TECHNOLOG

D C GREEN

aE [Rojetepastobel
mmm Scientific&)
om |Technical
RENEWALS
You may phone to renew this book
twice on 01429 857173. Please
note we are unable to renew books
which are overdue or reserved by
another borrower.
 Radio sys.ermis Te_nanology

URWy
2 &

BeiyRa
7, An
éSap,
‘oS,
f
Ny nh-O,
Radio Systems
Technology

D C Green
MTech, CEng, MIEE

> M>| Longman

eam scientific &
Technical
Copublished in the United States with
ons ., New Yor
Wace les

Longman Scientific & Technical,

Longman Group UK Limited,
Longman House, Burnt Mill, Harlow,
Essex CM20 2JE, England
and Associated Companies throughout the world

Copublished in the United States with

John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158

© Longman Group UK Limited 1990

All rights reserved; no part of this publication

may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording, or otherwise
without either the prior written permission of the Publishers or a
licence permitting restricted copying in the United Kingdom issued by
the Copyright Licensing Agency Ltd, 90 Tottenham Court Road,
London, W1P 9HE.

First published 1990

Second impression 1992

British Library Cataloguing in Publication Data

Green, D.C. (Derek Charles), 193] —
Radio systems technology.
1. Radio engineering
I. Title
621.384

ISBN O-58e-O0eb9?-0

Library of Congress Cataloging-in-Publication Data

Green, D. C. (Derek Charles)
Radio systems technology/D.C. Green.
De en:
ISBN 0-470-21672-7
1. Radio. _I. Title.
TK6550.G84 1990
621.384—dc20 90-5627
CIP

ISBN 0-470-21672-7 (USA only)

Produced by Longman Singapore Publishers (Pte) Ltd.

Printed in Singapore
Contents

Preface vii

1 Amplitude Modulation 1
Double-sideband amplitude modulation 1
Amplitude modulators 9
Detection of an amplitude-modulated wave 11
Double-sideband suppressed-carrier amplitude modulation 17
Single-sideband suppressed-carrier amplitude modulation 20
Single-sideband compared with double-sideband amplitude
modulation 22

2 Angle Modulation 24
Frequency modulation 24
Phase modulation 34
Signal-to-noise ratio in an f.m. system 36
Frequency modulators 42
Phase modulators 47
Frequency-modulation detectors 49

3 Transmission Lines 55
Mismatched transmission lines 56
Transmission lines as components 67
Matching 70
The Smith chart 71

4 Waveguides 85
Propagation in the rectangular waveguide 86
Impedance of a rectangular waveguide 93
Attenuation in a rectangular waveguide 93

5 Noise in Radio Systems 96

Sources of noise in radio systems 96
Signal-to-noise ratio 103
Noise factor 104
Noise temperature 108

6 Principles of Aerials 111

Radiation from an aerial 111
Effective length of an aerial 115
The monopole aerial 116
The /2 dipole 123
 Gain of an aerial 125
Effective aperture of an aerial 126
Long-wire radiator 128

Aerials 131
Arrays of driven dipoles 131
Height factor 142
Pattern multiplication 146
Mutual impedances between dipoles 149
The rhombic aerial 151
The log-periodic aerial 151
The Yagi aerial 155
The parabolic dish aerial 157

Propagation of Radio Waves 163

The ionosphere 163
The troposphere 166
Ground-wave propagation 169
Sky-wave propagation 171
Space-wave propagation 177
Scatter propagation 185
Propagation via a communications satellite 187

Communication Radio Receivers 189

Double-superheterodyne radio receivers 189
Intermodulation 192
Reciprocal mixing 197
Cross modulation 198
Blocking 199
Sensitivity 200
Selectivity 201
Noise factor 202
Stages in a radio receiver 203
Automatic gain control 210
Communication receivers 212

10 Radio Systems 216

Microwave radio-relay systems 216
Communications satellite systems 222
Noise performance of analogue microwave links 229
Bit error rate in digital systems 235
Land-mobile radio systems 235

Exercises 240

Answers to numerical exercises 247

Index 248
Preface

This book has been written to provide a comprehensive introduction

to modern radiocommunication systems. An attempt has been made
throughout to keep the treatment of each topic at a level appropriate
for a UK Higher Certificate/Diploma student or for an early-year
degree student. This means that the reader is assumed to possess a
prior knowledge of electrical principles, electronics, mathematics and
radio/transmission principles of at least the standard reached by the
Level III Business and Technician Education Council (BTEC) units.
The book discusses the principles of analogue modulation in
Chapters 1 and 2; but gives only a brief mention of digital modula-
tion since, at present, its radio applications are restricted to some
microwave systems. Transmission lines and waveguides are important
components in any radio system since they are used both to provide
feeders and to simulate components and tuned circuits. Lines and
waveguides are the subjects of Chapters 3 and 4. The important topic
of noise is considered in Chapter 5 before the next two chapters deal
with aerials. The basic concepts of aerials are introduced to the reader
in Chapter 6 before some of the more commonly employed aerials
are discussed in Chapter 7. Chapter 8 deals with the propagation of
radio waves from one aerial to another. The treatment of communica-
tion radio receivers, in Chapter 9, assumes throughout a previous
knowledge of radio receivers up to the BTEC Level III standard. No
circuits are given because of demands on space, but a mention is made
of some of the Plessey ICs that can be usefully employed in modern
receivers. Lastly, Chapter 10 provides an introduction to microwave
radio-relay systems, both terrestrial and satellite, and to land-mobile
systems.
Many worked examples have been provided throughout the book
to illustrate various principles and at the end of the book will be found
exercises together with answers to the numerical exercises.
I wish to express my thanks and gratitude to Eddystone Radio Ltd
for their permission to include the block diagrams of their 1650 and
1995 radio receivers, and to the Institution of British Telecommunica-
tion Engineers for their permission to use several diagrams from their
journal British Telecommunication Engineering.
DCG
 Lae es a e i ae

; .
.

i
fe
y
oy

suiubemn c1ahoaigdo2 & abiven 7a oot aot

sincin 5thasl narafhe ROLES aRtReoRE

| tee
ase Saalae a
&Fangs oan ruiiedao’ be @we

ft bosslasr ‘oe sewvitinlys poteai 0988 ae

o ote zahagovew bina sont noleaiterenT area,MENGE
o4 Moddeat ou, aeSante migieyE ibey yote a et a
iataesd baad tata ange uel WED Ie ol Testag ¥3
qmbodT + tins f > aireTo opine St A CAB say
peta emer sche Sot saoied & rsiyad> ot boaeeeies af aoe
Lobe subir re-visition teat OGHOD sion
2 ATTY, alae oe i
Ges dhomvt.: arom nse
- smtitie sveiad oy
pepsi
ory altshiw-cined 2 sarpal’d 0ita ob eke
palo Hpoeniieaas iT — br ithiay Seo me aqynw
® 1065basin? esreuesn2 COE. On perv ;
eHwes a = 1aae eb aw ereviien oihest

: a pniienia: fess rg 8 9 _ ‘obainicas ;

wovepe seed acttoton niet lesson ae ae
36pay)DS State
5 out ms battens ite adi
i 4a f
ns OF vodi te
» with,
meanttata out abst onia cies b
lepavng sau ol semisttiie sige 1
apeenial molt maura ¢
 Amplitude Modulation

Some form of modulation is always employed in a radio system to

frequency shift, or translate, a baseband signal from its original
frequency bandwidth to a specified part of the radio-frequency
spectrum. The various radio services have each been allocated
particular frequency bands by international agreement (see Transmis-
sion Principles for Technicians). Different stations operating in an
allocated frequency band must, of course, be positioned at different
points within that band so that a wanted signal can be selected from
all those signals that are simultaneously present at the receiving aerial.
Analogue modulation consists essentially of using the baseband,
or modulating, signal to vary one of the three variables of a sinusoidal
carrier wave. If the instantaneous voltage v, of the carrier wave is
ve = Vz sin (wet + 0) (1.1)
then: (a) the amplitude V, of the carrier may be varied, this is
amplitude modulation; (b) the frequency w,/277may be varied, this
is frequency modulation; and (c) the phase 6 can be varied to give
phase modulation. It will be seen in Chapter 2 that frequency and
phase modulation are very similar processes and they are often referred
to jointly as angle modulation. Amplitude modulation is still the most
widely used modulation method with both double-sideband and single-
sideband versions being common.

Double-sideband Amplitude For a sinusoidal carrier wave to be amplitude modulated the amplitude
Modulation of the carrier must be varied in the same way as the instantaneous
voltage of the modulating signal. If the modulating voltage is repre-
sented by V,,(t) then the instantaneous voltage of the modulated
carrier wave is
Ve = Ve + Vm(t)] sin wet (1.2)
in which the phase angle @ of the unmodulated carrier is assumed,
for convenience, to be zero.
When the modulating signal is a sine wave, i.e. V(t) = Vm sin
Wt, then equation (1.2) becomes

ve = [Ve + Vin SiN yt] SIN wet. (1.3)

This equation can be expanded, using the trigonometric identity

2 sin A sin B = cos (A—B) — cos (A + B), to give
2 RADIO SYSTEMS TECHNOLOGY

Fig. 1.1 Amplitude-modulation a <—_- —>

sidebands. Lower sideband Upper sideband

ve = V, sin wet + V/2 cos (we — Wm)t

eV ,,/2 COS. (@, + Wa)t (1.4)
Clearly, the sinusoidally modulated carrier wave contains components
at three different frequencies: (a) the lower side-frequency f. — fy,
(D) the carrier frequency f,, and (c) the upper side-frequency f, +
Jm- The modulating signal frequency f,, is not present. When the
modulating signal V,,(t) is not of sinusoidal waveform Fourier
analysis will show that it contains the sum of components at a funda-
mental frequency, plus one, or more, components at other, higher,
frequencies. Each of these component frequencies will modulate the
carrier to produce corresponding lower and upper side-frequencies
in the modulated waveform. Suppose, for example, that V,,(t) =
V, sinw,;t + V2 sin wt + V3 sin w;t. Then, the instantaneous
voltage of the modulated wave is
ve = Ve + Vy sin wt + V2 sin wt + V3 sin w3t] sin wet.
Cf)
Equation (1.5) can also be expanded to give
Ve = Ve sin wt + V;/2 cos (w, — w)t
+ V,/2 cos (we — w)t
+ V3/2 cos (w. — w3)t
— V,/2 cos (@, + @,)t
= V,/2 COs (W sie W)t
— V3/2 cos (a, + w3)t (1.5a)
which shows that the modulated wave now contains both lower and
upper sidebands that are symmetrically situated either side of the
carrier frequency. This is illustrated by Fig. 1.1.

Modulation Factor

The modulation factor m of an amplitude-modulated wave expresses

the degree to which the amplitude of the carrier is varied from its
unmodulated value. It is given by equation (1.6), i.e.

m= r.m.s. value of V,,(t)

—; (1.6)
r.m.s. value of unmodulated carrier
 AMPLITUDE MODULATION 3

When m is expressed as a percentage it is generally known as the

depth of modulation. The depth of modulation must never be allowed
to become greater than 100% because this situation would result in
excessive distortion of the modulation envelope. For a sinusoidally
modulated wave
m = (Vl J2)/(VolJ2) = Vmn/Ve, or from equation (1.3)

m =>
en ak ie = Ve)
V, (V, + Vin) is (V, e Vin)

_ maximum voltage — minimum voltage (1.7)

maximum voltage + minimum voltage po

Example 1.1

A carrier wave v, = 10 sin (8 X 10°) is amplitude modulated by the signal

4 sin (2 x 10° t) + 1 cos (6 x 10° f) volts. Calculate the depth of
modulation.

Solution
The r.m.s. value of the modulating signal voltage is
Fla Sg
partie 2 = 2.916 V,

and hence

2.916
m= = 0.412, or 41.2%. (Ans.)
10/2
The maximum depth of modulation that a practical amplitude
modulator is able to produce without generating distortion in excess
of a specified limit is generally restricted. Sometimes a carrier-
cancellation technique is employed to reduce the voltage of the carrier
component and in this way increase the depth of modulation.
The expressions for the instantaneous voltage of an amplitude-
modulated wave can be rewritten in terms of the modulation factor.
(a) For sinusoidal modulation
vy. = V1 + m sin wnt] sin wet. (1.8)

(b) For three-tone modulation

Vo = V1 + m, sin wt + mp sin wyt

+ my; Sin w3f] sin wet. (1.9)
(c) For a general complex modulating signal V,,(¢)

Ve = vi + Vin(t) sin wet. (1.10)

c
4 RADIO SYSTEMS TECHNOLOGY

The Root-mean-square Value of an Amplitude-modulated Wave

The r.m.s. value of an amplitude-modulated wave is the square root

of the sum of the squares of the r.m.s. values of each of its component
frequencies. Thus, for a sinusoidally modulated wave

ve GY Ge) Ge)
_F6-2)
V = r.m.s. carrier voltage Jd + m?/2). (1.11)

For a three-tone modulated wave,

V= Ls er mM ae maVe + maVs , OF
2 ‘ 4 4
2
2

where mp = |(m{ + m3 + m3).

Example 1.2

A carrier has an r.m.s. voltage of 10 V. It is amplitude modulated by a signal

having components at frequencies f, and f, when its r.m.s. voltage rises to
11.5 V. If the depth of modulation due to one of the components is 60%
calculate the depth of modulation caused by the other component.

Solution
From equation (1.12), 11.5 = 10/(1 + m2/2)

1.3225 = 1 + m3/2, or mp = 0.803.

Therefore

m, = (0.8037 — 0.6”) = 0.534 or 53.4%. — (Ans.)

If the modulating signal is of rectangular shape the number of its
frequency components is very large and an alternative approach will
give a simpler and more accurate result. Suppose that the rectangular
waveform has a peak value of V,, volts. The maximum value of the
modulated carrier will be V. + V,, volts and the minimum value will
be V. — Vj, volts (see Fig. 1.2). The r.m.s. value of the modulated
waveform is

V= IG |,»ar) = tC “ =) (1.13)
 AMPLITUDE MODULATION 5

Envelope

Fig. 1.2 Carrier amplitude modulated by a rectangular signal.

Example 1.3

An 8 V peak carrier wave is amplitude modulated by a square waveform of

peak value 5 V. Calculate (a) the r.m.s. value of the modulated waveform,
and (b) its depth of modulation.

Solution
(a) The maximum value of the modulated wave is 13 V and the minimum
value is 3 V. Hence
1 7/2 FE
lV = (| 13? sin? wt dt + | 3° sin? w,t a)
T 0 T/2

1 (7 169
= a — (1 — cos 2a,t) dt
yA ese)
came)
SP | —(Ieecos 2.@cr) a
7/2
1 7/2 Pr
= & (| 84.5 dt + | 4.5 ar)
T 0 7/2

(since the mean value of cos 2w,t over half a cycle is zero)

=
Melee *3
|[\—184.5 x — + 4.5(7 — 7/2)
rT
= 44.5 = 6.67 V.
2 i
(Ans.)

r=*
wooSs |(n a
2
or my = 0.884 = 88.4%. (Ans.)

Power Contained in an Amplitude-modulated Wave

The power developed by an amplitude-modulated wave is the sum

of the powers developed by the carrier component and by each of
the other components in both the lower and the upper sidebands.
6 RADIO SYSTEMS TECHNOLOGY

The total power developed in a resistance R is

2 2 2

R 2R 2
The carrier power is V2/2R watts and the total power in the side-
bands is m%V2/4R watts. The transmission efficiency n of an
amplitude-modulated wave is the ratio of the transmitted power that
conveys information, i.e. the total sideband power, to the total
transmitted power. Therefore,
272
n= male vd ; ee , x 100%
4R v2 + m2/2)

= my > x 100%.
Dak mT
(1.15)
The maximum value for the modulation factor my is unity and then
n = 33.3%. For any other value of m the sideband power will be
an even smaller percentage of the total power. This means that
d.s.b.a.m. is not a very efficient method of transmitting information
from one point to another. On the other hand, d.s.b.a.m. can be
demodulated by a relatively simple envelope detector that produces
an output voltage which is proportional to the modulation envelope.
Perhaps the simplest, and the most common, version of this is the
diode detector.

Example 1.4

A 10 kW carrier wave is amplitude modulated to a depth of 70%. Calculate

the total sideband power and determine what percentage of the total power it is.

Solution
From equation (1.14), Pr = 10 kW (1 + 0.77/2) = 12.45 kW. Hence, the
sideband power is 12.45 — 10 = 2.45 kW. (Ans. )
Expressed as a percentage of the total power = 19.68%. (Ans.)

Phasor Representation of an Amplitude-modulated Wave

The frequency spectrum of a sinusoidally-modulated d.s.b.a.m. wave

contains components at the carrier frequency f,, and at the lower, and
upper side-frequencies, f, + fm. Equation (1.4) can be rewritten as
ve = V, sin wt + V,/2 sin [(w, — @m)t + 2/2]
+ V/2 sin (a. + @_)t — 2/2] (1.16)
and this allows the spectrum to be represented by a phasor diagram,
Fig. 1.3(a). There are three phasors, one for each component, whose
length is proportional to the voltage. Each phasor rotates in the anti-
clockwise direction at its own angular velocity. The instantaneous
envelope of the modulated wave is given by the phasor sum of the
 AMPLITUDE MODULATION 7

Wo
+ W,

2
we ohVa! Viql2 Vnl2 V,/2
Fig. 1.3. Phasor representation of a
d.s.b.a.m. wave. (a) (6)

three phasors; in Fig. 1.3(a) the two side-frequency phasors are in

anti-phase with one another and so mutually cancel out. Consequently,
the instantaneous modulation envelope is equal to the unmodulated
carrier voltage. To obtain the envelope over one cycle of the
modulating signal it is more convenient to take the carrier phasor as
the reference and assume it to be both stationary and in the vertical
plane. This is shown by Fig. 1.3(b), the two side-frequency phasors
will then rotate in opposite directions with angular velocity w,.
Figure 1.4 shows the positions of the side-frequency phasors, and
of the phasor sum of all three phasors, for intervals of 7/8, where
T is the periodic time of the modulating signal. For each time interval
the phasor sum gives the instantaneous value, represented by R, of
the modulated wave; joining these together gives the positive envelope
of the wave. Obviously, the negative envelope can similarly be
obtained.

Distortion of the Amplitude-modulated Wave

A d.s.b.a.m. waveform will suffer a reduction in its modulation depth

and/or distortion if, in its transmission through a network or system,

PES R ——
re ay Envelope
R Pte? Ui, 5
Cc

2 Cc C=r t C

i ‘, U,L oe I
es ae | R=0

t=7/8 t=T/4 t=3T/8 Mf

(OPE
t=0

r
t=3T7/4

Fig. 1.4 Phasor diagram of a d.s.b.a.m. wave over one complete

cycle of the modulating signal.
8 RADIO SYSTEMS TECHNOLOGY

its side-frequencies are amplified, or attenuated, by a different amount

from one another and/or from the carrier. If, for example, a sinusoid-
ally modulated wave has a depth of modulation of 75% and is applied
to a tuned amplifier that has a gain of A, at the carrier frequency and
3 dB less gain at the two side-frequencies, then the output depth of
modulation will be 75/|2 = 53%.

Example 1.5

A 1 MHz carrier is sinusoidally modulated to a depth of 80% by a 5 kHz

signal. The modulated wave is passed through an amplifier which has a voltage
gain of A, at 1 MHz, 0.9A, at 995 kHz, and 0.8, at 1005 kHz. Calculate
the depth of modulation of the output waveform.

Solution
At the output of the amplifier: carrier voltage = V., lower side-frequency
voltage = 0.9 x 0.4V. = 0.36V,, and upper side-frequency voltage =
0.8 x 0.4V, = 0.32V,. The r.m.s. voltage
(%) (286) (2 “))
V= + 2
; 2 2 2
= —V. [1 + 0.1296 + 0.1024] = Jd + 0.232).
a2)
Hence, m*/2 = 0.232 or m = 68.1%. (Ans.)

The calculation of the percentage distortion is more difficult. To

simplify the algebra, suppose that the upper side-frequency is
completely suppressed. Then the modulated wave is given by

Ve = Ve, sin wet + mV,/2 cos (we — wWm)t

= V, sin w.t[1 + m/2 sin w,t] + mV,/2 cos wet cos Wmt.

This has an envelope given by

IVE. + m/2 sin w_t)? + (MV;/2 cos wmt)?]
= V1 + msin w,t + m7/4 sin? Wmt + m?/4 cos? weery?

= Vil + m*/4 + msin wt)"

m sin w,,t
= Ve faa)
ai eee eee
m? sin? wnt |
81 + m2/4y2
m? M SiN Wt
= V. | + m?/4) 1 ~
16(1 + m?/4)? 2(1 + m?/4)
m* COS 2Wmt
16(1 + m?/4)?
The percentage second harmonic distortion is equal to
 AMPLITUDE MODULATION 9

m? 2(1 + m7/4) m
2142 sas eae (1.17)
16(1 + m*/4) m 8(1 + m*/4)

Example 1.6

A carrier wave is sinusoidally modulated to a depth of 30% and has one of

its side-frequencies suppressed. Calculate the percentage second-harmonic
distortion of the modulation envelope.

Solution
From equation (1.17), the second-harmonic distortion
0.3
= SK 100'= 3.67%.% Ans.)
8(1 + 0.37/4)

Amplitude Modulators Probably the most commonly employed method of generating a

d.s.b.a.m. wave in a radio transmitter is the anode-, or collector-
modulated Class C r.f. tuned amplifier (see Radio Systems for Techni-
cians). Other d.s.b. modulators utilize the non-linear relationship
between the applied voltage and the resulting current of many
electronic devices. Essentially, there are two types of non-linear
characteristic: those in which the characteristic is continuous and can
be described by a power series of the form i = av + bv? + cv? +
..., and those in which the device acts as an electronic switch (this
means that the carrier voltage must be large enough, 2 V or more
usually, to turn the device ON and OFF).
If the carrier wave V, sin w,t and a modulating signal V,,(¢) are
applied in series to a non-linear device and the carrier voltage is not
large to switch the device, then the current flowing will be given by
i = a[V, sin wt + Va(t)] + b[V, sin wet + V(t)?
+ c[V, sin wt + V(t)’, ete.
The squared term will produce
i = bV2 sin? wt + bV2(t) + 2bV, sin wtV p(t)
and if a filter can be employed to pass only the terms
aV, sin wt + 2bV.V,(t) sin wet,
or aV,[1 + 2b/a V,,(t)] sin wt,

a d.s.b.a.m. wave will have been obtained. Unless the total input
voltage is fairly small the cubic term cv? (and perhaps even higher
terms) may also make a contribution to the filtered output signal and
so cause distortion. Suppose, for example, that V,,(t) = V; sin wt +
Vy sin wyt, then the cubic term gives cV;,(t) V, sin w,t which, when
expanded, shows the presence of components at frequencies w,, W,
+ 2w), @ + 2w2, and w, + (w; + w). These components may
fall within the pass-band of the wanted modulating signal and then
they cannot be removed by the filter. To avoid this effect it will often
 10 RADIO SYSTEMS TECHNOLOGY

be found necessary to use two non-linear devices in a balanced

arrangement, with respect to the carrier, so that the intermodulation
terms cancel out.
The switching-type d.s.b. modulator includes one, or more, tran-
sistors that are turned ON and OFF by the carrier voltage and which
effectively multiply the modulating signal by a square wave whose
amplitude is either +1 V or O V and 1 V. The Fourier series for a
+1 V square wave is

v = 4/r(sin wt + 1/3 sin 3a,.t + 1/5 sin 5a,t + ...)

and for a square wave whose voltage is either 0 V or 1 V is
v = 1/2-+ 2/r(sin w,¢ + 1/3 sin 3w,t + 1/5 sin5@,¢ + ...).
If the input signal is Vg + V,, sin w,¢ then, in either case, the output
voltage will be equal to K sin w,t + K sin wt sinw,yt +... .
When an integrated circuit modulator is used, some form of
transconductance multiplier is often employed. Figure 1.5 shows the

Carrier

Modulating
signal

Fig. 1.5 Transconductance

multiplier a.m. modulator.
 AMPLITUDE MODULATION 11

basic arrangement of such a circuit. Two differential amplifiers are

connected as shown and are driven by the carrier signal. The two
collector load resistors are externally connected to the appropriate
package pins and usually R; = R, = R. The differential amplifiers
have individual constant-current generators to supply their equal d.c.
emitter currents J;. In the absence of any input modulating signal,
or carrier, voltages the current relationships are: J; = 1, + L =
I, + l. The carrier voltage varies the mutual conductance of
transistors T,, T,, T; and T, so that J, + h = 15 + 8m,V- and
Peat dg = ds ing Pee The differential output current is
+ b) — & + ) = 28m, Ve.
When the modulating signal is applied to the circuit it varies the
emitter currents of the two differential amplifiers by g)Vm. Now,
the differential output current is 2 2,28m)VcVm and so the differential
output voltage is 22m,8m,RV-Vm volts.
If the carrier voltage is small the circuit acts as a linear multiplier
to produce an output proportional to V, sin w,tV,,(¢). If the carrier
voltage is greater than about 0.7 V it turns transistors T,/T, and
T,/T; alternatively ON and OFF so that the differential part of the
circuit switches the emitter current between the two output terminals.

Detection of an Amplitude- The methods used to detect, or demodulate, a d.s.b.a.m. signal fall
modulated Wave into one of three main classes: these are non-coherent or envelope
detection, coherent or synchronous detection, and non-linear detection.

Envelope Detection

Since the envelope of a d.s.b.a.m. signal has the same waveshape

as the original modulating signal, detection can be achieved by rectify-
ing the envelope. The basic circuit of an envelope, or diode, detector
is shown in Fig. 1.6. The input modulated signal should be of suffi-
ciently large amplitude (about 1 V) to ensure that operation is on the
linear part of the diode characteristic. Provided the time constant,
C,R, seconds, of the resistive load and shunt capacitor is long
compared with the periodic time of the modulating signal the voltage
that appears across R, will include the following components: (a) the

Fig. 1.6 Diode detector.

12 RADIO SYSTEMS TECHNOLOGY

modulating signal, (b) a d.c. voltage that is directly proportional to

the carrier amplitude, and (c) a number of high-frequency signals.
The d.c. voltage is blocked by capacitor C, and the components (c)
are filtered off.

Analysis

When the d.s.b.a.m. wave [V, + V,,(t)] sin wet is rectified the output
voltage consists of the envelope of half-sine waves at the carrier
frequency. The Fourier series for a half-wave rectified sine wave is
1 lie: | cos 2w,t cos 4w,t
Ve ie SINE hee te are ation ts
T 2 T 3 15
(1.18)
The output voltage of the detector is obtained by multiplying this series
by [V, + Vin(t)]. The product contains the terms V./a + V,,(t)/m +
...; the first term is a d.c. voltage and the second term is the wanted
detected audio signal. The detection efficiency n is the ratio
detected output voltage
peak input voltage
expressed as a percentage. Figure 1.7 shows two consecutive half-
cycles of the input voltage. If the input voltage is V, cos w,t the diode
will conduct when V. cos wt = Voy; then wt = 6 =
cos~! (Vou/V.) or cos 6 = Vour/V.. During the period 26 the current
that flows is
V. cos wet — Vou
C= I
r

where r is the forward resistance of the diode. The average value of

the diode current is

I; = 254 : idw.t
| dwt = —— j iid t
ri 2 Bereiktod 2m elke ms

Vo COS wot

Time
Fig. 1.7 Action of the diode detector.
 AMPLITUDE MODULATION 13

1 (9
== | i dwet
T Jo

a abe(V, sin 0 — Vy.9) =

V.
— (sin 6 — 6 cos 6).
ar ur
Hence, the output voltage
V.R
Voi = IncR) = — ! (sin 8 — 0 cos 0)
ur
and

Vow R .
n= ' = cos 6 = — (sin 0 — 0 cos 8).
re Tr
Then
@3 3 1/3
ny ee ry a ee eskl
R, 3 3 R,
and
3 ar \'3
n = COS os x 100%
R,
1 oF3 2/3
Spl ieee ese thin ed 0% (1.19)
Dial
For a low-voltage input signal the forward resistance of the diode
is not constant and so the diode detector efficiency will vary. For
voltages in excess of about 1 V the efficiency is constant at (usually)
90% or more.
The input resistance of a diode detector is related to its detection
efficiency. The diode conducts only when the input voltage is at, or
near, its peak positive value. Hence the average input power is

Pe =
Vc 2
Insist
V. 2eek SS
1
a R, (E) Ry
Therefore,
R
Rie = aa (1.20)
2n
In modern radio receivers the reactive part of the detector low-pass
filter is connected via a buffer amplifier so that the load seen by the
diode detector is always resistive.
Example 1.7

Calculate the input resistance of a diode detector that has a load resistance
of 8 kQ and a diode forward resistance of 100 Q.

Solution
From equation (1.19)
14 RADIO SYSTEMS TECHNOLOGY

Capacitor ;
voltage Time constant
too long
a +
=
=

“ Time

Fig. 1.8 Output voltage of a diode

detector, showing the effect of an Input signal
incorrect time constant.

1 /3mx x 100 \2"

n=]1-—- : x 100%
= 88%.
2 8 x 10
From equation (1.20)
8000
Rw = ———— = 4545 Q. (Ans. )
2 xX 0.88

Clipping Caused by Incorrect Time Constant

The diode detector shunt capacitor C; must be able to discharge

rapidly enough for the voltage across it to follow the modulation
envelope. This is most difficult when the envelope is decreasing in
amplitude, see Fig. 1.8. If the time constant C,R, is too long,
relative to the periodic time of the modulating signal, the capacitor
voltage will not be able to follow the troughs of the modulation
envelope; this is shown by the dotted line in Fig. 1.8.
As the capacitor C, discharges through R, its voltage is v =
Vie~"181, where V, is its initial, peak value which is equal to the
peak input voltage (minus the diode voltage drop). The rate at which
v falls is given by —dv/dt = v/C,R,. For v to follow the modulation
envelope without distortion
v d
= — [V1 + msin w,t)] = —mw,V.
Mm” ¢ cos wt V/s.
CR, ee mt)] m

Thus

V1 + m sin w,t) = —mw,V. cos Wmt

1 ~ MV. COS Wmt
or
(1.21)
CR, Vil + msinwpt) |
This inequality is most difficult to satisfy when its right-hand
side
has its maximum negative value. Differentiating and equating to zero,
 AMPLITUDE MODULATION 15

gives sin w,t = —m. Therefore, cos w,_t = —ja - m7”), and
substituting into inequality (1.21), gives
oe —monJl — m*) — — —Mwdp
Chic « 1 =m? Jd — m?)
1 = m?
and C,R < 4a = my (1.22)
MW

Example 1.8
A diode detector is to produce an audio output signal in the frequency band
100 to 4500 Hz. If the maximum modulation depth is 50% and the load resistor
is 10 kQ calculate the maximum possible value for the shunt capacitor.

Solution
From equation (1.22),
Ja = 0.25?)
C\amax) = = 6.85 nF. Ans.
EEG 5 o5C 9g" x A500 X10" ile

The diode detector has the advantage of simplicity but the disad-
vantages that: (a) at least 0.5 V is required for the diode to conduct
and this means that a high i.f. gain is necessary; (b) if the input signal-
to-noise ratio is low the output signal-to-noise ratio will fall more
rapidly than the input signal-to-noise ratio, this is known as the
threshold effect; and (c) it generates energy at harmonics of the
intermediate frequency of the radio receiver.

Non-linear Detection

If ad.s.b.a.m. signal is applied to a non-linear device detection will

occur because of the v* term. The device may well be the diode in
an envelope detector when the input voltage is small. If the input
voltage is V(1 + m sin w,t) sin wet the v? term will give
bv + m sin wnt)? sin? w,t. Expanding this gives the term
bV2m sin wt which is, of course, the wanted modulating signal.
There are a number of other, unwanted, components at other frequen-
cies also present and the most troublesome of these arises from the
terms bV2/2 m” sin? wmt. Expanding this gives —bV2/4 m* cos
2w,t which results in second-harmonic distortion and possible
intermodulation.
bV2m7/4
The second-harmonic distortion is or 0.25m%. If,
pe
c
for example, m = 30%, the percentage second-harmonic distortion
will be 7.5%. This second-harmonic component and the other, higher-
frequency components will probably be removed by a low-pass filter.
More serious are the intermodulation products that are generated.
 16 RADIO SYSTEMS TECHNOLOGY

Suppose that V,,(t) = V, sin w,;t + V> sin wt, then the squared term
contains the component bV,V> sin wt sin w)t which, upon expan-
sion, shows the presence of components at frequencies f, + f,. These
intermodulation products will fall within the bandwidth occupied by
the modulating signal and so they cannot be filtered out.

Coherent, Synchronous or Product Detector

A product detector multiplies together the d.s.b.a.m. wave to be

demodulated and the unmodulated carrier; it is therefore necessary
that the original carrier, at the correct frequency, is available at the
receiver. The output voltage of the detector is vy, = V. sin w.t X
V1 + m sin wpt) sin wt or V2/2(1 — cos 2w,t)(1 + m sin Wt) and
this contains the term V,V,,/2 sin w,t. This is the wanted modulat-
ing signal so that detection has been achieved. The block
diagram of a product detector is given by Fig. 1.9(a). The carrier
component must be extracted from the incoming d.s.b.a.m. signal
and one method of doing this is shown by Fig. 1.9(b). The d.s.b.a.m.
signal is hard-limited to produce a square wave of frequency f,. The
two signals are applied to an analogue multiplier whose output contains
the detected signal. Another carrier extraction method employs a
phase-locked loop (see Electronics IV), see Fig. 1.10. The phase
detector has inputs of V.(1 + m sin w,t) sin wt and the voltage
generated by the voltage-controlled oscillator (v.c.o.) and it generates
an output voltage that is proportional to the phase difference between
the two voltages. The output of the phase detector is fed, via a low-
pass filter, to the control terminal of the v.c.o., causing it to change
frequency in the direction which minimizes the error. Once lock has
been established the v.c.o. frequency will be equal to the carrier
frequency and this is also the output voltage of the circuit.

x(t)-y(t)
y(t)=V(1+m sin w,,t) sin wet
Product
detector

x(t)=V, SiN wet "YD

(2)

y(t) Analogue
multiplier
V,
any sin w,,t

x(t)-y(t)

Hard
Fig. 1.9 (a) Product detector. limiter
(0) Use of a hard limiter to obtain
the carrier component.
(6)
 AMPLITUDE MODULATION 17

V(1+m sin w,,t) sin w, Phase Voltage- V, sin wt

controlled
detector
oscillator

Fig. 1.10 Carrier extraction by

phase-locked loop.

The product detector is now in widespread use since it is particularly

convenient for implementation in an integrated circuit. Usually, the
product detector is within the same IC package as other radio receiver
circuitry. The product detector offers the advantage that the input
signal may be very small since its lower limit is set only by the wanted
signal-to-noise ratio. This will allow the i.f. gain to be up to 60 dB
less than if a diode detector is used. The disadvantages are (a) the
circuit will demodulate input noise if there is no signal and so a squelch
circuit is really necessary, and (b) the d.c. output voltage is small
and so a d.c. amplifier is needed to produce the a.g.c. voltage; this
is often on-chip, if not an op-amp or an a.g.c. generator chip can
be used.

Double-sideband Suppressed- Most of the power carried by a d.s.b.a.m. wave is developed by the
carrier Amplitude Modulation carrier frequency component and hence the transmission efficiency
is low. Since the carrier conveys zero information it is not necessary
that it be transmitted and it can be suppressed at the modulation stage
if a balanced modulator is used. The basic principle of a balanced
modulator is illustrated by Fig. 1.11. Each non-linear device, which
may be a diode or a transistor, has a current-voltage characteristic
given by i = ip + av + bv” + .... A FET is the best device for
this purpose since its mutual characteristic very nearly obeys square-
law. Assuming identical square-law devices, the voltage applied to
n.l.d. A is va = Vy sin wt + V, sin wt and the voltage applied
to n.l.d. Bis vy = —V,, sin w,t + V, sin w,t. Therefore

V,, SIN wt R.

Fig. 1.11 Balanced modulator. V, sin wt

18 RADIO SYSTEMS TECHNOLOGY

i, = aV_ SiN w_t + aV, sin wt + bV2, sin? wpt

+ 2bV,,V, sin wt sin w_t + bV?2 sin? w,t
and

i, = —aV,, sin w,t + aV, sin wt + bV2, sin? wt

— 2bV,,Vz sin wet sin w,t + bV2 sin wet.
The output current i, of the modulator is proportional to the
difference between i, and i, and so it is
lout = 2(AVmq SIN Wyt + 2bV~V. sin wt SiN Wt).
The term at the modulating signal frequency can be removed by a
filter, provided w, > w,, to leave the wanted d.s.b.s.c. signal. A
transconductance multiplier IC can also be used as a balanced
modulator; the carrier component voltage should be large enough to
switch the transistors and the modulating signal input should be
capacitor-coupled to remove any d.c. component. A number of IC
balanced modulators are also available such as the Motorola 1596 and
the Plessey SL 6401.

Phasor Diagram

The phasor diagram of a d.s.b.s.c. amplitude-modulated wave is

shown in Fig. 1.12. Clearly, the envelope of the modulated wave is
not sinusoidal, indicating that distortion has occurred. This means
that the signal cannot be demodulated at the receiver unless the carrier
component is re-inserted with both the correct frequency and phase.
If there is an error 6 in the phase of the re-inserted carrier the
instantaneous voltage of the wave will be
v = V, sin (wt + 6) + mV, sin wet sin wt
V.[cos wt sin @ + sin w,t(cos 6 + m sin w,,t))].
The envelope of this is V..|[sin? 6 + (cos 6 + m sin w,t)"] or

1 — cos 20
V. ne + cos* 6 + 2 cos 0m sin Wmt
, 1/2
+ m? sin? wrt

er: Ee elope oe ORs

op = oe _ or. a
ae ee
at
seas
OoSage
= a7 =O
ae
t=0 “pe | ee t=Tl2 t=T
t=T/8 t=37/8 t=5T7/8 t=7T/8
f=/4 t=3T7/4

Fig. 1.12 Phasor diagram of a d.s.b.s.c. amplitude-modulated wave.

 AMPLITUDE MODULATION 19

1 cos 26 1 cos 20
V. + + + 2m cos @ sin w,t
2 2 2) 2

m2 m2 1/2
Pe OSS Ot
2 pe

m 2 2 1/2
=V, c i oT + 2m cos @ sin w,t — = cos 2m!|

Vy, guts eee ee li

2 2

m? cos 2Wmt
(nep oI2) + at (1.23)

Equation (1.23) shows that the envelope of the reconstructed d.s.b.s.c.

a.m. waveform varies in a manner that is a function of both the
modulating signal and the phase error. If 6 = 90° cos 6 = 0 and
there will be no variation of the envelope at the signal frequency.
What amplitude modulation there is occurs at twice the signal
frequency and there is considerable phase modulation as well.

Example 1.9

A 20 V carrier is amplitude modulated to a depth of 60%. The carrier is

then removed and after a phase shift of 90° is re-inserted. Calculate the
resulting peak phase deviation.

Solution
The two side-frequency phasors are (0.6 X 20)/2 = 6 V. The peak phase
deviation is, see Fig. 1.13,

pier itaw 12/20 4231°. ~ (Ans.)

The requirement for the re-inserted carrier to be correct in both

frequency and phase is not simple to satisfy and requires the use of
complex circuitry. The d.s.b.s.c. system is not used for ordinary
radiocommunication purposes but it is employed for (a) the transmis-
sion of colour information in television broadcasting, and (b) the
transmission of stereo information in v.h.f. sound broadcasting.

Fig. 1.13
20 RADIO SYSTEMS TECHNOLOGY

Single-sideband Suppressed-carrier The information carried by an amplitude-modulated wave is contained

Amplitude Modulation in both sidebands and so it is not necessary for both sidebands to be
transmitted. Either sideband can be suppressed at the transmitter
without any loss of data. Figure 1.14 shows the phasor diagram of
an s.s.b.s.c. a.m. signal with its carrier component re-inserted with the
correct phase. The envelope can be seen to be sinusoidal. The maxi-
mum phase error is 90° and if the phasor diagram is redrawn with
this carrier re-insert error it will be found that the envelope is still
sinusoidal. This means that the phase of the re-inserted carrier is not
important for a speech system.
Two methods of suppressing the unwanted sideband are used: either
the output of the balanced modulator can be passed through a filter
of the appropriate bandwidth, or the phasing technique shown in Fig.
1.15 can be used. The use of a filter to remove the unwanted sideband
becomes more difficult at high frequencies because the frequency gap
between the sidebands becomes a small percentage of the filter’s centre
frequency. For this reason modulation is often carried out at a low
frequency, then the wanted sideband is shifted to the desired position
in the frequency spectrum.
Referring to Fig. 1.15, the two balanced modulators have both the
modulating and carrier voltages applied to them, but the upper
modulator has both its inputs phase shifted by 90° before they are
applied. The input signals to the upper modulator are V,, sin
(Wmt + 90°) and V, sin (wt + 90°) so that its output current
contains components
cos [(w,t + 90°) — (apt + 90°)] — cos [(w,t + 90°)
+ (wrt + 90°)]
= COS. (@.f.— Wmt)i— Cos (wt + w,t + 180°).

The inputs to the lower modulator are V,, sin w,t and V, sin w,t and
so its output current includes components cos (wt — wt) —
COS (w,t + wt). The output signals of the two modulators are added
together and, since the upper side-frequency components are in anti-
phase with one another they cancel, give an output of 2 cos
(w, — W,)t. If it is required to transmit the upper sideband instead

Seger ey iin
R

Cc . Cc
FRE mR
c as
pa)
R

t=7T/8 t=T/4 t=3T/8 t=T/2 ip [l, E t=T

t=57/8 L t=7T7/8
t=37/4

Fig. 1.14 Phasor diagram of an s.s.b.s.c. a.m. signal.

 AMPLITUDE MODULATION 21

Balanced
90° phase
modulator
shift
1

90° phase
shift

Modulating Carrier s.s.b.s.c. signal

signal © frequency
Vin(t)

Oscillator

Carrier
frequency

Balanced
modulator
2
Fig. 1.15 The phasing method of
producing an s.s.b.s.c. a.m. signal.

of the lower sideband either one of the 90° phase-shifting circuits

must be moved to the lower part of the circuit. The phasing method
of suppressing one sideband has the advantage that it is easy to switch
from transmitting one sideband to transmitting the other. One circuit
which can be used as the audio phase-shifting circuit is shown by Fig.
1.16; here w, = 1/C,R,; and w, = 1/C,R) specify the audio
bandwidth shifted.

Demodulation of s.s.b.s.c. Signals

An s.s.b.s.c. wave can be demodulated using either a product, or a

switching, detector. Suppose that the lower sideband is received. The
output of a product detector is proportional to the product of the lower
sideband signal and a locally generated carrier. Thus,
Vor = V sin (We — Wm)t Vo Sin wet
= WV (sin wt COS Wpt — SIN Wpt COS Wet) SIN Wet.
This includes the term VV,/2 cos w,t which is the demodulated
message signal and it can be extracted using a low-pass filter. If there
is a frequency error dw, in the re-inserted carrier the wanted output
of the detector will be VV,/2 cos (w,t + dw,t). The frequency error
that can be tolerated depends upon the type of signal involved. Since
every frequency contained in the modulating signal is shifted by the
same amount 6f, the harmonic relationship between the components
is lost. The recommendation of the CCITT is that the error should
not be in excess of +2 Hz. However, it is found that for speech an
22 RADIO SYSTEMS TECHNOLOGY

90° phase-advanced
Modulating modulating
signa signal
Vin(t)

Fig. 1.16 90° phase-shifting circuit.

error of +20 Hz is hardly noticeable and as much as +50 Hz is

tolerable.
If th e re-inserted carrier is of the correct frequency but there is a
phase error 6 the demodulated signal will contain a term VV,/2 cos
(Wnt — @). For speech and music this error is of little consequence
but it will matter for any system where the signal waveshape is of
importance.

Single-sideband Compared with Single-sideband operation of a radio system offers a number of

Double-sideband Amplitude advantages over d.s.b. operation.
Modulation
(a) The bandwidth required for an s.s.b.s.c. system is only one-
half of the bandwidth that must be allocated to a d.s.b.a.m.
system. The reduction in bandwidth allows a greater number
of channels to be accommodated within a given bandwidth.
(b) The signal-to-noise ratio at the output of an s.s.b.s.c. system
is higher than the output signal-to-noise ratio of a d.s.b. system
transmitting the same power. If the carrier is sinusoidally
modulated the d.s.b. system will have peak side-frequency
voltages of mV,/2 and a peak envelope voltage of V1 + m)
volts. The s.s.b.s.c. system will have a side-frequency voltage
of V.(1 + m) volts; this is V1 + m)/(mV,/2) or 201 + m)/m
times as great as each d.s.b. side-frequency voltage. If the
transmission path is free from distortion the two d.s.b. side-
frequency voltages add algebraically in the detection process.
Hence the s.s.b.s.c. side-frequency voltage is (1 + m)/m times
as great as the d.s.b. sum voltage. Quoted in decibels, this is
20 logio[(1 + m)/m] dB. In addition, the output noise power
will be reduced by 3 dB because the bandwidth has been halved.
Therefore, the increase in the output signal-to-noise ratio given
by s.s.b.s.c. operation is
 AMPLITUDE MODULATION 23

: : ae 1+m
signal-to-noise ratio increase = 3 + 20 len )dB.
m
(1.24)
The maximum possible value for the modulation factor m
is unity and this gives the signal-to-noise ratio improvement
as 9 dB. For any smaller value of m an even larger advantage
is obtained.
(c) An s.s.b.s.c. transmitter is more efficient than a d.s.b.
transmitter.
(d) Selective fading of d.s.b. radio waves may cause considerable
distortion when the carrier component fades relative to the side-
frequencies. This effect does not occur in an s.s.b.s.c. system
because the received signal is demodulated against a re-inserted
carrier of constant amplitude.

Example 1.10

A d.s.b.a.m. system radiates a carrier power of 30 kW. The system is changed

to operate using s.s.b.s.c. and a radiated power of 10 kW. Calculate the change
in the output signal-to-noise ratio of the system if the maximum depth of
modulation is 75%.

Solution
From equation (1.24), the increase in the output signal-to-noise ratio is
3 + 20 log,o(1.75/0.75) = 10.36 dB. The decrease in the transmitted power
is 10 log, (30/10) = 4.77 dB. Therefore, increase in output signal-to-noise
ratio
= 10.36 — 4.77 = 5.59 dB. (Ans.)

Peak Envelope Power

The output power of an s.s.b.s.c. radio transmitter is usually specified

in terms of its peak envelope power (p.e.p.). The p.e.p. is the power
developed by the peak value of the transmitted sideband and any pilot
carrier. When there is no pilot carrier, or it is of very small amplitude,
the term peak sideband power (p.s.p.) is often employed instead.
 Angle Modulation

When a sinusoidal carrier wave V, sin (w,t + 0) is angle modulated

the amplitude V, of the carrier is maintained at a constant value and
its instantaneous phase w,t + 6 is varied by the modulating signal.
There are two possibilities: either the frequency w,/2 a of the carrier
wave, or its phase @, can be made to vary in direct proportion to the
instantaneous voltage of the modulating signal. The differences
between frequency modulation and phase modulation are not obvious,
since a change in frequency must inherently involve a change in phase,
and they are listed in Table 2.2.
Frequency modulation is much more commonly used than phase
modulation and it possesses a number of advantages over amplitude
modulation, particularly if a wide bandwidth can be made available.
Frequency modulation is employed for sound broadcasting in the v.h.f.
band, for the sound signal of u.h.f. television broadcast transmissions,
for some land, sea and air mobile systems, and for u.h.f./s.h.f. multi-
channel telephony systems including those routed via communications
satellites. Phase modulation is not often used in analogue radio com-
munication and it finds its main application in the field of digital radio
communication. It is, however, often employed as a stage in the
generation of a frequency-modulated wave.

Frequency Modulation Frequency modulation of a carrier wave occurs when the instantaneous
deviation from the unmodulated carrier frequency is directly propor-
tional to the instantaneous amplitude of the modulating signal voltage.
The maximum amount by which the carrier frequency can be deviated
is known as the frequency deviation. Frequency deviation has no
inherent limit and for any particular f.m. system a maximum permis-
sible deviation must be specified. This maximum is known as the rated
system deviation f;. Once the rated system deviation has been deter-
mined it sets the maximum modulating voltage that can be applied
to the frequency modulator. Most of the time the modulating voltage
will be smaller than this maximum value and then the frequency
deviation is kf,, where
ie modulating signal voltage 2]
maximum allowable modulating signal voltage =
Clearly, k can have any value between zero and unity.
 ANGLE MODULATION 25

Instantaneous Voltage

If the modulating signal is V,,(¢) the instantaneous carrier frequency

f, will be
h=fe + havnt). (2.2)
The instantaneous angular velocity w; of the modulated wave is
w, = 2af, = dé/dt = 2af. + 2akfyV,,(t),

and
0 = \\wide = \ 2afde + \,2akfWVn(t) dt, or
0 = wet + 2akfy \, Volt) di. (2.3)
The instantaneous voltage v, of the frequency-modulated carrier
wave is vy. = V, sin @ or
Ve = Ve sin [wt + 2akfy \y Vint) de]. (2.4)
The peak phase deviation of the carrier depends upon the integral with
respect to time of the modulating signal voltage.
If the modulating signal voltage is of sinusoidal waveform, i.e.
Va(t) = V,, COS wt, then equations (2.2) and (2.3) become

=f t+ kfg cos ont (2:5)

and @ = wt + Kfa sin Wf, (2.6)

m

respectively, and the instantaneous voltage of the wave is

vy. = V. sin E So
Kfa sin ont (257)
m

Modulation Index

The peak phase deviation of the carrier is equal to kfq/f,, and this
factor is usually known as the modulation index m;. The modulation
index is equal to the ratio of the frequency deviation to the modulating
frequency. Very often the expression for the instantaneous voltage
of a sinusoidally modulated f.m. wave is written in terms of the
modulation index, thus

vo = V, sin (apt + m, Sin wy). (2.8)

Deviation Ratio

When an f.m. system is designed the maximum permissible values

for both the frequency deviation of the carrier and the modulating
26 RADIO SYSTEMS TECHNOLOGY

signal frequency must be used. Then the modulation index is known

as the deviation ratio D.

Bet falfm¢max) ‘ (2 2 9)

The deviation ratio of a particular f.m. system is fixed, whereas the

modulation index varies continuously with change in the modulating
signal voltage and/or frequency. In the BBC v.h.f. sound broadcast
system fy = 75 kHz and fimax) = 15 kHz so that D = 5.

Example 2.1

When a 2 kHz sinusoidal signal is applied to a frequency modulator the

90 MHz carrier is deviated by + 16 kHz. (a) Calculate the phase deviation
of the carrier. (b) Calculate the new frequency and phase deviations if the
modulating signal has both its voltage and its frequency doubled.

Solution
(a) Phase deviation = (16 x 10%)/(2 x 10°) = 8 rad. (Ans.)
(b) Phase deviation = (32 x 10°)/(4 x 10°) = 8 rad. (Ans.)
Frequency deviation = 32 kHz. (Ans.)

Frequency Spectrum of a Frequency-modulated Wave

The frequency spectrum of a frequency-modulated wave is much more

complicated than that of an a.m. wave. For small values of modula-
tion index, say ms < 0.25, the f.m. wave consists of a carrier
component f, and two side-frequencies f, + f,,, just like a d.s.b.a.m.
waveform. If, however, the modulation index is increased, second-
order side-frequencies f. + 2f,, will also appear.
Further increase
in the value of m, leads to the appearance of more,
and more, higher-
orders of side-frequencies, and the frequencyspectrum rapidly
becomes complex. Equation (2.8) can be rewritten
(using the identity
sin (A + B) = sin A cos B + sin B cos A) in the form

ve = Vy sin wet cos (mr sin wt)

+ V, COS wet sin (ms Sin wy f). (2.10)
If the value of m, is less than unity equation (2.10) can be expanded
using the series forms of sin 6 and cos 6, but if m, is larger than unity
expansion requires the use of Bessel functions.

Small values of my

The series forms of sin 6 and cos 6 are

3 5
sii @enadneeepet ;
3! 5!

gp
and cos? =1—-—+—.....
2! 4!
 ANGLE MODULATION 27

If mg is very small, say less than 0.25, m? and all higher powers of
m, will be negligibly small, and then
cos (my SIN wt) = 1
and sin (mr sin @,f) = Mr SIN Of.
Equation (2.10) can then be written as

vo = V, sin wot + Vemg cos wet Sin Opt

= V, sin wot + mV.p[sin (w. + &m)t
— sin (w, — @,)t], (2.11)

(using the identity 2 cos A sin B = sin (A + B) — sin (A — B).

This shows that the spectrum of a narrow-band frequency-modulated

(n.b.f.m.) system consists of the carrier f, and the lower and upper
side-frequencies f, + f,. Comparing with equation (1.4) it can be
seen that n.b.f.m. is equivalent to d.s.b.a.m. with the lower side-
frequency phase-shifted by 180°.

O25 mips T
If the modulation index m, is greater than 0.25 and less than unity,
m¢ eS:sin t
cos (m; sin wt) = 1 — — = om and

sin (ms SIN wt) = Mz SIN wpl.

This means that the instantaneous voltage of the f.m. wave is

ie
1 — m? sin? wnt
Ve V. sin wet

+ mV, COS Wot SIN &mt

: mV, sin wet
II Vesin at = aegis

mf V, sin w.t COS 2Wmt

+ meV, COS Wet SIN Wt.
4
Hence (using the identity 2 cosA sin B = sin (A + B) — sin(4 — B)),
2 mV. _.
; m
v. = V. sin wet 1 ~ ae + ee [sin (wp + Wm)t

— sin (W@ — W,)t] +

mV. [sin (Ww. + 20m)t

+ sin (@, '=92,,)t]. (2.12)

Example 2.2

A 10 V, 100 MHz carrier wave is frequency modulated by a 2000 Hz tone

28 RADIO SYSTEMS TECHNOLOGY

when the frequency deviation is 1000 Hz. Calculate the amplitudes of the
first-order, and of the second-order side-frequencies.

Solution eh
m, = kf,/f, = 0.5. From equation (2.12),
52
Carrier amplitude = 10 (:= )= 9.375 V. (Ans. )
4

0.5 x 10
First-order side-frequency amplitude = —————— = 2.5 V.
2 (Ans.)
0.5°>x.10
Second-order side-frequency amplitude = —————— = 0.3125 V.
8 (Ans.)

ms = 1

When the modulation index m, is equal to, or is greater than, unity

expansion of equation (2.10) requires the use of Bessel functions.
Using these shows the presence of a number of higher-order side-
frequencies in the f.m. wave. This is shown by equation (2.13).
Ve = V.Jo(m,) sin wet
+ V.Ji(m,) [sin (w, + Wm)t — sin (Ww — wm)t]
+ V.J2(m,) [sin (w, + 2@m)t + sin (@ — 20,)t]
+ V.J3(m,) [sin (w, + 3@_)t — sin (a — 30,)f]
+ V.Ja(m,) [sin (w, + 4@m)t + sin (we — 40,,)f]
FECOLG. 4 (2.13)
where Jo(m,), J\(m,), Jo(m,), etc., are Bessel functions of order 0,
1, 2, etc., and give the amplitudes of the carrier and side-frequency
components. Examination of equation (2.13) shows that the frequency
spectrum contains the carrier component at frequency f., the odd-
order side-frequencies f. + fn, fe + 3fn, fr + 5Sfm, etc., and even-
order side-frequencies f. + 2f,, f. + 4fnm, fr + 6f,,, etc. All the
odd-order lower side-frequcncies have a minus sign in front of the
lower side-frequency component, whilst all the even-order lower side-
frequency components are preceded by a plus sign.
The amplitudes of the carrier, and of the various side-frequencies,
vary with the modulation index but not in a simple manner. The carrier
and each order of side-frequency reach a positive maximum value
and then decrease and go cyclically through zero, reach a negative
maximum value and then rise towards zero, and so on. Side-
frequencies of the same order have the same amplitude so that the
frequency spectrum is symmetrical about the carrier frequency. Table
2.1 gives the values of Jo(m,), J,(m,), etc. for integer values of m,
from | to 12 and for side-frequencies up to the ninth order. The data
is also shown graphically by Fig. 2.1. Table 2.2 shows the values
of the modulation index, or the deviation ratio, at which the carrier,
the first-, second- and third-order side-frequencies are zero.
29
ANGLE MODULATION

23 bg I eee ee SS eee ee ES
vOes 0 680€°0 61L6¢c°0 6v1¢ 0 €9c1'0 68S0°0 ¢+20°0 a = “ aa a JOPJO UG
LSv0'0 0S2c'0 6ZLE'0 LSO€°0 S€ec'0 08ct'0 S9S0°0 v8to0 SH fz = = Jeplo Yig
€OZL'O- v8lo'0 29120 gZce'0 90ce'0 9e&c0 96¢1'0 veso'0 ceSt00 oT = ay 4apso YiZ
ZevzeO- 91020- SrvlOO- €v0c0 9ZEe'0 c6EE 0 8s7c'0 OLELO 1670°0 vLLO'O = sre JEpsJo UI9
sezoo-— esezo— lreco— OSsdOO0- 8S8t'0 6ZvE'0 Le9€°0 LL9¢°0 Leet 0 0€v0'0 = ore JOpJO UIG
Sc8l°0 OSIOO- g96120- ss9z0- vSOlO- sZsl°0 9ZS¢°0 cl6e°0 LL82°0 OZEL'O Ove0'0 =r Jepso Uiy
1S6L°O €22c'0 v8so'0 608l'0— -tl62O- 9Z91'0- 8rlt0 879e°0 c0er'0 L60€°0 68210 96100 JOp1O Pig
6v80'0— O6E1'0 97S¢'0 8r7rl'O O€lLtLO- vlOEO- 6erv20- 99470'0 Ly9E°0 19870 8cseo0 6rll0 J@pjo pug
veecO—- s89ZtO0- SEvO'O €Sve'0 9vEc 0 2¥00'0—;"29Zc0—— 92cS0-— 09900— , Léte0 29Z45°0 1lO0vv'0 4@pso 4s}
ZZv0'0 clLZLO-— 6Svc0-— €0600-— ZIZI0 L00€°0 90Sl'0 QJZL':0— 12Z6e°0- 109z0- 6Eez'0 25920 [(w)%7] sae
(eu) "r (LL) °r (o1) “r (6) "r (9)¢r Gyr (9)"r (s)"r (p)"r (e)"r er ae juauodwiog
ES

e1qeL
oh

be
 30 RADIO SYSTEMS TECHNOLOGY

aR oO

o©

o (o2}

o ~

°o 2) order+

Amplitude
relative
unmodulated
amplitude
carrier
to

Modulation index (or deviation ratio)

Fig. 2.1 Showing how the

amplitudes of the various
components of an f.m. wave vary
with the modulation index. Table 2.2

Modulation index m,

Carrier First-order Second-order Third-order

Rs OS ee a ee ee oa a eS
2.405 3.832 5.136 6.380
5.520 7.016 8.417 9.761
8.754 10.173 11.620 13.015
11.792 13.324 14.796 16.223
eee

Figure 2.1 can be employed to determine the amplitudes, relative

to the unmodulated carrier voltage, of each of the components of an
f.m. wave. The amplitude of a component, for a particular value of
m,, or D, is obtained from the figure by projecting vertically upwards
from the horizontal axis on to the required carrier, or nth-order side-
frequency, curve and thence on to the vertical axis. Negative signs
are (usually) ignored since only the magnitude of each component
is (usually) required. The number of significant side-frequencies, i.e.
those of amplitude +0.01 or greater, is equal to (ms + 1).
 ANGLE MODULATION 31

=a +f, m fOhatht 4h daloOhs

ieTiobete Sle bbt Shenton!) +h
Cc
fyt2hn
tS,
fo 4h fyotOim
f,+5tm
fot Thm

(a) (6)

Fig. 2.2 Spectrum diagrams for f.m.

Example 2.3
waves (a) m, = 0.5, (b) m; 4.
Plot the frequency spectrum diagrams of a frequency-modulated wave having
a deviation ratio of (a) 0.5 and (b) 4.

Solution
The required spectrum diagrams are shown in Figs 2.2(a) and (b) respec-
tively. The amplitudes of the components of Fig. 2.2(a) were calculated in
Example 2.2 and should be compared.

The Phasor Diagram of an f.m. Wave

The phasor diagram of an n.b.f.m. wave, that has first-order side-

frequencies only, has its lower and upper side-frequency phasors
positioned symmetrically about a quadrature carrier phasor. This is
shown by Fig. 2.3 for time intervals of 7/8, where T is the periodic
time of the modulating signal. As with the phasor diagram of the a.m.
wave it has been assumed that the carrier phasor is stationary in the
vertical direction so that the lower and upper side-frequency phasors
rotate in opposite directions with angular velocity +w,. The phase
of the resultant, relative to the unmodulated carrier phasor is varied
in a linear manner either side of the carrier. The amplitude of the
resultant phasor is not constant, which means that some amplitude
modulation is present.
The phasor diagram of an f.m. wave having both first-order and
second-order side-frequencies is more complex. The second-order

R C=R
!
\
! U
!
r=0

t=T/4 t=37/8 t=Tl2 t=57/8 t=37/4 t=77/8 t=T

t=0 t=T7/8

Fig. 2.3 Phasor diagram of an n.b.f.m. wave.

32 RADIO SYSTEMS TECHNOLOGY

side-frequencies have the same phase relationship to one another as

do the a.m. side-frequencies and they rotate with twice the angular
velocity, i.e. at +2. Figure 2.4 shows the phasor diagram. The
effect of the second-order side-frequencies is to reduce the variations
in the length of the resultant phasor and so reduce the amplitude
modulation of the carrier. The second-order side-frequencies do not
affect the phase deviation of the resultant phasor from the carrier.
In general, odd-order side-frequencies contribute to the phase devia-
tion of the carrier whilst even-order side-frequencies tend to keep the
amplitude of the modulated waveform constant. If all the side-
frequencies of an f.m. wave are included, the phasor diagram will
have a constant-length resultant phasor R which deviates either side
of the carrier phasor; this is shown by Fig. 2.5. The maximum phase
deviation @max Occurs when all the odd-order side-frequency phasors
are instantaneously in phase with one another and in phase quadrature
with the carrier phasor.

Complex Modulating Signal

When a carrier is frequency modulated by a complex signal the
resulting frequency spectrum is not the sum of the spectra produced
by each component in the complex signal acting alone. Consider a
carrier frequency modulated by the two-tone signal V,,(t) = V, sin
wt + V2 sin wt. Each component acting on its own will give a
certain modulation index m and will produce a carrier and various
side-frequency components. The magnitudes of these components can
be obtained from Table 2.1. If the modulation index due to V; sin wt
acting alone is m,, then the frequency components in the f.m. wave
are Jo, and J,,;; similarly V, sin wt will give a modulation index of
mp and components Jo, and J,2. For example, suppose that me, =
2, then from Table 2.1 Jo, = 0.2239 and J,,,; = 0.5767, 0.3528,
0.1289 and 0.0340.

t=T/4 t=37/8 t=T/2

Fig. 2.4 Phasor diagram of an f.m. wave with both

first-order and
second-order side-frequencies.
 ANGLE MODULATION 33

Locus of resultant Unmodulated carrier C

phasor R

Instantaneous resultant
phasor

Maximum phase deviation

Fig. 2.5 Resulting phasor diagram
of an f.m. wave when all side- Instantaneous
frequencies are included. phase deviation

When the carrier is modulated by the two-tone signal, the f.m. wave
will contain components at the following frequencies:
(a) carrier frequency f, of amplitude Jo, x Jo);
(b) side-frequencies f. + mr f, of amplitude J; X Jo;
(c) side-frequencies f, + mpf, of amplitude Jy. X Jo; and
(d) side-frequencies f. + mmf; + Mpf, of amplitude
A/a x Jm2-

Clearly, even two-tone modulation will give rise to a large number

of frequencies in the f.m. wave.

Bandwidth Required for an f.m. Wave

It is difficult to determine the exact bandwidth that is occupied by

an f.m. wave because of the large number of side-frequencies it may
contain. Usually, it is necessary to band-limit the wave and not
transmit all of the higher-order side-frequencies. Between 98% and
99% of the signal power will be retained if (m, + 1) pairs of side-
frequencies are transmitted. The minimum bandwidth required is
hence
bandwidth = 2f,,(m; + 1) (2.14)
or bandwidth = 2(kfy + fn)- (2.15)
An f.m. system will, of course, be designed to transmit the most
demanding modulating signal without excessive distortion. This signal
is the one that produces the rated system deviation and which contains
components at up to the maximum frequency fnymax) to be trans-
mitted. The minimum bandwidth that must be provided for the
transmission of this signal is given by equation (2.16), i.e.
system bandwidth = 2(f4 + fmmay)- (2.16)
34 RADIO SYSTEMS TECHNOLOGY

Restricting the system bandwidth to the value specified by equa-

tion (2.16) will cause more distortion to the sinusoidally modulated
wave than to a speech signal that contains random voltages at random
frequencies and a distributed spectrum. If an f.m. signal is passed
through a network whose bandwidth is smaller than specified by
equation (2.16) it is said to be over-deviated.

Example 2.4

The bandwidth occupied by an f.m. signal is 80 kHz when the modulating

signal is 5 V at 10 kHz. What will be the occupied bandwidth if (a) the signal
voltage increases to 10 V, (b) the frequency decreases to 5 kHz, or (c) both
(a) and (b) occur together?

Solution
From equation (2.15), 80 kHz = 2(kfy + 10) kHz or kf; = 30 kHz.

(a) New kf, = 10/5 x 30 = 60 kHz; B = 2(60 + 10) = 140 kHz.

(Ans.)
(b) B = 260 + 5) = 70 kHz. (Ans.)
(c) II: 2(60 + 5) =
B = 130 kHz. (Ans.)

Phase Modulation When a carrier is phase modulated the phase deviation of the carrier
is directly proportional to the instantaneous amplitude of the
modulating signal. Thus
O(t) = wot + k®yV,,(t), (2.17)
where ®, is the peak phase deviation permitted in the system, known
as the rated system deviation, and k (as for frequency modulation)
is the ratio (modulating signal voltage)/(maximum permissible
modulating signal voltage). The instantaneous voltage of a phase-
modulated wave is
Ve = V, sin [o,¢ + K®V,,(¢)] (2.18)
or vy. = Vz sin [wot + MVm(t)], (2.19)
where m, is the modulation index and is equal to the peak phase
deviation of the carrier.
The instantaneous angular velocity w; of the phase-modulated wave
is the rate of change of its phase, so
d dV,,(t
On ear [wt =e k®,V,,(t)] = We =e k®, oa)
dt dt
and the instantaneous frequency f, is
k®,dV,,(t)
it
f=
owes
fk 2adt
(2.20)
 ANGLE MODULATION 35

If the modulating signal is of sinusoidal waveform so that V,,(t) =

Vin SiN Wt, then
6 = wot + ky sin wyt (2.21)
ve = Vz sin (wot + k®g sin wt) (2.22)

and
fRl=aK + kGafn COS opt. (2.23)
The differences between frequency modulation and phase modula-
tion can now be seen by comparing equations (2.5) and (2.6) with
(2.21) and (2.23). The results are tabulated in Table 2.3.

Table 2.3

Modulation Frequency deviation Phase deviation

Frequency Proportional to the Proportional to the

voltage of the voltage and inversely
modulating signal proportional to the
frequency of the
modulating signal
Phase Proportional to both the Proportional to the
voltage and the voltage of the
frequency of the modulating signal
modulating signal

Equation (2.22) is of the same form as equation (2.8), where m, =

k®,, and it can be similarly expanded to show the presence of
various orders of side-frequencies. Although the two forms of angle
modulation are similar, frequency modulation is much more common.
The use of phase modulation is restricted to some mobile systems and
to high-speed data circuits. Phase modulation is more difficult to detect
since it requires an accurate reference phase; also it makes less effi-
cient use of an available bandwidth than does frequency modulation,
and so its signal-to-noise ratio is inferior.
Consider two systems, one frequency and the other phase modu-
lated, having a maximum modulating frequency of 15 kHz and a
modulation index of 5. The number of significant side-frequencies
is then 6 so that a minimum bandwidth of 6 x 15 X 2 = 180 kHz
must be provided. If the modulating frequency is reduced, in turn,
to 5, 3. and 1 kHz the modulation index will increase in the f.m. system
but will remain constant in the p.m. system. Consequently, the number
of significant side-frequencies, and hence the occupied bandwidth,
will vary as shown by Table 2.4.
 36 RADIO SYSTEMS TECHNOLOGY

Table 2.4

Frequency modulation Phase modulation

Modulating
signal Occupied Occupied
frequency Modulation bandwidth Modulation bandwidth
(kHz) index (kHz) index (kHz)

15 5 6x15x2 = 180 5 6x15x2 = 180

5 15 16x5x2 = 160 5 6x5 x25 = 60
3 25 26x3x2 = 156 5 6x3x2 = 36
1 75 VOR Ice —" oe 5 CxIx<2 = ae

Signal-to-noise Ratio in an During the transmission of a frequency-modulated signal both noise

f.m. System and interference will be picked up and will both amplitude and phase
modulate the signal. The first of these effects is easily removed by
the use of a limiter in the f.m. receiver. Phase modulation of an f.m.
signal, on the other hand, cannot be removed and it is only possible
to minimize its effects upon the output signal-to-noise ratio of the
system.
Consider the unmodulated carrier frequency and suppose that the
interfering signal is also of sinusoidal waveform. If the voltages of
the carrier and noise signals are V, and V, respectively, the phasor
diagram is given by Fig. 2.6. In general, the noise voltage will not
be at the same frequency as the unmodulated carrier and so its phasor
will rotate about the top of the carrier phasor with an angular velocity
Waitt, Where war is 2 times the difference in their frequencies. If
the noise voltage is at a higher frequency than the carrier voltage the
noise phasor will rotate in the anti-clockwise direction as shown.
The total voltage is
v = V, sin @t a _V., Sin (a... Wairpt
V, sin wt + V, sin wt cos Wairet
+ Vz COS wet SiN weiget
= (V. + V, cos wgiget) sin wt + Vi, COS Wet Sin Wgiret
V sin (wt + ¢),

So

Resultant ee os ar
phasor ofoan as

Maximum
Fig. 2.6 Interference in an f.m. phase deviation
system. Se Locus of
resultant phasor
 ANGLE MODULATION 37

where

V = [Ve + Vy cos wyiget)? + V2 sin? wait]

= [V2 + V2 + 2VV, cos waist]
and

¢ ke tan =! Vn sin Wairel |

V~ aie Vz cos Waitel

Usually, V, > V, and then

V,
yoy. =F oh cos vant
Cc

and

V,
@ = tan”! leesin vaat| = ZcSiN Weigel.

The magnitude V of the resultant voltage varies between the limits

V. + V, so that the f.m. signal is amplitude modulated. The resultant
voltage has a sinusoidally varying phase, relative to the unmodulated
carrier, with a peak phase deviation of V,/V.. The resulting
frequency deviation fgey is

fa ee ee La
ee 27 2a dt V~ oe

1 V,
Seas
oe Ware
diff oe
V. COS -Wa;diff

V,
or faey = faite ae COS Wgitet. (2.24)
Cc

The peak frequency deviation is fai¢¢-V,/V. and the r.m.s. frequency

deviation is fi¢¢Vo/J2 Ve-
When the carrier and noise voltages are at the same frequency their
difference frequency is zero and there is no output noise. The f.m.
receiver will produce an output that consists solely of the stronger
signal, the weaker signal being completely suppressed. This is known
as the capture effect.

Example 2.5

A 10 mV, 100 MHz carrier has a 25 pV, 100.1 MHz interfering signal
superimposed upon it. Calculate the peak phase and frequency deviations of
the carrier that are produced.

Solution
The peak phase deviation = (25 x 107°/(10 x 10~*) = 2.5 x 10~? rad.
(Ans.)
The peak frequency deviation = 2.5 x 107? x 0.1 x 10 = 250 Hz.
(Ans.)
38 RADIO SYSTEMS TECHNOLOGY

Noise Output Power

The noise power 6Np at the output of the detector in an f.m. receiver
in an audio bandwidth of B Hz is proportional to V2f%i¢,/2 V2.
Now V2/2 = No = Péf, where P is the noise power density spectrum
in watts/hertz. Hence

KPf aie Of
5No = a

(K is the transfer function of the f.m. detector, i.e. an input Af

produces KAf volts at the output.)
The total noise output power in the audio bandwidth is

_ ' KPFindf _ 2KP i

7” y2 y2
Pe
Sard f
0

_ 2KP |faisr |
vi Sxklo
2KPB?
or No (2.25)
3V2
Thus the noise output power is proportional to PB?/V2.

The Triangular Noise Spectrum

Random noise can be considered to consist of a large number N of

equal-amplitude, equally spaced, sinusoidal voltages in a bandwidth
of N Hz. The previous analysis can then be applied to demonstrate
that an f.m. system has a triangular noise voltage spectrum and a
parabolic noise power spectrum.
The output voltage of an f.m. detector is proportional to the
frequency deviation of the input voltage. This means (equation (2.24))
that the noise output voltage is proportional to frequency and hence
has a triangular spectrum. The output noise voltage rises linearly from
zero volts at zero frequency to a maximum, equal to V,/V., at the
frequency equal to the rated system deviation f; (see Figs 2.1):
Usually, the passband of the audio amplifier which follows the detector
will be less than the rated system deviation and it will remove all noise
voltages at frequencies higher than fm(max)- The output noise voltage
of an a.m. receiver with the same signal and noise voltages would
be proportional to V,/V,; thus the areas enclosed by the points ADE
and ABCD represent, respectively, the output noise voltages of f.m.
and a.m. systems having the same audio passband. Area ADE is
smaller than area ABCD, which is an indication that the f.m. system
has a smaller output noise than the a.m. system. This means that the
use of frequency modulation can provide an increase in the output
signal-to-noise ratio of a system.
 ANGLE MODULATION 39

|<
os

voltage
Noise
output

A fm(max) fs
Fig. 2.7. The triangular noise
spectrum. Frequency (kHz)

Signal-to-noise Ratio Improvement

The signal output power of an f.m. system is K (kf;)/2 and the noise
output power is equal to 2KPB 3/3V2 and so the output signal-to-
noise ratio is
Bibep Py Wap gun:
B 4PB (3m) 4PB
An a.m. system, modulated to a depth of m%, has an output signal-
to-noise ratio of
m°P, m?V2
2PB 4PB
For a proper comparison of the signal-to-noise ratios assume that both
systems are 100% modulated, i.e. mp = D and m = 1. Then
signal-to-noise ratio (f.m.) = 3D)? (2.26)
signal-to-noise ratio (a.m.)
or, in decibels,

Signal-to-noise ratio improvement = 20 logo(/3D). (2.26a)

Example 2.6

An a.m. system has an output signal-to-noise ratio of 30 dB. What output

signal-to-noise ratio could be obtained if frequency modulation with a devia-
tion ratio of 5 were employed instead and the transmitted power were reduced
by 50%?

Solution
From equation (2.26) the improvement in signal-to-noise ratio is
20 log,o(|3 x 5) = 18.75 dB. The 50% reduction in the transmitted power
is equivalent to 10 log,) 2 = 6 dB reduction. Therefore, the new signal-to-
noise ratio = 30 + 18.75 — 6 = 42.75 dB. (Ans.)
 40 RADIO SYSTEMS TECHNOLOGY

Pre-emphasis and De-emphasis

An improvement in the output signal-to-noise ratio of an f.m. system

can be achieved if the modulating signal is pre-emphasized before
it modulates the carrier. The signal is passed through a pre-emphasis
network, Fig. 2.8(a), which amplifies the high-frequency components
of the signal more than the low-frequency components. The voltage
gain of the circuit is A, = g,,(R, + jwL) = g,R(1 + wr), where
T is the time constant, L/R seconds, of the circuit. The signal is then
distorted and it will be necessary at the receiver to restore the various
components of the signal to their original amplitude relationships. This
step is achieved by passing the signal through a de-emphasis circuit
such as the one shown in Fig. 2.8(b). For this circuit,

Vin PGE. PAW,

™ MRE thee. - TEEGGER
and so

Vout = 1

V; JL ew R°C*)e

This circuit will have a 3 dB loss when

Vax 1 1
Ve* LP Nd + etR2Ch.
i.e. at the frequency f,, where
1
fi
aeRO”
Thus

air 1
Vi shal Ul ie

oO
=

&5 +15
12)

S=) +12
S +9
= Pre-emphasis
z +6 characteristic

eo
$
= 0
2 -3
fea]
2 6 De-emphasis
=S 49 characteristic
Fig. 2.8 (a) A pre-emphasis circuit,
(b) a de-emphasis circuit, and Set
Bails
(c) pre-emphasis and de-emphasis = 0 3182 6364 12 728
characteristics for 50 us time (eo)
Frequency (Hz)
constant.
(c)
 ANGLE MODULATION 41

At frequency f = 2f,
Vos 1
“| = or —7 dB;
Vin 1+4
at frequency f = 4f,
Lacs 1
a = —12 dB,
V; Inch 16
and so on. Therefore at higher frequencies than f; the attenuation of
the network falls at 6 dB/octave. To ensure that the component
frequencies are restored to their original amplitude relationships the
time constants of the pre-emphasis and de-emphasis networks must
be equal. In the UK sound broadcast system a time constant 7 = L/R =
CR = 50 us is used. Then
x 1
fi = 3183 Hz,
20 Ka 10
giving the pre-emphasis and de-emphasis characteristics shown in Fig.
2.8(c).
To determine the effectiveness of pre-/de-emphasis in increasing
the output signal-to-noise ratio of an f.m. system it must be compared
with an a.m. system with the same value of fimax). Assume a 1 Q
resistance; then for the a.m. system the output noise voltage V,, is
constant with frequency and hence
Sm(max) > df
a.m. noise power = |
0 i ee tL
eae
Using a standard integral
1
| i = — tan ‘(ax),
fa a
this gives V2, filtan~ (f/f, Yom), Therefore,

tan”! eel 227)

Novain) te Viol
1

The noise output voltage of an f.m. system is proportional to

frequency, i.e. V,. f/f. Hence
2 d
f.m. noise power = Vi, |
fm(max)
ia if |
0 ee Ee iis

5 fe faf
fa Jo i ie
Using another standard integral
2
nee = x7fax — tan~ '(ax)],
bax
gives
 42 RADIO SYSTEMS TECHNOLOGY

y2 3
f.m. noise power = ss1 Be faint = )
fa fi fi
(2.28)
If f,= 3183 Hz, finmmaxy) = 15 kHz, andf; = 75 kHz, the ratio

a.m. noise power _ 4334 _ 225.68 or 23.54 dB.

f.m. noise power 19:2
Without pre-emphasis the signal-to-noise ratio improvement of f.m.
over a.m. was 18.75 dB so that the increase provided by the use of
pre-emphasis is 23.54 — 18.75 or 4.8 dB.

Frequency Modulators A frequency modulator converts an input signal voltage into an output
frequency change. The change in the carrier frequency will be equal
to the rated system deviation when the signal voltage is at its maximum
value. The sensitivity of the modulator is equal to the ratio f,/ Fragnee)
in kHz/V. Ideally, there should be no amplitude modulation of the
carrier. Frequency modulators are classified as being either direct
or indirect types.

Direct-frequency modulators

Figure 2.9 illustrates the basic principle upon which the majority of
direct-frequency modulators are based. The frequency of oscillation
of an oscillator is determined by a parallel-tuned circuit whose total
capacitance (or, much less often, its total inductance) is provided by
the parallel connection of a physical capacitor C, and a voltage-
variable reactance. The modulating signal voltage is applied, usually
togeth>r with a bias voltage, to the voltage-variable reactance to vary
its capacitance. The change in capacitance will then alter the resonant
frequency of the tuned circuit and so vary the frequency of oscilla-
tion. In this way the modulating signal voltage is able to modulate
the carrier frequency.
When the modulating signal voltage is zero the voltage-variable
reactance has its average value C, of capacitance and the oscillation

+Voc

Ly

Variable-
Modulating reactance
Spee
signal circuit

Fig. 2.9 Principle of a frequency

modulator. = =
 ANGLE MODULATION 43

frequency will be the unmodulated carrier frequency, 1.e.

ie as H Ree ee :
var Gale oe (2.29)
;
be eer
IE (Cy tiC)) 2m \(L1 Cr)
When the modulating signal voltage is instantaneously equal to v,,
volts the total capacitance of the tuned circuit is varied by an amount
6Cy. Then
1
f. + Of, = (2.30)
2m |IL\(Cr + 6Cy)]
Divide equation (2.30) by (2.29) to give

eee ssf oe )- _ AZ
re Cy + 5Cy JQ. + 6Cy/Cy)
Since 6C; < Cy equation (2.31) can be written as
5
Ic 2Cy
5 oe (2.32)
or Yen =
ip 2Cy
This means that a fractional increase in Cy will produce a fractional
decrease in f, which is approximately half as large.

Reactance Frequency Modulator

The circuit of a reactance frequency modulator that employs a bipolar

transistor is shown in Fig. 2.10. R,, Ry, R3 and C; are bias and
decoupling components, L; is an r.f. choke and C; is a d.c. block.
The required reactance effect is provided by C, and Ry. The current
i flowing in the series C)R, circuit is i = Vie/(Rq + 1/jwC,) and this
develops a voltage

VJ wy Ry
V.. =
a pied+

Voc

Oscillator tuned circuit

7t i ! t 1 aiI J

Modulating
signal

Ro

Fig. 2.10 Reactance frequency

modulator.
 44 RADIO SYSTEMS TECHNOLOGY

across Ry. The collector current i, of T, is equal to g,,Vj, and so the

output admittance Y,,, of the modulator is

Yuout = i—— = J8m CrRy = wC, 244

Ve je joGR, J&m Ry. (2.33)

The circuit thus provides an effective capacitance of C. = gm CR,

across the oscillator’s tuned circuit that is directly proportional to the
mutual conductance g,, of the transistor. Since g,, can be varied by
applying the modulating signal to the input terminals of the circuit,
frequency modulation of the carrier is achieved.

Example 2.7

The reactance frequency modulator of Fig. 2.10 has C, = 5 pF and R, =

1 kQ. The total tuned circuit capacitance when there is no input signal is
500 pF and the unmodulated carrier frequency is 1 MHz. Calculate the devia-
tion of the carrier frequency when a modulating signal varies the mutual
conductance of T, by 2 mS.

Solution

6C = 2°80"? «5 « 107" x 10° = 10ipe.

From equation (2.32),
—1 x 10° x 10
of. = = —10 kHz. (Ans.)
I 2 x 500
Varactor Diode Frequency Modulator
The basic circuit of a varactor diode frequency modulator is shown
in Fig. 2.11. The varactor (or voltage-variable) diode is connected
in parallel with the tuned circuit C,L, of the oscillator to be modu-
lated.C, is a d.c. blocking component while L, is an r.f. choke.
When the modulating signal voltage is zero the varactor diode is
reverse-biased by the bias voltage Vz volts and it then has a capaci-
tance of Cy = Cp/.|Vg pF, where Co is the capacitance for zero
applied voltage. The frequency of oscillation is then

Modulating
signal

Oscillator
tuned circuit Bias voltage

—
Ly

Fig. 2.11 Varactor diode frequency

modulator.
 ANGLE MODULATION 45

k= : Hz
2a {IL(C, + Cal
When a modulating signal is applied to the circuit the reverse-bias
voltage becomes Vg + V,, sin w,t and

Cy + 6Cg =
Co
\ [Vg + Vm Si &nt)
= Co
V,
JV JO + — sin ont)
Vp

C
Cy aia 6Cq = : .

J be Aeeesino, f
Vp
Therefore,
6C, 1 V,
license aot = 1— — sin ont
Gs Vin 2Vz
(:Sit ont
Vp
and

een d = av, 1 og
d SiN m ge
Wt. (2.34)
;

Equation (2.34) can be substituted into equation (2.32) to obtain an

expression for the deviation of the carrier frequency. Better linearity,
with a consequent reduction in signal distortion, is obtained if two
back-to-back varactor diodes are employed (see Fig. 2.12). The effec-
tive capacitances of the two diodes are now in series and so each must
have twice the capacitance of the single varactor diode they replace.

Example 2.8

The circuit of Fig. 2.12 is a part of a frequency modulator which operates

at a centre frequency of 1 MHz. L, = 100 wH, C, = 200 pF and each
varactor diode has C, = 200 pF. Calculate the values of the bias voltage
and the peak modulating voltage if the frequency deviation is 1 kHz. Also
calculate the sensitivity of the modulator.

Solution
The necessary tuning capacitance is equal to
1
= 253 pF
4a? x 10x 100 x 107°
and so the varactor diodes must contribute 53 pF. Each diode must therefore
46 RADIO SYSTEMS TECHNOLOGY

Modulating
signal

Fig. 2.12 Improved varactor diode

frequency modulator.

have a capacitance of 106 pF. This means that the reverse bias voltage V,
must be

Vz = (200/106)? = 3.56 V. — (Ans.)

When the carrier frequency is deviated by 1 kHz
110° 5Cy
or 6C; = 0.5 pF
1x 10° 2x 253 x 10-2
and then C; is either 253.5 pF or 252.5 pF. Each varactor diode must then
have a capacitance of either 107 pF or 105 pF. Choosing the former
Vs — V,, = (200/107)* = 3.49 V and the peak modulating voltage is
3.56 — 3.49 = 70 mV. (Ans.)
On the other half cycle, Vz + V,, = (200/105)? = 3.63 V and the peak
signal voltage is 3.63 — 3.56 = 70 mV. (Ans.)
The sensitivity of the modulator is 1/(70 x 107*) = 14.286 kHz/V.
(Ans.)

The varactor diode and reactance frequency modulators produce

n.b.f.m. signals because 6Cy « Cy and if wideband modulation is
wanted some stages of frequency multiplication will be necessary.

Indirect Frequency Modulation

The inherent stability of the unmodulated carrier frequency when a

direct modulator is used is not good enough to meet modern standards.
There are two ways in which this difficulty can be overcome: (a)
direct frequency modulation with automatic frequency control applied
to the transmitter; and (b) indirect frequency modulation of a very
stable crystal oscillator.
The expressions for the instantaneous frequency and voltage of a
sinusoidally modulated phase-modulated wave are given by equations
(2.22) and (2.23), respectively. If the modulating signal v,, = V,, sin
Wmt is integrated before it is applied to the phase modulator, to give
Vin/Wm COS Wt, these equations become fi =f. +k qsin w,,t and

: k®,
Ve = V, sin.| w¢ + — cos wf }.
Wm
 ANGLE MODULATION 47

The frequency deviation is now proportional to the voltage of the

modulating signal only, and inversely proportional to the modulating
signal frequency. These are, of course, the characteristics of a
frequency-modulated waveform. Hence, an f.m. wave can be obtained
by integrating the modulating signal and then using a phase modulator.

Phase Modulators A phase-modulated wave can be generated using an a.m. balanced

modulator and a carrier source that has both in-phase and quadrature
outputs. Figure 2.13 shows the circuit of an Armstrong phase
modulator. The balanced modulator has two inputs: the modulating
signal and a 90° phase-shifted carrier voltage and it generates the upper
and lower sidebands of amplitude modulation but suppresses the
carrier. Assuming the modulating signal to be of sinusoidal waveform,
the output of the balanced modulator is mV, cos w.t sin w,t. This
is added to the non-phase shifted carrier voltage to give an output of
Vour = Ve sin wot + mV, cos wf SiN Wyt
= |V2+ mV2 sin? wat) sin (wt + 6),

where 0 tan~! (m sin w,,t). Hence,

Vour = Ve sin (wot + m Sin pf)

which is of the same form as equation (2.22).
The carrier can be produced by a stable crystal oscillator. The
maximum modulation index that this circuit can give is only about
0.25 and so several stages of frequency multiplication will be necessary
to obtain a wideband f.m. wave. If, for example, the modulation index
is 0.25 and it is to be equal to 5, frequency multiplication of 20 times
is required. The crystal oscillator frequency should then, of course,
be equal to f./N, where f, is the wanted carrier frequency and N is
the multiplication ratio.

Example 2.9

An f.m. transmitter has an unmodulated carrier frequency of 90 MHz and

a rated system deviation of 75 kHz. The Armstrong modulator used can only
produce a peak frequency deviation of 3 kHz. Calculate (a) the multiplication
ratio used, and (b) the frequency of the crystal oscillator.

Balanced
modulator
Modulating
signal

90° phase-
shifting circuit Phase-modulated
output

Fig. 2.13 Armstrong phase Crystal

oscillator
modulator.
 48 RADIO SYSTEMS TECHNOLOGY

Solution
(a) N = 75/3 = 25. (Ans.)
(b) Oscillator frequency = 90/25 = 3.6 MHz. (Ans.)

An alternative form of phase modulator is shown in Fig. 2.14. The

buffered output of a crystal oscillator is applied to a parallel-tuned
circuit whose resonant frequency can be varied by applying the
modulating signal to the varactor diode. The output voltage of the
modulator will then be phase modulated. Capacitor C, is an r.f.
bypass and C; is a d.c. block.

Example 2.10

The varactor diode used in the circuit of Fig. 2.14 has the capacitance-voltage
characteristic C; = 200/VV pF. When the d.c. reverse-bias voltage is 4 V
the circuit is resonant at 5 MHz and has a Q-factor of 20. Calculate the peak
phase deviation produced when a 50 mV peak sinusoidal modulating signal
is applied to the circuit.

Solution
When the modulating signal voltage V,, is zero C, = 200/V4 = 100 pF.
When the modulating signal is applied to the circuit the minimum and
maximum values of C, are

Caminy = 200/V4.05 = 99.38 pF

and Cymax) = 200/V3.95 = 100.63 pF. Hence 6C is either 0.62 pF or it is
0.63 pF; using the mean value in 6f./f. = —8C/2C gives
tS LO 0625
of. = 15.625 kHz.
200
The impedance Z of the tuned circuit at any frequency is given by

1 + jOB/,
where R, is the dynamic impedance, Q is the Q-factor, B is the bandwidth
considered and fq is the resonant frequency. The angle @ of the impedance
is ttan! (QB/fy) or

Vp

Crystal
oscillator
C. fV(t)
Phase-modulated
Fig. 2.14 Phase modulator.
output
 ANGLE MODULATION 49

20 x 31 250
a= stan | |= +0.124rad. (Ans.)
5 x 10

Frequency-modulation Detectors A frequency-modulation detector is required to produce an output

voltage that is directly proportional to the instantaneous frequency
of its input signal. It has a transfer function with units of V/kHz.
Before the widespread use of integrated circuits in radio receivers,
most f.m. detectors were based upon the principle of first converting
the f.m. signal into a signal whose amplitude varied in proportion
to the frequency deviation and then envelope-detecting in the converted
signal. The concept is illustrated, in block diagram form, by Fig. 2.15.
The two most commonly employed examples of this technique are
the phase discriminator and the ratio detector (see Radio Systems for
Technicians).
Most modern radio receivers make full use of the widely available
ICs and these receivers are most likely to employ either a quadrature
detector or a phase-locked loop (p.\.1.) (see Electronics IV).

The Quadrature Detector

The f.m. signal to be detected is split into two parts. One part is
directly applied to one of the two inputs of an analogue multiplier
circuit. The other part is passed through a capacitor having a high
reactance at the unmodulated carrier frequency and then to a parallel-
tuned circuit, see Fig. 2.16. The capacitor C, introduces a phase shift
of very nearly 90° and the tuned circuit a phase shift that depends
upon frequency. The impedance of the tuned circuit is

Rg
1 + jQB/
At the resonant frequency B = 0 and the circuit is purely resistive;
at any frequency off of resonance the circuit introduces a phase shift
of +tan~' [QB/fo]. The voltage across the tuned circuit is applied
to the other input of the multiplier and an output proportional to the
product of the two input signals is obtained. This output is passed
through a low-pass filter to obtain the original modulating signal
Vn(t).
If the unmodulated carrier frequency is w,/2a and the instan-
taneous frequency of the input signal is w = w, + dw, then the phase
shift @ between the two multiplier inputs is

Frequency-modulated Frequency-to- Envelope Detected signal

input voltage detector
converter

Fig. 2.15 Principle of an f.m. detector.

50 RADIO SYSTEMS TECHNOLOGY

Frequency-modulated
input Detected
Analogue output
multiplier

Fig. 2.16 Quadrature detector. =

0 = (n/2) + tan~! [205u,/w,] = (4/2) + o.

(Plus sign if w > w,, negative sign if w < w,.) If the input signal
is V sin wt the phase-shifted signal will be V sin (wt + (x/2) + 9),
or V cos (wt + ). Multiplied together these two signals give
V2
V? sin wt cos (wt + o) = ae [sin (2wt + ¢)

+ sin (+ ¢)]

(using the identity 2 sin A cos B = sin (A + B) + sin (A — B)).

The output of the low-pass filter is

Ve ye _ 1,205,
- Sas aeie hanes ea(2i === |
We
For 2Q6w, < w,,
; ( oa ae) 2 06w,
sin {|tan7 ©————}) = ———.
We We
This output voltage bears a linear relationship to dw,, i.e. to the
frequency deviation of the carrier, and it is therefore the desired
modulating signal.
The quadrature detector is widely employed in modern radio
receivers and it is usually incorporated within an IC that also performs
several other circuit functions. The detector can operate with small
amplitude signals (100 1 V or so) and it is easy to set up; it is only
necessary to tune the phase-shifting circuit to the incoming carrier
frequency. The circuit provides good linearity, with consequent small
distortion, as long as the frequency deviation is not more than about
1% of the unmodulated carrier frequency. The quadrature detector
tends to produce a noise output when there is no input signal and so
the IC usually includes a muting, or squelch, circuit, too.
If the analogue multiplier is replaced by an AND gate and the input
signal is hard limited to produce a variable-frequency pulse waveform,
a coincidence detector is obtained.
 ANGLE MODULATION 51

Frequency-modulated Detected
input Amplifier ~ output

Fig. 2.17 Phase-locked-loop

detector.

The Phase-locked-loop Detector

The phase-locked loop (p.1.1.) can be operated as a detector of f.m.

signals when it has more or less the same performance as the quadra-
ture detector. The p.1.1. detector does not need an externally connected
tuned circuit but, on the other hand, an external piezo-electric crystal
is required. In some cases the 10.7 MHz output of an i.f. amplifier
may be converted to a lower frequency in the region of 200 kHz before
it is applied to the p.1.1. detector. Sometimes the mixer used is on-chip.
with the p.1.1.
The block schematic diagram of a p.1.1. f.m. detector is shown in
Fig. 2.17. The free-running frequency fo of the voltage-controlled
oscillator (v.c.o.) is set to be equal to the unmodulated carrier
frequency f. of the signal to be demodulated. The phase detector
generates an output voltage that is directly proportional to the phase
difference, or phase error, between the f.m. signal and the v.c.o.
voltage. This error voltage is passed through a low-pass filter and
then amplified to produce both the output voltage and a control voltage
for the v.c.o. The free-running frequency of the v.c.o. is varied by
the voltage applied to its control terminal. The polarity of the control
voltage is always such that it varies the frequency of the v.c.o. in
the direction that reduces the difference frequency Af = f. — fo and
hence the phase error. This means that the frequency of the v.c.o.
is forced to try to become equal to the instantaneous frequency of
the input signal.
Suppose that, initially, the input signal is not modulated. The action
of the p.1.1. reduces the difference between the input carrier frequency
f; and the frequency fo of the v.c.o. Once the v.c.o. frequency has
moved to be very close to the carrier frequency the loop attains lock.
Then the v.c.o. rapidly attains the same frequency as the input signal
but there is always a phase difference between the two voltages. This
phase error must always exist in order to maintain the v.c.o. control
voltage at the required value. This voltage, and hence the output
voltage, is a d.c. voltage.
When the input signal is frequency modulated a similar action takes
place. As the input signal frequency deviates from the unmodulated
carrier frequency the error voltage varies too to ensure that the v.c.o.
tracks to minimize the phase error. As a result the instantaneous v.c.o.
frequency is always approximately equal to the frequency of the
incoming signal. The output of the low-pass filter is the detected
modulating signal voltage. The filter must have a cut-off frequency
52 RADIO SYSTEMS TECHNOLOGY

equal to the maximum modulating frequency fignax) Of the f.m. signal

in order to minimize the effects of noise and interference. Should the
free-running frequency of the v.c.o. not be set to be exactly equal
to the unmodulated carrier frequency the detected output voltage will
have a d.c. component, but this is (usually) unimportant.

Analysis

Once the loop is in lock the frequency of the v.c.o. is equal to the
carrier frequency. The f.m. input signal is

vs = Vs sin [wt + 2akfy |) Ve(O)]

Vs sin [wt + Ks |, V(t) dt] = Vg sin [wt + 0,(2)).
The output voltage of the v.c.o. is
Vo = Vo cos [wt + 6,(t)].

These two signals are the inputs to the phase detector which generates
an output voltage vy that is directly proportional to their product.
Thus
Va = Kp VsVo sin [wot tr 0, (t)] COS [wt ++ 6>(t)]

Kp VsV,
= SS Esin [01(0) — 0(0)]
+ sin [2 wt AIF 0,(t) a6 6,(t)] }.

Kp is the gain factor, in V/rad, of the phase detector. The low-pass

filter will only transmit the lower-frequency components of Vq So that
the voltage appearing at the output of the circuit is
Ky AVVV;
ies sini sin [,(t) — 0,(t)].
Since the loop is locked the phase error will be small and

sin [0,\(¢) — 0,(t)] = 0,(t) — 6,(t),

giving
Kj AVVV,
ye — [6,(t) — 0,(¢)). (2.35)

The instantaneous angular velocity w of the v.c.o. output voltage

1S, G6 = .Gig.ct Kovo(t), where wo is the free-running angular velocity
and Ko is the conversion gain of the v.c.o. The conversion gain Ko
relates the frequency of the v.c.o. to its control voltage and it is
expressed in rad/s/volt or in kHz/volt. Since w = d6/dt, d6,/dt =
Kovo(t); also

dé,/dt a KsV,,(t).

Equation (2.35) can be written as

 ANGLE MODULATION 53

KpAvVsV,
VIO see TL) Ke J volt) de],
and differentiating with respect to time,
dvo(t) KpAvyVsVo d6,(t)
= — Kovolt
dt 2 dt neo
dd, (¢ 2d vo(t
10) = olf) ai Ko vo(t) = Kovo(t).
dt dt(KpAyV5Vo)
Therefore,

spotless
3 —-
ditn 07
06
Vecl]1 14. NC
Input [2 13.) NC 5
Input L] 3 12|_]NC
v.c.o. output] 4 11 ]NC 4¢
Phase det. inputL] 5 10[ 1+ Vo
Ref. output L] 6 9{_]v.c.o. external C
Detected output L] 7 8) v.c.o. external R

(a) (6)

+ Voc

Frequency- Detected
modulated * ~ output
input

Fig. 2.18 The 565 p.l.i. IC: (a) pin

; —
connections, (b) internal block
diagram, (c) connection as an f.m.
detector. (c)
54 RADIO SYSTEMS TECHNOLOGY

KsV n(t) = Kovo(t),

K.
or “¢) = — VD? (2.36)
Ko
The ratio Ks/Ko has the dimensions of kHz/kHz/V or volts and so
the output of the detector is the wanted modulating signal V,,(t).

The 565 as a Frequency Demodulator

The 565 IC is a widely used p.1.1. for operation at frequencies up
to about 500 kHz. Figure 2.18(a) and (b) show, respectively, the pin
connections and the internal block diagram of the 565. The free-
running frequency of the v.c.o. is set by the external components C,
and R, according to the expression f = 0.3/C,R, Hz. Capacitor C,
can be of any value but R; should be somewhere in the region of
4000 Q. R, is an internal 3.6 kQ resistor which, together with an
external capacitor C,, forms the low-pass filter. Figure 2.18(c)
shows how the 565 should be connected to act as an f.m. demodulator.
The pins 4 and 5 must be connected together so that the v.c.o. output
is applied to the phase detector.
Transmission Lines

A transmission line consists of two conductors separated from one

another by a dielectric; the two conductors may be identical and form
a twin line or one conductor may be positioned inside the other (a
coaxial line). A transmission line provides a means for guiding an
electromagnetic wave from one point to another. The behaviour of
a transmission line is normally described in terms of the current and
voltage waves which propagate along it. The performance is deter-
mined by the values of its secondary coefficients. These are: the
characteristic impedance Zp, the propagation coefficient y, and the —
velocity of propagation v. These, in turn, are functions of the primary
coefficients of the line, i.e. the resistance R, the conductance G, the
inductance L and the capacitance C. The full expressions for the
secondary coefficients involve all four of the primary coefficients and
also the frequency (Electrical Principles IV), but at radio frequencies
these expressions can be reduced to simpler, and more convenient,
forms. At radio frequencies the characteristic impedance Zp of a line
is purely resistive and is given by Zp = \(L/C) Q. The propagation
coefficient y is given by

a = Me Se Len jaVLC.
2Zp Z
The real part of y is known as the attenuation coefficient a in nepers/
metre (or dB/m), and the imaginary part is the phase-change coeffi-
cient B in rad/m. The attenuation coefficient is not a constant quantity
but instead it increases with increase in frequency; this is because
the resistance R is proportional to frequency and the conductance
G is directly proportional to frequency. Usually,

oy babi
2Zo 2
and then the dielectric loss can be neglected. The attenuation
coefficient a is then equal to the conductor loss R/2Zp nepers per
metre.
The attenuation of an r.f. line is proportional to v frequency
while the wavelength is inversely proportional to frequency. This
means that the attenuation per wavelength decreases with increase
in frequency. At the higher radio frequencies, in particular, the
electrical length of a line is often small, perhaps only a fraction of
a wavelength, and then the line loss will be small. Such lines are often
 56 RADIO SYSTEMS TECHNOLOGY

described as being low-loss or even, if the losses are small enough

to be neglected, as loss-free.
At radio frequencies the phase and group velocities are of equal
value and given by
ERGO TYE /
- 8 VEC m/s.

At the lower radio frequencies twin conductors may be employed

since they are generally less expensive, but at higher frequencies their
losses may be excessively high because of radiated energy. At the
higher radio frequencies, therefore, coaxial lines are employed. If
an air-spaced line has a characteristic impedance of Zp then the use
of a continuous dielectric of relative permittivity €, will reduce the
impedance to Zp/Ve,. When the inner conductor of a coaxial line is
held in position by spacing discs of thickness ¢ and relative permit-
tivity €,, spaced distance d apart, the characteristic impedance is
equal to
7

*Staaras d

Example 3.1

The inner of a coaxial cable is supported by 0.64 cm thick discs spaced 5 cm

apart. The disc material has a relative permittivity of 2.3 and the air-spaced
value of the characteristic impedance is 70 ohms. Calculate the characteristic
impedance of the cable.

Solution

70
= = 64.8 0). (Ans.)
Jf Cee DD ast)
5
When a transmission line is used to transmit energy from one point
to another the load impedance is chosen, as far as possible, to match
the characteristic impedance of the line. This will ensure both the
maximum transfer of energy to the load and the absence of standing
waves on the line. The behaviour of a matched transmission line is
fairly straightforward and an understanding of it is assumed in this
chapter (see Electrical Principles IV).

Mismatched Transmission Lines Whenever the load terminals of an r.f. line are closed in an impedance
that is not equal to the characteristic impedance of the line, the load
will be unable to absorb all of the incident power. A fraction of the
incident power will be reflected by the load and transmitted back
towards the sending end of the line. If the sending-end terminals are
 TRANSMISSION LINES 57

Vi > — Vie” matched to the source impedance all of the reflected power will be
fo — ==ije7" dissipated in the source impedance. If, however, the sending-end
terminals are also mismatched some of the returning energy will be
further reflected and multiple reflections will take place.
Figure 3.1 shows a line of length / metres, having secondary
coefficients Z) and y, which is terminated in a load impedance Z,
py Vie
—2y/
Sas = py Ve
Sf
x
that is not equal to Z). The sending-end terminals of the line are
pile
—2y!
— + pile
= matched. When the source is first connected to the line the impedance
presented to it is the characteristic impedance of the line Zp). The
Fig. 3.1 Currents and voltages on a incident current J; and voltage V; into the line are therefore equal to
-mismatched transmission line. E,/2Z) and E;/2, respectively. The incident current and voltage
waves propagate along the line and experience both attenuation and
phase change as they travel. At the distant end of the line the magni-
tudes of the incident current and voltage waves are [e~™ and
Vie—V', respectively.
Since the load impedance Z, is not equal to the characteristic
impedance Zp both the incident current and the incident voltage
waves are reflected. The values of the reflected waves are determined
by the current, and voltage, reflection coefficients p; and p,, respec-
tively. The reflected current is equal to ple and the reflected
voltage is p,Vje ~The reflected power is, of course, equal to the
product of the reflected current and the reflected voltage.
The reflected waves propagate along the line towards its sending
end and are further attenuated and phase shifted as they travel. At
the sending end of the line the reflected current is p;/je ~27! and the
reflected voltage is p,V,e~*”'. Since the source impedance is equal
to the characteristic impedance of the line no further reflections occur.
At any point along the line the total current and voltage are the phasor
sum of the incident and the reflected waves. At a distance x from
the sending end of the line
V, = Ve-*% + p, Ve. (3.1)
This can be written as

Vi, = Vie? sae (3.2)

where V, is the reflected voltage at the receiving end of the line, i.e.
V, = py V.e—™'. Similarly, the current at any point distance x from
the sending-end terminals is
I, = Ie7*% + Le™ (3.3)
(3.4)
V. yx V. “yx
Ce ee ee
Zo Zo
The minus sign is necessary because the reflected current is always
in antiphase with the reflected voltage.
It is sometimes convenient to be able to express the current and
voltage on a mismatched line in terms of the hyperbolic functions
cosh x = (e* + e *)/2 and sinh x = (e* — e *)/2. At the sending
end of the line x = 0 and, from equations (3.2) and (3.4),
58 RADIO SYSTEMS TECHNOLOGY

VeeuKaittoy

and Is tee
V, Sok See
V,
Lo. La)
Adding these two equations together gives V; = (Vs + IsZp)/2, and
subtracting gives V. = (Vs — I;Zo)/2. Equation (3.2) can now be
written as

V, Es = oh IsZo Jem + be o gales

2 2
x —yx x —yx
¥ frog
- = = Igo] oa—>|
or V, = Vs cosh yx — IgZp sinh yx. (3.5)
Similarly

I, = Is cosh yx — = sinh yx. (3.6)

Equations (3.5) and (3.6) are known as the general line equations.
The current and voltage at any point on a mismatched line may also
be expressed in terms of the current and the voltage at the load
terminals. At a distance x from the load terminals
V, = Vie™ + Vee~™ (3.7)
V, yx V, seal C3
Te eA (3.8)
Zo Lo
where V, is the incident voltage at the load, i.e. V; = Vie~”’, and
Vp is the reflected voltage at the load, i.e. Vp = pyV.
At the load terminals x = 0 and the load voltage V;, is V, =
V; + Vp. Also, IL.Zy = V; — Vp, where fj, is the load current.
Following the same steps as before leads to
V, = V, cosh yx + [Zo sinh yx (3.9)
Ve.
[, = h, cosh yx + Zh sinh yx. (3.10)

Equations (3.9) and (3.10) are the alternative form of the general line
equations.

Voltage and Current Reflection Coefficients

The voltage reflection coefficient p, of a mismatched line is the ratio

(reflected voltage)/(incident voltage) at the load. Similarly, the current
reflection coefficient p; is the ratio (reflected current)/(incident
current) at the load. Putting x = 0 in equations (3.7) and (3.8) makes
it possible to obtain the load impedance in terms of the load current
and voltage. Thus
 TRANSMISSION LINES 59

Vy V; + Vp
ee
h Zh
Rearranging gives
ZL pa 1 + p,
Zo 1 = VR/V; 1 — Py

Therefore,
5 ie
in ee
ge (3.11)
Zt La
Since the reflected current is always in antiphase with the reflected
voltage

CP Ay Zi
Di Get Ay
Zt Z
There are two related terms that are sometimes referred to in the
literature. These are:
ane
Incident power 4B
(a) return loss = 10 logio
reflected power

= 20 logig B dB, (3712)

v

ree
LS ee lel |a
(b) reflection loss |= 10 logy
load power

1
10 logio eee dB. (3.13)
=p
Ley

Example 3.2
Anr.f. line has a characteristic impedance of 50 Q, 3 dB loss, and it is \/2
long. The line is terminated by a load of 100 + j20 ©. Calculate the voltage
reflection coefficient at the load. The line is fed by a source of 50 0 impedance
and 2 V e.m.f. Calculate the sending-end and load voltages.

Solution
From equation (3.11),
100° 420 — 50 ;
py = = 0.345 + j0.087 = 0.36 214°.
100 + j20 + 50 (Ans.)

The incident voltage is 1 V. 3 dB is a voltage ratio of {2:1 so that the incident

voltage at the load is 0.707 2 —180° V. The reflected voltage at the load
is 0.707 z2 —180° x 0.36 214° = 0.255 2 —166° V. The reflected voltage
at the sending end of the line is 0.255 z —166° x 0.707 2 —180° or
0.18 214° V.
60 RADIO SYSTEMS TECHNOLOGY

(a) Sending-end voltage = 1 + 0.18 214° = 1.175 + j0.044 =

1,176 )22.2°-Vi~ Ans.)
(b) Load voltage = 0.707 z-—180° + 0.255 2-166° =
02956° 2. — 176° V. .\ (Ans.)

Voltage Reflection Coefficient at any Point on a Line

At the mismatched load p, = V;/Vp. At any distance x from the load
Vere ™
Py) = Vier

OF Pyq) = pyen?™. (3.14)

Example 3.3
A loss-free line has a characteristic impedance of 60 0 and a load impedance
of 120 Q. Calculate the voltage reflection coefficient (a) at the load, (b) at
a distance of \/8 from the load, and (c) \/8 from the load if the line loss
is 8 dB per wavelength.

Solution 120 — 60
(a) p, = ——— = 1/3. (Ans.)
120 + 60
(P) Puy Geo oe es a ONS ne)
(c) Py (x)
—
=
|
Ze
,—2(0.115+j 7/4
sin = ze 97? » 90° = 0.267 2 —90°.
(Ans.)

Input Impedance of a Mismatched Line

The impedance at any point along a mismatched line is the ratio of

the total voltage to the total current at that point. The input impedance
of a line is the ratio (sending-end voltage)/ (sending-end current). Using
equations (3.2) and (3.4) with x = 0

Zs =
Vi S =
V+ V, Wit V,
y L = Zo ee

ie Vaal. Voy,
4 4
Now, V, = pyaVi = pyVie~™ = p,e~?”V; so that
1 + p,e~*7!
Ze = ie a G15)

Alternatively, from equations (3.9) and (3.10),

z= V, cosh yl + [, Zp sinh yl
—
V,
I, cosh yl + — sinh yl
Z
Z, cosh
ér = as SOS ahahyl + Zp sinh yl
Zominbil (3.16)
Zo cosh yl + Z, sinh yl
 TRANSMISSION LINES 61

If the load terminals are short-circuited so that Z, = 0, then

Zs = Zp tanh yl. (3.17)

Similarly, for an open-circuited line Z, = 0 and Zs = Zp cosh yl.

Example 3.4
A line is 3/2 long and has a characteristic impedance of 50 0 and 3 GB loss.
Calculate its input impedance when the load impedance is 100 Q.

Solution
100 — 50 :
(a) p,. = ———_—— = 5 SUB = 0.345 nepers.
100 + 50
Substituting into equation (3.15)
1 + be 7203454137) 14+ 1,-0@
Z, = 50|A a
Le 2(0.345 +j3: m) ]- | [Ge ?ie |= 70
(Ans.)

(b) Work from first principles, see Fig. 3.2. The total sending-end voltage
= —
R z-180°
is
V; 7
Vs=VWt—=—V,
V, >
_> —i = °
I p Z— Go
6 6
—§|_ @—_—3 V2"
and the total sending-end current is
3 dB loss Go anos
I,S = i ee
6 6 i

La 0° <_ ae z —180° Therefore,

ca 3/2
Vs TV; 6 7
I
pe Zz —180° <—
i
3/2 Z0° Z; = = x =—Z=700. (Ans.)
I; 6 SI; 5

Fig. 3.2
Low-loss Lines

A low-loss line is one whose attenuation al is small enough for the

approximations cosh a/ = 1 and sinh al = al to be valid. The input
impedance of a short-circuited line, see equation (3.17), can then be
written as
sinh (a + a
aa A (a + jB)l
a Z|sinh al cosh j8/ + sinh j@l cosh al
cosh al cosh j@/ + sinh al sinh jl

ve sinh al cos Bl + jcosh al sin Bl

cosh al cos Bl + j sinh al sin Bl}
(Sinh jx = j sin x and cosh jx = cos x.)
62 RADIO SYSTEMS TECHNOLOGY

_ n| Bl + j sin a
(3.18)
cos B/ + j sin Blal
(a) When / = 2/4, then cos B/ = 0 and sin B/ = 1. Then

Hee a|-1-|=2
j Z
jal al
(3.19)
(b) When 1 = 2/2, then cos Bl = —1 and sin 61 = 0. Now

Zs = 2|=>] = Zyal. (3.20)

—1

Loss-free Lines

Very often the attenuation of an r.f. line is small enough to be

neglected so that a! = 0 and y/ = j 1. Then equation (3.16) becomes
+ jZ sin Bl
Zo a ee (3.21)
Zp cos Bl + jZ, sin Bl

Short-circuited Line
If the load terminals of the line are short-circuited
Z_ = 0 and Z, = jZ tan Bl. G.22)
This means that the input impedance of a loss-free short-circuited line
is a pure reactance whose magnitude and sign are determined by both
the characteristic impedance of the line and the line’s length.

Open-circuited Line
If the load terminals of a line are left open-circuit, ZL = © and
Zs = —jZy cos Bl. (3.23)
h/4 Length of Line
When the electrical length of a loss-free line is exactly one-quarter
of awavelength y/ = j@/ = jx/2. Then cos Bl = 0 andj sin 61 = j
so that equation (3.21) becomes

_ 29
ZsS = —.
Zz, (3.24)

h/2 Length of Line

For a /2 length of loss-free line y/ = j8l = jx and so cos Bl
=
—1 and sin 6/ = 0, giving an input impedance of
=f
Zs; = Z| se — ZL:
(3.25)
 TRANSMISSION LINES 63

This means that the input impedance of a )/2 length of loss-free line
is equal to the load impedance.

/8 Length of Line
Now 6/1 = 7/4 and cos B/ = sin Bl = 1/V2. Hence

N
Zs =Z

|Sle
Sls
+

This means that the magnitude of the input impedance of a \/8 length
of loss-free line is equal to the characteristic impedance of the line.

Example 3.5

Calculate the impedance Z 4 of the /4 section of line shown in Fig. 3.3, if

the input impedance Z, is equal to 50 .

Solution
Since the length of line from the load to the point B is \/2 the input impedance
at B is 200 Q and the total load impedance for the \/4 section is 100 Q. The
impedance required at the point A, for the input impedance of the system
to be 50 Q, is also 50 Q. Hence,

Zi = |(50 x 100) = 70.79. —(Ans.)

Standing Waves and Voltage Standing Wave Ratio

At any point along a mismatched line the total voltage, or current,

is the phasor sum of the incident and the reflected waves at that point.
If the r.m.s. values of the total current and voltage are plotted against
distance from the load, standing waves will be obtained.
Consider a loss-free line. The maximum voltage V,,,, on the line
will occur whenever the incident and reflected waves are in phase
with one another, i.e. Vinax = Vi + Vi = Vil + |,/). The mini-
mum voltage Vin occurs at those points on the line where the
incident and the reflected waves are in antiphase with one another,
i.e. Vnin = Vi- Ve = Vil — |0,|). The points at which the

am@#§$_ 5 }} 2_=<+ 1/2

4°-e++—_
Fig. 3.3
 64 RADIO SYSTEMS TECHNOLOGY

maximum voltage occurs are known as antinodes, whilst the point

of minimum voltage are known as nodes. Figure 3.4 shows the
standing wave of voltage on a line of characteristic impedance 600 Q
and load impedance 200 (. The voltage reflection coefficient is equal
to (200 — 600)/(200 + 600) = —0.5 and so the maximum and
minimum voltages on the line are 1.5V; and 0.5V,, respectively. The
minimum voltage occurs at the load and at multiples of \/2 from the
load. The maximum voltage occurs )/4 from the load and then at
multiples of \/2 from that point.
The voltage standing wave ratio S is the ratio (maximum voltage)/
(minimum voltage) or the ratio (minimum voltage)/(maximum voltage).
Either definition can be used as required; no confusion should result
since the v.s.w.r. will always be either greater than, or less than,
unity. Using the first definition

Oak EEE py arr L reap

B20)
Vinin Vid th |p, |) Reis |p|
For the line whose standing wave is shown by Fig. 3.4, |p,| = 0.5
and § = 3.
There are a number of reasons why the presence of standing waves
on a line used to transmit energy from one point to another is
undesirable. These reasons are as follows.
(a) If the load is not matched to the line the maximum transfer
of oe to the load will not take place. The incident power
is V?/Z) and the reflected power is V}/Z. The power
dissipated in the load is
2 2
Vi ac Vr = (Vi me Vp) (Vi = Ve) Vsexs in
Py =

Zo Lo Zo
sei
Ory miPL =a
ee 3.28
(3.28)
(6) The reflected current and voltage waves are attenuated as they

2.0V,
o
S
15V, =
>
oO
£
10V; 3
S
2)
05V, =
oc

Fig. 3.4 Standing wave on a nN 3/4 M2 M4

0
mismatched loss-free line.
Distance from load
 TRANSMISSION LINES 65

travel back towards the sending end of the line. This means
that the total line loss is increased.
(c) At an antinode the voltage is Vax = Vi(1 + |,|) and it may
be anything up to twice as great as the incident voltage. Since
the breakdown voltage of the dielectric between the conductors
of a line must not be exceeded, this limits the maximum possible
peak value of the incident voltage, and hence the incident power
which the line can transmit.

v.s.w.r. on a Lossy Line

When the loss of a line is not negligibly small the incident and reflected
waves are attenuated as they travel. As the distance from the load
increases the incident voltage will get bigger and the reflected voltage
will get smaller, and so the v.s.w.r. will become smaller, see Fig.
3.5. The voltage reflection coefficient at distance x from the load is
Peeps pye”?™, and therefore the v.s.w.r. at this point is
voltage
line
total
R.M.S.
Distance from load
Seq
1+ (oe?"| _ 1+ lolen?a (3.29)
eee | 1 = lpple see
Fig. 3.5 Standing wave ona
mismatched lossy line.
Example 3.6

A line has an attenuation of 6 dB and the v.s.w.r. near to the load is 2.

Calculate the v.s.w.r. near the sending-end of the line.

Solution
From equation (3.29)

( = ie
Seni ~ (Seer
= tek
Six)
S—1\.,, G+ De*=@ - te
~a & + ye
ax = 6/8.686 = 0.69 nepers, soe™ = 2 ande “ = 0.5. Hence

IOP Si chsSo i SigUS) 6.5

= = 1.18. (Ans.)
Sq) =
3x2-1x 0.5 S25

Line Mismatched at Both Ends

If the impedance of the source connected to the sending-end terminals

of the line is also not equal to the characteristic impedance of the line
reflections will occur at both ends of the line. The voltage and current
waves reflected at the load will travel back to the sending end of the
line. Here some of the reflected energy will be reflected again to travel
over the line towards the load. Another reflection at the load takes
place and the reflected energy is transmitted back towards the source.
This will happen repeatedly until the effect of line attenuation reduces
the reflected waves to a negligibly small value. The situation on a
66 RADIO SYSTEMS TECHNOLOGY

pLp.Vje —p.p,Ve
PipsVye 1% = —pip,V5e”
Fig. 3.6 Line mismatched at both pipsVe —pipsVye°™
ends. etc. etc.

line mismatched at both ends is illustrated by Fig. 3.6, where py and

Ps are the voltage reflection coefficients at the load, and at the
sending-end terminals, respectively, and Vi = (EsZ)/(Zx + Zo).
The total voltage at any point along the line is the sum to infinity of
the incident and reflected voltages at that point. The voltages at any
point have values that form a geometric progression (or series), and
thus their sum to infinity is
initial term
Sis
1 — common ratio

Sending-end Voltage
Source-to-load direction
Initial term = V{, common ratio = pipse 7”. Hence

V; =
Vs ie
1 — py pse™
Load-to-source direction
Initial term = p, View common ratio = p; pse ~27 Hence
ay
V> =
pV se
LS pipe
The sending-end voltage is the sum of these voltages, i.e.

Vs =
Vs(1 + ppe~?”)
x 271 (3.30)
1 — pypse
Load Voltage

Source-to-load direction
Initial value = Vje~”, common ratio = pipse 7”. Hence
Vie
V, = Dy”
+ Seipepge 7
 TRANSMISSION LINES 67

Load-to-source direction
Initial value = py V§e~”, common ratio = pipse 2”. Hence
1, —
pLVse *
V> = Se
1 ALAS. |
The total load voltage V;, is the sum of V; and V2, or

Ki'= vse
pa “(1 + 78
py) (3.31)
PLPs&

Example 3.7
A line of characteristic impedance 600 © and 3 dB loss is )/2 in length. It
is fed by a source of e.m.f. 5 V and impedance 1000 © and it is terminated
by a load of 400 Q impedance. Calculate the load voltage.

Solution
1000 — 600 400 — 600
Ds = ——— = 0.25; p, = —— _ = — 0.25
1000 + 600 400 + 600

1 5 x 600
yo ee BIL,
600 + 1000

et, = 6 CTE! — 0°07. xz — 180° = .—0.707.

arte e 2atj By! = eSe = 60" =e (eye

Therefore,
1.875 x —0.707 x 0.75
Vy. = = —0.964V. (Ans.)
heal Oia 25x 0:5)

Transmission Lines as At the higher end of the v.h.f. band, and above, the use of discrete
Components components such as capacitors and, particularly, inductors becomes
increasingly difficult, and very often a transmission line is used to
simulate a wanted component. The input impedance Zz of a loss-free
line short-circuited at its load terminals is given by Zs = jZp tan I.
Such a line has an input impedance that is very nearly a pure reactance,
if the electrical length 6/ of the line is less than /4 the line will
simulate an inductor; if \/4 < Bl < \/2 acapacitor will be simulated.
Loriger lengths are not employed except at u.h.f. because the line
losses will no longer be negligibly small. If the line length is approxi-
mately equal to 4/4, or to \/2, its impedance will vary with frequency
in a similar manner to that of a series-, or parallel-tuned circuit.
Suppose, for example, that an inductance of 100 nH is wanted at
a frequency of 600 MHz and that it is to be simulated by a short-
circuited line of characteristic impedance 50 (. The reactance required
is j24 x 600 x 10° x 100 x 10~? = j377Q, so that 377 =
50 tan Gl or Bl = tan~! 7.54 = 1.439 rad. Hence, 2ml/h = 1.439
68 RADIO SYSTEMS TECHNOLOGY

or] = 1.439\/2a = 0.2292. The wavelength d is 0.5 m so that the

physical length of the short-circuited line should be 0.229 x 0.5 =
0.1145 m = 11.45 cm.
A similar calculation can be made to determine the length of line
needed to simulate a capacitive reactance. If the length thus calculated
is inconveniently long an open-circuited line will give a shorter length
but there may well be radiation problems from the end of the line.
When the electrical length of the line is very short, certainly less than
d/8, tan 61 = Bl and so the simulated inductive reactance xX =
Zl = VL/C x w VLC X 1 = LI. Then the simulated inductance
is given by L/ where L is the inductance per metre and / is the length
of the line.
The Q-factor of the simulated inductance is
ee oh _ wLlt _ NIC -@ a: 2) 520
r 2avVL/C 2a 2a 2ra’

orig _ (3.32)
Clearly, the Q-factor will increase with both frequency and with a
reduction in the attenuation of the line. If the attenuation coefficient
of the line is 8 x 1077 nepers per metre and } is 0.5 m then
Q = 785.

A Line as a Tuned Circuit

A length of loss-free line can also be employed to simulate either a

series-, or parallel-tuned circuit. The possible simulations are shown
in Table 3.1.

Table 3.1

Electrical
length Short-circuited Open-circuited
Se 8 ARES EET a)Es |Siena A Pa ssh dS
/4 Parallel-tuned Series-tuned
M2 Series-tuned Parallel-tuned

Consider a low-loss short-circuited line whose input impedance is

ae al cos 61 + j sin Bl
cos Bl + j sin Blal
(equation 3.18) again).

SS
EE EE eee
tFrom a = R/2Zp.
 TRANSMISSION LINES 69

(a) For 1 = )/4 Bl = 2/4, so that cos G/ is small and sin 6] = 1.

Therefore,

Zi = &Z } : = ~
cos Bl + jal al j cos Bl
al

Now

cos Gl = sin (90° — Bl) = 90° — Bl = Bol — Bl

By = at
and 8B = w/v, so

cos Bl = at = mall= pa]

W bent)
Therefore,

Zs = 4 ; (3.33)
i 2
Fealesiit EBo a |pew
0

The impedance of a parallel-tuned circuit can be written in the

form
Rg
1+ j99(om— =
9
(see Electrical Principles IV) and it is evident that the two
impedances vary with frequency in a similar manner.
Comparing the two equations
Bo 2% us
a 2a rere eee ah

(b) For / = /2, Bl = x, cos Bl = —1 and sin / is small. Hence

pn Z|—al + j sin 7 bs Za iat j sin =
—1 al

Now

sin Bl = —sin (—6l) = —sin (—z + Bl)

—sin (—Bol + Bl) = Bol — Bl

bent)
Hence
 70 RADIO SYSTEMS TECHNOLOGY

yy = @
1 = pa ae )
Wo
Zs = Zoal ?
Qa

which is of the same form as the impedance of a series-tuned

sorrel 22)
circuit, i.e.

Matching For the maximum transfer of energy from a line to its load the
impedance of the load must be equal to the characteristic impedance
of the line. Wherever possible the load impedance is selected to satisfy
this requirement. Very often, however, the two impedances cannot
be made equal to one another and then some form of matching device
may be used. The majority of line matching systems are of one or
another of the following forms: (a) \/4 low-loss matching sections,
(6) single, or double stubs, or (c) baluns (see Radio Systems for
Technicians).

\/4 Matching Sections

A quarter-wavelength (/4) length of loss-free line has an input

impedance Zg given by equation (3.24), i.e. Zs; = ZZ. This means
that a load impedance Z, can be transformed into a desired value of
input impedance Zs by the suitable choice of the characteristic
impedance Zp. The required value of Zp is easily obtained by
transposing equation (3.24) to give

ZL = VZ Zz. (3.34)
If the impedance of the load is purely resistive the \/4 matching
section can be connected between the line and the load as shown by
Fig. 3.7. If, however, the load has a reactive component the /4
section must be inserted into the line at a distance from the load at
which the impedance of the line is purely resistive. This is a calculation
that is best performed with the aid of a Smith chart (p. 80).

=— V4—~»

Line Z, Zi

Fig. 3.7 Use of a \/4 matching

section.
(a) (b)
 TRANSMISSION LINES 71

Main
1
transmission ——P | transmission
Yin=YoliB YL* Yo line Yo
l

Fig. 3.8 (a) Single-stub matching,

(b) double-stub matching.

Stub Matching

The admittance of a mismatched line varies with the distance from

the load. At some particular distance d, the admittance will be equal
to the characteristic admittance Yy = 1/Zp of the line, in parallel with
a susceptance B. At this point Y,, = Yo + jB, Fig. 3.8(a). If this”
susceptance can be cancelled out by another susceptance of equal
magnitude but of the opposite sign, the input admittance of the line
will become Y;, = Yo. The line will then be matched at this point.
The necessary susceptance is provided by the connection of a short-
circuited stub line in parallel with the line at distance d, from the
load. This is shown by Fig. 3.8(b). Not all values of load impedance
can be matched in this way and it will sometimes be necessary to
employ two stubs, as shown by Fig. 3.18. Stub matching problems
are also best tackled with the aid of a Smith chart.

The Smith Chart The Smith chart is a plot of normalized impedance against the
magnitude and angle of voltage reflection coefficient. It can be used
in the solution of many problems involving transmission lines (and
waveguides) since it can often greatly simplify a problem. The Smith
chart consists of: (a) a real axis with values which vary from zero
to infinity, with unity in the centre; (b) a series of circles centred
on the real axis; and (c) a series of arcs of circles that start from the
infinity point on the real axis. This is shown by Figs 3.9(a) and (b).
The circles represent the real parts of the normalized impedances,
i.e. R/Zp, and the arcs represent the imaginary parts of the normalized
impedances, i.e. +jX/Zp.
Figure 3.10 shows a full Smith chart. In addition to the circles and
the arcs of circles representing normalized resistance and reactance
the edge of the chart is marked with scales of (a) angle of reflection
coefficient (in degrees), and (b) distance (in wavelengths). Movement
around the edge of the chart in the clockwise direction corresponds
to movement along the line towards the source; conversely, anti-
clockwise movement around the chart represents movement along the
line towards the load. It should be noticed that a complete circle around
 72 RADIO SYSTEMS TECHNOLOGY

the edge of the chart represents a movement of one-half wavelength

(\/2) along the line.
Any value of normalized impedance can be located on the chart.
Suppose that Z = 100 + j100 0 and Z) = 50 Q. The normalized
impedance is then z= (100 + j100)/50 = 2 + j2; this is shown plotted
in Fig. 3.10 by the point marked as A. Similarly, an impedance
10 — j20 Q is normalized to (10 — j20)/50 = 0.2 — j0.4 and is
represented on the chart by the point B. Admittances can also be
plotted on the Smith chart; the admittance must first be normalized
by dividing it by the characteristic admittance Y of the line. Thus,
fo}
if Y = 0.02 — j0.03 S and Z = 50 Q then y = (0.02 — j0.03)/
(2) 0.02 = 1 — jl.5 and is plotted on the chart as the point C.

or!
xanc® 0 Posit; Ye te
Use of the Smith Chart

Essentially, the Smith chart deals with lines of negligible loss; the
effect of any line attenuation can be taken into account by the use
of a separate scale and this will be dealt with later.

Voltage Reflection Coefficient

To determine the voltage reflection coefficient produced by a load
co
impedance the impedance must first be normalized and located on
the chart. A straight line should then be drawn from the centre of
(b)
the chart (the point 1 + j0), through the plotted point z to the edge
Fig. 3.9 (a) Real parts, and of the chart. The distance of the point from the centre divided by the
(b) imaginary parts of normalized distance centre-to-edge is equal to the magnitude of the voltage reflec-
impedance Z/Z). tion coefficient. The phase angle of the voltage reflection coefficient
is read from the scale at the edge of the chart.

Example 3.8
Calculate the voltage reflection coefficient of the line in Example 3.2.

Solution
The load impedance Z, = 100 + j20 Q and the characteristic impedance
Z = 50 Q so that z, = 2 + j0.4. This point is plotted on the Smith
chart
as shown by Fig. 3.11. The line drawn from the point 1 + jO through
z,
passes through the voltage reflection coefficient angle scale at 14°.
The
distance from the point (1 + j0) to Zz, is 30.6 mm and the distance
from
(1 + j0) to the edge of the chart is 85 mm and so

|o,| = 30.6/85 = 0.36.

Therefore, voltage reflection coefficient = 0.36 2 14°. (Ans. )

The procedure is slightly different when a load admittance is involved.

First, locate the normalized admittance yy on the chart and then
draw
a straight line from y,, through the centre of the chart, to the
edge
 TRANSMISSION LINES 73

ard the \oad) Ba ickward

corward WOE Wavelengths (oward the Sou, 8)

SSA
| SSNS

NS
SOSA SA SOS

OS
E
SS
SS
oR
sesee ete et 32 SSK
ey

ene
OK
ee bei
Ed

SS
PAPAL

oS

ae' et
SSSS

e
BE eetne eeoeee
KOKA
Roe

Sx
LH

e,

\>
SRKKRE
a LAT
(Ty
wae
TE

ee

sw2
‘a,

0S
:

SSL
Se
se
eH

Loy
PAZ
>Oy

Fig. 3.10 The Smith chart; the

point A represents Z = 2 + j2, the
point B represents Z 0.2 — j0.4
and point C represents y fi jig: of the chart. Then the magnitude of the voltage reflection coefficient
|p,| is equal to the ratio (distance y, to centre)/(dis tance edge to
centre), and the angle of y, is read from the scale at th e edge of the
chart.

Voltage Standing-wave Ratio

To calculate the v.s.w.r. on a line locate the normalized impedance

z, (or the normalized admittance y,) on the chart and draw a circle,
 74 RADIO SYSTEMS TECHNOLOGY

Se ae Forward (toward the load)

Fig. 3.11 Use of the Smith chart to

determine the voltage reflection
coefficient. centred on the point (1 + j0), which passes through the point z,. The
V.S.W.r. is then equal to the value of the real axis at the point where
it is cut by the circle. Two values will be obtained: one greater than,
and the other less than unity and each will be the reciprocal of the
>

other. Referring to Fig. 3.11 the v.s.w.r. is about 2.1 or 0.47.

Input Impedance of a Length of Line

To determine the input impedance of a length of loss-free line locate
the normalized load impedance on the chart and draw an arc of a circle
>
 TRANSMISSION LINES 75

centred on the centre of the chart, i.e. the point (1 + j0), moving
clockwise. The length of the arc, measured on the wavelength scale,
should be equal to the electrical length of the line. The normalized
input impedance of the line is then given by the location of the end
of the arc. The procedure can be reversed if the input impedance is
known and the load impedance is to be determined.

Example 3.9

A 50 © line has a load impedance of 20 — j20 Q. Use the Smith chart to

find (a) the voltage reflection coefficient, (b) the v.s.w.r., (c) the input
impedance of a 0.2) length of this line, and (d) the lengths of line that have
a purely resistive input impedance and the values of these resistances.

Solution
The normalized load impedance is

A ay a4
: 50
Z SSeS “ —_ J a

and this is plotted on the chart shown in Fig. 3.12.

42.5 mm
(a). hos f= ee 055 2 67 SO TE
85 mm
Therefore, p, = 0.5 z —131°. (Ans.)
(b) S = 3 or 0.33.
(c) Travelling around the S = 3 circle a distance of 0.2) from z, towards
the source gives z,, = 0.65 + j0.85. Therefore,

Z,, = (0.65 + j0.85)50 = 32.5 + j42.50. (Ans.)

(d) Travelling from z, towards the source 0.068 gives z,, = 0.33 + j0,
and hence Z,, = 16.67 Q. (Ans.)
Travelling from z, towards the source (0.068 + 0.25)\ = 0.318
makes z,, = 3 + jO and Z,, = 150 Q. (Ans.)

Simulation of a Component

When a length of loss-free line is used to simulate an inductance or

a capacitance it is nearly always short-circuited at the load terminals.
Then Z; = z, = 0 and the input reactance of the line can be found
by moving clockwise around the outside of the chart.

Example 3.10

A length of loss-free short-circuited line has Zy = 50 Q and is to simulate

(a) an inductive reactance, and (b) a capacitive reactance, of 35 Q, at
600 MHz. Calculate the necessary lengths of line.

Solution
(a) X, = j35/50 = j0.7. From the Smith chart this is a distance of
0.0962 from the top of the real axis. At 600 MHz \ = 0.5 m and
so the length needed is 4.81 cm. (Ans.)
 76 RADIO SYSTEMS TECHNOLOGY

becit
mueEEE cSeE
tgencs
Bess
a
2

SSiss
te
SRE
COOL

Seng
SERSK
OLR
ees
Ly

o
AYA

etal
HEH
SPS
TY

Sx
XD
pe ze

TNS
ae

HIUAVSs

Fig. 3.12 Calculation of the v.s.w.r.

and input impedance of a line.
(b) x; In —j35/50 = —j0.7. Now the length needed is 0.30632 or
15.32 cm. (Ans.)

Determination of an Unknown Impedance

If both the length of the line and its input impedance are known the
method previously described can be used. If not , the procedure to
be adopted is as follows.
(a) Measure the v.s.w.r. with the unknown load connected to the
line and note the position of any voltage minimum.
 TRANSMISSION LINES 77

(b) Remove the load from the line and short-circuit the load
terminals. This will cause the noted position of the voltage
minimum to shift to a new point (which is less than \/4 away).
Note this new position.
(c) Measure the distance in centimetres between two adjacent
voltage minima — this corresponds to one-half a wavelength
on the line.
(d) Draw the v.s.w.r. circle.
(e) Starting from the point where the v.s.w.r. circle cuts the real
axis at a value less than unity, move around the v.s.w.r. circle
a distance equal to the distance moved by the voltage minimum
in (b) and in the same direction. The point reached is the
normalized load impedance.
If an unknown load admittance is to be determined follow the same
procedure but start on the Smith chart at the point where the v.s.w.r.
circle cuts the real axis at a value greater than unity.

Example 3.11
A 50 Q line has a v.s.w.r. of 2 when an unknown load impedance is connected
to its output terminals. Adjacent voltage minima are found to be 30 cm apart.
When the unknown load is removed from the line and is replaced by a short-
circuit the voltage minima moves by 7.5 cm towards the source. Calculate
the value of the unknown load impedance.

Solution
See Fig. 3.13. The v.s.w.r. = 2 circle has been drawn. 30 cm = A/2 so
that 7.5 cm = 0.125). Moving around the v.s.w.r. circle for this distance
in the clockwise direction gives z, = 0.8 + j0.58. Therefore

Z,, = 500.8 + j0.58) = 40 + j29Q. (Ans.)

Effect of Line Attenuation

The effect of the line attenuation upon the v.s.w.r. and the input
impedance can be taken into account by the use of the scales shown
in Fig. 3.14. The line attenuation scale is marked in 1 dB steps and
may be entered at any point since only distances along the scale are
of any significance. The use of the scales will be illustrated by an
example.

Example 3.12
A 50 Q line has 3 dB loss and is terminated in a 200 + j25 Q load. Calculate
its input resistance if the line is 0.3A long.

Solution
Refer to Fig. 3.15;
200 + j25 ¢
Pee ee a At 10,9,
50
78 RADIO SYSTEMS TECHNOLOGY

THA ALA SR
we,

e,
D
Lh
CXLo
ALAA o>
TAA

yeyo
YWeYO

f=)
— oo S
un +
asjueD

o
os5S D>= ro> o NS© =2 _
jo
ebp3

> S) oeo D o n w=SI c

Jo

T Toward iS
owards load—> Effect of line attenuation (1 dB steps) ~<— generator
Fig. 3.14 Scales used to take into
account any line attenuation.
 TRANSMISSION LINES 79

ad) Backw, ar
the 1080)
For ard yoward Wavelengths g (lowarg th
le Source )

RRS
PEI ORR
PKTA
we eee ee,es Re IL
LTT}

LKG
DAT
Lhy/

LLL
SZ
L}
a
®e,
s

[“

ALS
{)
x5)
SS
ete OK)
oe
Ber
Beree
eeuee
ESS
SEEK

ee
Se
Bk OO
XXX
OC
ST
KRYIS
|

This point is located on the chart and the loss free v.s.w.r. circle drawn to
give S 4. Locate S$ 4 on the v.s.w.r. scale of Fig. 3.14 and move across
onto the line attenuation scale; this point is slightly less than 2.25 dB. Move
J

alon 8 this scale towards the source a distance equal to 3 dB and then move
back onto the v.s.w.r. scale. This gives S = 1.89 (approximately). The
v.s.w.r. circle therefore becomes a spiral, as shown. The point reached is
0.54 + j0.22 so that Z,, = 27 + jll Q. (Ans.)
Zin
80 RADIO SYSTEMS TECHNOLOGY

Matching

(a) \/4 section

If the load impedance is not purely resistive the \/4 matching section
must be inserted at the distance from the load at which the line
impedance is wholly real.

Example 3.13

The load impedance Z, = 20 + j10 Q is to be matched to a 50 © line by

a X/4 length of loss-free line. Calculate the position and impedance of the
matching section.

Solution
The normalized impedance z, = (20 + j10)/(50) = 0.4 + j0.2. This point
has been plotted on the Smith chart shown in Fig. 3.16 and the v.s.w.r. circle
drawn. Moving around the v.s.w.r. circle towards the source until the real
axis is reached covers a distance of 0.214}. At this point z = 2.6; hence
Z = 50 X 2.6 = 1300 and Z) = |(50 x 130) = 80.6 Q. Therefore the
\/4 section should be of 80.6 2 impedance and be connected 0.214 from
the load. (Ans.)

(b) Single-stub Matching

When a single-stub matching system is employed a short-circuited
stub line is connected in parallel with the line at the distance from
the load where the normalized conductance of the line is unity. The
length of the stub must be such that its input susceptance is of equal
magnitude but opposite sign to the susceptance of the line at that point.
The design procedure is as follows.
(a) Plot the normalized load admittance on the Smith chart.
(b) Draw the v.s.w.r. circle.
(c) Move around this circle towards the source until the unity
conductance circle is reached and note the distance travelled.
(d) Determine from the chart the susceptance of the line at this
point, say +jB; the required susceptance of the stub is then
equal but opposite, i.e. —jB.
(e) Determine the necessary length of the stub by moving in the
clockwise direction from the point of infinite susceptance to
the wanted susceptance and note the distance travelled.

Example 3.14
A load has a normalized admittance of 0.4 + j0.6. Calculate the length and
position of a single stub used to match the load to the line.

Solution
The normalized load admittance y, is plotted on the Smith chart shown
in
Fig. 3.17 and the v.s.w.r. circle drawn. The distance that must be travelled
around the circle to reach the unity conductance circle is 0.078
Xr. (Ans. )
At this point the normalized susceptance of the line is +j1.35. To
find
the required length of the stub move from the point —j1.35 on
the edge of
 TRANSMISSION LINES 81

Fig. 3.16 /4 matching section

design.
the chart to the point of infinite susceptance. This gives the length of the stub
as 0.1022. (Ans.)

A single-stub matching system is satisfactory for single-frequency

operation with a constant value of load impedance but it is of limited
?

usefulness otherwise.

(c) Double-stub Matching

If two short-circuited stubs are connected across a mismatched line
82 RADIO SYSTEMS TECHNOLOGY

OORT
OOO
EERE
QDR
TT
ae
TH}
T
TH
SSX0 ry
0600
SSK
2Y SOY
Le

Fig. 3.17 Single-stub matching

design.
at two different locations a much greater range of load impedances
can be successfully matched together. The basic concept is illustrated
by Fig. 3.18, the distance d, of the stub nearest the load may
sometimes be zero. The stub nearest the source is only used to cancel
out the susceptance of the line at that point, so that the normalized
admittance of the line at this point is y, 1 + jO and hence Y,
Yo. The stub nearest the load must have an input susceptance B, such
that the total normalized admittance y, (y, = Y) + jB,/Yo) at the point
of connection, when transformed by the length of line d,, to become
Yin, falls somewhere on the unity conductance circle.
 TRANSMISSION LINES 83

LOOSE
Fig. 3.18 Double- Q E © iS)S ie4 D

PATELI
n ondPe]

TAILORS
8
elie (towarg th

R
ERR
TA
ALLEL

RPOK ORK
PORE EER
ROS REE
Lee
eee
ry
LOR
CELE
LOL ORK EEK
COE
eres:RL 408
SSX KI
LDQ IS RK
SSN
Yy we x?
O Oo
eeegal,

Y
aeae

SOS
.
[Ty
isaew ae
aap

Fig. 3.19 Double-stub matching

design.
84 RADIO SYSTEMS TECHNOLOGY

In most cases the distance d, between the two stubs is either \/8
or 3)/8. Any closer spacing will distort the field distribution so that
the two stubs would act, more or less, as though they were in parallel
with one another. Spacings between )/8 and 3/8 reduce the number
of loads that can be matched. Suppose that dy is chosen to be 0/8;
then the normalized admittance y, must lie on the unity conductance
circle rotated \/8 towards the load, see Fig. 3.19.
Suppose, for example, that the normalized load admittance is y, =
0.6 + j0.6 and that the second stub is connected at the load so that
d, = 0. There are two points, marked as A and B, at which the
addition of load susceptance will make the resultant load admittance
lie on the /8 rotated unity conductance circle. At point A, y, =
0.6 + j0.1 and at point B, y, = 0.6 + j1.96. When transformed by
the \/8 section of line the normalized admittances at A and B become
either y,; = 1 + j0.58 or y. = 1 — j2.76. Thus, the source-end stub
must provide a normalized susceptance of either —j0.525 or +j2.6.
The necessary lengths of the two stubs can then be found in the same
manner as for a single stub, i.e. 0.4 or 0.05A, respectively.
Waveguides

When a signal is propagated along a transmission line electromagnetic

energy is guided in the space between the two conductors. Some of
the energy is always lost because of the inevitable power dissipation
in the resistance of the conductors. Even with a coaxial cable these
losses become excessively high at frequencies above about 3 GHz.
At these frequencies it becomes necessary to employ rectangular
waveguides as the transmission medium. The rectangular waveguide,
shown in Fig. 4.1, consists of a hollow, rectangular metal tube that
has two horizontal walls of internal dimensions a and two vertical
walls of internal dimensions b. Usually, the waveguide is made of
copper, although sometimes other materials, such as aluminium or
brass, are used. The dimensions a and b of the waveguide must be
comparable with the wavelength of the signal and it is only at
frequencies in excess of about 1 GHz that these dimensions become
small enough for the use of a waveguide to be both practical and
economic.
The use of a waveguide is made possible by the skin effect which
limits the flow of current to the surface of a metal when the frequency
is high enough. Energy can therefore be completely confined to the
interior of the guide. Electromagnetic energy directed into the guide
cannot radiate sideways and so it is propagated down the waveguide
with little attenuation. An electromagnetic wave has both an electric
and a magnetic component that are always in phase quadrature, and
also mutually at right angles to the direction of propagation. This is
shown by Fig. 4.2. When an electromagnetic wave is incident upon
a perfect conductor no energy can be absorbed by the conductor and
so the wave must be completely reflected. At the surface of the
conductor the following boundary conditions must be satisfied.
(a) The electric field at the surface of the conductor must be
perpendicular to the surface: there can be no tangential compo-
nent. This must be so or else the field would be short-circuited
by the (assumed) perfect conductor.
(b) Any magnetic field at the surface of the conductor must be
parallel to the surface: there cannot be any perpendicular
component.
The polarization of an electromagnetic wave is the plane in which
the electric field lies; this is usually either the horizontal or the vertical
plane but less often the polarization may be circular.
86 RADIO SYSTEMS TECHNOLOGY

Fig. 4.1 Rectangular waveguide.

Electric
field

90°
Ashi A
Z Distance

Electric
field
+

¢, 907 90°
we AF , Distance
oo ae 4 90
Fig. 4.2 Electromagnetic wave.

Propagation in the Rectangular Figure 4.3 shows one point in the positive peak wavefront of an
Waveguide electromagnetic wave entering a rectangular waveguide. The wave
is incident upon one of the two vertical walls. The polarization of
the wave is assumed to be such that the electric field is perpendicular
to the plane of the paper and the magnetic field lies in the plane of
the paper. When the electric field arrives at a wall it will be completely
reflected with 180° phase reversal. This is indicated in the figure by
the + and — signs. The total field at the surface of the wall is the
phasor sum of the incident and reflected fields and it is zero. The
angle of reflection 6 is always equal to the angle of incidence 6. The
reflected wave travels across the waveguide to the other vertical wall
and here it is again totally reflected with another reversal in its polarity.
This reflected wave travels across the waveguide to the other vertical
wall, is again reflected, and so on. The electromagnetic wave will
therefore propagate down the waveguide by means of a series of
reflections from each of the vertical guide walls.
The wave travels from one vertical wall to the other with a velocity
equal to the velocity of light c, i.e. 3 x 10° m/s. This velocity can
be resolved into two components, as shown by Fig. 4.4. The group
velocity Vg is the component of velocity parallel to the walls of the
waveguide and it is the velocity with which the electromagnetic energy
propagates down the waveguide. From Fig. 4.4, Vg = c cos 6 and
this means that the group velocity is always less than the velocity of
 WAVEGUIDES 87

Wavefront

Fig. 4.3 Single point on a wavefront

propagating down a waveguide.
light. The component of velocity Vy normal to the guide walls is
equal to c sin 0. Figure 4.3 shows one particular point on the wave-
front of a propagating electromagnetic wave. If the complete wavefront
is considered the situation is somewhat more complex and it is illus-
trated by Fig. 4.5. As each successive peak of the wavefront travels
across the waveguide, one end of the wavefront will arrive at a wall
some time before the other end. As each part of the wavefront reaches
Vn c the wall it will be immediately reflected, with a change in its polarity,
whilst the remainder of the wavefront continues to move in its original
Fig. 4.4 Velocities in a waveguide. direction. Eventually, the entire wavefront will have reached the wall
and have been reflected to travel in the new direction. Figure 4.5
shows only the complete wavefronts; clearly, alternate positive and
negative wavefronts propagate in both directions across the waveguide.
The distance between two successive wavefronts propagating in the
same direction is equal to one-half of the free-space wavelength
dX (A = Cc/f). The distance between the intersection of two positive
wavefronts and the intersection of two negative wavefronts is equal
to one-half of the waveguide wavelength \g. These two different
wavelengths are indicated on Fig. 4.5.
At every point within the waveguide the electric and the magnetic
field strengths are the algebraic sum of the two components at that

Fig. 4.5 Wavefronts propagating down a waveguide (e = positive,

© = negative).
88 RADIO SYSTEMS TECHNOLOGY

(a)
ee N ee
(b)

—_— oO e ee °
°
Electric field line Py
—— ___ __ Magnetic field line S ineekeae pe!SS
ek i fel
PY kez eee
1 Tee
|°
@ Out of paper
Olnto paper
“4
°
pares | geek Dl ee te
Soh, SS : le | | —-—? | |o
° | Rion ontBw tithese) le| Roa Lis |0
( Aa )
Nop Bebe eel ects OF np tere NBUDSVaW
BA) Segoe ey Be

Fig. 4.6 Field configurations in a

rectangular waveguide, (a) end view,
(b) side view, and (c) top view.
point. Whenever the peak of one of the waves intersects with a peak
of another wave of the same polarity the resultant field will have its
maximum value. Wherever two peaks of opposite polarity coincide
they cancel out and the resultant field is zero; for the electric field
this only occurs at the vertical walls of the waveguide. In this way
the resultant electric and magnetic field configurations can be deduced
(see Fig. 4.6). Figure 4.6(a) shows the field configurations looking
into the end of the waveguide; the electric field consists of straight
lines between the two horizontal walls, and the magnetic field consists
of lines of force parallel to the horizontal walls. The electric field
strength has its maximum amplitude at the centre of the guide and
falls to zero at each of the vertical walls. The electric field is normal
to the horizontal walls and so it conforms to the boundary condition
(a). If the viewpoint is changed to the side of the waveguide, as shown
by Fig. 4.6(b), the electric field between the horizontal walls changes
its direction at regular intervals along the length of the guide. Lastly,
the top view of the waveguide, Fig. 4.6(c), shows only the magnetic
field and this can be seen to consist of a series of closed loops of lines
of force that are at right angles to the electric field. The field patterns
propagate along the waveguide, without any change in their shape,
with a velocity known as the phase velocity Vp, which is always
greater than the velocity of light. It can be seen from Fig. 4.6 that,
while the magnetic field has a component which lies in the direction
of propagation, the electric field does not. For this reason this mode
 WAVEGUIDES 89

of propagation is known as a transverse electric mode and it is labelled

as TE;9. The first subscript 1 denotes that there is only one half-cycle
in the field configurations in the a dimension of the waveguide. The
second subscript 0 denotes that there are no half-cycle variations in
the field strengths in the b dimension of the guide.
Other modes of propagation are also possible but the TE,) mode
is both the easiest to set up and of the lowest frequency. The TE;
mode is by far the most commonly employed and it is often called
the dominant mode. The higher-order modes may be generated
whenever energy is delivered to, or taken from, the waveguide and
also at any point where some field distortion occurs. Usually, the guide
dimension b is chosen to ensure that all modes other than the dominant
mode are suppressed. Usually, b is about one-half of a.
If the frequency of the signal is varied, the angle of incidence, and
hence of reflection, will also vary to ensure that the dimension a is
always equal to an integer number of half-wavelengths. If the
frequency is increased the angle @ becomes larger and eventually the
point is reached where only one half-wavelength is no longer possible —
and two become established. This would be the higher mode TEp.
On the other hand, if the frequency is decreased the angle 0 becomes
smaller and eventually the point is reached, when 6 = 0°, where the
wave merely moves perpendicularly between the two walls of the
waveguide. There is then no propagation of energy down the wave-
guide; this means that a rectangular waveguide acts like a low-pass
filter whose cut-off frequency is determined by the dimension a. The
free-space wavelength \c at which this happens is known as the cut-
off wavelength. The cut-off frequency fo = Cc/Nc is the lowest
frequency that the waveguide is able to transmit.
Refer to Fig. 4.7 which is an expanded version of the pattern
between a positive peak and a negative peak. From the triangle ABC
Ag/2
cos § =

e+ @)]

Fig. 4.7 Calculation of group

wavelength.
90 RADIO SYSTEMS TECHNOLOGY

from the triangle ADC

cos 6 = )/2a,

and from the triangle BCD

sin 8 = X/Xg.

Therefore,

Ag/2 Pleat (a? + ‘s.)

aie Ne 2 2a 4 4 4
2

2 4)2a2 2
2-2
na(« ag
NSE
a ‘a
SA 2 ; yo _
NG ee =e: 2 2

Bie 4a 2

Gey]
Oa hen= (4.1)

b-Ge)| 1] a aS)

If \/2a < 1, Ag is real and the wave is able to propagate along the
waveguide. If, however, )/2a > 1, \g will be an imaginary quantity
and this means that no propagation of energy takes place. The cut-off
wavelength Ac is the wavelength at which \g changes from a real
to an imaginary quantity, i.e. \c = 2a. The cut-off wavelength of
the next higher mode is equal to a. If the operation of the waveguide
is to be restricted to the dominant mode only, the waveguide must
be operated at some frequency between the cut-off frequencies of the
two modes.
The lowest frequency for which the dominant mode is able to
propagate down a waveguide is the frequency at which the wider
dimension a is equal to one-half a wavelength. Therefore
r
Ag = , (4.2)

J-
x2
1 ae eae iss

Equation (4.2) is very often written in the form

are 1 43
Nel Poem. beNeee ae

Example 4.1
The internal dimensions of a rectangular waveguide
are 0.7112 cm by
0.3556 cm. Calculate (a) the cut-off frequency and (b) the group
wavelength
at a frequency of 25 GHz.
 WAVEGUIDES 91

Solution
Ne = 2a= 2x 0.7112 = 1.4224 cm.
c 35°40"
(a) fo = = = 21.09 GHz. _—(Ans.)
xe 142410?
(b) From equation (4.3)
1 a5 10° 1
ae - _ 35 = 2000.
NG 3 x 10 (14.224 x 10°~)
Hence \%, = 5 X 10~* and Ag = 2.237 cm. (Ans.)

It is normally recommended that a rectangular waveguide should be

used with signals having a free-space wavelength within +20% of
4a/3. The nominal operating frequency should be 1.5f¢ with a
recommended bandwidth of 1.25f¢ to 1.9f¢. This will ensure that
only the dominant mode is transmitted. The next higher modes have
cut-off wavelengths of a (TE9) and 2b (TEo;). Hence, iAl2
a < and 2b < } these nearest modes will both be suppressed.
Usually, b is approximately one-half of a so that the TEo, mode is
suppressed and the attenuation of the waveguide is as low as possible.
The electromagnetic field components of the dominant mode at any
point in a rectangular waveguide are given by the following equations.
Pe =H 0
E, = Asin ete sin (wt — Bz)
a

AH, = —A
Hvaintalerel
Wo
sin
a
sin (wt — Bz) (4.4)

i= 4) is COs eae cos (wt — 8z),

aw Uo a

where x is distance in the a dimension, y is distance in the b dimension,

zis distance along the length of the waveguide, @ is the phase-change
coefficient

en cheng) mcs
and pio is the permeability.

Velocity of Propagation

in a
The phase velocity Vp is the rate at which any particular point
along the wavegui de. The group velocity Vg is
field pattern moves
is propaga ted along the wavegui de. From
the rate at which energy
along
Fig. 4.8 the point on a wavefront marked as E will propagate
the point F first, and then G. To keep the
the waveguide to reach
92 RADIO SYSTEMS TECHNOLOGY

Short time later

Fig. 4.8 Calculation of group and

phase velocity.

phase of the wave constant the point E must arrive at F at the same
time as any other point on the wavefront. The two wavefronts AE
and EJ move in the directions perpendicular to the wavefronts at the
velocity of light c. Their point of intersection E, a particular phase
in the field pattern, moves along parallel to the walls at phase velocity
Vp. For the two wavefronts and the point E to arrive at the point F
simultaneously
G
A === =sc0s @.
Vp EF
Therefore

aise ies, (4.5)

cos 0
Since cos 8 < 1 the phase velocity is always greater than, or equal
to, the velocity of light. Further
cy
Vay, = —-. (4.6)
r
The group velocity is the velocity with which energy is transmitted
and it is equal to
d
Vala eos bay (4.7)
\c
Also

VpVq = c-. (4.8)

Example 4.2

A rectangular waveguide has internal dimensions 3.485 cm and 1.580

cm.
Calculate the phase and group velocities at a frequency of 7 GHz.

Solution
Ac = 2a = 2 X 3.485 = 6.97 cm. Hence
 WAVEGUIDES 93

1 7 x 10°\? 1
Ren fex aa ~ 9.7 x 1072)?
or Ag = 5.434 cm.

From equation (4.6)

Vp = 7 x 10° x 5.434 x 107? = 3.804 x 10° m/s. (Ans.)

From equation (4.8)
(3.x 10%)?
1G = 23659 x. 10) mis. (Ans.)
3.804 x 10

Impedance of a Rectangular The ratio of the electric and the magnetic field strengths in a
Waveguide rectangular waveguide is known as the wave impedance Z,, and it is
constant at all points in the guide. Thus,
Ey Wo 2 mCpUo/d
A, B 27/g x

oe Aone es AG
Vioeo A €9 »

= 1207) /X. (4.9)

Since \g > \ the wave impedance is always greater than 377 Q. If

the wave impedance is matched to the waveguide’s load then maximum
power transfer from waveguide to the load will be obtained. If,
however, the load is mismatched, reflections and hence standing
waves, will be set up in the guide. The effect is exactly the same as
occurs in a transmission line, and problems can be solved similarly.

Attenuation in a Rectangular As a signal is propagated down a waveguide the magnetic field induces
Waveguide e.m.f.s into the walls of the guide and these cause currents to flow.
The direction of the current flow is always at right angles to the direc-
tion of the magnetic field adjacent to the wall. In the side walls
(dimension b) the current flows vertically but in the horizontal walls
the current distribution is as shown by Fig. 4.9. Because the walls
must possess some resistance this flow of current results in power
dissipation and, since this power can only be supplied by the propa-
gating wave, attenuation. Some more losses are also introduced
because the surfaces of the inner walls are not perfectly smooth. For
all sizes of rectangular waveguide the variation of attenuation with
frequency is of the form shown by Fig. 4.10. At the low-frequency
end of the range the attenuation is high because of the cut-off effect
mentioned earlier. There is then a relatively wide bandwidth over
which the attenuation has a fairly flat minimum value rising only
slowly with frequency, and then at higher frequencies the attenua-
tion rises rapidly. Figures for each size of waveguide are given in
Table 4.1.
94 RADIO SYSTEMS TECHNOLOGY

Fig. 4.9
walls.
Currents in waveguide
aA!
|
|
|
|

Fig. 4.10 Attenuation—frequency

characteristics of a rectangular
Attenuation
(dB/m)
:
[oO
Cut-off wavelength

fo)
waveguide. Free-space wavelength

Table 4.1

Waveguide dimensions (cm) Frequency Cut-off Power rating

Internal External limit (GHz) frequency ator Attenuation
a b a b Lower Upper (GHz) (MW) (dB/m) WG WR

16.51 8.255 16.92 8.661 1.14 1.73 0.908 13.47 5 x 40° 6 650
10.92 5.461 eos 5.867 Lede 2.61 1.373 5.90 9.4 x 107° 8 430
7.214 3.404 7.62 3.81 2.60 3.95 2.080 2.43 0.018 10 284
CoM os Me PaariS 5.08 2.54 3.94 0:99 3.155 1.04 0.034 12 187 .
3.485 1.580 3.81 1.905 5.38 8.18 4.285 0.544 0.056 14 137
2.850 1.262 O75 1.588 6.58 10.0 5.260 0.355 0.077 15 ae
2.286 1.016 2.54 1.270 8.2 12.5 6.560 0.229 0.106 16 90
1.580 0.7899 1.783 0.9931 1S 18.0 9.49 0.123 0.171 18 62
1.067 0.4318 1.27 0.635 17.6 26.7 14.08 0.048 0.357 20 42
0.7112 0.3556 0.9144 0.5588 26.4 40.1 re Mi 0.025 0.5767 22 28
0.569 0.2845 0.7722 0.4877 33.0 50.1 26.35 0.016 0.787 23 22
0.4775 0.2388 0.6807 0.4420 39.3 59.7 31.4 0.010 1.026 24 19
0.3759 0.1880 0.5791 0.3921 49.9 75.8 39.9 0.007 1.466 25 15
0.3099 0.1549 0.5131 0.3581 60.5 92.0 48.4 0.005 1.957 26 12

Power Handling Capability

The maximum power that can be transmitted by a rectangular wave-

guide is limited by the need to avoid voltage breakdown of the air
within the guide. This means that it is necessary to ensure that the
 WAVEGUIDES 95

maximum electric field strength is always less than the breakdown

value; for dry air at atmospheric pressure this is 3 X 10° V/m. The
maximum power that a waveguide can transmit using the dominant
mode is given by equation (4.10), i.e.
E2,,.ab rn \2 4.10
Pax = ——— 1 — (——] |.
mn 4807 | gt Ce

Sizes of Waveguides A range of standard sizes of rectangular waveguides has been

developed and these are given by Table 4.1. For each size there are
recommended upper and lower frequency limits, a power rating and
the attenuation at 1.5 times the cut-off frequency. The microwave
spectrum is often said to be divided into a number of bands each of
which is given a label. Unfortunately the labelling is not universally
agreed but Fig. 4.11 shows one version that is employed in the USA.

L band X band Qband Wband

K band
Beat V band
C band
ttt ne ieee fae
Gp tn
1 4.5 6-78 910 20 30 40 506070 100 200
Frequency (GHz)

Fig. 4.11 Microwave bands. L 0.39 — 1.55, S 1.55—5.2, X 5.2-10.9,

K 10.9 — 36, Q 36 — 46, V 46 — 56, and W 56 — 100 GHz.

(a) United
Two other microwave labelling schemes are also in use:
IEE; Cm0.5 - 1, D 1-2,E 2-3,F 3-4,G 4-6,H 6-8,
Kingdo
40 — 60, andM 60 — 100 GHz.
I 8 — 10, J 10 — 20, K 20 — 40, L
L1-2,S 4.C 4-8, X7- 127) 12—18 , K 18 — 26,
(b) NATO;
Q 26 — 40, V 40 - 60, and O 60 — 90 GHz.

eS
 Noise in Radio Systems

Noise and interference, which may be regarded as any unwanted signal

that appears at the output of a radio system, set a lower limit to the
usable signal level at that output. For the output signal to be useful
the signal power must be larger than the noise power by an amount
specified by the required minimum signal-to-noise ratio. This, in turn,
varies considerably depending upon the nature of the signal; two
examples of minimum signal-to-noise ratios are: (a) 30 dB for a
telephone circuit, and (b) 60 dB for a music circuit. The required
minimum signal-to-noise ratio determines the power that a radio
transmitter must supply to its aerial and the spacing of the relay stations
in a microwave radio-relay system. Economic factors, therefore,
require that the effects of any noise and interference are reduced to
as low a figure as is possible.
A noise source may consist of randomly occurring, non-periodic
voltages which have a mean value of zero but which may contain some
relatively large voltage peaks. Such noise sources have a uniform
power density over a given bandwidth and are said to be white; it
is possible to calculate the r.m.s. value of a white-noise source. Other
noise sources are impulsive in their nature: they usually originate from
various man-made sources such as electric motors, neon signs, and
so on. Interfering signals may be picked up by a receive aerial and
if inadequately filtered may appear at the output of the receiver; the
main culprits here are the adjacent-channel and image-channel signals
(p. 191). Other interfering signals may be generated within the radio
receiver itself as a direct result of intermodulation (p. 192).

Sources of Noise in Radio Systems The various sources of noise that may arise and degrade the perfor-
mance of a radio system may conveniently be divided into two groups.
These are: (a) noise generated within the radio receiver itself, and
(b) noise that is picked up by the receive aerial; this, in turn, may
originate from either natural or man-made sources.

Noise Generated within a Radio Receiver

Thermal Noise

When the free electrons in a conductor receive heat energy they will
be caused to move randomly about in that conductor. The movement
 NOISE IN RADIO SYSTEMS 97

of an electron constitutes a current which gives rise to a random noise

voltage that appears across the conductor. The r.m.s. value of this
voltage is
V, = \(4kTBR) V, (5.1)
where: k is the Boltzmann constant (= 1.38 x 10 ~?3 J/K); Tis the
absolute temperature in K (K = °C + 273); B is the noise bandwidth
in Hz; and R is the resistive part of the impedance in which the noise
is generated. |
The standard reference temperature is usually taken as being 290 K
and is given the symbol 7p. This temperature should always be
assumed unless some other value is given.
A resistance R ohms can be represented by a noise voltage generator
Noiseless V,, in series with a noiseless resistor of R ohms, as shown by Fig.
resistor
5.1. The noise bandwidth of a circuit, or device, is the width along
the frequency axis of a rectangular response curve whose area and
R

height are the same as the actual frequency response characteristic

of the circuit (see Fig. 5.2). The noise bandwidth of a circuit can
be determined graphically or by the use of the expression

Fig. 5.1 Representation of the

Neco 2
|AI? >of (5.2)
0 | Aveiax)
thermal noise generated in a resistor.
where A,(f) is the voltage gain of the circuit at any frequency f, and
Avmax) is the maximum voltage gain.

Example 5.1

Derive an expression for the noise bandwidth of the circuit shown in Fig.
5.3. Calculate the noise bandwidth if the output voltage falls by 3 dB at a
frequency of 12 kHz.

Solution
Vo ot 1/jwC € 1
Vin R + I/joC 1 + jwCR

Actual frequency
ee response
cot
@
D>
©D
&
ie)

i |
sean Noise bandwidth

Fig. 5.2 Noise bandwidth. 0 Frequency

 98 RADIO SYSTEMS TECHNOLOGY

Fig. 5.3

V.out 1
V,1 Ja + w2C?R%)
and this falls by 3 dB at a frequency wo/27, where w) = 1/CR. Hence
3 1
Af)
a + Pus)
From equation (5.2)

Spe) Mec
od Fo. aeees
et \ istWied. oa
four) As tan aie
NG cE hh x il
rahi A
(Ans. )

If fy= 12 kHz, B, = 15708 Hz. —(Ans.)

In similar manner it can be shown that for (a) a single-tuned circuit
B, = 1.57B3 gg, and (b) two cascaded single-tuned circuits B, =
1.22 B; gg. If the noise bandwidth of a circuit is not known the 3 dB
bandwidth will have to be used; this will introduce some error of
course but this error will be small if the gain—frequency characteristic
of the circuit has a rapid roll-off.

Available Noise Power

If a resistor R, at temperature 7, is connected to another resistor R,
at temperature T, as shown by Fig. 5.4, noise power will be
Van Vie transferred from each resistor to the other resistor.
The noise power P, delivered by resistor R; to resistor R, is

R, R, P, = V,)R
Hie s2 eT

and the noise power P, delivered from R, to R, is

-Q—

Fig. 5.4 Available noise power. Py = VioR

n24*1 21 :

R, +R) R,
The net noise power transferred from source to load is

d 1
tA standard integral is | ; # > = —tan! (
x“ +y y yy
 NOISE IN RADIO SYSTEMS 99

4kB(R,R)T; — R,RyT>)
ree P, — Pn =
(R, + Rp)?

4kB R, RT, — T-
or Py = ol ie (5.3)
(R; + Ry)
The maximum, or available, noise power P, occurs when the two
resistors are of equal value and 7, = 0 (i.e. resistor Rp is noise free).
Then
V/2\2 4kTBR (5.4)
Ny Be = kTBW.
R 4R
If P, is divided by the bandwidth B the available noise power ina
1 Hz bandwidth is obtained; this is known as the power density
spectrum (p.d.s.) and it is constant up to about 300 GHz.
Very often it is convenient to quote the available noise power in
decibels. Thus
P, = —174 dBm + 10 logo By. (5.5)

Available Power Gain

There are a number of ways in which the power gain, or loss, of a
network can be defined, but for noise calculations the available power
gain G should be used. This is the ratio (output power)/(input power)
when both the input and the output impedances of the network are
matched to the source and the load, respectively.
available output power
Available gain G = (5.6)
available input power
The expression for the noise bandwidth (5.2) can be written in terms
of the available power gain

laee ,
secilites
By .= |.sce (5.7)
0 Geax

where G(f) is the available power gain at any frequency f, and Gra
is the mid-band available power gain.

Noise Produced by Resistances in Series

If two, or more, resistances are connected in series the total mean-
square noise voltage is the sum of the mean-square noise voltages
generated by each resistance. For two resistances R, and Ry, the total
mean-square noise voltage V2 is Va= v2, + V2, = 4kB(R,T, +
R,T»). Usually, the resistances are at the same temperature, and then
V2 = 4kBT(R, + Ro), which means that the total mean-square noise
voltage is effectively generated in the total resistance of the circuit.

Noise Produced by Resistors in Parallel

If two resistors are connected in parallel and are at the same tempera-
ture the total mean-square noise voltage will be equal to 4kTBRy,
100 RADIO SYSTEMS TECHNOLOGY

where Ry = R,R,/(R, + R2). If the two resistors are not at the same
temperature the superposition theorem will have to be used to calculate
the total mean-square noise voltage generated.

Example 5.2

An amplifier has an input resistance of 2000 © and it is at a temperature of

290 K. It is connected to a source of 1000 © resistance at a temperature of
100 K. Calculate the total noise voltage across the input terminals of the
amplifier if the noise bandwidth is 1 MHz.

Solution
The thermal noise voltages V,, and V,; generated in the source and input
impedances respectively are
j
Vis = 44 x 1.38 x 10~% x 100 x 10° x 10%) = 2.35 pV,
2.35 nV 5.66 nV
and V,; = (4 x 1.38 x 10% x 290 x 2 x 10°) = 5.66 pV.
Vv
19 Referring to Fig. 5.5, the total noise voltage across the amplifier’s input
1kQ 2kQ terminals is

ae OS 2:35 x3\2 5.66

x 1\?
V4. = (AE) + Cae! = 2.45 uV. — (Ans.)
Fig. 5.5
Semiconductor Noise

All semiconductor devices may be subject to the following sources

of noise.
(a) Thermal agitation noise in the resistance of the semiconductor
material occurs mainly in the base resistance of a bipolar
transistor and in the channel resistance of an FET.
(6) Shot noise is caused by the random movement of electrons and
holes across each p—n junction in the device. The r.m.s. value
of the shot-noise current is given by J, = |(2eIB), where e
is the electronic charge = 1.602 x 107!9 C, 7 is the dic,
current flowing across the junction and B is the noise
bandwidth. Shot noise is white.
(c) Flicker or 1/f noise caused by fluctuations in the conductivity
of the semiconductor material; it is inversely proportional to
frequency.

Intermodulation Noise

A major cause of noise in radio systems carrying multi-channel

analogue telephony signals is intermodulation (p. 192). Intermod
ula-
tion occurs whenever a complex signal is applied to a non-linear
device; it results in the production of components at frequencies equal
to the sums and the differences of the frequencies, and at the harmonic
s
of the frequencies, contained in the input signal. When the input signal
contains components at several different frequencies the number
of
 NOISE IN RADIO SYSTEMS 101

intermodulation products may be very large; the effect at the output

of the circuit is then very similar to that of thermal agitation noise.
The main difference between them is that whereas thermal agitation
noise is constant, intermodulation noise is a function of the amplitude
of the input signal.

Noise Picked up by the Receiving Aerial

The noise picked up by a receive aerial is generated by a number of

sources of atmospheric, galactic, and man-made origin. At frequen-
cies in the low-, medium-, and high-frequency bands the external noise
is much larger than the noise generated within the radio receiver. At
higher frequencies (in the v.h.f., u.h.f. and s.h.f. bands) externally
generated noise falls to a relatively low level and the overall noise
performance of a system is mainly determined by the receiver itself.
Considerable attention is then given to reducing the noise factor of
the receiver to as small a figure as is possible.

Natural Sources of Noise

A receiving aerial will pick up noise from the sky and from the earth
itself as well as from various interfering signals. Sky noise has a
magnitude that varies both with frequency and with the direction to
which the aerial is pointed. Sky noise is normally expressed in terms
of the noise temperature T, of the aerial. This is the temperature at
which the aerial must be assumed to be for thermal agitation in its
radiation resistance (p. 115) to produce the same noise power as is
actually supplied by the aerial. Thus T, = (noise power)/kB. If the
aerial is used for terrestrial communications, so that its main beam
has only a small upwards inclination then its noise temperature is
effectively that of the earth, or of the lower atmosphere, and this is
usually taken as being 300 K. If, on the other hand, the aerial points
upwards to the sky its noise temperature may be that of space which
is only a few kelvin.
At medium and high frequencies atmospheric noise or static is
always present. Every time a flash of lightning occurs somewhere
in the world impulse noise is generated. Since thunderstorms are
always simultaneously occurring at different points around the earth
and the noise generated can propagate for very long distances, atmos-
pheric noise is always present. The combined effect of a great many
noise impulses makes it sound at the output of a receiver very like
thermal agitation noise. Atmospheric noise has its greatest magnitude
(approximately 10 »V/m) at about 10 kHz and it is negligible at
frequencies above about 20 MHz. The level of atmospheric noise
varies considerably with the location of the aerial, with the time of
day and year, and with the frequency.
The manner in which sky noise varies with frequency above
102 RADIO SYSTEMS TECHNOLOGY

Horizontal elevation

Oxygen molecule

100-4
Water vapour
Galactic
noise

Vertical elevation

Noise
(K)
temperature

100 10°
Fig. 5.6 Variation of sky noise with
frequency. Frequency (MHz)

100 MHz is shown, for both horizontal and vertical elevations, by

Fig. 5.6. Galactic noise, which is produced by radiation from distant
stars, has its maximum value at about 20 MHz and then it falls
inversely proportional to (frequency)”* to a minimum at about
1 GHz. At 50 MHz Ty, is in the region of 5000 K. Galactic noise is
negligible at frequencies higher than about 500 MHz and then the main
source of noise is the earth itself. Noise due to radiation from the
earth also decreases with increase in frequency and, typically, the
aerial noise temperature T, is about 300 K at 200 MHz. Another
source of sky noise is the sun. The sun has a noise temperature of
about 6000 K or more and occasionally produces bursts of noise which
are several times greater than this. The effect of solar noise can always
be minimized by pointing the aerial away from the sun. At higher
frequencies the noise temperature rises with peaks at 23 GHz, due
to water vapour molecules giving a peak in atmospheric absorption,
and at about 60 GHz due to oxygen molecules.

Man-made Sources of Noise

There are a large number of possible sources of man-made noise which
may be picked up by a receiving aerial. Whenever an electric current
is switched on or off one, or more, voltage spikes will be generated
because of the inevitable capacitance and inductance of all circuits.
The voltage spike may be a single transient, or several transients may
occur in rapid succession. Transient spikes occur whenever a circuit
is switched on or off by a mechanical or an electronic switch; examples
of such switching are vehicle ignition systems, electric light switches
and the brushes in electric motors. These cause audible clicks that
 NOISE IN RADIO SYSTEMS 103

a3 80
og Urban
o
E
2 60

2 Rural
8o 40
ao
Ss
B
ro} 20
co

gS
ro T i aL ae
0 10 20 30
Fig. 5.7 Typical variation of total
aerial noise in urban and rural areas. Frequency (MHz)

may adversely affect reception and it is difficult to specify the actual

circuit degradation. Continuous interference spectrums are generated
by circuits such as thyristor speed controls and switched-mode power
supplies.
The interference may either be electromagnetic (e.m.i.) or it may
be radio-frequency (r.f.i.). The former is of lower frequency (30 kHz
to 30 MHz) but it is usually of higher voltage and it may travel over
the mains wiring for a considerable distance. Radio-frequency inter-
ference is of high frequency (30 to 300 MHz) and is both conducted
over, and radiated from, the mains wiring which acts as a rather
inefficient aerial. The total man-made noise picked up by a receive
aerial cannot be separately identified from the atmospheric and galactic
noise that is also present, but obviously this source of noise will be
larger in an urban environment than in a rural area. Figure 5.7 shows
how, typically, the total of atmospheric, galactic and man-made noise
may vary in both urban and rural areas.
Man-made noise can often be suppressed at its source, although
it may not always be economical to do so. Alternatively, it is often
possible to position the receive aerial at a site which is remote from
interference.

Signal-to-noise Ratio As a signal is propagated over a radio link it will be subjected to

various forms of noise and/or interference and it will also suffer
attenuation. The radio receiver will have sufficient gain to compensate
for the path attenuation and give the required output signal power,
but the amplifier and mixer stages will add further noise to the signal.
The usefulness of the output signal for its intended purpose is expressed
by means of its signal-to-noise ratio. This is defined as
wanted signal power
signal-to-noise ratio = 3 ; (5.8)
unwanted noise power
 104 RADIO SYSTEMS TECHNOLOGY

Because signal-to-noise ratio is the ratio of two powers it is commonly

quoted in decibels.
The larger the output signal-to-noise ratio of a radio system the
greater will be its ability to satisfactorily receive weak signals. The
minimum permissible signal-to-noise ratio of a radio system is one
of its most important parameters. Different minimum signal-to-noise
ratio figures are specified for different kinds of signal and, for
example, the minimum figure for a mobile radio-telephony system
is about 15 dB.

Noise Factor Figure 5.8 shows a circuit that has an available power gain G and
that introduces an internally generated noise power at the output
terminals of Nc watts. The input signal to the circuit is S,, with an
associated noise power of N,, = kT)B watts. This means that the
input signal-to-noise ratio is S;,/Nin-
The noise factor, or noise figure, F of the circuit is defined as

total noise power at output

P=
amplified input noise power
; (5.9)
It is assumed that both the source and the circuit are at the standard
temperature of 290 K and that the circuit is linear. Therefore

F
_ GNin
+ Ne SWatye aN
(5.10)
GN; GNin
The noise factor is unaffected by the value of the load impedance
because any mismatch at the output terminals which might exist will
reduce both the noise and the signal powers equally, and so will not
alter their ratio. If the circuit were noise free Nc would be zero and
then the noise factor would be F = 1 or 0 dB. This is the theoretical
minimum figure for the noise factor. In practice, typical figures are
4 dB for a v.h.f. amplifier and 12 dB for a radio receiver.
It is often convenient to refer the noise generated within the circuit
to the input terminals. Then N, = N./G and

fo
GN,, + GN. N
ee eee (5.11)
GN; Nin
This expression can be rearranged to give N. = (F — 1)N;,,, which
expresses the internally generated noise as a function of both the noise
factor and the actual input noise. The internally generated noise may

Power gain G
In
Sout a GS;,
Nin =kToB Internal noise Nout = GNin + No
power No
Fig. 5.8 Noise factor.
 NOISE IN RADIO SYSTEMS 105

be regarded as being generated by thermal agitation in an equivalent

input resistance (see Electronics IV, Ch. 8).
When a coherent input signal is applied to the circuit the expression
for the noise factor can be written as

were of a
(aa
ene) (Gere)
(ie)
GSin
GNi, + Nc

input signal-to-noise ratio

or F 3 ase 6.12)
output signal-to-noise ratio
This expression shows that the noise factor of a circuit indicates the
extent to which an input signal-to-noise ratio is degraded by the noise
generated internally within the circuit.

Output Noise from a Circuit

The available input noise power to a circuit is kT)>B watts, and
putting this into equation (5.12) gives
_ sinlkigB= ~*~ ANo
GS;,/No GkT)B’
where No is the output noise power. Therefore
Ny = FGKTpB. (5.13)
In decibels this is

No = —174 dBm + 10 logig B + 10 logig F + 10 logio G.

(5.14)
This is the total noise power at the output of the circuit and it is the
sum of (a) the amplified input noise, and (b) the noise produced within
the circuit. Since the former is equal to Gk7B, the latter must be

N. = (F — 1)GkTB. (5.15)
In the case of a radio receiver which has different bandwidths at
different points in the circuit the narrowest bandwidth should be used.

Variation of Noise Factor with Frequency

The gain of a circuit and the noise generated within it per unit band-
width are often a function of frequency and this means that the noise
106 RADIO SYSTEMS TECHNOLOGY

factor of the circuit may also vary with frequency. Thus a distinction
between the noise factor at a single frequency and the full-bandwidth
noise factor may need to be made. If the bandwidth of the circuit is
narrow enough for any variations in the gain and/or generated noise
to be ignored, the spot noise factor is obtained. The spot noise factor
can be measured at a number of points in the overall bandwidth of
the circuit and then the average noise factor can be obtained.

Noise Factor of a Lossy Network

Suppose the network has an attenuation (power ratio) of X. The

available input noise power is k7)B and hence the available output
noise power is (kT)B)/X + Nc. This power is the same as the
available noise power k7)B from the output resistance of the
network. Therefore
kTpB
ee + No = kIpB

1 X-1
or Ne — kT) B (:= = i cT9B( X )

The same available output noise power would be obtained if a noiseless

network having the same attenuation was to be supplied by a source
that produces an available noise power of kT7)>B(X — 1). Hence
KT) B(X — 1) = (F — WkT)B
or F=X. (5.16)

Thus the noise factor of a lossy network, such as an attenuator or

a length of coaxial or waveguide feeder, at temperature 7p is equal
to its attenuation.

Overall Noise Factor of Circuits in Cascade

Figure 5.9 shows two circuits connected in cascade. The circuits have
noise factors of F, and F), and available gains of G, and G,, respec-
tively, and it is assumed that their bandwidths are the same. If the
overall noise factor of the combination is Fo then the available output
noise power Np will be equal to FyG, G,kT)B. For the first circuit,
the output noise power is No, = F; G,kTpB and this is the input noise
to the second circuit. Hence

Noo = GF, G,kT)B ae (F, Se 1) GokT)B.

Fig. 5.9 Noise factors in cascade.

 NOISE IN RADIO SYSTEMS 107

Since No = No;

Fy G, G,kT)B = G, G)F\ kT) B =f (Fy a 1)G,kToB

m=
and Fy = F, + O17)
1

This result can be extended to any number of circuits connected in

cascade. For n circuits
Fy, - 1 F; -— 1 F,- 1
Fy
= Fi + — Me £ af 2
G GG, iG caG
(5.18)

If the first, and later, circuits in Fig. 5.9 are amplifiers the noise
generated by the second, and following, amplifiers will be reduced
because of its division by the product of the gains of the preceding
amplifiers. If, however, the first circuit introduces a loss then the noise
introduced by the second circuit will be increased. This means that
to obtain a good, i.e. a low, overall noise factor, the first circuit must
introduce the minimum possible loss and should, if possible, be an
amplifier.

Example 5.3

A radio receiver has a noise factor of 6 dB. An amplifier with a power gain
of 10 dB is connected between the aerial and the receiver. The overall noise
factor is then 6 dB. Calculate the noise factor of the amplifier. Calculate the
overall noise factor if a 6 dB attenuator were to be connected (a) between
the aerial and the amplifier, and (b) between the amplifier and the receiver.
Assume the noise picked up by the aerial to have an effective noise temperature
of 290 K.

Solution
From equation (5.17)
3.98 — 1
36S Sh ee
10
or F, = 3.682 = 5.66 dB. (Ans.)
32082. cecal Seky == Ik
(a) Fy = 3.98 + ae = 15.84 = 12 dB.
1/3.98 10 x 1/3.98
(Ans.)

3.98 = 1 3.98 — 1
(b) Fy = 3.682 + + = 51 snd Be
10 10 x 1/3.98
(Ans.)

Effective Noise Factor when 7; 4 7p

The noise factor of a circuit is defined, measured and quoted with

reference to a temperature of 290 K. If the source temperature is not
108 RADIO SYSTEMS TECHNOLOGY

at 290 K but is at some other temperature the degradation of the input

signal-to-noise ratio will be different from that indicated by the noise
factor. If the source temperature is less than 290 K the degradation
will be greater than indicated, but if the source temperature is in excess
of 290 K the signal-to-noise ratio reduction will be less than expected.
To obtain the correct figure for the output signal-to-noise ratio the
effective noise factor F.¢ will have to be employed.
From equation (5.13), the noise factor of a circuit referred to the
standard temperature of 290K is F = N)/Gk7)B, and hence
the available output noise power due to the circuit alone is
(F — 1)GkT)B.
If, now, a source at temperature T; # Tp is applied to the input
terminals of the network, the available noise output power will be
No = GkTs;B + (F — 1)GkT)B. Now the effective noise factor
Figg = No/GkT;B, or
GkT;B + (F — 1)GkTpB To
Fig = a re
GkT;B Ts
(5.19)

Example 5.4

A radio receiver has a noise factor of 10 and it is connected by a feeder of

3 dB loss to an aerial. The noise delivered by the aerial is at an effective
noise temperature of 145 K. Calculate the noise power at the output of the
receiver if the receiver has a power gain of 50 dB and a bandwidth of 10 kHz.

Solution
The input noise power to the receiver system is

ETB E38 SOO OSTA S10 ao 1

The noise factor of the feeder is 3 dB, or 2, and so the overall noise factor
is 2 + (10 — 1)/% = 20. The effective noise factor at 145 K is
(20 — 1)290
Fy = 1 + —————— = 39
145
Therefore, the output noise power is

No = FGkT,B = 39 x 10° x 2 x 10°!’ = 78 pW. — (Ans.)

Noise Temperature The concept of noise temperature provides an alternative to noise

factor for the specification of the noise performance of a radio system.
The noise power Nc generated within a circuit and appearing at the
output terminals may be considered to be the result of thermal agita-
tion in the output resistance of the circuit. The output resistance must
be regarded as being at a noise temperature T, = 17), where t may
be greater than, or smaller than, unity. Thus

Ne — kT, B
 NOISE IN RADIO SYSTEMS 109

On 7s:
Ne
— (5.20)

If the internally generated noise is referred to the input of the circuit

it may be regarded as entering the input of the circuit as shown by
Fig. 5.10. The internally generated noise is now equal to kT,,B and
so the total input noise is kB(T, + T,), where Ty, is the noise
temperature of the aerial which also may, or may not, be equal to 7p.
The input signal-to-noise ratio of Fig. 5.10 is S;,/kT,B and the
output signal-to-noise ratio is
GS;,/[GkB(T, + T,)] = [input signal-to-noise ratio] [T,4/(T, + T,)].

Relationship between Noise Factor and Noise Temperature

From equations (5.15) and (5.20) the noise generated within a circuit
is (F — 1)GkT)B = GktT)B, or

p=Fr=e T, (5.21)
The concept of noise temperature, instead of noise factor, is usually
employed for expressing the noise picked up by an aerial and for low-
noise circuits. For a low-noise circuit the use of noise factor to specify
the noise performance will often lead to inconvenient numbers;
consider, for example, a noise factor of 1.068 97 is a noise tempera-
ture of 20 K. Table 5.1 shows equivalent noise factor and noise
temperature values.

System Noise Temperature

The system noise temperature T; of a radio receiving system is the

kT,B Power gain G

: No = GkB(T, +T,)

No internally
generated noise

Fig. 5.10 Noise temperature. kT,n B©

Table 5.1

Noise factor (dB) 0.29 0.56 0.82 1.06 1.29 2.28 3.00
Noise temperature (K) 20 40 60 80 100 200 290
110 RADIO SYSTEMS TECHNOLOGY

sum of the aerial noise temperature 7, and the overall noise tempera-
ture of the system, i.e. Ty = T, + Toy. The overall noise tempera-
ture T,, is easily obtained by combining equations (5.18) and (5.21).
Subtracting 1 from both sides of equation (5.18) gives
t t
fy = t + > + — (5.22)
G; GG,
T. T.
5 elHO cidPec ena ee (5.23)
G, G, G,
The system output noise power No is then given by equation (5.24)

This method of calculating the output noise power from a radio

system is particularly useful when dealing with a system in which
the aerial noise temperature is not equal to 290 K.

Example 5.5

Calculate (a) the system noise temperature, (b) the output noise power, and
(c) the output signal-to-noise ratio of the system shown by Fig. 5.11. The
available signal power from the aerial is 3 pW, and the bandwidth is 10 MHz.

Solution
(a) Converting the decibel figures into ratios, 1 dB = 1.26, 20 dB = 100
and —10 dB = 0.1. The feeder has a noise factor of 1.26 and so its
noise temperature is 75.4 K. The system noise temperature is
50 630 500
Te I 40 + 75.4 + " . = 186.4 K.
1/1.26 100/1.26 104/1.26
(Ans.)
1
(b) Ny = GkTsB — x 100 x 100 x 0.1 x 1.38 x 10-7 x
1.26
186.4 x 10’ = 20.4 pW. — (Ans.)
(c) Output signal power = 1000 x 1/1.26 x 3 x 10~" = 2.38 pW.
Therefore
2.48 x10"
output signal-to-noise ratio = ——————_—_
104S< 1052
51 dB. (Ans.)

T,=40K
Low-noise To detector
Amplifier
Waveguide amplifier
feeder
- 1 dB loss 20 dB gain 20 dB gain 10 dB loss
Fig. 5.11 T,, = 50 K T,, = 630 K T,,= 500 K
 Principles of Aerials

Transmitting and receiving aerials provide the link between the radio
transmitter and receiver, and the propagation path via the atmosphere.
All types of aerial are able both to transmit and receive radio-frequency
energy and for each purpose will have the same gain and radiation
pattern. It is customary to consider the operation of most aerials in
their transmitting mode and to use the principle of reciprocity to obtain
the receiving characteristics if, and when, required. Some aerials,
such as those employed in broadcast systems, are required to transmit
energy equally well in all directions in the horizontal plane. Other
aerials are required to concentrate their radiation in one direction,
these must have a directive radiation pattern to achieve the maximum
gain in the wanted direction. Thus the directivity of an aerial,
expressed graphically by means of its radiation pattern, is an important
parameter.
For the maximum efficiency in the radiation of energy an aerial
should be of resonant length; this means that its electrical length should
be one-half a wavelength (\/2). This requirement can be satisfied at
frequencies in the h.f., v.h.f., u.-h.f. and s.h.f. bands, and it is
possible, although more difficult, at some frequencies in the m.f. band.
It is not possible to obtain a )/2 aerial at frequencies in the low-, and
very-low-frequency bands because of the enormous physical structures
that would be necessary. Transmitting aerials for use in the v.1.f.,
1.f. and m.f. bands are always vertical structures that are mounted
upon the earth. The aerials are fed between the base of the aerial and
the earth; ground reflections then make the aerial appear to be up
to twice its physical height. An aerial whose electrical length is less
than one-quarter wavelength, including the ground-plane image, is
said to be an electrically short (or small) aerial.

Radiation from an Aerial When a radio-frequency current flows in a conductor a magnetic field
is set up around that conductor, the magnitude of which is directly
proportional to the instantaneous value of the current. As the current
varies with time the magnetic field will change as well. The changing
magnetic field produces a changing electric field, not only in the
vicinity of the magnetic field but also in the region surrounding it.
Similarly, the changing electric field produces a further changing
magnetic field in the next surrounding region, and so on. Radio-
frequency energy is, in this way, propagated away from the conductor
112 RADIO SYSTEMS TECHNOLOGY

in the form of an electromagnetic wave. Figure 4.2 shows an electro-

magnetic wave in which the electric and magnetic fields vary sinusoid-
ally with time. The two fields are always in phase quadrature and
are mutually at right angles to their direction of propagation.
The plane that contains the electric field and the direction of propa-
gation is known as the polarization of the wave. If, for example, the
electric field is in the vertical plane the wave is said to be vertically
polarized.
The electromagnetic wave propagates through the atmosphere with
a velocity equal to the velocity of light c, which is generally (for radio
work) taken as being equal to 3 x 10° m/s. At all times during its
travel the two fields of the electromagnetic wave vary in phase with
one another. The ratio of the amplitude of the electric field strength
to the amplitude of the magnetic field strength is a constant that is
known as the impedance of free space.

Impedance of free space = 2 ECU = 1207 YQ. (6.1)

H (A T/m)
The power density P, of the electromagnetic wave is equal to the
product of the r.m.s. values of the electric and the magnetic field
strengths, i.e.
2
P, = EH W/m = W/m. (6.2)
207
Since both the electric field and the magnetic field lie in a plane
that is at right angles to the direction of propagation a radio wave
is a transverse electromagnetic or TEM wave.

Isotropic Radiator

An isotropic radiator is a theoretical aerial, that cannot be realized

in practice, which is able to radiate energy equally well in all the
possible planes. The radiated energy will therefore have a spherical
wavefront with its power spread uniformly over the surface of a
sphere. The surface area of a sphere of radius r is 4ar? and so the
power density Py of the radiated energy is

P;
P,
cae W/m, (6.3)

where P, is the transmitted power and D is the distance from the

radiator. Clearly, the power density is inversely proportional to the
square of the distance from the radiator. The power density is also,
from equation (6.2), equal to E7/120z, and equating the two rela-
tionships gives the electric field strength at distance D from the
isotropic radiator as

E
_ (G0P) , /m. (6.4)
 PRINCIPLES OF AERIALS 113

This, of course, is inversely proportional to distance. If the trans-

mitted power is given in kilowatts and the distance D in kilometres,
then E = (173VP,)/D mV/m.
Although the isotropic radiator is not a practical aerial it is
commonly employed as a reference against which other, practical,
aerials can be compared.

Current Element

A current element consists of a current of uniform amplitude / amperes

which flows in a very short length d/ of conductor. It is sometimes
known as a Hertzian dipole and it also cannot be practically realized.
The current element is an extremely useful concept because the results
obtained from a study of it can be applied to practical cases if another
concept, that of the effective length of an aerial, is employed. Many
practical aerials may be conveniently regarded as consisting of a large.
number of current elements in cascade.
When a radio-frequency current flows in a very short electrical
length d/ of conductor (see Fig. 6.1) to form a current element, the
magnetic field that is set up will have a field strength given by
I dl sin 6 D
ie = le cos o(+ ~ 2)
4a cD c

D A T/m,
ele (‘- =| (6.5)
D c
where c is the velocity of light, D is the distance, in metres from the
current element, and @ is the angle shown in Fig. 6.1.
Unlike the isotropic radiator, the current element does not radiate
energy equally well in all directions but, instead, it produces a field
that is proportional to sin 9. Equation (6.5) indicates that the magnetic
field has two components which are known, respectively, as the radia-
tion field and the induction field. The induction field is inversely
proportional to the square of the distance from the current element
and, therefore, its amplitude rapidly falls to a negligible value. The
induction field represents the energy that is not radiated away from
the current element. The magnitude of the radiation field is directly
I sin ot
proportional to the frequency of the current flowing in the element
and inversely proportional to distance. It therefore represents the
Fig. 6.1 Current element.
energy that is radiated away from the current element.
The induction and radiation fields are of equal magnitude at the
distance from the current element at which

ef pe
cD D?

ON
Ob ie
w 27
114 RADIO SYSTEMS TECHNOLOGY

The distance D = )/27 marks the boundary between the near field
and the far field. At distances greater than \/27 the induction field
makes an insignificant contribution to the total field and so it is
neglected.
The magnetic radiation field has an r.m.s. value of

H= Idly sin? — Idi sin 6

(6.6)
4ncD 2D;
An expression for the r.m.s. electric field strength is obtained by
multiplying H by 1207 to give
_ 6071 dl sin 0
& ‘ (6.7)

If 7d/ = 1 mA and D = 1 km then the field strength is equal to

188 mV/m.
In the meridian plane (the plane that contains the current element)
the field strength varies with sin 6 and the radiation pattern is as shown
by Fig. 6.2(a). In the equatorial plane, where 6 = 90° and hence
sin @ = 1, there is equal radiation in all directions and so the radiation
pattern is a circle (see Fig. 6.2(b)).

Current element

Fig. 6.2 Radiation patterns of a

current element, (a) meridian plane
and (b) equatorial plane. (a) (6)

Radiated Power

The total power radiated by a current element is determined by finding

the power that passes through a small area and then integrating over
the surface of a sphere. Referring to Fig. 6.3, the area of the shaded
zone is 2mxD d@ = 2D? sin 0 d6. The power which passes
through this zone is
E? Se dy = 6071 7J7(d1)? sin? 6 dé ey
1207 ?
The total power P, radiated over the surface of the sphere is
a2
3 | [604 T 212%(d1)2
. sinsin3 "dp
P,
0

12077/7°(d1)? (2 (3 sin 6 — sin 36)

Fig. 6.3 Power radiated by a Se dé
current element. N ; 4
 PRINCIPLES OF AERIALS 115

120 4217(d1)? n/2

38 le euoe [=3cosa + —-
4)? 5

304 717(d1)?
eee <P
804 717(d1)?
of P= aisesat NY (6.8)
d
Expressing both equations (6.7) and (6.8) in terms of J 2 and then
equating them gives
a a Pd? E?)’*D*
8072(dl)? ss60. (1)?

E = {45P)
ie (6.9)

Radiation Resistance

The total power P, radiated from an aerial may be considered to be

equal to the product of the square of the input current to the aerial
and a non-physical resistance R,, i.e. P, = I°R,. This concept gives
the radiation resistance of a current element as
come 80m (dl)*
< (6.10)

Effective Length of an Aerial The equations previously obtained for the electric field produced by,
and the power radiated from, a current element cannot be directly
applied to a practical application. They can, however, be usefully
employed if another concept, that of the effective length of an aerial,
is introduced.
The effective length lg of an aerial is that length which, if it
carried a uniform current at the same amplitude as the input current
Ito the aerial, would produce the same field strength at a given point
in the equatorial plane of the aerial. This means that the product of
the physical length of the aerial and the mean current flowing in the
aerial must be equal to the product of the effective length and the
assumed uniform current. Thus

Dohy /mean = lege

Sri te eo
fot (6.11)
I
l 1 ‘ph |
Op fj aes re 3 ” 1(y”) dy (6.12)
fae Bepats be
116 RADIO SYSTEMS TECHNOLOGY

1 (pn
of be = Is” I(y) dy. (6.13)
0
Alternatively, the effective length of a receive aerial may be defined
Aerial current as leg = V,./E, where V,, is the voltage that appears at the open-
circuited terminals of the aerial when it is situated in an electric field
phy
of strength E V/m.
For an electrically short aerial, such as that shown in Fig. 6.4, the
current distribution can be assumed to vary linearly from its maximum
value J at the input terminals to zero at the top of the aerial. Clearly,
the mean value of the aerial current is //2 and, from equation (6.11),
UR <= lerp = Uphy/2. Using equation (6.12)
Input current
1 1 (phy: 7
Fig. 6.4 Current distribution on an
lege = —PE a | ey City ay]
13 phy JO lohy
electrically short aerial.
‘phy 1 y? |phy
Sa | iy 9) dy = bu ree
0 lohy 2 0

res k e = ee lphy
Lege ee IED 2
Using equation (6.13)

1 =
1
—
[fhe
os Ll a d =
hs
phy ;
eff I | diy ( phy yy) y 2

The radiation resistance of such an aerial is given by

R. = 8072 Loby_ Nei ony “oO.

: 2» d
For longer aerials the assumption of a linear current distribution is
no longer valid and it is customary to assume the distribution to be
sinusoidal.

The Monopole Aerial A monopole aerial is one which is mounted vertically upon the surface
of the earth and which is fed between the base of the aerial and earth.
This type of aerial is employed in the v.1.f., 1.f. and m.f. bands; in
the two lower bands the aerial must, of necessity, be electrically short,
but in the m.f. band it is possible to employ aerials whose electrical
length is \/4 or even longer.
Figure 6.5 shows an aerial of physical height Inny Which is mounted
vertically upon the earth. The aerial will radiate energy equally well
in all directions in the horizontal plane but it will exhibit some
directivity in the vertical plane. Some energy is directed upwards
towards the sky whilst some other energy is radiated downwards
towards the earth. The aerial site is chosen to ensure that the earth
in the neighbourhood of the aerial is both flat and of high conductivity.
 PRINCIPLES OF AERIALS 117

Actual aerial wave

b
>
NN
| NY mM Reflected
4
wave

Input tA es
terminals cee Ces ON.
~ a0 Vat)
z 7 7; 7, 77)
a4 mt

eee
4 oe
Pans Image wave
phy Pe 7 3
7
eo
ca

a Image
Fig. 6.5 Monopole aerial. aerial

Therefore the waves radiated towards the earth are totally reflected
with an angle of reflection equal to the angle of incidence. This makes
the ground-reflected waves travel in the same direction as the direct
waves. At a distant point P from the aerial energy is received by means
of both the direct and the ground-reflected waves. The total field
strength at this point is the phasor sum of the individual field strengths
produced by each of the two waves. From the viewpoint of an observer
at the point P it appears as though the ground-reflected wave has
originated from an image aerial beneath the earth. This effect makes
the aerial appear to be of twice its actual physical height and so doubles
the effective length of the aerial. This means that an electrically short
monopole aerial has an effective length equal to its physical length.
The electric field strength produced at the distant point P is hence
E= 120a1 leg sin 0 V/m. (6.14)
AD
The monopole aerial produces a similar radiation pattern, in both
the horizontal and the vertical planes, as a balanced dipole of twice
the physical length, that is situated in free space.
The power radiated by a monopole aerial is obtained by the use
of equation (6.8) but since there is no electric field beneath the ground
the actual power is only one-half of that predicted, i.e.
272 2 l »
pes nner. 5Gy oleh. (6.15)
nN
If both equation (6.14) and equation (6.15) are expressed in terms
of J” and then equated
om PdGor rn BPD?
160 171 2¢¢ (120
1)712¢¢
118 RADIO SYSTEMS TECHNOLOGY

E= LOOP), (6.16)
D
If P, is in kilowatts and D is in kilometres, then

_ {90 x 10°)VP, _ 300

E VP, mV/m. (6.17)
D

Example 6.1
A monopole aerial is 25 m high and it is supplied with a current of 100 A
at 200 kHz. Assuming the current distribution on the aerial is linear calculate
(a) the power radiated by the aerial, and (b) the field strength produced at
ground level at a point 100 km distant.

Solution

Sa 10°
y= Sie AGRROR oe 1500 m.
200 x 10
The electrical length of the aerial is 25/1500 = 0.0172.
sae y
(a) Po= 1600-\{ —~—] 10" =. 1097 W. (Ans.)
1500
(b) From equation (6.14)
120a x 100 x 12.5
E= oes 3.14 mV/m. (Ans.)
1500 x 100 x 10
Alternatively, using equation (6.16),
aol ias1=N12 J(90 x 1097)
i; = 3.14 mV/m. (Ans.)
100

As the electrical length of a monopole aerial is increased the assump-

tion of a linear current distribution can no longer be made. The current
distribution becomes an increasingly large part of a complete sine-
wave, and when the electrical length is one wavelength a complete
cycle is described. If the length is less than \/4 it is only necessary
to determine the effective length of the monopole. If the length is \/4
el
oYGopal pian, shih or more it will be necessary to consider the monopole to consist of
the cascade connection of a large number of current elements. Each
current element produces an electric field, due both to a direct and
ha Seapets |=N2 to a ground-reflected wave, given by equation (6.14), at a distant point
P. The total field strength at the distant point is then the phasor sum
of the field strengths due to each current element. This will result
Lt iE ee in the addition, or cancellation (perhaps complete), at different angles
in the vertical plane; a number of examples are shown by Fig. 6.6.
The vertical-plane radiation pattern changes only slowly with increase
rn See in the aerial height for heights up to about )/4, and then, for heights
between d/4 and \/2, becomes somewhat flatter in shape. The maxi-
Fig. 6.6 Radiation patterns of mum radiation at ground level occurs when the electrical length of
monopole aerials of different heights. the aerial is 5/8, although there is then some unwanted skywards
 PRINCIPLES OF AERIALS 119

radiation. If the electrical length is increased to more than 52/8 the

radiation of energy is concentrated in the direction of the sky.
The radiation patterns differ from that of a current element because
the distances from each part of the monopole to the distant point P
may differ by an appreciable part of a wavelength. This will not affect
Current the amplitude of the individual field strengths but it does introduce
distribution phase differences.

Input a
current

Effective Length of a Short Monopole

Earth

When the electrical length of a monopole is short the current distribu-

tion on the aerial can be assumed to be a part of a cosinusoidal wave.
As shown by Fig. 6.7 the current will be zero at the top of the aerial
and increases to a maximum value Jp at the origin which is at a
distance / = )/4 below the top of the aerial. Since the height of the
aerial is less than )/4 the origin is at a point situated (J — h) below
the surface of the earth. The current /(y) at any point from the base
of the aerial is given by
<— Origin current /, -
I(y) = Ip cos ra (6.18)
Fig. 6.7 Current distribution on a
monopole aerial. where y varies from (J — h) to J.

Example 6.2
A monopole aerial is 36 m in height and is supplied with an input current
of 30 A peak at a frequency of 833 kHz. Calculate (a) the effective height
of the aerial, and (b) the field strength produced at ground level at a distance
of 50 km.

Solution
At 833 kHz

3 x 10°
eS ee a 360 m.
833 x 10
Hence \/4 = 90 m. From equation (6.18),
27
I,
in
= 30 ll Ih cos ea (90 — 26)= I cos 54°.
360
Therefore, J) = 30/(cos 54°) = 51 A. Hence
=
I(y)(y) = 51 cos 46
0
and the mean value of the aerial current is

1 ae 2 ay
I,
mean
Sl Wb ieee cos dy
90 _— 54 54 360
120 RADIO SYSTEMS TECHNOLOGY

i ( 2 ay ) a
sin (———
360
=A = 81.17[sin 90° — sin 54°]
2 1/360 a

= 15.5A.
(a) Hence the effective length of the aerial is
36, Xe
Se a OR ORIN. (Ans.)
30
(Note that the error introduced if the approximation J. = Upny/2 =
36/2 = 18 m had been used is small.)
(b) From equation (6.14)

MA FESO 18.16
= 8 mV/m. (Ans. )
V2 x 360 x 50 x 10°

Top Loading

An increase in the effective length of an aerial could be achieved if

Loading the mean aerial current were to be increased without a corresponding
current “Top loading increase in the input current. This can only happen if the aerial current
is prevented from falling to zero at the top of the aerial by the use
of top loading. Top loading of a monopole aerial means that a horizon-
tal system of conductors, which has a relatively large capacitance to
earth, is fitted to the top of the aerial. The aerial current will flow
in the top loading and so it will not fall to zero at the top of the radiating
part of the aerial. The idea is illustrated by Fig. 6.8. The top loading
makes little, or no, contribution to the total field strength produced
at a distant point because it is only a small electrical distance from
its under-earth image. This image is of the opposite polarity to the
top-loading conductors and hence tends to cancel out any radiation
from the top loading.
The effective length of a top-loaded monopole is calculated in a
Fig. 6.8 Top loading of a monopole manner similar to that used before but the top loading must be taken
aerial. into account. The total length of the aerial must now include the length
of the top loading. Suppose, for example, that the aerial of Example
6.2 was fitted with a top-loading system in the form of a number of
radially spaced conductors of length 5 m. Then the distance y would
vary from / = 90
m to (J — h) = 90 — 41 = 49 mand Inca, =
19.9 A.

\/4 Monopole

When the length of a monopole aerial is increased to \/4, or more,

the electrical distance between a current element in the actual aerial
and the corresponding image current element leads to a phase differ-
 PRINCIPLES OF AERIALS 121

ence between the field strengths produced at a distant point P. Consider

Fig. 6.9, which shows a current element /dy, distance y from the
base of the aerial, which is radiating energy. The radiation from the
image current element /dy has an extra distance 2y cos @ to travel
before it reaches the point P. This extra distance produces a phase
lag @ = (27/X)2y cos 6; thus, referring to Fig. 6.10,
i) E/2
COS?= ==
a Ey

and EF = 2E, cos *

Hence, the total field strength dE; due to the two current elements is

2
dE; = 2E, cos £ = 2E, cos i y cos |

Direct wave

Aerial Reflected wave

Image
aerial

Fig. 6.9 Waves radiated by current

elements in both the monopole aerial
and its image.

¢l2

Fig. 6.10 Phasor diagram of the

field strengths at a distant point.
122 RADIO SYSTEMS TECHNOLOGY

where
_ 607/ dy sin 6
a xD :

The aerial current at distance y from the base of the aerial is

20
I(y) = I cos ee »|
IN
and hence

1207/ sin 0 27 27
diy 2 == nl08"| ——— yas y cos |
AD d rN
The total field strength produced at the point is then

120a/ sin 6 (4 21 24
k= eS cos {|— y } cos {|— ycos 6 ]| dy
dD 0 rN r

120a/ sin 0 (4 25
———— CoS.|-——_ ye. + cos 6)
2D 0 nN

+ cos im VC s=_.cos || dyf

2
sin | = yal + cos 0| sin oe y(1 — cos @) a
a 6071 sin 0 - mn

AD 2
=™ 1 + cos 0) sari — cos 6)
0

sin e=-- -=* cos

COS 0) sin i — RY oon 0
_ 607I sin 6 2 2

AD 26 é 25
+ cos 0) es (1 — cos @)

eee 1
sin — cos (
— cos 6 sin — cos {|— cos 6
6071 sin 6 2 (
AD a 20 iS
=® (1 + c05 8) are — cos @)

tcos A cos B = cos (A + B) + cos (A — B)

jsin (A + B) = sin A cos B + cosA sin B
tsin (—A) = —sin A, cos (—A) = cos A
 PRINCIPLES OF AERIALS 123

cos <i cos 6 cos = cos 0

607/ sin 0 2 2
+

AD 2 2
<= C1 + cos 8) sail — cos 0)

cos ee cos 0 cos we cos 6

= 30/7 sin 6 2, 2
+

D 1 + cos 0 1 — cos0

2 cos S COs )
_ 30/ sin 6 2
D 1 — cos? 6

cos 5 cos )
on. f= py : (6.19
DU GOD, sin 0 yey
a 2
A similar analysis for a \/2 monopole (using i = J sin Ee »))
results in

601 fa (x cos 0) +
E;y= (6.20)
D sin 0

The )/2 Dipole A dipole is a single conductor of length / that is centre fed as shown
by Fig. 6.11(a). Nearly always the length is such that the dipole is
resonant at the frequency of the input signal. This means that it is
one-half wavelength long and has the current distribution shown in
Fig. 6.11(0). In practice, the dipole length is often made slightly less
than )/2 in order to achieve a purely resistive input impedance. At
frequencies in the h.f. band, and particularly in the v.h.f. and u.h.f.
bands, the physical dimensions of the \/2 dipole make it the basic
element of many types of aerial array.
The electric field set up by a X/2 dipole at a distance D is given
by equation (6.19). When 6 = 90°, which defines the equatorial plane
of the dipole, the radiation pattern is a circle (see Fig. 6.12(a)). In
the plane of the aerial the radiation pattern is a figure-of-eight shape,
as shown by Fig. 6.12(b). The beamwidth of the pattern, i.e. the angle
(2) (6) subtended by the 3 dB points, is 129° — 51° = 78°.

Fig. 6.11 (a) 2 dipole, (6) current

distribution on a X/2 dipole. Effective Length

The effective length J. of a \/2 dipole is, from equation (6.13),

124 RADIO SYSTEMS TECHNOLOGY

Equal radiation
in all directions Vertical dipole

Vertical .
dipole No radiation in
Fig. 6.12 Radiation patterns of a direction of
N2 dipole: (a) equatorial plane, aerial axis
(b) meridian plane. (2) (6)

lee = 33
1

f= 2)
/4

—h/4

I 5y
2

nN/4 r
: sin E y =. (6.21)
/f 27/Xr r —n/4 Tv

Radiation Resistance

If the effective length \/z of a X/2 dipole is substituted into equation

(6.8), the value obtained for the radiation resistance will be 80 .
This figure is, however, incorrect. The correct value for the radiation
resistance of a d/2 dipole is 73.14 Q.

Example 6.3

Calculate the electric field strength produced in the equatorial plane of a \/2
dipole at a distance of 20 km if the input current is 3 A r.m.s. Also calculate
the power density of the wave at this point.

Solution
From equation (6.20)
60 x 3
Ee ae ee — SO MV Ms (Ans.)
20
2
Power density = = 215 nW/m. (Ans.)
1207
Alternatively, radiated power

P, = VR, = 9 X 73.14 = 658.3 W

: P.G,t 658.3 x 1.64
and power density = = = 215 nW/m.
47D? 4x 400 x 10°
(Ans.)
—_—_—_
EES

fSee p. 127.
 PRINCIPLES OF AERIALS 125

Gain of an Aerial The gain of an aerial indicates the extent to which the energy that
it radiates is concentrated in a particular direction, or the extent to
which the aerial receives signals better from one direction than from
all others. The gain of an aerial is defined relative to a reference aerial
which is usually either an isotropic radiator or a \/2 dipole. When
the gain is with reference to the isotropic radiator it is expressed in
dBi. The gain of an aerial is the same whether the aerial is employed
to transmit or to receive signals.
The gain of an aerial may be defined in two ways.
(a) The gain is the square of the ratio of the field strength pro-
duced at a point in the direction of maximum radiation, to the
field strength produced at the same point by the reference aerial,
both aerials radiating the same power.
(b) The gain is the ratio of the powers that the aerial, and the
reference aerial, must radiate to set up the same field strength
at a point in the direction of maximum radiation.

Current Element

From equation (6.4) and equation (6.9) the gain of a current element is

ee (an geste 2 = 1.5 or 1.76 dBi.

D D

Monopole Aerial
From equations (6.4) and (6.16)

G= os ) =3 or 4.77 dBi.
D=}iLD
d/2 Dipole
The power P, radiated by a )/2 dipole is equal to 73.14] ? and
substituting J? into equation (6.19) gives
60,/(P,/73.14)
= eee oe

Therefore, the gain of a X/2 dipole is

he etre avy 2
=" 164 or 2.15 dBi.
D D

Effective Radiated Power

The effective radiated power (e.r.p.) of an aerial is the power which

an isotropic radiator would have to radiate to produce the same field
strength at a point in the direction of maximum radiation. Numerically,
e.r.p. is equal to the product of the total transmitted power P, and
the gain G, of the aerial, i.e.
126 RADIO SYSTEMS TECHNOLOGY

e.r.p. = G,P. (6.22)

Using this concept, equation (6.4) can be adapted to give the electric
field strength produced by an aerial of gain G,, i.e.
V30P,G
E = ———. (6.23)
D

Effective Aperture of an Aerial The effective aperture A, of an aerial is the imaginary cross-sectional
area that would absorb the same power from an incident wave as does
the aerial when it is matched to its load. If the incident radio wave
has a power density of Py = E7/120a W/m, then the power P,
received by the aerial will be equal to P;A,. The gain G of an aerial
is directly proportional to its effective aperture; thus

G= : ; ;
Ae ; . (6.24)
effective aperture of isotropic radiator A,.(js0)

d/2 Dipole

The voltage V induced into a \/2 dipole by an incident radio wave

is the product of the electric field strength and the effective length
of the dipole. Thus V = EX/z volts. The dipole has an input resis-
tance that is equal to its radiation resistance of 73.14 Q, and for the
maximum transfer of power to the load, the load resistance must also
be equal to 73.14 Q. Then (see Fig. 6.13)
2 2 bite
Fig. 6.13 Determination of the E*Ac ges 4 cs E*
effective aperture of a /2 dipole. 12007. 4R, 427 x 73.14
30?
or A, = ——— = 0.132?. (6.25)
73.147

Current Element

Since the current flowing in the element is assumed to be of uniform

amplitude the voltage V induced into the aerial by an incident radio
wave is equal to Ed/. Therefore
EUG) SOV? . (edly
1207 4R, 2
320 3? (.)
d
3)?
or A, = —— = 0.12). (6.26)
8a

Isotropic Radiator

A current element has a gain of 1.5 relative to an isotropic radiator,

and so
 PRINCIPLES OF AERIALS 127

15 = eco. 3%”
Ac(iso) 8 TA (iso)

2
or Aigo) = ae (6.27)

Power Received by an Aerial

The power P, received by an aerial when it is sited in a field of

strength E V/m is
apy, E22
2G: = (BY,
2

1207(47) 30 \4a
From equation (6.23),

rr
_ (G0P,G)
D
and substituting this gives
r 2
P= O6
t Pp,
AG(=~). 6.28
(6.28)

Example 6.4
The transmit aerial of a 600 MHz radio link has a gain of 26 dB. The signal
is received by an identical aerial that is 30 km away. Calculate the overall
loss of the radio link.

Solution
From equation (6.28)
‘ a x 10° 2
PJP, = 398° x 3
4a x 30 x 10° x 600
x 10°
= 2.79 x 1077.
Therefore, the link loss = 65.5 dB. (Ans.)

Example 6.5
A /4 monopole aerial is supplied with a current of 30 A r.m.s. at 3 MHz.
Calculate (a) the field strength produced at ground level at a point 50 km
distant, and (b) the power received by an aerial of 10 dBi gain situated at
that point.

Solution

Qe a mM
3 x 10
and hence \/4 = 25 m. Since the monopole aerial is \/4 long the input
current is at the origin and so J = 30 cos (2 ry/100). The mean aerial
current is
128 RADIO SYSTEMS TECHNOLOGY

jee 21 30 sin 2 my \ |?
Dineen ra | 30 cos i by = ( )|
Dexlio 100 25 | 27/100 \ 1007 |

Ve 100 50
= ee xX sin eS = LORleAc
21 100
Therefore

19. xX 25 4
Lease SS cae ey en eS Re 15.9 m.
30
Substituting into equation (6.14)
120” x 30 x 15.9
= 36 mV/m. (Ans. )
100 x 50 x 10°
(b) The receive aerial has a gain of 10 dBi = 10 times, hence its effective
aperture is
107-10 x 10° :
A, = 10 x A. (iso) — = = 7958 m.
4a 4a
The power density at the distant point P is P, E*/120xc =
3.44 x 10~° W/m, and therefore the received power is

P, = A.P, = 7958 x 3.44 x 107° = 27.36 mW. = (Ans.)

Long-wire Radiator Figure 6.14 shows the basic arrangement of a long-wire radiator.
Essentially, it consists of a conductor, several wavelengths long at
the frequencies of operation, which together with the earth forms a
transmission line of characteristic impedance Zp. At its sending end
a radio-frequency source of e.m.f. E; and impedance Zp supplies a
current I; = Es/2Zp. The attenuation of the line is negligibly small
and so this current flows, with unchanged amplitude but varying phase,
Zo
to the matched load. The length of line / can be considered to consist
of the cascade connection of a large number of current elements. Each
current element will radiate energy, the amplitude of which is at its
maximum value in the equatorial plane and zero along the axis of
the line. The total field strength produced at a distant point P is the
phasor sum of the field strengths produced by each of the individual
Fig. 6.14 Long-wire radiator. current elements.
Consider Fig. 6.15. The current at a distance x from the sending
end of the line is J, = Ise ~)** and so it lags the sending-end current
I by angle Bx = 27x/) radians. The current element J, dx is closer
to the distant point P by x cos @ metres and so the field strength dE
produced by element J, dx lags the field strength due to the current
element at the beginning of the line by angle (2mx/A) (1 — cos 6)
radians. Now,

jO0aI dx sin 0
dE =
AD
and hence the total field at P is
 PRINCIPLES OF AERIALS 129

Fig. 6.15 Calculation of field

strength.

E= (j60m sin @ heresies ae dx

0 AD

e G2 ma/A)L — cos 8) |
j607/, sin 0
\D (j2 7x/A)(1 — cos 8) |o

_ 30s sin 0 fe ~ 2 Aya — cos 8) _ 4]

D 1-cos@

_ | 60/5 sin 0 uf Stl ‘3

or E = i eee - sin(< ad cos »)|4 (6.29)

The main lobes of the radiation pattern of a long-wire radiator for

five different lengths are shown in Fig. 6.16. As the length of the
wire increases the number of sidelobes increases also but these are
not shown. When the length of the wire is between 4) and 8A there
is a linear relationship between the angle of the main lobe and
frequency. The long-wire radiator can be used as an aerial in its own
right but, more often, it is a part of a rhombic aerial.

Example 6.6
Calculate the ratio of the voltages induced into a long-wire radiator of 100 m
length if the incident radio wave is at 15 MHz and the angle of incidence
is (a) 17°, and (b) 24°.

“eve @)l-2(8)
te-i* = cos x — j sin x, |e7/* — 1] = [(cosx — 1)? + sin? x]
130 RADIO SYSTEMS TECHNOLOGY

48° 35° 24° 17°

68°

Fig. 6.16 Radiation patterns for

long-wire radiators. N2 r 2 4nr 8X

Solution
From equation (6.29), with \ = 20 m,

sin 17° A ae LOO

sin (1 — cos 17°)
1 =coseiiy~ 20
ratio =
sin 24° az X 100
sin (1 — cos 24°)
1 — cos 24° 20

= 0.92. (Ans.)

Values of the lobe angle 6 for different values of wire length, in

wavelengths, can be obtained from equation (6.29) and are shown
plotted in Fig. 6.17. It can be seen that the lobe angle varies, more
or less linearly, from 24° at] = 4) to 17° at 1 = 8X.

60

40

Angle
of
(0°)
lobe
main 20

Fig. 6.17 Main lobe angle; variation 1 2 Ss FAP 56 I BB 10% Titeat2

with wavelength.
Length in wavelengths
 7 Aerials

For radiocommunication between two points, high-gain, directive

aerials are required for both transmission and reception. In the h.f.
band the rhombic aerial has been widely employed for many years
but it has now been superseded, for all but a few special applications,
by the log-periodic aerial. The log-periodic aerial can provide an
equal, or superior, performance in terms of both gain and bandwidth
for a smaller physical size and, hence, cost. All h.f. radio links employ
sky-wave propagation via the ionosphere, and for the best results the
aerial must be selected to suit the propagation conditions pertaining
at a given time. In particular, for reliable communications to be main-
tained, the system needs to be flexible in its usage of frequency and
this means that both the transmitting aerial and the receiving aerial
must be wideband. Nowadays, the selection of frequencies, eleva-
tion angle, etc., is made easier and quicker by the use of computer
control.
Omnidirectional h.f. transmitting aerials are required for some
purposes, and in the past were mainly vertical whip aerials. These
are of fairly low gain and of inadequate performance at high eleva-
tion angles, and modern systems probably use a vertical stack of loop
aerials fed in a log-periodic manner.
In the v.h.f. and u.h.f. bands the dimensions of a \/2 dipole are
small enough for linear arrays of dipoles to be the predominant type
of aerial. Alternatives that are also often employed are the corner
reflector, the Yagi, and the log-periodic aerials. For omnidirectional
radiation of energy a vertically stacked sleeve dipole is often used.
Mobile radio normally uses either the whip aerial or the helical aerial.
In the s.h.f. band high-gain parabolic dish aerials are the most
common types, although horn aerials are also employed.

Arrays of Driven Dipoles A N/2 dipole has a gain of 2.16 dBi, a circular equatorial-plane radia-
tion pattern, and a figure-of-eight radiation pattern in the meridian
plane. For many applications this gain and directivity is inadequate
and then two, or more, dipoles can be used in an array. A variety
of radiation patterns can be obtained by varying one, or more, of:
(a) the number of dipoles used, (b) the spacing between the dipoles,
and (c) the amplitudes and relative phases of the dipole currents. The
radiation pattern and the gain of an aerial may be affected by mutual
impedances between the dipoles in an array, but this factor will not
132 RADIO SYSTEMS TECHNOLOGY

be considered until page 149. In the past arrays of dipoles were often
employed as h.f. aerials, but today their main applications are in the
v.h.f. and the u.h.f. bands.

Two-dipole Array

Vertical Dipoles
Horizontal-plane Radiation Pattern
Figure 7.1 shows two vertical \/2 dipoles mounted in the same
horizontal plane and spaced apart by a distance of d metres. If the
dipoles are supplied with equal-amplitude, in-phase currents, both
dipoles will radiate energy equally well in all directions in the horizon-
tal plane, and the total field strength at any point in this plane will
be the phasor sum of the individual field strengths produced by each
aerial. If the distance to a distant point P is very much larger than
the dipole spacing d these individual field strengths will be of equal
amplitude. They will not usually, however, be in phase with one
another. The point 0 is equidistant from both dipoles and so at this
point the two fields are in phase and so will simply add. In general,
the energy radiated by dipole A must travel a further distance d cos 0
to reach the point P. Therefore, the field strength due to dipole A
will lag the field strength due to dipole B by an angle
21d
d= ae cos @ radians.

The phasor diagram of the field strengths at the point P is shown by

Fig. 7.2. From this figure, the total field strength Ey is equal to
d
Ey = 2E cos © = 2F cos Se (7.1)
If either \/2 dipole on its own was supplied with a power of P watts

Plan view
of 2 vertical
dipole

Fig. 7.1. Two-dipole array.

 AERIALS 133

ping! (Ex! =lE_!=1E|

lee ss

Fig. 7.2 Phasor diagram of the field

strengths at a distant point. Ey E,

the field strength produced at the point P would be Ey = 60//D. If

the same power is supplied to the two-dipole array each dipole will
receive a power of P/2 watts and so will produce a field strength of
E)/N2 at point P. Hence, the total field strength at the point P is
Eo amd cos 0 md cos 0
Ey = 2 — cos (———_] = V2Ey cos (——— }.
mee ( N ( nN G2)
The gain of the array, relative to a \/2 dipole, is 20 log,)V2 = 3 dB.
The array factor is the term V2 cos [(d cos 0)/N].

Example 7.1

Calculate and plot the horizontal radiation pattern of two vertical \/2 dipoles —
which are )/4 apart and supplied with equal-amplitude, in-phase currents.

Solution
DTN T
Opeee COS0 COS),
A 4 2
Hence the array factor is V2 cos [1/4 cos 6]. It is usually sufficient to
calculate the resultant field strength at 30° intervals over the range 0° to 180°,
see Table 7.1.
Using the figures given in the final row of Table 7.1 the radiation pattern
for the array has been plotted and it is shown by Fig. 7.3(a).

When the currents in the two dipoles are of equal amplitude but
there is a phase difference a between them the radiation pattern will
be altered. The angle a must be added to ¢ to obtain a new phase
difference y = @ — a radians if the phase of I, leads Jp, or y =
¢ + a if the phase of Jp leads J,. Very often, the value of a is made
equal to the distance between the dipoles. If, for example, the dipole
spacing is \/4, a will be equal to 90°.

Table 7.1
BOE ee ae ee, Se ee a
0 0° 30° 60° 90° 120° 150° 180°
S
eS Si ie e eS e
cos 6 1 0.866 0.5 0 —0.5 — 0.866 =a

45° cos @ 45° 39° 22.5° 0° 22.0" 39° —45°

{2 cos (45° cos @) 1 1.1 1.3 1.41 1.3 te 1

eH
ee
134 RADIO SYSTEMS TECHNOLOGY

Fig. 7.3 Radiation pattern of a 90°

two-dipole spaced )/4 apart with
(a) equal-amplitude in-phase
currents, (6) and (c) equal-amplitude
90° out-of-phase currents.

180°—__

270°

(a)

180°——_
te 180°

270° 270°

(6) (c)
 AERIALS 135

Example 7.2

Calculate and plot the horizontal plane radiation pattern for two vertical \/2
dipoles that are \/4 apart and are fed with equal-amplitude currents that are
90° out of phase.

Solution
If Ig leads I,,
2m
ea ee = 90°(1 + cos@)

and hence the array factor is V2 cos [45°(1 + cos 0)]. The results are
tabulated in Table 7.2. The radiation pattern is shown in Fig. 7.3(b).
Conversely, if J, leads J,,

Ww = 90°(cos 6 — 1)

and the array factor is equal to 2 cos [45°(cos 6 — 1)]. The results are
presented in Table 7.3. This radiation pattern is plotted in Fig. 7.3(c).

The currents supplied to the two dipoles may not always be of equal
amplitude; Fig. 7.4(a) shows the phasor diagram of the field strengths
at the distant point P when |J,| > |Z,|. Taking Eg as the reference
phasor and resolving E, into its horizontal and vertical components
gives the phasor diagram shown in Fig. 7.4(b). From this diagram
the total field strength Fy is

Ey = J[(Eg + E, cos y)* + EX sin? ¥]

= |(E% + 2E,E, cos y + E4 cos? y + E% sin’ p)
or Ey = (EX + EX + 2E,Eg os y). (7.3)

Table 7.2
at RI sR ca a ree Rete bE
6 0° 30° 60° 90° 120° 150° 180°
rr rts ee ly ee eee
1 + cos 6 2 1.866 1.5 1 0.5 0.134 0
45° (1 + cos 6) 90° 84° 67.5° 45° 22.5° 6° 0°
Array factor 0 0.15 0.54 1.0 1.31 1.41 1.41
lee ees alge eer) oil ie, vital aA) were eters

Table 7.3
Na nae cleSse ay oe el i Des NE ch er
6 ere 0° 30° 60° 90° 120° 150° 180°
Cn ee es ee ee ee
cos § — 1 0 —0.134 —0.5 ai1,0 2 — 1.866 2
45° (cos-§ — 1) 0° —6° 22/5° —45° —67.5° —84° —90°
1.41 1.41 1.31 1.0 0.54 0.15 0
Array factor ee
is nb pess cont cute ethnics wT
136 RADIO SYSTEMS TECHNOLOGY

Fig. 7.4 (a) Phasor diagram of En

unequal-amplitude field strengths, E, siny EY
(6) simplified version of (a). (a) (b)

Note that if Ex = Ep = E

Ey = |[2E7(1 + cos y)] = V2EJ(1 + cos y)

= V2E\(2 cos” ¥/2) = 2E cos y/2,
as before.

Vertical-plane Radiation Pattern

The vertical-plane radiation pattern of two vertical \/2 dipoles
mounted in the same horizontal plane is obtained by multiplying the
array factor by equation (6.19), i.e.
cos (7/2 cos @)
Ey = V2E cos y/2 (7.4)
sin 0

Co-linear Vertical Dipoles

Two co-linear vertical dipoles (see Fig. 7.5) have a circular radia-
tion pattern in the horizontal plane since, quite clearly, the energy
radiated from each dipole does not combine with that from the other.
In the vertical plane the radiation pattern is given by the product of
the array factor and the dipole’s meridian-plane pattern, i.e. by
equation (7.4).

ee
PY Dipole in Front of a Reflecting Plane

If a \/2 dipole is mounted in front of a reflecting plane the image

ge
.
h
oees
Pae
GET
ate
ENS of the dipole will act as though it were the second dipole in a two-
: , ; dipole array. The polarity of the image dipole depends upon the
Bee ee ee orientation of the physical dipole; Fig. 7.6 shows the two most
common cases. When, Fig. 7.6(a), the dipole is mounted parallel to
the reflecting plane the image dipole has the opposite polarity.
Conversely, Fig. 7.6(b), if the dipole is mounted normal to the reflect-
ing plane, both the dipole and its image have the same polarity. The
apparent spacing d between the two ‘dipoles’ is twice the distance
s between the dipole and the reflecting plane.
One application of this principle is the corner reflector. A /2 dipole
is placed in a position parallel to the intersection of two reflecting
planes as shown by Fig. 7.7(a). The reflecting planes are made of
 AERIALS 137

ener | eee pete fN(e

os Reflecting plane
Fig. 7.6 Dipole mounted in front of Reflecting plane
a reflecting plane: (a) parallel to, and
(6) normal to the plane. (a) (6)

Coaxial feeder

d/2 dipole

Reflecting (6)
plane

(a)

Fig. 7.7 (a) Corner reflector, and either solid metal or wire mesh or perhaps a system of conductors
(b) operation of the corner reflector. spaced apart at 0.1) or less, and they are at an angle of either 90°
or 60° to one another. Each plane must be at least one wavelength
long in each direction, i.e. / = \. The spacing of the /2 dipole from
the intersection of the planes is somewhere between 0.3 and 0.5
with the latter distance being the most common. Assuming that the
plane angle 0 is 90°, Fig. 7.7(b) shows that reflections from the corner
reflector produce image dipoles A, B and C; image A has the same
polarity as the actual dipole D, while both images B and C have the
138 RADIO SYSTEMS TECHNOLOGY

opposite polarity. Effectively, therefore, the corner reflector gives

a four-dipole array with two dipoles in each of two planes. The way
in which the array factor of such an array can be calculated is shown
on page 146.

Three-dipole Array

An increase in both the gain and the directivity can be obtained if

a third dipole is added to give a three-dipole array. If the centre dipole
B is taken as the reference then at the distant point P the field strength
due to dipole A will lag the field strength due to aerial B by an angle
21d
y= cos? +a;

while the field strength due to dipole C will lead by an angle

d
b= = cos 6 8.

Example 7.3

In a three-dipole array the dipole spacing is \/2 and the dipole currents are
21 20° for the centre dipole and J 2 180° for both the outer dipoles. Determine
the expression for the radiation pattern of this aerial.

Solution

T
= —— — cos §_= 7.cos0,
2
y = 7 — 7COS

and y' = x + mos0.

The phasor diagram of the field strengths at a distant point P is given by

Fig. 7.8(a). The diagram can be simplified by resolving the two smaller
phasors into their horizontal and vertical components (see Fig. 7.8(b)). From
Fig. 7.8(b), the total field strength is

Ey = 2E + Ecos [x(1 + cos 6)] + Ecos [x(1 — cos 6)]

= 2E[l — cos (a cos 6)] = 4E sin? [(a/2) cos 6]

4
OF Fe = A Ep sin? [(x/2) cos 0]. (Ans.)

Broadside Array

A broadside array consists of a number n of \/2 dipoles equally spaced

in one line and carrying equal-amplitude, in-phase currents. If the
total power supplied to the array is P, watts, each dipole will be fed
 AERIALS 139

E sin x(1+ cos @)

E cos 7(1—cos 6)

2E CE cos m1(1+COos 0)

Fig. 7.8 (a) Phasor diagram of a

E sin 1(1—cos @)
three-dipole array, and (b) simplified
_ version of (a). (a) (b)

Fig. 7.9 Broadside array.

with P,/n watts and will produce a field strength of E,/Nn at the
distant point P.
Suppose, for example, that the number n of dipoles is 5, as in Fig.
7.9. The radiation from dipole A has a further distance d cos 6 to
travel in order to reach the distant point P than has the radiation from
dipole B. The field strength due to aerial A will therefore lag the field
strength due to dipole B by angle ¢ = (2 1d/)) cos @ radians. In similar
manner, field strength Eg lags field strength Ec, Ec lags Ep and
Ep lags Ep, all by the same angle ¢. Figure 7.10(a) shows the phasor
diagram of the field strengths at point P.
If a line is drawn normal to the centre of each phasor the lines will
meet at the point marked as 0. Lines then drawn from point 0 to the
ends of each phasor subtend the angle ¢ (see Fig. 7.10(b)). The total
field strength Ey is the phasor going from the base of phasor Ex to
the tip of phasor E, as shown by Fig. 7.10(c). From this diagram,
Ey = 2AC sin (n@/2). To find the value of the length AC, consider
the triangle ACD. From this
BD
sin (/2) = ve
140 RADIO SYSTEMS TECHNOLOGY

E/2
or AC = ————.
sin (¢/2)
Therefore

= eee (7.5)
sin (¢/2)
This can be written as

ae (Eo/Vn) sin (nd/2) is EoVn sin (ng/2) 0.6

sin (¢/2) n sin (/2)
When the angle 6 is very nearly equal to 90°, cos 0, and hence ¢,
is small. Then
Vn sin (ng/2) _ Vn ng/2
nsin($/2) sn g/2
and the gain of the broadside array is (Vn)? or n, i.e. the gain is
equal to the number of dipoles in the array.
The radiation pattern of a broadside array has two main lobes in
the plane perpendicular to the line of the array. Its exact shape is best
determined by finding the angles at which maxima and minima occur.

Example 7.4
(6)
A broadside array consists of six vertical \/2 dipoles spaced \/2 apart, and
energized by equal-amplitude, in-phase currents. Calculate and plot the
horizontal-plane radiation pattern of the aerial.

Solution
Here n = 6 and

Pysh ON
= —— COS 0 = 1m COS.05
NGED
and so
_ Ey sin 37 cos @)
. V6 sin [(2/2) cos 6]
At 6 = 90°
E
(c) E, = — x 6 = V6Ey.
V6
Fig. 7.10 Phasor diagrams of the Nulls occur in the radiation pattern when the numerator of equation (7.6)
field strengths produced by a is zero and the denominator is not zero, i.e. when 3 cos 6 = 1, 27, 37,
broadside array. etc. Hence

(a) cos @ = 4,0 = +70.5°

(b) cos 0 = 4,0 = +£48.2°
(C)<cosi6n=ile bp=1020

The centre of each minor lobe occurs when sin (n¢/2) = 1 or no/2 =
+(2k + 1)x/2 or
 AERIALS 141

ap +(2k + 1)d
COS 0 Rat
2nd
In this case,
+(2k + 1)A (2k + 1)
cos @ = ;
2x6 xX X/2 6
(a) k= 1; 0 = cos! [+3] = +60°;

(b) k = 2;0 = cos"! [+3] = +33.6°;

(c) k = 3; 6 = cos”! [+4; which is, of course, not possible.

The radiation pattern is shown by Fig. 7.11.

Beamwidth

The beamwidth of an aerial is the angle subtended by the 3 dB points

on the major lobe(s) of the radiation pattern. Thus, in Fig. 7.12 the
angle 6 is the beamwidth. The maximum field strength produced by
a broadside array is Ey @nax) = VnE) = nE. Therefore
Er ___ sin (n@/2)
ae 2 Eximaxy 1 sin (9/2) |
If the angle @ that gives Ey = Exymaxy/V2 is large, cos 6 will be
small and then sin (¢/2) is small. Therefore
Ey - sin (nd/2)
Ev (max) n / 2

and this is equal to 1/V2 when n@/2 = 1.39 radians. Hence

nid d cos 0 me
dr
i 1.39
or cos @ = sin (90° — @) = —.
nnd
Since @ is large, sin (90° — 0) = 6 = 8/2, and so the beamwidth @ is

van 0.885 51d°

ys (7.7)
nd nd
In the case of the array in Example 7.4

Emax!2

270°
Fig. 7.11 Broadside-array radiation ZA es
pattern. 2

Fig. 7.12 Beamwidth of an aerial. Ena!{2

142 RADIO SYSTEMS TECHNOLOGY

51 <n
6 x d/2
The greater the number of )/2 dipoles employed in an array the more
directive the radiation pattern will become, although the number of
small secondary lobes will also increase.

End-fire Array

An end-fire array consists of a number n of /2 dipoles spaced \/4

apart and fed with equal-amplitude currents but with a progressive
phase difference equal to the spacing, 1.e.

a= = M4 = z/2 radians.

The same analysis as that used for the broadside array is applicable,
but with Y = @ + a replacing ¢, to give

Ey _ VnEp sin (ny/2) (7.8)

nsin (y/2)
Example 7.5

Calculate and plot the radiation pattern of a six-dipole end-fire array.

Solution

De ON T
y ry m/2 Pree + cos 9).

The maximum field strength Eqymax is V6Ey when y = 0, i.e. when

(1/2)(1 + cos 0) = 0 or 6 = 180°.
Nulls in the radiation pattern occur when

sin [(37/2)(1 + cos 0)] = 0

or |Bix/2)C"“"cos 0) = "ki.

(a) k = 1;cos8@=4— 1 = -—4 or = 109.5°.

(b) k = 2; cos =F -—1=4 ord = 705°.
() k =3;cs0 =F —1 = 1 or 0 = 0°.

The radiation pattern is shown in Fig. 7.13 and clearly there is only the
one main lobe.

Height Factor When an aerial is mounted above the earth, ground reflections will
produce an image aerial beneath the earth as shown by Fig. 7.14.
If the elements of the aerial are mounted vertically the image aerial
will have the same polarity as the actual aerial, but if the aerial
elements are in the horizontal plane the image aerial will be of the
opposite polarity.
 AERIALS 143

90°

. (0 0°

Fig. 7.13 End-fire array radiation

pattern. 270°

!
6 6

h h

Earth Earth

A h
a= | 1

6
i| @
See
Fig. 7.14 Aerial mounted above the |
earth: (a) vertical aerial, and
(b) horizontal aerial. (a) (b)

An aerial and its image behave in the same way as a two-dipole

array with a spacing equal to 2h, where h is the height of the aerial
above the earth.

Vertical Aerial

The aerial and image currents are in phase with one another, and so
the total field strength produced at a distant point P in the vertical
plane is (from equation (7.1) with d = 2h)
2ah
Ey, = 2E cos exe COS 0)
144 RADIO SYSTEMS TECHNOLOGY

The angle 0 is now the angle relative to the vertical line passing through
both the aerial and its image. If the angle of elevation + is used instead,
which would be more convenient in practice, y = 90° — 0, and
cos 0 = cos (90° — vy) = sin y. Therefore
27h si
Ey =)2E cos a (7.9)
r
The height factor H(y) is
2th
H(y) = 2 cos orem si

and it is the factor by which the free-space radiation pattern must be

multiplied to take account of ground reflections.

Horizontal Aerial

Now the apparent current in the image aerial is in anti-phase with

the current in the actual aerial, and hence

2th si
Ey, = 2E cos Gearon si 90°

or Ey = 2E sin ‘eeu! (7.10)

The angle of elevation y of the main beam is the one that makes
_ «| 27h si i
sin |——q sin ap ale

Then
27h
a= any = a2

Or .y = sins! iii
: ah |

Example 7.6

A transmitting aerial operates at 20 MHz and it is to have its maximum radia-

tion in the vertical plane at an angle of elevation of 14°. Determine the height
above ground at which the aerial should be mounted if it is (a) horizontally,
or (b) vertically mounted.

Solution

3 x 108
= ———— = 15m.
20 x 10°
 AERIALS 145

(a) From equation (7.10), the maximum field strength is obtained when
; 27h\ .
Sint ||) Sin. 14°o) = "1
r

2th
or Were sin. 14° = /2.

Therefore,
15
Noa es ae Se (Ans.)
4 sin 14°
(b) From equation (7.9), the maximum field strength occurs at the height
where
2ah\ .
cos | |—— } sin 14°} = —-1
r

2anN
or (———=} sin il4? = 7:
ON
Therefore

15
= ees a OTM, (Ans.)
2 sin 14°

Figure 7.15 shows the height factor H(7) plotted for heights of \/4,
/2, 3/4 and Xd above the earth for both horizontal and vertical aerials.

Wins ahaa ie Se
h=NN4 h=N2

h=3N4 h=x

(2)

ON ee
h=N4 h=N2

Earth
h=3N/4 h=
Fig. 7.15 Height factor for
(a) vertical and (b) horizontal aerials. (6)
146 RADIO SYSTEMS TECHNOLOGY

It is clear that with a horizontal aerial there is zero radiation at ground

level as well as at some angles of elevation.

Pattern Multiplication The use of pattern multiplication to obtain the radiation pattern of
a two-dipole array has already been touched upon (p. 136), but it was
not specifically labelled as such. The radiation pattern of a two-dipole
array can be drawn for each of the three planes shown in Fig. 7.16.

(a) In the plane xz, i.e. the equatorial plane, the array factor is
V2 cos [(2/4) cos 6]. When 6 = 0° this gives <2 cos (1/4)
= 1, and when 6 = 90° it gives V2 cos 0° = V2 (see Fig.
7.17(a)(i)). The radiation pattern of a single dipole is a circle
(Fig. 7.17(a)(ii)), and the overall radiation pattern in this plane
is the product of (a)(i) and (a)(ii) and this is shown by Fig.
7.17(a)(iii).
(b —
In the plane xy, the meridian plane, the array factor is the same
as for the equatorial plane. The dipole pattern is given by
equation (6.19) and it is shown by Fig. 7.17(6)(ii). The overall
radiation pattern is given by the product of these figures and-.
it is shown by Fig. 7.17(6)(iii).
(c) In the plane yz, once again each dipole has a radiation pattern
given by Fig. 7.17((b)(@ii)=(c)(@ii)) but in the array factor
equation the angle 6 has only the value of 90°. Hence, the array
factor is always equal to V2 (see Fig. 7.17(c)(i)). The overall
radiation pattern is the product of Figs 7.17(c)(i) and 7.17(c)(ii)
and it is given by Fig. 7.17(c)(iii).

The principle of pattern multiplication can always be employed to

determine the radiation patterns of complex aerial systems. Suppose,
for example, the radiation pattern of an array consisting of three rows
of four \/2 dipoles (Fig. 7.18) is to be determined. In each plane the
overall radiation pattern is the product of (a) the pattern of an
individual )/2 dipole, (b) a four-dipole array, and (c) a three-dipole
array.

>

Fig. 7.16 Showing the planes xz,

xy and yz of an aerial (4/4 spacing).
 AERIALS 147

(i) (ii) (iii)

ie ‘Z Z

x =

x x x (a)

y
y y

x =

x x x (b)

Y y y

Fig. 7.17 Pattern multiplication.

y Example 7.7

Calculate and plot the horizontal-plane radiation pattern of the four vertical
)/2 dipole array shown in Fig. 7.19. The dipoles are fed with equal-amplitude,
in-phase currents.
La
——JI ee Ge
oo.
Solution
The radiation pattern due to dipoles A and B, and to C and D, is given by
2m
E, = V2E, cos ven cos d = /2E, cos ee cos a}
4.x 2X 4
Sean
pl:
ae
Similarly, the radiation pattern due to the dipoles A and C, or B and D, is
given by

Ey = V2E, cos [(7/4) sin @].

The overall radiation pattern is therefore given by

Sees |
§
Dae
toa
=
Lae
ab
|
aes aS
Ee
ee
oe
3 |
Ey, = 2E} cos [(1/4) cos 0] cos [(7/4) sin 6].
N out of paper
This is tabulated in Table 7.4. The radiation pattern is shown in Fig. 7.20.
Fig. 7.18 4 x 3 array.
Example 7.8

A dipole array is receiving a strong interference signal 50° from the centre
of the main lobe of its radiation pattern. To eliminate this interference another
148 RADIO SYSTEMS TECHNOLOGY

Pe, a
Pcie 90°
a)
a |
a
ee He

he D1 yy Ya
%\

(2) [oho
180°— — 0°

Fig. 7.19

Fig. 7.20 270°

Table 7.4

0 0° 30° 60° 90? 120° 150° 180°

45° cos 6 45° 39° 22.5° 0° ~22.5° — 39° —45°
45° sin 0 0° 229° 39° 45° 39° 22.5° 0°
{2 cos (45° cos @) 1.0 at 1.31 1.41 1.31 104 1.0
J2 cos (45° sin @) 1.41 1.31 LeU 1.0 11 1.31 1.41
Overall 1.41 1.44 1.44 1.41 1.44 1.44 1.41

array is to be used to position a null at 50°. Determine the necessary separation

of the two arrays.

Solution
The principle of pattern multiplication states that if the two-dipole array factor
has a null at 50° then the overall radiation pattern will also have this null.
Hence

td
cos < cos 50° |= oO
IN

0.643 ad
or = 7/2
d
nN nN
and = = 0.78 Xd. (Ans.)
2 X 0.643 1.286
 AERIALS 149

Mutual Impedances between In the treatment so far of dipole arrays it has been assumed that there
Dipoles was zero interaction between the dipoles. This assumption is not
correct if the spacing between the dipoles is small. Each dipole that
carries a current will induce a voltage into every other nearby dipole
and this means that each pair of dipoles has a mutual impedance
between them. The magnitude and phase of each mutual impedance
is determined by the length and the diameter of the two dipoles and
by their spacing.
When a current flows in a dipole the e.m.f. it induces into another
dipole in the array tends to oppose the voltage that is applied to that
dipole. The voltage applied to a dipole must therefore be equal to
the sum of (a) the voltage needed to produce the dipole current if
there was zero mutual impedance, and (b) the voltages necessary to
balance the e.m.f.s induced into the dipole by currents flowing in the
other dipoles. Consider, as an example, the three-dipole array shown
in Fig. 7.21. For this array the voltage equations are:

Va = [nZan + IpZap + IcZac

Vz — IpZpp 4e In Zap ata Ic Zpc (Gig)

Vo = IcZec st In Zac ae IpZpc.-

The input impedance of a dipole is the ratio (input voltage)/(input

Fig. 7.21 Mutual impedances current); hence
between dipoles. Za = VialTy = ZAA at Zaplp/Ia + Zaclc/In
Zp = Vp/Tpz = ZBB ad Zaplallp = Zaclc/lpz (al2)

Zc = Velie = Zec ar Zaclallc ap Zpclp/Ic.

The self-impedance, Za,, Zpg, Zcc, of a dipole is equal to its radia-

tion resistance, which if the dipole is slightly shorter than /2, is
73.14 Q.
If the dipoles are fed with equal-amplitude, in-phase currents, as
for a broadside array, then J, = Jy = Ic and Za = Zan + Zap t+
Zac. If the dipoles are spaced )/4 apart and the currents have a
progressive phase difference of 90°, as for an end-fire array, then
Ig = jl, and Ic = jf = jh = —In, and, for example; Za =
Zan + jZap — Zac.
The power radiated by each dipole is equal to the square of the
dipole current times the real part of the dipole impedance. Thus, the
power Py, radiated by dipole A is equal to J 4 xX (real part of Z,).

Example 7.9

Two vertical \/2 dipoles spaced \/4 apart are supplied with equal-amplitude
currents having a phase difference of 90°. Each dipole has a radiation
resistance of 73 Q and their mutual impedance is 40 — j30 (2. Calculate (a) the
radiated power, and (b) the gain of the array in dBi, if the dipole currents
are both 100 mA.

Solution
(a) I, = 100 mA, Jy = j100 mA.
150 RADIO SYSTEMS TECHNOLOGY

j100 F
Ze Tao (A) — 130) Nea 103 + j40Q.
P, = (100 x 10°)” x 103: = 1:03. W-
—j100 ;
Zn ="73"+' (40"—"730): x = 43 — j40.0.

Pei S*(100UKCLOT? > x 43 = 0.43 W.

The total radiated power = 1.46 W. (Ans.)
(b wa In the end-fire direction the electric field strength is 2E, where E is
the field strength produced by either dipole. If the total power had been
supplied to one dipole only its current would have been (1.46/73) =
141.4 mA, giving a field strength of
141.4
Ba se
100
Therefore, the gain, relative to a \/2 dipole, is
2E x 100
20108751. | = 3:02 dB:
141.4E
Therefore, gain in dBi = 3.02 + 2.16 = 5.18 dBi. (Ans.)

Example 7.10

Determine the gain, relative to a \/2 dipole, of a four-dipole broadside array

that uses a dipole spacing of \/4. The mutual impedances between the dipoles
are Zap = Zc — Zep — —12 5 j30 Q, Zac — Zap — 4 ae j20 OF and

Zap = —2 — jl2 Q. Each dipole has a radiation resistance of 73 .

Solution
The aerial currents are of equal amplitude and phase. Hence,

Zy = 73 + (—12 + j30) + (4 + j20) + (—2 — jl2).

The real part of Z, is

1 — 12 4 =" 2° ="63 0
and so the radiated power is Py = 63/* W. Also,

Zz = 73. + (=—12 + j30) + (—12 +. j30) + (4 + j20),

giving a real part of 53 Q. The power radiated by dipole B is Pg = 53/7 W.

Zo = 73 + (4 + j20) + (—12 + j30) + (—12 + j30)

and the radiated power is Po = 5317 W. Lastly,

Zp = Z, = 63 + j382
and the radiated power Py) = 63/7 W.
The total radiated power is

Py = Py + Py + Po + Py = 23217 W.
If this power were radiated by one dipole only, its current would have to
be |(23217/73) = 1.7831. Therefore
 AERIALS 151

4E
the gain = 20 log, Pal = 7.02 dB. (Ans.)

If the mutual impedances had been ignored the gain of the array would have
been 20 logiy V4 = 6.02 dB.

It should be noted that when the dipoles are supplied with in-phase
currents the power radiated by each dipole can be obtained by taking
only the resistive part of each mutual impedance.

The Rhombic Aerial A long-wire radiator (p. 128) produces a radiation pattern that has
one main lobe and a number of small side-lobes. The main lobe is
at an angle @ to the wire axis and is a function of the electrical length
of the wire. The rhombic aerial uses four long-wire radiators to form,
in the horizontal plane, a rhombus shape as shown by Fig. 7.22. The
tilt angle B is chosen to ensure that (a) the lobes marked as X point
in opposite directions and their radiated energies cancel out, and
(b) the lobes marked as Y point in the same direction and hence their
radiations are additive. For this to occur two things are necessary:
(a) the distance x from the mid-point of one wire to the mid-point
of the next must be )/2 longer than the direct distance y between these
two points; (b) the tilt angle 6 should be equal to (90° — @). Since
the lobe angle varies with frequency the choice of the tilt angle must
be a compromise and it is generally calculated at the geometric mean
of the two extreme operating frequencies.
The gain of a rhombic aerial is a function of the length of each
Fig. 7.22 Rhombic aerial. wire, the tilt angle and the angle of elevation. The higher the gain
that is required the- smaller must be the angle of elevation but,
fortunately, this is a favourable situation for long-distance routes.
Typically, a rhombic aerial will operate over a 2:1 frequency ratio,
e.g. 7-14 MHz, with a gain that varies from about 3 dB to about
15 dB. Since the h.f. band covers a frequency ratio of about 4:1, three,
or four, rhombic aerials would be needed to give complete coverage.
Although the rhombic aerial was very widely employed in the past
it has now been superseded for most applications by the log-periodic
aerial. The log-periodic h.f. aerial can provide a wider bandwidth
than can the rhombic aerial and it is physically smaller.

The Log-periodic Aerial The log-periodic aerial (1.p.a.) is a type of aerial whose radiation
pattern and gain change very little over a wide frequency band. The
bandwidth of the aerial is restricted only by the physical size of the
elements at the low-frequency end of the aerial, and by the accuracy
of the construction at the upper-frequency end. Several different forms
of l.p.a. exist but perhaps the most common consists of a tapered
dipole array. Figure 7.23 shows the basic form of a dipole l.p.a.;
moving along the aerial from the feed point both the length of each
 152 RADIO SYSTEMS TECHNOLOGY

Fig. 7.23 Basic log-periodic aerial.

dipole, and the spacing between adjacent dipoles, increase by a

constant ratio. This ratio, 7, is known as the scale factor.

% 1 L ee (7.13)

This means, of course, that ) = 1,7, 1; = br = 1,77, and so on up

to 1, = 1,/(r"—'). The aerial is fed, not at the origin, but at a point
distance d, = d,/r to the left of the first dipole. The distance dy from
the origin to the first dipole is then equal to d,/(1 — 1/r). The
characteristic angle a is, from Fig. 7.24, obtained from

l l
tana =— = a etc.
dy dtd
The space factor o is the name given to tan w.
At any frequency within the operating bandwidth of the aerial only
three, or perhaps four, of the dipoles are at, or near, the
resonant
length of \/2. Only these dipoles will radiate appreciable power
and
—*) they are said to be in the active region. The active region forms
— a
———. 9 ————> =< d, ——_> radiation centre whose dimensions, in wavelengths, are
both constant
and independent of frequency. All other dipoles, which are either
Fig. 7.24 much
longer or much shorter than \/2 will radiate little, if any,
energy.
 AERIALS 153

As the frequency of the current supplied to the l.p.a. is varied the

position of the active region will move back and forth along the array.
The result is that both the gain and the radiation pattern of the I.p.a.
remain approximately constant with frequency.
The highest, and the lowest, operating frequencies coincide with
those wavelengths which make the shortest, and the longest, dipoles
)/2 in length. The difference in length may easily be as much as 10:1
and the ratio gives the corresponding aerial bandwidth. Thus the
maximum frequency is fax = C/Amin = C/41,, the minimum
frequency is fin = C/Amax = C/41, = c/(,/4r"—'). Hence

oe 2k 1

Sein nes

The practical bandwidth obtained is somewhat less than this ratio

would predict because otherwise the active region would sometimes
move right off the array.
The excitation of the dipoles in the l.p.a. is always achieved by
feeding the current in at the smaller-dimensioned end, and transposing
the dipoles along the line of the array. This is necessary to ensure
an end-fire radiation pattern in the direction of the origin. Figure 7.25
shows typical radiation patterns for both the horizontal and the vertical
planes at two different frequencies. The radiation pattern has a wide
main lobe and it is clearly not very directive (less so than a rhombic
aerial), but it does have small side-lobes.
A horizontally polarized 1.p.a. may be used in various configura-
tions, either supported from a tower or suspended from masts. Figure
7.26. shows two practical high-frequency log-periodic aerials; the
vertically polarized version in (a) has a very low angle of elevation
and this makes it suitable only for long-distance radio links. The aerial
gain is reduced because the ground in the vicinity of the aerial is not
of zero resistance. Because of their larger gain most h.f. 1.p.a.s are
horizontally polarized. The aerial cannot be mounted at a fixed
physical height above the ground because its electrical height would

aoe ee

(a)

C SJ ~ 0° safe)
Fig. 7.25. Log-periodic aerial
radiation pattern: (a) horizontal
plane, and (b) vertical plane. ()
 154 RADIO SYSTEMS TECHNOLOGY

Post

Conductor

Supporting stay

Supporting glass fibre/polyester rope

Supporting rope Dipoles

= SS 3322
ere ————
SS
Insulator

Rigid separators

Steel lattice
mast

Mounting post Rope anchor

Fig. 7.26 (a) Vertically polarized (b)

|.p.a., (6) Horizontally polarized |.p.a.

then vary with frequency. This is why horizontally polarized 1.p.a.s

slope relative to the ground to ensure that each dipole is at the same
electrical height above earth. Figure 7.26(b) shows a typical example.
Table 7.5 Typically, an h.f. l-p.a. can operate over most of the h.f. band with
a gain of between 10 dBi and 15 dBi, depending on the number of
Bandwidth 4 to 28 MHz dipoles. Table eS gives two sets of typical data for L.p.a.s.
Log-periodic aerials are widely employed for various services in
Number of dipoles 17 46 the h.f. band. They are particularly useful when a wide horizontal
a 18° 10° plane beamwidth is wanted, when a high elevation angle is needed,
T 0.85 0.95 or when site area is a problem. An increased gain can be obtained
Angle of elevation 2a Seon if two l.p.a.s are used in a broadside array. If a number of areas at
Beamwidth, vertical ae ae different azimuth angles are to be served a rotatable l.p.a. is often
Gain (dBi) horizontal S ee used. The log-periodic aerial is also often used in both the v.h.f. and
the u.h.f. bands and Fig. 7.27 shows a typical example.
 AERIALS 155

Fig. 7.27 v.h.f./u.h.f. log-periodic

aerial.

The Yagi Aerial The equatorial and meridian plane radiation patterns of a \/2 dipole,
shown in Figs 6.12(a) and (b), are not directive enough for many
applications. An increase in both the gain and the directivity can be
obtained by the addition of one, or more, parasitic elements. A
parasitic element is one that is not directly supplied with an exciting
current and that is coupled by mutual impedance to the driven )/2
dipole. If the parasitic element is longer than )/2 and it is mounted
behind the dipole, relative to the required direction of maximum radia-
tion, it is known as a reflector. Conversely, if the parasitic element
is shorter than )/2 and it is mounted in front of the dipole it is known
as a director. A Yagi aerial consists of a \/2 dipole, a reflector, and
one, or more, directors. Figures 7.28(a) and (b) show, respectively,
a /2 dipole with one reflector, or with one director. A parasitic
element can be directly fixed onto a common metal support since
its mid-point is a voltage node.
The spacing between the )/2 dipole and each of the parasitic
elements is small enough for there to be mutual impedances between
them. When a current is supplied to the dipole and causes it to radiate
energy an e.m.f. will be induced into each parasitic element. This
e.m.f. will make a current flow in the element so that the parasitic
element accepts power from the dipole and then re-radiates it. The
phase relationships between the energy radiated by the 4/2 dipole and
the energy radiated by each parasitic element depends upon both the
element spacing and the phase of the current in each element. In turn,
the phase of the current in an element is determined by the electrical
length of that element. The element spacings and lengths are chosen
to give the maximum radiation in the wanted direction and minimum
radiation in all other directions.
A director increases the radiated field on its side of the dipole while
a reflector concentrates the radiated field in the opposite side of the
dipole. Since a reflector is longer than )/2 in length it has an inductive
reactance, while, conversely, a director has a capacitive reactance.
Values of reactance for different lengths of element are shown by
156 RADIO SYSTEMS TECHNOLOGY

Reflector

Director

Insulation

Coaxial
feeder

——
Direction of
transmission
Fig. 7.28 (a) Dipole with reflector.
(b) Dipole with one director.

Fig. 7.29; note that when the length is \/2 the reactance of the dipole
is +j43 Q.
100 Capacitive
Consider a X/2 dipole and a reflector. Zero voltage is applied to
Inductive the reflector and hence equation (7.11) becomes
Vp = IpZpp ae IpZpr (7.14)

0 aS IpZpr 4° IpZpr- (7.15)

Input
(Q)
reactance
From equation (7.15), Jp = —IpZppr/Zpp, and hence
Z2

Vp = Ip Zon 7 |
RR
0.2 0.22 0.24 0.26 0.27 0.28 Z2

Electrical length (\)

OF aZpitet Zenon eee (7.16)
ZRR
Fig. 7.29 Reactance/electrical- If typical values for Zpp, Zpp and Zpp are put into equation (7.16)
length curves for a dipole. it will be found that the input impedance of the dipole has been
considerably reduced. This reduction, which is accentuated if one,
or more, directors are added, may lead to difficulties in matching the
aerial to the coaxial feeder. To overcome this difficulty a folded dipole
is often used (Fig. 7.30). This has an impedance of about 300 Q.

Example 7.11

An aerial consists of a \/2 dipole of 73 + j0 Q impedance and a parasitic

element whose impedance is 73 + j68 Q. If the mutual impedance between
the dipole and the parasitic element is 65 + jO Q, calculate (a) the input
impedance and (b) the front-to-back ratio of the aerial. The dipole—reflector
spacing is 0.13).
 AERIALS 157

Solution
(a) From equation (7.16)
65°
Lpiarls eae = 42 +4290. (Ans.)
73 + j68

—IpZ,
Os ieee 1
RR

2 and therefore
Tp —65
SS Se ee ator
Ip 73 + j68
Coaxinl In the direction from reflector to dipole there is a phase difference of
feeder 137° — 360° x 0.13 = 90° and hence the field strength is propor-
tional to ae + 0.65") = 1.194. In the opposite direction, i.e. from
dipole to reflector, the phase difference is 137° + 360° x 0.13 =
180°. Hence, the field strength in this direction is proportional to
Y 1 — 0.65 = 0.35. Therefore
1.194
Fig. 7.30 Folded dipole. the front-to-back ratio = 20 logo be = 10.66 dB. (Ans.)

The addition of more directors will increase the gain of the Yagi
aerial, and Fig. 7.31 shows the relationship between the number of
directors and the gain. The element spacing is most critical for the
dipole and the reflector, and for the dipole and the first director; the
former should normally be somewhere in the range 0.17) to 0.22.

= rs

— ie)

o — (=)
dipole)
2

relative
t(dB
gain
Aerial

: 0 2 4 6 8 10 12 14 16 18 20
4 . F Number of directors
Fig. 7.31 Showing the relationship
between the gain of a Yagi array and
the number of directors. The director spacings are usually chosen to be in the range 0.15)
to 0.4, however a large number of other combinations of lengths/
spacings are equally likely to be used.

The Parabolic Dish Aerial For point-to-point radio links in the upper part of the u.h.f. band and
in the s.h.f. band the usual aerial employed is the parabolic dish. The
158 RADIO SYSTEMS TECHNOLOGY

|
pt ee a ee =
Parabolic \ exe
reflector “a Plane
\ wavefront

ieee

Fig. 7.32 Principle of a parabolic

dish aerial.
|

parabolic dish aerial is essentially a large metal dish that is illuminated

by a source of spherical wavefront, and which converts the incident
radiation into a wave having a plane wavefront. It is necessary to
control the amplitude of the dish illumination from its centre to its
edge both to maximize the aerial gain and to minimize losses. The
illumination efficiency y is 100 times the ratio of the on-axis direc-
tivity for a given illumination to the directivity produced by a uniform
illumination with the same total radiated power. This means that a
uniformly illuminated dish has an illumination efficiency of 100%.
The basic concept of a dish aerial is illustrated by Fig. 7.32. The
focal point of the dish is the point at which incident parallel waves
which are reflected by the dish converge. The feed point is placed
at the focal point of the dish; energy is radiated from the feed point
with a spherical wavefront and is directed on to the dish. The geometry
of the dish is such that the distance from the focal point to the dish
and then to an arbitrary plane X on the other side of the focal point
is a constant, regardless of which point on the surface of the
dish
is considered. Thus, the distance RAX = distance RBX = RCX =
RDX = RFX. This ensures that the spherical wavefront produced
by the feed is converted into a plane wavefront at the plane X.
The
reflected waves are all parallel with one another and form
a highly
directive radio wave.
The effective aperture A, of the dish is, for an illumin
ation
efficiency of 100%, equal to the geometric aperture, i.e.
D\2

et
 AERIALS 159

where D is the diameter of the dish. The gain G of the aerial, with
respect to an isotropic radiator is, from equation (6.24),
a (D/2)? 1 *D?
G= re = x2 (7.17)

or, relative to a \/2 dipole

D/2)? 2 6(>)
riee D\? . (7.18)
Clearly, for a dish aerial to have a high gain its diameter must be
several times greater than the signal wavelength. This is the reason
why the dish aerial is only employed in the u.h.f. and s.h.f. bands.
The radiation pattern of a dish aerial is highly directive with a very
narrow main lobe in the direction of the dish axis. Consequently, the
usual way of drawing the radiation pattern gives insufficient detail.
Normally, therefore, the radiation pattern is drawn using Cartesian
co-ordinates as shown by Fig. 7.33.

0.75

0.5

0.25

relative
maximum
strength
Field
to

0 20° 10° 0 10° 20°

Fig. 7.33 Radiation pattern of a Angular distance from direction of
parabolic dish aerial. maximum radiation

The beamwidth of a dish aerial is approximately given by equation

(7.19), 1.6.
70°
beamwidth = ape (7.19)

Side Lobes

The radiated power contained in the side lobes of the aerial’s radia-
tion pattern is power that is radiated in unwanted directions where
it may well interfere with other systems. The net efficiency of a dish
aerial can be increased if the total energy in the side lobes is mini-
mized. For the ground aerial of a communications satellite system
160 RADIO SYSTEMS TECHNOLOGY

fo)o the side-lobe performance, in the direction of the geometric arc, is

specified by such bodies as CCIR, INTELSAT and EUTELSAT.
a
Ninety per cent, or more, of the side lobes are required to lie under
the envelope defined by 29 — 25 logio 0, where @ is the side-lobe
angle in degrees. This relationship gives the signal-to-noise ratio in
decibels for various satellite spacings shown by the graph of Fig. 7.34.
29
—25 log,6
The design of a ground-station aerial must reduce the side-lobe level
85 to the defined figure.:
There are a number of sources of side-lobe energy produced by
Signal-to-interference
ratio
(dB)
a dish aerial. These are:
: age Pa aa (a) feed spillover due to direct radiation from the feed, and/or
Satellite spacing (degrees) 3 i 4 ; ee
(for a Cassegrain or Gregorian aerial) direct radiation from the
Fig. 7.34 Side-lobe specification for
sub-reflector, spilling past the edge of the main reflector;
(b)
a ground-station aerial. blockage in which the feed point, or the sub-reflector, blocks
some of the energy reflected by the main reflector, thus
producing a hole in the energy distribution of the aerial’s
aperture; and
(c) reflector illumination, where, if the main dish is uniformly
illuminated, relatively large side lobes are produced, particu-
larly up to about 10° from the aerial’s axis.
The side-lobe amplitude can be considerably reduced by tapering
the amplitude distribution across the dish. If the amplitude distribu-
tion is varied from a maximum at the centre, to zero at the edge of
the dish the side-lobe level will be reduced. This reduction will,
however, be paid for by a reduction in the illumination efficiency.
The inverse distribution, i.e. varying illumination from a maximum
at the edge of the dish to zero at the centre, is not employed since
it both reduces the gain and increases the side-lobe level.

Example 7.12
Calculate the gain in dBi of a 5 m diameter dish aerial at (a) 4 GHz,
and
(b) 10 GHz if the illumination efficiency is 60%. (c) The gain of an
INTELSAT standard dish aerial is 60 dBi; calculate its diameter at
11 GHz
if the illumination efficiency is 60%.

Solution

3 x 108
(a) \ = ———_ = 0.075 m.
4 x 10
aa ee
The gain = 10 logy, jo.o(--*) = 44.2 dBi. (Ans.)
0.075

5a \2
Gain = 10 lo zw |
0.6| eal
—— = 52.2 dBi. Ans.
(Ans.)
 AERIALS 161

(c) \ = 0.027 m.
4 3 mD |?
60 dBi = -10° = 0:6) ———
0.027
or D = V123 = 11.1 m. (Ans.)

Feed Arrangements

The simplest form of primary feed is a dipole fed by a coaxial line

and, usually, with a parasitic element, or a plane reflector, to ensure
that the radiation is directed onto the dish. At frequencies greater
than a few gigahertz a waveguide feeder will be employed to connect
the aerial to the transmitter (or receiver). Then, it is more usual to
employ some kind of waveguide feed and one commonly employed
example is the horn radiator. The horn radiator is just a length of
waveguide whose open end is flared out to form a horn.
There are four main ways in which the main reflecting dish can
be illuminated by the feed: these are known, respectively, as front-
feed (or focus), Cassegrain, Gregorian and offset-feed. The basic
arrangement of each of these methods is shown in Fig. 7.35.

Front feed
Front feeding a dish aerial, shown by Fig. 7.35(a), is the simplest
method of illuminating the main reflector. To reduce the side-lobes
to an acceptable level the illumination efficiency is only about 55 to
60%. The feed and its supporting structure produces aperture blockage
and this increases the side-lobe level. There may also be some
re-radiation from the struts, which will degrade the cross-polarization +
of the aerial. A horn front feed can give a good gain and an acceptable
noise temperature and it is often used with small aerials.

Cassegrain

The Cassegrain method of feeding a dish aerial, shown in Fig. 7.35(b),

uses a sub-reflector to reflect the energy radiated by the primary feed
onto the main reflector. The system has the advantages over the front-
feed method in that the feed system is simpler and a shorter length
of waveguide feeder is necessary. Although aperture blockage and
strut re-radiation are still a problem the Cassegrain aerial is commonly
employed.

+Communications satellite systems often employ identical frequency bands to carry

different signals. Two signals are transmitted at the same frequency but one is horizon-
tally polarized and the other is vertically polarized. The cross-polarization isolation
is a measure of the degree of crosstalk immunity between the two signals.
 162 RADIO SYSTEMS TECHNOLOGY

Hyperbolic
sub-reflector

(2) (6)

\
Elliptical : ”
sub-reflector Si Ce Ni
Na eek
oN x x
SiMe cae
We

Fig. 7.35 Parabolic dish aerials: Offset

(c) sub-reflector
(a) front-feed; (b) Cassegrain;
(c) Gregorian; and (d) offset-fed.
(d)

Gregorian
The Gregorian aerial, Fig. 7.35(c), is a variant of the Cassegrain aerial
in which an elliptical sub-reflector is employed. This gives an
increased illumination efficiency of about 76% and also improved
cross-polarization isolation.

Offset Feed
Blockage of the transmitted beam may be avoided if the feed, or
the
sub-reflector, is mounted outside the area occupied by the main beam.
Figure 7.35(d) shows an offset Cassegrain dish aerial. A front-fe
ed,
or a Gregorian aerial can be similarly offset. The feed, and the
sub-
reflector, are mounted at the appropriate angles to the aerial axis
to
ensure that the main beam is directed along the axis. The overall
size
of the offset aerial is smaller, for a given gain at a given frequen
cy,
and the side lobes are smaller, but de-polarization effects are increase
d.
This type of aerial is becoming increasingly popular.
 Propagation of Radio
Waves

The radio-frequency energy radiated by a transmitting aerial in the

direction(s) determined by its radiation pattern will travel through the
atmosphere and arrive at the distant aerial by one, or more, different
modes of propagation. These are the ground wave, the sky wave, the
space wave, tropospheric scatter, and via a communications satellite.
When the transmit aerial is mounted on the surface of the earth
propagation will mainly be by means of the ground or surface wave.
There will also be some sky-wave radiation which may, or may not,
be returned to earth. The ground wave is used in the v.1.f. and Lf.
bands for long-distance, narrow-bandwidth communications, and in
the m.f. band for shorter-distance sound-broadcast signals. In the v.1.f.
band the received signal at long distances from the transmitter is
extremely stable and shows very little change over a day, or even
seasonally. The main disadvantage of the v.1.f. band is its very limited
bandwidth and the large physical structures that are necessary for the
transmitting aerials. The range of the ground wave is not quite as
large for signals in the 1.f. and the m.f. bands since the attenuation
suffered increases with frequency; the daytime range is typically a
few hundred kilometres at the low end of the m.f. band to about
100 km at the upper end of the m.f. band. The night-time range may
well be some thousands of kilometres because propagation can then
take place via the ionosphere.
At higher frequencies the transmit aerial can be mounted above the
earth, perhaps by several wavelengths, and then propagation is mainly
by either the sky wave or the space wave. The sky wave is mainly
employed for long-distance telephony links in the h.f. band but also
for some sound-broadcast services. The space wave is employed for
various services in the v.h.f., u.h.f. and s.h.f. bands, such as point-to-
point multi-channel telephony systems and land-, sea- and air-mobile
systems. The u.h.f. band is also utilized for television-broadcast
signals.
The scope of tropospheric scatter systems is much more limited
and it finds application for some beyond-the-horizon radio systems
where considerable terrain difficulties exist. Finally, communications
satellite systems are now employed to carry wideband telephony and
television systems over long distances.

The lonosphere Ultraviolet, and other, radiation from the sun ionizes large numbers
of the atoms which make up gas molecules in the upper atmosphere.
 164 RADIO SYSTEMS TECHNOLOGY

This ionization causes the earth to be surrounded by a belt of ionized

gases which is known as the ionosphere. The height of the ionosphere
above the surface of the earth varies considerably but it is normally
within the limits of 50 km to 400 km. Within the ionosphere the
density of the free electrons is much higher than it is at heights either
above or below the ionosphere. Below the ionosphere there are two
other regions: the troposphere, from ground level to about 10 km
above the ground; and the stratosphere, from about 10 km to the lower
edge of the ionosphere.
The electron density, measured by the number of free electrons per
cubic metre, within the ionosphere is not a constant quantity. During
the daytime the ionosphere contains four distinct regions, or layers,
within which the electron density exhibits a maximum value. These
four layers are known as the D, E, F, and F, layers, in order of their
increasing height above ground, and their increasing electron density.
There is also a fifth layer, labelled C, below the D layer, but it plays
no part in radio-communications. The heights of the four layers above
the ground vary daily, seasonally, and with the 11-year variations
in the 22-year sun-spot cycle. Approximate heights, for both
daytime and night-time, are shown by Figs 8.1(a) and (b), respec-
tively. In the daytime the D layer has a much lower electron density
than do the other layers, and during the night it disappears completely.
The E layer becomes weaker at night but it does not normally
disappear. In the daytime the F, layer is at a height of about 200 km
but the height of the F, layer varies considerably. Typically, the F,
layer is at about 250 km to 350 km in the winter, and between 300 km
and 400 km in the summer.

Refractive Index of the Ionosphere

As a radio wave propagates through the ionosphere its electric field

will exert a force upon the free electrons which causes them to move.
If the free electron density is N electrons per m?, each free electron
has a charge of e coulombs, and the field strength of the radio wave

€ =
~ 400 = 400
2
no}
no}
FE
2 300 5 300
cS) is
2 200 © 200
se)
Ee
2 fe)
oO 2
= 100 ~ 100
oO

ong
a
Bo) [o.)
o ra
x Na
10° 10° 10° 10” 190° 10°ND MOM EIO 16
Fig. 8.1 lonospheric layers (a) in Electron density (electrons/m*) Electron density (electrons/m*)
the daytime, and (b) at night.
(a) (b)
 PROPAGATION OF RADIO WAVES 165

is E sin wt radians, then the force exerted upon each free electron
is F = eE sin wt newtons. Each electron is given an acceleration of
F eE sin wt
,
m m

where m is the mass of an electron.

The consequent velocity v of an electron is obtained by integrating
its acceleration with respect to time; hence
v = (—eE cos wt)/wm m/s.

The current represented by this movement of electrons is

i = Ne x (—eE cos wt)/wm = (—Ne?E cos wt)/wm A/m?.

The current that flows in a dielectric of unity relative permittivity is
i = €) dE/dt = eqwE cos wt A/m?
and so the total current flow is
] Ne’E cos wt
PPEeEicos oh =
wm

Ne?
seg) Mirae wE cos wt
@ “MEG

From this, the relative permittivity e, of an ionospheric layer is

Ne?
Cg dr aa _
@ “ME

Substituting in the values for the charge and the mass of an electron
gives
81N
é=1-
anes
This means that the relative permittivity of an ionospheric layer is
smaller than that of free space.
The refractive index n of a layer is equal to the square root of its
relative permittivity, and hence
81N
Refractive index n = .(:_ oe: (8.1)
2
f
Also, of course, the refractive index is equal to the ratio (sine of angle
of incidence)/(sine of angle of refraction).

Behaviour of the Ionosphere at Different Frequencies

The effect of the ionosphere on an incident radio wave is very much
a function of frequency. In the v.1.f. and 1.f. bands the ionosphere
has a high conductivity and it reflects, with very little loss, any wave
that is incident upon its lower edge. In the m.f. band the D layer acts
 166 RADIO SYSTEMS TECHNOLOGY

like a lossy medium whose attenuation reaches its maximum value

at the gyro-frequency of 1.4 MHz. Signals in the m.f. band are
absorbed by the D layer and are not returned to earth. At frequencies
in the h.f. band the E and F layers refract radio waves and, if condi-
tions are correct, return the wave to earth. The D layer has little,
if any, refractive effect. Signals in the v.h.f., u.h.f. and s.h.f. bands
normally pass straight through the ionosphere.

Ionospheric Variations

The intensity of the ultraviolet, and other, radiation from the sun that
enters the earth’s atmosphere is continually fluctuating. Both regular
and irregular variations occur. The regular variations occur for two
reasons: (a) the intensity of the sun’s radiation varies with both the
time of the day and the month in the year; and (b) sun spots occur
at the sun’s surface which produce 11-year fluctuations in its radia-
tion that affect the ionosphere. The sun-spot cycle has an average
periodicity of twenty-two years.
Irregular ionospheric disturbances are also experienced. Solar flares
emit large amounts of radiation from the sun that produce a large
increase in the ionization of the D layer. This may cause the D layer
to absorb all h.f. signals, giving a complete blackout for anything
up to about two hours. Sometimes, ionospheric storms occur; this
is the name given to irregular fluctuations in the conductivity of the
ionosphere which cause rapid fading, particularly at the higher
frequencies. Ionospheric storms tend to occur at intervals of 27 days.
Sporadic E consists of a cloud of drifting electrons which suddenly,
and unpredictably, appear within the E layer. The electron cloud has
a much higher electron density than is usual and it is therefore able
to return to the earth waves that normally pass Straight through the
ionosphere. When sporadic E is present, the m.u.f. of a returned
wave
may be, typically, some 20—40 MHz, although even higher
figures
sometimes occur. Sporadic E is more likely to occur in the
summer
than in the winter.

The Troposphere The CCIR have adopted the following expression for the
refractive
index n of the troposphere.

pre aeulds
T § Pp ja 28g
7 x10 : 8.2
(8.2)
where T is the temperature in K, and e and p are
the water-vapour
and atmospheric pressures in millibars. Since all three
parameters are
functions of height this expression can be written in
the form
n(h) = 1+ ae~™,
(8.3)
where a and b are constants, e is the base
of natural logarithms,
 PROPAGATION OF RADIO WAVES 167

i.e. e = 2.7183, and h is the height above the ground in kilometres.

If a = 315 x 10~° and b = 0.136 (which refer to the CCIR
average atmosphere), the CCIR reference refractive index is obtained.
The refractive index, calculated from equation (8.3) gives most
inconvenient numbers, e.g. ath = 1 km, n = 1.000 27, and ath =
2km, n = 1.000 24. To obtain more convenient numbers it is
customary to use the refractivity of the troposphere instead. This is
merely equal to (n — 1) x 10°; thus for h = 1 km the refractivity
is equal to 270. The variation of the refractivity with increase in height
above the ground is shown by Fig. 8.2.
Since the refractive index falls with increase in height the radio wave
will follow a curved path through the troposphere. Provided the wave
is launched in a horizontal plane the radius of curvature of the path
is —dh/dn. In the British Isles, at a height of about 1 km, n is
normally about 40 parts in 10° lower than at ground level (see Fig.
8.2); hence —dh/dn = 2.5 x 10’ m. The radius of the earth is
6400 km and so the radius of curvature of the radio path is 25000/
6400 = 3.9 times the radius of the earth. When a radio link is planned
it is convenient to consider that the radius of the earth is k times its
actual value and that the space wave travels in a straight line (see
Fig. 8.3). Under normal atmospheric conditions k = 4/3, giving an
effective earth’s radius of (4 x 6400)/3 = 8500 km.
If the distance between the two aerials is not too great the earth
between them may be assumed to be flat. Little error is then introduced
into any calculations and, if required, a correction factor can be used.
Departures from normal atmospheric conditions are not usually large
enough to give significant fading on a line-of-sight radio path. |
Sometimes, however, abnormal conditions do arise that noticeably
affect space-wave propagation.

(km)
ground
above
Height

40 80 120 160 200 240 280 320 360

Fig. 8.2 Variation of refractivity with
height. Refractivity
 168 RADIO SYSTEMS TECHNOLOGY

Curved path Straight path

Earth’s radius Earth’s radius

6400 km k x 6400 km
Fig. 8.3 Space-wave propagation. (a) (6)

Temperature Inversion
In the early part of a sunny day the cloudless skies may result in the
air temperature being higher than the ground temperature, contrary
to the normal state of affairs. This is known as a temperature inversion.
Temperature inversions also occur because of the following.
(a) Subsidence: a mass of warm air may be further heated by
compression and then rises to a greater height, while cooler
air falls to a lower height to replace it.
(b) Dynamic: a mass of warm air may move on top of a mass of
cold air.
(c) Nocturnal: the air at the surface of the earth is rapidly cooled
after sunset.
(d) Cloud layer: the sun’s rays may be reflected from the upper
surface of a cloud and heat up the air above the cloud.

Water Vapour
For normal atmospheric conditions the humidity of the atmosphere
Standard refraction
falls gradually with increase in height, but sometimes an abrupt change
Super-refraction in the humidity gradient may occur. This is most likely to happen
above the sea during hot weather and it often occurs to the leeward
Earth’s radius = kR of land as warm air moves from the land out over the sea.
(a)

Super-refraction
{ Duct
Earth’s radius = kR A temperature inversion and/or a non-standard water-vapour gradient
(b) will produce non-standard refraction of the space wave. This
Sub-refraction
may
be either super-refraction or sub-refraction. When the refractiv
e index
Standard refraction of the troposphere decreases with height more rapidly than usual
the
wave is bent towards the earth to a greater extent than
normal and
a duct is formed. Super-refraction is shown by Fig. 8.4(a).
Earth’s radius = kR The wave
(c) may be reflected from the earth, again super-refracted,
again reflected
and so on to produce the ducting shown in Fig. 8.4(6). When
a duct
Fig. 8.4 (a) Super-refraction, is present the space wave may propagate for distances well
beyond
(6) ducting, and (c) sub-refraction the radio horizon. Sometimes an elevated duct may appear
at a height
of a space wave. of a kilometre or so above the ground.
 PROPAGATION OF RADIO WAVES 169

Sub-refraction

Sub-refraction of a space wave occurs when the refractive index of

the troposphere decreases with height at a smaller rate than is usual.
The space wave is then refracted to a lesser extent than normal, and
if the k factor of 4/3 has been employed in the design of the link,
the wave will appear to be bent upwards as shown by Fig. 8.4(c).

Ground-wave Propagation At very low, and low frequencies the transmitting aerial is an electri-
cally short monopole which radiates energy in the form of a ground,
or surface, wave. The ground wave is vertically polarized and it is
able to follow the undulations of the surface of the earth because of
diffraction.+ As the wave travels, its magnetic field cuts the earth
and induces e.m.f.s into it. In turn, these induced e.m.f.s cause
currents to flow in the resistance of the earth and so dissipate power.
Wavefront (tilted forward) This power can only be supplied by the radio wave and so there is
a continuous flow of energy from the wave into the ground. This
Forward component results in the wavefront having two components of velocity, one in
the forward direction and one downwards normal to the earth. This
Resultant
Downward is shown by Fig. 8.5; the resultant velocity of the wave is the phasor
component sum of the two components and this makes the wavefront tilt forwards.
Since the downward component of velocity is always normal to the
stotP25
eee eceseetetecs surface of the earth the wave is able to follow the undulations of the
ground.
Fig. 8.5 Ground-wave propagation. The wave is attenuated, for two reasons, as it travels. First, the
wavefront diverges as it travels so that the field strength is inversely
proportional to distance and, second, power is taken from the wave
to supply the ground losses. The calculation of the ground power losses
is complex and it depends upon such factors as the frequency of the
wave, and the conductivity and permittivity of the earth. The attenua-
tion is expressed by an attenuation factor K whose value can be
approximately predicted from published graphs. Thus, the electric
field strength Ep at a distance D kilometres from the transmitting
aerial is given by
KE
Ep = i V/m, (8.4)

where £, is the electric field strength 1 km from the transmitting

aerial. E, is equal to 300VP, (p. 118), where P, is the transmitted
power in kilowatts. Hence, if the distance D is expressed in kilometres
K 300VP,
= Te Ss M V/m. (8.5)

+Diffraction is a phenomenon which occurs with all wave motion. It causes a radio
wave to bend around any obstacle it passes. For the ground wave the earth itself is
the obstacle.
 170 RADIO SYSTEMS TECHNOLOGY

At low, and medium frequencies the attenuation factor K is inversely

proportional to the square of the frequency. Hence the attenuation
of the wave increases rapidly with increase in frequency. At high
frequencies and above the (frequency) factor reduces the amplitude
of the ground wave to a negligible value even over quite small
distances.

Example 8.1

A radio transmitter is to produce, using the ground wave, a field strength

of 5 mV/m at a distance of 200 km. If the attenuation factor is 0.2 calculate
the necessary radiated power. What would be the field strength if the frequency
of the transmitted signal were to be doubled?

Solution
res 0.2 x 300VP,
200
or P, = 278 kW. (Ans.)

If the frequency is doubled

Ex) = 1.25 mV/m. (Ans.)

At frequencies in the m.f. band the maximum field strength at

ground level is obtained when the height of the transmitting aerial
is 5/8. There is then appreciable radiation into the sky (p. 119).
During the daytime, the sky wave is completely absorbed by the D
layer and does not return to the earth. During the night, however,
the D layer has disappeared and then the sky wave will be returned
to earth via either the E or the F layer. The two cases of interest are
illustrated by Figs 8.6(a) and (b). In the first case, a transmitted signal
is received at a distant location by means of both the ground wave
and the sky wave. The total field strength at this point is the phasor
sum of the individual field strengths and this will vary because of
fluctuations in the length of the ionospheric path. This means that
the received signal will be prone to fading. In the second case, two
different signals, radiated at the same frequency by different trans-
mitters, are received together at a point. The quality of the received
wanted signal will be impaired and its reception may be so poor that

lonosphere lonosphere

Interfering
sky waves

Ground wave

Fig. 8.6 Medium-frequency

transmissions at night.
 PROPAGATION OF RADIO WAVES 171

it is unusable. This is, of course, the effect that makes the night-time
reception of medium-wave broadcast signals in Europe of such poor
quality.

Sky-wave Propagation The basic principle of a radio link operating in the h.f. band is illus-
trated by Fig. 8.7. The sky wave is directed into the ionosphere where
lonosphere it is continuously refracted. If, before it reaches the top of the F,
layer, it has been refracted to the extent that the angle of refraction
is 90° then the wave will be returned to earth. The intrinsic instability
of the ionosphere causes the length of a sky-wave path to vary continu-
ously in a random manner, and considerable fading may take place.
Particularly difficult in this respect are the periods around dawn and
Earth around dusk, when the electron densities of the ionosphere change
more rapidly than at other times.
The potential unreliability of an h.f. radio link has meant that h.f.
Fig. 8.7 Sky-wave propagation. radio has, in the past, lost considerable ground to both communications
satellite and terrestrial radio-relay systems. Nowadays, however, the
relative cheapness of h.f. radio systems plus the introduction of various
technical innovations, such as diversity, frequency synthesis, and new
modulation techniques, have revived interest in h.f. technology.
Congestion in the h.f. band has been partially alleviated by an
increased use of s.s.b./i.s.b. transmissions.
A radio wave that enters the E layer with an angle of incidence
¢; will continuously be refracted away from the normal. If the values
of the electron density N and the frequency f are such that sin ¢; =
Jd — 81N/f’) then, since the refractive index n = (sin ¢,)/(sin ¢,),
sin ¢, must be equal to unity. Then ¢, = 90° and the wave must then
be travelling in a horizontal direction. Any further refraction of the
wave will then return it back to earth. If no part of the E layer has
an electron density large enough for the sin ¢, = 1 relationship to
be satisfied, the sky wave will not be returned to earth but will escape
from the top of the layer. The wave will then be incident on the F,
(or, at night, the F) layer with an increased angle of incidence and
it will here be further refracted and may perhaps be returned to earth
by this layer. If not, the wave will leave the top of the F, layer and
pass on to the F, layer with an even larger angle of incidence and
now it may be returned to earth. If the wave is not returned by the
F, layer it will escape from the earth. The concept is illustrated by
Figs 8.8(a) and (b). In Fig. 8.8(a) a wave entering the E layer with
an angle of incidence ¢, is returned to earth but another signal, at
the same frequency, which is incident on the E layer with a smaller
angle of incidence ¢ , is not returned. This second wave travels on
to the F, layer and is from here returned to earth. Figure 8.8(0)
shows two waves of frequencies f, and f, incident upon the E layer
with the same angle of incidence $,, where f; > f|. The lower-
frequency wave is returned to earth by the F, layer but the
higher-frequency wave is not.
 172 RADIO SYSTEMS TECHNOLOGY

MLE:yy
Wa
MLE ee
Fig. 8.8 Showing the effect on a
sky wave of (a) the angle of
incidence, and (b) frequency.

Critical Frequency

The critical frequency {,,i: of an ionospheric layer is the highest

frequency that can be radiated upwards with an angle of incidence
lonospheric / aes ¢; of zero and be returned to earth. This will be the frequency of the
layer wave that travels up to the top of the layer, where the electron density
is at its highest value, before it is refracted to the extent that ¢,
becomes equal to 90°. From equation (8.1),
sin 0° = 0 = id 0 81 Ninax/f crit)

or Jo a NaN eae (8.6)

Each of the ionospheric layers will have its own value of critical
frequency. From the ground it appears as though the wave has
travelled in a straight line, has been reflected by the ionosphere, and
Earth has then returned to earth from the point of reflection in another
straight-line path. The virtual height of a layer is the height at which
Fig. 8.9 Virtual height of a layer. this apparent reflection takes place (see Fig. 8.9).

Maximum Usable Frequency

The maximum usable frequency (m.u.f.) of a layer is the highest

frequency that can be employed for a sky-wave path between two
points on the earth’s surface. For a wave to be returned to earth
sin ¢, = 1, and so

sin ¢, = Ja oa Sey,

1 — sin? 4; = ahi
2 .

Saas

or Smax ts
= m.u.f. =—e —“—
dextt Ay
= f,,. sec dj. (8.7)
cos ¢;
Since the electron density of each layer is subject to continuous
fluctuations, some regular and predictable and others not, the m.u.f.
 PROPAGATION OF RADIO WAVES 173

is not a constant figure. The m.u.f. of a sky-wave path between two

points on the earth’s surface will vary throughout the day, and graphs
of the forecast m.u.f. for various routes are commercially available.

Optimum Working Frequency

Because of the instability of the ionosphere, operating a radio link
at a frequency equal to the m.u.f. would not give a reliable system.
It is customary to work an h.f. route at a frequency of about 80%
of the m.u.f.; this lower frequency is known either as the optimum
working frequency or as the optimum traffic frequency.
Since the m.u.f. is not of constant value it will be necessary to be
able to change the frequency of a sky-wave link as, and when, the
propagation conditions alter. Usually, an h.f. radio transmitter is
allocated (although not exclusively) several different frequencies, and
any one of them may be in use at a given time. When the propagation
conditions are severe it may become necessary to use two, or more,
frequencies simultaneously and, perhaps, even retransmit when
conditions improve.

Example 8.2
The virtual height of a layer is 110 km and its critical frequency is 4 MHz.
Calculate the m.u.f. for two points on the surface of the earth that are 600 km
apart if (a) the earth is assumed to be flat, and (b) the radius of the earth
is 6400 km.

Solution
(a) From Fig. 8.10(a),

ae
@ = tan = 70°.aes
110
Therefore

the m.u.f. = 4 sec 70° = 11.7 MHz. (Ans.)

300
(a)

Fig. 8.10 (6)

174 RADIO SYSTEMS TECHNOLOGY

(b) From Fig. 8.10(6) and then Fig. 8.10(c), d = 300 = 64000. 0 =
300/6400 = 0.0469 radians = 2.69°.
x? = 64007 + 65107 — (2 x 6400 x 6510 x cos 2.69°),+
or x = 322 km. Therefore,
6400 5
ot
sin ¢; sin 2.69°

6400 sin 2.69°

sin ¢;
322
and ¢, = 68.9°, and

the m.u.f. = 4 sec 68.9° = 11.1 MHz. (Ans.)

Maximum Value of ¢;

The angle of incidence ¢$; with which a sky wave enters the
ionosphere cannot be increased without limit. The maximum possible
value $i(max) Occurs when the transmitted wave is tangential to the
earth’s surface, as shown by Fig. 8.11. Here R is the radius of the
earth, approximately 6400 km, and h is the virtual height of a layer.
From the figure,
R
ae
9} max): op sin
R+h : (8.8)

Also, 8 = 180° — (90° + imax) = 90° — sin~! [R/(R + h)] =

cos—! [R/(R + h)). Therefore, the maximum ground range is

2d = 2R0 = 2R cos! ( a ) (8.9)

R+h
The maximum distance using the E layer is about 4000 km; longer
distances are possible using the F layer(s). If a path length approach-
ing, or exceeding, the maximum range is wanted a multi-hop path
will be necessary.
Fig. 8.11 Maximum value of ¢;.

Example 8.3

At a particular time of day the E layer has a maximum electron density of

1.5 x 10'' electrons/m and it is at a virtual height of 150 km. Calculate
the m.u.f. and the maximum single-hop range.

Solution

fort = 9.5 x 10!) = 3.49 MHz.

aa sivas ie 7.7
IN Piqmax) = Sin Se] as
eo 6550
Sra ee ea a a ee Ee Sn ee a
yUsing the cosine rule.
+Using the sine rule.
 PROPAGATION OF RADIO WAVES 175

Therefore, the m.u.f. is 3.49 sec 77.7° = 16.4 MHz. (Ans.)

6400
The maximum ground range = 2 x 6400 x cos! es
6550
= 2745 km. (Ans. )

Skip Distance

There is a maximum usable frequency for any distance and this

distance is also the minimum distance at which that frequency can
be transmitted over the sky-wave path. This minimum distance is
called the skip distance.

Fading

PanoY. ony General fading, in which the complete signal fades to the same extent,
is produced by fluctuations in the ionospheric attenuation. Unless there
is a complete fade-out of the signal the effects of general fading can
be overcome by the use of a.g.c. in the radio receiver.
Selective fading occurs when the signal picked up by the receive
aerial has arrived via two, or more, different paths (see. Fig. 8.12).
The total field strength at the aerial is the phasor sum of the field
strengths produced by each signal. The phase difference between the
signals arriving via the two separate paths is equal to 2 7/) times the
difference between the lengths of the two paths. If this difference
lonosphere should vary, due to fluctuations in the ionosphere, the total field
strength will also vary in a frequency-dependent manner because of
the 1/X term. This means that the different frequency components of
ky wave a complex signal may fade to different extents.
Ground wave There are a number of ways in which selective fading may be
combatted. These include:
Bette TAOS NSS
EN ee
; Tee Earth “WEES
(a) the use of a highly directive transmitting aerial so that the
(0) number of possible propagation paths is minimized;
Fig. 8.12 Selective fading. (b) operation at a frequency as near to the m.u.f. as possible;
(c) the use of s.s.b./i.s.b. signals;
(d) Lincompex; and
(e) the use of space and/or frequency diversity.

Example 8.4
The signals received by an aerial arrive over two different paths, one of which
is 75 km longer than the other. At the carrier frequency the two signals cancel
out. If the carrier is amplitude modulated, at what side frequencies will
cancellation occur?

Solution
For the two signals to cancel 2 d/\ = nz, where d is the path-length difference
and n is an odd integer. Therefore
176 RADIO SYSTEMS TECHNOLOGY

2d
gens bina fe

nc
or 24 =n
= —
f
nx 3x 108
and f= 3 = 2n kHz.
2X
I5= x NO
Hence, the side-frequencies that cancel out are f. + 2000 Hz, fu +
6000 Hz, etc. (Ans.).

Frequency Diversity

Signals at different frequencies received by the same aerial very rarely

fade simultaneously. This fact is used as the basis of a frequency-
diversity system. A single aerial is connected to a number of radio
receivers, each of which is tuned to a different frequency, whose
outputs are commoned. The receiver circuitry is so arranged that the
receiver that is instantaneously receiving the strongest input signal
will provide the output signal. The obvious disadvantage of frequency
diversity is its use of two, or more, frequencies at the same time.

Space Diversity
Signals at the same frequency that are received by two aerials sited
several wavelengths apart rarely fade simultaneously. In a space-
diversity system two, or three (but rarely more), aerials are sited some
distance apart and are connected to two, or three, radio receivers.
Each receiver is tuned to the same frequency and has a commoned
output. As with frequency diversity the circuitry is arranged so that
the receiver that is receiving the strongest signal supplies the output.
The disadvantage of space diversity is the need for more than one
aerial and the large site area required.

Example 8.5
At the distant end of an 8 MHz sky-wave radio link signals are received via
two paths that are at angles of 12° and 24° to the ground. Calculate the
optimum distance between the aerials in a two-aerial space-diversity system.

Solution
Consider Fig. 8.13. The wavefront of signal 1 arrives first at aerial A and
then at aerial B, so that the signal at B lags the signal at A by angle
21d
l= te cos 12° radians.

Similarly, for signal 2 the phase difference between the signals at aerials A
and B is
21d :
2 = ae cos 24° radians.
 PROPAGATION OF RADIO WAVES 177

Wavefront Ss.
~
2 ~

Wavefront ; ~ ~ bare
1 & TORS
/
~

~~

: 8.13
Fig. a~— d ——+ e

For the optimum space diversity the difference between the two phase angles
should be equal to 7 radians. Therefore
21d
(cos 12° — cos 24°) = x

3 «40%
and d = ; = 290 m. (Ans.)
~ 8 x 10° X 2 (cos 12° — cos 24°)

Space-wave Propagation At frequencies in the v.h.f., u.h.f. and s.h.f. bands the main mode
of propagation between two points on the surface of the earth is the
space wave. Since the wavelength of the signal is small, both the
transmitting and the receiving aerials can be mounted at a height of
several wavelengths above ground. Figure 8.14 illustrates the prin-
ciple of space-wave propagation. A radio wave travelling in the
troposphere follows a slightly curved path because of tropospheric
refraction, and this results in the radio horizon being more distant
than the optical horizon. At distances less than the optical horizon
reception is by means both of a direct wave and of a ground-reflected
wave, but at greater distances, up to the radio horizon, only the direct
wave is received. Some signals are also received at distances greater
than the radio horizon because some diffraction takes place.
Figure 8.15 shows how, typically, the magnitude of the reflection
coefficient |o | of the earth may vary with the angle of incidence for
both horizontally polarized and vertically polarized waves. For a
horizontally polarized wave |p| is always equal to unity, but for a
vertically polarized wave |p| varies considerably with the angle of
incidence. The angle of the reflection coefficient 2 p is always 180°
for a horizontally polarized wave, but for a vertically polarized wave
Zp varies from about 180° to about 10° as the angle of incidence
is increased from zero. In practice, the angle of incidence is always

Direct wave

Slight curvature
; due to Radio horizon
Straight-line path
refraction
|<~Diffract
ion region

Optical
Reflected horizon Earth (radius = 6400 km)
Fig. 8.14 Space-wave propagation. wave
 178 RADIO SYSTEMS TECHNOLOGY

Horizontal polarization

0.8
Vertical polarization
0.6
{
0.4

0.2
Magnitude
reflection
of
coefficient

Fig. 8.15 Reflection coefficient of 1 2 3 4 5 6 it, 8 9 10

the earth. Angle of incidence (degrees)

small (and it is often called the grazing angle) and little error results
if Zp is assumed to be 180°.
The relative merits of using horizontal, or vertical, polarization are
as follows. The total field strength at the receive aerial is the resultant
of the field strengths due to several components, including some
diffracted energy. A wave that has been diffracted over a treeless hill
will suffer less attenuation if it is vertically polarized. Conversely,
if the hill is tree-covered a vertically polarized wave will be scattered
to a greater extent and will thus suffer the greater attenuation.
Reflected signals arrive at the receive aerial after reflection from the
earth in front of the aerial, and from objects either side of the radio
path. Reflecting objects in the vertical plane, such as hills, will produce
a stronger reflected signal if the wave is vertically polarized.
In general, it is found that vertical polarization gives a larger
received field strength at low heights above the ground but the
probability of fading is greater. In hilly and/or wooded areas horizontal
polarization is probably the better but vertical polarization is preferred
for links that pass over flat countryside. For horizontal polarization
the received field strength falls to zero at heights below about 4 metres.
This does not matter for point-to-point links since the receive aerial
is always mounted at greater heights than that, but it does mean that
mobile land systems must employ vertical polarization.

Direct wave
Reflected wave
Received Field Strength

Figure 8.16 shows a line-of-sight radio link; the k-factor is 4/3 so

that the radio wave travels in a straight-line path and the earth is
assumed to be flat. Two aerials, one at height h, above ground, and
the other at height h,, are D kilometres apart. The total field strength
Fig. 8.16 A line-of-sight radio link. produced at the receive aerial is the phasor sum of the individual field
 PROPAGATION OF RADIO WAVES 179

strengths produced by the direct wave and by the reflected wave. The
magnitudes of these waves are inversely proportional to the distance
they have travelled. Since the extra length of the reflected path is
negligible compared to the distance D between the two aerials any
difference in amplitude due to this factor is negligibly small. Hence
if |p| = 1, |Ep| = |Exl = £,/D; if |p| # 1, then |ER| =
|p| £,/D. The amplitude of the resultant field strength will be a
function of the phase difference between the direct and the reflected
waves. This phase difference exists because of the angle of the ground-
reflection coefficient and because of the difference in the direct and
reflected path lengths. For small grazing angles the phase change upon
reflection is approximately constant at 180°. The phase difference
¢ due to the path length difference is @ = 2 7/) times that difference.
The phasor diagram of the field strengths at the receive aerial is shown
by Figs 8.17(a) and (b). Consider Fig. 8.17(a) in which |p| = 1
so that |Ep| = |Ep]; this figure has been re-drawn in Fig. 8.17(c)
from which

20 eelpa alld
< — ¢ | @
AB
= 2 cos) ——_ |, = Ee sin:
|—

2E
and Ey = 2AB = 2Ep sin CR be SR: (8.10)
2 D 2
Figure 8.17(b) is the phasor diagram when |Ep| # |Ep|; resolv-
ing Ep into its horizontal and vertical components gives the diagram
shown in Fig. 8.17(d). From this

Ep PE cos(r~4)

pEp sin(x—¢) E,

Fig. 8.17 Phasor diagrams of the

received field strength in a line-of-
sight link.
180 RADIO SYSTEMS TECHNOLOGY

Ey = \{{(Ep + |e|Ep cos (x — )/°

+ |p |Ep sin? (x — $)}
= Ep|[1 + |p|? + 2] | cos (4 — 4)]
or Ey = i + |o|* — 2|p| cos 4]. (8.11)

IG:iio beri »
Ey = a {[2d. — cos $)] = = a sin? 6)

= 21 sin (5)
D Z
as before.
It is now necessary to determine the angle @. Figure 8.18 is an
extension of Fig. 8.16. From this figure

Nate isis Grice BAIS De (h,

— h,)?
D-__—_

Fig. 8.18 Calculation of the angle ¢.

h, + h,)?
D3 =D? + (+h)? = D+

The difference between the lengths of the direct and the reflected paths
is

D, — D, =
(+ hy)” — (Cy — hy)? _ Ahh
2D D
and hence the phase angle
2% 2hh, ~~ “4ahh,
Height
ground
above
d D AD
Therefore, from equation (8.10),
2E 27h,h
A et i leShe (8.12)
D AD
At a fixed distance from the transmitting aerial the received field
Received field strength strength goes through successive maxima and minima as the height
Fig. 8.19 Variation of field strength h, above ground is increased. This is shown by Fig. 8.19. In prac-
with height above ground at the tice, the reflection coefficient of the earth at u.h.f. and higher frequen-
receive end of a line-of-sight radio cies is never exactly equal to 1 z 180° because of surface roughness
link. and the minima do not quite reach zero.

Example 8.6

An aerial is mounted 250 m above flat earth. Determine the minimum height
at which the receive aerial should be mounted if it is to receive the maximum
field strength. The distance between the aerials is 22 km and the frequency
is 600 MHz.
 PROPAGATION OF RADIO WAVES 181

Solution
From equation (8.12) the maximum field strength is obtained when
2ahh, nw
ee:
>

where n is an odd integer, or

n\D
4h,
T

The minimum aerial height occurs when n = 1 and therefore

But
Sex 22 «10?
h, = = 11m. (Ans.)
600 x 10° x 250 x 4

The distance D between the two aerials is always much larger than
the heights h, and h, of the aerials. This means that 2h,h,/ND is a
small angle and so equation (8.12) can be written as
_ 2E, 2ah,h,
TOF 4 FCAALD

or aa
Ey = afte :
(8.13)

This means that the field strength at a fixed height h, above the
ground is inversely proportional to the square of the distance from
the transmitter, and directly proportional to frequency. As the distance
is increased and nears the optical horizon the field strength tends to
be equal to that produced by the direct wave alone and to be
independent of frequency.

Curvature of the Earth

The assumption made, in deriving expressions for the received field
strength, that the earth between the aerials is flat is, of course, not
true. The error implicit in the assumption, however, is small unless
the length of the link is approaching the optical horizon. If the curva-
ture of the earth is to be taken into account a divergence factor F must
be introduced that allows for the wave to be reflected from a curved
surface. The divergence factor is employed by multiplying it and the
reflection coefficient together, i.e. Ep = pFEp.

Maximum Distance between Aerials

Because of the curvature of the earth there is a maximum distance

from the transmitting aerial at which the receive aerial can be sited
and still be able to receive the direct wave. This distance is known
as the radio horizon. It is shown in Fig. 8.20 in which the effective
radius of the earth is KR = 8500 km, h, and h, are heights of the
Fig. 8.20 Maximum distance transmit and receive aerials, and D, and D, are the distances from
between two aerials. each aerial to the point of grazing incidence. At this point the direct
182 RADIO SYSTEMS TECHNOLOGY

wave is tangential to the surface of the earth (when there will not be
a reflected wave). From the figure
(h, + kR)? = Di + (kR)?
or D, = |(2h,kR).
Similarly, D, = \(2h,kR). The maximum distance between the
aerials is

Dmax = D, + Dy = |(2h,kR) + |(2h,KR). (8.14)

If the heights of the two aerials are measured in metres and the
distance D is in kilometres,

h
Datt= af x 8500) + as x 8500 },
1000 1000
or Dmax = 4.13(Vh, + Vh,). (8.15)

Fresnel Zones

For a good, reliable signal to be received by an aerial there must be

adequate clearance between the direct path and any obstacles. The
necessary clearance is usually expressed in terms of Fresnel zones.
A Fresnel zone is the locus of the points from which the sum of the
distances to the transmitting and receiving aerials is equal to the direct
distance between the aerials plus an integral number of half
wavelengths.
The first Fresnel zone is the locus of the points from which the
sum of the distances to each aerial is \/2 longer than the direct path
between the aerials. Since the ground-reflected wave experiences
approximately 180° phase change upon any reflection that takes place
at a point on the first Fresnel zone, it results in a signal which, at
the receive aerial, is in phase with the direct wave.
The second Fresnel zone is the locus of the points from which the
sum of the distances to each aerial is \ longer than the direct path
between the aerials. Reflection from a point on this zone will produce
a signal which, at the receive aerial, is in anti-phase with the wanted
signal.
In similar fashion, the third, fifth, etc., Fresnel zones produce in-
phase signals, and the fourth, sixth, etc., Fresnel zones produce anti-
phase signals. The first, second and third Fresnel zones are shown
by Fig. 8.21.
The design of a line-of-sight radio link involves a choice of the
Fresnel zone clearance deemed to be necessary since, if the direct
path is too near the surface of the earth extra diffraction losses must
be expected. To avoid this, the clearance above ground of the direct
ray should be equal to about 0.6 times the radius of the first Fresnel
zone. Account must also be taken of possible sub-refraction effects
 PROPAGATION OF RADIO WAVES 183

Third Fresnel zone

Second Fresnel zone

Fig. 8.21 The first, second and

third Fresnel zones.

when the atmospheric conditions are such that the k-factor becomes
smaller than unity. In the British Isles the k-factor rarely falls below
0.7 and so this is the value that is generally used in link design.
Reflections from more distant objects tend to cancel out.

Radius of a Fresnel Zone

In Fig. 8.22 the distance T—A—R is )/2 longer than the distance T—R
so that the point A is on the first Fresnel zone. The radius of this
zone is r and hence
(D} + r*) = (D5 +r’) =D, + Dy + 02
r2 r2

2D? 2D3

Fig. 8.22 Calculation of the radius or r= : (8.16)

of the first Fresnel zone. Z J(Bi
The radius of any of the higher-order Fresnel zones is simply found
by multiplying equation (8.16) by Vn.

Example 8.7
A 3 GHz radio link has transmitting and receiving aerials at equal heights
above the ground, and it is 36 km long. There is a 30 metre high obstacle
midway between the two aerials. Determine the minimum height at which
the aerials ought to be mounted for the radio path to be unobstructed. Assume
the k-factor to be 0.7.

Solution
The minimum height of the two aerials must be equal to h; + h, + hg,
where h, is the height for grazing incidence, h, is the height of the obstacle,
and h, is 0.6 times the radius of the first Fresnel zone.
seer no ys
oon 10")
From equation (8.14),

18 x 10° = j(2h, x 0.7 x 6400 x 10°)

 184 RADIO SYSTEMS TECHNOLOGY

or h, = 36.16m

h, = 30 m.

From equation (8.16),

0.1 x 18 x 10°
h, = 0.6 : = 18m.

Therefore, the minimum aerial height is

h = 36.16 + 30 + 18 = 84.16 m. (Ans.)

If the two aerials are not at equal heights and/or the obstacle is not
at mid-path the problem is more complex. It is best approached by
assuming the earth between the aerials to be flat and then increasing
the effective height of the obstacle(s) by the amount necessary to
account for the curvature of the earth. The effective increase in height
hin, Of the obstacle is given by equation (8.17), i.e.

inc ees (8.17)

2kR
where D, is the distance from the transmitting aerial to the obstacle
and D, is the distance from the obstacle to the receive aerial.

Example 8.8
A 3 GHz signal is transmitted from an 80 m high aerial towards a receiving
aerial that is 39 km away. A 50 m high obstacle is 25 m from the transmitter.
Calculate the necessary minimum height of the receive aerial. Assume the
k-factor to be 0.7 and allow a clearance equal to 0.6 times the radius of the
first Fresnel zone.

Solution
When flat earth is assumed the obstacle must be given an effective height of
25 x 14 x 10°
= 39.1 + 50 = 89m.
2 x 0.7 x 6400 x 10°
The first Fresnel zone clearance is

0.1 x 25 x 14 x 10°
0.6 : = 18m.
39 x 10
From Fig. 8.23
_ 82 18) — 807 h, — 80
—w—_§_25 km ———>=—_14 km —> tan 0
od wok lO. BO CTO
Fig. 8.23 or h, = 122m. (Ans.)

Height Diversity

If the radio path crosses over a wide stretch of water, such as a

tidal
estuary, or the sea, the geometry of the link may vary with the
tide.
 PROPAGATION OF RADIO WAVES 185

It is often necessary to employ some form of diversity reception in

order to obtain a reliable system. To economize with the usage of
frequency it is usual to employ height diversity in which two receiving
aerials are used, one mounted above the other. For the optimum results
the lower aerial should be mounted at the height corresponding to
the first maximum in the field strength/height characteristic (Fig.
8.19), and the other aerial at the height of the null immediately above.

Scatter Propagation A tropospheric-scatter radio link operates with its distant terminal well
beyond the radio horizon. A large amount of radio-frequency energy
is radiated, by a highly directive aerial, towards the horizon. A very
small proportion of the radiated energy is forward scattered by the
troposphere and is directed downwards towards the receive aerial.
Most of the transmitted energy continues upwards, passes through
the ionosphere and is radiated into space. Figure 8.24 shows the path
geometry of a tropospheric-scatter radio link. The solid angles formed
by the narrow radiation patterns of the two aerials intercept one another
to form a common volume that is known as the scatter volume. The
scatter volume is typically only one, or two, kilometres above earth
and it is here that the useful energy is returned to earth. Both the
transmitter Jaunch angle and the scatter angle should be small, usually
less than about 4°.

RY
Scatter volume ~~ Rey)
cs

Launch
angle

Fig. 8.24 Tropospheric scatter link. Transmitter

Tropospheric-scatter propagation is possible at most frequencies

higher than 500 MHz, but it is employed mostly at frequencies in
the region of 900 MHz, 2 GHz and 5 GHz. It provides a bandwidth
of several megahertz and is employed to carry wideband telephony
systems, with between 38 and 132 channels. The main application
for a tropospheric-scatter system is for communication over hostile
terrain, such as sea or mountains, where a communications satellite
system would not be economically viable. It is the principal broadband
communication technology employed for communication between the
mainland and the North Sea oil rigs; these operate at 2.25 GHz with
a transmitted power of 1 kW.
The path attenuation between transmitter and receiver is much
greater than the free-space attenuation (given by equation (6.28)), by
an over-the-horizon loss, or scatter loss. The scatter loss accounts
186 RADIO SYSTEMS TECHNOLOGY

for the fact that only a small fraction of the transmitted power arrives
at the receive aerial. Since this loss is high, typically 70 dB or more,
a tropospheric-scatter system must be provided with high-power trans-
mitters, high-gain aerials, and low-noise factor, highly sensitive radio
receivers. The received signal is further reduced by the aperture-to-
coupling loss. This expresses the loss of gain of a parabolic dish aerial
when it receives a scattered signal and it is typically about 10 dB;
it occurs because the received signal is made up of a large number
of components bearing a random phase and amplitude relationship
with one another. The scatter loss is continually varying and this leads
to substantial fading of the received signal. To counter this some kind
of diversity reception is usual and perhaps the most commonly
employed is quadruple space diversity. This involves the use of four
aerials: two at each end of a route. The two transmitting aerials radiate
energy simultaneously, with their signals polarized at right angles to
one another. The two receive aerials pick up signals in both the
horizontal and the vertical planes and feed receivers whose outputs
are commoned.

Example 8.9

A 200 km tropospheric-scatter system operates at 2 GHz with a transmitted

power of 1 kW and a receiver of noise factor 5 dB. The system uses frequency
modulation with a maximum modulating frequency of 100 kHz and a deviation
ratio of 4.5.
Calculate the dimensions of the identical transmitting and receiving aerials
required to give an output signal-to-noise ratio of 60 dB. The over-the-horizon
loss is 70 dB and the aerial noise temperature is 290 K. Assume the aerials
to have an illumination efficiency of 60% and an aperture-to-medium loss
of 9 dB.

Solution
The frequency deviation of the signal is 4.5 x 100 = 450 kHz. Hence the
required bandwidth is 2(450 + 100) x 10° = 1.1 MHz.

60 dB = P, dBW + 20 log,) (V3D) dB — kTB dBW — FdB

= P. dBW + 17.84 dB — 143.6 dBW + 5 dB
= —96.4 dBW = 2.29 x 107" w.
From equation (6.28)

2.29 x 10°" = GG. x 100( 3 x 108 2

Se )
2x 10° x 4m x 200 x 103
¥107 3 0.495

GG
6 br22doe 10° ‘
Oe ee, = SO
AASt AOF
G,= Gr=70 718" or ©48-6 dB.
Therefore,
 PROPAGATION OF RADIO WAVES 187

aD \2 aD \?
(Amo e—OLOle———)- =. 6 ==
oN 0.15
Of De. 1G. mk (Ans.)

Propagation via a Communications The power received by the ground station of a communications satellite
Satellite link can be determined using equation (6.28). The term (4 mD/d)?
is known as the transmission loss and it accounts for the way in which
the radio wave diverges as it travels. This relationship can be
alternatively expressed as
effective radiated power x gain of receive aerial
r
transmission loss
(8.18)

The received power is often referred to as the carrier power.

In practice, further losses are experienced because of climatic
conditions such as rain, snow, etc.

Example 8.10
A communications satellite is 40 000 km from a point on the surface of the
earth and it transmits a power of 2 W from an aerial of 20 dB gain. Calculate
the power received at the earth station by an aerial of effective aperture 10 m
if the frequency is 11 GHz.

Solution
Method (a)
3.x 10°
C—O ame
1x. 402
The flux density at the receive aerial is
P.G, 2 x 100 =.
P, = = = 9/95.x 10S. Wim:
4D? 4n(40 x 10°)?
The received power
P. = P,A, = 9.95 x 107 x 10 = 9.95 x 10°" W. —(Ans.)
Method (b)
4r x 40 x 10° 2,
Transmission loss = ( )= 3.466 x 107.
0.027

: ; . 4mA, 4x x 10
Gain of the receive aerial = mae a 1723 78.
r 0.027
Therefore
2 x 100 x 173 278
the received power ll
3.466 x 107°
9.95 x 10°'*W. — (Ans.)
188 RADIO SYSTEMS TECHNOLOGY

Figure-of-Merit or G/T Ratio

In a communications satellite system the received power levels are

always very small because of the extremely large distances involved.
This means that the noise generated within the receiving system must
be reduced to the minimum possible value in order to achieve a satis-
factory carrier-to-noise ratio. There are two main requirements that
must be satisfied to achieve this. First, the bandwidth B of the receiver
must be as narrow as possible, and second, the system noise
temperature must be as low as possible.
If the system noise temperature (p. 109) is 7; then the noise power
P;,, at the input to the demodulator is given by P;, = GkT; 5B, where
G is the gain of the receiver from the r.f. input to the demodulator
input. Since the received carrier power is P, watts the carrier-to-
noise ratio at the demodulator input is
CNY GRA 2G ( rd y
N GkIsB kT.B \4aD
= PGE) RORRG.
(8.19)
kB \4nD/) Ts
The carrier-to-noise ratio is proportional to the ratio (gain of receive
aerial)/(receiver system noise temperature). Since, for a given system
P.G, ‘oun wee
it tee is a constant, the performance of a communications
kB \4xD
satellite system can be improved by increasing its G/T ratio. A typical
figure for the G/T ratio, or figure-of-merit, is 40 dB~'K.

Example 8.11

The parabolic dish aerial used by the ground station of a communications

satellite link has a diameter of 28 m and an illumination efficiency of 70%.
The link operates at a frequency of 4 GHz. Calculate the G/T ratio of the
system if the system noise temperature is 80 K.

Solution

3° x 100
dS 0075 at
BSE 10
The gain of the receive aerial is
n4aA, 0.7 x 4a? x 14?
= 962 922.5 Of ~ 5958 dB!
Ae 0.0757
Also, T = 80 K = 19 dB~'K. Therefore,
G/T ratio = 59.8 — 19 = 40.8dB~'K. —(Ans.)
 Communication Radio
Receivers

The function of a communication radio receiver is to select the wanted

signal present at the receive aerial, which may be of very small
amplitude, from the background noise and to reject a large number
of, possibly stronger, unwanted signals. The radio receiver must then
amplify and demodulate the received signal to provide an output base-
band signal with at least the minimum required signal-to-noise ratio.
The main problem is not amplifying the received signal to the wanted
level but overcoming the adverse effects of both the received noise
and internally generated noise and interference.
Most m.f./h.f. communication radio receivers are able to receive —
different kinds of signal, such as d.s.b. amplitude modulation,
s.s.b.s.c., c.w. and data, and in some cases frequency modulation
also. Receivers designed for use in the v.h.f./u.h.f. bands usually
receive either amplitude-modulation or frequency-modulation signals.
Modern communication receivers are mostly of the double super-
heterodyne type; less often single-, or triple-superheterodyne receivers
are employed. There are two main reasons for this trend: first, the
first intermediate frequency (i.f.) can be high, giving a wide separa-
tion between the wanted signal frequency and the image-channel signal
frequency; and, second, the second i.f. can be low, making good
adjacent-channel selectivity easier to obtain. Since many applications
for communication radio receivers require the receiver to be remotely
controlled and/or able to change frequency rapidiy and accurately,
broadband front-ends and frequency synthesis are commonly
employed.

Double-Superheterodyne Radio The block diagram of a typical h.f. communication radio receiver is
Receivers shown by Fig. 9.1. The first i.f. of 45 MHz is above the 2—30 MHz
tuning range of the receiver. The second i.f. is at the (more or less)
standard value of 1.4 MHz. The first, and the second, local oscillator
frequencies which must be supplied to the two mixers are both derived
from a frequency synthesizer. Signals picked up by the aerial are
passed through the 30 MHz low-pass filter to remove all signals at
frequencies above the tuning range, but especially any signals at, or
near, the image-channel frequency, and the first intermediate
frequency. The r.f. stage should transfer the maximum r.f. power
from the aerial to the receiver and this means that it should be matched
to the aerial. This, however, is not always possible since the receiver
may be used in conjunction with more than one type of aerial. In some
190 RADIO SYSTEMS TECHNOLOGY

Aerial Output

30 MHz 45 MHz Abies

low-pass Variable ;
if Band-pass First if. Second bandwidth Second if, a.m. af.
filter filter amplifier i mixer i filter i amplifier detector amplifier

a.g.c.
foo = 46.4 MHz Product
f,, = 47 MHz detector
to 75 MHz Peale)
synthesizer
Carrier
re-insert

Fig. 9.1 An h.f. communication

radio receiver. older types of h.f. receiver the r.f. stage consists of one, or more,
r.f. amplifier stages that are ganged with the first local oscillator. Some
modern receivers employ a bank of sub-octave band-pass filters
any one of which may be switched into circuit, quite likely by
microprocessor-controlled digital circuitry.
The signals are then passed through an attenuator whose attenua-
tion is varied by the automatic gain-control (a. g.c.) system. The r.f.
attenuator is followed by a broadband amplifier and the first mixer
(an r.f. amplifier is not provided in some receivers). Besides ampli-
fying the received signal and improving the noise factor of the
receiver, the amplifier also isolates the first mixer from the aerial
and helps to prevent local oscillator radiation from the aerial. The
first mixer mixes the amplified signal with the first oscillator frequency
and generates a component at their difference frequency fo, — fs =
45 MHz. The frequency supplied by the frequency synthesizer must
therefore be variable from 47 MHz to 75 MHz. The difference-
frequency component is then selected by the first if. band-pass filter
and it is then amplified. The selected signal is then mixed with the
fixed second local oscillator frequency of 46.4 MHz to shift it to the
second intermediate frequency of 1.4 MHz. The wanted 1.4 MHz
component of the output of the second mixer is selected by one of
a number of band-pass filters. Each filter has a centre frequency of
1.4 MHz but a bandwidth appropriate for the different kinds of signals
handled by the receiver, e.g. 8 kHz for a.m., 2.4 kHz for
S.s.b.,
1 kHz or 400 Hz for c.w. After filtering, the signal is passed to either
one of the two detectors. The a.m. detector demodulates d.s.b.a.m.
signals and also generates the a.g.c. voltage. The product detector
is used to demodulate all other kinds of signal. Finally, the
demodulated signal is amplified by the audio amplifier.
The quality of the output signal is affected by a number of factors,
amongst which are: (a) adjacent-channel, co-channel and image-
channel signals, (b) reciprocal mixing, and (c) cross-modulation
and
intermodulation. In the design of a modern communication receiver
every effort is made to minimize the adverse effects of these
factors;
this means that particular attention is paid to r.f. linearity and
filtering
and to i.f. selectivity. Increasingly, modern designs make full
use of
readily available ICs to provide such circuit features as amplific
ation,
detection, frequency synthesis and digital tuning and
displays.
 COMMUNICATION RADIO RECEIVERS 191

The block diagram of a v.h.f./u.h.f. communication radio receiver

follows a broadly similar scheme to Fig. 9.1. Typically, the receiver
might cover the frequency band 20 MHz to 470 MHz with a first inter-
mediate frequency of 515 MHz. The second intermediate frequency
is nearly always 10.7 MHz. The received signals may be either
amplitude, or frequency modulated.

Choice of the Local Oscillator Frequency

The first intermediate frequency of a communication radio receiver

is equal to the difference between the first local oscillator frequency
fo; and the wanted signal frequency fg. It is usual to make the
oscillator frequency higher than the signal frequency since the ratio
foi(max)/fo1mmin) is then smaller. Assuming this, there are two possi-
bilities: either the wanted signal can be shifted to a lower intermediate
frequency, known as down-conversion, or it can be shifted to a higher
intermediate frequency, known as up-conversion. For modern h.f.
communication radio receivers up-conversion is the more common
technique since it ensures that all possible image-channel signals lie
above the tuning range of the receiver. The image-channel signal can
then be suppressed by a low-pass filter whose cut-off frequency is
equal to the highest frequency to which the receiver can tune. A further
advantage of up-conversion is that the ratio fo\(max)/fo1cminy 18 smaller
than for down-conversion.

Image-channel Interference

When a superheterodyne radio receiver has been tuned to a frequency

fs there will always be another frequency, known as the image-
channel frequency fim, that will also produce the first intermediate
frequency fir if it is allowed to reach the first mixer. Now,

fir = for — Ss = fim — Sor:

fist = fim — Git. + fs)
or fim = Js + 2fin-
Thus, the image-channel signal is separated from the wanted signal
by a frequency gap equal to twice the first intermediate frequency.
The r.f. stage must include sufficient selectivity to stop the image-
channel signal reaching the first mixer.
The frequency of the image-channel signal will vary as the receiver
is tuned to receive different signals. The r.f. stage must therefore be
able to reject image-channel signals at different frequencies. Tradi-
tionally, the necessary front-end selectivity has been provided by
variable-tuned resonant circuits ganged with the first local oscillator.
Increasingly, modern communication receivers either employ a single
 192 RADIO SYSTEMS TECHNOLOGY

low-pass filter (as in Fig. 9.1), or a bank of sub-octave band-pass

filters.
Any vestige of the image-channel signal that reaches the first mixer
will cause an unwanted signal to appear at the first intermediate
frequency. This signal cannot be rejected by the first i.f. filter and
so it will produce crosstalk at the output of the receiver. The image-
response ratio is the ratio, in decibels, of the r.f. input voltages at
the wanted signal frequency, and at the image-channel frequency, that
are necessary to produce the same audio output power. Typically,
an h.f. communication receiver might have an image-response ratio
of 100 dB and a v.h.f./u.h.f. receiver of some 70 to 90 dB.

Choice of the Intermediate Frequencies

The main factors to be considered when choosing the intermediate

frequencies for a communication receiver are: (a) interference signals,
(b) adjacent-channel selectivity, (c) if. breakthrough, and (d) the
availability of crystal and SAW filters.
The first intermediate frequency should not be in, or near, the tuning
range of the receiver. Then the front end of the receiver can include
filtering to prevent any signals at, or near, the first intermediate
frequency reaching the first mixer. Usually, the first i-f. stage is
isolated from the r.f. input by at least 80 dB, but preferably 120 dB.
The smaller the frequency separation between the wanted signal
and the image-channel signal the harder it is to achieve adequate
suppression of the image-channel signal. This factor requires the first
intermediate frequency to be as high as possible. Conversely, good
adjacent-channel selectivity is easier to obtain if the second inter-
mediate frequency is low. The actual frequencies chosen are decided
by the frequencies at which crystal and SAW filters are readily
available.

Co-channel Interference
Co-channel interference is caused by an unwanted signal at
the same
frequency as the wanted signal. Clearly, it cannot be elimina
ted either
by filtering or by the selectivity of the receiver. In a v.h.f./u
.h.f.
frequency-modulation receiver co-channel interference
is not
important since it is eliminated by the capture effect.

Intermodulation When two, or more, signals at frequencies f, and f; are

applied to
a non-linear characteristic they will generate intermodulatio
n products.
These may be either second-order f, + Sz products or third-o
rder
2f, + fr, 2 + fi, products. The more important
of these are the
third-order products since they tend to have frequencies
that are within
the passband of the first i.f. stage. For example, suppos
e that a wanted
 COMMUNICATION RADIO RECEIVERS 193

signal at 3.01 MHz is accompanied by two unwanted signals at

3.02 MHz and 3.03 MHz. The third-order intermodulation product
(2 x 3.02) — 3.03 is at the same frequency as the wanted signal.
A non-linear device has a transfer characteristic of the form
v=at bv + cy? + dv? + :.., (9.1)

where a, b, c and d are constants.

The intermodulation products are generated by both the square and
the cube terms. Consider the square term, with inputs V, cos wt and
V, cos wt,

Vout = C(V; cos wt + V2 cos ant)

= cV{ cos? w,t + cV5 cos” wt

+ 2cV,V2 cos w,t COS wt. (9.2)

The first two terms give the second harmonics of the two input signals
but these will not pass through the i.f. filter. The third term can be
expanded to give

cV, V2[cos (w, — w2)t + cos (w, + w»)t]

and this shows that the output voltage contains components at

frequencies f; + f;. These components are known as the second-
order intermodulation products.
In general, second-order intermodulation products fall outside of
the passband of the first i.f. filter and are suppressed. If, however,

Aa fA+ ©
then

A-f-
and the second harmonic of this is equal to the first intermediate
frequency. This is known as half-i.f., or repeat-spot, interference.
Alternatively, if the receiver is tuned to frequency fs and there is an
unwanted signal at frequency fs — fir/2, the second harmonic of the
unwanted signal will mix with the second harmonic of the local
oscillator frequency to give the intermediate frequency.
Third-order intermodulation products are generated by the cubic
term in equation (9.1). Thus

Vor = a(V; cos wt + Vz cos wt)?

= d(V} cos? wit + V3 cos? wot
+ 3V2V>
V2 cos? wt COs wot2
+ 3V,V3 cos w;t cos” wt)
194 RADIO SYSTEMS TECHNOLOGY

V3
= “ (3 cos w,t + cos 3a ft)

dv3
Se pir (3 cos wt + cos 3af)

oe dV {V3 COS Wot + + cos (2, + w)t

+ dcos (2a; — a)t]

te dV, V 53 COS w,t + 3 cos (2w. + w))t
+ dos (2w) — w,)r]. (9.3)
Either of the 2w; — w) or 2w) — w, components may be at such
a frequency that it falls within the passband of the first i.f. filter.
Higher, even-order products are generally out-of-band while higher
order, odd-order products are usually of negligible amplitude.
The third-order intermodulation level is the level, in decibels,
relative to 1 uV (dBuV) of two unwanted signals, respectively 10 kHz
and 20 kHz off-tune that generate an unwanted third-order output
equivalent to that produced by a wanted signal at 0 dBuV e.m.f. A
good communication receiver would have a level of about 85 dB BV.
To reduce intermodulation, a pre-selector stage, such as a bank of
sub-octave filters, is often employed before the first non-linear stage.
Alternatively (or in addition), an r.f. attenuator will reduce the third-
order intermodulation level by three times the r.f. attenuation in dB.
If, for example, a 6 dB attenuator is fitted it will reduce each inter-
fering signal by 6 dB but, since the amplitude of the third-order
products is proportional to V7V>, or VV%5, it will reduce the inter-
modulation level by 18 dB. Hence, the use of a 6 dB r.f. attenuator
would increase the signal-to-intermodulation level by 18 — 6 or
12 dB.

Intercept Point

Reference to equations (9.2) and (9.3) shows that if the amplitudes

of the two unwanted signals are equal to one another, i.e. Vj = V> =
V, the amplitudes Vip of the second-order and the third-order
intermodulation products are given by

Vip = k,V", (9.4)

where k, is the nth-order constant, and n is the order of the
intermodulation products.
This means that whereas the output voltage due to the wanted signal
increases in direct proportion to its input voltage V the intermodula-
tion output voltage increases in proportion to V”. There must
therefore be a level at which the two output voltages are equal
to one
another. This voltage, Vpy, is known as the nth-order intercep
t point.
The intercept point is a purely theoretical level because the
amplifier,
 COMMUNICATION RADIO RECEIVERS 195

Third-order intercept point

Second-order intercept point

Wanted
output , ‘
Third-order intermodulation output

Output
(dBm)
power
Second-order intermodulation output

Fig. 9.2 Second- and third-order

intercept points. Input power (dBm)

or mixer, would saturate before the level is reached. The intercept

point is determined by extending, on a graph of output power in dBm
plotted against input power in dBm, the wanted signal and the nth-
order intermodulation product outputs to the point where they intercept
each other. Figure 9.2 shows how output power varies with increase
in the wanted signal, and for both the second-order and third-order
intermodulation products. Since both axes are logarithmic the second-
order plot has twice the slope, and the third-order plot has three times
the slope, of the wanted signal output plot. The points at which the
curves cross are known as the second-order intercept point and the
third-order intercept point. The third-order intercept point is a measure
of the ability of a receiver to reject large-amplitude unwanted signals
whose frequency is near to that of the wanted signal. The third-order
intercept point is primarily determined by the linearity of the first
mixer and it should be as large as possible.
For the third-order intermodulation products n = 3 and equation
(9.4) can be written in the form
V ial 0

P3p = (5) Ye alee (9.5)

where P = V/2 and is the power due to one signal on its own. This
means that the intermodulation power is proportional to the cube of
the input power, i.e. a 1 dB increase in the input power gives a 3 dB
increase in the third-order intermodulation output power.
The ratio D of the intermodulation power to the wanted signal power
is D = P3p/P5y. P31p is proportional to P? and Py is proportional
to P. Hence
D = (KP)’. (9.6)
At the third-order intercept point P3rp = Pour and so D is equal to
unity. Equation (9.6) becomes 1 = (KP)” = (KPsp)- 08 Kk =
P3,;p_'. Consequently, equation (9.6) can be written as
196 RADIO SYSTEMS TECHNOLOGY

D= Ba. (9.7)
P31p
Example 9.1
Calculate the third-order intercept of a receiver if, for an input power
of 0 dBm, the ratio (intermodulation power)/(wanted power) at the output
is —40 dBm.

Solution
From equation (9.7)
Pe
— 40 a 2 logio Fe — 2 x (O ae P31p) = SON Bs
3IP

Therefore, the third-order intercept point is +20 dBm. (Ans.)

Note from this example that for a 0 dBm two-tone signal the third-
order intercept point is equal to -} times the magnitude of the
third-order intermodulation products.
When the two signals are not at 0 dBm they should both be norma-
lized to 0 dBm. The third-order intermodulation level then increases
by 3 dB for every 1 dB increase in the two-tone signal level. If the
two signals are at different levels then subtract one third of their level
difference in dBm from the larger level and take the result as being
their common level.

Example 9.2

When two 0 dBm tones are applied to a mixer the level of the third-order
intermodulation products is —60 dBm. The mixer has a conversion loss of
6 dB. Calculate, in dBm, the third-order intermodulation output power
when
the level of the two tones is (2) —10 dBm, (6) +10 dBm, (c) +30
dBm,
and (d) +20 dBm and +11 dBm.

Solution
The level of the wanted signal at the mixer output is —6 dBm.

(a) The third-order output power is (—60) + 3 x (—10)

= —90 dBm.
(Ans.)
(b) The third-order output power is (—60) + 3 x 10 =
—30 dBm.
(Ans.)
(c) The third-order output power is (—60) + 3 x 30 = +30
dBm.
(Ans.)
This answer means that the third-order output level is equal to
the input level
so this is the third-order intercept point.

(d) The equivalent input level is

(20 — 11)
20 — pareesian? = +17 dBm.

The third-order output power is (—60) + 3 x 17 = —9 dBm.

(Ans.)
 COMMUNICATION RADIO RECEIVERS 197

Typical figures for an m.f./h.f. communication receiver are

+15 dBm second-order and +20 dBm third-order intercept points.
For a v.h.f./u.h.f. receiver typical figures are second-order intercept
point +5 dBm up to 470 MHz, and 0 dBm from 470 MHz to
1.1 GHz.

Dynamic Range

The dynamic range of a radio receiver is the range of input levels

that produce output powers lying in between the noise floor of the
receiver and the input level that makes the total intermodulation
product power equal to the noise floor. The dynamic range is given by
dynamic range = 2{P3p — noise floor]

= Psp — FkT)B (dB)). (9.8)

Example 9.3
Calculate the dynamic range of a receiver that has a third-order intercept point
of +20 dBm, a noise factor of 6 dB and a bandwidth of 8 kHz. kT =
—174 dBm.

Solution
From equation (9.8)

dynamic range = 3{20 + 174 — 6 — 10, log 8000] = 99 dB.

(Ans.)

Reciprocal Mixing When a large-amplitude off-tune signal appears at the input to the
first mixer it will mix with the noise sidebands of the first local
oscillator and may produce in-band noise. This process is known as
reciprocal mixing and it is illustrated by Fig. 9.3 in which both the

Local oscillator
Noise sidebands
Unwanted frequency 1

Wanted signal Up-converted signals

if.
signal
Voltage bandwidth
—

Fig. 9.3 Reciprocal mixing. Frequency

 198 RADIO SYSTEMS TECHNOLOGY

wanted signal and the unwanted signal are assumed to occupy narrow
bandwidths that are much less than the first i.f. bandwidth. The first
local oscillator has upper, and lower noise sidebands, sometimes
known as phase noise, and these produce unwanted signals which lie
within the passband of the first i.f. filter. Reciprocal mixing is defined
as the amount of noise introduced by a 20 kHz off-tune signal that
will produce an output equivalent to that produced by the wanted signal
when its voltage is 1 »V e.m.f. Suppose, for example, that the
frequency of the wanted signal is 12 MHz and that the first inter-
mediate frequency is 45 MHz. The first local oscillator frequency is
then 57 MHz. If there is a 45.01 MHz, 3 kHz slice of oscillator noise
93 dB down on the oscillator voltage then an unwanted signal at
12.01 MHz would be converted to a 3 kHz noise band at 45 MHz
and spuriously received.
The reciprocal-mixing performance of a receiver affects its ability
to reject off-tune signals and it means that the effective selectivity
of the receiver is not as good as the selectivity defined by the i-f.
filters. The effect of reciprocal mixing on selectivity is shown by Fig.
9.4; clearly, the selectivity characteristic is widened.

Second i.f. bandwidth

v4 Reciprocal mixing
Voltage
iN Less reciprocal mixing

First i.f. bandwidth

Fig. 9.4 Effect of reciprocal mixing

upon selectivity. Frequency

Cross Modulation Cross modulation is the transfer of the amplitude modulation of an

unwanted carrier, that appears at the input to the first mixer, on to
the wanted carrier. Equation (9.3) contains the term dV, V3 cos wt;
if V, is the wanted signal and V, is the unwanted amplitude-
modulated signal, i.e. V,(1 + mcos w,t), then the amplitude of the
unwanted signal is
3dV, V3 + 2m cos w,t + m? cos? Wink): (9.9)
igen the modulation factor m is small enough for the term
m? cos? Wm to be neglected. Then the total output signal at the
wanted frequency w, is, from equations (9.1), (9. 3) and (9.9)
bV, + ZdV} + 3dV,V3 + 3dV,V3m cos wml.
The second and third terms are small enough compared with the first
term, since d < b, to be neglected and hence the output at the wanted
 COMMUNICATION RADIO RECEIVERS 199

frequency w, consists of the wanted component bV, plus an unwanted

cross-modulation component 3dV,V5m cos w,,t. The signal-to-cross-
modulation ratio SCMR is
3dV,Vim
SCMR = 20 loeio(———+— |dB
1

20 lowo(3dV5al dB. (9.10)

The cross-modulation factor is the ratio of the modulation factors
of the superimposed modulation and of the unwanted signal, i.e.
2
cross-modulation factor = otalOs = AEE (9.11)
bm b
Cross modulation is specified as the level of the 20 kHz off-tune
signal, 30% modulated at 1 KHz, that produces an output signal 30 dB
down on the level produced by the wanted signal, 30% modulated
at 1 kHz, at 60 dBuV e.m.f.
Since both cross modulation and third-order intermodulation arise
because of the cubic term in a non-linear characteristic it is to be
expected that there is a relationship between them. Approximately,
this relationship is
SCMR
Voy(dBuV) = 2Vsp(dBuV) — 6(dB) — dB. (9.12)

Blocking Blocking is an effect in which the gain of a radio receiver is reduced

when a large-amplitude unwanted signal either overloads a stage, or
excessively operates the receiver’s a.g.c. system. When blocking
occurs the level of the wanted output signal falls each time the inter-
fering signal is received. The blocking ratio is the ratio
(response to a signal at one frequency when there is a
simultaneous excitation at another frequency)/(response
to the one signal above). (9.13)

The blocking ratio of a receiver depends both upon the magnitudes

of the two signals and on their frequency difference. It is specified
as the level of an interfering 20 kHz off-tune signal that gives a change
in the wanted output signal of 3 dB, with the a.g.c. system of the
receiver inoperative.
If the level of the two signals, in dByV, to give | »V inter-
modulation product is V3zp then, approximately, the voltage Vg of the
blocking signal is
Vz = 2V3p(dBuV) — 3(GB). (9.14)

If, for example, V31p = 82 dByV the level of the blocking signal will
be + x 82 — 3 = 120 dByV or | V. Because of the high levels
200 RADIO SYSTEMS TECHNOLOGY

involved it is quite possible that the fifth-order intermodulation

products may also be of significant amplitude; if so, Vg =
2Vspp = 4.5 dByV.

Sensitivity The sensitivity of a communication radio receiver is the smallest input

signal voltage that is required to give a specified output power with
a specified output signal-to-noise ratio. It is necessary to include signal-
to-noise ratio in the definition because otherwise the output could
consist mainly of noise and be of little use. The lower limit to the
sensitivity of a receiver is set by the input thermal noise plus some
contribution from the noise factor of the receiver.
For an amplitude-modulated receiver typical figures might be:
(a) 2 pV sensitivity with 30% modulation at 400 Hz and 20 dB
signal-to-noise ratio (h.f. receiver);
(b) 1 pV sensitivity with 50% modulation at 1 kHz for 12 dB
SINAD (p. 203) and 7.5 kHz selectivity (v.h.f. receiver); and
(c) —99 dBm sensitivity with 30% modulation at 1 kHz for 12 dB
SINAD and 8 kHz selectivity (h.f. receiver).
Ans.s.b. receiver will have a better sensitivity than a d.s.b. receiver
because there is no carrier power and the bandwidth is narrower.
Typically, sensitivity = 0.5 wV with 1 kHz output for 12 dB SINAD
and 3 kHz selectivity.
For a frequency-modulation receiver the sensitivity is quoted with
a specified r.m.s. or peak frequency deviation. Typically, this might
be sensitivity 0.5 uV with 2.1 kHz r.m.s. frequency deviation (this
is 3 kHz peak deviation) for 12 dB SINAD and 15 kHz selectivity.
If a receiver employs one, or more, tuned r.f. Stages its sensitivity
will vary with frequency as shown by Example 9.4.

Example 9.4

A radio receiver can be tuned to receive signals in the frequency band 4 to

20 MHz with the tracking error given by Table 9.1. At 20 MHz the sensi-
tivity of the receiver is 2 wV. Calculate its sensitivity at (a) 4 MHz,
(b) 8 MHz, and (c) 14 MHz and plot the sensitivity curve of the
receiver.
Assume the r.f. stage to have a Q-factor of 50.

Table 9.1
Eb ns nels a 2 OS I ae
Signal frequency (MHz) 4 8 14 20
Tracking error (kHz)
80 20 100
ee ee eee
0

Solution
When the receiver is tuned to any particular frequency any tracking error
appears in the r.f. stage.
 COMMUNICATION RADIO RECEIVERS 201

(a) When the wanted signal is at 4 MHz the r.f. stage is tuned to
4.02 MHz. Hence
R
sensitivity 2xR :
Ali i aie is
507(40 x 10%)?
2 ff+ aS: |pV = 2.23 pV.
(Ans.)
(b) The r.f. stage is tuned to 8.08 MHz. Hence

ory) 507(160 x 10%)?

sensitivity = 2 ae = 2.81 pV.
(8.08 x 10%)?
(Ans.)
(c) The r.f. stage is tuned to 14.1 MHz. Hence
ie 507(200 x 10%)?
sensitivity = 2 eet = 2.45 pV.
(44x 10°)2
(Ans.)

The sensitivity curve of the receiver is shown plotted in Fig. 9.5.

3.0

Ss
=
>
2 2.5
2
®
”

2.0
0 4 8 12 16 20 24
Fig. 9.5 Sensitivity curve of a radio
Frequency (MHz)
receiver.

Selectivity The selectivity of a communication radio receiver is its ability to select

a wanted signal from all the signals that are simultaneously present
at the aerial terminals and to reject all other signals. Selectivity may
be quoted graphically, showing the output of the receiver in decibels
relative to the maximum output, plotted against frequency off-tune.
Alternatively, several points on such a graph may be quoted; e.g.
6 dB down at 3 kHz bandwidth, 60 dB down at 12 kHz bandwidth.
Most of the selectivity of a radio receiver is provided by the i.f. filters.
The 6 dB and 60 dB bandwidths are known, respectively, as the nose
and the skirt bandwidths. The nose bandwidth is the band of frequen-
cies over which a signal can be received with little loss of strength.
 202 RADIO SYSTEMS TECHNOLOGY

The skirt bandwidth is the band of frequencies over which it is possible

to receive a strong signal. The shape factor is the ratio (skirt
bandwidth)/(nose bandwidth) and typically it is about 4.
The adjacent-channel ratio of a receiver is the ratio, in decibels,
of the input voltages at the wanted, and at the adjacent-channel
frequencies, necessary for the adjacent channel to produce an output
power of 30 dB down on the wanted signal power.
In modern h.f. communication receivers the selectivity is provided
mainly by crystal filters, whereas v.h.f./u.h.f. receivers often employ
SAW filters. With the use of either of these types of filter it is possible
to achieve almost any desired selectivity.

Noise Factor The output of a radio receiver must always contain some noise, partly
because the input signal is not noise free and partly because the
receiver itself generates some noise. The noise factor F of a radio
receiver is a measure of the degradation of the input signal-to-noise
ratio caused by the receiver, i.e.
Fe= input signal-to-noise ratio
: (9.15)
output signal-to-noise ratio
At all frequencies up to about 30 MHz the noise picked up by an
aerial is generally larger than the noise internally generated by the
receiver. There is then little to be gained by the receiver having a low
noise factor and often an r.f. amplifier is not provided. At higher fre-
quencies the aerial noise is much smaller than the internally generated
noise and then a low noise factor is advantageous. Consequently,
v.h.f./u.h.f. receivers always employ r.f. gain.
The noise factor of an m.f./h.f. radio receiver is typically in the
region of 10 to 12 dB; v.h.f./u.h.f. receivers have a typical noise
factor of 8 to 10 dB but, in some cases, it may be only about 3
dB.
Since the sensitivity of a radio receiver is defined in terms of a
specified output signal-to-noise ratio it is evident that sensitivity and
noise factor are related. If the sensitivity of a receiver is N nV for
an output signal-to-noise ratio of S (as a ratio) in a bandwidth of B Hz,
then the noise factor of the receiver is
F = 61 + 20 logi9 N — 10 logio (S — 1) — 10 logio B dB.
(9.16)

Example 9.5
A radio receiver has a sensitivity of 1.5 LV for an output
signal-to-noise ratio
of 20 dB in a bandwidth of 3 kHz. Calculate its noise
factor.

Solution
From equation (9.16)

F.=6hest, 20slog,9- 15: —10 logig 99 — 10 log; 3000 = 9.8 dB.

(Ans.)
 COMMUNICATION RADIO RECEIVERS 203

SINAD Ratio

Sometimes it is more helpful to consider the total distortion at the

output of a receiver as well as the noise. The SINAD ratio is given
by equation (9.17), i.e.
(signal power + noise power +
distortion power)
SINAD = (9.17)
(noise power + distortion power) ;
The sensitivity of a radio receiver is often specified in terms of
SINAD.

Stages in a Radio Receiver The Radio-frequency Stage

The radio-frequency stage, or front end, of a communication radio

receiver has several functions to perform.
(a) It must couple the aerial to the receiver in an efficient manner.
A wide variety of coupling circuits can be used and Fig. 9.6
shows just a few of them. The tuning capacitances will probably
be provided by voltage-tuned varactor diodes.

Fig. 9.6 Some r.f. coupling circuits.

(b) It must suppress signals at the image-channel and intermediate

frequencies. For less stringent requirements tuned circuits will
be able to provide sufficient selectivity ,but for optimum perfor-
mance a bank of sub-octave filters will be necessary. The bank
of filters will divide up the tuning range of the receiver into
at
sub-bands and any one of the filters is switched into circuit
a given time. Some h.f. receivers which employ up-conve rsion
may use only a single 30 MHz low-pass filter. Ina v.h.f./u.h.f.
receiver the band-pass filters may be helical resonators. When
 204 RADIO SYSTEMS TECHNOLOGY

filters are used to obtain the r.f. selectivity any r.f. amplifiers
employed will be wideband circuits.
(c) Older-type communication receivers tended to employ an r.f.
amplifier at frequencies in excess of about 5 MHz. The modern
tendency is not to employ r.f. gain below about 30 MHz,
because little, if any, improvement in the noise performance
of the receiver results, and an r.f. amplifier is a source of both
cross modulation and intermodulation.
When an r.f. amplifier is employed it must be designed to have
a low noise factor, to operate linearly for even the strongest anticipated
input signals, and to generate the minimum intermodulation products.
If the amplifier is a broadband type it will be susceptible to second-
order intermodulation products as well as to third-order ones. The
gain of the r.f. amplifier may be varied by the a.g.c. system of the
receiver, or the gain may be constant and the amplifier preceded by
an a.g.c.-controlled r.f. attenuator.

The Mixer Stage

The function of the first mixer stage is to convert the wanted signal
frequency into the first intermediate frequency. Similarly, the function
of the second mixer is to convert the first intermediate frequency into
the second intermediate frequency. Double-balanced mixers are
increasingly used for several reasons.

(a) They give a high degree of isolation between the first local
oscillator and the r.f. stage which minimizes unwanted radiation
from the aerial.
(b) The local oscillator voltage fed into the i.f. filter is suppressed
by at least 30 to 40 dB and this reduces noise.
(c) The even-order intermodulation performance is good.

Probably the most commonly employed discrete-component double-

Oscillator
Fig. 9.7 Diode ring mixer.
voltage
 COMMUNICATION RADIO RECEIVERS 205

balanced mixer is the diode ring, shown in Fig. 9.7, largely because
it has a large dynamic range. Ring mixers are commercially available
as a complete package ready for use. The circuit does have a number
of disadvantages; these are: (i) the local oscillator power must be fairly
high; (ii) the circuit has a conversion loss of at least 6 dB; (iii) the
intermodulation performance depends upon the source and load
impedances; and (iv) the rejection of the carrier component depends
upon the balance of the circuit. Most of the difficulties can be over-
come with the use of a transistor-tree balanced mixer whose basic
circuit is shown in Fig. 9.8. This circuit is employed in several ICs
and an example is the Plessey SL 6440.

Output

Oscillator
voltage

is
ignal voltage

Fig. 9.8 Transistor-tree balanced

mixer.

The Local Oscillator

The first local oscillator must be capable of tuning to any frequency

in the tuning range of the receiver plus the first intermediate frequency.
The second local oscillator has to provide only one frequency, this
is equal to the sum of the first and the second intermediate frequencies.
The first local oscillator must have: (a) high spectral purity,
(b) frequency agility, so that it can quickly change frequency;
(c) small increments offrequency (generally, for the frequency bands
below 30 MHz a frequency resolution of 1 to 100 Hz is required,
with s.s.b. receivers no worse than 10 Hz; v.h.f./u.h.f. receivers
usually have a quoted frequency resolution of about 1 kHz); and
(d) frequency stability.
The above requirements are difficult to satisfy with an LC oscillator
and if the receiver is to operate at a few fixed frequencies a crystal
oscillator with switched crystals may be used. Increasingly, modern
 206 RADIO SYSTEMS TECHNOLOGY

communication receivers employ a frequency synthesizer to generate

the two local oscillator frequencies. A frequency synthesizer is able
to generate a large number of precise frequencies which are derived
from a single, stable, reference source. The disadvantage that is
introduced by the use of a frequency synthesizer is that the tuning
of the receiver is not continuous but can only occur in discrete steps.

Frequency Synthesizers

A frequency synthesizer is a circuit that derives a large number of

discrete frequencies, singly or simultaneously, from an accurate, high-
stability crystal oscillator source. Each of the derived frequencies has
the accuracy and the stability of the reference source. A frequency
synthesizer must be able to cover a wide frequency band so that the
receiver can work over the whole of the tuning range. Most of the
frequency synthesizers employed in modern communication receivers
are of the phase-locked loop (p.1.1.) type and Fig. 9.9 illustrates the
basic concept. The system consists of a very stable crystal oscillator
which acts as the reference source, a phase detector, a low-pass filter,
and a voltage-controlled oscillator (v.c.o.). The phase detector
produces an output d.c. control voltage, the magnitude and polarity
of which is determined by the phase difference between the crystal
oscillator and v.c.o. voltages. The control voltage is filtered to remove
all a.c. components and it is then applied to the v.c.o. to vary its
frequency. The action of the p.1.1. ensures that the frequency of the
v.c.o. changes in the direction that reduces any difference between
the crystal oscillator frequency and the v.c.o. frequency. Once lock
has been achieved the two inputs to the phase detector are at the same
frequency, but there is always a phase difference between them in
order to maintain the controlling d.c. voltage.

Reference
Phase Voltage- Output
crystal Low-pass
detector controlled
oscillator filter
oscillator

Fig. 9.9 Phase-locked loop.

To obtain more than one output frequency a frequency divider,

which may be a programmable type, must be connected in the position
shown in Fig. 9.10(a). The signals applied to the phase detector
are
then at frequencies of fg and fc/N,, and the v.c.o. runs at a frequenc
y
of fo = Nifg. The p.1.l. with a frequency divider in the loop
allows
a large number of frequencies to be obtained by altering the
division
ratio N,. Each of the possible output frequencies is an integral
multiple of the reference frequency. If, for example, fp =
1 MHz
and N; = 3 the v.c.o. frequency will be 3 MHz, but if N,
= 20 the
frequency of the v.c.o. will be 20 MHz. Clearly, the incremen
ts in
frequency that can be obtained are equal to the reference
frequency
 COMMUNICATION RADIO RECEIVERS 207

Crystal: Phase lows Voltage- oOutput fo = N,f,

reference Aistacs ae poe controlled ile
oscillator Serer Mier oscillator

FIN, Programmable
divider N,

(a)

Crystal i
oy eke Phase Low-pass ids Soe - Output fo = N, fa/N2
reference divider datos filt controlled
oscillator No ae oscillator

Programmable
divider N,

Fig. 9.10 Frequency synthesizers:

(a) using a p.l.l., and (6) with
improved frequency resolution. fx. To improve the frequency resolution the reference frequency can
also be divided, as shown by Fig. 9.10(b). If, now, fp = 1 MHz and
N>, = 100 the output frequency will be fo = Nifp/N2 with a
frequency resolution of 1 kHz. The method is relatively simple, fully
digital, and can be integrated, but it is slow to change from one
frequency to another.
To obtain both rapid frequency changes and small frequency resolu-
tion a multiple-loop frequency synthesizer is often employed, and Fig.
9.11 shows one example of the technique. The top loop produces an
output at frequency frN,/N, and this is divided down by the ratio
N, to give frequency fxN,/N,N>. The lower loop produces an output
at frequency fgNp/N, and this is applied to a mixer, along with the
output of the system at frequency fo. The band-pass filter selects the
difference frequency fy — fgNp/N; so that the inputs to the phase
detector C are at frequencies fpNa/NiN> and fo — faNp/N1; these
are locked by the output loop to become equal to one another. Then
N (9.18)
f= aah oi Ng.
N, | No
Example 9.6

In the frequency synthesizer of Fig. 9.11fg = 1 MHz, N, = 10 and N, =

100. Determine the range of output frequencies of the synthesizer if N, is
variable from 200 to 300 and Ng from 350 to 400.

Solution
From equation (9.18)
208 RADIO SYSTEMS TECHNOLOGY

oscillator

falN, Phase
detector Output f
C

f= fo—faNpIN,
Phase
detector
B

Fig. 9.11 Multiple-loop frequency

synthesizer.

1 x 10° { 200
Gar + 350] = 35.2 MHz
10 100
1 x 10° [ 300
SR 2ee + 400} = 40.3 MHz.
10 100
Therefore, the range is from 35.2 MHz to 40.3 MHz. (Ans.)

Ganging and Tracking

Older types of communication receiver employ mechanical control

of the tuning of the receiver. The rotation of a tuning-control knob
simultaneously varies tuning capacitors in both the r.f. stage and the
first local oscillator. The tuning capacitors are mounted on a common
spindle and are said to be ganged. The maintenance of the correct
frequency difference (the first intermediate frequency) between the
frequencies to which the r.f. stage and the oscillator are tuned is called
the tracking. Usually, identical capacitors are used with different
values of inductance in each circuit. Tracking errors are inevitabl
e
and they result in a variation of both the sensitivity (see Example
9.4)
and the image-channel rejection of the receiver. More modern
radio
receivers replace the tuning capacitors with voltage-tuned varactor
diodes; this allows rapid frequency changing to take place,
often under
microprocessor control. Modern communication receivers
often
employ a frequency synthesizer and up-conversion and are
then able
to avoid tracking problems by using either no r.f. selectivi
ty at all
or a bank of switched band-pass filters.
 COMMUNICATION RADIO RECEIVERS 209

Intermediate-Frequency Amplifier Stages

The function of the i.f. amplifier in a communication receiver is to

provide most of the gain and the selectivity of the receiver. Each i.f.
amplifier stage must shape and select a relatively narrow bandwidth
at the mixer output and reject adjacent-channel signals. The second
i.f. stage must amplify the signal to the level necessary for the detector
to operate satisfactorily. In the past, double-tuned coupled circuits
were commonly employed to provide a desired loss—frequency
characteristic but, since they have a poor shape factor, they are rarely
used today. Most modern communication receivers employ either
crystal filters or SAW filters, to give the desired selectivity. These
filters offer the considerable advantages of requiring no i.f. ampli-
fier alignment, and having a selectivity that is not affected by the
application of a.g.c. to the stage.

Crystal Filters

Standard crystal filters (see Radio Systems for Technicians) are readily
available at a number of fixed frequencies, e.g. 100 kHz, 1.4 MHz,
10.7 MHz and 35.4 MHz, with a bandwidth of between 0.01% to
1% of the centre frequency. The insertion loss of a crystal filter is
between 1 dB and 10 dB, the shape factor is very good and the
generation of spurious responses is small.

SAW Filters

A surface acoustic wave (SAW) filter is a four-terminal structure that

has a pair of comb-like transducers deposited onto a piezo-electric
substrate (see Fig. 9.12). The input transducer converts an electrical
signal into a surface acoustic wave, while the output transducer
converts the acoustic signal back to electrical form. The ‘fingers’ of
each transducer are spaced apart by a common distance d. When an
input voltage is applied to the SAW filter a surface acoustic wave
is excited that propagates along the substrate. The maximum excita-
tion is obtained when the comb spacing is equal to the wavelength
d of the signal. Half of the acoustic power is transmitted to the absorber
and is completely lost, the other half is sent in the opposite direction
towards the output transducer. This means that the minimum loss of

Input Output
Absorber

Absorber
Substrate

= =

Fig. 9.12 SAW filter. SAW

 210 RADIO SYSTEMS TECHNOLOGY

the SAW filter is 6 dB. When the surface acoustic wave passes through
the output transducer it is converted into an electrical signal.
When the frequency of the signal is not at the nominal centre
frequency, the comb spacing is no longer equal to the acoustic signal
wavelength and the electric/acoustic conversion efficiency is reduced.
The roll-off of the conversion efficiency is very rapid and so a highly
selective loss—frequency characteristic is obtained. The characteristic
can be tailored by the manufacturer by varying the length and/or the
number of the comb teeth and/or their spacings in each transducer.
Like crystal filters SAW filters are offered by the manufacturers at
a number of set frequencies, e.g. 45 MHz, 100 MHz and 405 MHz,
with a bandwidth of between 1% and 10% of the centre frequency.
Their insertion loss varies from 6 dB to about 28 dB. The SAW filter
is frequently used to provide i.f. selectivity because it is small,
lightweight, very reliable and it requires no adjustments.

The Detector Stage

The function of the detector stage is to recover the information

modulated on to the received carrier, and often also to generate the
a.g.c. voltage. Most a.m. receivers that employ discrete circuitry in
the detector stage still use the diode detector because of its simplicity
and its good performance. The main problem is that the input signal
level must be several times larger than the threshold level of the diode,
otherwise considerable signal distortion will occur. The demodula-
tion of an s.s.b. or c.w. signal requires the use of a product detector.
Very often the detection process is carried out within an IC that also
provides a number of other circuit functions. Frequency-modulation
receivers tend to use the ratio detector in discrete-component designs,
and either the quadrature detector or the phase-locked loop detector
if ICs are employed.

Automatic Gain Control The amplitude of the wanted carrier that appears at the input of a radio
receiver may fluctuate widely, by perhaps 100 dB or more. Automatic
gain control (a.g.c.) is applied to a receiver to maintain the carrier
level at the detector input at a more or less constant value. The level
chosen is such that overload of, and consequent intermodulation
product generation in, the final r.f. stage and/or the first mixer is
minimized. The application of the a.g.c. voltage is distributed over
a number of stages. Usually, the gain of the i.f. stages is reduced
first and the r.f. gain is only reduced when the level of the input signal
is large enough to ensure a good output signal-to-noise ratio. The basic
idea of an a.g.c. system is illustrated by Fig. 9.13. A d.c. voltage
is generated in the detector stage (or in a separate a. g.c. generator)
that is directly proportional to the amplitude of the carrier at the input
to the detector. This d.c. voltage is applied to each of the controlled
 COMMUNICATION RADIO RECEIVERS 211

First i.f. Second Second if. Detector

amplifier nai mixer ee amplifier

Auxiliary (derived from either the Manes:

a.g.c. detector or the first i.f. amplifier)

Fig. 9.13 Application of a.g.c. toa

sagio Fecalver: stages to vary their voltage gains. If the carrier level should rise the
a.g.c. voltage will also increase and will reduce the gain of each
controlled stage. This will, of course, reduce the overall gain of the
receiver and so tend to restore the carrier level at the detector input
to its original value. Conversely, if the carrier level should fall the
a.g.c. system will increase the overall gain of the receiver.
Whenever the r.f. input signal is large and is likely to cause over-
loading of one, or more, stages the auxiliary a.g.c. will come into
action and reduce the gain of the r.f. stage. If the a.g.c. voltage is
used to vary the gain of an r.f. amplifier problems may arise with
regard to both its dynamic range and the production of distortion
because of shifts in the operating point of the amplifier(s). An alterna-
tive, that overcomes these problems, is the use of an a.g.c.-controlled
r.f. attenuator. The r.f. attenuator may either have a continuous loss
that is varied by the a.g.c. voltage, or have fixed values of loss that
are switched into, or out of, circuit by the a.g.c. voltage. The r-f.
attenuators are often fitted in front of, and in between, the stages of
r.f. gain as shown by Fig. 9.14.

rat: rf. fei A:

attenuator feamplifier onattenuator amplifier

a.g.c. voltage

Fig. 9.14 Use of r.f. attenuators to All a.g.c. systems have an inherent delay in their response to a
vary the gain of a receiver. change in the input signal voltage. It is not desirable for the a.g.c.
system to have too rapid a response or it will respond to impulsive
noise. The attack time is the time taken for the a.g.c. voltage to rise
to a predetermined percentage of its final value when the carrier level
falls. The decay time is the time taken for the a.g.c. voltage to fall
to a predetermined percentage of its original value when the carrier
level rises. The choice of the attack and delay times depend upon the
kind of signal being received.
(a) For a d.s.b.a.m. signal the constant-frequency carrier can be
used to generate the a.g.c. voltage. The attack and delay times
need only be fast enough to allow the a.g.c. system to respond
to fading, but slow enough to avoid it responding to low-
frequency modulation. Typical figures are in the region 0.1 s
to 0.2 s.
 212 RADIO SYSTEMS TECHNOLOGY

(b) When an s.s.b. signal is received the absence of a carrier means

that the a.g.c. voltage must be derived from the peak signal
level. The a.g.c. system must therefore be able to respond
quickly when a modulated signal appears. To prevent a transient
occurring at the end of each syllable of speech the gain of the
receiver must only slowly increase at the end of each syllable.
If the attack time is too long the system may not be able to
follow rapid fades but, on the other hand, if the attack time
is too short each new syllable will be accompanied by a ‘roar-
ing’ sound. Typically, the attack time should be from 2 to 10 ms
and the decay time 500 ms to 1 s.
There are three ways in which the a.g.c. system for an s.s.b.
receiver can be improved. These are: (i) the use of a pilot carrier
(which would also be useful for automatic frequency control), (ii) the
use of hang a.g.c., in which envelope-derived a.g.c. is sustained for
about 0.6 s after the signal has fallen to zero, and (iii) a combination
of envelope a.g.c. with a fast-acting squelch circuit that operates
during the inter-syllable intervals.
The performance of an a.g.c. system is quoted in manufacturer’s
literature in the form, ‘less than 6 dB change in output voltage for
90 dB increase from threshold’. Many frequency-modulated receivers
are not provided with an a.g.c. system but instead rely solely upon
limiting to keep the detected output signal at a constant level.

Communication Receivers The most important parameters of a communication receiver, or a

transceiver, are its selectivity, its sensitivity and its dynamic range,
since these determine both the noise factor and the intermodulation
performance. Transceivers must be designed to ensure that none of
the power transmitted can be received by its own receiver. To make
this requirement easier to satisfy different frequencies are usually
employed for reception and for transmission. Also, the noise generated
within the transmitter should not increase the level of the received
noise. Low-power transmitters often have several stages of r.f. filter-
ing before the r.f. power amplifier output stage to reduce off-tune
noise.
Figure 9.15 shows the block diagram of the signal circuitry of the
Eddystone 1650 m.f./h.f. communication radio receiver, which
operates over the frequency range of 10 kHz to 30 MHz in5 Hz steps.
The receiver can operate with d.s.b. amplitude-modulated, s.s.b.
(both
lower- and upper-sideband), and c.w. signals. Six different second
i.f. 6 dB bandwidths are provided, namely 300 Hz, 1 kHz,
2.4 kHz,
3 kHz, 8 kHz and 16 kHz, by switching into circuit the appropri
ate
band-pass filter. The r.f. a.g.c. voltage is generated by an
a.g.c.
generator IC and this voltage is used to vary the loss of the pin
diode
r.f. attenuator. When the product detector is switched into
circuit an
audio a.g.c. generator is used to produce the a. g.c. voltage
for the
i.f. amplifier.
 COMMUNICATION RADIO RECEIVERS 213

Joyydwe

Joyydwe
4 =

urew
—v08' by

(‘yquds)

oipne
eury

oipne
O-<
ZHW
bb
908'

yuse

\semouseu Ayanoejes
Joydwe

jINOJIO UEYM ,"M'9, POW

UO
,"wW'e, epow
BANOW“MD 404)

(Ajuo
soyenueye
epoip urd
@P Ss/0

suonisod
e6e0n

peyejes
JO
9014)
‘o-6e
ZHW S02'9b

Ul)
JOU
6uyoo.
~
nee

oipny ‘o'6'e
“ys

Joyesoueb <a
euBis

ZHW

©)
Joxiw

—9L2'or
902°92 eynw
soyesedwoo < |
1s}

(‘yyuAs) 02
@P aye6
1eAe7

soyydwe
yndjno

‘o'6'e ‘ue6
ya
peourjeg
4

(uuhs)
OOPL
6'€*
ZH
yenuew 0} jenbe)

sdojeaug
(eGey0a web 41

<I

pue
=

<i
“yep
eS)

pueqepiM pueqmoueu
IEA
Jol
a6e0A

feuondo
40}D9|eSe1d

= | | | | | | | | | | | Y | /'}W OS9} euojsAppy ey,

|
|
| 4oyesoue6

a4
‘268

ZH

<< tb
Your
y1

Joyenueye
+1
Joyyjdwe
06"
| aBeyon
|
feuee
yeuBis 1 ZHW tb

jy6reyjs yGnoiu)
suonisod
feIskuoJoy

Aan eters aa 82 }
PEYydIMS
ZHW
0€
St'6 “bls
pue
g

Asayinog)
jo
auojsAppy
olpey
(‘p17
OIPCJ
“JOAIONOI
*}y
UO!PEOIUNWIWOD
 Joyidue
jeubis Aouanbay02¢—02
Z7HW O22 ZHW
214

026

,
ys
ZHW

OZ¢
w

ZHW

002
~

§=ZHW
OZb ZHW

Joyyjdue
ZHW
026
b0S ZHW

re b0S ZHW Ov8 ZHW 8001 ZHW

Heew= Pe}<H=H
280uonoeful ajqeuny
y!a
Aouenbasy benny) Aguenbey
Ore Sez—-O0Le

8001Ore

9°S)
eouUsIAj\a1 JO}BII19SO

Wo4}
88—€6E

ZHIN
RADIO SYSTEMS TECHNOLOGY

092-08

aug
yi vt
‘ISO
JExi- SIS ZHW Jayydwe
SIS ZHW SIS ZH Joydwe
SLS ZHI
= JOXIN SL ZHW "|"! Joyidwe janaj-mo7 puegapim
Jeiidwie °y1 ZHW OZb
a picaal =] HEH
= KH HEHE
= D—~<H : ‘1 IndjnoZk ZHW
<a ~ ~~ “uae ~N x Se, ~
|< e-¢ |=
008 ZHW yu Jo-o6e 1980 40}e
|\19S04012 El jenuew
ureB saying “dwe
4eyng V7 mz jonu09 \7 anit
sayijdwe JO}E|IIOS
tales!pga
“OS aul
uojoelul | | =| vO0S—2°v0S
S2' ZHW
PES S6'P86—S6
ZHW
Ut OL ZH» sdajs
ul OS ZH} sdajs
puegepin
pos, Jodie
;me ag Z°OL HW pueg-moueN
|H J@A8]
sieuy
©
Of"y1
20h
ZHW
INd\nNo
ZH¥ = ‘qm Pajsajapyndjno
=“mq
one ate Z°OL ZHW“}1 siayidwe ase imac
=" > -yyeg: ZHAOL L ZH
°

P |= | “
= ‘
ow! ; X S
o%od fe
Se}
xf "MOL ainw
2
91e6
urey, olpne
“dwe
‘we'gm

wy rel= HSLwp ‘wo ay “q's’sq°u wy j e

ye ure6
yeeq ep O— wequ eur]
oipne ‘dwe
/

auy Jeng]
Y <Y

yu
A

aul)
o'6'e
c1-0
d/O

oo

JAXIWy
ZHY
SSP
Jaue9 uOIesuU!
“9SO

JO}EI|INSO
Jojilul|
‘Wy
qru,
ain ZH4 86 FZHW Z'OL

pueJoe}ap
sdejs ZH 001 u!

ZHW ZOOL puegeapim

oe “*BI4 91°6 ey, euojsAppz
S661
“Wy pue soyIWUI|
40)98)ap
UTA “yUNn/y uoHFedIUN
OlpesWWOD
“19819084 Asayinoy)
jo auojsAppz
ain

olpey (‘PIT
 COMMUNICATION RADIO RECEIVERS 215

The block diagram of the Eddystone 1995 v.h.f./u.h.f. communica-

tion radio receiver is shown by Fig. 9.16. The blocks marked as ‘bite’
are built-in test equipment oscillators that are provided to aid fault
location. The receiver covers the frequency band 20 MHz to 1.1 GHz;
for frequencies between 470 MHz and 1.1 GHz a front-end down-
converter is used to shift the signal to the frequency band 80 to
393 MHz. A cheaper variant of the receiver covers only the frequency
band 20 MHz to 470 MHz and this does not need the down-converter.
The main part of the receiver has a first i.f. of 515 MHz with a roofing
bandwidth (maximum overall bandwidth) of 6 kHz, and a second i.f.
of 10.7 MHz. The bandwidth of the second i.f. filter can be set in
the range 3 kHz, 7.5 kHz, 15 kHz, 30 kHz or 60 kHz by choosing
one of the five crystal filters, to 250 Hz using a ceramic filter, or
to 600 Hz using a roofing filter. The band-limited output of the second
i.f. filter is then applied to the appropriate detector for the type of
signal being received; this may be a.m., or f.m. narrowband or wide-
band. For amplitude modulation detection may be achieved by an
envelope, or a product detector; n.b.f.m. detection is carried out at
455 kHz and wideband f.m. detection at 10.7 MHz. Audio muting,
or squelch, is provided to reduce noise whilst tuning the receiver from
one signal to another. Complete muting is not used so that low-level
signals are not missed. The muting is signal level-derived for all
reception modes other than f.m. when it is noise derived.
Both the Eddystone 1650 and the 1995 series of communication
radio receivers are operated with microprocessor control and they
provide scanning, sweeping and channel-storage facilities. Each
receiver can be human operated, or computer operated, or operated
from a distance by a remote-control unit.
 Radio Systems

The public telecommunications network of a country is used for the

transmission of speech, telegraphy, data and sound/television broad-
cast signals. All the circuits employed, except for those in the local
distribution network, are routed over multi-channel telephony systems
or high-bit-rate data systems which, in turn, are routed via some
combination of copper cable, optical fibre, or microwave radio-relay
system. Line-of-sight radio-relay systems are extensively used to
provide a wide range of communication services. Their traffic capacity
varies from a few, to several thousand, speech channels using either
analogue or digital techniques. Besides the public network there are
also a number of private telecommunication networks operated by
such organizations as the railways, gas and electricity companies.
International telecommunication networks involve the use of both
copper and optical-fibre cable, both underground and submarine,
terrestrial radio-relay and communications satellite systems. The tradi-
tional application of satellite communications has been the international
telecommunications network but this is increasingly being challenged
by optical-fibre systems. Communications satellite systems cannot
compete effectively with optical fibre but they will continue to carry
a large proportion of the total long-distance traffic for a long time
to come. New satellite applications, including television broadcasting
and point-to-point business services, such as British Telecom’s
SATSTREAM, are emerging all the time.
Land-mobile systems are commonly employed both for private net-
works, such as ambulance, police and taxis, and for communication
via the public switched telephone network (p.s.t.n.). Comprehensive
mobile systems also exist for both air and maritime communications.

Microwave Radio-relay Systems A microwave radio-relay system employs line-of-sight space-w

ave
transmissions in both the u.h.f. and the s.h.f. bands. Nearly
always,
the length of a route is much longer than the maximum possible
distance between two aerials and then a number of radio-relay
stations,
or repeaters, are necessary. The basic idea of a microwave
radio-
relay system is illustrated by Fig. 10.1. The radio signal radiated
by
the transmitting aerial is received by the first relay station
where it
is amplified before it is re-transmitted to the next relay station,
and
so on until the signal arrives at the receiving station. The
r.f. signal
must be amplified before its amplitude has fallen to such
a level that
the minimum required signal-to-noise ratio cannot be
obtained. This
 RADIO SYSTEMS 217

Fig. 10.1 Microwave radio-relay

system. —

factor determines the spacing between, and hence the number of, the
relay stations.
Most of the relay systems presently in use operate in frequency
bands below 10 GHz and employ analogue techniques with frequency
modulation. Analogue systems suffer from a progressive deterioration
in the signal-to-noise ratio with increase in the length of the route
and this limits the performance of a system. The more modern radio-
relay systems employ digital techniques using either quadrature ampli-
tude modulation or some form of phase-shift modulation. Since a
digital signal can be regenerated at each relay station a required signal-
to-noise ratio can be maintained throughout the length of a system.
Table 10.1 lists the frequency bands that are in use for microwave
radio-relay systems in the UK.

Table 10.1

Frequency
(GHz) Name Use

1.7-1.9 2 GHz spur Television and 960 channel

analogue. Spurs and low-
capacity links.
1.9-2.3 2 GHz main 960 channel analogue
(baseband 316—4188 kHz).
3.7—4.2 4 GHz band 1800 channel analogue
5.85-6.425 Lower 6 GHz band ) (baseband 316—8204 kHz).
140 Mb/s digital.
2 x 34 Mb/s digital.

6.425—7.11 Upper 6 GHz band ( 960 channel analogue.

10.7—11.7 11 GHz band 140 Mb/s digital.
14.0—14.5 14 GHz band 2 x 34 Mb/s digital.
Television.

17.7-19.7 19 GHz band 8 Mb/s digital medium-

capacity feeders. 140 Mb/s
digital.

Notes:
A 140 Mb/s system gives 11 520 64 kb/s channels.
The 2 x 34 Mb/s systems give 720 64 kb/s channels.
The 2 x 34 Mb/s systems are used for short-spur and junction connections.
 218 RADIO SYSTEMS TECHNOLOGY

Analogue Radio-relay Systems

The block diagram of one r.f. channel in an analogue radio-relay

system is shown by Fig. 10.2. The input baseband signal is first pre-
emphasized and is then used to frequency modulate a 70 MHz carrier.
The 70 MHz intermediate frequency is then amplitude limited before
it is applied to a mixer together with the output of a microwave oscil-
lator. This oscillator runs at a frequency equal to (f, — 70) MHz,
where f, is the frequency at which the signal is to be radiated from
the aerial. The upper sideband [(f, — 70) + 70] MHz = f, is
selected by the sideband filter and is then amplified by either a
travelling-wave amplifier (t.w.a.) or a solid-state amplifier. The
amplified signal then passes through first an isolator and then another
filter, before it is routed to the transmit aerial via one, or more, circu-
lators. The circulators allow the odd-numbered r.f. channels to be
combined together and radiated as a horizontally polarized wave.
Similarly, the even-numbered r.f. channels are also combined together
and are radiated with vertical polarization. At the receiver the signal
is selected by the appropriate band-pass channel filter and is then mixed
Fig. 10.2 One r.f. channel in an with the output of a microwave oscillator. This oscillator runs at a
analogue radio-relay system. (From frequency fo of (f, + 70) MHz and so the mixer output contains a
British Telecommunication component at (f, + 70) — f, = 70 MHz; this component is selected
Engineering.) by the i.f. filters. The 70 MHz signal is first amplified and equalized

Vertical
(or horizontal) Sequence of channels often
reversed at alternate stations to
Horizontal equalize multiplex losses
(or vertical) Ferrite
r.f. channel
multiplex
circulators

channel Waveguide channel

branching filter branching filter
Ferrite
Ferrite
isolator
isolator
t.w.a. or solid-
state amplifier

mae Microwave
; Seebane
filter
Baseband output
1800-channel ©SCillator Up-
telephony converter i.f. amplifier 1800-channel
— 37 dBr or mixer with a.g.c. telephony — 28 dBr
960-channel 70 MHz 960-channel
telephony — 45 dBr i.f. amplifier telephony — 20 dBr
television at 1 Vee dB
Limiter
Pre- Frequency
emphasis modulator Limiter, demodulator de-emphasis

TRANSMIT TERMINAL REPEATER RECEIVE TERMINAL

 RADIO SYSTEMS 219

before it is demodulated to recover the baseband signal. Finally, the

baseband signal is de-emphasized to obtain the original amplitude
relationships between the low-, and the high-frequency components.
The block marked as c.r.u. is the carrier re-insertion unit. Its func-
tion is to inject a noise-free signal into the signal path whenever the
incoming carrier fades completely. This prevents the a.g.c. system
of the receiver increasing the gain and producing a large output noise.
The development of solid-state Gunn oscillators and their use with
microwave cavities has led to the disappearance in radio-relay
equipment of many low-power microwave thermionic devices. The
travelling wave tube (t.w.t.) or amplifier (t.w.a.), which can provide
a high-power output with a relatively high efficiency is, however,
still employed in both analogue and digital systems. The power output
that semiconductor amplifiers can develop at gigahertz frequencies
is steadily increasing and such circuits are increasingly employed as
the r.f. power amplifier. At present, a solid-state amplifier can only
handle some tens of watts compared with a t.w.a.’s hundreds of watts.
In modern equipment all the other electronic circuitry uses semi-
conductor devices. Microwave Ga or InP ICs have been developed
which can work up to about 30 GHz in microstrip circuitry.

Digital Radio-relay Systems

The main disadvantage of an analogue radio-relay system is that the

noise powers generated in each link are additive and so the signal-to-
noise ratio inevitably decreases with increase in the length of a route.
The problem can be overcome by the use of a digital system, since
a digital signal can be regenerated at each relay station and then the
noise is no longer cumulative. Further advantages arising from the
use of digital techniques are as follows.

(a) The ever-increasing availability of |.s.i. and v.1.s.i. digital

ICs allows cheaper, smaller-sized, equipment to be designed
and used.
(b) The widespread use of ICs reduces power consumption and
increases reliability.
(c) Converting a system from analogue to digital operation can
considerably increase its channel capacity.
(d) The bit error rate for data signals sent over a digital system
is much lower than if the same signals are transmitted over an
analogue system.

Digital radio-relay systems operate in a number of frequency bands

(see Table 10.1) with bit rates of either 2 x 34 Mb/s or 140 Mb/s.
The former can give 720, and the latter 11 520, 64 kb/s telephone
channels. The 140 Mb/s rate has been standardized by the CCIR
because it is compatible with the fourth hierarchical digital multi-
plexing level of 139.264 Mb/s. In the UK the standard digital radio-
 220 RADIO SYSTEMS TECHNOLOGY

relay systems operate in the 11 GHz band and they provide six
140 Mb/s r.f. channels. This gives a capacity of 6 x 1920, or 11 520,
telephone channels.

Digital Modulation
Digital modulation of a carrier may modulate either the amplitude,
the frequency, or the phase of that carrier, but usually some form
of phase-shift keying is employed. With bi-phase p.s.k., or b.p.s.k.,
the modulating signal puts the carrier phase into either one of two
possible states. With quaternary p.s.k., or q.p.s.k., the signal is coded
into the dibits 00, 01, 11 and 10 and these are each represented by
one of the four possible phases of the carrier. A combination of
amplitude and phase modulation, known as quadrature amplitude
modulation (q.a.m.) is also used in some systems.
With b.p.s.k. the carrier is transmitted with the reference phase,
i.e. 0°, to indicate binary 1, and with the opposite phase to indicate
binary 0. Differential b.p.s.k. means that the reference phase is the
phase of the last bit, or dibit, received; it has the advantage that the
receiver does not need an absolute phase reference. With q.p.s.k.
phase shifts of +45°, +135°, +225° and +315° relative to the phase
of the previous symbol represent, respectively, the dibits 00, 01, 11
and 10. This is shown by what is known as a constellation (see Fig.
10.3(a)). A q.a.m. signal has 2” phases, each of which can have
more than one amplitude. Figure 10.3(b) gives the constellation of
a q.a.m. signal with eight different phases and two different ampli-
tudes; this gives 16 different states and it will allow combinations of
four bits, e.g. 0101 or 1010, to be coded.
Figures 10.4(a) and (b) show, respectively, the block diagrams of
the transmitter and the receiver of an 11 GHz, 140 Mb/s, digital radio-
relay system. As with an analogue system, a relay station, or repeater,
consists of the back-to-back linking of a receiver and a transmitter,
and the necessary links are shown by the dashed lines.
The input signal is modulated by a pseudo-random digital sequence
that scrambles its frequency spectrum. The scrambling process is used

Fig. 10.3 Constellations of

(2) q.p.s.k. and (b) g.a.m. signals.

Previous Previous
phase phase

01

(a) (b)
 RADIO SYSTEMS 221

because: (a) it improves the spectral and power distributions of the

signal; (b) it reduces jitter; (c) it makes the system appear trans-
parent; and (d) it provides timing information. The scrambled signal
is split into two 70 Mb/s streams and each of these is applied to a
four-phase modulator. Also applied to the phase modulator is an
11 GHz carrier signal produced by the frequency multiplication of
the output of a 1 GHz oscillator. The output of the modulator is a
four-phase p.s.k. signal with a symbol rate of 70 Mbaud. This signal
is then both amplified and filtered before it is combined with the other
r.f. channels and fed to the transmit aerial. The final stage of amplifi-
cation may be provided with either a t.w.a. or a solid-state circuit.
At the receiver the incoming 11 GHz signal is selected by the appro-
priate channel filter and it is then applied to a balanced mixer. Here
it is mixed with an 11 GHz signal to produce upper and lower side-
bands. The lower sideband, centred on 140 MHz, is selected by the
i.f. filter; it is then amplified, before it is applied to the phase-locked
loop demodulator. The reference carrier for demodulation is obtained
by multiplying the received i.f. signal by 4; this process removes the
phase modulation and gives a reference carrier at 4ffp.
Fig. 10.4 The 11 GHz, 140 Mb/s
Multiplexing the outputs of the six r.f. channels is carried out using
digital radio-relay system, a combination of circulators and horizontal/vertical polarization. This
(a) transmitter and (b) receiver. is shown by Fig. 10.5.
(From British Telecommunication 18 GHz equipment operating at 2 Mb/s and 8 Mb/s is also
Engineering.) employed and is often used to provide Megastream circuits.

Modulator
coder Four-phase Ae Power Channel
Scrambler ad and driver modulator amplifier monitor _ filter
To multiplex
Input and aerial
140 Mb/so
Variable
Oscillator attenuator Low-power
i, filter output
From regenerator” ,
alarm
at repeaters
Transmit
oscillator

(a) To modulator
, at repeaters
i ae
Channel Pre-amplifier Main Regenerator aad
filter Mixer and i-f. filter i.f. amplifier \ gy Combiner
From aerial and
demultiplex
Four-phase Differential Descrambler
coherent decoder
Oscillator
demodulator
filter Signal Eye-
level closure
monitor alarm
Local
oscillator

(6)

+Jitter is the short-term variation of the significant instants of a digital signal from
their ideal position in time. Effectively it is phase modulation of the signal timing.
222 RADIO SYSTEMS TECHNOLOGY

Fig. 10.5 Multiplexing r.f. channels

using circulators. (From British
Telecommunication Engineering.)

Communications Satellite Systems The basic principle of a communications satellite system is illustrated
by Fig. 10.6. The satellite is in an orbit 35 800 km above the earth
in an equatorial plane and it is geostationary. The earth stations are
fully integrated with their national telephone networks and, in Europe,
with each other. The North American earth station transmits on
frequency f, and receives on frequency f,, the European stations
transmit on frequency f; and receive on frequency f,. Depending
upon the system these frequencies are in the bands 4 GHz/6 GHz or
11 GHz/14 GHz. The higher frequency band is used for the up-link
and the lower frequency band is used for the down-link.
Each of the earth stations transmits its traffic to the satellite on the
particular carrier frequency allocated to it in the frequency band

Fig. 10.6 Communications satellite Communications

system. satellite

Atlantic Ocean
 RADIO SYSTEMS 223

5.935—6.425 GHz, or 14—14.5 GHz. This is a bandwidth of

500 MHz and it allows the simultaneous use of a satellite by more
than one earth station. The capacity of a satellite is shared between
its various users using either frequency-division multiple access
(f.d.m.a.) or time-division multiple access (t.d.m.a.). The number
of telephone channels thus provided to a user varies from 12 in a
2.5 MHz bandwidth to 1872 in a bandwidth of 36 MHz. All the
signals transmitted to earth by a satellite are received by every earth
station; each station selects the particular carrier frequencies that it
has been allocated, in the frequency bands of either 3.7—4.2 GHz
or 10.95—11.2 GHz and 11.45—11.7 GHz.
Communications satellites that form an integral part of the inter-
national telephone network are operated by the Communication
Satellite Corporation (COMSAT) on behalf of an international body
known as INTELSAT (International Telecommunication Satellite
Consortium). The capacity of an INTELSAT satellite has increased
considerably since the first satellite was put into service in 1965. This
is shown by Table 10.2.
The Western European countries have launched their own satellite,
known as EUTELSAT, which is employed to carry relatively local
traffic.
Figure 10.7 shows the block diagram of the communication equip-
ment of an INTELSAT V satellite. The satellite has seven receiving
aerials: east and west hemi, east and west zone, and global at 6 GHz,
and east and west spot at 14 GHz. The signals received by each aerial
are amplified and are then applied to a mixer to be shifted to the
common 4 GHz band. The amplified mixer output is then applied to
various band-pass filters to obtain the required transponder band-

Table 10.2
Ge ee eee ee eee ee
INTELSAT Date of Bandwidth
No. first launch (MHz) Capacity
as il net ee bee le ee ee ee en A ee

| 1965 50 240 channel

HI 1967 130 or TV
Ill 1968 300 1500 channel
or TV
IV 1971 500 4000 channel
or TV
IVA 1975 800 6000 channel
or TV
V 1980 2144 12 000 channel
or TV
VA 1985 2250 15 000 channel
or TV
Vi 1987 3300 30 000 channel
or TV
ee
De
 224 RADIO SYSTEMS TECHNOLOGY
4 GHz 11 GHz
4 GHz transmitters up-converter. 4 :
Chi-2 Input filter multiplex switch | 4.5W and 4 Output filter multiplex eu-2
a matrix b transmitters
as Ch1-2
i e.
eal resem Chi-2
ua
x Hv} | xX
oe

4 Sh
Ch1i-2 Chi-2
=
Bo
aw
~J Ch1-2
i.
;

AE
Chi-2
ae ES
Ch1—2 < rote
Be His ep ~
== Chi =2 Ch1-2

bro =
ee mx

anh
NY

|
NS

a
OO ;

~ =F

as |Sea REEF
~»
r t lea brad
~w ro lav)
BY 1 | +4 + + t 21 |
> bh3—2 OS
™N

Ch3— x 8.5
ED
mw freee ths{ olay1 +P}
1
TN]
|
ow
By
Ch3-—4 1 1 Ch3—4)
8.5 W =
ce | ie ow
ay a Pa
h5—6 4.5 W
eae Ch5—
=
x | Sree f iS 2 ~
~I bhs—¢ | 45W |
Shaap Phos eet
res
ye
z
Pe,
a
pow
ining [bearot '
7

4 DI x
Kas iS a
t—
1

Mg

| Ch J

mw
™N
ww

Ch
~w
mw
mw

bh7—<
mw
Q=4 ~
h7-8 \
2 x
my,
}

©) t
1

Chg
AG a
a 4=
x !
Sue
Chg
b
1

©
1

pe
PO

Receive aerial Receivers I.L.c. Ch9

aaa Transmiti aerial
i
i
feed assemblies GH Bes
mies 1 feed assemblies
West zone 6 A B5W | Ch10 @)—+ West zone
(rhc. pol.) L<I4 [> oe) (.h.c. pol.)
East zone 6 85 WwW Ch11 @ » East zone
h pol.)|
(r.h.c. C A L 2 ae
He > BS (I.h.c. pol.)
West hemi i 6 ee nis
85W iid
Chi2 West hemi
(I-h.c. pol.) Leila Pe (r.h.c. pol.)
East Hemi > 6 > > nm East hemi
(I.h.c. pol.) 4
a oe
5 14 S 6)
‘est spo
cation ss eS (linear pol.)
; — pol.: polarization . Channel
East spot we H{>+-+6) se
r.h.c.: right-hand circular S: switch (eres er
(linear pol.) a I.h.c.: left-hand circular \.l.c.: low-loss coupler
6 ais
(I.h.c.
nanpol.) bs
(r.h.c. pol.)
Fig. 10.7 INTELSAT V satellite: communication circuitry. (From British
Telecommunication Engineering.)
 RADIO SYSTEMS 225

widths. (Transponder is a term used in satellite technology to denote

a wideband channel.) The band-limited signals are then directed to
a switch matrix which is controlled from the ground. Most of the
outputs of the switch matrix are amplified by a t.w.a. Some switch
matrix outputs, however, are first frequency shifted to the 11 GHz
band and then they are amplified, again by at.w.a. The t.w.a.s have
saturated output powers of either 4 W or 8.5 W at 4 GHz and of 10 W
at 11 GHz, but they are normally operated at a lower power level
in order to reduce intermodulation. The output of each t.w.a. is then
applied to the output filter multiplex and hence to one of the seven
transmit aerials.

Multiple Access

Multiple access is the name given to the way in which the traffic-
carrying capacity of a satellite can be considerably increased. It allows
a large number of earth stations to be given access to a satellite and
it may employ either frequency-division, or time-division techniques.
The number of telephone channels made available to an earth station
may be fixed (this is known as fixed access) or the allocated number
may be assigned in response to the traffic demand (this is known as
demand access).

Frequency-division Multiple Access

In frequency-division multiple access (f.d.m.a.) all the earth stations
which share a communication satellite do so at the same time, each
station being allocated its own unique frequency bands. The f.d.m.a.
system is always employed in conjunction with both frequency modu-
lation and frequency-division multiplexing. Each earth station is
allocated one, or more, carrier frequency(ies) and it modulates all
of its out-going traffic, regardless of its destination, on to that carrier.
Every earth station in the network must therefore be able to receive
at least one carrier from all the other earth stations.
Large blocks of telephone channels are allocated to individual earth
stations on a semi-permanent basis. The number of circuits required
on each route can be calculated from traffic studies. For some routes
which carry little traffic only a few circuits are required and this results
in their having a poor availability. On some other routes there may
not be enough traffic to economically justify the provision of even
one circuit. In any case, because the telephone traffic varies with the
time of day it is unlikely that all the allocated channels would be in
use for all of the time. This means that the fixed access version of
f.d.m.a. is not efficient. On the routes with lower traffic density the
transponder’s bandwidth can be divided into a large number of carriers
and a particular channel allocated to each one. This system is known
as single-channel per carrier (s.p.s.c.) working.
226 RADIO SYSTEMS TECHNOLOGY

If demand access is employed a pool of channels is made available

to all the earth stations. A channel is only assigned to a particular
route between two earth stations as the demand arises. The receiving
station is notified which channel is to be used before transmission
starts, and when the connection is no longer required the channel is
returned to general availability status. The control of a demand-access
f.d.m.a. system is exercised by a control channel and computers.
The number of demand-assignment telephone channels needed to
carry a certain telephone traffic can be further reduced by the use
of speech interpolation. A digital speech interpolation (DSI) equip-
ment monitors the telephone conversations and fills any gaps in the
speech with speech from other calls. The activity is monitored by the
DSI equipment and every time a speaker pauses the equipment may
take away the channel and assign it to another call. A speaker will
not be disconnected unless the channel is wanted for another active
call. When the speaker speaks again a new channel is assigned to him.
Since, on average, more than 50% of a conversation consists of silent
intervals, the circuit capacity can be at least doubled.

SPADE
A digital system, known as single-channel per carrier p.c.m. multiple-
access demand-assignment equipment (SPADE) is a demand-
assignment system which uses a separate r.f. carrier for each telephony
channel. The bandwidth of the satellite’s transponder is divided into
800 channels which are associated in pairs to give 400 two-way
circuits. Control of the setting-up of a connection between two earth
Stations is vested in the demand assignment and switching unit
(d.a.s.s.) that is provided at each station. The necessary communica-
tion links between d.a.s.s. units is provided by a common signalling
channel which is shared between all the earth stations on a time-
division basis.
When a call request is received the d.a.s.s. unit selects a pair of
carrier frequencies and informs the destination earth station that an
incoming call is imminent and which frequencies are to be used. The
analogue signal is then applied to a p.c.m. encoder to produce a
56 kb/s p.c.m. signal. This signal is then placed into a memory and
is read out at the higher rate of 64 kb/s. This leaves some gaps in
the signal which are filled, using a process known as bit stuffing, with
preamble and start-of-message synchronization bits. The 64 kb/s signal
is then applied to a four-bit p.s.k. modulator to produce a q.p.s.k.
signal. This q.p.s.k. signal is processed by the earth station’s
transmitting equipment and is radiated by the aerial.

Time-division Multiple Access

Time-division multiple access (t.d.m.a.) is a technique which allows
a number of earth stations to have access to a common satellite
trans-
 RADIO SYSTEMS 227

Fig. 10.8 Time-division multiple Communications satellite

access.

ponder at different times. Each earth station takes its turn to transmit
data through the transponder for a small fraction of the total time,
as shown by Fig. 10.8. The bursts of data from each of the n users
arrive at the satellite in a pre-assigned sequence. The bandwidth
occupied by a burst depends upon the type of modulation used and
it may, or may not, occupy the whole of the bandwidth of the trans-
ponder. As long as each earth station is able to maintain the correct
instants for its bursts to start and to finish its bursts will not overlap
the bursts generated by other earth stations. In the INTELSAT systems
the p.c.m. digital signals that are transmitted have an 8 kHz sampling
rate so that the pulses are 125 ws wide. The frame size is 2 ms and
hence each traffic burst includes (2 x 10 Wes 105°) = 46
samples from each of the telephone channels being transmitted.
The basic block diagram of the equipment needed at each t.d.m.a.
earth station is shown by Fig. 10.9. The input digital data is fed into
a buffer store at 12 Mb/s and is read out of the store at the much
higher rate of 120 Mb/s. Since the output bit rate is much higher than
the input bit rate it is clear that the data can only be transmitted in
short bursts. The bursts are repeated every 2 ms and for the same
input and output information 12 x 10° = (120 x 10°T)/(2 x 10:73)
and so the time duration T of a burst is 200 us. Before it is transmitted
each traffic burst must have a preamble added; the preamble is the
name given to a number of bits which are added to the traffic burst
to enable a distant earth station to correctly receive the signal.
Preamble bits are generated in the t.d.m.a. equipment and added to
228 RADIO SYSTEMS TECHNOLOGY

Fig. 10.9 Earth-station t.d.m.a. 420 Mb/s bursts t.d.m.a. terminal

equipment. (From British 12 Mb/s Burst i.f. to
200 us long
Telecommunication Engineering.) From constant
every 2ms transmitter
terrestrial Burst Modulator
network assembler

Transmitter
preambles

Preamble
generator

12 Mb/s
To constant Burst i.f.
terrestrial Preamble from
network receiver receiver

the traffic burst by the burst assembler. The assembled burst is then
applied to a q.p.s.k. modulator to produce a burst at the intermediate
frequency. The i.f. signal is then applied to the transmitting equip-
ment to be frequency shifted to the allocated frequency band and it
is then amplified before it is radiated from the aerial. A receiving
earth station must demodulate the received carrier and then recover
the synchronization pulses before it will be able to identify the
beginning of each frame. It will then be able to assemble the original
telephony signal and pass it on to the destination.
A t.d.m.a. system must be synchronized to make sure that the traffic
bursts originating from different earth stations do not overlap one
another. The necessary synchronization is obtained by the use of
reference bursts. The reference bursts are transmitted by some of the
earth stations and they are received by all of the earth stations in the
network. Each earth station must then ensure that its traffic bursts,
with added preamble, are transmitted to the satellite at fixed times
relative to each reference burst.

Radiation Techniques

The traffic capacity of a communication satellite can still further be

increased by the use of (a) spatial polarization, and (b) polarization
diversity. Spatial polarization means that separate transponders are
employed for (i) wide area coverage, and (ii ) for localized spot
coverage, of the earth. Polarization diversity means that the same
frequency band can be used without mutual interference by two
separate signals if one signal is transmitted using horizontal polariza-
tion and the other signal uses vertical polarization. Alternatively, left-
and right-hand circular polarizations are sometimes used.
 RADIO SYSTEMS 229

Communication Satellite Orbits

Most communications satellites are in orbit at a height of 35 800 km

above the earth in an equatorial plane. At this height the orbiting
satellite keeps pace with the rotation of the earth and so it appears
to be stationary above a particular point on the earth’s surface. This
orbit is therefore known as the geo-synchronous orbit. Satellites in
the geo-synchronous orbit must be sufficiently far apart for inter-
ference from up-links to adjacent (in the orbit) satellites not to be a
problem. This means that the earth stations must use highly directive
aerials, and the CCIR have specified the maximum sidelobe level
which is allowable. The specified aerial gain G is
G = 29 — 25 logig 6 dBi, (10.1)
where @ is the angular distance in degrees between satellites in the
geo-synchronous orbit. Each degree corresponds to a span of about
700 km in that orbit.

Example 10.1
An earth station requires an e.r.p. of 80 dBW in order to obtain the required
signal-to-noise ratio at the satellite. Two parabolic dish aerials are available,
one aerial has a gain of 40 dBi and, hence, a 40 dBW signal, the other aerial
has a 60 dBi gain and a 20 dBW signal. Determine which aerial should be
used if 6 = 2°.

Solution
Since both of the aerials are able to supply the wanted e.r.p. the choice of
aerial is based upon the sidelobe levels of their radiation patterns. The first
aerial will produce an interference e.r.p. of

40 + 29 — 25 logy 2 = 61.5 dBW.

The second aerial produces an interference e.r.p. of

20 + 29 — 25 log, 2 = 41.5 dBW.

Assuming the adjacent satellite also receives a wanted 80 dBW signal the
interference is 18.5 dB down for the first aerial and 38.5 dB down for the
second aerial. Therefore the second aerial should be chosen. (Ans.)

Noise Performance of Analogue An analogue microwave system, be it terrestrial or satellite, carries

Microwave Links a number of 3.1 kHz bandwidth telephony channels that have been
assembled, using frequency-division multiplex, to form a baseband
signal. The number of telephony channels that are active, i.e. carrying
speech, at any instant, and hence the mean baseband power, varies
continuously and reaches its maximum value in the busy hour. (Note
that on some routes the ‘busy hour’ may be as much as 2 or even 2.5
hours long.) If the system is heavily loaded, i.e. most of its channels
are carrying speech, there will be an increase in the level of inter-
modulation noise. On the other hand, if the system is lightly loaded,
230 RADIO SYSTEMS TECHNOLOGY

Fig. 10.10 Showing how the noise-

to-carrier ratio varies with the
frequency deviation.

Noise-to-carrier
ratio

Optimum
Frequency deviation

the mean baseband power will be low and, since the thermal noise
is at a constant level, the carrier-to-noise ratio will fall. This means
that the carrier-to-noise ratio increases with increase in the baseband
signal level, and hence with the number of active channels, up to an
optimum level above which increased intermodulation noise ensures
that the carrier-to-noise ratio falls. In turn, this means that there is
an optimum value for the frequency deviation of the microwave carrier
in the transmitter. This is shown by Fig. 10.10. The CCIR state that
the mean power of a single telephone channel in a system having 240
channels, or more, is — 15 dBm0+ or 32 w»W. The CCIR also give
the mean power of N multiplexed channels as
P = —15 + 10 logy
N dBm0, (10.2)
where N > 240.
If there are fewer than 240 channels then the N channels have a
mean power of
P = —1 + 4 logiy N dBm. (10.3)
A microwave link is set up so that a sinusoidal 1 kHz, 0 dBm0 test
tone produces a standard value of frequency deviation. This standard
frequency deviation is 200 kHz for a 960 channel system and 140 kHz
for a 1800 channel system. The relationship between the test-tone
frequency deviation and the frequency deviation produced by the base-
band signal must be known in order to ensure that the i.f. bandwidth
is wide enough. The i.f. bandwidth must be able to accommodate the
i.f. signal produced by the peak level of the baseband signal.
The signal-to-noise ratio at the output of a telephone channel is the
ratio of the level of the standard test tone (0 dBm0 at 1 kHz) to the
noise in the 3.1 kHz bandwidth. The worst-case telephone channel
is the top channel, i.e. the one with the highest f.d.m. carrier
frequency. The r.m.s. test-tone frequency deviation Siar.m.s.) is related
to the r.m.s. baseband frequency deviation Sac.m.s.) by the loading
factor L, which is given by equation (10.2) or by equation (10.3).

jdBm0 = dBm — dBr. Ifa signal level is measured at any point in a system it can
be related to the zero transmission level point by the relative level in dBr of that point.
 RADIO SYSTEMS 231

The loading factor is also equal to

r.m.s. baseband aaree
L 20 logio oe
r.m.s. test-tone deviation

20 logig I. (10.4)
The ratio of the peak frequency deviation produced by the baseband
signal to the r.m.s. frequency deviation is known as the peak factor p.
For systems with more than about 25 channels the peak factor is
approximately equal to 3.2. The peak baseband frequency deviation
Ffapeak) is equal to pfair.m.s.) = Phac.m.s.)» and hence the bandwidth
occupied by the microwave signal is given by equation

Bos 2(Pidcans.) a Ten) (10.5)

Example 10.2
A microwave carrier is frequency modulated by the baseband signal repre-
senting 1800, 3.1 kHz bandwidth, telephony channels arranged in frequency-
division multiplex. The top channel is at 8204 kHz and the r.m.s. test-tone
frequency deviation is 140 kHz. Calculate (a) the r.m.s. frequency devia-
tion produced by the baseband signal, and (b) the minimum r.f. channel
bandwidth necessary. Assume the peak factor to be 3.16.

Solution
(a) From equation (10.2)

L = —15 + 10 log,) 1800 = 17.6 dBm0.

Therefore,

Sac m.s.) )
17.6 = 20 lo SSS Se
awl x 103
and

fucms) = 7-59 X 140 x 10° = 1.063 MHz. — (Ans.)

(b) From equation (10.5)

Minimum bandwidth = 2(3.16 x 1.063 x 10° + 8.202 x 10°)

= 23.1MHz. (Ans.)

Test-tone Analysis

When a frequency-modulated carrier of peak voltage V, has a single-

frequency noise voltage V, superimposed upon it the carrier will be
both frequency, and phase modulated (p. 36). The peak frequency
deviation of the carrier is given by equation (2.23) as Vifaite Ve,
where fyig¢ is the separation between the carrier frequency and the
interfering signal frequency. In a 1 Q resistance V, = \(KT) and the
carrier power is P = V2/2. Provided the carrier-to-noise ratio
is at least 10 dB, the peak frequency deviation of the carrier is
232 RADIO SYSTEMS TECHNOLOGY

LRT/P Dfsige- Hence, the r.m.s. frequency deviation is given by

kT
Sidmiss a Alera es (10.6)

Because of the triangular noise spectrum (p. 38) the worst-case

channel is the channel that is located at the highest carrier frequency
fmax in the multiplexed baseband signal. If the bandwidth of each
telephony channel is b Hz then the noise power Np at the output of
the top channel is given by

fr kT
No = 2K |ne (sp. fa)df. (10.7)
fmax — 2P.

The factor 2 is necessary because the input noise is at frequencies

both above and below the carrier frequency, and K is the transfer
constant of the frequency detector.
KkT KkTf?,,,0
No0 3P, 4ax ee (f. a
NAMAX. aie b )3} = P.
me

since

3b
ay ne? b)? = reas (:ax rae, ae ca a af
max
The output signal power is proportional to the square of the r.m.s.
frequency deviation of the carrier, i.e. Sy = Kiem ee and so the
output signal-to-noise ratio is

aes = Ficms) BrP,

hte, KTBypf max D

: : neu Bg Jide) ‘ Bir

or output signal-to-noise ratio =
Vinee b

times the carrier-to-noise ratio. (10.8)

Usually, the output signal-to-noise ratio is quoted in decibels. Thus
output signal-to-noise ratio = input carrier-to-noise ratio (dB)

+ 20 logio - B
+ 10 logo a dB. (10.9)
max

Sometimes the frequency deviation of the carrier due to the test

tone is smaller than the maximum baseband frequency and then the
f.m. advantage will be negative. The last term is always positive and
it expresses the fact that a single telephony channel does not occupy
all of the i.f. bandwidth.
The output signal-to-noise ratio is increased by the use of pre-
emphasis in the transmitter and de-emphasis in the receiver. Figure
10.11 shows the pre-emphasis characteristic that is recommended by
the CCIR; it gives a signal-to-noise ratio improvement of 4 dB.
 RADIO SYSTEMS 233

+6

+5

+4
+3

+2

+1

(db)
kHz
200
relative
deviation
to
frequency
r.m.s.

0.1 0.2 1 2 10
0.01 0.02
Normalized frequency (frequency/highest frequency in baseband)

Fig. 10.11 CCIR pre-emphasis Furthermore, the use of psophometric weighting effectively gives
characteristic. another 2.5 dB increase in the output signal-to-noise ratio.

Example 10.3
The carrier power at the input to a radio-relay receiver is — 87 dBW. The
receiver has an i.f. bandwidth of 40 MHz, a noise factor of 10 dB, and the
carrier frequency deviation due to a 1 kHz 0 dBm0 test tone is 200 kHz. The
carrier is frequency modulated by a baseband signal consisting of the
frequency-division multiplex of 960 3.1 kHz bandwidth telephony channels.
Calculate the test-tone signal-to-noise ratio at the output of the top telephone
channel at 4188 kHz.

Solution
F = 10 dB = 10; P, = —87 dBW = 2 nW.

The input noise power is

kTyB = 1.38 x 107? x 290 x 40 x 10° = 160 x 107" W.

The noise at the output of the detector is

FGRT,B =.160 x 10~"G.

Therefore, the carrier-to-noise ratio is
ax 102G
= 1249.4 = 31 GB.
160 x 10°"G
From equation (10.9), the output signal-to-noise ratio is
 234 RADIO SYSTEMS TECHNOLOGY

31 ote
+ 201 oO en SU a 2)0 oreart 6;
510 |4188 x 102 ne Ue |
= 31 — 26.42 + 41.11 + 6.5 = 52.2 GB. (Ans.)

Noise Power Ratio

When a fully loaded multiplex signal is applied to a microwave system

the noise performance of the system is determined by both inter-
modulation noise and thermal noise. The noise power ratio (n.p.r.)
is a measure of the intermodulation noise in a multiplexed system and
it is measured by using a white noise input signal to simulate the fully
loaded multiplex signal. This means that the necessary white noise
power is given by equation (10.2) or by equation (10.3). The method
employed to measure the n.p.r. is shown by Fig. 10.12. Briefly, the
output of the white noise generator is band limited to the i.f. band-
width of the system under test and a notch filter is used to remove
the noise from a single channel before the modulation stage. The only
noise in this channel will then be due to intermodulation and it will
partly occupy this channel. The notch filter can be switched into, or
out of, the circuit to vary the noise in that particular channel. At the
receiver a noise receiver is used to measure the ratio of the noise in
the test channel, with and without the notch filter in circuit. The n.p.r.
is then the difference between the two power levels, quoted in decibels.
Thus, referring to Fig. 10.12, the n.p.r. is equal to 10 log9 P,/P> dB.
Typically, the n.p.r. of a microwave system is in the region of 50 dB.
A measured value of n.p.r. can be converted into a corresponding
value of signal-to-noise ratio by the use of equation (10.10) or equation
(10.11).
Signal-to-noise ratio = n.p.r. + 61 logy N + 4.6 dB
(10.10)
for N < 240 channels.

Signal-to-noise ratio I n.p.r. + 18.8 dB (10.11)

for N > 240 channels.

Fig. 10.12 Measurement of noise

power ratio.

White noise
generator
 RADIO SYSTEMS 235

Example 10.4
Calculate the n.p.r. of a 960 channel system that has a worst-case channel
signal-to-noise ratio of 50 dB.

Solution
From equation (10.11),

50L= Dpsb ise So.

or n.p.r. = 31.2 dB. (Ans.)

Bit Error Rate in Digital Systems The performance of a digital microwave system, be it terrestrial or
satellite, is expressed in terms of its bit error rate (b.e.r.). The b.e.r.
is equivalent to the output signal-to-noise ratio of an analogue system
and it is the probability that a transmitted bit will be incorrectly
received. The b.e.r. is quoted as a number, e.g. 1 x 107° means
that, on average, one bit in every 100 000 will be in error.
An error in the correct reception of a bit may occur because noise
picked up by the system has corrupted the signal waveform to such
an extent that the decision circuitry in the receiver cannot accurately
determine whether a bit is a 1 or a0. The CCITT recommendations
for a digital radio-relay link are: (a) a b.e.r. of 1 xX 10~° over a
one-minute period must not be exceeded for more than 0.4% of any
month; and (b) a b.e.r. of 1 x 1073 over a one-second period must
not be exceeded for more than 0.054% of any month.
Expressions are available which relate b.e.r. to signal-to-noise ratio
for the various forms of digital modulation but they are beyond the
scope of this book.

Land-mobile Radio Systems Many organizations, both private and public, depend upon mobile
radio systems for their successful operation. Examples are many and
include: (a) emergency (the ambulance, fire and police services);
(b) public utilities (gas, water and electricity); and (c) private (delivery
vans, service technicians, mini-cabs and taxis). Despite the inclusion
of emergency services and the public utilities these land-mobile
services are normally lumped together and referred to as the private
land-mobile radio (PMR) service. This is to distinguish services that
do not have access to the public switched telephone network (p.s.t.n.)
from the public land-mobile services which do have such access.
Nowadays, this latter service is provided by cellular radio, and it now
provides national coverage of the UK. Paging systems are also
commonly employed, particularly in hospitals, to alert a person that
he is required to contact his office.
236 RADIO SYSTEMS TECHNOLOGY

Private Mobile Radio

The channel spacing for a PMR system has been standardized at

12.5 kHz, and several users may have to share a channel in over-
lapping service areas. Both amplitude and frequency modulation are
used in conjunction with double-frequency simplex operation. This
term means that two frequencies are used for each call, one for each
direction of transmission but that only one person may speak at a time.
Since frequencies in either the v.h.f. or the u.h.f. band are employed,
the area covered by a base station is limited by both the e.r.p. and
the height of its aerial. The coverage area is made as large as possible
to reduce costs to a minimum, but often a number of interconnected
base stations are necessary to cover a large area. Figure 10.13 shows
a typical arrangement. The mobiles are not given any particular
frequencies but are allocated a pair of frequencies on demand.

Fig. 10.13 Land-mobile radio To Omni-directional -

To
5
system. posal mobiles ee a _-£—* mobiles
BASE
STATION

MASTER BASE
STATION
CONTROL
©) / STATION

In the older PMR systems a car, or van, driver is required to keep

his receiver turned on and he must listen for his call-sign to be broad-
cast. The driver can then respond to the call-sign and receive his call.
Modern PMR systems employ five-tone selective calling. An indivi-
dual mobile can be contacted by means of its identifying code which
is a unique combination of five tones. The mobile receiver recognizes
its code and automatically responds to the call. With so many organiza-
tions, big and small, joining the ranks of land-mobile users the demand
for frequencies has become intense. One solution to the problem of
the limited available frequency spectrum is the use of trunking. Trunk-
ing is the name given to the use of a number of channels that are shared
between a larger number of mobiles. When a mobile wishes to make
a call the radio searches the allocated frequency band for a free
channel, and when one is found it is seized and used to contact the
base station. Once a channel has been allocated to a particular mobile
that allocation lasts for the duration of the call. Immediately the call
 RADIO SYSTEMS 237

is terminated the channel is released and is returned to the ‘pool’ for

re-use. A trunked system can handle up to six times as much traffic
as a basic PMR system whilst appearing to the users as though each
had the exclusive use of a channel.

Cellular Radio

A cellular radio system divides up the geographic area to be covered

into a large number of much smaller areas or cells. Each cell has its
own low-power base station sited somewhere within the cell. The cells
are grouped together in clusters, as shown by Fig. 10.14, and each
cluster has the available radio channels allocated to it in a regular
pattern that is repeated over the entire area. Each base station is
Base station
allocated a particular set of channels for telephonic communication.
Each set is a proportion of the total number of channels allocated to
the system; the remaining channels are allocated to other cells in the
cluster. This process allows the same channel frequencies to be re-used
many times within the other clusters. The number of cells in a cluster
must be such that the clusters can fit together into contiguous areas,
in practice, this means 4-cell, 7-cell, 12-cell and 21-cell clusters.
Figure 10.15(a) shows how three 7-cell clusters fit together; adjacent
cells in different clusters clearly do not share the same frequencies.
Fig. 10.14 A cell cluster in a The frequency allocation for each cluster is shown in Fig. 10.15(b);
cellular radio system. transmissions from base to mobile are in the frequency band 890 MHz
to 935 MHz and in the reverse direction they are in the band 935 MHz
to 980 MHz. When the number of cells in a cluster is small there
will be more channels allocated to each cell, and this means that a
cell can carry a greater telephone traffic. Unfortunately, reducing the
number of cells in each cluster also reduces the distance between two
cells using the same channel frequencies, and this increases co-channel
interference. The 7-cell cluster is probably the best compromise
between these two conflicting factors and it is for this reason that the
7-cell cluster is the most often employed. Not all the channels are

890 MHz 935 MHz

Fig. 10.15 (a) Three 7-cell clusters.
(b) Frequency allocation for each
< 45 MHz sal

cluster in (a).
Pecan eee]
25 kHz Base-to-mobile

45 MHz

cle r[c 5
|SA
ei
A|\8 E

Mobile-to-base

(6)
238 RADIO SYSTEMS TECHNOLOGY

used for telephone conversations: 21 channels are used for the

provision of common signalling facilities.
The geographic size of a cell is varied in accordance with the antici-
pated telephone traffic. In a rural area a cell may have a diameter
of 30 km or more, but in an inner-city area the cell diameter may
be less than 2 km. Reducing the size of a cell provides more cells,
and hence more channels, in an area of a given size but, since the
Six-sectored
cell separation is then reduced, it also increases co-channel inter-
aerial
ference. Co-channel interference can be reduced by the use of sectored
aerials at the base station. A three-sectored aerial has a coverage angle
of 120°, a six-sectored aerial has a coverage of 60°, compared with
the 360° coverage provided by an omni-directional aerial. A cell is
Sector
then divided into three, or six sectors, each of which effectively
Fig. 10.16 Use of a six-sectored becomes a new cell with its own set of channel frequencies. Each
aerial. cell is now corner-excited as shown by Fig. 10.16.
The base stations are interconnected by 2.048 Mb/s p.c.m. land-
line links to mobile switching centres (m.s.c.) either directly or via
a nodal point. The m.s.c. are fully interconnected with one another
and with the p.s.t.n. so that a mobile user has access to both the land
public telephone network and the cellular network. The cellular radio
network is illustrated by Fig. 10.17.
The operation of the cellular radio network is as follows. The
network is organized into a number of traffic areas and each m.s.c.
keeps track of the location of each mobile. Whenever a mobile is not
active it continuously receives from the nearest m.s.c. a code that
identifies the traffic area in which it is travelling. If the received code

Fig. 10.17 Cellular radio network.

 RADIO SYSTEMS 239

is not error free, indicating a low signal level, the mobile will search
for another signal with a higher signal level. Once the new signal
has been found the mobile will check whether the traffic area code
has changed. If it has, it means that the mobile has moved into another
traffic area and the mobile will then register its new location by
identifying itself to the new base station. The network then ensures
that all the m.s.c. have recorded the new location of the mobile.
To originate a call a mobile transmits a message over a common
signalling channel to the nearest base station and it is then allocated
a channel. If, during the duration of a call, the mobile moves from
one cell to another the call is automatically taken over by the new
base station. A base station continuously monitors the signal level
received from all the mobiles in its cell and if this level falls below
a threshold value it informs the m.s.c. The m.s.c. then commands
all the surrounding base stations to measure the signal level that they
are receiving from the mobile. The m.s.c. will then transfer the call
to the base station that is receiving the highest signal level from the
mobile, and it notifies both base stations of the new channel frequency.
The original base station then sends a control signal to the mobile
which switches its transceiver to the new channel. The changeover,
known as hand-off, takes place so quickly that the user is only aware
of a very brief break in transmission.

Table 10.3

Peak
Maximum Channel frequency Signalling
Frequency band number spacing deviation bit rate
(MHz) of channels (KHz) (KHz) (kb/s)

890—935
(base to mobile) 1800 25 9.5 8
935-980
(mobile to base)

The cellular radio system used in the UK is known as the Total

Access Communication System (TACS) and it has the parameters
shown in Table 10.3.
 Exercises

Chapter 1 1.8 Show that the improvement in system signal-to-noise

ratio obtained by converting a d.s.b. system to s.s.b.s.c.
1.1 A carrier wave has an r.m.s. value of 4 A when working is given by 3 + 20 log, [(1 + m)/m] GB.
unmodulated which rises to 4.4 A when it is sinusoidally The signal-to-noise ratio at the output of a radio system
modulated. Calculate the depth of modulation. The carrier is 30 dB when an 80% modulated d.s.b. 10 kW transmitter
is then modulated by a bipolar square wave having the same is used. Calculate the output signal-to-noise ratio when the
peak value as the sine wave. Determine the r.m.s. value of system is converted to operate as s.s.b.s.c. with a transmitted
the modulated wave. power of 4 kW.
1.2. A carrier wave has an r.m.s. value of 10 V when it 1.9 A sinusoidal signal has an r.m.s. value of 10 V before
is unmodulated and of 10.8 V when it is sinusoidally modu- it is amplitude modulated and 10.4 V after modulation. The
lated. Calculate its depth of modulation. The carrier is then modulated signal is passed through a network that attenuates
modulated by a signal that contains components at two the lower side-frequency by 10 dB but leaves the carrier and
different frequencies and its r.m.s. value goes up to 10.8 V. upper side-frequency unchanged. Calculate the r.m.s. value
If the depth of modulation due to one frequency is 40% of the wave.
calculate the depth of modulation produced by the other 1.10 Show that a square-law device can be used for either
frequency. What is the overall depth of modulation? amplitude modulation or demodulation.
1.3. A 1 MHz carrier wave is amplitude modulated to a
depth of 80% by a5 kHz sinusoidal signal and is then applied
to a single tuned circuit that is resonant at 1 MHz and of Chapter 2
Q-factor 100. Calculate the depth of modulation of the output
signal. 2.1 The r.f. power output of a radio transmitter is 1 kW
1.4 A carrier 12 sin (8 x 10° xt) volts is amplitude when it is not modulated. Calculate the power output when
modulated by the signal v = 6 sin (2000 zt) + 3 cos (4000 at)
volts. Calculate the depth of modulation. If the modulated
Vo C
wave is applied across a 100 2 resistance calculate the power
dissipated.
1.5 A tuned circuit has a Q-factor of 60, a capacitance of
100 pF and is resonant at 455 kHz. It is effectively connected
in parallel with the input resistance of a diode detector having
a load resistance of 250 kQ. If the detection efficiency is 86%
calculate the effective Q-factor of the tuned circuit.
1.6 A carrier wave is amplitude modulated to a depth of
60%. One of its sidebands is then completely suppressed. Carrier

Determine the percentage second-harmonic distortion of the

modulation envelope.
1.7 Figure Q.1 shows the circuit of an amplitude modulator
whose output tuned circuit is resonant at 10’ rad/s and has Modulating
an effective Q-factor of 50. The carrier frequency is signal

10’ rad/s and the carrier voltage is large enough to switch

T, and T, alternately ON and OFF. Calculate (a) the d.c.
current of T,, (b) the amplitude of the unmodulated carrier
output, and (c) the peak audio input voltage which will give
a modulation depth of 100%. For all transistors hp; = 100,
Vor = 0.65 V. Fig. Q.1
 EXERCISES 241

the transmitter is modulated (a) in amplitude with a depth of 600 , 6 dB loss and it is terminated by a load impedance
of modulation of 50%, and (b) in frequency with a modula- of 1200 + 300 ©. Calculate the input impedance of the line.
tion index of 0.5. For each case draw frequency spectrum 3.2 List the factors that limit the use of inductors and
diagrams showing the relative amplitudes of the components. capacitors at very high frequencies. Calculate the length of
2.2 A single-channel a.m. radio-telephony system has an loss-free line that would simulate an inductance of 20 nH
output signal-to-noise ratio of 30 dB. If the system is changed at 600 MHz. Z = 500.
to frequency modulation with a rated system deviation of 3.3 A 2/8 length of loss-free line has a characteristic
10 kHz and the transmitted power is doubled, what will then impedance of 600 0 and a load of 800 {. Calculate the input
be the output signal-to-noise ratio? The maximum modulating impedance of the line. Also calculate the ratio, in dB, of the
frequency is 3 kHz. powers in the load with and without the insertion of a N/4
2.3. Explain the effect that pre-emphasis and de-emphasis section of loss-free matching line of the appropriate
have on the output signal-to-noise ratio of an f.m. system. impedance.
If the de-emphasis network has a time constant of 75 ps, 3.4 A 75 Q transmitting aerial, operating at 500 MHz, is
calculate the improvement in signal-to-noise ratio for a base- connected to the transmitter by a 6 m length of coaxial cable
band signal having a maximum frequency of 15 kHz and a of 75 Q characteristic impedance. A cable fault causes an
rated system deviation of 75 kHz. effective capacitance of 4.244 pF to be in series with the line
2.4 Derive an expression for the instantaneous voltage of at a distance of 2 m from the aerial. Calculate the value of
a 100 MHz carrier that is frequency modulated by a 5 kHz the voltage-reflection coefficient on the cable (a) at the fault,
5 V tone if a signal voltage of 10 V produces the rated system and (b) at the input to the line. Also calculate the power
deviation. Use Table 2.1 to determine the amplitudes of the delivered to the aerial before and after the fault occurs if the
various components in the f.m. wave and then find the transmitter power output is 30 W.
percentage power in the first (m + 1) side-frequencies. 3.5 A load of 210 — j180 © is connected to the output
2.5 Explain what is meant by the triangular noise spectrum terminals of a 150 © loss-free line at 100 MHz. Find the posi-
of an f.m. system. A carrier of 1.5 mV at 100 MHz is tion and the length of a single matching stub made of the
received together with an interfering signal of 50 pV at same cable. Calculate the voltage across the resistive part
100.1 MHz. Calculate the peak frequency deviation of the of the load if the sending-end voltage is 100 pV.
carrier. 3.6 Anunknown load presents an impedance of 80 + j110 0
2.6 A varactor diode modulator operates with a reverse- when supplied via a line of characteristic impedance 100
bias voltage of 8 V and has a capacitance of 30 pF. The and length 0.154). Use a Smith chart to find (a) the load
oscillator has a tuning capacitor of 60 pF and oscillates at impedance; (b) the v.s.w.r., (c) the shortest distance from
6 MHz. Calculate the peak value of the modulating signal the load at which a single matching stub could be connected,
voltage required to deviate the carrier frequency by 16 kHz. and (d) the length of this stub.
2.7. Discuss the relative merits of frequency and phase 3.7. A load of 100 — j100 © is to be matched to a 50 2
modulation and explain how they differ from one another. feeder by connecting a short-circuit stub across the line 0.1
Draw spectrum diagrams for a carrier wave (a) frequency- from the load, and then using a /4 section to connect this
modulated by a 2 MHz signal with a rated system deviation point to the rest of the feeder. Use a Smith chart to calculate
of 4.8 MHz, and (b) phase-modulated by a 2 MHz signal the length of the stub and the impedance of the \/4 section.
with a peak phase deviation of 2.4 rad. 3.8 A coaxial cable has an attenuation of 0.2 dB per metre.
2.8 The output of a phase modulator is a carrier of 1 MHz Calculate the Q-factor of a piece of this cable that is resonated
and a frequency deviation of 1 kHz. The modulated signal at a frequency of 900 MHz.
is passed through three frequency multipliers in turn that 3.9 A loss-free line divides into two sections A and B. The
have, respectively, multiplication factors of 6, 8 and 6. The length of section A is 10 cm and that of section B is 14.5 cm.
signal is then applied to a mixer together with a 200 MHz The impedances of the loads connected to the two sections
tone. Calculate the centre frequency and the deviation of the of line are 166 + j60 © on to A and 37.5 + j0 Q on to B.
output of the mixer. Deduce a combination of multipliers The frequency of operation is 258 MHz and the characteristic
and mixers that will turn the modulator output into a 96 MHz impedance is 100 ©. Calculate the voltage-reflection coeffi-
carrier with 75 kHz frequency deviation. cient on each line and also the v.s.w.r. on the source side
of the junction.
3.10 A 50 Q slotted line is used to measure the impedance
Chapter 3 of an unknown load. With the load connected the v.s.w.r.
on the line was S = 2.0. Adjacent voltage minima were found
3.1 Anr.f. line is \/2 long, has a characteristic impedance to be at 41.2 cm and 71.2 cm from the load. With the load
 242 RADIO SYSTEMS TECHNOLOGY

disconnected and replaced by a short-circuit the voltage (a) the free-space wavelength, (b) the cut-off frequency,
minima were found to be at 33.7 cm and 63.7 cm from the (c) the phase velocity, (d) the group velocity, and (e) the
load. Use a Smith chart to calculate the impedance of the load. guide wavelength.
4.8 A rectangular waveguide has internal dimensions of
2.8 cm and 1.3 cm and it is operated at a frequency equal
Chapter 4 to twice the cut-off frequency of the dominant mode. Calcu-
late (a) the guide wavelength, and (>) the angle at which
4.1 Explain why a transverse electromagnetic wave cannot reflection from the waveguide wall takes place.
be propagated down a rectangular waveguide. Show how a 4.9 A rectangular waveguide has wide dimension a of
TE wave can be regarded as being the resultant of two TEM 2.3 cm. A wave travelling down this guide makes an angle
waves. A waveguide of dimensions 2.8 cm by 1.3 cm has of reflection of 60° at each wall. Calculate (a) the phase
a group wavelength equal to the cut-off wavelength. Calculate velocity, (b) the group velocity, (c) the frequency of opera-
the frequency of the propagating signal. tion, (d) the guide wavelength, and (e) the cut-off
4.2 Explain why the cut-off frequency is an important wavelength.
parameter of a waveguide. Also explain why the power-
handling capability of a waveguide is limited by dielectric
breakdown. A waveguide operating at 3.2 GHz has a break- Chapter 5
down electric field strength of 3 x 10° V/m. If a safety
factor of 2 is used calculate the maximum power that can 5.1 Define the terms ‘noise factor’ and ‘noise temperature’
be transmitted. and derive the relationship between them. Figure Q.2 shows
4.3 A rectangular waveguide has internal dimensions of a radio system. Calculate the required available signal power
0.569 cm and 0.285 cm. Calculate (a) the cut-off frequency, delivered by the aerial for the output signal-to-noise ratio
(b) the guide wavelength, and (c) the phase and group to be better than 30 dB. The bandwidth is 10 MHz.
velocities if the frequency is 40 GHz. 5.2 Explain how galactic noise and atmospheric noise vary
4.4 A rectangular waveguide has internal dimensions of with frequency. Hence explain why certain frequencies are
1.067 cm and 0.4318 cm. It is connected to another similar used for communications satellite systems.
waveguide that is filled with a dielectric of relative permit- An aerial is connected to a radio receiver by a feeder of
tivity 2. This second waveguide is matched to its load. If 1.46 dB loss, and a low-noise amplifier of 20 dB gain and
the frequency of operation is 20 GHz calculate the v.s.w.r. a noise temperature of 89.9 K. If the noise factor of the radio
in the air-filled waveguide. receiver is 4.0 dB calculate (a) the overall noise factor of
4.5 What is meant by the term ‘dominant mode’ in a the feeder, amplifier and receiver, and (b) the system noise
rectangular waveguide? What are higher modes and why are temperature.
they generally undesirable? How can the higher-order modes 5.3. A radio receiver has a noise factor of 5 dB and it is
be suppressed? A rectangular waveguide has dimensions connected to an aerial of noise temperature 133 K by a feeder
1.580 cm and 0.7899 cm and transmits a signal whose of (a) O dB loss, and (b) 1.76 dB loss. Calculate the output
frequency is twice the cut-off frequency. Calculate the angle noise power if the bandwidth is 10 MHz, and the gain of
at which reflection from the walls occurs. the receiver is 60 dB.
4.6 Ina test on a 31 cm length of mismatched waveguide 5.4 An amplifier has a noise factor of 7 dB and its output
the measured v.s.w.r. pattern has adjacent minima of 2.9 cm terminals are connected to a radio receiver whose noise factor
and the normalized input impedance was 3 — j29. A is 10 dB. Calculate the minimum gain the amplifier must have
matching iris having an admittance of —j1.43 S is then for the system noise temperature to be less than 1500 K.
connected in parallel with the mismatched load. Use a Smith 5.5 Three amplifiers have the parameters shown in Table
chart to find the v.s.w.r. on the waveguide. Q.1. Determine the order in which the amplifiers ought to
4.7 A rectangular waveguide has internal dimensions of be connected in order to give the lowest overall noise factor.
1 cm and 2.3 cm and it is operated at 10 GHz. Calculate What is the value of this minimum noise factor?

Fig. Q.2
Pre-amplifier Amplifier Mixer
T, = 400K

lL =6dB F = 10dB
G = 20dB
 “EXERCISES 243

Table Q.1 ground level 30 km away, (c) the power received by an aerial
of gain 12 dBi, and (d) the transmitted power.
6.2 Explain the meanings of the terms ‘gain’, ‘effective
Gain Noise factor
aperture’ and ‘directivity’ as applied to an aerial. Two aerials
Amplifier (dB) (dB)
that operate at 600 MHz have effective apertures of 4 m and
1 20 i 2 m, respectively. Calculate the gain of each aerial with
2 10 5 respect to (a) an isotropic radiator, and (b) a \/2 dipole.
3 6 2 6.3 What is meant by the terms the ‘induction field’ and
the ‘radiation field’ of an aerial? At what distance from an
aerial operating at 3 MHz are the two fields of equal ampli-
tude? If, at a much greater distance, the radiation field has
5.6 An aerial is connected to an amplifier that has a gain a magnetic field strength of 265 x 10~° AT/m what is
of 26 dB and a noise factor of 3 dB. The noise temperature (a) the electric field strength, and (b) the power density at
of the aerial is 300 K and it delivers a signal power of 1 nW this point?
to the amplifier. If the available noise power from the aerial 6.4 100 kW power is radiated by an aerial whose effective
is —110 dBm calculate (a) the input signal-to-noise ratio, length is 100 m. Calculate the field strength produced at a
(b) the output signal-to-noise ratio. distance of 100 km if the frequency is 60 kHz.
5.7 An aerial whose noise temperature is 20 K is connected 6.5 Calculate the electric field strength at ground level at
by a waveguide feeder of 0.3 dB loss to a low-noise amplifier a point 10 km from a \/2 monopole. The aerial is supplied
of noise temperature 10 K and gain 28 dB. The output of with a 3 A r.m.s. current at a frequency of 10 MHz.
the low-noise amplifier is connected to a TWT that has a 6.6 Explain what is meant by the effective height of an
noise factor of 10 dB and a gain of 25 dB. If the noise band- aerial. An earthed monopole has an effective height of 0.12
width of the system is 2.8 MHz, calculate what signal power and it is supplied with an r.m.s. current of 100 A. Calculate
must be delivered by the aerial to give an output signal-to- the total radiated power and the effective radiated power.
noise ratio of 30 dB. Also find the field strength at ground level 50 km distant.
5.8 The noise generated within an r.f. amplifier can be 6.7 The transmitting aerial shown in Fig. Q.3 is supplied
assumed to be generated in an equivalent resistance of 56 kQ with current at 80 A peak and at frequency 666.66 kHz.
across the input terminals. The noise temperature of this resis- Calculate (a) the effective height of the aerial, and (b) the
tance is 290 K and the bandwidth of the amplifier is 10 kHz. electric field strength produced at ground level 40 km away.
Calculate the output signal-to-noise ratio of the amplifier
when the input signal is a 100 nV voltage sinusoidally
modulated 30% at 1 kHz.
5.9 Explain why the first stage of a radio receiver should
have a low noise factor and a high gain. Why, for the latter
stages, may gain be more important than noise factor? A radio
receiver has a noise factor of 9 dB and a bandwidth of
100 kHz. Calculate the input signal power needed to give
an output signal-to-noise ratio of 20 dB.
Fig. Q.3
5.10 (a) Show that the r.m.s. noise voltage generated in 6.8 Show that the power received by an aerial of gain G,
the parallel combination of a resistor R and a capacitor C is given by P, = P,G,G(\/4aD)*. Calculate the total
is given by JAT/C ) V. (b) Calculate the variation in the noise attenuation if G, = G, = 30 dBi, D = 30 km, and the
temperature of a circuit as its noise factor is varied from 0 dB frequency of operation is 600 MHz.
to 2 dB. 6.9 Calculate the radiation resistance of a \/16 aerial. Deter-
mine its radiated power when the input current is 100 A
r.m.s. If the loss resistance is 25 () calculate the efficiency
Chapter 6 of the aerial.
6.10 An aerial is 100 m in height and carries a current at
6.1 Show that the radiation resistance of an earthed \/4 100 kHz whose amplitude decreases linearly from 200 A at
vertical aerial is approximately 40 (). the base to zero at the top. Calculate the effective height of
An aerial of this type is supplied with a current of 10 A the aerial when it is mounted upon perfectly conducting earth.
r.m.s. at 4 MHz. Calculate (a) the effective height of the Calculate the field strength produced at a distance of 20 km
aerial, (b) the electric field strength produced at a point at from the aerial.
 244 RADIO SYSTEMS TECHNOLOGY

Chapter 7 Derive an expression for the field strength produced at a

distant point.
7.1 Two vertical \/2 dipoles are in the same horizontal 7.10 Show that the vertical-plane radiation pattern of a
plane and spaced }/4 apart. The dipoles are supplied with horizontal /2 dipole mounted at a height of h above
currents of Jz 0° and 1.51.2 90°, respectively. Calculate and the ground will have maxima at angles of elevation of
plot the horizontal-plane radiation pattern of the array. sin” '(\/4h), sin~'(3/4h), sin—'(5/4h), sin~'(70/4h),
7.2 An aerial array consists of four vertical \/2 dipoles in etc. Sketch the radiation patterns if h = d, 2X, and 3X.
the same horizontal plane spaced )/2 apart. The currents fed
to the aerials are of the same magnitude but have a progres-
sive phase difference of 90°. Derive the expression for the Chapter 8
horizontal-plane radiation pattern of the array and show that
maximum field strength occurs at 120° to the line of the 8.1 A 2 GHz radio link operates over a 48 km stretch of
array. sea with aerials at equal heights. Calculate the necessary
7.3 A Yagi aerial consists of a dipole of 73 + j0 2 imped- heights of the aerials if the k-factor is 0.65 and the direct
ance and a parasitic element of 81 + j108 © impedance. The wave is to clear the sea by a distance equal to 0.577 times
mutual impedance between the two elements is 40 — j30 Q. the first Fresnel zone. If a second aerial is fitted to give height
Calculate the ratio of the powers radiated in the directions diversity, calculate its height also.
dipole to parasitic and parasitic to dipole. Hence state whether 8.2 An ionospheric layer has a maximum electron density
the parasitic element is a reflector or a director. of 1.6 x 10'' electrons/m? and it is at a virtual height of
7.4 An aerial array consists of three \/2 dipoles A, B and 162 km. Calculate its m.u.f. if the radius of the earth is
C that are fed with equal-amplitude, in-phase currents. Each 6400 km and the sky-wave link is 2000 km long.
dipole has a radiation resistance of 73 Q and their mutual 8.3 Ata distance of 1 km from a 1 MHz radio transmitter
impedances are Zag = Zgc = —12.5 — j302, Zac = the daytime field strength is 200 mV/m. Calculate the field
5 + j17.5 Q. Calculate the gain of the array. strength 110 km from the transmitter if the ground-wave
7.5 Draw the radiation pattern of an h.f. log-periodic aerial attenuation is 9 dB greater than the free-space loss. Also
for both the horizontal and vertical planes. Why is the 1.p.a. calculate the field strength at the same point during the night
generally preferred for long-distance radio h.f. systems? when there is also a sky wave received. Assume the sky wave
Draw a sketch of an l.p.a. suitable for use at about 5 to to be reflected from a height of 120 km and that the earth
20 MHz. What is the effect on the main lobe of the radiation is flat.
pattern of changing frequency? 8.4 Explain how the field strength of a u.h.f. transmitter
7.6 Calculate the necessary aperture for a parabolic-dish varies with the distance from the transmitter.
aerial to have a gain of 60 dBi at 5 GHz if the illumination Calculate the maximum range for a u.h.f. radio link if the
efficiency is 60%. transmitting aerial is at a height of 110 m and the receive
What are the effects on the gain, the beamwidth and the aerial height is 60 m. Take the radius of the earth as 6400 km.
side-lobe level of this aerial if the amplitude distribution 8.5 A transmitting aerial has a gain of 6 dBi and is mounted
across the dish aperture varies (a) uniformly, (b) from 180 m above flat earth. The transmitted power is 5 W at
maximum at the edge to zero at the centre, or (c) from zero 50 MHz and the 16 m high receive aerial is 18 km distant.
at the edge to maximum at the centre? Calculate the total field strength at the receive aerial if the
7.7 What is meant by the principle of pattern multiplication? reflection coefficient of the earth is —0.8.
An aerial array consists of 15 vertical \/2 dipoles arranged 8.6 A communications satellite is in orbit 35 800 km above
in three rows of five dipoles. Derive an expression for the an earth station. The down path from satellite to earth station
horizontal-plane radiation pattern. operates at 4 GHz. At the earth station the receiver has an
7.8 What is meant by the gain of a transmitting aerial? An effective input noise temperature of 60 K and a bandwidth
aerial is operated at 11 GHz; calculate its effective aperture of 30 MHz, the parabolic dish aerial has a gain of 60 dBi,
if its gain is 56 dBi. Explain what is meant by the term ‘beam- and the aerial noise temperature is 40 K. If the e.r.p. of the
width’ and calculate its value for this aerial. Why are side satellite is 30 dBW, calculate (a) the carrier-to-noise ratio
lobes undesirable in (a) a transmitting aerial, and (b) a
of the receiver, and (b) the G/T ratio.
receiving aerial? 8.7 A communications satellite link has the following data:
7.9 Three vertical \/2 dipoles are mounted 3) apart in the operating frequency 6 GHz, gain of satellite aerial 6 dBi,
same horizontal line and are fed with currents of J, = gain of earth station aerial 50 dBi, noise temperature of earth
0.257 sin (wt — 90°), Jz = I sin wt and I, = 0.25] sin station aerial 290 K, noise factor of receiver 3 dB, bandwidth
(wt + 90°). of receiver 20 MHz, and link length 36 000 km.
 EXERCISES 245

The carrier-to-noise ratio at the earth station is to be 30 dB. 9.5 Draw the block diagram of a double superheterodyne
Calculate (a) the minimum signal power density at the earth radio receiver that can tune over the frequency range 3 to
station receive aerial, (b) the minimum field strength at the 30 MHz. State typical figures for the first and second inter-
receive aerial, and (c) the minimum power that must be mediate frequencies, and label the frequencies and typical
transmitted by the satellite. voltages of signals at each block in the diagram if the receiver
8.8 Discuss how the propagation of h.f. radio waves is is tuned to 10 MHz.
affected by sun spots. 9.6 Why is a.g.c. always applied to an amplitude-
The maximum electron density in the ionospheric layer modulation radio receiver but not always to a frequency-
is 1.6 x 10'! electrons/m* and occurs at a height of modulation receiver? Explain the problems associated with
162 km. Calculate the skip distance if the radius of the earth the application of a.g.c. to (a) a d.s.b.a.m. receiver, and
- is 6400 km and the m.u.f. is 5.6 MHz. (b) an s.s.b. receiver. How may the gain of the front-end
8.9 Explain briefly the way in which a tropospheric scatter be controlled by a.g.c. and what are the relative merits of
system works. the methods?
A 2.5 GHz tropospheric scatter system has its terminals 9.7 A mixer has the input—output voltage characteristic
200 km apart and uses aerials whose diameters are both given by v = 2 + 1.4v + 0.8v7 + 0.4v° + ... and has
6.1 m. If the transmitted power is 20 kW and the over-the- two 0 dBm signals at different frequencies applied to it.
horizon loss is 62 dB calculate the signal power supplied to Calculate the magnitude of the second- and third-order inter-
the receiver. modulation products. Plot the characteristic for the wanted
8.10 What is meant by the term ‘Fresnel zone’ and what signal and both the second- and third-order intermodulation
is the use of this zone in radio-communications? products and hence determine the second- and third-order
A 600 MHz signal is transmitted from a 75 m high aerial intercept points.
towards the receive aerial 30 km away. There is a 42 m high 9.8 Briefly explain the function of each block in the radio-
obstacle 20 m from the transmitter which must be cleared receiver circuit given in Fig. 9.15.
by the direct ray by a distance equal to 0.577 times the radius
of the first Fresnel zone. Assuming the k-factor to be 0.7,
determine the minimum height for the receive aerial. Chapter 10

10.1 A 600 channel microwave system uses receivers with

Chapter 9 an i.f. bandwidth of 30 MHz, a frequency deviation of
200 kHz, and a maximum channel frequency of 2.54 MHz.
9.1 An amplifier has a gain of 15 dB and has a two-tone Calculate the carrier-to-noise ratio corresponding to a
input signal (each tone at — 15 dBm) applied to its terminals. weighted output signal-to-noise ratio of 71.58 dB.
The third-order intercept point is +22 dBm. Calculate the 10.2 A digital radio-relay link operates at 11.2 GHz and
magnitude of the third-order intermodulation products. is 45 km long. Atmospheric losses are equal to 1.5 dB. The
9.2 A radio receiver has a gain of 20 dB and a third-order transmitted power is 10 W with 9 dB feeder losses and both
intercept point of +20 dBm. The input signal consists of two the transmit and receive aerials have a gain of 49 dB. The
tones, one at 0 dBm and the other at — 12 dBm. The third- receiver has an i.f. bandwidth of 80 MHz, a noise factor of
order intermodulation products are at — 12 dBm. Calculate 8 dB and feeder losses are 8 dB. The minimum carrier-to-
the third-order intercept point. Also calculate the noise floor noise ratio required to obtain a b.e.r. of | X 10~° is 18 dB.
if the bandwidth is 3000 Hz and the noise factor is 8 dB. Calculate the fade margin.
9.3. Explain why the noise factor of a radio receiver is an 10.3 List all the important sources of noise in an analogue
important feature at v.h.f. and u.h.f. but not at h-f. radio-relay link carrying 960 channels. For each source state
A v.h.f. radio receiver has a noise factor of 9 dB, an a.f. whether the source is thermal or intermodulation in its nature.
bandwidth of 3 kHz and an output signal-to-noise ratio of 10.4 Discuss the effects of transmission delay on telephone
25 dB. Calculate (a) the input signal-to-noise ratio, and signals routed via a communications satellite system.
(b) the sensitivity of the receiver. Two earth stations are 9650 km apart and are equidistant
9.4 (a) Explain, with reference to a modern communica- from a satellite that is in orbit at a height of 35 880 km. The
tion receiver, why the double superheterodyne principle is up-link operates at 6 GHz and the down-link at 4 GHz. If
often used. (b) What is meant by the selectivity of a radio the satellite introduces a gain of 113 dB and the earth-station
receiver and how is it provided? (c) Explain what is meant aerials are each 16.4 m in diameter calculate the overall path
by reciprocal mixing and how it affects the selectivity of a loss, and the transmission delay.
radio receiver. 10.5 Draw the block diagram of the equipment used in the
246 RADIO SYSTEMS TECHNOLOGY

terminal stations of a terrestrial radio-relay link. Explain the deviation due to the test tone and then calculate (a) the power
function of each block. received by the aerial, (b) the carrier-to-noise ratio, and
10.6 Describe the operation of an INTELSAT system with (c) the output signal-to-noise ratio in the top channel.
particular reference to the orbit used, the carrier frequencies 10.8 Explain the principle of white-noise testing of a
and the on-board processing. Explain why cross polariza- frequency-division multiplex radio-relay system. State the
tion is often employed. Why is it that earth stations are located meaning of the term ‘noise power ratio’. In a 60 channel
well remote from large towns or cities even though most of system the test-tone level is set to be —12 dBm in each
the traffic handled is destined for the town or city? channel at the input of the link. Calculate the noise power
10.7 An analogue 960 channel radio-relay system operates required to simulate busy-hour loading.
at 6 GHz. The transmitter output power is 5 W, feeder losses 10.9 Explain the meanings of the terms ‘white noise testing’
are 5 dB, and both aerials have a gain of 41.5 dB. The and ‘noise power ratio’. Calculate the r.m.s. frequency devia-
receiver has a noise factor of 10 dB, an i.f. bandwidth of tion due to the white noise signal in a 960 channel system.
40 MHz and is 56 km distant. State the r.m.s. frequency Also calculate the i.f. bandwidth necessary.
Answers to Numerical Exercises

1.1 64%,4.77A La 3,2,1; 3.47 dB

1.2 57.7%, 41.6%, 57.7% 5.6 50 dB, 47.1 dB
13 56.6% 5.7 2.17 x 10°"w
1.4 55.9%, 832.5 mW 20 dB
1.5 24.55 5.9 3.18 x 10°°Ww
16 7.5% 5.10 (b) 0 K to 169.6 K
1.7 (a) 1.05 mA, (6) 6.4 V, (c) 4 V
1.8 36dB 6.1 (a) 11.94 m, (b) 20 mV/m, (c) 7.5 mW, (d) 757 W
| am a pe 6.2 (a) 23 dBi, 20 dBi, (b) 20.85 dB, 17.85 dB
6.3 (a) 100 mV/m, (b) 26.5 uW/m?
2.1 (a) 1125 W, (6) 1000 W 6.4 30 mV/m
2.2 48.23 dB 6.5 18 mV/m
2.3 6.26 dB 6.6 157.9 kW, 473.7 kW, 75.4 mV/m
2 Brod D5 TAZ, 6.7 (a) 42.46 m, (b) 71.1 mV/m
2.6 ~0.25 V 6.8 57.55 dB
2.8 488 MHz, 288 kHz 6.9 1.54.0, 15.4 kW, 5.8%
6.10 100 m, 126 mV/m
3.1 7160
3:2mm 7e8Dsem 7.3 7.2 dB. Director
3.3 600 z —30° 2, 0.09 dB 7.4 7.13 dB
3.4 (a) 0.45 2 —63°, (6) 0.45 z —178°, 7.6 478 m?
30 W, 23.93 W 7.8 23.56 m*, 3.5°
3.5 0.307 m, 0.357 m, 118.2 pV
3.6 (a) 30Q, (b) 3.2, (c) 0.08A, (d) 0.40A 8.1 93.71 m, 113 m
$7 0335, 26.5.0 8.2 15.36 MHz
3.8 409 8.3 0.23 mV/m, 0.53 to 0.99 mV/m
3.9 0.285 —j0161, —456, 1.7 8.4 75.3 km
3.10 25 + joa 8.5 0.49 mV/m
8.6 28.27 dB, 40 dB K™'
4.1 7.576 GHz 8.7 (a) 8 x 10- W/m’, (b) 55 »V/m, (c) 32.9 kW
4.22 5.57MW 8.8 453 km
4.3 26.36 GHz, 1.151 cm, 4.604 x 10? m/s, 19.55 x 8.9 14.6 nW
10° m/s 8.10 109 m
44 1.4
4.5 29.8° 9.1 —44 dBm
4.6 2.4 9.2 2 dBm, — 131.2 dBm
(b) 6.52 GHz, (c) 3.957 x 10° m/s, 9.3 (a) 34 dB, (b) 2.44 pV
4.7. (a) 3m,
(d) 2.274 x 10° m/s, (e) 3.957 cm 9.7 —32 dBm, —70.5 dBm, +18 dBm, +26 dBm
4.8 (a) 3.23 cm, (b) 30°
4.9 (a)6 x 108 m/s, (b) 1.5 x 108 m/s, 10.1 47.3 dB
(c) 13.04 GHz, (d) 2.65 cm, (e) 4.6 cm 10.2 42 dB
10.4 170 dB, 0.25 s
5.1 101 pW 10.7 (a) 1.6 W, (b) 60 dB, (c) 81 dB
5.2 (a) 2.68, (b) 248 K 10.8 —16 dBm
5.3. (a) 105 nW, (6) 112 nW 10.9 1102 MHz, 15.34 MHz
5.4 8.85 dB
 index

Adjacent-channel ratio 202 sectored 238 Amplitude modulators 9

Aerial 110, 131 top loading 121 Analogue microwave radio-relay
array, see arrays of dipoles Yagi 155 systems 218, 229
beam width 123, 141 director 155 loading factor 230
corner reflector 136 folded dipole 156 noise power ratio 234
current element 111, 113, 116, gain 157 test tone analysis 231
125, 126 reflector 155 Angle of incidence 171, 177
directivity 116, 133, 138, 142, 155 Amplitude detectors 11 maximum value of 174
effective aperture 126, 158 coherent 16 Aperture-to-coupling loss 186
effective length 115, 119, 123 diode 11 Array factor 133
effective radiated power 125 detection efficiency 12 Arrays, of dipoles 131
electric field strength 112, 114, incorrect time constant 14 broadside 138
PEAS 12126 input resistance 13 end-fire 142
front-to-back ratio 155 non-linear 15 three-dipole 138
gain 125, 154, 157, 159 product 16 two-dipoles 132
half-wave dipole 123, 125, 126, synchronous 16 reflecting plane 136
131 Amplitude modulation 1 Attack time 211
height factor 142 double sideband 1 Attenuation, ground wave 169
Hertzian dipole 113 bandwidth 2, 22 sky wave 175
image 117, 121, 136, 142, 144 carrier wave 1 transmission line 55
induction field 113 depth of modulation 3 waveguide 93
isotropic radiator 112, 126 envelope 7 Attenuation coefficient 55
log periodic 151 instantaneous voltage 1, 2, 3, 7 Attenuation factor 169
active region 152 lower sideband 2, 5 Attenuator, noise factor of 106
characteristic angle 152 lower side-frequency 2, 6 Automatic gain control 210
gain 154 modulating signal 1, 2 attack time 211
scale factor 152 modulation factor 2, 4, 6, 8 delay time 211
space factor 152 phasor representation 6 hang 212
long wire radiator 128 power content 5 Available noise power 98
monopole 116, 120, 125 r.m.s. value 4 Available power gain 99
mutual impedances of 149 transmission efficiency 6 Average noise factor 106
parabolic dish 157 upper sideband 2, 5
feed methods 161 upper side-frequency 2, 6 Balanced modulator 17
illumination efficiency 158 double sideband suppressed carrier Bandwidth
sidelobes 159 17 amplitude modulated wave 2, 22
parasitic element 155 distortion 18 frequency modulated wave 33
pattern multiplication 146 envelope 18 noise 97
polarization of 111 single sideband suppressed carrier nose 201
power received by 127 0 roofing 215
radiated power 114, 117 distortion 20 skirt 201
radiation field 113 envelope 20 Baseband 218, 229
radiation resistance 115, 124 signal-to-noise ratio 22 Beamwidth 123, 141
rhombic 151 versus double sideband 22 Bessel function 28
 INDEX 249

Blocking 199 time-division multiple access 223 Envelope detector 11

blocking ratio 199 transponders 224 EUTELSAT 160, 223
Component, transmission line as_ 67,
Capacitance, transmission line as 67 75 Fading, of radio signals 175
Capture effect 37, 192 COMSAT 223 Figure-of-merit 188
Carrier wave 1, 24, 34 Constellation 220 Filter, crystal 209
Cassegrain dish aerial 161 Corner reflector 136 SAW 209
CCIR 160, 166, 167, 219, 229, 230, Critical frequency 172 Flicker noise 100
232. Cross modulation 198 Folded dipole 156
CCIED™ 210 235 Cross modulation factor 199 Fourier series 2, 10, 12
Cellular radio 237 Cross polarization 161 Frequency bands, of microwave
TACS 239 Crystal filter 209 systems 217
Characteristic impedance 55, 56 Current element 113, 125, 126 Frequency detectors 49
Circulator 218 Current reflection coefficient 58 coincidence 50
Co-channel interference 192 Curvature, of the earth 167, 181 phase locked loop 51
Coherent detection 16 Cut-off frequency 89 quadrature 49
Coincidence detector 50 Cut-off wavelength 89 Frequency deviation 24
Communication radio receivers 189 Frequency diversity 176
adjacent-channel ratio 202 De-emphasis 40 Frequency division multiple access
automatic gain control 210 Delay time 211 225
attack and delay times 211 Depth of modulation 3 Frequency modulation 24
hang 212 Detectors, amplitude 11 bandwidth 33
blocking 199 coherent 16 Bessel function 28
co-channel interference 192 diode 11 capture effect 37
cross modulation 198 envelope 11 de-emphasis 40
crystal filter 209 non-linear 15 deviation ratio 24
down conversion 191 product 16 frequency deviation 24
dynamic range 197 Ses-Desic. 21 frequency spectrum 26
frequency synthesizer 206 synchronous 16 instantaneous voltage 25
front end 203 Detectors, frequency 49 modulation index 24
ganging 208 coincidence 50 noise 36
high frequency 189, 212 phase locked loop 51 noise output power 38
image-channel interference 191 quadrature 49 phasor diagram 31
image response ratio 192 Detector, stage in radio receiver 210 pre-emphasis 40
intercept point 194 Deviation ratio 24 rated system deviation 24
intermediate frequency 192 Diffraction 169, 177 side-frequencies 26
intermediate-frequency stage 191, Digital microwave radio relay systems signal-to-noise ratio improvement
209 219 over a.m. 39
intermodulation 192 bit error rate 235 triangular noise spectrum 38
local oscillator 191, 205 Digital modulation 220 versus phase modulation 35
mixer stage 204 Diode detector 11 Frequency modulators 42
noise factor 202 detection efficiency 12 direct 42
noise floor 197 incorrect time constant 14 reactance 43
radio-frequency stage 203 input resistance 13 varactor diode 44
reciprocal mixing 197 Distortion, of a.m. wave 7, 15, 18 indirect 46
SAW filter 209 Diversity, frequency 176 Frequency multiplication 47
sensitivity 200 height 184 Frequency spectrum 26
shape factor 202 space 176, 186 Frequency synthesizer 206
signal-to-cross modulation ratio 199 Double-sideband amplitude modulation Fresnel zones 182
SINAD ratio 203 ; 1 radius of 183
tracking 208 Double-sideband suppressed carrier Front end 203
ultra-high frequency 215 apie el Front-to-back ratio 155
up conversion 191 Double-superheterodyne radio receiver
very-high frequency 215 189, 212 Gain, of aerial 125, 154, 157, 159
Communication satellite 187, 222 Down conversion 191 Galactic noise 102
propagation 187 Ganging 208
figure-of-merit 188 Effective aperture, of aerial 126, 158 General fading 175
G/T ratio 188 Effective length, of aerial 115, 119, General line equations 58
Communication satellite systems 222 123 Geo-synchronous orbit 229
frequency-division multiple access Effective radiated power 125 Grazing angle 177
223,225 Electric field strength 112, 114, 117, Gregorian dish aerial 161
orbits 229 1118, 121, 126; 2169 Ground wave 169
radiation techniques 228 Electromagnetic wave . 85, 91 Group velocity 56, 86, 92
SPADE 226 End-fire array 142 Group wavelength 90
speech interpolation 226 Envelope, of a.m. wave 7, 18, 20 G/T ratio 188
 250 RADIO SYSTEMS TECHNOLOGY

Gyro-frequency 166 Mobile radio system Peak sideband power 23

cellular 237 Phase locked loop,
Half-i.f. interference 193 PMR 236 f.m. detector 51
Half-wave dipole 123, 125, 126 Mobile switching centre 238 frequency synthesizer 206
Hang a.g.c. 212 Modulating signal 1, 2 Phase modulation 34
Height factor 142 Modulation, instantaneous voltage 34
Hertzian dipole 113 amplitude 1 rated system deviation 34
cross 148 Phase modulator 47
Illumination efficiency 158 digital 218 Phase noise 198
Image aerial 117, 121, 136, 142, 144 frequency 24 Phase shift keying 220
Image channel interference 191 phase 34 Phasor diagram,
Image response ratio 192 Modulation factor 2, 4 d.s.b.a.m. wave 6
Impedance, of free space 112 Modulation index 24, 26, 30 d.s.b.s.c. wave 18
of waveguide 93 Modulator, s.s.b.s.c. wave 20
Inductance, transmission line 67 amplitude 9, 17 f.m. wave 31
Induction field 113 frequency 42 Polarization, of electromagnetic wave
Input impedance, of a line 60, 74 phase 47 85: Lieb s
Instantaneous voltage Monopole aerial 116, 120, 125 Power, a.m. wave 5
a.m. wave 1, 2, 3, 6 Mutual impedance, between dipoles radiated from an aerial 114, 117
f.m. wave 24 149 Pre-emphasis 40, 232
p.m. wave 34 Private mobile radio 236
INTELSAT 160, 223 Natural noise 101 Product detector 16
Intermediate frequency 192 Noise 96 Propagation 163
Intermediate frequency stage 209 bandwidth 97 communication satellite 187
Intermodulation 192 f.m. system 36 figure-of-merit 188
intercept point 194 galactic 102 G/T ratio 188
noise 100 intermodulation 100 ground wave 169
Ionosphere 163 man-made 102 attenuation factor 169
layers 164 natural 101 diffraction 169
refractive index 164 output power 38, 98, 105 ionosphere 163
variations in 166 phase 198 gyro frequency 166
Isotropic radiator 112 power ratio 234 layers 164
semiconductor 100 refractive index 164
Jitter 221 sky 102 sporadic E 166
temperature 101, 108 variations in 166
k-factor 167 thermal 96 scatter 185
triangular spectrum 38 sky wave 171
Launch angle 185 Noise factor (or figure) 104 angle of incidence 171
Loading factor, of microwave system average 106 maximum value of 174
230 in cascade 106 critical frequency 172
Local oscillator 191, 205 effective 107 diversity 176
Log-periodic aerial 151 floor 197 fading 175
Long-wire radiator 128 lossy network 106 maximum usable frequency 172
Loss, radio receiver 202 optimum traffic frequency 173
aperture-to-coupling 186 spot 106 optimum working frequency 173
ground wave 169 system 109 skip distance 175
scatter 187 variation with frequency 105 virtual height 172
transmission line 187 Noise output power 38, 105 space wave 177
waveguide 93 Noise temperature 101, 108 angle of incidence 177
Lower sideband 2 relation with noise factor 109 curvature of the earth 181
Lower side-frequency 2 system 109 Fresnel zones 183
Non-linear detector 15 height diversity 184
Man-made noise 102 Nose bandwidth 201 maximum distance between aerials
Matching, with transmission line 70, 181
80 Optimum traffic frequency 173 received field strength 178
Maximum usable frequency 172 Optimum working frequency 173 reflection coefficient 177
Microwave radio relay system 216 Orbits, of communication satellites stratosphere 164
analogue 218, 229 229 surface wave 169
loading factor 230 troposphere 164
Over-the-horizon loss 185
noise. power ratio 234 CCIR reference 167
k-factor 167
test-tone analysis 231 Parabolic dish aerial 157 refractivity 167
digital 219 Parasitic element 155 sub-refraction 169
bit error rate 235 Pattern multiplication 146 super-refraction 168
Mixer stage 204 Peak envelope power 23 temperature inversion 167
 INDEX 251

Propagation coefficient, of line 55 Side-frequency 2, 26 group velocity 56

Signal-to-noise ratio 103, 109 input impedance 60, 74
Quadrature amplitude modulation 220 a.m. system 22 loss-free 61
Quadrature detector 49 f.m. system 36 low-loss 62
Quadruple space diversity 186 SINAD ratio 203 matching, using 70, 80
Q factor, of a line 68 Single-channel per carrier working multiple reflections 65
Quarter-wave line transformer 70, 80 225, 226 phase velocity 56
Quaternary p.s.k. 220 Single-sideband suppressed carrier propagation coefficient 55
a.m. 20 Q factor, of 68
Radiation, from an aerial 110 Skip distance 175 reflection loss 59
Radiation field 113 Skirt bandwidth 201 return loss 59
Radiation resistance 115, 124 Sky noise 102 Smith chart 71
_ Radio-frequency stage 191, 203 Sky wave 171 standing waves 63
Radio horizon 177 Smith chart 71 voltage reflection coefficient 58,
Radio receiver, see communication Space diversity 176, 186 60, 72
radio receiver Space wave 177 voltage standing wave ratio 64, 73
Radio relay system, see microwave SPADE 226 Triangular noise spectrum 38
radio relay system Speech interpolation 226 Troposphere 164
Rated system deviation 24, 34 Sporadic E 166 Tropospheric scatter propagation 185
Reactance frquency modulator 43 Spot noise factor 106
Reciprocal mixing 197 Standing waves 63 Up conversion 191
Rectangular waveguide 85 Static noise 101 Upper sideband 2
Reflection coefficient, Stratosphere 164 Upper side-frequency 2
current 58 Stub matching 71, 80
earth 177 Sub-refraction 169 Varactor diode frequency modulator
voltage 58, 60, 72 Sunspot cycle 166 44
Reflection loss 59 Super-refraction 168 Velocity 56, 86, 88, 91
Refractive index, Surface wave 169 Virtual height 172
ionosphere 164 Synchronous detector 16 Voltage reflection coefficient 58, 60,
troposphere 166 System noise temperature 109 2,
Repeat spot interference 193 Voltage standing wave ratio 64, 73
Return loss 59 TACS 239
Rhombic aerial 151 Temperature inversion 168 Waveguides 85
r.m.s. value, of a.m. wave 4 Thermal noise 96 attenuation 93
Time-division multiple access 223, cut-off frequency 89
Satellite system, see communication 226 cut-off wavelength 89
satellite system Top loading 121 field components 91
SAW filter 209 Tracking 208 group velocity 86, 90, 92
Scatter propagation 185 Transponder 224 impedance 93
Sectored aerial 238 Transconductance multiplier 10 modes of propagation 89
Selective fading 23, 175 Transmission efficiency 6 phase velocity 88, 91
Selectivity 198, 201 Transmission lines 55 power handling 94
Semiconductor noise 100 attenuation coefficient 55 propagation in 86
Sensitivity 200 characteristic impedance 55, 56 sizes 95
Shape factor 202 components 67, 75
Shot noise 100 current reflection coefficient 58 Yagi aerial 155
Sideband 2 general line equations 58
 “6. u
7 _ = 7
fac Oya
, ue : be
‘
’

Oo. ek es "trac
oe eS "e
. , ;

¢tiooiee diets
e at hg f &, vest KY air:

bd “gether
lps ‘ -
Bir 2a] P|
e. ‘ ‘MEO +gesing- ee fodaee me rae outer Asie i
is a> nF uty e 7 Hew OETLS 79 heal algniz
: ¥ bt ar Mons: mevibeanaes me) bein Risietid baie sign a -
an Ty aribael | OE ee i
Hinenante <Uisge heh andssabes rr, ovenipod oe
eee RP! Yeo? trea a ie isi wha eee jude Faget
vege cane , reyes Eitire ey cereal Mata “7
: - ~? ee she ti Bex a ue * WSF ave edz
, hn oat WiBrEaT
IE:=, . PO rsaihy eirae
: te: ‘ J ‘ 4 ue x i oa mis i On Tevil Sede
"= ES PSG Sree ee gehtleys —S Wi giana some
Bsc engi basiSohne ssi0s Pe wh asi tie re
’ inna ih deatiaid ae) . ie
AIM states. 1
a ee, aa
, cSt qottalonea dpe?
aa
fie Hay
wae t alee% hodyaogest Sit TT clronep?
‘ te t
: i

SET eT Shan. eee

aij 5: dere, geefirent
1G : 1Gi -maieus gate
vie ‘ Al. suAeaeel Prclney dette’: 3 i
2 ,1 BNR tne - at th eased - an
bis, aise 17 St. nouesk sia? AAS ES ne
Hee At.
be
ghaleV
5+ 6Ol she sop a akHS
‘ a ' coe CT he ge
oTh.. ttle Geani'¥ P “Ohl see ‘agabee yaad Save Ue,
wie $603'd af aout epee pietabioscatiie
bia a
{ J - f eetans
coat
ae
rT)
> gra tee
ney
a ye
sae ~
quips: (el gies? wr Sail
ioeTuas! ¥ .
“2 shea Seared? -
ie 9 a eT sistaG menaivit, cw
3 +9 (iS
OA i isiovand Th-te oe
iv RS RRAP INS rm
awe, green) apt
~ Ste
anideer?
AStel Bs reheiguant
G i *hotignlur squetuhsostutr
ak Ms
tqeaslie sonata |.
ove
PSS AS pepik eee amt
VER Ee eG erhry s oY 4. wmretiitbect @.4 keene
: ¢: “2 esciz ld al ieheqen suatyernsd)
Fi Se ae — EP a ite
"
: 1 : )re : Prag
Pras bee).
ey 5 a= £ on eaTadd
! sultegr a a
Td See 5
ie iy
ss
cc

Sow At wo Pe
: oie

iat u
. her. ARR Oe
iP 5 i
aa le
4 ie
=
- iar
1s
fe
aNedy
:
ite
.
| : :
=
Alas 2R9 ; =
; a a
cee Naa
oi att et
2 ie ae 3
° 2

1? ce
9 :
 RADIO SYSTEMS TECHNOLOGY
=

Baicrelle communication is playing an increasingly important role in

commercial and industrial life and there is a growing need for engineers
and technicians with radio knowledge. Although specialised literature is
available on specific aspects of the topic, until now there has been
nothing which covered this important field as a whole. <

This book gives details of radio systems currently in use, including both
established and recently introduced systems, providing the reader with a
firm grounding in modern techniques.

Topics covered include:

® modulation © transmission lines and WEN
fe[Uo fens
© noise ™ aerials = © microwaves and radio-relay
© satellite systems

A large number of worked examples have been provided to aid

understanding, and the book concludes with exercises summarising the
content of each chapter. ; .

Students of radio communications at BTEC Higher Certificate or Diploma

and first year degree level will find D C Green's latest text a
comprehensive introduction to the subject.

Also available by the same author:

Electronics II Fourth Edition | Digital Electronics for Technicians
Electronics Ill Fourth Edition Digital Electronic Technology
Electronics IV Second Edition — Transmission Principles for Technicians

Cover photograph: “Telefocus; a British Telecom photograph.”

0 582 02697 0
> Rp
nen|
Longman ISBN O-582-02697-0
nlenlen scientific &|
Tel ebeblool@
Copublished in the United States with John .| |
Wiley & Sons, Inc., New York 0470-21672-7
(USA
ee
only) in : 9 "7805 82"026971
a
7 —
Se . a