2 would be required. -100

400
For the measurement of ambient temperature conditions the TEMPERATURE °C ----.
resistances, apart from the temperature measuring resistance, Fig.1.3 RESISTANCE-TEMPERATURE CHARACTERISTICS
would have to be made of a metal whose resistance does not vary
with temperature. A metal which nearly fulfills this requirement Thermistors are made of semi-conducting materials, they are
is constantan. manufactured by sintering (i.e. heating under pressure) powder
The term pyrometer is used for temperature measuring ins- mixtures of metallic oxides such as manganese, nickel, cobalt,
truments operating above 500°C, the term thermometer being copper, iron or uranium. The size and configuration can be
used for operation below 500°C. controlled so that rods, beads, discs and washer shapes can be
Resistance thermometers can be exceedingly accurate. produced as desired. Figure 1.4 shows an ellipsoid of thermistor
Platinum is the most suitable sensing wire element but copper material, wires about 0·25 mm apart are firmly embedded in the
and nickel wire are used in the range -100°C to 200°C and material making good electrical contact. The whole assembly
tungsten, molybdenum and tantalum are used to 1200°C, in may be coated with glass to give strength and protection.
protective atmospheres. The platinum element usually has a A washer shape of thermistor is particularly useful, it could be
resistance of 100 ohms at O°C in which case resistance of wires is fitted over bolts or studs. Several washers could be used together
limited to about 3 ohms. Use up to 600°C with twin wires is often with the terminals connected in series or parallel arrangements.
acceptable with the three wire method used for higher accuracy;, Beads or rods of thermistor material are suitable for use as
measurement is by Wheatstone, Kelvin or Mueller bridges or
probes.
potentiometric methods.

~"
16 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TEMPERATURE MEASUREMENT 17

SINTERED increases and this enables them to move from inner tightly bound
THERMISTOR MATERIAL
orbitals to the outer conduction bands of the atom. With more
EMBEDDED IN
THERMISTOR MATERIAL free electrons available to carry current the resistance to current
flow reduces.
In metals, where there are many free electrons, increase in
temperature leads to "traffic jams" of electrons hence the resis-
tance to current flow increases with increase of temperature.
The very high sensitivity, small size, high energy and rapid
PROTECTIVE response makes thermistors very useful detecting devices par-
GLASS COATING ticularly for use in computers and scanner installations. Range is
Fig.1.4 THERMISTOR BEAD -100°C to 300°C but special compositions can extend this much
further, as high as 1600°C. The small thermal mass can lead to
Advantages of thermistors are: self heating and coupled with high sensitivity and exponential
characteristic means instability must be carefully watched. The
(1) Relatively small and compact, the bead arrangement
thermistor merely replaces the resistance element on one limb of
shown could have a diameter up to 2·5 mm with a resistance up to
a bridge circuit in the measuring unit.
about 100 megohms.
(2) Low specific heat, hence the thermistor does not take very THERMOCOUPLE
much heat away. Whenever the junctions formed of two dissimilar homogene-
ous materials are exposed to a temperature difference, an emf
(3) Physically strong and rugged.
will be generated which is dependent on that temperature differ-
(4) Relatively high temperature coefficient of resistance, it ence, also on the temperature level and the materials involved.
could be as high as ten times that of some metals. This thermoelectric emf is called Seebeck effect (Seebeck disco-
vered in 1821) and is an algebraic sum of two other effects
(5) They can be used for extremely low temperature meas-
discovered by Peltier and Thomson. The two materials, usually
urement with great accuracy.
metals, form the thermocouple.
The mathematical relationship for thermistors is given by Figure 1.5 shows a thermocouple consisting of two wires, one
Po = poef3 (1/8 - 1/80) iron, one constantan (i.e. a copper nickel alloy), with a milli-
voltmeter coupled to the copper wire. If the junctions A and B
where; are maintained at the same temperature no current will flow
Po is the specific resistance at temperature {} around the circuit since the emfs in the circuit will be equal and
po is the specific resistance at temperature {}o opposite. If however A is heated to a higher temperature than B
f3 is a constant which depends upon the material used in the then current will flow since the emf at one junction will be greater
construction (f3"" 4000). The characteristic is shown in Fig 1.3. than the opposing emf at the other junction.
The large negative temperature coefficient of resistance of A third wire can be introduced as shown in Fig. 1.5, where AB
thermistors may be explained by considering the number of and AC form the couple wires. Providing the junctions Band C
electrons available for carrying current. are maintained at the same temperature, the introduction of the
Few electrons are available at low temperatures but as the third wire BC will not affect the emf generated. Hence A will be
temperature increases the kinetic energy of the electrons the hot junction and B with C will form the cold junction. Couple

m-i1,;...~h:"';:·Y;;~'ijji
18 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TEMPERATURE MEASUREMENT 19

wires AB and AC shown as iron and constantan respectively can In a practical thermocouple system the cold junction Band C
be made of various metals and alloys, choice depends upon may be at a relatively high temperature due to the environment.
temperature of operation, the wire BC would generally be long This would mean that the temperature difference between the
compared to the couple wires and could be made of copper. hot and cold junctions would be small and mV similarly. The
Figure 1.5 also shows the device in detail. indicator itself could then become the cold junction if the wires
from terminals Band C to the indicator are of the same material,
or material with similar characteristics, to the couple wires. The
wires are then called compensating wires and the col<:i junction
temperature would be reasonably constant if the indicator is
within an air conditioned control room or immersed in a block of
metal of large thermal capacity. Alternatively cold junction com-
pensation signal by separate means or bi-metallic instrument
components is arranged.
A copper (+) constantan (-) couple is used up to about
350°C, constantan being a 40% Ni 60% Cu alloy. Up to 850°C an
iron-constantan couple is used with a chromel (90% Ni 10% Cr)
and alumel (94% Ni 2% AI) couple up to 1200°C. Average emfis
0·05 mV/ °c which compares with about 18 mVrc for a thermis-
tor. Platinum-platinum plus 10% rhodium couples have been
used to 1400°C.
The emf generated is usually given by an expression of the
form:
A e = A + BO + C0 2 + D0 3 + ...
where 0 is the temperature and A, B, C, D are constants of
diminishing order. Fig. 1.6 shows an electronic thermocouple
with operational amplifier (A) giving 0·1 vrc. Y is for calibra-
tion at cold junction temperature and X for other temperature
calibration.

-6V••- - - - - - - - - .

x
y

+6V. '

Fig.1.5 THERMOCOUPLE Fig. 1.6 ELECTRONIC THERMOCOUPLE

 20 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TEMPERATURE MEASUREMENT 21

RADIATION PYROMETRY Note:

Many of the measuring devices for temperature considered
When temperatures are above the practical range of ther-
previously, particularly electrical types, could also be classified
mocouples, or the "target" is not accessible, or an average temp-
as telemetering or transmitting units as the signal is readily
erature of a large surface is required then radiation pyrometers
conveyed over a considerable distance to a remote measuring,
are used. Theory is generally based on black body radiation and
recording or display station. This applies in many cases to other
the work of Stefan"Boltzmann and Planck with amended factors
such devices in the following chapters. The detecting element
of emissivity to allow for variation from the ideal black body
(sensor) is inherently a transducer in operation for many instru-
radiator. Types of radiation pyrometer are optical, radiation and
two colour. The former will be considered. ment units.

OPTICAL PYROMETER
TEST EXAMPLES 1
Referring t<:> Fig. 1.7: S is the source and rays enter lamp box L
after passing through the lens, aperture and absorption filter.
1. Describe, with the aid of simple sketches, three types of
The lamp is electric and current and voltage are measured at G.
temperature measuring device. State how they are graduated
Rays leaving L pass through a red filter, lens and aperture to eye
E. and where they are used.
2. Sketch and describe an electrical instrument used for read-
ing temperature at a remote distance. Stattt the usual tempera-
ture range and the materials used in construction.
E

t 3. Give a reasoned explanation of how Wheatstone Bridge

networks are employed in circuits of electrical resistance ther-
mometers, explosive gas sampling devices or any similar applica-
tion. Sketch the circuit for such a device indicating the function of
the Wheatstone Bridge.
Fig.1.7 OPTICAL PYROMETER

The device is often known as a disappearing filament unit.

Both source and reference, the latter being the filament of a
small vacuum lamp, are observed through the microscope. The
power to the lamp is adjusted until the reference source just
disappears into the main source. Power is calibrated to give a
temperature reading directly. Correction factors apply for the
filters used and the device is a selective radiation pyrometer as
only a narrow band of radiation wavelength is utilised. Radiation
pyrometry is particularly useful for furnace, molten metal, pro-
cess control, etc. evaluation of temperature.
 l"~ r-.
')00.

CHAPTER 2

PRESSURE MEASUREMENT

WATER MANOMETER
This instrument is used for measuring pressures of a low order
such as fan pressures, etc. Fig. 2.1 illustrates a U tube water
manometer, one limb of which is connected to the system whose
pressure is to be measured, the other limb is open to the atmos-
phere. The pressure reading is the difference of the water levels
read from the scale.

!
PRESSURE
SOURCE

n ! ATMOSPHERE

II II r--
40

30 ...,

20 -l

10 -l
~lomm
1-

Fig. 2.1 WATER MANOMETER (U TUBE)

24 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PRESSURE MEASUREMENT 25

Note: 1 m 3 of fresh water has a mass of 1 Mg and weighs 103 x This instrument measures pressures of a higher order than that
9·81 N. measured by the water manometer, such as scavenge or super-
Hence: charge air pressure for Ie engines. The uniform bore glass tube is
1 m of fresh water exerts a pressure of 103 x 9·81 N/m2 , 1 mm small in diameter so that when mercury is displaced from the well
of fresh water exerts a pressure of 9·81 N/m2 , the reading indi- into the tube, the fall in level of the mercury in the well is so small
cated in Fig. 2.1 is therefore equivalent to 20 x 9·81 = 186·2 it can be neglected. Hence the pressure reading is indicated
N/m2 , i.e. 0·1862 kN/m 2 , above atmospheric pressure (1 bar = directly by the level of the mercury in the glass tube. The relative
105 Pa = 105 N/m 2 ~ 1 atm). A wide cistern manometer (well density of mercury is 13·6 hence 1 mm of mercury is equivalent
2
type) is used for lower pressure differentials (Fig. 2.2 as mercury toapressureof9·81 x 13·6N/m2 ,i.e. 134N/m2 0rO·134kN/m •
manometer). A variation is the inclined tube manometer - the A special application is the vacuum gauge (kenotometer) which
small bore limb is set at a small angle to the horizontal and the is a combined barometer and manometer with the scale on the
longer scale parallel to it gives even smaller differential pressure right hand side calibrated in absolute pressure.
readings.

MERCURY MANOMETER MERCURY BAROMETER

A mercury manometer of the well type is shown in Fig. 2.2. A sketch would be similar to Fig. 2.2 but with the top of the
glass tube sealed at a vacuum and the pressure source would be
the atmosphere.
tTO
ATMOSPHERE
_ _ RETAINING
If we assume the atmospheric pressure is supporting a column
PLUG of mercury 760 mm in the tube, then:
atmospheric pressure = 760 x 0·134
2
I" I SCALE = 102 kN/m
2
= 1·02 x 105 N/m
GLASS
TUBE = 1·02 bar
For an atmospheric pressure of 760 mm the equivalent water
MERCURY barometer would be 760 x 13·6, i.e. 10336 mm or 10·336 m.
The value of the atmospheric pressure varies slightly with
PLUG
,
FILLING climatic conditions hence to ascertain true absolute pressures the
barometer reading should be taken at the same time as a gauge
pressure is taken.
If for example we wished to obtain the absolute pressure in a
condenser and the readings were:
condenser gauge reading 742 mm
barometer gauge reading 762 mm
t
PRESSURE
SOURCE
then, condenser pressure = (762 - 742) x 0·134
Fig.2.2 MERCURY MANOMETER (WELL TYPE) = 2·68 kN/m2
 26 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
PRESSURE MEASUREMENT 27

Materials used in the construction of the gauge are solid drawn

phosphor bronze or stainless steel for the pressure relay tube.
DIAL Bronze or stainless steel for the quadrant, gear and linkage.
, -- I Case, brass or plastic.
] DRUM --'
FINE CHAIN

BELL
--CRANK

CORRUGATED CYLINDER
(EVACUATED)

Fig. 2.3 ANEROID BAROMETER

ANEROID BAROMETER
The aneroid barometer is shown in Fig. 2.3. It consists of a
corrugated cylinder (detecting element) made of phosphor
bronze or other similar material, a steel spring, bell crank,
pointer, dial and case (measuring element). The corrugated
cylinder is completely evacuated hence the pressure of the
atmosphere tends to collapse it. The centre of the corrugated
cylinder deflects downwards if atmospheric pressure increases i PRESSURE
SOURCE

and the spring causes deflection upwards if atmospheric pressure

decreases. Cylinder motion is transmitted to the instrument Fig.2.4 PRESSURE (OR VACUUM) GAUGE - BOURDON
pointer.

PRESSURE GAUGE (BOURDON)

The Bourdon movement is frequently used in transducers and
A pressure relay tube is the principal working component controllers to vary output signals in pneumatic or electrical form.
(detecting element). This tube which is semi-elliptical in cross
section is connected to the pressure source. When the tube is
PRESSURE GAUGE (SCHAFFER)
subjected to a pressure increase it tends to unwind or straighten This type utilises a strong flexible metal diaphragm (detecting
out and the motion is transmitted to the gauge pointer through element) which moves up as pressure increases. The device is
the linkage, quadrant and gear (measuring element). If the tube shown in Fig. 2.5. Again this device can also be used as a trans-
is subjected to a pressure decrease it winds, or coils, up and the ducer (pneumatic or electric) in telemetering or control with an
motion is again transmitted to the pointer. This gauge is there- output signal proportional to diaphragm movement. Similar
fore suitable for measuring pressures above or below atmos- remarks apply to most detecting (sensing) devices (as detailed in
pheric pressure. A diagrammatic sketch is shown in Fig. 2.4.
Chapter 6).
 28 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
PRESSURE MEASUREMENT 29

As an alternative to a one membrane diaphragm a sealed

capsule (twin membrane) can be inserted in the cell body and
secured so that different pressures are applied at each side. The
capsule can be filled with a constant viscosity fluid (for a fairly
wide temperature range) which also damps oscillation. Silicone is
such a fluid. Again mechanical movement of the capsule is prop-
ortional to differential pressure.
Another type of dIp cell utilises two separate bellows. Such a
design, often called a Barton cell, is shown in Fig. 2.6. Refer now
to this sketch.
Pressure increase on the high pressure side displaces liquid
(via an adjustable damping plug) and expands the right hand low
pressure bellows. This bellows is connected to the horizontal
spindle, one end of which has a flexure strip and the other end is
Fig.2.5 PRESSURE (OR VACUUM) GAUGE - SCHAFFER fastened to flat plate springs. Equilibrium exists when spring
force equals differential pressure. Mechanical travel is via the
DIFFERENTIAL PRESSURE CELL (DIp CELL) flexure strip, lower seal diaphragm and vertical spindles to indi-
The sketch of Fig. 2.5 illustrates a single diaphragm subject to cator or transducer. A bi-metallic strip adjusts bellows fluid
differential pressure. The dIp cell is often used in direct differen- capacity to allow for volumetric expansion. This device is shown
tial pressure recording as well as flow and level applications. The dotted within the left hand bellows. The bellows are often made
detecting element of the cell is a bellows, or diaphragm, whose of welded stainless steel discs and inner faces butt if excess
mechanical movement is used to indicate or transduced to elec- pressure is applied, so protecting the bellows.
trical or pneumatic signal output.
PIEZOELECTRIC DETECTING ELEMENT (SENSOR)
With certain solid crystals having an asymmetrical electric

(I
w
charge distribution, any deformation of the crystal produces

~
w equal external unlike electric charges on the opposite faces of the
I~ =
0::
::>
(/)
(/)
crystal. This effect is known as the piezoelectric effect.
0::
a..
---+_
I
I
I

I
.........
_
---.
J W
0::
a..
-+-
Deformation of the crystal can be caused by pressure and the
charges produced can be measured by. means of electrodes
I I ~
3:
~ ~ ~ ,
-
......,
1 0
-.J
attached to the opposite surfaces of the crystal. r

Crystal materials can be naturally occurring or man made.

Quartz (i.e. SiO z) is a material that can be used in temperature
environment up to 550°C, it is stable mechanically and ther-
mally. Barium titanate, a ceramic produced commercially, is also
used but tourmaline is principally used because of its good elec-
trical properties. The output voltage from tourmaline crystal,
acting as a detecting element (sensor) and transducer, is a linear
Fig. 2.6 DIFFERENTIAL PRESSURE CELL
function of the pressure applied and the charge sensitivity is
 30 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PRESSURE MEASUREMENT 31

approximately 2 x 10- 10 coulomb/bar, it can be used for pres-

sures varying from 1·03 to 800 bar. The only drawbacks are (1)
its sensitivity to temperature change (i.e. its pyroelectric effect)
hence ideally it should only be used under controlled tempera-
ture conditions (2) it is more suited to pressure variation than to ~««('««<'«'««('b4 i 1+
static pressure.
Figure 2.7 shows crystals in series and parallel. The series
arrangement gives a higher output voltage for the same pressure
applied whereas the parallel arrangement reduces output impe-
dance and hence is a more stable system. Figure 2.8 is a commer-
cial piezoelectric sensor using tourmaline crystals in parallel, the
sensor is small and is coated with insulating material to help
minimise the effect of temperature variation in the medium
whose pressure is being measured.
Fig. 2.8 PIEZOELECTRIC SENSOR
PRESSURE CAUSING OPPOSITE
CHARGE BUILD UP AT CRYSTAL
SURFACES STRAIN GAUGE

i! ! ; ; ; ; , ,+
A strain gauge is fundamentally a resistive wire of about 0·01
mm diameter subject to strain by pressure (force or acceleration)
CRYSTAL CONDUCTIVE with electrical resistance change proportional to strain. Bonded
COATING
elements of the wire wound type are either flat grid or helical
wound on a former. The wire is fixed to a backing material such
as paper, resin or plastic which is glued to the surface under test,
wires are soldered or spot welded. One alternative is the foil type
where the grid is etched from thin metal foil using printed circuit
" , !, '+ ,! ; !, ,+ techniques. Another alternative is a p or n doped silicon semi-
conductor which can be very small and is extremely sensitive.
+ + + The unbonded strain gauge is essentially a pressure sensor and
+ + + a typical design is shown in Fig. 2.9.
+ + + + + + The detecting (sensing) element can be fastened directly to the
diaphragm as shown, or alternatively, a central force rod can
transmit diaphragm movement to the detecting (sensing) ele-
CRYSTALS IN PARALLEL CRYSTALS IN SERIES
ment consisting of plate springs with posts on the periphery upon
Fig.2.1 PIEZOELECTRIC CRYSTALS which the strain gauge wire is wound.
In all strain gauges, to minimise resistance change due to
Pyroelectricity, like piezoelectricity, makes use of the spon- temperature effects, it is usual to employ materials with a low
taneous polarisation of materials such as ferro-electrics. Temp- temperature coefficient of resistance for the wire, such as con-
erature variations causing changes in polarisation can be utilised stantan. Alternatively a second compensating wire loop can be
in very accurate calorimetric recording. incorporated.

';:;~i):!J'!j!jH'h§l~~'jij~a~!'
 32 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

WIRES TO
MEASURING ELEMENT
DIAPHRAGM WITH
STRAIN GAUGES
ATTACHED CHAPTER 3
t= { V~

LEVEL MEASUREMENT

A80Ul-
1cm
Liquid level sensors are usually classified under two headings,
LI---.:---===::11-----
i.e. direct methods and inferential methods.
Fig.2.9 STRAIN GAUGE PRESSURE SENSOR
DIRECT METHODS
The measuring element for strain gauges is generally a Wheat-
stone bridge circuit with temperature compensating resistance FLOAT OPERATED
and strain gauge resistance arranged as two, of the four, resis- The float is generally a hollow cylinder or ball working on
tances and a central galvanometer and constant dc voltage direct action or displacement principles. Level variation is
source. Null balance methods can also be used, or for high transmitted by chain or wire and pulley or torque tube (usually
sensitivity the bridge may be fed by ac voltage and the gal- with counterweights fitted) to the indicator. High or low level
vanometer replaced by a transistor amplifier and detector. The alarm contacts are easily arranged. Pulley movement can also be
measuring unit is effectively a transducer, i.e. displacement- arranged to operate a contact arm over an electrical resistance so
electrical. varying current or voltage to indicator or receiver.

SIGHT GLASSES
TEST EXAMPLES 2 Various types are in use dependent on working conditions.
The simple boiler water glass gauge with toughened glass and the
1. Sketch and briefly describe three types of pressure measur- plate type of water gauge for high pressures are typical.
ing device.
PROBE ELEMENTS
2. Sketch and describe a Bourdon type of pressure gauge. State Floatless types of level sensors can be arranged where the
the materials used in construction. Discuss briefly how the Bour- liquid is a conductor. Sensing electrodes, rods or discs, vary
don movement can be utilised for telemetering devices. electrical circuits when they are in contact with liquids. A typical
3. Describe, with the aid of a sketch, any type of differential example is detection of the fluid level in a tank by capacitive
pressure cell. Detail three applications in instrumentation of the ~echniques (see Fig. 3.1). This technique is used in oil-water
use of the dIp cell. Interface detection on oily-water separators - see Chapter 13.
 34 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
LEVEL MEASUREMENT 35

where Bo is the free space permittivity, B r relative permittivity

" hI" "head of fluid being measured
" h z " "distance from fluid surface to tank top
" 2r" "diameter of the probe
x " "separation between the circular plates
N
.c ...
W
N
" 10 " "natural logarithm.

~
z
UJ
~
oUJ
I-~
.£ ...
w I-
UJ
INFERENTIAL METHODS
UJ
II) :::>UJ
o a..Q::
1-:::>
Q:: :::>lIl PRESSURE ELEMENTS
a.. o<l
UJ
~
The static-pressure method is extensively used.
ac p= pgh
-(~
SUPPLY

p is pressure, p density of fluid, g gravitational acceleration,

Fig. 3.1 LEVEL SENSOR (CAPACITIVE) h fluid head. Any of the sensing and measuring devices
described in Chapter 2 are therefore applicable to level meas-
urement. In particular a pressure gauge, calibrated in height
units, is probably the most simple indicator. Bellows or dia-
phragms are used with the pressure bulb located inside or outside
Detection of level is measured by variation of capacitance the vessel with suitable correction factors applied for correct
which is accomplished by alteration in dielectric strength. One datum and density of fluid. The differential pressure cell is often
plate of the capacitor is a probe, possibly made of stainless steel, used in level measurement. Telemetering for remote reading is
the other is the shell and both are connected to an ac bridge readily applied, pneumatically or electrically, to the displace-
which is supplied with high frequency low voltage alternating ment movement of the level (pressure) sensor.
current. As the interface moves the dielectric strength (relative
permittivity B r ) alters.
MANOMETER TYPES
An approximate expression for the capacitance in an ac bridge The static pressure equation is applicable and devices have
circuit (Fig. 3.1) is:
been described in Chapter 2. Three types of remote level indi-
c= Bo[Brlh l + Brzh z ] cator are now described.
2/ [1 + x/r]
0
For the Electroflo electrical type as sketched in Fig. 3.2 the
difference in level h is directly proportional to the difference
and since Brz is approximately unity for air (Brl > 1) between the level of the liquid in the tank and a datum. The
C = Bo[Brlh l + h z] transducer element contains resistances immersed in mercury to
21 [1 + x/r] form an electrical circuit with transmission to remote indicator
0
• receivers.
 36 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LEVEL MEASUREMENT 37

WATER
LEVEL

rl~~~
hT I I 1111' I -..' I I-I

(XI_ "'-
-
h
H n

tJ~
1- -

Fig. 3.2 LEVEL SENSOR/TRANSDUCER (RESISTIVE)

INDICATING FLUID
Figure 3.3 is a diagrammatic arrangement of the Igema remote OF RELATIVE
water level indicator. The lower portion of the U tube contains a DENSITY P

(red) coloured indicating fluid which does not mix with water and
has a density greater than that of water.
The equilibrium condition for the gauge is H = h + px where p
is the relative density of the indicating fluid. H, h and x are
variables.
If the water level in the boiler falls, h will be reduced, x will be
increased and H must therefore be increased. The level of the
, Fig.3.3 IGEMA REMOTE WATER LEVEL INDICATOR

r
water in the condenser reservoir being maintained by condensing Figure 3.4 is another type of remote water level indicator. In
steam. this case the operating fluid is the boiler water itself. The opera-
If the water level in the boiler rises, h will be increased, x will tion of the gauge is as follows:
,~
be reduced and H must therefore be reduced. Water will there- If we consider a falling water level in the boiler, the pressure
fore flow over the weir in the condenser reservoir in order to difference across the diaphragm h will increase, causing the
maintain the level constant. diaphragm to deflect downwards. This motion of the diaphragm
A strip light is fitted behind the gauge which increases the is transmitted by means of a linkage arrangement (see insert) to
brightness of the (red) indicating fluid, which enables the tJ., the shutter which in turn moves down pivoting about its hinge,
operator to observe at a glance, from a considerable distance, causing an increase in the amount of (red) colour and a decrease
whether the gauge is full or empty. in the amount of (blue) colour seen at the glass gauge.
 38 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LEVEL MEASUREMENT 39

CONDENSER RESERVOIR
PURGE SYSTEMS
For small air flow rate, about one bubble per second, a pres-
PRESSURE DIFFJ::::- "1Z ' sure equal to that in the dip tube will be applied to the indicator
ACROSS DIAPHRAGM 'h' as shown in Fig. 3.5. This simple bubbler device is an arrange-
-1-. ment that is similar to the well known pneumercator used for
determining depths of water and oil in tanks. Air supply to the
l:=:=::oo::t --
open ended pipe in the tank will have a pressure which is directly
proportional to the depth of liquid in the tank.

~
HIqH LEVEL ELECTRICAL
CoNTACTS I

GLASS SHCiUTHINGE
I E R DIAPHRAGM 0 GAUGE
GAUGE
0

., f
~ ,." 0 L_~
SHUTTER LINKAGE ARRANGEMENT DEPTH

L·:v.~.
SWITCH/
COCK

VENT
Fig.3.4 REMOTE WATER LEVEL INDICATOR

Fig.3.5 PNEUMERCATOR LEVEL INDICATOR

It will be clearly understood that if the water level now rises
then the (red) will be reduced and the (blue) increased. TEST EXAMPLES 3
Separating the (blue) and (red) colours, which are distinctive
and can clearly be seen from a considerable distance, is a loose 1. Sketch and describe a distant-reading boiler water level
fitting black band which moves with the shutter, giving a distinct gauge.
separation of the two colours. Explain its principle of operation.
An adjustment screw and spring are provided to enable the Give two ways in which the gauge would give incorrect readings.
difference in diaphragm load to be adjusted. Hence correct State the routine maintenance necessary to ensure maximum
positioning of the shutter and band in relation to the reading of a reliability at all times.
glass water gauge fitted directly to the boiler is possible.
Both devices shown in Figs. 3.3 and 3.4 are also capable of 2. Sketch and describe any type of level sensor. Discuss the
being observed by closed circuit television systems to extend the modifications necessary to utilise the device as a transmitter in a
distance between transmission and reception. Alternatively telemetering system.
telemetering by pneumatic or electrical transmitters is readily 3. Describe, with a sketch, an instrument used to indicate the
arranged. A dIp cell would be very suitable for the former and level of liquid in a tank. Can the result be used to determine
any displacement transducer, such as a differential transformer, pressure above a given datum and, if so, state the other variables
could be used for the latter. likely to be involved.

'~1i!m:rJlT:~;:;
 CHAPTER 4

FLOW MEASUREMENT

Many of the techniques utilised in flow measurement employ

principles also used in pressure and level measurement.

BASIC PRINCIPLES
Flowmeters are generally divided into two fundamental types,
i.e. quantity meters and rate of flow meters, use of the word
flowmeter generally implies the latter.

QUANTITY METERS
These devices measure the quantity of fluid that has passed a
certain point. No time is involved. Types are usually classified as
positive or semi-positive. A typical positive type utilises the flow
to drive a reciprocating piston and a counter is attached. The
meter acts like a conventional engine with fluid pressure supply-
ing motive power. Stroke length and cylinder dimensions fix the
quantity delivered per cycle. Semi-positive types are usually
rotary. A form of gear pump, or eccentrically constrained rotor,
can be used which is driven by the fluid. Quantity is measured by
number of rotations (cycles) and fluid per cycle. .

RATE OF FLOW METERS

Measure the velocity of fluid passing a certain point at a given
instant. From this rate of flow (quantity per unit time) can be
determined from velocity multiplied by area of passage. They are
therefore classified as inferential, i.e. volume inferred from veloc-
ity. There are two fundamental components of the rate of flow

.",j~mi~-=~!10;;~~~~h
 42
5.S. BEKUL
REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASURE OFF. No. 0 0 8

meter. The primary element is that portion of the instrument increments of flow rate. From the aspect of display and continu-
which converts the quantity being measured into a variable to ous recording of flow rate the square root characteristic is not an
operate the secondary element, for example, orifice and pressure embarrassment. If however the differential pressure signal has to
tappings from a venturi. The secondary element measures the be used in a control system the square root is usually extracted to
variable created by the primary element, for example, a differen- give a signal which is directly proportional to the fluid flow rate.
tial pressure cell.
Square root extraction is described later in this chapter under
differential pressure inferential devices. It is difficult to integrate
INTEGRATORS
readings of a system when the square root extraction is not
Quantity meters are more expensive and less suited to deal applied.
with large fluid quantities than rate of flow meters. Rate of flow
meters are often used as quantity meters by fitting an integrator.
As a simple example consider variable flow. The rate of flow can
be measured at set time intervals and a graph plotted, the area of
which gives quantity over the time period required. In practice
this is performed mechanically or electrically by an integrator. INFERENTIAL-ROTATIONAL
One type is based on the planimeter principle and another type Basically these can be considered as mechanical or electrical.
(escapement) utilises mid-ordinate techniques from a heart-
shaped cam drive. Flat faced cam drive, or worm and wheel MECHANICAL TYPE FLOWMETER
designs, can be used with a turbine wheel or helix type of primary Designs are usually of the "turbine wheel" type with speed of
element having a counting mechanism secondary element, rotation directly proportional to linear flow velocity and, with
incorporating the integrator to interpret quantity. The integrator area of passage fixed, the volume rate is inferred. A wheel, fan or
is often included within the receiving unit of telemetering sys- helix is inserted in the pipe or duct, mounted vertically or hori-
tems and a typical device is illustrated and described in Chapter zontally, gear trains are used to interpret the movement. This is
6. Integration is readily performed electrically by use of a con- the principle of vane anemometers.
ventional watt-hour meter. It should be understood that integra-
tion is a general instrumentation operation whose use is not Rotormeter
restricted only to flowmeters. The measuring principle is illustrated in Fig. 4.1 where the
meter, of the duplex rotor positive displacement type, is shown in
SQUARE ROOT EXTRACTION
three specific positions. This rotary flowmeter operates on the
When inferential devices are used, with velocity sensors utilis- displacement principle, the measuring system consisting of a
ing differential pressure techniques, the velocity is not directly casing with two rotors. Bearing bushes are provided on either
proportional to pressure difference, or head. Velocity is related side of each rotor so that the rotor runs clear of the casing and
to the square root of pressure, or head, i.e. a curve of flow rate spindle in a radial direction. The rotor is located in an axial
plotted against pressure, or head, is of parabolic form. direction by means of end bearings so that it runs clear of the
This means that if a pressure difference is used in a sensor bearing plates. Both the end bearings and the spindle are fixed to
device connected to a manometer, or pointer through a linear the shaft whilst the bearing bushes are fixed to the rotor and
mechanism, the rate of flow scale on the manometer or traversed rotate relative to the end bearings and the shaft. Each rotor
by the pointer would have to be a square root function. The scale carries a gear wheel at the rear through which the rotors are
divisions would increase in square root increments for equal coupled together. The rotary movement of the rotors is transmit-

~"l!l!'ll11!!l!'.nMJafa~~lJ<3$;·
44 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASUREMENT 45

ted to the outgoing shaft of the meter through a gearwheel fitted causing this rotor to rotate in clockwise direction. Since both
at the front of one of the rotors. Consider Fig. 4.1: rotors are coupled together through gears, the left-hand rotor
In position "A", the left-hand rotor is fully relieved from load will rotate in anti-clockwise direction.
whilst the liquid pressure acts on one side of the right-hand rotor In position "B", the liquid in the displacement chamber is
pressed by the right-hand rotor to the outlet.
In position "C", the right-hand rotor is entirely relieved from
IN load whilst the liquid pressure now acts on one side of the
left-hand rotor so that it takes over the task of the right-hand
rotor.
A SURFACE CONTACT To reduce leakage losses to a minimum, it is essential to
SEAL BETWEEN
ROTOR AND CASING provide for effective sealing between inlet and outlet. To this
end, the rotors seal off against the casing by surface contact.
In the rotormeter, the motion of the rotors is transmitted to the
FLUID OUT
external parts by the attraction between two permanent mag-
nets, an inner and an outer magnet. The maximum torque trans-
mitted by this system is 0·4 N m. This arrangement offers the
ROTOR
following practical advantages;
B (1) Perfectly leakproof transmission, which means that it is
impossible for corrosive or hazardous liquids to leave the meter.
(2) Protection of attached parts or instruments. If the mechan-
ism of external parts or instruments should be blocked for what-
ever reason, the permanent magnetic coupling will slip, thus
avoiding any damage to the instruments.
Although the torque to be transmitted by the magnets may
C vary with the external parts or the type of instrument used, it will
hardly ever exceed a value of 0·05 N m with the meter running at
constant speed. Hence, at ambient temperature the magnetic
coupling has an 8-fold safety margin which will, in most
instances, be amply sufficient to take accelerations and decelera-
tions (which are the most frequent operating conditions) without
STAINLESS STEEL CAP
INNER MAGNET
UTER MAGNET
slipping of the coupling.
PERMANENT Although the torque transmitted by the magnets decreases
~~~~~G~~;7i"-r;~ with rising temperature, experience has shown that even with a
liquid temperature of up to 250°C the available torque is still
amply sufficient.
OUTGOING SHAFT OF
ROTORMETER ELECTRICAL TYPE FLOWMETER
Fig. 4.1 FLOWMETER (MECHANICAL) One type utilises rotating vanes with a small magnet attached
to one vane and a coil in the pipe wall. The electrical impulse can

lfat~~~t~~~~~jjp1j~
 46 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASUREMENT 47

be counted on a digital tachometer calibrated in flow rate. The

design now described has no moving parts.
FLOAT STOP

Electro-magnetic flowmeter
This type is shown in Fig. 4.2. RATE OF FLOW
SCALE
The principle utilised is that of a moving conductor (the
liquid) in a magnetic field generating a potential difference. In GLASS TUBE
the simple arrangement shown the two electro-magnets are sup-
plied with current (ac is preferred to dc to reduce polarisation of
the dielectric). There are two sensor electrodes. If B is the flux FLOAT
density of the field, v velocity of flow, d pipe diameter, then in
suitable units the emf generated at any instant is given by:
e = Bvd
For constant Band d, e is directly proportional to v. t
Fig.4.3 ROTAMETER

This is a variable area meter. A long taper tube is graduated on

its vertical axis. The float moves freely in the tube and by an
arrangement of shaped flutes in the float it slowly rotates. As
flow rate increases the float rises in the tube, so that the annular
area increases, which means that the differential pressure across
the tube is at a constant value. The float can be arranged with a
magnet attachment and a follower magnet outside will transmit
motion to a pointer via linkage if required.

ELECTRO-MAGNETS
PIPE CARRYING
CONDUCTIVE FLUID
INFERENTIAL-DIFFERENTIAL PRESSURE
PRIMARY ELEMENTS

Fig.4.2 FLOWMETER (ELECTRICAL) The orifice and the venturi will be described and these are
mainly used although flow nozzles and special inserts, such as a
Dall tube, are also employed. Fig. 4.4 shows both the orifice
ROTAMETER plate and venturi sensors using energy conversion to produce a
This type does not strictly fit into the classifications given but a pressure difference which can be utilised by the secondary ele-
brief description is appropriate at this stage. A sketch is given in ment to provide a signal for direct reading, telemetering or
Fig. 4.3. control.
48 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASUREMENT 49

Using the venturi flow sensor as an example the theory where KE is kinetic and PE is pressure energy.
involved is as follows: The equation for continuity of flow for area A is;
Assuming unit mass, and energy at points 1 and 2 being the
vlAl = v2A2 (b)

,,-ORIFICE PLATE
!!!III
By substituting for V2 from (b) in (a) and using mass flow rate m
VENA CONTRACTA as equal to pv 1 A 1 it can be shown that
FLOW
• in = k'"!i
where p is the pressure difference (p 1 - P2) and k is a meter
constant in terms of areas and density which includes a discharge
coefficient factor. Frictional losses are greater for orifice than for
venturi meters.
'P2
SECONDARY ELEMENTS
Any differential pressure device can be used as a secondary
element including the manometer, diaphragm and dip cell as
described in Chapter 2. The measure scale will be non-linear for
~ ,,:ENTURI TUBE

o ~ J
direct recorders due to the square root relation and telemetering,
control and integration will be generally unsatisfactory unless a
- - - .0 ~ -- --- --
correcting unit is fitted. When manometers are used various
compensations can be used. The simple manometer can utilise a
curved measuring limb and the well type manometer can be
THROAT arranged with a shaped chamber or may include a parallel tube
and a shaped displacer. Other direct measuring devices utilise a
Pl' t P2 cam incorporated in the mechanism, an example is the ring
balance. In the electrical resistive sensor described in Chapter 3
Fig. 4.4 FLOW SENSORS
(Fig. 3.2), when used for flow measurement, the electrode tips
immersed in mercury are arranged in a parabolic curve with each
other which gives the compensation. Three flow sensor square
root extraction devices are now considered in more detail.
same, i.e. neglecting friction and shock losses as small, then from
Bernoulli for incompressible flow of fluid of density p: SQUARE ROOT EXTRACfORS
KE at 1 + PEat 1 = KE at 2 + PEat 2 Figure 4.5 is one type of square root extractor using a parabola
shaped bell which can be connected through linkage to mechani-
t + pl/p = t + p2/p
Vl 2 V22 (a) cal, pneumatic or electrical display and control systems.

~jW."!lF;~~~~~~~r.
 50 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASUREMENT 51

I LOW Figure 4.6 is a diagrammatic sketch of another type of square

,f. PRESSURE
root extractor utilising the pneumatic flapper nozzle position
balance principle (as described in Chapter 6).
GLAND

OUTPUT AIR TO
RECORDER OR
CONTROLLER

SPECIAL
SHAPED BELL

MERCURY

~NOZZLE

j'HIGH
PRESSURE
, -VANE
LEVER 'B' WITH PIVOT

Fig.4.5 SQUARE ROOT EXTRACTOR (MECHANICAL)

LEVER MOVEMENT
CAUSED BY. DIFFERENTIAL
PRESSURE

Fig.4.6 SQUARE ROOT EXTRACTOR (PNEUMATIC)

The differential pressure is applied with the high pressure

inside and the low pressure outside the bell.
With its changing cross sectional area and buoyancy, due to
change in differential pressure, the bell movement is made to be Refer to Fig. 4.6:
directly proportional to the square root of the differential pres- The differential pressure from a flow sensor acts on a horizon-
sure. Hence the bell movement is directly proportional to the tal lever B, this effects the amount of air escaping from the nozzle
fluid flow being sensed by the orifice plate or venturi sensor. and hence the pressure in the bellows. Pressure alteration in the
This device is often called a Ledoux Bell. For air flow meas- bellows causes movement of the vertical lever A. The very small
urement, instead of the bell being shaped, a shaped displacer is relative motion between levers A and B provides square root
arranged in a separate chamber. The displacer is connected to extraction. The output air signal is directed to the measure
rise with the bell. Such a device is used for measuring steam and element (recorder) or controller.
air flow, and for controlling steam/air flow ratio, the characteris- Fig. 4.7 shows a square root extraction technique using electri-
tic is linear. cal force balance (as described in Chapter 6). The differential

'~ii.m~~illlli:.tE-jfj~rID1~~~~~-~
52 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FLOW MEASUREMENT 53

pressure, in electrical signal form represented as variable x input,

is applied to.the left hand side. With the force balance beam in
equilibrium the output signal is variable j x. Variation in input ORIFICE PLATE
signal, causing unbalance, can be arranged to be re-balanced 1--
by suitable adjustment to output signal. This can be done in
various ways, one method could be to connect the right hand of
FLOW . .
the beam to a differential capacitor in the output circuit (see Fig.
6.6). This is a closed loop but without electrical connection
between input detected signal and output measured signal. The
force balance system is a simple analogue computer - output can
also be easily arranged for squaring, summing, etc. by suitable
wiring combinations.

A B C
FORCE BALANCE BEAM

=Bo~
o
MAJNET
PIVOT = rx
:
OUTPUT AIR
PRESSURE '·2 <
TO 2·05 bar. TO
DIFFERENTIAL
PRESSURE
TRANSMITTER

I
AIR
4 SUPPLY
RECEIVER
Fig.4.7 SQUARE ROOT EXTRACTOR (ELECTRICAL)

I
Figure 4.8 shows two arrangements of flow sensor/transducer
- - - - - l..~ FLOW
units, each with square root extraction. For the pneumatic
system the valves marked A, Band C would be used in the
sequence, open B, close A and C, when taking the differential
pressure transmitter out of operation. This valve sequence would DIFFERENTIAL
PRESSURE
also apply to the electric system. TRANSMITTER
With steam flow measurement, the pressure tappings from the / AMPLIFIER
flow sensor are led to cooling reservoirs wherein the steam
condenses. Water only then acts on the square root eliminator or
recorder thus preventing damage.
The pneumatic system would utilise square root extraction as
shown in Fig. 4.6 (the technique shown in Fig. 4.5 could also be
50 Hz SUPPLY
used).
The electrical system can utilise mechanical movement (from a Fig.4.8 FLOW SENSORS/TRANSDUCERS
device such as Fig. 4.5) as shown to give square root extraction.
Variation of the position of the soft iron core of the inductor unit

~jj",~:"j,~~~~~
54 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

governs output signal to the amplifier. Alternatively direct sign-

als to the amplifier can be taken from electrical devices shown in
Fig. 3.2 and Fig. 4.7.

TEST EXAMPLES 4 CHAPTER 5

1. Sketch the apparatus and describe the method of measuring

steam flow through a pipe.
OTHER MEASUREMENTS
Explain why results obtained using this apparatus may differ
from theoretical results.
2. Describe, utilising a sketch, any type of flowmeter. For
inferential types of device how is the measured variable related From an engineering viewpoint the variables discussed in the
preceding chapters are the most significant. However other vari-
to flow rate?
ables have a particular specialist interest and a selected number
3. Inferential flowmeter devices using differential pressure are now presented with the appropriate instrumentation meas-
techniques for velocity sensing exhibit non-linear characteristics. urement.
Discuss these characteristics and describe a device to restore
linear characteristic to measure or control output signal. ELECTRICAL
Such instrumentation covers a wide field, generally well
described in electrical text books, and on this basis is not consi-
dered in this text unless specifically necessary. The measurement
of capacitance, current, voltage, resistance, frequency, phase,
energy, etc. are all very important in instrumentation. The reader
will be familiar with such as moving iron and moving coil meters,
etc. The bridge (ac or dc) and potentiometer are freq uently used in
telemetering and control components. Electronic measuring
devices are being increasingly used and rectifiers, amplifiers,
oscillators, etc. are an integral part of telemetering and control
systems.
SPEED-TACHOGENERATOR
The dc tachogenerator is a small precision generator driven by
the shaft whose rotational speed is required. Output voltage is
directly proportional to speed. The tacho is best geared to run at
maximum speed so giving maximum output signal and improved
signal to noise ratio.
When used as a tachometer in measuring systems the output
voltage from the tachogenerator is measured on a conventional
 5.S. BEKULAN
56 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
L
OTHER MEASUREMENTs .
OF F. No. 0 0 8 9 j
5

voltmeter calibrated in terms of rotational speed (an analogue As the (ferrous) toothed wheel rotates each tooth alters the air
device). gap and flux in a pick up coil (P) whose output pulses are
Mechanical type tachometers are based on centrifugal action, amplified (A). Pulses pass through a timing gate (G), say one
linked to produce lateral travel. second opening period, and are counted on a digital counter (D)
A conventional ac generator for use as a tachogenerator or which scales (related to teeth number per revolution) and dis-
tachometer is generally not satisfactory because frequency (and plays as revolutions per second. Alternatively revfmin readings
phase) as well as voltage amplitude are proportional to speed of can be arranged with different gate, or scale, settings.
rotation. In this case a "drag-cup" device is used. The rotor is a
thin aluminium cup rotating around a fixed iron core. The stator TORQUE-POWER
is wound with two coils at right angles, one ac supply the other ac Indicated power can be measured by a conventional mechani-
output. With the cup stationary there will be no output as the cal indicator although modern practice is tending towards oscil-
windings are at right angles. Induced emf with cup rotation, due loscope display with integration for power. Shaft power of
to cutting of flux of supply winding, links with the output winding engines is measured by a torsionmeter in conjunction with a
so giving a signal proportional to rotational speed; frequency and tachomete~ (power proportional to product of torque and rota-
phase being that of input signal. The device can be used for rate tional speed). Specific fuel consumption is readily achieved from
of change detection. With dc supply at constant speed no emf is these readings with a flowmeter calibration of fuel consumption.
induced in the output coils but angular acceleration or decelera- Various types of torsionmeter are available but those giving a
tion induces a voltage proportional to this change in the output continuous reading are usually of the electrical type. One design
coils. A "velocity voltage" applied to a differentiation (rate of in common use is based on differential transformer operation
change) circuit (CR series) will give a voltage across the resistor (Chapter 6) which is illustrated in Fig. G of the specimen·ques-
which is approximately proportional to acceleration (especially tions at the end of the book. Another design is based on magnetic
with a small time constant). Alternatively as force is proportional stress sensitivity and is termed a torque inductor-torduc-
to acceleration a simple spring accelerometer can be used. A tor-and is now described.
digital tachometer (counter) is shown in Fig. 5.1. The torductor is, as the name implies, a torque inductor, it is a
stress transducer that is eminently suited to the measuring of
o torque in rotating shafts. It gives a high power output and
requires no slip rings or other shaft attachments since it operates
without any contact. Figure 5.2 shows a ring torductor. It consists
of one primary ring which carries four poles, marked N,S, that is
supplied with (50 Hz) alternating current. Two outer secondary
rings have four poles each, arranged at 45 0 to the primaries, all of
which are connected in series with mutually reversed windings.
No contact exists between the poles and the shaft, there being
a 2 to 3 mm air gap provided to ensure this.
When no torque is applied to the shaft there are no stresses in
de
the shaft and the magnetic fields between NS poles induced in the
Fig.5.1 DIGITAL TACHOMETER shaft will be symmetrical, the equipotential lines are then situated
symmetrically under the secondary poles SIt Sz, as shown and
secondary flux and voltage will then be zero.
 58 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 59

without any amplification. If this signal is now married to a speed

signal, from a tachogenerator perhaps, then the power being
developed could be displayed directly on to one dial.

VISCOMETER
Newton investigated the viscosity of fluids and postulated, for
most fluids under prescribed conditions, that flow rate is propor-
tional to applied stress, more exactly that applied shear stress is
proportional to velocity gradient.

F dv
- =7]-
A dx

where 'TJ is a constant called the coefficient of viscosity. Applying

this equation to the case of flow through a small bore tube of
radius r and length I gives:

N TTpr 4
~ ~ 7] = 81V

~ ~Sl S1
For a constant flow rate (V):

~
L1N~~~TENTIAL
EQU
L..:.J r~ 7] = p X a constant
WHEN AND
ZERO TORQUE EQUIPOTENTIAL L1NES-
STRESS WHEN TRANSMITTING
TORQUE where p is differential pressure.
Fig.5.2 RING TORDUCTOR
Figure 5.3 shows the operational arrangement of the sensor
When a torque is being transmitted the equipotential lines element of a viscometer (Viscotherm). A small gear pump driven
form an asymmetrical pattern, as shown, due to the mutually at constant speed, by an electric motor through a reduction gear,
perpendicular unlike stresses, acting at 45 0 to the shaft axis, forces a constant fluid quantity from the housing through a small
causing increased permeability in one direction and decreased bore tube (capillary). Fluid flows through the capillary without
permeability at 90 0 in the other direction. This causes the 8 1 pole turbulence, i.e. streamline (laminar) flow prevails and pressure
to become magnetically slightly positive and the 8 2 pole slightly differential is proportional to viscosity of fluid. The pressure
negative. differential can be measured by any of the means previously
The output from the secondaries of the ring torductor is of the described. The device, within a control system, is described in
order of a few milliwatts, which is large enough to be used Chapter 13 (Fig. 13.19).
60 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 61

GEAR PUMP LOW PRESSURE

SUCTION CONNECTION

I
•
LIGHT
'4

---f.I-l.~ OUT LET

I • v "I
Fig. 5.4 PHOTOCELL
OIL INLET
Fig. 5.3 VISCOSITY SENSOR the anode and in this way create a potential V which can then be
amplified and used for alarm and control, etc.
In the vacuum cell all current is carried by photo electrons to
PHOTO-ELECTRIC CELLS the positive anode. In the gas filled cell emitted electrons ionise
Photo-conductive cells are constructed with a thin layer of the gas, producing further electrons, so giving amplification.
semi-conductor material and their resistance varies with the Secondary-emission (photo-multiplier) cells utilise a series of
incident light energy. They are used as sensors in many situations increasingly positive anodes and give high amplification.
such as oil-water content, smoke density, oil mist, flame indi-
cator, etc. detection as described later in this chapter.
Photo-emissive cells rely on the light energy providing energy i i • 6V
to release electrons from a metallic cathode.
If visible light, which is radiation and hence energy, falls upon
certain alkali metals - such as caesium - electrons will be emitted
from the surface of the metal. Metals in general exhibit this I-
~
characteristic but for most materials, the light required has a ll..
I-
threshold wavelength in the ultra violet region so that visible ~
o
light does not cause electron emission.
Light energy comes in packages called photons and the energy i iQV

of the photons is used in doing work to remove the electrons and Fig. 5.5 PHOTO-TRANSISTOR
to give the electrons kinetic energy after escape from the metal.
Figure 5.4 shows a simple photocell, visible light falls on the
metal cathode from which electrons are emitted, they collect at
62 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 63

Photo-transistors exhibit similar characteristics and small size A light scatter photo-cell separated from a semi-conductor
and high amplification make their use particularly attractive intermittently flashing light source is shown in Fig. 5.6. The
especially when applied to counting systems, i.e. digital housing enclosure allows smoke but not light inside. With smoke
tachometry. This device is shown in Fig. 5.5. Optical focus inci- present in the container light is scattered around the barrier on to
dent light on to the base increases the base current, hence collec- the photo-cell and an alarm is triggered. The light obscuration
tor current, and output voltage falls. type is used in oil mist detection for IC engine crankcases and the
combination type is used for detection in CO 2 flooding systems.
OIL IN WATER SENSOR
A useful application of photo-cells is in detection of oil-water OIL MIST DETECTOR
interface (as an alternative to the method described in Chapters The photo-cells of Fig. 5.7 are normally in a state of electric
3 and 13). Fluid passing through glass is exposed to long balance, i.e. measure and reference tube mist content in equilib-
wavelength light from an ultra violet lamp which causes fluores- rium. Out of balance current due to rise of crankcase mist density
cence if oil particles are present. This light can be detected by the can be arranged to indicate on a galvanometer which can be
secondary element photo-cell unit which produces a signal for connected to continuous chart recording and auto visual or aud-
amplification. The amount of fluorescent light is dependent on ible alarms. The suction fan draws a large volume of slow moving
the amount of oil in the oil-water mixture and this affects the oil-air vapour mixture in turn from various crankcase selection
amount of visible light detected by the photo-cell.

SMOKE DENSITY DETECTOR

For fire warning and exhaust gas indication a photo-cell in
conjunction with an amplifier and alarm or indicator is used.
Three types are in use, those which operate by light scatter, by
ROTATING
light obscuration and a combination. SELECTOR
VALVE

~-,
NEON LAMP
•• I
' ,-

•
V/4

OLAMP
STANDARD BASE ,i I I I • 9

ALARM AND
~11- T, un,

CONTROL
CIRCUIT
ADJUSTABLE
BARRIER MIRRORS

i U'I LJ-t-FLASHING LAMP

II
IL
J I
JII
Fig.5.7 CRANKCASE OIL MIST DETECTOR

PHOTO ELECTRIC 'ENCLOSURE

CELL (ALLOWS SMOKE TO ENTER
BUT NOT LIGHT)
Fig.5.6 SMOKE DENSITY DETECTOR
64 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 65

points. Oil mist near the lower critical density region has a very
high optical density. Alarm is normally arranged to operate at
2·5% of the lower critical point, i.e. assuming 50 mg/l as
lower explosive limit then warning at 1·25 mg/I.

pH SENSOR
The pH value of a solution is the logarithm of the reciprocal of
the hydrogen ion concentration in the solution. Its value ranges I .. I PLATINUM WIRE
between 0 to 14, neutrality being 7 anything from 7 to 14 is
alkaline and from 0 to 7 is acidic. pH measurement and control is __ I MERCURY
extremely important, being primarily used for feed water
GLASS
analysis.
POTASSIUM CHLORIDE SOLUTION (SATURATED)
The method of pH measurement is by means of a conductivity
MERCUROUS CHLORIDE (CALOMEL)
cell consisting of two electrodes and a temperature sensor, pH AND POTASSIUM CHLORIDE PASTE
value varies with temperature hence it is important that this be POROUS PLUGS
controlled by means of a sensor/compensator. One of the elec-
trodes is a reference electrode which has a fixed potential irres- REFERENCE ELECTRODE
pective of the variation of hydrogen ion concentration of the
solution. The other electrode produces a potential dependent
mainly upon the difference in hydrogen ion concentrations be-
tween the buffer solution and the solution whose pH has to be GLASS STEM

measured (across the membrane). In this way the potential dif- I. I SILVER/SILVER CHLORIDE
ference between the glass measuring electrode and reference ELECTRODE

electrode is a measure of the pH value of the solution. Elec-

trodes, with sensor/compensator, are inserted in the fluid flow
path.
Figure 5.8 shows the two types of electrode used in the BUFFER SOLUTION OF CONSTANT pH
conductivity cell. (SAY 0·' N KCI)

THIN WALLED GLASS BULB (MEMBRANE)

HEAT (FIRE) DETECfOR MEASURING ELECTRODE

Detector heads are generally one of three types. At a set heat
Fig. 5.8 pH SENSOR
(temperature) condition the increasing pressure on the pneuma-
tic diaphragm bulb type closes electrical alarm contacts,
increased differential temperature on the bi-metallic type acti-
vates alarms, increased heat fractures a quartzoid bulb (contain-
ing a highly expansive fluid) releasing water sprinkler supply and
pressure alarm. A typical fire detection-alarm circuit is shown in
Fig. 5.9.
66 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 67

FLAME DETECTOR
Fig. 5.10 illustrates the infra red type of device. Flame has a
characteristic flicker frequency of about 25 Hz and use is made of
this fact to trigger an alarm. Flickering radiation from flames
reaches the detector lens/filter unit, which only allows infra-red
rays to pass and be focused upon the cell. The signal from the cell
goes into the selective amplifier, which is tuned to 25 Hz, then
ZONE into a time delay unit (to minimise incidence of false alarms, fire
INDICATORS
....... has to be present for a pre-determined period), trigger and alarm
ZONE
ON/OFF circuits.
SWITCHES
SYSTEM FAILURE ALARM
SUPPLY '-'
SYSTEM TEST SWITCH GAS EXPLOSION-DETECTOR METER
The instrument illustrated in Fig. 5.11 is first charged with
fresh air from the atmosphere using the rubber aspirator bulb
(A). On-off switch (S 2) is closed together with check switch (S J
CIRCUIT FOR A ZONE and the compensatory filament (C) and detector filament (D)
allowed to reach steady state working temperature. The zero
adjustment rheostat (F), can now be adjusted so that galvanome-
ter (G) reads zero. Voltage is adjustable from battery (B) by the
DETECTOR HEADS
rheostat (E). Switch S2 is now opened.
Fig. 5.9. FIRE DETECTION - ALARM CIRCUIT The instrument is now charged from the suspect gas space and
while operating the bulb, the switch S 2 is again closed. If a
flammable or explosive gas is present it will cause the detector
NEON LAMP filament to increase in temperature. This disturbs the bridge
~ balance and a current flows. Galvanometer (G) can be calibrated
STANDARD
BASE so that the scale is marked to read "% of Lower Limit of Explo-
TIME DELAY sive Concentration of Gas".
UNIT
An alternative design has two ionising chambers, one refer-
ence (air) and the other sample, each containing a radioactive
AMPLIFIER AND
FREQUENCY ionising source. Combustion particles when ionised are more
UNIT TUNED
TO 25Hz I PHOTO-ELECTRIC bulky and less mobile than normal gas molecules so they are
CELL readily neutralised. This results in higher resistance and voltage
change at the sample chamber - which activates alarms.

I GAS ANALYSIS
Fig. 5.10 FLAME DETECTOR For detailed gas analysis an Orsat apparatus is used. However
a number of measurements require to be continuously recorded.
Two representative examples can be considered namely an
oxygen analyser and a carbon dioxide analyser.
68 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 69

ac SUPPLY
~

FILTER
GAS
SAMPLE INLET . < l « « « « « « " "

;t)
C COMBUSTION CHAMBER
(ALUMINIUM DIE CASTING)

\l
MAGNETIC
FIELD

F \.VWVV) BAFFLE
E
7----)- -- -- .-}--)-
Fig.5.11 GAS EXPLOSION - DETECTOR METER ---. ---.
• GAS IN GAS OUT

Fig.5.12 OXYGEN ANALYSER

Two platinum wire resistances are heated by current from an

ac bridge and the gas to be measured enters the resistance
Oxygen analyser chamber via a diffuser. One of the resistance wires is placed in a
Gases can generally be classified as either diamagnetic or magnetic field hence oxygen is drawn towards this resistance,
paramagnetic, the former seek the weakest part of a magnetic thus convection currents are set up around this resistance which
field and the latter the strongest. Most of the common gases is then cooled relatively to the other resistance. The bridge is
are diamagnetic but oxygen is paramagnetic and use is made of then unbalanced, the amount of unbalance is a measure of the
this in the oxygen analyser shown in Fig. 5.12. oxygen content and this is displayed on the galvanometer.
70 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 71

CO 2 analyser RELATIVE HUMIDITY

Referring to Fig. 5.13. A hair element will react to changes of humidity and provide a
Approximate thermal conductivities are in proportion: linear movement, with negligible force, which can be converted
CO 2 = 1, H 2 0 = 1, CO = 4, O 2 = 2, N2 = 2 to electrical or pneumatic signal and amplified as required.
The sample enters via a filter and drier, water vapour must be
removed as it has the same conductivity as CO 2 • The wire cell WATER ANALYSIS
resistance is proportional to heat dissipation, proportional to The measurement of pH has been considered previously.
thermal conductivity of gas in the cell, proportional therefore to Two other measurements are commonly required, i.e. electrical
CO 2 content. Air is used in the reference cell. Thus the only
difference between gas sample and air, from the thermal conduc-
tivity viewpoint is CO 2 (as H 2 0 removed and O 2 and N2 same
j conductivity meter for dissolved solid assessment and dissolved
oxygen meter.

value). This assumes no CO or H 2 , if these are present (normally "Dionic" water purity meter
only very small proportions) they will be registered as CO 2 unless Specific conductivity mho/cm 3 is the conductance of a column
the sample is first passed over a burner and these two gases of mercury 1 cm 2 cross sectional area and 1 cm long. This is a
burned off before the reading. large unit and micromho/cm 3 (reciprocal megohm) is used and
Thus the Wheatstone bridge electrical unbalance is dependent when corrected to 20°C is called a dionic unit. Conductivity of
on CO 2 content and the unbalance electrical current is measured
by the potentiometer.
Chemical absorption and mechanical types are also used.
,
~
\
pure distilled water is about 0·5 and fresh water about 500 dionic
units. The sensor, shown in Fig. 5.14, measures conductivity of

TO ac SUPPLY
INSULATING PLUNGER TO VARY
VOLTS TEMPERATURE COMPENSATOR THE WATER CROSS SECTION
COMPENSATING BI-METAL STRIPS ~ AUTOMATICALLY
COIL BY TEMPERATURE COMPENSATOR
IN OR OUT TO GIVE AUTOMATIC
GUN METAL CASING CORRECTION TO 20°C
' CONNECTED TO NEGATIVE

a:::
w PLATINUM ELECTRODE
~
w INSULATING TUBE }tj
WIRE RING (POSITIVE)
:E
o
~
z
w
~
o
a.. INSULATING TUBE
u
o
o
~
PLATINUM ELECTRODE
INSULATING TUBE ~tt WIRE RING (POSITIVE)

GUN METAL CASING

ZERO
'-.....ADJUSTMENT
CONNECTED TO NEGATIVE ~A
-
.. AIn
FLOW OF SAMPLE
(WATER)

Fig. 5.13 THERMAL CONDUCTIVITY TYPE C02 RECORDER

RHEOSTAT
r
Fig.5.14 DIONIC WATER PURITY METER
72 REED'S INSTRUMENTATION AND CONTROL SYSTEMS OTHER MEASUREMENTS 73

two water columns in parallel, i.e. between positive platinum INSTRUMENT CALffiRATION. TESTING AND ADJUSTMENT
rings and negative gunmetal collars. The insulating plunger, Generally a specialist subject. Pneumatic instruments would
operated by bi-metallic strip, varies cross sectional area for be tested by master gauge, standard manometer or deadweight
automatic correction to 20°C. The measurement is by conven- tester. Electrical instruments are tested by standard resistors,
tional ohmeter. The device should be used with de-gassifying potentiometers, capacitors, etc.
units to avoid errors due to occlusion of carbon dioxide. Using a Bourdon pressure gauge as example:
1. Zero (error) adjustment changes base point without changing
Dissolved oxygen meter the slope or shape of the calibration curve. It is usually achieved
The unit is shown in Fig. 5.15. The sample water flows via a by rotating the indicator pointer relative to the movement, link-
chamber which surrounds the katharometer (Wheatstone bridge age and element.
circuit) and receives pure hydrogen. Some hydrogen is taken into 2. Multiplication (magnification) adjustment alters the slope
solution and this releases some dissolved oxygen (in air). This without changing base point or shape. This is effected by altering
mixture passes to atmosphere across one side of the bridge whilst the drive linkage length ratios between primary element and
the other side is in pure hydrogen. The cooling effect is different indicator pointer.
on the two sides of the katharometer, depending on air (oxygen) 3. Angularity adjustment changes the curve shape without alter-
present, an9 resultant unbalance current operates indicator or ing base point and alters scale calibration at the ends. This error
recorder calibrated directly in ppm oxygen. For very low oxygen is minimised by ensuring that link arms are perpendicular with
content it is often necessary to utilise an electro-chemical cell in the pointer at mid scale.
place of this meter. Fig. 5.16 shows calibration curves and adjustment for the
Bourdon link type of instrument mechanism.

M Z
SAMPLED WATER
[II] OXYGEN
METER 100------------

COOLING
COIL

, ..
CONSTANT
HEAD
n DEVICE
i
. =r Yr-;
It
HYDROGEN FROM
LOCAL GENERATOR
CELL

DRAIN SCRUBBING
CHAMBER
o T--+ 100

Fig.5.15 DISSOLVED OXYGEN METER Fig.5.16 INSTRUMENT CALIBRATION AND ADJUSTMENT

 74 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

Instrument readings (I), true values (T), desired result (D).

Zero error and adjustment (Z), multiplication error and adjust-
ment (M), angularity error and adjustment (A) - error curves or
lines for actual values.
Over the design range pointer movement bears a linear rela-
tionship to pressure, and the scale is calibrated accordingly. CHAPTER 6
Hysteresis - a vibration phenomena. Best eliminated by cor-
rectly meshed gearing and fitted pivots to reduce backlash, etc.
TELEMETERING

TEST EXAMPLES 5

1. Explain the principle of operation of a carbon dioxide recor- The objective of this chapter (and Chapter 7) is to link ins-
der for monitoring the uptake gases. trumentation in preceding chapters, primarily concerned with
State what normal maintenance it requires. measuring devices, with control elements as described in subse-
State what action is taken if the carbon dioxide content is unac- quent chapters. Some repetition may inevitably result in the
ceptably low. presentation.
If this action does not alter the carbon dioxide reading, explain There are a wide range of variables to be measured. Detecting
how the accuracy of the recorder is checked and adjusted. and measuring elements are mainly electrical but a significant
number are displacement operated mechanical types. Chemical
2. Explain the term "photo-electric effect" and describe and electronic devices are also used.
equipment suitable for crankcase monitoring and fire detection Transducers can generally be simplified into three basic
in which this phenomenon is utilised. reversible types namely:
3. Describe, with the aid of sketches, a torsionmeter. mechanical displacement~pneumatic
Explain the principle of operation and briefly describe the con-
struction of the shaft and indicator units. mechanical displacement~electrical

pneumatic~ electrical
4. Discuss the errors liable to be exhibited by link type instru-
ments. Describe how such an instrument could be calibrated and
adjusted to reduce these errors to a minimum.
I
%;~ !
Pneumatic principles are invariably flapper nozzle; orifice, dia-
phragm. Electrical principles include resistance change, variable
inductance, variable capacitance, current or voltage, with fre-
quency and phase used to a limited extent. Conversion of electri-
cal signal is also used, i.e. resistance-current, voltage-current, etc.
and such modern transducers often incorporate electronic oscil-
lators and amplifiers.
Telemetering may be defined as signal transmission over a
considerable distance. The device at the measure point, usually a
transducer, is then often called a transmitter with the receiver
located at the recording or control centre.
 76 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERING 77

The material in this chapter will be covered in three sections mechanical advantage of the flapper lever, i.e. altering the x
namely pneumatic transmitters, electrical transmitters, receiv- dimension to the right or the left. A typical output pressure-
ers. flapper travel characteristic is included in Fig. 6.1. The device is
obviously a displacement-air pressure transducer, displacement
variation from flow, level, etc. variables. A pneumatic relay can
PNEUMATIC TRANSMITTERS be fitted on the air input.
POSITION-BALANCE TRANSDUCER (PNEUMATIC)
Displacement of a mechanical linkage gives variation in FORCE-BALANCE TRANSDUCER (PNEUMATIC)
pneumatic signal output pressure. The flapper-nozzle is the basis This is essentially a null method, i.e. equal and opposite forces
of many pneumatic mechanisms and the position (motion) bal- (torques) which eliminates inherent errors of the position-
ance is essentially a balance of positions (see Fig. 6.1). balance device. Consider Fig. 6.2, the flapper is a constrained bar

~ SET
'+I ACTUATING
DISPLACEMENT
SPRINGrt t
2

I_ x I ·¥ :-+ a
.. * .D

~ w
e:::
FLAPPER ::>
en
en
w
e:::
a.. ........---.

CONSTRICTION- 2·5 3·' 37 4·3 4'9

FLAPPER TRAVEL~
mm x 10- 2 tMEASURE
t
INPUT
AIR
SIGNAL
t
Fig. 6.1 POSITION BALANCE TRANSDUCER (PNEUMATIC) Fig.6.2 FORCE BALANCE TRANSDUCER (PNEUMATIC)

Ideally equal increments of flapper movement should produce pivoted about a fulcrum (adjustment of which varies mechanical
equal increments of pressure output, i.e. linear proportionality. advantage and thus changes proportion of input to output
In practice this is only achieveq over a limited flapper travel. To change). Bellows have equal effective area. With the device in
ensure increased sensitivity and linearity negative feedback is equilibrium assume an increase in measured signal pressure
used via a bellows. Linear output over the pressure range is which will produce a net up force and clockwise torque on the
obtained for an effective flapper travel range near the nozzle of bar. The flapper movement towards the nozzle will continue
about 0·015 mm. Output signal pressure is proportional to until the increased output pressure in the feedback bellows pro-
actuating link travel and the device is adjustable by varying the duces an anti-clockwise torque, to balance the actuating torque,
78 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERING 79

and at this point equilibrium is restored. Again a displacement ELECTRICAL TRANSMITTERS

variable to air pressure transducer. A relay may be fitted.
Consider first 3 examples of position (motion) balance conver-
ters namely variable resistance, inductance and capacitance.
ELECTRO-PNEUMATIC TRANSDUCER
The unit shown in Fig. 6.3 is based on the force balance VARIABLE RESISTANCE TRANSDUCER
principle with input variable of current (10-50 rnA dc is usual). In this simple and early form of transducer the mechanical
Electrical current signal variation causes a torque motor to pro- movement of the measuring element connection varies the elec-
duce a variable force (F) which is balanced by the feedback trical circuit resistance. The device is restricted in use due to
pneumatic bellows force (B) at equilibrium. The bar is circular contact problems, variations in wire connection resistance and
and between the poles of the permanent magnet acts as an the need for a constant stabilised voltage supply. Two types are
armature when excited by a dc current. Consider an increase in shown in Fig. 6.4, i.e. mV measurement and rnA measurement
armature current; the strength of the armature poles will locally or remote, the latter is a potentiometer technique.
increase accordingly. The S pole will move up as unlike poles Displacement may be due to any variable variation, i.e. level,
attract and produce a clockwise moment about the pivot whilst pressure, etc. via Bourdon tube, diaphragm or similar displace-
the N pole will move down producing an anti-clockwise moment. ment device. Provision of a cross coil measuring meter, or remote
The moment arm of the S pole force is greater so there is a net receiver, makes indication virtually independent of supply vol-
clockwise moment. This action closes in on the nozzle giving a tage variations. One coil is directly energised from supply (cur-
higher output pressure and increasing the feedback bellows force rent proportional to voltage) and a cross coil is deflected by
until equilibrium is achieved, i.e. a direct acting transducer. An measurement current so that deflection of the meter within the
alternative torque motor design utilises a suspended coil within permanent magnet is proportional only to resistance of the
the magnetic field of a permanent magnet. Fig. 6.3 is an input transmitter. This is a two-coil ratiometer.
electrical (current) to output pneumatic pressure transducer
(I/PP).

LlNK FROM

V MEASURING

r---,- ELEMENT

i
I
ri .....
Z
4(1)
..........
I ::>'11'. (rnA (I)~
zO
0>
I U
I
I
IL _
de CONSTANT L ...J
INPUT VOLTS

Fig.6.3 ELECTRO-PNEUMATIC TRANSDUCER Fig.6.4 VARIABLE CONTACT RESISTANCE TRANSDUCER

 s.s. BEKULAN
OFF. No. 008
80. REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERIl" I a ' I

Other Resistance Systems With the core moved right the induced voltage in secondary 2
The Wheatstone bridge principle is used in many cases, both winding increases and that in secondary 1 decreases so giving a
dc and ac, with direct or null balance techniques. The bridge differential output. Similarly movement left gives a voltage dif-
together with electrical resistance sensors for resistance ther- ference but 180 out of phase. The characteristic can be linear
0

mometers (Chapter 1) and strain gauges (Chapter 2) have been with zero volts at mid travel. Input displacement, from such as
described in detail already. Another electrical resistance trans- Bourdon tube or diaphragm is converted to an electrical signal
mitter has also been discussed previously in Chapter 3 namely for telemetering to indicators, recorders, data processing or elec-
the type utilising electrical resistors in mercury for measurement tronic controllers.
of level. Many electrical transducers are combined in the transmitting
unit with oscillators and amplifiers of solid state modular assemb-
VARIABLE INDUCTANCE TRANSDUCER lies. The oscillator supply in Fig. 6.5 from a power supply unit is
The unit shown in Fig. 6.5 is a differential transformer with commonly at 12 V dc and incorporates a chopper unit, ac amp-
three coils fully wound on a cylindrical former (only half the lifier and output-feedback stage. A stabilised current to the
winding is shown on the sketch for simplicity). The core, which is primary winding of about 5 rnA is often used from say a 1·6 kHz
moved laterally by displacement of a sensor element, provides oscillator. The amplifier itself, 12 V dc supply, accepts ac input
the magnetic linking flux path between coils. The primary ac from the differential transformer via ac bridge circuits at up to 2
voltage induces secondary voltages and as the two secondary rnA and gives output via demodulator-filter circuits at about 50
windings are in series opposition the two outputs are opposite in rnA maximum, dc. It is effectively a dc input dc output system.
magnitude and phase with the core laterally in the middle of the Components mentioned above are considered in more detail in
former. the next chapter.
This inductance ratio system can employ any type of receiver.
For a simple system direct ac supply, without oscillator and
amplifier, can be used and ac output passed across a bridge
rectifier for each coil to a two-coil ratiometer with pointer indi-
cator.
In an inductance balance system the receiver is identical to the
FORMER transmitter with secondary windings interconnected. Unbal-
r ---woo----I.,
----r--I- -~
PRIMARY
I
I
I
anced emfs due to displacement and inductance change at the
I I _.J I transmitter result in corresponding displacement at the receiver
- l_i~~OREl-r---
I I ---- I

~ ~Ir~J
so as to maintain current in each part of the circuit constant.
1----4i I I
(The lever and gearbox are mechanical transformers).
I I _I

L __ -=: - - --- - - "OIFFERENT~~lR

TRANSFOR
VARIABLE CAPACITANCE TRANSDUCER
The capacitance of a parallel plate capacitor is given by
OUTPUT
c= As
Fig.6.5 VARIABLE INDUCTANCE TRANSDUCER d
82 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERING 83

INPUT FORCE
A is plate area, e absolute permittivity and d plate separation. FROM MEASURE
Change of capacitance utilisation in conjunction with an ac ELEMENT

bridge circuit has been considered previously with reference to

level probes (Chapter 3). For displacement measurement a para-
llel RL resonant circuit can be utilised. Alternatively a differen-
tial capacitor principle can be used for displacement-current PIVOT FEED
MAGNET 0
~
BACK[1W.r1illOW
0
conversion, and is now described. UNIT 0 0
(MOTOR)

DIFFERENTIAL INPUT

r , LOAD
I

t• I I I
• OSCILLATOR
MEASURE
SENSOR
SIGNAL A
I
I
I
I
.:: I
AMPLIFIER
RECTIFIER
Fig. 6.7 ELECTRONIC FORCE BALANCE SYSTEM

I
I
IL.. _ _ _ _ _ _ _ _
I
...JI
VOLTAGE-CURRENT TRANSDUCER
OUTPUT It is often necessary to use amV/1 converter when dealing with
Fig. 6.6 DIFFERENTIAL CAPACITOR TRANSDUCER
thermocouple or resistance thermometer inputs. Such a device is
shown in Fig. 6.8.
Deviation between input mV and a standardised zero suppres-
sion voltage, from a zener diode power pack and bridge, is
Consider Fig. 6.6: algebraically added and passed through a filter network (to avoid
The central plate of the differential capacitor is moved verti-
stray ac pick-up). This signal is algebraically added to the feed-
cally by the displacement of a sensor device. The outer plates are
back stabilising loop of the amplifier and passed through the
connected to a combined oscillator, amplifier, rectifier unit. amplifier to output. Adjustable resistors A, C, D allow tempera-
Movement of the centre plate, towards one fixed plate and
ture correction, zero adjustment and span control of feedback; B
away from the other gives a change in capacitance to the
oscillator-amplifier. A change in output current to receiver
results. OUTPUT
demA
----4

ELECTRONIC FORCE-BALANCE SYSTEM

An electro-pneumatic converter has been already described INPUT
de mV
and an alternative pneumatic (pressure, force, displacement,
etc.) electric converter can now be considered (Fig. 6.7)
When movement varies the inductance, coupled to the oscil-
lator amplifier, let us assume amplifier output current increases.
This will continue until the feedback current on the force motor
produces equilibrium. Effective full scale beam travel is only
about 25 microns. Input may be from Bourdon tube or dia-
phragm. The device is referred to as a PP/I transducer. Fig. 6.8 VOLTAGE-CURRENT TRANSDUCER
 84 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERING 85

measured value. A dc amplifier can be used. If an ac amplifier bar (F) into nozzle (N). Increased pressure from relay (R) acts on
is preferred a chopper input and synchro-rectifier output is a 60 "tooth" turbine wheel (T) to cause rotation and equilibrium
needed, output and feedback isolated by a transformer. is obtained when up force due to centrifugal force (proportional
to square of turbine wheel speed) through thrust pin (P) on bar
balances down force. Adjustment is via spring (S) causing
RECEIVERS movement of weights (W) (up increases feedback force and
reduces counts per unit input pressure).
The variation of receiver types is very large ranging from direct Turbine wheel speed is directly proportional to flow, and by
measuring meters, recorders, display units, controllers and gear reduction to the counter, indication is of total flow.
analysing units. In many cases if a transducer is used it is merely a
form of converter device as already described in this chapter. POTENTIOMETRIC PEN RECORDER
Frequently it is not possible to sensibly separate transmitter and The potentiometer is used a great deal in instrument systems
receiver because they are inherently linked in operating princi- and also control systems (position control, Chapter 14). For
ple. With these provisos in mind a selection of units not previ- recording of small dc voltage it is usual to convert to a suitable
ously, or subsequently, described are now presented. frequency ac and amplify although dc amplification can be used.
Conversion to ac is achieved in a dc chopper amplifier with
SIGNAL output direct to servo-amplifier pen drive motor.
~ Figure 6.10 is a simplified sketch of a continuous balance
system.

THERMOCOUPLE

cr\--I-~lr~-I----~
~SLIDE
AIR BALANCING WIRE
~ CONTINUOUS
MOTOR
BALANCE
UNIT
~ -- --- --- - r~GEARING
/

~
'" Lv
---, 1
Fig.6.9 RECEIVER INTEGRATOR
Fig.6.10 POTENTIOMETER CIRCUIT

Input dc voltage, from say a thermocouple, is measured

RECEIVER INTEGRATOR against slidewire voltage at B with a constant voltage bridge
Refer to Fig. 6.9: source. Difference between A and B is amplified at the continu-
This receiver is for flow recording. Down force on bellows (B) ous balance unit so energising the balancing motor to move pen
from increased (above datum) input signal (proportional to arm and B until the voltage difference is zero. Similar arrange-
square of flow measurement) and about fulcrum (X) closes force ments utilise conductivity, ratio, bridge, etc. circuits. The
86 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TELEMETERING 87

balancing motor is two phase with a reference winding and a Figure 6.11 shows the arrangement in equilibrium with equal
control winding from the balance unit. Input to the balance unit currents (I) in line Band C and zero current in line A. The
incorporates a converter and centre tap of an input transformer. receiver rotor is locked by equal and opposite torques from the
The vibrating reed converter, in moving between two contacts, attractions on unlike pole faces. Assume the transmitter to be
allows current to pass alternately through each half of the trans- moved 30° clockwise. Current flows to receiver from line C,
former. Secondary ac voltage is amplified and fed to the control subdivides at point X and equal currents return through lines A
winding of the balancing motor, so timed with ac supply to give and B of magnitude 1/2. This creates a strong N pole at fixed
the correct restoring action. The chart is driven by a constant magnet X and two weak S poles at the other two fixed magnets.
speed geared motor. A damping feedback tachogenerator driven The receiver indicator will therefore turn to the corresponding
from the balance motor is often fitted. equilibrium position, i.e. 30° clockwise.

XYRECORDER
POSITION MOTORS (ac)
Used to measure a quantity Y, varying withX, where X is not a These devices are usually known by trade names such as
function of time. Two servo systems, perpendicularly connected, synchro, resolver, magslip and for larger powers, selsyn. Figure
cause the pen to move to any area position on the chart. Inputs 6.12 shows a transmitter and receiver of a synchro system.
cause perpendicular travel related to X and Y.

POSITION MOTORS (dc)

One technique is to feed current into a toroid resistor transmit-
ter with three tappings connecting to a three phase star winding
enclosing a two pole rotor receiver. This is sometimes referred to
as a Desynn transmission link. A similar principle is used in the
position indicator (electric telegraph).

+- "v

Fig. 6.12 SYNCHRO SYSTEM

A
Both rotors are supplied from the same ac source and stators
are linked in star. With rotors in the same angular position emfs
from transmitter and receiver stators balance and there is no
circulatory current. If the transmitter rotor is moved, induced
emfs are unequal and current circulates so producing a torque to
c bring the receiver rotor into line and restore equilibrium. Zero
receiver torque exists at alignment and maximum occurs at 90°
Fig. 6.11 POSITION INDICATOR (ELECTRIC TELEGRAPH) out of alignment. The resolver system is similar but utilises two
88 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

phase and is used for both fine control and data processing
systems. An intermediate synchro (follow through, hunter diffe-
rential, etc.) can be arranged, with three phase rotor and stator
connections, so that summing or differential control outputs are
possible.
CHAPTER 7

TEST EXAMPLES 6
ELECTRONIC DEVICES
1. In a pneumatic telemetering system explain how the air flow
through a nozzle in a transmitter is varied. How does this varia-
tion cause a change in the signal to the receiver?
2. Sketch and describe a variable inductance and a variable Electronics has developed greatly in this century and espe-
capacitance type of transducer. Discuss typical applications, with cially in recent years. The subject is specialised and a full presen-
examples, of these devices in measuring or control systems. tation is available from a wide range of books. For the purpose of
this short chapter an extreme rationalisation is necessary as
3. Describe a remote telemetering system. Detail on a suitable follows:
diagram both the indicator and receiver unit and the connections
between. (1) Components described are only those active devices with a
direct application to instrumentation and control and the basic
introductory electronics theory is reduced to a minimum.
(2) Description is related to semi-conductor devices only. All
the theory and practice of vacuum valve electronics including
diode, triode, tetrode, pentode, etc. is generally omitted as obso-
lete.
The contents of this chapter are given in 5 sections namely:
semi-conductors, rectifiers, amplifiers, oscillators and other
devices. It applies generally to analogue circuits in the frequency
domain, i.e. concerned with sinusoidal (or similar) waveform and
frequency analysis (current and voltage). Chapter 16 includes
digital circuits in the time domain, i.e. concerned usually with
square waveforms where the criteria is whether current or vol-
tage is present or not i.e. logic devices.

SEMI-CONDUCTORS
ATOMIC THEORY
The basic theory of electronics relates to atomic physics with
negatively charged electrons orbiting around the positively
 90 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 91

charged nucleus. Bonding decides conductivity, for example through the lattice, this can be regarded as causing the conduc-
copper as a good conductor and mica as an insulator. Conven- tion. Majority carriers, electrons or holes in n-type and p-type
tional current flow is positive to negative. When considering impurities, greatly outnumber minority carriers (which are
electron drift movement, which takes place from the negative temperature excited electrons). The temperature (maximum)
terminal through the conductor to the positive terminal, it must for solid state devices is usually fixed about 75°C (above which
be noted that the flow is opposite to the conventional. thermal runaway occurs).
Germanium and silicon in pure form hav~ a diamond lattice When ap-type and ann-type material are made into a specific
formation, each of four valence (binding) electrons per atom, junction electron flow to the holes in the p-type occurs. A nega-
and electron flow is only possible by partial lattice breakdown tive charge exists in the p-type and a positive charge in the
due to thermal energy. Arsenic and antimony have a five valency n-type, thus giving a potential difference across the junction
shell readily fitting into the diamond lattice and leaving a surplus which stops further migration of electrons. This n-p junction can
conduction electron. Aluminium or indium have a three valency act for example as a rectifier, with the positive side connected to
shell and whilst bonding into the lattice occurs a missing electron the n-type the potential difference is increased, with the negative
creates a conduction hole. side connected to the n-type the potential difference is reduced.
A few random electrons (leakage current) can go against the
ELECTRON CONDUCTION bias. When reverse biased the holes and electrons are drawn
With conductors the electrons, which constitute the current away from the junction leaving a depletion layer with virtually
flow, are capable of drifting through the material. The electrons no current carriers.
can be imagined to lie in a valence band and energy supply, heat Further consideration leads to transistors in which a pnp or
for example, is sufficient to excite the electrons sufficiently to npn sandwich exists. For npn the emitter emits electrons, collec-
allow them to jump across a narrow non-conducting band into tor collects electrons, and the base controls the flow of electrons
the conducting drift band.
With insulators a non-conducting band is wider and electron e e e
~

m cfib
jump does not readily occur, so that no current flows. ~
r--i~-,
I I I
With intrinsic semi-conductors, such as germanium or silicon,
the properties are midway between conductors and insulators.
Some electrons can break from the crystal lattice bond structure,
the gap created can then be filled by another electron, hence
electron movement does occur and this increases with tempera-
D +
~
hi

----e+
EQUIVALENT CIRCUIT
-
nip
+ I -
I
--+
e
REVERSE BIAS
nIp
+ I -

---+
e
FORWARD BIAS
ture increase. Extrinsicsemi-conductor materials contain slightly
impure or doped material. If arsenic or antimony is added to

~
germanium or silicon there is a surplus of electrons in the crystal I
-----+ C
lattice, such donor atoms give an n-type conductor. If aluminium I I
•t
p l n l p B
or indium is added there is a positive gap or hole in the crystal I I COLLECTOR
lattice, such acceptor atoms give a p-type conductor. ~~ -5V E
4-- 4-- B
Conduction by electrons is similar to that in metals. For a e e BASE
-0·2V
p-type material, an electron moves through the lattice, being FORWARD BIAS THEN REVERSE BIAS: E son. t C 1Mil t AMPLIFIER pnp
attracted by the positive hole, to fill the hole, this creates another
hole, and so on. The hole then appears to move to the negative Fig. 7.1 SOLID STATE JUNCTIONS, EXAMPLES
 92 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 93

by controlling the charge concentration in the base region. For E

0,
pnp the polarity is reversed and flow is as shown on Fig. 7.1. A
relatively larger emitter-collector current can be controlled by a
small base-collector current and voltage.
Current gain (collector to base current ratio) can easily be in
the range 10 to 200. Power gain may be as high as 50 000.

RECTIFIERS F
O2 ~
Metal rectifiers, dependent on barrier layer properties, such as
Fig. 7.3. CENTRE TAP TRANSFORMER RECTIFIER
copper-oxide and selenium are in general use. In the heavy
electrical industry mercury arc and ignitron units are employed.
For electronics the valve diode is well known as a rectifier and for
When E is positive with respect to F current flows through D 1 to
larger power, in conjunction with control applications such as the load and returns to G, O 2 is reverse biased. When F is positive
energising (firing) relays, etc., the gas-filled grid design is used- with respect to E flow is via O 2 to the load, returning to G, D 1
thyratron. The latter two have generally been superseded by the
reverse biased.
semi-conductor diode (pn junction rectifier) and the thyristor
(silicon controlled rectifier). More complex devices such as tun-
APPLICATIONS
nel, variable capacitance, microwave and four layer diodes are Low frequency (audio) signals cannot be efficiently radiated.
not considered in this book.
A high frequency (radio) carrier wave has a low frequency signal
The bridge rectifier is shown in Fig. 7.2.
impressed on it at the transmitter by varying or modulating the
amplitude of the carrier wave in sympathy with the low fre-
quency signal. At the receiver the signal information is recovered
A. A
from the carrier wave by a process known as demodulation or
detection. The wave at the receiver is first rectified (detected)
ae and the signal information recovered by passive networks
SUPPLY > i .+
(resistance-capacitance-inductance). These networks are usually
C de known as smoothing or filter circuits. Figure 7.4 shows a
s. ) v LOAD

CARRIER
UNMODULATED (hf)
Fig. 7.2 BRIDGE RECTIFIER

When A is positive with respect to B current flows through 0 1 to

the load and returns through O 2 to B. When B is positive with
~&:fJ~ c, C2
de
LOAD

respect to A flow is through 0 3 and returns through 0 4 ,

• I ,

The centre tap transformer rectifier is shown in Fig. 7.3. Fig.7.4 DEMODULATION CIRCUIT
94 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 95

transmitter-receiver unit with full wave bridge rectification- If the reverse bias is increased beyond a certain value, called
detection by semi-conductor diodes and a high frequency the breakdown voltage, the reverse current will increase sharply
capacitor (C 1) filter with smoothing choke inductor (L) and filter and this is known as avalanche or zener current. Rectification
capacitor (Cz). This is a capacitor input filter system, C 1 is often diodes never operate in the breakdown region but the zener
termed reservoir and C z smoothing capacitor. A capacitor pro- diode, used in voltage stabilisation and reference circuits, is
vides high impedance to dc and an inductor high impedance to operated in this region. Note in particular the symbols shown in
ac. The function of the filter may also be explained by regarding Fig. 7.5 to represent semi-conductor rectification, zener diodes,
it as an integrating network. In general conversion between ac thyristor (silicon controlled rectifier) and the different scales on
and dc is often required in instrument-control systems. Modula- the axes. Current flow is in the direction of the arrow, the bar to
tion, demodulation and re-modulation are frequently used in the indicate non-reversal (see also Fig. 7.2, bridge rectifier circuit).
electronic systems to utilise the best component for a particular
duty of sensing, amplifying, control, etc. Low voltage supplies ZENER DIODE
are often required for transistorised equipment. Typically a Figure 7.6 shows two applications of zener diodes. As a vol-
mains transformer, bridge rectifier, smoothing circuit and diode tage regulator (stabiliser) the reverse connected zener diode
(zener) volts stabilisation is used. conducts if input voltage is above breakdown voltage and current
from the supply is the sum of diode and load current. For input
SEMI-CONDUCTOR DIODE voltage increase then current increases through both R and the
When ac current is applied to the pn junction a large current diode but the diode resistance decreases and current through the
will flow when forward biased polarity applies and conduction diode further increases. A larger volt drop across R will occur but
stops when reverse biased and the diode acts like any other output voltage across the diode remains reasonably constant.
rectifier. This is confirmed by the characteristic shown in Fig. 7.5. Variations of input or output cause shunt of more or less current
through the diode resulting in constant voltage, i.e. across the
diode regulation circuit (see also Fig. 11.9).

~(TRIGGER)
40

THYRISTOR (SCR) R
RECTIFIER
DIoDE +1 /IItNo, , +1 I

20

40

BREAKDOWN
20

: 20 ~
0·5 * ',0 --+V
VOLTAGE REGULATOR"'t"'

Fig.7.6
~
..J
TRIGGER

ZENER DIODE APPLICATIONS

... I
,
'len
40~::L
w~

,
I W!
>
/: ~ The second diagram illustrates use as a trigger safety device.
ZENER DIODE ~
The relay will be held and will not fire until a certain prescribed
Fig.7.5 pn JUNCTION CHARACTERISTIC voltage is reached.
 96 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 97

AMPLIFIERS
Amplifiers are an essential part of instrumentation and con-
trol. The pneumatic amplifier (relay) is described in Chapter 10.
Thermionic valve amplifiers (triodes) have been extensively

---I
p
used although the solid state junction transistor is now preferred.
Feedback is an inherent addition but a detailed discussion on this
topic is reserved until the end of this chapter. Devices are active,

~)
l VARIOUS VALUES
i.e. utilising an external power source - as distinct from passive.
GATE I npn rI OF GATE CURRENT

ROTATING ELECTRICAL AMPLIFIERS

I p TRANSISTOR H
The separately excited generator is often used in control sys-
l /n tems. The Ward Leonard type unit is generally well known and

~
an application and description are given in Chapter 14. For
"',/....
",'"
higher sensitivity the cross-field dc generator, such as amplidyne
Si- VA- B - - . 0
and metadyne, can be used. A pair of brushes at 90 to the main
Fig. 7.7 THYRISTOR (SCR) brushes provides output and the main brushes are short cir-
cuited. A small input signal current to the control winding gives a
large change of current in the short circuit and hence the load
circuit.

THYRISTOR MAGNETIC AMPLIFIER

From Fig. 7.7 it will be noted that this is a pnpn sandwich. An amplifier of interest because of its use in electronic compo-
Consider the section shown as an npn transistor which requires a nents, often in series with transistor amplifiers. A high permea-
base current from the gate to cause conduction from collector to bility ferromagnetic core is wound with an ac current coil. The
e·mitter. With A negative both top and bottom junctions are
reverse biased. With A positive and nQ gate voltage the middle
INPUT
junction is reverse biased. As shown in Fig. 7.7, and with a pulse TRANSFORMER
of positive current injected into the gate, then as the pn junction ~i - - - - - - - . - - - . , 'I~
(top) is already forward biased the device is turned on. When
conduction exceeds about 10 rnA an avalanche (zener) effect
occurs and gate current is not required to maintain flow. Opera-
a::
tion is similar to the grid thyratron except that it is fired by de
INPUT
wi de
~ OUTPUT
current and not voltage. To stop conduction, voltage has to be u::
reduced to zero and the device requires another signal to fire.
Applications include inversion, stabilisation, regulation. High
currents can be controlled and the device is a very fast, high
power gain, switch. An application of the thyristor is given in
Chapter 14. Fig. 7.8 MAGNETIC AMPLIFIER
 98 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 99

inductance opposes increases of current until the induced With transistors the device is usually utilised in one of two
magnetic flux saturates the core when the reactance now behaves modes linear and non-linear. In the former sinusoidal signals are
as a resistor, i.e. saturated reactor and large increases in current amplified without distortion usually in two main types, i.e. small
can occur. A dc energised control winding on one limb of the core signal voltage amplification and power amplification. The non-
brings about, and varies, the degree of saturation and hence ac linear mode utilises switching from off to saturated condition
current flow in the gate limb. Bias and feedback windings are also very rapidly. Typical applications include: oscillators, i.e. square
incorporated on gates to improve flexibility and stability, and ac (or saw) wave supply generators, bistable (flip-flop) devices used
output can be rectified to dc which is filtered and appreciably in counting circuits, etc.; static switching and hold (memory)
amplified above the input. The typical unit is shown in Fig. 7.8; circuits, logic devices, etc. These applications are described later
ac input is from a transistorised oscillator and dc input is stabil- in this chapter and in Chapter 16.
ised.
Flux due to dc is unidirectional through gate windings A and
B but flux due to ac is, as shown, in opposite directions. The ac JUNCTION TRANSISTOR
output can be passed through a bridge rectifier and filter. Posi- The pnp junction is shown in Fig. 7.1. Due to initial diffusion,
tive feedback is utilised for gain adjustment without oscillation one junction (emitter-base) is forward biased and the other
problems because this unit is inherently very stable. The princi- junction (base-collector) is reverse biased. The former conducts
ple can also be used in a transducer with sensor core to vary heavily and positive carriers (holes) diffuse across the n region. If
inductance by movement. this region is arranged to have few electrons little combination
occurs and holes are attracted across the next junction by the bias
CLASSIFICATION voltage. Emitter current is the sum of collector current and base
Generally this depends on duty. If frequency is the criterion current. The collector current is much greater than the base cur-
the range is from zf (zero frequency - dc amplifier) through If rent, and proportional to it over a wide range, so that collector
(audio), rf, vhf to uhf (900 MHz). Generally the first two are of current can be controlled by the base current. Power for
interest here (see Fig. 7.20). emitter-collector current does not come from the amplifier input
Another classification depends on the equipment to be con- current, i.e. base current, but from an external power source. This
trolled by the amplifier. Voltage types give undistorted voltage amplifier is an active device and in system terms can be treated as
output and power amplifiers are to provide drive power. The a "black box", i.e. input, output and power source relationships
latter is used to provide power to recorders, controllers, etc. required without details of internal "box" arrangements. (For
A third criterion depends on the position of the bias point in npn, biasing voltages are reversed and conduction is mainly due
relation to the characteristic curve of the device. If operated near to electrons). pnp is used in this text.
the middle of the linear characteristic there is no distortion and Factors of importance are transfer characteristic (function)
this is termed a Class A amplifier, which is ideal for voltage (comparison of input and output), dynamic range (of power),
amplification but has a low power efficiency. If biased at or near efficiency (on power basis), amplification (gain) (magnitude of
"cut off' this is termed Class B and has a good power efficiency input and output voltages or currents) and frequency response
but severely distorted output - often half wave. Two such ampli- (transit times).
fiers can be matched as a push-pull amplifier, which via a trans- The transistor acts like a triode valve, except that no current
former gives undistorted output. Class C amplifiers are biased flows in the grid to cathode circuit of the valve but current does
past cut off and have the highest efficiency and greatest distor-
tion. { flow in the base-emitter circuit of the transistor. The input circuit
is biased to produce steady base current flow. Changes in input
100 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 101

current cause much larger changes in output current, i.e. current rent, therefore the input resistance is about 100 times higher and
amplifier. without the resistor the current gain is 99. The common collector
has similar characteristics to the latter. Currents shown in Fig.
Circuit Configurations 7.9, in rnA are illustrative.
Means to connect input, supply power, and to make use of
output are necessary. There are three ways to connect, differing TronsfrrCharocrer~ti~
in the way input and output are connected to the transistor, and These depend on circuit arrangement adopted as well as upon
each gives different characteristics which will depend on duty. the transistor itself. Consider the junction transistor as a current
Power must be consumed in the external circuit (load), usually by amplification device. The current transfer ratio (a) compares
resistors, and the collector current passes through the transistor collector and emitter currents and is less than unity (due to
and load resistor. leakage). Current gain ({3) compares output and input currents,
its value depends on the circuit configuration and is obtained by
COMMON BASE applying Kirchoffs law to the base junction and expressing the
result in terms of a. Current gain is less than unity for common
base and typically between 10 and 200 for the other two con-
figurations. Strictly current changes are compared. Transfer
0 01
0
..J0ogg characteristic is preferably linear. Equivalent circuits and "h"
parameters are used in transfer analysis.
~UTPUT The pnp transistor circuit (common emitter) is shown in Fig.
POWER (HIGH R)
CIRCUIT SOURCE CIRCUIT
7.10. In practice the common junction is often earthed, output
COMMON EMITTER

Bf: ~:-. J099

voltage tapped off after C to earth (via capacitor), bias battery
replaced by RC circuits and supply is from voltage power lines
not battery.

COMMON COLLECTOR

BIam C10099 Jl 00
0

C . =r •
Cl
Ie

Fig.7;9 CIRCUIT CONFIGURATIONS

Configurations are as shown in Fig. 7.9. These are common

base, common emitter and common collector. This depends on +
the electrode which is common to the input and output signal; fHOLES
common emitter is the most often used. From Fig. 7.9, for the
common base connection, assuming an ac input of say 45 mV
and 1 kD resistor in the collector circuit, then 990 mV will be .\ T II I I

+
developed across the resistor, i.e. voltage gain 20. For the com-
mon emitter connection the input circuit has a much lower cur- .1 Fig.7.10 TRANSISTOR CIRCUIT (COMMON EMITTER)
 102 REED'S INSTRUMENTATION AND CONTROL SYSTEMS l
!
I
ELECTRONIC DEVICES 103

Characteristics are as shown in Fig. 7.11 where variations of Parameters

base current (at input) give variations of collector-emitter cur- The common emitter circuit gives medium input and output
rent projected down from the load line. impedance, good current or voltage or power gain, inversion.
Common base gives low input and high output impedance, no
current gain but good voltage and medium power gain, in phase
operation. Common collector gives high input and low output
impedance, no voltage and power gain but good current gain, in
phase operation. The latter is often used for buffer stages con-
necting high impedance source to low impedance load (impe-
r
~
18 D-'5 mA
dance matching). The common emitter mode, pnp or npn, is
most frequently used for current amplification and especially for
E 18 0·' mA voltage amplification because a low input signal is required.
U
1-1 (Q
'10
Equivalent T-Circuit
(~
~,('- Presents a simple representation of the transistor to allow
O~ calculation of current, voltage and power gains. Such a circuit is
k shown in DoT Class One specimen examination question number
veE mV---..
18 at the end of the book (common base configuration).

Fig. 7.11 COLLECTOR CHARACTERISTICS (CE) UNIJUNCTION TRANSISTOR

A bar of lightly doped n type material, with base connections
B1 and B 2 at each end, has a p type material contact (emitter)
made between the two. If the emitter junction is reverse biased
Transistor Voltage Amplifier the bar acts as a potential divider. When input voltage is suffi-
It is necessary to include a high resistance in series with the cient to drive an emitter current it flows in the B 1 emitter region.
collector and tap off voltage variations across it, 250 gain can be This decreases the region resistance and emitter voltage (nega-
achieved. If connected as a common emitter a positive pulse into tive resistance characteristic). The device is used for trigger
the base makes it less negative and fewer holes flow from emitter operation, pulse oscillators, time delays and particularly for
to base. Collector current therefore decreases and the volt drop pulse generation to fire thyristors. In Fig. 7.12, C charges until
across the resistor decreases which causes collector voltage to voltage can eject emitter current. At a certain value of this
0
increase. There is 180 phase reversal between base input and
collector output voltage signals. Voltage characteristic slopes r------.... +
down, so collector voltages decrease for base voltage increase RL
but it is the amplification, not the negative slope, that is impor-
tant. If it is necessary not to have inverted output two amplifiers
in series are used - inversion is useful for negative feedback.
v
• , rei 1 • OUTPUT
vv
Common base connections are desirable for voltage amplifica-
tion but are difficult to design so that common emitter configura-
tions are usually found. Fig.7.12 UNIJUNCTION TRANSISTOR (OSCILLATOR)
104 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

current the voltage V collapses giving a negative spike output

pulse with rapid discharge of C. The capacitor recharges and
oscillation continues. This device is therefore a simple relaxation
I
~I
ELECTRONIC DEVICES

increased feedback which can be achieved by omitting C E and C

105

the interstage capacitor (direct coupling). Other methods can be

used such as emitter follower but this adds to complexity.
oscillator but is described here in the general context of the
transistor. POWER AMPLIFIER
Power dissipation must be kept low so a high efficiency is
SMALL SIGNAL JUNCTION TRANSISTOR AMPLIFIER important. Class B types are preferred but single transistors can
A linear mode device, Fig. 7.13, shows apnp common emitter not be used because the operating characteristic range results in
configuration multi (two) stage unit, the dotted line illustrates half waveform output so that a push-pull twin arrangement is
the staging, extension to more stages is achieved in the same utilised. Such amplifiers require high input currents and a power
additive way. In a triode valve auto grid bias is used and in (driver) stage is used, usually in Class A, the combination being
transistors a similar principle is required. A suitable bias point classified as AB (linear mode). A typical arrangement is shown
which is stable with temperature variation is essential. Several in Fig. 7.14.
ways, all dependent on feedback can be used. In Fig. 7.13 a
resistor R E in the emitter lead maintains a constant base voltage
DRIVER STAGE POWER STAGE

i i i " 6V

EpUT
INPUT
OUTPUT
INPUT

+. I I I' '
• I i I I I I I IOv
I
Fig.7.14 PUSH/PULL POWER AMPLIFIER STAGE AND DRIVER
Fig.7.13 TRANSISTOR 2 STAGE AMPLIFIER (SMALL SIGNAL)

and any undue rise in the leakage (temperature induced) current The driver transistor feeds to the transformer phase splitter.
causes the emitter voltage to fall. The emitter-base junction Inputs to the power stage are of equal amplitude and 180 out of
0

approaches reverse bias so reducing current through the unit. phase. Each power transistor conducts for half a period and the
The potential divider R 1 + R z keeps base voltage constant when complete waveform is restored in the output transformer. This
there is no input signal. Capacitor C E acts as a bypass for ac transformer has its primary ends connected to the transistor
components of emitter current. collector leads and centre tap connected to the more negative
Coupling between stages can either be by a transformer lead. Signal flux is in the output transformer core throughout the
method or, as shown, resistance-capacitor. Whilst current amp- whole period, with the complete waveform in the secondary
lifier design is easily arranged, voltage amplification requires a winding. Power supplies are commonly ± 6V, as an order of
high input impedance to the first stage. Essentially this needs magnitude.
106 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

dc AMPLIFIER
Most instrument signals require dc amplification of low vol-
tages. This is difficult to arrange as drift variation, largely caused
I
t
ELECTRONIC DEVICES

The FET is an n type bar (lightly doped) with electron input

107

(source) and output (drain) at its ends. On both sides of the bar p
type (heavily doped) material electrodes act as the gate (rever'se
by temperature variation in transistors, is amplified and passed biased relative to source). A narrow channel through the centre
on. With ac amplifiers the coupling capacitor excludes variations is controlled by the gate and impedance variation can arrange
but with dc units the coupling is usually direct. Differential two conditions, saturation and "pinch off' output voltage.
amplifiers have been used but perfect matching is difficult. Zener In Fig. 7.15 the FET varies its drain and source resistance
diode stabilisation and feedback circuits are also employed. according to voltage level and polarity of the input signal. Essen-
The solution often adopted is to use an inverter input to derive ·1 tially it is working as a make and break to produce square
I

ac from dc, direct ac amplifiers and a converter output to give dc I waveform. Output, in the absence of any restoring mechanism,
from ac. The converter is essentially a rectifier with smoother would be symmetrical about a OV line as dc level is lost in the ac
circuit, as described in the demodulation circuit of Fig. 7.4, or a amplifier. The output can be periodically shorted at B with
transistorised feedback integrator. another FET as at A, so that negative going portions are
The inverter is called a chopper and converts steady input into removed and dc level restored. A smoothing circuit would be
square waves, height proportional to signal strength and easily fitted at output. Source line could well be earthed. A FET can
amplified, by mechanical or transistor switching devices. When also give large current variation for small changes in gate voltage,
transistors are used voltage and temperature stabilisation are i.e. amplifier operation, if required.
required. The transistor switch occupies four successive states,
i.e. off (voltage applied between E and C with leakage current
OPERATIONAL AMPLIFIER
only flowing), on-transition (C current rising and voltage falling),
A combination of a high gain dc amplifier, i.e. chopped, ac,
on (C current flowing and saturation voltage only between E and
smoothed, together with feedback and input impedances is
C), off-transition (C current falling and voltage rising). Rating
required in control systems. The complete amplifier (with a gain
conditions require careful design and a small offset voltage,
of -A) is usually diagrammatically represented as a triangle.
temperature related, arises which causes problems. A field effect
Such devices are described in detail in Chapter 11. They are
transistor is very suited to chopping and a circuit is shown in Fig.
essentially voltage amplifiers, in cascade multi-stage form, gain
7.15, such transistors have negligibly small voltage offset with
10 5 - as an integrated circuit on a silicon chip, used in analogue
the distinct advantage of a very high input impedance.
computation and digital logic.

FEEDBACK ANALYSIS
This aspect has been mentioned previously but it is now neces-
IN ~ ~ c~OUT AMP I sary to consider a more detailed analysis before concluding the
work of this section. Consider Fig. 7.16 of a closed loop consist-
I
DRAIN I ing of a feedback amplifier of forward gain G and the monitored
fraction F of output fed back. Voltages are taken as the variables,
VI input, e error, Vo output (the more general terminology would
be OJ, 0, (
0 ), Subtraction of Vo from VI is arranged with an odd
number of amplifier stages, giving 180 phase shift, and adding
0

Fig. 7.15 FET CHOPPER CIRCUIT V o to V I; this is negative feedback.

 108 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

F
I ELECTRONIC DEVICES

Assume a change in G due to say variation of supply voltage,

ageing of transistors, etc. of as much as 50% reduction.
. 5 x 10 3
109

Overall Gam = = 98
FVo 1 + 10- 2 X 5 X 103

i.e. fall of 50% amplifier gain results in about 1% fall in overall

VI Vo
G gain. Negative feedback gives stability and accurate control.
Consider now the case of positive feedback.
Fig. 7.16 CLOSED LOOP (FEEDBACK AMPLIFIER BLOCK DIAGRAM) Vo=Ge
e = VI + FVo
Vo= Ge this, for electrical, is feedback voltage in phase with input vol-
e = VI-FVo tage.
Vo G
Vo = G(VI-FVo)
VI - 1 - FG
Vo = G(1 _ F VO ) Again there is increase of overall gain compared to "open loop"
VI VI
gain. However there is a marked disadvantage in that a small
Vo (l + GF)= G change in G causes a much greater change in overall gain and also
VI overall gain increases rapidly as FG increases so that when
Vo G
FG = 1 the overall gain is infinite. Output is available with no
VI - 1 + FG input, i.e. the amplifier is an oscillator. Positive feedback use
therefore causes instability and self oscillation. Oscillators are very
Now if the "open loop" gain FG is high compared to (i.e. for useful in instrument electronics but oscillatory situations in con-
a high value of amplifier gain G): trol systems must be avoided. Instability and oscillation shows as
practical hunting or cycling on the control system which is most
Overall Gain = ~ unsatisfactory.

This is independent of G, which can vary, and only dependent on

F which is fixed bY' accurate stable resistors. The resistive input OSCILLATORS
impedance is increased by a factor of 1 + FG. The marked
improvement in performance over the open loop amplifier is at An oscillator is an instrument for producing voltages that vary
the expense of additional amplifier stages made necessary by in a regular fashion. The output may be sinusoidal, i.e. a sinewave
increasing G to make FG much greater than unity. generator (harmonic oscillator) or square, triangular, sawtooth
Consider an amplifier with a gain of 10 4 and negative feedback in shape, i.e. relaxation oscillators. The latter are mainly impor-
fraction F one-hundredth, i.e. 10- 2 • tant in control system equipment. A wide range of harmonic
100l
oscillators exist from high frequency (LC types) to low frequency
Overall Gain = 4 = 99 (RC types) but apart from basic principles only the relaxation
oscillator will be described because of its practical applications.
110 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 111

BASIC THEORY RELAXATION OSCILLATORS

An LC parallel circuit in which inductive reactance equals Such oscillators are used in the non-linear mode where rapid
capacitive reactance is resonant oscillatory, resonant frequency switching of transistors from off to saturated (bottomed) is
equals l/27T j LC, negligible supply current and appreciable loop arranged. For sawtooth waveform a unijunction transistor, as
current (for an LC series circuit maximum supply current at this previously described and illustrated in Fig. 7.12, can be used.
frequency). In practice resistance is present and oscillations Alternatively a gas discharge tube with RC circuit is suitable
would diminish. If sufficient energy, correctly timed, is supplied usually with a thyratron (or thyristor) controlling time base strike
to the LC circuit utilising a transistor to make up losses the voltage. A more sophisticated method is to use a single amplifier
oscillation can be maintained indefinitely. This energy is sup- connected as an integrator (see Chapter 11). High current pulses
plied as correct phase and magnitude positive feedback. of short duration can be obtained from a blocking oscillator.
Referring to Fig. 7.17, LC is the frequency determining tuned For rectangular waveform generators a push-pull blocking
circuit and the remainder is oscillation maintaining. By mutual oscillator can be used as shown in Fig. 7.18. The transformer core
induction atL' positive feedback is arranged from the transistor. material has a square form hysterisis loop. When supply is

+ , •

INPUT (
t~
Fig.7.17 TRANSISTOR OSCILLATOR

HARMONIC OSCILLATORS Fig.7.18 PUSH/PULL BLOCKING OSCILLATOR

An amplifier providing its own input functions as an oscillator
and if the feedback is small is a pure sinewave generator. Typical
refinements for high frequency harmonic work include Hartley switched on (high starting resistance R) assume transistor 1 starts
and Colpitts oscillators. For lower frequency work in harmonic conduction so that a voltage is developed across the collector
generation the principle is similar to that above but the feedback winding (C) and induced across base winding (B). Polarity is such
line includes resistors and capacitors. A three RC section input that 1 is forward and 2 reverse biased so that the positive
gives 180 phase shift and as collector voltage in a transistor is
0
feedback turns 1 full on and 2 is off. Collector current from 1
180 0 out ofphase with base voltage this technique gives in phase rapidly builds up and the core of C saturates and voltage col-
between the two. A phase shift oscillator or Wien bridge network lapses which cuts off base drive to 1 and collector current starts to
is regularly used. Linear mode signal generators are of two types fall. A voltage is now induced in B to try and maintain this
namely constant voltage which is a low impedance source and current but reversed in sense. 2 is now biased on and 1 off and
constant current which is a high impedance source. the half cycle is repeated.
 ~
112 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC DEVICES 113

HEATER
Two state circuits, with abrupt transition from one state to the CATHODE
other, can be arranged withRC coupled amplifiers in place of the
transformer coupling described. The basic circuit has the output IMODULATOR
I ELECTRON LENS
of one transistor connected to the input of another and vice
versa. Such devices are called free running (astable) multivib-
rators. If instead of two unstable states, only one is arranged, the
circuit is called a monostable triggered multivibrator or univib- '-y----J
ELECTRON
rator (flip-flop). The effect of a pulse is to flip the circuit into the GUN

unstable state whereupon normal multivibrator action allows

flop back into the stable state. A flip-flop circuit is described in Fig.7.19 CATHODE RAY OSCILLOSCOPE
Chapter 16. Another method is to "clip" sinusoidal waves using
zener diodes. Power supplies often ±6V, i.e. of this order of
magnitude. Applications in analysis are numerous. A relevant practical
application of the CRO is display of a pressure volume diagram
from an IC engine. Screen scales can be calibrated directly in the
OTHER DEVICES required units and the application of integration gives power and
torque characteristics on display.
Electronic digital devices are greatly used in logic circuits and
computing, typical applications are considered in Chapter 16. RADIO COMMUNICATION

The speed of response and reliability of transistor switching The production, transmission and detection of sound or picture
greatly enhance its use in annunciator and control circuits. information is by electromagnetic waves - the range and spec-
Certain specialist equipment, mainly used in analysis, such as trum of such radiation is given in Fig. 7.20 (see also Fig. 7.4).
ultra violet recorders, wave and transfer function analysers, etc.
is best described from specialist literature if required. Perhaps
the most important in this respect is the cathode ray oscilloscope
FREQUENCY Hz
(CRO).
3 6
• 9 12 15 18 21
10 10 10 10 10 10
"j
10
CATHODE RAY OSCILLOSCOPE 10 3 10° 10-3 10-6 10-9 10-12
The oscilloscope is shown in Fig. 7.19. The full circuit includes .. WAVELENGTH m
power supply packs and amplifiers. In addition a relaxation RADIO I X
lOIlG SHORT VHF UHF RAOAR I
oscillator timebase (multivibrator saw tooth generator) is
required.
I
I
I RED U VIOLET
~
In Fig. 7.19 the electron beam from the gun is focussed on to a VISIBLE
Fig.7.20 ELECTROMAGNETIC RADIATION SPECTRUM
fluorescent screen and the glass tube is under high vacuum
(pressure 10- 6 mm Hg). Input signals represented as voltages are
connected to the Y plates and deflections produced are shown on
the vertical axis of the screen. The timebase circuit is connected
to the X plates and the beam is deflected horizontally from left to
right with uniform speed and then returned in almost zero time.
114 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

TEST EXAMPLES 7

1. Explain why electronic devices are inherently suitable for

display systems and generally unsuitable for actuating systems.
Describe any method by which a small electrical signal may be
amplified. CHAPTER 8

2. State the purpose served by transistors. Sketch the circuit

diagram for an npn voltage amplifier. Explain the precautions FINAL CONTROLLING ELEMENTS
which should be taken regarding care, handling and operating
environment for transistorised devices.
3. Describe a half wave rectifier. Sketch the output wave form:
(i) without, (ii) with a capacitor across the purely resistive load The element which acts directly on the controlled body, pro-
terminals. What modifications are necessary to give full wave cess or machine. In a servo-mechanism this is a servo-motor,
rectification? rectilinear or rotary, receiving amplifier output and driving the
load. Correcting unit (motor and correcting element) is applic-
able to process control systems. An actuator is a motor with
limited rotary or rectilinear motion.

CORRECTING UNITS
Such devices may be pneumatically, hydraulically or electri-
cally operated or a combination and the word motor is applicable
to all.

DIAPHRAGM OPERATED CONTROL VALVE

This unit, as described, is pneumatically operated. Description
can be given in three parts namely motor, valve and positioner.
The sketch given (Fig. 8.1) refers to a fuel variable but is applic-
able to any similar variable.

Motor element
Air pressure acts on top of a synthetic rubber diaphragm and is
opposed by upward spring force, oil flow is right to left, hand
regulation is possible and the fail-safe position is shut (up). The
pressure-stroke characteristic is based on linear which requires a
large constant area diaphragm, minimum friction and a linear
spring force-deflection characteristic. A limited travel motor-
actuator-reverse action.
116 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FINAL CONTROLLING ELEMENTS 117

contoured valve plugs can be designed for proportional control.

Characteristics are then a true linear relation between flow and
, , t"I ' ,HAND JACK valve travel for the former and near parabolic equal percentage,
i.e. equal increments of valve travel give equal percentage
change in existing flow, for the latter.
· Maximum controllable flow %
T urn D own R atlO = -----,.---_~~~::--_..,,--:-
Mimmum controllable flow %
i.e. each a percentage of the theoretical (100%) flow.
Positioner
Such devices are necessary when:

(b) the valve is remote from the controller,

(c) the medium being controlled is viscous,
(d) there are high gland pressures required.
Essentially each of the above effectively increases friction,
hysterisis or unbalanced forces acting on the valve spindle. The
positioner provides extra power to position the valve accurately
and speedily to offset these effects.

Fig. 8.1 FUEL OIL CONTROL VALVE WITH POSITIONER

~lY ~%t,
Correcting element (valve)
The valve can be single seated reverse action as shown, or
direct action single seated, or double seated (direct or reverse) I

which give balanced valve forces and less operating energy and
VALVE
are widely used. Materials for all components depend on the MOVEMENT
medium being controlled. The overall flow characteristic
requires to be assessed for the piping system as a whole, as well as
for the valve, to achieve design conditions. In general valves may
be simplified into three types of variable % Flow-% Valve Lift f
CONTROLLER
characteristics. Mitre valves with wings (bevel or poppet) usually OUTPUT PRESSURE
P
give inverted near parabola characteristics best suited to on-off
Fig.8.2 VALVE POSITIONER
operation. Vee port (in wings) high lift or modified parabolic
118 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FINAL CONTROLLING ELEMENTS 119

A motion feedback device from the valve spindle senses devia-

tion between the desired value position input signal and the
actual valve position and supplies extra correcting power. A
flapper is connected at one end to the valve spindle and to a
pressure bellows at the other, with the nozzle between the two.
Increase of p increases pressure on the diaphragm until valve
movement restores the flapper to the throttle position and
equilibrium is restored. Further speed boost can be added, as
shown, by inserting a pneumatic relay to apply full air pressure to
the diaphragm. Both positioner and relay (as described in Chap- MOTOR
ter 10) act as pneumatic amplifiers. Valve positioners are very
often used in sequence operation of valves in a system, i.e. split
range control (Chapter 13). Double seated valves (two inlets
through ends, outlet at centre) are often used giving three way AIR
mixing and bypass arrangements on engine coolant systems.
--.----, I
I

PISTON OPERATED CONTROL VALVE

Double seated valves with balanced valve forces are invariably Fig.8.3 TORQUE ACTUATED VALVE
utilised with diaphragm control valves because of the charac-
teristics of this actuator. However the piston type actuator gives
powerful valve forces, long stroke and accurate positioning so
that single seated valves can be used which often have a more
desirable flow pattern and require less maintenance. In the
direct acting type a loading pressure on top of a piston (down
An alternative design utilises electric motor gear drive to the
force) is maintained constant (supply air via combined
spindle by a similar arrangement. The motor must be reversible,
reducing-relief valve). Actuating air on the bottom of the piston
variable speed and fitted with limit switches.
(up force) is controlled in pressure by a small relay pilot valve,
diaphragm operated from input signal, and connecting to supply
(open up) or vent (close down). ROTARY CYLINDER CONTROL VALVE
The valve consists of a semi-rotary shutter operating in the
square throat section of the valve passage. Operating torque is
TORQUE ACTUATED CONTROL VALVE transmitted to the shutter gate via a spindle perpendicular to flow
There are two air motors, one for each direction of rotation of direction. The "swing through" design of butterfly valve requires
the valve spindle. Motors are transverse to spindle and rotate the a good spindle seal and tight shut off is best arranged by closure
spindle by a ratchet and toothed wheel. Motors are standard of shutter against a flexible disc. Such valves are best suited to
diaphragm or piston mechanisms. High torques for large valve throttling operation between about 15° and 60° for reasonable
forces can be achieved by multiple diaphragms or duplication of tum down ratios. Torque requirement is not high and linear
motors. A diagrammatic sketch is shown in Fig. 8.3. actuators of any design are suitable.
120 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FINAL CONTROLLING ELEMENTS 121

WAX ELEMENT TEMPERATURE CONTROL VALVE SERVO-MOTORS

This type has a copper capsule containing wax whose expan-
May be rectilinear or rotary; operated by air, fluid or electri-
sion varies with temperature. Movement is transmitted via a
city; applied in either process or kinetic control systems.
diaphragm, plunger and linkage to vary the position of a shutter
which rotates in the valve body. The valve is best suited to mixing
or bypass conditions, control normally being limited to a range of dc ELECTRIC MOTORS

temperature of about 10°C, fail-safe inherent in design. While Essentially the servo-motor is a conventional motor, series
the self contained simple design is attractive for many duties it is shunt or compound, with control of field current or armature
a fixed control. The sketch of Fig. 8.4 is a diagrammatic represen- voltage by the controlling device. High torque and low inertia is
tation. Temperature rise causing down-movement of the required so that armatures are reduced in diameter and leng-
plunger, clockwise rotation, closing right and opening left outlet thened. Good commutation over a wide range of speeds is neces-
connection. sary and design must allow for peak transient changes. Perfor-
mance is limited by heating caused by high armature currents and
magnetic saturation of iron paths. Reversal is arranged by revers-
ing the current through the field or armature via the controlling
device, which is generally satisfactory, although split field motors
can be used.

t
INLET
ac ELECTRIC MOTORS
The three phase induction motor is a most desirable machine
in electrical work, being cheap and reliable. Unfortunately start-
ing torque is low and the torque-speed characteristic is non-
linear so that control is difficult. For servo use the torque charac-
teristic can be improved by using high resistance rotors which
unfortunately generate extra heat and cost. Commutator motors
are available but add to complexity and cost. Thyristor circuitry
offers the most possibilities for improvement.
The two phase induction motor is used in low power systems
~ especially for position control. Applications include instrument
OUTL~~ potentiometers, bridges and pen recorders. Such a motor has two
stator coils wound at right angles, which are fed with alternating
currents 90° out of phase, to produce the rotating magnetic field.
Fig.8.4 WAX ELEMENT ACTUATED VALVE For reasonable modulation, torque is proportional to the two
currents. If a fixed voltage and frequency is applied to one
reference winding, then torque is proportional to the voltage of
the other winding, which is connected to represent the amplitude
of the control signal. Characteristics, especially with a high resis-
tance rotor, are reasonable linear over a limited range. Heat
generation at reference field and rotor are high.
122 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

Single phase motors for servo systems are unsatisfactory

, FINAL CONTROLLING ELEMENTS 123

Slipper pads bear against the swash plate face and plungers are
except for on-off control and special starting arrangements, such driven in and out axially for each revolution of the rotor. The
as split phase, are required. swash plate movement varies effective stroke and can reverse the
Synchronous motors can be used for low power drives such as
pen recorders, etc. where synchronism of timing is required.

HYDRAULIC RAM SERVO

Used for linear actuation, i.e. ram or jack type, but can be
utilised as a torque device with multiple rams. Generally a
I
t
flow. A number of axial plungers are used in the rotor. Delivery
can be to an identical motor with fixed swash plate.
Other types, particularly using a booster supply pump, have a
similar design and can give discharge pressures over 140 bar.
Higher pressures give smaller components and very positive
action.
medium control performance system used in position devices.
Normally short stroke but can operate with long stroke and high
pressures (ship steering gear). There is little in the construction HYDRAULIC ROTARY VANE SERVO
calling for special attention. Details are as sketched in Fig. 8.6. Rotation depends on which
side of the vane is connected to the pump pressure feed, this
HYDRAULIC VARIABLE DELIVERY PUMP should be clear from the plan view as sketched. The large rotary
Not strictly a final controlling element but has important vane unit is normally designed for a maximum pressure of about
applications in hydraulic control systems with output to hyd- 90 bar as distortion and leakage are liable to occur at higher
raulic servo-motor (linear or rotary). The swash plate pump is a
good example with alternatives such as Hele-Shaw eccentric
slipper, tilting trunnion, cam operated ball piston, etc. The hyd-
t t'
pressures. The design is simple and effective. In fact the apparent
space and weight saving is not as great as may be imagined due to
the higher pressures and integrated construction utilised in mod-
raulic motor is virtually a reverse pump and has similar construc- ern hydraulic ram designs. There is however, a definite space
tion. Consider Fig. 8.5 illustrating the swash plate type: saving but the first cost is higher. Absorption and transmission of
torque relief is essential to avoid excess radial loading of vanes.
The three vane type is used for rudder angles to 70 0 • Steel sealing
strips backed by synthetic rubber are fitted into grooves along
the working faces of rotor and stator vanes.

VALVE PLATE
PLUNGER t
:'(
PNEUMATIC PISTON SERVO
Refer to Fig. 8.7:
ROTOR
The pilot valve has two outlets, one to the top and one to the
bottom of the servo-piston. If the valve is displaced from its
neutral position then pressure at one port increases whilst at the
other port it decreases, so causing piston movement. The move-
----- I I ment of the piston is arranged, via linkage gear and cam, to vary
the tension on a spring giving an opposing moment to the signal
pressure on the bellows. When these two moments balance the
SWASH PLATE
pilot valve is at mid or neutral position and the pressures on each
Fig.8.5 VARIABLE DELIVERY PUMP side of the servo-piston balance, so locking the piston.
124 REED'S INSTRUMENTATION AND CONTROL SYSTEMS FINAL CONTROLLING ELEMENTS 125

(In the case of a diaphragm valve the pilot valve would only
have one outlet to the diaphragm top, the valve stem movement
via the gear linkage and cam would alter the spring restoring
movement until equilibrium existed.) An incoming air signal

OIL MANIFOLD
ti inflates the bellows causing the balance beam to pivot about the
right hand end and operate the pilot valve. This produces a
second air signal whose relationship to the first signal is depen-
dent on the tension in the spring attached to the beam end point.
The mechanical movement of the valve, or other control
device being operated, is transmitted by a driving rod to the cam
t ELEVATION
and linkage, thereby adjusting the tension in the spring. By using
a suitably shaped cam the position of the regulating unit to which
(PART SECTION DETAIL)

4
the positioner is attached may be given a predetermined rela-
tionship with the incoming air signal.
1/ 120 bar change in signal pressure will give full travel to the

pilot valve which could give 1 1/ 3 bar variation in output pressure.

FULL PLAN
(PART SECTION DETAIL)
l~

1
-I
BELLOWS I
~

Fig.8.6 HYDRAULIC ROTARY VANE SERVO

Fig.8.7 PNEUMATIC PISTON SERVO AND POSITIONER

1
126 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

OTHER SERVO-MOTORS
Variable couplings and clutches can be used, either hydraulic
or electric, for rotary speed control. Electromagnetic solenoid
devices, linear and rotary, are available for many purposes
including electrical contact and relay operation and incremental
digital stepping motors. CHAPTER 9

TEST EXAMPLES 8 PROCESS CONTROL THEORY

1. Make a detailed sketch of a simple diaphragm operated
control valve, such as a reducing valve, although any other type
of a similar control valve will be accepted. Analyse the action of
the interconnecting elements, that is, those parts effecting con- TERMINOLOGY
trol. The following definitions should be considered with Fig. 9.1 to
Explain how load changes are sensed and the command signals which they mainly refer round the loop system. Other definitions
are transmitted to the actuator. of terms are given at the appropriate place in the text to cover
2. An air damper is controlled in position by variable air pres- and clarify essential points. Process control is concerned with
physical quantities involving variables such as temperature, pres-
sure to a pneumatic actuator. Sketch and describe the system,
sure, flow, level, etc.
describe the actuator and explain its mode of operation.
About twenty terms are exactly defined immediately below as
3. Sketch and describe a control valve of the wax element type from British Standards which together with Figs. 9.1 and 9.4 are
and show how it is incorporated into a coolant system. from the same source.
These extracts are from B.S. 1523: Part 1: 1967, a Glossary of
Terms used in Automatic Control and Regulating Systems (Part 1.
Process and kinetic control). They are reproduced by permission
ofthe British Standards Institution, 2 Park Street, London, W 1 Y
4AA, from whom copies of the complete standard may be
obtained.
(

CONTROL SYSTEM
An arrangement of elements (amplifiers, converters, human
operators, etc.) interconnected and interacting in such a way as
to maintain or affect in a prescribed manner, some condition of a
body, process or machine which forms part of the system.

;'J PROCESS CONTROL SYSTEM

A control system, the purpose of which is to control some
41 physical quantity or condition of a process.
128 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 129

COMMAND SIGNAL CONVERTED DEVIATION

The quantity or signal which is set or varied by some device or A physical quantity related only to the deviation, and normally
human agent external to and independent of the control system proportional to it, but of a different physical kind suitable for
and which is intended to determine the value of the controlled operating the amplifier element. It may also be used for operat-
condition. (Symbol 'Yj.) ing other elements in the system. (Symbol (J = (Jo - (Jj.)

SET VALUE(Set point) (Author's note: often written (J = (Ji - (Jo, 'Y = 'Yi - 'Yo)'
The command signal to a process system.
SIGNAL PROCESSING
DESIRED VALUE The processing of the information contained in a signal by
The value of the controlled condition which the operator modulating, demodulating, mixing, gating, computing or filter-
desires to obtain. ing.

INPUT ELEMENT CORRECTING UNIT

The element which is included, when necessary, to convert the Of a process control system. The single unit containing the
actual command signal into a converted command signal suitable motor element and correcting element in a process control sys-
for operating the comparing element. tem.
Note. In some systems there is no input element as the command
signal is taken, without conversion, direct to the comparing ele- MOTOR ELEMENT
ment.
The element which adjusts the correcting element in response
to a signal from an automatic controller.
CONVERTED COMMAND SIGNAL
A physical quantity related only to the command signal, and
CORRECTING ELEMENT
normally proportional to it, but of a different physical kind
The final controlling element in a process control system.
suitable for operating the comparing element or the co-
ordinating element. (Symbol (Ji')
Note. In definitions where there is no ambiguity the term )! CONTROLLED CONDITION

"command signal" will be used to imply the command signal itself The physical quantity or condition of the controlled body,
or the converted command signal. ( (
process or machine which it is the purpose of the system to
control. (Symbol 'Yo.)
COMPARING ELEMENT
The element which accepts, in physically similar form, the DETECTING ELEMENT
command signal and the controlled condition, or their equiva- The element which responds directly to the value of the con-
lents, and determines the deviation or the converted deviation. trolled condition.

DEVIATION ~j
MEASURING ELEMENT
The difference between the measured value of the controlled The element which responds to the signal from the detecting
condition and the command signal. (Symbol 'Y = 'Yo - 'Yi') 'Il element and gives a signal representing the controlled condition.
130 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

a
W z
LOAD ::::lo
0;::
The rate at which material or energy is fed into, or removed ~C5rh°
zz
a
from, the plant (on a process control or regulating system). 1. 00
uu w
...J
The controlled device. 2. The properties (e.g. inertia, friction) of ...J
o~
the controlled device that affect the operation ofthe system (o/a ~z
Z z<{
kinetic control system). o
_ O...J
Ull.
~
U

MEASURING UNIT ( .. L ---11 '-... ·--ll

A unit which gives a signal representing the controlled condi-
tion. It comprises a detecting element and measuring element.
,!::
:z
uZ'
WW I .
I <.!)
z~

Note. Such a unit is used as the monitoring element ofa process I~ ~§

, - ,z uw ,
I: 5~
w~
~w Ii ~oo
control system. :@I';: ~jj wI-
0:: ~ I ~ I-Z
ooW
~ o::w :~~ ,
>~
:I u ~~ I IZ~ fil: oow

..-~ ~~ ~ ~5 I
CONVERTED CONTROLLED CONDITION 0...J I ::>w ...JZ ...J...J
~w OW
A physical quantity related only to the controlled condition
and normally proportional to it, but of a different physical kind
fila
w~z
L_..
'
<.!)<.!)

,0::0::
69

zz
Q:LL
1-0
w0::9 ,::>0 00 Zoo
suitable for operating the comparing element or the co- ww~ i~F uu Oz
Uo
u>~ ~ ~

11~_~~m_- --:.~---. !~~-_. -~ om

ordinating element. (Symbol 80 ,) u-
Note. In definitions where there is no ambiguity the term con- ~~
trolled condition will be used to imply the controlled condition I :' ~­
I-~
>=~ii5~ ~ II
II <.!)~<.!) <.!)
Z Zw z :::>0
itself or the converted controlled condition. <u Ii:
~u;l3~ ~~
L UII u;e"
: " wW
...J~
it a g~ <{ ~ :
~ZO::...J 0..0
I

AUTOMATIC CONTROLLER (Automatic regulator) a

wz <{<{Il.W a W...J
~w ~oo
~o Ow ,
A portion of an automatic controlling or regulating system in 0::- W...JZ --. --- Z
~~ w-:§ ~...JQ
which a signal representing the controlled condition is compared z> w~!:::
ow >~a ~O
with a signal representing the command signal and which oper-
ates in such a way as to reduce the deviation. 11
ua <.!)
~~
o::z
zzz
000
uuu I
• c:D:I:
i!OO
~~
Note. The two functions o/an automatic controller or regulator, ~w crP
(,
namely to determine the deviation and to generate the control awa
~z
8jj I'
signal dependent on the deviation, are in many devices carried out
by two separate parts, the comparing element and the controlling ~~~cD-
z~<.!) ~ 0::w
~ '

element respectively.
00-
uuw ~ I ~ I
MONITORING FEEDBACK
.
I L-_~-==-.__-=w_~_J
~w

~§
I

II

.____
6
~
8 .
J'
The feedback of a signal representing the controlled condition
a
along a separate path provided for that purpose, for comparison z
<{...J
~<{
with a signal representing the command signal to form a signal ~z ?>---
j

o~
representing the deviation. ~
uw
132 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 133

PROCESS ( I
Consider a detector element which is directly inserted in a
The act of physically or chemically changing, including com- pipeline. The fluid flowing increases in temperature at a uniform
bining, matter or of converting energy. rate of say 10°C in one minute.

OFFSET (Droop)
Sustained deviation.

OVERSHOOT
The difference between the maximum instantaneous value of
~) ----------r------- kT
the step function response and its steady state value.
f
u ---------- --
o I
DEAD TIME (/)
...J I
The time interval between a change in a signal and the initia- «
z I
tion of a perceptible response to that change (a dead band region C) I
U5 I
may exist on a controller). u..
o
INPUT SIGNAL" I
eF I
w
=>
...J
I OUT PUT SIGNAL!
CASCADE CONTROL SYSTEM :) ~
I eE I
A control system in which one controller provides the com- I I
I I
mand signal to one or more other controllers. .-, I... ..,. .1
I I
I I
SETTLING TIME I I
The time taken to approach a final steady state within specified I
2 6
limits. TIME, minutes ~

Note. The settling time depends on the limits specified and is

Fig. 9.2 LINEAR RESPONSE OF DETECTOR ELEMENT
meaningless unless these limits are specified. Example. In the
particular case of a series LR circuit subject to a step function of
voltage, the settling time of the current to within 1 per cent of its .i
final value is approximately five times the time constant L/R. Referring to Fig. 9.2:
( i
The first indication of temperature change at the detector
!
element may be after about say 2 minutes and there may be a
constant lag at a given reading of about 6 minutes (ramp input).
If OF is the fluid temperature and OE is the element tempera-
RESPONSE OF DETECTION ELEMENTS
ture:
Time lags obviously occur in a plant due to the individual lags OF - OE oc CERF
of components and transmission of signal lags. The lags must be
C E is thermal capacity of element (mass times specific heat)
fully evaluated before the control design can be established. As ~
R F is liquid to element thermal resistance to heat flow
an illustration the lag of a temperature detector element can be
considered: OF - OE = kCERF
134 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 135

where k is a constant. -:L Considering the case of a temperature detector element in a

Measuring lag is OF - OE (say DC) pocket then to reduce lags the time constant (CR) must be
C E R F is the time -constant T (say minutes) reduced to give quick response and the following would be aimed
at:
()F - ()E = kr
This lag consideration is based on linear variation, if the variation 1. A close fitting thermometer in a pocket with immersion in a
was exponential then the measure lag is usually arbitrarily high conductivity fluid.
defined in terms of the time it takes for the output signal amp-
litude to reach 63 ·2% of the input signal amplitude. The lag time 2. Clean fittings and a high velocity for the fluid to be meas-
on the sketch is given in minutes but should preferably be ured (turbulent flow).
reduced to seconds in practice. Fig. 9.2 assumes fairly heavily
damped response. Under-damped response would show in oscil- 3. Light, good conductivity pocket material using deep immer-
lation curves about the line OE. sion into the flow of fluid.
If the disturbance causing the variation is a continuous sine
form variation the appearance would be as in Fig. 9.3, note that 4. Reduced piping distances, friction and inertia effects.

PHASE DISTANCE VELOCITY LAG

LAG That time interval between an alteration in the value of a signal
INPUT tP
SIGNAL and its manifestation unchanged at a later stage arising solely

it
\ OUTPUT SIGNAL
from the finite speed of propagation of the signal.
/(~S~N~)
For example the time it takes for a heating effect to travel with
the fluid from heat source to detection element along a lagged
o -_
pipeline. Lag = Distance/Velocity. Causes phase lag. Theoreti-
(i; - ....... .... cally no magnitude change.
g, TIME- ........
.......
~
-l
~-
~ TRANSFER LAG
That part of the transmission characteristic, exclusive of
GAIN (AMPLIFICATION) = die distance-velocity lag, which modifies the time-amplitude rela-
ATTENUATION (LOSS) = e/d tionship of a signal and thus delays the full manifestation of its
ep = tan-1 2 ~.,. radians influence.
For example the measure lag as given previously for the detec-
Fig.9.3 STEADY STATE RESPONSE TO SINUSOIDAL SIGNAL tion element which is dependent on Rand C. Causes phase lag
and reduces amplitude.
The aim is to keep inherent lags as small as possible, together
the response has a reduced amplitude and has a phase lag. T is the with reducing inertia and increasing stiffness (or their equival-
time lag of detecting element, n is the period of process distur- ents), for the system. The alternative is to increase system gain
bance, attenuation applies as gain is less than unity. but this can create instability.
136 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 137

TYPES OF CONTROL ACTION A typical example of two step controller action would be liquid
level control in a tank with a varying supply and a required
Can be illustrated by any variable; level is selected for this continuous and steady outflow. Overlap, which could be adjust-
chapter. able, allows working between predetermined limits (differential
gap or overlap) and gives less irregular action, this would be
TWO STEP CONTROLLER ACTION (Basic action. 1.) suitable, as another example, for refrigeration motor cut in and
The action of a controller whose output signal changes from out control by room temperature.
one predetermined value to another when the deviation changes Lag tends to allow overshoot and this should normally be
sign. reduced to reasonable limits. The closeness of control is influ-
It should be noted that the limits are not necessarily on-off enced by the capacity of the system and also that property of an
although this is often used (see Fig. 9.4), especially in digital uncontrolled system to reach equilibrium for a fixed set of condi-
systems (Chapter 16). tions (inherent regulation).
Such action of two step control is a simple but most useful
method with numerous applications in practice. By arranging
more chosen values and corresponding correcting signal steps
the control can be made much closer with less overshoot, this
~
method is termed multi-step controller action.
z )'01 --
II -~T-
o --- --
i= I Note. The more simple the control principle the better. Addi-
oz tions or refinement such as integral action, derivative action, etc.,
-T----- I OVERLAP
8
Q
)'02
II I ---- ---T
I ---- 1- as covered later, should only be applied where the requirements of
UJ the process control definitely require these modifications.
-J
-J
I I I
o
a::
I-
z
I I
I ,II
o TIME PROPORTIONAL CONTROL ACTION (P) (Basic action. 2.)
u I
I The action of a controller whose output signal V (or <l» is pro-
I
I portionate to the deviation O.
I-
::J
I
I
ois the difference between the measured value of the control-
a...
I-
I led condition 00 and the command signal OJ.
::J
0>
I
~I VOC - 8
a::-J
UJ«
-Jz
V= - KIB
-J(!)
0-
a::(/)
I-
z
I The negative sign denotes that the correction signal is opposite in
o direction to the deviation. K b a constant depending on the
u
controller characteristic, is called the proportional action factor.
TIME
Potential correction <l> (change of actual controlled condition
Fig.9.4 THE ACTION OF A TWO STEP CONTROLLER WITH OVERLAP 'Yo) is proportional to the movement ofthe correcting unit (which
depends on V).
<l>ocV
Q)=CIV
138 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 139

where C 1 is a constant depending on the correcting unit charac- increases to 11 m the valve movement to shut in is 0·05 (11 - 10)
teristic. = 0·05 mm, i.e. the new valve position is 0·3 mm from the seat.
Now There is no controller gain here, much the reverse in fact.
V= - KIf)
Proportional band
<I> = - KICIf)
That range of values of deviation corresponding to the full
<I> = -fLf) operating range of output signal of the controlling unit, from
proportional action only.
JL = <I> / (J and is numerically the proportional control factor, or
This band can be expressed as a percentage of the range of
simply the controller gain, a typical value, in pneumatics, may be
values of the controlled condition which the measuring unit of
about 15.
the controller is designed to measure (see Figs. 9.6 and 9.7).
SET
l.p.l. q .1 VAWE

-K 1 = t = 0·05 mm/m CONTROLLER

RATIO.9.. ~ lxl00 ~~OPORrlo
P O'OS ADJUSTING
...~..... 100 I\t~
'" '"
_ 2000 SCREW ___ c.,v' ~ .<~
--1-
/22{/ \ ~00'\'1;
INDICATOR

1
CORRECTING
•
SIGNAL

MEASURING INSTRUMENT
FOR VARIABLE

25m

~~ ,1
PROPORTIONAL BAND IS PER CENT POINTER
TRAVEL AT INSTRUMENT TO GIVE TOTAL
VALVE TRAVEL AT CONTROL VALVE
100 PER CENT PB. POINTER FULL SCALE FOR FULL V/V TRAVEL
20 PER CENT PB. POINTER Ysth FULL SCALE FOR FULL V/V TRAVEL
, -'-V" / / V / / / / / /,c L 200 PER CENT PB. POINTER FULL SCALE FOR 'Y2 FULL V/V TRAVEL

Fig. 9.6 PROPORTIONAL BAND (NOT TO SCALE)

t
Fig.9.5 LIQUID LEVEL CONTROL BY SELF OPERATING CONTROLLER
(Total valve span) x 100
Proportional band = (Total measure span) x KI
Referring to Fig. 9.5 and assuming linear characteristics. It is
desired to maintain a fixed height h in the tank, the outflow
demand varies. As this is self operating (no intervening medium For the example for the given level controller, if the full meas-
such as compressed air) then we can utilise V, K 1 and (J as sym- urement scale is from 0 to 20 m head, i.e. 100%, and the full valve
bols. If h is set for 10 m, assume the valve is then 0·25 mm from stroke is 0·5 mm, i.e. 100 % then 10 m fully strokes the valve (i.e.
the seat, K 1 = 0·05 mm/m (decided by leverage), then if h 0·05 x 10 = 0,5).
140 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 141

Proportional band = ~~ x 100 = 50 % Note:

For good control the following are essential:
alternatively: 1. A high deviation reduction factor (hence high p,), i.e. small
deviation from set value after a disturbance. High p" means high
0·5 K h highly sensitive, narrow proportional band, etc.
Proportional band = x 100 = 50 %
0·05 x 20 2. Minimum offset.
If the proportional bandwidth is narrow then a big controlling 3. Low value of subsidence ratio at short oscillating period, i.e.
movement is required for a small deviation, the control is sensi- quick return to set value after a disturbance.
tive, i.e. high value of K 1 and a small offset results (see later). Too These are achieved by plant analysis. Widening a proportional
narrow a proportional bandwidth can however cause instability band causes an increase of offset, of damping and of period of
and hunting. The practical result is a compromise, the set value oscillation.
must of course be within the band.
fI
o
INITIAL LOAD

9
20%
PROPORTIONAL
BAND
SUBSIDENCE RATIO= alb
(FOR DAMPED OSCILLATION)
~

100 I f
w
I ::::)
...J
I ~
w I Z
o
~ 80 I i=
FINAL OFFSET
<{ 5
a:::
.....
I ~ 1'_ ",,)1-1 .""'LV~ r -DESIRED VALUE
::::) I u
Q..
I o
~ 60 I
W
...J
o ...J
I oa:::
a::: .....
w I z
...J

g..... 1,.0 8 LI __:-:-:"=--"'"'"':'""-----------

TIME - - .
z
o Fig. 9.8 PROPORTIONAL CONTROLLER ACTION
u
~
20

Offset
Is sustained deviation due to an inherent characteristic of
60 80 100 proportional control action. (It should be noted that with all pro-
UNIT RANGE portional controllers the set value differs from the desired value
Fig.9.7 PROPORTIONAL BAND by varying amounts depending on the given load conditions.) If
K 1 is large, for a given deviation. the offset will be small, K 1 is depen-
dent on the proportional band of the controller (see Fig. 9.8).
Consider the following analogy (as shown in Fig. lOA):
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 143

spring hanging from a support, with a mass of say 10 kg on

nd, will have a certain extension and the mass will be at a
ain vertical position X. If the mass is pulled down and r,SUMMATION SIGNAL ~~.;g:VAlVE I S.. '---~-~=---
S B'EKUT l-..N
released it will oscillate but will finally return to X. This is an I POSITIONER
" OFF N 0 OL.J . o. 80
example of proportional action in that the spring restoring force
I 6...:-,...-- L
-\~
is proportional to the extension (also opposite in direction). The ;{......- -.............. -- -
desired value position X is always reached. - - - -'~EGRAL
' ,
.
-- ~
-::;-INTE-;;RAlJ
I ACTION 1.

SUMMING&tPROPORTIONALn.-..J~i
Now if the mass is changed to say 20 kg then after oscillation it I
L , SIGNAL-V' UNIT
-, ;READING 1l
will not return to X but will return to a new position of equilib-
rium Y. Offset will be the difference between the X and Y -SENSOR 1. \
~-
positions.
This means, with proportional control, where a set value (the
UNIT
2.TERM CONTRQl~. ~
SIGNAL- -

PROPORTIONAL I
"-
~
~---
ACTION UNIT
command signal to a process control system) at a given load DEMAND
occurs it will only give coincidence with the desired value at that ......... ......... ~~
..........
load. At any other load offset exists. If this offset is acceptable to
Fig. 9.10 PROPORTIONAL PLUS INTEGRAL CONTROL LOOP
a plant the proportional controller is satisfactory. If the offset is
too great additional refinements have to be incorporated into the
controller (see Fig 9.10, Fig. 9.11 and integral action later). In For a human control loop. A man regulates the water inlet
many cases ofpractical description in this book and elsewhere, for valve to a tank to maintain gauge level with variable outflow
simplicity, desired value, set value or set point, etc., are all used for demand. He is told the level required (desired value), will see the
the same thing and no distinction is made on a fine point of level (measured value), after a change will compare the two and
principle. decide on adjustment (correcting action), finally there is amplifi-
Fig. 9.9 illustrates another analogy using level control, i.e. cation for muscle action to operate valve (correcting element).
equilibrium before and after a demand change with two different Proportional control will arrest the change and hold it steady but
heads and valve settings. at a point different from the original set value due to the load
change. The human operator would bring the level back to the
desired value after arresting the level, i.e. he would apply re-set
(integral) action to remove offset. Overshoot would not occur
because the operator would not, while adjusting for offset, go on
--a-DESIRED VALUE altering the valve right up until offset was gone. He would ease
-- r -, ~ f-LOAT -OLD SET VALUE
.
J \
---- -NEW SET VALUE down valve adjustment rate as desired value was approached, i.e.

"'-
,-....... '"""-
............
........... --- apply a damping action, based on rate (derivative).
""-
INTEGRAL CONTROL ACTION (I) (Basic action. 3.)
..-- ........,
"'-- The action of a controller whose output signal changes at a rate
........... ............ which is proportional to the deviation.
-... ------ ,....., ----..
DEMAND
Note:
Fig. 9.9 SIMPLE PROPORTIONAL ACTION CONTROL LOOP The object of integral control action is to reduce offset to zero.
 144 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PROCESS CONTROL THEORY 145

dV oc - 8
dt

by the definition given above 4

DEVIAiION.
I ~~
DESIRED VALUE
- - - - - - - - - - -
dV
- = -K28
dt
V= - K2f8dt
!
I I
I : II AT TIME ", V, = v,

K 2 is called the integral action factor. I 1:

<I>=ClV I I I

<I> = - ClK2f8 dt
Ii I I I
A
I
I
I
I I I
I
<I> = - pf8dt CONTROLLER - - - - t - - - ~~~/N7EGRALACTIONJ- - - .-; -
OUTPUT ~ i ~ ADJUSTMENT VI = -K 2 S 8dt
P is called the integral control factor.
SIGNAL I -- - -j[r~I~'.t _I_BETWEEN" AND-,- - - ~ -
Thus the potential correction <I> at a given time t is propor-
i:- f IMMEDIATE f
tional to the area between the desired and recorded values f t 8
PROPORTIONAL ACTION V. = -K 8 1

ADJUSTMENT AT t l
i ,--S---I------------
dt is a mathematical way of writing that area). Rate of changeOof
potential correction with respect to time is proportional to the ~ r~~~~GRAL Acrl~N
deviation{d<l>/dt expresses mathematically rate of change of t. t t2

<I> with respect to t}. TIME-

Now Fig. 9.11 PROPORTIONAL AND INTEGRAL CONTROLLER ACTION

p. = ClKl
and P = ClK2

. Kl =!!:..=
.• K2 P
s

tion from integral action to the potential correction. S is

increased by increasing resistance.
Integral action time S
Fig. 9.11 is given assuming an instantaneous deviation change.
By definition: This change is referred to as step function input. Deviation under
In a controller having proportional plus integral action, the proportional action alone would constantly increase, giving big-
time interval in which the part of the output signal due to integral ger offset, but the addition of integral action maintains a constant
action increases by an amount equal to the part of the output deviation.
signal due to proportional action, when the deviation is unchang- A more likely short term action would give integral action until
ing. deviation ceased, i.e. no offset. Integral action reduces a previ-
After time S then Vp = VI and the controller output change ously fixed controller gain JL and introduces extra lag, so is
equals 2Vp- Note that S = JLI p, i.e. integral action reduces undesirable, it should not be used unless the use of a wide
controller gain. The larger the S time setting the less the contribu- proportional band gives too big offsets. See Fig. 9.12.
 PROCESS CONTROL THEORY 147
146 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

A. The only time the valve is not moving is at the desired value,
CD CD offset will never be possible.

/
LOAD CHANGE
,
OSCILLATORY
Note:
Proportional plus integral (P + I) two term can be applied by
HUNTING
CHART including the link shown dotted in Fig. 9.13. For a rising float,
above set point, both act in the same direction downwards to
LEVEL RECORDER
close the valve, pivot P can be moved to vary the individual
TOO LITTLE INTEGRAL TOO MUCH INTEGRAL CORRECTION INTEGRAL
ACTION EFFECT, I.E. ACTiON EFFECT I.E. ACTION EFFECT, I.E.
actions. For a falling float above set point the actions are in
RESET TAKES A LONG OVERSHOOT AND LONG VERY SLIGHT OVERSHOOT opposition. For a falling float below set point both act to open the
TIME TIME TO DAMP OUT AND QUICK DAMPING
valve. Integral action is always tending to reduce offset. Integral
Fig. 9.12 INTEGRAL ACTION TIME
action is not used alone, if it was the characteristic would be
similar to two step action. Note the use here of a hydraulic
Referring to Fig. 9.13:
If the level rises, the small piston moves up and high pressure
controller (small and slave piston) for the integral action itself.
Proportionally controlled first order open loop systems, with an
fluid flows through the top port and returns through the port B,
this action via the link closes the valve to reduce the inflow. This inherent integration characteristic (for example shaped lavatory
movemp.nt will always continue as long as a deviation exists and cisterns), respond exponentially to a stop function when the loop
the rate of travel depends on the area of the top port opening is closed.
which is proportional to the deviation. Conversely fall in level
causes oil to flow in at the bottom port and return through port DERIVATIVE CONTROL ACTION (D) (Basic action. 4.)
The action of a controller whose output signal is proportional
p to the rate at which the deviation is changing.
r-----*1 • ,
ADJUSTING Note:
SCREW The object of derivative control action is to give quicker
response and supplement inadequate proportional control
damping.
Voc - dO by definition above.
dt
dO
V= -K a -
dt
K 3 is called the derivative action factor.
<I> = CIV
dO
<I> = - CIKg-
dt

+ dO
<I> = - TJ dt
Fig. 9.13 INTEGRAL ACTiON ON LIQUID LEVEL CONTROL
 PROCESS CONTROL THEORY 149
148 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

In Fig. 9.14 the derivative action is assumed instantaneous.

'TI is called the derivative control factor. After time T then V D = Vp and total controller output signal
Now change equals 2Vp. Use of derivative, action increases IL and
I-' = ClKl gives phase lead, which are desirable, but T is limited as too much
derivative action may cause instability and hunting. Note that T
'YJ = ClK3 = K 3/K 1, the larger the T setting the greater the derivative action
. Kl _I-' _ 1 contribution to the potential correction (increased by increasing
.. K3 -~-1'
resistance).

Referring to Fig. 9.15:

Now from mechanics, rate of change of displacement with
Derivative action time T respect to time is velocity. At reasonable velocities damping
resistance is proportional to velocity, so one would expect some
By definition: form of damping device in this rate response action.
In a controller having proportional plus derivative action, the Derivative control is not used alone, it is a transient condition
time interval in which the part of the output signal due to propor- which must be combined with proportional control.
tional action increases by an amount equal to the part of the
signal due to derivative action, when the deviation is changing at
P
a constant rate. Such deviation change is referred to as ramp
function input.

-
ADJUSTING
SCREW

i
DEVIATION - - - - - - - - r ....QESIR~~VALU~
~ ~ !de SLOPE- ~~
I dt

I
I
PISTON
I AT TIME t 2 THEN Vo = Vp CYLINDER
I SPRING
I
I
t
CONTROLLER
I
-------~I-------PROPORTIONAL--
I I ACTION ADJUSTMEN~ Vp =-K,8
-+ 0 ,,~-< / / / / V / / / / / <( /1

OUTPUT
SIGNAL ______ _ BETWEEN t, AND t2
lMMEDIATE } - - - - - -

I ' j
DERIVATIVE ACTION I V --K de
ADJUSTMENT AT 0- 3 dt t,: +
I DERIVATIJE
TIME
ACTION I Fig.9.15 DERIVATIVE ACTION ON LIQUID LEVEL CONTROL

t, t2
TlME--+
Fig.9.14 PROPORTIONAL AND DERIVATIVE CONTROLLER ACTION
150 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
PROCESS CONTROL THEORY 151
If the level rises at a certain rate the piston (in the dashpot)
moves down at a certain velocity, proportional to this velocity is a LEVEL
down force on the cylinder which acts to close the valve in, the RISE
cylinder motion is resisted by a spring. Valve displacement is
proportional to down force. Whenever the float stops changing
f
SET VALUE~L1NEAR~SINUS\OIDAL9TANK LEVEL
position the down force ceases and the springs return the cylinder
~ I I (VARIABLE) K:
to its original position. LEVEL
FALL

Note: VALVE
OPEN
Proportional plus derivative (P + D) two term can be applied
by including the link shown dotted, proportional plus derivative t
VALVE
PROPORTIONAL
ACTIO~GNAL
plus integral (P + D + I) three term can be applied by including CLOSE
the link shown dotted and the link shown chain dotted. For a
rising float, for (P + D + I) with the arrangement shown, above
VALVE
set point, all act in the same direction downward to close the OPEN
valve, pivot P can be moved to affect the value of all control
factors.
t
VALVE
CLOSE
INTEGRAL
ACTiON SIGNAL

VALVE
THREE TERM CONTROLLER OPEN

\ t
VALVE
DERIVATIVE
ACTlUlIrSTGNAL
CLOSE

Si
B
4 LEVEL SENSORI

:il-
TRANSDUCER

4EJJt-+-
4 TIME ---+ I TlME---+
P+I+D .
Fig.9.17 DISTANCE-TIME GRAPHS OF CONTROL ACTIONS
I
I
I
I
r-..,I DISTANCE TIME GRAPHS OF CONTROL ACTIONS
I
•
I
I Refer to Fig. 9.17:
L_._
I I
IVALVE
- POSITIONER
Such analysis gives a clear pictorial representation. Slope of a
distance time graph is velocity; an inclined straight line is con-
DEMAND stant velocity as the slope is constant, a curve of increasing slope
represents acceleration, a curve of decreasing slope represents
Fig.9.16 THREE TERM CONTROLLER
deceleration. The top two diagrams should be self explanatory,
relative heights depend on proportionality factor. For integral
action note that whenever the variable is away from desired
THREE TERM CONTROLLER value the integral effect is always moving to correct. For the
P + I + D actions, combined, are illustrated for level control in value at any instant on the third sketch think of area developed at
Fig. 9.16. that instant on the first sketch, applied on the opposite side of the
axis, and to a suitable scale factor. For derivative action note that
 PROCESS CONTROL THEORY 153
REED'S INSTRUMENTATION AND CONTROL SYSTEMS
152
from tank B (the controlled variable) controls the input supply
it opposes the motion of the variable irrespective of the desired
quantity to tank B, i.e. the level in tank A, and is a single
value. The value of the signal on the fourth sketch is the change
capacitor control loop for tank B. This is achieved because the
of slope of the first sketch, again opposite side of axis, i.e. slope
master controller signal controls the set point of the slave con-
only changes at four points on sketch one and at such points the
troller. A two capacitor system has therefore been simplified to
derivative effect is acting almost instantaneously. •
two single capacitor systems which are more easily controlled.
Alternatively sensing for the slave could be flow rate at supply
CASCADE CONTROL rather than tank A level. The process can be extended to multi-
Consider a multiple capacity system for level control, as an
capacity systems with control of any desired variable. The princi-
example a two capacitor tank system illustrated in Fig. 9.18.
ple is utilised very often in practice, for example IC engine
Single capacitor systems respond quickly to load changes and are
coolant and Butterworth heating, as described in Chapter 13. If a
easy to control utilising correct proportional band and reset
certain pressure range of controller output is divided up by relays
action but interaction occurs with multiple systems. Tank A acts
or controllers for different functions, in for example a sequence,
as a lag effect on the controlled process from tank B so the
this may be termed split range control. This is also utilised in
combination is less sensitive, especially to supply variations, this
practice, for example exhaust range pressure control (Chapter
is an inherent problem with large inertia (mass, heat capacity,
13).
etc.) systems as for example IC engine coolant circuits.
EXAMPLE- LEVEL CONTROL
When considering numerical questions it is often best to utilise
0
a tabular approach as the following example illustrates:
A I I B I~...J The sketch (Fig. 9.19) shows a single element boiler water
...J
w
~~ level control system. Assuming that the system has been adjusted
~...J
0 VARIABLE so that the level is at the qesired value of 16 cm ("half glass")
~
~ pENSOR u

SUPPLY -
DEMAND when the load is 500 kg/min, determine:

B 8
(a) the offset if the load is reduced to zero,
CONTROL I " . t
~~E I I (b) the proportional band setting required such that the offset
•I LEVEL
CONTROLLER LEVEL
CONTROLLER
is limited to 8 cm if the load changes from 500 kg/ min to 100
kg/min.
L-.
I SLAVE MASTER
I
L __ J
SET POINT
....J 1..-_.- SET POINT
Steam flow load (kg/min) 500 0 100
(VARIABLE) (FIXED) Level(cm) 16 32 24
Fig.9.18 CASCADE CONTROL (LEVEL) Controller input (kN/m 2) 60 100 80
Range change (kN/m2 ) 40 20
Controller output (kN/m2 ) 70 20 30
Consider Fig. 9.18:
There are two variables, supply and output demand, affecting Range change (kN/m 2) 50 40
the controlled variable which is level in tank B. The slave control-
i.e. offset 16 cm, proportional band 50%. Unless stated to the
ler with level sensing from tank A controls the input supply
contrary, a linear proportionality is assumed between the indi-
control valve according to the set point and is a single capacitor
control loop for tank A. The master controller with level sensing cator/controller variable scale ranges in such cases. A similar
 PROCESS CONTROL THEORY 155
154 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
Similar questions are included at the end of the book, includ-
r • STEAM FLOW (LOAD) ing those requiring graph plots of controller signals-after
a -600 kg/min analysis by a method such as just described. When the controller
includes such as integral action the approach is similar but it must
be remembered that integral action time (S) elapses while the
signal changes (by integral action) by an equal amount to the
immediate proportional action signal. See Higher; TEC/
SCOTEC specimen examination questions number 16
(CGLI Advanced METC) and number 18 (OND Supple-
mentary).

TEST EXAMPLES 9
LEVEL TRANSMITTER
20 - 100 kN/m2 1. Explain the meaning of the following terms relating to pro-
-16 - +16 em cess control:
(a) desired value,
(b) error signal,
DESIRED
(c) detecting element,
REVERSE ACTING
VALUE
CONTROLLER (d) feedback,
PROP: BAND=80% (e) reset action,
(f) servo-motor.
2. Explain the meaning of the following terms, using suitable
diagrams where appropriate:
(a) potential correction,
REGULATING UNIT (b) proportional control,
2
20 - 100 kN/m (c) integral control,
o- 800 kg/min
(d) integral action time,
(e) derivative control,
(f) derivative action time.
EXAMPLE- LEVEL CONTROL
3. Draw simple diagrams showing the response of a detecting
question is included at the end of the book - specimen examina- element suffering from a distance velocity lag equivalent to 5s
tion question number 15, Higher; TEC/SCOTEC (& OND Supp- and a single transfer lag, when subjected to disturbances in the
lementary). An alternative method of solution to that previous is: form of:
System proportional control factor J.L equals multiple of propor- (a) a step,
tionality characteristics/coefficients, i.e. 31·25 = 2·5 x 1·25 x (b) a ramp.
10 when ell = - J.L () and as ell = - 500 kg/min so () = + 16 cm. Illustrate, on two simple diagrams, the effect of an increase in
Similarly working in the reverse direction from () = + 8 cm the frequency on "phase lag" and "attenuation" for a detecting
controller characteristic/coefficient, or proportional action element suffering from transfer lag.
factor K 1 is now 2 (gain), proportional band is 50%.
 CHAPTER 10

PNEUMATIC CONTROL PRINCIPLES

Note:
The bar is used as the unit of pressure in this section. 1 bar =
105 pascal (Pa) = 10 5 N/m 2 •

PNEUMATIC TWO STEP CONTROL TECHNIQUE

Referring to Fig. 10.1:
The constriction may be about 0·2 mm bore and the nozzle
about 0·75 mm outlet bore, these sizes are largely fixed by air
purity condition, i.e. particle filtration size.
PIVOT
~
I
ATMOSPHERE

FLAPPER ?/NOZZLE (N) CONSTRICTION IC)

IF) - ) / ,
<I .. AIR PRESSURE
SUPPLY

--6- l.
DEVIATION

TO MEASURE
UNIT
I ·I I

------+
PLANT
FLUID
Fig. 10.1 TWO STEP PNEUMATIC CONTROL

With the flapper, or baffle, moved away from the nozzle

full nozzle pressure drop occurs, pressure p may be about 1·2 bar
or less. With the flapper almost closing the nozzle, pressure p
may be near 2 bar, i.e. almost supply pressure. Two values of
pressure p can be arranged, which will depend on the flapper
158 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
PNEUMATIC CONTROL PRINCIPLES 159

position, which is in turn decided by the measure signal move-

ment. An on-off operation, or low-rate and high-rate operation,
can be utilised with these two pressures.
Average flapper travel between two limits is often less than JN VENT DIAPHRAGM\.

2 bar INPUT
0.05 mm, the nearer the flapper to the nozzle the stronger the PRESSURE
measure signal force required, this is a limitation. The relation DOUBLE
SEAT VALVE
between flapper travel and pressure p is non-linear, i.e. equal
increments of flapper travel do not give equal increments of lk- 1t bar OUTPUT PRESSURE
pressure p, but over a fairly wide range of travel, say 30 to 70%,
the relation is reasonably linear, i.e. linear between 0·015 and
0.035 mm in 0·05 mm travel. II n ,..-FLOW

THE RELAY Fig.10.3 ALTERNATIVE DESIGN OF RELAY VALVE

Provides pneumatic amplification, proportional movement,
and reduced time lag. Equivalent in action to an electronic
amplifier.
N C
PNEUMATIC PROPORTIONAL CONTROL TECHNIQUE
a+--AIR SUPPLY

I
.,...-1
This utilises adjustable (negative) feedback due to the bellows
and flapper linkage. Input (from the measure signal) is compared

~
~TROLLING UNIT (OPEN)
to output (from the relay signal) and the action is to reduce this
_ __ BLEED PORT (SHUT)
difference, so matching input to output. Any desired ratio be-
BELLOWS
tween input and output can be achieved by adjusting the linkage
'FIXED ratio a: b shown in Fig. 10.4. For a 50% proportional band then
P
the measurement change is 50% of scale for full valve stroke, i.e.
Fig.10.2 RELAY VALVE (CONTINUOUS BLEED)
under ideal conditions the control should operate to maintain
Ifp increases (see Fig. 10.2) then the bellows acts to close the measured value and desired value together at 50% valve stroke.
bleed port and supply air passes, conversely if p decreases a Varying load means the controller keeps conditions stable within
continuous bleed to atmosphere occurs. Amplification by a fac- the proportional band, but not at the desired value, maximum
tor of 16 can easily be arranged, for example a flapper travel of offset cannot exceed half band width.
0.01 mm causing a change ofp on the bellows of 0·05 bar could
give output from 1·2 to 2 bar. Flapper travel is approximately Referring to Fig. 10.4:
proportional, by a linear relation, to output signal in this throttle Consider the measure link moving right, this decreases the
position of 0 ·05 bar bellows pressure range. An alternative type nozzle escape and pressure p increases. Pressure p acts, via the
of relay is given in Fig 10.3 for comparison. relay, on the bellows so tending to act in the opposite direction to
Various alternatives and refinements can be added, for exam- the initial movement with proportional action against the spring
ple, a bellows connected to the relay output will give action (just as for the simple spring analogy, p. 142 earlier). This
proportional to output utilising negative feedback, etc. The relay decreases the sensitivity (flapper travel near nozzle). The ratio
could introduce further non-linearity if not properly matched in a:b (which is adjustable) decides the bandwidth, this action is the
design. simple lever principle (feedback can never exceed deviation).
 PNEUMATIC CONTROL PRINCIPLES 161
160 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

i.e. potential correction is proportional to the deviation and

equals the proportional control factor multiplied by the devia-
tion. Negative sign to indicate opposite direction.
SPRINGlr FLAPPER
· I I
BELLOWS (B,)
I
F=-s'x t ..- m f
e - --.;..--
STACK TYPE CONTROLLER PRINCIPLE (P ACTION)
~
. . ~=-J..Ie
......FIXED x m Referring to Fig. 10.5:
...... "I
t .P DECIDES
L_J V and 4> VALUE The construction is of air chambers stacked on top of each
MECHANICAL ANALOGY other separated by diaphragms and incorporating relay valves,
a
~CI \ !! &+-AIR SUPPLY
nozzles and restrictor control valves.
• I I r
N p
b
F

TO MEASURE
UNIT ---+
+-e+ PLANT
FLUID
y
Fig.10.4 PNEUMATIC PROPORTIONAL CONTROL WITH
NEGATIVE FEEDBACK

This cancellation whereby a pressure increase moves the flap- x

per to lower the pressure means that a greater movement of the
measure unit, for a given change in control line pressure, is
required so ensuring proportional action. This also gives a wider
proportional band without increasing mechanical linkages which
would reduce accuracy. The relay shown in all the sketches is not
in practice necessarily fitted to all controllers. Without the bel-
lows the proportional flapper travel region is very small.
Bellows movement is proportional to pressure p.

x = mO - np
Fig.10.5 STACK TYPE CONTROLLER PRINCIPLE (P ACTION)
where x is movement of flapper next to the nozzle, m and n are
proportionality constants (including the adjustments a andb) for
the deviation () and the negative feedback pressure P (which also
decides V and <1> ). The measured value (controlled condition) enters at chamber
C and the set value (command signal) set up at the reducing valve
x is negligible compared to other movements.
enters at chamber B. Variation between these two values causes
:. np = mO
m the diaphragm arrangement to move up or down vertically so
:.p=--O that air flow through the nozzle to the chamber D controls
n
cancellation of the deviation caused by the pressure variation.
:. <1> = - fLO
 PNEUMATIC CONTROL PRINCIPLES 163
162 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

CAPACITY TANK (C)

Pressures at A and B would be equal if control valve X were ADJUSTABLE

INTEGRAL
closed, hence pressures at C and D would equalise as D pressure I RESTRICTOR

(controller output signal) changed, this means 100% propor-

tional band.
Conversely if X were opened fully pressures at A and D would
be equal so that deviation from set value would cause the nozzle
to be fully opened or closed. This is two step control action.
Valve Y can act as an adjustment but essentially it prevents direct N
.-
AIR
connection between output and set value air lines. SUPPLY
F
Variation of the setting of X between open and closed gives
proportional band variation between 0% and 100%. po
By utilising different stacking arrangements, e.g. capacity bel- TO MEASURE UNIT t CORRECTING
+-6---+ UNIT
lows, different connections to control valves (restrictors), vari-
Fig.10.6 PNEUMATIC PROPORTIONAL PLUS INTEGRAL CONTROL
able bellows areas, etc., then the correct proportional, derivative WITH NEGATIVE FEEDBACK
or integral actions can be incorporated as required. P, P + D,
P + I actions separately generated and combined in an addition
unit give three term action without interaction problems. Deviation, under proportional action alone, would become grea-
ter, more offset would occur. However bellows B 2 exerts positive
feedback to raise pressure P2 at a rate dependent on the devia-
PNEUMATIC PROPORTIONAL PLUS INTEGRAL CONTROL TECHNIQUE tion, this maintains constant deviation. Integral action would
Integral (reset) action can be regarded as a slow cancellation of continue until deviation ceased and there would be no offset, i.e.
the sensitivity reduction provided by the negative feedback of the restoration to desired value. A repeat is accomplished when the
proportional system. amount of change in air pressure in the reset (integral or floating)
Referring to Fig. 10.6: bellows equals the amount of original change in output pressure
Without the needle valve adjustable restrictor I, the propor- to the proportional bellows.
tional negative feedback bellows B 1 effect would be completely
x = mf) - n(p - P2)
cancelled by the proportional positive feedback bellows B 2 effect
i.e. negative feedback due to p and positive feedback due to P2.
(assuming equal bellows sizes and form), simulated two-step or
Again taking x = 0:
near proportional action for limited flapper travel would result
(depending on flapper travel utilised). Similarly in the steady p - P2 = '!!. f) for whole action
n
state with the needle valve as there would be zero pressure
difference across it.
When a disturbance causes a deviation to occur (say p
now d:t
2 oc p - P2 for integral action only
increases) then the rate ofp2 change is proportional to the devia-
i.e. rate of change of pressure P2 is proportional to the difference
tion effect p - P2' of pressure.
If the measure unit moves right under a constant deviation
increase then p increases giving near proportional action Vp dp2 1
immediately. Negative feedback to bellows B 1 reduces sensitivity
di = CR (p -:- p2)
giving wider proportional band and true proportional action.
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
PNEUMATIC CONTROL PRINCIPLES 165
164

PNEUMATIC PROPORTIONAL PLUS DERIVATIVE CONTROL TECHNIQUE

CR is a time constant dependent on the capacity C of the tank
Derivative (rate) action increases sensitivity by restricting the
and the resistance R of the restrictor I, CR is the integral action
negative feedback provided by the proportional system so that
time S. during the change high speed sensitivity occurs but when the
SdP2 = m (J
dt n change ceases ordinary proportional action occurs.

integrating Sp2 = ~f (Jdt

CAPACITY
P2 = mf(Jdt TANK IC)
nS
Thus for whole action
m + -mf (Jdt
p = -(J ADJUSTABLE
n nS ,.., DERIVATIVE
RE5TRICTOR I D)

applying negative direction sign, with p equivalent to <I> AIR

SUPPLY

<I> = - fL ((J + ~f (J dt) for (P + I) F

~ CORRECTING
TO
TO MEASURE UNIT UNIT
Note: +--- e ---+
Consider the distinct analogies: Fig.10.7 PNEUMATIC PROPORTIONAL PLUS DERIVATIVE
CONTROL WITH NEGATIVE FEEDBACK
dP2 I
Pressure: di = CR (p - p2)

where rate of pressure pz with respect to time is proportional to

excess pressure p - pz- CR time constant, C tank capacity and R Referring to Fig. 10.7:
restrictor flow resistance- Without the needle valve adjustable restrictor D, proportional
• dV2 I action results, similarly for the steady state with the needle valve.
ElectrIcal: - = - (V - V2)
dt CR When a disturbance causes a deviation to occur (say p increases)
where rate of change of voltage V z with respect to time across a then the rate of p I change is proportional to the deviation effect
condenser is proportional to excess voltage V - V z. CR time p - Pl. Derivative action stabilises more quickly after a change.
constant, C capacitance of the condenser and R current resistor Considering the inner bellows, then the smaller area gives a less
resistance, the latter often negligibly small. force per unit pressure change of p, a narrower proportional
band, higher sensitivity and less feedback occurs than if the same
d(JE I pressure effect acted on the larger outer bellows.
Temperature: - = - ((JF - (JE)
dt CR If the effective area of the inner belows was one quarter of that
where rate of change of temperature BE across a detector element of the outer bellows then an instantaneous deviation produces a
with respect to time is proportional to excess temperature finite response equivalent to four times the normal proportional
BE - (JF- CR time constant, C is thermal capacity of element and action with negative feedback for the same actuating signal
R thermal resistance to heat flow. applied to the outer bellows.
166 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PNEUMATIC CONTROL PRINCIPLES 167

Considering the outer bellows which gives the derivative Rearrange the last expression:
action then the change of pressure via the relay across the resis-
tance and tank (D and C) gives a pressure drop proportional to p = PI (T ~ + 1)
the rate of change of activating signal deviation. This means the
sensitivity reduction due to negative feedback is adjusted in line P= m ()
n
(T!!:-
dt
+ 1)
with the rate of change of the deviation.
The combination bellows means that when movement starts
(say to the right) to increase, the narrow proportional band, m T-+()
p=- (d() )
n dt
caused by P only on the small bellows, gives high output relay
signal pressure. Such exaggerated output is then amended for
p =ml - T- +() (d() )
derivative action by P I on the large bellows until the measure nl dt
movement ceases and pressure in the outer bellows equals pres- where ml and nl are new proportionality constants to allow for
sure in the inner bellows (no pressure drop). This means the the proportionality feedback effect of the combined bellows.
control valve operates sooner for the same rate of change from Now applying the negative sign to indicate the opposite direc-
the measure unit. tion and with P equivalent to potential correction <1>.
Double bellows are not always utilised, strictly the derivative
action is on the outer bellows only. This arrangement has disad-
vantages as phase lag.to the derivative action occurs (see com- <1> = - fL (() + T~~) for (P + D)
pound controllers later).
The inner bellows gives proportional action only, for simplicity
PNEUMATIC COMPOUND CONTROLLER (P + I + D)
regard this bellows as omitted.
Three term (P + I + D) or two term (P + I or P + D). A
x = me - npl controller action in which the output signal from the controller is
the result of more than one operation on the deviation.
negative derivative action feedback due to Pl.
K2f Kad())
Taking x = 0 as previously: V = - Kl (e+ -Kl () dt + K- -dt 1
for three term
m ll
PI = -
n
u
•
for whole actIon.
<I> = - fL (() + ~f dt () + T~~) for three term

Now The sketch of Fig. 10.8 shows the compound pneumatic con-
troller, the action should be clear from previous diagrams.
dPI 1 ..,
dt = CR (p - pI) for denvatIve actIOn only.
Interaction
With the pneumatic arrangement shown in Fig. 10.8, but D at
X, adjustment of either lor D affects each of the three actions.
CR is the time constant, dependent on capacity C of the tank Thus the effective action times (5 and T) differ from the nominal
and resistance R of the restrictor D, CR is derivative action time (dial set) action times. D can be moved from X to Y which may
T. improve performance but still gives interaction. For truly inde-
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS PNEUMATIC CONTROL PRINCIPLES 169
168

pendent adjustments then derivative and integral actions should when delivering output of one cubic metre (referred to standard
be generated separately based on proportional action and com- conditions) for up to 10 instruments.
bined in a relay (see also Chapter 15). Air lines should have a gradient of at least 1:50, with moisture
traps at lowest points, and instrument tappings are taken from the
CAPACITY TANK (C)~ top of headers. Final stage filters are often of the ceramic type
and silicone impregnation makes them water repellant.
I
Annealed seamless copper tubing, pickled inside and out, is
often used-especially for single instrument loads. Polythene
82
and PVC tubing is resistant to corrosive atmospheres and is also
much cheaper for larger installations. For the supply to say 10
instruments the hp line (4 to 7 bar) would be about 12 mm bore,

I
e!
N -
)
- - -
I i
!!..
,.
\- AIR
I ~SUPPLY
delivery through 18 mm reducing valve and filter to a 25 mm
bore lp header (at 1·5 bar).

F TEST EXAMPLES 10
.
TO MEASURE UNIT
"'-8-
t TO
CORRECTING
UNIT
1. Describe the operating principle of a pneumatic controller.
Explain what is meant by the term "proportional action". Show
Fig. 10.8 COMPOUND (P + I + D) PNEUMATIC CONTROLLER by means of a simple sketch how the controller functions to
maintain a particular system in equilibrium.
2. (a) Make a diagrammatic sketch of a two term (P + I)
AIR SUPPLIES pneumatic controller of the nozzle-flapper type and briefly
An adequate supply of clean, dry compressed air is required describe its operation.
with well designed, installed and maintained air line systems. (b) With reference to the diagram describe briefly how the
Quantity is defined under standard intake conditions, i.e. 15°C controller could be set before commissioning.
and 1 bar, which relate size, capacity and consumption. (c) With reference to the diagram describe briefly how the
Quality requires that filtration removes solid particles, oil and proportional band adjustment could be calibrated.
water. If dew point can be reduced at high temperature, below
any likely ambient temperature of the system, the installation can 3. (a) Make a diagrammatic sketch of a three term nozzle
be kept dry. High compression, with interstage and after cooling is flapper controller which has provision for receiving pneumatic
effective especially when large delivery receivers allow cooling desired value and measured variable signals.
under pressure. Absorber filters such as silica gel or activated (b) Explain how the proportional action is generated and
alumina should be fitted at low level system points to act as how the gain may be varied.
moisture traps-such traps should also be fitted adjacent to (c) Describe how the integral action is generated and how
reducing valves. the degree of integral action may be varied.
Compressors are either arranged to run continuously, fitted (d) Describe how the derivative action is generated and
with unloading devices to allow running light when pressure how the degree of derivative action may be varied.
supply is reached, or have pre-set cut in and cut out pressure (e) Explain why a relay valve is necessary and what addi-
switches. Machine capacity should be such that 50% time loading tional benefits may result from its use.
only is required; a 3 kW unit (4 bar) would meet this criteria
 CHAPTER 11

ELECTRONIC CONTROL PRINCIPLES

Note:
V has been mainly used previously as the symbol for controller
output signal. In this chapter V is used as a general symbol for
voltage, suffix i or 1, etc for input (alternative E or e, for V,
sometimes used elsewhere).

THE OPERATIONAL AMPLIFIER

Amplification, generally voltage, is readily achieved for ac or
dc. It is easier to amplify ac than dc using simple circuits and dc
amplifiers suffer from drift due to voltage variations. Measure
signals are however dc so that many process controllers utilise dc
amplifiers. Amplifier gain, i.e. ratio of output to input, can reach
108 , without negative feedback from the last stage, and 104 is
common. Inclusion of such feedback allows reduced, and fixed,
ratios between output and input utilising resistive and capacitive
components to generate control actions. Essentially there are
input and feedback networks and the amplifier adjusts output
voltage so that the amplifier takes negligible current, i.e. equal
currents in the two networks. There must be a voltage polarity
change to balance currents, i.e. 180 degrees out of phase, which
requires an odd number of amplifier stages. Gain of circuit
depends only on resistance ratio which is accurate and indepen-
dent of any variation in amplifier gain. The amplifier inverts and
multiplies; it is an active component as distinct from a passive
electrical network. The high gain amplifier can use negative
feedback to reduce gain to what is actually required and the
 ELECTRONIC CONTROL PRINCIPLES 173
172 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

ELECTRONIC PROPORTIONAL CONTROL TECHNIQUE

negative feed back gives stability. Energy input is required. (See
also Chapter 7, operational amplifier, feedback analysis). Apply Ohm's law to the circuit of Fig. 11.1.
VI- Vx VI .
Figure 11.1 illustrates the operational amplifier circuit (power h = = - as Vx IS zero
RI RI
supplies to amplifier omitted, earths shown). The signal input
voltage is usually applied at one terminal (V 1) and the other input Vo - Vx Vo .
IF = = - as VXIS zero
terminal is usually earthed, similarly for output and amplifier. RF RF
h= -IF
because amplifier current is zero.
R, This is Kirchhoffs law for currents at a point.
VI IF
I, Vo
RF
- Vo=- VI
RI
Action is scalar multiplication (V o proportional to Vl) with the
~ --..e- multiplying factor R F / R 1 and the negative sign indicating inver-
sion. Ajdustable circuit gain (proportional band) is achieved by
Fig. 11.1 OPERATIONAL AMPLIFIER altering R F with R , fixed (RF/R l = 1 is 100% bandwidth). For
Alternatively two inputs can be fed into A and V o will then multiplying factors below unity an adjustable potentiometer
represent the difference. 1 1 is negligibly small and zero potential (attenuator) can ~e used. V o = tV l where t =r/R the tapping
effectively applies at X, i.e. infinite input impedance as an ideal. ratio. Removal of the minus sign can be achieved by using two
Due to inversion, feedback is essentially negative. Input resistance amplifiers in series. G = R F / R 1 is commonly 1 or 10, gain.
may well be of the order of one megohm, output resistance some Consider now two inputs as shown in Fig. 11.2.
tens of ohms, amplifier gain say 10 5 , overall gain say 100. The
upper lead to A (often marked - inside triangle) is commonly
RF
called the inverting input, lower lead ( +) is then the non invert- NEGATIVE
ing input. Magnitude of power supplies for the circuit depends on FEEDBACK
the power requirements of the amplifier itself. Response to R,
positive and negative inputs is required and amplifier power V~
, MEASURED
needs require say + 6V, -6V and OV connections (these are VALUE
e
omitted from sketches for simplicity). Voltage used depends on ERROR
RZ SIGNAL
conditions, inputs will be less. Vz~"-­
DESIRED
VALUE
ELECTRONIC TWO STEP CONTROL TECHNIQUE
Fig.11.2 ELECTRONIC PROPORTIONAL CONTROL
On-off devices find a wide range of application using a simple
switch technique. A typical example is room temperature control
in which a bi-metallic strip closes or opens electrical connections
leading to energy input. Small permanent magnets ensure snap
action and differential gap is adjustable. Digital devices are often
used.
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
ELECTRONIC CONTROL PRINCIPLES 175

YO=-RF
T.T (VI -+-
RI
V2)
R2

where VI and V z are negative inputs. This is essentially asummer

(and scalar) action which can be extended to further inputs as CF
required. IfR I = R z = R 3 = etc. = R F then V o = -(VI + V z + V 3
+ etc.), i.e. summation only. Ratio control of inputs can be ~\"'.£'/vj
;' I

achieved by adjustment of the respective input resistances. A IF

controller must produce an output signal proportional to the R1
V1 11 Va
deviation between desired and measured values. The two signals
can be compared (comparing element), perhaps elsewhere, by }I I
I ' ,~~ . I
. . - - \')'',1\1'- - - -
opposition flow through a common resistor the voltage across ;'

which now represents deviation (error) signal, for transmission Fig.11.3 ELECTRONIC INTEGRAL (RESET) CONTROL
to the amplifier input. If one input voltage is applied as negative
to a summer the result is effectively subtraction. For Fig. 11.2 if
VI and V z are regarded as measured value and desired value (in
opposition) this gives error input and proportional to deviation
i.e. output voltage is the integral of input voltage with the time
control action, bandwidth adjustment at R F • Amplifier power
constant (reset rate) dependent on C F and R 1. If, as for the two
supplies and earth lines are omitted for simplicity.
previous sketches, there is a modification to two inputs, via
resistors representing measured and desired values, then amp-
ELECTRONIC INTEGRAL (RESET) CONTROL TECHNIQUE lifier input voltage corresponds to error voltage, i.e. output vol-
Consider the circuit of Fig. 11.3: tage is the integral of error input voltage. A feedback resistor R F
By placing a capacitor CF in the feedback circuit a limit is is necessary to give proportional addition and make adjustment
placed on the amplifier response rate to change of input signal. more easy, with a fixed capacitor, alternatively potentiometer
adjustment could be provided. Integral action is essentially rate
For a capacitor C= Q/V control in the feedback network of the circuit by capacitance. R F
and R z additions shown dotted. Integral action is very rarely
:. QF = CFVO applied on its own.
IF = dQF = CF dVo
dt dt ELECTRONIC DERIVATIVE (RATE) CONTROL TECHNIQUE
Consider the circuit of Fig. 11.4 in which it is necessary to
f dVo = ~Ff IFdt consider P + D combination:
A capacitor Co is in the input circuit, together with a resistor
VI R h to produce a rate of change component.
but I I = -IF andh = - In the steady state there is no current through CD.
RI
h= -IF
Vo = - -I-fVI dt
CFRI RF
Vo= - - VI
RI
 176 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC CONTROL PRINCIPLES 177

In the transient (changing) state: ELECTRONIC COMPOUND CONTROLLER (P + I + D)

A controller action in which the output signal from the control-
11 = VI ler is the result of more than one operation on the deviation
RI
(error signal), in this case three, i.e. three term controller.
Ie = dV-I
CD
dt
Vo = - R F ( -VI
RI
+-1-
CFRI
J VI dt + CD-
dVI)
dt
II + Ie = -IF
leading to:
II +
Vo
Ie = - RF
Vo = - -RF ( VI
RI
+ -CF1 JVI dt + RICn-
dVI)
dt
Vo = - R F (II + Ie)
Vo = -- RF -VI
( RI
+ Cn dVI)
_ The relation with the proportional, integral and derivative fac-
dt tors given previously and equations relating to Fig. 10.8 are
The output voltage therefore has the two desired components, obvious.
i.e. proportional to input and proportional to rate of change of
input. The feedback resistor is necessary to give proportional
CONTROL FUNCTION GENERATOR
addition and adjustment. If, as before, measured value and r- ------------.
I I
desired value inputs through resistors are applied then input
voltage is deviation error signal and output voltage is signal to
final control element. The phase advance network ahead of the DESIRED
4f'~
II
IL-
T _
-GAIN
_
-RESET
_
- RATE
II
..JI
amplifier gives attenuation across the CR circuit which requires VALUE

compensation with increased gain at the amplifier. The R z

desired value resistor is shown dotted. Derivative action is never
applied on its own. OUTPUT

RF

IF

V1 R1

I MEASURED
I VALUE
I
tI I OUTPUT
I DESIRED
.. __...A,/,,,:-,..<.,...,,- _ ..I VALUE
7 RZ
/
Fig.11.5 COMPOUND (P+ I + D) ELECTRONIC CONTROLLER
Fig.11.4 ELECTRONIC PROPORTIONAL PLUS DERIVATIVE
CONTROL
 ELECTRONIC CONTROL PRINCIPLES
179

178 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

FROM INPUT
FROM OUTPUT BRIDGE (MEASURE)
BRIDGE (FEEDBACK)
Referring to Fig. 11.5:
The sketch shows the compound electronic controller, the •• FEEDBACK INPU!
VOLTS T VOLTS
action should be clear from previous diagrams and should be
compared with Fig. 10.8. Similar remarks about interaction
apply as for the pneumatic case. The upper sketch illustrates
grouping to controller (note the summer, and the potentiometer (f)
...... ,(f)

*C o
(f) :::::l ......

gain adjustment) whilst the lower sketch is basic operational ~~

1l.....J

U ~I~
amplifier configuration. w a
e
Figure 11.6 is a typical response curve for (P + I + D) action.
Figure 11.7 is a simplified form of electrical circuit which can be
compared with the previous sketch and to the complete elec-
tronic controller diagram given later in Chapter 12.
Fig.11.7 ELECTRONIC CONTROLLER CIRCUIT
(COMPOUND P + I + D)

SINUSOIDAL
SIGNAL Note:
/
With electric-electronic controllers:
INPUT
1. For output voltage to be proportional to input voltage it is
necessary to add a resistor (R 1) in the feedforward path to the
amplifier (i.e. in series) and a resistor (R F , adjustable) in the
feedback path (i.e. in parallel).
2. If output volts are to be proportioned to deviation, repre-
sented by error input volts (e), then error signal is applied as
COMBINED
input to the circuit of 1.
3. For the addition of integral action to 2 above it is necessary
to include a capacitor (C F ) in the feedback circuit in series with
TIME ---+
resistor R F •
Fig.11.6 RESPONSE TO SINUSOIDAL SIGNAL 4. For the addition of derivative action to 2 above it is neces-
(COMPOUND P + I + D) sary to include a capacitor (CD) in the feedforward circuit to the
amplifier in parallel with the resistor R 1·
5. If P + I + D action is required to combine 2, 3, 4, above,
then
Vo =- -RF e
( Rl
RF
+ R1CF
-- f
edt de)
+ RF CD-
dt
.
180 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ELECTRONIC CONTROL PRINCIPLES 181

BLACK BOX ANALYSIS this is particularly important for integrating circuits, terminals A
It is useful to summarise some of the preceding work in this and H are used to give offset null. They connect a tapped (resis-
chapter utilising this analysis approach, concerned with external tor) potentiometer across the ends of the amplifier to a supply
relationships and not internal circuitry (which includes opera- lead (usually negative) so giving compensation for inherent off-
tional amplifier with negative feedback loop). Amplifier power set at amplifier input. ~
suppliesiand eaTthing, are important and actual voltage used . S.S. BEKULAN
depends. on the amplifier and conditions of working.
OFF. No. 0 0 8 0
Referrin~ to Fig. 11.8 for illustrative connections: Consider now various configurations applied to Fig. 11.8:

~ _E 1. Inverter V o = -VI
A~ Input voltage signal V I through a resistor R I to terminal B,
I r- .... terminal C earthed through a resistor R E • R I = R F. Output
I I -- I---e F
B .--\- .. -; voltage signal V o at terminal G (also see Fig. 11.1, upper lead to
I I A is inverting input, lower lead to A is non inverting
I I
I -4-1---e G input-shown earthed) .
C .--\---;
I .... ....
1,.. ....
2. Inverter Summer V o = -(VI + V z)
0- ~H As for inverter but input signals VI and Vz in parallel to B.
R I = R z = R F (also see Fig. 11.2).
Fig.11.8 BLACK BOX ANALYSIS (OPERATIONAL AMPLIFIER)
3. Scalar Inverter Summer V o = -(aV I + bVz)
As for inverter summer but a = RFIR}, b = RFIR z, R I 4= R z,
R F usually adjustable (see previous sketches).
A is null offset B amplifier inverting input ( - ) 4. Inverter Multiplier V o = -at VI
C amplifier non-inverting As for inverter but input potentiometer (resistance R) con-
input (+) D supply ( -6V) nected across VI and earth, with a tapping lead to R I then to B.
E spare terminal F supply ( + 6V) a = RFIR},R I =1= R z =1= R F, tapping ratio t = rlR, which is less than
G output (feedback to B) H null offset unity.
Only two input signals (maximum), are considered to either B
5. Inverter Divisor V o = - Vt!t
or C for simplicity of illustration. Terminology for Fig. 11.8 As for inverter but output potentiometer (resistance R) con-
should be clear except for null offset, which needs elaboration. In nected across V o and earth, with a tapping lead to make the
derivation of equations the input current to the operational feedback loop through R F to B, tapping ratio t = rlR, which is
amplifier is assumed negligibly small, and where resistances are
less than unity.
specified equal that these are exactly so. This is practically not
possible. To balance input resistances, and reduce such offset, 6. Non Inverting Summer V o = VI + Vz
terminal C (when not in use) is earthed through a resistance (R E) VI input throughR I in parallel with Vz input throughR z into C,
of magnitude equal to the equivalent resistance of feedback and which is led to earth via resistor R 3 , B earthed through resistor
input circuits in parallel. To balance any remaining offset, and R 4 • R I = R z, R 4 = R F, R 3 = R I /2 (as two inputs).
 ELECTRONIC CONTROL PRINCIPLES 183
REED'S INSTRUMENTATION AND CONTROL SYSTEMS
182
TEST EXAMPLES 11
7. Subtractor Vo = VI -Vz
VI throughR I to C, earthed via resistor R 3 • V z throughR z to B. 1. Describe a three term electronic controller. Show on a diag-
All resistances equal. ram the variation in controller output due to:
8. Non Inverting Multiplier Vo = cV 1 (a) proportional action,
V 1 through R I to C. B earthed through R 4 • R I = R 4 • (b) integral action,
C = (R F + R 4 )/R 4 •
(c) derivative action,
9. Integrator (d) combination P+ I + D.
The circuit has already been sketched in Fig. 11.3. The amp- 2. Describe, in detail, an electronic operational amplifier. Dis-
lifier C terminal could be earthed through R E and terminals A cuss the various methods of adjusting gain and comment on the
and H connected. uses of this device in control or measuring systems.
The analysis could be further extended to two term and three 3. Show, by means of simple sketches, how an electronic con-
term controllers troller arranged for proportional control action can be modified
to include:
TRANSISTOR STABILISATION CONTROL (a) integral action,
The constant current (shunt) stabiliser utilising a zener diode (b) derivative action.
has been described previously (Fig. 7.6). Voltage (or series)
stabilisation based on an integrated circuit, i.e. transistor, gives
better results. The principle is to use the transistor (T) as a
variable resistor by a feedback control loop (see Fig. 11.9).
Transistor resistance depends on emitter-base potential which is
controlled by output voltage. Output (Vo) and standardised
(zener) voltage (V) are compared and error voltage (e) is fed to
the transistor base through a difference amplifier (A).

1--1 i

v,
Fig.11.9 TRANSISTOR STABILISATION CONTROL
 CHAPTER 12

ACTUAL CONTROLLER TYPES

There are numerous forms of controllers produced by a large

number of specialist instrument and control equipment manufac-
turers. As a representative selection of present well proven
practice twelve types are given; one mechanical, one mechanical-
hydraulic, one electro-pneumatic, three electronic and six
pneumatic. In some controllers the action is given as (P + I + D).
It should be remembered that the most simple control suitable
ought to be used and the controllers described are also available
as simple (P) or (P + I) or (P + D). The description so given is
mainly to cover the more involved case which can then be easily
simplified.

DRIVE A( /iI'O

Fig. 12.1 MECHANICAL CONTROLLER

186 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ACTUAL CONTROLLER TYPES 187

(1) MECHANICAL CONTROLLER Due to CD rotation pilot valve X moves down allowing R to
Utilises a lever principle extended to a mechanical differential. move down due to oil escape to drain. As R moves down to a new
A drives (1) anticlockwise and B drives (2) clockwise. If A and B equilibrium position, AB pivots about B and rotates CD clock-
have the same speed then C revolves but the shaft has no linear wise so closing X, locking R and restoring D to its original
motion. If A tends to speed up then C follows (1) and the shaft position. The engine is now running at original speed but at a
moves linearly to alter the supply and increase the revolutions of different load. The governor is P + I, i.e. isochronous. To reduce
B. A proportional action exists, synchronism reduces movement the governor to simple proportional pivot A ,is fixed, R and X
to zero, a distinct anti-hunt characteristic exists (see Fig. 12.1). eliminated, and link CD removed. The conical form spring gives
linearity to the speed measuring system.
(2) MECHANICAL-HYDRAULIC CONTROLLER (GOVERNOR)
A similar device is shown in Fig. 14.9 and an electrical alterna-
The device is illustrated in Fig. 12.2. tive in Fig. 14.10.
(3) ELECTRO-PNEUMATIC CONTROLLER (PHILIPS)
A
The controller shown in Fig. 12.3 is a converter. A sliding
C contact resistor 8 is moved by the valve spindle to form the
position feedback and the input measure signal is at contact M. If

~I
.... ---,
::2''' , • Vs ., .~~ :
I
I
I
~ QC
I
DRIVE SUPPLY
I
P
VM
•
I
I
I
I
t
I
I
A I
t . --~-7I-:H--------:l I
OIL OIL TO FUEL
l~'~' :
;t~~ J ?,: ~J L~~~J
INLET DRAIN CONTROL

Fig. 12.2 MECHANICAL-HYDRAULIC CONTROLLER (GOVERNOR)

For sayan increase in engine load the flyweights move radially

inwards and pilot valve Y moves down so allowing the servo- Fig.12.3 ELECTRO-PNEUMATIC CONTROLLER
power amplifier (8) to move up under admitted oil pressure. This say V M + V s > 0 then relay A is energised and supply air flows to
increases fuel control setting and also rotates feedback link AB the diaphragm top to move the spindle down. If V M + V s < 0
and reset link CD in an anti-clockwise sense about pivot A relay B is energised and air from the diaphragm top is vented, so
(initially fixed as reset piston R has equal pressures on each side). allowing the valve spindle to move up. Movement ceases as soon
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
ACTUAL CONTROLLER TYPES 189
188

X and Y. This current is arranged to be opposite in direction to

as V M and V s equate to zero when both nozzles are closed and no
current flows. Proportional band, for the valve positioner, is the current between A and B. The fixed resistance R therefore
has opposing current flow and when the volt drop across the
adjustable at P.
output bridge equals the volt drop across the input bridge (i.e.
feedback equals input) the voltage across the amplifier is zero,
(4) ELECTRONIC CONTROLLER (KENT)
Referring to Fig. 12.4 (and also refer back to Fig. 11.7). motion stops. Therefore for each A position there is a propor-
tional X position.
Considering first just the proportional action. The deviation (by
the transducer) causes movement of the adjustable rheostat at A For output to equal zero, feedback volts equals supply volts
so that A and B now have a voltage difference, i.e. the input and is opposite in direction, i.e. E F = -E s or, as E s is adjustable,
Wheatstone bridge circuit is unbalanced, current therefore flows due to the adjustable rheostat for proportional band, and intro-
ducing linkage and rheostat proportionality factors:
between A and B.
REV: CORRECTING <I> = - J-L0
I MOTOR SIGNAL

~r=ll:---~
BELLOWS
I To add integral action replace R by an adjustable resistor R 1
--~~ and a condenser of capacitance C 1 • Thus there is a continuous
II
.- 8
~
II
: I
MEASURE
adjustment of feedback. Current flow charges the capacitor
DEVIATION which then resists further current flow.
INPUT
OUTPUT
To add derivative action retain R and add an adjustable resis-
INPUT BRIDGE
BRIDGE, /' tor R D and a condenser of capacitance CD' The capacitor must be
charged at a rate equal to the rate of change of input supply E s.
PROPORTIONAL
BAND To give (P + I + D) omit R and replace by the dotted circuit
ADJUSTMENT
RHEOSTAT containing CD, R D, C b R 1 as shown in the sketch. A more modern
design of this type utilises photo cells and electronic valves (or
transistors) but the principle is much the same as above.

Note:
I
I
I CD
I
I
Mathematics could be introduced to show:
HH
I I E= E F + Ee
:R1fRD
I
~I E = EF + _1_JEF dt (integral)
L_-U-;.l RICI
C1
and
Fig. 12.4 ELECTRONIC (P + I + D) CONTROLLER E= E F + ED
This current produces an input voltage which is an indication
of the error (deviation) magnitude from the desired value. Cur- E = EF + RDCD dEF (derivative)
dt
rent flows to the amplifier, then to the electric motor. As the
E is voltage on output bridge.
motor moves to alter the controlling element it moves the adjust-
E c is volt drop across capacitor C I.
able rheostat at X so that X and Y have a voltage difference, i.e.
ED is volt drop across resistor RD'
the output bridge is unbalanced; current therefore flows between
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
ACTUAL CONTROLLER TYPES 191
190

Possibly a more simplified sketch arrangement of an alterna- procedure is usually necessary as dc amplifiers are subject to drift
tive type of electronic controller (Philips) is shown in Fig. 12.5. which at zero true input would affect the deviation and hence the
controller output.
The voltage developed across the correcting unit signal from
(5) ELECTRONIC CONTROLLER (PHILIPS)
the actuating unit is fed back through the derivative and integral
Referring to Fig. 12.5.
networks in cascade. Some interaction occurs which is reduced
by an auto-resistor in the derivative unit. Note that the final
positioner action gives signal feedback whereas an alternative
described in all controllers given previously has the feedback
E,
signal straight back after the amplifier itself. An obvious advan-
tage of electronic controllers is their flexibility, high speed of
signal transmission and high gain. As automatic controis develop
more complex controller actions, further than P + I + D, will
DESIRED certainly be required and electronic controllers will provide the
VALUE
most convenient method. Voltage difference e 1 a () and feedback
r---'"IIr-----, \ I E2 voltage is ez, equilibrium exists when el = ez. I and D actions are

If
I C I I RO I generated by resistance capacity systems C I R I and CDR D. Action
1 II
I
I
I
R1 II
II
Teo I
I
I
times are adjusted by variable resistors R 1 and RD.
I I: I With R D equal to zero then P + I action is generated only,
L JL----J i.e. integral action shunted via CD.If R 1 is made infinite then P
INTEGRAL DERIVATIVE
+ D action is generated, CD R D and C 1 are in series.
Fig. 12.5 ELECTRONIC (P + I + D) CONTROLLER (2)
(6) ELECTRONIC CONTROLLER (FOXBORO)
The detector element transmits a dc signal (measured value) The circuit is much simplified to facilitate the description of
this modern unit. Refer to Fig. 12.6:
which is usually amplified and may then be in the range 0-40 V,
0-10 rnA. Similarly the desired value from a fixed resistor is The derivative (rate) circuit is shown at A, B, C. Time adjust-
ment is by adjustable resistor-capacitor (A) to the measure
supplied in opposition to the first signal. The two signals are
compared and the difference between them is obtained by pas- input. Rate amplifier (B) is transistorised (solid state), two stage,
npn, dc type, of overall gain about 10, and introduces a signal to
sing the two currents in opposite directions through a common
the bridge network (D) proportional to the rate of change of
resistor, thus voltage across the resistor is proportional to devia-
tion. The resistor acts as a sensitivity control potentiometer in measurement. Separate derivative addition gives no interaction
that the proportional band can be varied by varying the ratio and rapid action as the unit is in the forward loop and not in the
between the two parts of the potentiometer, the variation is 2 to feedback loop. A filter C gives smoothing (RC network).
200% range, or more. Thus pure proportional action is obtained The dc measure signal (10-50 rnA) and set point signal are
with direct action from the measured value and negative feedback developed across resistors and any error signal will unbalance the
bridge until the amplifier output and negative feedback rebal-
from the desired value.
The signal is then passed through a transistorised chopper ance. A resistor (K) decides the gain by controlling feedback so
(capacity modulator) to give ac which is then amplified in an ac proportional action and variable bandwidth are achieved. The
amplifier and (if necessary) rectified for use at the actuato~. This function of the bridge is to impress the resultant of error and
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
ACTUAL CONTROLLER TYPES 193
192

feedback signals on the main amplifier (H) until equilibrium

occurs.
The adjustable resistor and capacitor across the bridge deter-
mine the rate at which output changes (hence feedback) to drive
the measurement to coincidence with set point, i.e. integral
(reset) action.
~
The controller amplifier part of the circuit is E, F, G, H, I, J.
The main amplifier is similar to the rate amplifier but is four
stage, ac, gain about 2000. The main amplifier bridge circuit
S-u ~
r-=====--: u~
~S
e: 7 -' ~ consists of two diodes (E) arranged so that their capacitance
6~~
I I 1_0
change is proportional to dc voltage change across them, initial
I....... ---_-_...J
L_-=-=-
~ set bias and arranged in opposition for large unbalance to small
r----.., a: signal, plus the split inductive transformer winding (F). Error
I I w

i!
ttl
lJ.. I
I ~
~
):1
...J
...J
oa:
I-
Z
o
u
input causes amplifier bridge unbalance (due to capacitance
change) and an ac mV signal will enter H due to the oscillator (I).
H is providing a positive feedback oscillating circuit for bridge
excitation developed in the oscillator loop, tuned resonant circuit
wi CD
~I ~
Ci for bridge (G).
4 I Wa::: +
.... Output level is demodulated and raised by a two stage output
~ 1>-tlJ.. 8
Z I w« + amplifier (J) to an output signal in the range 10 to 50 rnA, dc
I ~j ~
I (i)u
1--
u power supply, external load (L) and negative feedback (2-15 V)
I f{~ I Z via diodes and resistor K. Note transformer couplings to isolate
oa:
I- controller input and output circuits. External switching is

! ]11
u provided with internal circuitry for set point change, limits,
W
...J
L ~ '-----+-'
W automanual transfer, etc. Essentially the unit described is a
CD chopper-type dc amplifier fully transistorised. It consists of a
N
w- transistor input chopper, a high gain ac amplifier, and a transistor
=
u::: output chopper (demodulator) with feedback.

(7) PNEUMATIC CONTROLLER (RATIO TOTALISER-HAGAN)

This unit gives (P), (P + D), (P + I), (P + I + D) control
actions and also provides addition (or subtraction), multiplica-
tion (or division) and averaging computing actions, as may be
required. This controller is an ideal example to illustrate all the
basic actions in as simple a manner as possible.
Referring to Fig. 12.7.
The principle is that of force-balance using the simple lever
principle. Four bellows act on the beam (lever) and variations of
 194 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

I! ACTUAL CONTROLLER TYPES

bellows forces or level fulcrum ratio (a:b) will affect the mag-
nitude of the output signal.
(p4 - pa)a = (pI - p2)b
195

:.P4 = ~(pI-
a
p2) + pa

variations of a, b, pI, pz or P3 obviously affect the value of P4

---+ OUTPUT (output).
When used for proportional action:
PRINCIPLE
P4 = ~(pi - P2) withpa = 0
a

•
Pz SET VALUE t z SET VALUE INTEGRAL
ACTION
P3
.---.C
i.e. difference between set value and measure value yields a
proportional output signal which is adjustable by the a:b ratio.
When integral action is added:
, b :L a L b I a the restrictor I and capacity tank C give the necessary integral
P, MEASURE
t VALUE
OUTPUT P4
VALUE t
P, MEASURE
t VALUE
OUTPUT
VALUE t154 action via P 3 bellows.
PROPORTIONAL IP) P+I
When derivative action is also added:
the restrictor D and capacity tank C give the necessary derivative
action via an extra p 5 bellows on the extra (lower) totaliser. The P
+ +
Pz SET VALUE INTEGRAL P3 + I output from the upper totaliser P4 bellows is fed to the P3
P3 ACTION
bellows of the lower totaliser to give P + I + D output from the
b
- -a lower totaliser p bellows.
L
MEASURE OUTPUT P
P, VALUE P MEASURE
, VALUE VALUE t4 For averaging:
t AVERAGING
Taking a = b then as pz = P4
P4 = HPl + P3)
t
P3 Ps P3
For addition:
DERIVATIVE
ACTION
Taking a = b then as pz = 0
t
P, MEASURE
b I a t P4 P,
b
P4 = Pl + P3
t VALUE
OUTPUT
VALUE t Subtraction can be arranged utilising the other bellows pz in
ADDITION lOR SUBTRACTION) P+ I + D place of bellows P3.
For multiplication:
f b :L a L.-.. withp3 = pz = 0
P, MEASURE OUTPUT tP4
t VALUE VALUE pib = P4Q
MULTIPLICATION (OR DIVISION)
b
P4 = -PI
Fig.12.7 PNEUMATIC RATIO TOTALISER a
 ACTUAL CONTROLLER TYPES 197
196 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

and ratio (a : b) decides multiplication factor if over unity. Divi-

sion can be arranged by making the ratio (a : b) less than unity.
REDUCING TUBE
(8) PNEUMATIC CONTROLLER (FOXBORO)
This unit gives (P + I + D) actions (Fig. 12.8).
Forces exist due to four bellows on the force balancing floating
RELAY
PROPORTIONAL BAND DIAL disc which acts as the flapper, the resultant of moments of bel-
fULCRUMS lows forces determines the throttle position. With the fulcrums

~
PROPORTIONAL BAND NOZZLE over the proportional and reset bellows there is no feedback
"'EASURE"'EN~
ADJUSTING LEVER tr= fORCE BALANCING
effect and the distance between the centre line of the adjusting
SET (O} OR FLOATlNG DISC
lever and the other two bellows is a maximum so giving zero per
(A) BELLOWS ~2~~ RESET BELLOWS
PROPORTIONING BELLOWS - MEASUREMENT (BI cent proportional band (see A). With the centre line of the
OR SET (A) BELLOWS adjusting lever about 1 unit from the proportional bellows and
,\ RESET
CAPACITY about 4 units from the set bellows gives a 25% proportional band
CONTROLLER ITANK
ACTION
(see B). 100% proportional band exists for C. Infinite propor-
RE~rATE tional band exists for D. Note the reversed controller action
available if required. Derivative addition gives delayed feedback
with differential across the resistor where flow is proportional to
- -

_'~
r=
-

=
DECREASING
~EASUREMENT
INC.R
.. EASES OUTPUT
\
INCREASING
ADJUST ABLE
RESTRICTOR
(RESET)
rate of change of deviation. Integral addition gives the usual
delayed feedback, on the reset bellows.

... - ,'f (9) PNEUMATIC CONTROLLER (DRAYTON)

~ 1Q
L jD··.···~
z
Referring to Fig. 12.9.
The sketch is for a proportional controller but other type units
I-OJ. \~_~ : \ /
CONTROLLER MANifOLD
are available with integral or derivative action added. The sensing
element, in this case thermo-sensitive system, on a change of

~@J
conditions will, via the Bourdon tube, alter the flapper position.
This decides the control pressure, this pressure is led back via the
proportioning orifice and sensitivity adjustment to the propor-
tioning bellows. For maximum sensitivity this pressure feedback
(A) 0'1, (B) 25'1,
is vented and no pressure acts on the bellows. If sensitivity is
decreased the resulting pressure build up on the proportional
bellows contracts the bellows and moves the nozzle away from

~~ the flapper. This means a greater flapper travel is required to

~~
close the nozzle, i.e. wider proportional band. The relay pilot
valve assembly shown has a null position non-bleed action.
(C) 100'1, (D) CD
When the flapper approaches the nozzle the control pressure
Fig.12.8 PNEUMATIC CONTROLLER
increases, this lifts the primary diaphragm causing the valve to
open and allow air pressure above the secondary diaphragm to
 ACTUAL CONTROLLER TYPES 199
198 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

the outlets (3). Increase of pressure above the secondary dia-

phragm will tend to balance pressure under the primary dia-
phragm, at balance the diaphragms return to the original position,
the valve closes on both seats, and balanced pressure acts on the
control valve. For flapper travel away from the nozzle then
CONTROL
pressure above the secondary diaphragm will vent until balance
PRESSURE GAUGE is restored.
/
SENSITIVITY ADJUSTMENT (10) PNEUMATIC CONTROLLER (STANDARDISING RELAY - BAILEY)
\ The relay (Fig. 12.10) receives a proportional signal as input
and modifies it to proportional plus integral. The relay consists of
four chambers isolated from each other internally by metallic
diaphragms connected to a central post, spring loaded at one end
and operating inlet and exhaust valves through a beam at the
opposite end. Alteration in chamber A pressure will affect a

SETTING
ADJUSTMENT

\.
, "-
t A --.;;;;;;:::;-
'\ FLAPPER
PRIMARY
NOZZLE
B

~ 1
OPERATING AIF
CONTROL
AT 2-4. bar
o
VALVE

Fig.12.9 PNEUMATIC DIALSET CONTROLLER

AIR
SUPPLY
Fig.12.10 PNEUMATIC STANDARDISING RELAY
 "I
200 REED'S INSTRUMENTATION AND CONTROL SYSTEMS ACTUAL CONTROLLER TYPES 201

similar change in chamber D due to the repositioning of the FINE MESH CHAIN DRIVE

valves caused by beam movement from the bellows. Flow also NEUTRAL
I • POSITION I ,0
occurs to C via the restrictor throttle valve, as C is at a lower • INDICATOR

pressure, so giving a similar effect to the initial down movement

but at a rate dependent on the deviation, i.e. integral action
(adjustable by throttling valve). This regenerative effect con-
tinues until there is a return to the required setting when the DIFFERENTIAL
AIR PRESSURE
supply proportional controller restores A chamber to balance BELLOWS I
RATIO
I SCALE
giving a balance of C and D chambers and repositioning at the I
I I
I I
new required position. The spring maintains a given set value, at I PILOT:
equilibrium both valves are slightly open. I VALVE,
I I
The relay is easily modified to an averaging relay, i.e. taking REGISTERS hp
I I
two signals, combining and giving a resultant output. Loading
pressure comes into A as before and also in through the throttl-
A1R-TAPPiNG-

~~N~~E_~ __ J
I
CONTROL
AIR
t
l I
,SUPPLY
,AIR
10
AIR TAPPING OUTPUT I ,INPUT
I
ing value to C directly. Thus two effects are combined, the C I I

signal can employ any time delay dependent on the valve orifice ."'I,\·\·\""",,·u RAT I0
ADJUSTMENT
opening. C and D are not connected.
Such a relay controiler is characteristic of many similar designs Fig. 12.11 FUEL-AIR RATIO CONTROLLER
usually called stack type (often without use of beams, i.e. valve
and diaphragm action in one line only). They are sometimes The correct fuel-air ratio can be maintained irrespective of the
termed' blind controllers as they are often sited close to the number of burners in use provided that air registers are closed on
control function to minimise distance-velocity lags. burners which are not in use.

(11) FUEL-AIR RATIO CONTROLLER (BAILEY) (12) VISCOSITY CONTROLLER (VAF VISCOTHERM)
Change in combustion air flow is measured in terms of pres- The viscosity sensor has been described previously (Fig. 5.2).
sure difference across the air register, and is transmitted via the The high pressure connection ( +) and low pressure connection
large bellows to the ratio beam (Fig. 12.11). Change in fuel oil (-) is led to a dp cell. Consider now Fig. 12.12:
pressure, caused by the master pressure controller due to varia- Differential pressure is applied across the diaphragm (D) of
tions of steam pressure, is fed to the smaller OF bellows. These the transmitter (cell). Increasing differential pressure (increasing
two signals are in opposition when applied to the beam system. viscosity) causes the diaphragm and balance beam to move to the
Between the beams there is a movable roller fulcrum the left. The inlet supply nozzle (B) is opened by the flapper (F)
movement of which, by the ratio adjustment screw, gives differ- which allows build up of air pressure in the feedback bellows
ent equilibrium conditions and the ratio is indicated on the ratio (B). This gives a restoring action on the balance beam until
scale. Beam lever position operates a linkage to the pilot valve equilibrium is again reached. Discharge nozzle (A) is shut. Air
which varies control air output signal. This output signal is fed to pressure in the feedback bellows is output signal of the controller
the averaging relay where it "trims" the signal being fed through through C to a diaphragm valve regulating steam to the oil fuel
to the air damper actuators. The adjustable proportional band heater.
and negative feedback bellows should be noted. This type of For decreasing viscosity, discharge nozzle (A) is opened giving
controller utilises proportional control only. air bleed, and inlet nozzle (B) is closed. At equilibrium nozzles
202 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

- +
,
1
I

Lr=-1
;
0

c
CHAPTER 13

TYPICAL CONTROL CIRCUITS

A The number of different control loops utilised is large. Appli-

B
cations for each of steam, motor (IC engine), general, engineer-
ing knowledge sections will now be considered. Much process
control is still pneumatic, illustrations are mainly biased to this
Fig. 12.12 VISCOSITY CONTROLLER type. Electronic sensing and control devices can easily be substi-
tuted but the final power control element is often pneumatic.
are virtually closed which reduces air wastage. Zero adjustment Control of displacement, velocity and acceleration using
is at Z and range adjustment at R. The proportional action is electrical-electronic servo-mechanisms are detailed in the next
readily extended to include integral action by adjustable reset chapter.
control on the feedback bellows. Sensor and controller are
described as a control circuit, see Fig. 13.19. STEAM PLANT
Auto-combustion and feed system control have been used for
many years. Modern sophisticated sub-systems for temperature,
pressure, level, flow, etc., are interlinked into boiler and turbine
overall control systems suitable for remote and transient condi-
TEST EXAMPLES 12
tions.
1. Sketch and describe an instrument to maintain the viscosity (1) EXHAUST RANGE PRESSURE CONTROL
of a fuel at a constant value. Explain how it corrects any deviation This utilises sequence operation with valve positioners, see
of the viscosity from the desired value. Fig. 13.1.
Range pressure (1 bar) is sensed and converted to a pneumatic
2. Explain why load sensing governors are usually fitted to
pressure signal by the transducer. This signal is transmitted and
engines driving alternators. Sketch a governor for this duty and
compared with the set value and any deviation 'causes the con-
explain its action. troller to give an output signal change proportional to the devia-
3. Describe, with the aid of a sketch, any type of three term tion. Output signal 2 to 1·75 bar gives dumping to condenser,
electronic controller. Discuss how the separate control actions 1.75 to 1·5 bar 1p bleed is fed in, 1·5 to 1·25 bar desuper-
are generated and adjusted.
 1
204 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TYPICAL CONTROL CIRCUITS 205

EXHAUST STEAM RANGE

(3) SOOTBLOWER CONTROL
The sootblower system utilises air for both control and blow-
ing. Operational rotation of the blower head is achieved by
means of an air piston ratchet mechanism. The sequence of
operation is governed by the distributor operated by a similar air
DESUP:
STEAM piston ratchet mechanism. The supply air to both pistons comes

AIR SUPPLY

AIR SUPPLY

Fig. 13.1 EXHAUST RANGE PRESSURE CONTROL

heated steam is fed in whilst bleed remains full open. Thus

exhaust range is constant pressure maintained. On the diagram R
is for relay and VP is for valve positioner. This is split range
control.
Fig.13.3 SOOTBLOWER CONTROL
(2) TURBINE GLAND STEAM PRESSURE CONTROL
Such an arrangement is given in Fig. 13.2 from which it will be from the pilot valve whose operation is dependent on the charg-
noted that gland steam pressure is sensed and supply either ing and discharging rate of the air reservoir. Adjustable control
increased or dumped. orifices are provided at entry to reservoir (charging) and on the
atmosphere line (discharging). Each impulse or air puff blast
rotates both ratchet gears by one tooth and gives a blowing blast
• AIR SUPPLY
for a few seconds (Fig. 13.3).
1
I
I (4) CONDENSER CIRCULATING WATER TEMPERATURE CONTROL
1
I This system utilises constant pump speed with water recircula-
VALVE ---I tion. For reduced power or low sea temperature operation the
POSITIONER L_ I I

condenser may be operating far from design conditions. Exces-

sive high vacuum results in possible turbine erosion, and low
TO CONDENSER I I .. TO GLAND temperatures give excessive undercooling of condensate which
OR HEATER STEAM BOX
reduces plant efficiency.
Referring to Fig. 13.4:
, • GLAND STEAM SUPPLY
A fall of sea temperature would generally be arranged to
Fig. 13.2 TURBINE GLAND STEAM PRESSURE CONTROL decrease the air loading pressure, giving bellows expansion,
206 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

r-----------------,
I
I
DISCHARGE
OVERBOARD ~ ~
~ I
I'~~
J
, TYPICAL CONTROL CIRCUITS 207

I
£
-------, I
I I
I
+ PRIMARY
SUPERHEATER
,
I
I
I
AIR I
I
SUPPLY : I
I I
I
I
I
I CONTROL "00
O
I LOADING
I PRESSURE
SECONDARY SET J~
.-@__• ~
i
SUPERHEATER

I
CONTROLLER J J
I
STEAM FLOW
: VALVE
,THERMAL ELEMENT Fig.13.5 STEAM TEMPERATURE CONTROL
CONDENSER

Fig.13.4 CONDENSER CIRCULATING WATER TEMPERATURE CONTROL

which through the positioner allows air to the top of the servo
II positioners) could be used in place of two linked valves from one
positioner.

piston so allowing more of the scoop into the discharge pipe and Note.
giving more recirculation. Three term usually means P + I + D and three element usually
means three variable combination in a controller e.g. pressure,
(5) STEAM TEMPERATURE CONTROL level, flow.
Referring to Fig. 13.5:
Superheat control is based here on the amount of steam flow (6) STEAM FLOW/AIR FLOW RATE CONTROL
through the attemperator. The sensing element input signal to Refer to Fig. 13.6:
the steam temperature transmitt€r (T) is directed to the Steam flow rate is sensed at the orifice plate with tappings to a
recorder-controller (C), which is often three term. Output signal steam flow transducer consisting of a dp cell (1) usually supplied
from this controller is combined in an adding relay (A) with the with condensed water. The cell would incorporate a square root
output signal from a steam flow transmitter (F). Relay output eliminator, perhaps mercury well type with mechanical linkage
signal passes through the control station (S) (hand-auto) to oper- to variable inductance (2) operated amplifier (3). The air flow
ate the valve positioner with linked control valves to vary attem- rate is similarly sensed. Outputs from both dp cells are fed to an
perator flow rate. electronic computing relay whose output signal is related to
Two element control allows more effective operation during required air flow for the measured steam demand. A tapping can
transients. Increased steam flow would reduce temperature be arranged to the burner fuel supply controller. The computing
without the second element action which would be reducing flow relay output signal enters a three term electronic controller. The
through the attemperator. Split range control (two valves and electrical signal output from the controller operates the final
208 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

~ TYPICAL CONTROL CIRCUITS 209

II
II
r-
r I !' ,--_ .. __ .. ,
-+-D¥
BRIOGE c=J
o
1-
II

1 --- ---- ---- --- --- ---- ----

I
I ENGINE
I IPRINTERI
ROOM
I
AIR~
FLOW •
I
I
~ ICONTRo::n..J..~MANOEUVRING
BRIER POSITION
TRANSMITTER 1
I
I
,
I INDICATOR __
I I
2131 : I
--"'---
II -g'
T COMPU ING
BRIDGE OR 1-•
ENGINE
ROOM
SELEC~
..J

r • ..,
I
ELECTRO-
PNEUMATIC
CONVERTER
'-------+-----1'"
-D RELAY I I
--- - I
I
THREE TERM
CONTROLLER
L.
I
I
I
IL II I
• ..J
ESSENTIAL SAFETY
Fig. 13.6 STEAM FLOW/AIR FLOW RATE CONTROL
,
I
I
CIRCUIT LOCKS

control element. This element is a damper in the inlet air duct.

-·-AHD. -- ...AST.

An electro-pneumatic converter and power cylinder are

required. Note direct signal measures of the four variables at pen
recorder.

(7) BRIDGE CONTROL (TURBINE MACHINERY)

Fig. 13.7 BRIDGE CONTROL (TURBINES)
Instrumentation and Alarms.
For the bridge console the least instrumentation and alarm
indicators the better, alarms should be essentials only and
instruments only those vitally necessary.
 ~ ;
..•.•...•........

TYPICAL CONTROL CIRCUITS 211

210 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
'I
Suggested alarms could be: Timing Relays
1. High salinity, 2. low feed pump suction pressure, 3. high To prevent excessive speed changes, by too rapid signals,
condenser water level, 4. low vacuum, 5. lubricating oil pressure, which would endanger engines and boilers. For example half
6. tank contents low level. For direct instruments opinion is speed in say! minute, emergency swings from say full ahead to
divided but no more than say another six indications should be full astern allow astern braking steam usage, rotation slightly
necessary. Engine console and alarms would obviously provide ahead and astern at say 3 minute intervals of a sustained stop of
full instrumentation. A typical simplified system is given in Fig. engines, etc. (manoeuvring valves now usually cam operated
13.7. This system has direct control at the steam manoeuvring sequential).
valves. Essential Safety Locks
This is essentially a combination electro-pneumatic although Override on timing by such elements as low boiler water drum
all-pneumatic or all-electric can easily be arranged. level, low turbine inlet steam pressure, etc.
The following points with reference to Fig. 13.7 should be Emergency Control
noted: Direct hand control of manoeuvring valves camshaft.
1. It is assumed that all normal safety protection devices and Local Control
control are provided, e.g. loss of lubricating oil pressure, high or Independent power control at the actuators themselves.
low water level in boilers, electrical failure, etc. Feedbacks
2. Subsidiary control loops have been omitted, e.g. The steam pressure feedback gives accurate positioning for
evaporators, generator, etc. pressure of the actuator. Speed feedback is arranged so that a
Consider now the individual aspects relating to Fig. 13.7: difference of speed between measured and desired values causes
an additional trimming signal to the controller. This may be
Selector necessary as pressure and speed are not well correlated at low
Bridge or engine room control can be arranged at the selector
speeds.
in the engine room. When one is selected the other is ineffective.
Outline Description
Duplication The following is a brief description of one type of electronic
Both transmission control systems are normally identical and bridge control for a large single screw turbine vessel to illustrate
operation of the one selected gives slave movement of the other. the main essentials. Movement of a control lever modifies the
Manoeuvring output of an attached transmitter (electronic signal 0-10 rnA dc).
A separate sequence is arranged, for example opening astern The transmitted signal is passed, via override, alarm and cut out
master valve, opening turbine drains, etc., but if this separate units, to the desired flow module which is connected to a time
control is not applied separately it will be automatically applied relay and feeds to the controller. The electronic controller com-
from the main transmitter. pares desired speed with actual speed as detected by a tachome-
Interlocks ter generator and dc amplifier. The correct controller signal is
Essential blocks are necessary such as no valve opening with passed to the manoeuvring valve positioner from which a return
turning gear in, etc. signal of camshaft position is fed back to the dc amplifier, thus
Programme Relays giving the command signal to the actuator reversing starter. The
The correct sequence operation is arranged for the various two control levers are independent and do not follow each other,
actions for example, manoeuvring conditions satisfied before an engine room override of bridge control is supplied. The rate of
moving on to say valve opening sequences, etc. valve opening is controlled by the actuator so that too rapid valve
 TYPICAL CONTROL CIRCUITS 213
212 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

cooler and functions as a mixing valve. This valve may be of the

opening is prevented by a time delay, full normal valve operation
rotary cylinder type or diaphragm operated type. Integral action
shut to open, or vice versa, occurs in about 1 minute, this can be
is usually incorporated in the controller as offset may be appreci-
reduced to about 20 seconds in emergency by full movement of
able otherwise (up to goC). This system is often preferred. An
the telegraph from full speed direct to or through stop. A near
alternative arrangement is to throttle the sea water supply for the
linear rev/min to control lever position exists. Auto-blast refers
cooler. This gives a big capacity lag in the system as one variable
to the automatic time delay opening of the ahead manoeuvring
(sea water) controls the other variable (fresh water). Valve selec-
valve for a short period after a certain length of time stopped -
tion is most important. Maximum pressure and temperature,
this has an override cut out for close docking, etc.
maximum flow rate, minimum flow rate, valve and line pressure
drops, etc., must be carefully assessed so that valve gives the best
IC ENGINE PLANT results. The controller shown in Fig. 13.8 is pneumatic but an
IC engine plant now utilises general (mainly thermal) auto- electronic controller with a rotary valve is also common. Engine
control to a great extent and manoeuvring systems, etc., have coolant is shown as fresh water but could equally well be oil.
always involved fairly detailed devices. The efficiency is virtually Most loops on a motorship are single element, for example
inherent in the design, however auto-control can still give temperature control of lubricating oil, fuel, air, water, etc.
improvements to efficiency of operation.
(2) TWO ELEMENT COOLING LOOP
(1) JACKET (OR PISTON) TEMPERATURE CONTROL - SINGLE ELEMENT
More sophisticated loops are sometimes found on auxiliary
boiler controls (pressure, level, etc.) as well as main engine
Referring to Fig. 13.8:

CONTROLLER 1-.. -AIR SUPPLY __@Q:v.

SW INLET
.--------
•
I
-.
I
I
L_...,
FW RETURN
'~..:.-=-----
TL_- 1 II

•
FRESH WATER
I~ ~ I .1 ~ RETURN FROM
i I I ENGINE FW SUPPLY _ I
I SENSING ELEMENT

__ ~~vJ
: (THERMAL)
COOLER
FRESH WATER
I ~ SUPPLY TO
ENGINE

I
I
I
SW OUTLET J
TUATOR
I
Fig. 13.8 JACKET (OR PISTON) TEMPERATURE CONTROL - I >-...,I
SINGLE ELEMENT I
II I
The sensing element may be on the supply or return line to the L______ _ . OVERBOARD ..J l
engine, there are certain advantages for each case. The 3 way
Fig. 13.9 TWO-ELEMENT COOLING LOOP
valve (2 entry and 1 exit) varies the re-circulation or supply to the
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
TYPICAL CONTROL CIRCUITS 215
214

coolant. In the latter case a two element type may be advantage-

TO TO
ous as heat transfer rates are high and sea water temperature REFRIGERATOR GENERATORS
becomes more critical than in smaller single element loops. The
two variables involved are engine load and sea water tempera-
ture.
If the engine is considered to be at a fixed load then by
reference to Fig. 13.9 it is seen that the water inlet temperature is SPARE AND
fixed by the set value of the controller A which accounts for OTHER SERVICES:-

water temperature changes (controller A is "slave"). AIR CONDITIONING}

DISTILLER ETC.
Now if the engine load changes the inlet water temperature AUX:CONDENSER
should change, i.e. the lower the load the higher the water
temperature. This is achieved by changing the desired value of
controller A according to the engine load variations. Controller
B (the "master") provides an indication of engine load by
measuring the return water temperature. Controller B signal TO AIR
changes the set value of controller A (cascade control). A fresh COMPRESSORS
water heater may be placed in the engine· supply line with heat
input controlled by controller A. Split range (level) control
allows heat input at low coolant temperatures and cold input at HARBOUR
CIRCULATION
high coolant temperatures. The system works equally well for oil
coolant. The actuator can be operated by local or remote control
with the controllers out of operation.
Fig. 13.10 OVERALL COOLANT SYSTEM CONTROL (RING MAIN)
(3) OVERALL COOLANT SYSTEM CONTROL
The sketch given previously in Fig. 13.8 for jacket (or piston)
coolant temperature control is typical. Similar systems can apply divided with say two constant speed and two controllable speed
almost exactly for lubricating oil, turbo blowers, scavenge to allow flexibility, harbour circulation is best arranged by a feed
air,fuel valve, etc., coolants. Each system is separately controlled in to the main from say the ballast pump. The individual control
so that adjustment can be made to individual systems without is shown on the jacket cooler only for illustration. The control
here is direct control of the sea water quantity utilising two-way
affecting the others.
One obvious arrangement for sea water supply is to utilise diaphragm control valves, all main engine circuits shown in the
independent tappings (both main engines and auxiliaries) from a diagram are so controlled, any auxiliary circuits can be similarly
continuously circulated ring main - in this design individual sizes controlled if required. Certain discharges, where convenient, can
and flow rates require very careful investigation. A typical ring be combined to reduce the number of shipside valves.
main system is given in Fig. 13.10. For simplicity no duplication
of coolers, indication of individual valves, strum or strainer Another alternative is a series or parallel arrangement
boxes, etc., is shown. Note that only main engine circuits are The two diagrams given in Fig. 13.11 illustrate series circula-
controlled in this given case. It is advisable to have say four tion with three way valves with bypass, and parallel circulation
pumps capable of operation on the main, the pumps preferably with two way valves with direct supply, the latter arrangement
 TYPICAL CONTROL CIRCUITS 217
REED'S INSTRUMENTATION AND CONTROL SYSTEMS
216

OVERBOARD
,OPERATING
I

I
AIR
DIRTY BOILER OIL TANKS

I
OVERBOARD n-c;r-_..J
lFI~---~-~
~~~1~OL II
SWITCH
I
I
I

r
I
I
I

~
CLEAN
BOILER
!I :
I
!!
OIL I
OVERBOARD TANK I
I
I 1= RELIEF TO
DIRTY
BOILER
OIL TANKS

I
DIAPHRAGM
CONTROL
VALVE
~

Fig. 13.11 OVERALL COOLANT SYSTEM CONTROL

(SERIES & PARALLEL)
Fig. 13.12 BOILER OIL PURIFICATION CONTROL
SYSTEM FOR IC ENGINE

requires a satisfactory system pressure control such as control

flow from pump or dump of excess water (as shown). Grouping
of the coolers and choice or combination of systems can be
arranged dependent on flow and temperature considerations. No
(5) WASTE HEAT FLASH EVAPORATOR CONTROL
duplication, etc., is shown. This system is shown in Fig. 13.13. Waste, or low grade heat,
from engine coolant has a good energy potential. Fresh water
(4) BOILER OIL PURIFICATION CONTROL SYSTEM FOR IC ENGINE
jacket (or piston) coolant evaporates sea water in the second
This system shown in Fig. 13.12 is designed to maintain a
stage heat exchanger which is condensed at about 0·1 bar in the
working level in the boiler oil service tank to the main engine.
first stage pre-heater and removed by the distillate pump.
The oil supply from the dirty oil tanks continuously passes
When the pressure sensor-transducer (P) allows the controller
through a self cleaning purifier. The oil fuel heater (at purifier, or
(C) to operate the sea and coolant inlet valves, vapour produc-
if also provided on the main engine supply rundown) can easily
tion starts. Control is by measurement of made water flow at the
be arranged to give fixed oil temperature or controlled viscosity
flow sensor transducer (F) and the signal allows the controller to
in a similar manner to the previously described temperature flow
regulate water inlet valves accordingly.
control system.
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
TYPICAL CONTROL CIRCUITS 219
218

SW
IN ----------MR---'
r-
I EJECTOR AIR
OUT
I
I
I
I
lI
SPEED FEEDBACK
1-----

SCRUBBER_~
DE-MISTER

111
,,,,,,,,,,,,

BRINE
EJECTOR

Fig.13.13 WASTE HEAT FLASH EVAPORATOR CONTROL

(6) BRIDGE CONTROL (DIRECT REVERSING IC ENGINE)

It is suggested that the section on bridge control for turbines
should be read again as there are many obvious similarities. I L Y
.......---...
I
ENGINE
RUNNING IN
CORRECT
DIRECTION
ENGINE
RUNNING
ABOVE
FIRING
Instrumentation and Alarms SPEED
Minimum usage, suggested alarms (bridge console): --.--
1. Low starting air pressure, 2. lubricating oil discharge pres-
sure and temperature, 3. cooling water discharge pressure and
temperature, 4. tank contents level gauge, 5. fuel oil discharge F
pressure and temperature, 6. scavenge belt pressure. TACHO
ADJUSTED
TO CORRECT
A further six instruments could be provided. GENERATOR SETTING
I
t..-o-------- -------.GOVERNOR
Note:
All normal protective devices are assumed, subsidiary control Fig. 13.14 BRIDGE CONTROL (DIRECT REVERSING IC ENGINE)
loops are not considered.
Refer to Fig. 13.14 and consider the following:
 _ TYPICAL CONTROL CIRCUITS 221
220 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

4. Movement of control lever for fuel for a speed out of a

Selector critical speed range if the bridge speed selection is within this
Bridge or engine room control is in the engine room. With one
range.
selected the other is inoperative. 5. Emergency full ahead to full astern, etc., actions, must have
time delays to allow fall of speed before firing revolutions, astern
Duplication air open, engine stop, correct astern timing and setting.
Both transmission control systems are identical master and
slave functions as selected. Outline Description
The following is a brief description of one type of electronic-
Interlocks pneumatic bridge control for a given large single screw direct
Essential blocks are necessary, such as no action with turning
coupled IC engine to illustrate the main essentials. The IC engine
gear in, etc. lends itself to remote control more easily than turbine machin-
ery.
Programme and Timing Relays Movement of the telegraph lever actuates a variable trans-
Consider say a requirement of "half ahead". The programme
former so giving signals to the engine room electronic controller
must satisfy a sequence such as: which transmits, in the correct sequence, a signal series to oper-
1. Fuel admission check for zero. ate solenoid valves at the engine. One set of solenoid valves
2. Air lever to ahead. controls starting air to the engine while a second set regulates
3. Sufficient air and camshaft direction checks. fuel supply, the latter via the manual fuel admission lever, is
4. Air admitted. coupled to a pneumatic cylinder whose speed of travel is gov-
5. Adjustable delay period for firing speed. erned by an integral hydraulic cylinder in which rate of oil
displacement is governed by flow regulators. This cylinder also
6. Fuel admitted. actuates a variable transformer giving a reset signal when fuel
7. Delay checks for air off, correct direction, rotation above lever position matches telegraph setting.
firing speed. With the engine on bridge control the engine control box
8. Adjust fuel for set value speed. starting air lever is ineffective and the fuel control rack is held
clear of the box fuel lever. Normal fuel pump control is elimi-
Essential Safety Locks nated and fuel pressure, in a common rail, is automatically
Override on timing by such elements as low lubricating oil adjusted to speed and load by a spring loaded relief valve. Engine
pressure, low cooling water pressure, etc. override of bridge control is provided.
The function of the electronic controller is to give the follow-
Emergency and Local Control ing sequence for say start to half ahead: Ensure fuel at zero,
Directly at the engine controls themselves. admit starting air in correct direction, check direction, time delay
to allow engine to reach firing speed, admits fuel, time delay to
Further Protective Considerations
cut off air, time delay and check revolutions, adjust revolutions.
1. Speed governor.
Similar functions apply for astern or movements from ahead to
2. Non-operation of air lever during direction alteration.
astern directly. Lever travel time to full can be varied from stop
3. Failure to fire requires alarm indication and sequence to full between adjustable time limits of! minute and 6 minutes.
repeat with a maximum of say four consecutive attempts before Fault and alarm circuits and protection are built into the system.
overall lock.
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TYPICAL CONTROL CIRCUITS 223
222

ENGINE CONTROL BRIDGE

(7) REMOTE CONTROL (DETAIL - IC ENGINE) ROOM ROOM
Pneumatic reversing and starting systems have been an
integral part of IC engine equipment for many years. Figure 3\ r4

,.~
13.14 and description are fairly general and it is now appropriate
to extend this to a more detailed arrangement. I

Consider Fig. 13.15 and the following nomenclature: I

I
1. telegraph receiver; 2. speed lever; 3. start lever; 4. shut off I

lever; 5. bridge maximum speed lever; 6., 7., 8., 9., 10., 11.
. . I I

,:
selector pilot valves; bridge or control room, ahead, astern,
-~--~~~~=~~=r~,"~a~ ! :~I:
I 4 L4-i~- -r-~I'
~.~--f-~-f----~-~-, t: : :::
speed, start, maximum speed; 12. telegraph transmitter; 13. I I _._ _.J

speed cam; 14. programme motor drive; 15. speed cam (fine L-----f'11 I
control); 16. control cam ahead-astern; 17., 18., 19. selector I
l
t._-t I
u J
--4---~----~=t---F--f
: L~.J I I U L
-7--
J
pilot valves; speed, astern, ahead; 20., 21., 22. control valves; L
"--7-~ r·...J
l I
ro4-, :l I
I
I
I

-- dX!__t}.J
x.. ---rr
~r
flow, speed, flow; 24. solenoid valve; 25. timing volume; 23., 26., l,.... • ~ --.-
27. double check valves. There are ten relay valves, two each i F' I' I II

ahead and astern plus U, V, W, X, Y, Z. A is air supply, B outlet

20
___ .J
: I
I -deL_._>-~""I I
L~
I ~:L~ II
I
r-ID---t- ..- - z ~ ~
- .... - - I

lifts engine speed lever handgrip out of gear to allow remote II ... __ I
Ir~ ~
23 . - - - - ' I~r:::":-l
I~~r I
operation, C outlet to speed set, D from and to direction inter-
locks, E outlet to reversing interlock, F outlet to starting servo, G A. I )'
I
I
22 4I
- -
II
L_,.... ..., • AHO· -
rrB i
-"AST
.r-"
LL I
I
I

4
I
I
I
I
I
,I
I
I
t
outlet to reversing servo, L to solenoid valve.
With the engine telegraph set at remote control position the
control air at a pressure of about eight bar flows through line A to
G...-
0-"-- -- -t ®25 Ijt_-.--f26l--4 -t.J. I
-- - -.-I
G4----.--4~~1~=~~~~~~~~~~J
I,I

- -. s=-' , T,
III

I ,

all nine pilot valves (as shown full lines) if lever 4 is opened. 0-"-- - --I L ...... --f¥J'--._l
There is an additional section on the engine telegraph for remote
t
control and also on the control room telegraph for bridge control I
I
position. As 4 is open air passes through B, C, D, E to operate I
interlocks. Air also passes to the upper ahead or astern direction F+--j---+------I
BCoE+- ....J

selector relay valves, partly clears speed set relay valve X for C.. - --------------
operation (via 21), and clears the starting routine at X and Z via Fig.13.15 PNEUMATIC REMOTE CONTROL SYSTEM
24 (from line L).
Operation of 1 selects either control room or bridge and
output from 6 loads either U or V accordingly. Ahead lines 7, 19
lead to V and one as appropriate feeds top ahead relay valve and Start from 3, 10 through 23 and either of lower ahead-astern
goes out via line G to position the reversing servo at the engine relay valves past 27 via line F to engine starting cylinder servo
(astern similarly through 8, 19 via U and top astern relay valve to (depending on loading signal from either line D direction inter-
locks). 23 is also cleared from 26, through Y and Z, with built in
other line G).
Speed set from 2, 9 leads to W (directly loaded from control flow and timing control at 22 and 25.
room) (via 6 or from bridge via 17,5,11,20) thence to X and line The engine will for example reverse, start and gradually reach
selected speed in an automatically programmed sequence. Two
C to act on speed set servo.
 TYPICAL CONTROL CIRCUITS 225
224 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

fine set speed buttons allow speed variation from fixed values.
The system has built in avoidance of critical speed ranges and
acceleration to full speed when on bridge control. This pneuma-
tic system is readily adaptable to electronic signals to servos.

GENERAL PLANT
General control applications have always been used and the
ship steering gear control system is perhaps a classic example.
Early control components included the safety valve and Watt
governor. A selection of typical systems is now given.
AIR FLOW
(1) AUTO-COMBUSTION AND ATIEMPERATOR CONTROL SYSTEM
For simplicity no feed water control or similar controls are
,I
--,
I I
CONTROLLER

I I I
shown in this system but they are commonly fitted in practice I
II 'I
'FEEDBACK ~IR AVAILABLE SIGNAL
- - - . - - - - ......:
:
IFEEDBACK
(Fig. 13.16). I
For main propulsion steam turbine drive wide range fuel oil I

I
I
burners are a very desirable feature with automatic combustion lAIR FLOW

control as they allow manoeuvring without the need to change

burners. Burners must have a high turn down ratio. Mechanical
atomisation burners have a turn down ratio of about 2: 1 whereas
J"lfd
\-+ c ") {
~=: {:1
ITRANSMITTER

steam atomisation assisted burners have a turn down ratio of

about 20: 1. The "steam" burners however induce a water loss of Fig.13.16 AUTO-COMBUSTION AND ATTEMPERATOR CONTROL
about 0·75 per cent of steam flow. The ideal arrangement is to SYSTEM
utilise "steam assisted" burners during manoeuvring and
"mechanical" burners at full power. This gives full automatic
control over the manoeuvring range. An automatic dumping
The common signal at the adding relay for pressure and flow
valve to the condenser ensures a nominal steam flow when the
adjusts the desired values of both air and oil flow systems. As
main turbines are stopped. Until recently automatic combustion
both these loops are independent sub-groups of a closed nature
control was only used at full power, in fact the main need and
they will keep changing respective flow rates until the desired
modern application is when load changes, i.e. manoeuvring
value (or set value) is reached. Under normal operation steam
transients. flow signals are in advance of steam pressure signals and this
For IC engine plant especially in view of the Clean Air Act,
anticipatory action is good control practice. The fuel limiting
combustion control of boilers is virtually essential.
relay in the oil sub-loop (fuel-air ratio control) receives the
Fig. 13.16 is fairly simple in principle and can be considered
feedback signal from the air flow rate and compares the required
suitable for auxiliary boiler practice in motorships. The detail
fuel rate with air availability. Fuel rate cannot be increased until
should be taken in conjunction with the following system on the
an increased air flow signal allows the alteration of the desired
more fully automated auxiliary automatic boiler. value setting at the fuel controller.
Referring to Fig. 13.16.
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS TYPICAL CONTROL CIRCUITS 227
226

Attemperation also has anticipatory control in that the load 1. The pressure switch initiates the start of the cycle. The
change signal acts before the temperature change signal. The switch is often arranged to cut in at about 1 bar below the
function of the attemperator is to limit steam temperature at low working pressure and cut out at about 1/5 bar above the working
boiler loads, there is a consistent relation between the two. pressure (this differential is adjustable).
Command signal choice between fuel pressure, fuel flow, fuel 2. The master initiating relay now allows "air-on". The air
flow/steam flow etc. -show significant variations in response feedback confirms "air-on" and allows a 30 second time delay to
characteristics and research still proceeds. proceed.
3. The master now allows the arc to be struck by the electrode
(2) AUXILIARY BOILER AUTOMATIC FUEL AND CONTROL SYSTEM relay. The "arc made" feedback signal allows a 3 second time
Refer to Fig. 13.1 7 for the lighting sequence. delay to proceed.
4. The master now allows the fuel initiating signal to proceed.
The solenoid valve allows fuel on to the burner. The "fuel on"
feedback signal allows a 5 second time delay to proceed (this may
be preceded by a fuel heating sequence for boiler oils).
5. The master now examines the photo electric cell. If in order
the cycle is complete, if not then fuel is shut off, an alarm bell
rings and the cycle is repeated.
BOILER
MASTER
INITIATING
RELAY I Refer to Fig. 13.1 7 for emergency devices:
I
I
Obviously failure of any item in the above sequential cycle
...J I causes shut down and alarm. In addition the following apply:
<t
Z
C)
I a. High or low water levels initiate alarms and allow the master
I

NON _
CONDUCTING
u;
a::
o
~
•
I
I
to interrupt and shut down the sequential system.
b. Water level is controlled by an electroflo type of feed reg-
E ulator and controller. Sequential level resistors are immersed in
CHARGING
~
I
FLUID I
a:: conducting mercury or non-conducting fluid, so deciding pump
~ I
speed by variable limb level. The fixed limb level passes over a
FAN weir in the feed box.

NOTE.
FEEDBACK (3) DRUM LEVEL CONTROLLER (FEED REGULATOR)
SIGNAL LINES Robot feed regulators are proportional controllers (single
SHOWN DOTTED
term) working on a fairly sensitive proportional band. Due to
drum contents "swell" and "shrinkage" during manoeuvring
load changes, the action is temporarily in the wrong direction.
This wrong action is very severe due to the narrow bandwidth
Fig.13.17 AUXILIARY AUTOMATIC BOILER CONTROL SYSTEM
and hand operation of feed checks was often necessary. Propor-
tional action is made less sensitive, this reduces the severity of the
228 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

short term wrong way action but introduces offset. Offset is got
rid of by the addition of integral action, i.e. the control is two
term for the single element action.
I TYPICAL CONTROL CIRCUITS

manoeuvring. This signal would give an anticipatory action,

which is usually desirable in all control systems. The level signal
would act as a trimmer and has a wide proportional band so as
229

A two element action is obtained by incorporating a steam not to affect the system during swell and shrinkage.
flow measurement to reduce severe feed flow variations when Three element control gives the highest value of performance.
Feed flow is compared to steam flow for the correct 1: 1 ratio. If
the ratio is incorrect then an out of balance signal is given to the
controller. Drum level again acts as a trimming device on a wide
proportional band with integral action. Single, two and three
element actions are illustrated in Fig. 13.18.

(4) VISCOSITY CONTROL

Refer to Fig. 13.19:
A continuous sample of oil is passed across a capillary tube.

~ED
SINGLE ELEMENT
The measurement of viscosity has been considered previously
(Chapter 5) with a detailed description of a viscometer. Flow is
laminar in the tube so that viscosity is directly proportional to
STEAM FLOW pressure drop. Pressure difference is sensed by dp cell transmit-
/ TRANSMITTER
ter and the signal passed to a controller and recorder. The
controller if supplied by air can transmit a direct power signal to

CONTROLLER

--D' I
~---
TWO ELEMENT
-r
r------4----- -
OJ I 1

RATIO
/ RELAY
I
I
I
I
L__
j «
~ RECORDER , ' -

!
I
·
I
I
dp
CELL
_0-
I
jVISCOMETER

~~ILL~Ryi

<,
OIL
HEATE~>
<.:.-

THREE ELEMENT

Fig. 13.19 VISCOSITY CONTROL

Fig.13.18 FEED REGULATORS
 TYPICAL CONTROL CIRCUITS 231
230 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

TO
STEAM
1
_
~1;9NTROLLER
I 1,-----,
operate a diaphragm control valve. This valve controls steam
input to an oil fuel heater. P control is generally adequate, rate
and/or reset are easily added. The sensor has been described
ZONE
r-CiRE'~A~'t--:-~~ILOT previously (Fig. 5.3) and so has the controller (Fig. 12.12).
t I
I
I'
I ,..
REDUCING I
I
J I
~-l
If ~ALVE
(5) REFRIGERATION CONTROL
l II~I VALVE. Refrigeration and air conditioning utilise some considerable
HUMIDISTAT
lll t-.J. . B INE
T__
_ _ _--t
±
AIR
:::
I III /II L
RECIRC.
==_FRESH DUCT
degree of control. Two examples are now given (Fig. 13.20):
Air Conditioning (upper sketch)

Ic:~UMIDIFIER
Air (fresh and recirculated) is control sprayed with steam to fix

I
I HEAT humidity. Air is heated (steam grid) or cooled (brine grid) with
t EXCHANGER
steam (or brine) quantity controlled by temperature. The air is
TO SUB-MASTER MASTER

ONE +-1rl---1~1.=AIR now passed to the various zones where sub-units adjust the air
temperature to the thermostat setting of the zone. This is
I I ~AMBIENT achieved by controlling the steam supply to a calorifier.
I, THERMOMETER
'_5-
I
Refrigeration Chamber (lower sketch)

J~~~AL6RIFIER
The brine quantity, for adjusting air temperature, is controlled
irrespective of fan direction (suction or delivery) with controller
STEAM bulb in bypass pocket sensing air delivery temperature to
chamber.

-~\PI ::?'ffi~~:~:AT ~--~-AIR

(6) ALTERNATOR CONTROL
Consider the sketch of Fig. 13.21 and regard this as a main
+
CABIN system (A-J) and sub system (K-M).

Main system (load sharing)

One alternator is shown (A) of say a four set installation.
Current, voltage, power factor sensors give power computation
at relay (B) which is amplified (C). Total electrical load is com-
puted at relay (E), fractioned off to enter relay (D). The two
signals at D are compared for load sharing, error signal triggers
(F) and thyristor switches (G) for increase/decrease speed signal
to governor controller (H) with feedback loop to alternator input
power. Total computed power is fed to relay (1) which functions
CONTROLLER
to start another alternator at say 75% maximum rated load of
those alternators in operation. The signal is also fed to stop relay
.AIR
SUPPLY (J) which is arranged to shut down an alternator when computed
load is say 60%, 40% and 20% of maximum rated load for the
four alternator unit.
Fig. 13.20 REFRIGERATION CONTROL
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
TYPICAL CONTROL CIRCUITS 233
232

similar to K -M, is required in which individual alternator off

A loading takes place and circuit breaker opening at 10% full load
occurs.

(7) BUTTERWORTH HEATING CONTROL

On Fig. 13.22, PI and P 3 are pressure sensors, P z pressure
OUTPUTS INPUTS controller; T 1 temperature sensor and T z temperature controller;
i ~ L 1 level sensor and L z level controller; R 1 and R z relays; A 1 and
A z actuators. Signals of pressure and temperature enter R 1 and
N the lower value is passed on to R z. Control is inherently on
temperature with pressure over-ride. Signals from R 1 and the
water-pressure sensor enter R z and the lower value is passed to
A 1. This provides protection against water supply failure. Con-
1
I L.r-r-T-' _
L- !.l_ +- - - -- 3r----- ------,:

G G
{D I---fu
T11-- T2
-,
I
~
II
JR:"'L--JR2L--,
I

~_J I I
:
I
I
.J
P1

Fig. 13.21 ALTERNATOR CONTROL

HEAT
EXCHANGER

Sub-system (alternator start)

This particular sub-system is shown dotted on Fig. 13.21.
Obviously a start signal for another required alternator will need HEAT
EXCHANGER
a pre-start routine relay (K). This would initiate air, lubrication,
cooling, input power, etc., and start the alternator up to running
speed. Relay (L) will arrange synchronisation (voltage, fre- CONDENSATE
quency, phase) and initiate a signal to close circuit breaker (M).
The main control s~itchboard (N) would be arranged to
Fig.13.22 BUTTERWORTH HEATING CONTROL
handle all monitored inputs from individual alternators. Stan-
dard protection from engine faults would be provided. Normal
electrical protection is required, for example, reverse power densate level is controlled as shown. The system can be pneuma-
trips, overload alarm (105%) and trip, preferential tripping, etc. tic, electronic or a combination of the two. The control problem
These could be regarded as module systems within a particular is essentially the difficulty of handling large quantities of water,
sub-system. Obviously a shut down arrangement sub-system, with high velocities, maintaining close temperature control and
 TYPICAL CONTROL CIRCUITS 235
REED'S INSTRUMENTATION AND CONTROL SYSTEMS
234
water pressure opens the oil discharge valve (through 3) and the
controlling one variable by means of another variable. The sys-
water discharge valve closes. Each probe and valve has a signal
tem shown has utilised cascade control principles, effectively
indicator lamp and an alarm bell operates when the lower probe
temperature master reset to pressure slave, the latter being a
bridge is unbalanced. A third probe at a low level can be
pressure control system. arranged to cut out the pump if oil falls to that point.
(8) OILY-WATER SEPARATOR INTERFACE LEVEL CONTROL (9) CONTROLLABLE PITCH PROPELLER
The level sensor has been described in Chapter 3. The control
Use of these propellers has increased with the greater use of
system of Fig. 13.23 utilises two probes with the lower probe
unidirectional gas turbine and multi-diesel drives and bridge
(shown) giving a balanced electrical bridge in water and the
control. Engine room (or bridge) signal is fed to a torque-speed
upper probe (not shown) giving balance in oil or air. selector which fixes engine speed and propeller pitch - feedbacks
apply from each. Consider Fig. 13.24.

.
INPUT
SIGNAL
BLADE ,
I

..
ROTATION
I
I j .... __ DRAIN
AHEAD
I ~ -----
I
THR~ST )I( .. - .- I
I
I
:
BEARING RING L
CRANK \

~~e=t=
--~ ~
INTERFACE SERVO
PISTON
BOSS ....
\,//
PROBE
/

ac
I-SHELL SUPPLY THRUST PISTON TUBE
ROD

Fig.13.23 OILY-WATER SEPARATOR INTERFACE LEVEL CONTROL Fig.13.24 CONTROLLABLE PITCH PROPELLER

With the pump started and supplying water to the separator to

The input fluid signal acts on the diaphragm in the valve
rise to the lower probe level, the bridge is balanced and the housing and directs pressure oil via one piston valve through the
solenoid de-energised. When water rises to the upper probe its
tube to one side (left) of the servo piston or via the.other piston
bridge is unbalanced and the output signal is amplified which
valve outside the tube (in the annulus) to the other side (right) of
energises a "left hand" solenoid (not shown) which moves the
the servo piston. Movement of the servo piston, through a crank
pilot valve to the left. This allows clean water pressure to pass
pin ring and sliding blocks rotates blades and varies pitch.
(from 1) to close the oil discharge valve (through 2). Sheil
The feedback restoring signal, to restore piston valves to the
pressure rises and a spring loaded water discharge valve is
neutral position at correct pitch position, is dependent on
opened. As oil build up occurs the oil-water interface moves
springes) force (i.e. servo piston position) which acts to vary the
down de-energises the left hand solenoid and then energises the orifice (0) by control piston (c) so fixing feedback pressure
"right hand" solenoid, the pilot valve moves right (as shown) and
236 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

loading on the pilot valve (p) in the valve housing. The central
part of the tailshaft includes a shaft coupling and pitch lock
~":-~
x-
l
~"" I
TYPICAL CONTROL CIRCUITS

TEST EXAMPLES 13

1. Sketch and describe a fully automatic, oil fired, packaged

237

device (not shown).

(10) AUTO STEERING (BLOCK DIAGRAM)

l steam boiler.
Explain how it operates.
The ship steering gear (on auto pilot) utilises classic control State what attention is needed to ensure safe operation.
principles best illustrated by a block diagram. Give two advantages possessed over conventional boiler installa-
Fig. 13.25 shows a block diagram for auto steering. The con- tions.
troller will be three term with adjustment for beam sea (or wind) 2. Sketch and describe a system of control for manoeuvring a
and dead band operation to reduce response to small random main engine from the bridge.
signals. Both rudder and ship are acted upon by external forces. Explain how control is transferred to the engine or control room
upon failure of bridge control.
FEEDBACK
3. Describe with a block diagram, an automatic combustion
FEEDBACK control system, with particular reference to the methods of
measuring each of the following data:
(a) pressure,
(b) level,
(c) flow,
(d) temperature.
EXTERNAL
INPUT 4. Define the terms:
(a) cascade control,
Fig. 13.25 AUTO STEERING (BLOCK DIAGRAM)
(b) split level control.
Discuss the application of such principles within a description,
utilising sketches, of an automatically controlled lubricating oil
system and an automatically controlled cooling water system of
the type used in auxiliary diesel driven generators.
,
I

CHAPTER 14

KINETIC CONTROL SYSTEMS

KINETIC CONTROL SYSTEM

A control system, the purpose of which is to control the dis-
placement, or the velocity, or the acceleration, or any higher time
derivative of the position of the controlled device.
(It should be noted that forces and torques are involved in the
above definition).

SERVO-MECHANISM
An automatic monitored kinetic control system which includes
a power amplifier in the main forward path. (Includes continu-
ous, discontinuous, on-off, multi-step, etc. actions.)

POSITION SYSTEMS
The control of position (displacement) in a system (linear or
angular).

POSITION CONTROL SERVO-MECHANISM (dc)

Consider the electrical example of Fig. 14.1 in which the
output servo-motor shaft is required to follow the input shaft. As
long as there is a difference of angular position between these
shafts, measured by toroidal potentiometers, a difference of
potential will cause current to flow in the required direction
through the amplifier and servo-motor shunt field. The armature
of the servo-motor carries a constant voltage (stability with large
ballast resistor R) and hence a torque and motor shaft rotation
occurs as soon as the field is excited.
240

R
REED'S INSTRUMENTATION AND CONTROL SYSTEMS

1
I

SYSTEM RESPONSE
KINETIC CONTROL SYSTEMS

The effects of energy transfer including inertia, friction, etc.,

need to be considered whether mechanical or electrical, etc.
241

Descriptive analysis is often presented electrically for any system

due to ease of diagram circuitry and the mechanical, etc., equi-
valents can then be simply derived.
Consider a step input to the servo-mechanism of Fig. 14.1. At
the instant of applying input, maximum deviation exists, and the
servo motor first accelerates rapidly; when deviation ceases the
motor stops. This is an ideal situation because in practice inertia
of the motor causes overshoot, reverse current and rotation, with
Fig. 14.1 POSITION CONTROL SERVO-MECHANISM (de)
oscillation. Such oscillation would continue but for the frictional
effects, static and viscous, which damp out oscillations.

(A constantly excited motor series field can additionally be /

/

connected across the mains if the inertia of the servo-motor is ,y....//

~/
high.) As an alternative shunt field current can be arranged to be 0·,//
constant and armature voltage varied. The system as considered //2
/
however has the true relationship that torque is proportional to /
actuating signal and is generally independent of rotational speed.
The principle is effectively used in practical electrical ship steer-
ro
<D
r
<Do /
/
/
/
;!l

/
ing gears (Ward-Leonard system). Reference should also be /
/

made to Fig. 14.2 the simple block diagram of the system. Whilst /
/
/
a detailed analysis of the dynamics of such a control system is /
/
given later, at this stage it is clear that output torque is propor- /
/

tional to deviation, i.e. proportional action. Field control is used t_ t ------+

for small powers and armature control for higher powers, the
STEP INPUT HAM? JNPUI
latter utilising a series ballast resistor (dc) or rectifier from ac
supply. Fig. 14.3 SYSTEM RESPONSE

Referring to Fig. 14.3 for the step input 8i , curve one repres-
MONITORING FEEDBACK ents the undamped oscillation (of natural frequency Wn)
response at output 80 • Curve two represents light damping
(damping factor k < Wn), curve three critical damping (minimum
~ t<-..-/\ ei-eo I A l-J SE~V~- I I eo time without oscillation to equilibrium k = Wn) and curve four
heavy damping, i.e. aperiodic (k > Wn). As a first assumption
viscous friction is assumed to account for all frictional effects,
Fig.14.2 BLOCK DIAGRAM (ONE TERM POSITION) with resisting friction force (or torque) proportional to velocity.
242 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

Da!llping, to prevent overshoot, limits amplifier gain and speed

1
I
(
KINETIC CONTROL SYSTEMS

and reverse amplifier output will provide a braking torque to

243

response. bring the motor to rest. For a ramp function (linear displacement
Also given in Fig. 14.3 is output response 00 to ramp input Oi variation, constant velocity) velocity misalignment exists
for the three types of damped condition. 00 does not equal Oi in because with the input shaft stopped in position a zero input
the steady state, i.e. error with output lagging input. This is a voltage to the amplifier is required before shaft alignment. This
position lag (offset) due to velocity (rate) which is termed a requires an equal and opposite input voltage to the tachometer
velocity misalignment (friction proportional to velocity). A ramp output voltage available in the main feedback circuit, to balance.
input of position equals a step input of velocity. Motor and tachometer unit is often called a velodyne.

OVERSHOOT
This occurs with proportional control and is due to inertia
effects. It can be reduced by any of three methods:

Friction damping
Friction is non-linear, regarded in two components, static STABILISING FEEDBACK
r--:------ - - - - ,
(Coulomb) and viscous, the former giving steady state error and
usually neglected in simple analysis. Viscous friction is propor- ....-,
-JT2~~)(' \
- _1
," - A r-EJ--[J I
A M
r.: L,
T,YG
L '<.Y yrJ \(.2r
. +\1
tional to velocity and gives damping due to absorption of kinetic L1J l-_J
energy. Such damping, when utilised, is achieved by increasing
RESET FEEDFORWARD L
load torque as velocity is increased. Damping devices employing
viscous friction are not often used because of the following
disadvantages: increased response time to achieve steady state,
increased losses, increased energy input with a larger velocity Fig. 14.4 BLOCK DIAGRAM (THREE TERM, POSITION)
misalignment to produce this input. Viscous friction, and static
friction, are of course always naturally present to an extent.

Stabilising feedback Rate network

Modifying feedforward or feedback is to minimise any ten- From the above it is obvious that a voltage proportional to
dency to oscillate. In principle the object is to decrease servo- deviation plus a voltage proportional to rate of change of devia-
motor output torque as speed increases by tachogenerator feed- tion (i.e. velocity) is required, which is P + D action. The deriva-
back, which is preferable to viscous friction damping as the effect tive (rate) circuit has been covered in detail for process control. It
is more linear and no extra energy input is required. is indicated at D and is a phase advance whose voltage requires
Refer to Fig. 14.4 in which the relevant part of the loop is extra amplifier gain.
shown dotted, also shown is a gearing (G) and load system (L). For a step input, peak overshoot, settling time and rise time
Stabilising feedback from the tachogenerator (T 1), proportional tolerances may be specified as a given percentage of the step
to velocity, reduces the voltage input error signal, proportional change. With a ramp function it is necessary to specify allowable
to deviation, hence the amplifier input is reduced. No input will velocity misalignment (function of rate of input change) as a
exist before the shafts are aligned due to the tachogenerator given percentage of maximum input velocity (rate of ramp input
feedback voltage reduction. Inertia will move the output shaft displacement).
 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
KINETIC CONTROL SYSTEMS 245
244

Refer now to Fig. 14.5:

OFFSET
This has been discussed in process control and is a characteris-
tic of proportional control at different loads. Obviously in the
position system being discussed it is velocity misalignment. This
is a steady state position error due to viscous torque (propor-
tional to velocity) additive to load so that offset must occur. It can
be eliminated by one of two methods, the former preferential:

Reset network
A voltage proportional to deviation plus a voltage increasing
with time at a rate proportional to deviation is required, i.e. P + I Fig. 14.5 POSITION CONTROL SERVO-MECHANISM (ae)
action. This has been discussed previously and the reset network
The input shaft is connected to the single phase ac rotor of the
(I) is shown on Fig. 14.4. The capacitor stores enough delayed
transmission (T) and the output shaft to a similar rotor of the
charge to feed the amplifier to reduce deviation to zero against
controller (C) whose excitation is amplified (A) for supply to the
the extra loading of frictional effect. It can assist against initial
main drive servo-motor (M) to bring the shafts into alignment.
inertia and friction loading. Three phase stators of transmitter and controller are directly
connected. A null point of relative rotor positions exists, varia-
tion from which gives a proportional voltage in one phase or the
other. This is not a synchro transmission link because ~he output
Reset feedforward
A second identical tachogenerator (T 2) is driven off the input
shaft rotor is not mains excited.
The two transducers are three phase induction motors (as
shaft and supplies a feedforward additive voltage proportional to
synchros-magslips) and usually the main output servo-motor is
speed. In the steady state for a ramp function the feedforward
two phase (one fixed, one control).
from T 2 balances the feedback from T 1• hence no velocity mis-
Stabilising feedback can be used with an ac tachogenerator.
alignment, offset removed. When the input shaft stops (T 2 zero
The stator is wound with an input reference field which acts as
output) the following output shaft rotation provides the usual
excitation and an input field wound at right angles. Rotor cutting
stabilising feedback rate action from T 1 to reduce overshoot.
of reference field induces an emf in the output field proportional
to speed and in phase and frequency with input signal.

HYDRAULIC POSITION CONTROL SERVO-MECHANISM

POSITION CONTROL SERVO-MECHANISM (ac)
Generally limited to small powers and although cheap have a Hydraulic position control has many applications. A typical
example would be an electro-hydraulic steering gear. The con-
low start torque and relatively low performance. Essentially two
trol function acts to vary plunger travel, often by radially-
categories of ac system exist, i.e. demodulator-modulator and all
displaced or "swash plate" piston operated devices, and so give a
ac. The principles involved in the former have been discussed in
variable delivery pump. Linear or angular operation to any posi-
Chapter 7, with application to electronic controllers in Chapters
tion control system, particularly hydraulic, is easily arranged.
11 and 12.
246 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
KINETIC CONTROL SYSTEMS 247

SPEED SYSTEMS A speed set voltage from the input potentiometer has an input
The control of speed (velocity), linear or angular, in a system. volts signal VI- Monitored feedback from the tachogenerator
Most of the principles of speed control are applicable to posi- gives a voltage (Vo) proportional to output shaft speed. The error
tion control which has been covered so that only a brief analysis is signal e = VI - Vo is amplified and fed to main drive motor.
required. A derivative (rate) and integral (reset) circuit could be added.
Rate of velocity change is acceleration. Gyros measure accelera-
SPEED CONTROL SERVO-MECHANISMS (dc) tion in many control systems.
The essential equations relating to dc motors are:
SPEED CONTROL (ac)
f oc <t>la The actual control unit requires modulation, demodulation
where T is output torque, <t> flux and fa armature current. and compensation networks, as discussed previously, and
Noc V - laRa devices are readily available utilising ac amplifiers. However the
<t> servo-motor itself poses real difficulties as the torque-speed
where N is speed, V applied voltage and R a armature resistance. characteristic is non-linear. High resistance rotors and thyristor
circuitry gives improvements, with higher power induction and
P oc VIa - f~Ra synchronous motors, but has disadvantages. In general, except
where P is output power. for low power position systems and ac amplifier control units, the
Obviously speed control can be effected by varying armature all ac system is not greatly used as yet. However a mixed ac-dc
voltage or field flux, the former being shown on Fig. 14.6. arrangement for speed control has many advantages. The power
amplification-conversion device utilised to what is essentially a
dc drive system from ac supply is an important consideration.
Vacuum valves (thermionic) are limited to very small powers.
Magnetic amplifiers, with main and field solid state rectifiers, are
reliable and can be used directly with three phase for larger
powers or as field controllers in a motor generator set. The motor
v, generator set of the Ward-Leonard type is often used for very
-,... high powers but is a large unit, subject to time delays in opera-
e =v,-vo -_J tion.

WARD-LEONARD SPEED CONTROL

Consider the arrangement shown diagrammatically in Fig.
14.7:

Vo Terminology and principle of operation should be clear from

the sketch. The input motor (DM) is a constant speed induction
motor with standard dc generator (G) and servo-motor (M), the
Fig. 14.6 SPEED CONTROL SERVO-MECHANISM (de)
latter with a constant excitation field. Generated supply voltage
to the servo-motor depends on error signal between desired and
measured speed values, as proportional signal voltages, which is
248 REED'S INSTRUMENTATION AND CONTROL SYSTEMS KINETIC CONTROL SYSTEMS 249

and A on the other cycle. Full wave rectification with thyristor

trigger control on each cycle. Adequate overload protection is
required in the bridge circuit.

Fig.14.7 WARD-LEONARD SPEED CONTROL

amplified and fed to the generator control field. For small powers
a valve or transistor amplifier could be used with extension to
medium powers with a magnetic amplifier. High gain and stabil-
ity from drift can be obtained utilising such ac amplifiers suitably Fig. 14.8 THYRISTOR SPEED CONTROL
modulated and demodulated. For high powers it is usually neces-
sary to replace these amplifiers with either a rotating amplifier-
auxiliary generator (exciter) or amplidyne (metadyne, cross field
dc generator) - or modern thyristor control amplifier.
Controlled rectifier units can also be used. The mercury arc
GOVERNOR SYSTEMS
rectifier and ignitron can be used for large powers but are gener-
ally limited to specialist applications. Thyratron devices, with Many engines, turbine and reciprocating, are still fitted with
transformer coupling to anode and grid, have been successfully t'''1 trip devices to allow full energy supply under normal conditions.
used. Thyristors are being increasingly used either directly or as If revolutions rise about 5% above normal the energy supply is
field control devices. cut off until normal conditions are restored - lock out occurs at
about 15% excess which can only be unlocked by hand. Aspinall
THYRISTOR SPEED CONTROL types come into this category.
Consider the arrangement shown in Fig. 14.8: Smaller engines, such as electric generator drive often use
centrifugal governors based on the Watt principle, the Hartnell
The sketch has the familiar layout used previously, field con- governor is typical. Control is essentially proportional action
trol is by thyristor. The bridge circuit consists of two rectifiers C with sensed output (rotation speed) controlling energy input and
and D and two controlled rectifiers (thyristors) at A and B. The offset (exemplified by no load to full load speed droop) occurs.
gate of each thyristor is triggered by pulses from P representative Modern engine governors are isochronous devices but reset
of error speed input signal. Alternating current supply is rectified action is applied to eliminate hunting. One such P + I governor
for output to field by passage through Band C on one cycle and D has already been described in Chapter 12.
250 REED'S INSTRUMENTATION AND CONTROL SYSTEMS KINETIC CONTROL SYSTEMS 251

Two designs will now be considered namely Mechanical- Refer to Fig. 14.9:
Hydraulic (similar in principle to Fig. 12.2) and Electrical- With the engine running at constant speed under a steady load
Hydraulic (utilising principles discussed previously in this Chap- the up-force due to centrifugal force from the flyweights is bal-
ter). anced by the down-force of the speeder spring. The plunger is
central with the regulating plug covering the regulating ports in
MECHANICAL-HYDRAULIC SPEED CONTROL SERVO-MECHANISM
the sleeve. The plunger moves vertically but does not rotate and
Many engine units employ these servo devices incorporating
the opposite applies to the sleeve. The power piston is stationary
in-built safety for such as oil or water failure with IC engines. The
and the buffer piston central under these conditions.
design now considered is similar in principle to that of Fig. 12.2.
Consider a load increase on the engine for which condition Fig.
14.9 is applicable. Speed reduces and the plunger moves down
with pressure oil flow to the right of the buffer piston, which
moves left. The power piston will move up and admit more fuel
to the engine. Pressure oils also acts on the compensating piston
under side which will exceed the pressure on this piston upper
side so that the plunger will be restored up. The power piston will
now stop. As engine speed returns to normal oil is leaking
through the needle valve, to restore equal pressures on each side
of the buffer piston and compensating piston. The buffer piston is
returned to mid-position by the springs. This gradually reduces
the up-force on the compensating piston but the increasing
engine speed is also increasing this up-force due to centrifugal
force. The compensating piston is designed to be balanced
gradually so that the rate of leakage at the needle valve (unload-
ACCUMULATORS ing) equals the rate of loading due to extra centrifugal force

-,! 1= .,-~ ~
tlRE~L:E~ caused by higher engine revolutions. The engine will now run at
normal speed but with increased load and higher fuel setting. The

iG
....""
1_-

PLUG
-,
t
L

REGULATING'"
I
. . ..... -

I
TO Il
• SUMp
I +
needle valve should be screwed in sufficiently to prevent hunting
but without making the operation sluggish. Oil pressure is shown
dotted with the slightly lower pressure chain dotted.
The bypass arrangement ensures that for a large speed change
TO 1
SUMP r--'-- 1 the power piston only moves as far as the bypass. Pressure oil
I
I I flows directly to the power cylinder without further increasing
BUFFER !
!
TO FUEL
PISTON
I I
the pressure differential on the compensating piston. After suffi-
cient governor movement and speed return to near normal the
CONTROL
differential pressure acts as usual.
In the event of a large load decrease the power piston is at fuel
minimum and blocks the needle valve connection. This gives a
higher speed setting than normal and reduces a tendency to
Fig.14.9 MECHANICAL-HYDRAULIC SPEED CONTROL under speed.
SERVO-MECHANISM
252 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

ELECTRICAL HYDRAULIC SPEED CONTROL SERVO-MECHANISM

, KINETIC CONTROL SYSTEMS 253

Rotational speed is sensed by tachogenerator, or as shown in

Fig. 14.10 by ac alternator, with frequency pulses converted in
the rectifier to dc voltage proportional to speed. Set value is
applied to the controller and the two input voltages, opposite in
polarity, are compared. Error signal (if any) is amplified, con- TEST EXAMPLES 14
verted to hydraulic signal, and operates to alter energy supply
through a servo-motor. To reduce hunting, and offset, the con-
troller has reset (integral) action via a feedback loop. The unit is
also anticipatory i.e. load changes are fed back to the controller
to amend energy input before speed change occurs - speed
,I
1. Explain the difference between "open loop" and "closed
loop" systems of control.
Draw a circuit diagram for a system in which the Ward-Leonard
arrangement with feedback control is used to regulate the speed

~
control is virtually a fine trimming operation. of a dc motor and explain the mode of operation.
2. (a) Sketch a clearly-labelled circuit diagram for a simple
electrical remote position control servo-mechanism with zero
damping.
(b) With the aid of wave form sketches describe the action
of the system when subjected to a step input.
(c) Compare any advantages and disadvantages of the fol-
lowing methods of damping such a servo-mechanism:
i. Viscous friction,
ii. Output velocity feedback.
3. Describe a thyristor control arrangement for the speed con-
trol of a large electrical fan. Show how a zener diode can be used
I TO
ENERG to stabilise the voltage supply to the load and include the neces-
SUPPL'
sary protection to safeguard the diode against overload.
- - - -. ELECTRO- - . OIL SUPPLY
HYDRAULIC
SERVO
MOTOR
CONVERTER
-.OIL DRAIN
I:
Fig.14.10 ELECTRICAL-HYDRAULIC SPEED CONTROL
SERVO-MECHANISM
..
~

CHAPTER 15

CONTROL SYSTEM ANALYSIS

This subject is generally complex and the objective of this

chapter is to introduce the basic principles so as to allow an initial
appreciation which could be further developed, if required, at a
later stage. Consideration is given to the systems approach, the
order of linear systems, performance of systems, component
interaction and adjustment.
..

THE SYSTEMS APPROACH

SYSTEM
Capable of many and varied definitions. A general definition
could be: a functional assembly with components linked in an
organised way and affected by being within, and changed if
removed from, the boundary.
f May exist as sub-systems within a larger system so that a
hierarchy exists, for example the biological cell, within the heart,
of an animal, within a social human system, of a universe.

"I STATE
A system may be discrete, i.e. exist in one only clearly defined
state at a given time, or may be continuous in change. Can be
deterministic in operating to a fixed sequence, or probabilistic
with random change or subject to external influence. A closed
system always tends to seek equilibrium and ideally has no
energy transfer with surroundings outside the boundary whereas
256 REED'S INSTRUMENTATION AND CONTROL SYSTEMS CONTROL SYSTEM ANALYSIS 257

an open system tends to approach a steady state of balance with general systems theory and philosophy based on the techniques
the surrounding environment. The black box philosophy is appl- outlined in principle above.
icable, internal form unknown and sealed, with the only factors
of interest being the output and input variables and their rela- SYSTEM ORDER
tion.
Systems considered in this context are linear, i.e. equations
SYSTEMS APPROACH
with constant coefficients. In practice non-linearities exist but
There are barriers, sometimes artificial, between specialist unless complicated theory is utilised are difficult to analyse.
subjects which are in many cases being eroded by the unifying Many cases exist where the effect of non-linearity can be neglig-
theme of technology now being adopted through a systems ible by correct design so that linear theory can be applied.
approach to problem analysis and solution. Mathematical mod- A systems approach requires a unification between similar
els can be used to describe a variety of systems ranging from quantities and should result in a generalised mathematical model
mechanical control to organisational man management. whose equations are applicable for simulation and evaluation.
Analogies between electrical and mechanical systems can easily
ANALOGUES
be illustrated but the principles established are often more
widely ranging. The aspect is too broad for detailed considera- It should be remembered that rate of change of a variable with
tion at this stage but the technique of moving from the particular respect to time can be written dldt in calculus notation. Thus, for
(say engineering), by analogue and commonality, to the general example, translational velocity (rate of change of translational
system is a desirable aim. It is therefore usual to demonstrate a displacement with respect to time) can be written dildt; an
systems approach to an engineering situation and use this as a alternative is .i. Similarly, for example, rotational (angular)
" acceleration is d~ldt; or cii; and as this is the second rate deriva-
vehicle to suggest that this systems approach applies throughout
and a general theory may be attainable. tive of rotational displacement it also equals d 2 (Jldt 2 ; or 8.
Essentially the systems approach can be detailed as: Numerous other examples can be quoted.
Variables in common use include force, torque, voltage (drop),
1. Specify aims and objectives of the problem or analysis.
pressure (drop), displacement, velocity, acceleration, current,
2. Establish system and sub-system boundaries. flow rate, etc. Parameters include stiffness, damping coefficient,
3. Devise functional conceptual models of the problem lead- mass, inertia, resistance, capacitance, inductance, etc (see Table
ing to block diagrams with attendant mathematical models ,~

15.1).
(equations) allowing for interaction between component units, Consider the following:
feedback analyses, etc. ..
4. Scale system variables and construct analogue and circuit Generalised Impedance (Z) = ~cross Vari~ble (X)
diagrams.
5. After evaluation and iteration the synthesis can be estab-
lished and tested for final appraisal. Through variables are such as velocity, current and flow rate;
(The reader is strongly aclvise-ifat-this stage to consider Fig. across variables are such as force, voltage and pressure; impe-
16.1A of the next chapter for a systems approach applied to a dance parameters are such as inertia, resistance and capacity.
very basic engineering mechanism, Le. the simple pendu~um and ~;.;
Elements in a system can be classified as dissipative, where X OC'Y
its analogue.) Development to transport, banking, education, (such as resistors); as storage, whereXoc Y (such as capacity); or
manufacturing, community systems, etc., is an essential part of as storage, where X oc'Y (such as inductance). Table 15.1
258 REED'S INSTRUMENTATION AND CONTROL SYSTEMS CONTROL SYSTEM ANALYSIS 259

compares the translational mechanical, electrical and fluid sys- I R

tems. Extension to pneumatic, thermal, and rotational mechani-
cal systems gives a similar result in every case. B is damping
coefficient, S' spring stiffness, the dot above variables indicating
v
rate of change, e.g. will be acceleration. 4--Bv
--,
l
I
i
I
I
I--~

I!.J
F
E3Jv o

MECH ElEC
Fig. 15.1 FIRST ORDER SYSTEMS
x X X
Gen. X Y z=-
y
z=-
y
z=-:-y
F F The constant applied force (F) is resisted by the spring force
Mech. Force Velocity DamperB= - Sprmg
. S' =-
F Massm =-:- (S' x) and the dashpot damping force proportional to velocity
v v v
(Bv). Now v = dxldt = Dx so that the equation is written:
. V . 1 V V
Elec. Voltage Current ReSistor R = - Capacitor -= - Inductor L = -:-
F= S'x+ BDx
I C I 1
I . and after re-arrangement this becomes
Fluid Pressure Flow Resistive R' =.e Capacity - =.!!.... Inertive l' = ~
f C' f f B
-Dx+x=-
F
.., S' S'
Table 15.1
which is a typical first order equation, time constant T = ;,

SYSTEM ORDER For the electrical resistance-capacity system shown:

The response of any system, or a component within the system, TDVo + Vo = Vi
can be described by a mathematical equation. The order of the
equation, and hence the system or component, is fixed by the
(in charge terms, TDQ + Q = ViC when DQ is current)
highest derivative "power". In a mechanical system, for exam- where T = RC is a time constant.
ple, velocity is the first derivative of displacement, i.e. v = dxldt For a heat resistance-capacity system:
,-;;"

so that a system with such a term as the highest derivative is TDOE + OE = OF

classified as a first order equation. All such first order systems are
defined by having one form of energy storage component. Simi- • where T = RC, fh and OF element and fluid temperatures.
larly acceleration as the second derivative of displacement, i.e. a For a fluid restrictor-capacity tank system:
= d 2x Idt 2 is a second order. All such second order systems are TDpo+ po =Pi
classified by having two forms of energy storage components and
one form of dissipative energy component. It is often convenient
(in quantity terms, TDq +q= PiC' where Dq is flow)
to write D for dldt in calculus notation (note D- I is a first Such equations can be extended to economic, management,
integration ). etc. systems and a general equation arising from the above
analogous cases may be written down as follows (y any variable):
FIRST ORDER SYSTEMS
Consider the mechanical translation system of Fig. 15.1: .. TDy +Y = bf(t)
 CONTROL SYSTEM ANALYSIS 261
260 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

CLOSED LOOP TRANSFER FUNCTION

where T is the system time constant and b is a constant. It is Figure 15.2 illustrates the usual arrangement. The transfer
possible that the input may not be a step or ramp form of function of forward elements is G and of feedback elements F.
constant, such as a dc voltage, but may be varying with time, such
as an ac voltage. The f(t) forcing function term is a general way of
writing an input function dependent on time, for example
sinusoidally, to allow for such variations. If the input is not so
F
varying the right hand side of the equation is a constant, as
covered in the analogous cases given above, wheny is made up of
a constant and a variable, with a system time constant in the
solution i.e. y = b (1 - e- t1r ). 60
G

TRANSFER FUNCTION
Fig.15.2 CLOSED LOOP TRANSFER FUNCTION
The transfer function of an element is the ratio of its output
signal to its input signal.

Transfer Function = ~
fh 1. Consider the loop of feedback to be opened:
Its use, together with block diagrams, simplifies analysis using 11

s-plane in place of differential equations. In some cases output is Open Loop Transfer Function (G) 7= ~
merely amplified or attenuated input, e.g. gearbox, whilst in " for forward path elements.
other cases the signals may be in different physical form with
different amplitude and phase. 2. Consider closure of the feedback loop and F = 1 (unity
Consider the RC network of Fig. 15.1: feedback):
. 00 00
Vi = IR + Vo Closed Loop Transfer FunctIOn = Oi = 0 +0 0

dVo
1= C - = CDVo 00 10 _ G
dt /1
1 + 0010 - 1 + G
Vi = RCDVo + Vo
.. 3. Consider the system as sketched:

(this is the first order equation given previously) f= FO o; 0 = Oi - f; 00 = GOo

combining these equations gives:
OolG = Oi - FOo
Vi = Vo(RCD + 1)
Vo 1 Oi= 00 ( 1 +GFG)
Transfer Function = Vi = 1 + T D i
. 00 G
Closed Loop Transfer FunctIOn = e; = 1 + FG
This result will be characteristic of all such first order equations.
262 REED'S INSTRUMENTATION AND CONTROL SYSTEMS CONTROL SYSTEM ANALYSIS 263

1. The open loop transfer function is very useful in stability 1. RCL series electrical network.
testing (see later).
Applied volts = Resistor volts + Capacitor volts
2. Note that for unity feedback the closed loop transfer func-
+ Inductor volts
tion is (open loop transfer function) divided by (one plus open
loop transfer function). Increasing gain G reduces offset.
3. This is a simple example. Practical cases with more involved
V=lR+- If
C
·df
Idt+L-
dt
transfer functions are more complicated and difficult to solve. 2

The response output 80 will depend on the form of the input

dV
_ = R _dl +-f + L d- 12 (f
a ter d'ffi ..)
I erentIatIon
dt dt C dt
signal. For elements in series the overall transfer function is the
product of the individual transfer functions, assuming there is no '.I 2. Mechanical translational damper, spring, mass assembly.
interaction between the components. There will be an overall (See Fig. 15.10).
gain (or attenuation) and if input is a variable with time there will Applied force = Damping force + Spring force + Inertia force
be a need to algebraically add phase shifts. For a given open loop
transfer function there is only one closed loop transfer function,
hence open loop analysis gives closed loop analysis automati- F= Bv + S'x + ma (Bdampingcoefficient)
cally. This fact is very important as an unstable closed system
dx
cannot be measured but by opening the system an analysis is F= B
dt
+ S'x + m dd2X
t 2
(S' spring stiffness)
possible allowing stability compensation to be made.
If the system under consideration was itself a component of a 2
d x + 2k dx + W 2X = !-
control system it could be represented as one block diagram dt2 dt n m
enclosing its transfer function. Such a second order system is
shown in Fig. 15.3. 2k = !!. where k is the damping factor
m

w~ = -
S' where Wn is the natural (undamped) frequency
m
F
D 2x + 2cwn Dx + w~x = m-
ej
•
I 1
1+2 c 1,0+1,2 0 2
I eo
• c = -k h
were 'd . .
C IS a ampmg ratio
Wn

Fig.15.3 TRANSFER FUNCTION BLOCK DIAGRAM

(constant changing is purely for mathematical convenience, m is

the effective mass of the spring and its load).
3. Position control servo-mechanism.
SECOND ORDER SYSTEMS Reference should be made to Fig. 14.1, Chapter 14, showing
Such systems have analogues as detailed previously. A review this unit. Output drive torque is dependent on inertia, viscous
in depth is not required but three examples are given: and stiffness (static and load error) torques.
264 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

KIK2K3((Ji - (JO) = Ia

(I is moment of inertia, H is damping coefficient)

+ Hw I CONTROL SYSTEM ANALYSIS

SYSTEM PERFORMANCE
265

I H
- - - D2(JO
K 1 K2 K3
+K K2 K3
D(Jo + (Jo = (Ji
Ideal response is where output is identical to input command.
1

(Inertia torque is for motor and load, direct drive assumed, This cannot be obtained because viscous friction and measure
viscous torque proportional to angular velocity; K 1 , K 2 , K 3 are delays result in output lag. Effects such as inertia produce oscilla-
respectively potentiometer bridge, motor torque-current and tion about the steady state. The mathematical approach gives an
amplifier scaling constants.) Simplifying constants gives: estimate of likely performance which can be improved by
(D2 + 2cwn D + w~)8o = W~(Ji experimentation rigs, usually electrical analogues. The objective
is attainment of desired value fairly quickly and accurately, with
1 stability. Response to simple input signals has already been
Transfer function = 2 T D .
T 1 D2 + C 1 +
2
covered in the text but a quick resume can now be presented.
where T 1 = I/wn and is periodic time of undamped natural oscilla- Performance results are in two distinct parts:
tion divided by 21T. For sinusoidal input the transfer function is as 1. Transient response when the system is responding and may
above but with fw replacing D, i.e. if (J = Ae iwt then D (J = iw(J; hunt.
both D and i are operators. 2. Steady state response when the transient has died away.
It is necessary to introduce mathematical solutions to equa-
Note: tions, especially second order, in the following sections. Such
A general second order equation, allowing for input variation solutions can be readily verified if required but the object in
with time is:
presenting the work is purely to introduce techniques and a
(D2 + 2cwn D + w~)y = a(.rJ~f(t) general appreciation is all that is required.
where a is a constant,y the variable andf(t) the forcing function. STEP INPUT RESPONSE
For a first order system
HIGHER ORDER SYSTEMS T Dy + Y = bf(t) is the general equation
Commonly arise but are not considered here except to note
three term control action (P + I + D), action factors K b K 2 , K 3 •
T D(Jo + (Jo = (Ji in control terminology

V = - Kl ( (J + -K2
Kl
f (Jdt -3-
+ KKl d(J)
dt
The steady state is a constant step input (Ji. The transient solution
is (Jo = -(Jie- tIT . The complete solution is (Jo = (Ji(1 - e-tI'T). To
"slow down" such a system increase resistance, damping or
=- Kl (1 + SiD + TD) (J capacity but decrease stiffness.
Reference to Fig. 15.4 shows an exponential curve. The time
S integral action time, T derivative action time, V controller constant 'T is the time to reach steady state if the initial slope was
output. Application to say a position control servo-mechanism maintained. Output actually only reaches 63·2% of this value.
gives a transfer function of a third order character obtained by There is no overshoot. The curve is characteristic of the LR
equating the system second order equation in terms of (Jo to the electrical "growth" response. The CR electrical circuit gives a
above equation in which (J = (Ji - (Jo. decay characteristic.
266 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

, CONTROL SYSTEM ANALYSIS

For a second order system:

Reference to Fig. 14.3 of Chapter 14 illustrates response for
underdamped (c < 1 curve 2), critical damping (c = 1 curve 3)
267

8i at t = 0 and overdamped (c > 1 curve 4). The inclined line of steady state
- - - /1-- - - - - - - - - - - (Jo, as appropriate to the degree of damping, will start at 2cT 1 on

r
8j
II
/ I

/
/
\
I
_!.
8o=8j(1-e r) the horizontal axis (see also Fig. 9.2 of Chapter 9).
/
/
/,
0'6328 j
SINUSOIDAL INPUT RESPONSE
This is the case with the forcing function related to time, i.e.
f(t) such as (Jj sin wt or (Jj cos wt, for sayan alternating voltage
i. T .i t-
input; first or second order equation. The system is subjected to
Fig. 15.4 STEP RESPONSE (FIRST ORDER)
varied frequency sine wave inputs and response output is noted
in magnitude and phase. Such analysis is also useful for evalua-
tion of higher order systems.
The first order solution to TDO o + 00= Oi cos wt results in one
For a second order system: equation (magnitude) and another equation (phase).
(D2+ 2cwn D + w;)y = aw'ftf(t) general
(D2 + 2cwn D + w'ft)Oo = w'ftOi control 00 I
Again the steady state is a constant step input Oi' (after dividing 0; - VI + w2Tr
through by w£). The transient solution is obtained by setting the tan¢> = WTl
left hand side to zero and three resulting equations are possible
depending on roots obtained. Reference should be made to Fig.
14.3 of Chapter 14 where curves representing the three solution Consider as example the RC network covered previously (Fig.
equations are shown. Curve 2 (c < 1) is one solution for under- 15.1) with the object of obtaining the steady state sinusoidal
damped and is oscillatory. Curve 3 (c = 1) is another solution for response. The method, with sinusoidal type inputs, is to replace
critical damping. Curve 4 (c > 1) is the third possible solution for operator D by iw using the complex number notation which is
overdamped (aperiodic). The oscillatory case illustrates over- very useful in ac networks or polar control plots. Symbol i (some-
shoot, a settling time, etc. For the ~lectrical (series) system times j) is an operator which rotates a vector by 90° in an
critical damping occurs when R = 2-vYfC, for mechanical (trans- anticlockwise direction, without altering its length, its numerical
lational) when B = 2ysni. value is.y=I; i 2 is 180°, numerically -1; i 3 is 270°, numerically
--y-CI, i.e. -i; i 4 is 360°, or 0°, i.e. numerically 1. This is illus-
RAMP INPUT RESPONSE trated by an Argand Diagram of Fig. 15.5 which shows a vector A
For a first order system: whose modulus, or amplitude (length), is 5, i.e. V3 2 + 4 2 and
The complete solution is 00 = w[t - T(1 - e- t / T )]. Reference arguement (phase angle) is tan- 1 4/3. The vertical (Y) axis is
to Fig. 14.3 of Chapter 14 would indicate exponential approach referred to as imaginary and the horizontal (X) axis as real and
to an inclined line representing steady state 00 , similar to curve the number composed of real and imaginary parts is called a
3 (or 4), starting at T on the horizontal axis (see Fig. 9.2, Chapter 9). complex number. The diagram is often called the "s-plane".
 268 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

y
iA
, CONTROL SYSTEM ANALySIS

The second order solution is more complicated but shows the

same characteristic result of a response of different magnitude
269

and phase to input. Solutions are best illustrated graphically, Fig.

15.6 (also Fig. 9.3). Measurement is by a cathode ray oscillo-
scope or transfer function analyser. Resonant frequency is
x 1/27TyILC electrical (series) and 1/27TVS[m mechanical (transla-
-A A tional).

FREQUENCY RESPONSE ANALYSIS

Referring to Fig. 15.6; the dotted line shows first order
response and the full lines are second order response curves for
-iA

-------
Fig.15.5 ARGAND DIAGRAM

From the circuit of Fig. 15.1: lol - ...........-c=0·05 -~

c=03
Vi=~
1=
Z R - ijwC t c=1·0

(Capacitive reactance (Xc) is i/wC where w = 27Tf, impedance

Z.) (-i indicates 270°, i.e. a current leading voltage, capacitive
effect.)
Ul-
Va = (-i/wC)Vi
R-UjwC) c=0'05
-¢
The next step is to evaluate the transfer function:
~ 90° ........ ~ ......... _---
Va _ -i/wC IjiwC
Vi -- R - i/wC R +
l/iwC

I
I
+ iwCR - I
1
+ iwr
J 180°
_ _ _ _I '- --
I
7?r --

Fig.15.6 FREQUENCY RESPONS.E CURVES

This is the typical first order transfer function with D replaced by
iw. This expression gives the modulus, which is termed mag-
nitude ratio (M) in control terms, and argument (phase shift). various degrees of damping. In the former case magnitude ratio
(M the ratio between output and input amplitude) decreases
M= 1 steadily with input frequency and phase lag (c/» increases, due to
V1 + w
2 2 2
r the output being unable to follow input. In the latter case results
c/> = tan- wr 1 are similar with a high resonance peak at Wn for low damping
which reduces with increased damping, as does w~. Lag is inher-
phase of Vo relative to Vi. ent in control systems. A frequency response diagram of M'""-' c/>
270 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

with M a logarithmic axis or decibels (dB = 20 loglo iJo/fh), both

effectively the same, is used in open and closed loop analysis.
~ CONTROL SYSTEM ANALYSIS

The Nyquist stability criteria is that the closed loop system is

stable if the open loop frequency response locus traced from
271

Response diagrams as in Fig. 15.6, but with a logarithmicw (or w = 0 to w = 00 does not enclose (pass to the left of) the point
wT 1) axis and a logarithmic M axis (or dB), are similarly used. ( - 1, 0). The locus of 80 for w increasing can be seen to be stable
Such diagrams if used in open loop frequency response analysis from the diagram. The marginal state through (- 1,0) is shown,
are called Nichols and Bode diagrams respectively. with instability beyond that. The circle is for unity gain (M = 1).
Locus curves starting at various gain values on an extended axis
to the right (where w = 0), can be plotted for fixed gain with
STABILITY RESPONSE increasing values of w to determine correct gain for complete
The main aim of a frequency response test is to assess stability. l stability.
I
One common method is to open the feedback loop and inject a (Also shown is a vector sum fJ t = fJo + fJ being the input signal to

I
small sinusoidal constant magnitude input signal (fJ) to the for- give fJ = 1; this gives indication of closed loop response but is not
ward path elements only and obtain a polar plot of this open loop part of the normal Nyquist criteria.)
frequency response. The input is usually made unity and the If fJo = 1 and fjJ = 180° for a particular frequency w, i.e. tip of 00
polar plot is a Nyquist diagram obtained by measuring magnitude locus vector is point (- 1, 0) then fJi to a closed loop system is
:Ii
ratio and phase angle of output for increasing values of frequency zero. This is unity feed back (positive) and slight change of fJ i will
from zero to infinity. M and fjJ could of course be calculated but cause oscillation which would grow with increased amplitude
this is obviously pointless, at least at this stage. A typical Nyquist oscillations for open loop gain over unity. This is the basis of the
diagram (open loop polar plot) is shown in Fig. 15.7. Nyquist criteria.
Phase margin (a), as shown, should exceed 30° (usual range
30°-60°) and gain margin (sometimes expressed in dB) should
exceed 0·3 (usual range 0·3 to 0,6) to ensure stability. The curve
of Fig. 15.7 is a typical fourth order.
Illustrative response curves are shown in Fig. 15.8 (M = 1);a is
typical of a first order system, b second order, c is combination of
a and b (third order effectively). Curve a is typical of a passive LR
circuit response and is a 4th quadrant semicircle with output
lagging input and amplitude decreasing as w increases. (A pas-
sive CR circuit is plotted in the 1st quadrant with output leading
input and amplitude increasing as w increases.) The first and
second order systems sketched can not be made unstable with an
increase in gain factor but the third order system can (as can the
fourth order of Fig. 15.7). In such a case feedback from a
tachogenerator would induce oscillation and increasing amp-
litude, i.e. instability. This can be prevented by reducing the gain,
which however reduces accuracy. Other methods of stabilisation
include adding a passive phase lag network (P+ I) or phase lead
network (P + D) or a stabilising feedback CR circuit with the
object of increasing stability without reducing gain. Curve d

'cl!'f;i'i,'4 ,..,~..:.i (., ~

272 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

moved to curve e of Fig. 15.8 illustrates the objective. It should

be noted that good stability and high accuracy are incompatible
~
J
24.
CONTROL SYSTEM ANALYSIS

i i i
b c
i I
273

and a compromise between the two is desired for best perfor-

mance. 18~""""'--- I I ---"""'l Y "' I ~ I

z
<i
~

!l..
o
g
z
W
!l..
o

Fig. 15.8 RESPONSE CURVES (NYQUIST)

-1ao -150 -120 -90 -60

OPEN LOOP PHASE, degrees - - -

Fig. 15.9 NICHOLS CHART

Contours of open loop response (Figs. 15.7 and 15.8) can be On Fig. 15.9 curve a (unstable) shows 150 phase lead at 11 dB,
used to evaluate closed loop response. For any point on the open curve b (stable) 32 0 phase lag at -3 dB and curve c (improved
loop, vector values of output to input ratio and the angle between stability) 63 phase lag at - 8 dB, the operating frequency to be
0

them give points for the closed loop response. These can be near 1 rad/s.
plotted on another similar harmonic response diagram or on a The closed loop characteristic (direct feedback systems) can
frequency response diagram. be derived on this chart from the intercepts of the open loop
The Nichols chart (Fig. 15.9) is frequently used in frequency locus with the gain and phase contours. This requires a simple
response and stability analysis. Point plots of constant gain and calculation for values (which includes a feedback fraction).
phase (derived by calculation due to the logarithmic scale) give
contours on which open loop response can be plotted, minimum FURTHER ANALOGUES
contours are shown on the sketch to simplify the illustration. In some cases approaches are used in which electrical compo-
Note that 20 10g10 1 = 0 dB, 20 log 10 10 = 20 dB and 20 10g10 0·1 nents in parallel are regarded as equivalent to mechanical com-
= -20 10g10 10 = -20 dB for the dB plotting so that negative dB ponents in series, and vice versa. Certainly current, which has a
corresponds to amplitude ratio (gain) of less than 1 (attenua- common value through series electrical components, is analog-
tion). Magnitude ratio 0·7 to 0·4 (-3 to -8 dB) corresponds to ous to velocity, which has a common value through linked
gain margin 0·3 to 0·6 (3 to 8 dB) which with a phase margin mechanical translational components. Capacitors in parallel are
about 45 is a typical simple design specification.
0
equivalent to springs in series, and vice versa.
 CONTROL SYSTEM ANALYSIS 275
274 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

1
L

Vi C R Vo r---:- - FACTORY
- - -SYSTEM
- - -BOUNDARY
-,---..,
I FEEDBACK

Fj ~ J flsl
JVVVV~
1'=-0 SALES
MANAGEMENT I
CONTROL I
Fig. 15.10 SECOND ORDER SYSTEMS
I
I
I
I
L -.J
The lower diagram of Fig. 15.10 illustrates a mechanical transla-
Fig. 15.11 MANUFACTURING SYSTEM
tional damper, spring, mass system which has been described
previously (the left hand end is often fixed, F i = 0, F o = F). It is
typical of an anti-vibration mounting in which high frequency
input oscillations will be damped out (see analogue circuit diag-
ram, Fig. 16.1B).

The upper diagram of Fig. 15.10 is the equivalent electrical COMPONENT INTERACTION
circuit. High impedance to high frequency inputs results in a By this heading is meant the internal interference within a
filter system in which only low frequency components are controller of the interference effect between connected control-
applied to R. The ratio of V to Vi can be found at a given
0
lers.
frequency.
INTERNAL INTERACTION
Fig. 15.11 illustrates the systems approach to a factory man- The three term equation for a controller has been considered
agement system which should be self explanatory. Management and the equation in transfer function terms is:
includes finance, development, etc. sub-systems and techniques
include critical path analysis, O.R., quality control, 0 & M, v= - Kl (1 -I- s~-I- TD)O
queuing theory, etc.
276 REED'S INSTRUMENTATION AND CONTROL SYSTEMS CONTROL SYSTEM ANALYSIS 277

This assumes each control action can be generated separately ADJUSTING CONTROLLERS TO PLANT
which may not be possible, and resulting interaction fan occur For initial commissioning the controller must be set up exactly
which will affect output signal. Let consideration be applied to a to the manufacturers instructions and all maintenance must fol-
three term pneumatic controller-see Fig. 10.7 of Chapter 10. As low similarly from makers advice. Integral resistance is usually
stated the position of integral and derivative adjustment can set at maximum and derivative resistance at minimum. Propor-
affect output in various ways. Consider the derivative adjust- tional band can now be set for minimum stabilisation time.
ment placed at X on Fig. 10.8. The control output signal can be Derivative resistance can now be increased to reduce this time a
shown to be: . little, integral resistance now being adjusted to the same as
derivative resistance. There is a definite relation between TandS
v= - K, [ (I + 2:) + s~ +TD ] e settings; even with independent settings T can never exceed S.
Considering now setting and adjustment in more detail. This is
Integral or derivative adjustment affects controller gain (K 1)' a skilled operation requiring time and a knowledge of plant
In this case K 1 is altered by the interaction factor (1 + 2T/S). characteristics so that the following, for a P + 1+ D controller, is
Analysis of each controller design is required to establish the obviously a condensed simplification.
exact output signal. Similar remarks apply for electronic control- The object is to critically damp the signal to rest in the
lers. minimum time without overshoot and oscillation. Instability may
occur for too narrow a proportional band, too short integral
EXTERNAL INTERACTION action time or too long derivative action time. Stability with
Consider two first order controllers in series with no interac- underdamping gives oscillation with too long a stabilisation time;
tion between the stages. The overall transfer function becomes: overdamping gives no oscillation but too long stabilisation time
(Jo I usually due to a too wide proportional band, too large integral
e; (1 + nD) (1 + T2 D) action time or too short derivative action time.
For proportional action band only, it is best to start at say
which is a non oscillatory type of second order function. For a 200% bandwidth and move the dial away from and then back to
sinusoidal input the gain (attenuation) of individual elements are the set value, noting the settling time. This is repeated at step
multiplied and phase angles added algebraically, utilising as reductions of bandwidth until the oscillations do not reduce to
usual iw for D. zero (too much reduction would cause instability with increasing
The control units must be non interacting otherwise the trans- oscillations). A slight increase in bandwidth now gives the cor-
fer function of one controller will be modified by the loading of a rect value for minimum offset and stabilisation time.
following controller. This is usually avoided by inserting buffer For P + I the proportional band would be set as for Paction
amplifiers (or stages) of unit gain, without phase shift, between above with integral action time at maximum. The integral action
the controllers in series. time is then reduced (using big steps initially) until hunting
oscillation starts. A slight increase in integral action time now
gives the correct value and stability.
For P + D the proportional band is narrowed as above until
COMPONENT ADJUSTMENT
hunting is occurring and it is held at that value. The derivative
The adjustment of controllers, especially P + I + D, in a plant action time, which had been set at minimum, is increased to
is usually done empirically by generally well established experi- remove hunting. The proportional band is again narrowed
ence criteria. slightly and hunting removed by adding to the derivative action

l~rdRI1C:;f Ig:i:i ;;
 CONTROL SYSTEM ANALYSIS 279
278 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

TEST EXAMPLES 15
time. This process is continued until hunting cannot be removed
by the derivative action time. The proportional band is now
1. Sketch the harmonic response (Nyquist) diagram for fre-
widened slightly for correct setting.
quency response tests on:
P + I + D controllers may be adjusted in practice as for a P +
D controller, noting the derivative action time, add the same (a) a stable system,
integral action time then adjust for minimum offset. Interaction (b) a critically stable system,
is always rather a problem except in a well designed controller, (c) an unstable system.
well matched to plant characteristics. 2. A thermocouple at 10°C is placed in a fluid at a temperature
of 60°C and the reading after 4s is 40°C. Assuming exponential
EMPIRICAL SETIING METHOD (P + I + D) delay response evaluate the time constant of this instrument. If
I and D terms are reduced to zero and proportional band is the thermocouple were then used to measure a temperature
narrowed until continuous cycling occurs. This may require a rising steadily at 2°C per second what would be the steady-state
small step input on the desired value setting to start the oscilla- error of the reading? (4· 37s, 8·74 °C)
tion. Whilst continuous oscillation at constant amplitude is tak-
ing place the periodic time is measured. With this proportional 3. A step change of 2·5% is applied to the input of a P + I
band (W) and periodic time (To) it is possible to empirically set controller and the output gives a sudden step change of 5%and
the controller thus: after two minutes the total output change is 12·5%. Determine
proportional bandwidth and integral action time. A ramp change
% Bandwidth for P = 2W of 1% linearly is applied to the input of a P + D controller and the
% Bandwidth for (P + I) = 2.2 W output gives a sudden step change of 5% and after this the output
changes linearly at 3% per minute. Determine proportional
Sfor(P + I) = To minutes
1·2
bandwidth and derivative action time. (50%, 80s, 33%, 100s).

0/0 Bandwidth for (P + I + D) = 1·67 W

S for (P + I + D) = ~o minutes
T for (P + I + D) = To minutes
8
Signals are then trimmed for optimum perform. The above is
satisfactory for a continuous process but not for auto-start with
no overshoot.

--===~;~--"---~~=,-=....:-.,-"~~,":--=:~~-_.---=-",:
- --=:.=...-_~._.-

,;;5jli!!li.!!!JlOil!i';,~-' c
 CHAPTER 16

LOGIC AND COMPUTING

In this chapter we will be concerned with analogue computers,

switching logic circuits, digital computers, data processing and
computer control. Each section is obviously a specialist area of
study and it is only possible in this chapter to give an introduction
to each.

ANALOGUE COMPUTERS
Many of the principles involved have already been covered in
the immediately preceding chapters. It is now proposed to sum-
marise this information.

ANALOGUE
That which has correspondence or resemblance (analogous) to
something else which may be otherwise entirely different in
form. Analogies between resistance-damping, inductance-mass,
capacitor-spring, etc. have been considered. This is extremely
useful for simulation. Similar relationships exist for pneumatic,
thermal and fluid systems.

ANALOGUE COMPUTER
Essentially a device to represent continuous measures of phys-
ical quantities in numerical form. Provides concurrent, fast,
inherently graphical and reasonably accurate simulation and
investigation of mathematical models of dynamic systems. In
most cases electrical analogues are used with voltage signals
representing system variables.
282 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
LOGIC AND COMPUTING 283

Basic elements b. Evaluate the mathematic model equations, in this case:

These have all been considered previously in Chapters 10 and
11 utilising both pneumatic and electrical transducer analogues
of variables. Elements include the operational amplifier, feed- . 8 - m B' w
back, averaging, ratio, summer, scalar multiplier, inverter, m dw
dt = -mg sm
proportional-derivative-integral devices, flow, power, root
extraction components, etc. incorporated as required to form a d8
1. dt =w
simple analogue computer.

Control engineering
The application of basic elements to form control devices in
hence 8= f: w dt + 80 i.e. 8 = 80 at t = 0

process and kinetic systems is a direct application of small dw . 8 - B' w

2. -=-gsm
analogue computation. These aspects have been described pre- dt
viously.
hence w= -g lsin8dt-B'fwdt+wo
Data processing
The large analogue computer is used, certainly at input inter-
face, of data recording, display, scanning, alarm, etc. systems.
This aspect will be considered later in this chapter. c. Draw block diagram, starting with integrators, adding sca-
lars, etc. as required.

System analysis
d. Scale variables, draw analogue programme, draw circuit
The use of the analogue computer to solve mathematical equa-
tions for system analysis is a valuable engineering tool. Variables diagram (see Fig. 16.1A).
can be varied readily and simulated results quickly obtained.
Outputs may be conveniently recorded on pen chart devices, two e. Analyse by test for interaction, evaluation, synthesis;
axes plotters (X - Y), cathode ray oscilloscopes and ultra violet refine-iteration.
recorders. Obviously the mathematical analysis is usually On the block diagram the multipliers, summers, integrators,
involved but the selection of a simple analysis of a basic equation inverter will be noted. It should be remembered that there is a
should effectively illustrate principles which can be applied to sign change on the operational amplifier. Interconnection of
complex systems. components is termed patching.
On the scaled analogue programme (patching) diagram, some-
times called flow diagram, the unity ratio of amplifier resistance
Simple pendulum analogue simulation
is indicated by a 1. Voltages Vb V 2 , V 3 and V 4 are analogues
The procedure is an example of the systems approach to
proportional to 8, w, 80 and W o respectively. Potentiometers
engineering problems. The procedure could be analysed from
shown P. Adjustment of potentiometer tap ratio k and amplifier
objectives as follows:
gain K gives required ratio scaling for gain (F and G in feedback
a. Sketch the conceptual (physical) model- see Fig. 16.1A.
terms).

':<I'El;iiii c;
 284 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LOGIC AND COMPUTING 285

On the circuit diagram all factors should be clear from preced-

ing diagrams. ResistancesR 1 and R 2 are proportional to g and B'.
For a typical first order equation:

Dx = - ax + f( t)

~ mgp -w I J '<-' 1I
X = -fax-f( t )dt

~
-
AIR RESISTANCE
PER UNIT MASS
B' 0< W
ra 8----,
IJ
iNS
"'-../
~
I
I

L-__
I -
... _

INV
_ _I

I This is readily patched with two potentiometers (one input, one

I
mg I I feedback) and an inverter integrator with suitable scaling for
i
I _ -wolI constants in a similar way to that shown in this pendulum exam-
i
CONCEPTUAL MODEL
ple.
I For a typical second order diagram:
I
I (D2 + B'D + g)(J = 0 pendulum
r~ l (D2 + 2cwn D + w~)x = aw~f(t) general
L +- J
BLOCK DIAGRAM The approach is similar utilising, for example, three opera-
tional amplifiers in series (Fig. 16.1B). This example is an
analogue for an anti-vibration mounting involving linear
damper, spring, mass. The model diagram has been considered
previously (see Fig. 15.10) and the equation, linear simple har-
monic second order, was derived in Chapter 15 and is repeated in
a different form (pendulum and general case).
Net force (mass times acceleration) acting in the opposite
direction to displacement and velocity is given by:
d2x
m -2 = -S'x-Bv
dt
(where S' is spring stiffness and B is damping coefficient)
~ • eV2 V2

-ex _Ddx
2
dx =
ANALOGUE PROGRAMME CIRCUIT DIAGRAM dt 2 dt
Fig. 16.1A ANALOGUE SYSTEMS APPROACH
d2X
x = _E dx _ F
dt dt 2

!'!1l!=m-' mnflU~I:~!-,ii;f
286 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
LOGIC AND COMPUTING 287

Referring to Fig. 16.1B: LOGIC CIRCUITS

The components used will decide values of constants. Scale Logic is a systematic approach to problem solving and decision
factors have to be analysed relating voltage to variables - time making. Fig. 16.2 illustrates a simple logic flow chart analysis for
constants allow real time scale variations. Output must equal marketing a new product (yes decisions full lines, no decisions
input, so both are connected. An input force is necessary to dotted lines i.e. binary algorithm).
provide essential acceleration and for this analogue is usually by
square wave pulses from a signal generator. Output characteris-
tics are analysed on an oscilloscope. Analogue variations (dam-
ping/mass and spring force/mass) are arranged at potentiometer
and feedback resistor respectively. The circuit can be regarded
from input to output as integration, or differentiation in the other
direction. The last amplifier is a scale factor (which can also be
arranged to have adjustable feedback). ...------
It is now convenient to move from analogue (frequency
domain to digital (time domain) as used in logic circuits.
...-
.-

~r
I
I
Il-
l:::>
Q.. Ie: Fig. 16.2 LOGIC FLOW CHART (ALGORITHM)
~ 16
I
I
I
ov t
I
2
I
t
I
, +
-EQ!-F d x G dx -Hx x
dt dti dt

Fig.16.18 SHM ANALOGUE CIRCUIT DIAGRAM

(LINEAR VIBRATION DAMPER)

E:ii:::1
288 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LOGIC AND COMPUTING 289

BOOLEAN ALGEBRA The truth table is given (Table 16.2) where "true" is referred
Is the algebra of logic. Has tended to the development of set to as state 1, i.e. relay closed, closed circuit, current flows, voltage
theory, Venn diagrams, sentence logic and the logic of switching across load; "false" is referred to as state 0, i.e. relay open, open
circuits. It is not possible to consider these aspects in any detail circuit, no current flows, no voltage across the load. In electron-
because notation and laws require involved consideration. How- ics power supply connections are often +6V, -6V, OV. Voltage
ever the table given (Table 16.1) indicates the obvious relation- used depends on devices and circuitry requirements.
ship and the logic of switching circuits can be developed some-
what to illustrate principles. Only two input variables (A and B) A B A.B A+B A.B=A+B A+B=A.B
are considered for simplicity but extension is easily arranged.
0 0 0 0 I 1

Set Theory Sentence Logic Switching Circuits 1 0 0 1 I 0

anb aAb A.B 0 I 0 1 1 0

Intersection of a and b a and b "and"
1 I 1 1 0 0
aUb a-vb A+B
Union of a and b Either a or b or both "or" INPUTS AND OR NAND NOR
a' "'a A Table 16.2
Complement of a Not of a "not"

Table 16.1
• Note:
1. A series circuit is the AND function for output, i.e. output
signal is the same sense as inputs when all inputs are the same
The laws of Boolean Algebra and the use of Truth Tables only.
greatly facilitate the simplification of electrical logic circuit 2. A parallel circuit is the OR function for output, i.e. output
design. signal is the same sense as input change for anyone or all input
changes. Inclusive disjunction (gate) means either or all for the
SWITCHING CIRCUIT LOGIC OR function and exclusive means either not all (symbol EB).
Devices, using electromagnetic relays in control systems, have 3. The single relay, or switch, is the NOT function, i.e. contact
been used over a long period of time for such functions as
sequential starting, protection interlocks, counting circuits, etc.
.- closed gives output (closed circuit, state 1, A) and contact open
gives no output (open circuit,state O,A). Similarly the unity ratio
A simple application of electrical circuits is shown in Fig. 16.3. operational amplifier (inverter) is a NOT function.
4. In electronics the provision of a unity ratio operational

~~ ~2SJ
amplifier (inverter) in series with an AND circuit gives a NAND
circuit (NOT-AND) with output opposite in sense to input when
all inputs are the same only. Similarly the inverter in series with
A B
an OR circuit gives a NOR circuit (NOT-OR) with output oppo-
B
site in sense to input change for anyone or all (inclusive) input
Fig. 16.3 SWITCHING CIRCUIT LOGIC changes. Compare output states in the previous table. Logic

,.... or..{~.:i1;~m lP::;~~1iitit-j3~~1:::::~~_E.~~1S!m5:.! 1:~!.1t.rl

 290 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
.,(
•.•.•.••...••..•..
LOGIC AND COMPUTING 291
.. ".-

;1#
symbols vary, two conventions are shown in Fig. 16.4 and other
conducting, output voltage is small, near 0 V (state 0 V). A diode,
forms are shown in the text to illustrate the variations. Small
to earth across output, is sometimes fitted to ensure output 0 V.
circles at output change AND to NAND and OR to NOR.
Reversing polarity of both input and output signal requirements
gives an OR circuit, i.e. A or B or both at 0 V (state 0), with any
diode conducting, means output is 0 V (state 0). Resistors are
sometimes fitted at inputs (see Fig. 16.10).
~ ~ ~
B B -6V

C=A.B C=A+B C=A A

AND OR NOT

~ j)-; ~
B B
: B -
: I R! OUTP~I
Fig. 16.4 LOGIC SYMBOLS ov

+6V
Fig.16.5 DIODE GATE (AND)
SOLID STATE LOGIC
Such logic circuits (gates) are being increasingly used in place
of relays and thermionic valves. The convention for relays of 0
OR gate (diode) negative true logic
(off) and 1 (on) has to be modified because input and output
Refer to Fig. 16.6:
have non clearly defined states less simple than on-off. 1 may
Output volts are zero until one or both diodes conduct, when
represent higher (or more positive) voltage and 0 lower (or less
-6 V is applied to either or both inputs, the output is then -6 V,
positive) voltage. This is positive logic which is usually used with
i.e. output state 1 for any combination of state 1 inputs.
npn transistors because collector potential (and output) becomes
more positive when the transistor is cut off. The reverse applies
with pnp transistors and negative logic is used; this convention is -6V
adapted in this chapter, i.e. negative true logic, logical 1 negative •
with respect to logical 0 from a voltage level aspect. Digital logic A R
functions can be achieved by the use of diodes and transistors,
the former simpler and the latter more effective.
OUTPUT
AND gate (diode) negative true logic
Refer to Fig. 16.5:
With A and B at say -6 V (state 1) no current flows through R OV

and output is -6 V, i.e. output state 1 for coincidenceof state Is

+6V
at input. With any input at say 0 V (state 0), i.e. any diode
Fig.16.6 DIODE GATE (OR)

"i:~!rr~Z!:::~~--
292 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LOGIC AND COMPUTING 293

Again reversing polarity reverses role (becomes AND) and resis- Referring to the time delay switch of Fig. 16.8 where operation
tors are sometimes fitted at inputs (with or without diodes - see is similar to the above. With no input volts, output volts approx-
Fig. 16.7). Diodes ensure that inputs cannot affect each other. imately equal negative supply volts. When the input signal is
applied and causes the base to become negative then output
NOT gate becomes zero as the transistor activates and the capacitor (CAP)
Refer to Fig. 16.7: charges. When the input signal is removed the capacitor dis-
charges through the transistor emitter circuit as the rectifier (A)
blocks any outlet (anode is negative) through the input circuit. So

pnp TRANSISTOR OUTPUT

INPUT
COMMON pnp TRANSISTOR OUTPUT

+
COMMON
Fig. 16.7 NOT ELEMENT ARRANGED AS STATIC SWITCH
,,'
p
-+-

• Q
Fig.16.8 STATIC SWITCH WITH TIME DELAY
This is the inverter amplifier, single input, with output for no
input and if an input no output. Input say - 6 V (state 1) gives
output 0 V (state 0), i.e. antiphase, or input 0 V output -6 V. For
the configuration shown (pnp transistor) when there is 0 V input there is a time delay before the output reappears, delay depends
signal the base is slightly positive with respect to the emitter on the resistor and the capacitor values (i.e. time constant). A
(reverse bias) and no current flows. The transistor is then essen- variable resistor between P and Q allows shorter time delays and
tially a high resistance impedance-resistance between emitter a capacitor between X and Y allows longer time delays, resistor
Jj
and collector with an output voltage approximately equal to the
negative supply voltage. When an input signal of sufficient nega-
tive magnitude is applied the base swings to negative with respect
. '
and capacitor both being adjustable. Supplies may be ±6 V for
power source, common line may be earthed.

to the emitter and current flows to the collector. If such current is NOR gate (inclusive)
arranged to cause saturation of the transistor then the resistance See Fig. 16.9:
across emitter and collector is negligible so that the volt drop is The gate has one output and two or more inputs. Output
negligible and output volts are almost zero. Thus the bistable changes if one, or more, input states change and the output
amplifier with common emitter connection (as distinct from an change is antiphase potential to input change. This is a negative
alternative common base connection sometimes utilised) acts as sign output ORgate. If -6 Vis applied toA or B or both, thepnp
a switch circuit with on-off limits. The emitter could be regarded / transistor base is negative so that a large collector current flows
as earthed. This device is often used in annunciator systems (see and output volts are almost zero (small resistance across
Fig. 16.18). Supplies may be ±6 V. emitter-collector) if appropriate circuit values are designed in.

·a~a;r;'5;:;:;;
294 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

:: ~ ·0
1
/J
:1
NAND gate
LOGIC AND COMPUTING

Shown in Fig. 16.10 with symbol sketch above.

295

Single output changes only if all inputs change. With inputs at

6V
o V output volts are - 6 V and transistor is biased at cut off. If all
inputs are -6 V then bias is removed and the output is 0 V.
Output and input are antiphase. This is a negative sign output
AND gate.
OUTPUT
Note:
1. Circuit simplification will be apparent if only one (or two)
gates are used for all duties. NAND or NOR gates are so used.
OV
I 6V
2. Reversing polarity of gates reverses the role of the device.
+ 3. NAND gates can act as inverter NOT gates when only one
Fig.16.9 (INCLUSIVE) NOR GATE input is applied. If a NAND gate output is fed to another single
input NAND gate in series final output is in phase, i.e. AND gate.
Combinations of NAND gates can also provide OR, NOR,
HOLD, circuits.
With 0 V applied to inputs the output is -6 V (transistor biased
to cut off at near +6 V). A symbol sketch is shown at the top of 4. Similar remarks to 3. apply in general for NOR gates.
the diagram. Notation on such sketches often varies. An alterna-
tive often used is a semi-circle, base to inputs, with small circle HOLD (memory) circuit
(sometimes blacked in) or bar line on circumference leading to A time delay circuit has been considered previously (Fig.
output to indicate inversion (Fig. 16.4). 16.8). The remaining logic circuit commonly used is hold, or
memory, which can now be considered. Feedback holds in the
circuit even after input signal is removed. A reset signal then
restores the state of the system.
Consider Fig. 16.11:
:: :@ 0

6V FEEDBACK

START
_ -L
6V-. I
OUTPUT
OUTPUT

RESET
-6V-
T
STOP
OV
6V Fig. 16.11 HOLD CIRCUIT
+
Fig.16.10 NAND GATE

"'4--;
296 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

I 3
LOGIC AND COMPUTING 297

::1J--3 ::1- +
Closing the start button momentarily allows -6 V input and "j
the first NOR gate conducts with output 0 V. The second NOR
gate gives output -6 Vand this signal is fed back to the first NOR
gate. This circuit output is maintained when the start button is
AND OR NOT
released, i.e. held or remembered from the initial signal instruc-
tion. Closure of the reset (say stop) button changes the 0 V line
input from reset to -6 V. The second NOR gate, previously with
two inputs at 0 V, now has one input at -6 V and conducts, so
output is 0 V. Final output becomes 0 V and feedback at 0 V
means both inputs to the first NOR gate are 0 V and its -6 V H
output maintains the non-energised state when the stop button is
released.
Another alternative example is given in Fig. 16.12 following.
Electronic transistor gates are similar to relays. Functions can
G
also be achieved using diodes powered by the input signals.
Fig. 16.12 LOGIC UNITS
However transistor gates, as amplifiers, have the decided advan-
tage of utilising a separate circuit of supply for output power
which increases the scope greatly.

Note:
In logic circuitry entire circuits are packaged and it is not signals in the AND circuit and cuts off output H." Strictly the
necessary to know the exact circuit configuration of a particular combined sketch is NOT (A), NOR (D), NAND (CF) but
device (chip) because it is encapsulated. Signal tracing is imposs- redundant items (BCEF) simplify to OR (D) and AND (CF) and
ible and it is only necessary to understand the relation between the shading section on D would then be crossed.
overall input and output signals and repair is by replacement (the
black box philosophy). Flip-flop circuit
NAND and NOR are obviously combinations of the three Multivibrator circuits have been discussed (Chapter 7). The
given actions and various "tree" type circuits can be quickly built univibrator circuit as sketched in symbolic form in Fig. 16.13 is
up for otherwise complicated functions. used in computers. With inputs X and Yat state 1, say -6 V, the
For example the logic illustrative circuit shown as a combina- feedbacks at state 0 (0 V) and state 1 (-6 V). For the lower gate
tion in Fig. 16.12. The circuit may be interpreted as follows: "if feedback (0) and input Y (1) through NAND gives state 1
the off signal is not interrupted at the button and the on signal or (-6 V). For the upper gate feedback (1) and input X (1) through
the feedback (or interlock) signal G is energised there will be an NAND gives state 0 (0 V). Connecting inputs to state 0 will give
output at H. Pressing the on button gives inputs at A and D, stable reversing. A positive pulse to one input reverses potentials
hence an AND function and output atH (the two NOT functions at output and another pulse to the other input reverses output
at Band C cancel, i.e. like two negatives make a positive; simi- potentials again. For counting both inputs are connected so that
larly E and F). Release of the on button still allows output to be two positive input pulses are necessary for each positive output at
maintained through the feedback, i.e. the alternative input of the the outer transistor, i.e. each successive binary changes at half the
two element (0R) circuit. Pressing the offbutton cuts off one of the speed of the one driving it.
298 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LOGIC AND COMPUTING 299

x. double inverters I (single input NOR or NAND gates) gives a

(1)
ov • A delayed fuel supply pulse (delay by capacitors in circuit) at 0 V.
(0)
All inputs to NOR gate at 0 V. If any feedback input to NOR gate
becomes -6 V (due to stop button operation or fault on circuits
2, 5, 6) its output becomes 0 V to gate B. Gate B will stop
conducting and its output signal will become -6 V so shutting off
fuel by signal change. A time delay, not shown, can easily be
arranged to shut off only the air start signal after use. Many
----
.1'
-(1) versions of the above circuit can obviously be built with various
Y.
(1) logic gates.
Fig.16.13 FLIP-FLOP CIRCUIT
FLUIDIC LOGIC
Devices use fluid flow in sensing, logic computation and actua-
LOGIC SEQUENCE ENGINE SYSTEM tion. A commonly utilised principle is the Coanda effect, i.e. the
Figure 16.14 is a practical illustration of applied logic. All tendency of a jet stream to attach itself to the pipe surface (wall).
inputs, points 1 to 7, can be regarded as -6 V through input A complete range of devices can be built up and are widely used
resistors. With all essential start criteria 2 to 6 satisfied the in industry. It is not possible to cover all designs but to illustrate
transducers will have closed these input relays and operation of principles two types are chosen. The bi-stable multivibrator
the start button 1 gives all 0 V inputs (earth) to gate A whose (flip-flop) has been considered electronically and fluidic princi-
inverted output is -6 V. Gate B inputs are -6 V from A and ples are illustrated in Fig. 16.15.
-6 V from the inverted feedback NOR gate (whose inputs are all
oV). Gate B output at 0 V supplies the air start signal pulse and via

OUTPUT B

AIR SUPPl~
FLUID" ..-

00h OUTPUT A

INPUT A

INPUT B ~ I
.,--...J7
i \ --.......:JOUTPUT A

MANUAL
STOP
Fig. 16.14 LOGIC SEQUENCE ENGINE SYSTEM Fig. 16.15 FLUIDIC LOGIC DEVICES

'"r~~JIWa&i:1£,Hg'.!e!~a~i~~~im'::~~:UlS~:~_~-~~1 ~~:,::i!l!llrlt~ ~ ii;:

 300 REED'S INSTRUMENTATION AND CONTROL SYSTEMS LOGIC AND COMPUTING 301

Referring to the top sketch the fluid can be regarded as 19 denary in the scale 2.
attached to the wall of tube A at output. Input signal at B 19 = 1 x 2 4 + 0 X 23 + 0 X 22 + 1 X 21 + 1 x 1
separates attachment and diverts flow to tube B output. This
condition will hold (memory) if input B signal is removed and can
= 10011 (base 2)
be reset by an input A signal to the initial state. The latter, base 2, is the binary scale.
The lower sketch illustrates an exclusive OR device, i.e. input
A or B will give output A or B but not both. If inputs are zeros
output is zero and if inputs are applied together the jets impinge, BINARY SCALE
flow goes to vent, output is zero. The reader should first verify for practice:

101011 base 2 = 43 base 10

39 base 10 = 100111 base 2
DIGITAL COMPUTER
This scale has the advantage of expressing any number in terms
It is first necessary to consider simple computer-counting in
of two symbols, 0 and 1. If 0 is taken to represent current off and
binary terms.
1 to represent current on (or two voltage states) then it is possible
to use many electrical currents to perform calculations. This is the
SCALES OF NOTATION
basis of the digital computer counting in digits, pulses or bits.
The denary scale (base 10) is in general use but any number
base can be used in a scale. Consider the following:
1 1 1
Scale 10 ... 10 3 10 2 10 1 1 10 3 ...
10 10 2
Addition and subtraction
1 1 1
Scale 3 ... 3 3 32 3 1 1 ht u 2 6 2 5 24 2 3 22 2 1 1
3 32 33
-
1 43 1 0 1 0 1 1
Scale n ... n 3 n2 n 1 -
1
n n2 n3 + 39 + 1 0 0 1 1
-
51 denary in the scale 3. 1 Carry Figures 1 1 1 1 1
51 = 1 x 3 3 + 2 X 3 2 + 2 X 3 1 + 0 x 1 = 1220 (base 3)
82 1 0 1 0 0 1 0
248 denary in the scale 5.
248 = 1 x 53 + 4 X 52 + 4 X 51 + 3 x 1 = 1443 (base 5)

1475 denary in the scale 12.

1475 = ten x 12 2 + 2 x 12 + eleven x 1 = t2e (base 12)
Addition is as illustrated above. If the digit sum in any column
(t and e to be new digits representing ten and eleven in the denary totals 2, carry 1 to the next column and leave 0; if 3 carry 1 leave
scale)
1; if 4 carry 1 two columns and leave 0; etc.

'''1~~~, -~~--
~~~!;:r- ii:j
302 LOGIC AND COMPUTING 303
REED'S INSTRUMENTATION AND CONTROL SYSTEMS

COMPONENT UNITS
Subtraction is as illustrated below. When subtracting 1 from 0
A brief introduction to constituent elements of the digital
borrow (10) i.e. 2 from next column, change 0 to 1 working to the
computer can first be considered. Consider Fig. 16.16:
left until a 1 is reached, change this to a 0, and continue the
subtraction.
Input unit
Functions to accept input data coded information in the form
of either punched or magnetic tape or converted analogue signals
ht U 26 25 24 23 22 21 1 and transmit the electronic signals in digital form to the arithme-
tic unit.
Borrow Figures 1 (10) (10) 1 1 (10)
-
88 J- .e-- ~ ~ ;r Z 0 Arithmetic unit
-63 - 1 1 1 1 1 1 Processes inputs to desired arithmetic functions such as addi-
- tion, averaging, etc. Assembly of bistable devices for binary
25 1 1 0 0 1 operation including binary counter (flip-flops in series, extend-
ing range), shift register {flip-flops driven by clock), accumulator
and buffer store, etc. all pulse operated combinational and mem-
ory devices.

Multiplication and division Memory unit

Procedure exactly as denary using final addition or subtraction Store information with basic two stable state element. Storage
techniques above. can be paper or magnetic tape spools, magnetic drum or magne-
tic core stores. Magnetic cores are of two state ferrite core rings,
grouped into lines (words) with up to 36 digits (bits) per word. A
store with 12 bit (cores) per word and 2 12 words (4096) has
almost 50 k cores and is known as a 4 k memory store. This would
be classified as small with 64 k as large.

Control unit
Essentially the brain element of the computer. This is some-
times referred to as programme unit. The programme input is
decoded, addressed and so internal transfer involving memory
and arithmetic functions is carried out. Programmes involved in
the central processor may be specific language for a particular
type of machine or generalised language code such as Algol,
Cobol, Fortran, which is processed to machine code. Input and
output units are also controlled.

'Il,llilll ml?l'm"'
304 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

Output unit
Receives computed outputs as electrical signals, transmits
directly to control functions of plant, operates digital logging
l Reference sets
LOGIC AND COMPUTING

Amplified transducer inputs from scanners are compared to

manually set high-low level limits by potentiometers arranged as
305

units, display and alarm devices and presents output for typewri- multi-position rotary switches before analogue inputs to the
ter or as tape. Outputs may require reconversion digital to converter. Alternatively comparison is arranged digitally by
analogue for control action to analogue computers, plant con- direct programme control on desired values or manual set by pins
trollers, etc. inserted on a matrix patch board.

Peripheral units Scanner

In this case is meant outside the digital computer itself i.e. Measurements are sequentially selected by transistor circuits.
outside the input-output boundaries shown chain dotted on Fig. The transistors are operated by an input signal from the scan
16.16. control device regulated by pulses from a pulse unit digital clock.
Scan speeds can reach 400 points per second if required.

I I
I Analogue to digital converter
PROGRAMME .1
I
CONTROL r-- TYPE Analogue display is illustrated by a car speedometer, i.e. con-
tinuous, and digital display by the distance device, i.e. discrete
I
~ ~TAPE steps. One design of converter has a reference voltage potentially
~ I divided by resistance binary steps 2°, 2 1 , 22 , etc. which are com-
TAPE .1 ~OISPLAY
pared to the input voltage signal in sequence, until parity, when
output from tapped resistances is then digital. A similar principle
~ALARM can be applied by a balance bridge comparing input and output
I resistances. There are many types of converter available and a
I typical digital instrument is shown in Fig. 16.17. The comparator
~LOGGING
provides an output when the reference voltage is greater than the
MEMORY I integrated input voltage. This output opens a gate circuit allow-
~CONTROL ing the oscillator pulses (proportional to input voltage) to the
I I
L·OAC
counter. Range variable, with RC time constant.
Fig.16.16 DIGITAL COMPUTER C

For general purpose computers requiring high speeds and used

mainly for data processing in business systems inputs and prog- V· R
ramme inputs would be tape and outputs tape and type. The i--w..
programme input and digital pulse timing clock are obviously
essential peripheral equipment to any digital computer.
Engineering computers require reference sets, scanners and
analogue-digital converters. Outputs would be type logging on
paper tape, display, alarm and control functions. Fig.16.17 DIGITAL INSTRUMENT

fl5,~;{;:':::':':""~ ;_:r~;r:I~~~;~~1ti.~;.H~T~!~:;,~~~~!~~:~:5m_~!:;! :\."1

306 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

DIGITAL COMPUTER
Units and sketch have been considered. In large computers it is
necessary to have maximum utilisation of computer time. Batch
1
(
I ALARM
RECEIPT
LOGIC AND COMPUTING 307

1
processing utilises storage until a suitable time for handling the
programme. Real time operation requires instant response to
priority input when required. Time sharing to consumers is gen-
erally necessary in present digital systems.

DATA PROCESSING
Main application is in business systems. In engineering, digital
data processing equipment provides an integrated system of
monitoring plant process and includes alarm scanning, central- N
ised display and data recording.

DATA LOGGER
T-, ...
ALARM~
CONTACT
This is essentially centralised instrumentation only. As such
has now only a secondary function of measuring and logging with -6V
display. Useful for continuous monitoring and documentation of
Fig.16.18 ANNUNCIATOR SYSTEM
records of, for example, large refrigeration storage plant. Gen-
erally refined and superseded by an integrated system to include
alarm scanning which provides malfunction protection in com-
plex machinery installations.

ANNUNCIATOR SYSTEMS
Are central surveillance - collection points. An off normal Under normal conditions the alarm contact is closed, D output
condition results typically in the following sequence: (1) which operates the green light N, G output (0) and J output
(0) so klaxon L and red light M are off.
At fault, alarm contact opens, D output (0) and N off, E output
Condition Green Lamp Red Lamp Klaxon
(1) which feeds back to hold D. Input to F is (0) and signal
Normal On Off Off generator K also enters pulses into F so that M flashes. Outputs
Fault Off Flashing On from D and B are both (0) so output from Gis (1) and operates L.
Receive Off Steady Off When the alarm receipt button is operated a (1) signal is fed
Normal On Off Off into A and B, output is (1) which feeds back to hold A; also
overides flash input to H so M exhibits constant red light and
Fig. 16.18 shows a diagrammatic circuit for a single point annun- switches L off via G.
ciator system. State (1) is -6 V and State (0) is OV, NOR gates When the alarm contact returns to normal and closes, both
are used (except J-NOT). inputs to Care (0) and its (1) output resets the hold circuit DE,
Consider Fig. 16.18: then AB, so M is extinguished and N lights up.

tr::;tm~-trr.:~~~~
 LOGIC AND COMPUTING 309
308 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

INTEGRATED SYSTEM Signal selection

Includes data logging as one function. Essentially a digital A scanner unit. Sub-units may be used at local points in the
computer but of a fairly simple design, peripherals are as already installation for say 40 points which reduces cable runs to the
described. The programme is designed in and fixed so that prog- central control room and relieves space requirement of the pro-
ramme input, control, arithmetic and memory units as shown in cessor.
Fig. 16.19 are simplified and combined into one central proces-
sor. A very brief outline description will now be 'given and Signal processor
reference to Fig. 16.16 and Fig. 16.19 should be made. Provides amplification, analogue to digital conversion, limit
sets analogue or digital, scaling, linearising, etc. with outputs to
digital (or analogue with d-a conversion) display, routine logging
and alarm circuits.

ANALOGUE SECTION DIGITAL SECTION

Output devices
Visual and audible alarms are required. A logging typewriter
records alarm conditions and events, until cleared. The typewri-
ter can log by routine all points on demand or at time intervals.
Display requires careful consideration with a minimum of gauges
o---,d', restricted to essential observation. Mimic diagrams of plant cir-
....
::)
I
I cuits with indicator lights to indicate key points in the plant are
CL. I
~(/) I
I
often used. A central control room needs careful design for
oJ
w~
::)z
I
I
comfort, correct instrumentation and alarm indication plus good
C)C) I
0 -
....J(/) I effective lighting. Construction of components is modular with
~ I
z I
I
plug-in circuit cards for fault rectification. Process plant subject
~ TO
0--1-0""1 to such surveillance provides accurate and regular records and
ALL

INPUT
/L . iI I UNITS
machinery protection with reduced watchkeeping staff.
SELECTION
RELAYS INPUTS
FROM
TRANSDUCERS
COMPUTER CONTROL
Fig.16.19 SCHEMATIC DIAGRAM OF ELECTRONIC
ALARM SCANNING AND DATA LOGGING SYSTEM This envisages a full digital computer so programmed to main-
tain output variables at the design condition by providing outputs
to controllers. At present most controllers work on analogue
inputs so the digital computer requires an output digital-
analogue converter. Direct digital control is however coming
into use where computer output acts directly as control action on
Primary inputs
Analogue signals from transducers represent variables such as the final correcting element. Simple computers generally include
pressure, level, flow, etc. Each transducer has a plug-in printed controllers, data processing is now well established, it follows
circuit module for measurement range, limits, etc. and is that computer control is very relevant. In marine practice a
switched in by the scan control unit. computer can be so programmed to provide for example com-

..~~~~'7~~~- ':EF~;~J~:~lm:m:=:~~~t;,
 LOGIC AND COMPUTING 311
310 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

plete preparation of machinery plant for sea together with com-

puter control on passage. Built in emergency action is provided.
Such a large computer could of course cope adequately with
cargo handling in tankers and navigational route placing with
suitable addition to store and provision of correct programming. r' r----
~ --------~---*
I I -----
The analogue computer is faster on a complex path run but the
digital computer is accurate, repeatable and suited to short runs.
There are many fairly cheap analogue-hybrid units and parallel
hybrid refers to the addition of digital logic for automatic prog-
ramming of variables. Interface matching between analogue and
~
I ,--
----1
digital can often be very difficult. A true hybrid computer is I I
digital: interface: analogue which usually requires a suitable
T -'- r--
l L-I L I I B
L __
compiler processor language for operation such as Actran.
Recent developments have led to increased automated mass ....t,....

production of miniaturised and highly sophisticated functional F

complexes of circuits at a decreasing cost and volume (size). Mini
computers oflarge potential computing power are now used with
time sharing links to central large main frame computers. Micro
processors, with a memory chip (storage, programme) and input/
output devices linked to the central processing unit chip, are s
increasingly used virtually as micro computers dedicated to one
particular control system and function.

COMPUTER SIMULATION
x
Overall system performance, including interaction between
components and loops, at the initial design stage is becoming Fig.16.20 BOILER WATER LEVEL CONTROL AND SIMULATION
increasingly important. Analysis and simulation of the dynamic
(transient) performance, as well as steady state behaviour, is
required. Mathematical models, based initially on linear analysis
for frequency response, and by computer simulation for non-
linear systems are applied in design studies.
The upper sketch of Fig. 16.20 illustrates a boiler water-level
control loop. Boiler (B), steam (S) and feed (F) flow transmitter
signals are to the P computing relay (X) with output, joined by
level transmitter signal (L), arranged to the P + I computing
relay (Y). The output signal from Y operates the feedwater
control valve and positioner (V).
A computer simulation can be set up and a suitable patch diag-
ram is shown in the lower sketch of Fig. 16.20.
312 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
,
TEST EXAMPLES 16

1. Describe, with sketches, an electrical panel for use in

monitoring the alarm conditions of a set of diesel generators. SPECIMEN EXAMINATION QUESTIONS
2. With reference to data logging systems, explain the meaning Note: Questions on instrumentation and control systems are set
of the following terms: within a wide range of subject titles such as Engineering Know-
(a) sensing device, ledge (General, Motor, Steam), Instrumentation, Electrotech-
(b) scanner, nology, Power Plant, etc. The specimen questions following are
(c) transducer, based on expected 1/2 hour answers, which is the most common
(d) scaling unit. practice. In some examinations however only short answer ques-
Make a diagrammatic sketch of the components interconnected tions are set, e.g. DoT Class Three - 10 minutes, and in others a
in the logger. combination of sections involving 1/2 hour and short answer
3. Write down in logic symbols, for input and output, the questions is used, e.g. some TEC and SCOTEC, previously
following functions, each with two inputs A and B: CGLI/OND.
(a) OR gate, DoT CLASS THREE
(b) NOR gate, At the time of publication of this third edition very few speci-
(c) AND gate, men questions for this relatively new grade of Certificate of
(d) NAND gate. Competency are available. However the questions following
Sketch, in logic symbol form, each gate. Draw a circuit diagram might reasonably be expected to be typical of the standard of
for the 0 R and AND gates and briefly describe the modifications questions to be set at this level.
to give NOR and NAND functions respectively.
1. Sketch a thermometer suitable for remote reading - indica-
tion.
2. Sketch a valve for automatically controlling fluid flow in a
pipeline.
3. Describe, with a simple sketch, any level indicating device.

4. An engine alarm sounds intermittently. How would you

determine if it is a genuine or nuisance alarm?
5. List any five essential instrument readings, indicating the
running condition of machinery, for display on a centralised
instrument panel.
314 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

6. Describe briefly any two types of pressure measuring device.

7. Sketch a moving coil type of electrical measuring instru-

ment.

8. Briefly describe any method of speed control for an electric

motor.
\
SPECIMEN EXAMINATION QUESTIONS
9. Describe, with a simple sketch, how a flapper-nozzle device,
controls pneumatic pressure in an air line. t
,! DoT CLASS TWO
10. Describe how remote smoke/fire indication can be located
at a central observation station. 1. Describe with sketches, two methods for increasing the
strength of a signal in a control system.
11. List four essential parameters of machinery in operation 2. Describe instruments used for measuring the temperature in
requiring alarm indicators. the following spaces:
(a) refrigerated hold,
12. Describe, with a sketch, anyone type of flame or fire or (b) engine room,
smoke detector. (c) boiler uptake.
3. Describe with sketches, two methods for remotely determin-
ing the quantity of liquid in a tank. Compare the accuracy of
these methods and explain how the degree of accuracy can be
maintained. State one possible source of error for each of the
methods described.
4. Sketch and describe three methods of sensing temperature
change.
Describe how the signals are converted and fed into an automatic
control system.
5. Sketch a compressed air system for use with pneumatic
controls.
Write notes on each component shown.
6. Explain how a temperature sensitive element installed in a
refrigerated stores locker can be used for starting and stopping a
domestic refrigeration compressor.
7. Describe, with sketches, how the inlet temperature of lub-
ricating oil to main engine or gearing may be automatically
monitored and controlled.
8. Describe a bridge/engine room telegraph interconnecting
gear. Explain how the system may operate a "wrong way" alarm.
 316 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 317

9. State why a pneumatic control system requires clean dry air. 14. The following terms may be used to describe a boiler feed
Explain how the following air pollutants are dealt with: water controller:
(a) water, (a) detecting element,
(b) oil, (b) servo-motor,
(c) dust and dirt. (c) desired value,
(d) difference element.
10. Draw a line diagram of an arrangement whereby the pres- Relate the terms to practical components and describe their
sure of oil delivered to a main lubricating oil system by a constant operation in the controller.
speed, positive displacement pump is pneumatically controlled
within set limits. Trace the sequence of events upon deviation in
oil pressure.
15. Sketch a simplified circuit diagram of an npn transistor
11. Sketch a compressed air system for pneumatic controls illustrating its use as an amplifier. Give a reasoned account of its
labelling all the principal items. operation in terms of the electron theory.
Describe with sketches an automatic drain on the air compressor.
State what routine maintenance and tests are needed to keep the
16. Sketch and describe a master controller, operated by varia-
system fully operational.
tion of pressure in the boiler, for regulating the air and fuel
12. Sketch a pneumatically operated valve for regulating cool- supply to the furnace by a pneumatic control system.
ant flow. Explain how "hunting" of the system is prevented.
Explain how the pneumatic system controls valve movement.
State how valve position is indicated at the control station. 17. The left hand side of a small bar of crystalline silicon
contains a small proportion of the element phosphorus as an
13. Fig. A illustrates how a transistor amplifies a signal from a impurity and the right hand side contains a small proportion of
thermocouple to operate a relay. Describe the principles under- aluminium. Silicon has four valence electrons, phosphorus five
lying this amplification and explain why it is necessary. and aluminium three.
Sketch the bar, and indicate the n-type region and the p-type
region. Draw a battery connected between the opposite ends of
the bar and show the polarity of its terminals which will bias the
p-n junction in the forward direction.
Explain why no current will flow through the bar when this
polarity is reversed.

18. With reference to medium or slow speed engines, describe

Fig. A transducers suitable for producing electrical or pneumatic signals
to indicate:
(a) lubricating oil pressure,
(b) jacket cooling water temperature.
Explain how each of these transducers can be tested.
318 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
SPECIMEN QUESTIONS 319

19. Fig. B illustrates a three phase full wave bridge rectifier 24. Draw a line diagram of a pneumatic system for control of
circuit. Using this diagram explain how an ac voltage is converted ballast, labelling the principal components.
to a dc voltage. Draw the graph of output voltage against time for Describe how a ballast valve is remotely operated.
one cycle. State where such a circuit could be used in marine State why driers, filters and automatic compressors are prefer-
engineering. able to "bleed off' from main or auxiliary air receivers.

25. Fig. C shows a simplified circuit of an automatic voltage

regulator. Explain how the arrangement maintains a constant
voltage output.
THREE PHASE
ae SUPPLY
ae GENERATOR

OUTPUT
de

Fig. B
t
20. With reference to automatic combustion control, explain
with sketches how: I
(a) and why pressure differentials across air registers are
measured,
(b) air/fuel ratios are adjusted,
(c) the boiler can continue to operate upon failure of the
fuel flow regulating valve.

21. Sketch in detail a section through a hydraulic governor as

Fig.C
fitted to medium speed unidirectional engines. Explain how it
operates under frequent and wide fluctuations of load.

22. Describe with sketches an automatic supervisory and alarm

arrangement for the control of bilge water accumulation under 26. Describe, with sketches, integrated alarm and control sys-
periodically unattended conditions. tems for monitoring the following:
Explain how the control system is prevented from giving a false (a) air pressure,
(b) tank contents,
alarm due to ship motion.
(c) oil temperature.
23. Explain the term "semi-conductor" and describe what is
meant by "p" and "n" type materials. 27. Describe with sketches, two different methods for the
With the aid of a simple sketch discuss the conditions at the remote measurement of fluid flow through a pipe. Compare the
junction between a p-type and an n-type semiconductor. accuracy of the methods. Give one cause of error in each method.

----------------------------------- -
~2~1~~~'!EriS~~~~~~~;~Il:'im!!~g21~,!"'T~·:~
320 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

28. Draw a line diagram of a system to operate widely distri-

buted valves from one point.
Describe how anyone valve is remotely manipulated.
Suggest, with reasons, a shipboard system which lends itself to
such control.

29. Describe how a crankcase oil mist detector operates.

Sketch a mist detector system showing the run of the sampling SPECIMEN EXAMINATION QUESTIONS
pipes.
State how sampling is controlled.
DoT CLASS ONE
30. Draw a line diagram of an automatic sootblower system.
Describe the sequential operation of the blowers. Give reasons 1. Describe how the supervisory equipment for the control of
machinery in a periodically unattended engine room is itself
for the order of operation followed.
monitored for defects on individual channels and as a complete
State what safety devices are incorporated in the system and why
unit.
they are fitted.
2. Explain the problems in controlling the upper and lower
limits of temperature of lubricating oil supplied to the main
machinery.
Describe with sketches how this supply temperature can be
automatically monitored and controlled.
3. State two advantages and two limitations in the use of elec-
trical signals and also in the use of pneumatic signals as transmit-
ting media in data transmission systems.
Explain with the aid of diagrams the principle of operation of a
force balance transmitter employing either electrical or pneuma-
tic signals as the transmitting medium.
4. Describe, with sketches, how electrical signals are converted
to pneumatic signals in control systems.
Suggest a shipboard application featuring this conversion and
state the defects to which the arrangement is susceptible.
5. Sketch and describe a system for indicating remotely the
propeller shaft speed.
Explain how, for the system selected, inaccuracies occur and are
kept to a minimum.
6. Sketch and describe a method of measuring the pressure
differential for fluid flow systems.
State what are the effects of altering the orifice plate size or the
position of the tapping points.

'ilf'.dlmrnr,.;: fa'5$:;t;t
322 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
SPECIMEN QUESTIONS 323

7. Describe the construction and principle of operation of a 12. State why the temperature of lubricating oil supplied to an
Bourdon pressure gauge. engine needs close control.
State the factors upon which its operation depends. Sketch and describe an arrangement and explain the principle of
Outline a sequence of tests and adjustments applied to such a operation of instrumentation and control equipment for auto-
gauge known to be inaccurate and in particular mention at least matically maintaining the temperature of lubricating oil supplied
three of the following: to an engine at its desired value.
(a) leakage,
(b) hysteresis, 13. Sketch and describe in detail the construction and opera-
(c) non-linearity, tion of one of the following:
(d) magnification, (a) an electric torsionmeter,
(e) zero error. (b) a preferential trip,
(c) an electric telegraph.
8. State what is the purpose of each of the following items in a
machinery control system: 14. Define the term "cascade control" as applied to control
(a) portable mercury manometer, engineering.
(b) portable inclined-tube manometer. Describe, with sketches, cascade control as applied to an engine
(c) portable temperature potentiometer, coolant system. Show on the sketches how pressure and tempera-
(d) compressor and vacuum pump. ture varies at the cardinal points in the system.
Describe in detail any two of these items. Give one advantage and one disadvantage of this control
arrangement.
9. With reference to a gyro-controlled hydraulic steering gear,
explain its action using control engineering terms making specific 15. Fig. D shows a circuit of a common-emitter amplifier. The
reference to: transistor has a high current gain so that its base current is small.
(a) desired value, If the current through the emitter resistor is 0·4 mA determine
(b) feedback, the battery voltage. Assume that when the transistor is conduct-
(c) actuator. ing, the voltage between the base and emitter is 0·2 V. (90·2 volts
State the part played by the cylinder relief valves in the automatic - this is abnormally high, with a 2 kn resistor in the emitter
control system. circuit the supply voltage is 11 V which is a much more accept-
able value).
10. Sketch and describe a fuel meter used with high viscosity
fuel.
Explain how it operates.
Explain the value of the readings obtained and how they are
used.

11. Sketch and describe how coolant flow may be measured on

a linear scale.
Explain the principle of operation of the instrument concerned. , " +
Explain why the values recorded may vary from those expected Fig. 0
from calculations.

.>-<
324 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 325

16. With reference to boiler combustion control explain: 19. Describe with sketches how the pressure of a fluid is con-
(a) the operation of the master controller following varia- trolled by a pneumatic controller incorporating proportional and
tion in steam pressure, integral (reset) action.
(b) the importance of pressure drop across the air regis- Give reasons for instability in the controller action. State how
ters, this instability is overcome.
(c) how the air-fuel ratio is adjusted,
(d) how the boiler operation can continue upon failure of
20. Give a detailed diagrammatic sketch of a Woodward gov-
the fuel flow regulating valve.
ernor. Explain how this governor operates. State what advan-
17. The common-emitter output characteristics for a transistor
tages it possesses over inertia governors.
are as follows:

Collector Collector Current (rnA) 21. Draw in detail a diaphragm operated control valve. Ana-
Voltage (V) I b = 30~ Ib = 60~ I b = 90~ lyse the action of the inter-connecting elements, that is, the parts
3·0 1·0 2·1 3·2 affecting control.
7·0 1·29 2·55 3·9 Explain how the load change is communicated to the actuator.
10·0 1·5 2·9 4·4 State where such a valve may be used in an engine room.
Draw the graph of collector current against collector voltage
and construct load lines to show the operation from a 6·5 V 1 22. Explain why the simple float control feed regulator is
battery with load resistors of 1000 0 and 1500 0 respectively. inadequate for the present generation of main boilers.
If a suitable value of base bias current is 60 f.LA for an input signal Describe with sketches, or block diagrams, a feed control system
of ± 30 ~, determjne the current amplification for each load. in which it is possible to programme a set point for various loads.
(33.3, 31.6).
18. The figure shows the equivalent T-circuit of a transistor
used in a common-base circuit. The resistances presented to the 23. Give the advantages and disadvantages of data-logging
alternating components of the current by the emitter (r e ), the systems used in connection with ships' machinery.
base (rb) and the collector (r c ) are 30 0, 0·6 kO, and 1MO Explain the value of recorded data and how it is interpreted and
respectively. usefully employed.
r;
If the current amplification factor (a) is 0·98 and the load resis-
tance (R) is 9 kO, calculate the current, voltage and power gain <it, 24. With reference to automatic voltage regulators discuss the
and the input resistance. (0'97, 186, 180; 48 ohms). function of the following basic elements:
(a) error detecting element,

Irr-:J
oc: Iere
(b) correcting element,
(c) stabilising element.
Clv
v.• ,. 25. A pressure controller is fitted in a fuel line where ultimate
8" state error is minimal and first response is necessary. Sketch and
B describe such a controller making reference to components giv-
Fig. E ing the desired characteristics.
 326 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 327

26. Fig. F. shows a simplified circuit of a transistor employed as 29. Draw a line diagram of a boiler combustion control system
a switch in a relay operation. Describe how this is accomplished. labelling the principal items.
Explain how the system functions and in particular how feed
water supply, fuel supply and air/fuel ratio are regulated to
match steam pressure and flow variation.
Explain how these controls can be tested for alarm conditions
without upsetting the balance of the system.
R

30. Sketch and describe a hydraulic servo system associated

with a controllable pitch propeller.
OUTPUT State how loss of fluid pressure occurs and warning is given of
impending failure. Explain what happens upon loss of oil pres-
sure and why.
State what routine maintenance is necessary to ensure trouble
e I , I e+ free operation of the propeller.

Fig. F

27. For the purposes of remote recording give two advantages

and two disadvantages of:
(a) electrical signals,
(b) pneumatic signals.
Explain which system is best suited for sensing and remote
indication of:
i. lubricating oil pump discharge pressure,
ii. engine bearing temperature,
iii. fuel tank liquid level.
State what precautions are necessary when sensors are located in
hazardous areas.

28. Sketch and describe a three element feed water control

giving reasons for its location.
Explain how unity relationship is maintained between the iden-
tified variables.
Explain why three element control is superior to two element
control.
 I SPECIMEN QUESTIONS 329

6. Describe, with the aid of a suitable circuit diagram, a method

of obtaining the output characteristics of a junction transistor.
Sketch the characteristics and from them show how it is possible
to draw the transfer characteristic and hence determine the
current amplification factor of the transistor.
7. (a) Draw diagrams of instruments capable of measuring:
SPECIMEN EXAMINATION QUESTIONS i. carbon dioxide content in a gas sample,
ii. oxygen content in a gas sample.
TEC/SCOTEC (& METC, OND) (b) Describe in each case, for these two instruments, the
1. Draw a labelled diagram of a flow measuring device to give a principle of operation and indicate typical readings expected in a
remote reading on an instrument panel. sample of gas from a boiler uptake.
Explain how the device operates to measure and indicate a
8. (a) State three essential requirements for any instrument
change of flow.
suitable for use in a modern marine power plant.
2. (a) For any expansion type of thermometer, give four fac-
(b) Explain, with reference to instrument display, the terms
tors involved in obtaining a high speed of response.
"analogue" and "digital".
(b) List three methods of temperature measurement and (c) Give two advantages of the oscilloscope over moving
sketch and describe one method in detail, giving advantages and iron instruments for the measurement of ac quantities.
disadvantages.
3. Describe, with a sketch, the static characteristics of a junc- 9. (a) Draw the circuit diagram, indicating clearly the correct
tion diode. polarity of the electrical supplies, for a pnp junction transistor
Sketch the forward and reverse characteristics and write short connected in common-emitter mode.
notes on: (b) Draw the circuit diagram, indicating clearly the direc-
1. the forward resistance, tion of current through the load, for four silicon diodes
ii. the leakage current, connected in a bridge configuration to produce full wave rectifi-
Iii. zener action. cation of current in a resistive load.
With the aid of waveform diagrams, illustrate the action of a
capacitor connected in parallel with the output of a half-wave 10. Sketch and describe a pneumatic diaphragm operated con-
rectifier. trol valve, clearly label and describe each part.
4. (a) List four different methods of measuring liquid level. Explain what is meant by "Fail Open" and "Fail Closed".
(b) Sketch and describe a remote reading liquid level indi-
cator suitable for a high pressure boiler. 11. Describe fully how you would carry out a calibration test on
5. (a) Draw a fully labelled sketch illustrating the principle of a Bourdon pressure gauge.
operation of a differential pressure transmitter. Sketch a typical calibration curve and comment on possible
Describe its operation for a change in the measured differential errors and methods of correction.
pressure.
12. Describe, with the aid of a clearly labelled diagram, the
(b) State how zero and range adjustments are made on dp operation of a direct acting pneumatic relay.
transmitters and name two different marine applications for such
State, within a measuring system, where such a device is usually
a transmitter.
used and give the reason for its use.
 330 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 331

13. Explain, with the aid of sketches: 20. Sketch a purge or "bubbler" system which could be used to
(a) an electrical method, measure the level of fluid in a tank.
(b) a non-electrical method, What other information regarding the fluid would be necessary
of measuring fluid flow, listing advantages and disadvantages of in order to infer a mass measurement from a level measurement?
each method. How could such an instrument be developed to measure the
draught of a ship?
14. What is meant by the term "system response"? Give three
examples of system response.
21. (a) Draw a circuit diagram to show how junction diodes
How is the "time constant" or "system lag" measured?
may be used to give full-wave rectification from a single-phase ac
Why is it beneficial in a control system to reduce the time supply.
constant of that system?
(b) A npn transistor is to be connected in common emitter
15. Explain carefully, with the aid of a clearly labelled diagram, mode for use in voltage amplification. Draw a simple circuit
the principle of operation of either: diagram to illustrate the configuration.
i a motion balance transmitter,
22. Give an example of level measurements which utilise the
or ii. a force balance transmitter.
following physical principles:
Give two examples where either i. or ii. may be found in marine (a) hydrostatic head,
practice.
(b) float movement,
16. Using V-tube manometer diagrams, explain how: (c) displacement.
(a) absolute pressure, Illustrate your answers with suitable diagrams and descriptive
(b) gauge pressure, data.
(c) differential pressure, 23. Change of fluid level in a tank can be detected and meas-
is measured.
ured by the change of capacitance of a capacitance probe. Sketch
Explain briefly how one such measurement may be transmitted such a system and describe its operation in detail, from change in
to a remote recording station. level to change in indication of measured level.
17. Explain, with the aid of sketches, the principle of opera-
24. Make a clearly labelled circuit diagram of a "bridge" ther-
tion of a nozzle-flapper system. Show graphically the relation
mometer which has ambient temperature compensation, and
between flapper clearance and output pressure. What
also has provision for "zero or standardising test" selection.
reasons are there for fitting a feedback bellows in the sys-
Explain the operating sequence followed to ~nake the zero/stan-
tem.
dardising test before using the thermometer.
18. Sketch and describe a strain gauge. Describe how the ambient temperature compensation functions.
What is the principle of operation and for what purpose can the
strain gauge be utilised?

19. Describe what is meant by ramp, step and sinusoidal

response when applied to a system. Illustrate each with a simple
sketch.
 SPECIMEN QUESTIONS 333

5. Make a diagrammatic sketch of a three term controller suit-

able for use in a feed water control system. The sketch should
include measure and regulating units.
Briefly explain why it is necessary to incorporate:
i. a negative feedback arrangement,
ii. a relay (amplifier).

SPECIMEN EXAMINATION QUESTIONS 6. (a) State three advantages which solid state devices have
HIGHER; TEC/SCOTEC (& CGLI ADVANCED METC) compared to thermionic valves.
Why is silicon preferred to germanium for use in solid state
1. Describe, with the aid of a suitable diagram, the construc-
rectifiers?
tional details and operational working principle of one of the (b) State the function of the gate in a transistor.
following devices: Sketch waveforms of ac supply voltage, load voltage and gate
(a) oil mist detector, pulse for a thyristor when the trigger angle is 90°.
(b) torsionmeter,
(c) oxygen analyser.

7. Make a diagrammatic sketch of a steam flow plant incor-

2. Draw a diagram of a system for fully automatic main engine
porating a flowmeter utilising an orifice plate and a differential
jacket temperature control, incorporating steam preheating of
pressure transmitter.
the coolant, utilising split range control.
Explain the operating principle of the transmitter.
Comment on the type of control action to be used and the "fail
Sketch the output signal graph with and without square root
safe" arrangement adopted.
extraction.

3. (a) Construct a "truth table" for a two input NOR logic

gate.
(b) Sketch the circuit symbols for a pnp and npn transistor 8. Explain the meaning of the following terms used in control
and clearly label the leads. terminology:
(c) Draw a block diagram for part of a marine data logger (a) offset,
and describe in detail the function of one particular block unit. (b) fail safe,
(c) proportional band,
(d) derivative action time,
4. Automatic control valves are available with the following (e) cascade control.
features:
(a) linear or equal percentage characteristic,
(b) power to open or power to close,
(c) single or double seated. 9. Explain, with the aid of a diagram, a controller utilizing
Discuss the factors that determine the choice of each of the above proportional plus integral action.
features. Describe how such a controller is "tuned" within a system.
 334 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 335

10. A boiler pressure is to be maintained at 5 bar by a pneuma- 14. An electric torsionmeter, shown in Fig. G. is fitted to the
tic proportional controller. The pressure element has a range output shaft of a marine engine.
from 3 bar to 6 bar and the proportional band is set at 20%.
(a) Calculate maximum and minimum pressures corres- (a) With the aid of this diagram explain in detail the prin-
ponding to no load and to full load. (5·3 bar, 4·7 bar). ciple of operation.
(b) If the controller output pressure range is from 1·2 bar
to 2·0 bar, draw a graph of boiler pressure and controller output
pressure, and from the graph determine the boiler pressure at a (b) Explain how the instrument would be used to deter-
controller output pressure of 1·44 bar. (5 ·16 bar) mine the output power using the appropriate formula. Detail
carefully how shaft constants are obtained.
11. (a) Sketch typical forward and reverse characteristics for a
zener diode and briefly explain the action of this device when a
reverse voltage is applied to it. FLANGE FACE
ATTACHED TO
(b) State one application of the zener diode in marine CORES \ PRIMARY
equipment. .:..-..+------+
~ULL L1N~~fUIT
SHAFT UNIT
(c) A 10 V, 500 mW zener diode used as a voltage stabil- AIR GAPS
iser is supplied at 40 V through a series resistor of 500 ohms. If
ALTERED BY
TORQUE i-l-:-l ~ I L, f::r --- --"1
I ,.-t;:] . t:J--, I
the load voltage is stable for diode currents greater than 5 rnA,
draw the circuit diagram and calculate: / L..~--=-:....-_-::...-::...t-=--:::':~~---...J-i ~ SLIP RINGS
FLANGE FACE
~~ON SHAFT
i. the maximum and minimum load current for stable
operation, (55 rnA, 10 rnA),
ATTACHED TO
IRON PIECE •~
MAINS
I:
I I
THUS

I I
ii. the minimum power rating of the series resistor. (1800 I I SECONDARY
mW). INTERUPTERS l :- m~~~~TD
rl MOTOR h tA Xr - - -.-Q--J
J I LINES)
12. Describe, with the aid of sketches, the construction and
operational features of one of the following devices:
4'--_----'~ ---1 I i -....c-GALVANOMETER
r---- J I
(a) a buoyancy tube liquid level transmitter, :0 I

(b) a pneumatic differential pressure transmitter, r----...J

I r-----
(c) a vapour pressure thermometer utilised for remote I NDICATOR UNIT
AIR GAPS
I
I
I
L --
indication. ALTERED BY
MICRO SCREW
I
L_ t ~

13. With the aid of simple diagrams illustrate the response of a

proportional controller, when the system is subject to a load
change if: Fig.G
i. the proportional band is too narrow,
ii. the proportional band is too wide,
iii. the proportional band is at the optimum setting.
Briefly describe why a relay valve is used with such a controller
and state what secondary function relay valves usually perform.
336 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

15. Sketch a thyristor speed control system for a de motor.

Discuss briefly how the system will respond to an increase in the
speed set point potentiometer and explain the motor protection
I SPECIMEN QUESTIONS

17. Fig. I shows the arrangement of a spray-type of attem-

perator in which the outlet steam temperature is controlled by
varying the amount of feed water injected (1 kg feed is required
337

features included in the system. to attemperate 10 kg of steam).

Assuming that the system has been adjusted so that the meas-
16. (a) A reverse-acting two term pneumatic controller is sub- ured variable (outlet steam temperature) is at the desired value
ject to an input which varies as shown in Fig. H. when the load is 40 kg of steam per minute determine:
Construct a diagram showing the variation in controller output i. the offset if the load changes to 20 kg of steam per
due to: minute, (-16°C),
ii. the limits of load that can be accommodated, assuming
1. proportional action, that the offset shall not be more than ±30°C. (2·5-77·5
ii. integral action, kg/minute).
Iii. combined (total output).
(b) State wJ1ether the diagram constructed indicates that PROPORTIONAL CONTROLLER
2
the load on the plant has been restored to the original value after INPUT AND OUTPUT RANGE 20-100 kN/m
PROPORTIONAL BAND 100 %
the disturbance and give a brief reason for your answer. ..-----,
DESIRED VALUE
Proportional band 80%, integral action time 120 seconds, con- 400°C

troller output before disturbance 70 kN/m 2 • (K I = 1·25 then

use a tabular method; at three minutes, for example, using
V = K I (e + K 2 /K I Ie dt) the total change in controller out- DIRECT ACTING VALVE
DIRECT ACTING
20-100 kN/m 2
out is -37'5, i.e. -37,5 = -1·25 (20 + 1/2 X 20). Controller 0-10 kg FEED/min TEMPERATURE
~ TRANSMITTER
output is not restored to its original value so neither is the 360 -440 °C
load). 20 -100 kN/m 2

t 100
I--J
I
~ 90
E 80
STEAM r-+ 500°C 1_-
I
z INLET 1
=
~
::> 60
70 I
I
- .,---
LOAD RANGE
0-100 kg STEAM! min
I
I
~ I
-
a:::
50 I
I
-:---1--
I

I
I

I
I
UJ 40 I I I I I
6....J 30 __ II I __ I
L
1I
I _ I
a::: I I I I I
~ I I I I I
Z 20 I I I I I
8 10: : I I :
I :: l I Fig.•
00 1 2 3 4 5 6 7 8 9
TIME (min) -----...

Fig. H
338 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

18. (a) Sketch, and state, the type of logic gate which obeys the
following:
"If all inputs are one the output is zero; for all other combina-
I SPECIMEN QUESTIONS

21. (a) Draw a system for the automatic control of main tur-
bine engine gland steam pressure. The system is to use two
control valves, one make up and one rejection operating on an
339

tions the output is one." (NAND). equal split range with 4 kN/m 2 underlap in the signal.
(b) Fig. J. (left hand sketch) shows the circuit of an electri- (b) State, or show on the system drawing:
cal logic gate. By inspection or otherwise, state what type of gate i. the control valve actions and fail safes,
the circuit represents and write down the equation for the output ii. the operating signal range for each valve,
at Z. (OR, A + B + C). iii. the reason for underlap in the signal.
(c) Construct a truth table for the logic circuit shown in (c) Describe clearly the actions throughout the control
Fig. J. (right hand sketch) and give the outputs at Z from each loop as the gland steam pressure rises.
logic unit. (For example, ABC condition 011, outputs 01100). 22. An oily water separator is fitted with two teflon coated
capacitance probes, each with a separate measuring bridge and
output:

~:
A i. an oil depth probe vertically, to operate the oil
8
discharge valve,
= t z, ii. an interface probe horizontally, to operate an alarm
and oily bilge pump trip,
(a) Sketch the arrangement and explain clearly in terms of
capacitance change, the operation of each probe as oil level
varies.
(b) If the separator is completely empty all alarm condi-
Fig. J
tions will operate when starting up. Explain the reason for this.
23. (a) In a calibration check on a P + D controller the ramp
input is increased linearly at 1% per minute. This produces an
19 (a) Draw a fully labelled diagram of a pneumatic tempera- immediate 4% step change in output, after which the output
ture transmitter operated by a liquid or gas filled sensor. changes linearly at 2% per minute.
(b) Describe the operation of the transmitter in response Sketch input and output characteristics and determine:
to a temperature increase, making particular reference to the i. derivative action time, (2 minutes),
need for, and operation of, negative feedback. ii. proportional band. (50%).
(b) A step change of 6% is applied to the input of a P + I
20. (a) Define, and illustrate with simple diagrams, the follow- controller and the output undergoes a sudden step change of 4%
ing: and after a time interval of 2 minutes the total output change is
i. distance-velocity lag, 10%.
ii. transfer lag, Sketch input and output characteristics for this calibration test
Iii. potential correction. and determine:
(b) Define, and state the mathematical representation of i. proportional band, (150%),
the following modes of control: ii. integral action time. (1,333 minutes).
i. proportional, (c) Define the terms distance-velocity lag and exponential
ii. integral. lag.
340 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

24. Fig. K shows a boiler arranged for manual control of the

drum steam pressure.
(a) Complete the diagram by adding the necessary com-
ponents and connections to convert to an automatic closed loop
combustion control system.
(b) Describe the operation of the automatic system.

SPECIMEN EXAMINATION QUESTIONS

FD FAN

, HIGHER; TEC/SCOTEC (& OND SUPPLEMENTARY)

/
, 1. With the aid of a diagrammatic sketch describe a type of
/

AIR CONTROL DAMPER

controller which incorporates proportional, integral and deriva-
···._L~~'·· tive actions. Explain clearly how each action is generated and
comment briefly on possible interaction that may occur.

STEAM 2. Sketch circuit diagrams of resistance-capacitance networks

,. . .------l101~'b~~~¥~~ which will provide an output signal that approximates to:
(a) proportion plus derivative of input signal,
(b) proportion plus integral of input signal.
Show, in each case, how these networks could be incorporated
into a closed loop control system so as to improve the operating
J ,
~0AIRFLOW
;:: INDICATOR
performance of the system.
3. For a closed loop position control system:
(a) make a sketch to illustrate how the components are
connected,
(b) describe the sequence of events following movement
of the input shaft,
FUEL
CONTROL (c) explain the term "overshoot" in relation to such a system,
VALVE
(d) describe how damping feedback could be incorporated.

4. (a) Define the terms proportional and integral action.

Fig. K (b) Sketch the open loop characteristic response for propor-
tional and integral action when a step input changes is applied in
each case.
(c) Calculate the integral action time if a step input change
of 0·04 bar applied to a pneumatic controller set with a 100%
proportional bandwidth gives a response change of 0·2 bar in 3s.
(0,75 s).
342 'I
REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 343 !

5. (a) Explain, by means of a block diagram or otherwise, the 10. (a) Describe the effect of:
fundamental principle of a dc chopper amplifier. i. positive feedback,
(b) Sketch a circuit/block diagram to show how a chopper ii. negative feedback,
amplifier is used in an instrument servo-mechanism which on a closed loop speed control system.
records dc potentials by means of a self balancing potentiometer. (b) For voltage at input et, output eo and error actuating
Briefly describe the operating principles utilised. input e to a negative feedback electronic amplifier of forward
gain G and feedback fraction F, derive the following:
6. (a) Explain by means of a two transistor analogy, or other-
wise, how a gate input pulse will cause a transistor to conduct. G
eo
(b) A thyristor stack is used to regulate the heating power to = 1 +FG
et
a cargo hold in order to provide automatic temperature control.
Draw a block diagram to show the basic components for such a e 1
- = 1 +FG
control system. et
(c) Using switches and a lamp as an example, explain with
the aid of simple diagrams what is meant by the following logic
terms: If the "open loop" gain is infinite, determine F if the overall gain
i. the AND function, is 25. (0'04).
ii. the OR function.
11. (a) Draw a block diagram of a remote position control
7. With the aid of a diagram describe the construction and servo-mechanism suitable for controlling the angular position of
operating principle of a valve positioner. the ship's rudder. Clearly label the inputs and outputs for each
Give three reasons why such a device may be utilised in a control block and use them to explain what is meant by proportional
system. control.
Explain how a positioner may be adjusted to alter the valve (b) Explain with the aid of diagrams why damping needs to
stroke from 1·2-1·8 barto 1·3-2·0 bar and state why gain should be introduced into the system, when a step change of input is
be as high as possible. applied, and state one method of introducing damping into such
a system.
8. Explain the principles of viscosity measurement and detail
the type of instrumentation used.
12. (a) Explain the meaning of the following terms, using suit-
Sketch an oil viscosity control system and describe the operating able diagrams where appropriate:
principle.
i. distance-velocity lag,
Give reasons for which control actions you would incorporate in ii. transfer lag,
the controller.
iii. time constant,
iv. thermal capacity.
9. Describe with a block diagram the operational construction (b) A thermometer bulb is housed in a pocket. Show by
of a data logger suitable for marine use. means of a response curve an estimation of the effect of the
Clearly indicate the functions of alarm annunciation recording of pocket and state what factors influence the design of the pocket
input signal information, and analogue to digital conversion. to minimise this effect.
344 REED'S INSTRUMENTATION AND CONTROL SYSTEMS
SPECIMEN QUESTIONS 345

13. (a) Briefly describe why it is desirable to employ a multi-

15. The air pressure control system shown in Fig. M has been
element system in the control of boiler water level.
adjusted such that a load of 6 kg/min the equilibrium value of the
(b) Make a diagrammatic sketch of a three-element boiler controlled condition is at the desired value of 30 bar.
water level control system, naming all the parts and giving a brief
description of its operation.
Determine:-
(c) Sketch a diagram showing the variations in boiler water
level and feed pump load that might be expected if a water tube i. the proportional control factor (J-t} of the system if
boiler is subject to a step increase in load, with: the proportional band of the controller is 50%, (0'5),
i. single-element control, ii. the offset of the load changes to 8 kg/min, ( - 4 bar),
ii. three-element control.
Ill. the proportional band setting required such that
for a load change from 6 kg/min to 2 kg/min the
14. (a) Fig. L shows the block diagram of a low power remote offset is limited to 2 bar, (12'5%),
position control servo system with velocity feedback damping.
State the function of each component in the system. iv. the smallest load that the system can deal with if
(b) Explain briefly how the system will respond to a step the proportional band of the controller is set to
input and sketch the corresponding response curve for: 100%, (3,5 kg/min),
i. switch S open,
ii. switch S closed. v. the proportional band of the controller such that
the system can deal with any load within the range
0-10 kg/min. (41'7%).

PRESSURE TRANSMITTER
(OIRECT ACTING)
RANGE 0 -40 bar
OUTPUT 20-100 kN/m 2
SERVO POWER
AMPLIFIER SUPPLY
PNEUMATIC (P)
AIR CONTROLLER
RECEIVER INPUT AND OUTPUT
RANGE 20-100 kN/m2

REGULATING UNIT (AIR TO OPEN)

.--
LOAD
RANGE 0 -10 kg Imin
INPUT 20 -100 kN/m 2

Fig.M.

Fig. L
346 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 347

16. (a) Sketch the forward characteristic for a thyristor and 18. Fig. 0 shows the arrangement of a damper position control
indicate clearly: system for a boiler employing damper control of the final steam
i. the hold current level, temperature, and the variation in steam temperature, during a
ii. the effect of increasing the gate trigger pulse upon transient condition. Draw an accurate diagram showing the
the forward break-over voltage. damper position during the same transient period. The damper
may be assumed to be in the 60% open position immediately
(b) State three advantages of a thyristor compared to other before the transient condition commences. (System proportional
types of controlled rectifier. control factor JL = multiple of proportionality characteristics,
coefficients, i.e. 2 = 2 x 0·8 x 1·25 then use a tabular method;
(c) Fig. N shows the circuit of an ac controller:
at two minutes, for example, using <I> = - JL(O+ liS 10 dt) the
i. explain briefly how the power in the load is control-
damper is 20% open, i.e. -40 = -2 (10 + 1/0·5 x 5). Damper
led,
is closed between 2·5 and 5 minutes).
ii. sketch the wave forms of load voltage and gate DAMPER

~I
pulse for half maximum load power. PNEUMATIC ACTUATOR IAIR-TO-CLOSE)
RANGE: 20-100 kN/m 2
: DAMPER POSITION: '00% OPEN-SHUT
I
R, 6------

f
18kfi

~)
GAS FLOW PNEUMATIC CONTROLLER
1 PROPORTIONAL BAND=125%
I ---+1-----=- INTEGRAL ACTION TIME=30s
50 Hz I

M:J I
ac I

c,
O·lfJF
Zo
SUPERHEATER
TEMPERATURE TRANSMITTER
LOAD
RANGE: 460 -SOooC
OUTPUT: 20-100 kN/m 2

Fig. N f 500

u
~
.-.. a..: 490
17. (a) Define the following terms: ~
I I
w
i. integral action, I- I I
~ I I
ii. derivative action. <{ 480 r--- 1 - - -
~ : I I I
(/) - - - 1" - - - - r - - - IL - - _1- -
I
(b) With the aid of suitable diagrams derive expressions I-

for: ~ 470 ---~- --~ - - i ---i--~-~-

~ I I I I : I :
i. integral action in terms of controller gain and UJ I I I I I I I
(L I I I I I I I
~ 460
integral action time, fJ o 1 2 3
I

4. 5
I

-
ii. derivative action in terms of controller gain and TIME lmin)~

derivative action time. Fig. 0

348 REED'S INSTRUMENTATION AND CONTROL SYSTEMS SPECIMEN QUESTIONS 349

19. (a) Draw a detailed diagram of an instrument suitable for 23. Fig. P shows the main components of a jacket cooling water
converting electrical control signals into pneumatic signals. system for a marine diesel engine, outlet temperature is control-
(b) Explain clearly the operation throughout the instru- led condition.
ment in response to an increased input signal.
(c) Discuss briefly the relative advantages of electro- (a) Assuming that it has been decided to modify the sys-
pneumatic control. tem to provide a split level, two element, cascade closed loop
pneumatic control system, complete the diagram naming clearly
all the additional components required.
20. (a) Draw the circuit for a two input diode-transistor logic (b) Describe the operation of the modified system giving
OR gate suitable for operating an alarm system. Construct the typical operating ranges and stating clearly the modes of action
truth table and describe the operation of the gate.
employed e.g. direct acting proportional.
(b) Construct a logic flow diagram fora sootblowing sys-
tem which incorporates: (c) Describe the effects on the system of:
i. a warming through period, i. an increase in sea water temperature,
ii. retractable blowers which blow in each direction,
ii. a decrease in engine load.
iii. provision to bypass selected blowers,
iv. suitable operation checks.
Show the details for one blower only and draw a block for the
others.
HEADER
TANK

21. (a) Sketch a fully labelled diagram of a two element CW

CONTROL
superheated steam temperature control system which uses a VALVE
three term controller and two control valves, a steam to attem-
perator and an attemperator bypass valve.
(b) Explain the reason for using two element control in this
system and describe the actions throughout the loop following a
load increase.
(c) State the reasons for using integral and derivative MAIN
terms in the controller. ENGINE

22. Show in block diagram form the basic requirement for a JACKET
bridge control system for a main diesel engine. The control WATER
PUMP
programme unit may be shown as a block.
Briefly describe the sequence of events throughout the system
Fig.P
following movement of the bridge telegraph from stop to full
ahead. What emergency arrangements must be incorporated in
the system and fitted on the bridge?
350 REED'S INSTRUMENTATION AND CONTROL SYSTEMS

+~
lD u
0::
W ~"U
I- en
W
~

°z
~

W
I-

°a.. INDEX

~
OIl .,.. I I I
C'.,
• A Arithmetic unit 303
a.. ac electric servo-motor 121,245 Atomic theory 89
0- I ~ rn"\u..~ I I ( I Z~I \ '.t
::::> " position control 244 Attenuation 134,173
en " speed control 247 Auto-combustion/ attemperator
u
0 Across variable 257 control 224
Action, control types 136, 151 Automatic controller 130
derivative control 147, 149 Auto-steering
integral control 143, 146 (block diagram) 236
proportional control 137,141 Auxiliary boiler control 226
P+D 148, 150 Averaging 194
P+I 143, 147 B
~ ~ .~ ] .9 -g
~
time 144, 148 Bailey controllers 199
rJ"loCl~ "'0 ~ two-step control 136,157
u~ I (]) •. 5 ~ I-<
Bandwidth 139
Actual controller types 185 Barometer, aneroid 26
0«
~3~(])
I-<~"t)~
~
<+-0
&
0..
Actuated torque valve 118
" mercury 25
Adjustment of controllers 276
uo::
0::
° ~~S~
....
~ ,gu " " instruments 73
Basic control actions
Bi-metallic thermometer
136
12
00::
W c(]) & ou
e .....~ . c. .
0..
(]).
. . (]) I-<
Air-fuel ratio control 201 Binary arithmetic 301
" -steam rate control 207
~
«
Uo::
...l
e=' 2(])~
~
Cc ""-'0£)0
>.~~
~ ~ .....
" supplies 168
Black box
Blind controller
8,180,256
200
~ ZW I-< (]) 0 0 ~ rJ"l en Alarm circuit, fire 66 Block diagrams 236,240,243
1/)0::«0- ~.do..~ Q~C Algorithm 287
zOIO- C ~ O£)"'O =' ~
rJ"l Blocking oscillator 111
231
~ ii ~u
frl~ .- "'0 C C I-<o...c Alternator control Bode diagrams 270
C I-< ..... 0 0 C Amplification 134,158
I-
~OUo.. ~ ..... ~ Boiler oil purification
Amplifiers 97,171
OO-J ~Io::
<+-0
U C en
O(])~(])
I-< 'a
~ 0 .D S
I-<
.- (]) 0..
rJ"l.dS
C ~
~ I-< .~
Analogies
Analogue
142, 164
8,281
system
Boolean algebra
216
288
Wi
LiJ ~~::::~ . . ..s~ circuits 89
Bourdon gauge, pressure
Bridge control, IC engine
26
218
z z[ e "'0~enc
(]) ~ >. I-< I-<
.d(])O
computer 9,281 " turbine 208
+b~ W I~ ~ =' ~ (]) ~ 0..<+-0 C -digital converter 305
°Zi:$~
:::.c
U ....
oE (])
(])
I <+-0
0
S
='
cU ~ 0.. .~ .9
I-< ~ 0 =' ~
simulation
tachometer
282,286
55
rectifier
BS terminology
92
127
...l ~QI
Z
::::> °
> 0::
W
.dl-<enl-<='O.du~
~ 2 C ~ ~ ~ u.= ~
Analogues 257,273
Butterworth heating control 233
I Analysis, carbon-dioxide 70 C
t:. ;;>
o
e S(]).....(]) 0..(])
:>(])~CI-<(])'aUo..
I-< U (]) 0
......... control system 255 Calibration of instruments 73
.dO .dSOC~rJ"l dissolved oxygen 72 Capacitance transducer 81
rJ"l',c >. ~ ~ ~ .~
feedback 107 Capacitive level sensor 34
O.~(])O(]»(])"t)
c.D:>2 .... .d.d(])
(]) u u .D
;;> (]) frequency response 269 Carrier wave 93
a ....
m .::!J
0 o...d 0...-
S ~ '1-<
~ o..=, >.=,tJ I en ~
oxygen
stability
68
270
Cascade control system
Cathode ray oscilloscope
132,152,214
112
u:: O£)ot+:lo 0 (]) water 71 Cell, differential pressure 28
-CU(])U-.c-"'O
~..... O'C 0.D u u >. AND gate 289 " photo 60
--g ~
S.D S--~--t+:l
I-< en I-< (]) (])
~.,,:
Aneroid barometer 26 " ,,-transistor 61
Annunciator systems 306 Characteristics, control valve 116
~ 'a (]) ~ (]) C 'C
~ Argand diagram 267 pn junction 94
N.D-5u-5 .D
•
 352
353
Characteristics, resistance-temperature 15 Control valve (diaphragm) 115
thyristor 94, 96 Detector, level (capacitive) 34 Electron conduction 90
(piston) 118 " (resistive)
zener diode 94 36 Electronic control principles 171
(rotating cylinder) 119 oil in water 34,62,234
Chopper circuit 106 techniques 172
(torque) 118 oil mist
Circuit configurations, transistor 100 63 controller types 188,190
(wax element) 120 oxygen
fire alarm-detection 66 72 controllers 171, 188
Controllable pitch propeller 235 pH
Circuits, logic 287 Controlled condition 129 64 one term 173, 174
Circulating water control loop 205,212 piezoelectric 29 three term 177
Controller action P 141, 160, 173 pressure
Classification of amplifiers 98 32 two term 175
Closed loop system 7, 108
P+D 148,165,175 relative humidity 71
P +I 145, 162, 186 devices 89
" "transfer function 108, 261 smoke density 62 force-balance 82
types 185 speed (tachogenerator)
CO2 analyser-recorder 70 Controllers, electronic one term 173 55 thermocouple 19
Collector characteristics (CE) 102 4 strain gauge 31 two step control 172
three term 177 temperature 11
Combustion/attemperator control 224 Element 5,42,115,128
two term 175 torque 58
Command signal 128 Empirical controller settings 278
Common base transistor
collector transistor
100
100
master
pneumatic, two step
153,214,234
157 . Deviation
viscosity 59
128
Engine logic system
Equivalent T-circuit
298
103
one term 160 Diaphragm control valve 115
emitter transistor 101 Error 175
Comparing element 128 three term 167 Differential capacitor transducer 82
two term 162, 165 Evaporator flow control 217
Comparison telemetering systems 6 equations 259,263 Examination questions, specimen 313
setting, adjustment 277 inductor transducer
Complex numbers 267 80 Example, level control 138,152,153
slave 153,214,234 pressure cell 28
Component adjustment 276 Exhaust range pressure control 203
stack type 161 Digital
"interaction 167,275 9 Explosion meter detector 67
Controlling elements 115 circuits
Compound controller P + 1+ D 167,177 Converted command signal 128
89 Extraction, square root 42,49
Computer, analogue 281 computer 9,300,304 Extrinsic semi-conductors 90
controlled condition 130 instrument
control 309 305
deviation 129 logic
digital
simulation
300, 304
310
Conversion, decimal-binary 300 ti tachometer
288
56
F
Converter, analogue-digital 305 Feedback 107
Computers 9 Diode gate, AND, OR 291 amplifier 107
" pneumatic-electronic 82
Computing and logic 281 Coolant control, IC engine 212 ... semi-conductor
zener
94 analysis 107
Condenser water control 205 95 negative 107, 158, 160
Correcting elements 115,129 Dionic water purity meter
Conduction electrons 290 71 positive 109
"units 115, 129 Direct digital control 309
" holes 290 Cybernetics 2 pneumatic 158
Configurations (transistor) 100 level measure 33 stabilisation 242, 271
Control action graphs 151 Displacement-pneumatic transducer 77 Feedforward 244
types 136 Dissolved oxygen meter 72 Feed regulators, level 227
D
cascade 132, 152 Distance-velocity lag 135 Field effect transistor 106
Damping 242, 266 Divisor 181, 194
circuits, general 224 Data logging 306 Filled system thermometers 11
motor (IC engine) 212 DoT specimen examination Filter circuit 93

J
" processing 282,306 questions 313
steam 203 dc amplifier 106 Final controlling elements 115
factors 138, 144, 147 Drayton controller 197 Fire detection-alarm circuit 66
" electric servo-motor 121 Driver amplifier
level, example 138,152, 153 " position control 239 105 " (heat) detector 64
loops 7,142 Drum level controller 227 First order equation 258
" speed control 246
position servo 239, 245, 263 Dead time 132 " "system 258
speed servo 246 Decimal-binary conversion 300
et E Flame detector (sensor) 67
split range 153,204 Economy 2 Flapper and nozzle 67, 76, 157
Demodulation circuit 93 Electrical flowmeter 45
system, analysis 255 Derivative action 145,147,165,175 Flash evaporator control 217
cascade 132, 152 measurements 55 Flip-flop oscillator 297
factor 147 square root
kinetic 239 time 148 Flow measurement 41
process 127 extractor 52 Flowmeter, electrical 45
Desired value 128 telegraph 86
technique 157,172 Detecting element, response electro-magnetic 46
theory 127 133 thermometry 13
" ,,(sensor) 5,129,133 mechanical 43
three element 207, 228 transmitters 79 Flow sensors 48,53
Detector-alarm circuit, fire 66 Electroflo level transducer 35
term 150,167,177 Detector, CO2 Fluidic logic 299
transistor stabilisation 182 70 Electro-hydraulic speed control 252
flame
I

Force-balance system (electronic) 82

two element 213,228 67 ",:J -magnetic flowmeter 46
flow 48 " transducer (pneumatic) 77
"term 162, 165, 175 " radiation spectrum 113 Foxboro controllers 191, 197
gas explosion (meter) 67 -pneumatic controller 187
Control unit 303 heat (fire) Frequency response analysis 269
64 ~ transducer 78 Friction damping 242
 354 355

Fuel-air ratio control 200 Input element 128 Magnitude ratio 268,272 Nichols chart diagram 270,273
" oil control valve 116 ramp signal 33,241,266 Manoeuvring relays 220 Non-linear mode amplifier 99
Further analogues 273 step signal 241,265 Manometer, inclined tube 24 NOR gate 289,293
unit 303 mercury 24 NOT gate 289,292
G Instrument calibration, testing, water 23 np junction 91
Gain 134 adjustment 73 Manufacturing system 275 npn junction 91
" amplifiers 108 Instrument, digital 305 Master controller 153,214 n type material 90
" margin 271 Integral action control 143 Mathematical models 282,310 Null balance potentiometer 85
Gas analysis 67 factor 144 Measurement of electricity 55 Nyquist diagram 270
" explosion meter detector 67 time 144,146 " flow 41 stability criteria 271
" filled thermometer 11 control action 143 " gas 67
Gates, logic 290 factor 144 " level 33 o
Gauge, pressure 26 Integrated system 308 " oil in water 34,62 Offset 132,141,244
" strain 31 Integrators 42, 182 " oil mist 63 Oil mist detector 63
Generalised impedance- 257 Interaction 167,275 " pH 64 " purification system 216
Gland steam control 204 Interface, oil-water 34,62,234 " pressure 23 " -water detector (sensor) 34,62
Governor 185 Intrinsic semi-conductors 90 " smoke density 62 interface 34,62,234
electro-hydraulic 252 Inverter 181 " speed 55 control 234
mechanical-hydraulic 186,250 " temperature 11 separator 234
systems 249 J " torque-power 57 Open loop system 7,261,270
Graphs, control action 151 Jacket water control, single " viscosity 59 " "transfer function 261
element 212 Measuring element 5, 129 Operational amplifier 107,171,180
H Jacket water control, two element 213 " unit 5 Optical pyrometer 20
Hagan controllers 193 Junction transistor amplifier 99, 104 Mechanical controller 185 OR gate 289,291
Harmonic oscillator 110 flowmeter 43 Order of system 257,258
response 269 K governor 186 Orifice sensor 48
Heat (fire) detector 64 Kenotometer 25 -hydraulic speed· Oscillators 109
Higher order systems 264 Kent controller 188 control 186, 250 Oscilloscope, cathode ray 112
TEC/SCOTEC specimen Kinetic control systems 239 square root extractor 50 Other devices, electronic 112
examination questions 332,341 thermometry 11 measurements 55
Historical 1 L Memory circuit 295 servo-motors 126
Hold (memory) circuit 295 Lag 271 unit 303 Output unit 304
Human control loop 143 Level control, cascade 132,152,214 Mercury barometer 25 Overall coolant systems 214
Hybrid computer 9 example 138,152, 153 in steel thermometer 12 Overlap 137
Hydraulic-mechanical governor 186 oil-water separator 234 manometer 24 Overshoot 132,242
position system 245 regulator 227 Meter, carbon dioxide 70 Oxygen analyser 68
pumps 122 indicator, remote 37,39 dissolved oxygen 72 Oxygen (dissolved) meter 72
ram servo 122 measurement 33 explosive gas 67
rotary vane servo 123 " sensor, capacitive 34 quantity 41 p
servo 122 " resistive 36 rate of flow 41 Parallel-series cooling system 216
variable delivery pump 122 simulation-control 310 water purity 71 Parameters, transistor 103
Linear mode amplifiers 99, 104 Microprocessors 310 Pen recorder 85
I Liquid in glass thermometer 11 Mimic diagrams 309 Peripheral units 304
IC engine control circuits 212 " level control 138, 142, 146, 149 Misalignment, velocity 242, 244 pH sensor 64
oil purification 216 Linear vibration damper 286 Models 282 Phase advance network 271
remote control 218,222 Load 130 Modulation 93 lag 134,269,271
Igema gauge 37 Logic and computing 281 Monitoring element 5,130 margin 271
Impedance, generalised 257 circuits 287 feedback 130 Philips controller 187, 190
Inclined tube manometer 24 engine control 298 Motor element 115, 129 Photo-electric cells 60
Index 351 flow chart 287 position 86, 240, 245 " -transistor 61
Inductance balance 81 fluidic 299 servo ac 121,245,247 Piezoelectric sensor 29
ratio 81 solid state 290 dc 121,240,263 Piston control valve 118
transducer 80 switching 288 Multiplier 181, 194 water temperature
Inferential-differential pressure 47 symbols 290 Multivibrator 112,297 control 212,213
measures 35,41 Loops, control 7, 108,213,261 p type material 90
rotational 43 N pn junction 91,94
Infra-red flame detector 67 M NAND gate 289,295 Pneumaticcontrolprincip~s 157
Inherent regulation 137 Magnetic amplifier 97 Negative feedback 1W, 158, 160 p 159,161
356 357

Rotameter 46 Signal, processing 129,309

Pneumatic control P + D 165 Proportional control factor 138 Rotary cylinder valve 119 " selection 309
P +I 162 P+ I action 143,144,162 " vane hydraulic servo 123 Silicon controlled rectifier 96
P + 1+ D 167 P + D action 148, 150, 165 Rotating electrical amplifiers 97 Simple pendulum analogue simulation 282
techniques 157 P + I + D action 150,167,177 Rotor meter 43 " "computer 285
controller types 193,197,199 Pump, variable delivery 122 Rototherm 12 Simulation, computer 310
one term 160, 161 Purge level system 39 Single element control 212,228
three term 167 Purification control system 216 S " term control 160, 173
two term 162, 165 Push-pull power amplifier 105 Safety requirements 2 Sinusoidal response 134, 178,267
P 160, 161 " blocking oscillator 111 Scalar multiplier 173 Slave controller 152,214
P + D 165 Pyrometer 14,20 Scanner 305 Small signal amplifier 104
P +I 162
Schaffer pressure gauge 27 Smoke density detector 62
P + 1+ D 167
Q SCOTEC/TEC specimen examination Smoothing circuit 93
Pneumatic-current transducer 82
Quantity meters 41 questions 328 Solid state devices 91
-electronic transducer 82
Questions, specimen examination 313 Second order equations 263 Soot blower control 205
force-balance transducer 77
systems 258,262 Specimen examination questions
piston servo-positioner 123
Secondary elements 42,47 DoT Class One 321
position-balance transducer 76 R Self balancing bridges 14 Specimen examination questions
relay 158 Radiation pyrometry 20 " " potentiometers 85 DoT Oass Three 313
remote control system 223 spectrum 113 Semi-conductor diode 94 Specimen examination questions
square root extractor 51 Radio communication 113 Semi-conductors 89 DoT Class Two 315
stack controller 161 Ram, hydraulic servo 122 Sensor, CO 2 70 Specimen examination questions
standardising relay 199 Ramp input response 133,241,266
transmitters 76
(detecting element) 5 Higher; TEC/SCOTEC (and
Rate 165,175 flame 67 METe) 332
two-step technique 157 " air-fuel 201 flow 48,53 Specimen examination questions
Pneumercator 39 " -steam 207 gas explosion (meter) 67 Higher; TEC/SCOTEC (and
pnp junction 91 " of flow meter 41 heat (fire) 64 OND supp.) 341
" "transistor 99 " network 243 level (capacitive) 34 Specimen examination questions
Position-balance ac motor system 87 Receiver integrator 84 (resistive) 36 TEC/SCOTEC (and METC, OND) 328
dc motor system 86 Receivers 5,84 oil in water 34,62,234 Spectrum electromagnetic 113
indicator (telegraph) 86 Recorder, pen 85 " mist 63 Speed control 246
transducer (pneumatic)76 XY 86 pH 64 elec-hydraulic 252
control 239 Rectifiers 92 piezoelectric 29 mech-hydraulic 250
motor ac 87, 244 Reference sets 305 pressure 32 thyristor 248
dc 86,239 Refrigeration control loops 231 relative humidity 71 Ward Leonard 247
Positioner 116, 125 Relative humidity detector 71 smoke density 62 measurement (tachogenerator) 55
Positive feedback 109 Relaxation oscillator 111
Potentiometer circuit 85
speed (tachogenerator) 55 Split range control 153,204,214
Relay, manoeuvring 220 strain gauge 31 Spring-mass-damper analogue
Potentiometric pen recorder 85 " pneumatic 158 temperature 11 simulation 286
Power amplifier 105 Remote control, IC engine 222 torque 58 Square root extraction 42, 49
measurement 57 transmission systems 6,75 viscosity 59 " extractor, electrical 52
Pressure control 204 water level indicator 37 Series-parallel cooling loop 216 mechanical 50
-current transducer 32, 82 Reset 162, 174 Servo-governor 249 pneumatic 51
elements 35 network 244 electro-hydraulic 252 Stabilising feedback 242,271
gauge, Bourdon 26 feedforward 244 mechanical-hydraulic 250 Stability response criteria 270,271
Schaffer 27 Resistance level sensor 36 mechanism 239 Stack controller 161, 199
measurement 23 thermometer 13 motor ac and dc 121,240,245 Standardising relay (Bailey) 199
sensor 29, 32 -temperature characteristics 15 position control ac 244 Steam-air flow rate control 207
Primary elements 42,47 transducer 79 " "dc 239,263 temperature control 206
Probe, level 33 Response analysis 269 pneumatic piston 123 Steering control 236
Process 132 curves 133,241, 266, 270 ram hydraulic 122 Step input response 241,265
control system 127 detecting elements 132 rotary vane hydraulic 123 Strain gauge 31
" theory 127 Nyquist 270, 272 speed control ac 247 Study programme 3
Processor 308 sinusoidal input 134, 178, 267 dc 246 Subsidence ratio 141
Propeller pitch control 235 stability criteria 270 Set value 128 Subtractor 182,194
Proportional action 137, 142, 160, 173 step input 241,265 Settling time 132 Summer 174,181,194
" factor 137 Ring main cooling system 214 SHM analogue circuit diagram 286 Supplies, air 168
band 139 " torductor 58 Switching circuit logic 288
Signal, command 128
control action 137, 141, 159 Root extractors 42,49
 358 359

Synchro system 87 Transducers, variable resistance 79 Vane servo 123 Water analysis 71
System 8, 255 " "voltage-current 83 37
analysis 25~282 Transducers 5, 75
Vapour pressure thermometer 12 " level indicator, remote
Variable capacitance transducer 81 " manometer 23
equations 259, 263 Transfer characteristics (transistor) 101 71
first order 258 function 8, 260, 268
delivery pump 122 " purity meter
inductance transducer 80 " temperature control 205,212
manufacturing 275 block diagram 262 120
resistance " 79 Wax element control valve
order 257, 258 lag 135 24
Velocity misalignment 242,244 Well type manometer
performance 265 Transformer rectifier (centre tap) 93 13,32,80
Venturi sensor 48 Wheatstone bridge circuit
response 241 Transient response 265
second order 258, 262 Viscometer 59
Transistor 94 229
state 255 Viscosity control
amplifier (small signal) 104 controller (VAF) 209 X
Systems approach 255,256 " , voltage 102 X plot 282
comparison, telemetering 6 sensor 60 86
characteristics 102 XY recorder
Voltage amplifier transistor 102
configurations 100 -current transducer 83
T field effect 106
Tachogenerator stabilisation 95,182 y
55 junction 99
Tachometer Yplot 282
55 oscillator 110
Tachometer, digital W
56 parameters 103 247
T-circuit, transistor Ward Leonard speed control
103 photo-cell 61 Z
Technique, control Waste heat flash evaporator
157, 172 stabilisation control 182 217 Zener diode 95
TEC/SCOTEC specimen control
T-circuit 103
examination questions 328 unijunetion 103
Telegraph 86 voltage amplifier 102
Telemetering 5, 75 Transmitters 5, 76, 79
" systems, comparison 6 Trigger pulse 95
Temperature measurement 11 Truth tables 289
" control loops 205,207,212 Turn down ratio 224
Terminology BSI 5, 127 Two element control 213,228
Testing, instruments 73 " step action 136,157,172
Thermal conductivity CO 2 recorder 70 " term control 162,165,175
Thermistor 15 Types of control action 136
Thermocouple 17 " "controller 185
" electronic 19 Typical control circuits
Thermometer, bi-metallic 12 General 224
filled system 11 IC Engine 212
liquid in glass 11 Steam 203
resistance 13
Thermometry U
11,13
Three element control Unbonded strain gauge 31
207,228
term control Unijunction transistor 103
150,167,177
Through variable Units, correcting 115, 129
257
Thyristor " measuring 130
96 Univibrator 112
" speed control 248
Time constant 132, 134 Unmanned machinery spaces (UMS) 4, 210
delay Utilisation 2
293
derivative action U tube manometer 23
148
integral action 144 v
lags 134 VAF viscotherm controller 201
Torductor 57 Value, desired 128
Torque actuated valve 118 " set 128
" measurement 57 Valve, diaphragm actuated 115
Transducers, electromagnetic 78 piston actuated 118
flow sensor 53 pneumatic actuated 116
force-balance 77 positioner actuated 116
position-balance 76 relay actuated 158
pressure-current 82 rotating cylinder actuated 119
variable capacitance 81 torque actuated 118
inductance 80 wax element actuated 120
 REED'S MARINE ENGINEERING SERIES

The series covers the full range of subjects for all grades of the
Department of Transport Certificates of Competency in Marine
Engineering. The books will be extremely useful to marine
engineer cadets and other engineering students studying on
Technician Education Council (and SCOTEC) Courses. Materi~1 is
presented from first principles with many diagrammatic sketches
and worked solutions to examples.

Vol. 1. MATHEMATICS
Vol. 2. APPLIED MECHANICS
Vol. 3. HEAT & HEAT ENGINES
Vol. 4. NAVAL ARCHITECTURE
Vol. 5. SHIP CONSTRUCTION
Vol. 6. BASIC ELECTROTECHNOLOGY
Vol. 7. ADVANCED ELECTROTECHNOLOGY
Vol. 8. GENERAL ENGINEERING KNOWLEDGE
Vol. 9. STEAM ENGINEERING KNOWLEDGE
Vol. 10. INSTRUMENTATION & CONTROL SYSTEMS
Vol. 11. ENGINEERING DRAWING
Vol. 12. MOTOR ENGINEERING KNOWLEDGE

These books a,e obtainable f,om all Nautical Opticians,

Chartselle,s and Bookselle,s and a,e published by:

THOMAS REED PUBLICATIONS LIMITED