Telecommunication Circuit Design
Circuit Design
Circuit Design
Second Edition
Patrick D. van der Puije
A Wiley-Interscience Publication
Designations used by companies to distinguish their products are often claimed as trademarks. In all
instances where John Wiley & Sons, Inc., is aware of a claim, the product names appear in initial capital or
ALL CAPITAL LETTERS. Readers, however, should contact the appropriate companies for more complete
information regarding trademarks and registration.
Copyright # 2002 by John Wiley & Sons, Inc., New York. 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 or mechanical, including uploading, downloading, printing, decompiling, recording
or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act,
without the prior written permission of the Publisher. Requests to the Publisher for permission should be
addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New York, NY
10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: [email protected]
This publication is designed to provide accurate and authoritative information in regard to the subject
matter covered. It is sold with the understanding that the publisher is not engaged in rendering professional
services. If professional advice or other expert assistance is required, the services of a competent
professional person should be sought.
ISBN 0-471-22153-8
This title is also available in print as ISBN 0-471-41542-1.
For more information about Wiley products, visit our web site at
Chapter 1
Chapter 2
The History of Telecommunications
1.1 Introduction 1
1.2 Telecommunication Before the Electric Telegraph 1
1.3 The Electric Telegraph 2
1.4 The Facsimile Machine 4
1.5 The Telephone 6
1.6 Radio 8
1.7 Television 9
1.8 The Growth of Bandwidth and the Digital Revolution
1.9 The Internet 11
1.10 The World Wide Web 13
References 15
Bibliography 16
Amplitude Modulated Radio Transmitter
2.1 Introduction 17
2.2 Amplitude Modulation Theory 18
2.3 System Design 21
2.3.1 Crystal-Controlled Oscillator 22
2.3.2 Frequency Multiplier 22
2.3.3 Amplitude Modulator 22
2.3.4 Audio Amplifier 22
2.3.5 Radio-Frequency Power Amplifier
2.3.6 Antenna 23
Radio Transmitter Oscillator 23
2.4.1 Negative Conductance Oscillator 24
2.4.2 Classical Feedback Theory 26
2.4.3 Sinusoidal Oscillators 28
2.4.4 General Form of the Oscillator 28
2.4.5 Oscillator Design for Maximum Power Output 31
2.4.6 Crystal-Controlled Oscillator 35
2.5 Frequency Multiplier 37
2.5.1 Class-C Amplifier 40
2.5.2 Converting the Class-C Amplifier into a Frequency
Multiplier 40
2.6 Modulator 47
2.6.1 Square-law Modulator 47
2.6.2 Direct Amplitude Modulation Amplifier 49
2.6.3 Four-Quadrant Analog Multiplier 51
2.7 Audio-Frequency Amplifier 54
2.7.1 Basic Device Characteristics 54
2.7.2 Class-A Amplifier 55
2.7.3 Class-B Amplifier 60
2.8 The Radio-Frequency Amplifier 67
2.9 The Antenna 71
2.9.1 Radiation Pattern of an Isolated Dipole 72
2.9.2 Monopole or Half-Dipole 72
2.9.3 Field Patterns for a Vertical Grounded
Antenna 74
2.10 Classification of Amplitude Modulated Radio-Frequency
Bands 75
References 76
Problems 76
Chapter 3
Amplitude Modulated Radio Receiver
Introduction 79
The Basic Receiver: System Design 79
The Superheterodyne Receiver: System Design 82
Components of the Superheterodyne Receiver 85
3.4.1 Receiver Antenna 85
3.4.2 Low-Power Radio-Frequency Amplifier 86
3.4.3 Frequency Changer or Mixer 90
3.4.4 Intermediate-Frequency Stage 99
3.4.5 Automatic Gain Control 101
3.4.6 Demodulator 103
3.4.7 Audio-Frequency Amplifier 105
3.4.8 Loudspeaker 105
3.5 Short-Wave Radio 107
References 108
Bibliography 108
Problems 108
Chapter 4
Frequency Modulated Radio Transmitter
4.1 Introduction 111
4.2 Frequency Modulation Theory 111
4.3 The Parameter Variation Method 115
4.3.1 Basic System Design 115
4.3.2 Automatic Frequency Control of the FM
Generator 117
4.3.3 Component Design with Automatic Frequency
Control 118
4.4 The Armstrong System 122
4.4.1 Practical Realization 125
4.4.2 Component Circuit Design 127
4.5 Stereophonic FM Transmission 137
4.5.1 System Design 137
Bibliography 138
Problems 139
Chapter 5
The Frequency Modulated Radio Receiver
5.1 Introduction 143
5.2 Component Design 146
5.2.1 Antenna 146
5.2.2 Radio-Frequency Amplifier 146
5.2.3 Local Oscillator 147
5.2.4 Frequency Changer 147
5.2.5 Intermediate-Frequency Stage 147
5.2.6 Amplitude Limiter 148
5.2.7 Frequency Discriminator 148
5.3 Stereophonic Frequency Modulated Reception
5.3.1 Synchronous Demodulation 158
5.3.2 Stereophonic Receiver Circuit 158
References 158
Problems 159
Chapter 6
Television Transmitter
Introduction 161
System Design 162
Component Design 163
6.3.1 Camera Tube 163
6.3.2 Scanning System 167
6.3.3 Audio Frequency and FM Circuits
6.3.4 Video Amplifier 170
6.3.5 Radio-Frequency Circuits 179
6.3.6 Vestigial Sideband Filter 179
6.3.7 Antenna 179
6.3.8 Color Television 180
References 184
Bibliography 184
Problems 185
Chapter 7
The Television Receiver
7.1 Introduction 187
7.2 Component Design 189
7.2.1 Antenna 189
7.2.2 Superheterodyne Section 189
7.2.3 Intermediate-Frequency Amplifier 189
7.2.4 Video Detector 190
7.2.5 The Video Amplifier 191
7.2.6 The Audio Channel 191
7.2.7 Electron Beam Control Subsystem 191
7.2.8 Picture Tube 202
7.3 Color Television Receiver 203
7.3.1 Demodulation and Matrixing 203
7.3.2 Component Circuit Design 205
7.3.3 Color Picture Tube 207
7.4 High-Definition Television (HDTV) 209
Bibliography 210
Problems 210
Chapter 8
Telephone Network
Introduction 213
Technical Organization 213
Basic Telephone Equipment 216
8.3.1 Carbon Microphone 216
8.3.2 Moving-Iron Telephone Receiver 218
8.3.3 Local Battery – Central Power Supply 220
8.3.4 Signalling System 220
8.3.5 The Telephone Line 221
8.3.6 Performance Improvements 222
8.3.7 Telephone Component Variation 224
8.4 Electronic Telephone 224
8.4.1 Microphones 225
8.4.2 Receiver 225
8.4.3 Hybrid 226
8.4.4 Tone Ringer 229
8.4.5 Tone Dial 230
8.5 Digital Telephone 243
8.5.1 The Codec 244
8.6 The Central Office 253
8.6.1 Manual Office 253
8.6.2 Basics of Step-by-Step Switching 253
8.6.3 The Strowger Switch 256
8.6.4 Basics of Crossbar Switching 257
8.6.5 Central Office Tone Receiver 259
8.6.6 Elements of Electronic Switching 261
References 263
Problems 263
Chapter 9
Signal Processing in the Telephone System
9.1 Introduction 267
9.2 Frequency Division Multiplex (FDM) 267
9.2.1 Generation of Single-Sideband Signals 268
9.2.2 Design of Circuit Components 269
9.2.3 Formation of a Basic Group 269
9.2.4 Formation of a Basic Supergroup 270
9.2.5 Formation of a Basic Mastergroup 270
9.3 Time-Division Multiplex (TDM) 272
9.3.1 Pseudodigital Modulation 273
9.3.2 Pulse-Amplitude Modulation Encoder 273
9.3.3 Pulse-Amplitude Modulation Decoder 279
9.3.4 Pulse Code Modulation Encoder=Multiplexer 286
9.3.5 Pulse-Code Modulation Decoder=Demultiplexer 286
9.3.6 Bell System T-1 PCM Carrier 286
9.3.7 Telecom Canada Digital Network 290
9.3.8 Synchronization Circuit 290
9.3.9 Regenerative Repeater 291
9.4 Data Transmission Circuits 295
9.4.1 Modem Circuits 298
References 300
Problems 301
Chapter 10 The Facsimile Machine
10.1 Introduction 305
10.2 Systems Design 307
10.2.1 The Transmit Mode 308
10.2.2 The Receive Mode 308
10.3 Operation 310
10.3.1 ‘‘Handshake’’ Protocol 310
10.4 The Transmit Mode 311
10.4.1 The CCD Image Sensor 311
10.4.2 The Binary Quantizer 315
10.4.3 The Two-Row Memory 317
10.5 Data Compression 318
10.5.1 The Modified Huffman (MH) Code 318
10.5.2 The Modified READ Code 318
10.6 The Modem 319
10.7 The Line Adjuster 319
10.8 The Receive Mode 319
10.8.1 The Power Amplifier 319
10.8.2 The Thermal Printer 321
10.9 Gray Scale Transmission: Dither Technique 322
References 323
Bibliography 323
Glossary 323
Problems 324
Chapter 11 Personal Wireless Communication Systems
11.1 Introduction 325
11.2 Modulation and Demodulation Revisited 326
11.3 Access Techniques 327
11.3.1 Multiplex and Demultiplex Revisited 327
11.3.2 Frequency-Division Multiple Access (FDMA) 328
11.3.3 Time-Division Multiple Access (TDMA) 328
11.3.4 Spread Spectrum Techniques 329
11.4 Digital Carrier Systems 331
11.4.1 Binary Phase Shift Keying (BPSK) 333
11.4.2 Quadrature Phase Shift Keying (QPSK) 334
11.5 The Paging System 338
11.5.1 The POCSAG Paging System 338
11.5.2 Other Paging Systems 341
11.6 The Analog Cordless Telephone 343
11.6.1 System Design 343
11.6.2 Component Design 343
11.6.3 Disadvantages of the Analog Cordless
Telephone 347
11.7 The Cellular Telephone 347
11.7.1 System Overview 348
11.7.2 Advanced Mobile Phone System (AMPS) 348
11.8 Other Analog Cellular Telephone Systems 355
11.8.1 Disadvantages of Analog Cellular Telephone
Systems 357
11.9 The CDMA Cellular Telephone Systems 358
11.9.1 System Design of the Transmit Path 358
11.9.2 Component Circuit Design for the Transmit
Path 359
11.9.3 System Design of the Receive Path 361
11.9.4 Component Circuit Design for the Receive
Path 362
11.10 Other Digital Cellular Systems 362
References 363
Bibliography 363
Problems 364
Abbreviations 364
Chapter 12 Telecommunication Transmission Media
12.1 Introduction 367
12.2 Twisted-Pair Cable 367
12.2.1 Negative-Impedance Converter 370
12.2.2 Four-Wire Repeater 374
12.3 Coaxial Cable 375
12.4 Waveguides 376
Optical Fiber 376
Free Space Propagation 377
12.6.1 Direct Wave 379
12.6.2 Earth-Reflected Wave 379
12.6.3 Troposphere-Reflected Wave 379
12.6.4 Sky-Reflected Wave 379
12.6.5 Surface Wave 380
12.7 Terrestrial Microwave Radio 380
12.7.1 Analog Radio 381
12.7.2 Digital Radio 384
12.8 Satellite Transmission System 384
References 389
Bibliography 389
Problems 390
Appendix A
Appendix B
Designation of Frequencies
VHF Television Frequencies 397
UHF Television Frequencies 398
Appendix C
The Electromagnetic Spectrum
Appendix D
The Modified Huffman Data Compression Code
Appendix E
Electronic Memory
E.1 Introduction 405
E.2 Basics of S-R Flip-Flop Circuits 405
E.3 The Clocked S-R Flip-Flop 407
E.4 Initialization of the S-R Flip-Flop 409
E.5 The Shift Register 409
E.5 Electronic Memory 411
E.6 Random Access Memory (RAM) 411
Appendix F
Binary Coded Decimal to Seven-Segment
Introduction 391
The Ideal Transformer 392
The Practical Transformer 394
The first edition of this book was published in 1992. Nine years later it had become
clear that a second edition was required because of the rapidly changing nature of
telecommunication. In 1992, the Internet was in existence but it was not the
household word that it is in the year 2001. Cellular telephones were also in use
but they had not yet achieved the popularity that they enjoy today. In the current
edition, Chapter 1 has been revised to include a section on the Internet. Chapter 10 is
new and it covers the facsimile machine; I had overlooked this important telecommunication device in the first edition. Chapter 11 is also new and it describes the
pager, the cordless telephone and the cellular telephone system. These are examples
of a growing trend in telecommunications to go ‘‘wireless’’.
This book is about telecommunications: the basic concepts, the design of subsystems and the practical realization of the electronic circuits that make up
telecommunication systems. The aim of this book is to fill a gap that exists in the
teaching of telecommunications and electronic circuit design to electrical engineering students. Frequently, courses on electronic circuits are taught to students without
a clear indication of where these circuits may be used. Later in their career, students
may take a course in communication theory where the usual approach is to treat
subjects such as modulation, frequency changing and detection as mathematical
concepts and to represent them in terms of ‘‘black boxes’’. Thus the connection
between the function ‘‘black boxes’’ and the design of an electronic circuit that will
perform the function is glossed over or is completely missing.
The approach followed in this book is to take a specific communication system,
for example the amplitude modulated (AM) radio system, and describe in mathematical terms how and why the system is designed the way it is. The system is then
broken down into functional blocks. The design of each functional block is
examined in terms of the electronic devices to be used, the circuit components
and requirements for power. The effectiveness of each functional block is determined. In most cases, more than one circuit is presented, starting from the very
elementary which usually illustrates the principles of operation best, to more
sophisticated and practical varieties. The order in which the signal encounters the
functional blocks determines the order of the presentation, so new information is
presented at an opportune moment when the interest of the student is optimal.
Examples are provided to emphasize the link from concept, to design and realization
of the circuits. Systems examined in this text include commercial radio broadcasting,
television and telephone, with sections devoted to personal wireless communication,
satellite communications and data transmission circuits.
This book was written with the final-year engineering undergraduate student in
mind, so a clear and explicit style of writing has been used throughout. Illustrative
examples have been given whenever possible to promote the active participation of
the student in the learning process. However, the practical approach to electronic
circuit design will no doubt be useful to people involved in the telecommunications
industry for updating, review and as a reference.
Prerequisites to the course in telecommunication circuit design are university
mathematics, basic electronics and some familiarity with communication theory
(although this is not strictly necessary). In every case where communication theory
has a direct impact on the design, enough background has been given to gain an
understanding of the topic.
A number of specialized topics have been excluded in the interest of brevity.
These include antennas, filters and loudspeakers. Many books are available on these
subjects. Rather than presenting a cursory treatment of these very important subjects,
I opted for a qualitative description of the operation and design of antennas, filters
and loudspeakers in the hope that the reader can develop an appreciation of the
outstanding features of these devices which can be built upon if they are of special
interest. A list of available reading material has been given in the appropriate
Although most of the circuits discussed in this book can be found in integrated
circuit form, I have, in general, avoided detailed discussion of integrated circuit
design. This is because the ‘‘rules’’ of integration aim to reduce the area of the chip
to a minimum and thus tend to increase the number of transistors or active
components, which take up little area, at the expense of passive components such
as resistors and capacitors, which take up relatively large areas. Furthermore,
because the integrated circuit process can produce very closely matched transistors,
the integrated circuit designer often uses symmetry to achieve circuit functions not
possible with discrete devices. An explanation of how an integrated circuit works is
therefore more complicated than its discrete counterpart. In every case where the
simple and the modern have clashed, I have chosen the simple. However, integrated
circuit design techniques have been discussed whenever they are relevant and do not
distract the reader from a good understanding of the basic principles of circuit
In Chapter 1, a brief history of telecommunication is given. The last 150 years has
been a time of tremendous growth and change in telecommunications, more than
enough change to qualify as a ‘‘revolution’’, perhaps the greatest revolution in the
history of mankind — the Information Revolution.
Chapters 2 and 3 describe the amplitude modulated (AM) radio system and the
electronic circuits that make it possible, from the design of the crystal-controlled
oscillator in the transmitter to the loudspeaker in the receiver. Chapters 4 and 5
repeat the process with the frequency modulated (FM) radio and include sections on
stereophonic commercial broadcast and reception.
Television—the transmission and reception of images—is discussed in Chapters
6 and 7. The design of the circuits involved in the acquisition of the video signals,
the processing, transmission, coding, broadcasting, reception and decoding are
In Chapters 8 and 9, the growth of the telephone system is traced from its humble
beginnings to the world-wide network that it is today. The need to open up the
system to an increasing number of subscribers has led to the development of
sophisticated signal processing techniques and circuits with which to implement
Chapter 10 describes the facsimile machine as a system, as well as the design of
its component parts.
In Chapter 11 the pager, the cordless telephone and cellular telephone systems are
Chapter 12 covers the development of channels in new transmission media, such
as satellites and fiber optics, as well as improvements to hard wire connections made
possible because circuit designers have produced the hardware at the right time and
at the right cost. The growing traffic of ‘‘conversations’’ between machines of
various descriptions has accelerated the trend towards ‘‘digitization’’ of signals in the
telephone network. The design of circuits capable of accepting data corrupted by
noise, restoring and retransmitting them is discussed.
This book started off as lecture notes for a senior college course in electrical
engineering called ‘‘Telecommunication Circuits’’. At the time I proposed the
course, it was becoming increasingly clear that the knowledge of our graduating
students of communication systems left much to be desired. Students seemed to
think that anything analog (including radio, television and the telephone) was passé.
Digital circuits (computers and software development), on the other hand, were
considered ‘‘cutting edge’’. It was necessary to bring some balance into this situation
and I hope that this book helps to restore some semblance of symmetry. The
seemingly simple task of changing a set of lecture notes into a textbook turned out
not to be quite as simple as I had imagined. However, I have learned a lot from it and
I hope the reader does, too.
The material contained in this book is more than can be presented in the normal
13-week term. However, the organization of chapters is based on the three major
telecommunication networks: radio, television, and the telephone. It is therefore
convenient to organize such a course around a group of chapters with minimal
rearrangement of the material and still maintain coherence.
Ottawa, Ontario, Canada
September 2001
John G. Proakis, Editor
Northeastern University
Introduction to Digital Mobil Communications
Yoshihiko Akaiwa
Ditigal Telephony, 3rd Edition
John Bellamy
ADSL, VDSL, and Multicarrier Modulation
John A. C. Bingham
Biomedical Signal Processing and Signal Modeling
Eugene N. Bruce
Elements of Information Theory
Thomas M. Cover and Joy A. Thomas
Practical Data Communications
Roger L. Freeman
Radio System Design for Telecommunications, 2nd Edition
Roger L. Freeman
Telecommunication System Engineering, 3rd Edition
Roger L. Freeman
Telecommunications Transmission Handbook, 4th Edition
Roger L. Freeman
Introduction to Communications Engineering, 2nd Edition
Robert M. Gagliardi
Optical Communications, 2nd Edition
Robert M. Gagliardi and Sherman Karp
Active Noise Control Systems: Algorithms and DSP Implementations
Sen M. Kuo and Dennis R. Morgan
Mobile Communications Design Fundamentals, 2nd Edition
William C. Y. Lee
Expert System Applications for Telecommunications
Jay Liebowitz
Polynomial Signal Processing
V. John Mathews and Giovanni L. Sicuranza
Digital Signal Estimation
Robert J. Mammone, Editor
Digital Communication Receivers: Synchronization, Channel Estimation, and Signal
Heinrich Meyr, Marc Moeneclaey, and Stefan A. Fechtel
Synchronization in Digital Communications, Volume I
Heinrich Meyr and Gerd Ascheid
Business Earth Stations for Telecommunications
Walter L. Morgan and Denis Rouffet
Wireless Information Networks
Kaveh Pahlavan and Allen H. Levesque
Satellite Communications: The First Quarter Century of Service
David W. E. Rees
Fundamentals of Telecommunication Networks
Tarek N. Saadawi, Mostafa Ammar, with Ahmed El Hakeem
Meteor Burst Communications: Theory and Practice
Donald L. Schilling, Editor
Digital Communications over Fading Channels: A Unified Approach to Performance Analysis
Marvin K. Simon and Mohamed-Slim Alouini
Digital Signal Processing: A Computer Science Perspective
Jonathan (Y) Stein
Vector Space Projections: A Numerical Approach to Signal and Image Processing, Neural
Nets, and Optics
Henry Stark and Yongyi Yang
Signaling in Telecommunication Networks
John G. Van Bosse
Telecommunication Circuit Design, 2nd Edition
Patrick D. van der Puije
Worldwide Telecommunications Guide for the Business Manager
Walter H. Vignault
According to UNESCO statistics, in 1997, there were 2.4 billion radio receivers in
nearly 200 countries. The figure for television was 1.4 billion receivers. During the
same year, it was reported that there were 822 million main telephone lines in use
world-wide. The number of host computers on the Internet was estimated to be 16.3
million [1]. In addition to this, the military in every country has its own communication network which is usually much more technically sophisticated than the
civilian network. These numbers look very impressive when one recalls that
electrical telecommunication is barely 150 years old. One can well imagine the
number of people employed in the design, manufacture, maintenance and operation
of this vast telecommunication system.
The need to send information from one geographic location to another with the
minimum of delay has been a quest as old as human history. Galloping horses,
carrier pigeons and other animals have been recruited to speed up the rate of
information delivery. The world’s navies used semaphore for ship-to-ship as well as
from ship-to-shore communication. This could be done only in clear daylight and
over a distance of only a few kilometres. The preferred method for sending messages
over land was the use of beacons: lighting a fire on a hill, for example. The content
of the message was severely restricted since the sender and receiver had to have
previously agreed on the meaning of the signal. For example, the lighting of a
beacon on a particular hill may inform one’s allies that the enemy was approaching
from the north, say. In 1792, the French Legislative Assembly approved funding for
the demonstration of a 35 km visual telegraphic system. This was essentially
semaphore on land. By 1794, Lille was connected to Paris by a visual telegraph [3].
In England, in 1795, messages were being transmitted over a visual telegraph
between London and Plymouth – a return distance of 800 km in 3 minutes [4].
North American Indians are reputed to have communicated by creating puffs of
smoke using a blanket held over a smoking fire. Such a system would require clear
daylight as well as the absence of wind, not to mention a number of highly skilled
A method of telecommunication used in the rain forests of Africa was the
‘‘talking drum’’. By beating on the drum, a skilled operator could send messages
from one village to the next. This system of communication had the advantage of
being operational in daylight and at night. However, it would be subject to operator
error, especially when the message had to be relayed from village to village.
The first practical use of electricity for communication was in 1833 by two
professors from the University of Goettingen, Carl Friedrich Gauss (1777–1855)
and Wilhelm Weber (1804–1891). Their system connected the Physics Institute to
the Astronomical Observatory, a distance of 1 km, and used an induction coil and a
mirror galvanometer [4].
In 1837, Charles Wheatstone (1802–1875) (of Wheatstone Bridge fame) and
William Cooke (1806–1879) patented a communication system which used five
electrical circuits consisting of coils and magnetic needles which deflected to
indicate a letter of the alphabet painted on a board [5]. The first practical use of
this system was along the railway track between Euston and Chalk Farm stations in
London, a distance of 2.5 km. Several improvements were later made, the major one
being the use of a coding scheme which reduced the system to a single coil and a
single needle. The improvement of the performance, reliability and cost of communication has since kept many generations of engineers busy.
At about the time when Wheatstone and Cooke were working on their system,
Samuel Morse (1791–1872) was busy doing experiments on similar ideas. His major
contribution to the hardware was the relay, also called a repeater. By connecting a
series of relays as shown in Figure 1.1, it was possible to increase the distance over
Figure 1.1. The use of Morse’s relay to extend the range of the telegraph.
which the system could operate [5]. Morse also replaced the visual display of
Wheatstone and Cooke with an audible signal which reduced the fatigue of the
operators. However, he is better known for his efficient coding scheme which is
based on the frequency of occurrence of the letters in the English language so that
the most frequently used letter has the shortest code (E: dot) and the least frequently
used character has the longest code (‘–apostrophe: dot-dash-dash-dash-dash-dot).
This code was in general use until the 1950s and it is still used by amateur radio
operators today.
In 1843, Morse persuaded the United States Congress to spend $30,000 to build a
telegraph line between Washington and Baltimore. The success of this enterprise
made it attractive to private investors, and Morse and his partner Alfred Vail (1807–
1859), were able to extend the line to Philadelphia and New York [6]. A number of
companies were formed to provide telegraphic services in the east and mid-west of
the United States. By 1851, most of these had joined together to form the Western
Union Telegraph Company.
By 1847, several improvements had been made to the Wheatstone invention by
the partnership of Werner Siemens (1816–1892) and Johann Halske (1814–1890) in
Berlin. This was the foundation of the Siemens telecommunication company in
The next major advance came in 1855 when David Hughes (1831–1900) invented
the printing telegraph, the ancestor of the modern teletype. This must have put a lot
of telegraph operators out of work (a pattern which was to be repeated over and over
again) since the machine could print messages much faster than a person could
write. Another improvement which occurred at about this time was the simultaneous
transmission of messages in two directions on the same circuit. Various schemes
were used but the basic principle of all of them was the balanced bridge.
In 1851, the first marine telegraphic line between France and England was laid,
followed in 1866 by the first transatlantic cable. The laying of this cable was a major
feat of engineering and a monument to perseverance. A total of 3200 km of cable
was made and stored on an old wooden British warship, the HMS Agamemnon. The
laying of the cable started in Valentia Bay in western Ireland but in 2000 fathoms of
water, the cable broke and the project had to be abandoned for that year. A second
attempt the following year was also a failure. A third attempt in 1858 involved two
ships and started in mid-ocean and it was a success. Telegraphic messages could then
be sent across the Atlantic. The celebration of success lasted less than a month when
the cable insulation broke down under excessively high voltage. Interest in
transatlantic cables was temporarily suspended while the American Civil War was
fought and it was not until 1865 that the next attempt was made. This time a new
ship, the Great Eastern, started from Ireland but after 1900 km the cable broke.
Several attempts were made to lift the cable from the ocean bed but the cable kept
breaking off so the project was abandoned until the following year. At last in 1866,
the Great Eastern succeeded in laying a sound cable and messages could once more
traverse the Atlantic. By 1880, there were nine cables crossing the ocean [6].
The telegraph was and remained a communication system for business, and in
most European countries it became a government monopoly. Even in its modernized
form (telex) it is essentially a cheap long-distance communication network for
In 1843, the British Patent Office issued a patent with the title ‘‘Automatic
electrochemical recording telegraph’’ to the Scottish inventor Alexander Bain
(1810–1877). The essence of the invention is shown in Figure 1.2. Two identical
pendulums are connected as shown by the telegraph line. For simplicity, we assume
the ‘‘message’’ to be sent is the letter H and it is engraved on a metallic plate and
shaped to the appropriate radius so that the ‘‘read’’ stylus makes contact with the
raised parts of the plate as the pendulum sweeps across it. On the far end of the
telegraph line, the stylus of the second pendulum maintains contact with the
electrosensitive paper which rests on an electrode shaped to the same radius as
before. The electrosensitive paper has been treated with a chemical which produces a
dark spot when electric current flows through it.
To operate the system, both pendulums are released from their extreme left
positions simultaneously. Since they are identical, it follows that they will travel at
the same speed, one across the ‘‘message’’ plate and the other across the electrosensitive paper. At first no current flows, but as the transmitter pendulum makes
contact with the raised portion of the plate, the circuit is complete and the resulting
current causes the electrosensitive paper to produce a dark line of the same length as
the raised metal segment. The original patent included the functions: (a) an
electromagnetic device to keep the pendulums swinging at a constant amplitude
Figure 1.2. The configuration of the Bain ‘‘Automatic electrochemical recording telegraph.’’ To
keep the diagram simple, additional circuits required for synchronization, phasing and scanning
are not shown.
(synchronization), (b) a second electromagnetic arrangement to ensure that the two
pendulums start their swings at the same instant (phasing), and (c) a mechanism to
move the message plate and the electrosensitive paper simultaneously one step at a
time after each sweep at right angles to the direction of the pendulum swing
(scanning). When several sweeps have occurred, the lines produced will form an
exact image of the raised metal parts of the ‘‘message’’ plate. Figure 1.3(a) shows the
letter H scanned in 20 lines and Figure 1.3(b) shows the corresponding current
waveforms. Figure 1.3(c) shows the reproduced image.
All the facsimile machines since the Bain patent have the three functions listed
above. In modern facsimile machines, the first two functions have been replaced by
electronic techniques which ensures that the transmitter and the receiver are
‘‘locked’’ to each other at all times. The mechanism for scanning the message is
also largely electronic, although in most machines it is still necessary to move the
page mechanically as it is scanned.
In 1848, Frederick Bakewell, an Englishman, produced a new version of the fax
machine in which the ‘‘message plate’’ as well as the image were mounted on
Figure 1.3. (a) Shows the letter H scanned in 20 lines, (b) shows the current waveforms for
each line scanned and (c) shows the reproduced H. Note the effect of the finite width of the
receiver stylus on the image.
cylinders which were turned by falling weights, similar to a grandfather clock. To
ensure that the cylinders turned at the same speed he used a mechanical speed
governor. The scanning head (stylus) was propelled on an axis parallel to that of the
cylinder by a lead-screw. This was an example of ‘‘spiral scanning’’. Unlike Bain, he
used an insulating ink to write the message on a metallic surface. But, as before, the
paper in the receiver was chemically treated to respond to the flow of electric current
and it was mounted on an identical cylinder with the ‘‘write’’ head driven by an
identical lead-screw. The main difficulty with this design was the necessity to keep
the two clock motors in remote locations starting and running at the same speed
during the transmission.
In 1865, Giovanni Caselli (1815–1891), an Italian living in France, patented an
improved version of Bain’s machine which he called the ‘‘Pantelegraph’’. He then
established connections between Paris and a number of other French cities. His
machine was a combination of the insulating ink message plate of Bakewell, the
pendulum of Bain’s transmitter, and the Bakewell cylindrical receiver. The pantelegraph was a commercial success and it was used in Italy and Britain for many years.
By the end of the 1800s it was possible to send photographs by fax. The picture
had to be etched on a metallic plate in the form of raised dots (similar to the
technique used for printing pictures in newspapers). The size of the dots represented
the different shades of gray; small for light and large for dark gray. The transmitter
stylus traced lines across the picture making contact with the raised dots and thus
producing corresponding large and small dots at the receiver.
In 1902, Arthur Korn demonstrated a scanning system which used light instead of
physical contact with a metallic plate and the resultant flow of current. His method
was far superior to all the previous techniques, especially in the transmission of
photographs. He wrapped the photographic film negative of the picture on the
outside of a glass cylinder which was turned at a constant rate by an electric motor.
An electric lamp provided the light and a system of lenses were used to focus the
light onto the negative. The light that passed through the film was reflected by a
mirror onto a piece of selenium whose resistance varied according to how much light
reached it. The selenium cell was used to control the current flowing in the receiver.
The receiver recorded the image directly onto film. To ensure that the transmitter and
receiver cylinders were in synchronism at all times, he used a central control system
with a tuning fork generating the control signal.
In the 1920s the large American telecommunication companies, American
Telephone and Telegraph (AT&T), Radio Corporation of America (RCA) and
Western Union, became interested in fax machine development and they used new
techniques, materials and devices such as the vacuum tube, phototubes and later
semiconductors to produce the modern fax machine.
In 1876, Alexander Graham Bell (1847–1922) was conducting experiments on a
‘‘harmonic telegraph’’ system when he discovered that he could vary the electric
current flowing in a circuit by vibrating a magnetic reed held in close proximity to an
electromagnet which formed part of the loop. By connecting a second electromagnet
together with its own magnetic reed in the circuit, he could reproduce the vibration
of the first reed. Using a human voice to excite the magnetic reed led to the first
telephone for which he was granted a patent later that year. He went on to
demonstrate his invention at the International Centennial Exhibition in Philadelphia
and before the year ended, he transmitted messages between Boston, Massachusetts,
and North Conway, New Hampshire, a distance of 230 kilometres. Few people
realized the potential of the new invention and in 1878, when Bell tried to sell his
patent to the Western Union Telegraph Company, he was turned down [7].
The early telephone system consisted of two of Bell’s magnetic reed-electromagnet instruments in series with a battery and a bell. Bell’s instrument worked very
well as a receiver, in fact so well that it has survived almost unchanged to this day.
As a transmitter, however, it left a lot to be desired. It was soon replaced by the
carbon microphone (one of the many inventions of Thomas Edison (1847–1931))
which was, until recently, the most widely used microphone in the telephone
In the early telephone system, each subscriber was connected to a central office
by a single wire with an earth return. This led to cross-talk between subscribers. At
about this time, electric traction had become very popular which resulted in
increased interference from the noise generated by the electric motors. The earthreturn system was gradually replaced by two-wire circuits which are much less
susceptible to cross-talk and electrical noise. The rapid growth of the telephone
system was based almost entirely on the fact that the subscriber could use the system
with the minimal amount of training. The ease of operation of the telephone
outweighed the disadvantages of having no written record of conversations and
the requirement that both parties have to be available for the call at the same time.
The basic central office responds to a signal from the subscriber (calling party)
indicating that he wants service. A buzzer excited by current from a hand-cranked
magneto was the standard. The telephone operator answers and finds out whom
(called party) the calling party wants to talk to. The operator then signals to the
called party by connecting his own hand-cranked magneto to the line and cranking it
to ring a bell on the called party’s premises. When and if the called party responds,
he connects the lines of the two parties together and withdraws until the conversation
is over, at which point he disconnects the lines. In order to carry out his function, the
operator had to have access to all the lines connected to the exchange. This was not a
problem in an exchange with less than fifty lines but as the system grew, more
operators were required for each group of fifty subscribers. If the calling party and
called party belong to the same group of fifty, the above sequence was followed. If
they belong to two different operators, it was necessary for the two operators to have
a verbal consultation before the connection could be made. The errors, delays and
misunderstandings in large central offices led to a re-organization whereby each
operator responded to only fifty incoming lines but had access to all the outgoing
lines. Another improvement in the system was to replace all the batteries on the
subscribers’ premises with one battery in the central office.
The motivation for the changeover from manual switching to the automatic
telephone exchange was not, as one would expect, the inability of the central office
to cope with the increasing volume of traffic. It was because the operators could
listen to the conversations. The inventor of the automatic exchange, Almon B.
Strowger (1839–1902), after whom the system was named, was an undertaker in
Kansas City around the 1890s. There was another undertaker in the city whose wife
worked in the local telephone exchange; whenever someone died in the city, the
telephone operators were the first to know and the wife would pass a message to the
husband, giving him a head-start on his competitor [8]. The automatic exchange
certainly improved the security of telephone conversations; it was also one more
example of machines replacing people.
The success of the telephone system led to a large number of small telephone
companies being formed to service the local urban communities. Pressure to interconnect the various urban centres soon grew and techniques for transmission over
longer distances had to be developed. These included amplification and inductive
loading. Since these transmission lines (trunks) were expensive to construct and
maintain, techniques for transmission of more than one message (multiplex) over the
trunk at any one time became a matter of great concern and an area of rapid
In 1864, James Maxwell (1831–1879), a Scottish physicist, produced his theory of
the electromagnetic field which predicted that electromagnetic waves can propagate
in free space at a velocity equal to that of light [9]. Experimental confirmation of this
theory had to wait until 1887 when Heinrich Hertz (1857–1894) constructed the first
high-frequency oscillator. When a voltage was induced in an induction coil
connected across a spark gap, a discharge would occur across the gap setting up a
damped sinusoidal high-frequency oscillation. The frequency of the oscillation could
be changed by varying the capacitance of the gap by connecting metal plates to it.
The detector that he used consisted of a second coil connected to a much shorter
spark gap. The observation of sparks across the detector gap when the induction coil
was excited showed that the electromagnetic energy from the first coil was reaching
the second coil through space. These experiments were in many ways similar to
those carried out in 1839 by Joseph Henry (1797–1878). Several scientists made
valuable contributions to the subject, such as Edouard Branly (1844–1940) who
invented the ‘‘coherer’’ for wave detection, Aleksandr Popov (1859–1906) and
Oliver Lodge (1851–1940) who discovered the phenomenon of resonance.
In 1896, Guglielmo Marconi (1874–1937) left Italy for England where he worked
in cooperation with the British Post Office on ‘‘wireless telegraph’’. A year later, he
registered his ‘‘Wireless Telegraphy and Signal Co. Ltd’’ in London, England to
exploit the new technology of radio. On the 12th of December 1901, Marconi
received the letter ‘‘S’’ in Morse code at St, Johns, Newfoundland on his receiver
whose antenna was held up by a kite, the antenna which he had constructed for the
purpose having been destroyed by heavy winds. He had confounded the many
skeptics who thought that the curvature of the earth would make radio transmission
impossible [10].
Up to this point, no use had been made of ‘‘electronics’’ in telecommunication:
high-frequency signals for radio were generated mechanically. The first electronic
device, the diode, was invented by Sir John Ambrose Fleming (1849–1945) in 1904.
He was investigating the ‘‘Edison effect’’ that is, the accumulation of dark deposits
on the inside wall of the glass envelope of the electric light bulb. This phenomenon
was evidently undesirable because it reduced the brightness of the lamp. He was
convinced that the dark patches were formed by charge particles of carbon given off
by the hot carbon filament. He inserted a probe into the bulb because he had the idea
that he could prevent the charged particles from accumulating by applying a voltage
to the probe. He soon realized that, when the probe was held at a positive potential
with respect to the filament, there was a current in the probe but when it had a
negative potential no current would flow: he had invented the diode. He was granted
the first patent in electronics for his effort. Fleming went on to use his diode in the
detection of radio signals – a practice which has survived to this day.
The next major contribution to the development of radio was made by Lee
DeForest (1873–1961). He got into legal trouble with Marconi, the owner of the
Fleming diode patent, when he obtained a patent of his own on a device very similar
to Fleming’s. He went on to introduce a piece of platinum formed into a zig-zag
around the filament and soon realised that, by applying a voltage to what he called
the ‘‘grid’’, he could control the current flowing through the diode. This was, of
course, the triode – a vital element in the development of amplifiers and oscillators.
Shortly after the establishment of the telegraph, the transmission of images by
electrical means was attempted by Giovanni Caselli (1815–1891) in France. His
technique was to break up the picture into little pieces and send a coded signal for
each piece over a telegraph line. The picture was then reconstituted at the receiving
end. The system was slow, even for static images, but it established the basic
principles for image transmission; that is, the break up of the picture into some
elemental form (scanning), the quantization of each element in terms of how bright it
is (coding), and the need for some kind of synchronization between the transmitter
and the receiver. Subsequent practical image transmission schemes, whether
mechanical or electronic, had these basic units.
The discovery in 1873 by Joseph May, a telegraph operator at the Irish end of the
transatlantic cable, that when a selenium resistor was exposed to sunlight its
resistance decreased, led to the development of a light-to-current transducer.
Subsequently, various schemes for image transmission based on this discovery
were devised by George Carey, William Ayrton (1847–1908), John Perry and others.
None of these was successful because they lacked an adequate scanning system and
each element of the picture had to be sent on a separate circuit, making them quite
In 1884, Paul Nipkow (1860–1940) was granted a patent in Germany for what
became known as the Nipkow Disc. This consisted of a series of holes drilled in the
form of spirals in a disc. When an image is viewed through a second disc with
similar holes driven in synchronism with the first, the observed effect was scanning
point-to-point to form a complete line and line-by-line to cover the complete picture.
This was a practical scheme since the point-to-point brightness of the picture could
be transmitted and received serially on a single circuit. The persistence of an image
on the human eye could be relied on to create the impression of a complete scene
when, in fact, the information is presented point-by-point. Nipkow’s scheme could
not be exploited until 1927 when photosensitive cells, photomultipliers, electron
tube amplifiers and the cathode ray tube had been invented and had attained
sufficient maturity to process the signals at an acceptable speed for television.
Several people made significant contributions to the development of the components
as well as to the system. However, two people, Charles Jenkins (1867–1934) and
John Baird (1888–1946), are credited with the successful transmission of images at
about the same time. They both used the Nipkow disc. Mechanical scanning
methods of various forms were used with reasonable success until about 1930
when Vladimir Zworykin (1889–1982) invented the ‘‘iconoscope’’ and Philo Farnsworth (1906–1971) the electronic camera tube, which he called the ‘‘image
dissector’’. These inventions finally removed all the moving parts from television
scanning systems and replaced them with electronic scanning [11]. The application
of very-high-frequency carriers and the use of coaxial cables have contributed
significantly to the quality of the pictures. The use of color in television had been
shown to be feasible in 1930 but would not be available to the general public until
the mid-1960s. By the 1980s, satellite communication systems brought a large
number of television programs to viewers who could afford the cost of the dish
antenna. By the beginning of the 21st century, the dish antennas had shrunk in size
from over 3 m to less than 70 cm and the signal had changed into digital form.
Electrical telecommunication started with a single wire with a ground return, but, as
the system grew, the common ground return had to be replaced with a return wire,
hence the advent of the open-wire telephone line. The open-wire system with its
forests of telegraph poles along city streets strung with an endless array of wires
eventually gave way to the twisted pair cable. The twisted pair cable owes its
existence to improved insulating materials, mainly plastics, which reduced the space
requirements of the cable. The bandwidth of an unloaded twisted pair is approximately 4 kHz and it decreases rapidly with length. This can be improved by
connecting inductors (loading coils) in series with the line at specific distances
and by various equalization schemes to about 1 MHz. However, the twisted pair has
found a niche in the modern telephone system where its bandwidth approximately
matches that required for analog audio communication. This is still the dominant
mode of telephone communication up to the central office. Beyond the central office
the network of inter-office trunks use a variety of conduits for the transmission of the
Increased bandwidth alone was not an answer to the expanding telecommunication traffic. High-frequency carriers had to be developed in order to exploit fully the
bandwidth capability of new telecommunication media such as coaxial cables,
terrestrial microwave networks and fiber optics. The development of the coaxial
cable, which confines the electromagnetic wave to the annular space between the two
concentric conductors, reduced significantly the radiation losses that would otherwise occur. As a result the bandwidth was increased to approximately 1 GHz and
attenuation was reduced. Terrestrial as well as satellite microwave communication
systems have further expanded the bandwidth into the terrahertz range and, for those
who can afford the dish antenna and its associated equipment, it has increased the
number of television channels available to over 800. The application of fiber optics
to telecommunication has extended the channel bandwidth to that of visible light
(1 1012 Hz). It is now possible for one optical fiber to carry as many as 300 109
telephone channels at the same time.
An increasingly dominant factor in telecommunication is the enormous popularity of digital techniques. The information is reduced to a train of pulses (binary
digits; 1s and 0s) and sent over the channel. The limited bandwidth, phase change
and the noise in the channel cause the signal to deteriorate so it is necessary to
‘‘refresh’’ or regenerate the signal at various points along the channel. This is
accomplished by using repeaters whose function is to determine whether the digit
sent was a 1 or a 0 and to generate the appropriate new digits and transmit them. At
the receiving end, the digits are converted back into an analog signal. The compact
disc music recording system is a common example of this technique. Although the
need for information transfer between computers spurred on the development of
digital communication, speech signals increasingly are being converted into digital
form for telephone transmission.
The use of personal computers as a means of communication gained enormous
popularity in the last decade of the 20th century. However, computer science experts
have used the ARPANET (Advanced Research Projects Agency of the U.S.
Department of Defense) for communication between computers since 1969. The
basic idea was to enable scientists in different geographic locations to share their
research results [12] and also, as a money saving scheme, their computing resources.
The first four sites to be connected were the Stanford Research Institute, the
University of California at Los Angeles, the University of California at Santa
Barbara, and the University of Utah in Salt Lake City. The messages traveling
between these centers were over 50 kbps telephone lines. In 1962 when the
ARPANET was being designed, the Cold War was in full swing and so one of the
specifications for the design was that the network should survive a nuclear attack in
which parts of it were knocked out [13]. Needless to say, this feature of the design
was never tested! The ideal structure was a network with every node connected to
every other node (high redundancy) so that, if a part of the network went down for
whatever reason, the traffic could be routed around the trouble spot. Moreover, in
such a design all nodes are of equal importance, hence there is no one node the
destruction of which would cripple the network. The configuration of the network is
shown in Figure 1.4. This is similar to the electric power grid which was designed to
provide electric power to consumers with a maximum reliability service.
Another design feature of the ARPANET which further improved its robust nature
was the use of ‘‘packet’’ switching. In packet switching, the incoming message is first
divided into smaller packets of binary code. Each packet is labeled with a number
and the address of its destination and then transmitted to the next node when the
local router can accommodate the packet. Each packet, in theory, can travel from the
source to the destination by a different route and arrive at different times. At its
destination the packets are re-assembled in the proper order ready for the recipient.
The strength of packet switching is the fact that, if a number of nodes are put out of
operation, the packets will still find their way to their proper destination by way of
the remaining operational nodes, in perhaps a longer time. Moreover, errordetection codes can be included with each packet and, when errors are present in
a packet, that packet can be re-sent.
The ARPANET grew so that by 1983 there were 562 sites connected. By 1992,
the number of ‘‘host’’ or ‘‘gateway’’ computers connected to it had reached one
million. Four years later, the number was 12 million. It has been estimated that by
the year 2000 the number with access to the Internet worldwide will be 100 million
[14]. The term ‘‘Internet’’ came into use in 1984, and this was also the time when the
Figure 1.4. Each node of the Internet is connected to all the neighboring nodes. The increased
redundancy implies a high level of reliability. The design of the electric power grid follows the
same principle for the same reason.
United States Department of Defense handed over the oversight of the network to the
National Science Foundation. The Internet is currently run in a very loose fashion by
a number of volunteer organizations whose membership is open to the public. Their
main activity is centered around the registration of names, numbers and addresses of
the users of the system.
The Internet is a collection of a large number of computers connected together by
telephone lines, coaxial cables, optical fiber cables and communication satellites
with set protocols to enable communication between them and also to control the
flow of traffic.
What factors have contributed to the unprecedented growth of the Internet?
Personal computers have been in common use in scientific laboratories and in
universities since the mid-1980s but they were mostly used for calculation,
information storage and retrieval. Many businesses acquired desk-top computers
for preparing invoices, word-processing and general book-keeping. Some enthusiasts owned their own personal computers and some belonged to clubs for the
exchange of computer software which they had developed. The growth of computing
power of the personal computer was one of the pivotal developments that made the
Internet possible. In 1971, the Intel Corporation produced its first microprocessor,
the Intel 4004. It was used in a calculator and its clock frequency (an indication of
how fast it operates) was 108 kHz. The following year the Corporation produced the
Intel 8008 which was twice as fast (200 kHz) as the 4004 and it was used in 1974 in
a predecessor of the first personal computer. Also in 1974, Intel produced the 8080
which was clocked at 2 MHz. The 8080 was marketed to computer enthusiasts as
part of a kit and it very quickly became the ‘‘brains’’ of the modern personal
computer. By the year 2000, the Intel Pentium III processor had achieved a clocking
speed of 1.13 GHz (over four orders of magnitude faster than the original 4004).
This phenomenal increase in speed was coupled to an equally incredible decrease in
price which made the personal computer affordable to the general public. A
hypothetical comparison with the automobile industry in 1983 was as follows:
‘‘If the automobile business had developed like the computer business, a Rolls-Royce
would now cost $2.75 and run 3 million miles on a gallon of gas’’ [15].
Even this comparison was considered conservative fifteen years later. In the Spirit
of the Web, Wade Rowland amends the statement as follows:
‘‘That Rolls-Royce would now cost twenty-seven cents and run 300 million miles on a
gallon of gas’’.
The telephone system was already in place, although its capacity would have to
be expanded to carry the digital data in addition to the voice signals for which it was
designed. Most of the main communication lines carrying Internet data have been
updated, for example the copper cables (coaxial) laid across continents and on the
sea-bed are capable of 2.5 Mbps and fiber optic cables can go as high as 40 Gbps.
The building of the throughways for large volumes of data was the easy part of the
problem. The more difficult part was to get the data to its destination in the
workplaces and into the living quarters of the owners of personal computers.
Unfortunately, the cost of wiring houses with optical fiber cable or even coaxial
cable cannot be justified on economic grounds. Currently, the speed limit to data
flow is determined by the analog telephone line (sometimes referred to as a ‘‘twisted
pair’’) between the central office (or its equivalent) and the wall socket to which the
most personal computers are connected. This is popularly known as the-last-mile
problem. New circuits have been developed to speed up data transfer on the existing
twisted pair cable. These include the T1, the Integrated Services Digital Network
(ISDN), the High-bit-rate Digital Subscriber Line (HSDL) and the Asymmetrical
Digital Subscriber Line (ASDL).
In the early 1990s, cable television service had reached into a large number of
homes in North America and Europe and there was talk of them providing highbandwidth (mainly coaxial cable) conduits to subscribers for access to the Internet.
Unfortunately, the television cable network had millions of amplifiers and one-way
traps installed which restricted signal flow in only one direction. Connection to the
Internet required a bilateral flow of information and the cost of the conversion was
considered prohibitive. At the time this book was going to press, the cable television
companies were in the process of converting their networks for bilateral flow of
information (cable modems). The speed of transmission was predicted to be from
10 Mbps to 400 Mbps [16].
Another serious impediment to the free flow of data from one computer to another
was the almost incomprehensible commands required to effect computer communication. The ‘‘spoken language’’ of the computers was UNIX and this was quite
unfriendly to the uninitiated. The ‘‘point-and-click’’ feature of the computer
‘‘mouse’’ and the development of browsers, such as Netscape Navigator and the
Internet Explorer, finally lowered the threshold to a level where even computer
neophytes could successfully access information from the Internet. These ‘‘facilitators’’ were all available by 1989 and they would have a very profound effect on the
popularity of the Internet. The Internet can be seen as a network connecting various
sites where information is stored. The stored information and the technology for
transferring the information back and forth is the World Wide Web.
The World Wide Web in its infancy carried only text. Later the transmission of
graphics, in color, was added. With the increasing speed and sophistication of the
personal computer, sound and video have been added subsequently. What made the
Web particularly useful is the ability of the Internet browsers and search engines to
provide a list of Internet sites where the requested information may be located. It is
necessary to prompt the system with a set of keywords for the search to begin. At the
Web site there are ‘‘links’’ to other sites so the search can ‘‘fan out’’ in very many
An important factor that stimulated the growth of the Internet was the decision of
the United States government to turn over the running of the Internet to commercial
interests. This happened in stages. The National Science Foundation’s NSFnet was
superceded by the Advanced Networks and Services network, ANSnet, a non-profit
organization. However, in 1998 the Federal Communications Commission (FCC)
insisted that subscribers to the telephone system be billed at the same rate for voice
services as for data. The way was now open for commercial organizations to offer
their services on the Internet. The Internet Service Providers (ISP) collect money
from their subscribers for access to their servers. They in turn have to pay for access
to the long-distance, high-speed Internet backbone.
The decentralized nature of the Web and its ability to transfer information
bilaterally meant that its users could add their own contribution to the vast
amount already present in the form of their ‘‘personal home page’’. Commercial
organizations, special-interest groups and even governments would take advantage
of the possibilities offered by the Web. But so would people and groups with hidden
and not-so-hidden agendas to propagate their own distortions. As there is no
authority to monitor the content of the Web sites, only the criminal laws of the
country in which the Web site is registered can be used to control information on the
A number of services are available to people who have access to the Internet.
Newsgroups can be found on practically any topic. These newsgroups run roundthe-clock and anyone can join in the discussion from his or her keyboard. The
participants are free to use assumed names and identities. The number of people in
any given newsgroup can vary from zero to several thousand, so it is possible to
reach a very large audience from one’s keyboard. Chat rooms are similar to
newsgroups except that the number of people ‘‘in’’ a chat room is likely to be
much smaller. It is essentially a conversational mode of communication from one’s
keyboard. One of the more popular features of the Internet is electronic mail
(e-mail). It is the nearest thing to mailing a letter to a correspondent, and although it
is not as secure as the service provided by the Post Office, it is much faster. To send
e-mail, one has to register a unique e-mail address and choose a password.
1. Statistical Yearbook 1999, UNESCO, Paris.
2. Berto, C., Telegraphes et Telephones de Valmy au Microprocesseurs, Le Livre de Poche,
3. Stumpers, F. L. H. M., ‘‘The History, Development and Future of Telecommunications in
Europe’’, IEEE Comm. Magazine, 22(5), 1984.
4. Fraser, W., Telecommunications, MacDonald & Co, London, 1957.
5. Tebo, Julian D., ‘‘The Early History of Telecommunications’’, IEEE Comm. Soc. Digest,
14(4), pp. 12–21, 1976.
6. Osborne, H. S., Alexander Graham Bell, Biographical Memoirs, Nat. Acad. Sci., Vol. 23,
pp 1–29, 1945.
7. Smith, S. F., Telephony and Telegraphy A, 2nd Ed., Oxford University Press, New York,
8. Bernal, J.D., Science in History, Vol. 2, Penguin Books Ltd, Middlesex, 1965.
9. Carassa, F., ‘‘On the 80th Anniversary of the First Transatlantic Radio Signal’’, IEEE
Antennas Propagat. Newsl., pp. 11–19, Dec., 1982.
10. Knapp, J. G. and Tebo, J. D., ‘‘The History of Television’’, IEEE Comm. Soc. Digest, 16(3),
pp 8–21, May 1978.
11. Licklider, J. C. R. and Clark, W., ‘‘On-line Man-Computer Communications’’, Massachusetts Institute of Technology, Aug., 1962.
12. Baran, P., ‘‘On Distributed Communications Networks’’, RAND Corporation, Washington
DC, Sept., 1962.
13. Zakon, R. H., ‘‘Hobbes’ Internet Timeline v5.1’’ at, Oct.,
14. ‘‘The Computer Moves In’’, Time, January 3, 1983, 10.
15. Rowland, W., Spirit of the Web, Somerville House Pub., Toronto, 1997.
Jones, C. R., Facsimile, Murray Hill Books, New York, 1949.
Costigan, D. M., Fax: The Principles and Practice of Facsimile Communication, Chilton
Book Co., Philadelphia, 1971.
McConnell, K., Bodson, D., and Urban, S., Fax: Facsimile Technology and Systems, 3rd Ed.,
Artech House, Boston, 1999.
Dodd, Annabel, Z. The Essential Guide to Telecommunications, 2nd Ed., Prentice-Hall,
Englewood Cliffs, NJ, 2000.
Lehnert, Wendy, G., Internet 101: A Beginner’s Guide to the Internet and the World Wide Web,
Addison-Wesley, Reading, MA, 1998.
A radio signal can be generated by causing an electromagnetic disturbance and
making suitable arrangements for this disturbance to be propagated in free space.
The equipment normally used for creating the disturbance is the transmitter, and the
transmitter antenna ensures the efficient propagation of the disturbance in free space.
To detect the disturbance, one needs to capture some finite portion of the electromagnetic energy and convert it into a form which is meaningful to one of the human
senses. The equipment used for this purpose is, of course, a receiver. The energy of
the disturbance is captured using an antenna and an electrical circuit then converts
the disturbance into an audible signal.
Assume for a moment that our transmitter propagated a completely arbitrary
signal (that is, the signal contained all frequencies and all amplitudes). Then no other
transmitter can operate in free space without severe interference because free space is
a common medium for the propagation of all electromagnetic waves. However, if we
restrict each transmitter to one specific frequency (that is, continuous sinusoidal
waveforms) then interference can be avoided by incorporating a narrow-band filter at
the receiver to eliminate all other frequencies except the desired one. Such a
communication channel would work quite well except that its signal cannot
convey information since a sinusoid is completely predictable and information, by
definition, must be unpredictable.
Human beings communicate primarily through speech and hearing. Normal
speech contains frequencies from approximately 100 Hz to approximately 5 kHz
and a range of amplitudes starting from a whisper to very loud shouting. An attempt
to propagate speech in free space comes up against two very severe obstacles. The
first is similar to that of the transmitters discussed earlier, in which they interfere
with each other because they share the same medium of propagation. The second
obstacle is due to the fact that low frequencies, such as speech, cannot be propagated
efficiently in free space whereas high frequencies can. Unfortunately, human beings
cannot hear frequencies above 20 kHz which is, in fact, not high enough for free
space transmission. However, if we can arrange to change some property of a
continuous sinusoidal high-frequency source in accordance with speech, then the
prospects for effective communication through free space become a distinct
possibility. Changing some property of a (high-frequency) sinusoid in accordance
with another signal, for example speech, is called modulation. It is possible to
change the amplitude of the high-frequency signal, called the carrier, in accordance
with speech and=or music. The modulation is then called amplitude modulation or
AM for short. It is also possible to change the phase angle of the carrier, in which
case we have phase modulation (PM), or the frequency, in which case we have
frequency modulation (FM).
In order to simplify the derivation of the equation for an amplitude modulated wave,
we make the simplification that the modulating signal is a sinusoid of angular
frequency os and that the carrier signal to be modulated (also sinusoidal) has an
angular frequency oc.
Let the instantaneous carrier current be
i ¼ A sin oc t
where A is the amplitude.
The amplitude modulated carrier must have the form
i ¼ ½A þ gðtÞ sin oc t
gðtÞ ¼ B sin os t
i ¼ ðA þ B sin os tÞ sin oc t
is the modulating signal. Then
The waveform is shown in Figure 2.1.
The current may then be expressed as
i ¼ ðA þ kA sin os tÞ sin oc t
Figure 2.1. Amplitude modulated wave: the carrier frequency remains sinusoidal at oc while
the envelope varies at frequency os .
The factor k is called the depth of modulation and may be expressed as a percentage.
Simplification of Equation (2.2.5) gives
i ¼ A sin oc t þ
½cos oc os Þt cosðoc þ os Þt
The frequency spectrum is shown in Figure 2.2.
From Equation (2.2.7) it is evident that modulated carrier current has three
distinct frequencies present: the carrier frequency oc, the frequency equal to the
difference between the carrier frequency and the modulating signal frequency
Figure 2.2. Frequency spectrum of the AM wave of Figure 2.1. Note that there are three
distinct frequencies present.
Figure 2.3. Frequency spectrum of the AM wave when the single frequency modulating signal
is replaced by a band of audio frequencies. Note that the information in the signal resides only in
the sidebands.
(oc os ), and the frequency equal to the sum of the carrier frequency and the
modulating signal frequency (oc þ os ). The difference and sum frequencies are
called the ‘‘lower’’ and ‘‘upper’’ sidebands, respectively.
To make the situation more realistic, let us assume that the modulating signal is
speech which contains frequencies between os1 and os2 . Then it follows from
Equation (2.2.7) that the sum and difference terms will yield a band of frequencies
symmetrical about the carrier frequency, as shown in Figure 2.3.
Figure 2.4 shows how two audio signals which would normally interfere with
each other, when transmitted simultaneously through the same medium, can be kept
separate by choosing suitable carrier frequencies in a modulating scheme. This
method of transmitting two or more signals through the same medium simultaneously is referred to as frequency-division multiplex and will be discussed in detail
in Chapter 9.
Figure 2.4. The diagram illustrates how two audio-frequency sources, which would normally
interfere with each other, can be transmitted over the same channel with no interaction.
The choice of carrier frequency for a radio transmitter is largely determined by
government regulations and international agreements. It is evident from Figure 2.4
that, in spite of frequency division multiplexing, two stations can interfere with each
other if their carrier frequencies are so close that their sidebands overlap. In theory,
every transmitter must have a unique frequency of operation and sufficient
bandwidth to ensure no interference with others. However, bandwidth is limited
by considerations such as cost and the sophistication of the transmission technique to
be used so that, in practice, two radio transmitters may operate on frequencies which
would normally cause interference so long as they propagate their signals within
specified limits of power and are located (geographically) sufficiently far apart. The
location as well as the power transmitted by each transmitter is monitored and
controlled by the government.
Once the carrier frequency is assigned to a radio station, it is very important that
it maintains that frequency as constant as possible. There are two reasons for this: (1)
if the carrier frequency were allowed to drift then the listeners would have to re-tune
their radios from time to time to keep listening to that station, which would be
unacceptable to most listeners; (2) if a station drifts (in frequency) towards the next
station, their sidebands would overlap and cause interference. The carrier signal is
usually generated by an oscillator, but to meet the required precision of the
frequency it is common practice to use a crystal-controlled oscillator. At the heart
of the crystal-controlled oscillator is a quartz crystal cut and polished to very tight
specifications which maintains the frequency of oscillation to within a few hertz of
its nominal value. The design of such an oscillator can be found in Section 2.4.6.
Figure 2.5 is a block diagram of a typical transmitter.
Figure 2.5. Block diagram showing the components which make up the AM transmitter.
Crystal-Controlled Oscillator
The purpose of the crystal oscillator is to generate the carrier signal. To minimize
interference with other transmitters, this signal must have extremely low levels of
distortion so that the transmitter operates at only one frequency. As discussed earlier,
the frequency must be kept within very tight limits, usually within a few hertz in
107 Hz. It is difficult to design an ordinary oscillator to satisfy these conditions, so it
is common practice to use a quartz crystal to enhance the frequency stability and to
reduce the harmonic distortion products.
The quartz crystal undergoes a change in its physical dimensions when a potential
difference is applied across two corresponding faces of the crystal. If the potential
difference is an alternating one, the crystal will vibrate and exhibit the phenomenon
of resonance. For a crystal, the range of frequency over which resonance is possible
is very narrow, hence the frequency stability of the crystal-controlled oscillators is
very high. In general, the larger the physical size of the crystal, the lower the
frequency at which it resonates. Thus a high-frequency crystal is necessarily small,
fragile, and has low reliability. To generate a high-frequency carrier, it is common
practice to use a low-frequency crystal to obtain a signal at a subharmonic of the
required frequency and to use a frequency multiplier to increase the frequency.
Figure 2.5 shows that the crystal-controlled oscillator is followed by a frequency
Frequency Multiplier
The purpose of the frequency multiplier is to accept an incoming signal of frequency
fc =n, where n is an integer, and to produce an output at a frequency fc. A frequency
multiplier can have a single stage of multiplication or it can have several stages. The
output of the frequency multiplier goes to the carrier input of the amplitude
Amplitude Modulator
The amplitude modulator has two inputs, the first being the carrier signal generated
by the crystal oscillator and multiplied by a suitable factor, and the second being the
modulating signal (voice or music) which is represented in Figure 2.5 by the single
frequency fs. In reality, the frequencies present in the modulating signal are in the
audio range 20–20,000 Hz. The output from the amplitude modulator consists of the
carrier, the lower and upper sidebands.
Audio Amplifier
The audio amplifier accepts its input from a microphone and supplies the necessary
gain to bring the signal level to that required by the amplitude modulator.
Radio-Frequency Power Amplifier
The power level at the output of the modulator is usually in the range of watts and
the power required to broadcast the signal effectively is in the range of tens of
kilowatts. The radio-frequency amplifier provides the power gain as well as the
necessary impedance matching to the antenna.
The antenna is the circuit element that is responsible for converting the output power
from the transmitter amplifier into an electromagnetic wave suitable for efficient
radiation in free space. Antennae take many different physical forms determined by
the frequency of operation and the radiation pattern desired. For broadcasting
purposes, an antenna that radiates its power uniformly to its listeners is desirable,
whereas in the transmission of signals where security is important (e.g. telephony),
the antenna has to be as directive as possible to reduce the possibility of its reception
by unauthorized persons.
Perhaps the simplest way to introduce the phenomenon of oscillation is to describe a
common experience of a public address system going unstable and producing an
unpleasantly loud whistle. The system consists of a microphone, an amplifier and a
loudspeaker (or loudspeakers) as shown in Figure 2.6. The amplified sound from the
Figure 2.6. The diagram illustrates how acoustic feedback can cause a public address system
to go unstable, turning the system into an oscillator.
loudspeaker may be reflected from walls and other surfaces and reach the microphone. If the reflected sound is louder than the original then it will in turn produce a
louder output at the loudspeaker which will in turn produce an even louder signal at
the microphone. It is fairly clear that this state of affairs cannot continue indefinitely;
the system reaches a limit and produces the characteristic loud whistle. Immediate
steps have to be taken to ensure that the sound level reaching the microphone is less
than that required to reach the self-sustained value. If, on the other hand, we are
interested in the generation of an oscillation, then the study of the characteristics of
the amplifying element, the conditions under which the feedback takes place, the
frequencies present in the signal and the optimization of the system to achieve
specified performance goals are in order.
The electronic oscillator is a particular example of a more general phenomenon of
systems which exhibit a periodic behavior. A mechanical example is the pendulum
which will perform simple harmonic motion at a frequency determined by its length
and the acceleration constant due to gravity, g, if the energy it loses per cycle is
replaced from an outside source. In the case of the pendulum used in clocks, the
source of energy may be a wound-up spring or a weight whose potential energy is
transferred to the pendulum. The solar system with planets performing cyclical
motion around the sun is another example of an oscillator, although this time there is
no periodic input of energy because the system is virtually lossless.
Three theoretical approaches to oscillator design are presented below. The first is
based on the idea of setting up a ‘‘lossless’’ system by canceling the losses in an LC
circuit due to the presence of (positive) resistance by using a negative resistance. The
second is based on feedback theory. The third is based on the concept of embedding
an active device and the optimization of the power output from the oscillator.
Negative Conductance Oscillator
Consider the circuit shown in Figure 2.7. The externally applied current and the
corresponding voltage are related to each other by
I ¼ G0 þ Gn þ sC þ
ðV Þ
Figure 2.7. The negative conductance oscillator has a negative conductance generating signal
power which is dissipated in the (positive) conductance. The components L and C determine the
frequency of the signal. An alternate statement is that the negative conductance cancels all the
losses in the circuit. It then oscillates losslessly at a frequency determined by L and C.
where G0 is the load conductance, Gn is the negative conductance, I is current, V is
voltage, s is the complex frequency, C is capacitance, and L is inductance. If the
circuit is that of an oscillator, the external excitation current must be zero since an
oscillator does not require an excitation current. Hence
0 ¼ G0 þ Gn þ sC þ
ðV Þ:
For a non-trivial solution, V is non-zero, therefore
G0 þ Gn þ sC þ
which gives the quadratic equation
s2 CL þ sLðG0 þ Gn Þ þ 1 ¼ 0:
The solution is then
ðG þ Gn Þ
ðG0 þ Gn Þ2
s1 ; s2 ¼ 0
4C 2
jGn j ¼ G0
that is, the system is lossless. Equation (2.4.5) becomes
¼ jo pffiffiffiffiffiffiffi
s1 ; s2 ¼ LC
which is the resonant frequency for the tuned circuit. The circuit will continue to
oscillate at this frequency as if it were in perpetual motion.
A number of devices exhibit negative conductance under appropriate bias
conditions and may be used in the design of practical oscillators of this type.
These include tunnel diodes, pentodes (N-type negative conductance), uni-junction
transistors and silicon-controlled rectifier (S-type).
The voltage–current characteristics of N- and S-type negative conductances are
shown in Figures 2.8(a) and (b), respectively.
Figure 2.8. (a) Characteristics of an N-type negative conductance device. The device has a
negative conductance in the region where the slope of the curve is negative. Examples of
practical devices which have such characteristics are the tunnel diode and the tetrode. (b)
Characteristics of an S-type negative conductance device. The device has a negative conductance in the region where the slope of the curve is negative. Examples of practical devices which
have such characteristics are the four-layer diode and the silicon controlled rectifier.
Classical Feedback Theory
Consider the system shown in Figure 2.9 where A is the gain of an amplifier and b
represents the transfer function of the feedback path. Es is the signal applied to the
input and Eo is the output of the system [1].
In the derivation that follows, it is necessary to make the following assumptions:
(1) the input impedances of both the amplifier and the feedback network are
infinite and their output impedances are zero,
(2) both A and b are complex quantities.
Figure 2.9. Classical feedback system with gain A and feedback factor b.
The gain of the amplifier alone is
Application of Kirchhoff’s Voltage Law (KVL) at the input gives
Eg ¼ Es þ bEo :
Substituting Equation (2.4.8) into Equation (2.4.9) gives
Eo ¼ AðEs þ bEo Þ
Es 1 bA
from which we obtain
Since the Es and Eo are the input and output, respectively, of the system as a whole,
we can define this as A0 where
A0 ¼
Es ð1 bAÞ
Three separate conditions must be considered that depend on the value of the
denominator of Equation (2.4.12)
(1) Positive feedback. If the modulus of ð1 bAÞ is less than unity, then the gain
of the system A0 is greater then the gain of the amplifier A and therefore the
effect of the feedback is said to be positive.
(2) Negative feedback. If the modulus of ð1 bAÞ > 1, then A0 < A.
(3) Oscillation. If the modulus of ð1 bAÞ ¼ 0 then the gain A0 is infinite
because with no input ðEs ¼ 0Þ there is still an output. In fact the system is
supplying its own input and
bA ¼ 1:
It must be noted that the waveform of the signal need not be sinusoidal and in fact it
can take any form so long as the waveform of the signal that is fed back, bEo , is
identical to the signal Eo . However, the object of this exercise is to generate a carrier
for a telecommunication system and therefore only sinusoidal signals are acceptable
– any other waveform will generate other carriers (harmonics of the fundamental)
and cause interference with transmissions of other stations.
Sinusoidal Oscillators
Since both A and b are complex quantities, condition (3) implies
jbAj ¼ 1:
Stated in words, the magnitude of the loop gain must equal unity, and
ffbA ¼ 0; 2p; 4p; etc.
Again, in words, the loop-gain phase shift must be zero or an integral multiple of 2p
radians. The condition given in Equation (2.4.13), which implies Equations (2.4.14)
and (2.4.15), is known as the Barkhausen Criterion.
These two conditions must exist simultaneously for sinusoidal oscillation to
General Form of the Oscillator
An oscillator circuit shown in Figure 2.10 [2] and an equivalent circuit is as shown in
Figure 2.11, where the amplifying element is replaced by a voltage-controlled
voltage source in series with a resistance Ro to simulate the output resistance of the
element. The amplifying element may be a tube, a transistor or an operational
The load seen by the amplifier is
ZL ¼
Z2 ðZ1 þ Z3 Þ
ðZ1 þ Z2 þ Z3 Þ
Figure 2.10. Circuit diagram for a more generalized form of the oscillator.
The amplifier gain without feedback is
ðRo þ ZL Þ
ðZ1 þ Z3 Þ
and the feedback constant is
The loop gain is
bA ¼ Av Z1 ZL
ðRo þ ZL ÞðZ1 þ Z3 Þ
Figure 2.11. The equivalent circuit of the generalized form of the oscillator. Ro represents the
output resistance of the amplifier.
Substituting for ZL as defined in Equation (2.4.16), Equation (2.4.19) becomes
bA ¼ Av Z1 Z2
½Ro ðZ1 þ Z2 þ Z3 Þ þ Z2 ðZ1 þ Z3 Þ
For simplicity, we may assume that the impedances are lossless; hence
Z1 ¼ jX1 ;
Z2 ¼ jX2
Z3 ¼ jX3
Av X1 X2
½jRo ðX1 þ X2 þ X3 Þ X2 ðX1 þ X3 Þ
Then Equation (2.4.20) becomes
bA ¼ Recall that for oscillation to occur
1 bA ¼ 0:
This means that bA must be real and hence,
X1 þ X2 þ X3 ¼ 0
X2 ¼ ðX1 þ X3 Þ:
that is,
The expression for the loop gain becomes
bA ¼ Av
Since bA ¼ 1, it follows that X1 and X2 must have opposite signs; that is, if one of
them is inductive, the other must be capacitive and X3 can be capacitive or inductive,
depending on the sign of (X1 þ X2 ). The two possibilities are shown in Figures 2.12
and 2.13, respectively.
The circuit shown in Figure 2.12 is better known as a Colpitts oscillator. The
circuit is redrawn in Figure 2.12(b) to emphasize the symmetrical structure of the
The circuit shown in Figure 2.13 is better known as a Hartley oscillator.
From the point of view of the structure of the circuits, it can be seen that they are
the same. It should be noted that the operational amplifier can be replaced by a tube
or a transistor.
Figure 2.12. (a) The generalized form of the oscillator with two of the impedances replaced by
capacitors and the third by an inductor to form a Colpitts oscillator. (b) The diagram in (a) has
been redrawn to emphasize the symmetry of the circuit.
Oscillator Design for Maximum Power Output
A major flaw in the two previous designs is that they do not anticipate the necessity
for the oscillator to supply power to a load. The theory of the design for maximum
power output from an oscillator [3] is based on the characterization of the amplifying
element (‘‘active device’’) as a two-port. A discussion of two-ports is beyond the
scope of this book but may be found in any standard text on circuit theory.
A two-port can be described in terms of its terminal voltages and currents by four
parameters: impedances, admittances, voltage ratios, and current ratios under
constraints of open or short-circuit. Without limiting the generality, assume that
the active device has been characterized in terms of the short-circuit admittance
Figure 2.13. (a) The generalized form of the oscillator with two of the impedances replaced by
inductors and the third by a capacitor to form a Hartley oscillator. (b) The diagram in (a) has been
redrawn to emphasize the symmetry of the circuit.
parameters, or Y parameters, for short. Figure 2.14 shows the two-port and its
terminal voltages and currents, which are assumed to be sinusoidal.
The Y parameters are functions of frequency and bias conditions, and in general,
complex so that
Y11 ¼ g11 þ jb11 :
The total power entering the two-port is
P ¼ V1 I1 þ V2 I2 :
The Y parameters and the terminal voltages and currents are related by
I1 ¼ Y11 V1 þ Y12 V2
I2 ¼ Y21 V1 þ Y22 V2 :
Substituting for I1 and I2 in Equation (2.4.28) gives
P ¼ Y11 jV1 j2 þ Y22 jV2 j2 þ Y12 V1 V2 þ Y21 V1 V2 :
The ratio of the output voltage, V2 , to the input voltage, V1 , can be defined as
¼ A ¼ AR þ jAI :
The real power entering the two-port is
PR ¼ jV1 j2 ½g11 þ g22 ðA2R þ A2I Þ þ ðg12 þ g21 ÞAR ðb12 þ b21 ÞAI :
Figure 2.14. A two-port representation of an active device to be used in the design of an
oscillator. Short-circuit admittance (Y) parameters are used in the design for convenience. Other
parameters could be used in the description.
This can be rearranged as follows:
2 2
ðg21 þ g12 Þ
ðb21 b21 Þ
¼ AR þ
þ AI þ
g22 jV1 j2
4g11 g22 ðg21 þ g12 Þ2 ðb21 b12 Þ2
This equation is of the form:
z ¼ ðx aÞ2 þ ðy bÞ2 þ c
and therefore it is that of a paraboloid in space with axes PR ðg22 =V1 =2 Þ, AR and AI as
shown in Figure 2.15.
It was assumed that real, positive power was supplied and dissipated in the twoport; therefore, it follows that negative values of power, as shown in Figure 2.15,
must represent power generated by the two-port and dissipated in the surrounding or
embedding circuit; that is, above the A plane, real power is supplied to the two-port,
and below it the device supplies real power to the embedding circuit. Because the
object of the exercise is to generate and supply real power to an external circuit, the
most interesting part of Figure 2.15 is the part below the A plane. It is clear that
movement towards the apex of the paraboloid represents increasing levels of power
supplied by the ‘‘active’’ two-port and that the maximum power supplied occurs at
the apex. We shall return to this remark when we consider the optimization of the
power output.
The most general embedding circuit for the two-port is as shown in Figure 2.16
with each branch made up of a conductance in parallel with a susceptance. The
susceptances can be considered as the tuned circuit which will determine the
Figure 2.15. Three-dimensional representation of the output power of the oscillator as a
function of the complex parameter A.
Figure 2.16. The general passive embedding circuit for a two-port.
frequency of oscillation and the conductances as the destination of the power
generated by the active two-port.
The embedding network can also be described in terms of a two-port as follows:
I10 ¼ ðY2 þ Y3 ÞV1 Y3 V2
I20 ¼ Y3 V1 þ ðY1 þ Y3 ÞV2 :
When the active device and the embedding are connected as shown in Figure 2.17,
the composite circuit can be described by the two-port equations which are
[Equations (2.4.29) þ (2.4.36) and Equations (2.4.30) þ (2.4.37)]:
I1 þ I10 ¼ ðY11 þ Y2 þ Y3 ÞV1 þ ðY12 Y3 ÞV2
I2 þ
¼ ðY21 Y3 ÞV1 þ ðY1 þ Y3 þ Y22 ÞV2 :
Figure 2.17. The active two-port is shown with the passive embedding connected.
For an oscillator, no external signal current is supplied at port 1 and therefore
I1 þ I10 ¼ 0. Similarly I2 þ I20 ¼ 0. From Equation (2.4.32) we have
V2 ¼ V1 ðAR þ jAI Þ
From Equation (2.4.38) we have
V1 ½Y11 þ Y2 þ Y3 þ ðAR þ jAI ÞðY12 Y3 Þ ¼ 0
and from Equation (2.4.39) we have
V1 ½Y21 Y3 þ ðAR þ jAI ÞðY1 þ Y3 þ Y22 Þ ¼ 0:
For non-trivial values of V1 , real and imaginary values of Equations (2.4.41) and
(2.4.42) are separately equal to zero; that is,
g11 þ G2 þ G3 þ AR ðg12 G3 Þ AI ðb12 B3 Þ ¼ 0
b11 þ B2 þ B3 þ AR ðb12 B3 Þ þ AI ðg12 G3 Þ ¼ 0
g21 G3 þ AR ðG1 þ G3 þ g22 Þ AI ðB1 þ B3 þ b22 Þ ¼ 0
b21 B3 þ AR ðB1 þ B3 þ b22 Þ þ AI ðG1 þ G3 þ g22 Þ ¼ 0:
Equations (2.4.43) to (2.4.46) can be written in the form of a matrix as follows:
6 AI
4 0
ð1 AR Þ
2 3
3 G1
6 G2 7
6 7
6 G3 7 6 b21
ðAR 1Þ 7
76 7 ¼ 6
56 B1 7 4 g11
6 7
ð1 AR Þ 4 B2 5
ReðAy22 Þ
ImðAy22 Þ 7
ReðAy12 Þ 5
ImðAy12 Þ
All the terms in the matrix are known except G1 , G2 , G3 , B1 , B2 and B3 ; that is, there
are six unknowns but only four equations so a unique solution cannot be found
unless arbitrary values are chosen for at least two of the unknowns. Fortunately, an
oscillator normally has only one conductive load and therefore two of the three
conductances can be set to zero. The matrix equation can then be solved for one
conductance and three susceptances.
Crystal-Controlled Oscillator
The oscillator used in a transmitter has to have a very tight tolerance on the stability
of its frequency. This is necessary if interference between radio stations is to be
avoided. The drift of the frequency of an ordinary LC oscillator, for example, makes
it unsuitable for this purpose. Greater frequency stability can be achieved by using a
crystal as a part of the oscillator circuit [4]. In Section 2.3.1, the behavior of the
crystal when it is excited by an ac signal was discussed. It is evident that, since the
crystal reacts to electrical excitation, it must be possible to devise an electrical circuit
made up of inductors, resistors and capacitors whose frequency characteristics are
approximately those of the crystal. Such a circuit is shown in Figure 2.18. The
approximate circuit is reasonably accurate at frequencies close to the resonant
frequency. Over a larger frequency range a more complicated equivalent circuit has
to be used.
Typical values of the components of the equivalent circuit are C ¼ 0:0154 pF,
R ¼ 8 O, L ¼ 0:0165 H, Co ¼ 4:55 pF. The capacitance Co is due largely to the
electrodes which are attached to the crystal. The crystal will therefore resonate in the
series mode at a frequency os where
o2s ¼
which gives fs ¼ 9:984 106 Hz. It will resonate in the parallel mode at an angular
frequency given approximately by
o2p ¼ CCo
ðC þ Co Þ
which gives a resonant frequency, fp ¼ 10:001 106 Hz – a change of less than
0.2%. The corresponding quality factor of the crystal is then Qo ¼ 130;000.
Figure 2.18. (a) The equivalent circuit of the crystal and its package. (b) The electrical symbol
for the crystal.
Figure 2.19. The reactance characteristics of the crystal. Note that this is not to scale.
Figure 2.19 shows the reactance of the crystal plotted against frequency. It should
be noted that the reactance of the crystal is inductive over a narrow band of
frequency and also that both reactance and frequency are not to scale.
Figure 2.20 shows a typical crystal-controlled oscillator. The crystal is substituted
for one of the inductors in what would otherwise be classified as a Hartley oscillator.
This type of crystal-controlled oscillator is called a Pierce oscillator. Similarly, the
crystal-controlled oscillator corresponding to the Colpitts variety is called a Miller
oscillator. In the circuit shown in Figure 2.20, the active element is a field-effect
transistor whose gate-to-drain capacitance plus stray capacitance constitute C3 .
The very high Qo of the crystal ensures that the oscillator has an extremely
limited range of frequencies in which it can continue to oscillate. Various other
measures may be taken to improve the frequency stability, such as placing the crystal
in a temperature-controlled environment and the Q factor can be enhanced by
evacuating the glass envelope which protects it.
High precision oscillators are invariably connected to their load through a buffer
amplifier. This ensures that variations in the load do not affect the operation of the
The purpose of the frequency multiplier is to raise the frequency generated by the
crystal-controlled oscillator to the value required for the transmitter carrier. As
explained earlier, it is not possible to obtain physically robust crystals at high
Figure 2.20. (a) A Hartley oscillator with one of the inductors replaced by a crystal. This circuit
is called a Pierce oscillator. The field-effect transistor may be replaced by any other suitable
active device. (b) The equivalent circuit of the Pierce oscillator demonstrating its symmetrical
frequency since their physical size gets smaller as the frequency of oscillation gets
higher. The standard technique is therefore to use a crystal to generate a signal at a
frequency which is a subharmonic of the required carrier frequency and then to raise
the frequency up to the required value using a cascade of frequency multipliers.
A useful analog of a frequency multiplier is a child’s swing. With a child on the
swing, the adult must give it a push to get the swing into operation. Subsequent to
that, further supplies of energy must take place at a frequency determined by the
length of the swing and the gravitational constant of acceleration, g. It is also
necessary to supply the energy at a point in time when it enhances the swinging
action rather than oppose it; on average, the adult will have to supply energy equal to
that lost during the cycle to maintain a constant amplitude. If the energy supplied per
cycle is less than the energy lost, the amplitude of the swing will decrease to a
smaller value so as to restore the energy balance. If the energy supplied per cycle
is greater than that lost per cycle, the amplitude will grow to a new steady-state
value. The motion of the child will be very nearly a simple harmonic one if the total
energy stored in the system is large compared to the energy supplied by the adult,
that is, the system Q has to be large if the child is to execute a near-sinusoidal
motion. The most important point of this analog is that the energy does not have to
be supplied at the same frequency as the swing: it can be supplied at a subharmonic
frequency, that is, the push can be given every other cycle of the swing or every third
cycle or higher so long as enough energy is supplied to maintain the energy balance.
When the push occurs every other cycle, it is clear that the output of the system is at
twice the frequency of the input – this is a frequency multiplier with a multiplication
TABLE 2.1 Swing Analogy for a Frequency Multiplier
Swing, Including Child
Frequency Multiplier
Adult (timing)
Input signal source
Adult (energy transferred to child)
DC power supply
Length of swing, l, and gravitational constant, g Inductance, L, and capacitance, C
Air resistance and bearing friction
Energy loss in R
Frequency f ¼
Amplitude of swing
Frequency f ¼
2p LC
Amplitude of voltage or current in tank circuit
factor of two. When the energy is supplied every third cycle, the multiplication
factor of three is obtained, and so on. Evidently, there is a limit on how high the
multiplication factor can be and it is determined by the amount of variation in the
amplitude of the swing which can be tolerated. Table 2.1 shows a comparison of the
swing and the frequency multiplier.
Figure 2.21 shows a typical frequency multiplier. Energy is fed into it by applying
a suitable positive pulse to the base of the transistor. This causes the transistor to
conduct momentarily, that is, current flows from the direct current (dc) power supply
through the inductor and a finite amount of energy is stored in the inductor. The
current flow is shut off when the input pulse ends and the transistor is essentially an
open-circuit. The energy stored in the magnetic field of the inductor is transformed
into energy stored in the electric field of the capacitor. The transformation of energy
from one form to another and back again would continue indefinitely in a sinusoidal
form if the system were lossless and this would take place at a frequency determined
by the values of the inductance and capacitance. The resistance RL represents the
losses in the system – the amplitude of the sinusoid will decay with time. The
steady-state amplitude of the voltage or current will be determined by the equaliza-
Figure 2.21. A class-C amplifier to be used, with minor modifications, as a frequency
tion of the energy input and the energy output (loss). Subsequent to the initial input
pulse, all input pulses must be timed to enhance rather than oppose the stored energy
of the system. It is, of course, not necessary to have an input pulse for every cycle of
the output; the input pulse can be supplied once for every two, three or greater
number of cycles. When the input frequency is the same as the output, the frequency
multiplier (multiplication factor ¼ 1) is simply a class-C amplifier. Since a class-C
amplifier represents the simplest frequency multiplier, the design of a class-C
amplifier will be discussed next.
Class-C Amplifier
In a class-C amplifier, the current in the active device flows for a period much less
than p radians of the output waveform. The active device current waveform is
therefore highly non-sinusoidal. In a class-A or -B amplifier this would give a
correspondingly non-sinusoidal output. However, in the class-C amplifier the
collector load consists of a parallel LC circuit which is tuned to the frequency of
the input signal. The tuned circuit (sometimes referred to as a tank circuit) presents a
very high impedance at the resonant frequency to the collector of the transistor and
hence a high gain is obtained at this frequency. Other frequencies, such as harmonics
of the input frequency, are attenuated. Therefore, the output is very nearly sinusoidal.
In a practical class-C amplifier, the base of the transistor is held at a voltage such
that the transistor is in the off state. The input signal brings the transistor into
conduction at its positive peaks and causes enough current to flow in the inductor to
store energy equal to that dissipated in the load resistor RL and at other sites in the
circuit. When the input signal drops below the threshold of the transistor, it switches
off and the LC tank oscillates freely with a sinusoidal waveform at the resonant
frequency of the tank circuit. Because of the dissipation in the circuit, the waveform
is actually an exponentially damped sinusoid as shown in Figure 2.22.
The collector voltage, the base voltage and collector current waveforms are
shown in Figure 2.23. Note that the quiescent value of the collector voltage
waveform is Vcc and that its maximum amplitude is 2Vcc .
The conversion efficiency of the amplifier is given by:
Z ¼ (ac power output)=(dc power input):
A class-C amplifier can have a relatively high conversion efficiency, usually about
85%, because the dc current flows for a very short part of the cycle and this happens
when the collector voltage is at its lowest value. Thus the power lost in the transistor
is minimal.
Converting the Class-C Amplifier into a Frequency Multiplier
To convert a class-C amplifier into a frequency multiplier with a multiplication factor
of 2, the L and C of the tank circuit are chosen to resonate at 2oo when the input
signal frequency is oo . Successful operation of the system demands that the Q factor
Figure 2.22. The output waveform of a class-C amplifier after a single pulse excitation. Note
the sinusoidal waveform and the exponential decay of the envelope.
Figure 2.23. The collector voltage (vc ), base voltage (vbe ) and collector current (ic ) of the classC amplifier. Note that the base voltage need not be sinusoidal for the collector voltage to be
of the circuit is sufficiently high so that the amount of damping at the end of the
second cycle is negligible. Higher multiplication factors are possible with the
damping problem progressively getting worse as the multiplication factor increases.
Figure 2.24 shows the input current and output voltage waveforms of a frequency
multiplier with a multiplication factor of 2. Note the effect of damping on the
amplitude of the output voltage.
Example 2.5.1 Frequency Multiplier. A frequency multiplier with a multiplication factor of 2 is driven by an input current whose waveform can be assumed to be a
half-sinusoid with a peak value of 200 mA at a frequency of 2=p MHz. The input
current flows for a period corresponding to 5 of one cycle of the output waveform
and the average input impedance over this period is 750 O resistive. The following
data applies to the circuit:
multiplier load ¼ 50 O resistive
transformer turns ratio ¼ 15 : 1
transformer coupling, k ¼ 1
dc voltage supply ¼ 15 V
current gain of the transistor ¼ 100
loaded Q factor of the transformer primary ¼ 50 (min).
Figure 2.24. The output voltage waveform of a times-3 frequency multiplier and its driving
current. The decay is exaggerated to show the effect of low Q factor.
Calculate the following:
the value of the primary inductance
the value of the tuning capacitance
the impedance of the collector load at the output frequency
the ac power dissipated in the load
the power gain of the multiplier.
Solution. A frequency multiplier is essentially a class-C amplifier whose input is
driven at a frequency which is a subharmonic of the output frequency. A suitable
circuit is shown in Figure 2.25
Input frequency ¼ 2pf ¼ 4 106 rad=s:
Output frequency ¼ 8 106 rad=s:
The input current and the output voltage waveforms are shown in Figure 2.26.
When the load is transferred to the primary the collector circuit is as shown in
Figure 2.27 with the equivalent resistor having a value n2 RL ¼ ð15Þ2 50 ¼
11:25 kO.
(a) Assuming that the tank circuit is lossless, we can assign the loss in n2 RL to
the inductor and determine the series equivalent RL circuit (see Figure 2.28).
Equating the impedances,
joLp Rp
¼ Ro þ joLo :
Rp þ joLp
Figure 2.25. Circuit diagram for the frequency multiplier example.
Figure 2.26. Collector and base current waveforms of the frequency multiplier example.
Rationalizing the left-hand side and equating real and imaginary parts gives
Ro ¼
o2 L2p Rp
R2p þ o2 L2p
Lo ¼
Lp R2p
R2p þ o2 L2p
Figure 2.27. The frequency multiplier with the transformer load transferred to the primary.
Figure 2.28. Transformation of a parallel RL to a series RL circuit.
The loaded Q factor of the tank circuit has a minimum value of 50; therefore
Qo ¼ 50 ¼
n2 R L
The primary inductance is
Lp ¼
n2 R L
11:25 103
¼ 28:13 mH:
50o 50 8 106
(b) For the tank circuit at resonance, then
Lp C
¼ 555 pF
o2 Lp
o2 ¼
(c) Because the parallel LC tank circuit is at resonance, it is actually an opencircuit. Therefore the load seen by the collector is n2 RL ¼ 11:25 kO.
(d) The base and collector currents are related by
ic ¼ bib ¼ 20 mA peak:
Current in the secondary ¼ n 20 mA peak. Since the system is in steady
state, the energy dissipation associated with the current n 20 mA peak
flowing in the load can be averaged over one cycle of the input signal (2
cycles of the output) and the ac output power calculated on that basis.
The collector current waveform is shown in Figure 2.29.
Figure 2.29. This diagram illustrates the various stages involved in the calculation of the
output power of the frequency multiplier.
The instantaneous power in the collector circuit is
pðyÞ ¼ i2c RL ¼ ðn 20 103 sin yÞ2 RL
and the average power is
i2c RL dy:
The total energy stored in the tank circuit during n radians (10.9 ns) of the
input current is
W ¼ ð10:9 nsÞ 1
i2c RL dy:
This amount of energy must be dissipated during a period corresponding to
4p radians (equivalent to 720 or 1.57 ms in time) of the output waveform
(two cycles) since the system is in equilibrium. Therefore,
10:9 109 1
Po ¼
1:57 106 p
i2c RL dy ¼ 33:75 mW:
(e) The input power is
Pin ¼
ib Rin dy ¼ 15 mW:
(f) The power gain is
G ¼ 10 log10
¼ 33:5 dB:
In Section 2.3, the basic principles of modulation were discussed in general, and
amplitude modulation was treated at some length. Briefly then, modulation is a
means of achieving multiplexing, that is, providing a common conduit in which
several messages can be transmitted at the same time without interference. In terms
of amplitude modulation, the message (voice, code or music) is used to control the
amplitude of a high-frequency sinusoidal signal (carrier). In this section, several
circuits which can be used to generate amplitude modulation will be presented with
the derivation of the appropriate design equations.
Square-law Modulator
From Equation (2.2.5) it can be seen that the modulated current has two components:
(1) A sin oc t
(2) kA sin os t sin oc t.
These suggest that the circuit required to achieve amplitude modulation must
generate the product of the carrier and the modulating signal frequency and add
to this a suitably scaled carrier signal. The multiplication can be approximated by a
non-linear element, such as a diode, and the addition by connecting the two sources
in series. Since the diode has a voltage–current characteristic which is approximately
a square-law relationship, it is commonly used in the square-law modulator. The
modulator is sometimes called the diode modulator.
Consider the circuit shown in Figure 2.30.
vs ¼ Vs cos os t
vc ¼ Vc cos oc t
and assume the following:
Figure 2.30. Using the nonlinearity of the diode to obtain amplitude modulation. The circuit is
tuned to oc with sufficient bandwidth to select the carrier and sidebands and attenuate all
spurious signals.
(1) Operation is at the resonant frequency of the circuit; hence
o2 ¼
(2) The diode current is given by
i ¼ a1 v þ a2 v 2 þ :
Since the two voltage sources are connected in series,
v ¼ Vc cos oc t þ Vs cos os t:
Substituting Equation (2.6.5) into Equation (2.6.4) gives
i ¼ a1 Vc cos oc t þ a1 Vs cos os t þ a2 Vc2 cos2 oc t þ 2a2 Vc Vs cos oc t cos os t
þ a2 Vs2 cos2 os t þ :
cos2 f ¼ 12 ð1 þ cos 2fÞ
into Equation (2.6.6) gives
i ¼ a1 Vc cos oc t þ a1 Vs cos os t þ
a2 Vs2
ð1 þ cos 2os tÞ:
a2 Vc2
ð1 þ cos 2oc tÞ þ 2a2 Vc Vs cos oc t cos os t
The LC tank circuit is tuned to oc and acts as a bandpass filter, and since oc os ,
signals of frequency os, 2os and 2oc will be filtered out, leaving
i ¼ a1 Vc cos oc t þ 2a2 Vc Vs cos oc t cos os t:
When the load RL is reflected into the primary, the voltage across the primary is then
given by
2a2 Vs
vp ¼ a1 n RL Vc 1 þ
cos os t cos oc t:
This is an amplitude modulated voltage wave, where the index of modulation is
2a2 Vs
A ¼ a1 n2 RL Vc :
v ¼ Að1 þ k cos os tÞ cos oc t:
Direct Amplitude Modulation Amplifier
Consider an amplifier into which a signal of frequency oc is fed [5, 6]. If the dc
power supply of the amplifier is varied as a function of another signal of frequency
os , where oc os , then it is evident that the amplitude of the output voltage of the
amplifier will be high when the dc supply voltage is high and low when the dc
supply voltage is low. The basic circuit for a modulator operating on this principle is
shown in Figure 2.31.
The circuit shows a class-C amplifier with its power supply connected through the
secondary winding of a transformer. Node A is connected to ground by a capacitor
whose value is chosen so that, at frequency oc, it is a short-circuit. From the point of
view of the frequency oc, node A is at ground potential, as it should be for any
amplifier. Assuming that T1 is a 1 : 1 transformer, then with vs applied to the primary,
a voltage vs will be developed in series with the dc supply voltage Vcc , such that
vcc ¼ Vcc þ vs :
vs ¼ Vs cos os t
Figure 2.31. The circuit diagram of a direct AM amplifier. Note that, in this type of modulator, a
high level of audio-frequency power is required to drive the circuit.
vc ¼ Vc cos oc t:
vcc ¼ Vcc ð1 þ m cos os tÞ
where m ¼ Vs =Vcc .
Since the output voltage is proportional to the input and the supply voltage,
vo ¼ kvcc Vc cos oc t
vo ¼ kVc Vcc ð1 þ m cos os tÞ cos oc t
where k is a constant
This is an amplitude modulated voltage. It is clear the C2 must appear to be an opencircuit at the frequency os because the dc supply source Vcc is, in fact, an ac shortcircuit. The two constraints on C2 namely thay it must be a short-circuit at oc and an
open-circuit at os are easily satisfied because, in general oc os .
Four-Quadrant Analog Multiplier
An amplitude modulated signal can be obtained by adding the carrier signal to the
product of the modulating (audio) signal and the carrier. Thus,
vo ¼ A cos oc t þ kA cos oc t B cos os t
There are a number of techniques which can be used to achieve this. One of the more
practical of these is the variable transconductance method [7]. The output current of
a common-emitter transistor amplifier depends on the input signal voltage at the
base-emitter junction and the magnitude of the emitter resistance that can be
controlled by varying the emitter current. The collector current is therefore proportional to the product of the input signal and the emitter current. Consider the circuit
shown in Figure 2.32.
Assuming that the collector currents are approximately equal to the emitter
currents, when Vk increases positively, the current I1 increases at the expense of I2
since the total current in the transistors Q1 and Q2 are fixed by the ideal current
source, I .
I1 ¼ ð1 kÞI
where k is a current gain and 1 > k > 0.
Figure 2.32. The long-tailed pair used as a variable transconductance element in an analog
Clearly, the ideal current source, I , must have two components: a dc (bias)
component and an ac (signal) component:
I ¼ Idc þ Ic cos oc t
I2 ¼ kI ¼ kðIdc þ Ic cos oc tÞ:
Similarly, the voltage source Vk , when it is varied sinusoidally, must have two
components: a dc (bias) and an ac (signal) component
k ¼ Vdc þ Vs cos os t:
The output current is
I2 ¼ m1 Idc Vdc þ m2 Vdc Ic cos oc t þ m3 Vs Idc cos os t þ m4 Ic Vs cos oc t cos os t
where the ms are constants. To convert I2 to an amplitude modulated wave, the term
m1 Idc Vdc þ m3 Idc Vs cos os t
will have to be removed from Equation (2.6.26) to give
I2 ¼ Að1 þ B cos os tÞ cos oc t
A ¼ m2 Ic Vdc
m4 Vs
m2 Vdc
Assuming that the circuit is to be realized in integrated circuit form, where the
transistor characteristics can be very closely matched, the removal of Equation
(2.6.27) is most easily achieved by using the concept of ‘‘balancing’’, that is, by
duplicating the circuit and interconnecting the two parts as shown in Figure 2.33.
Note that the ideal current source I 0 has a dc current equal to the dc component of
I . A practical realization of the circuit with various bias arrangements, temperature
compensation and a number of other features is shown in Figure 2.34 and it is
manufactured by Motorola Semiconductor Products, Inc. and designated ‘‘MC1595 Four-quadrant Multiplier’’:
Figure 2.33. The connection of a second long-tailed pair as shown, removes the unwanted
signals by using the concept of ‘‘balance’’. This is possible when the circuit is integrated on a
single chip because circuit integration produces matched transistors which can be relied on to
track together with temperature.
Figure 2.34. The circuit diagram of the four-quadrant multiplier (MC1595) manufactured by
Motorola Semiconductor Products Inc. Reprinted with permission from the Linear and Integrated
Circuits Handbook, 1983.
There are two types of amplifiers which can be used to boost the output of the
microphone to the level required by the modulator – class-A and class-B. Both of
these amplifiers are ‘‘linear’’ in the sense that the output is nominally an exact replica
of the input. The difference is that a class-A amplifier has current flowing in the
active device at all times, and generally its output power and conversion efficiency
are modest. It is most often used with low-level signals. Class-B amplifiers have
current flowing in the active device for only one-half of the cycle. This situation
would normally lead to an unacceptable level of distortion so two active devices are
used in a ‘‘push–pull’’ configuration. The advantage of this type of amplifier is that it
has a higher power output level than the class-A amplifier and a higher conversion
efficiency as well. The audio-frequency amplifier used in the transmitter will be
determined by the output of the signal source (microphone) and the signal level
required at the input of the modulator. It could be a combination of the two, that is, a
class-A followed by a class-B amplifier. The two types of amplifier will be discussed
in detail.
Basic Device Characteristics
Figure 2.35 shows the characteristics of a bipolar junction transistor (BJT) in the
common emitter mode. This is used as the basis of the design and can be applied
with slight modifications to other devices and other modes.
It is evident that the VCE Ic characteristics are controlled by the base current, IB ,
supplied to the transistor. When the transistor is connected as shown in Figure 2.36
and the value of IB is reduced then it follows that the collector current Ic will
decrease and so will the voltage drop across Rc until finally when Ic ¼ 0, VCE ¼ Vcc .
Figure 2.35. The collector characteristics of a junction bipolar transistor showing collector
current (Ic ) plotted against collector-emitter voltage (VCE ) for various values of base current (iB ).
Figure 2.36. The basic common-emitter amplifier with load Rc .
The transistor is said to be ‘‘cut-off’’. On the other hand, if IB is increased, the
collector current Ic will increase and so will the voltage drop across Rc until finally
VCE ¼ 0 and VRc ¼ Vcc . The transistor is said to be ‘‘in saturation’’. The collector
current is no longer under the control of the base current but determined by the value
of Rc . In practice, the value of VCE is not quite zero and, for most discrete devices,
VCE has a value of about 0.5 V.
Applying KVL to the collector current path,
Vcc ¼ VCE þ VRc ¼ VCE þ Ic Rc
Ic ¼
V þ V :
Rc cc
When Equation (2.7.3) is plotted on the characteristics of the device shown in Figure
2.37 it gives a straight line with a slope equal to ð1=Rc ) and intercepts on the y axis
equal to Vcc =Rc and on the x axis equal to Vcc . This line is known as the load line and
determines the behavior of the device when it is connected to the collector load, Rc ,
and the supply voltage Vcc .
Class-A Amplifier
Because the current in a class-A amplifier flows for the entire 2p radians of the input
ac cycle, its analysis depends on a simultaneous consideration of the dc as well as the
ac conditions of the circuit. Base-Current-Biased Class-A Amplifier. The intersection of the
load line and the characteristic representing the value of the base current, Ibq ,
establishes the quiescent or operating point of the transistor. Figure 2.37 shows how
the collector voltage and collector current can be made to change sinusoidally by
superimposing a sinusoidal current source, ib , on the quiescent value.
Figure 2.37. The operation of the amplifier is illustrated on the Ic VCE characteristics by the
load line with corresponding current and voltage waveforms.
The circuit can be made to provide amplification by a suitable choice of a
collector resistor Rc, a base resistor Rb, the supply voltage Vcc and by connecting
them as shown in Figure 2.38.
The design of an amplifier is a series of judicious choices and their modification,
when necessary, in order to meet the design specification. The design is therefore not
unique. The technique is best illustrated by a design example.
Figure 2.38. The common-emitter amplifier with the quiescent point determined by VBB
and Rb .
Example 2.7.1 Base-Current-Biased Class-A Amplifier. The following data
apply to the circuit shown in Figure 2.38.
Collector supply voltage, Vcc ¼ 10 V
Base supply voltage, VBB ¼ 3 V
Current gain of transistor, b ¼ 100
Base-emitter voltage drop, VBE ¼ 0:7 V (silicon)
Input voltage ¼ 0.5 volts peak-to-peak
Required voltage gain ¼ 10
Amplifier load, RL ¼ 10 kO.
Solution. The first step is to locate the point corresponding to Vcc on the VCE axis of
the transistor characteristics. The slope of the load line is determined by the value of
Rc and the choice of Rc depends largely on the current required to drive the load of
the amplifier. With an input voltage of 0.5 V peak-to-peak and a voltage gain of 10,
the output voltage will be 5.0 V peak-to-peak. The current required to drive the load
is therefore 0.5 mA peak-to-peak. To ensure that the operation of the transistor is not
adversely affected when the load draws off 0.5 mA peak-to-peak from the transistor,
we have to allow about 10 times the load current in the collector of the transistor in
the form of direct current, that is, Ic ¼ 5 mA in the quiescent state.
The next step is to determine the value of Rc . For the amplifier to handle the
maximum possible signal, it is necessary to place its quiescent point in such a
position that the collector voltage can swing as far in the positive direction as it can
swing in the negative direction. For the maximum signal, the transistor goes into
saturation at the minimum, and it is cut off at the maximum value of the collector
voltage. It then follows that the required point on the Ic axis is 10 mA. A straight line
can now be drawn to join Vcc ¼ 10 V to Ic ¼ 10 mA. From the slope of load-line,
Rc ¼ 1 kO.
The next step is to choose the value of Rb to establish the appropriate value of Ib
so that the quiescent point is in the middle of the load line. From the diagram, the
required base current
Ib ¼ Ic =b
¼ 50 mA:
Neglecting the resistance of the base-emitter diode,
Rb ¼ ðVBB VBE Þ=Ib
¼ ð3:0 0:7Þ=50 106 O
¼ 46 kO:
The diode resistance is given approximately by the empirical formula:
Rd ¼ 26 mV=ðdiode current in ampsÞ
¼ 26 103 =50 106 O
¼ 520 O:
Rd is indeed negligible compared to Rb .
The input voltage source has to be adjusted to give the required input voltage at
the base of the transistor and the coupling capacitor Co is chosen so that, at the
lowest frequency of operation, its reactance is negligibly small compared to the
load RL .
The amplifier, as designed, has the following disadvantages:
(1) It has two dc voltage supplies.
(2) Its input impedance is very low – that of a forward biased diode.
(3) Its bias point is largely determined by VBE which is sensitive to changes in
(4) The voltage gain of the amplifier is sensitive to the current gain of the
transistor. The current gain of two transistors fabricated next to each other on
an integrated circuit chip can vary by as much as a ratio of three to one so it is
necessary to design amplifiers which are essentially independent of the
current gain of the device. Resistive Divider-Biased Class-A Amplifier. The amplifier circuit
can be modified to that shown in Figure 2.39. The base of the transistor is then held
at a voltage determined by the resistors R1 and R2 . A resistor Re is connected
between the emitter and ground. The purpose of this is two-fold; it raises the input
impedance of the amplifier and redefines the voltage gain as the ratio Rc =Re .
The amplifier circuit is called ‘‘common-emitter’’ although the emitter is not
connected directly to ground. A proper common-emitter amplifier has characteristics
which are dependent on the parameters of the transistor, the absolute value of the
collector load and other factors such as the temperature at which it is used. These
disadvantages make the use of Re attractive.
As before, the steps in the design are illustrated by the following example:
Example 2.7.2
Resistive Divider-Biased Class-A Amplifier
DC supply voltage, Vcc ¼ 10 V
Current gain of transistor, b ¼ 100
Base-emitter diode voltage drop, VBE ¼ 0:7 V (silicon)
Input voltage ¼ 0:5 V peak-to-peak
Required voltage gain ¼ 10
Amplifier load resistance ¼ 10 kO.
Figure 2.39. The resistive-divider biased ‘‘common-emitter’’ amplifier. The use of the emitter
resistor Re increases the input resistance and reduces the dependence of the amplifier gain on
the transistor parameters.
Solution. To keep the transistor in conduction, the base voltage must be kept at
0.7 V above the emitter voltage. The point in the cycle when the transistor is most
likely to be cut off is at the negative peak (0.5 V peak) of the input voltage. The
minimum quiescent base voltage is therefore (0:7 þ 0:5 peak) ¼ 1.2 V.
The steps leading to the choice of a quiescent collector current remains the same
as in Example 2.7.1 and gives the value 5 mA. Assuming that the current gain of the
transistor is sufficiently high, then Ic is approximately equal to Ie . Under quiescent
conditions, the emitter is held at 0.5 V and the current flowing in Re is 5 mA. Re has a
value of 100 O.
The ac voltage at the base of the transistor (0.5 V peak) is the same as what
appears at the emitter, except for the dc shift of 0.7 V. Since the voltage gain of the
amplifier is 10, it follows that the ac signal at the collector must be 10 times as large
as what appears across the emitter. Keeping in mind that the emitter and collector
currents are approximately equal, it follows that the collector resistance Rc must be
10 Re ¼ 1 kO.
The base current required to maintain the collector current of 5 mA is given by
the relationship:
Ic ¼ b Ib
Ib ¼ 50 mA:
The current Ib is drawn from the resistive chain R1 and R2 which also define the
voltage at the base. To keep the base voltage reasonably constant for changes in Ib
due to variations in b, the current in the resistive chain should be about 10 Ib , that
is, 500 mA. The chain has 10 V across it so (R1 þ R2 Þ ¼ 20 kO. Since the base
voltage must be at 1.2 V, it follows that R1 ¼ 17:6 kO and R2 ¼ 2:4 kO. The
Figure 2.40. The emitter, base and collector voltage waveforms of the common-emitter
amplifier with dc bias voltages shown.
waveforms at the emitter, base and collector are shown in Figure 2.40. Note that any
increase in the amplitude of the input signal will cause the amplifier to clip the
output signal in both the cut-off and saturation modes.
The input signal is coupled to the amplifier by a capacitor Ci and the output to the
load by a capacitor Co : The values of Ci and Co are chosen so that at the lowest
frequency of operation their reactances are less than 10% of the impedance of the
load to which they are connected.
The advantages of this configuration over that presented in Example 2.7.1 are as
(1) Only one dc power supply is used.
(2) The input impedance is given by the parallel combination of R1 , R2 and bRe
which gives 1.74 kO.
(3) The bias point is determined by the ratio of the resistive chain R1 and R2
which is independent of temperature.
(4) The voltage gain of the amplifier is determined by the ratio of Rc =Re and it is
not a function of the current gain or any other parameter of the transistor.
Class-B Amplifier Complementary Symmetry Amplifier. The basic circuit for a
complementary symmetry amplifier is shown in Figure 2.41. To keep the circuit
simple, no biasing circuits are shown.
Figure 2.41. Basic class-B complementary symmetry amplifier. Note that the circuit is, in fact,
two emitter followers (NPN and PNP) with a common load resistor.
It is necessary to make the following assumptions:
(1) The amplitude of the input voltage is much larger than VBE . This assumption
ensures that cross-over distortion is kept at a minimum.
(2) The transistors Q1 and Q2 are exact complements of each other. This
condition (not strictly necessary since we are dealing with emitter followers)
ensures that symmetry is maintained.
(3) The dc power supply voltages, Vcc1 and Vcc2, are equal.
When the input voltage goes positive, the base-emitter junction of Q1 becomes
forward biased and current ic1 flows as shown. The base-emitter junction of Q2 is
reverse-biased and Q2 is cut off. Essentially, the circuit acts as an emitter follower.
When the input voltage goes negative, Q2 conducts a current ic2 in the direction
shown. Q1 is cut off. If ic1 and ic2 are equal, the voltage developed across the load RL
will be sinusoidal.
The operation of the circuit, in terms of the device characteristics and the load
line, can be seen in Figure 2.42. Note that the y axis of the PNP transistor, Q2 , has
been shifted to make the point Vcc coincide with the point Vcc . The advantage
gained is the simplicity of the diagram.
Figure 2.42. Composite characteristics of the complementary symmetry transistor pair with
the load line, collector voltage and current shown. The solid lines refer to Q1 , the dotted to Q2 .
The ac power output is
I^ V^
I^ V^
Pout ¼ pcffiffiffi pccffiffiffi ¼ c cc :
2 2
The current flowing in each transistor is a half-sinusoid whose average value (dc
equivalent) is I^c =p. The average current for the two transistors is (2I^c Þ=p. Since this
current is flowing against a potential difference of Vcc , the dc power input is
Pdc ¼
2Ic Vcc
The conversion efficiency is
ac power p
¼ ¼ 78:5%:
dc power 4
The complementary symmetry class-B amplifier has the following disadvantages:
(1) It requires two equal dc power supplies.
(2) Complementary pairs of transistors are not easy to manufacture, especially at
high power levels.
(3) The circuit cannot have a voltage gain since it is basically two emitter
followers with a common load.
To avoid the above disadvantages, the transformer-coupled class-B amplifier can be
used. Transformer-Coupled Class-B Amplifier. The transformer-coupled
class-B amplifier uses two center-tapped transformers, as shown in Figure 2.43. The
only assumption necessary is that the two transistors are matched. Since they are
both NPN (or PNP) this is not, in general unduly restrictive.
Figure 2.43. Circuit diagram of the transformer-coupled class-B amplifier. The waveforms of
the currents and voltages at various points are indicated. The solid lines refer to Q1 , the dotted
line to Q2 .
When the input voltage goes positive, two voltages vb1 and vb2 appear across the
secondary of the input transformer as shown. The voltage vb1 biases Q1 on and vb2
cuts off Q2 . The current ic1 , which flows in Q1 and one-half of the primary of the
output transformer, is a half-sinusoid. The voltage developed across the secondary of
the output transformer is therefore a half-sinusoid as shown.
When the input voltage goes negative, vb1 biases Q1 off and vb2 brings Q2 into
conduction. The current ic2 , that flows in Q2 and the other half of the primary of the
output transformer, is a half-sinusoid but it flows in the opposite direction to ic1 . The
voltage that is developed across the output transformer is then a negative-going halfsinusoid. The resultant current in the load iL is a complete sinusoid.
It should be noted that at any point in time only one-half of the primary of the
output transformer is coupled to the secondary.
Transformers are bulky, heavy and bandwidth limited so class-B amplifier
designs which do not use transformers are available [8, 9].
Example 2.7.3 Class-B Amplifier. The idealized output characteristics of a
transistor to be used in a class-B push–pull amplifier are given in Figure 2.44.
The following parameters are used:
(1) Supply voltage Vcc ¼ 50 V.
(2) The amplifier load resistance ¼ 8 O.
(3) The emitter-base diode characteristics can be approximated by a straight line
as shown in Figure 2.45.
Figure 2.44. The transistor characteristics have been idealized and the maximum collector
power dissipation curve superimposed on it.
Figure 2.45. To keep the example simple, the nonlinear emitter-base diode has been made
Calculate the following:
the turns ratio of the output transformer,
the conversion efficiency,
the input power,
the output power,
the power gain in decibels.
Solution. The collector characteristics has a rectangular hyperbola drawn on it to
indicate the maximum dissipation capability of the device. The load line may then
touch the locus of the maximum dissipation line but may not intersect it. In the
absence of any requirement on the maximum power output of the amplifier, it may
be assumed that it should provide the maximum power. The load line can then be
drawn through VCE ¼ 50 V and tangential to the maximum dissipation curve. This
gives the apparent collector load resistance value of 250 O.
A suitable circuit for the class-B amplifier is shown in Figure 2.46. At any point
in time only one-half of the transformer primary is linked to the secondary. The turns
ratio a : 1 refers to this link and not to the whole of the primary which is, in fact,
2a : 1. The load ‘‘seen’’ by the collector of the transistor is to be 250 O but this
represents a load RL transformed through the transformer. Hence
250 ¼ a2 RL
a ¼ 5:59:
Figure 2.46. The circuit diagram for Example 2.7.3.
The collector characteristics of the ‘‘composite’’ transistor and the ac load line,
together with the collector current and the corresponding collector voltage, are
shown in Figure 2.47.
The following parameters can be derived:
Current gain of the transistor is approximately equal to 50.
Output voltage ¼ Vcc = 2 V (rms).
Output current ¼ I^c = 2 A (rms).
Output power, Po ¼ VI cos fðwhere f ¼ p radiansÞ:
Po ¼ ð50 0:2Þ=2 W
¼ 5 W:
Figure 2.47. The composite transistor collector characteristics have been idealized with the
load line, collector current and voltage waveforms shown. The solid lines refer to Q1 , the dotted
line to Q2 .
The negative sign can be ignored since it simply indicates that the transistor is
generating the power, not dissipating it.
(e) The base current required to drive the maximum collector current I^b ¼ 4 mA.
Using the approximation Ic ¼ IE and the graph of IE Veb , for IE ¼ 200 mA,
V^ eb ¼ 1:0 V
Input power; Pi ¼ ðV^ eb I^b Þ=2
¼ 0:002 W
Power gain ¼ 10 log10 Po =Pi ¼ 33:9 dB:
The choice of an amplifier to boost the power coming out of the modulator to the
level required to drive the antenna can be made when the following points have been
(1) The amplifier must be linear so as to preserve the nature of the modulated
(2) The output power of the amplifier can vary from a few watts to hundreds of
(3) A high conversion efficiency is a necessity when high power is a requirement.
The portion of the dc power supplied to the amplifier which is not converted
into useable ac signal is dissipated in the amplifier components. Suitable
arrangements have to be made to remove the heat. This can add to the
complexity and the cost of operation of the transmitter.
(4) The signal to be amplified consists of the carrier and the two sidebands which
normally constitute a narrowband signal.
The usual choice is a class-B (actually a class-AB) amplifier with a tuned load.
The class-B is chosen for its linearity, high output power and high conversion
efficiency and the tuned load ensures that it is a narrowband amplifier. For relatively
low power requirements, transistors may be used in the output stage of the amplifier.
When the power output required is in excess of a few hundred watts, vacuum tubes
are used.
A suitable circuit for the radio-frequency power amplifier is shown in Figure 2.48.
Its operation is the same as that of the class-B audio-frequency amplifier discussed in
Figure 2.48. Class-B amplifier with a tuned input and output. The amplifier is designed to
amplify the carrier and sidebands only.
Example 2.8.1 Class-B Tuned Amplifier. The final stage of a class-B radiofrequency power amplifier in a radio transmitter has a tuned circuit in which
C ¼ 2500 pF. The carrier frequency oc ¼ 5 105 rad=s and the modulating signal
has frequencies ranging from os1 ¼ 20 rad=s to os2 ¼ 10 103 rad=s. The power is
coupled through a transformer of turns ratio 10 : 1 to an antenna whose impedance is
assumed to be 250 O resistive.
Calculate the following:
(a) The inductance in the tuned circuit.
(b) The Q factor of the tuned circuit if the extreme edges of the sidebands are to
be within 3 dB of the carrier.
(c) What is the maximum winding resistance that the coil in the tuned circuit can
have under the condition in (b)?
(d) What is the Q factor of the coil alone?
(a) o2c ¼ 1=ðLCÞ
L ¼ 1=ð25 1010 2500 1012 Þ H
¼ 1:6 mH:
(b) Now
Qo ¼
o2 o1
where o1 and o2 are the two 7 3 dB frequencies
Q ¼ ð5 105 =20 103 Þ ¼ 25:
Figure 2.49. (a) The tuned collector circuit of the amplifier with the resistive load transferred to
the primary of the transformer. (b) The collector circuit when the series RL has been replaced
with the parallel equivalent. (c) The collector equivalent circuit when the two resistive elements
have been combined.
(c) There are two sources of loss in the tuned circuit which contribute to give a
loaded Q of 25. They are:
(1) The antenna resistance1 coupled into the tuned circuit by the transformer
(n2 RL ).
(2) The winding resistance of the coil, r.
The equivalent circuit of the load is shown in Figure 2.49(a–c) in three stages. Stage
a shows the reflected resistance in parallel with the tuned load and the winding
resistance in series with the coil. Stage b is the equivalent circuit when the winding
resistance, r, is converted into a resistance R0 in parallel with the inductance. Stage c
shows the equivalent circuit when the resistances n2 RL and R0 are combined to form R00 .
Now, if the two circuits shown in Figure 2.50 are equivalent then
r þ joL ¼
joLp Rp
Rp þ joLp
Figure 2.50. The conversion of the parallel Rp Lp circuit to the series RL equivalent.
Antenna resistance is not the physical resistance of the antenna; it is an equivalent resistance in which the
radiated power is dissipated.
Rationalizing and equating real and imaginary parts:
o2 L2p Rp
R2p þ o2 L2p
Lp R2p
R2p þ o2 L2p
o2 L2p
The Q factor is
When Q 1; RP oLP , then
L ’ Lp :
From equivalent circuit c,
R00 ¼ 25 oL0 ¼ 20 kO:
The resistance, R00 , has two components as shown in circuit b, n2 RL , in parallel with
R0 :
n2 RL ¼ 102 250 ¼ 25 kO
¼ 20
ð25 þ R0 Þ
which gives
R0 ¼ 100 kO:
Using Equation (2.8.1) when Rp ¼ R gives
r ¼ 6:4 O:
(d) The Q factor of the coil only is
Q0 ¼
¼ 125:
Antenna structures can take many physical forms, from the ‘‘whip’’ antenna used
mostly for in-car radios, through ‘‘rabbit ears’’ for television to the microwave
parabolic ‘‘dish’’ used in satellite communication. These are only the most common
examples of antennas which are used for reception of broadcast signals. The antenna
is the last processor of the signal at the transmitting end and the first at the receiving
end. Antennae vary widely in shape, size and complexity, depending on the
frequency of operation and the desired field pattern.
Radio communication in free space is possible because, when an alternating
current flows in a conductor, part of the energy is lost in the form of electromagnetic
radiation into free space. When the frequency of the current is low, the radiation
‘‘loss’’ is very small, but as the frequency increases, substantial losses can occur. In
designing an antenna the object is to construct a structure which will maximize the
radiated energy in a given direction or over a geographic area.
A useful analog of how an antenna radiates energy is given by a body which is
floating in the middle of a pond on a windless day. For the purposes of this
description we may assume that the body is a beach ball. If we can get a man
suspended from a crane directly above the beach ball to push the ball very slowly
into the water and release it equally slowly then the energy expended in pushing the
ball into the water is recovered when it is released and very little will be lost. If the
man increases the frequency at which he pushes and releases the ball, an increasing
amount of energy will be lost in creating waves which will radiate from the ball
outwards. The amount of radiated energy at any given point on the pond will bear an
inverse relationship to the distance from the ball. It is necessary to bear in mind that
the waves on the pond are surface waves (two dimensional) whereas those produced
by an antenna are three dimensional.
The speed of propagation of the radio wave is the same as that of light
(c ¼ 3 108 m=s). The wavelength and the frequency are related by
lf ¼ c
where l is the wavelength in meters and f is the frequency in hertz. Electromagnetic
radiation has two vectors; the E and H which are orthogonal to each other and to the
direction of propagation.
A detailed study of antenna design is beyond the scope of this text and the
interested reader may consult a suitable textbook on the subject. The discussion will
therefore be qualitative.
Radiation Pattern of an Isolated Dipole
A simple way to gain insight into the operation of a dipole is to start from an opencircuit balanced transmission line. Such a transmission line will have a standingwave distribution of currents in opposite directions so that the net radiation from the
structure will be quite small. Figure 2.51(a) shows the transmission line and the
standing-wave pattern. It is assumed that the diameter of the conductors is
infinitesimally small compared to the wavelength of the signal to be transmitted.
If the transmission line is bent at right angles as shown in Figure 2.51(b), the
structure is called a dipole and it has the current distribution shown. Since the
oppositely directed currents are now far apart, they do not counteract each other and
hence the dipole is a good radiator of electromagnetic waves.
If it is assumed that 2l ¼ l=2, the dipole is then called a half -wavelength dipole.
The E field of the half-wavelength dipole when plotted gives a circle tangential to the
axis of the dipole touching the axis at the mid-point. In three dimensions, the pattern
is a doughnut, as shown in Figure 2.52.
Now assume that l l=2: the dipole is said to be a short dipole and the field
pattern is also a doughnut but the cross section is slightly distorted as if compressed
Monopole or Half-Dipole
The equations which describe the behavior of the isolated dipole can be used, with
slight modifications, for the monopole or half-dipole by assuming that the Earth’s
Figure 2.51. (a) The current standing wave pattern on an open-circuit balanced transmission
line. (b) The current standing wave pattern when the line is bent at a point where l ¼ l=2 to form
a dipole.
Figure 2.52. The radiation pattern of an isolated half-wavelength dipole.
surface, on which the antenna is to be placed, is a perfectly flat conductor. This
permits the treatment of the monopole and its image on the ground plane as a dipole.
Figure 2.53 shows the monopole and its image. This representation of the antenna is
closer to reality than the isolated dipole.
Field Patterns for a Vertical Grounded Antenna
The E field pattern generated by a vertical grounded antenna resembles that of the
monopole closely even though the Earth’s surface is not exactly a perfectly flat
Figure 2.53. This diagram shows how the half-wavelength dipole can be replaced with a
monopole and a flat conducting surface (ground plane).
conducting surface. Figures 2.54(a)–(e) are a series of diagrams which present the
expected approximate field pattern as the height of the vertical grounded antenna
changes. It is important to note that the height is measured in terms of the
wavelength of the signal to be radiated.
Figure 2.54. (a) Radiation pattern when h is approximately l=10. (b) Radiation pattern when h
is approximately l=4. Note the slight elongation of the pattern along the ground plane. (c) The
radiation pattern has undergone further distortion as h increases to approximately l=2. (d) As h
approaches 5l=8, the ground wave gets more elongated and the sky wave appears. (e) With h
approximately 3l=4, the sky wave has grown considerably at the expense of the ground wave.
Figure 2.54. (continued )
Amplitude modulated radio frequencies are grouped into three bands according to
the wavelength of their carrier frequencies. The carrier frequency chosen depends to
a large extent on the distance between the broadcasting station and the target
1. Long wave (low frequency). All transmission whose carrier frequencies are
less than 400 kHz are generally classified as long wave. At a frequency of
100 kHz, a quarter-wavelength antenna is 750 meters high. Such an antenna
poses several problems such as vulnerability to high winds and danger to low
flying aircraft. Long wave broadcasting stations therefore use an electromagnetically short antenna which necessarily limits their reach to a few tens of
kilometers because the short antenna has only the ground wave.
2. Medium wave. Carrier frequencies in the range 300 kHz to 3 MHz are
regarded as medium wave. The height of the antenna becomes more manageable and the possibility of using the sky wave to reach distant audiences is a
reality. Generally, it is used for local area broadcasting.
3. Short wave. Short wave generally refers to carrier frequencies between
3 MHz and 30 MHz. The wavelengths under consideration are between 100
meters and 1 meter. Antenna structures can be constructed to give specified
directional properties. Most of the energy can be put into the sky wave and the
signal can be bounced off the ionosphere (the layer of ionized gas that
surrounds the Earth) to reach receivers halfway round the world. A very severe
problem is encountered in short wave transmission, that is, the signal tends to
fade from time to time. This phenomenon is caused by the multiple paths by
which the signal can reach the receiver. It is clear that if two signals reach the
receiver by different paths such that their phase angles are 180 apart they will
cancel each other. The ionosphere sometimes experiences severe turbulence
due mainly to radiation from the Sun. Short wave transmission is therefore at
its best during the hours of darkness.
1. Nyquist H., Regeneration Theory, Bell System Tech. J., 11, 126–147, Jan. 1932.
2. Millman, J. and Halkias, C. C. Integrated Electronics: Analog & Digital Circuits and
Systems, Chp. 14, McGraw-Hill, New York, 1972.
3. Vehovec, M., Houselander, L. and Spence, R., ‘‘On Oscillator Design for Maximum
Power’’, Trans. IEEE, CT-15, 281–283, 1968.
4. Buchanan, J. P., Handbook of Piezoelectric Crystals for Radio Equipment Design, WADC
Tech. Report 56-156, U.S. Air Force, Oct. 1956.
5. Ryder J. D., Electronic Fundamentals & Applications: Integrated and Discrete Systems,
5th Ed., Prentice-Hall, Englewood Cliffs, NJ, 1976.
6. Fraser, W., Telecommunications, MacDonald, London, 1957.
7. Kaye, A. R., ‘‘A Solid-State Television Fader-Mixer Amplifier’’, J. SMPTE, 74, 602–606,
July 1965.
8. Bray, D. and Votipka, W., 25-Watt Audio Amplifier with Short-Circuit Protection,
Application Bulletin, Fairchild Semiconductor, August 1967.
9. Ruehs, R., High Power Audio Amplifiers with Short-Circuit Protection, AN-485, Motorola
Semiconductors, 1968.
An ideal voltage source has a signal consisting of a sinusoidal carrier of
frequency 1 MHz amplitude modulated by a sinusoidal signal of frequency
20 kHz to a depth of 100% and is applied to an LC series tuned circuit. The
capacitor has a value 100 pF and may be assumed to be lossless. The Q factor
of the circuit is 100. Calculate the following:
(a) The value of the inductance required to tune in the signal.
(b) The value of the coil resistance.
(c) The depth of the modulation of the current which flows in the circuit.
(d) Discuss the reason(s) for the change of the depth of modulation in (c), if
In an AM transmitter, the current flowing in the antenna is 7.5 A when the
carrier is not modulated. When a sinusoidal modulation is applied, the current
changes to 8.25 A. Calculate:
(a) The depth of modulation.
(b) The antenna current when the depth of modulation is 65%.
A feedback circuit with a b ¼ ð8:0 þ j4:0Þ 103 is connected to an amplifier
and the resulting system oscillates sinusoidally. Calculate the gain and phase
shift of the amplifier. Design a circuit which can perform the feedback function
and show how it can be connected to the amplifier. The frequency of the
oscillation is 100=2p kHz and it may be assumed that the amplifier has a high
input impedance and a low output impedance.
Show that when a source of emf E volts and internal impedance Zs ¼ Rs þ jXs
is connected to a load ZL þ jXL, maximum power is transferred to the load
(a) Rs ¼ RL
(b) Xs ¼ XL .
A load resistance 1150 O is connected across the secondary winding of a
transformer. The inductance of the winding is 80 mH and the Q factor is 32.
The primary winding has ideal characteristics and it is connected to a voltage
source in series with a capacitor C1. The frequency of the voltage source is
20 103 Hz and the internal resistance is 142 O. Determine the value of C1
and the mutual inductance between the two coils for maximum power to be
delivered to the load.
The wiper of a rheostate is moved linearly with respect to time between the
25% and 75% points repetitively at a frequency of 6 Hz. A voltage signal
Vin ¼ 10 cos oo t is connected to the terminals of the rheostat. Sketch the
waveform of the voltage across the wiper with respect to one of the terminals
when oc ¼ 2p 60 rad=s. Derive a general expression for all the frequencies
present and evaluate the amplitude of the largest component present other than
oc .
A source of sinusoidal signal can be represented by an ideal voltage source of
value 1 V peak in series with a 600 O resistance. This source is to be used to
drive a single transistor class-A amplifier coupled to a 10 O resistive load by
means of a transformer. Design a suitable amplifier given that a 12 V dc source
is available and the maximum possible power output is required. The b of the
transistor is 100 and the lowest frequency of interest is 20 Hz. The dc current in
the collector of the transistor is not to exceed 2 mA. Calculate the power gain in
decibels, assuming that 0 dB refers to the situation when the source is
connected directly to the load.
Show that an amplifier with a voltage gain A and feedback factor b will
oscillate when jbAj ¼ 1. What other condition(s) have to be met for sinusoidal
oscillation to occur? The feedback circuit to be used in an oscillator is as
shown in Figure P2.1.
Assuming an ideal voltage amplifier, derive an expression for:
(a) the frequency of oscillation
(b) the gain of the amplifier.
Evaluate (a) and (b) if C1 ¼ C2 ¼ 1000 pF and R1 ¼ R2 ¼ 20 kO.
The load of a class-C amplifier is coupled through a lossless transformer of
turns ratio 10 : 1, and it is tuned to a frequency 250=2p kHz. The average power
dissipated in the load at this frequency is 250 mW. The amplifier load is 50 O
and the loaded Q factor of the primary with the load transferred to the
secondary is 25. Calculate the following:
(f )
the primary inductance
the secondary inductance
the capacitance required to tune the primary
the frequency at which the output power will equal 125 mW
the 3 dB bandwidth of the amplifier
the minimum dc power supply voltage required.
Describe how you would convert the amplifier into a frequency multiplier with
a multiplication factor of 2. What conditions have to be met in order to obtain a
larger multiplication factor?
The electromagnetic disturbance created by the transmitter is propagated by the
transmitter antenna and travels at the speed of light as described in Chapter 2. It is
evident that, if the electromagnetic wave encounters a conductor, a current will be
induced in the conductor. How much current is induced will depend on the strength
of the electromagnetic field, the size and shape of the conductor and its orientation to
the direction of propagation of the wave. The conductor will then capture some of
the power present in the wave and hence it will be acting as a receiver antenna.
However, other electromagnetic waves emanating from all other radio transmitters
will also induce some current in the antenna. The two basic functions of the radio
receiver are:
(1) to separate the signal induced in the antenna by the transmission which we
wish to receive from all the other signals present,
(2) to recover the ‘‘message’’ signal which was used to modulate the transmitter
In order to separate the required signal from all the other signals captured by the
antenna, we use a bandpass filter centered on the carrier frequency with sufficient
bandwidth to accommodate the upper and lower sidebands but with a sufficiently
high Q factor so that all other carriers and their sidebands are attenuated to a level
where they will not cause interference. This is most easily achieved by using an LC
tuned circuit whose resonant frequency is that of the carrier.
Figure 3.1. (a) The envelope detector circuit. The diode ‘‘half-wave’’ rectifies the AM wave and
the RC time-constant ‘‘follows’’ the envelope with a slight ripple. (b) The input signal to the
envelope detector. (c) The output signal of the envelope detector. Note that (1) when the voltage
is rising the ripple is larger than when the voltage is falling. A longer time constant will help
reduce the ripple; however, it will also increase the likelihood that the output voltage will not
follow the envelope when the voltage is falling causing ‘diagonal clipping’. (2) In practice, the
carrier frequency is much higher than the modulating frequency, hence the ripple is much smaller
than shown.
Figure 3.1. (continued )
To recover the ‘‘message’’ we require a circuit which will follow the envelope of
the amplitude of the carrier. Such a circuit is called an envelope detector and it
consists of a diode and a parallel RC circuit as shown in Figure 3.1(a).
The input signal to the circuit is most appropriately represented by an ideal
current source connected to the primary of the transformer. This ideal current source
represents all the currents induced in the antenna by all the radio stations broadcasting signals in free space. The signal is coupled to the parallel-tuned LC circuit
which selectively enhances the amplitude of the signal whose carrier frequency is the
same as the resonant frequency of the LC circuit. In Figure 3.1(b), only the enhanced
modulated signal is shown at the input of the envelope detector. Because the diode
conducts only when the anode has a positive potential compared to the cathode, only
the positive half of the signal appears across the output resistor. Because the
capacitor is connected in parallel with the resistor, when the diode conducts the
capacitor must charge up to the peak value of the voltage. When the input voltage is
less than the voltage across the capacitor, the conduction is cut off and the capacitor
starts to discharge through the resistor with the voltage falling off exponentially.
With the proper choice of time-constant RC, the output voltage waveform will have
the form shown in Figure 3.1(c). This waveform is essentially the envelope of the
carrier signal with a ripple at a frequency equal to the carrier frequency. A low-pass
filter can be used to remove the ripple.
The circuit shown in Figure 3.1(a) has been used with success as a practical
receiver with the resistor R replaced by a high impedance headphone. Needless to
say, such a simple circuit has its limitations. The power in the circuit is supplied
entirely by the transmitter and naturally it is at a very low level, especially as the
distance between the transmitter and the receiver increases. Secondly, the ability of
the LC tuned circuit to suppress the signals propagated by all the other transmitters is
limited and therefore such a receiver will be subject to interference from other
stations. These limitations can be overcome by using the superheterodyne configuration described below.
The superheterodyne receiver takes the incoming radio-frequency signal whose
frequency varies from station to station and transforms it to a fixed frequency called
the intermediate frequency (IF). It is then easier to do the necessary filtering to
eliminate interference and, at the same time, to provide some power gain or
amplification to the desired signal.
A normal AM superheterodyne receiver block diagram is shown in Figure 3.2.
The antenna has induced in it currents from all the transmitters whose electromagnetic propagation reach it. The first step is to use an LC tuned radio-frequency
amplifier to enhance the desired carrier and its sidebands. The radio-frequency
amplifier is tuneable over the frequency for which the receiver is designed by
varying the capacitor in the tuned circuit. This capacitor is mechanically coupled or
‘‘ganged’’ to another capacitor which forms part of the local oscillator circuit. The
local oscillator frequency and the frequency to which the radio-frequency amplifier
is tuned are chosen in such a way that, as the value of the ganged capacitors change,
they maintain a fixed frequency difference between them. The outputs from the local
oscillator and the radio-frequency amplifier are used to drive the frequency changer
or mixer. The frequency changer essentially multiplies the two inputs and produces a
signal that contains the sum and difference of the input frequencies. Because of the
fixed difference between the incoming radio-frequency and the local oscillator
frequency, the difference frequency remains constant as the value of the ganged
capacitor is changed. The output of the frequency changer is then fed into the
intermediate-frequency amplifier. The intermediate-frequency amplifier is designed
to select the difference frequency plus its sidebands and to attenuate all other
frequencies present. Since the difference frequency is fixed (for domestic AM radios
the intermediate frequency is 445 kHz) the filters required are relatively easy to
Figure 3.2. The block diagram of the superheterodyne receiver. The capacitor which tunes the
radio-frequency amplifier is mechanically ganged to the capacitor which determines the
frequency of the local oscillator. In the normal AM receiver, the oscillator frequency is always
455 kHz above the resonant frequency of the radio-frequency amplifier throughout the range of
design with sharp cut-off characteristics. The output of the intermediate-frequency
amplifier which then goes to the envelope detector consists of the intermediate
frequency and its two sidebands. The envelope detector removes the intermediate
frequency, leaving the audio-frequency signal which is then amplified by the audiofrequency amplifier to a level capable of driving the loudspeaker. It is clear that there
will be a very large difference between the signal from a powerful local radio station
and a weak distant station. To help reduce the difference an automatic gain control
(AGC) is used to adjust the signal reaching the envelope detector to stay within
predetermined values.
The most interesting signal processing step in the system takes place in the
frequency changer or frequency mixer or simply the mixer [1]. There are two basic
types of mixers: the analog multiplier and the switching types. The analog multiplier
frequency changer simply multiplies the radio-frequency signal and the local
oscillator so that when the modulated carrier current is
im ðtÞ ¼ Að1 þ k sin oS tÞ sin oC t
and the local oscillator signal is
io ðtÞ ¼ B sin oL t
the output of the mixer is
iðtÞ ¼ Að1 þ k sin oS tÞ sin oC t B sin oL t
iðtÞ ¼
2 ABð1
þ k sin oS tÞ½cosðoL oC Þt cosðoL þ oC Þt
iðtÞ ¼ 12 AB½cosðoL oC Þt cosðoL þ oC Þt þ k sin oS t cosðoL oC Þt
k sin oS t cosðoL þ oC Þt
iðtÞ ¼ 12 ABfcosðoL oC Þt cosðoL þ oC Þt
þ 12 k½sinðoL oC oS Þt þ sinðoL oC þ oS Þt
12 k½sinðoL þ oC oS Þt þ sinðoL þ oC þ oS Þtg:
The spectrum of Equation (3.3.6) is shown in Figure 3.3. It should be noted that this
has been simplified for clarity. The product formation in Equation (3.3.3) is not a
precise process and tends to create a large number of frequencies due to sub- and
higher harmonics present in both the radio-frequency and local oscillator signals.
The radio-frequency and local oscillator signals are usually present in the output as
well. It is important to keep all the unwanted signals outside the frequency band of
the intermediate frequency and, failing that, to reduce their amplitude to a very low
It can be seen that the mixing operation gives two additional carriers and their
sidebands at frequencies corresponding to the sum (oL þ oC ) and difference
(oL oC ) of the local oscillator and carrier frequencies. The required signal at
Figure 3.3. A simplified spectrum of the output from a frequency changer which uses a
nonlinear device.
the difference frequency (intermediate frequency) can now be filtered out by the
intermediate-frequency stage of the receiver. It should be noted that the mixing
operation does not affect the sidebands. To clarify the changes in frequency that take
place as the signal proceeds through the system, the AM broadcast band (600 kHz–
1600 kHz) is used as an example in Table 3.1.
The frequency changer or mixer presents two immediate problems: the choice of
the local oscillator frequency and the design strategy of the mixer itself.
(1) It can be seen from Table 3.1 that the local oscillator frequency has been
chosen to be higher than the incoming radio-frequency signal. There is a very
good reason for this. The ratio of the maximum to the minimum capacitance
Radio frequency (kHz)
Incoming signal, fc fs
Local oscillator, fL
Intermediate frequency, fk
Image frequencya, fim
Output, intermediate-frequency amplifier, fk fs
Envelope detector, fs
Low-frequency end
High-frequency end
600 5
600 þ 455 ¼ 1055
1055 þ 455 ¼ 1510
455 5
1600 5
1600 þ 455 ¼ 2055
2055 þ 455 ¼ 2510
455 5
The image frequency is the frequency of the unwanted signal which, when combined with the local oscillator
frequency, will give the intermediate frequency. Normally the radio-frequency amplifier should suppress the
image frequency but this may be difficult if the signal from the desired station is very weak and the image
signal is very strong.
required to tune the local oscillator across the broadcast band is 3.79
when a higher local oscillator frequency is chosen. If the lower local
oscillator frequency had been chosen, the ratio would have been 62.4.
Such a variable capacitor would be difficult to manufacture with reasonable
(2) The mixing operation was treated earlier as an analog multiplication.
However, the realization of a precise analog multiplier is a non-trivial
problem. A crude analog multiplication can be achieved by using a device
whose voltage–current characteristics are non-linear. An ordinary p–n junction diode can be used to perform the task. The derivation of the output signal
is similar to that given in Section 2.6.1 and will therefore not be repeated
The switching type of mixer uses a device such as a diode or transistor carrying a
current proportional to the radio-frequency signal and switches it from one state to
another at the local oscillator frequency.
Receiver Antenna
The AM receiver antenna can take many different forms such as the ferrite bar found
in most portable receivers, the whip antenna found on automobiles, and the outdoor
wire type consisting of several metres of wire strung between two towers. In general,
the longer and higher off the ground the antenna is, the more likely it is that it will
have a strong signal induced in it by the electromagnetic signals propagated by the
transmitters. The level of signal induced in the antenna may vary from a few
microvolts to a few volts, depending on the proximity of the transmitter, its radiated
power output, the size of the receiver antenna, and its orientation to the transmitter.
Because of the tremendous variation in the input signal a fixed gain amplifier will
very often either not provide enough signal to the frequency changer or will overload
it and consequently generate a large number of undesirable frequencies. To ensure a
reasonable reception of the largest number of broadcasting stations, the gain of the
amplifier is controlled automatically by the incoming signal – the weaker the signal,
the higher the gain of the radio-frequency amplifier.
The antenna signal is coupled by a radio-frequency transformer to the input of the
radio-frequency amplifier. The transformer is made up of two coils, each containing
several turns of wire wound on a coil former which may or may not have a ferrite
core. The major consideration in the design of the transformer is that the primary
inductance be sufficiently high to ensure that signals at the lowest frequency of
interest are not unduly attenuated. Since the signal frequency can vary from 600 to
1600 kHz, the transformer is not tuned.
Low-Power Radio-Frequency Amplifier
Since the input voltage of the amplifier is of the order of microvolts and the signal to
be delivered to the demodulator is usually in volts, the amplifier must have a high
gain. A multi-stage amplifier has to be used to realize the necessary gain. Some of
the stages of gain can be placed before the frequency changer, in which case they are
referred to as the radio-frequency amplifier stage, or after the frequency changer, in
which case they are called the intermediate-frequency amplifier stage. It is usual to
design the radio-frequency amplifier stage for a modest gain and the intermediatefrequency stage for high gain. Both the radio-frequency and intermediate-frequency
amplifiers are narrow-band amplifiers. This is evident from calculating the Q factor
for the two types of amplifiers. Considering that the normal bandwidth of the AM
radio is 0–5 kHz, both radio-frequency and intermediate-frequency amplifiers have
to have a bandwidth of at least 10 kHz.
The Q factor of the radio-frequency amplifier at the low end of the broadcast band
(600 kHz) is 60 and at the high end (1600 kHz) is 160. The Q factor of the
intermediate-frequency amplifier (centre frequency 455 kHz) is 45. However, operation of the radio-frequency amplifier with such a high Q factor will cause serious
tracking problems with the local oscillator and will also lead to excessive attenuation
at the edges of the sidebands. For practical purposes, a Q factor of about 10 is used
in the radio-frequency amplifier, leaving the major part of the filtering problem to the
intermediate-frequency stage. The design of the intermediate-frequency filter about a
fixed center frequency is a much easier process and can be achieved with greater
precision than in the radio-frequency stage, where the center frequency of the
bandpass filter changes when the tuning capacitor is changed. In spite of the
difference in Q factor of the radio-frequency and intermediate-frequency amplifiers,
they have enough similarities for the same general principles to be used for their
The wide variation of the radio-frequency input signal level and the need for
automatic gain control in the radio-frequency amplifier was discussed earlier. It is
usual to amplify the incoming radio-frequency signal by a fixed amount in order to
derive a control signal for the gain of a subsequent variable gain amplifier. A typical
fixed gain radio-frequency amplifier is shown in Figure 3.4. Although a bipolar
transistor is shown, a field-effect transistor can be used. The collector load is an LC
tank circuit in which the capacitance is variable. The variable capacitance is
mechanically ganged to the capacitance which controls the frequency of the local
oscillator so that, as the capacitance is changed, the resonant frequency of the LC
tank circuit tracks the local oscillator frequency with a constant difference equal to
the intermediate frequency (455 kHz).
It can be seen that the above circuit bears a striking resemblance to the frequency
multiplier circuit given in Figure 2.21. The difference is that the frequency multiplier
operates in class-C while the radio-frequency amplifier operates in class-A.
The load driven by the amplifier may be coupled to the collector circuit by a
transformer, in which case the inductor in the collector circuit becomes the
transformer primary. The load may also be coupled by a capacitor. In both cases,
Figure 3.4. A typical radio-frequency amplifier. The load RL represents the input resistance
(impedance) of the circuits which are driven by the amplifier.
the load can be represented by a resistance RL in parallel with the tuned circuit. To
simplify the analysis of the circuit, the winding resistance r in series with the
inductance is transformed into an equivalent shunt resistance (refer to Figure 2.50)
Rp where
Rp ¼
o2 L2p
Lp ¼ L
when Q 1.
The amplifier load RL combined in parallel with Rp is now the resistive part of the
collector load. The new equivalent circuit is shown in Figure 3.5, where
Req ¼ Rp kn2 RL :
It can be seen from Figure 3.5 that:
(1) The emitter resistor Re has not been bypassed to ground with a capacitor.
(2) At the frequency of resonance, the parallel LC circuit in the collector circuit
will behave like an open circuit. The equivalent collector load is Req .
(3) Because the inductor is connected directly between þVcc and the collector,
the dc voltage on the collector is þVcc .
Figure 3.5. The amplifier shown in Figure 3.4 with the transformer load transferred to the
primary and combined with the winding resistance r .
The major advantage of not bypassing Re is that the gain of the amplifier is
determined by the ratio of the collector-to-emitter load impedance, which, in this
case, is Req =Re and it is essentially independent of the transistor parameters such as
current gain and transconductance. The design steps are illustrated in the following
Example 3.4.1 Low-Power Radio-Frequency Amplifier. The antenna of an AM
radio receiver (600 to 1600 kHz) supplies 100 mV peak to the input of the radiofrequency amplifier when the modulation is a sinusoid, the modulation index is
unity, and the radio-frequency is 600 kHz. The dc supply voltage is þ6 V and the
required gain is 20. The amplifier load represented by the input impedance of the
automatic gain control circuit is 10 kO resistive and it is capacitively coupled. The
variable capacitor used in the tuned circuit (and mechanically coupled to the
capacitor used in the local oscillator) has a maximum value of 250 pF and a
minimum value of 25 pF. The Q factor of the coil is expected to be about 50 and
the current gain of the transistor is 100. Design a suitable amplifier.
Solution. The inductance of the tuning coil is given by
L ¼ 1=ðo2 CÞ:
When o ¼ 2p 600 103 and C ¼ 250 pF
L ¼ 281 mH:
When o ¼ 2p 1600 103 the capacitance required to tune the amplifier
C ¼ 35:2 pF:
The combination of L and C can be used to tune the amplifier to any frequency in the
AM broadcast band.
The winding resistance of the coil
r ¼ oL=Q
¼ 21:2 O:
The equivalent parallel resistance
Rp ¼ o2 L2p =r
¼ 52:9 kO:
Combining Rp with the load resistance of 10 kO gives
Req ¼ 8:41 kO:
The loaded Q of the collector circuit
QL ¼ Req =ðoLÞ
¼ 7:94:
The relatively low Q should ensure that the sideband ‘‘edges’’ are not subject to
severe attenuation.
At the resonant frequency, the parallel LC circuit in the collector behaves like an
open circuit. The equivalent collector load is therefore Req .
The emitter resistance
Re ¼ Req =gain
¼ 420 O:
The output voltage ¼ 100 mV 20 ¼ 2:0 V (peak) and the current drawn by the
10 kO load is 0.2 mA (peak).
To ensure that the amplifier is capable of supplying 0.2 mA ac current to the load,
the dc current in the collector may be set at ten times the load current, that is,
Ic ¼ 2 mA. The dc voltages at the emitter and base are then Ve ¼ 0:84 and
Vb ¼ 1:54 V, respectively. The dc voltage on the collector is still 6 V. It is clear
that the amplifier will go into saturation when the collector voltage drops to a
minimum value of 1.34 V (Ve þ 0:5) and to cut off when the collector voltage is
12 V. Since the collector signal is only 2 V, about a quiescent value of 6 V, there is
no danger of clipping at the collector.
The dc base current
Ib ¼ Ic =b
¼ 20 mA:
The values of R1 and R2 are chosen so that a dc current of 10Ib will flow in the chain
but with a voltage of 1.54 V at the base of the transistor. This gives
R1 ¼ 22:3 kO
R2 ¼ 7:7 kO:
The coupling capacitor is chosen so that, at the lowest frequency of interest, its
reactance is negligibly small compared to the load.
Frequency Changer or Mixer
Two distinct approaches can be used in the design of a mixer. The first is based on an
analog multiplication of the radio-frequency and the local oscillator signals. The
second uses the local oscillator signal to switch segments of the radio-frequency
signal positive and negative. In this case the local oscillator must produce a square
wave. The Analog Mixer. As discussed earlier, a crude analog multiplication
can be achieved by using a non-linear device such as a p–n junction diode which can
be approximated by the equation
i ¼ a1 v þ a2 v 2 þ :
A mixer using diodes produces an output signal with considerable loss. Various
schemes exist, some employing several diodes with a single-ended or differential
If an ‘‘active’’ mixer is used, considerable gain can be obtained in the mixing
process [5]. The preferred analog active mixer uses a dual-gate metal-oxide
semiconductor field-effect transistor (MOSFET). The advantages of this design
includes a lower power requirement from the local oscillator and improved isolation
between the local oscillator and the receiver antenna. The isolation between the local
oscillator and the antenna will ensure minimum radiation of the local oscillator
signal and hence minimize interference with other electronic equipment.
To understand the design process, it is necessary to begin with the drain current
(iD ) – drain-to-source voltage (vDS ) characteristics of the MOSFET. A typical nchannel depletion mode MOSFET is shown in Figure 3.6.
It can be seen that the characteristics are similar to those of a BJT except that the
drain current is controlled by the gate-to-source voltage. An elementary commonsource amplifier is shown in Figure 3.7.
Figure 3.6. Typical characteristics of an n-channel, enhancement-mode metal-oxide semiconductor field effect transistor (MOSFET).
Applying KVL to the drain current path,
VDD ¼ iD RD þ vDS
iD ¼
VDD v :
When Equation (3.4.6) is plotted on the FET characteristics shown in Figure 3.6, it
gives a straight line with a slope of (1=RD ), an intercept on the x axis given by VDS
equal to VDD and on the y axis ib ¼ VDD =RD . This is the load line which describes
the behavior of the amplifier.
Figure 3.7. Typical biassing arrangement for the common-source MOSFET amplifier.
When designing an amplifier, it is necessary to select a bias point VGS along the
load line and the input signal vgs so that the device will remain in the ‘‘active’’
region. This is achieved by ensuring that the device is biased above its threshold Vth .
vGS ¼ VGS þ vgs
and the relationship between iD and the vGS can be approximated by
iD ¼ IDSS 1 Vth
where the Vth and IDSS are defined in Figure 3.8.
Substituting Equation (3.4.7) into (3.4.8) gives
1 GS
2 1 GS
2 #
The first term, ð1 VGS =Vth Þ2 , represents the dc component of the drain current. The
second term, 2ð1 VGS =Vth Þðvgs =Vth Þ, is an ac current proportional to the input
voltage and represents the normally desired output. The third term, ðvgs =Vth Þ2 ,
represents a non-linearity which is normally undesirable. In terms of designing a
mixer, however, this is the desired output. The relative value of this term can be
increased by making the input signal vgs large. However, since Equation (3.4.8) is an
approximation, making vgs too large can produce spurious signals which may
interfere with the required signal.
Figure 3.8. A typical iD vGS characteristic of an n-channel depletion-type MOSFET showing
the threshold voltage, Vth , and the saturated drain-to-source current, IDSS .
A more practical version of the MOSFET amplifier is shown in Figure 3.9.
The MOSFET used in the circuit is an n-channel enhancement device and the
resistive chain R1 and R2 is chosen to hold the gate at a specific potential above that
of the source. Rs is used partly to stabilize the dc bias point and partly to reduce the
dependence of the gain on the parameters of the device. In general, semiconductor
device parameters vary widely from one device to another and it is necessary to build
some controls into the design. When semiconductor devices are used in the design of
circuits whose specifications must be held to very tight tolerances, it is a good idea to
use design strategies which rely on ratios of passive components, such as resistances,
rather than on the values of the device parameters. In this case it can be shown that
the gain of the amplifier is equal to the ratio of the drain impedance ZD to the source
resistance Rs .
Since the mixer has two input signals of different frequencies several problems,
such as frequency ‘‘pulling’’ and local oscillator feed-through to the antenna, can be
avoided by ensuring that the two sources are well isolated from each other. It is
possible to achieve a high level of isolation by using a dual-gate MOSFET. The
design process is best illustrated by an example.
Example 3.4.2 The Mixer. Design a mixer for an AM radio using the dual-gate
n-channel depletion MOSFET whose characteristics are given in Figure 3.10. The
following are specified:
Supply voltage, VDD ¼ 12 V.
Drain bias current, ID ¼ 5 mA.
Primary inductance of the drain transformer, Lp ¼ 250 mH.
Centre frequency of the output (intermediate frequency) ¼ 455 kHz.
Figure 3.9. A more practical version of the MOSFET amplifier shown in Figure 3.7.
Figure 3.10. The drain characteristics of the dual-gate n-channel MOSFET used in Example
(5) 3 dB bandwidth ¼ 20 kHz.
(6) Transformer turns ratio is 10 : 1.
What is the value of the resistive load that the mixer must ‘‘see’’, assuming that both
the primary and secondary winding resistances are negligibly small?
Solution. A suitable circuit for the mixer is as shown in Figure 3.11.
The capacitance required to tune the drain to 455 kHz (intermediate frequency) is
given by
o2 ¼
Lp Cp
Cp ¼
¼ 489 pF:
o2 Lp ð2p 455 103 Þ2 250 106
The bandwidth Df is related to the center frequency f0 by
Q0 ¼
455 103
¼ 22:75:
20 103
Figure 3.11. A typical MOSFET mixer using the dual-gate n-channel device.
The resistive load transferred to the primary n2 RL will be in parallel with Lp . Q0 for a
parallel LR circuit is given by
Q0 ¼
n2 RL
¼ 22:75
n2 RL ¼ 2p 455 103 250 106 22:75 ¼ 16:26 103 O:
RL ¼ 162 O:
From the device characteristics given in Figure 3.10, locate VDS ¼ VDD ¼ 12 V, and
using a straight edge pivoted at this point, determine a point along the line given by
ID ¼ 5 mA which will give a wide dynamic range for the drain current. A good point
is given by the intersection of VDS ¼ 6 V and ID ¼ 5 mA. The load line can now be
drawn in as shown. From the slope the required load is 1.2 kO. In a common-source
amplifier, this load would normally be connected in series with the drain. However,
from the dc point of view, the drain is connected to VDD by a short-circuit (through
the inductor Lp ). The device can be correctly biased by connecting the 1.2 kO
resistor in series with the source, that is, choose
Rs ¼ 1:2 kO:
With a drain current of 5 mA, the voltage of the source will be
VS ¼ 1:2 103 5 103 ¼ 6:0 V:
From Figure 3.10 it can be seen that the required gate-source voltage on gate 1 is 0 V.
Gate 1 must therefore be biased at 6.0 V by the resistive chain so that
R1G1 ¼ R2G1
Since no current flows into the gate, the value of the two resistances is arbitrary and
can be made as large as practicable. Let
R1G1 ¼ R2G1 ¼ 100 kO:
From Figure 3.10 it can be seen that the required gate-source voltage on gate 2 is
4.0 V. Because the source is biased at 6.0 V, gate 2 must be biased at 10 V. Again no
current flows into gate 2 and therefore the resistance can be made as large as
practicable. However
R1G2 : R2G2 ¼ 2 : 10:
Choosing R2G2 ¼ 100 kO makes R1G2 ¼ 20 kO. The coupling capacitors C1 and C2
are chosen so that they present negligible impedance to the radio-frequency and local
oscillator, respectively.
The drain tank circuit is tuned to resonate at the intermediate frequency; therefore
the drain load is
n2 RL ¼ 16:26 103 O:
The gain of the stage is approximately equal to
n2 RL
¼ 13:6 ¼ 22:6 dB:
This is not the same as the mixer gain which is defined as
10 log10
ðintermediate frequency powerÞ
ðradio-frequency powerÞ
In a practical circuit, the relative signal levels of both the radio frequency and the
local oscillator will be adjusted to optimize the intermediate-frequency signal. Switching-Type Mixer. The circuit diagram of one of the simplest
switching-type mixers in shown in Figure 3.12 [2]. The radio-frequency signal Vs is
Figure 3.12. The circuit of a switching-type mixer.
a sinusoid. The local oscillator output VL is a square wave at a frequency which is
higher than the radio frequency. The square wave is defined as
gðtÞ ¼ 1
gðtÞ ¼ 1 for
> t > T:
Assuming ideal diodes, and that VL is larger than Vin , then when VL > 0, D1
conducts and D2 is off:
Vo ¼ VL þ Vin
and when VL < 0, D2 conducts and D1 is off:
Vo ¼ ðVL þ Vin Þ:
Vo ¼ VL þ Vin gðtÞ:
The output voltage is then
This is evident from an examination of Figure 3.13.
The square wave gðtÞ can be expressed in terms of its Fourier components as
gðtÞ ¼
1 sinð2n þ 1Þo t
p 0
ð2n þ 1Þ
Figure 3.13. Waveforms of the inputs Vin and VL and the output Vo .
Vin ¼ A sin oC t:
Vin gðtÞ ¼
1 cos½ð2n þ 1Þo o t cos½ð2n þ 1Þo þ o t
2A P
p n¼0
ð2n þ 1Þ
The output of the mixer consists of the local oscillator frequency and an infinite
number of sums and differences of the local oscillator harmonics and the radio
frequency. The desired frequency components can be filtered in the intermediatefrequency stage that follows the mixer.
If the AM carrier equation is used in place of Equation (3.4.25), it can be
demonstrated that the mixing operation maintains the relationship between the
desired intermediate frequency and its sidebands.
The major disadvantages of this type of mixer are as follows:
(1) The large signal required from the local oscillator to switch the diodes calls
for considerable power output from the oscillator and this makes the design
of the local oscillator difficult.
(2) The large local oscillator signal is present in the output of the mixer and it can
interfere with the filtering process, especially when the local oscillator
frequency is much higher than the radio frequency. Hence the desired sum
and different frequencies are close to each other and to the local oscillator.
This part of the output can be removed by changing the basic circuit from a
single ended to a differential output.
Intermediate-Frequency Stage
The output of the mixer contains a multitude of frequencies made up of the sums and
differences of the local oscillator frequency and the radio-frequency signal and their
various harmonics. The task at hand then is to select the frequency ( flo frf ),
together with its sidebands, and to amplify it if necessary before it is demodulated. A
filter is required to achieve this. An ideal filter for this purpose would be one with
rectangular characteristics – a flat response across the frequency band and infinitely
steep ‘‘skirts’’ with infinite attenuation beyond. A practical filter will, of course, be
much less exotic than that.
The only type of frequency selective circuit discussed so far has been the LC
tuned circuit. A single tuned LC circuit can have steep skirts and high attenuation for
out-of-band signals when the Q factor is high but a high Q factor also means that the
in-band frequency is very narrow. The design of a filter which can select the
intermediate frequency and its sidebands and suppress to an acceptable level all the
other frequency components present in the output of the mixer is beyond the scope
of this book. The interested reader will find a short list of sources for more
information in the bibliography.
In general, a filter is placed between a resistive source and a resistive load which
may or may not have the same value. For frequencies in the passband, the filter
‘‘matches’’ the source to the load so that the reflection of the signal from the load is
minimal, that is, maximum power is transferred to the load. In the stopband, the filter
input presents such a severe mismatch to the source that most of the signal power is
reflected with very little reaching the load.
Filters can be classified as follows:
1. Passive LC Filters. These filters are made up of only inductors and capacitors. They are considered to be lossless. In general, the closer the filter
characteristics are to the ideal, the more Ls and Cs required. Discrete LC filters
are used at frequencies from as low as 20 Hz to as high as 500 MHz. The lowfrequency limit is set by the low Q factors of the inductors and at the highfrequency end by the circuit strays of both L and C. Passive LC filters can be
used at frequencies as high as 40 GHz but the components have to be
considered to have distributed parameters. Inductors and capacitors are then
‘‘replaced’’ by open or short-circuited transmission lines and tuned LC circuits
by resonators. The physical appearance of the components of these very high
frequency filters bear no resemblance to ordinary Ls and Cs.
Crystal Filters. The very high Q factors which can be obtained from crystals
offer the filter designer the possibility of high selectivity and steep skirt
gradients. However, crystal parameters are, in general, not under the direct
control of the filter design and consequently have not been very popular.
Crystal filters have an upper frequency limit of a few megahertz.
Active LC Filters. These filters are made up LC sections separated by
amplifiers. The advantages to be gained from this is that the filter can then
have an overall gain instead of a loss and the buffering action of the amplifiers
between various segments of Ls and Cs can reduce the amount of interaction
between them, making the tuning of the filter considerably easier. The
bandwidth of the amplifier used can limit the range of application of these
filters to an upper range of about 300 MHz.
Active RC Filters. These filters are made up of resistors, capacitors and
operational amplifiers. The introduction of integrated circuits prompted the
development of these filters. Unfortunately, they can be used only at low
frequencies with an upper limit of a few megahertz.
Digital Filters. These filters have become practical since the development of
cheap and fast microcomputers with large memories. The signal is sampled by
an analog-to-digital converter. The samples are converted into a digital code
and multiplied by a function of the desired output characteristics. The resulting
signal is fed into a digital-to-analog converter to give an analog output. Even
with the fastest microcomputers digital filters are limited to an upper
frequency of about 50 kHz.
Mechanical Filters. In these filters, the electrical signal is converted into a
mechanical vibration by a magnetostrictive transducer. A number of mechanical resonators in the form of discs, plates, and rods determine which
frequencies will be propagated through the device to reach the output and
which will not. At the output another transducer converts the vibration back
into an electrical signal. Mechanical filters normally operate in the frequency
range 50–600 kHz. These filters have the attraction of being quite accurate and
yet inexpensive.
Surface-Acoustic-Wave Filters. These filters belong to a more general class
called transversal filters. In surface-acoustic-wave (SAW) filters, two transducers are fabricated at opposite ends of a highly polished piezzoelectric material
such as quartz or lithium niobate. When an electrical signal is applied to the
input transducer, the material changes its physical shape and, at the appropriate
frequency, will cause a travelling wave to be propagated on the surface of the
material. At the output, the travelling wave is converted back into an electrical
signal. The mechanical shape, size, and placing of the metallization on the
piezoelectric material determine which frequencies are propagated and which
are attenuated. SAW filters normally operate in the range 20–500 MHz.
8. Switched Capacitor Filters. These filters have been made possible by the
ease with which MOS (metal-oxide semiconductor) technology can be used to
realize capacitors, switches, and operational amplifiers on the same silicon
chip. Two switches and a capacitor can be used to simulate a resistor. It is then
possible to construct equivalent circuits for most of the filter structures used in
active RC filters. Switched capacitor filters are generally used where accurate
filtering is not required. However, the characteristics of switched capacitors
continue to improve as they continue to be the subject of intense research
interest. The need to switch the capacitors at a much higher frequency (clock
frequencies between 30 and 40 MHz) than the highest frequency present in the
signal limits these filters to operation below 3 MHz.
Automatic Gain Control
The function of the automatic gain control is to ensure that the signal reaching the
demodulator is sufficiently high and within the limits for efficient demodulation. It
does this by sensing the level of the signal at the input to the modulator and adjusting
the gain of a variable gain amplifier to keep the level constant. In practice it is not
possible to boost all signals to a constant level and, in any case, it is undesirable to
amplify noise to the same level when that is all that is available. The desirable
characteristics of an automatic gain control (AGC) circuit is as shown in Figure
3.14(a). Below the ‘‘knee’’, the signal is amplified by a constant factor. Above the
knee the output is kept constant. A practical characteristic is shown in Figure 3.14(b)
where the knee is rounded and the output continues to rise but at a limited rate.
The AGC subsystem consists of a variable gain amplifier whose gain is controlled
by a voltage or current derived from the output signal. A block diagram is given in
Figure 3.15.
Figure 3.14. (a) The ideal input–output characteristics of an automatic gain control (AGC)
circuit. (b) The input–output characteristics of a practical AGC circuit.
Figure 3.15. A block diagram of a feedback-type AGC circuit.
The rectifier produces a dc signal which is proportional to the signal which
appears at the output of the variable gain amplifier. The dc signal is fed back to the
variable gain amplifier and attempts to keep the output signal constant above a
predetermined level. It is usual to place the variable gain amplifier before the mixer
so that the radio-frequency signal reaching the mixer does not vary too widely.
A suitable amplifier for the AGC circuit is the four-quadrant analog multiplier
discussed in Section 2.6.3. In the scheme shown in Figure 3.15, the four-quandrant
amplifier will be part of the radio-frequency stage with the radio-frequency signal
applied to one of its two inputs. The second input is a dc signal derived from the
rectified and smoothed intermediate-frequency signal so that, when the intermediatefrequency signal is high, the constant of multiplication is low, and when the signal is
low the constant is high.
Because the four-quadrant analog multiplier was discussed at length earlier, only
the rectifier and its associated time-constant will be discussed.
A common characteristic of all AGC circuits is that they have a fast ‘‘attack’’ time
and a somewhat slower ‘‘release’’ time. This means that the system can capture
Figure 3.16. A circuit which provides a ‘‘fast attack’’ and ‘‘slow release’’ for the AGC control
sudden increases in signal level but take a longer time to adjust the gain upwards for
low signals. A suitable rectifier and time-constant are shown in Figure 3.16.
The first positive-going signal that appears at the base of the BJT causes it to
switch on and very rapidly charge the capacitor C with its emitter current which is
much larger than the base current of the BJT. The dc signal required to control a
sudden increase in signal is therefore produced very quickly. When the positivegoing voltage starts to drop, the charged capacitor holds up the voltage of the emitter
and the falling base voltage ensures that the BJT is cut off. The discharge of the
capacitor is then controlled by the resistor in parallel with it. The attack time can be
slowed down by connecting a small resistor in series with the capacitor. The correct
attack and release times have to be determined by experiment.
The input to the demodulator is a carrier of frequency 455 kHz (intermediate
frequency) with an amplitude envelope determined by the audio signal. The circuit
used for demodulation is therefore the envelope detector which is shown in Figure
3.1. The operation of the circuit was discussed in Section 3.2. An example of an
envelope detector follows.
Example 3.4.3 Envelope Detector. The amplitude of the signal applied to the
input of an envelope detector is 4 V peak when the modulation index is zero and the
intermediate frequency is 455 kHz. Choose a suitable time-constant for the detector
to avoid diagonal clipping when the modulation index is 0.8 and the highest audio
frequency component in the input signal is 10 kHz. Assume that the diode is ideal.
Solution. The output waveform with the capacitor removed will be as shown in
Figure 3.17.
Figure 3.17. The output voltage waveform of the envelope detector used in Example 3.4.3.
Because B=A ¼ 0:8, B ¼ 3:2 V peak. The equation for the envelope is
vS ðtÞ ¼ B cos oS t
where os ¼ 2p 10 103 . The slope of the envelope is
¼ Bos sin os t:
From the diagram, it is clear that the maximum slope occurs at t ¼ T =4, when
sin os t ¼ 1. Hence
¼ Bos :
dt max
When the capacitor is connected to the circuit and the diode is not conducting, the
output voltage decays according to the equation
vc ðtÞ ¼ V et=t
The slope of the decaying capacitor voltage is
V et=t :
From the diagram it is clear that, when t ¼ T =4 or when os t ¼ p=2, V ¼ A and
Aet=t :
The maximum value occurs at t ¼ 0:
¼ :
To avoid diagonal clipping, the capacitor voltage must decay faster than the envelope
voltage, that is,
j j j Bos j
¼ 19:89 106 s:
Bos 3:2 2p 10 103
Let R ¼ 10 kO, then C ¼ 1989 pF.
Audio-Frequency Amplifier
The audio-frequency amplifier was discussed in Section 2.7. A few minor changes
may be necessary in the design of the final stage of the amplifier so that it can drive a
loudspeaker efficiently.
So far, the audio-frequency signal is in an electrical form. To convert it to an acoustic
signal, an electrical-to-acoustic transducer is required. This can take two basic forms:
the loudspeaker and the headphone. The more common of the two devices is the
loudspeaker [3, 4]. There are two basic types of loudspeakers: the direct radiation
and the horn loudspeaker. The diaphragm of the direct radiation loudspeaker is
coupled directly to the air whereas that of the horn loudspeaker (as the name
suggests) is through a horn. The horn loudspeaker is generally more efficient due to
the better matching between its diaphragm and the air provided by the horn.
However, it is usually large, heavy, and expensive. A direct radiation loudspeaker,
by contrast, has a simple construction, occupies a small space, and has a relatively
uniform response over a moderate frequency band. The design of loudspeakers is a
highly specialized field, and therefore outside the scope of this book. However, a
qualitative description of the behavior of loudspeakers is a subject that should be of
interest to radio engineers.
A cross-section of the direct radiation loudspeaker is given in Figure 3.18.
It consists of a paper cone, the apex of which is mechanically connected to a
cylinder on which the voice coil is wound. The voice coil is located in a magnetic
field provided by a permanent magnet. The outer rim of the cone is corrugated to
permit the cone to move along its axis with the minimum of mechanical resistance.
The whole cone may be viewed as the diaphragm of the loudspeaker. The
Figure 3.18. A cross-sectional view of a typical loudspeaker.
corrugation is referred to as the suspension of the cone. The outer edge of the
suspension is fastened to the frame of the loudspeaker and the frame is secured to the
An ideal loudspeaker should produce a uniform sound pressure at all frequencies
with non-directional characteristics, have a good dynamic range, high efficiency, and
produce no distortion. The characteristics of the practical loudspeaker are quite
different from the above and the reason is that the properties required to produce a
low-frequency vibration are quite different from those required to produce a highfrequency sound. Furthermore, the environment in which the loudspeaker operates
has a significant effect on its performance.
When a current is passed through the voice coil, a force is exerted on the voice
coil by the interaction of the flux due to the permanent magnet and the flux generated
by the current in the voice coil. The result is that the voice coil moves and so does
the cone to which it is attached. When an ac signal is applied, the cone will vibrate in
accordance with the signal, producing a sound. The cone, the voice coil, and the
cardboard former on which it is wound have a finite mass, as has the air which the
cone has to displace as it moves back and forth. Although the suspension is designed
to offer the minimum mechanical resistance to motion, this resistance is not zero.
When suitable assumptions have been made about the stiffness of the cone and the
linearity of the suspension system and so on, a differential equation describing the
motion of the cone can be written and used to study the effects of changing the
parameters of the system on its performance. The following broad conclusions can
be drawn from both theory and practice.
(1) The force driving the cone is proportional to the magnetic flux density in the
air gap (B), the length of the conductor in the voice coil (l ) and the current
flowing in the coil (i):
f ¼ Bli:
A high flux density can be obtained by using a large powerful magnet in
conjunction with a small air gap. The minimum air gap is dictated by the
thickness of the voice coil former and the diameter of the conductor used to
wind the coil and the need to keep the voice coil from touching the magnet
assembly. The longer the wire in the coil the greater its mass and, since it is
part of the cone, a larger force will be required to drive it. A large coil current
makes demands on the amplifier required to drive it.
(2) A good low-frequency response requires a large (area) cone. A large cone
will, in general, have a relatively large mass. However, a good high-frequency
response requires a cone of very small mass. It would appear that it is not
possible to use one cone for the complete range of frequencies in the audio
signal. A reasonable compromise, one which comes with a suitable price tag,
is to use a filter to divert the low-frequency components of the signal to a
large loudspeaker (woofer) and the high frequency to a small loudspeaker
(3) When the frequency is low, the cone moves as one lumped mass, that is, all
points on the cone remain in the same position relative to each other as the
cone moves back and forth. As frequency increases, the cone starts to behave
like a distributed mass and a phase difference between different parts occur.
The result is that the response becomes irregular, with peaks and troughs
depending on whether the radiated signals from the different parts of the cone
add or subtract from each other at a given point within its field of radiation.
(4) The sound wave from the front of the loudspeaker must be separated from the
back wave. Otherwise, at low frequency the high pressure created by the
forward movement of the cone is neutralized by the low pressure created at
the back of the cone. The energy put into the loudspeaker simply moves the
air around the edge of the loudspeaker instead of being converted into
radiated energy. This can be prevented by mounting the loudspeaker on a
large baffle whose dimensions are at least greater than one-quarter the
wavelength of the lowest frequency. Similar results can be obtained by
enclosing the back of the loudspeaker in a closed box.
(5) The cone, as a mass supported on its suspension, is a system which can have
a resonant response when suitably excited. It so happens that the resonant
frequency is usually in the low-frequency end of the loudspeaker response.
By adjusting the mass of the cone, the characteristics of the suspension and
the damping, the peak of the resonant response can be moved to a frequency
that improves the low-frequency performance of the loudspeaker.
The loudspeaker will generate harmonic distortion when the force–displacement
characteristics of the cone suspension are non-linear. Another source of distortion is
the non-uniformity of the flux density in the air gap, especially when the voice coil is
at the extremities of its excursion.
Frequencies between 3 MHz and 30 MHz are set aside for use in long distance radio
communication. This frequency band is generally referred to as the short-wave band.
The wavelength that lies between 1 m and 100 m is short compared to that of the
medium wave. It was pointed out in Section 2.9 that the increased frequency of the
carrier used in short-wave communication gives its electromagnetic transmission
wave increasingly directional properties. Transmission antennae can be designed to
concentrate the power of the transmitter in a given direction. The ionosphere and the
Earth’s surface are used as reflectors to direct the signals from the transmitter to the
receiver when very long distances have to be covered.
The superheterodyne arrangement is used in short-wave radio receivers. However,
because the radio frequency is much higher the intermediate frequency used is also
high. Intermediate frequencies in use with short-wave radio are usually between 1.5
and 28 MHz. In some special cases, a double superheterodyne system is used. The
radio frequency is translated to the first intermediate frequency, a high frequency,
and then translated the second time to a lower frequency, between 455 kHz and
5 MHz. The advantage of this system is improved selectivity and image rejection.
Lovering, W. F., Radio Communication, Longmans, Green & Co. Ltd., London, 1966.
Smith, J., Modern Communication Circuits, McGraw-Hill, Inc., New York, 1986.
Olson, H. F., Elements of Acoustical Engineering, Van Nostand, Inc., New York, 1947.
Hutchinson, C. L. (Ed), The ARRL Handbook for the Radio Amateur, The American Radio
Relay League, Newington, CT, 1988.
Johnson, D. E., Introduction to Filter Theory, Prentice-Hall, Englewood Cliffs, NJ, 1976.
Herrero, J. L. and Willoner, G., Synthesis of Filters, Prentice-Hall, Englewood Cliffs, NJ,
Humphreys, D. S., The Analysis and Synthesis of Electrical Filters, Prentice-Hall Inc.,
Englewood Cliffs, NJ, 1970.
Zverev, A. I., Handbook of Filter Synthesis, John Wiley, New York, 1967.
Van Valkenburg, M. E., Analog Filter Design, Holt, Rinehart and Winston, New York, 1982.
Huelsman, L. P. (Ed). Active Filters: Lumped, Distributed, Integrated, Digital and Parametric,
McGraw-Hill, New York, 1970.
Haykin, S. S., Synthesis of RC Active Filter Networks, McGraw-Hill, London, 1969.
Sedra, A. S. and Brackett, P. O., Filter Theory and Design: Active and Passive, Matrix, Inc.,
Champaign, IL, 1978.
An AM receiver is tuned to 1000 kHz but two other stations (the first operating
at 1010 kHz and the second operating at 1200 kHz) are causing interference.
The intermediate frequency of the receiver is set at 455 kHz. Explain with the
aid of suitable calculations which of the two stations is most likely to cause the
interference. The intermediate frequency is now set at 100 kHz. Will this
change improve the situation? Explain your answer fully.
A diode which has characteristics described approximately by
i ¼ 5 þ v þ 0:05v2 mA
is to be used in a demodulator of an AM receiver. The carrier frequency is
1 MHz with a 100% modulation by a 5 kHz sinusoidal signal. Assuming that
both the carrier signal source and the local oscillator have a 5 V output on open
circuit and a 1 kO internal resistance, design a suitable (single-tuned LC)
circuit using the diode, transformers (inductors) and capacitors to select the
desired intermediate-frequency signal and to supply maximum power to a
resistive load 550 O. Provide a suitable circuit diagram and specify the values
of all the circuit elements used such that the sideband signals lie within the
3 dB bandwidth. Calculate the mixer gain in decibels.
A switching-type mixer, such as shown in Figure 3.12, is driven from an ideal
sinusoidal voltage source of value 5 V peak. The local oscillator voltage
(squarewave source) is assumed to be sufficient to cause instantaneous
switching of the diodes. Calculate the amplitudes of the output signal for
n ¼ 1, 2, and 3 and their frequencies if the sinusoidal source has a frequency of
1 MHz and the squarewave is 8 MHz.
The secondary winding of a transformer has self-inductance L ¼ 500 mH and
resistance of 15 O. The primary current is an amplitude modulated wave with a
carrier frequency of 500 kHz modulated by a single-frequency sinusoidal
signal of 7.5 kHz to a depth of 50%. Calculate the value of the capacitor
required to tune it and the amplitude of the upper or lower sideband signal
relative to the carrier in the secondary winding.
The collector load of a bipolar transistor amplifier consists of a coil of selfinductance L1 ¼ 1 mH whose Q factor is equal to 100 when tuned to resonance
at 500 kHz by a parallel capacitor which may be assumed to be lossless. The
amplifier drives a resistive load through a perfectly coupled coil, the mutual
inductance being equal to 0.5 mH. If the resistive load has a value 6.25 kO,
determine the equivalent impedance of the collector load. What is the 3 dB
bandwidth of the circuit? Calculate the value of the emitter resistance for a
voltage gain of 25.
Using the MOSFET whose characteristics are given in Figure P3.1, design a
class-A amplifier with a voltage gain of 10. Allow for a drain current of 3 mA
when the dc supply voltage is 12 V. What is the maximum sinusoidal input
voltage that the amplifier can take without clipping? Calculate the input
impedance of the amplifier.
Repeat the design using the device whose characteristics are given in Figure
Using the device of Figure P3.1, design a transformer coupled class-A
amplifier to operate at 500=2p kHz with a voltage gain of 10 and a 3 dB
bandwidth of 20=2p kHz. The load resistance is 100 O and the dc supply
voltage is 15 V. Specify the values of all components used. You may assume
that the windings of the transformer have negligible resistances.
Figure P3.1.
Figure P3.2.
An AM signal applied to an envelope detector has carrier frequency
oc ¼ 250 krad=s. The upper envelope is a triangular waveform which goes
from þ 10 V to þ 2 V and back up to þ 10 V at a constant rate in 0.2 ms (the
lower envelope is similar!). Assuming that the diode is ideal, calculate the
maximum time-constant of the detector load to avoid diagonal clipping. Justify
any approximation that you make.
In Chapter 2, the amplitude of a high-frequency (carrier) sinusoidal signal was varied
in accordance with the waveform of an audio-frequency (modulating) signal to give
an amplitude modulated (AM) wave which could be transmitted, received, and
demodulated to recover the original audio frequency signal.
In frequency modulated (FM) radio, the frequency of the carrier is varied about a
fixed value in accordance with the amplitude of the audio frequency. The amplitude
of the carrier is kept constant. The waveform of a sinusoidal carrier modulated by a
saw-tooth wave is shown in Figure 4.1.
All signals carried on any transmission system will sooner or later be contaminated by noise so the susceptibility of the communication system to noise is an
important consideration. The noise can be defined as a random variation superimposed on the signal. In AM systems, the information to be transmitted is
contained in the envelope of the carrier signal. The noise therefore appears on the
envelope and has a direct role in corrupting the signal. In FM systems the
information to be transmitted is contained in the variation of the frequency of the
carrier about a pre-set value. The amplitude of the FM signal is kept constant and,
indeed, if there are changes in the amplitude of the FM signal, they are removed by
clipping before demodulation. By comparison, FM systems are less susceptible to
degradation by noise.
While a saw-tooth modulating signal provides a simple picture of the FM signal, a
sinusoidal modulating signal is the simplest for the derivation of the mathematical
expressions to describe the FM signal.
Figure 4.1. The sawtooth waveform vs frequency modulates a carrier to give the output vfm .
Note that the relative change in frequency has been exaggerated for clarity. In normal FM radio,
the change in frequency relative to the carrier is less than 0.15%
A sinusoidal voltage can be expressed as:
v ¼ A sin ot
v ¼ A sin yðtÞ
where o is a constant representing the angular velocity of the sinusoid and y is a
phase angle with respect to an arbitrary datum. In general, the relationship between
the phase angle and the angular velocity is given by
¼ oðtÞ:
In a frequency modulated system, o is varied about a fixed value oc , in accordance
with the modulating signal which is assumed in this case also to be a sinusoid:
vs ¼ B cos os t:
The instantaneous angular velocity
oi ¼ oc þ Doc cos os t
where oc is the long-term mean angular velocity, Do o, and Doc is the
maximum deviation of the angular velocity about oc .
Substituting instantaneous values into Equation (4.2.3)
oi ðtÞ ¼
dyi ðtÞ
Substituting Equation (4.2.5) into Equation (4.2.6) and integrating
yi ðtÞ ¼ f þ
oc dt þ
Doc cos os tdt
where f is the initial value of the phase angle which without loss of generality can
be set equal to zero. Then substituting Equation (4.2.7) into Equation (4.2.2) gives
vfm ðtÞ ¼ A sinðoc t þ
sin os tÞ
where kDoc ¼ B, the amplitude of the modulating signal.
The term koc =os is called the modulation index, where
max: frequency deviation of the carrier
modulating signal frequency
vfm ðtÞ ¼ A sinðoc t þ mf sin os tÞ:
mf ¼
vfm ðtÞ ¼ A½sin oc t cosðmf sin os tÞ þ cos oc t sinðmf sin os tÞ:
The terms cosðmf sin os tÞ and sinðmf sin os tÞ can be expanded in the form of Fourier
series with coefficients which are Bessel functions of the first kind Jn ðmf Þ where n is
the order and mf is the argument, to give
cosðmf sin os tÞ ¼ J0 ðmf Þ þ 2
J2n ðmf Þ cos 2nos t
J2nþ1 ðmf Þ sinð2n þ 1Þos t:
sinðmf cos os tÞ ¼ 2
Substituting Equations (4.2.12) and (4.2.13) into Equation (4.2.11) and using
cos x sin y ¼ 12 ½cosðx þ yÞ þ cosðx yÞ
sin x sin y ¼ 12 ½cosðx þ yÞ þ cosðx yÞ
the result is
vfm ðtÞ ¼ AfJ0 ðmf Þ sin oc t
þ J1 ðmf Þ½sinðoc þ os Þt sinðoc os Þt
þ J2 ðmf Þ½sinðoc þ 2os Þt sinðoc 2os Þt
þ J3 ðmf Þ½sinðoc þ 3os Þt sinðoc 3os Þt þ g
where Jn ðmf Þ is plotted against mf for various values of n in Figure 4.2.
From Equation (4.2.16) it can be seen that:
(1) The carrier frequency oc is present and its amplitude is determined by the
modulation index mf .
(2) The next term represent two frequencies which are the sum ðoc þ os Þ and
difference ðoc os Þ of the carrier and the modulating frequency with an
amplitude J1 ðmf Þ.
(3) The next two terms have an amplitude J2 ðmf Þ and frequencies ðoc þ 2os Þ and
ðoc 2os Þ.
(4) There are an infinite number of sums and difference of the carrier and integer
multiples of the modulating signal.
Figure 4.2. A plot of Bessel functions of the first kind, Jn ðmf Þ against mf for n ¼ 0; 1; 2 and 3.
The values of Jn ðmf Þ are used to calculate the amplitudes of the side-frequencies present in the
FM signal.
It would appear that, in order to transmit a simple sine wave in an FM system, an
infinite bandwidth is required to accommodate all the multiple sidebands. However,
from Figure 4.2 it can be seen that, as n increases, the amplitudes of the sidebands
decrease and their contribution to the signal power falls off rapidly. A second aspect
of the bandwidth requirements of the FM system can be seen from Equation (4.2.8).
Unlike the AM system, where a modulation index greater than unity causes severe
distortion, the modulation index in FM does not appear to have an upper limit except
that, for a fixed modulating signal frequency os, increasing the modulation index mf
means a greater deviation of the FM signal from the carrier frequency. This implies a
larger bandwidth. With no apparent technical limits to the bandwidth requirements
and to permit the maximum number of FM stations to function with minimum
distortion and interference, the maximum frequency deviation for commercial FM
radio is set at 75 kHz. However, since there is substantial signal power beyond the
75 kHz limit, the actual bandwidth is set at 200 kHz, with the highest modulating
frequency limited to 15 kHz.
Frequency modulated communication channels can be found from approximately
1600 kHz to 4000 MHz. Parts of this spectrum are reserved for the use of police,
VHF and UHF television sound channels, VHF mobile communications, and pointto-point communication. The commercial FM radio broadcast band is from 88 to
108 MHz.
Basic System Design
A simple method for generating an FM signal is to start with any LC oscillator. The
frequency of oscillation is determined by the values of C and L. If a variable
capacitor DC is connected in parallel with C and the capacitance variation is
proportional to the modulating signal then an FM signal will be obtained.
Consider a negative conductance oscillator as shown in Figure 4.3. The negative
conductance may be obtained from a tunnel diode, a suitable biased pentode, or a
bipolar junction transistor with a suitable feedback circuit.
The frequency of oscillation is given by
o2c ¼
When a variable capacitor DC is connected in parallel with C, the frequency of
oscillation will be:
ðoc Doc Þ2 ¼
Any p–n junction diode in reverse bias can be used to realise DC. However the
relationship between DC and the bias voltage is non-linear and appropriate steps will
Figure 4.3. The basic FM generator; the value of the variable capacitor DC is controlled by the
modulating signal which in turn determines the frequency of the oscillator.
have to be taken, such as using two diodes in push–pull or specially fabricated
diodes (e.g. varactor diodes or diodes with hyper-abrupt junctions, both of which
come under the classification voltage-controlled capacitors).
The direct generation of FM signals presents further difficulties. Assuming that
the transmitter were to operate at 100 MHz, the required inductance and capacitance
would be approximately 160 nH and 16 pF, respectively. Stray inductances and
capacitances associated with the circuit will render it impractical. A less impractical
idea is to generate the FM signal at a lower frequency, where larger values of L and C
are required, and then to use a cascade of frequency multipliers to raise the frequency
to the required value. Supposing that the low-frequency FM signal is generated at
200 kHz. Suitable values for L and C are 1.27 mH and 500 pF. Since the frequency
has to be multiplied by 500 (actually 29 ¼ 512) to place it within the commercial
FM frequency band, the maximum frequency deviation has to be
75 kHz=500 ¼ 150 Hz so that the frequency deviation will not exceed the legal
limit. The required change in capacitance DC ¼ 2:5 pF. This is within the range of
capacitance change that can be obtained from a p–n junction diode. A block diagram
of the system is given in Figure 4.4.
Figure 4.4. An improved FM generator design in which the FM signal is generated at a lower
frequency and then multiplied by a suitable factor to give the required frequency.
The design of all the blocks shown in Figure 4.4 was discussed earlier. However,
frequency stability requirements of the system makes the approach shown in Figure
4.4 impractical.
One of the requirements of a transmitter is that its frequency will remain at the
assigned value at all times so that its frequency does not drift and cause interference
with other channels of communication. In the scheme shown in Figure 4.4 there is
no built-in mechanism to ensure that the long-term average frequency will be
constant. A modified scheme is shown in Figure 4.5, in which the carrier frequency
of the FM signal is compared with that of a crystal-stabilized oscillator and the
resulting error signal is used to correct the carrier frequency.
Automatic Frequency Control of the FM Generator
The main oscillator is an LC-type similar to the negative conductance oscillator
discussed earlier. The variation DC generates the FM signal. The output is fed into a
buffer amplifier which provides isolation between the oscillator and its load so that a
changing load will have minimal effect on the oscillator operation. The signal
proceeds to the limiter which removes any AM which may have occurred and also
filters out the harmonics generated by the limiter. Part of the amplitude-limited FM
signal is then fed to a mixer whose other input is from a crystal-controlled oscillator.
The difference frequency is suitably amplified, filtered and then fed into a
discriminator to produce a dc signal proportional to the difference between the
required and the crystal-controlled oscillator frequency. The dc or error signal is
Figure 4.5. The circuit in Figure 4.4 has been improved by the addition of an automatic
frequency control (AFC) circuit to stabilize the carrier frequency.
used to keep the main oscillator at the required frequency. Using this scheme, it is
possible to impose the stability of the crystal-controlled oscillator on the main
oscillator. It should be noted that the time-constant of the error signal should be
chosen so it can correct for the long-term frequency deviation of the main oscillator
without affecting the short-term frequency deviation due to the modulating signal.
The output of the limiter=filter is fed into a chain of multipliers to bring the
frequency to its operating value. A power amplifier then boosts the signal power to
the desired value and drives the antenna which radiates the signal.
Component Design with Automatic Frequency Control
Two new component blocks were introduced into the system to realize the parameter
variation method. These were the limiter and the discriminator. The circuit design of
these component blocks follow. The Amplitude Limiter. An FM signal, by definition, must have a
constant amplitude. In practice, circuit non-linearities cause variations in the
envelope of the FM signal related to the modulating signal, that is, some amplitude
modulation takes place. It is evident that any AM present in an FM system will
interfere with the signal during the demodulation process since most FM demodulators convert the variation of frequency to a variation of amplitude before detection.
A second reason for limiting the amplitude of the FM signal is that the noise present
in the communication channel generally rides on the envelope of the signal and
therefore by clipping the amplitude of the signal some of the noise can be removed.
The ideal amplitude limiter can accept an input signal of any amplitude and
convert it to one of constant amplitude. A practical approximation to the performance of the ideal limiter is obtained by amplifying the input by a large factor and
clipping a small portion symmetrically about the time axis. The output waveform
will then be a square wave. In order to get back the sinusoid, the signal must be
passed through a bandpass filter which will remove all the harmonics leaving only
the fundamental. The amplitude limiter has three identifiable parts: the pre-amplifier,
the symmetrical clipper, and the bandpass filter. The sub-system block diagram is
shown in Figure 4.6.
In Section 4.2, it was shown that an FM signal has an infinite number of
sidebands but that it was not necessary to preserve all the sidebands in order to
maintain a high level of fidelity. Working with the permitted modulation index,
mf ¼ 5, and retaining sidebands with coefficients Jn ðmf Þ greater than 0.01 (i.e., 1%
of the unmodulated carrier) it can be shown that the first eight sidebands must be
Figure 4.6. A block diagram showing the details of the limiter=filter shown in Figure 4.5.
preserved. Since the sidebands are spaced os apart, and os has a maximum value of
15 kHz, the required bandwidth is 240 kHz.
The (3 dB) Q factor of the pre-amplifier with the required bandwidth will be
just over 400. However, this would mean that the outermost sidebands will be
subjected to a 3 dB attenuation compared to those nearer the carrier, causing further
variation in the amplitude of the signal. In a practical situation, the modulating signal
is made up of a large number of discrete frequencies and this causes further
complications in the calculation of the required bandwidth. An amplifier with a flat
frequency response could be used but it will have the disadvantage of amplifying the
noise below and above the spectrum of interest. A compromise would be to use a
tuned pre-amplifier with a much lower Q factor. It will have the advantage of a
minimal variation in its response over the spectrum occupied by the significant
sidebands and yet attenuate the noise present in the rest of the spectrum. A loaded Q
factor of about 20 should be adequate.
It is accepted practice to generate the FM at a lower frequency and to use a
cascade of frequency multipliers to bring it up to the operating value. If the FM is
generated at 200 kHz, the appropriate frequency deviation will be about 150 Hz. The
corresponding bandwidth to accommodate the first 8 sidebands will be 470 Hz.
Figure 4.7 shows a circuit diagram of the pre-amplifier and its load, the
symmetrical clipper. The detailed design of a similar amplifier was presented in
Section 3.4.2 and will not be repeated here. However, it is worth noting that the
diodes can be considered as short-circuits and hence the resistance R will be in
parallel with the tuned circuit. Symmetrical Clipper. A simple symmetrical clipper is shown in
Figure 4.7. The resistance R is chosen so that the Q factor of the tuned circuit is
Figure 4.7. The circuit diagram of the pre-amplifier and the symmetrical clipper.
approximately equal to 20. The details of the design of a parallel tuned circuit with a
specified Q factor was discussed in Section 3.4.2.
The output of the clipper will be about 0.7 V when single silicon diodes are used
or multiples of 0.7 V, depending on how many diodes are connected in series to
determine the level of clipping. It is evident that the more severe the clipping is, the
more likely it is that the output of the clipper will be a square wave but then more
gain will be required in the succeeding stages.
A more elegant design when the clipping level is high is to use back-to-back
Zener diodes as shown in Figure 4.8. The Zener diode provides a much sharper cutoff than an ordinary diode when it is reverse biased. In the forward direction, it is an
ordinary diode. With this arrangement, only one of the two Zener diodes is operating
as a Zener diode while the other one is conducting current in the forward direction.
They switch roles when the applied voltage changes polarity. Bandpass Filter. It is reasonable to assume that the output of the
clipper is a square wave. The frequency of the square wave varies but it is centered at
the carrier frequency. Fourier analysis shows that the most significant harmonic in
the square wave is the third. So with the sub-carrier at 200 kHz, the third harmonic
will be at 600 kHz. A simple tank circuit tuned to the carrier frequency is all that is
necessary to filter out the harmonics. But, as before, the choice of the Q factor of the
tuned circuit must be made to ensure that all the significant sidebands are within the
passband of the filter and are subjected to minimum attenuation. The factors that
affect the choice of the bandwidth were discussed in Section Discriminator. The purpose of the discriminator is to convert the
variation of frequency to a variation of amplitude. A frequency-to-amplitude
convertor followed by an envelope detector is used to recover the message contained
in the modulating signal. The transfer characteristics of two circuits that could be
used for the frequency-to-amplitude conversion are shown in Figure 4.9.
The simplest circuit with the characteristics shown in Figure 4.9(a) is a simple RC
high-pass circuit with its corner frequency much higher than the carrier frequency of
the FM signal. Figure 4.9(b) shows the response of a simple RC lowpass circuit with
its corner frequency chosen to be much lower than the carrier frequency. Both
Figure 4.8. The use of two Zener diodes connected back-to-back is an improvement on the
symmetrical clipper shown in Figure 4.7.
Figure 4.9. The required characteristics of a frequency-to-amplitude converter: (a) is a highpass circuit and (b) is the low-pass version.
circuits should, in principle, convert the FM signal into an AM signal. However, in
practice the sub-carrier frequency of the system shown in Figure 4.5 will be about
200 kHz, while the variation will be limited to a maximum of 150 Hz. Since both
of these circuits have a slope of 6 dB=octave, the variation in amplitude of these
circuits will be exceedingly low. The use of a high gain amplifier following such a
circuit will lead to increased noise. A better approach is to find a circuit which has a
much greater amplitude-to-frequency slope.
A tuned RLC circuit has a rapid change of amplitude with frequency on both sides
of the resonance frequency, especially when the Q factor of the circuit is high. The
circuit and its response are shown in Figure 4.10. It is evident that this circuit will be
Figure 4.10. A single LC tuned circuit used as a frequency-to-amplitude converter. In this
case, the skirt with the positive slope (high-pass) has been used. The skirt with the negative
slope (low-pass) would be equally effective. Note that because the nonlinearity of both skirts is
predominantly a second-order function, the distortion products of the circuit will be mainly even
prone to produce distortion, especially of even harmonics (the non-linearity is
A variation on the circuit in Figure 4.10 is shown in Figure 4.11. The two tuned
circuits made up of L1 –C1 and L2 –C2 are tuned to two different frequencies equally
spaced from the carrier frequency. The individual and composite characteristics are
shown in Figure 4.11. It can be seen that this circuit has a greater dynamic range
than the previous circuit and, when properly tuned, should have no even harmonic
distortion products in the output. Envelope Detector. The basic envelope detector was discussed in
Section 3.4.6. The importance of the choice of the time-constant in order to avoid
distortion of the envelope was discussed at some length. The purpose of this
envelope detector is to correct the long-term frequency deviation of the main
oscillator while allowing the short-term frequency deviation caused by the modulating signal. Its time-constant is therefore chosen on the basis that the lowest
frequency present in the modulating signal, os;min , will not cause the frequency
correction system to go into operation. This condition is satisfied when the timeconstant of the detector, tdet , is chosen so that
os;min tdet 1:
The use of automatic frequency control in the generation of FM signals results in a
practical scheme but one which depends on feedback. For proper operation of the
system all of the circuits must be made to track each other. Furthermore, the voltagecontrolled capacitor used in the automatic frequency control circuit has a nonlinearity as well as a temperature characteristic which must be compensated for.
Evidently, a different scheme with less complexity is needed. The Armstrong
system, named after its inventor, is one such scheme. The theoretical basis of the
Armstrong system is given by Equation (4.2.7). It is clear that, in commercial FM
systems, the frequency deviation must be small compared to the carrier frequency,
that is:
oc ðtÞ k
j sin os tj ¼ mf j sin os tj
The following approximations can be applied to Equation (4.2.11):
cosðmf sin os tÞ 1
sinðmf sin os tÞ mf sin os t
Figure 4.11. The use of a double-tuned circuit provides increased dynamic range and improves the
linearity by balancing the non-linearity of one LC circuit with the other. Note that the ‘‘push-pull’’
arrangement shown here will generate predominantly odd harmonics.
to give
vfm ðtÞ ¼ A½sin oc t þ ðmf sin os tÞ cos oc t
vfm ðtÞ ¼ Afsin oc t þ 12 mf ½sinðoc os Þt þ sinðoc þ os Þtg:
It should be noted that the above approximation eliminates all the sidebands except
the two closest to the carrier. The modulation is then said to be a narrow band
frequency modulation ðNBFM Þ. The advantage it offers is a simple geometric
representation of the FM signal as a phasor.
The term ½ðmf sin os tÞ cos oc t is a double-sideband-suppressed carrier (DSBSC) signal. The addition of the carrier term (sin os tÞ would normally produce an AM
signal but, in this case, the carrier is 90 out of phase with the DSB-SC signal. The
difference between AM and FM is shown in the phasor diagrams in Figure 4.12.
Figure 4.12. (a) A phasor diagram for an NBFM showing the two phasors of amplitude 12 Amf
rotating in opposite directions at angular velocity os in quadrature to the phasor A at t ¼ 0. (b) A
phasor diagram for an AM signal showing two counter-rotating phasors of amplitude 12 kA and
angular velocity os . Note that the amplitude of the resultant phasor varies from Að1 kÞ to
Að1 þ kÞ.
Assume a coordinate system in which the phasor A sin oc t rotates anticlockwise
at an angular velocity oc. Suppose that the rotating phasor is used as a new reference
system so that it is represented by a horizontal phasor of value A. In terms of this
reference system, the term ð12 Amf Þ sinðoc þ os Þt is represented by a phasor of value
2 Amf rotating in an anticlockwise direction with an angular velocity of os . Similarly,
the term ð12 Amf Þ sinðoc os Þt is represented by a phasor of equal value rotating in a
clockwise direction with an angular velocity of os . At time t ¼ 0, the component of
each rotating phasor in the horizontal direction is zero. Therefore the two phasors
must be parallel to each other but at right angles to the horizontal phasor A. This
situation is depicted in Figure 4.12(a). The resultant is a phasor which apparently has
a maximum value
A ð1 þ m2f Þ
and a maximum angular displacement of
c ¼ tan1 ðmf Þ:
c mf :
Since mf is small, c is small and
The apparent variation of the amplitude of the FM signal is due to the fact that this
analysis has not taken into account the many sidebands present in FM.
For comparative purposes the AM system is treated similarly. An AM signal is
represented by Equation (2.2.7):
vam ðtÞ ¼ A sin oc t þ
cosðoc os Þt cosðoc þ os Þt:
Again the carrier A sin oc t is represented by a horizontal phasor of value A and the
two sidebands by two counter-rotating phasors of value kA=2 rotating at os . At time
t ¼ 0, the components of both of the sideband in the horizontal direction are zero
since they are cosine terms and A is a sine term. The phasors representing the two
sidebands must be at right angles to the A phasor as shown in Figure 4.12(b). Note
that one is positive and the other negative. At some other time t > 0, the phasor
representing the combination of the sidebands subtracts from or adds to the phasor
A. The amplitude of the resultant therefore varies from Að1 kÞ to Að1 þ kÞ but at all
times it is in the horizontal position.
The conclusion is that in a narrow-band (Doc oc ) frequency modulated
system, the phase difference between the carrier and the DSB-SC signal is 90 .
This is the basis of the Armstrong system. A block diagram of the FM transmitter is
shown in Figure 4.13.
Practical Realization
A crystal-controlled oscillator generates the carrier frequency A cos oc t. This signal
is fed to two blocks. The first is a 90 phase shift circuit which, as the name suggests,
Figure 4.13. Block diagram of the basic Armstrong system FM transmitter. This is an example
of an indirect generation of FM signal.
shifts the phase by 90 and therefore its output is A sin oc t. The second block is a
balanced modulator. The other input to the balanced modulator is, of course, the
modulating signal. But according to Equation (4.2.8), mf ¼ Doc =os . The modulating (audio) signal contains a band of frequencies from approximately 20 Hz to
15 kHz and therefore as os increases, Do must increase or the modulation index mf
must decrease. However, in FM the frequency of the carrier changes proportionately
to the amplitude of the modulating signal, not to the frequency of the modulating
signal. In order to satisfy this condition, the amplitude of the modulating signal has
to be reduced by one-half whenever the frequency doubles. This amounts to putting
the signal through a low-pass filter with a slope of 6 dB=octave. An integrator with
its corner frequency below the lowest modulating frequency is required. The
amplitude of the modulating signal frequency band is thereby ‘‘predistorted’’ by
the integrator to conform to the above condition. The need for the integrator is
evident from Equation (4.2.7).
If the modulating signal is B sin os t the output of the integrator will be
ðB=os Þ cos os t. This signal is fed to the balanced modulator which produces as an
output the product of its two inputs, namely, ðk 0 AB=os Þ sin oc t cos os t, where k 0 is a
constant. This signal and the output of the 90 phase shifter go to the adder which
gives an output
vo ðtÞ ¼ A sin oc t þ
k 0 AB
sin oc t cos os t:
The signal goes through a series of multipliers to bring the frequency up to the
operating value. A power amplifier raises the signal to the proper level for radiation
by the antenna.
In a practical FM transmitter, the carrier is generated at a lower frequency and the
modulation process carried out. The frequency is then multiplied up to the operating
value. If it is assumed that the sub-carrier frequency is again 200 kHz – a frequency
at which a suitably robust crystal can be found – the required multiplication factor to
place it in the middle of the commercial FM frequency band is then about 500
(29 ¼ 512). The maximum frequency deviation for a sub-carrier at 200 kHz was
calculated in Section 4.3.4 to be 150 Hz.
The ratio of the sub-carrier frequency to the maximum frequency deviation may
not be the most convenient in a practical system. In order to produce a high-fidelity
signal, that is, maintain a high level of linearity, it is an advantage to make the carrier
frequency independent of the frequency deviation. This can be done by splitting the
multiplication operation in two and inserting a mixer between them so that the
carrier frequency can be changed without affecting the frequency deviation. To help
explain the process, an example is given in Figure 4.14.
Suppose the sub-carrier frequency is 200 kHz and the maximum deviation is
20 Hz. To get the sub-carrier frequency up to the normal operating frequency of
about 100 MHz, a cascade of frequency multipliers equal to 29 (512) will be
required. Multiplication of the carrier also multiplies the frequency deviation so
that the corresponding frequency deviation will be ð512 10 Hz) ¼ 10,240 Hz. This
is much lower than the limit allowed, which is 75 kHz. On the other hand, in order to
convert the 20 Hz frequency deviation into the required 75 kHz, the multiplication
factor is 3750 (approximately 3 210 ¼ 3072). However, multiplying the subcarrier by this factor will give a carrier frequency of 614.4 MHz, which is outside
the commercial FM radio band. To correct this, the multiplication process is split
into two stages and a mixer is inserted between them. The first stage of multiplication may be 96 (3 25 ). The carrier frequency is then 19.2 MHz and the
corresponding frequency deviation is 1.92 kHz. The signal is now mixed with a
‘‘local oscillator’’ whose output frequency is 16.2 MHz to produce a carrier
frequency of 3.0 MHz but the frequency deviation remains unchanged at
1.92 kHz. The second stage of multiplication (32 ¼ 25 ) produces a carrier frequency
of 96.0 MHz and a frequency deviation of 61.4 kHz. This is not quite at the allowed
limit but close enough for practical purposes. The ‘‘local oscillator’’ signal may be
derived from the 200 kHz crystal-controlled oscillator by using a suitable frequency
multiplier (81 ¼ 34 ).
Component Circuit Design
In the block diagram of the Armstrong system given in Figure 4.14, four new
components were added, namely the 90 phase shifter, the balanced modulator, the
integrator, and the adder. The design of these circuit components now follow. The 90 Phase Shift Circuit. In the circuit shown in Figure 4.15(a),
the resistances RC and RE are equal. Since the collector and emitter currents are, for
all practical purposes, equal, it follows that the voltages vC and vE must be equal in
magnitude. But the phase angle between them is 180 . The voltages vC and vE are
Figure 4.14. Block diagram of a practical Armstrong system FM transmitter.
represented on the phasor diagram shown in Figure 4.15(b). The voltage which
appears across R and Z is ðvC þ vE Þ. If the impedance Z is purely reactive (capacitive
or inductive), it follows the phasors vR and vZ will always be at right angles to each
other. It follows then that the point A must lie on a circle whose radius is equal to the
magnitude of vC as shown. The output voltage is then vo . As frequency goes from
zero to infinity, point A will trace out a semicircular locus. The output voltage vo
changes its phase angle from zero to 180 but its magnitude remains constant.
It can be seen that, with the appropriate choice of R and Z at a given frequency, vo
can be made to lead or lag vin by 90. One major disadvantage of this circuit is that
its load must be a very high impedance, preferably open-circuit, otherwise the
magnitude of the output voltage vo will not remain constant for all frequencies. This
circuit is a simple example of a class of networks known as all-pass filters.
Example 4.4.1 90 phase shift circuit. Design an all-pass circuit with a phase
shift of 90 at 200 kHz. The dc supply voltage is 12 V and it may be assumed that the
circuit drives a load of very high impedance. The b of the transistor is 100.
Solution. A suitable circuit for the phase shifter is shown in Figure 4.16 with the
impedance Z replaced by a single capacitance C.
Since the gain is unity vC ¼ vE and the stage will have the maximum dynamic
range when the transistor is biased with VE at Vcc =4 and Vc at 3Vcc =4, that is, at 3 and
9 V, respectively. The maximum signal voltage that can be applied to the RC series
circuit is then 12 V peak-to-peak. The equivalent circuit is then as shown in Figure
Let the peak current in the RC circuit be approximately equal to 1 mA. Then
R ¼ 4:2 kO and the reactance of the capacitor must also be 4.2 kO so that at
200 kHz, C ¼ 190 pF.
Figure 4.15. (a) An active 90 phase shift circuit. (b) Phasor diagram for the circuit in (a).
Figure 4.16. Circuit diagram used in Example 4.4.1.
The dc collector current in the transistor must be chosen so that the 1 mA taken
by the RC circuit will have negligible effect on the transistor operation. A collector
current of 10 mA will be adequate. Under quiescent conditions, the emitter voltage is
3 V. Hence RE ¼ 300 O and RC ¼ 300 O. Since the b of the transistor is 100, the
base current is 100 mA. Allowing a current 10 times the base current in the resistive
chain (i.e. 1 mA) makes ðR1 þ R2 Þ ¼ 12 kO. But the voltage at the base is 3.7 V and
therefore R2 ¼ 3:7 kO and R1 ¼ 8:3 kO. The coupling capacitor has to be chosen so
that it is essentially a short-circuit at 200 kHz; a 0.1 mF capacitor is adequate. The 90 Phase Shift Circuit: Operational Amplifier Version.
When the frequency of operation is within the bandwidth of an operational amplifier,
the circuit shown in Figure 4.18 can be used to realize a 90 phase shift.
The analysis of the circuit is simplified by the application of superposition. The
first step is to separate the inverting input from the non-inverting input. With the
non-inverting input connected to ground, and vin connected to the inverting input, it
can be shown that the output is
v01 ¼ R2
v :
R1 in
Figure 4.17. An equivalent circuit of the 90 phase shift circuit shown in Figure 4.15(a).
Figure 4.18. (a) An operational amplifier version of the 90 phase shift circuit. The advantage of this
circuit is that its output impedance is low. (b) The circuit in (a) when superposition is applied.
With the inverting input grounded and vin applied to the non-inverting input, it can
be shown that the output is
v :
R1 in
R3 þ
The output when vin is connected to both inputs is
v0 ¼ v01 þ v02
¼ þ
R1 1 þ joCR3
When R1 ¼ R2 and oCR3 ¼ 1,
¼ j1:
The gain is unity and the output leads the input by 90. The main advantage of this
approach is that the output impedance of the circuit is low enough to drive the
balanced modulator directly. Balanced Modulator. A balanced modulator is a circuit which accepts
two input signals of different frequencies, a carrier A cos oc t and a modulating signal
B cos os t, and produces an output proportional to the product of the two inputs
v0 ðtÞ ¼ kAB cos oC t cos oS t
The balanced modulator is therefore a two-input analog multiplier. It is called a
balanced modulator because it produces an AM output with the carrier and
modulating signals suppressed. It is also described as a double-sideband-suppressed
carrier modulator (DSB-SC).
In Section 2.6.3, the four-quadrant analog multiplier was used to obtain an
amplitude modulated wave. From Equation (2.6.26) the output current
I2 ¼ m1 Vdc Idc þ m2 Vdc IC cos oC t þ m3 VS Idc cos oS t þ m4 VS IC cos oC t cos oS t
where the ms are constants.
To obtain the required output, the first three terms of Equation (4.4.17) have to be
removed. The four-quadrant analog multiplier is realized only in integrated circuit
form and one of the basic properties of circuit integration is that the process
produces transistors and other components whose characteristics match very closely.
Common-mode signals can be removed from the output by providing two identical
circuits in ‘‘parallel’’ and taking the output as the difference between two corresponding points in the two circuits. The basic balanced modulator circuit is shown in
Figure 4.19.
It should be noted that the major difference between the circuit of Figure 4.19 and
that of Figure 2.33 is that in Figure 4.19, both transistor combinations Q1 –Q3 and
Q2 –Q4 have a collector load RL and the output is taken differentially. For specific
biasing and detailed information the reader may refer to the specification of the MC
1595 (four-quadrant multiplier produced by Motorola Semiconductor Products,
Inc.). The Integrator. The integrator is, in fact, a simple low-pass filter with a
6 dB per octave slope (20 dB=decade) with its corner frequency chosen to be below
the lowest frequency present in the modulating signal. The basic integrator circuit is
shown in Figure 4.20(a) with the time-constant t ¼ RC chosen so that
oS; min t > 1:
Because the modulating signal contains frequencies spanning almost three decades,
the highest frequency present will be attenuated by approximately 60 dB. Moreover
the very long time-constant required for this application makes the simple RC lowpass circuit unattractive. A more attractive form of the low-pass filter is shown in
Figure 4.20(b).
Figure 4.19. The four-quadrant multiplier used as a balanced modulator (DSB-SC). Note the
difference between this and Figure 2.33.
Figure 4.20. (a) A simple ‘‘integrator,’’ that is, a low-pass filter. (b) A practical integrator can be
realized by connecting a suitable feedback RC circuit to an operational amplifier.
The output voltage of the operational amplifier is
v :
Z1 in
joCR1 in
v0 ¼ When Z1 ¼ R1 and Z2 ¼ 1=ðjoCÞ,
v0 ¼ It can be shown that in the time domain
v0 ðtÞ ¼ 1
vin ðtÞdt:
The name ‘‘integrator’’ is therefore appropriate.
For an ideal operational amplifier (open-loop gain ¼ infinity), the gain of the
circuit shown in Figure 4.20(b) has to be infinite at dc since Z2 is an open-circuit at
dc. In a practical operational amplifier the open-loop gain, although large (approximately 10,000), is not precisely defined. In order to define the dc gain a resistance R2
is connected in parallel with C so that the dc gain is equal to ðR2 =R1 Þ. The output
voltage is then modified to
v0 ¼ Example 4.4.2
v :
R1 1 þ joCR2 in
The Integrator. Design an integrator to satisfy the following
(1) Zero dB gain at 1 kHz,
(2) Corner frequency at 10 Hz,
(3) What is the gain at 10 Hz and 15 kHz?
Solution. The required response is as shown in Figure 4.21.
The relationship between two frequencies f1 and f2 on the basis of decades of
frequency is given by
f1 10n ¼ f2
where n is the number of frequency decades. When f1 ¼ 10 Hz and f2 ¼ 1 kHz,
n ¼ 2.
The gain at 10 Hz ¼ 2 20 dB ¼ 40 dB.
The voltage gain of 40 dB corresponds to a voltage ratio of 100. Therefore,
R2 =R1 ¼ 100.
Choosing R2 ¼ 1 MO makes R1 ¼ 10 kO. From Equation (4.4.22) it follows that
at the corner frequency oCR2 ¼ 1 and C is therefore equal to 0.016 mF. Using the
results for 1 kHz, 15 kHz corresponds to 1.18 decades of frequency. Therefore, the
gain at 15 kHz is
20 1:18 ¼ 23:6 dB: The Adder. The addition of two signals is most easily accomplished by
using a resistive T network as shown in Figure 4.22.
Figure 4.21. The required characteristics of the integrator specified in Example 4.4.2.
Figure 4.22. A simple resistive adder. The use of such a circuit would lead to severe signal
This circuit has two major disadvantages: the output signal is a fraction of the input
signal, and there will be a tendency for one signal to ‘‘pull’’ the other. To increase the
isolation between the two signals, and hence reduce the possibility of pulling, R1 and
R2 must be increased relative to R3 . This aggravates the loss of signal power at the
output of the circuit.
An adder circuit that can scale as well as amplify the signals to be added is shown
in Figure 4.23.
Equation (4.4.19) can be extended to give
v0 ¼ Rf
v fv fv:
R1 1 R2 2 R3 3
The relative values of R1 ; R2 and R3 determine the proportion of v1 ; v2 and v3
present in v0 , respectively, while the ratio Rf to R1 determines the voltage gain for the
voltage v1 and so on.
Figure 4.23. An operational amplifier type adder. This circuit enables the circuit designer to
choose the gain required as well as the multiplication constants.
In an effort to create a realistic sound presentation from recorded music, two
microphones are used, one to record the sound as it is perceived on the right side and
the second on the left side. During playback the right- and left-hand side signals have
to be fed to the right- and left-hand side loudspeakers, respectively. The listener
sitting in front of the speakers can distinguish the sounds of the different instruments
as coming from their proper positions when the music was recorded. When this
technique is applied to the cinema and television, a car approaching from one side of
the screen can be heard in the proper position presented in the picture and the sound
appears to move with the picture as the car moves across the screen. In a normal
recording process, two separate channels are required for a stereophonic system. So
it is expected that a stereophonic FM system will require twice the bandwidth of a
monophonic channel. In fact, it requires more bandwidth because the system had to
be designed in such a way that listeners who own a monophonic FM receiver can
tune in the stereophonic transmission and receive the same performance as they
would if the original program had been monophonic.
System Design
The system diagram shown in Figure 4.24 is used to pre-process the modulating
signal before it goes to frequency modulate a sub-carrier using the Armstrong
technique discussed in Section 4.4.
Figure 4.24. A block diagram showing the stages of processing of a stereophonic modulating
signal for the Armstrong system FM transmitter.
Figure 4.25. The spectrum of the composite modulating signal used in the stereophonic FM
The left-hand side signal LðtÞ and the right signal RðtÞ are fed to an adder and a
subtractor (the signal is first inverted and then added) to produce ½LðtÞ þ RðtÞ and
½LðtÞ RðtÞ, respectively. The ½LðtÞ þ RðtÞ signal is fed into one of the inputs of a
three-input adder as shown in Figure 4.24. The ½LðtÞ RðtÞ signal is one of the two
inputs to the balanced modulator. The other input is a 38 kHz signal originally
generated at 19 kHz ( pilot frequency) and then doubled. The output of the balanced
modulator is then the product
½LðtÞ RðtÞA cos 2op t
where 2op is the angular frequency of the doubled pilot oscillator frequency and A
its amplitude. The other two inputs of the three-input adder are the pilot signal and
the output of the balanced modulator. The output of the adder is then
M ðtÞ ¼ ½LðtÞ þ RðtÞ þ ½LðtÞ RðtÞA cos 2op t þ B cos op t
where B is the amplitude of the pilot oscillator. The design of the circuits in all the
component blocks shown in Figure 4.24 have been discussed in this and earlier
chapters. The baseband spectrum of the stereophonic signal is shown in Figure 4.25.
The processing of the stereophonic signal and its separation into right and left
signals will be discussed in Chapter 5.
1. Taub, H. and Schilling, D. N., Principles of Communication Systems, McGraw-Hill, New
York, 1971.
2. DeFrance, J. J., Communications Electronics Circuits, Holt, Rinehart and Winston, New
York, 1966.
3. Stremler, F. G., Introduction to Communication System, 2nd Ed., Addison-Wesley, Reading, MA, 1982.
4. Clarke, K. K. and Hess, D. T., Communication Circuits: Analysis and Design, AddisonWesley, Reading, MA, 1978.
5. Norwood, M. H. and Shatz, E., ‘‘Voltage Variable Capacitor Tuning: A Review’’, Proc.
IEEE, 56(5), 788–798, May 1968.
6. Moline, R. A. and Foxhall, G. F., ‘‘Ion-Implanted Hyperabrupt Junction Voltage Variable
Capacitors’’, IEEE Trans. Electron Devices, ED-19(2), 267–273, Feb., 1972.
The power output of an FM generator operating at 100=2p MHz is 10 W when
no modulation is applied. What would you expect to see on a spectrum
analyzer connected across its load? A sinusoidal modulating signal is applied
and increased slowly until the amplitude of the carrier goes to zero. What is the
average power in the remaining side frequencies?
(1) the average power in the side frequencies next to the carrier (first-order
side frequencies),
(2) the average power in the second-order side frequencies,
(3) the average power in the third-order side frequencies,
(4) the average power in the remaining side frequencies.
A direct FM generator uses a varactor diode for which
C ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 þ 2V
where V is the reverse-bias voltage across the p–n junction. Assuming that the
oscillator has a parallel LC tuned circuit such as shown in Figure 4.3:
(1) write an expression for the frequency of operation as a function of V,
(2) write another expression for the frequency of oscillation fo when the dc
voltage applied to the p–n junction is Vo .
(3) Assuming that v is a small time-varying voltage superimposed on Vo ,
calculate the slope (Hz=V) of the frequency–voltage characteristics of the
circuit about the oscillating frequency fo when fo ¼ 25 MHz and Vo ¼ 5 V.
A discriminator (frequency-to-voltage converter) is to be designed to operate at
10=2p MHz. You are provided with a center-tapped transformer whose
secondary self-inductance is L and series resistance r for each half. Choose
two capacitors of suitable values to tune the secondary for minimum second
harmonic distortion, given that L ¼ 0:2 mH and r ¼ 40 O (some iteration may
be necessary).
In an indirect FM transmitter, the carrier frequency fc is 250 kHz with the
maximum frequency deviation Df ¼ 50 Hz. This signal is to be frequency
multiplied in two stages (with a mixer between them) to bring the carrier
frequency up to 100 MHz. Stage one has a multiplication factor of 64. Choose
a suitable local oscillator frequency and the multiplication factor of the second
multiplier. Calculate the center frequencies and bandwidths as the signal
progresses through the frequency multiplier=mixer. Will this transmitter meet
the frequency deviation specifications for commercial FM stations?
In the circuit of Figure 4.18, C and R3 are interchanged. Under what conditions
will the circuit still behave like a 90 phase shifter?
If, in an FM transmitter operating at 100 MHz, a sinusoidal modulating signal
of 15 kHz causes the carrier frequency to deviate by 3 kHz, derive an
expression for the modulated wave, taking into consideration only the first
two pairs of side frequencies. Calculate the amplitudes of the side frequencies
relative to the unmodulated carrier.
In the circuit shown in Figure P4.1, a sinusoidal signal
vs ¼ 1:5 sin os t
is applied to terminals 1 and 10 of the transformer T1, where os ¼ 5 krad=s. A
second signal
vc ¼ 10 sin oc t
is applied to the terminals A and B, where oc ¼ 50 krad=s. Assuming that the
diodes have ideal characteristics:
(1) Sketch the waveform of the signal which appears across terminals 2 and 20
of the transformer T2 in relationship to vs and vc .
(2) Derive an expression for the output voltage vo in terms of vs and vc .
Figure P4.1.
(3) Calculate the amplitude of the component of vo at angular frequency noc
where n is an integer.
(4) Comment on your results from (3).
(5) Suggest an application for the circuit.
You may assume that for a square wave
f ðtÞ ¼
1 1
1 2P
sin not:
2 p n¼1 n
In amplitude modulation, the frequency of the carrier is kept constant while its
amplitude is changed in accordance with the amplitude of the modulating signal. In
frequency modulation, the amplitude of the carrier is kept constant and its frequency
is changed in accordance with the amplitude of the modulating signal. It is evident
that, if a circuit could be found which will convert changes in frequency to changes
in amplitude, the techniques used for detecting AM can be used for FM as well.
In Section, three frequency-to-amplitude conversion circuits were
discussed and their performance in terms of linearity and dynamic range were
examined. It therefore follows that the FM receiver must have the same basic form as
the AM receiver. The structure of the FM receiver is as shown in Figure 5.1.
The superheterodyne technique is used in FM for the same reasons it is used in
AM; it translates all incoming frequencies to a fixed intermediate frequency at which
the filtering process can be carried out effectively.
The antenna is responsible for capturing part of the electromagnetic energy
propagated by the transmitter. The basic rules of antenna design apply but, because
in commercial FM radio the frequency of the electromagnetic energy is between 88
and 108 MHz, it is practical to have antennas whose physical dimensions are within
tolerable limits.
The radio-frequency amplifier raises the power level to a point where it can be
used in a mixer or frequency changer to change the center frequency to a lower
frequency – the intermediate frequency (IF). The mixer in conjunction with the local
oscillator translate the incoming radio frequency to an intermediate frequency of
10.7 MHz. There is nothing special about an intermediate frequency of 10.7 MHz
except that it is a relatively low frequency at which the required values of Ls and Cs
are large enough to reduce the effects of circuit strays. It is at this fixed frequency
that filtering to remove the unwanted products of the mixing process and other
Figure 5.1. The block diagram of the domestic FM receiver showing frequency ranges and bandwidths.
interfering signals and noise takes place. The filtered signal then proceeds to the
amplitude limiter. The need for the limiter becomes evident when one recalls that the
FM signal is usually converted into an AM signal in the discriminator before it is
detected. This means that any variation in the amplitude of the FM signal will be
superimposed on the proper signal from the discriminator and hence will cause
distortion. The amplitude limiter very severely clips the signal to a constant
amplitude and also filters out the unwanted harmonics that are produced by the
limiter. The signal then proceeds to the frequency discriminator (frequency-toamplitude convertor) and onto the envelope detector. The audio-frequency amplifier
raises the output of the envelope detector to a level suitable for driving a
Although the structures of AM and FM receivers are similar, there are very
important differences which require different design and construction approaches.
These are the following:
1. The higher carrier frequencies (88–108 MHz) used in FM requires small
values of both L and C in the tuned circuits used. This means that stray
inductances and capacitances will constitute a larger percentage of the
designed value and hence have a much greater effect on all tuned circuits.
Although measures can be taken to incorporate the effects of fixed circuit
strays in the design, there are other changes in element values, due to such
factors as temperature and vibration, which can cause sufficient drift to
necessitate retuning of the receiver during a program. The local oscillator is
most vulnerable to stray elements, especially because it has to operate at a
frequency 10.7 MHz above the carrier frequency. To ensure the stability of the
oscillator, high circuit Q factors, negative temperature-coefficient capacitors
and automatic frequency control (AFC) are used. Some radio-frequency
amplifiers for FM front-ends use distributed parameter circuit components
such as coaxial and transmission lines.
2. With the intermediate frequency set at 10.7 MHz, the band from which image
interference can originate is from 109.4 to 129.4 MHz. This frequency band is
reserved for aeronautical radionavigation systems. It follows that one FM
station cannot cause image interference for another but the aeronautical
radionavigation systems can. As in AM, image interference can be reduced
by differentially amplifying the desired signal relative to the image signal. A
high Q factor tuned radio-frequency amplifier is used for this purpose.
3. The use of a high Q factor tuned amplifier in the radio-frequency stage
requires a very stable local oscillator frequency which will accurately track the
incoming radio frequency and produce a minimum variation from the selected
intermediate frequency. The local oscillator by itself is not capable of this but,
used in conjunction with the AFC and other stabilizing measures, it can
perform satisfactorily.
4. The ideal intermediate-frequency filter should have a bandpass characteristic
which is flat with infinitely steep sides. The flat top is required to avoid any
frequency dependent amplitude variation, the steep sides to eliminate interference from the unwanted products of the mixing process and from adjacent
channels. Two techniques, both involving a number of successive stages of
tuned circuits, are used to obtain an approximation to the ideal filter
characteristics. In the first, all the tuned circuits have the same resonant
frequency. This is known as synchronous tuning. In the second the resonant
frequencies are placed at different points in the passband. This is known as
stagger tuning.
The above qualitative as well as quantitative changes from the AM system discussed
in Chapter 3 make a separate consideration necessary.
An important point to remember is that an antenna is a reciprocal device, that is, it
can be used both for transmitting signals as well as for receiving them. An antenna
structure that produces a good ground wave radiation pattern will have a good
response to the same ground wave radiation when used in the receiving mode.
Commercial FM receivers commonly use two types of antennas: the vertical whip
antenna, most commonly used with automobile radios, and the dipole and folded
dipole antennas used with other types of portable and non-portable FM radios.
Assuming that the vertical whip antenna approximates a vertical grounded
antenna, a half-wavelength antenna for the middle of the FM band will be about
1.5 m long. Such an antenna can be conveniently mounted on a vehicle. The field
patterns of vertical grounded antennas are given in Figure 2.54. It can be seen that
when the height h is less than l=2, the response is limited to the ground wave.
Commercial FM stations are designed to operate within a local area and therefore
have antennas which ensure that most of the radiated power goes into the ground
wave. An FM receiver with an antenna whose height is equal to or less than l=2 will
have a good response.
The dipole and its variation, the folded dipole, are commonly used with nonportable FM receivers. They can be used in conjunction with directors and=or
reflectors to increase the gain of the antenna. Such an arrangement is called a Yagi–
Uda antenna.
Antenna design is beyond the scope of this text. However the interested reader is
encouraged to refer to any standard text on antennas.
Radio-Frequency Amplifier
The purpose of the radio-frequency amplifier is to boost the power of the incoming
signal relative to all the other signals picked up by the antenna to a level which can
be used in the frequency changer. A second function of the radio-frequency amplifier
is to act as a matched load to the antenna so that the antenna signal is not reflected at
the interface leading to the loss of efficiency.
The bandwidth of an FM signal in commercial radio was calculated to be
approximately 240 kHz in Section With a carrier frequency of about
100 MHz, the required Q factor is therefore about 400. Such a high value of Q
factor cannot normally be achieved in a simple tunable radio-frequency amplifier
and the practical approach is to use a lower Q circuit and to correct for this in the
intermediate-frequency stage that follows. An alternate technique uses a number of
stages in cascade separated by buffer amplifiers. For the superheterodyne system to
work, the local oscillator frequency must be set equal to the radio frequency plus the
intermediate frequency; for the commercial FM band, the standard intermediate
frequency is 10.7 MHz. The frequency of the local oscillator must be variable from
98.7 to 118.7 MHz. This is normally not a problem since the ratio of the high
frequency to the low frequency is only 1.2 : 1. The more important point is that the
center frequency of the radio-frequency amplifier and the frequency of the local
oscillator must maintain the difference of 10.7 MHz throughout the FM frequency
band. In the case of AM, the frequencies were low and some drift could be tolerated
without serious deterioration of the signal. In the FM system, the frequencies are
much higher and small percentage changes in one or both radio frequency and local
oscillator frequency can cause large changes in the intermediate frequency. To
overcome this problem, a system for automatic control of the frequency (AFC) of the
local oscillator is used. The basic operation is similar to that described in Section
Local Oscillator
The local oscillator can take any of the usual oscillator forms with a bipolar
transistor as the active element. It must produce enough power to drive the mixer.
The values of the inductors and capacitors have to be chosen to minimize the effects
of circuit strays. As mentioned earlier, the local oscillator incorporates an AFC
circuit for stable tracking with the (tunable) radio-frequency amplifier.
Frequency Changer
The basic frequency changer was discussed in Section 3.4.3. For this application the
dual-gate FET mixer has the advantage of low leakage of the local oscillator signal
to the antenna via the radio-frequency amplifier. Such a leakage and its subsequent
radiation can cause interference with other communication and radionavigation
Intermediate-Frequency Stage
The intermediate frequency for commercial FM radio is 10.7 MHz. The required
bandwidth of the filter is about 240 kHz centered at 10.7 MHz, giving a Q factor of
about 45. It is usual to realize the filter in two or more stagger-tuned stages with
suitable buffer amplifiers between them.
Amplitude Limiter
The radio-frequency amplifier, the mixer, and the intermediate-frequency amplifier,
in theory, should have a flat amplitude response in their pass bands. In practice, this
is not so. The result is that the signal emerging from the intermediate-frequency
amplifier has some variation of amplitude with respect to frequency. This is a form of
AM and it must be removed if distortion is to be avoided.
The amplitude limiter was discussed in Sections and It is worth
noting that sometimes the amplitude limiter is preceded by an automatic gain
control. This reduces the severity of the clipping action and hence the required signal
power and the spurious harmonics produced.
Frequency Discriminator
The purpose of the frequency discriminator is to convert relatively small changes of
frequency (in a very high-frequency signal) to relatively large changes in amplitude
with respect to time. The signal can then be demodulated using a simple envelope
detector which was discussed in Section 3.4.6. Two basic frequency discriminators
were discussed in Section and these serve to illustrate the concepts. In
practice, a number of more sophisticated discriminators are used. Some of these will
now be discussed. Foster–Seeley Discriminator. The Foster–Seeley discriminator [1,2]
is similar to the balanced slope discriminator shown in Figure 4.11. The major
difference is that it has two tuned circuits instead of three and both are tuned to the
same frequency. This is a major advantage when the receiver is being aligned. A
secondary advantage is that it has a larger linear range of operation than the slope
The basic circuit of the Foster–Seeley discriminator is shown in Figure 5.2.
Connected to the collector of the transistor, L1 and C1 are tuned to resonate at the
frequency fo (the intermediate frequency of the receiver). The inductance L1 is
mutually coupled to a center-tapped inductor L2. The center tap is connected by a
coupling capacitor Cc to the collector of the transistor. The inductor L2 and C2 are
tuned to resonate at the frequency fo. Two identical circuits consisting of a diode in
series with a parallel combination of a resistance and a capacitance (D3 –R3 –C3 and
D4 –R4 –C4 ) are connected across L2 to form a symmetrical circuit. A radio-frequency
choke (RFC – a high-valued inductance which can be considered to be an opencircuit at high frequency but a short-circuit at low frequency) connects the center tap
to the ‘‘neutral’’ node of the D3 –R3 –C3 and D4 –R4 –C4 circuits.
The circuit can be divided up into two parts by the line connecting X –X 0 . The
parts of the circuit to the left of X –X 0 operate at a high frequency fo with a relatively
small deviation Df . The circuit to the right of X –X 0 are two envelope detectors, as
Figure 5.2. The Foster–Seeley discriminator. The line X –X 0 divides the circuit into high (radio) and
low (audio) frequency.
discussed in Section 3.2. The high-frequency input voltages to the envelope
detectors are rectified by the diodes D3 and D4 and the time-constants R3 –C3 and
R4 –C4 are chosen to smooth out the half-wave pulses but follow any slow changes in
the envelope (amplitude) of the half-wave pulses. These slow changes represent low
(audio) frequency.
Before proceeding to the analysis of the circuit, it is in the interest of simplicity, to
make three assumptions:
(1) The impedance of the coupling capacitance Cc is small enough to be
considered as a short-circuit, at the frequency of operation.
(2) The impedance of the RFC is an open-circuit at the high frequency fo but a
short-circuit at the low (audio) frequency.
(3) The neutral node of the envelope discriminators can be considered to be
grounded since the secondary circuit, including the envelope discriminators,
is symmetric.
Now, all that needs to be demonstrated is that, when a signal of frequency fo Df is
applied to the circuit, the amplitude of the voltage appearing across the inputs of the
envelope detectors will vary proportionally to Df .
The tuned circuits L1 –C1 and L2 –C2 are both high Q factor circuits but their
mutual coupling coefficient, M, is low. This means that the secondary load coupled
into the primary circuit is negligible. The primary current is
I1 V1
The voltage induced in the secondary is
2V2 ¼ joMI1
where the sign depends on the relative directions of the primary and secondary
windings. Assuming the positive sign and substituting for I1
2V2 ¼
Since the secondary circuit is tuned to resonance, the secondary current is
I2 ¼
R2 L1
where R2 is the series resistance of the secondary circuit.
The voltage across the capacitor C2 is
jo0 C2
jo0 C2 R2 L1
2V2 ¼
Substituting for I2,
2V2 ¼
The secondary voltage applied to one envelope discriminator is given by
V2 ¼
V :
j2o0 C2 R2 L1 1
It is now clear that, at resonance, the primary voltage V1 is at right angles to the
secondary voltage V2 . The phasor diagram of the voltage applied to the inputs of the
envelope discriminators is as shown in Figure 5.3(a). This can be modified by
reversing the direction of one of the phasors representing V2 as shown in Figure
The discriminator input voltage phasors V3 and V4 are equal in magnitude and
since the outputs of the envelope discriminators are proportional to the magnitude of
the applied voltages, when the output is taken differentially, it is zero. This means
that the Foster–Seeley discriminator has zero volts output at the resonant frequency.
The impedance of the secondary tuned circuit, at any given frequency o, is
Z2 ¼ R2 þ j oL2 :
Figure 5.3. (a) The basic phasor diagram of the discriminator. (b) The phasor diagram when
the FM carrier is unmodulated.
But at resonance
C2 ¼
o20 L2
Eliminating C2 from Equation (5.2.8) gives
o2 L
Z2 ¼ R2 þ j oL2 0 2 :
If we now define the Q factor at resonance as
Q0 ¼
o0 L2
then Equation (5.2.10) can be written as
Z2 ¼ R2 þ jo0 L2
o o0
o0 o
Now consider relatively small changes of frequency about the resonant frequency oo
and define ‘‘fractional detuning’’ d as
o o0
o o0
o0 o
For a high Q factor circuit at a frequency near the resonance, the fractional detuning
d is much smaller than 1, therefore
o o0
o0 o
Substituting into Equation (5.2.12)
Z2 ¼ R2 ð1 þ j2QdÞ:
Figure 5.4. (a) The phasor diagram when signal frequency is lower than carrier frequency. (b)
The phasor diagram when signal frequency is higher than carrier frequency.
Replacing R2 in Equation (5.2.7) by Z2 gives
V2 ¼
V :
j2o0 C2 L1 R2 ð1 þ j2Q0 dÞ 1
It is evident that V1 and V2 are no longer at right angles to each other. The angle
between them depends on the magnitude and sign of d. When d is positive, the angle
between V1 and V2 is less than 90 , and when d is negative the angle is greater than
90 or vice versa. The phasor diagrams for positive and negative values of d are
shown in Figure 5.4(a) and (b), respectively.
It is clear from these that the magnitudes of the input voltages V3 and V4 are
unequal when d has any value other than zero. The characteristics of the Foster–
Seeley discriminator are shown in Figure 5.5.
A variation on the Foster–Seeley discriminator which combines the functions of
the amplitude limiter and frequency discriminator is called the ratio detector [3]. Its
performance however leaves much to be desired.
Figure 5.5. The amplitude–frequency characteristics of the Foster–Seeley discriminator.
Figure 5.6. The circuit diagram of the quadrature detector with an emitter follower to give a low
impedance output. Quadrature Detector. The basic circuit diagram of the quadrature
detector is shown in Figure 5.6. All biasing circuit components have been omitted
for clarity of its operation.
The circuit consists of a tuned amplifier Q1, with a very high Q factor collector
load. The input to the circuit is the output from the amplitude limiter which is a
frequency modulated square wave (i.e. a square wave of fixed frequency with
relatively small deviations in its zero crossings). Due to the high Q factor of the
tuned circuit, the output from the amplifier is a sinusoid at the fixed frequency. The
same square wave is fed to the base of Q4 which is a constant current source for the
differential pair Q2 –Q3 . Therefore current flows in the differential pair only when Q4
is switched on. The sinusoid applied to Q2 determines what proportion of the
constant current in Q4 flows through Q2 as opposed to Q3 . It can be seen from Figure
5.7(a) that, when the input signal is unmodulated, that is, when the phase difference
between the sinusoid and the square wave is fixed, the circuit can be adjusted so that
Q2 and Q3 conduct equal currents, giving a constant voltage across C3 . The timeconstant R3 C3 hold the base of Q5 at a dc value and the output remains constant.
When the input to the circuit is modulated, the sinusoid driving Q2 is no longer
coincident with the square wave and Q3 now conducts for a period proportional to
the ‘‘phase shift’’ between the two signals. This can be seen in Figure 5.7(b). The
voltage across C3 is a slowly varying direct current and the output is in fact the audio
frequency which was used to frequency-modulate the radio-frequency.
It should be noted that:
Figure 5.7. (a) The phase difference between the sinusoid and the square wave when the
carrier is unmodulated is such that currents of equal magnitude flow through Q2 and Q3 giving a
dc output. (b) When the carrier is modulated the relative phase between the sinusoid and the
square shifts and the currents in Q2 and Q3 are no longer equal; the dc changes its value – the
changing dc is the audio-frequency signal.
(1) the amplifier Q5 provides a low output impedance for the circuit,
(2) the structure of the circuit is suitable for realization in integrated circuit form,
(3) the time-constant R3 C3 is chosen to ‘‘follow’’ the changes in the amplitude of
the audio-frequency signal,
(4) when the two signals are 90 out of phase (in quadrature), Q2 and Q3 conduct
equal currents and this condition may be used as a datum. Phase-Locked Loop FM Detector. The phase-locked loop [4] FM
detector is the most complex of FM detectors but it has the advantage that it can be
realized in integrated circuit form where complexity is not necessarily a disadvantage. The basic system is as shown in Figure 5.8.
It consists of a phase detector that generates an output signal which is proportional to the difference between the phases of the two input signals (‘‘error’’ signal).
The output signal is amplified and low-pass filtered and used to control a voltagecontrolled oscillator (VCO) which usually operates at a higher frequency than the
input signal. The output of the oscillator is divided by a suitable factor N to bring it
to the same frequency as the input signal. This is the second input to the phase
Figure 5.8. The block diagram of the phase-locked loop FM detector. Note that the relatively
slow-varying dc required to keep the loop in lock is the audio-frequency signal.
The error signal fed to the VCO causes it to change frequency so that fd moves
closer to fr . When the two frequencies are close to each other, the system locks, that
is, the two frequencies become equal and their phase difference is zero. The control
voltage from the low-pass filter is then dc. When the incoming signal changes its
frequency, and hence phase, the control voltage will change its value to keep the
system in lock. The excursions of the control voltage is, in fact, the required
demodulated output.
The phase-locked loop FM detector has the advantage of having no LC tuned
circuits. In its integrated-circuit form, it requires a number of external resistors and
capacitors for its proper operation. This information is usually provided by the
The baseband frequency spectrum of the stereophonic FM signal was described in
Section 4.5.2 and shown in Figure 4.25. The use of the signal to frequency modulate
a suitable carrier remains essentially the same as in the case of monophonic
transmission except for the fact that the stereophonic spectrum has a larger
bandwidth – more than three times larger. At the receiver, the signal is demodulated
and the baseband information is recovered. Figure 5.9 shows a scheme for separating
the left, LðtÞ, and right, RðtÞ, information and passing them on to their respective
The first bandpass filter has a passband 23–53 kHz and it is used to separate the
double-sideband-suppressed carrier (DSB-SC) signal which contains the left-minusright signal, ½LðtÞ RðtÞ. The second bandpass filter is a narrow-band filter centered
on 19 kHz and it separates the pilot carrier. The third filter is a low-pass filter with a
Figure 5.9. The block diagram of the stereophonic FM receiver showing the ideal characteristics of
the various filters required and the access point for listeners with monophonic receivers.
cut-off frequency of 15 kHz and it separates the left-plus-right, ½LðtÞ þ RðtÞ (i.e. the
monophonic) signal from the other two.
The DSB-SC signal cannot be demodulated unless the carrier is reinstated.
Because the DSB-SC signal was obtained, in the transmitter, from the pilot oscillator
followed by a frequency doubler, the process is repeated in the receiver to obtain the
carrier at 38 kHz. The carrier is then fed into the synchronous demodulator to
produce a baseband signal containing the ½LðtÞ RðtÞ information and, after an
amplification factor of two, it is combined with the ½LðtÞ þ RðtÞ signal in the adder
and subtractor, respectively, to produce LðtÞ and RðtÞ.
Synchronous Demodulation
Synchronous demodulation of a DSB-SC signal can be achieved simply by multiplying the DSB-SC signal ½LðtÞ RðtÞ cos 2op t by the synchronized signal
cos 2op t. The result is
vðtÞ ¼ ½LðtÞ RðtÞ cos2 2op t
vðtÞ ¼
2 ½LðtÞ
RðtÞð1 þ cos 4op tÞ:
The required baseband signal 12 ½LðtÞ RðtÞ is easily separated from the signal at
4op .
Stereophonic Receiver Circuit
Filter design is beyond the scope of this book. A list of references on filter design
can be found at the end of Chapter 3. The design of frequency multipliers was
discussed in Section 2.5. The design of the synchronous demodulator (two-input
analog multiplier) was discussed in Section The adder=subtractor was the
subject of Section
1. Foster, D. E. and Seeley, S. W., ‘‘Automatic Tuning Simplified Circuits and Design
Practice’’, Proc. IRE., 25, 289, 1937.
2. Seeley, S. W., Radio Electronics, McGraw-Hill, New York, 1956.
3. Seeley, S. W. and Avins, J., ‘‘The Ratio Detector’’, RCA Review, 8, 201, 1947.
4. Best, R. E., Phase-Locked Loops, McGraw-Hill, New York, 1984.
5. DeFrance, J. J., Communication Electronics Circuits, 2nd Ed., Rinehart Press, San
Francisco, 1972.
6. Stark, H. and Tuteur, F. B., Modern Electrical Communications, Prentice-Hall, Englewood
Cliffs, NJ, 1979.
A bipolar transistor operating at a frequency of 100 MHz has the following Y
parameters: y11 ¼ 9:84 þ j9:1; y12 ¼ 0:01 j0:68; y21 ¼ 60:7 j95:8 and
y22 ¼ 0:60 þ j1:8, measured in mSi in the common emitter configuration
with 0.6 V peak-to-peak appearing across the base emitter junction. Using
this device, design an oscillator to supply maximum power to a suitable
conductive load. What modifications would you make to the oscillator to
obtain a local oscillator for an FM receiver operating in the range 88–
108 MHz?
An FM receiver is tuned to a transmitter at 100 MHz. The modulation is
sinusoidal at a frequency of 15 kHz. The intermediate frequency is 10.7 MHz
and the Q factor of the intermediate-frequency amplifier is 100. Estimate the
maximum allowable drift in the local oscillator frequency, assuming that the
J3 ðmf Þ terms must lie within the 3 dB bandwidth of the intermediatefrequency amplifier.
Using the characteristics of the FET given in Figure P3.7, design an amplifier
tuned to 100 MHz with a voltage gain of 15. The dc supply voltage is 15 V and
you may allow 2 mA dc to flow through the drain. The 3 dB bandwidth
should be no larger than 250 kHz when the amplifier is connected to a second
stage whose input impedance is 250 O resistive.
Give two reasons why a double-sideband-suppressed carrier (DSB-SC) modulation system is not commonly used in communication systems. Name one
communication system in which it is used, giving the relevant details.
An incoming signal in a communication receiver is given by
vðtÞ ¼ 12 b½cosðoc om Þt cosðoc þ om Þt
where oc is the carrier angular frequency and om is the modulating signal
angular frequency.
What do you need to demodulate the signal and how would you carry out
the demodulation? Support your answer with suitable diagrams and derived
formulae. Can you think of a second demodulation technique for the signal?
Describe with the aid of a block diagram the superheterodyne technique for the
reception of commercial FM signals. You have an AM=FM receiver and you
wish to make changes to the configuration of the receiver by introducing a
second mixer to translate the intermediate frequency of the FM system from
the standard 10.7 MHz to the standard AM system frequency of 455 kHz.
Redesign the block diagram, giving all the necessary details to achieve your
wish. What would be the advantages and disadvantages of your new receiver?
The dual-gate FET (whose characteristics are given in Figure P5.1) is to be
used in the design of a frequency changer in a commercial FM receiver whose
intermediate frequency is 10.7 MHz. The output of the radio-frequency
amplifier is applied to gate 1 and it is 0.4 V peak-to-peak at a frequency of
100 MHz. The signal from the local oscillator applied to gate 2 is a square
wave whose voltage varies from þ4 V to a negative value sufficient to cut off
the drain current. Design a suitable mixer with the following specifications:
(1) direct current power supply, 12 V,
(2) average drain current, 4 mA,
(3) 3 dB bandwidth, 240 kHz.
Choose all the other values and in every case justify your choice. Calculate the
ratio of the intermediate-frequency power output to the radio-frequency power
Figure P5-1.
The purpose of the discriminator in an FM receiver is to convert variation in
frequency to variation in amplitude. Describe the operation of the Foster–
Seeley discriminator and demonstrate its frequency-to-amplitude conversion
ability. What measures should be taken prior to the signal entering the
discriminator to reduce distortion and why?
The transmission of video images depends on a scanning device that can break up
the image into a grid and measure the brightness of each element of the grid. This
information can be sent serially or in parallel to a distant point and used to reproduce
the image. It is evident that the smaller the size of the grid element, the better the
definition of the image.
One of the simplest devices which can measure the brightness of light is the
phototube. It consists of a cathode which is coated with a material which gives off
electrons when light is shone on it and an anode which can collect the emitted
electrons when a suitable voltage is applied to it. The cathode and anode are
enclosed in an evacuated glass envelope. The number of electrons emitted by the
cathode is proportional to the intensity of the light impinging on it. Assuming
complete collection of the electrons, the current in the resistor R shown in Figure 6.1
will be proportional to the light intensity and so will the voltage across R.
A primitive video signal can be generated by using a 3 3 matrix made up of
phototubes as shown in Figure 6.2.
For simplicity we assume that the tree is black and its background is white. A
suitable lens focuses the image of the tree onto the matrix of phototubes. It is clear
that the voltage output from phototubes (1,1), (1,3), (3,1) and (3,3) will be high; all
others will be low. The voltages so obtained can be transmitted and used to control
the brightness of a corresponding 3 3 matrix of lights at a distant point giving a
vague idea of what the tree looks like! The picture detail can be improved by
increasing the number of elements in the matrix so that each element corresponds to
the smallest area possible. The assumption of a black tree on a white background is
no longer necessary since, with increasing detail, different shades of grey can be
The information may be sent along individual wires linking the phototube to the
light matrix (parallel transmission) but this would be very expensive and impractical
Figure 6.1. The phototube with its added circuitry to convert light intensity to voltage.
for any system other than the simple one described here. A better system would be
one in which the voltage from each phototube is scanned in some given order and the
voltage and position of each phototube are sent on a single wire to the receiving end
for reconstruction (serial transmission). The price to be paid for reducing the number
of wires is the increased complexity introduced by the scanner and a system for
coding and decoding the voltage and position information at the transmitter and
receiver, respectively.
Figure 6.3 shows the basic components of a television transmitter. A system of
lenses focus the image onto a camera tube which collects and codes the information
about the brightness and position of each element of the matrix forming the picture
by scanning the matrix. A pulse generator supplies pulses to the camera to control
the scanning process. The output from the camera goes to a video amplifier for
amplification and the addition of extra pulses to be used at the receiver for decoding
Figure 6.2. Generation of a primitive video signal using a 3 3 matrix of phototubes.
Figure 6.3. A block diagram of the television transmitter.
A microphone picks up the sound associated with the picture and after
amplification the signal is fed to the audio terminal of a frequency modulator. The
carrier signal supplied to the modulator is a 4.5 MHz signal, generated by a crystalcontrolled oscillator at a lower frequency and multiplied by an appropriate factor.
The FM signal carrying the audio information is added to the video signal. The
output of the video amplifier consisting of the video signal, receiver control pulses,
and the frequency modulated signal is fed to the amplitude modulator. The carrier of
the amplitude modulator is supplied by a second crystal oscillator and associated
multiplier which produce a signal of frequency which lies within the band 54–
88 MHz (VHF). The radiofrequency power amplifier boosts the power to the legally
determined value and the vestigial sideband filter removes most of the lower
sideband signal before it goes to the antenna for radiation.
Camera Tube Iconoscope. The first practical video camera tube invented by the
American scientist Vladimir Zworykin [4,5] is best viewed as a progression from
the primitive phototube arrangement discussed earlier. In Zworykin’s iconoscope he
Figure 6.4. A cross-sectional view of the iconoscope. The two parts of the anode are cylinders
formed on the inner wall of the tube and neck. Reprinted with permission from H. Pender and
K. McIlwain (Eds) Electrical Engineers Handbook, 4th Ed., Wiley, 1967, pp. 15–21.
replaced the matrix of phototubes with what he called the mosaic. This was made up
of a very large number of droplets of photosensitive material on one side of a sheet
of mica. The other side of the mica sheet was covered with a very thin layer of
graphite – the signal plate. A cross-section of the iconoscope is shown in Figure 6.4.
The mosaic and the signal plate constitute a large number of tiny capacitors which
share one common plate. When the image is projected onto the mosaic, each
individual droplet of photosensitive material emits electrons proportional to the light
intensity. These electrons are collected by an anode placed close to the mosaic. The
mosaic is now a picture ‘‘painted with electric charge’’. The transformation of the
charge into a voltage output is carried out by the electron gun and the associated
The electron gun produces a very narrow beam of electrons focussed on the
mosaic. The arrival of these new electrons have the effect of ‘‘discharging’’ the tiny
capacitor on which they fall. Since the number of electrons required to discharge the
capacitor is proportional to the charge induced by the light intensity, the electron gun
current is a function of the charge present and hence of the light intensity. The
electron gun current is then proportional to the voltage which appears across the
resistor R.
In order to transform the picture into a video signal, it is necessary to make the
beam of electrons sweep across the mosaic in a series of orderly lines. The electron
gun has two deflection systems for scanning the picture. The first moves the electron
beam at a constant speed along a horizontal straight line and returns it very quickly
to the start ready for the next sweep (horizontal trace). The second system controls
the vertical position of the beam and ensures that each line is swept before returning
the beam to the top of the picture ready for the next frame (vertical trace).
Two major disadvantages of the iconoscope are the high light intensity required to
obtain acceptable quality images and the production of secondary electrons from the
photosensitive material during the scanning process. The secondary electrons cause
noise (false information) in the video signal. In due course, the iconoscope was
replaced by the image orthicon, another invention of Zworykin. The Image Orthicon. In the image orthicon [6] the two functions of the
mosaic in the iconoscope, which are (a) the production of the image in terms of
charge and (b) the target for the electron beam scan, are separated. Figure 6.5 shows
a cross-section through the image orthicon.
The camera lens focusses the image onto the photocathode. Every small segment
of the photocathode emits electrons proportional to the amount of light falling on it.
The electrons are attracted to the target by the positive voltage applied to it. The
focus coil creates a magnetic field which ensures that the electrons travel to the target
in straight parallel paths. The target is a very thin sheet of glass which has a high
conductivity through its thickness but low conductivity across points on its surface.
The electrons striking the target dislodge secondary electrons from the photocathode
side of it and are immediately captured by the very fine wire mesh, called the target
screen, which has a relatively low positive voltage on it. This leaves a positive charge
on the photocathode side of the target. Due to the high conductivity through the
thickness of the target an identical pattern of positive charge is produced on the other
side of it.
The electron gun positioned at the other end of the evacuated tube produces a
narrow stream of electrons which are accelerated towards the target by the high
positive voltage on the accelerator anode. The accelerator anode is a graphite
coating on the inside of the neck of the tube. If the electrons were allowed to strike
the target without any control of their speed, it is clear that they will produce
secondary electrons. To control the speed of the electrons on arrival at the target, a
second ring of graphite coating, called the decelerator grid, is provided. By adjusting
the voltages on the accelerator anode and the decelerator grid, the speed of the
electron stream at the target can be controlled to ensure that no secondary electrons
are emitted. The electron stream striking an elemental area of the target uses up some
of the electrons to neutralize the positive charge left there by the image. The
remaining electrons are attracted in a backward direction by the accelerating anode
but they take a different path. By placing an electron collector in the appropriate
position in the neck of the tube, the returning electrons can be collected and
amplified by an electron multiplier. The output of the electron multiplier is converted
into a voltage inversely proportional to the amount of light impinging on the
elemental area of the photocathode. By using a scanning system in conjunction with
the electron gun (deflection coil), the complete image can be produced at the output
in the form of a varying voltage which is a function of the brightness of all the
elements of the image projected on the photocathode.
Figure 6.5. A cross-sectional view of the image orthicon. Reprinted with permission from V. K.
Zworykin and G. Morton, Television, 2nd Ed., RCA, 1954, p. 367.
167 Vidicon. The vidicon is not as sensitive as the image orthicon but it is
generally smaller in size and therefore well adapted to field applications such as
surveillance, where high resolution is not critical. Its principle of operation is
different from both the image orthicon and the earlier iconoscope since it relies on a
change of photoconductivity as a function of light intensity. An electron scanning
system is used to extract the video information.
Scanning System
In Section 6.3.1, the role of the electron beam scanning system in the production of
the video signal was discussed. A series of pulses are generated which control the
initiation of the horizontal sweep of the electron beam across the target (horizontal
trace) from the left-hand side to the right. A blanking pulse is used to cut off the
beam while it is returned to the left-hand side ready for the next sweep (harmonic
retrace). During this period, the vertical trace circuit moves the beam down just the
right distance for the second line to be swept. When the whole frame has been
scanned, a reset pulse returns the beam to the top left-hand corner ready to repeat the
process. The beam is blanked during the reset. All the control pulses are added to the
video signal and used at the receiver to synchronize the receiver to the transmitter.
All the pulses are derived from the 60 Hz power supply or from a 31.5 kHz crystal
controlled oscillator and divided down to the appropriate frequency.
Black-and-white television in North America uses 525 horizontal lines of scan to
cover the image. If the lines were scanned sequentially, the persistence of the
phosphor on the first line would have faded before the last line would have been
scanned. This would lead to an annoying flicker on the picture tube, especially at
high levels of brightness. The scan is therefore interlaced, that is, all odd numbered
lines are scanned first from 1 to 525 and then the beam is returned to line 2 and all
even numbered lines are then scanned. So each frame has 525=2 lines and they are
scanned 60 times per second. This means that there are 262:5 60 lines scanned per
second. The horizontal scan therefore operates at 15.75 kHz. It must be pointed out
that not all the 525 lines are used for picture production; some are used for vertical
retrace and other controls as well as equalization and synchronizing purposes.
The period for the horizontal scan is 63 ms and it is divided up into 56 ms for the
trace and 7 ms for the retrace. The beam deflection system is therefore required to
move the beam from the left of the picture to the right in 56 ms and return it in 7 ms.
The mechanism for deflecting the electron beam relies on the fact that, when a
current-carrying conductor is placed in a magnetic field, the conductor experiences a
force mutually orthogonal to the field and the direction of the current. Motion will
occur if the conductor is free to move. In the camera tube, the electron gun and its
associated components produce a stream of electrons. Each of these electrons carries
a charge and a moving charge is a current. The electron beam therefore experiences a
force on it and, since it is free to move, it will be deflected from its original path.
Figure 6.6(b) illustrates the situation.
Consider the current in the conductor (the electron beam) to be flowing into the
plane of the page. The ‘‘right-hand rule’’ gives the direction of the magnetic flux Bi
Figure 6.6. (a) The magnetic field, Bi due to a current, i flowing perpendicular to the page
superimposed on a uniform magnetic field, Bf , with no interaction. (b) The combined effect of the
two magnetic fields generates a force mutually at right angles to the conductor and magnetic
associated with the current in the conductor as concentric circles in a clockwise
direction. Assuming the conductor is placed in a magnetic flux Bf which is in the
vertical upward direction, the two fluxes will interact to produce the distorted field
shown in Figure 6.6(b). The flux on the left of the conductor is strengthened while
that on the right is weakened; the conductor therefore moves to the right. When
trying to determine the direction of electron deflection, it is necessary to remember
that electron flow is opposite to current flow.
The deflection of the electron beam therefore depends on designing a circuit to
supply the appropriate current to the horizontal deflection coil so that it produces a
magnetic field which is a linear ramp with respect to time. The vertical deflection
system also has to be supplied with a current which produces a linear ramp magnetic
field and its operation has to be synchronized to the horizontal trace so that it repeats
the process after 262.5 cycles.
The video signal, the control pulses, and the output of the FM modulator are
combined in the video amplifier. A typical composite video signal is shown in Figure
6.7. Note the following:
(1) Only horizontal synchronizing pulses are shown ( 5 ms); vertical synchronization pulses ( 190 ms) are distinguished from the horizontal ones by the
difference in their width.
(2) The blanking pulse starts before and ends after the horizontal synchronization
(3) The video signal contains the radio-frequency carrier of 4.5 MHz frequency
modulated by the audio signal.
Figure 6.7. A typical composite video signal with blanking and horizontal synchronization
The spectrum of the composite signal at the output of the video amplifier is
shown in Figure 6.8.
Audio Frequency and FM Circuits
Various audio frequency amplifiers were discussed in Section 2.7. The design of the
crystal oscillator was discussed in Section 2.4.7 and the frequency multiplier in
Section 2.5. Design details for the FM modulator can be found in Section 4.4.
Figure 6.8. The spectrum of the composite video showing the FM carrier used for the
transmission of the audio signal.
Video Amplifier Calculation of Bandwidth. Ideally, a video signal consists of frequencies from zero (dc) to some high frequency. The dc response is required when large
areas of black, white or other intermediate shades have to be transmitted. The limit
of the high-frequency response is determined by the resolution required for a single
vertical black line on a white background or vice versa. The minimum visible
horizontal line has a height (width) equal to the width of the horizontal trace. To
keep the same resolution for the horizontal as for the vertical line, the video signal
must go from white to black in
56=525 ms ¼ 0:11 ms:
The lowest frequency that can change from peak positive to peak negative is a
sinusoid as depicted in Figure 6.9. Note that the period of the sinusoid is 2 0:11 ms
or 0.22 ms.
It has been determined from psychological experiments that a picture that has a
ratio of horizontal-to-vertical dimension less than 4 : 3 is not pleasing to look at. This
ratio has been adopted as standard for all video equipment. It is referred to as the
aspect ratio. Taking the aspect ratio into account, one cycle of the sinusoid shown in
Figure 6.9 must take 0:22 3=4 ms, (i.e., the period T ¼ 0:16 ms). The lowest
frequency that must be present in the video signal for it to be able to reproduce the
vertical black line on a white background with the same resolution as a horizontal
black line on a white background is given by
f ¼
¼ 6:25 MHz:
This means that all the video circuits in the system must have bandwidths equal to or
greater than 6.25 MHz. Note that the bandwidths of the sub-circuits must be greater
than 6.25 MHz because cascading reduces the system bandwidth compared to that of
the individual sub-circuit.
Figure 6.9. The sine wave shown represents the highest frequency required to follow the
transition from black to white level in 0.11 ms.
Figure 6.10. The equivalent circuit of a common-source FET. Video Amplifier Design. Because the video signal contains frequencies from zero to some high frequency, a video amplifier must be designed to have a
flat response from dc to the high frequency in question. Ideally, all circuits in the
video chain must be directly coupled. In practice, an amplifier with a bandwidth
from approximately 30 Hz to approximately 4.0 MHz is considered to be a video
Consider a FET video amplifier [1] in which the FET is modelled in the commonsource mode as shown in Figure 6.10.
Since rd RL and jXCds j rd, both rd and Cds can be neglected as shown in
Figure 6.11.
From Figure 6.11
I1 ¼ joCgs Vgs
I2 ¼ joCgd ðVgs Vo Þ
I3 ¼ I1 þ I2 ¼ joVgs ½Cgs þ ð1 Vo
ÞC :
Vgs gd
Figure 6.11. A simplified equivalent circuit of the FET with rd and Cds excluded.
Now I1 I2, therefore
Vo ¼ gm Vgs RL :
Substituting into Equation (6.1.4),
I3 ¼ joVgs ½Cgs þ ð1 þ gm RL ÞCgd Vgs
jo½Cgs þ ð1 þ gm RL ÞCgd I3
The FET can now be the represented by the equivalent circuit shown in Figure 6.12
in which
Cg ¼ ½Cgs þ ð1 þ gm RL ÞCgd :
Consider the two-stage video amplifier shown in Figure 6.13.
Figure 6.12. A modified form of the FET shown in Figure 6.11.
Figure 6.13. A two-stage video FET amplifier showing Cg for both devices.
Figure 6.14. The amplifier of Figure 6.13 with the inductor L connected in series with RD .
The resistor R1 is used to bias the gate of Q1 and it is usually very large (MOs).
The time-constant Cg R1 is therefore very large compared to Cg RD . To a first
approximation, the time-constant Cg RD determines the high-frequency response of
the amplifier. The 3 dB frequency is
o2 1
The bandwidth o2 can be increased by connecting an inductance in series with the
drain resistance RD as shown in Figure 6.14.
The equivalent circuit is shown in Figure 6.15.
The load is
ZL ¼
ðRD þ sLÞ1 þ sCg
Figure 6.15. A simplified equivalent circuit of the first stage of the FET amplifier with the Cg of
the second stage connected across the output.
s ¼ s þ jo
Vo ¼ gm Vgs ZL :
The gain is
¼ gm ZL
sþ D
Cg 2
s þs
þ o2o
o2o ¼
The gain function has two poles at
s1 ; s2 ¼ D 2L
o2o :
o2o ¼
the gain function poles are conjugate on the imaginary axis of the s plane and these
increase the gain at the resonant frequency oo.
The critical value of L is
Lcrit ¼
R2D Cg
The response is then as shown in Figure 6.16.
Figure 6.16. The frequency response of the amplifier after the addition of the inductor. Note
that the resonant effect increases the gain at high frequency, thereby increasing the bandwidth.
The shape of the gain response can be controlled by varying the Q factor of the
circuit. Let
o2 L
¼ 2
RD Cg 4Lcrit
Substituting q and s ¼ jo into Equation (6.1.14) gives
jAj ¼ A u
4 :
1 þ ð1 2qÞ
The gain response can, in general, be represented by the ratio of two polynomials in
o2 :
HðoÞ ¼
1 þ a1 o2 þ a2 o4 þ 1 þ b1 o2 þ b2 o4 þ ð6:3:21Þ
which gives a flat gain–frequency response when a1 ¼ b1 and a2 ¼ b2 and so on.
Equating the coefficients of ðo=o2 Þ2 in Equation (6.3.21) gives
q ¼ 2 1 ¼ 0:414
Figure 6.17. The frequency response of the amplifier for various values of q.
The negative value has no physical meaning and it is neglected. Substituting into
Equation (6.3.19), we get
q ¼ 0:414 ¼
4Lcrit R2D Cg
L ¼ 1:656Lcrit ¼ 0:414R2D Cg :
The frequency response for three values of q are shown in Figure 6.17.
The new 3 dB frequency is
o02 ¼ 1:72 o2 :
Example 6.3.1 Video Amplifier. A video amplifier uses FETs and has two stages
as shown in Figure 6.18.
Figure 6.18. The video amplifier used in Example 6.3.1.
The FETs have the following parameters:
gm ¼ 0:004 S;
Cgs ¼ 1:0 pF;
Cgd ¼ 1:5 pF:
Assuming that only the Miller effect capacitance affects the high-frequency
response, calculate the 3 dB frequency of the amplifier when the gain is 15.
Calculate the value of an inductance which, when connected in series with RD , will
extend the 3 dB cut-off frequency as high as possible while keeping the gain
response flat. Calculate the new cut-off frequency.
Solution. Neglecting all capacitance effects, the equivalent circuit of the amplifier is
shown in Figure 6.19, where
Vo ¼ gm Vgs RD
¼ gm RD ¼ 15:
RD ¼
¼ 3750 O:
D ¼ 15:
Vgs Rs
Rs ¼ 250 O:
and the voltage gain is
The voltage gain
Figure 6.19. The equivalent circuit of stage 1 of the amplifier when all capacitances have been
Figure 6.20. The equivalent circuit of stage 1 of the amplifier with the Cg of stage 2 included.
Rg can be made arbitrarily large since no current flows in it. The Miller effect
capacitance of Q1 will have negligible effect on the frequency response, especially if
the internal resistance of the driving source is small. The Miller effect capacitance of
Q2 will appear across RD as shown in Figure 6.20 where
Cg ¼ Cgs þ ð1 þ gm RD ÞCgd ¼ 1:0 þ ð1 þ 15Þ1:5 ¼ 25 pF:
The 3 dB frequency o2 is determined by
o2 t ¼ 1
t ¼ RD Cg
f2 ¼
¼ 1:67 MHz:
2pRD Cg
When the inductance L is connected in series with RD and L has the critical value,
L ¼ 4qLcrit ¼ qR2D Cg :
L ¼ 145:5 mH:
When q ¼ 0:414,
The new 3 dB cut-off frequency is
f20 ¼ 1:72f2 ¼ 2:87 MHz:
Radio-Frequency Circuits
The composite video signal from the output of the video amplifier is used to
amplitude modulate the radio-frequency carrier obtained at a lower frequency from a
crystal oscillator and multiplied by a suitable factor by a frequency multiplier.
Amplitude modulators are discussed in Section 2.6, frequency multipliers in Section
2.5, and crystal-controlled oscillators in Sections 2.3 and 2.4.
Vestigial Sideband Filter
From Figure 6.8, it can be seen that the composite signal occupies a bandwidth of
approximately 0–4.5 MHz. If both sidebands were transmitted, it would require a
bandwidth of just over 9.0 MHz. Not only would that consume scarce bandwidth but
it is also unnecessary because the information in the upper and lower sidebands are
the same. Single-sideband (SSB) transmission could be used but the filter needed to
remove one of the sidebands is fairly sophisticated and the demodulation equipment
is complex and difficult to maintain. A compromise is to remove most of the lower
sideband using a bandpass filter. As shown in Section 7.2.4, a simple envelope
detector is adequate for the demodulation of such an AM signal if the index of
modulation is low. The frequency spectrum at the output of the vestigial sideband
filter is shown in Figure 6.21.
Television broadcast frequencies are either in the very-high-frequency (VHF) band
which is from 30 to 300 MHz or in the ultra-high-frequency (UHF) band which is
from 300 to 3000 MHz. At these frequencies, antennas have highly directional
properties. To get the circular radiation pattern in the horizontal plane normally used
for broadcasting television signals, several arrangements of antenna arrays can be
Figure 6.21. The spectrum of the video signal after amplitude modulation and vestigial
sideband filtering. Note that the AM carrier signal occupies the zero frequency position after
used. The most popular is the turnstile array. The basic principle is that, when two or
more radiating elements such as dipoles are placed in close proximity to each other,
they interact to produce a radiation pattern that is the vector addition of the
individual elements. By varying the relative physical positions of the elements
and the phase angle of the signal, it is possible to use the interactive properties to
create a radiation pattern which is approximately circular in the horizontal plane.
Color Television
The transmission of video signals in color is a subject which can take up several
volumes. However, because color television is so common, a simplified explanation
of how it works is now offered.
The first step is to discuss some of the properties of color and the results of
mixing them. There are three primary colors: red, blue and green, and by using
appropriate proportions of these, all other colors perceived by the human eye can be
obtained. A simple framework which makes it easy to understand the properties of
color, the color triangle [2,3], is shown in Figure 6.22.
The primary colors occupy the apices of the triangle. A mixture of equal
proportions of red and blue produces the color magenta; that of red and green
produces yellow, and finally blue and green produce cyan. These fit into the scheme
as shown. The center of the triangle represents white since equal proportions of red,
blue, and green produces white. Two pieces of information are required in order to
code a video signal in color. The first is brightness; the better technical term is
Figure 6.22. The color triangle with the I and Q axes.
luminance, which is a measure of the energy in the light. Luminance is the only
information required for monochromatic (black-and-white) television. The second is
chrominance, or simply color. Chrominance is best represented on the diagram by a
polar coordinate scheme whose origin is at the center (white) of the triangle. When
one moves along a straight line from the origin towards the red apex one sees at first
white light followed by a slowly increasing presence of red (pink). At the apex (pure
red light) the color red is said to have reached saturation. The distance from the
center of the triangle (the magnitude of the complex number) is then a measure of
the saturation of the particular color. The angle of the complex number, with respect
to some arbitrary reference line, is called the hue of the light. The combination of
chrominance and luminance will produce a cylindrical coordinate system in which
luminance will be on the axis of the cylinder. In this scheme, the coding of color
must have three coordinates: luminance, saturation, and hue.
The above scheme does not take into account the sensitivity of the human eye
across the color spectrum and it needs to be modified since the human eye is the final
arbiter (detector) of the video signals transmitted. The response of the human eye is
shown in Figure 6.23.
A mixture of equal proportions of the primary colors is therefore perceived by the
human eye as a greenish-yellow color. In fact, to obtain white light the mixture has
to contain 59% green, 30% red, and only 11% blue.
The simplest system for coding video information for television transmission is
simply to scan the picture for the three primary colors and process each one as if it
were monochromatic. Some multiplex system could then be worked out to
accommodate all three signals. Such a scheme was tried in the United States in
the late 1940s and early 1950s but was found to be incompatible with the black-andwhite system then in existence.
In 1953, a scheme proposed by the National Television System Committee
(NTSC) was accepted by the Federal Communications Commission (FCC) and it has
since been adopted as a standard by television broadcasting authorities in many
Figure 6.23. The relative sensitivity of the human eye to the wavelength of light.
countries. The NTSC scheme separates the signal into chrominance and luminance
(the chrominance information is further separated indirectly into I and Q components, which correspond approximately to hue and saturation). This will be
explained later. The actual system acquires the luminance information by taking
the weighted sum of the three primary colors to obtain white light as perceived by
the human eye as:
Lw ¼ 0:59Lg þ 0:30Lr þ 0:11Lb :
Lw then contains all the information for black-and-white picture transmission. For
color pictures, the three signals Lw ; ðLr Lw Þ and ðLb Lw Þ are sent. The third
possible signal ðLg Lw Þ is not used because it is the smallest of the three difference
signals and therefore more likely to be corrupted by noise. The two difference
signals ðLr Lw Þ and ðLb Lw Þ are quadrature multiplexed, a form of DSB-SC
modulation, with the color sub-carrier of frequency oo. The two difference signals
have the form:
ðLb Lw Þ cos oo t þ ðLr Lw Þ sin oo t:
Lw þ ðLb Lw Þ cos oo t þ ðLr Lw Þ sin oo t
The signal
is used to amplitude modulate a suitable carrier in the usual vestigial sideband
At the receiving end, the two difference signals are de-multiplexed by multiplying
them by the output of a local oscillator which is synchronized to the color sub-carrier
of frequency oo in the transmitter. The color sub-carrier signal is sent in ‘‘bursts’’
during the trailing portion of the horizontal blanking pulse called the back porch.
This is shown in Figure 6.24.
The demodulation system is shown in Figure 6.25.
The recovered signals Lw ; ðLb Lw Þ and ðLr Lw Þ are used to reconstruct the
missing difference signal ðLg Lw Þ as follows:
ðLg Lw Þ ¼ 0:30
ðLr Lw Þ ðL Lw Þ:
0:59 b
The four signals are used to control the appropriate grids on the picture tube to
reproduce the original scene.
In the above discussion, the bandwidth requirements of the three signals
Lw ; ðLb Lw Þ and ðLr Lw Þ were assumed to be the same. This is, in fact, not
true. The human eye has a higher resolution in black-and-white than it does in color.
It is adequate to present the uniform, large areas of the picture (low frequency) in
color; the fine details (high frequency) can be presented in black-and-white. The
luminance information Lw must have the whole bandwidth available. The difference
signal containing the chrominance information requires much less bandwidth.
Figure 6.24. A modified back porch of the blanking pulse used to transmit the color subcarrier.
Further advantage can be taken of the greater sensitivity of the human eye to
green-yellow light by de-emphasizing the green-yellow component by using a
smaller bandwidth for transmitting that part of the spectrum. This is called the Q
signal. The blues and the reds, to which the human eye is less sensitive, are given a
relatively larger bandwidth. This is called the I signal. The comparative spectra for
the luminance, Q and I signals are shown in Figure 6.26.
Because the Q signal occupies a total bandwidth of 1 MHz, and it fits within the
overall bandwidth limits of the signal, both sidebands are transmitted. The upper
sideband of the I signal lies outside the overall bandwidth and part of the upper
sideband is removed by a vestigial sideband filter.
Figure 6.25. A scheme for demultiplexing the color information.
Figure 6.26. The full spectrum of the color TV signal. Note that zero frequency corresponds to
the AM (video) carrier frequency.
1. Ryder, J. D., Electronic Fundamentals and Applications: Integrated and Discrete Systems,
5th Ed., Prentice-Hall, Englewood Cliffs, NJ, 1976.
2. Lambert, J. H., (a) Beschreibungen einer Farbenpyramide, Berlin, 1772. (b) ‘‘Mémoire sur
la Partie Photometrique de l’Art de Peinture’’, Preussiche Akademie der Wissenschaften,
Berlin; Histoire de l’Acad. Royal des Science et Belles Lettres, 24, 80–108, 1768.
3. Maxwell, J. C., ‘‘On the Theory of Three Primary Colours’’ in Niven, W. D. (Ed), The
Scientific Papers of James Clerk Maxwell, Cambridge University Press, Cambridge,
England, 1890.
4. Zworykin, V. K., Morton, G. A. and Flory, L. E., ‘‘Theory and Performance of the
Iconoscope’’, Proc. IRE, 25(8), 1071–1092, 1937.
5. Zworykin, V. K. and Morton, G., Television, 2nd Ed., John Wiley, New York, 1954.
6. Rose, A., Weimer, P. K. and Law, H. B., ‘‘The Image Orthicon – A Sensitive Television
Pickup Tube’’, Proc. IRE, 34(7), 424–432, 1946.
1. Shure, A., Basic Television, J. F. Rider Publisher Inc., New York, 1958.
2. Fink, D. G. and Christiansen, D., Electronics Engineers’ Handbook, 2nd Ed., McGraw-Hill,
New York, 1982.
3. Stover, W. A. (Ed), Circuit Design for Audio, AM=FM and TV, McGraw-Hill, New York,
4. Fink, D. G., Television Engineering Handbook, McGraw-Hill, New York, 1957.
5. Stremler, F. G., Introduction to Communication System, 3rd Ed., Addison-Wesley, Reading,
MA, 1990.
What is a video signal? Explain why a video amplifier has to be used in
conjunction with a television camera tube.
Explain with the help of a block diagram the configuration and function of
each block of a typical television transmitter. Indicate the type of signal you
expect at the output of each block in terms of waveform, frequency, and
Explain with the help of a suitable diagram the operation of the iconoscope.
What conditions have to be met for it to operate optimally?
Explain with the help of a suitable diagram the operation of the image
orthicon. Discuss its performance relative to the iconoscope.
A television system has the following features:
625 lines per frame with interlacing,
25 frames per second,
5% of the trace period is used for the retrace,
the aspect ratio is 4 : 3.
Assuming equal resolution on the vertical and horizontal axes of the picture
tube, calculate the bandwidth required to handle the signal.
The voltage applied to the anode of a picture tube accelerates an electron to a
velocity of 10 106 m=s along the axis of the tube (x axis). The electron then
passes between two parallel (deflection) plates of length 2.5 cm (in the x
direction). The distance between the plates in the y direction is 2.0 cm and a dc
voltage of 270 V is maintained between them. Assuming that the electrostatic
field is uniform and confined to the space between the plates, calculate the
(1) the velocity of the electron in the y direction as it emerges from the
electrostatic field,
(2) the displacement of the electron in the y direction when it hits a target
placed 5 cm away from the end of the deflection plates.
You may assume the following:
(a) the force (in newtons) on the electron due to the electrostatic field is given
F ¼ eE
where e is the charge (coulombs) on the electron and E is the electric field
intensity (in volts per meter),
(b) the mass of the electron is 9 1031 kg and the charge is 1:6 1019 C.
The electrostatic field described in Problem 6.5 is replaced by a uniform
magnetic field of width 2.5 cm in the x direction. The magnetic field is oriented
in the z direction. The flux density of the field is 1:44 103 T (Wb=m2 ).
Assuming no fringing effects, calculate the following:
(1) the velocity of the electron as it emerges from the magnetic field,
(2) the displacement of the electron when it hits the target placed 5 cm away
from the edge of the magnetic field.
You may assume that:
(a) the force F (newtons) on a conductor of length l (in meters) carrying a
current i (in amperes) in a magnetic field of flux density B (in
teslas ¼ Wb=m2 ) is given by
F ¼ Bli
(b) an electron carrying a charge e (in coulombs) at a velocity of v (in meters
per second) is, in fact, a current i ¼ e (amperes ¼ coulombs=s) flowing in
a conductor of length l (meters).
A video amplifier with a voltage gain of 10 is to be designed in two stages such
that the first stage provides the required voltage gain and the second provides a
current gain with a low output impedance. Design such a circuit and estimate
its bandwidth using the hybrid-p model given in Figure P6.1.
Figure P6.1.
Calculate the value of an inductor that, when connected in series with the collector
load of the voltage–gain stage, will optimize the bandwidth of the amplifier.
In Chapter 6, the coding of video signals in a form suitable for transmission over a
telecommunication channel was discussed. In this chapter, the techniques for
decoding the signals and their presentation on a cathode ray tube will be examined.
The television receiver is almost identical to the AM radio receiver in its use of
the superheterodyne principle. There are a few differences in the details of the signal
processing due to the greater complexity of the system. Figure 7.l shows a block
diagram of a typical television receiver.
The antenna picks up the electromagnetic radiation from the transmitter and feeds
it to the radio-frequency amplifier. After amplification and filtering to attenuate other
incoming signals from other transmitters, the signal goes to the mixer where it is
mixed with the output from the local oscillator. As before, the local oscillator and the
radio-frequency amplifier are tuned to track each other with a constant frequency
difference equal to the intermediate frequency. The intermediate frequency for the
television receiver is usually 45.75 MHz. The signal is subjected to further filtering
before it proceeds to the video demodulator for the recovery of the baseband
information in the signal. The next stage is to separate the composite video signal
into its three components, namely the video proper, the FM sound subcarrier and its
sidebands, and the vertical and horizontal control pulses. The video signal is
amplified to the level required to drive the picture tube by the video amplifier and
the vertical and horizontal control pulses are suitably conditioned and used in the
deflection systems of the receiver to synchronize it to the transmitter – a condition
that must be met for proper reproduction of the images sent. The FM signal is
amplified, amplitude limited and detected and after some amplification it is used to
drive the loudspeaker.
Figure 7.1. The block diagram of the television receiver.
Antenna design is outside the scope of this book. However, a brief qualitative
discussion can be found in Section 2.9. Further discussion of antennas for
commercial FM reception is presented in Section 5.2.1. Frequencies for commercial
FM (88–108 MHz) occupy the spectrum between channels 6 and 7 of the VHF
television frequencies (54–88 MHz and 174–216 MHz, respectively). Except for
slight differences in the physical dimensions, the antennas tend to take the same
form. These are frequency ranges in which a half-wavelength dipole antenna has
reasonable physical dimensions (0.7–2.7 m).
Superheterodyne Section
The radio-frequency amplifier is tunable over the VHF frequency range. This is
accomplished with a variable capacitor which is mechanically ganged to the variable
capacitor which tunes the local oscillator. The objective is to generate a local
oscillator frequency which is equal to the radio-frequency amplifier center frequency
plus the intermediate frequency. In this case the range of the radio frequency is from
approximately 57 MHz (channel 2) to approximately 85 MHz (channel 6) and from
approximately 177 MHz (channel 7) to approximately 213 MHz (channel 13) for
VHF television. The bandwidth is nominally 6 MHz and the radio-frequency gain is
between 20 and 50 times. Since the video intermediate frequency is normally
45.75 MHz, the local oscillator has to be tunable from 102.75 to 258.75 MHz. From
Figure 6.21, it can be seen that there are two carrier frequencies in the composite
video signal: the video carrier and the voice carrier. The voice carrier is 4.5 MHz
above the video carrier, after mixing the video intermediate frequency is at
45.75 MHz and the voice intermediate frequency is at (45.75 4.50) or
41.25 MHz. In general parlance, the radio-frequency amplifier, the local oscillator
and the mixer are called the television front end or simply television tuner. These
components are generally housed in a separate shielded container in an attempt to
control the effects of stray electromagnetic fields and stray capacitive and inductive
elements. The principles to be followed in the design of the circuits in the television
front end are the same as discussed earlier in connection with radio. The only things
that have changed are the frequency of operation and the bandwidth requirements.
Greater attention must be paid to the physical layout of the practical circuits. Radiofrequency amplifiers were discussed in Section 2.8. Oscillator design can be found in
Section 2.4 and mixer design in Section 3.4.3.
Intermediate-Frequency Amplifier
Like all superheterodyne receiver systems, the detailed selection of the desirable and
the rejection of the undesirable frequencies take place at the intermediate-frequency
stage. At the same time, some parts of the spectrum may be emphasized to equalize
the quality of the low-frequency video (large uniform areas) and the high-frequency
video (areas with fine details). The exact frequency response of the video intermediate-frequency amplifier is of no importance at this point except to point out that
it is designed to compensate for the frequency response of the vestigial sideband
filter in the transmitter. To understand the techniques used to achieve this objective
requires a good understanding of the theory of filter design which is beyond the
scope of this book. A list of books on filter design is provided in the bibliography at
the end of Chapter 3.
The output of the intermediate-frequency amplifier must be of the order of several
volts to drive the video detector that follows. Intermediate-frequency amplifiers were
discussed in Section 3.4.4.
Video Detector
It will be recalled that the video signal is amplitude modulated and, in theory, it
requires a simple envelope detector to demodulate it. However, the situation is
complicated somewhat by the fact that the input signal to the detector is a vestigial
sideband signal. Demodulation of Vestigial Sideband Signals. When a carrier of
frequency oc is amplitude modulated by a signal of frequency om, the result is
f1 ðtÞ ¼ A½1 þ m cos om t cos oc t
½cosðoc þ om Þt þ cosðoc om Þt:
f1 ðtÞ ¼ A cos oc t þ
In vestigial sideband modulation, one of the sidebands is removed; it could be either
of them but in this case it is assumed that it is the lower sideband. Also complete
removal is assumed. For simplicity, we have
cosðoc þ om Þt
cos oc t cos om t sin oc t sin om t
f2 ðtÞ ¼ A cos oc t þ
sin oc t sin om t:
f2 ðtÞ ¼ Að1 þ cos om tÞ cos oc t 2
f2 ðtÞ ¼ A cos oc t þ
The amplitude of the carrier signal is
2 mA
jf2 ðtÞj ¼ A 1 þ cos om t þ
sin om t
jf2 ðtÞj ¼ A2 1 þ
þ A2 m cos om t :
When the depth of the modulation is low, m < 1 and m2 =4 1,
jf2 ðtÞj Að1 m cos om tÞ1=2 :
Using the binomial expansion, we get
jf2 ðtÞj A 1 þ cos om t :
This contains the modulating frequency om as well as its higher harmonics whose
amplitudes diminish very rapidly.
The conclusion is that a vestigial sideband signal can be demodulated using a
simple envelope detector so long as the modulation index is much less than unity. It
is worth noting that the amplitude of the output signal is one-half of what it would
have been if both sidebands had been present. Envelope detectors were discussed in
Section 3.2. An example is given in Section 3.4.6. Figure 7.2 shows the input and
output waveforms of a typical television envelope detector.
The Video Amplifier
The design of video amplifiers was discussed in Section 6.3.4. In the television
receiver, the load of the video amplifier is the grid of the cathode ray tube (usually
called the picture tube). This requires voltages between approximately 50 V and
100 V and, in theory, no current flows in the grid circuit. However, the grid represents
a capacitive load and a capacitance requires the movement of charge (current) to
change the voltage across it. The output stage of the video amplifier must be capable
of providing the necessary current and hence power. Another way of saying the same
thing is that the output stage of the video amplifier must have a low output resistance
so that the grid capacitance can be charged much faster than the fastest change in
voltage present in the video signal.
The Audio Channel
From the output of the video detector, a bandpass amplifier selects and boosts the
FM signal centered at 4.5 MHz. The limiter is described in Sections and, and the FM detector is identical to that described in Section The
audio-frequency amplifier was described in Section 2.7. The loudspeaker was
discussed in Section 3.4.8.
Electron Beam Control Subsystem
In the television transmitter, the pulse generator output was used to control the
vertical and horizontal sweeps of the electron beam which scanned the mosaic in the
camera tube. The same pulses were added to the video signal together with the
4.5 MHz FM voice carrier to make up the composite video. Figure 7.2 shows two
Figure 7.2. A typical intermediate-frequency television signal before and after detection.
such pulses (horizontal sync pulses only shown). The horizontal synchronization
pulses are used to control the initiation of the horizontal sweep of the electron beam
in the picture tube so that synchronism with the horizontal sweep of the electron
beam in the camera tube is maintained. Similarly, the vertical synchronization pulses
are used to keep the camera and picture tubes in step in the vertical direction. It is
very important to keep the camera and picture tubes in synchronism in both
directions, otherwise no meaningful image appears on the picture tube. The first
step is to channel the timing information in the sync pulses into a separate circuit for
further processing. Figure 7.3 is a block diagram of the electron beam control
The composite video signal is fed into the sync-pulse separator which takes out
both vertical and horizontal sync pulses. The output is used to drive the two separate
branches of the system. The vertical branch has the vertical sync separator which is
designed to produce an output only when a vertical pulse is present at the input. The
two timing signals undergo essentially identical process steps, namely synchronization to the oscillator, generation of the sweep signal, and amplification to obtain
enough power to drive the deflection coils. It is important to remember that the
vertical sync pulses have a frequency of 60 Hz whereas the horizontal runs at
15.75 kHz.
The sync pulses have the following timing characteristics:
Field period: 16.683 ms
Blanking period: 1.335 ms
Scan period: 15.348 ms
Line period: 63.556 ms
Blanking period: 10.5–11.4 ms
Scan period: 52.156–53.056 ms Sync Pulse Separator. The basic sync pulse separator is a simple
transistor invertor such as shown in Figure 7.4.
The ratio of R1 to R2 is chosen so that the transistor remains in cut off until the
applied voltage exceeds the blanking level (black level of the video signal). The
transistor then conducts and, with an appropriate value for R3, it goes into saturation.
The output of the circuit is then a series of rectangular pulses coincident with the
sync pulses, but inverted. These pulses are used directly to control the horizontal
deflection oscillator. Vertical Sync Separator. It must be recalled that the horizontal sync
pulses are 5 ms long while the vertical are 190 ms. They are easily separated by using
a simple low-pass RC filter with a suitable time constant. Figure 7.5 shows a typical
vertical sync separator circuit.
The RC low-pass filter shown in Figure 7.5 has three sections which should give
it a higher rate of amplitude change with frequency. The difference between the two
frequencies to be filtered (60 Hz and 15.75 kHz) make the filter design simple.
Figure 7.3. The block diagram of the electron beam scan control system.
Figure 7.4. The circuit used for separating the horizontal sync pulse. The threshold is set by
the relative values of R1 and R2 . Vertical Deflection Oscillator. The vertical deflection oscillator is an
astable multivibrator which is synchronized to the vertical sync pulses. Figure 7.6(a)
shows the circuit diagram of the astable multivibrator.
When the dc power is first switched on, current is supplied to the bases of both
transistors and they will both tend to conduct. In general, it can be assumed that one
of the two transistors (say Q1 ) will conduct a little bit better than the other. The
voltage at the collector of Q1 will therefore drop a little faster than that of Q2 .
Because it is not possible to change the voltage across the capacitor C2
instantaneously, the voltage at the base of Q2 will be forced downwards. This will
have the effect of reducing the forward bias on the base-emitter junction of Q2 . The
Figure 7.5. The vertical sync pulse separator. The low-pass filter ensures that it does not react
to the horizontal sync pulse.
Figure 7.6. (a) The circuit diagram of the astable multivibrator. (b) The voltage waveforms for
the bases and collectors.
collector current of Q2 will be reduced and this will make the collector voltage of Q2
go up. Again the voltage across C1 cannot change instantly so the collector voltage
of Q2 will tend to force the base of Q1 upwards. This will cause the base-emitter
junction of Q1 to become more forward biased than it was. More collector current
will flow and the collector voltage of Q1 will drop even more. We are now back to
where we started and the downward change which started the chain of events is now
magnified. This is evidently a regenerative (positive feedback) process which ends
with Q1 in full conduction and Q2 cut-off. With the appropriate choice of Rc1 and
Rb1 , Q1 goes into saturation, essentially its collector voltage is zero. This phase of
the operation is called the regenerative phase and, because regenerative phenomena
are basically unstable, they tend to happen very fast.
The next phase of the operation, called the relaxation phase, starts as C2 begins to
charge up through Rb2 . The voltage at the node B2 will start to rise exponentially, and
eventually it will reach a value sufficient to bias the base-emitter junction of Q2 in
the forward direction. A new regenerative phase starts. Q2 starts to conduct and its
collector voltage starts to drop. The voltage drop will be passed by C1 to the base of
Q1 which will now conduct less, its collector will come out of saturation (that is, go
positive) and will be passed by C2 to the base of Q2 . Q2 will conduct more heavily
and its collector voltage will drop even faster than before. Again the change that set
the chain of events in motion has been magnified by the gain of the transistors. Q2
will eventually go into full conduction (saturation) and Q1 will be cut off. The
appropriate choice of the values of Rb2 and Rc2 will ensure that Q2 goes into
saturation driving the base of Q1 to a negative value equal to the dc supply voltage.
C1 starts to charge up exponentially headed for the positive dc supply voltage, but
when it gets to the value required to forward bias the base-emitter junction of Q1 , Q1
will start to conduct again and the next regenerative phase is repeated followed by
the next relaxation phase ad infinitum.
The voltage waveforms at the collector and bases are shown in Figure 7.6(b).
A simple way to synchronize the astable multivibrator to the vertical sync pulse is
to couple the leading edge of the sync pulse to the base of one of the two transistors
using a differentiator circuit to get a sharp clock pulse. In order to get a stable
synchronization, the astable multivibrator should operate at a frequency slightly
lower than the vertical sync pulse frequency when free-running. The design
principles discussed above are best illustrated by an example.
Example 7.2.1 Vertical Deflection Oscillator Design. Design an astable multivibrator with a mark-to-space ratio equal to 15.35=1.34 (see scan and blanking
periods in Section 7.2.7) and synchronize it to the leading edge of a 60 Hz square
wave derived from the vertical sync pulse of a television receiver. The following are
(1) supply voltage, 12 V dc,
(2) two NPN silicon transistors, b ¼ 100 and Vbe ¼ 0:7 V,
(3) load current ¼ 10 mA.
Solution. Assuming that:
(a) the collector loads of the transistors are the loads of the oscillator,
(b) the collector voltage is zero when the transistor is in saturation,
the collector resistors are
Rc1 ¼ Rc2 ¼
¼ 1:2 kO:
10 mA
The time-constant must be chosen so that one complete cycle takes slightly longer
then 1=60 s, for example 17.0 ms. This means that one transistor will be in
conduction for 15.6 ms and the other for 1.36 ms.
From Figure 7.6(b), it can be seen that the equation of the voltage on the base of
Q1 is
vb1 ¼ 2V ð1 et=t Þ V :
The time taken by the base voltage of Q1 to reach 0.7 V is t ¼ 15:6 ms. The timeconstant is
t ¼ 20:71 ms:
Because the transistor has a b ¼ 100, the base current required to cause saturation in
each transistor is 10 mA=100 ¼ 100 mA. The base resistor is
Rb1 ¼ Rb2 ¼
ð12 0:7Þ
¼ 113 kO:
100 mA
The time-constant t1 ¼ C1 Rb1 ; therefore,
C1 ¼ 0:183 mF:
Similarly, t2 ¼ 1:79 ms ¼ C2 Rb2 , where
C2 ¼ 0:016 mF:
To produce the clock pulse coincident with the leading edge of the square wave, the
differentiation circuit shown in Figure 7.7 is used. The condition for approximate
differentiation is that the time-constant CR T , the period of the input voltage. The
diode clips off the unwanted negative-going spike.
The base voltage of Q1 is now an exponential curve with the positive clock pulse
superimposed on it as shown in Figure 7.8. It is clear from the diagram that, when
the clock pulse occurs at the time represented by the position a, the oscillator will
not synchronize to the clock pulse. However, when the clock pulse occurs in the
Figure 7.7. The differentiation circuit with its input and output waveforms.
position b the voltage at the base of Q1 will exceed 0.7 V, Q1 will go into the
regenerative mode and the oscillator will be synchronized. Vertical Sweep Current Generator. A very simple sweep current
generator with a fairly linear output current with respect to time is shown in Figure
7.9. The circuit also act as a class-A amplifier and drives the yoke coil of the vertical
deflection system.
Figure 7.8. The base voltage of the transistor used in the astable multivibrator showing the
synchronizing pulses.
Figure 7.9. The circuit diagram of the driver for the vertical deflection yoke.
The transistor is biased for class-A operation. The collector is connected to the dc
power supply by a large inductance with negligible resistance, that is, it is an opencircuit at the frequency of operation but a short-circuit to direct current. When the
square wave from the multivibrator is applied to the base of the transistor, the
collector current of the transistor will be a square wave. During current transitions a
voltage will be developed across Lc ; otherwise; the yoke inductor Ly has the dc
power supply connected directly across it. Because the dc voltage is a constant, di=dt
in Ly is a constant. The capacitor C serves as a block to dc flowing to ground.
Because of the low frequency of operation, this circuit does not consume much
power, neither is there a large reactive power circulating in the circuit. Horizontal Deflection System. From Figure 7.3, it can be seen that
the horizontal deflection system has the same modules as the vertical. The
description of the modules and their design will not be repeated. In practice, the
vertical and horizontal deflection systems are different and the differences arise
(1) The vertical system operates at 60 Hz while the horizontal runs at 15.75 kHz,
(2) Deflection in the horizontal and vertical directions are in the ratio 4 : 3 (aspect
(3) The horizontal deflection system is very sensitive to phase modulation and
hence the horizontal oscillator has to have an automatic phase control loop to
keep phase modulation as low as possible,
(4) High power (50 W) is required to run the horizontal deflection coil if the
energy stored in it is dissipated at the end of each cycle. By using extra
circuitry, it is possible to return most of the energy to the dc source by
resonating the inductance of the yoke with a suitable capacitor to cause
ringing at approximately 60 kHz,
(5) The high dc voltage (10,000 V) required to run the picture tube can be
obtained by using a transformer coupling (auto-transformer) to step up the
voltage followed by a rectifier (tube type). A simplified circuit diagram of the
final stage of the horizontal deflection system is shown in Figure 7.10.
The horizontal oscillator is synchronized to the output of the sync separator. The
oscillator then drives the base of Q2 with a string of rectangular pulses at
approximately 15.75 kHz. When Q2 is on, it draws current through the primary of
the transformer T1. This causes a current to flow in the secondary, biassing Q1 on
and putting it into saturation. The dc supply voltage V is therefore connected directly
across the primary terminal P of the auto-transformer and ground. Q1 and the
winding of the autotransformer between node P and ground form an emitter
follower. The horizontal deflection yoke inductance, Lx , which is connected to the
auto-transformer at node N , has a constant voltage, higher than V , connected across
it. The current that flows in Lx is the linear ramp required to produce a linear
horizontal deflection. The capacitor Cc blocks dc from flowing in Lx , and it is a
short-circuit at the frequency of operation. At the end of the linear ramp, Q2 is cut
off, a voltage is induced in the secondary of T1 , and this is in the appropriate
direction to shut off Q1 abruptly. Normally, the sudden termination of the current
flow in the auto-transformer and Lx will generate a huge voltage spike in the attempt
to dissipate the energy stored in the inductances. However, the diode D1 conducts
Figure 7.10. The driver of the horizontal deflection yoke with the extra-high voltage (EHV)
generator required for the anode of the picture tube.
providing a path for the current to flow into the tuned circuit made up of Lx and Cr .
Lx and Cr resonate (‘‘ring’’) at approximately 60 kHz (four times the horizontal
deflection frequency). If the timing is set up correctly, the energy stored in the
resonant circuit will be flowing from the capacitor into the yoke just as the next
current ramp is starting. This arrangement substantially reduces the amount of power
taken from the dc supply to drive the yoke. As no diode is connected across the
M N portion of the auto-transformer, a large voltage spike appears across it. This is
rectified and used to bias the anode of the picture tube.
Picture Tube
The picture tube is an example of a cathode ray tube (CRT) used in a wide variety of
display systems. It consists of a glass cylinder which flares into a cone with a nearly
flat base. Figure 7.11 shows a cross-section of the typical CRT used for the display of
television images.
At the end of the glass cylinder, there is an electron gun. The electron gun is
made up of a cylinder, closed at one end and coated with oxides of barium and
strontium which give off electrons when heated. The CRT has control grid(s),
focussing and deflection electrodes or coils. An electric current is passed through the
heater coil to bring the temperature of the cathode to the appropriate value, electrons
are given off and, under the influence of a positive voltage placed on the anode, the
electrons travel down the tube and strike the screen at high velocity. The grid(s)
carry a negative voltage which can be varied to control the number of electrons that
can leave the cathode and eventually reach the screen. The grid(s) therefore control
the intensity of the light forming the image. The electron lens is either an
electrostatic or a magnetostatic means of focusing the electrons, usually to produce
a small spot on the screen.
Figure 7.11. A cross-sectional view of the cathode ray tube (CRT). Reprinted with permission
from T. Soller, M. A. Starr and G. E. Valley, Cathode Ray Tube Displays, Boston Tech. Pub.
Lexington, MA, 1964.
The rest of the tube is made up of the deflection electrodes or coils which cause
the electron beam to move on the screen on the application of the appropriate voltage
or current. The anode is a conductive thin film on the inside wall of the flare as
shown and requires several thousand volts positive for proper operation. The screen
is coating with silicates, sulfides, fluorides and alkali halides which emit light when
bombarded by electrons. The color of the light emitted is determined by the
characteristic wavelength associated with each chemical element present in the
screen material. For example, copper-activated 85% zinc sulfide and 15% cadmium
sulfide gives off yellow light while 93% zinc sulfide and 7% cadmium sulfide
produces a blue-green phosphor. The persistence (or afterglow) of the phosphor can
be varied to suit different conditions.
Most modern CRTs used in television receivers have an electrostatic focussing
system because it is simpler and less expensive to manufacture. The deflection
system is, however, magnetostatic because of the tendency to have shorter tubes and
larger screen areas.
Demodulation and Matrixing
In Section 6.3.8, it was established that black-and-white and color television signals
had many common features. Their basic differences are:
(1) in addition to the luminance information (Lw ), the chrominance information
(Lr Lw ) and (Lb Lw ) is transmitted using a quadrature modulation (DSBSC) scheme with a (color) subcarrier of frequency approximately 3.58 MHz,
(2) on the back porch of the horizontal blanking pulse, an eight-cycle bursts of
the color subcarrier is superimposed.
One scheme used for the recovery of the original red, blue, and green signals is
shown in Figure 7.12. It is described as pre-picture tube matrixing.
From the output of the video detector, the color burst separator picks up the eight
cycles of the 3.58 MHz subcarrier and passes them onto the color subcarrier
regenerator. The regeneration circuit is essentially a very high Q-factor circuit
tuned to 3.58 MHz. The eight-cycle burst of signal causes the high Q circuit to go
into a free-running oscillation mode. The subcarrier must have limited variation in
both amplitude and phase, for proper synchronous detection, and therefore the decay
of the amplitude during the free-running period must be controlled. To maintain the
amplitude variation to 90% of the initial value for the time required to sweep one
horizontal line (63 ms) a Q-factor of 7000 is required. Crystals are used in the
regenerator because only crystals have the necessary high Q-factor.
The regenerated 3.58 MHz signal is split into two branches. The first branch goes
through a 90 phase shifter to produce the quadrature signal required for the
synchronous demodulation of the ðLr Lw Þ and ðLb Lw Þ signals contained in
the output of the color bandwidth filter. After demodulation, the matrix reproduces
Figure 7.12. The schematic diagram for color separating in the receiver. This is described as prepicture tube matrix because the colors are separated outside the tube.
the third difference signal, namely (Lg Lw ). The composite video signal containing
the luminance information Lw is now combined in the adders to give the original red,
green, and blue information. After suitable amplification, they drive the picture tube
cathodes. A variation on this scheme, called picture tube matrixing, produces the
(Lg Lw ) signal and the addition of the Lw is carried out in the picture tube by
driving the grids and the cathodes differentially. This is shown in Figure 7.13.
When the I and Q signal scheme is used in the transmission, it must be decoded
into (Lr Lw ) and (Lb Lw ) after the synchronous detectors and before the signal is
applied to the matrix.
Component Circuit Design
The 90 phase shift circuit was discussed in Sections and The
synchronous detector (four-quadrant multiplier) was discussed in Sections 2.6.4 and The design of the adder was described in Section Video amplifier
design was discussed in Section Color Burst Separator. The eight cycles of the color burst subcarrier
riding on the back of the horizontal blanking pulse are shown in Figure 6.24. Since
this signal is required for the synchronous detection of the chrominance signal, no
part of the video signal must be allowed to get through the separator. The most
popular technique is to use the falling edge of the horizontal sync pulse to trigger a
gate open and to close the gate immediately after the time it takes the eight cycles of
3.58 MHz to come through (2.23 ms). The circuit used is shown in Figure 7.14.
Q2 is an amplifier biased in the off state by the resistors R3 ; R4 ; R5 , and R6 . Its
emitter is coupled through the tuned circuit LC to the collector of Q1 . The base of Q1
is driven by the output from the horizontal oscillator (see Figure 7.3). The resonant
frequency of the LC circuit is adjusted so that the rise-time of the rectangular
waveform from Q1 is delayed long enough to let only the eight cycles of the
3.58 MHz signal through to the collector of Q2 . Color Subcarrier Regenerator. A typical circuit of the color subcarrier regenerator is shown in Figure 7.15.
The output from the transistor Q1 is taken from both the collector and the emitter.
C2 is adjusted to cancel the shunt capacitance of the crystal (see Figure 2.18). The
crystal is chosen so that its series resonant frequency is 3.58 MHz and it resonates
when it is excited by the subcarrier. In the series mode resonance, the crystal
impedance is essentially a short-circuit (ideal voltage source). The losses in Re and
R3 in parallel with R4 are therefore insignificant. The circuit has a high enough Qfactor to perform about 217 cycles of free-running oscillations between bursts of
color subcarrier signals. Q2 is used as a buffer amplifier. In some earlier models, the
signal from the regenerator was used to synchronize an LC oscillator whose output
was then used for the demodulation. This is an expensive way to solve the problem
in view of the satisfactory performance of the free-running system.
Figure 7.13. The block diagram for color separation in the tube. Note that the color differences are
applied to the tube directly.
Figure 7.14. The color subcarrier burst separator. The output is used to synchronize the color
subcarrier regenerator. The Matrix. The function of the matrix is to reproduce the third
difference signal (Lg Lw ). Equation (6.4.4) was derived for this operation and it
is repeated here.
ðLg Lw Þ ¼ 0:30
ðLr Lw Þ ðL Lw Þ:
0:59 b
It is clear that this can be achieved with a simple resistive network and a phase
reversing buffer amplifier.
Color Picture Tube
In Section 7.2.8, it was stated that different phosphor materials coatings on the
screen of the picture tube give different colors of light when bombarded by electrons.
A number of different techniques for producing images in color on the CRT exist; all
of them rely on this phenomenon. The most successful of these uses the shadow
mask technique. Figure 7.16 illustrates the basic concept.
Three electron guns representing the three primary colors are mounted symmetrically on a plane perpendicular to the axis of the tube. Electrons from each gun will
therefore bombard different areas of the screen after going through the holes in the
shadow mask. If the area of the screen corresponding to the blue gun is coated with a
phosphor that produces blue light and so on, then it is clear that, by depositing the
different phosphors materials in dots grouped in threes at the appropriate points on
the screen, the different colors can be reproduced by modulating the relative
saturation of each component.
The shadow mask technique has the following disadvantages:
Figure 7.15. The color subcarrier regenerator. Note that, because of the high Q factor of the crystal in
the circuit, it is virtually free-running between bursts of the subcarrier.
Figure 7.16. The color picture tube showing details of the shadow mask. Reprinted with
permission from H. B. Law, ‘‘Shadow Mask’’, Proc. IRE (now IEEE), 39(10), 1951.
(1) The shadow mask has to be very carefully aligned for good color
(2) The presence of the mask reduces the average brightness (luminance) of the
image but this can be compensated for by increasing the electron emission of
the guns.
High-definition television is an attempt to produce images of quality comparable to
that of a 35 mm motion picture film. The proposed system has 1125 lines of scan
with 60 fields per second. Japanese HDTV uses analog techniques and requires a
bandwidth of 20–25 MHz. In North America, HDTV uses digital technology which
permits the compression of the data to fit into a 6 MHz bandwidth. The standard for
digital television has the acronym MPEG-2 and it can achieve a bit compression
ratio of 55 : 1. It does this by retaining most of the data and presenting only changes
in a picture frame. Despite the improvement in picture definition achieved by HDTV
the viewing public has been slow to switch over to the new format. The high price
tag for the HDTV receiver means that only a very select number of viewers are
willing to adopt the new system. Equally, the broadcasting organizations are
unwilling to invest large amounts of money in HDTV to re-equip their studios
Total no. of lines
Actual no. of lines
Aspect ratio
Max. resolution
6 MHz
720 486
2 channels (stereo)
6 MHz
16 9
1920 1080
5:12 channels
The 0.1 channel refers to a subwoofer for bass ‘‘sounds you can feel’’.
when they are not sure of reasonable returns on their investment. Table 7.1 compares
the current standard (NTSC) to HDTV.
1. Kennedy, G., Electronic Communication Systems, McGraw-Hill, New York, 1970.
2. Smith, J., Modern Communication Circuits, McGraw-Hill, New York, 1986.
3. Parker, N. W., ‘‘Television Image-Reproducing Equipment’’, in Electronics Engineers’
Handbook, Fink, D. G. and Christiansen, D. (Eds), 2nd Ed., McGraw-Hill, New York,
4. Sherr, S., Fundamentals of Display System Design, Wiley-Interscience, New York, 1970.
5. Soller, T., Starr, M. A. and Valley, G. E., Cathode Ray Tube Displays, Boston Tech.
Publishers, Inc., Lexington, MA, 1964.
6. Law, H.B., ‘‘Shadow Mask’’, Proc. IRE, 39(10), 1187, 1951.
Give two advantages of vestigial sideband modulation systems. What are their
disadvantages? A transmission consists of a carrier and its upper sideband.
Show that, when the modulation index is small compared to unity, an ordinary
envelope detector can be used to demodulate it. Calculate the signal power in a
vestigial sideband signal if the peak value of the carrier voltage is 5 V when the
load resistor is 2:5 kO and the modulation index is 0.15. You may assume that
the diode is ideal.
A television receiver is tuned to receive a transmission with a carrier frequency
67.25 MHz (Channel 4). If the intermediate frequency is 45.75 MHz, what is
the possible frequency(ies) at the local oscillator terminals of the mixer? If
there are two possible frequencies, does one of them offer an advantage over
the other? The local oscillator signal for the television receiver is generated at a
lower frequency and multiplied by a factor of 512. Calculate the frequency of
the oscillator and design a Hartley-type oscillator given that the tuning
capacitor has the value 144 pF. Discuss the precautions you would take to
minimize changes in the frequency of oscillation.
Using suitable block diagrams and derived equations, explain how quadrature
multiplex is used to transmit and receive the chrominance information in a
color TV system. State the conditions which must be satisfied for demultiplexing to be possible. How are these conditions met in the domestic TV receiver?
Describe the construction and operation of a color TV picture tube. Explain
why a color tube requires a higher (EHV) dc voltage than the monochrome for
equal luminance.
Describe with the help of a suitable diagram the operation of a color subcarrier
Show that for a series-tuned RLC circuit at resonance:
(1) The total energy stored in the circuit is a constant,
Q0 ¼ 2p
Energy stored
Energy dissipated per cycle
In a crystal-controlled color subcarrier regenerator operating at 3.58 MHz, the
total energy remaining in the circuit after 218 cycles of free-running oscillation
is 90% of its original value. Calculate the peak value of the voltage across the
capacitor at the end of the free-running cycle and estimate the Qo of the circuit.
What is a video signal? Explain why a video amplifier has to be used in
conjunction with a television camera tube.
A television system has the following features:
625 lines per frame with interlacing,
25 frames per second,
5% of the trace period is used for the retrace,
the aspect ratio is 4 : 3.
Assuming equal resolution on the vertical and horizontal axes, calculate the
bandwidth required to handle the signal.
The early history of the telephone system has been outlined in Chapter 1. The
growth of the telephone system has been truly phenomenal and forecasts show a
continuing growth as new services such as data transfer, facsimile and mobile
telephone are added.
The telephone differs from the broadcasting system in two basic ways:
(1) In broadcasting, a few people who, in theory, have information send it out to
the many who are presumed to want the information; it is one-way traffic. The
communication link provided by the telephone is two-way traffic.
(2) The basic idea of broadcasting is to make the message available to anyone
who has the equipment and the interest to tune in. This is in contrast to the
norm in the telephone system where the privacy of the message is guaranteed
by law.
Because of these differences, the two systems handle very different types of
information – public versus private – and their patterns of development have been
For a telephone system to work, there must be a minimum of two people who wish to
communicate. It is then possible to install the circuit shown in Figure 8.1.
This would be quite adequate except for the fact that these two people may not
want to talk to each other all the time and therefore some additional system has to be
set up for either person to indicate to the other that they wish to talk. What was
Figure 8.1. The basic elemental telephone system.
added was a bell on the called party’s premises which can be rung from the calling
party’s premises.
Presumably the success of this prototypical communication system would soon
attract the attention of other people who would want to set up similar systems. It is
clear that soon the situation depicted in Figure 8.2 would develop where every
subscriber would have to be wired up to every other subscriber. That would be
prohibitively expensive and quite impractical.
Evidently, the way to deal with the situation is to connect every subscriber to a
central location and arrange to have an attendant to interconnect the various
subscribers in whatever combination that is required. That central location is, of
course, the central office which has and continues to have a central role in the
telephone system. The system configuration would then be as shown in Figure 8.3.
Assuming that the system has six subscribers then each subscriber has access to
the other five subscribers. But in the meantime, a group, also of six, in the next town
have heard of the success of the system and have set up a similar system of their
Figure 8.2. The connection diagram for six subscribers showing all the 15 possible telephone
Figure 8.3. The concept of a central office reduces the number of lines to six.
own. Now there is a possibility of reaching eleven other subscribers if a connection
can be made between the two systems. This brings up two very important points:
(1) The greater the number of people on the communication network, the more
attractive it is for other people to join.
(2) There has to be some level of compatibility between the two systems.
The system would have evolved, as shown in Figure 8.4.
Continuing with the story, the distance between the two towns is quite long and
the initial cost and upkeep are high but if this line can be made to carry more than
one conversation simultaneously, the cost per conversation will be substantially
Figure 8.4. The connection of a trunk or toll line between two central offices increases the
number of subscribers that can be reached from 5 to 11.
reduced. It is also very likely that because of the length of the line, the quality and
the reliability of the system may be degraded. This brings up an important point:
The greater the number of communication channels that can be established over the
same link, the less the cost per message. Multiplex and the conservation of bandwidth
will become goals of several generations of communications engineers.
The telephone system that started with people talking to each other has acquired
more than people for a clientele. Increasingly, the network is being used to supply
services to machines such as computers, facsimile devices, security guard services
and access to the Internet.
The basic telephone has surprisingly very few parts. These are shown in Figure 8.1.
When the microphone is connected in series with the battery, it produces a current
proportional to the pressure of the sound impinging on it. The transformer eliminates
the dc and sends the ac portion of the current through the line. The earphone at the
receiving end changes the variation of the current into sound. Obviously, the system
works in the reverse direction.
Carbon Microphone
Figure 8.5 shows a cross-section of the carbon microphone [1].
It has a light-weight aluminum cone with a flexible support around the periphery
so that it will deflect (vibrate) due to the changing sound pressure level. Attached to
the apex is a disc which acts as a piston when the cone deflects. A plastic housing
with an electrode attached to the bottom contains a loose pile of carbon granules.
When the pressure on the cone is increased, the carbon granules become
Figure 8.5. A cross-sectional view of the carbon microphone.
compressed, the resistance goes down and more current flows. The opposite happens
when the pressure is released.
Assuming that the sound pressure level on the carbon microphone is a sinusoid
then the resistance of the device is
rðtÞ ¼ r0 ð1 þ k sin otÞ
where r0 is the mean resistance, k is a coefficient less than unity, and o is the
frequency of the sound pressure.
When the microphone is connected to a battery of electromotive force E (volts) in
series with a load R, as shown in Figure 8.6, we have
R þ r0 ð1 þ k sin otÞ
R þ r0
sin ot
R þ r0
ð1 þ k
sin otÞ1 :
R þ r0
R þ r0
Using the binomial expansion,
½1 sin2 ot . . .:
sin ot þ
R þ r0
R þ r0
R þ r0
I0 ¼
R þ r0
Figure 8.6. The carbon microphone with dc supply E and load resistance R.
I ¼ I0 I0
cos 2ot þ : ð8:3:7Þ
sin ot þ 0
R þ r0
2 R þ r0
2 R þ r0
Because kr0 =ðR þ r0 Þ is smaller than unity higher order terms can be ignored. If it is
desirable to reduce second harmonic distortion, kr0 =ðR þ r0 Þ can be reduced, but in
doing so the amplitude of the fundamental will be reduced as well. A compromise
between distortion and signal amplitude has to be made.
The carbon microphone has the following attractive properties:
(1) It is simple and therefore inexpensive to manufacture.
(2) It is robust; it is not likely to need attention even in the hands of the public.
(3) It acts as a power amplifier; under normal bias conditions, (the electrical
power output far exceeds the acoustic power input. It does not normally
require additional amplification.
(4) Its input–output characteristics are shown in Figure 8.7. The non-linearity at
low input levels helps to suppress background noise and that at high levels
acts as an automatic gain control.
Moving-Iron Telephone Receiver
A cross-section of the moving-iron telephone receiver is shown in Figure 8.8. It
consists of a U-shaped permanent magnet that carries a coil as shown. In front of the
open face of the U, a thin cobalt iron diaphragm, is held by an annular ring support
with a short distance between them. With no current in the coil, the diaphragm has a
fixed deflection towards the magnet.
The signal current is passed through the coil and, assuming that it is sinusoidal,
then for one-half of the cycle the flux generated by the current will aid the pull of the
Figure 8.7. The input–output characteristics of the carbon microphone.
Figure 8.8. A cross-sectional view of the moving-iron telephone receiver.
permanent magnet on the diaphragm and it will deflect accordingly. During the other
half of the cycle, the coil flux will oppose that of the magnet and the diaphragm will
deflect much less.
The force between two magnetized surfaces is given by
ðN=m2 Þ
where B is the flux density in teslas ðTÞ and m0 is the permeability of free space, that
is, 4p 107 . Let A be the area of the pole face ðm2 Þ, B0 the flux density due to the
permanent magnet (T), and b0 sin ot the flux density due to the current (T).
The force in newtons is then
ðB þ b0 sin otÞ2
2m0 0
A 2
ðB þ 2B0 b0 sin ot þ b20 sin2 otÞ
m0 0
A 2 1 2
ðB þ b þ 2B0 b0 sin ot 12 b20 cos 2otÞ
m0 0 2 0
The second harmonic component can be reduced by making B0 large compared to
b0 . This will increase the direct component of the force, which is likely to cause the
diaphragm to touch the magnet. Note that when B0 is zero (no permanent magnet),
the device produces only the second harmonic. This is to be expected since both the
positive and negative halves of the sinusoid will exert an equal force of attraction on
the diaphragm.
Local Battery – Central Power Supply
The system as depicted in Figure 8.1 is powered from batteries that are located on
the customer’s premises. The batteries are of interest because they are a hazard to the
customer and they pose a very difficult problem for the maintenance staff.
Furthermore, their reliability is questionable because of their location among other
The solution to the problem is to have a common power supply located at the
central office (out of the way of the telephone subscriber) and readily available to the
maintenance personnel. The reliability of the service can then be improved by
installing a backup power supply. The scheme for achieving this end is illustrated in
Figure 8.9.
The central office battery in series with two inductors is connected to the lines of
the calling and called party as shown.
The inductors have a high inductance and therefore appear to be open-circuits at
audio frequency but short-circuits at dc. Every call requires two such inductors to
complete the connection.
Signalling System
The signalling system consisted of a magneto and a bell which responded to high ac
voltage input. The magneto was a hand-operated alternator whose flux was produced
by a permanent magnet. The calling party turned the crank to produce about 100 V
ac. The current travelled down the telephone line and caused the bell at the called
party’s end to ring. To avoid damage to the telephone receiver and to conserve
Figure 8.9. The local batteries are replaced by a central power supply.
Figure 8.10. The elemental telephone with signalling devices (magneto and bell) shown.
battery power, the hook switch was disconnected them from the line when the
telephone was not in use. A simplified diagram of the signalling system is shown in
Figure 8.10.
The Telephone Line
Physically, the telephone line consists of a pair of copper wires supported on glass or
porcelain insulators mounted on wooden poles. Electrically, an infinitesimally short
piece of line can be modelled as shown in Figure 8.11 [2,3]. The elemental series
resistance and inductance are represented by dR and dL and the elemental shunt
capacitance and conductance are represented by dC and dG, respectively.
The analysis of the model is beyond the scope of this book. However, the analysis
shows that the telephone line, at voice frequencies, can be approximated by an RC
Figure 8.11. (a) The equivalent circuit of the telephone line showing series resistance R and
inductance L and shunt capacitance C and conductance G. (b) An elemental equivalent circuit of
the telephone line.
low-pass filter whose cut-off frequency is a function of its length. The longer the line
is, the lower the cut-off frequency. The frequency response of a typical telephone
line is shown in Figure 12.1.
Performance Improvements
From Figure 8.9 it can be seen that the dc required to power the carbon microphone
has to flow through the receiver. This is not a good idea since it will make B0 , the
flux density of the permanent magnet (see Equation (8.3.11)), larger or smaller than
it should be. A second disadvantage is that all the ac current generated by the carbon
microphone has to flow through the receiver. This produces a very loud reproduction
of the speaker’s own voice in her receiver. The psychological effect is that the
speaker lowers her voice, making it difficult for her listener to hear what she is
saying. This phenomenon is called sidetone. The two problems can be solved by
using the circuit shown in Figure 8.12. It is an example of a hybrid. The Hybrid. The carbon microphone is connected to the centre-tap of
the primary of the transformer. One end is connected to the telephone line and the
other to an RC network which approximates the impedance of the line. The
secondary is connected to the receiver. There is still a path for dc from the central
office battery to flow through the carbon microphone.
In the transmit mode, the ac produced by the microphone divides up equally, with
one half flowing through the telephone line and the other half in the line-matching
impedance. Since these currents are in opposite directions in the primary of the
transformer, no net voltage appears across the secondary. The speaker cannot hear
himself. In the receive mode, the current IR flows through the first half of the primary
winding and then splits at node X with Im flowing through the carbon microphone
where the energy is safely dissipated. The remainder (IR Im ) flows through the
second half of the primary into the line-matching impedance. This time, the
Figure 8.12. The use of a hybrid transformer to control sidetone.
directions of the two currents in the primary are the same; a net voltage appears
across the receiver.
In practice, the level of sidetone fed back to the speaker has to be carefully
controlled. When it is too low, the telephone appears dead to the speaker and her
normal reaction is to raise her voice. When the sidetone is too high, it has the
opposite effect. The Rotary Dial. The rotary dial came with the invention of the
automatic central office. Evidently, the automatic central office offered several
advantages over the manual office. There was increased security of the messages
since there was no human interface in setting up calls. The time for setting up and
releasing a call was substantially reduced and the probability of operator errors
decreased. It guaranteed 24-hour service with fewer more highly trained personnel.
The dial is simply a method of issuing instructions to the central office and it does
this by producing a binary coded message by mechanically opening and closing a
switch in series with the circuit. The basic dial is as shown in Figure 8.13. It has a
finger wheel with ten finger holes and it is mechanically coupled by a shaft to a
second wheel which has ten cam lobes as shown. The shaft is mounted so that both
wheels can rotate about the axis. The wheel assembly is spring loaded so that, when
it is rotated in a clockwise direction, it will return at a constant speed under the
control of a mechanical governor.
Figure 8.13. The essential features and operation of the rotary dial.
To operate the dial, the caller inserts his or her index finger into the hole
corresponding to the number and pulls the finger wheel to the finger stop and then
releases it. While the finger wheel is rotating in the clockwise direction, the lever X
is free to move out of the way of the cam lobes without disturbing the switch lever Y.
When the wheel assembly is rotating in the counter-clockwise direction, every cam
lobe that passes X will cause the switch lever Y to open the switch. If current is
flowing through the switch, the current flow will be disrupted the number of times
corresponding to the number of the finger hole. The current pulses can be used to
operate a device (to be discussed later) at the central office to effect the required
connection. The return spring, cam lobes and mechanical governor are designed to
produce 10 pulses per second with approximately equal mark-to-space ratio. Since it
is bound to take the subscriber much longer then 1=10 seconds to rotate the finger
wheel again, a pause longer than 1=10 seconds can be recognized by the central
office as an inter-digit pause. It is then possible to send a second and subsequent
string of pulses to effect a connection which requires a multi-digit code. Note that
when the digit ‘‘0’’ is dialled, ten pulses are produced. Telephone Bell. The telephone bell has two brass gongs with a clapper
which is operated by an electromagnet. It is mechanically and electrically tuned to
respond optimally (resonance) to current at 20 Hz. It is left connected to the
telephone line at all times but the high impedance of its electromagnet coil ensures
minimal effect at voice frequencies. Also the 10 Hz pulse from the rotary dial has no
significant effect on it. Nominally, it operates on 88 V, 20 Hz ac supplied to it from
the central office in the ring-mode.
Telephone Component Variation
The telephone components described in this section are meant to be a representative
sample of what can be found within the territory of any telephone operating
authority or company. For each component there are several possible variations,
some made to get around patents rights granted to others, and some to lower cost and
improve reliability.
The subscriber telephone instrument has changed in its physical appearance and
electrical characteristics since it was first put into service. However, in broad terms, it
remained basically the same until the introduction of electronics in the form of
semiconductor devices. The availability of amplifiers at very low cost offered various
options such as new microphones, electronic sidetone control, tone ringers and tone
dialling. Some of these will now be discussed.
By the late 1960s a number of electronics research and development organizations
were working on the development of electronic telephone sets. Manufacturing cost
reduction, improved performance and the possibility of offering the subscriber a
number of new uses for their telephones were incentives for this development. The
first truly electronic component to emerge was the tone dialler, more popularly
known by its trade name TOUCH-TONETM. Telephone sets with other electronic
components were not far behind so that by the early 1980s there were a number of
telephones on the market with none of the well established components described
The features of the carbon microphone that made it indispensable in the telephone
set for a very long time were low cost, acoustic-to-electrical power amplification,
suppression of low level background noise, and high level signal compression. Its
drawbacks were distortion, high dc current requirements, and changes in its acoustic
sensitivity due to dc current flowing in it. All of its advantages can be obtained with
none of the disadvantages by using other microphones in conjunction with a suitable
amplifier. The new microphones could be made considerably smaller than the carbon
microphone. Examples of such microphones include the following:
(1) Electret Microphone. This is a variation on the capacitance microphone. The
incident sound causes the distance between two plates of a capacitance to change,
resulting in a change of voltage. The output voltage and power are very low. The
output impedance is very high (10 pF at audio frequency). The normal biassing of
capacitor microphones is averted by a built-in charge that is placed on the
capacitor during the manufacturing process. It has an excellent frequency
response and is normally used in acoustical measuring instruments. [4]
(2) Ribbon Microphone. A thin aluminum ‘‘ribbon’’ is suspended in the field of
a small powerful magnet. The incident sound causes the ribbon to vibrate in the
field, causing a voltage proportional to its velocity to be induced in it. It has a very
good frequency response but a very low output power.
(3) Crystal Microphone. A crystal of Rochelle salt (sodium potassium tartrate),
quartz and other piezo-electric materials produce a voltage when subjected to
mechanical deformation. The crystal is cut into a thin layer with suitable
conductors connected to the faces. The incident sound pressure causes a voltage
to appear across the conductors. The microphone is very often made up of several
layers of crystals.
The receiver is one of the few components that has successfully resisted change
since Alexander Graham Bell patented it in 1876. The materials used for making the
magnet and diaphragm and the actual mechanical construction have changed but the
basic principle of operation remains the same.
The function of the ideal hybrid is to direct the signal from the microphone on to the
telephone line without loss and to direct the incoming signal on the line to the
receiver with no loss. The operation of the hybrid is therefore similar to the operation
of a circulator – a well known device in microwave engineering. The two devices are
compared in Figure 8.14.
There are two major differences:
(1) There are two critical paths in the hybrid, (transmit and receive) but three
critical paths in the circulator. In a normal circulator, the transmitter cannot
have a path to the receiver.
(2) Circulators are realizable in reasonable physical dimensions at microwave
frequencies. A direct application of circulator theory at audio frequency
predicts a device several kilometres in diameter.
The problem of realizing a circulator at audio frequency can be solved by using a
gyrator [5]. Gyrators are better known for their ability to invert impedances [6,7].
They were the subject of intense interest at a time when the micro-electronics
industry was looking for a micro-miniaturized version of the inductor.
Consider the two-port shown in Figure 8.15 with the port voltages and currents as
Figure 8.14. The telephone hybrid (a) compared to the circulator (b).
Figure 8.15. A two-port with its defining voltages and currents.
Figure 8.16. The ideal voltage-controlled current source.
The two-port is described by the matrix
Y11 Y12 V1
Y21 Y22 V2
Applying this description to the ideal voltage-controlled current source shown in
Figure 8.16 gives
0 0 V1
gm 0 V 2
where gm is the transconductance.
Consider a second ideal voltage-controlled current source with a 180
phase shift.
The matrix equation is
when the two ideal voltage-controlled current sources are connected back-to-back as
shown in Figure 8.17.
The matrix equation is
Figure 8.17. Two ideal voltage-controlled current sources connected back-to-back to form a
The two-port can be converted into a three-terminal element by lifting the ground. Its
matrix equation is then
3 2
4 I2 5 ¼ 4 gm
32 3
gm 54 V2 5:
Figure 8.18 shows the three-terminal circuit in which the transconductance gm has
been replaced by the more general transadmittance Y0 and port 1 has been terminated
in an admittance Y1 , port 2 in Y2 and port 3 in Y3 .
Consider that an ideal current source I1 is connected across port 1 so that a
voltage V1 appears across it. This will induce voltages V2 and V3 across ports 2 and
3, respectively. The voltage ratio is
V2 Y1 ðY3 Y0 Þ
Y2 Y3 þ Y02
V3 Y1 ðY2 þ Y0 Þ
Y3 Y2 þ Y02
When Y1 ¼ Y2 ¼ Y3 ¼ Y0 , Equation (8.4.6) gives
Figure 8.18. The gyrator when properly terminated behaves like a circulator.
and Equation (8.4.7) gives
¼ 1:
This means that voltage across port 1 appears across port 2 but no voltage appears
across port 3, that is,
V2 ¼ V1
V3 ¼ 0:
When the process is repeated with the ideal current source connected across port 2,
V3 ¼ V2
V1 ¼ 0:
Finally, with the ideal current source across port 3,
V1 ¼ V3
V2 ¼ 0:
It is clear that the circuit is behaving like a circulator oriented in a clockwise
To get the desired effect, the transadmittance of the gyrator amplifiers have to be
adjusted to fit the line admittance. The admittances of the microphone and receiver
circuits have to be tailored so that the gyrator sees an admittance equal to its own
admittance connected to each port. In practice, less than a perfect match can be
achieved and therefore some sidetone is obtained. This property can be exploited to
adjust the level of the sidetone to a comfortable level.
Several electronic telephones use the audio-frequency circulator concept and its
various manifestations as hybrids.
Tone Ringer
In most electronic telephone sets, the bell has been replaced by some kind of tone
ringer which usually emits an attractive musical note or notes to signal an incoming
call. Quite often amplitude and frequency modulation are used to enhance the tone.
The tone ringer must satisfy the following conditions:
(1) The input impedance must be high so as not to interfere with the signal on the
line to which it is permanently connected.
(2) It has to operate on the 20 Hz, 88 V ac ringing signal that was used with the
electromagnetic bell.
In terms of circuit design, what is required is one or two audio-frequency oscillators,
a frequency and=or amplitude modulator, a power supply fed from the 20 Hz ring
signal, an audio-frequency amplifier and a loudspeaker. The design of all these items
has been discussed earlier. Oscillators were discussed in Section 2.4, modulators in
Section 2.6, audio-frequency amplifiers in Section 2.7, and loudspeakers in Section
3.4.8. The power supply for the ringer would require a rectifier and a capacitor to
smooth the output of the rectifier to form a suitable dc supply for the ringer.
Tone Dial
Instead of producing current pulses to signal the number dialled to the central office,
the tone dial produces a pair of audio-frequency tones. The frequencies of these
tones are carefully chosen so they are not harmonically related. This reduces the
probability of other tones or signals being recognized as dialled numbers. The dial
pad and the corresponding frequencies produced are shown in Figure 8.19. Touch-ToneTM – DigitoneTM Dial. This dial consists of two essentially identical oscillators, one of which produces the high-frequency and the other
the low-frequency tones. The change in frequency is achieved by switching in
resistors of appropriate values when the dial push button is depressed. Each button
has a unique pair of tones associated with it. The central office equipment recognizes
the number dialled by the two frequencies present.
Figure 8.19. The push-button dial and its corresponding frequencies.
The early versions of the TOUCH-TONETM dial used discrete bipolar transistors
in conjunction with an inductor and capacitances in a Colpitts configuration. Both
the low- and high-frequency groups were produced by a single oscillator. One of the
capacitances in each group was changed as the various buttons were depressed to
produce the required frequencies. Later versions used an integrated-circuit amplifier
with an RC twin-tee feedback circuit to produce the tones. The basic configuration of
the circuit is shown in Figure 8.20.
Two conditions have to be met for oscillations to occur, according to the
Barkhausen criterion:
(1) The closed loop gain must be equal to unity. In practice, the loop gain must
be slightly larger than unity for sustained oscillation.
(2) The change in phase around the loop must be an integer multiple of
2p radians:
The classical RC twin-tee filter has values of Rs and Cs as shown in Figure 8.21(a)
and the amplitude and phase responses in Figure 8.21(b).
Its transfer function is given by
1 o2 C 2 R2
1 o2 C 2 R2 þ j4oCR
Rationalizing and equating the imaginary part to zero gives the frequency at which
the phase angle is zero or 180
o0 ¼
Figure 8.20. The modified twin-tee feedback oscillator. The closing of one of the four switches
connects a different R=a to produce the tones. Two such circuits are used in the dial, one for the
low frequencies and the second for the high frequencies.
Figure 8.21. (a) The classical twin-tee notch filter showing its configuration and circuit element
ratios. (b) The gain and phase characteristics of (a).
Under this condition the output voltage v2 ¼ 0; the circuit has a null in its frequency
response with a very high Q factor. The high Q factor can be exploited for high
stability of oscillating frequency if the oscillator is designed to operate at the
frequency of the null. However, using the classical twin-tee values of Rs and Cs in
oscillator design will be self-defeating since an amplifier with infinite gain will be
Departure from the standard ratios of Rs and Cs produces lower values of Q
factor. The amplitude–frequency responses of the modified twin-tee filter (Figure
8.20) with different values of a are shown in Figure 8.22(a). The frequencies at
which the phase shift of the filter is zero or 180
coincides with the frequency at
which the output voltage is a minimum (null). It can be seen that the notch frequency
for the particular configuration and circuit element ratios shown in Figure 8.20
changes with the parameter a. It can also be observed that the depth of the notch
varies as a changes, with the ‘‘deepest’’ notch occurring when a is equal to
approximately 2.25. For values of a less than 2.25, the modified twin-tee circuit
has gain and phase characteristics as shown in Figure 8.22(b). When a is equal to or
greater than 2.25, the gain and phase response is as shown in Figure 8.22(c). The
oscillator design technique described here therefore works only when a is greater
than 2.25.
The oscillator design exercise consists of identifying the value of a which has its
notch at the required frequency. The gain needed from the amplifier at the null or
where the 180
phase shift occurs can be identified. The amplifier can then be
designed to have the required gain and a phase shift of 180
Figure 8.22. (a) The gain response of the modified twin-tee for various values of a. (b) The
gain and phase characteristics of the modified twin-tee when a is less than 2.25. (c) The gain and
phase characteristics of the modified twin-tee when a is greater than 2.25.
Figure 8.22. (continued ) Digital Tone Dial. The digital tone dial attempts to exploit the low cost
of digital integrated circuits. The design of the system is shown in Figure 8.23.
The crystal-controlled oscillator generates a signal at 3.58 MHz. This frequency
was chosen because a crystal designed to operate at that frequency was readily
available and inexpensive (it is used in the color burst carrier of television sets; see
Section The actual frequency is not important so long as it is sufficiently
high that division by an integer will produce frequencies which lie within the
permitted error margin of the tone frequencies. The oscillator frequency is first
divided by 16 to give a clock frequency of 223.75 kHz. The push-button pad has a
number of contacts which are used to send a binary logic statement of 1s and 0s to
the N coder. The function of the N coder is to generate its own set of 1s and 0s as
input to the divide-by-N . The divide-by-N consists of a set of resetable binary
counters with additional logic circuitry to reset the counters so that the number N
can be changed according to the output of the N coder. The function of the eightstage shift register is to produce eight sequential pulses at the output which have
1=8th the period of the required sinusoidal signal. These pulses are used to drive the
digital-to-analog converter to produce a crude eight-step approximation to the sine
wave. The low-pass filter attenuates the unwanted harmonics before the signal drives
the telephone line through the line driver. Except for minor differences in the N
coder, the two halves of the circuit are the same. The basic circuit blocks used are the
NOR, NAND, NOT or inverter and the resetable bistable multivibrator, known
collectively as logic gates. These gates are used in large quantities but each gate
occupies such a small area on an integrated circuit chip that the cost is minimal. For
example, this circuit had 10 NOR, 4 NAND, 44 NOT, and 23 resetable bi-stable
Logic gates can be realized in different forms, each with its own mode of
operation. They are grouped into families such as complementary metal oxide
semiconductor (CMOS), transistor–transistor logic (TTL), resistor transistor logic
Figure 8.23. The block diagram of the digital tone dial.
(RTL), and so on. In explaining the design of logic gates, the logic family whose
mode of operation appears to the simplest is chosen.
Before discussing the design of the digital tone generator it is best to digress to
explain the design of the basic building blocks shown in Figure 8.23.
Design of the NOT gate. The basic NOT gate is shown in Figure 8.24(a). When
the input is a 0 (or essentially zero volts), the transistor is cut off and no current flows
through the collector resistor Rc. The collector is then a 1 which is equivalent to the
system voltage Vcc . When the input is a 1, current flows through the resistor R1 and
forward biases the base-emitter junction. Collector current flows and, given an
appropriate value for Rc, the transistor goes into saturation, that is, the output
is a 0.
Two components are added on as shown in Figure 8.24(b) to improve the
performance of the gate. A capacitor is connected across R1 . Since the voltage
across a capacitor cannot change instantaneously, the leading edge of a positive
pulse will cause the base voltage to rise immediately, causing the transistor to
conduct with a minimum of delay. A capacitor used for this purpose is called a
‘‘speed-up’’ capacitor. The second component is the resistor R2 which is connected
to a dc source VBB . The purpose of this circuit is to ensure that the base-emitter
voltage is normally kept at a slightly negative value so that the probability of the gate
switching due to noise is reduced.
The design of this circuit is best illustrated by an example:
Example 8.4.1 The NOT-gate. A NOT gate drives a load which requires
IL ¼ 1 mA. The dc supply voltage Vcc ¼ 10, VBB ¼ 5 V and the transistor is a
silicon NPN bipolar with a b ¼ 100. Determine suitable values for R1, R2 and Rc .
Figure 8.24. (a) The basic NOT gate and its transition diagram. (b) The NOT gate with ‘‘speedup’’ capacitor and negative biassing to improve noise immunity.
Solution. To prevent the load current from interfering with the operation of the gate,
it is necessary to make the collector current about 10 times the load current
IC ¼ 10 IL ¼ 10 mA:
The transistor goes into saturation when the voltage drop across Rc is equal to
VCC ¼ 10V:
RC ¼ VCC =IC ¼ 10 V=10 mA ¼ 1 kO:
Base current required for 10 mA collector current is
IB ¼ IC =b ¼ 10 mA=100 ¼ 100 mA
R1 ¼ ðVCC VBE Þ=IB ¼ ð10 0:7Þ=100 mA ¼ 93 kO:
In practice, the value of R1 may be reduced to half the calculated value to increase
the margin of safety.
To improve the noise immunity, let the base voltage be reverse biassed by 3 V
when the input is 0 V. This means that R2 has to be chosen so that
R2 =R1 ¼ 2=3
R2 ¼ 62 kO:
It is necessary to verify that, when the input is at 10 V (VCC ), R2 and VBB will not
hold the base voltage below 0.7 V. This step may be carried out by considering the
transistor to be disconnected from the rest of the circuit. The voltage at the node of
R1 and R2 can then be calculated. Any voltage greater than 0.7 V ensures that the
transistor will indeed be biassed on when the input is a 1. In this case the base
voltage would have been 1.0 V but the base-emitter diode will hold it at 0.7 V.
The value of the speed-up capacitor depends on the frequency of operation of the
gate. A reasonable choice is 50–100 pF.
Design of the NOR gate. Figure 8.25 shows a two-input NOR gate with its truth
When both inputs are 0s no current flows in either transistor and the output is a 1.
When either A or B is a 1, current flows in the corresponding transistor which goes
into saturation and the output is a 0. Finally when both inputs are 1s both transistors
conduct and the output is a 0.
The NOR gate may be viewed as two NOT gates sharing a common collector
resistor. The design of the NOR gate follows the same procedure as the NOT gate.
Design of the NAND gate. A modification of the NOT gate by the addition of an
extra resistor and two diodes gives a NAND gate. This is shown in Figure 8.26
together with its truth table.
Figure 8.25. The NOR gate with its truth table.
When either A or B or both are 0s, current flows from Vcc through R3 and the
appropriate diode(s) to ground. The voltage at node X is held at 0.7 V and Q remains
in cut-off. Only when both A and B are 1s can base current flow through R3 and R1
and cause Q to go into saturation, producing a 0 at the output.
(The design of the NAND gate follows from that of the NOT gate. The procedure
for calculating the values of Rc , R2 and C remain the same. R1 from the NOT gate
Figure 8.26. The NAND gate and its truth table.
example has to be split in two to form R3 and R1 as shown in Figure 8.26. A
reasonable compromise might be to make R3 ¼ 33 kO and R1 ¼ 60 kO. The circuit
that drives the NAND gate will then not have to sink large currents through the
Design of the Bistable Multivibrator. The easiest way to understand the operation
of the bistable multivibrator is to consider it to be two NOT gates connected back-toback as shown in Figure 8.27.
When the dc power is first switched on, suppose Q1 draws more current than Q2 .
The collector voltage of Q1 will therefore start to drop faster than that of Q2 . The
base-emitter junction of Q2 will be a little less forward biassed and its collector
current will be reduced at a rate determined by the current gain of Q2 . Its collector
voltage will therefore tend to rise pulling up the base of Q1 , thereby causing the
base-emitter junction to become more forward biassed. More current will then flow
through the collector of Q1 due to the current multiplication in the transistor, causing
the collector voltage to drop faster than before. The event that started the process has
become magnified many times over. This is a regenerative process and comes to a
stop when Q1 goes into saturation and Q2 is cut off. Unlike the astable multivibrator
described in Section 7.2.7, there is no mechanism to cause the two transistors to
switch back and forth. Q1 stays in saturation and Q2 is cut off until some external
event causes them to change states. This event could be a small positive-going pulse
or trigger applied to the base of Q2 . If the trigger is sufficient to bring Q2 into partial
conduction momentarily, this will cause current to flow in the collector of Q2 . Its
Figure 8.27. The basic bistable multivibrator.
collector voltage will drop, causing the base-emitter junction of Q1 to be less
positively biassed. Less collector current will flow and the collector voltage of Q1
will start to rise causing the base-emitter junction of Q2 to become even more
forward biassed. Eventually, Q2 goes into saturation and Q1 is cut off. This is the end
of the regenerative cycle in the opposite direction and the circuit will remain in the
current state until another external event causes it to change states.
The bistable multivibrator can be made much more interesting and useful by the
addition of the ‘‘steering circuit’’ shown in Figure 8.28.
The diodes D1 , D2 and D3 are connected and their common node is coupled to
a source of negative-going clock pulses by the capacitor C3. Assuming that Q1 is
conducting, its collector voltage will be essentially at ground potential while that
of Q2 will be at Vcc . When the negative-going pulse arrives, D2 will conduct
momentarily causing the voltage at the collector of Q2 to drop. The regenerative
cycle is set in motion and the cycle ends with Q1 off and Q2 on. The next negativegoing pulse will cause the transistors to change states yet again. This is illustrated in
Figure 8.29.
It should be noted that the diode D3 guarantees that static charge build-up on the
capacitor C3 cannot exceed Vcc þ 0:7 V.
Figure 8.28. The bistable multivibrator with the steering diodes is called a flip-flop.
Figure 8.29. A timing diagram for the flip-flop showing the input trigger and the collector
Two very important conclusions can be drawn from Figure 8.29:
(1) The frequency of the output signal taken from either transistor is one-half of
the input frequency. This is a divide-by-two circuit.
(2) The circuit has a memory; if the current state is known, one can deduce the
previous state.
In terms of the current discussion, the first conclusion is the more relevant. By using
two or more of these bistable multivibrators in cascade, it is possible to get a divideby-four (22 ), three of them will perform a divide-by-eight (23 ) and so on. In Figure
8.23, a cascade of four bistables are used to realize the divide-by-sixteen. The
bistable multivibrator with its steering circuit is popularly known as a flip-flop.
A set and a reset feature can be added to the circuit whereby the state of the flipflop can be set before the arrival of a chain of pulses. For example, a 1 can be applied
momentarily to the base of Q1 to ensure that it is conducting current. It is then
possible to predict the state it will be in after any given number of input pulses. A
flip-flop with the set and reset features is called an SR flip-flop.
Yet another useful feature can be added so that, when the clock pulse arrives, the
flip-flop switches states only when a 1 or a 0 is present at a designated terminal. The
flip-flop is then called a J K flip-flop. Flip-flops are discussed in almost any
textbook on digital circuits and the interested reader is advised to refer to such texts
for more details.
The Crystal-controlled Oscillator. The crystal-controlled oscillator is shown in
Figure 8.30.
With the two inputs to the NOR gate tied together, it becomes a NOT gate which
is, in effect, a high gain amplifier with phase inversion. The crystal, together with the
Figure 8.30. The complementary metal-oxide semiconductor (CMOS) transistor version of the
crystal-controlled oscillator. The output is a square wave.
two capacitors provide the feedback path. The output waveform is not quite a square
wave. The two inverters in cascade amplify and clip the signal to give a good square
The Divide-by-16 Circuit. As mentioned earlier, this is a cascade of four flip-flops
connected in a binary ‘‘ripple counter’’ arrangement. It divides the crystal oscillator
frequency by 16 to give an output of 223.75 kHz.
The N-coder Circuit. The N -coder uses the information from the dial pad to set up
a pattern of 1s and 0s and sends these to the divide-by-N circuit. The particular
pattern decides what the value of N has to be. The N -coder uses only NOR and NOT
The Divide-by-N Circuit. A combination of the set and reset features and the
exploitation of other logic circuits such as NOR, NAND, and NOT gates provide the
possibility to reset a cascade of flip-flops after n pulses of input where n takes on
values other than 2m , where m is an integer. This is the scheme used in the divide-byN circuit with the N -coder supplying the necessary 1s and 0s to the appropriate flipflops to set and=or reset them.
Using the lowest frequency (697 Hz) as an example, the input frequency to the
divide-by-N circuit is 223.75 kHz. The selected value of N is 40. This gives an
output frequency of 5.594 kHz which is used to clock the eight-stage shift register.
The Eight-Stage Shift Register. The eight-stage shift register is made up of a
cascade of flip-flops, NOR, and NOT-gates to set up a sequence of pulses which have
a width equal to 1=8th the period of the required frequency. One-eighth the period is
used because the required sine wave is to be approximated by an eight-step function
as shown in Figure 8.31.
Figure 8.31. The eight-stage shift register closes each switch in turn. The resistor values are
chosen to produce eight current levels which approximate a sinewave.
The Eight-Level Digital-to-Analog Converter. The sequential pulses from the
shift register are used to drive a bank of eight field-effect transistors switches
connected in parallel as shown in Figure 8.32.
The output is an eight-level approximation to a sinusoidal current. Going back to
the example (under divide-by-N circuit), the fundamental frequency of the approximation to the sinusoid is 5.594=8 ¼ 699.3 Hz. The error is 0.3%.
The Low-Pass Filter. The low-pass filter attenuates the higher harmonics of the
waveform before the signal reaches the line driver. With the exception of very short
lines this filter is not required; the telephone line is a good enough low-pass filter.
The harmonic content can be reduced by using more than eight steps to represent the
Table 8.1 shows the target frequency f0 ; the required divisor n, the nearest integral
divisor N, the actual frequency f00, and the error for all 12 tone-dial signals. The
errors indicated here are within the 1:5% allowed.
Until the introduction of the digital telephone all signals from the station set were
transmitted to the central office in analog form over the subscriber loop, which in
Figure 8.32. The switches in Figure 8.31 are replaced by FETs.
most cases, is the twisted pair of plastic-insulated copper wire. In the central office it
may be subjected to a number of sophisticated signal processing techniques,
depending on the routing of the message and the medium of transmission. In the
digital telephone, the analog signal is converted into digital form by a codec (coderdecoder) using pulse-code modulation (PCM). The pulses are sent along the twisted
pair to the central office where they may be further processed before transmission to
their destination.
The Codec
The codec is available on an integrated circuit chip and it is installed in the station
set. Figure 8.33 shows a possible configuration of the system.
In this configuration, the analog part of the telephone is left unchanged. The
physical distance from the telephone hybrid to the 2-to-4 wire hybrid can vary from
Target frequency
f0 (Hz)
16 8f0
f0 (Hz)
Error %
Figure 8.33. The block diagram of the codec used in the digital telephone.
essentially zero, when the codec is installed in the station set, to a few kilometres,
when the codec is in the central office. In the transmit mode, the analog signal is fed
to the input of the analog-to-digital (A=D) converter which gives a set of binary
outputs in PCM in a parallel format. A parallel-to-serial (P=S) converter changes the
parallel format to a serial format so it can be sent down the transmit leg of the digital
line. In the receive mode, the input to the decoder is in PCM serial mode and the
serial-to-parallel converter changes the signal into the parallel mode before going
onto the digital-to-analog (D=A) converter. The analog output then goes to the
receiver. The design of the circuits in the boxes shown in Figure 8.34 now follow.
The order of presentation has been changed to take advantage of the fact that an A=D
converter makes use of a D=A converter. Digital-to-Analog Converter. As the name suggests, this circuit takes
a digital input and converts it to the equivalent analog output. The digital input is
presented in the form of 1s and 0s from the least significant bit (LSB) to the most
significant bit (MSB). For example the number 26 is presented as:
Figure 8.34. Details of the parallel-to-series PCM converter for both transmit and receive
Consider the operational amplifier connected as shown in Figure 8.35.
This is a summing amplifier (Section whose output is given by
v0 ¼ V þ R2
½a0 20 þ a1 21 þ a2 22 þ a3 23 þ :
The coefficients an are digital bits which are either 0 or 1 when the switch Sn is down
or up, respectively. The voltage V þ is the reference voltage and R1 and R2 are chosen
to give a suitable level of signal at the output.
One obvious disadvantage of the circuit is that the larger the number of bits
presented, the larger is the ratio of the largest resistor to the smallest. In integrated
circuits, large resistors occupy large areas of the chip and are therefore undesirable.
However, the function of the R1 resistors is to scale the current flowing into the
summing point of the amplifier according to the significance of the bit. Current
scaling can be achieved by changing the circuit as shown in Figure 8.36.
The switches are either connected to the inverting input of the amplifier or they
are connected to ground but the inverting input is a virtual ground so the resistive
network can be redrawn as the ladder shown in Figure 8.37.
Looking to the right at point X , the resistance seen is R1 . Again looking to the
right at point Y , the resistance seen is still R1 . It follows that looking at all points
corresponding to X gives a resistance R1 and hence the input resistance of the ladder
Figure 8.35. The circuit diagram of the digital-to-analog converter.
Figure 8.36. An improved version of the digital-to-analog converter which uses the R 2R
is R1 . This type of ladder is described as R2R ladder. Consider the current flowing
into the ladder, I . It follows that I will divide equally at node 1 as shown. At node 2,
the process is repeated with the current I =4 flowing in each branch so that at node n
the current in each branch is I =2n . All the currents are summed at the input of the
operational amplifier to give
Ii ¼
V þ vi
½a1 21 þ a2 22 þ a3 23 þ :
vi v0
Ii ¼
Figure 8.37. The R 2R ladder.
When the operational amplifier gain is very large
v0 ¼ V þ R2
½a1 21 þ a2 22 þ a3 23 þ R1
where the as determine whether a particular bit is a 1 or a 0. Note that the error in the
conversion is 12 LSB (Least Significant Bit). Analog-to-Digital Converter. Several different strategies have been
used in the design of A=D converters, each with its own advantages and disadvantages. For this discussion, the Ramp Counter A=D converter has been chosen
because its operation is relatively easy to understand. The ramp counter A=D is
shown in Figure 8.38.
Two inputs are applied to the comparator; the analog signal and a voltage from
the resistive network. The comparator output is a 0 if the analog signal is greater than
the voltage from the resistive network and a 1 if the opposite is true. The comparator
output and a clock drive a two-input NAND gate so that the NAND gate permits the
clock pulses to reach the resetable binary counter only when the comparator output
is a 0, that is, the NAND gate input is a 1 due to the inverter. The binary counter
counts the clock pulses and simultaneously closes or opens the switches S1 to Sn ,
depending on the bit presented (1 or 0) that it applies to the switches. Which
switches are closed or open determine the output voltage of the resistive network.
This is in fact a D=A converter! So long as the output voltage of the resistive network
is lower than the input analog voltage, the binary counter continues to count and
continues to increase the output voltage of the resistive network. Finally, the voltage
from the resistive network exceeds the analog input voltage and the comparator
output changes to a 0. The NAND gate stops the clock pulses from reaching the
binary counter and it therefore stops counting. The output from the binary counter
can now be read. The convert enable clears the registers of the binary counter ready
for the next sample of the analog input voltage to be converted.
The design of the NOT gate, the NAND gate, the binary counter, and the switches
were discussed in Section 8.4.5. The design of the resistive network plus operational
amplifier (D=A converter) was the subject of Section The comparator, the
parallel-to-serial convertor, and the 4-to-2 wire hybrid will now be discussed. Comparator. The circuit of the comparator is shown in Figure 8.39(a).
The ideal operational amplifier is assumed to have infinite gain so that, if the input
varies infinitesimally about 0 V, a positive-to-negative change will cause the output
to go from saturation at the negative dc supply voltage to the positive. The
operational amplifier is therefore a comparator which compares its input to 0 V.
To stop the output from saturating at the dc supply rails, a Zener diode may be
connected as shown in Figure 8.39(a). For a very small positive voltage input, the
output will go negative but it will be clamped when it drops to 0:7 V by the
forward biassed Zener diode. Equally, a very small negative voltage input will cause
the output to go positive but again it will be clamped, this time by the Zener voltage
Figure 8.38. The block diagram of the analog-to-digital converter. Note that it contains within it the
digital-to-analog converter.
Figure 8.39. (a) The zero-voltage comparator and its output characteristics. (b) The comparator for vi and vi0 and its output characteristics.
Vz of the diode. Evidently the clamping voltages were chosen so that a 0 0:7
volts and Vz V þ ¼ 1 – the system voltage. The input–output characteristics are
given in Figure 8.39(a).
Now consider the circuit in Figure 8.39(b) where there are two input voltages vi
and v0i connected by equal resistances R to the inverting input of the operational
amplifier. Evidently, the only way to keep the output from going positive or negative
(clamped condition) is that the current supplied by vi must be equal to the current
absorbed by v0i : The two voltages must be equal in magnitude. When they are not
equal, which they are not most of the time, the output is clamped at one extreme or
the other (1 or 0). The input–output characteristics show the effect of v0i as an offset
voltage. In the context of the A=D converter, the output of the comparator changes
state when the output of the resistive network exceeds the analog input voltage by, at
the most, the value of the LSB. Parallel-to-Serial Converter. The output of the A=D converter, which
is in pulse-code modulation (PCM parallel), has to be converted into serial form so
that it can be sent along a single communication channel. A universal asynchronous
receiver=transmitter (UART) performs this function. This is an elaborate, largescale-integration (LSI) integrated circuit which can convert data from parallel to
serial and vice versa. However, the basic operation of an eight-digit parallel-to-serial
converter is illustrated in Figure 8.40. The parallel information is available at the
nine (eight plus the start=stop) terminals. The moving switch sweeps past and
‘‘reads’’ the bits in serial form. Note that one or more of the bits can be used for the
Figure 8.40. A mechanical analog of the parallel-to-serial converter.
‘‘stop=start’’ code to tell the receiver when a segment of the code starts and stops.
Other bits may be used as a parity code for correcting the signal code in the presence
of noise. Examples of parallel-to-serial shift registers are the MC14O14B (Motorola
Semiconductor), the MM54HC165 (National Semiconductor Corp.) and the
SN54166 (Texas Instruments) integrated circuits and are listed in most integrated
circuit handbooks. Serial-to-Parallel Converter. The operation of the serial-to-parallel
converter is much simpler than its opposite number. Figure 8.41 shows an eightstage serial-to-parallel converter.
The stop=start pulse is used to clear the register and to trigger the input enable
circuit after a specified delay. The serial input data can then start to enter the shift
register moved on by the clock. The arrival of the next stop=start pulse signals the
end of the data stream. The binary outputs are read and the shift register is cleared
for the next stream of data to arrive. An example of an integrated circuit serial-toparallel converter is the SN54164 (Texas Instruments). 4-to-2 wire Hybrid. The basic 4-to-2 wire hybrid is shown in Figure
There are two identical but separate transformers, T1 and T2, connected to
conform to the dot notation as indicated. When the number of turns in each coil
Figure 8.41. A block diagram of the serial-to-parallel converter.
Figure 8.42. The 4-to-2 wire hybrid. Reprinted with permission from, Transmission Systems
for Communication, 5th Ed., AT&T, Bell Labs, 1983.
is as shown, a voltage source V1 in series with an impedance Z1 connected to port 1
will cause a voltage equal to
N1 1
to appear across the other two coils of T1 , each of which has N2 turns. Assuming that
Z1 ¼ Z2 and Z3 ¼ Z4, it follows that the current Ia will be equal to Ib and they will
flow in the directions indicated. In transformer T2, the magneto-motive force (mmf)
of the two current will cancel each other and the net voltage across port 2 will be
zero. The signal input power from port 1 is divided equally between ports 3 and 4. If
the voltage source had been connected across port 2 instead of port 1, the same effect
would be observed except there will be no voltage across port 1.
Consider the situation when a voltage source is connected across port 3. The
voltages across the two halves of port 4 will cancel each other. However, voltages
2N2 3
will appear across ports 1 and 2.
This arrangement can be used as a station set hybrid if the transmitter is
connected across port 1, the receiver across port 2, and the line across port 3 so
long as an impedance equal to that of the line is connected across port 4. Its normal
application, however, is as a 4-to-2 wire hybrid, the four wires being connected to
ports 1 and 2 and the two wires being connected to port 3.
Manual Office
The manual central office is largely historical although one sometimes hears stories
of a few survivors in very remote places. A description of its operation was given in
Chapter 1. It is evident that it was just a matter of time before automation would
eliminate the telephone operator’s job. The surprise was that it was not the human
errors due to voice communication problems, delay in getting service, or rising labor
costs that was the motivation for automation; it was the lack of security of the
message (see Section 1.5).
Basics of Step-by-Step Switching
As described in Section, the request for service from the telephone subscriber
to the central office comes in the form of a series of current pulses. These current
pulses are at a frequency of 10 Hz with a variable, but certainly much longer, inter-
digit pause. The engineering problem was to use these current pulses to move a
mechanical switch into the required position to effect the connection.
It is most convenient to start a description of the process with a small and
fictitious central office with 10 subscribers, each with the pulse dial described in
Section Under these conditions, switching would take place when a single
string of dial pulses arrived from the calling subscriber. Figure 8.43 shows an
electromechanical contraption that would perform the switching function. The
serrated wheel is attached to the switch wiper arm and it is spring loaded to keep
it in its rest position as shown.
When a pulse is applied to the electromagnet A, the pawl pulls the serrated wheel
the appropriate distance to cause the switch wiper arm to move from one contact to
the next. The wheel is prevented from returning to its rest position after every pulse
by the detent and the electromagnet B.
Consider the circuit shown in Figure 8.44. To initiate a call, the subscriber lifts
the handset, thereby closing the hook switch. Direct current flows in the coil of
electromagnet B and operates the detent to stop the serrated wheel from rotating in a
clockwise direction; it is, however, free to rotate in an counter-clockwise direction.
The subscriber now operates the dial and short-circuits the line seven times, say. The
ac current generated flows through electromagnet A and causes it to operate seven
times, moving the switch wiper arm onto the seventh fixed contact. There is no
known telephone system that operates this way! The conversation can now proceed.
When the call is over, the replacement of the handset opens the dc current loop, the
Figure 8.43. An electromechanical switch for a ten-subscriber central office.
Figure 8.44. The switch shown in Figure 8.43 connected to the station set.
electromagnet B is de-energized, the detent moves away from the serrated wheel and,
under the force of its return spring, the switch wiper arm returns to its rest position.
Each of the ten subscribers would be connected to such a switch and all ten
switches connected as shown in Figure 8.45.
A switching system such as this could be extended to accommodate 100
subscribers by making the following changes:
(1) Dialling will now consist of two trains of pulse with a suitable pause between
pulse trains.
(2) Each of the 10 selector switches will be connected to 10 other identical
(secondary) switches.
Figure 8.45. The complete ten-subscriber telephone system.
Figure 8.46. The Strowger switch showing horizontal and vertical banks as well as wipers.
Reprinted with permission from B. E. Briley, Introduction to Telephone Switching, AT&T, Bell
Labs, 1983.
(3) Some electrical or mechanical arrangements will be made to transfer the
second train of pulses to the bank of secondary switches.
Further extension to 1000 and 10,000 subscribers is possible by suitable changes in
the central office and the introduction of three- and four-digit dialling codes
respectively. Such a system is possible and would operate quite well, except for
the fact that it would be very large, prohibitively expensive to manufacture, install,
and operate.
The Strowger Switch
If you can imagine a new version of the fictitious electro-mechanical contraption
shown in Figure 8.43 with movement in two directions, then you have a good idea of
how the Strowger switch works. The actual Strowger switch is shown in Figure 8.46.
There are two sets of electromechanical drives: one vertical and the other
horizontal. There are ten vertical positions and ten horizontal ones. Such a switch
can accommodate up to 100 subscribers and therefore requires two trains of pulse.
The first moves the switch wiper upwards – to the third level if there were three
pulses in the train. During the inter-digit pause, control is switched to the horizontal
drive. The second train of pulse move the switch wiper horizontally to the seventh
position – if it had seven pulses, to connect to the number 37. At the end of the
conversation, the switch must be returned to its initial position.
Replacing the 10-position switches shown in Figure 8.45 with Strowger switches
would convert a 10-line office into a 100-line office. It would appear at first that the
two-dimensional switch would solve the economic problem; in fact, it is more
complicated than the 10-position switch and therefore more expensive. The solution
to the economic problem was discovered along with the fact that it is most unlikely
that in a 100-line office 50 calls will be in progress at the same time. It was therefore
not necessary to design the system for maximum capacity operation. The study of
traffic statistics revealed that the system could be designed to handle a maximum
number of calls (usually about 10% of subscribers) and if it happens that there is one
more than this maximum then service would be denied. This led to a redesign of
the central office so that the subscribers would share a considerable amount of
common equipment. Every new incoming call initiates a hunt for an idle path
through the system and, if no idle paths can be found, a busy signal is returned. The
precise details of the inner workings of the central office are not necessary since
the object of the exercise is to develop a general appreciation of how a call is
connected. Figure 8.47 shows a simplified connection using line finders, selectors,
and connectors.
Basics of Crossbar Switching
The basis of the crossbar switching system is a fallback to the manual central office
and is shown in Figure 8.48.
In a hypothetical nine-subscriber central office, each subscriber is represented in
the vertical and horizontal elements of the matrix. To connect any two of the
subscribers it is necessary to have 81 contact points. In the crossbar system, it
actually takes the operation of two relays to close one contact point – one for the
vertical element and the other for the horizontal element. As in the other central
office organization schemes, it is possible to reduce the number of contact points by:
(1) Reorganizing the interconnection as shown in Figure 8.49(a). This shows that
the number of contacts have been reduced to 54.
(2) Since it is most unlikely that every subscriber will be using her telephone at
the same time, let the maximum number of conversations passing through
this office be three. By using three links as shown in Figure 8.49(b), it is still
possible to connect every subscriber to every other subscriber so long as no
more than three calls are in progress simultaneously.
The process of using these and similar schemes to reduce the number of switches is
referred to as concentration.
Figure 8.47. A simplified connection through the step-by-step central office. Reprinted with permission from B. E. Briley, Introduction to Telephone Switching, AT&T, Bell Labs, 1983.
Figure 8.48. A representation of the cross-bar switch for a nine-subscriber central office
showing all 81 possible connections. Reprinted with permission from S. F. Smith, Telephony
and Telegraphy A, 2nd Ed., Oxford University Press, London, 1974.
Central Office Tone Receiver
So far the discussion of switching in the central office has been based on the 10 Hz
pulse signal generated by the rotary dial or imitations of it. The use of tone dialling
in the telephone system is becoming increasingly more important because it is faster
and it offers the possibility to use the telephone system for purposes other than
talking. There are already a number of household appliances and office equipment
on the market which can be connected to the telephone system and controlled from a
remote location by dialling special codes.
As discussed earlier, the tone dial generates two distinct frequencies when a
single button is pressed. By identifying the two frequencies present the number of
the button can be decoded. The system for decoding the signal is shown in Figure
The tone signal goes through a band split filter which separates the low-frequency
group from the high one. Each group is amplified and hard limited to produce square
waves. The signal is filtered again by the channel filters which are tuned to the
frequencies indicated. The output signals from the channel filters are rectified by the
detectors to indicate the presence (1) or absence (0) of the frequencies produced by
the dial. For an output to be valid, it has to have two frequencies present, one from
each frequency group. The system has several built-in features which prevent
mistaken identification of other audio signals as valid dial signals.
Figure 8.49. (a) A switching scheme for a nine-subscriber central office with the number of contacts
reduced to 54. (b) The use of three links which reduces the number of conversations that can be carried
out simultaneously to three also reduces the number of contacts to 27. Reprinted with permission from
S. F. Smith, Telephony and Telegraphy A, 2nd Ed., Oxford University Press, London, 1974.
Figure 8.50. The block diagram of the tone receiver at the central office.
Elements of Electronic Switching
The basis of electronic switching is the same as the crossbar technique but with the
electromechanical switches replaced by electronic components. The advantages
gained were speed of operation, increased reliability and lower cost. During its
development, several schemes were tried but none of them worked because the
replacement of the metal-to-metal contact in the electromechanical switches with
semiconductor switches proved to be unsuitable. The concept of electronic switching
had to undergo a fundamental change before success could be achieved. The
problem was that the crossbar technique was based on space-division multiplex
(SDM) – an approach that relied on finding multiple paths for the different
conversations passing through the central office by separating them in space, that
is, by assigning each path to a space or position in the matrix. Time-division
multiplex (TDM), which uses the same path for all conversations but separates them
in time, was more successful. Time-division multiplex is discussed in Chapter 9.
So far, in the discussion of what takes place in the central office from the
initiation of a call to its end, it has been assumed that the system was under the direct
control of the pulses or tones as they arrived. It is certainly advantageous to have a
memory in the system so that the information dialled in by the subscriber can be
stored, analyzed and an optimal path determined for the call. In the language of the
telephone engineer, the memory device is a register (see Appendix E). The pulses
from the dial are fed to a set of shift registers or decade counters, one set for each
decade of the number dialled. With the tone dial, the digits are first decoded and then
stored in binary form in a shift register. What we have then are banks of hard-wired
switches and a memory in which instructions can be stored, retrieved, and acted
upon in due course. To help drive this point home, consider the activities which take
place in the central office in order to connect a call. This is shown in the form of a
flow diagram in Figure 8.51.
Figure 8.51. A simplified flow chart for setting up a telephone call. The similarity to a computer
program flow chart is evident.
The flow diagram looks suspiciously like the flow diagram for a computer
program. It is indeed a computer program and the instructions can be carried out by
a general-purpose computer with the appropriate peripheral equipment. The modern
central office has a dedicated computer which controls every aspect of placing a call.
Humans simply supervise the machine and intervene only when something goes
wrong or when they work out a better set of instructions for the machine.
1. Fraser, W., Telecommunications, Macdonald & Co., London, 1957.
2. Plonus, M. A., Applied Electromagnetics, McGraw-Hill, New York, 1978.
3. Sadiku, M. N. O., Elements of Electromagnetics, Holt, Rinehart and Winston, New York,
4. Sessler, G. M. and West, J. E., ‘‘Electret Transducers: A Review’’, J. Acoust. Soc. Am., 53,
1589–1599, 1973.
5. van der Puije, P. D., ‘‘Audio Frequency Circulator for use in Telephone Sets’’, IEEE Trans.
Comm. Tech., 1267–1271, Dec. 1971.
6. Tellegen B. D. H., ‘‘A General Network Theorem with Applications’’, Philips Res. Rept., 7,
259–269, 1952.
7. Shenoi, B. A., ‘‘Practical Realization of a Gyrator Circuit and RC-gyrator Filters’’, Trans.
IEEE, CT-12, 374, 1965.
8. Sheahan, D. F. and Orchard, H. J., ‘‘High Quality Transistorized Gyrators’’, Elec, Lett., 2,
274, 1966.
9. Smith, S. F., Telephony and Telegraphy A, 2nd Ed., Oxford University Press, London, 1974.
10. Briley, B. E., Introduction to Telephone Switching, Bell Telephone Laboratories, Inc.,
11. Members of the Technical Staff., Transmission Systems for Communication, 5th Ed., Bell
Telephone Laboratories, Inc., 1983.
Describe the attractive features of the carbon microphone that make it well
suited to the telephone system. A carbon microphone has a static resistance of
75 O and it is connected in series with a 12 V battery and a load resistor R. A
sound wave impinging on the microphone causes a sinusoidal variation of the
microphone resistance with a peak value equal to 15% of the static value. What
is the value of the load resistor R if the second harmonic component of the
current is 7.5% of the current at the fundamental frequency? Calculate the
signal power delivered to R at the fundamental frequency and comment on it.
A successive approximation A=D converter has an output consisting of 8 bits.
The input to the converter is 0.82564 VR , where VR is the reference voltage.
The A=D converter is clocked at 1 MHz.
(1) Determine the total error in the conversion.
(2) Calculate the acquisition time.
If a change is made in the A=D converter so it counts from zero and advances
by the value of the LSB for every clock pulse (ramp counter A=D converter),
what would be the conversion error in the time given by (2) when the
LSB ¼ 0.001 VR ? How long will it take the ramp counter A=D to
achieve the same conversion accuracy as the successive approximation A=D
Design a bipolar transistor bistable multivibrator using a 12 V dc power supply.
The multivibrator is to drive a load which requires 1 mA. Indicate how you
would convert the multivibrator into a flip-flop. Show that the flip-flop
performs a divide-by-two function.
Using standard digital gates, design a circuit which will divide a clock
frequency by the following integer numbers: 7, 19, 25, 47, 77, and 92.
What is a shift register and how does it work? Illustrate your answer with a
design of a four-stage shift register. Describe two applications where a shift
register may be used.
What is sidetone and how can it be used to cultivate good speaking habits in
telephone users? Does sidetone control offer any other advantages? With the
aid of a suitable circuit diagram, describe the operation of a typical sidetone
suppression circuit.
Figure P8.1.
The circuit of Figure P8.1 shows a transformer hybrid in the receive mode
where Rm is the resistance of the microphone. Assuming that
ZL ¼ ZB ¼ 900 O, ZR ¼ 1000 O, and N1 ¼ N2 ¼ 200 turns, calculate the
power dissipated in ZR when Vin ¼ 1 V rms and the reactance of N1 is
1200 O at the frequency of operation. [Hint: transfer the load ZR to the
primary of the transformer].
In the design of digital circuits it is common to find two NOT gates connected
in cascade. What is the purpose of such a connection?
Using only two-input NOR gates, show how you can realize a:
(1) NOT gate,
(2) two-input OR gate, or
(3) two-input AND gate.
Until the introduction of the digital telephone, there was virtually no signal
processing on the subscriber loop. Indeed, there was no need for it. The majority
of subscriber loops were able to transmit voice signals with no particular difficulty
and in cases where the lines were longer than usual, line ‘‘loading’’ was used with
Signal processing has two major aims:
(1) To improve the quality of signal transmission over the telephone communication channels.
(2) To lower the cost of communication by improving the efficiency of channel
In general, the quality of a communication channel tends to deteriorate with distance.
In addition, long distance channels are expensive to establish and maintain. It
follows that the more messages that can be transmitted in a given time, the lower the
cost per message. It is therefore on the long-distance channels (trunks or tolls) that
signal processing techniques have proven to be most successful. In this chapter the
common signal processing techniques used in the telephone system and some of the
circuits employed will be examined.
Frequency division multiplex (FDM) is a technique in which a number of signals can
be transmitted over the same channel by using them to modulate carrier signals with
different and appropriate frequency so that they do not interfere with each other. The
assignment of specific carrier frequencies to radio stations for broadcasting and other
purposes is, in fact, FDM. It can be used with amplitude modulation as well as other
forms of modulation. In the context of the telephone, FDM is used in conjunction
with amplitude modulation.
In normal amplitude modulation, the carrier, upper and lower sidebands are
transmitted. When the depth of modulation is 100%, the amplitude of the carrier
voltage is twice that of the sidebands. The power in the carrier is therefore 23 of the
total. Unfortunately, the carrier has no information content. Each of the sidebands
contains 16 of the total power. In radio, 100% modulation is almost never used so the
power content of the sidebands is much less than described. It is noted that the
information content is duplicated in the two sidebands. It is clear that one way to
beat the corrupting influence of noise on the information content of the transmission
is to put as much as possible, if not all, of the available power into one of the
sidebands. An added advantage to this scheme is that the required bandwidth is
reduced to one-half of its original value. Clearly, this would allow twice as many
messages to be sent on the same channel as before. The transmission of only one
sideband in an AM scheme is called single-sideband (SSB) modulation. The price to
be paid for this advantage is that to demodulate an SSB signal, it is necessary to
reinstate the carrier at the receiver. The reinstated carrier has to be in synchronism
with the original carrier, otherwise demodulation yields an intolerably distorted
signal. Providing a synchronized local oscillator requires complex equipment at the
transmitter as well as at the receiver. In SSB radio, an attenuated form of the carrier
is transmitted with the signal. This is used to synchronize a local oscillator in the
receiver. In the telephone system, a centrally generated pilot signal is distributed to
all offices for demodulation purposes. In some cases, a local oscillator without
synchronization is used. If the frequency error is small (approximately 5 Hz),
successful demodulation can be achieved [7].
Generation of Single-Sideband Signals
A block diagram of the SSB generator is shown in Figure 9.1.
Figure 9.1. The block diagram for SSB modulation.
The signal and the carrier are essentially multiplied by the balanced modulator to
give a DSB-SC output. The bandpass filter removes either the lower or upper
f ðtÞ ¼ A cos os t cos oc t
f ðtÞ ¼ ½cosðoc þ os Þt þ cosðoc os Þt:
Assuming the upper sideband is eliminated, we get
f1 ðtÞ ¼
cosðoc os Þt:
To eliminate the upper sideband, it is necessary to have a bandpass filter with a very
sharp cut-off at the carrier frequency. This is not easy to achieve in practice, but the
task is made simpler when the modulating signal os has no low-frequency
components. Under this condition, crystal and electromechanical filters can be
designed to suppress the upper sideband. This is the case for a telephone voice
channel which is nominally from 300 to 3000 Hz.
From Equation (9.2.3) only the lower sideband is transmitted. At the receiving
end, the signal is demodulated (multiplied) by (the recovered) carrier, cos ot. The
result is
cosðoc os Þt cos oc t
f2 ðtÞ ¼ cos os t þ cosð2oc os Þt:
f2 ðtÞ ¼
A low-pass filter is used to separate the required signal at frequency os from that at
ð2oc os Þ.
Design of Circuit Components
The balanced modulator was discussed in Section Filter design is outside the
scope of this book but a representative list of books on filters is provided in the
bibliography at the end of Chapter 3.
Formation of a Basic Group
In the trunk or toll system, 12 channels form a basic group. The basic group is
formed by SSB modulation of 12 subcarriers at 64, 68, 72; . . . ; 108 kHz. These
carriers are generated from a 4 kHz crystal-controlled oscillator and multiplied by
the appropriate factor. The upper sidebands are removed and they are added together
to form the group. Figure 9.2(a) shows a block diagram for channel 1. Figure 9.2(b)
shows the spectrum of the basic group.
Figure 9.2. Formation of the basic group with spectra. Reprinted with permission from
Transmission Systems for Communications, 4th Ed., AT&T, Bell Labs, 1970.
For a small-capacity trunk, the basic group may be transmitted without further
processing. The transmission channel can be a twisted pair or coaxial cable.
Formation of a Basic Supergroup
For higher capacity channels, five basic groups are combined to form a basic
supergroup. Figure 9.3(a) shows the block diagram of the basic supergroup 1. Note
that to make the filtering problem easier, the carrier frequency is chosen to be
420 kHz. Figure 9.3(b) shows the frequency spectrum of the basic supergroup.
Table 9.1 shows the carrier frequencies and bandwidths for each basic supergroup.
For a 60-channel trunk, the signal can be transmitted in this form. Again a twisted
pair with coil loading or amplification and coaxial cable may be the medium of
By organizing the 12 basic groups into a basic supergroup of 5 it clear that the
subcarrier frequencies, the balanced modulators. and bandpass filters can all be
duplicated five times over. If the basic group had been made larger, new subcarrier
frequencies would have had to be generated and bandpass filters of different
characteristics would have been necessary.
Formation of a Basic Mastergroup
To create a 600-channel trunk, 10 basic supergroups are combined to form a basic
mastergroup. The frequency spectrum of the basic mastergroup is shown in Figure 9.4.
Figure 9.3. Formation of the basic supergroup with spectra. Reprinted with permission from
Transmission Systems for Communications, 4th Ed., AT&T, Bell Labs, 1970.
Note that there are gaps of 8 kHz between each basic supergroup spectrum. These
gaps are designed to make the filtering problem easier.
The carrier frequencies and bandwidths of the 10 basic supergroups are given in
Table 9.2.
The basic supergroup can be transmitted over coaxial cable or it can be used to
modulate a 4 GHz carrier for terrestrial microwave transmission or even sent over a
satellite link.
Supergroup number
Carrier frequency (kHz)
Bandwidth (kHz)
Figure 9.4. Formation of the basic mastergroup with spectra. Reprinted with permission from
Transmission Systems for Communications, 4th Ed., AT&T, Bell Labs, 1970.
Basic supergroup number
Carrier frequency (kHz)
Bandwidth (kHz)
Other larger groups can be formed, for example 6 mastergroups may be combined
to form a jumbogroup with 3600 voice channels.
To recover the original baseband signals from the various groups, the appropriate
number of filtering=demodulation processes will have to be carried out. At each
stage of the demodulation process, the correct carrier will have to be reinstated for
this to be possible.
In FDM, voice signals were ‘‘stacked’’ in the frequency spectrum so that many such
signals could be transmitted over the same channel without interference. In timedivision multiplex (TDM), each voice signal is assigned the use of the complete
channel for a very short time on a periodic basis. The theoretical basis of this
technique is the Sampling Theorem. An informal statement of the sampling theorem
If the highest frequency in a signal is B Hz, then the signal can be reconstructed from
samples taken at a minimum rate of 2B samples per second (Nyquist sampling rate or
The proof of this theorem is beyond the scope of this book. However, there are a
number of practical problems which arise in the application of the theorem:
(1) The theorem assumes that the samples have infinitesimally narrow pulse
widths. This is clearly not so in a practical circuit. The sampling rate is
usually chosen to be higher than the Nyquist frequency since it is the
minimum; it is discrete to avoid extreme conditions when dealing with an
imperfect situation.
(2) The theorem assumes that an ideal low-pass filter is used to remove all
frequencies above B Hz ahead of the sampler. When using a practical filter, it
is necessary to sample the signal at a higher rate (oversampling) to avoid
distortion due to aliasing.
A TDM system with two input signals is illustrated in Figure 9.5. The samplers or
commutators are shown here as switches which are driven in synchronism.
The TDM system shown in Figure 9.5 is an example of pulse amplitude
modulation (PAM). Practical TDM systems based on PAM have been built and
used in the telephone system (No. 101 ESS-PBX) [8].
Pseudodigital Modulation
To code an analog signal in pulse form one can use the height of the pulse, the width
(or duration), or the position of the pulse relative to standard position. When the
height is used, it is called pulse-amplitude modulation (PAM). When the coding is in
terms of the width it is called pulse-width modulation (PWM) and when the position
is used it is called pulse-position modulation (PPM). Pulse height, width, and
position are analog quantities which in turn can be quantized and represented by a
binary code where the digits are present, 1, or absent, 0. When this has been done the
modulation scheme is called pulse-code modulation (PCM). Although PCM is
qualitatively different from the other modulating schemes, they are compared in
Figure 9.6.
These schemes would work equally well in a noiseless environment. When noise
is present, and it always is, PCM has a clear advantage over the others. In the case of
PAM, PWM and PPM the receiver has to determine what the original amplitude,
width, and position were respectively in order to reconstruct them. In PCM, the
decision is simplified to whether the digit sent was a 1 or a 0. In all cases, it is
necessary to transmit timing information with the signal so that the receiver knows
where the bit stream starts and stops.
Pulse-Amplitude Modulation Encoder
To illustrate the design principle of a PAM communication channel, a four-channel
PAM system has been chosen. The coder or commutator is shown in Figure 9.7.
The master clock drives the four-phase ring counter. The ring counter drives four
sampling gates on and off in the correct sequence. When one of the four outputs is
on, 1, all the others are off, 0; so only the sampling gate with the 1 is connected to
the adder. Note that the second input to the fourth sampling gate are connected to the
master clock. This means that channel 4 will always produce a positive pulse. The
amplitude of this pulse is adjusted to be higher than the most positive value of the
analog input voltage. This is called the synchronization pulse or sync pulse for short.
It is used to identify and time the other channels.
The design of the component circuits now follows. Four-Phase Ring Counter. The four-phase ring counter and its timing
diagram are shown in Figure 9.8.
Figure 9.5. A mechanical illustration of time-division multiplex (TDM) with pulse amplitude modulation
(PAM). Reprinted with permission from B. P. Lathi, Modern Digital and Analog Communication
Systems, CBS College Pub., New York, 1983.
Figure 9.6. A comparison of PAM, PWM, PPM and PCM. Note that PAM, PWM and PPM are
not truly digital since they convey information by the variation of analog quantities, that is,
amplitude, duration and position in time. Reprinted with permission from B. P. Lathi, Modern
Digital and Analog Communication Systems, CBS College Pub., New York, 1983.
Figure 9.7. A block diagram for a four-channel PAM system. Note that channel 4 is used for timing
Figure 9.8. A block diagram of a four-phase ring counter with its timing diagram.
It can be seen from the diagram that, in the time taken by one frame, the output
pulses go through one cycle. The outputs are used to drive the sampling gates. Series Sampling Gate. The configuration of the series sampling gate
is shown in Figure 9.9.
Figure 9.9. The series sampling gate using a FET.
The transistor is an open-circuit when the control signal is a 0 and a short-circuit
when it is a 1. The output is as shown. Shunt Sampling Gate. The shunt sampling gate is shown in Figure
The transistor acts as a switch and short-circuits the output when the gate voltage
is a 1. When the gate voltage is a 0, it is an open-circuit and a path exists between the
input and the output. Series-Shunt Sampling Gate. The two circuits shown above have an
inherent deficiency because, when the transistor is on, its source-to-drain impedance
is low but not equal to zero. To improve the performance, the action of the two gates
can be combined as shown in Figure 9.11.
Figure 9.10. The shunt sampling gate.
Figure 9.11. A combination of the series and shunt sampling gates improves performance. Operational Amplifier Sampling Gate. The circuit is shown in
Figure 9.12.
The operational amplifier is connected to give a gain of R2 =R1 when the transistor
is in the off state. When the transistor is on, R2 is short-circuited and the gain is
reduced to unity. Multiplier Sampling Gate. A PAM sampler can be seen as a multiplication of the analog signal and a train of pulses. The process is illustrated in
Figure 9.13.
One of the best methods for accomplishing analog multiplication is use the fourquadrant analog multiplier. This circuit was described in Section 2.6.3. A practical
integrated circuit realization of this is the MC1595 (four-quadrant multiplier
manufactured by Motorola Semiconductor Products, Inc.). The Adder. The adder was discussed in Section
Pulse-Amplitude Modulation Decoder
The first step in the recovery of the original three signals is to reverse the action of
the commutator by separating them into their respective channels. Low-pass filters
Figure 9.12. A sampling gate using an operational amplifier.
Figure 9.13. The multiplier sampling gate illustrates that sampling is equivalent to the multiplication of two signals.
are then used to reconstruct the analog waveform from the PAM pulses. The PAM
decoder system is shown in Figure 9.14.
The incoming signal is fed into the Schmitt trigger. The trip level of the Schmitt
trigger is set so that only the large sync pulse will trigger it. The output of the
Schmitt trigger is then used to synchronize an astable multivibrator. The astable
multivibrator then runs in synchronism with the master clock in the PAM coder. The
output of the astable multivibrator drives a four-phase ring counter which produces
four sequential output pulses in synchronism with the ring counter in the encoder.
These pulses are used to drive the control gates of the sample-and-hold (S=H)
circuits. The incoming signal is also fed to the analog inputs of the sample-and-hold
circuits. Since the control gate of only the S=H-1 is open during the period allotted to
channel 1, the pulse amplitude of the signal in channel 1 is passed onto S=H-1. The
other channels follow in sequence. The low-pass filters remove the high-frequency
components of the PAM signals, producing a replica of the original analog signals. Schmitt Trigger Design. The circuit diagram of the Schmitt trigger is
shown in Figure 9.15.
The circuit is designed so that, with no input, Q1 has no base current and it is
therefore off. Q2 is supplied with base current from the resistive chain Rc1 , R1 and R2
and it is therefore on. Current flows in Q2 and with the correct choice of Rc2 , Q2 will
Figure 9.14. A block diagram of the PAM decoder. Note that in Figure 9.7, the first three channels
had a signal and the fourth was used for timing purposes.
Figure 9.15. The circuit diagram of the Schmitt trigger.
be in saturation. The emitter current of Q2 flows in RE and sets up a voltage VE . If an
increasing positive voltage is applied to the input, when it reaches the value
VE þ VBE , Q1 will start to conduct. Current is drawn through Rc1 causing the
voltage on the base of Q2 to drop. Q2 conducts less vigorously and the voltage across
RE tends to drop. But this drop in voltage at the emitter causes the base-emitter
voltage of Q1 to increase rapidly. This is a form of regeneration and proceeds very
fast, ending with Q1 conducting and in saturation and Q2 cut off.
When the input voltage is decreasing, there comes the point when it is slightly
below the value VE þ VBE . Q1 conducts less current, causing its collector voltage to
tend to rise and its emitter voltage to tend to drop. The rising trend at the collector of
Q1 is passed onto the base of Q2 by R1 and C. The combined effect of a decreasing
VE and a rising base voltage causes Q2 to switch on regeneratively and go into
saturation. The Schmitt trigger reacts to a slowly changing input voltage by
producing a voltage step when its trip level is exceeded. This happens for increasing
as well as decreasing voltages. The design of the Schmitt trigger is best illustrated by
an example.
Example 9.3.1 Schmitt Trigger Design. Design a Schmitt trigger circuit so that it
triggers when a voltage in excess of 3.0 V is applied at the input. The dc supply
voltage is 10 V and two NPN silicon bipolar transistors with b ¼ 100 are provided.
VCEðsatÞ ¼ 0:5 V and VBE ¼ 0:7 V. The load driven by the Schmitt trigger requires a
current of 1 mA.
Solution. The transistors are made of silicon; therefore, VBE ¼ 0:7 V. For the circuit
to trigger at 3.0 V, VE must be designed to be equal to ð3:0 0:7Þ ¼ 2:3 V. Since the
Schmitt trigger is to drive a load that requires 1.0 mA, it is good design practice to
allow about 10 times this current to flow in the collector of Q2 . Collector current of
Q2 is then 10 mA; therefore, the emitter current is also 10 mA, and
RE ¼ 2:3=10 kO ¼ 230 O:
Since Q2 is in saturation, the collector voltage must be equal to ðVE þ 0:5Þ ¼
ð2:3 þ 0:5Þ ¼ 2:8 V. The voltage drop across Rc2 is ð10:0 2:8Þ ¼ 7:2 V. Since
10 mA flows through Rc2 ,
Rc2 ¼ 7:2=10 kO ¼ 720 O:
The base voltage of Q2 must be at the voltage ðVE þ 0:7Þ ¼ ð2:3 þ 0:7Þ ¼ 3:0 V.
The base current of Q2 is
IB ¼ IC =b ¼ 10=100 mA ¼ 100 mA:
To maintain a reasonably stable trigger point, it is necessary to allow about 10 times
the base current to flow in the resistive chain Rc1 , R1 and R2 . The current in the
resistive chain is therefore 1 mA and its total resistance is 10 V=1 mA ¼ 10 kO. Let
R01 be equal to ðRc1 þ R1 Þ so that
R01 =R2 ¼ ð10:0 3:0Þ=3:0 ¼ 2:33:
R01 þ R2 ¼ 10 kO:
R01 ¼ 7:0 kO
R2 ¼ 3:0 kO:
When Q1 is in saturation, it must draw the same current as when Q2 was in
Rc1 ¼ Rc2 ¼ 720 O:
R1 ¼ ð7:0 0:72Þ ¼ 6:28 kO:
The capacitor C is a speed-up capacitor which helps the transition of Q2 from the on
state to the off state. A reasonable value is 50–100 pF. Sample-and-Hold Circuit. The circuit diagram of the ideal S=H
circuit is shown in Figure 9.16.
The switch S closes and the ideal voltage source charges capacitor C instantaneously. When S opens, the capacitor retains its charge indefinitely. In practice, S is
Figure 9.16. The ideal sample-and-hold circuit.
not an ideal short-circuit when closed and the voltage source has an internal
resistance Rs . This means that C charges up with a time-constant t ¼ CRs . So
long as the pulse width of the driving source is significantly longer than t the error
can be regarded as insignificant. Another practical consideration is that, when S is
open, C can lose its charge due to leakage in the dielectric of the capacitor and the
finite impedance of the load driven by the S=H.
A simple but practical circuit for the S=H is shown in Figure 9.17.
The operational amplifiers are connected as voltage followers with a gain of unity.
The output impedance of A1 is low enough for it to drive the required charge into the
capacitor. The N -channel JFET is switched on by a pulse applied to the gate and the
capacitor charges up to the value of the input voltage. When the JFET switch is
turned off, the high input impedance of A2 drains minimal current from C. The
design of the S=H circuit is best illustrated by an example.
Example 9.3.2 Sample-and-Hold Circuit. The S=H circuit shown in Figure 9.17
uses a JFET as a series sampling gate. The voltage follower A2 takes a current of
500 nA. The JFET has a source-to-drain resistance of 25 O when it is in the on state
and may be considered to be an open-circuit when it is in the off state. The signal
Figure 9.17. A practical sample-and-hold circuit.
amplitude is 2.0 V, the sample-time is 5 ms and the hold time is 500 ms. The capacitor
has a value 0:4 mF. Calculate the error in the output at the end of the hold time.
Assume that the signal source has negligible resistance.
Solution. When the sampling gate is closed, assume that the current flowing into the
capacitor i 550 nA (the leakage current taken by A2 ):
t ¼ CRs ¼ 25 0:4 106 s:
For the RC circuit shown in Figure 9.18,
vc ¼ Vs ð1 et=t Þ:
When t ¼ 5 ms, and Vs ¼ 2:0 V, vc ¼ 1:987 V.
Charge on C at t1 ¼ 5 ms,
Q ¼ CvC ¼ 0:4 106 1:987 ¼ 0:795 mC:
Charge lost by C in time t2 ¼ 500 ms,
DQ ¼ I Dt ¼ 500 109 500 106 ¼ 25 1011 C:
Charge on C at time t2 ,
Q DQ ¼ ð0:795 106 25 1011 Þ ¼ 0:794 106 C:
Voltage across C at time t2 ,
vc2 ¼
¼ 1:986V
Percentage error in the output is 1.4%. Other Circuit Blocks in the Decoder. The synchronized astable
multivibrator was discussed in Section and the ring counter in Section The design of the filters is beyond the scope of this book.
Figure 9.18. The RC circuit for the Example 9.3.2 showing the exponential response.
Pulse Code Modulation Encoder=Multiplexer
As mentioned earlier signals coded in PAM tend to be susceptible to corruption by
noise and circuit non-linearities since the information is contained in the amplitude
of the pulses. Greater noise immunity can be obtained if the amplitude of the pulse is
coded in binary form. In the presence of noise and other forms of pulse degradation
it is much easier for the receiver to distinguish between the presence (1) or absence
(0) of a pulse as opposed to the height of a pulse. PCM is the preferred technique in
all modern telephone systems.
To illustrate the basic principles of the design, a four-channel PCM carrier system
has been chosen and it is shown in Figure 9.19(a). The four voice channels are
sampled in sequence so the sampler output is a time-division multiplexed PAM
signal. The PCM encoder converts the amplitude of each of the four samples into an
eight-bit binary code. Figure 9.19(b) shows the sampled analog signals in the fourchannel system, the interleaved PAM samples, and the PCM output. The 32 (4 8)
bits are transmitted over a twisted-pair telephone wire. One bit (frame bit) is added
for synchronization purposes; it may be wider than the other pulses or it may have a
specially coded sequence of 1s and 0s which can be easily recognized by the
Comparing Figures 9.7 and 8.34 with Figure 9.19 shows that the PCM is a PAM
system feeding into the coder of the codec (A=D followed by a P=S converter). The
design of the ring counter and the sampling gates were discussed in Section 9.3.2.
The A=D and P=S converters were discussed in Section 8.5.1.
Pulse-Code Modulation Decoder=Demultiplexer
The system diagram of the decoder=demultiplexer is shown in Figure 9.20. The
frame bit extractor is designed to recognize the frame bit and send out a pulse to
synchronize the local oscillator to the master oscillator in the PCM encoder. The
output of the local oscillator is used to drive the S=P converter, the A=D converter,
and the four-phase ring counter. The output from the D=A converter and the ring
counter are fed to the S=H circuits. If the timing is correct, the decoder reconstructs
the four samples from the eight-bit codes. The circuit assigns each sample to its
correct channel (demultiplex) and the outputs of the low-pass filters follow the
envelope of the PAM signal, producing a replica of the original audio-frequency
Comparing Figures 8.34 and 9.14 with Figure 9.20 shows that the decoder=demultiplexer is, in fact, the decoder of the codec followed by the PAM
demultiplexer. The design of the S=P and D=A converters was discussed in Sections and, respectively. The design of the circuit components for the PAM
demultiplexer was discussed in Section 9.3.3.
Bell System T-1 PCM Carrier
The Bell T-1 carrier system uses an eight-bit PCM in 24 voice-channel banks. The
sampling rate is 8 kHz. The number of bits generated for one scan of the channels
Figure 9.19. (a) The block diagram of a four-channel PCM system. Note that this is equivalent to a
PAM circuit feeding into the coder of the codec. Reprinted with permission from Transmission Systems
for Communications, 5th Ed., AT&T, Bell Labs, 1982.
Figure 9.19. (b) An illustration of the signals in the four channels, the interleaved PAM and the
8-bit PCM equivalent. Reprinted with permission from Transmission Systems for Communications, 5th Ed., AT&T, Bell Labs, 1982.
(frame) is 24 8 ¼ 192. One bit (frame bit) is required for synchronization so the
total number of bits per frame is 193. The analog signal is band-limited to
approximately 3.5 kHz and, because the sampling rate is 8 kHz, the bit rate is
(193 8000Þ ¼ 1:544 Mbit=s. The minimum bandwidth required to transmit the
signal is 1.5 MHz. The use of the eight-bit code means that the voice signals are
quantized at 256 (28 ) levels. Some of the less significant bits may be robbed and
used for signalling purposes such as dialling and detection of hook switch on=off.
The quantization error resulting from this is considered to be tolerable, although
several compression schemes are used to minimize its effect [9].
When the signal is sent over a twisted-pair telephone wire, it suffers considerable
degradation from noise, bandwidth limitation, and phase delay. It is therefore
necessary to place repeaters and equalization circuits at intervals of approximately
Figure 9.20. The block diagram of the PCM decoder=demultiplexer.
6000 ft to restore the pulses. 6000 feet (approximately 2000 m) of 22 gauge twistedpair non-loaded wire has a 3 dB bandwidth of approximately 4 kHz.
Telecom Canada Digital Network
The Telecom Canada Digital Network (DS-system) is similar to the American
Telephones and Telegraph (AT&T) T -system. In the Telecom Canada digital network
[1], 24 voice channels each operating at 64 kbit=s are multiplexed into a
1.544 Mbit=s channel, designated DS-1. Four DS-1 channels are multiplexed into
a DS-2 which operates at 6.312 Mbit=s. DS-3 has seven DS-2 feeding into it and it
has a bit rate of 44.736 Mbit=s. Six DS-3 channels are multiplexed to form DS-4. Its
bit rate is 274.176 Mbit=s and it can handle 4032 voice channels. The signals from
DS-1 and DS-2 may be transmitted on number 22 shielded twisted pairs. DS-3 and
DS-4 use coaxial cable. All DS-system output pulses are put into the alternate mark
inversion (AMI) format. This format eliminates the dc content of the signal and it
can be used to detect errors in transmission as well as provide substitution codes for
time synchronization. The elimination of the dc content of the signal means that the
dc power required to operate the repeater can be sent on the same line as the signal.
The output from these PCM switches can be used to modulate a suitable carrier
for transmission over terrestrial microwave links, fiber optic cables or satellite links.
Newer and more elaborate digital systems have come on the market offering more
channels, higher bit rates, greater flexibility, higher voice fidelity, and lower cost.
The basic idea of a hierarchy of systems made up of modules which are interchangeable remains (see Section 9.2.4).
Synchronization Circuit
For large switches such as the DS-2 or DS-3, the simple timing circuit used in the
example of the four-channel system is evidently inadequate. A centrally located
cesium beam atomic clock (oscillator) generates the primary signal. The frequency
stability of this oscillator is guaranteed to be less than one part in 1011 during its
lifetime. The output signal is distributed to all nodes of the network where timing
information is required according to a hierarchy determined by the importance of the
node. At each node, the signal is used to synchronize a local crystal-controlled
oscillator. The local oscillator (nodal clock) then has the accuracy of the primary
cesium beam atomic clock. The nodal clock must be accurate on its own (one part in
1010 per day) so that, at times when the primary source is lost, the system can
function satisfactorily. All DS-1 signals are therefore synchronized indirectly to the
cesium clock. The synchronization of the nodal clocks is carried out as shown in
Figure 9.21.
This is a phase-locked loop (see Section in which the frequency of the
local voltage-controlled crystal oscillator (VCXO) is adjusted by a voltage derived
from the frequency difference between the two oscillators. When the difference is
zero no adjustment is applied and the VCXO stays in synchronism.
Figure 9.21. A block diagram of the system for the synchronization of the master clock to the
nodal clock.
Regenerative Repeater
It was mentioned earlier that the output of the DS-1 switch was transmitted on the
bandwidth-limited twisted-pair telephone line. The result of this is a rapid degradation of the signal with other factors such as noise contributing to it. It is therefore
necessary to install regenerative repeaters approximately 1800 m apart to detect the
degraded pulses and send new ones [3,4]. The action of a regenerative repeater
includes amplification, equalization, detection, timing, and pulse generation. The
regenerative repeater does not accumulate noise the way a repeater (amplifier) on an
analog line would; it generates a new clean pulse. A block diagram of the basic
regenerative repeater is shown in Figure 9.22.
The incoming signal is fed into a pre-amplifier=equalizer to boost the level of all
frequencies present in the signal to a suitable value. The output of the amplifier then
goes to the input of the frame bit detector (not shown) and the full-wave clock
rectifier. The frame detector identifies the start=stop of each frame and the rectifier
changes the bipolar form of the signal to unipolar, thus producing a discrete
frequency component at the signalling rate. A high Q factor LC-tuned amplifier
selects the sinusoidal frequency component at the clock rate. The resulting sinusoid
is amplified and fed to a phase shifter. The phase shifter output is further amplified
and then goes to the amplitude limiter which produces a square wave. The square
wave is fed into a differentiator to generate positive- and negative-going clock pulses
Figure 9.22. The block diagram of the regenerative repeater. The bit frame detector is not
shown. Reprinted with permission from Transmission Systems for Communications, 5th Ed.,
AT&T, Bell Labs, 1982.
at the zero crossings. A phase shift network is used to adjusts the phase of the clock
pulses so that the positive clock pulses coincide with the maximum points (see
Figure 9.26: this is referred to as ‘‘maximum eye opening’’) on the incoming signal
(amplified and equalized). The resulting pulse and the output of the preamplifier=equalizer go to the regeneration repeater where a decision has to be made, in
each time slot, whether the received bit was a 1 or a 0. If the decision is that the
received bit was a 1, the regenerator produces a pulse of appropriate dimensions. Pre-amplifier=Equalizer. As mentioned in Section 9.3.6, the PCM
signal transmitted over the twisted-pair telephone line degenerates quite rapidly due
to electrical noise, both man-made and natural. The restricted bandwidth and delay
characteristics of the line attenuate the higher frequency components of the signal,
reducing the rise and fall times of the pulses. The amplifier=equalizer has an
appropriate high-frequency response boost built into it to compensate approximately
for 1800 m of 22-gauge twisted-pair telephone line. This circuit may include an
automatic gain control to keep the signal amplitude within specified limits. Frame Bit Detector. The frame consists of 192 bits (8-bit code 24
channels). Since the sampling frequency is 8 kHz, each frame must be scanned in
125 ms. It is necessary, however, to have a marker to indicate where the frame starts
or ends so that each 8-bit word can be identified and steered to the correct channel at
the receiving end [6]. To provide such a marker, one extra bit, called the framing bit,
is added at the end of each frame. By changing the framing bit from a 1 to a 0 and
back again at the end of every frame, it is possible to identify it at the receiving end
and to use it for synchronization of the channels. Positive identification is possible
because the alternating value of the framing bit represents a frequency of 4 kHZ but
all the signals in the other channels were pre-filtered to remove all frequencies above
3.5 kHz. It follows that none of the other bits derived from signals in the channel can
have a pattern of alternating bits from frame to frame.
The incoming pulses are scanned for the alternating bit in the 193rd position, and
when it is found the system latches onto it. When a specified number of errors occur
in the 193rd time slot within a given time period, the system reactivates itself and
starts the scanning process over again. While this is going on information is lost and
the telephone user might hear a click or two.
The circuit of the frame bit detector and the reframing process when an error
occurs is quite extensive and will not be discussed in greater detail. The Clock Rectifier. The full-wave rectifier restores the timing
frequency to its original value and reverses the alternate mark inversion (AMI)
process. This can be seen in Figure 9.23(a).
The rectification process is best achieved by a center-tapped transformer and two
diodes as shown in Figure 9.23(b). The LC-Tuned Amplifier. The Q factor of the amplifier is designed to
be in the range 75–100. This ensures a narrow bandwidth for the precise recovery of
Figure 9.23. The bipolar signal is rectified by the center-tapped transformer and diodes.
the timing frequency and the so-called flywheel effect, that is, the circuit continues to
operate even when some of the pulses are missing. The output is sinusoidal. Narrow
bandwidth LC-tuned amplifiers were discussed in Section 2.5.2. Phase Shifter, Limiter and Differentiator. The phase shifter, limiter
and differentiator are shown in Figure 9.24.
The phase shifter shown in the diagram provides an output which leads the input
signal; R1 and C1 can be interchanged if a phase lag is required. The time-constant,
R1 C1 , determines the amount of phase shift.
The limiter is an operational amplifier with two back-to-back Zener diodes in its
feedback path. The output of the operational amplifier is clamped in the positive
direction by D2 acting as an ordinary diode and D1 acting as a Zener diode. Their
roles are reversed when the output voltage goes negative. The resulting square wave
output goes to the differentiator.
The choice of the time-constant for the ‘‘differentiator’’ is governed by the
condition R2 C2 T , the period of the input frequency.
Figure 9.24. The phase shifter, amplitude limiter and differentiator. Pulse Regenerator. The circuit of the pulse regenerator [4] is shown in
Figure 9.25.
The pulse regenerator consists of a bistable multivibrator whose operation is
controlled by the clock pulse and the received signal pulse (amplified and equalized)
through the transformer T1 and diodes D1 D6 .
The received signal pulse is applied to the primary of the transformer T1 whose
secondary center-tap is grounded. Assuming that the voltage at node X is positive
and also that this happens to coincide with a positive clock pulse, the diodes D6 and
D4 are therefore reverse biased. Assume further that Q1 is on and in saturation and
Q2 is off; the base of Q1 is then at þ0:7 V. Diode D2 will conduct current through
R04 due to the dc supply, VBB . There will he a tendency for Q1 to conduct a little less
current than before. Meanwhile, the voltage at node Y is negative and D5 is forward
biassed and the voltage at node S is more positive than the voltage on the base of Q2 .
D1 is therefore reverse biassed; consequently the base voltage of Q2 is free to
change. Returning to Q1 , its collector voltage will tend to rise which will cause the
base voltage of Q2 to rise too. The collector voltage of Q2 will start to drop and this
will cause the base voltage of Q1 to fall even faster. This is the regenerative phase of
the multivibrator action which ends with Q1 cut off and Q2 in saturation.
The signal from the pre-amplifier=equalizer is still in the AMI format so the next
signal pulse input to the regenerator will cause node Y to go positive and if this
happens to coincide with a positive clock pulse, the multivibrator will change states
with Q1 back on and Q2 cut off.
Since the collectors of Q1 and Q2 are coupled through the primary winding of
transformer T2 to the positive dc supply, Vcc , whenever they change states, a pulse of
current will flow in opposite directions in the primary winding of T2 , causing a pulse
of opposite polarity to occur across the secondary. The output pulses are therefore in
the AMI format (bipolar).
The time-constant of the transformer inductance, its associated load and the
current that flows in the transformer when the multivibrator changes states determine
Figure 9.25. The bipolar pulse regenerator.
the width of the pulse. The pulse shape is primarily controlled by the B H
(magnetic) characteristics of the transformer core material. A boxlike hysteresis
loop produces a pulse with fast rise and fall times. This is because the core flux
saturates early and any further increase in the current has no corresponding effect on
the flux.
There are a number of different circuits which can be used in a regenerative
repeater. Several versions use a blocking oscillator to restore the degraded PCM
To optimize the operation of the regenerative repeater, the phase shifter is
adjusted so that the positive-going clock pulse coincides with the maximum value
of the pre-amplifier output. This is shown in Figure 9.26.
Data transmission over metallic loops predated voice traffic by at least 30 years. In
1837, Wheatstone and Cooke in England, and Morse in the United States,
established communication systems which used binary information exclusively.
Figure 9.26. The clock pulse is adjusted to coincide with the peak of the degraded pulse.
Reprinted with permission from J.S. Mayo ‘‘Bipolar Repeater for Pulse Code Modulation
Signals’’, Bell System Technical Journal, 41, 25–97, 1962.
Later, much improved forms of the telegraph (Teletype or Telex) worked side by side
with the telephone into the 1970s. One of the major reasons for its continued use was
the need for the written word in business transactions. The advent of fast electronic
computers capable of handling data at an ever faster rate hastened the development
of transmission systems designed to cope with high speed data. The evolution of the
teletype into the integrated services digital network (ISDN) will now be traced
The printing telegraph was invented in 1855 and it was the first attempt to
transmit data automatically. In its practical form (Teletype) it had a keyboard based
on the typewriter. In the transmit mode, the pressing of a key would cause a start
code to be transmitted followed by a 5-bit message code followed in turn by a stop
code. It was therefore not necessary to synchronize the transmitter to the receiver.
This was all done by a set of mechanical cams which opened and closed contacts to
produce the necessary pulses. In the receive mode, the current pulses were decoded
and used to operate a suitable electromechanical system to print the required
character. The information was transmitted over telegraph lines. This system was
capable of speeds of 50–150 words per minute. It was usual practice to use a
punched paper tape to compose and edit the message before transmission.
The next significant improvement came with the introduction of the Teletypewriter Exchange (TWX) [1]. This used a voice channel on the telephone system,
including the switching system and represented an early example of service
integration. It ran at about 100 words per minute.
The addition of ‘‘store-and-forward’’ capability to the TWX provided a new
service called the Message Switched Data Service (MSDS). Store and forward, as
the name suggests, was the ability to accept and store a message and to transmit it
only when a channel was available and the receiver was ready. It also had the
possibility to send the message to more than one recipient terminal.
The next advance made data transmission compatible with voice telephony; it
used the analog telephone channel by interfacing through a modem (modulatordemodulator). It was called Dataphone. The digital signal was used to modulate an
analog carrier in a frequency-shift keying (FSK) scheme. The digit 1 or mark is
coded as one audio frequency and the digit 0 or space is coded as a different audio
frequency. A typical modem generates 980 Hz for a mark and 1180 Hz for a space.
These frequencies were chosen to be compatible with the bandwidth limitations of
the channel and the speed of the transmission. At the receiving end the frequencies
were decoded into binary form. Data transmission rates up to 9600 bit=s have been
achieved using this technique.
The need to reduce the cost of data transmission led to the introduction of an
essentially digital network called Dataroute. It operates at speeds up to 56 kbit=s and
it can handle both synchronous and asynchronous data traffic at various speeds by
using suitable conversion codes. The data may be inserted into the DS-1 PCM
system and transmitted over the 1.544 Mbit=s channel.
The advantages offered by the use of computers in the telephone system are
exploited in the Datapac system. The system consists of nodes interconnected by
56 kbit=s transmission trunks. The nodes have dedicated computers which receive,
check for errors, package, address, and transmit data to other nodes in the network.
This is an example of a packet switched network. It has facilities for retransmission
of data when errors are detected.
The next significant improvement in data transmission was provided by Datalink.
This has the capability of transmission rates of 2400, 4800 and 9600 bit=s. It is
independent of code sets and protocols and compatible with the digital telephone
Integrated Services Digital Network (ISDN) is the result of international efforts to
standardize the format to be used in the existing and growing number of new
services offered by the telephone system and other organizations. It uses packet
switching techniques in which the data stream is separated into packets of modest
size and stored in a buffer until a channel becomes available for its onward
transmission. It is quite likely that different packets of a message reach their
destination at different times and by different routes. It is therefore imperative that
every packet be labelled so the message can be reconstructed at the receiving end.
Current international agreements assume that ISDN would carry a whole range of
telecommunication services including digital video signals. The suggested classes of
service are:
simple voice-channel service,
complex telephone service (smart telephone),
message storage and delivery,
low bit rate data,
8–64 kbit=s data,
N 64 kbit=s data services, where N is large.
One of the more prominent data transfer systems under development is the
asymmetrical digital subscriber line (ADSL). It belongs to a class of techniques
generally referred to as xDSL which aim to create a high-speed data link between
personal computers and the Internet using the existing twisted-pair telephone line. It
is asymmetrical because, in general, the flow of data from the Internet to a personal
computer far exceeds the data flowing in the opposite direction. ADSL is supposed
to open up the bottleneck formed by the twisted pair between the telephone wall
socket and the central office, often referred to as ‘‘the last mile’’. The name ‘‘the last
mile’’ is deceptive because the length of the twisted-pair telephone line can vary
from a few meters to approximately 6 km. The variability of the line impedance and
its frequency response are problems that all xDSLs have to deal with. The nominal
bandwidth of the-last-mile is approximately 4 kHz. The bandwidth required for data
from the central office to the subscriber is approximately 1.1 MHz for the transfer of
acceptable video information for video conferences and the downloading of complex
graphical displays from web pages. To achieve this objective high-gain, low-noise
amplifiers with built-in frequency equalizers are required. However, amplifiers can
generate gain for signals travelling in one direction only. It is therefore necessary to
employ 2-to-4 wire hybrids so that signals can be transmitted in both directions.
Modem Circuits
The block diagram of a typical modem is shown in Figure 9.27. Typical frequencies
f1 ¼ 980 Hz;
f3 ¼ 1650 Hz;
f2 ¼ 1180 Hz
f4 ¼ 1850 Hz:
Figure 9.27. The block diagram of a typical modem.
Incoming Data 1 is fed into the input port of modem 1 where a FSK transmitter
converts it into f1 and f2 for mark and space, respectively. It is then coupled to the
telephone line through a 4-to-2 wire hybrid (in older models, the carbon microphone
and the telephone receiver were used in conjunction with an acoustic coupler; the 4to-2 wire hybrid would then be the station set hybrid). At the receiving end modem 2
hybrid directs Data 1 to the FSK decoder which converts f1 and f2 into mark and
space, respectively. Modem 2 communicates with modem 1 in the same way except
for the difference in the frequencies used.
The 4-to-2 wire hybrid was discussed in Section Frequency-Shift Keying Modulator (Transmitter). The FSK
modulator of the modem has the block diagram shown in Figure 9.28. A crystalcontrolled oscillator provides an accurate clock frequency which is fed into two
divide-by-N circuits. The outputs are passed though narrow bandpass filters tuned to
the required frequencies f1 and f2 . The outputs of the bandpass filters go to the inputs
of the multipliers. The multiplicands are the binary signal to be FSK-coded and its
complement. Note that the binary signal and its complement are used to control the
flow of signal from the bandpass filters to the adder. When the binary input is a 1, a
signal of frequency f1 is fed to the adder; f2 is cut off. When it is a 0, f1 is cut off and
f2 appears at the input of the adder.
The crystal-controlled oscillator was discussed in Section 2.4.6. An inexpensive
and readily available crystal designed to operate at 3.58 MHz (originally designed
for the color burst carrier of television receivers; see Section may be used in
which case we have:
Target frequency, f
Required divisor
Integer N
Actual frequency, f 0
Error %
Figure 9.28. The FSK modulator with a crystal-controlled oscillator driver.
Figure 9.29. The FSK demodulator.
The divide-by-N circuits can be realized by using two 12-stage ripple counters
(212 ¼ 4096) with suitable resets to get the required values of N . The bandpass
filters can be simple LC-tank circuits with modest Q factors tuned to 980 Hz and
1180 Hz, respectively. The analog multiplier was discussed in Section 2.6.3, the
inverter in Section, and the adder in Section
There are several other schemes which can be used to realize the FSK coder. Frequency-Shift Keying Demodulator (Receiver). A simple FSK
demodulator of the modem is shown in Figure 9.29. Two narrow bandpass filters are
tuned to the two frequencies present in the FSK signal. Their outputs are used to
drive the envelope detectors. Note that the envelopes of the two frequencies f1 and f2
are both square waves. The output of one of the envelope detectors is inverted and
added to the other to give the required binary output.
The narrow bandpass filters can be LC parallel-tuned amplifiers with modest Q
factors. The envelope detector was discussed in Section 3.4.6, the inverter in Section, and the adder in Section The decision threshold circuit is a Schmitt
trigger which was discussed in Section
There are several modems on the market which use sophisticated phase-locked
loop techniques for both the modulator and demodulator; most use integrated
1. Telecom Canada, Digital Network Notes, Ottawa, Canada, 1983.
2. Members of the Technical Staff, Bell Telephone Laboratories Inc., Transmission Systems
for Communications, 4th Ed., Bell Telephone Labs Inc., 1970. Publisher: Western Electric
Co. Inc., Winston-Salem, NC.
3. Mayo, J. S., ‘‘Bipolar Repeater for Pulse Code Modulation Signals’’, Bell System Technical
Journal, 41, 25–97, 1962.
4. Andrews, F. T., ‘‘Bipolar Pulse Transmission and Regeneration’’, U.S. Patent 2,996,578.,
Aug. 15, 1961.
5. Gray, P. R. and Messerschmitt, D. G., ‘‘Integrated Circuits for Local Digital Switching
Lines Interfaces’’, IEEE Comm. Mag., 18(3), 12–23, 1980.
6. Cirillo, A. J. and Thovson, D. K., ‘‘D-2 Channel Bank: Digital Functions’’, Bell System
Technical Journal, 51(8), 1972.
7. Stremler, F. G., Introduction to Communication Systems, 3rd Ed. Addison-Wesley, Reading, MA, 1990.
8. Briley, B. E., Introduction to Telephone Switching, Addison-Wesley, Reading, MA, 1983.
9. Dammann, C. L., McDaniel, L. D. and Maddox, C. L., ‘‘D-2 Channel Bank: Multiplexing
and Coding’’, Bell System Technical Journal, 51(8), 1972.
10. Members of the Technical Staff, Transmission Systems for Communications, 4th Ed., Bell
Telephone Labs. Inc., 1970. Publisher: Western Electric Co. Inc., Winston-Salem, NC.
11. Members of the Technical Staff, Transmission Systems for Communications, 5th Ed., Bell
Telephone Labs. Inc., 1982. Publisher: Western Electric Co. Inc., Winston-Salem, NC.
12. Lathi, B. P., Modern Digital and Analog Communication Systems, CBS College Publishing, New York, 1983.
What are the advantages of single-sideband suppressed carrier transmission
system? What are its disadvantages?
An AM radio transmitter delivers a maximum of 12 kW of radio-frequency
power to its antenna when the modulating signal is sinusoidal and the index
of modulation is 0.5. Calculate
(1) the power in the carrier,
(2) the power in the side frequencies.
Calculate the maximum power that can be radiated if the transmitter were
used in a single-sideband-suppressed carrier (SSB-SC) system. What is the
power gain offered by the SSB-SC system?
A telephone system uses frequency–division multiplex for 12 channels in a
basic group configuration (see Figure 9.2), where two tones of frequencies
f1 ¼ 400 Hz and f2 ¼ 2:5 kHz are applied to channel 5 (76–80 kHz). Unfortunately, the amplifier in channel 5 has a distorted characteristic which can be
represented by
i0 ¼ avi þ bv2i þ cv3i :
Derive an equation showing all the frequencies present in the output current.
Which channels in the basic group are likely to be disturbed by the distortion
caused by the channel 5 amplifier?
What are the advantages and disadvantages of a pulse-code modulation
system for the transmission of telephone signals?
A telephone system uses frequency-division multiplex for 600 channels in
a basic mastergroup configuration (see Figure 9.4). The output is to be pulsecode modulated using 128 levels. Calculate the minimum sampling frequency
allowing for one guard space between PCM samples.
Twelve audio channels in a telephone system are each band-limited to 3.5 kHz
and sampled at 8 kHz for transmission in a time-division multiplex PAM
(1) Calculate the minimum clock frequency, allowing for a suitable synchronizing pulse. Where would you place the synchronizing pulse?
(2) Describe how the output of the PAM system can be used to amplitude
modulate a carrier and calculate the bandwidth required to transmit the
signal, assuming that it is necessary to include up to the third harmonic of
the pulse train.
A pulse amplitude modulated (PAM) wave has the waveform shown in Figure
where the mark-to-space ratio is 1. Derive an expression for the PAM
Figure P9.1.
waveform given that, for a square wave,
f ðtÞ ¼
1 1
1 2P
sin not:
2 p n¼1 n
(1) The PAM waveform is applied to the input of an ideal low-pass filter with
a cut-off frequency equal to 10 kHz. Find an expression for the output of
the ideal low-pass filter. Sketch the waveform.
(2) The PAM waveform is now applied to the input of an ideal bandpass filter
of bandwidth 13 kHz to 19 kHz. Find an expression for the output of the
filter. Sketch the waveform.
(3) The bandpass filter in (2) is replaced by another ideal bandpass filter with
bandwidth 45 kHz to 51 kHz. Find an expression for the output of the
filter. Sketch the waveform.
Which of the above outputs would you choose in order to recover the
information in the original PAM waveform and why?
In the telephone system, for both FDM and TDM, channels are grouped to
form hierarchies for transmission. Discuss the advantages of this method of
organizing the channels.
What is a ring counter? Illustrate your answer with the design of an eightphase ring counter. Suggest an application for such a ring counter.
Design a Schmitt trigger to trigger when a voltage in excess of 4.5 V is
applied to its input. The dc supply voltage is 12 V and the load driven by the
Schmitt Trigger draws a current of 0.5 mA. Two enhancement-mode N channel MOSFET with threshold voltage equal to 3 V are provided.
Design an astable multivibrator to drive a sampling circuit in which the
sample time is 1 ms and the hold time 15 ms. The dc supply voltage is 6 V and
the sampling gate draws a current of 1 mA when it samples the signal. What
special precautions would you take in view of the large difference between the
sample and hold times?
Design a series sampling gate using a bipolar transistor. The source of the
signal to be sampled has a voltage of 5 V peak amplitude with an internal
source resistance of 100 O. The load ‘‘seen’’ by the sampling gate is 10 kO.
The transistor is NPN and has b ¼ 100, VCEðsatÞ ¼ 0:5 V and a base leakage
current (when the transistor is off) IEðoff Þ ¼ 50 nA. Calculate the error in the
Although the facsimile machine was invented in the 1840s, it remained largely a
device used in the newspaper industry for the transmission of pictures until the mid1980s. There were several reasons for this; some were technical and the others
The technical problems which held up the development of the fax machine are
illustrated in Figure 10.1. For simplicity we use the letter H and assume that
scanning is carried out horizontally from the top left side to the right. The scanning
head then moves to the second line and the process is repeated. Again, for simplicity,
the scanned field is divided into a matrix 20 by 20. Thus 400 pieces of information
have to be sent to the receiver in order to reconstruct the H. The first task is to
measure the level of light reflected or produced by each square and to assign a value
of 1 or 0; we assume here that a 1 is assigned when a square is white and a 0 when it
is black (the opposite would work just as well). Figure 10.1(b) shows the result. Each
one of these pieces of information is called a pel (which is a pixel with its gray scale
or color information placed in two categories, black or white, depending on its
relative brightness). For the transmission to be successful, the transmitter has to
‘‘tell’’ the receiver precisely which squares are to be left white and which are to be
made black. In other words, the ‘‘read’’ head in the transmitter and the ‘‘write’’ head
of the receiver must be exactly on their corresponding squares at the same time, that
is, they must be in synchronism and in phase. To obtain synchronism, two
pendulums of the same length (with a mechanism for keeping them in phase)
were used [1]. The pendulum was not very practical because it had to be made quite
large to store enough energy so that the losses during the scanning and printing
processes would be negligible. Improved synchronization was obtained when the
tuning fork replaced the pendulum but this new technique did not become accurate
enough for the purpose until the 1940s. The problem of synchronization was never
Figure 10.1. (a) The letter H showing white and black pels. (b) Binary representation of the pels.
satisfactorily solved; indeed, the problem disappeared eventually when digital
techniques were applied to fax machine development.
Another technical problem which accounts for the slow development of the fax
machine was the speed at which the information could be transmitted. We recall
from Chapter 1 that the initial attempts to construct fax machines took place before
the telephone was invented. The telegraph lines at the time used single wires with
ground returns. These were subject to electrical noise mostly generated by electric
street vehicles, which were very popular at the time. To compound the problem, the
telegraph lines used relays to extend their reach (Morse’s relay) and these were
inherently too slow to convey the volume of information required to make the fax
machine a success. Note that at the minimum rate for scanning a 8.5 11 inch
(21.6 28 cm) page (200 lines per inch) the number of pels generated is 2.86 106.
Even with modern coding schemes the telegraph lines could not have handled the
sheer volume of information in a reasonable time to make this a success.
The next technical problem that had to be solved was the adoption of a suitable
method of coding the information so as to reduce the high level of redundancy. We
observe from Figure 10.1(b) that row 1 is completely white and hence it is
represented by a row of twenty 1s. A shorter code made up of a few 1s and 0s
could be used to signal to the receiver to insert twenty 1s in row 1. Similarly, in row
2, there are two transitions from 1 to 0 and two transitions from 0 to 1. To convey
this information it is possible to devise a code word with less than twenty bits to tell
the receiver where the transitions occur and whether they are from 1 to 0 or vice
versa. It can also be seen from Figure 10.1(b) that, in our example, row 3 is the same
as row 2. A further reduction in the bits required can be achieved by sending a
relatively short signal to the receiver to repeat row 2.
At least in the United States, commercial rivalry discouraged cooperation
between the engineers who were working on the development of the fax machine.
Each manufacturer developed their own standards, different from those of their
rivals. The effect of this was to keep costs high; only the military, the police, and the
large news organizations could afford to own and operate fax machines. It was not
until the mid-1950s that the Institute of Radio Engineers came up with a set of
standards which were accepted by the industry.
The development of the modern fax machine starts with the adoption of the
recommendations of the Consultative Committee of the International Telegraph
and Telephones (CCITT) of the standards which came to be known as Group 3 (G3).
Group 1 and 2 standards (which incidentally were analog) had been essentially
ignored by the North American fax industry. The discussion of G3 standards started
in 1976 during the General Assembly of the CCITT which had in the meantime
transformed itself into the International Telecommunication Union (ITU), a special
agency of the United Nations. By 1980 a full complement of recommendations
which form the G3 standards had been adopted. These recommendations, with a lot
of options (to satisfy various special interest groups), covered the parameters for
‘‘handshaking’’ between send and receive machines, modem speeds, scan densities,
coding schemes and a system that was completely digital. The decision to adopt
digital techniques banished the problems associated with synchronization and
phasing. Compromises were made and sometimes these were influenced by the
possibility of royalty payments. Two coding schemes were accepted. The first
because the patent had expired [2] and the second because the owners of the
patent offered it free of charge to users.
The Transmit Mode
Figure 10.2 shows a block diagram of a typical G3 fax machine when it is operating
in the transmit mode. The scanner has 1728 charge-coupled devices (CCD) arranged
in straight line array. Each element of the array reads the brightness of the first row of
pixels and converts these to an equivalent voltage. An amplifier strengthens the
signal to a suitable level for the next stage of processing. A Schmitt trigger circuit
assigns a value of 1 (white) or 0 (black) to each pixel, thus changing them into pels.
This information is stored and the next row of pixels are read by the CCDs and
likewise converted and stored. The signal is then coded to reduce the high level of
redundancy that is present by using the Modified Huffman (MF), modified relative
address, or simply modified READ (MR) or modified modified READ (MMR). The
choice of which of these coding systems is used for the transmission depends on the
capabilities of the fax machines involved. A table of the modified Huffman codes is
given in Appendix D [3]. The coded information is accumulated in the memory and
sent to the modem at a time determined by the microprocessor. The modem converts
the signal into analog form for transmission along the telephone line to the receiving
fax machine. The line adjuster may be used to modify the impedance of the line, the
frequency response and=or to minimize echo on the telephone line. The stepper
motor drives the mechanical system which advances the sheet of paper through the
The Receive Mode
Figure 10.3 shows a block diagram of the fax machine when it is in the receive
mode. The modem receives its input from the telephone line, converts the analog
signal into digital form and stores it in the memory ready for decoding. The decoder
reconstructs the original message and it is suitably amplified. The output of the
power amplifier drives a thermal printer. A roll of specially treated paper is drawn
past a set of hot wires spaced at approximately 200 per inch (same as the resolution
of the scanner). The roll is driven by the stepper motor. When the output of the
decoder indicates that a pel is black, current flows through the corresponding hot
wire and this causes the paper to produce a black spot. All the blocks represented in
the diagram are under the control of the microprocessor. The current trend is away
from thermal to xerographic and carbon film transfer printers which use ordinary
Figure 10.2. The block diagram of the ‘‘send’’ portion of the facsimile machine.
Figure 10.3. The block diagram of the ‘‘receive’’ portion of the facsimile machine.
It is clear that every stand-alone fax machine has within it all the components
needed for both the transmit and receive modes.
‘‘Handshake’’ Protocol
To send a fax message, one inserts the page into the send machine. The paper is
caught between two rollers and it is immediately pulled part-way into the machine.
The machine is ready to read the first line of the message. The next step is to dial the
telephone number to which the receiving fax machine is connected. The number
dialled is stored by the send machine. On pressing the ‘‘start’’ button, the following
events take place:
1 The dial tone comes on.
2. The send machine dials the number stored and, if the number is not busy, the
‘‘ring-back’’ tone can be heard. Usually, the receive machine needs four rings
before it responds.
3. The receive machine goes ‘‘off-hook’’ (connects itself to the line) and sends a
2.1 kHz signal lasting approximately 3 s to the send machine to identify itself
as a fax machine.
4. The receive machine follows up by sending its identification code to the send
machine. This code tells the sending machine what the capabilities of the
receive machine are. The following information is vital and many other
options may be included:
(a) the speed of the modem,
(b) the scan density (number of lines per inch or mm),
(c) the type of decoding (MH, MR or MMR) it is programmed to perform,
(d) the size of its memory.
5. The send machine then sends a command signal which locks the receive
machine into conformity with the chosen attributes from the list in (4).
6. The send machine sends a standard test (training) signal to the receive
7. The receive machine sends a confirmation signal that the test signal was
correctly received.
8. The send machine sends the message.
9. If the test signal fails to arrive correctly, there may be options such as
telephone line equalization, change of modem speed, etc., or the call may be
10. At the end of the message a special code is sent to indicate this to the
received machine.
11. The receive machine then sends back a code indicating that the message was
successfully received.
12. The send machine goes back ‘‘on-hook,’’ terminating the call.
13. The receive machine also goes ‘‘on-hook,’’ ready for the next message.
The handshake protocol is typical of an older version of the fax machine. Newer
models may read a complete page at a time, process and store the information before
sending it.
In this section we examine the components that make up the circuit following the
order in which they are encountered by the signal.
The CCD Image Sensor Semiconductor Theory. To understand the operation of the CCD
scanner, a short overview of semiconductor theory is necessary. There are a number
of materials which can be made into semiconductors. The most widely used of these
is silicon. Pure silicon does not conduct electricity under normal conditions because
its electrons do not have enough energy to break away from the crystalline structure.
When thermal energy, an electric field or light is incident on a piece of silicon, the
electrons may acquire enough energy to escape the influence of the nucleus and
become ‘‘free electrons’’ within the material. The creation of free electrons in silicon
is facilitated by changing the material from a nonconductor to a semiconductor. In
order to turn pure silicon into a semiconductor, it is necessary to introduce an
‘‘impurity’’, that is, another chemical element. This process is called doping. A
common element used for doping silicon is phosphorus. Silicon has a chemical
valency of four and that of phosphorus is five. The phosphorus therefore introduces
an extra electron into the structure of the silicon. The extra electron is then available
for conduction of electricity under the right set of conditions. Phosphorus-doped
silicon is referred to as n-type (donor) silicon. In n-type silicon, the majority carriers
are electrons and the minority carriers are holes.
A second element commonly used for doping silicon is boron. Boron has a
chemical valency of three and it therefore produces a deficit of one electron. Instead
of talking about a deficit of one electron, we call it a hole. The boron-doped silicon is
referred to as p-type (acceptor) silicon. In p-type silicon, the majority carriers are
holes and the minority carriers are electrons [4].
Figure 10.4(a) shows a cross section of p-type silicon overlaid with silicon oxide
(an insulator) and then an aluminum electrode. With 0 V on the electrode, nothing
happens, but as the voltage is increased, the positive voltage on the electrode repels
the positive holes, creating a depletion layer as shown in Figure 10.4(b). As the
electrode voltage is increased further, a point is reached where the effect of the
electrode voltage has become so strong that it starts to accumulate electrons just
below the surface of the silicon. An inversion of the p-type to n-type silicon,
immediately under the electrode, has taken place. This layer is called the inversion
layer. The value of the voltage at which inversion takes place is called the threshold
voltage, Vth . Beyond the threshold voltage an increasing number of electrons
accumulate under the positive electrode. Indeed, a capacitor has been formed
between the electrode and the p-type silicon. This is shown in Figure 10.4(c) and
(d). The structure is capable of generating (when the electrode voltage, VE , is
sufficient) and holding a charge similar to a capacitor.
When light is incident on a piece of semiconductor material, electron–hole pairs
are generated. This happens when the energy of the photon is absorbed by the
material causing the excitation of bound carriers into mobile states. Given suitable
structures and the appropriate biasses, the minority carrier (that is, the electrons in ptype material in the mobile state) can be collected. The effect of the light is
essentially the same as that of an electric field or a rise in temperature [5]. This is the
basis for a large number of sensors such as phototransistors and infrared detectors. Semiconductor Light Sensor. Figure 10.5 shows a typical structure
of an image sensor [6]. The p-type silicon is formed into islands (wells). Each island
is a cell and, using a suitable mask, the cells can be selectively exposed to light. Note
that the light enters the cell from the back side of the silicon chip where the electrode
is not in the way. Using the appropriate thickness of silicon, the correct level of
doping in the silicon, and applying the correct voltage to the electrode, the electrons
can be captured near the surface of the silicon just below the electrode. The number
of electrons captured is almost linearly related to the intensity of the incident light.
We have an electronic image of the mask ‘‘painted’’ in the form of charge. The ‘‘Bucket Brigade’’. The role of the ‘‘bucket brigade’’ is to
transport the charge which forms the electronic image to an external circuit. The
structure of the device shown in Figure 10.6 is similar to that in Figure 10.5. The
difference is that there are four electrodes which are very closely spaced. We use the
analog of a well filled with water to illustrate the action of the ‘‘bucket brigade’’ [6].
Figure 10.4. (a) The structure of the p-type silicon device with no voltage applied to the electrode.
(b) The formation of the depletion layer when a positive voltage (VE < Vth ) is applied to the electrode.
(c) With increasing value of VE ðVE > Vth Þ the inversion layer is formed. (d) The comparison of the
device to a capacitor.
Figure 10.5. The image sensor showing the 1728 cells. The electrons shown in the cells were
generated by the incident light.
When a positive voltage of sufficient magnitude is applied to one of the electrodes, a
‘‘well’’ (inversion layer) of charge is formed. The ‘‘well’’ is filled with water
(negative charge, that is, electrons). We assume that there is water in the well
under ‘‘electrode b’’ because of its higher voltage. The water stays there because
there are no wells under a and c. Suppose we dig a well under electrode c (that is,
change the voltage on electrode c from 2 V to 10 V). Clearly the water will flow into
a well whose width has doubled, hence the level of the water will drop to one-half of
the original (the quantity of water is assumed constant). Finally, we fill in the well
under electrode b. The water level in the well under electrode c will go up to the
original level. We have succeeded in moving the water one step to the right.
The charge stored under electrode b can be shifted to electrode c using the voltage
waveform shown in Figure 10.6(e). Note that, to ensure that the charge is shifted
unambiguously from left to right, the two electrodes on opposite sides of electrode b
must be connected together. Similarly, every third electrode must be connected
together as shown in Figure 10.7.
A combination of the light sensor and the ‘‘bucket brigade’’ (or analog shift
register) makes it possible to ‘‘measure’’ the light incident on the image sensor and to
transmit the information to an external circuit. The CCD Image Sensor Unit. A combination of the light sensor and
the ‘‘bucket brigade’’ is shown in Figure 10.7. to avoid the production of stray
signals, a metal mask (not shown) shields all structures other than the light sensor.
To start the operation, a positive voltage is applied to the photogate and the
transfer gate is ‘‘closed’’ by the application of a negative voltage. Charge proportional to the level of illumination and also to time of exposure accumulates in each
cell. After the so-called integration time, the transfer gate is ‘‘opened’’ and the
charges accumulated in the cells flow through to the electrode which is most positive
of the set of three. This is a parallel process. The transfer gate is then closed and the
three-phase clock drives the accumulated charges from left to right. As each packet
of charge arrives at the terminal, it is converted into a voltage and suitably amplified.
This is a serial operation. Mechanical and Optical Systems. Figure 10.8 shows a typical
optical and mechanical system used for scanning the page. A cylindrical lens
concentrates the light from a fluorescent lamp onto the document. A spherical lens
brings into focus a row of pixels formed from a line 0.005 inches (0.127 mm) wide
(200 lines per inch). The document is moved one row at a time by the stepper motor
until the page is completely scanned.
The Binary Quantizer
The voltage output from the CCD image sensor is a series of pulses whose
amplitudes are proportional to the light incident on each cell. These are pixels
and they must be converted into pels because the Group 3 fax standards are designed
to operate with black and white levels only. How the G3 fax standard deals with
Figure 10.6. The hydraulic analog of the ‘‘bucket brigade’’ showing how the charge is transferred from
left to right using the three-phase clock. Reprinted with permission from J. D. E. Beynon and D. R.
Lamb, Charge-Coupled Devices and their Applications, McGraw-Hill, London, 1980.
Publisher’s Note:
Permission to reproduce
this image online was not
granted by the copyright
holder. Readers are kindly
requested to refer to the
printed version of this article.
Figure 10.7. The CCD image sensor unit is the combination of the light sensor and the ‘‘bucket
brigade’’. Reprinted with permission from J. D. E. Beynon and D. R. Lamb, Charge-Coupled
Devices and their Applications, McGraw-Hill, London, 1980.
shades of gray will be discussed in Section 10.9. The binary quantization is achieved
by using a Schmitt trigger circuit. The design of Schmitt triggers is described in
The Two-Row Memory
The two-row memory is required when a differential type coding system, such as the
Modified READ, is used. There are many types of memories in which the pels can
be stored. Because of the serial nature of the information coming from the binary
Figure 10.8. A typical optical and mechanical system used for scanning a page.
quantizer, we choose the shift register. A shift register is a string of flip-flops or
bistable multivibrators which are connected such that the states of the individual flipflops contain the information stored. In this type of shift register the data to be stored
is entered one bit at a time in response to a clock pulse. Similarly, the information
has to be ‘‘clocked out’’ one bit at a time.
The design of flip-flops or bistable multivibrators was described in Section Briefly, a flip-flop is a digital circuit whose output(s) can take only two
possible values, 1 or 0. Whichever value it takes remains until the arrival of a clock
pulse, at which point it switches to the other state. We can say that this circuit
‘‘remembers’’ its state before the clock pulse arrived. The flip-flop is the building
block for memories of this type. In Section, the flip-flop was described as two
NOT gates connected back-to-back. That was the non-resettable flip-flop. In order to
build a shift register, we require a resettable flip-flop. The design of the resettable
flip-flop and its application in the construction of a memory are presented in
Appendix E.
The Modified Huffman (MH) Code
As mentioned earlier, the data obtained from the scanner contains a huge amount of
redundant information. The MH code is based on the lengths of the run of white and
black pels [2]. It is relatively easy to count the number of white pels before a
transition to black occurs and then to count the black ones to the next white pel. The
number of white pels in the run is coded in the form of a n-bit word. Similarly, the
number of black pels is given a different n-bit word. Further to this, a number of
typical documents were scanned and the frequencies of the lengths of black and
white runs plotted in the form of a histogram. The most frequently occurring run
lengths were assigned shorter codes. This is similar to the approach used by Morse
when he developed his code; he chose the most frequently occurring letter in the
English language (e ¼ dot) and assigned it the shortest code. The MH code is given
in Appendix D [5]. Table D.1 gives the codes for run lengths from 0–63 and Table
D2 gives the codes from 64–1728. If the fax machine scanned a row that was
completely white (that is, 1728 white pels), it would transmit the code 010011011
(from Table D.2) followed by the terminating (that is, a run of length zero) code
00110101 (from Table D.1). Thus a white row with 1728 bits is reduced to a 17-bit
code (plus a few more for ‘‘houskeeping’’ purposes such as when to start a new line
and so on). Note that, because a page is likely to have more white spaces than black,
in general the codes for the white runs are shorter than those for black.
The Modified READ Code
The Modified READ (MR) code takes advantage of the fact that in typescripts and
manuscripts there are a lot of vertical or near-vertical lines. It therefore compares one
row of pels to the one immediately above it and codes the differences. Offsets of
zero, 1, 2 and 3 are allowed. This is described as two-dimensional coding. The
modified modified READ (MMR) is an extension of the modified READ with an
error correction option.
The design of a modem was discussed in Section 9.4.1. G3 fax standards have
several options for the speed (bit=s) of the modem. The two machines involved in a
transmission ‘‘negotiate’’ the speed at which the data will be sent. The following are
standard speeds used in fax transmission: 4.8 kbps, 9.6 kbps, 14.4 kbps, 21.6 kbps.
As mentioned earlier, during the ‘‘handshake’’ between the two fax machines, a
standard test (or training) signal is sent from the sender to the receiver. If errors
occur during the training, one of several parameters can be adjusted such as the
frequency characteristics of the telephone line. This is referred to as equalization.
The telephone line has the frequency characteristics of a low-pass filter and the loss
of gain at the higher end of the spectrum can cause errors during transmission. This
can be corrected with an amplifier which boosts the high-frequency gain relative to
the low end. Other telephone line characteristics such as echo can be modified to
improve the quality of the transmission.
The design of the modem was described in Section 9.4.1, and electronic memory
was discussed in Appendix E. The MH=MR decoder is a reversal ‘‘look-up’’ table of
that given in Appendix D.
The Power Amplifier
The power amplifier has to produce enough power to heat the temperature-sensitive
paper to about 110 C to make a dark mark. The print head is a resistor with very low
thermal capacity. With a scanning resolution of 200 lines per inch (both vertical and
horizontal), a page 8.5 11 inches has 1700 elements across (the actual number
used is 1728) and 2200 elements from top to bottom. Assuming a print rate of a page
per minute, each row of pels must be printed in less than 27 ms. The temperature of
the resistors have to reach their operating value and drop down (low enough not to
cause a smudge on the line below it) in much less time.
As the input signal to the amplifier is a pulse, the amplifier does not have to be
linear. A simple emitter follower is adequate. Figure 10.9(a) shows an emitter
follower, also called a common collector amplifier. The resistor Re is the resistance
of the printer head, R1 controls the base current of the transistor.
Example 10.8.1 The emitter follower. This is a special case design of the emitter
follower. A normal linear emitter follower is biased to handle sinusoidal waveforms.
In this case the signal is a pulse so it is quite adequate for the transistor to be in the
off state until the pulse is applied to its base, then current flows in Re . The capacitor
C is called a ‘‘speed-up’’ capacitor because it helps to reduce the transition time from
on to off and vice versa.
The power required to heat up the print head is usually about 0.5 W. Let
Vcc ¼ 12 V. The transistor is a silicon NPN device with current gain b ¼ 100.
Determine suitable values for Re and R1 .
Solution. Assume that the input pulse goes from 0 to 12V and also that output pulse
goes from 0 to 10 V. Since the power is 0.5 W it follows that the current
Ie ¼
¼ 50 mA:
The base current
Ib ¼
¼ 0:5 mA:
b 100
Since the emitter voltage is 10 V, it follows that the base voltage must be 10.7 V
(silicon transistor).
Figure 10.9. The power amplifier (emitter-follower) used for driving the thermal paper printer
The voltage drop across R1 is therefore 1.3 V. The value of
R1 ¼
¼ 2:6 kO:
0:5 103
A suitable value for the capacitor C is 100 pF.
The base current of 0.5 mA may be excessive for the output of the decoder to
provide. In that case, the circuit shown in Figure 10.9(b) may be used. This
combination of transistors is known as a Darlington pair and the effect of the
second transistor is to increase the input resistance (decrease the current required to
drive the amplifier by a factor of approximately b2 ) of the amplifier. The design steps
follow from the one above.
The Thermal Printer
The most significant part of the thermal printer is the print head which consists of
1728 heating elements spaced at 0.005 in (0.13 mm) apart. A cross section of the
printer showing the essential parts is illustrated in Figure 10.10. Details of the print
head are also shown.
The stepper motor advances the thermal sensitive paper a distance of 0.13 mm at
a time. Current is supplied to heat the elements of the print head corresponding to
the places where there are dark spots.
Figure 10.10. A cross section of the printer showing some of the essential parts. Details of the
print head are also shown.
The print head heating elements can be made of thick or thin film. They are
protected from abrasion by a layer of glass, as shown.
The impressions on a thermal fax paper tend to fade when exposed to sunlight
and=or moderate temperatures. Higher quality models of fax machines have other
types of printers. Some use xerography, carbon film transfer and ink-jet systems.
The G3 fax standards do not take into account the need to transmit images with gray
scales. The result is that the transmission of documents with print and line drawings
produce excellent results, but those with gray-scale information come out with high
levels of distortion. To produce an acceptable gray-scale reproduction while using a
binary form of coding, a technique known as dither was developed. This technique
relies on the human eye to perform an ‘‘averaging’’ function when it looks at areas of
varying densities of black pels; the more black pels per square, the darker the area
Figure 10.11. (a) Shows three discrete gray level signals and the constant threshold used to
digitize them. (b) Shows the output signal from (a). (c) Shows the same three discrete gray level
signals and the variable threshold used to digitize them. (d) Sector x has a lower density of black
pels than sector z and therefore sector x appears to be a lighter shade of gray than sector z.
appears. This is similar to the use of dots of varying sizes to produce pictures in
In a typical G3 fax machine, the scanner produces an analog output, with each
pixel proportional to the brightness of the image. The signal is converted into a
binary form by setting a threshold of brightness above which the signal is given the
value 1 (white) and below which it is 0 (black). Figure 10.11(a) shows three discrete
gray level signals and the threshold to be used to digitize them.
In the dither technique, the threshold for the binary coding system is varied so
that its average value is proportional to the brightness (gray scale), as shown in
Figure 10.11(c). The output signal is shown in Figure 10.11(d). For the gray scale
represented by sector x, the black pel density is 70% while that for sector z is only
40%. Therefore, sector x appears darker than sector z.
1. Costigan, D. M., Fax: The Principles and Practice of Facsimile Communication, Chilton
Press, Philadelphia, 1971.
2. Huffman, D. A., ‘‘A Method for the Construction of Minimum Redundancy Codes’’ Proc.
IRE, 40, Sept. 1952.
3. McConnell, K., Bodson, D. and Urban, S., Fax: Facsimile Technology and Systems, 3rd Ed.,
Artech House, Boston, MA, 1999.
4. Mauro, R., Engineering Electronics: A Practical Approach, Prentice-Hall, Englewood
Cliffs, NJ, 1989.
5. Chilian, P. M., Analysis and Design of Integrated Electronic Circuits, Harper & Row, New
York, 1981.
6. Reynon, J. D. E. And Lamb, D. R., Charge-Coupled Devices and their Applications
McGraw-Hill, London, 1980.
Dennis, P. N. J., Photodetectors: An Introduction to Current Technology, Plenum Press, New
York, 1986.
Quinn, G. V., The Fax Handbook, TAB Books, Blue Ridge Summit, PA, 1989.
Pixel. A picture element that has more than two levels of gray-scale information or
Pel. A picture element which has only black and white information (that is, a pel is a
pixel in which the average gray-scale level has been used to quantize it into a
binary form; black or white).
Bit. A binary digit, either 1 or 0.
Word. A set of bits; usually there are 8, 16, 32, 64 or more bits.
Read. Retrieve information stored in a memory.
Write. Store data in a memory.
Address. Location of data stored in a memory.
Compare and contrast the state of facsimile technology before and after the
development of inexpensive electronic memory.
Compare and contrast the state of facsimile technology before and after the
application of digital techniques.
Why is it necessary to code the output of the scanner in a fax machine before
What modifications would you make to the scanner of a fax machine so it can
be used as a television camera?
Describe briefly what happens when two fax machines are involved in a
How do the G3 fax machine standards deal with different shades of gray?
What are the advantages and disadvantages of the thermal printer? Describe
other technologies that have superior performance.
The idea that a person could carry around with him a telephone booth of his own is a
very attractive one. However, the technology required to make the telephone booth
small and light enough for this to be possible, and furthermore convenient to carry,
has been available only in the last 30 years. Mobile radio has been available in North
America since the 1930s but they were exclusive in the hands of the police. Later
taxicab operators installed radios in their vehicles. These mobile units were large,
heavy and electrical power hungry. They used amplitude modulation which is
notorious for poor performance in the presence of electrical noise and there was
more than enough noise generated by the ignition systems of the vehicles in which
they were installed. Moreover, they were operational in either the transmit or the
receive mode at any given time; they were of the ‘‘push-to-talk’’ type.
The invention of the transistor and its progression to integrated circuits made it
possible to reduce the weight and size of circuits and, at the same time, increase their
capability and flexibility. These advances were accompanied by an enormous
reduction in the amount of power required to operate transistor circuits. The stage
was set for the introduction of the ‘‘personal telephone booth’’.
There are many applications in which radio plays a vital part. These go from
‘‘remote keys’’ for the automobile, garage door openers, pagers, walkie-talkies,
cordless telephones to cellular telephones with access to the Internet. In this chapter,
we limit ourselves to a discussion of paging systems, cordless and cellular
telephones. The paging system was designed to send information from a base
station to a mobile terminal. The mobile terminal has no capability to transmit
information in the opposite direction. A serviceman for home heating furnaces, for
example, only needs to know the address of his next assignment; in general, he does
not have to contact his base office. A communication system in which information
travels in only one direction is described as simplex and the pager is an example.
Radio systems with a push-to-talk button use the same channel in both the forward
and reverse directions. It is therefore necessary for each person to indicate when they
have finished talking with the familiar word ‘‘roger’’. They are described as halfduplex. A full-duplex system uses two channels simultaneously, the first for
transmission and the second for reception. The cellular telephone is an example
of a full-duplex system.
In Section 9.2.1 we discussed the generation of a single-sideband-suppressed carrier
(SSB-SC) signal using a balanced modulator and a bandpass filter. Fig. 11.1(a)
shows the circuit configuration of the SSB-SC as well as the frequency spectrum of
the input and output. The output shows that the baseband signal has experienced an
upward frequency shift equal to the carrier frequency, oc and an ‘‘inverted’’ version
of it appears at a lower frequency and the two are symmetrically spaced about the
position of the carrier. A bandpass filter is used to select the upper sideband. Clearly,
the upper and lower sidebands contain the same information and only one of them
should be required for the recovery of the original signal.
Figure 11.1(b) shows a circuit in which upper sideband is multiplied (balance
modulated) with the carrier signal, oc . The corresponding spectrum shows that the
Figure 11.1. (a) The structure of the modulator and the spectrum of the corresponding
frequency shift. (b) The structure of the demodulator and the spectrum of the corresponding
frequency shift.
output has two ‘‘sidebands’’. The first is at a frequency 2oc and the second occupies
the position of the original baseband signal in the spectrum. The original signal is
recovered by using a lowpass filter. Equations (9.2.1 to 9.2.5) are the relevant
This modulation and demodulation technique can be used with amplitude,
frequency and phase or angle modulation schemes. When demodulation is carried
out using the carrier signal in a balanced modulator as shown in Figure 11.1(b), it is
referred to as coherent demodulation or coherent detection. The use of an envelope
detector to demodulate an AM signal is known as non-coherent demodulation.
Multiplex and Demultiplex Revisited
When modulation is used to accommodate a number of signals on a single channel
we refer to it as multiplexing. Figure 11.2 shows five baseband signals, each of
which occupies the frequency band 300 Hz to 3 kHz.
By choosing suitable carrier frequencies for each one, they may be transmitted
over the same cable or the airwaves by radio and subsequently demodulated with no
interference between them. When different carrier frequencies are used to multiplex
the baseband signals, it is referred to as frequency-division multiplex (FDM). Other
methods of multiplexing are described below.
The success of personal wireless communication systems is in part due to the
development of techniques which allowed a large number of signals to share a
limited spectrum. One of the boundaries of the spectrum available for personal
wireless communication is dictated by the size of the antenna for the radio interface.
Efficient transmission of radio signal at low frequency requires antennas several
thousand meters tall. Clearly, this is not possible as portability of the device is
essential. The boundary on the other end of the spectrum is set by the character of
high-frequency transmission which increasingly takes on the properties of visible
light which requires line-of-sight. Clearly, the modern environment (cities) in which
most of the potential subscribers live and work make line-of-sight communication
Figure 11.2. Five baseband signals occupying the same bandwidth can be separated by using
frequency-division multiplex.
devices inadmissible. Between these two boundaries we have other systems in
competition for the spectrum, such as air and sea navigation, satellite communication, radio and television broadcasting.
It so happens that national governments have arrogated to themselves the power
to assign portions of the spectrum for specific purposes within their territories and to
negotiate international treaties which govern their use. This, in short, brings us to the
assigned frequency bands of 824–849 MHz and 869–894 MHz for personal wireless
communication. For the large number of anticipated subscribers to be accommodated in such a restricted bandwidth it is necessary to develop techniques which
reduce the possibility of interference with each other.
One major factor working in our favor is that, provided we keep the radiated
power below a given level, and we are separated sufficiently by distance, we can
reuse the spectrum over and over again. We shall now discuss the techniques which
enable us to share the spectrum available.
Frequency-Division Multiple Access (FDMA)
Frequency-division multiple access is a fancy name for what is commonly done with
AM and FM radio broadcasting and TV stations; they are assigned different carrier
frequencies with suitable separation between them to ensure minimal interference
with each other. They are required by law to keep their carrier frequencies constant.
They also have a limited bandwidth and radiated power. The division of the spectrum
according to frequency was discussed in Section 9.2 under the heading ‘‘FrequencyDivision Multiplex’’ (FDM). Figure 11.3 shows a representation of the channels
spaced by their assigned carrier frequencies and separated by limited bandwidth and
appropriate guard bands.
Time-Division Multiple Access (TDMA)
In FDMA, a frequency band is dedicated to a particular channel for as long as it is
required. In TDMA, several channels share the same bandwidth but each channel
has the use of that bandwidth for a fraction of the time. TDMA was discussed in
Section 9.3 under the other name used to describe this technique: ‘‘Time-Division
Figure 11.3. In frequency-division multiple access (FDMA), channels are spaced by their
assigned carrier frequencies and separated by limitation on bandwidth and appropriate guard
Multiplex’’ (TDM). The basis of this technique is the ability to reconstruct a signal
from samples taken from it. Figure 11.4 shows how each channel is structured in
time to form frames and the sequences of the content of each channel. In TDMA, it
is necessary to synchronize the transmitter to the receiver so that bits from one
channel do not end up in another channel, hence the synchronizing bits.
Spread Spectrum Techniques
In spread spectrum communication systems the radio-frequency carrier is changed
very rapidly in a pseudo-random fashion over a bandwidth which is much wider than
the minimum required to transmit the signal. Potentially it should cause interference
with other users of the airwaves but, in fact, because the carrier operates for such a
short time at any given frequency, its effect is almost imperceptible. The average
perceived power on any given channel is very low and it therefore behaves like a
low-power noise source spread across the bandwidth it uses. Many communication
channels can operate in this fashion without interfering with each other. Spread
spectrum technology has been of particular interest to the military because it is
almost impossible to predict the next frequency of the transmission; they like to stay
away from eavesdroppers and to avoid the jamming of their communication systems
by the enemy. The real challenge in spread spectrum communication is to keep the
receiver synchronized to the transmitter. We shall return to the problem of
synchronization later.
There are two major types of spread spectrum techniques. They are frequency
hopped and direct sequence spread spectrum technologies. Frequency Hopped Multiple Access (FHMA). In FHMA transmission, the information is first digitized and then broken up into short passages.
Each passage is transmitted on a different carrier frequency determined by a pseudorandom number generator. Because the modulation used is either narrow band FM or
frequency-shift keying, at any instant, a frequency hopped signal occupies a single
narrow channel. However, because the carrier frequency hops around, it makes use
of a much wider bandwidth. Figure 11.5 shows a representation of a system that uses
FHMA. Clearly, in an FHMA the receiver has to have prior access to the sequence of
the carrier frequencies transmitted as well as the timing to be able to follow the hops
(synchronize). It is quite likely that two or more transmitters will at some time try to
use the same frequency. Code Division Multiple Access (CDMA). In CDMA transmission,
the information is first digitized and then multiplied by a binary pseudo-random
sequence of bits (called chips) with a bit rate much higher than that of the digitized
information [1]. Figure 11.6 shows the binary message signal, bits of a pseudorandom code, and the spread spectrum (coded) signal.
Note that, because the bit rate of the pseudo-random sequence is much higher
than that of the message signal, it requires a much larger bandwidth for its
transmission. The spread signal is used to modulate a carrier (usually FM or PM)
Figure 11.4. In time-division multiple access (TDMA), each channel is structured in time to form
frames and the sequence of the content of each channel.
Figure 11.5. A representation of a frequency hopped multiple access (FHMA) system.
Although no instances of two or more transmission on the same frequency and at the same
time are shown, there is a clear possibility that this can happen. Note that for simplicity, all
channels have equal bandwidth and occupy that bandwidth for the same length of time. Neither
of these conditions apply in practice.
and then transmitted. At the receiving end the spread signal is demodulated then
decoded using a locally generated pseudo-random bit sequence in a process called
correlation. Because the number of chips representing a message bit (1 or 0) is large,
the correlation does not have to be perfect; it has to correctly recognize the majority
of the chips as representing that message bit (1 or 0). In a system which is subject to
multipath fading, this is an advantage. It is clear that the receiver has to have prior
knowledge of the pseudo-random code to be able to decode the message. To other
receivers not using the identical code, the message appears to be just noise. The
attraction of this technique is that it can be used to accommodate a large number of
subscribers with different codes and they will not even know that they are sharing
the same bandwidth. A by-product of CDMA is improved security of the message.
One disadvantage of CDMA is that the power of individual transmitters has to be
controlled very carefully. A strong signal from one of the transmitters within the
wideband can overwhelm the sensitive front-end of the system and prevent the
reception of other signals. The transmit power control system for all the mobiles
adds complexity and costs.
So far, we have discussed carrier systems in which the message signal is in analog
form. Increasingly, electronic systems are using a digital format. For example, it has
Figure 11.6. The binary data and its equivalent bipolar form, the pseudo-random sequence (chips)
and the resulting spread spectrum data.
taken the music recording industry less than 15 years to replace the analog vinyl
record with the digital compact disc. There are technical as well as economical
advantages to be gained from this move. Moreover, the advent of integrated circuit
technology with its ability to fabricate extremely large numbers of circuits on
minuscule pieces of semiconductor has made the move to digital systems seem
To transmit a baseband (message) signal over a radio channel, it is necessary to
change some property of the radio-frequency signal using the baseband signal. We
can change its amplitude, its frequency, or its phase angle. In Chapter 2 we discussed
the modulation of a radio-frequency signal by a message signal in which the
amplitude of the RF signal varied according to the amplitude of the message signal
(amplitude modulation; AM). In Chapter 4 we discussed how to change the
frequency of the RF about a fixed value using the message signal (frequency
modulation; FM). It is now time to discuss the modulation scheme in which we vary
the phase of the RF signal according to the message signal (phase modulation; PM).
It must be pointed out that frequency and phase modulation are, in fact, the same.
The only difference is that in PM the phase of the modulated waveform is
proportional to the amplitude of the modulating waveform, while in FM it is
proportional to the integral. Both schemes are sometimes referred to as angle
modulation. Phase modulation, when the message signal is a continuous (analog or
tone) function, does not appear to have any practical applications. When the message
signal is digital, it has distinct advantages such as improved immunity to noise.
Binary Phase Shift Keying (BPSK)
When the modulating (message) signal is in binary form we refer to it as keying.
This is a left-over from the days when telegraph operators opened and closed a
circuit (presumably, using a ‘‘Morse key’’) to generate Morse code.
Figure 11.7 shows a comparison of the waveforms of the three modulating
It should be noted that, for clarity, the RF has been chosen to be only four times
the data rate. In practice, the RF is much higher than the data rate.
If we represent the digit 1 by the binary pulse pðtÞ ¼ 1, the digit 0 by pðtÞ ¼ 1
and the carrier by cos oc t, then after modulation we have
sðtÞ ¼ pðtÞ cos oc t
sðtÞ ¼ pðtÞ cos oc t ¼ pðtÞ cosðoc t þ pÞ
for the digit 1 and
for the digit 0.
Demodulation of a BPSK signal requires a balanced mixer and an exact replica of
the carrier.
Figure 11.7. (a) Binary data, (b) its bipolar equivalent, (c) the amplitude-modulated waveform,
(d) the frequency-modulated waveform, and (e) the phase-modulated waveform.
Quadrature Phase Shift Keying (QPSK)
In quadrature phase shift keying, the message signal is separated into in-phase (I)
and quadrature-phase (Q) components and are then modulated separately by two
carriers of the same frequency but with a phase difference of 90 . QPSK is used
because twice the information can be carried in the same bandwidth as when BPSK
is applied [2].
Figure 11.8 shows the structure of the QPSK modulator. It has been assumed that
the pulses used for modulating the carrier are rectangular. In fact, rectangular pulses
are quite undesirable since, in a limited bandwidth channel, they tend to smear into
the time intervals of other pulses [3]. The pulse shaping filter is used at the baseband
or at the IF stage to limit adjacent channel interference.
Figure 11.9 shows the waveforms of the original data, the I and Q components,
the I cos ot and Q sin ot as well as the QPSK signal. Note that the QPSK signal is a
combination of the waveforms of the I cos ot and Q sin ot components.
The demodulation of the QPSK signal is done coherently as shown in Figure
11.10. After down-conversion the received signal is split into two parts and each part
is demodulated using a carrier signal derived from the received signal by a carrier
Figure 11.8. Block diagram of the QPSK modulator.
Figure 11.9. The waveforms of (a) the data, (b) the non-return-to-zero in-phase (I) component,
(c) the non-return-to-zero quadrature (Q) component (note that the waveform shown in (a) is not
coincident in time with those shown in (b) and (c) ), (d) I cos ot, (e) Q sin ot, (f ) the QPSK signal,
and (g) the QPSK signal with the phase shifted by þp=4.
Figure 11.10. Block diagram of the QPSK demodulator.
recovery circuit. The low-pass filters remove the undesirable products of the
multiplication process. Two circuits make decisions on whether the bit that was
sent was a 1 or a 0. The I and Q components are passed to a multiplexer which
reconstitutes the original binary signal.
There are a number of occupations in which the professional has to move around
from one job to the next and essentially is almost never available at a wireline
telephone. The paging system is designed to receive and store information until the
professional is ready to read it. They are most commonly used by home appliance
servicemen, office equipment servicemen, doctors, photographers, and in the last
few years they have become very popular with teenagers. Different paging systems
have varying capabilities. The message received and stored may be as simple as
‘‘Call the paging center to pick up your message’’ or it may give the number of the
caller or an alphanumeric message, or in some of the more sophisticated systems the
caller can leave a voice message. The important difference between paging and other
systems of communication is that in paging, there is no need for an immediate
The POCSAG Paging System
In this section, we discuss the design and operation of one of the simpler paging
systems currently in use. This is the Post Office Code Standardization Advisory
Group (POCSAG) system. This system was introduced in the early 1980s as a
standard for the manufacture of pagers for the British Post Office. It can handle up to
2 million addresses per carrier and supports tone only (alert only), numeric, and
alphanumeric pagers. There are three speeds at which the POCSAG system transmits
its messages; they are 512, 1200, and 2400 bps. These data rates would normally be
considered to be slow but this is deliberate because, combined with high transmitter
power (hundreds of watts to a few kilowatts), it improves reliability. The message is
‘‘broadcast’’ over the entire area of operation and it is supposed to reach the
recipients whether they are in a building, on a highway, or in an airplane. Typical
carrier frequency of operation of the transmitters is around 150 MHz. Each message
is preceded by a CAP code which is a unique 7 or 8 digit code recognizable to only
one paging receiver in the geographic area of operation. The Paging Transmitter. Figure 11.11 shows a block diagram of the
transmit portion of the paging system. Most of the messages come in over the
telephone system. The source of the message can be from a Touch-tone1 telephone
whose keypad can be used to enter the information to be transmitted. It can be a
voice message, in which case a human dispatcher in the paging center has to
intervene and key in the appropriate message. The message can also come from a
computer with the appropriate software and a modem. Whatever its source or form,
Figure 11.11. Block diagram of the transmit portion of the paging system.
the message goes into an A=D converter. The digital output is used to drive a
frequency shift keying encoder in which the digit 0 is assigned a frequency of, say,
1200 Hz and the digit 1 is represented by a tone of frequency 2400 Hz. The message
is placed in a queue with other messages. The appropriate CAP code is inserted
ahead of each message frame. The dual tone signals may be sent over landlines or
wireless systems to a large number of frequency modulated transmitters distributed
over a geographic area. Component Circuit Design. The function of the ‘‘processor’’ is to
condition the analog signals coming over the telephone line into the paging center
for the A=D converter. The human dispatcher plays the same role. The design of the
A=D converter is described in Section The frequency shift keying encoder is
a form of modem. Modem circuits are described in Section 9.4.1. The frequency
modulated (FM) radio transmitter was the subject of Chapter 4. The Paging Receiver. The block diagram of the basic paging
receiver is shown in Figure 11.12 [4]. The paging receiver is typically a small
device which can be worn on a waist belt. It is basically an FM receiver with a fixed
carrier frequency. It has an internal antenna typical of portable radios. Each receiver
has a unique CAP code programmed into it and when a message arrives with the
appropriate CAP code, the message is saved in the memory and the controller
triggers the alert generator which sends a signal to the alert transducer. All other
messages are ignored. The alert may be in the form of a sub-audio vibration, a
Figure 11.12. Block diagram of the paging receiver.
chime, a beep, or a short excerpt of a well known tune. For a ‘‘tone only’’ paging
receiver, the wearer is simply alerted and has to place a call to a messaging center to
get the message. For numeric and alphanumeric receivers, the stored message can be
retrieved by pushing the appropriate buttons on the front of the device. On receiving
the appropriate commands from the keypad, the controller causes the output data
control to send the stored information to the decoder which converts it into a form
suitable for display on the liquid crystal display (LCD). The messages remain in the
memory until they are cleared. Component Circuit Design. The frequency modulated radio receiver was the subject of Chapter 5. A frequency shift keying decoder was discussed in
Section 9.4.1 (modems). Memory circuits, their control, coding, and decoding are
discussed in Appendix E. Liquid Crystal Display. Certain chemical compounds, such as the
cyanobiphenols, have the property that causes the rotation of polarized light passing
through them. These compounds are normally transparent to visible light but when
seen in a container they appear to be translucent. This is because the axes of the
molecules are normally randomly oriented and hence they scatter light in random
directions. When an electric field is applied to the compound, the axes of the
molecules line up and, depending on the orientation of the incident polarized light,
they allow the light to go through or stop it [5]. It is possible in some of the most
commonly used liquid crystals to make the light ‘‘twist’’ or gradually change its
orientation as it travels through the liquid. This phenomenon is known as twisted
nematic and the degree of twist can be set during manufacture. This is the basis of
the common liquid crystal display which is used in watches, pagers, and many other
electronic consumer goods.
Figure 11.13 shows two parallel transparent electrodes with backings of polarizing film, one vertically oriented and the other horizontally oriented. The space
between the electrodes is filled with the liquid crystal compound.
When the light enters the vertically polarized film, only the vertical component
will pass through and enter the liquid crystal. As it travels from the right-hand
electrode to the left-hand electrode its orientation is changed from the vertical to the
horizontal. If the left-hand side polarizing film is oriented horizontally, the light will
pass through it.
When an electric field is applied across the electrodes the axes of the molecules
are lined up such that they do not affect the orientation of the light. The vertically
polarized light cannot go through the horizontally polarized left-hand side film. The
contrast created by the presence or absence of the electric field is exploited in the
application of the liquid crystal as a display transducer.
In the display of alphanumeric characters, the most commonly used units are the
seven-segment and the dot-matrix displays [6]. The dot-matrix display is the more
versatile of the two but its decoding system is more complex. The decoding of
information in a binary format for display by the relatively simple seven-segment
display is not as simple as might be expected and an explanation of how it is
designed will be an unnecessary diversion at this point. The seven-segment
LCD and its driver, the MC5400=7400 series integrated circuit, are presented in
Appendix F.
Other Paging Systems
Since the early 1980s when the POCSAG system was introduced, a number of new
systems have been developed. They have increased the speed of transmission,
lowered the current drain on the battery, increased the number of addresses per
carrier, improved reliability, and made the system more difficult to tamper with. Two
of these systems are described very briefly below.
(1) ERMES (European Radio Message System). This system was introduced
in the early 1990s by the European Community. The data rate is fixed at
6250 bps and it is capable of operating over multiple radio-frequency
channels. The pager can scan all the channels when the subscriber is away
from his home base.
(2) FLEXTM (Flexible wide-area paging protocol). This system was introduced in 1993 as a high performance multi-speed paging protocol (1600,
3200 and 6400 bps). FLEX can support over 5 109 addresses and
conserves the pager battery life by sending data in specified time slots only
At the end of the 20th century it was estimated that there were 192 million pages
in use world-wide [8].
Figure 11.13. An illustration of the operation of the liquid crystal as a display transducer. Reprinted
with permission from W. C. O’Mara, Liquid Crystal Flat Panel Display, Van Nostrand Reinhold, New
York, 1993.
The cordless telephone was designed to liberate the telephone user from the tether
that the handset cord is. Before cordless telephones appeared on the market, long
handset cords were used to increase the distance between the handset and the base
but the longer the cord got the more clumsy it became. The cordless telephone not
only increased the distance from the base, it made the handset completely portable.
System Design
Figure 11.14 shows the configuration of the cordless telephone. The base station is
connected directly to the Public Switched Telephone Network (PSTN) and, from the
point of view of the PSTN, it is just another telephone set. In fact, part of it is a
transceiver which provides a two-way link to the handset. It can transmit signals to
the handset and receive signals from the handset for onward transmission to the
PSTN. The telephone part of the system is the same as any wireline telephone set
(see Chapter 8) and the radio part of it uses frequency modulation (see Chapters 4
and 5) in the 900 MHz band. The base station transmitter operates at one frequency
(say, 925.997 MHz) while the handset transmitter operates at another frequency (say,
902.052 MHz). This is an example of two simplex systems which form a frequencydivision duplex (FDD). A device called a duplexer provides a coupling between the
antenna and both the transmitter and the receiver. Other frequencies were used in the
past and a new generation of cordless telephones, using digital technology, have
been assigned spectra in the 2.4 GHz band in North America.
The handset antenna is coupled to both the FM transmitter and receiver by the
duplexer. Separate amplifiers condition the signal from the microphone and the
signal going to the speaker appropriately.
Component Design
The designs of all the components in Figure 11.14 were discussed earlier with the
exception of the duplexer. The Radio-Frequency Duplexer. The role of the RF duplexer is to
couple the strong signal from the transmitter to the antenna with minimal loss but
prevent the transmitter signal from reaching the input of the receiver. In many
applications, such as radar and the wired telephone system, a hybrid performs this
function (see Sections and 8.4.3). The duplexers in both the base station and
the handset have to ensure that the path from their transmitter to their receiver has
the highest possible attenuation so that no significant local feedback is possible. At
the same time they must ensure that the path from the transmitter to the antenna and
that from the antenna to the receiver have minimum attenuation. There are two
important factors to consider. The fact that the transmit and receive frequencies are
separated by a fairly wide margin of approximately 25 MHz is a great help in the
design. The fact that the signal power from the transmitter of the handset to its
Figure 11.14. Block diagram of the cordless telephone: (a) the base station, (b) the handset.
antenna is several orders of magnitude greater than the signal power reaching the
handset antenna from the base station makes the design of the duplex filters critical.
The same is true for the base station duplexer.
The connection of the antenna to the duplex filters is shown in Figure 11.15(a).
The attenuation characteristics of typical transmit and receive filters are shown in
Figure 11.15(b). Note that at the center of the transmit filter passband (BPF(a);
forward) there is a loss of about 2.5 dB but at this frequency the receive filter
(BPF(b); reverse) has an attenuation in excess of 50 dB [9]. This means that the path
of the transmitter signal to the antenna (double-headed arrow) has minimal loss
while the path of the transmitter signal to the input of the receiver has very high
attenuation. This is the critical path because, if the transmitter signal was allowed to
leaking into the input of the receiver, there would be a possibility of instability in the
system. It is also critical that the weak signal from the antenna reaches the input of
the receiver with minimal attenuation (single-headed arrow). Leakage of this signal
into the output of the transmitter is not critical.
Figure 11.15. (a) A block diagram of the transmit and receive surface acoustic wave (SAW)
filters, (b) the frequency characteristics of the two filters.
There are three types of RF duplexer technologies used in cordless telephones.
They are ceramic, Surface Acoustic Wave (SAW), and Film Bulk Acoustic
Resonator (FBAR). Ceramic RF duplexers are considered to have superior operating
characteristics compared to SAW devices but they are, in general, bulkier. FBAR is a
new technology that is supposed to match the performance while occupying only
about 10% of the space of ceramic devices. Surface Acoustic Wave Radio Frequency Duplex Filter. The
design of filters is outside the scope of this book; however, the basics of a surface
acoustic wave (SAW) device are easy to appreciate. A typical SAW device is shown
in Figure 11.16. It consists of a piezoelectric substrate such as quartz, lithium
niobate (LiNbO3), lithium tantalate (LiTaO3) and lithium borate (Li2B4O7). These
crystalline materials are cut at specific angles relative to the crystal axes and highly
polished. Thin-film inter-digital electrodes are deposited on the surface as shown in
Figure 11.16. Piezoelectric materials have the property of undergoing a mechanical
deformation when an electric field is applied to them. Equally, they produce an
electric field when they are deformed mechanically. When an alternating voltage is
applied to the input electrode, it causes a mechanical deformation on the surface of
the crystal which propagates in both directions, left and right. Part of the wave
travelling to the left will be reflected at the end of the substrate and there will be
losses associated with it. The wave travelling to the right will reach the output
electrode where the mechanical wave will induce a voltage in the receive electrode.
The amount of coupling between the two electrodes depends on a large number of
factors such as the frequency of the signal, the crystal material, the angle of the cut
of the crystal relative to its axes, and the geometry of the inter-digital transducers.
The most widely used SAW devices in cordless and cellular telephones are not
considered to operate in a truly surface acoustic wave mode. They are generally
described as leaky surface acoustic wave (LSAW) devices. This nomenclature is
Figure 11.16. A typical surface acoustic wave (SAW) filter showing the input and output
transducers. Reprinted with permission from C. K. Campbell, Surface Acoustic Wave Devices for
Mobile and Wireless Communication, Academic Press, San Diego, CA, 1998.
appropriate because the mechanical deformation that takes place in LSAW devices
does not limit itself to the surface of the crystal; it penetrates the material.
Disadvantages of the Analog Cordless Telephone
The range of the cordless telephone is limited to less than 200 m because its radiated
RF power is limited to less than 0.75 mW. The distance limitation is not necessarily a
disadvantage since the frequencies on which it operates can be reused by other
cordless telephone users outside the 200 meter radius with minimal or no interference. The real disadvantage of the system is that, because it uses a radio link with
no attempts at encryption, anyone with an FM radio capable of operating in the
frequency bands assigned to cordless telephones can tune in and listen to the
conversation. As mentioned earlier a new frequency band at 2.4 GHz has been
assigned to cordless telephones in North America. These cordless telephones are
digital, have a longer range (400 m), and use Spread Spectrum technology to ensure
that private conversations remain private.
The cellular telephone system can be considered to be an ‘‘enlarged’’ form of the
cordless telephone. A number of steps have to be taken to make such an expansion
(1) To increase the range of the mobile or portable the number of base stations
will have to be increased from one to many. This also establishes a need for
the mobile=portable to be switched from one base station to the next
seamlessly when it is on the move. This process is called handoff.
(2) To make the system economically feasible, it will have to handle a large
number of telephone conversations simultaneously. This means that the
number of channels in the base station must be increased from one to
many and interference between the channels must be avoided.
(3) Due to the variability of the path between the base station and the mobile,
fragments of the signal will find their way to the receiving antenna over
different paths, resulting in different amplitudes and phases. The vector sum
of the fragments may result in an augmentation or, with equal probability, a
total cancellation of the signal. This process is called fading. It is necessary to
take steps to keep the signal strength within acceptable limits.
A cellular telephone system which has the above attributes is the Advanced Mobile
Phone System (AMPS).
System Overview
Figure 11.17 shows the configuration of the cellular telephone system. The
hexagonal cell from which the system derives its name happens to be a convenient
way to divide a geographic area without overlap. The radiation pattern of the antenna
at the center of the cell is, in theory, a circle but using circles on a map will produce
overlaps and=or voids. The effects of geographical features such as hills, buildings
and other structures make the radiation pattern of the antenna difficult to predict. At
the center of each cell there is a base station which is connected to the Mobile
Telephone Switching Office (MTSO). The MTSO is connected to the PSTN.
The MSTO is the nerve center of the cellular system. It controls every parameter
of the conversation from when a subscriber requests service to when she terminates
the call.
Advanced Mobile Phone System (AMPS)
It has been estimated that at the end of 1997 there were 50 million cellular
telephones in the United States [3]. In Canada the number was 7.5 million. With
a growth rate of over 50% a year, the estimate for the year 2000 is close to 200
million cellular telephones in use in North America. Although there has been a shift
towards digital cellular technology, the majority of cellular telephones in use in
North America are based on AMPS. The AMPS cellular system, like the cordless
telephone, uses FM radio as part of the interface between the subscribers and the
PSTN. The AMPS uses frequency-division multiple access (FDMA). FDMA, when
applied to AMPS, means that each mobile telephone is assigned two carrier
frequencies for voice; one for communication from the base station to the mobile
(forward or downlink) and the other from the mobile to the base station (reverse or
uplink). In addition there are channels set aside for establishing calls and for
monitoring and control of the system. These are called traffic or control channels
and are shared by many mobiles.
AMPS uses two bands of frequencies: the forward channels from 869 to 894 MHz
and the reverse channels from 824 to 849 MHz. The following are part of the system
Figure 11.17. The configuration of the cellular telephone system.
(a) There are two simplex channels combined to form a duplex.
(b) For each mobile, the forward carrier frequency is 45 MHz higher than the
(c) Each channel occupies a bandwidth of 30 kHz.
(d) The total number of channels is 832. Reuse of the Channels. Figure 11.18 shows the cells clustered in
groups of seven [3]. With such an arrangement, it is possible to reuse the carrier
frequencies in all cells marked (a), (b), (c) and so on, so long as the radiated power
of each mobile in the cell is kept below the appropriate level considering the distance
between the respective cells (a), (b), (c), etc. The minimum frequency reuse distance
can vary quite widely. In densely populated centers, the cells are made smaller and
there are more of them. The radiated power levels are much lower. In more sparsely
populated areas, the cells are larger and the radiated power is higher. Setting up the Call. The steps described below have been simplified
considerably in the interest of brevity and clarity.
When a cellular telephone is turned off, the system has no way to reach the
subscriber. If the telephone is on but not yet engaged in a conversation (idle), the
Figure 11.18. The formation of seven-cell clusters which permit frequency reuse under
specified radiated power and minimum distance of separation.
mobile scans the forward control channels and determines the strongest of the group.
It keeps checking the particular forward control channel until the signal strength falls
below an acceptable level, at which point it starts to scan the other forward control
channels again to find the strongest signal. The forward and reverse control channels
are normally standard throughout the system so the mobile knows which frequencies
to scan when it is in the idle mode.
When a call comes into the MTSO, it broadcasts the mobile identification number
(MIN) of the called mobile (a 10-digit number derived from the telephone number
assigned to the device) over all the forward control channels in every base station.
The mobile recognizes its MIN and responds by identifying itself via the reverse
control channel which it has been monitoring all along. The MTSO recognizes the
MIN as a legitimate one and instructs the appropriate base station to move the
mobile’s center frequencies (forward and reverse channel pair) to an available pair of
channels. On the basis of the signal received at the base station, the power level for
the transmission from the mobile is set. A ‘‘ring’’ signal is sent to the mobile which
results in an alert in the form of a sub-audio vibration, a beep, or a pleasant short
tune. The subscriber answers the telephone and the conversation can start. The
MSTO continues to monitor the call and will initiate a handoff if and when the
quality of the channel falls below an acceptable level. A handoff may be necessary
because the mobile has moved away from the original base station resulting in a
weak signal. When the conservation is over the subscriber presses a button which
puts the mobile into the idle state and sends a message to the MSTP to terminate the
When the call originates from the mobile (the subscriber has dialed the desired
number and pressed the ‘‘talk’’ button) the mobile sends the following pieces of
(a) its MIN,
(b) a station class mark (SCM) (this is an indication of the power level of the
(c) its electronic serial number (ESN) (this is a unique 32-bit number
programmed into the unit during manufacture),
(d) the destination telephone number.
through a reverse control channel. The base station receives the information and
passes it onto the MSTO for validation. If everything is in order, the MSTO instructs
the base station and the mobile to move to an available pair of voice channels, dials
the destination telephone number, and connects the system to the PSTN. When the
call is terminated, the mobile sends an ‘‘end-of-call’’ code and goes into the idle
state. The MSTO dismantles the connection and goes into standby mode waiting for
the next call. Handoff Process. Handoff occurs when the signal reaching the base
station from the mobile gets too weak for adequate voice quality. Very often this
happens because the user of the mobile or portable is on the move away from the
current base station. The MSTO detects this from a set of measurements on the
received signal strength of the mobile and follows an algorithm which polls the base
stations in adjacent cells for various parameters such as overload conditions. It then
tests to determine whether the signal strength of the mobile in the intended new cell
is adequate before initiating the handoff. There is a built-in hysteresis so that at the
cell boundary the tendency to switch back and forth between the two base stations is
eliminated. One has to remember that there is no frequency reuse in adjacent cells.
The MTSO then sends a command to the mobile to change its frequency, select a
new SAT, and set its initial transmitter power for the new cell. The mobile responds
by turning its transmitter off and adjusting its parameters to the new values and
switching its transmitter back on. All of these steps are carried out without the user
being aware of what is going on. System Design. Figure 11.19 shows the block diagram of a voice
(downlink) channel in an AMPS cellular telephone mobile.
The Compander: The voice signal captured by the microphone is amplified and
then fed into the speech compressor. The compressor reduces the difference of
amplitude between the highest and lowest. This is necessary because the low
amplitude sounds are likely to be swamped by the noise in the system while the
high amplitudes are likely to cause the system to go into the non-linear portions of
its characteristics and generate distortion (noise) [10, 11]. The input–output
characteristics of the compressor is shown in Figure 11.20(a).
It can be seen that for low amplitude input signals the slope of the curve is quite
high, that is, the voltage gain is high while for high amplitude input signals the gain
is low. At the receiving end of the communication channel, it is necessary to reverse
the effect of the compressor on the signal by passing it through a circuit with the
opposite characteristics shown in Figure 11.20(b). The name of that circuit is the
expander. The application of a compressor followed by an expander is called
companding. The music recording industry uses this method to reduce noise; it is
more commonly known as the Dolby system. (See Section 3.4.5 for circuit design
Frequency Pre-emphasis: The energy in the spectrum of human speech tends to
be concentrated at the low-frequency end. This makes the high-frequency content of
speech vulnerable to the corrupting influence of noise. It is advantageous to raise the
level of the higher frequencies relative to the low frequencies before transmission
and restore them at the receiving end.
The frequency characteristics of the pre-emphasis filter are shown in Figure
11.21(a) together with a suitable circuit. Note that the slope of the filter is
6 dB=octave. The frequency characteristics of the de-emphasis filter together with
its circuit are shown in Figure 11.21(b).
Figure 11.19. A simplified block diagram of the mobile cellular telephone.
Figure 11.20. (a) The input–output characteristics of the compressor in the transmitter. (b) The
input–output characteristics of the expander in the receiver.
Deviation Limiter: The purpose of the deviation limiter is to limit the maximum
excursion of the incoming signal so that after frequency modulation the maximum
deviation of the FM signal will not exceed 12 kHz. If this is not done, there will be
adjacent channel interference.
Low-pass Filter: The deviation limiter restricts the signal by clipping it at its
extremities, thus producing higher harmonics of the signal. The low-pass filter
removes the harmonics.
Figure 11.21. (a) The circuit and frequency characteristics of the pre-emphasis filter used in
the transmitter. (b) The circuit and frequency characteristics of the de-emphasis filter used in the
The Supervisory Audio Tone Generator (SAT): The generator can produce
three audio signals at 5970, 6000, and 6030 Hz. A base station superimposes one of
the three tones on both the forward and reverse voice channels when it is in use. The
choice of which one to use is determined by the MSTO. The purpose of this signal is
to identify a particular mobile with a given base station during the conversation. The
base station sends the SAT to the mobile as soon as the voice channel is established
and the mobile has to re-transmit the same signal to the base station for the call to
continue. The frequencies are chosen to be outside the 3 kHz bandwidth but well
within the hearing range of the subscriber. The SAT is attenuated suing a suitable
bandpass filter so it is barely audible to the subscriber.
Frequency Modulator: The frequency modulator produces the modulated signal at
an intermediate frequency (IF) using the output of the IF oscillator. The IF carrier is
generated by a frequency synthesizer whose frequency is controlled over the forward
control channel. The frequency modulator may have a frequency multiplier before
and after it to bring it to the appropriate carrier frequency (see Figure 4.14).
Upconverter: The upconverter shifts the carrier frequency to its proper value using
the output of the RF oscillator. The RF carrier is generated by a synthesiser whose
frequency is controlled over the forward control channel. Here again, a frequency
multiplier as well as a number of filters may be necessary.
Transmitter: The transmitter consists of an RF power amplifier which is coupled to
the antenna. Note that the base station usually uses two different antennas, one for
the forward and the other for the reverse channels. No duplexer is required in that
Figure 11.22 shows the block diagram of the voice (uplink) channel.
Low-noise Amplifier: The low-noise amplifier brings the antenna signal up the
level required to operate the downconverter. In any communication system noise is a
constant threat to the integrity of the information transmitted on it. This is especially
true of analog systems such as the AMPS. The LNA typically must have a gain
greater than 16 dB with a noise figure less than 2.5 dB. It has been demonstrated that
the noise generated by the first stage of a chain of amplifiers has the most significant
(damaging) effect on the system [12]. The low-noise amplifier is the first of a chain
of amplification stages in the path of the signal coming from the radio antenna.
Considerable effort has therefore gone into the development of low-noise amplifiers
for this application. The LNA is usually a single chip integrated circuit using bipolar
junction transistor (BJT), metal-semiconductor field-effect transistor (MESFET),
complementary metal-oxide semiconductor (CMOS), and N-type metal-oxide semiconductor (NMOS) technologies. A detailed discussion of the design of low-noise
amplifiers falls outside the scope of this book. The interested reader is encouraged to
refer to the bibliography for further reading material on the subject.
Figure 11.22. A simplified block diagram of the voice (uplink) channel for the code-division multiple
access (CDMA) cellular telephone.
The Downconverter: The downconverter is responsible for removing all the
vestiges of the upconversion. The downconversion is typically carried out in two
steps: the first to an IF and the second to the baseband. It uses the output of the
frequency synthesizer.
The Demodulator: The demodulator recovers the baseband signal.
The SAT Filter: The SAT filter attenuates the audio tone to a level where it is
almost imperceptible to the subscriber.
The Reverse-emphasis Filter: The 6 dB=octave slope that was added is now
equalized by the simple RC low-pass circuit shown in Figure 11.21(b).
The Expander: This circuit reverses the effects of the compressor. The characteristics are shown in Figure 11.20(b).
The Power Amplifier: This is an audio amplifier designed to drive the circuits of
the PSTN. Component Circuit Design for the Transmit Path. The preamplifier is an audio-frequency amplifier designed to boost the signal from the
millivolt level to about 1 V. The design of audio-frequency amplifiers was discussed
in Section 2.7.
The compressor is an amplifier with a variable gain; high for low-level signals
and low for high-level signals. This is the same as an automatic gain control (AGC)
amplifier. This was the subject discussed in Section 3.4.5.
The design of the frequency pre-emphasis filter is given in Figure 11.21(a). Filter
design is beyond the scope of this book but the interested reader may refer to [13] or
any other basic circuit analysis book for a discussion of Bode plots.
The deviation limiter is essentially a voltage amplitude clipper. The amplitude
clipper was discussed in Section
The low-pass filter following the limiter is responsible for attenuating the higher
harmonics of the signal which appear at the output of the clipper. The design of
filters is outside the scope of this book. The interested reader is encouraged to refer
to a specialized text on the subject.
The supervisory audio tone (SAT) is a sinusoidal audio frequency oscillator
operating at approximately 6 kHz. The design of sinusoidal oscillators was discussed
in Section 2.4.3.
The design of the balanced frequency modulator was discussed in Section
The upconverter consists of balanced mixers, local oscillators, frequency multipliers, and various band-pass filters. The design of these components (with the
exception of filters) have been discussed in Section 4.4. For a more general
discussion of up- and downconverters, refer to Section
The discussion of the design of the FM transmitter (power amplifier and filters)
was in Chapter 4.
357 Component Circuit Design for the Receive Path. A description
of the operation of the duplexer was given in Section The low-noise
amplifier is a highly specialized integrated circuit device whose design falls outside
the scope of this book. All the other components are represented or have counterparts in the transmit path of the system described in Section
The ETACS (European Total Access Communication System) is almost identical to
AMPS. They both use FM transmissions for the air interface but there are
differences in some of the specifications, such as the bandwidth of the channels
and the system for the identification of subscribers. These differences are enough to
stop AMPS-based units from operating in Europe and vice versa. Some of the
specifications of the two systems are presented in Table 11.1
Disadvantages of Analog Cellular Telephone Systems
Early projections suggested that a digital cellular telephone system could yield a
20 : 1 improvement in spectrum efficiency relative to AMPS [10]. Later calculations
produced less optimistic projections of 5–10 times improvement.
Messages sent on analog cellular systems are subject to interception by anyone
with a frequency scanner and an elementary knowledge of electronics. The system
can be further compromised by the unauthorized collection of the access codes by
persons and=or organizations who use these to create ‘‘clones’’ with ‘‘free’’ access to
the telephone system.
Conversion to digital format can improve privacy if only because the equipment
necessary to intercept digitally coded messages is much more complex and
expensive. Secondly, it is relatively easy to use encryption codes with digitized
messages. Digital techniques also have the potential to identify fraudulent use of the
cellular telephone system more quickly and to deal with it by denying service
TABLE 11.1
Multiple access
Channel bandwidth
Traffic channel per RF channel
Reverse channel bandwidth
Forward channel bandwidth
Voice modulation
Peak deviation: voice channels control=wideband data
Data rate on control=wideband channel
Number of channels
30 kHz
824–849 MHz
869–894 MHz
12 kHz=8 kHz
10 kbps
25 kHz
890–915 MHz
935–960 MHz
10 kHz=6:4 kHz
8 kbps
Source: Ref. [3]
The code-division multiple access cellular telephone system was designed to
improve the capacity of the system and enhance the security of the messages as
well as prevent fraudulent access to cellular telephone services [14]. The projected
improvement in spectrum efficiency was a very powerful economic incentive for the
service providers. Since no new bandwidth was assigned to the CDMA cellular
system, it had to coexist with AMPS. CDMA is a wideband system requiring
1.25 MHz in each of the two cellular frequency bands (forward and reverse). Thus
10% of the available bandwidth used by AMPS or 41 adjacent channels had to be
vacated to make room for the CDMA system. By the end of the 20th century it was
estimated that there were about 200 million cellular telephone sets in use in North
America. The majority of these operate under AMPS and it is unlikely that the
subscribers will be willing to switch over to the CDMA system in a short time span.
Switch over to a completely digital system implies a huge investment in new
infrastructure by the cellular telephone service providers. In the meantime it became
clear that new subscribers were mainly consumers who used the system during nonbusiness hours and were very conscious of costs. It was therefore necessary for
cellular telephone manufacturers to produce portables=mobiles and base stations
capable of dual-mode operation.
System Design of the Transmit Path
In what follows, only the part of the system concerned with the path of the message
signals is included. Even then much has been excluded in the interest of brevity and
simplicity. The readers who need more detailed information may refer to the
bibliography and the references at the end of the chapter.
A simplified block diagram of the path taken by the message through the CDMA
cellular telephone system is shown in Figure 11.23. The voice is sampled at
8000 Hz and converted into digital form by the analog-to-digital converter
(ADC) and, at 8 bits=sample, produces a 64 kbps data stream. It is then used to
digitally modulate (multiply) the output of the pseudo-noise (PN) generator in a
Figure 11.23. A simplified block diagram of the transmitter of a code-division multiple access
(CDMA) cellular telephone system.
balanced mixer. The output of the mixer is then used to modulate the output of the IF
carrier (sinusoidal). The resulting signal is upconverted by a second sinusoidal
carrier, amplified and radiated by the antenna. Note that it is assumed that the
appropriate filters are placed in the correct positions to make the process work
Component Circuit Design for the Transmit Path
Analog-to-Digital Converter: The design and operation of the ADC was described
in Section
Pseudo-noise Sequence Generator: A more accurate name for this circuit is
Linear Recursive Sequence (LRS) generator. Consider a set of flip-flops connected
as a shift register shown in Figure 11.24(a). Assume that there are four flip-flops and
that each stage of the shift register is set arbitrarily as shown. If we connect the
output to the input and apply the clock pulse, the effect will be to shift the data 1100
one stage for each clock pulse. We have a ring counter which will continue to
circulate the data (....001100110011....) so long as the clock pulses are supplied. In
this configuration the sequence repeats itself after four clock pulse, the same as the
number of stages in the ring counter. Suppose we modify the circuit by adding the
Figure 11.24. The four flip-flop linear recursive sequence (LRS) generator whose truth table is
shown in Table 11.3.
TABLE 11.2
two-input exclusive OR (XOR) as shown in Figure 11.24(b). Now the digit fed back
to FF0 is determined by the outputs of FF2 and FF3 through the XOR. This circuit
configuration is arbitrary, one of several that could have been used [15]. However,
the object of this exercise is to generate the longest sequence for the minimum
number of flip-flops in the ring counter. Using the truth table for the two-input XOR
given in Table 11.2, we can construct the truth table of the four flip-flop LRS shown
in Table 11.3.
Table 11.3 shows that the sequence repeats itself after the 15th cock pulse. In this
case the initial states of the four flip-flops were carefully chosen to give the sequence
its maximum length. Other initial states may end prematurely with 0s at the outputs
of all the flip-flops. The LRS cannot recover from the all-zero state. It can be shown
that, for a four-stage LRS, the maximum length of the sequence before it repeats
itself is given by (2n 1 ¼ 15) where n ¼ 4. The CDMA system sues a 42-stage
LRS and hence repeats itself after (242 1) cycles! Anyone looking at the possibility
TABLE 11.3
of breaking the code by the observation of repeated patterns has a long wait. Code
breaking is made more difficult because knowledge of the chip clock rate, the
structure of the LRS (number of flip-flops and feedback circuit) and the initial
settings of the shift registers are a few of the parameters required to reproduce the
pseudo-noise sequence.
Other Components in the Transmit Path: The IF modulator using QPSK was
discussed in Section 11.4.42. The upconverter was discussed in Section 10.7.1 and
the RF power amplifier in Section 2.8.
System Design of the Receive Path
A block diagram of the receiver is given in Figure 11.25. The LNA receives its input
from the antenna and conditions it for downconversion from the RF to the IF stage.
After the second balanced mixer, the signal then goes to the QPSK demodulator
which produces a replica of the message data with the chips from the transmitter PN
present [16]. The PN must be extracted from the digitally modulated signal in order
to recover the original (digitized voice) signal. The local LRS generates chips using
identical settings used in the transmitter. It is then phase-aligned with the received
signal by sliding (sliding correlator) the local code past the received code. The
microprocessor keeps track of the number of bits that match and stops the sliding
action when a peak occurs (delay lock). The locally generated LRS code is then
mixed (multiplied) with the digitally modulated signal (demodulation). This step
separates the LRS from the digitized voice signal. A low pass filter removes that
unwanted products of the demodulation. The DAC converts the digitized voice to an
analog signal which is amplified and fed to the speaker.
Figure 11.25. Shows a simplified block diagram of the voice (downlink) channel for the codedivision multiple access (CDMA) cellular telephone.
Component Circuit Design for the Receive Path
All the components required for the downconversion and the QPSK demodulation
are the same as those used in the upconversion and the QPSK modulation. The
difference is in the type of filtering used following the two process. The correlation
and the locking action are implemented in software and do not have an equivalent
USDC (United States Digital Cellular): This system was designed to increase the
capacity of the cellular telephone service. In one AMPS channel, USDC supports
three subscribers. USDC is a time-division multiple access (TDMA) system and it
uses the same FDD for the downlink and uplink separated by 45 MHz. Like the
system based on CDMA, conversion to USDC has to be carried out within the
bandwidth reserved for cellular telephone services. It uses p=4 differential quadrature phase shift keying in a 30 kHz per channel FDD. Service providers are obliged
to continue service in the AMPS format when they introduce the USDC. The USDC
uses the same frequency reuse plan and operates from the same base stations as
AMPS. Some of the latest mobiles are designed to operate with both USDC and
AMPS. The plan is to fade out AMPS slowly and increase the presence of its digital
counterparts. The problem facing USDC is that the CDMA system seems to have
gained a wider acceptance and it may well become the dominant system.
GSM (Global System for Mobile Communications): The standards adopted for
the development of the second-generation cellular telephone system was put together
by the Group Special Mobile (GSM) a subsidiary of the Conference of European
Postal and Telecomunications Administrations (CEPT). The system took on the
French name of the group that proposed it. It primarily uses the frequency bands 890
to 915 MHz and 935 to 960 MHz but has also been assigned spectrum between 1.7
and 1.9 GHz. It is a digital system designed to be used throughout Europe and to
replace the many incompatible systems in existence. GSM uses both FDMA and
TDMA but it does not accommodate dual operation with any of the many analog
systems currently in use in Europe. GSM is incompatible with all of the North
America cellular systems and it is not likely that North Americans visiting Europe or
Europeans in North America will be able to use their own cellular telephones any
time soon. In the meantime, cellular telephone manufacturers have started to market
a ‘‘triple mode’’ telephone which can work with the AMPS (analog), the USDC or
CDMA (both digital) as well as the GSM (digital) systems. These telephones are
essentially three different telephones housed in the same handset (mobile) with an
automatic switch-over feature. Some of these cellular telephones can be used as
pagers, they can receive voice mail and have wireless access to the Internet.
1. Groe, J. B. and Larson, L. E., CDMA Mobile Radio Design, Artech House, Boston, MA,
2. Garg, V. K. and Wilkes, J. E., Principles and Applications of GSM, Prentice-Hall, Upper
Saddle River, NJ, 1999.
3. Rappaport, T. S. Wireless Communications: Principles and Practice, Prentice-Hall, Upper
Saddle River, NJ, 1996.
5. O’Mara, W. C., Liquid Crystal Flat Panel Display, Van Nostrand Reinhold, New York,
6. Diefenderfer, A. J., Holton, B. E., Principles of Electronic Instrumentation, Saunders
College Publishing, Philadelphia, 1994.
9. Campbell, C. K., Surface Acoustic Wave Devices for Mobile and Wireless Communication,
Academic Press, San Diego, CA, 1998.
10. Goodman, D. J., Wireless Personal Communications Systems, Addison-Wesley Longman,
Reading, MA, 1997.
11. Spanias, A. S. ‘‘Speech Coding; A Tutorial Review’’, IEEE Proc. Oct. 1994.
12. Smith, J., Modern Communication Circuits, McGraw-Hill, New York, 1986.
13. Irwin, J. D., Basic Engineering Circuit Analysis, 5th Ed., Prentice-Hall, Upper Saddle
River, NJ, 1996.
14. Telecommunications Industry Association=Electronic Industry Association Interim Standard-95, ‘‘Mobile Station—Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System’’, July 1993.
15. Dixon, R. C., Spread Spectrum Systems, John Wiley, New York, 1976.
16. Stremler, F. G., Introduction to Communication Systems, 3rd Ed., Addison-Wesley,
Reading, MA 1990.
Gilberg, B., ‘‘The Design of Bipolar Si=SiGe LNAs from the Ground Up’’, BCTM97 Short
Course Notes.
Ghapurey, R., and Viswanathan, T., ‘‘Design of Front End RF Circuits’’, Proc. 1998 Southwest
Symp. On Mixed-Signal Design, pp. 134–139.
Shaefer, D. K., and Lee, T. H., ‘‘A 1.5 V, 1.5 GHz CMOS Low-Noise Amplifier’’, IEEE J.
Solid-State Circuits, May 1997, pp. 745–759.
Steele, R., Mobile Radio Communications, IEEE Press, New York, 1992.
Tuttlebee, W. H. W., Cordless Telecommunications Worldwide, Springer-Verlag, Berlin, 1997.
Brodsky, I., Wireless: The Revolution in Personal Telecommunications, Artech House, Boston,
MA, 1995.
A pager is a one-way communication system. Why is it so popular when
cellular telephones can provide two-way communication?
Compare the cordless telephone to the cellular telephone in terms of
transmitted power, mobility, and privacy.
What are the differences between a conventional wireline telephone system
and a personal wireless telephone system?
Why is it necessary to control the transmitted power of each mobile unit in a
cellular telephone system?
Describe the procedures followed in a cellular telephone system when a
mobile is handed over from one base station to another. What is the role of
hysteresis in the handoff process?
In AMPS each mobile is assigned a bandwidth of 30 kHz but only approximately 3 kHz is required for a normal telephone conversation. Explain why?
Explain the following: frequency-division multiple access (FDMA), timedivision multiple access (TDMA), frequency hopped multiple access
(FHMA), and code-division multiple access (CDMA).
Analog-to-Ditigal Convertor
Automatic Gain Control
Amplitude Modulation
Advanced Mobile Phone System
Binary Frequency Shift Keying
Bipolar Junction Transistor
Band Pass Filter
Binary Phase Shift Keying
Code-Division Multiple Access
Conference of European Postal and Telecommunications Administrations
Complementary Metal-Oxide Semiconductor Transistor
Central Office
Digital-to-Analog Convertor
Differential Quadrature Phase Shift Keying
Digital Signal Processing
Electronic Industry Association
European Radio Message System
Electronic Serial Number
European Total Access Communication System
Film Bulk Acoustic Resonator
Frequency-Division Duplex
Frequency-Division Multiple Access
Frequency Hopping
Frequency Hopped Multiple Access
Frequency Hopped Spread Spectrum
Frequency-Shift Keying based Paging System
Frequency Modulation
Frequency Shift Keying
Global System for Mobile Communication
Institute of Electrical and Electronics Engineers
Intermediate Frequency
Integrated Services Digital Network
International Telecommunications Union
Liquid Crystal Display
Low Noise Amplifier
Linear Recursive Sequence
Leaky Surface Acoustic Wave
Lower Single Side Band
Metal-Semiconductor Field-Effect Transistor
Mobile Identification Number
Mobile Telephone Switching Office
N-type Metal-Oxide Semiconductor Transistor
Non-return to Zero
Pulse Code Modulation
Personal Communication System
Post Office Code Standardization Advisory Group
Phase Shift Keying
Public Switched Telephone Network
Quadrature Amplitude Modulation
Quadrature Phase Shift Keying
Radio Frequency
Supervisory Audio Tone
Surface Acoustic Wave
Station Class Mark
Space Division Multiple Access
Signal-to-Noise Ratio
Spread Spectrum
Single Sideband
Signaling Tone
Total Access Communication System
Time-Division Duplex
Time-Division Multiple Access
Telecommunications Industry Association
Upper Single Sideband
Voltage Controlled Oscillator
In this chapter the characteristics of the media in which the transmission of signals
takes place will be discussed. It so happens that we humans basically communicate
through speech=hearing and by sight. Human hearing is from 20 Hz to 20 kHz and
we can see only the portion of radiation spectrum from about 4:3 1014 Hz
(infrared; l ¼ 7 107 m) to approximately 7:5 1014 Hz (ultraviolet;
l ¼ 4 107 m). These communication channels occupy only small portions of
the detectable frequency spectrum which has no lower boundary but has an upper
boundary of about 1022 Hz (gamma rays). Acoustic radiation in the frequency range
20–20 kHz is attenuated quite severely in our environment even when attempts are
made to guide it along a conduit. It is therefore quite inefficient to transmit an
acoustic signal over any distance which would qualify as telecommunication. The
same observation can be made about visible light. To communicate over distances
greater than what we can bridge by shouting, or see reliably, it is necessary to
convert the signal into another form that can be guided (by wire, waveguide, or
optical fiber) or which can be radiated efficiently in free space.
Wire, coaxial cables, waveguides, optical fiber, and free space transmission have
characteristics which vary as frequency changes. A medium may be efficient in one
frequency range but quite unsuitable for another frequency range. But efficiency is
not the sole criterion for the choice of the frequency to which audio and video
signals have to be translated for transmission. To help keep some order and to
minimize interference among the various users of communication services, it is
necessary to assign various frequency bands for specific uses and governments
arrogate to themselves the right to demand a licensing fee for the use of these bands.
For example, satellite communication has been assigned 4–6, 12–14, and 19–
29 GHz but there is no technical reason why they cannot operate at frequencies in
between these frequencies or indeed outside them.
This consists of two insulated wires twisted together to form a pair. Several to many
hundred pairs may be put together to form a cable. When this is done it is usual to
use different pitches of twist in order to limit electromagnetic coupling between
them and hence cross-talk. The conductor material is copper, usually numbers 19,
22, 24, and 26 American Wire Gauge (AWG), and the insulation is usually
polyethylene. Wax-treated paper insulation was used in the past but the ingress of
moisture into the cable was a problem in most applications; it is still a problem even
with polyethylene insulated cables which are sometimes filled with grease-like
substances to take up all the air spaces and thus discourage moisture from entering.
Such cables may be suspended from poles where they are easy and inexpensive to
service but are aesthetically undesirable, or buried which make them expensive and
difficult to repair.
The frequency characteristics of a BST 26-gauge non-loaded cable terminated in
900 O are shown in Figure 12.1. It can be seen that the twisted pair has a low-pass
characteristic. It should be noted that, contrary to expectation, the primary constants
of the twisted pair (series resistance, shunt capacitance, series inductance and shunt
conductance, all per unit length) change with frequency. The bandwidth of the
twisted pair can be extended to a higher frequency by inductive loading of the line.
Lumped inductances are connected in series with the line at specified distances. The
best results are obtained when the interval is kept short and the value of the lumped
inductance is kept low, thus minimizing the discontinuities introduced by loading.
The frequency responses of a 12,000 ft (3.7 km) number 26-gauge with 900 O
terminations for the loaded and unloaded cases are shown in Figure 12.2.
Figure 12.1. Frequency characteristics of 26 gauge BST non-loaded cable terminated in
900 O.
Figure 12.2. Comparison of loaded and unloaded 12,000 feet (3.7 km) number 26 gauge cable
terminated in 900 O and 2 mF. Reprinted with permission from Transmission Systems for
Communications, 5th Ed., AT&T, Bell Labs, 1982.
It can be seen that, while loading solves the problem of limited bandwidth for the
typical subscriber loop voice channel, it is quite inadequate for analog (the basic
supergroup requires 552 kHz bandwidth) and digital (DS-1 requires 1.5 MHz
bandwidth) carrier applications, for which it is used. In both these cases, the line
has to be equalized by placing amplifiers or repeaters at specific distances along its
length that emphasize the high-frequency response or regenerate the pulses. Lines
used for digital transmission require phase equalization as well, otherwise pulse
degradation due to dispersion takes place. Dispersion causes the rate of rise of the
leading and trailing edges of the pulse to slow down and the base to spread out over a
much longer time than the original pulse.
It can be seen from Figure 12.1 that there is a flat loss at lower frequencies, so it is
usual to combine the equalizer with an amplifier. An amplifier used for this purpose
is called a repeater. A repeater can take a number of forms. In a two-wire system
where signals flow in both directions, a negative impedance converter is coupled in
series and=or in shunt with the line through a transformer. The configuration of the
negative-impedance converter and its connection to the line are shown in Figure
Measures have to be taken to ensure that the negative impedance does not
overwhelm the line impedance resulting in oscillation. The introduction of repeaters
Figure 12.3. The connection of the negative impedance converter (NIC) to the telephone line.
Reprinted with permission from Transmission Systems for Communications, 4th Ed., AT&T, Bell
Labs, 1970.
into the cable causes an impedance mismatch at the point of connection and this can
cause echo problems. Severe echo on the cable can impair the speech of most
telephone users. There are circuits built into the cable or at the terminations to cancel
the echo.
Negative-Impedance Converter
The negative-impedance converter is a two-port which converts an impedance
connected to one port into the negative of the impedance at the other port. Consider
the two-port shown in Figure 12.4 terminated at port 2 by an impedance ZL .
If the two-port is a negative impedance converter (NIC) then
Zin ¼ kZL
where k is a constant.
Figure 12.4. A two-port with its defining voltages and currents.
Such a two-port is best described by a chain matrix equation
This gives
V1 ¼ AV2 BI2
I1 ¼ CV2 DI2 :
I2 ZL ¼ V2
From the termination we have
Substituting into Equations (12.2.3) and (12.2.4) gives
V1 ¼ AV2 þ
I1 ¼ CV2 þ
But for a two-port
Zin ¼
V1 AZL þ B
For ZL to be equal to kZL , B ¼ C ¼ 0, then
Zin ¼
so that
There are two possibilities:
Both matrices satisfy the condition for a negative-impedance converter, namely
Zin ¼
Z ¼ kZL
k2 L
where k ¼ k1 =k2
From the first matrix,
V1 ¼ k1 V2 :
This is called the voltage negative impedance converter or VNIC [5].
From the second matrix,
I1 ¼ k2 ðI2 Þ
This is called the current negative-impedance converter, INIC or CNIC.
Without loss of generality, we can make k1 ¼ k2 ¼ 1 so that
The NIC is an example of what is described as a degenerate two-port, that is, it
cannot be described by open-circuit impedance [Z] nor short-circuit admittance [Y ]
parameters. However, it has chain and hybrid parameters (both ½h and ½k).
The VNIC may be described in terms of its hybrid k parameters as follows:
0 1
The basic form of the VNIC is shown in Figure 12.5.
Figure 12.5. The basic form of the voltage negative-impedance converter.
Figure 12.6. The T-equivalent model of the bipolar transistor.
The transistors Q1 and Q2 may be represented by the low-frequency T-equivalent
model shown in Figure 12.6.
Assuming that C1 and C2 are short-circuits at the frequency of operation, the
equivalent circuit of the VNIC is as shown in Figure 12.7.
The hybrid k parameters of the circuit in Figure 12.7 are
1 a2
6 R þ r þ r ð1 a Þ
6 1
a2 R2
re1 þ ðrb1 þ R2 Þð1 a1 Þ
R1 þ re2 þ rb2 ð1 a2 Þ
Figure 12.7. The basic VNIC when the transistors have been replaced with the T-equivalent
Figure 12.8. The circuit of the voltage negative-impedance converter.
When R1 ¼ R2 ¼ R and re is small compared to R and a1 ¼ a2 1, the circuit
behaves like a VNIC. The practical version of the VNIC circuit is shown in Figure
Four-wire Repeater
In a four-wire system, the forward and return paths are different and ordinary
amplifiers may be used. This is shown in Figure 12.9. Again precautions have to be
taken to counteract the possibility of instability through the hybrid-to-hybrid feedback path.
As frequency increases, the twisted pair has the tendency to lose signal power
through radiation. Ultimately, its usefulness is limited by cross-talk between pairs.
Figure 12.9. The use of ordinary amplifiers on the telephone line with 2-to-4 wire hybrid.
Reprinted with permission from Transmission Systems for Communications, 4th Ed., AT&T, Bell
Labs, 1970.
In a coaxial cable, one conductor is in the form of a tube with the second running
concentrically along the axis. The inner conductor is supported by a solid dielectric
or by discs of dielectric material placed at regular intervals along its length. A
number of these cable are usually combined together with twisted pairs to form a
multi-pair cable.
The structure of the coaxial cable ensures that, at normal operating frequencies,
the electromagnetic field generated by the current flowing in it is confined to the
dielectric. Radiation is therefore severely limited. At the same time, the outer
conductor (normally grounded) protects the cable from extraneous signals such as
noise and cross-talk.
The primary constants of the coaxial cable are much better behaved than those of
the twisted pair. The inductance, L, capacitance, C, and conductance, G, per unit
length are, in general, independent of frequency. The resistance, R, perpunit
ffiffiffi length is
a function of frequency due to skin effect; it varies as a function of f .
The frequency characteristics of a 0.375 inch (9.5 mm) coaxial cable are shown in
Figure 12.10. As expected, the coaxial cable has a much larger bandwidth than the
twisted pair. However, it still requires repeaters and frequency equalizers for analog
lines and phase equalization for digital signal transmission. The characteristics of the
repeaters are usually adaptively controlled to correct for changes in temperature and
other operating conditions.
Coaxial cable is used for transmitting data at 274.176 Mbit=s in the LD-4 (BellCanada) and T4M (Bell System in the USA) systems. They have 4032 voice
Figure 12.10. The insertion characteristics of a terminated 0.375 inch coaxial cable. Reprinted
with permission from Transmission Systems for Communications, 4th Ed., AT&T, Bell Labs,
channels or the equivalent video or digital data traffic. Its regenerators are spaced at
1.8 km intervals and the total length of the line can be 6500 km [1].
Specially constructed coaxial cables with repeaters of very high reliability are
used for submarine cable systems. Because of the very high cost of these cables, they
are used to transmit messages in both directions by assigning separate frequency
bands to each direction. In spite of the development of satellite communication
channels, submarine cables are still viable for trans-Atlantic and trans-Pacific traffic.
Because of the propagation delay involved in the signal travelling to the satellite and
back, most trans-Atlantic telephone conversations use the satellite link in one
direction only; cable is used in the opposite direction. In 1976, the TAT-6 (SG)
trans-Atlantic cable system was installed. It used a 43 mm diameter coaxial cable
with a 4200 voice channel capacity over a distance of 4000 km [2].
A waveguide may be viewed as a coaxial cable with the central conductor removed.
The outer conductor guides the propagation of the electromagnetic wave. In its most
common form it has a rectangular cross section with an aspect ratio of 2 : 1. The
wider dimension must be about one-half the wavelength of the wave which it will
transmit. Therefore the waveguide has a low-frequency cut-off. There are a number
of modes in which the wave can propagate but in every case the electric and
magnetic fields are orthogonal. When the electric field is at right angles to the axis of
the waveguide, it is described as transverse electric (TE) mode. When the magnetic
field is at right angles to the axis it is called transverse magnetic (TM) mode.
The mechanical structure of the waveguide disqualifies it from being used for
long-haul transmission. Irregularities on the walls, such as projections, holes, lack of
a perfect match at joints, bends, twists and imperfect impedance matching at the
terminations, can cause reflection and spurious modes to be generated, all of which
result in signal loss.
Waveguides are used mainly as feedlines to antennas in terrestrial microwave
relay systems and for frequencies above 18 GHz they are superior to all other media
in terms of loss, noise and power handling.
The use of optical fiber as a medium for telecommunication was made possible by a
coincidence of the development of a number of technologies.
(1) The laser which is a coherent frequency source of the order of 1014 Hz and it
can be modulated. A light emitting diode (LED) which produces noncoherent light can also be used.
(2) A low-loss glass fiber which can be used as a waveguide for the light.
(3) A detector for the signal at the receiving end.
The laser can be modulated at a rate in the range of 109 bit=s while the LED can
operate at 108 bit=s. The information bearing capacity of the system is enormous.
On-going research continues to increase the bit rate limits.
The optical fiber is essentially a high quality glass rod of about 50 mm diameter
for multi-mode propagation and 8 mm for single-mode propagation. The mechanical
properties of a glass rod that small will make the system impracticable. In practice, a
second layer of glass concentric with the optical fiber proper is deposited in the
outside, bringing the overall diameter to 125 mm. The outer glass sheathing, referred
to as cladding, has a different refractive index and the signal is therefore confined to
the core. The core may be of uniform refractive index or it may be graded. These
techniques have resulted in optical fibers that have attenuation less than 0.2 dB=km.
Various types of protective covering may be put on the fiber and several fibers put
together to form a cable.
The optical receiver is a reverse-biassed semiconductor junction diode and it is
coupled to the fiber so that the incoming light falls on the junction. The energy in the
light is transferred to the electrons in the semiconductor lattice, causing them to
break away and move into the conduction band. The high electric field sweeps the
electrons out of the junction into the external circuit where they can be detected as a
current. Silicon diodes have been used for 1 mm wavelength detectors. For longer
wavelengths, such as 1.3 and 1.5 mm, germanium, InAs and InSb are used [3].
In 1988, a trans-Atlantic optical fibre communication system went into service
(TAT-8). It spans a distance of 6500 km and provides the equivalent of 40,000
telephone channels. It operates on the 1.3 mm wavelength; repeater separation is
35 km and the bit rate is 274 Mbit=s.
The transmission media discussed earlier had one thing in common; the propagation
of the signal was guided by a twisted pair, a coaxial cable, a waveguide, or an optical
fiber. We now consider transmission systems which rely on propagation through free
In 1873 Maxwell showed that electromagnetic waves can propagate through free
space. It took three decades to demonstrate experimentally that this was possible,
when Hertz constructed the first high-frequency oscillator – this was the famous
spark-gap apparatus.
A large number of factors have to be taken into account when designing a free
space propagation communication system including the following:
the distance between transmitter and receiver,
the carrier frequency of the transmission,
the physical size of the antenna to be used,
the power to be radiated
the effect of the transmission on other users of the same and adjacent
Figure 12.11. Free-space transmission showing various paths. Reprinted with permission from B. P.
Lathi, Modern Digital and Analog Communication Systems, CBS College Publishing, New York, 1983.
Publisher’s Note:
Permission to reproduce
this image online was not
granted by the copyright
holder. Readers are kindly
requested to refer to the
printed version of this article.
Figure 12.11 shows different paths by which a signal radiated from a transmitter
can reach a receiver.
Direct Wave
As the name suggests, the signal travels directly from the transmitter antenna to the
receiver antenna. This requires that the two antennas be in line-of-sight. The direct
wave is the major mode of propagation for medium wave AM radio (540–
1600 kHz), commercial FM broadcasting (88–108 MHz), terrestrial microwave
relay systems, used for long-distance telephone and television signals (2, 4, 6, 11,
18, and 30 GHz), and satellite transmission systems (4, 6, 8, 12, 14, 17–21, and 27–
31 GHz).
Earth-Reflected Wave
Part of the propagated wave is reflected off the surface of the Earth and may arrive at
the antenna with a different phase from the direct wave. Depending on the magnitude
of the reflected wave, it can cause signal fluctuation and sometimes even complete
cancellation of the direct wave. This phenomenon has the most noticeable effect on
terrestrial microwave relay systems where automatic gain control, diversity protection, and adaptive equalization may be used to counteract it.
Troposphere-Reflected Wave
At a distance of approximately 10 km above the Earth’s surface, there is an abrupt
change in the dielectric constant of the atmosphere. This portion of the upper
atmosphere is called the troposphere. High-frequency signals, such as those used in
terrestrial microwave relay systems, may be reflected from the troposphere and have
the same effect as the Earth-reflected wave at the receiver. On the other hand, the
troposphere may be used as part of the communication channel in cases where direct
line-of-sight conditions are not possible. This is called tropospheric scatter propagation and the best frequencies for this are 1, 2, and 5 GHz. It is most commonly used
by the military for communication over difficult terrain, typically over a distance of
300–500 km.
Sky-Reflected Wave
Surrounding the Earth at an elevation of approximately 70–400 km is a layer of
ionized air caused by constant bombardment of ultraviolet, a, b and g radiation from
the Sun as well as cosmic rays. It is called the ionosphere and it consists of several
layers which have different reflective, refractive, and absorptive effects on radio
When a radio signal reaches the ionosphere, a number of things can happen
depending on the frequency and the angle of incidence. The wave may be reflected
back to Earth, it may undergo refraction and eventually be returned to the Earth or it
may pass through the ionosphere and escape into outer space. With the correct
choice of transmission frequency and angle of incidence it is possible to establish
communication between two points on the Earth’s surface where line-of-sight does
not exist. This is the basis of short-wave (3–30 MHz) transmission.
The ionosphere changes its nature from day to night and with such phenomena as
sunspot cycles and other cosmic events. The turbulent nature of the ionosphere
makes communication by short-wave rather unreliable since the signal is subject to
fading – both long and short term. This is due to cancellation and=or reinforcement
of the different parts of the signal arriving at the receiver by diverse routes. To
improve the reliability of communication via the ionosphere, automatic gain control
and diversity protection techniques, such as the utilization of two or more carrier
frequencies, are used.
The signal returning to the Earth after reflection from the ionosphere may be
reflected from the Earth’s surface up to the ionosphere once more. On its second
return to Earth it may be received by a suitably placed receiver. This phenomenon is
described as multiple hop and can be used to reach places beyond the single-hop
Surface Wave
At very low frequencies (VLF: 3–30 kHz), the ionosphere and the Earth’s surface
form two parallel conducting planes and can act as a waveguide. VLF signals are
used for worldwide communication and navigational aids. At higher frequencies,
temperature inversions and other local phenomena can generate a surface wave
which can add to the problem of fading.
Terrestrial microwave radio is a relatively inexpensive medium for long-haul
telephone and television signals. Its assigned frequencies are 2, 4, 6, 11, and
18 GHz, which make the system a line-of-sight operation. The distance between
repeater stations is approximately 40 km at the lowest frequency and 3 km at the
highest frequency, where rain can cause severe attenuation. It is well suited to
difficult terrain where the cost of burying or stringing up on posts any form of cable
would be prohibitively expensive. The repeater stations can be placed in strategic
positions such as hill tops with easy access for maintenance personnel.
The block diagrams of a microwave transmitter and receiver are identical to that
used in any other radio system. There are intermediate-frequency amplifiers,
oscillators, modulators (upconverters), demodulators (downconverters) filters, equalizers, and automatic gain control (AGC). However, because of the higher frequencies
involved, the hardware used to realize the various functions look very different.
What appears to the untrained eye to be a piece of printed circuit board may be, in
fact, tuned circuits, transmission lines, open and short-circuits and so on.
At microwave frequencies, the electromagnetic wave behaves increasingly like
light. Therefore antennas for microwave radio take the form of parabolic dishes or
the horn reflector (hog-horn); these structures focus the electromagnetic wave into a
narrow beam for optimal transmission. Parabolic dishes are usually used for one
frequency band only and the signals may be polarized in the vertical or horizontal
direction. The horn reflector type are multi-band and may also be polarized.
A detailed discussion of the design of microwave, antennas, oscillators, amplifiers, modulators, frequency changes, and so on, is outside the scope of this book. A
limited list of books on the subject are given in the bibliography.
Analog Radio
In analog radio, a signal made up of a large number of voice frequency telephone
channels or its equivalent is formed into a basic group, supergroup, etc. The
formation of the signal to be transmitted was discussed in Section 9.2. The
appropriate subcarrier frequencies were given. The next step is to use the composite
signal to modulate an intermediate-frequency carrier of 70 or 140 MHz before the
final upconversion to microwave frequencies. Frequency modulation is the preferred
method. Terminal Transmitter and Receiver. The block diagram of the
transmitter is shown in Figure 12.12(a). The modulating signal, 70 MHz, itself
modulated by 3600 voice-frequency telephone channels (jumbo group) or its
equivalent, is amplified and used to drive the modulator, mixer or upconverter.
The other input to the upconverter is from the microwave oscillator. The output of
the upconverter has to be bandpass filtered to remove undesirable products of the
modulation process. After amplification by the radio-frequency amplifier, the signal
goes to the transmit antenna for radiation.
The receiver is shown in Figure 12.12(b). The received signal is bandpass filtered
to eliminate all but the required signal. It is then mixed with the output of the local
oscillator to produce an intermediate-frequency signal. After amplification, equalization and the application of AGC, it goes to a demodulator where the original
modulating signal (70 MHz) is recovered. It should be noted that, in the case of the
3600 voice-frequency telephone channels, several levels of demodulation have to be
carried out before the voice frequency signals are recovered. Repeater. A typical analog microwave radio repeater is shown in
Figure 12.13. The antenna is very highly directive and it must be secured in a
position where it faces the transmitter antenna as directly as possible; small deviation
can cause significant loss of signal power. Antennas 1 and 2 are used simultaneously
for transmission and reception. The received signal, RF1, goes to the receive
circulator which directs it to the bandpass filter where all signals other than those
desired are attenuated. The local oscillator and the mixer downconvert the signal to
an intermediate frequency. After amplification, equalization, and the application of
AGC, the signal is upconverted by the microwave oscillator and the upconvert mixer
Figure 12.12. A typical analog microwave radio transmitter and receiver.
Figure 12.13. A typical analog microwave repeater. Note that the antennas are used for transmission
and reception simultaneously.
to a different frequency, RF2 (changing the frequency reduces the possibility of
instability in the system). The RF amplifier boosts the signal and feeds it to the
transmit circulator which passes it on to the antenna for onward transmission.
The signal travelling in the opposite direction, RF3 (different from both RF1 and
RF2), follows an identical path in the opposite direction. The intermediate frequency
used in this path is, in general, different from the one used earlier. The signal leaves
the system at a frequency RF4, different from all the others.
Digital Radio
All the functional blocks of the analog radio shown in Figure 12.13 are present in
some form or another in the digital radio. The baseband signal to be transmitted may
be the output of a digital switch operating at a rate of 1.544, 6.312, 44.736 or
274.176 Mbit=s. The binary output of the switch is used in a modulation scheme
similar to that of the modem discussed in Section 9.4.1. The digit 1 is assigned a
frequency f1 and 0 is assigned a different frequency f2. Usually f1 and f2 are equally
displaced from the carrier frequency f0, that is, ðf0 f1 Þ ¼ ðf2 f0 Þ. At the receiving
end, the demodulation process converts f1 and f2 to 1s and 0s. Regenerative Repeater. The block diagram of the regenerative
repeater is shown in Figure 12.14. The system is identical to that shown in Figure
12.13 except that the signal is demodulated to the baseband and then regenerated,
modulated, filtered, upconverted, and amplified for onward transmission. The signal
travelling in the opposite direction is subjected to the same processing steps, with the
exception that the carrier radio frequencies are selected to minimize the possibility of
instability. The regenerative repeater is described in Section
A satellite transmission system is just another microwave relay system with a single
repeater located in outer space. Because of its height, it can cover a large area of the
globe, making it possible to cover the entire surface with three geostationary
satellites. It has applications in point-to-point communications as well as broadcasting and it can reach remote parts of the Earth where other systems cannot reach
without large expenditures of money and effort. All that is required is the satellite
and the terminal equipment in the Earth stations. Satellites are, however, not cheap.
The cost of launching them and the difficulty of making repairs should anything go
wrong dictate very high levels of reliability.
To avoid the complexity of tracking the satellite from rising to setting and to
maintain continuous communication, it is best to ‘‘park’’ the satellite at a point
directly above the equator and to choose the correct speed in the direction of rotation
of the Earth so it stays in a relatively fixed position above a reference point on the
Earth’s surface. In general, satellites rotate in an elliptical orbit with the Earth at one
of the two foci of the ellipse (apogee and perigee). The two foci can be at the same
Figure 12.14. The regenerative repeater used in digital radio.
point, in which case the orbit is circular. There are distinct advantages to having a
satellite in a circular orbit since this fixes the distance travelled by the message and
hence the delay. To satisfy these conditions, an Earth satellite has to be 42,230 km
from the center of the Earth. But it is not enough to park the satellite in this position,
as it tends to drift away slowly due to the non-spherical shape of the Earth, the
gravitational influence of the Sun, the Moon and the other planets. A small rocket is
provided for the correction of minor deviations from the nominal position. The
useful life of the satellite is therefore determined by how long the rocket fuel lasts.
The satellite is made to spin or it has a wheel in it which spins. The gyroscopic effect
of this helps to further stabilize the satellite. If the satellite itself spins, it is necessary
to spin the antenna in the opposite direction to keep it pointing at the Earth at all
For about 30 minutes a day for several days around the equinoxes (March 21 and
September 21) the Sun is directly behind the satellite and the electrical noise
generated by the Sun makes operation impossible. It is necessary to switch to an
alternate satellite or rely on more Earth-bound means of communication. Once a day,
the satellite is eclipsed by the Earth and it not only loses its source of power from the
solar panels, it experiences a drastic change in temperature. To overcome these
problems, a battery is provided and all components are designed to operate in the
extreme temperature conditions of outer space.
Placing the satellite above the equator means that Earth stations in the extreme
north and south are not well illuminated by the satellite radiation. Furthermore, the
low angle of elevation of the Earth station antenna makes it vulnerable to atmospheric fading and interference from terrestrial manmade noise. One solution to this
problem is to use several satellites in non-equatorial, non-stationary orbits so that, as
one satellite sets, the next one is rising. This means that tracking is necessary. The
Russian domestic communication satellite system (Molniya) was designed on this
basis since most of the territory it is designed to serve is in the northern part of the
Northern Hemisphere.
The frequency bands assigned for satellite communication are 4–6, 12–14 and
19–29 GHz. In the higher frequency bands, there is increasing attenuation due to
signal power dissipation in rain droplets and water vapor in the atmosphere.
Frequency reuse is possible so long as the satellites are sufficiently far apart in
orbit for an Earth station to focus on only one of them at a time.
The repeater on a satellite is usually referred to as a transponder. The basic
structure of a typical transponder is shown in Figure 12.15. The antenna is used for
both transmission and reception. The path for a signal arriving from Earth is through
the circulator to the bandpass filter where all but the desired band of frequencies
containing the carrier are eliminated. Some gain is provided by an amplifier and the
mixer and the local oscillator change the carrier frequency to a different value. After
further bandpass filtering to remove the unwanted products of the mixing process,
the power amplifier boosts the signal power to the appropriate level for the return trip
to Earth through the circulator and antenna. System instability is prevented by the
correct choice of circulator characteristics and the change of the carrier frequency.
Publisher’s Note:
Permission to reproduce
this image online was not
granted by the copyright
holder. Readers are kindly
requested to refer to the
printed version of this article.
Figure 12.15. The communication satellite transponder. Reprinted with permission from V. I.
Johannes, ‘‘Transmission Systems for Telecommunications’’ in Electronic Engineers’ Handbook,
Fink, D. G. and Christiansen, D. (Eds), McGraw-Hill, Inc, New York, 1982.
The loss of signal power between an Earth station and the satellite is about 200 dB
[10]. The transponder power output is usually about 10 W; higher power is not
desirable because of the possibility of interference with other Earth-based communication systems. Besides, it is expensive to provide the dc power to run a highpower amplifier in outer space. To overcome these limitations, Earth stations use
large antennas and high power transmitters and very low-noise amplifiers in the
receiver. Sometimes the amplifiers are cooled with liquid nitrogen to achieve the
necessary low-noise performance. Figure 12.16 shows the gains and losses of power
in a television signal as it makes its way up to the satellite and back down to Earth.
The great distances traversed by the signal from one station to the other causes
enough delay (240 ms) to make a telephone conversation over the satellite channel
somewhat difficult; one person starts to talk when the other is already talking,
without realizing it. This leads to a series of interruptions in the conversation. In
general this is not intolerable but the situation certainly gets much worse when the
delay is doubled for a two-way conversation. To get around this problem, the normal
practice in a trans-Atlantic telephone conversation is that when one link is via
satellite, the other must be via submarine cable. The satellite system also generates
an echo that has the psychological effect of making people with no speech
impediment stutter. The circuits used in satellite transmission usually have an
echo cancellation feature.
To give some idea of how quickly satellite communication systems are changing,
it is worth noting that, in early 1990, there were about 94 communication satellites in
operation with another 194 launchings planned in the near future. By the end of the
year 2000, there was a total of 2271 satellites in orbit (Russia had 1322, the United
States 658, Japan 55 and 51 were international) [11]. It should be noted that not all
of the satellites listed are for civilian communication. The first of a series of the
International Telecommunication Satellites (INTELSAT I) was launched in 1965 and
it had 240 two-way telephone voice channels. In 1980 INTELSAT V went into
Figure 12.16. Power levels in the transmission of a TV signal via satellite. Reprinted with
permission from B. I. Edelson, ‘‘Global Satellite Communications’’, Scientific American, February
service with 12,000 two-way voice and two television channels. By the end of the
year 2000 there were 19 Intelsat satellites in orbit with plans for 22 new launches.
The latest operational satellite, the INTELSAT VIII, has 22,500 two-way telephone
voice and three television channels, and with the use of digital circuit multiplication
equipment (DCME) the number of voice channels can be increased up to112,500
1. Telecom Canada, Digital Network Notes, Ottawa, Canada, 1983.
2. Easton, R. L., ‘‘Undersea Cable Systems – A Survey’’, IEEE Comm. Magazine, 13(5), 12–
15, 1975.
3. Jones, W. E., Introduction to Optical Fiber Communication Systems, Holt, Rinehart and
Winston, New York, 1988.
4. Members of the Technical Staff, Transmission Systems for Communications, 5th Ed., Bell
Telephone Laboratories Inc., 1982. Publisher: Western Electric Co. Inc., Winston-Salem,
5. Linvill, J. G., ‘‘Transistor Negative Impedance Converter’’, Proc. IRE., 41, 725–729, 1953.
6. Li, T. ‘‘Advances in Optical Fibre Communication: an Historical Perspective’’, IEEE J.
Selected Areas in Comm. SAC-1(3), 356–372. 1983.
7. Schwartz, M. I., ‘‘Optical Fiber Transmission – From Conception to Prominence in 20
Years’’, IEEE Comm. Magazine, 22(5), 38–48, 1984.
8. Pritchard W. L., ‘‘The History & Future of Commercial Satellite Communications’’, IEEE
Comm. Magazine, 22(5), 22–37, 1984.
9. Johannes, V. I., Electronics Engineers’ Handbook, Fink, D. G. and Christiansen, D. (Eds),
McGraw-Hill, Inc, New York, 1982.
10. Edelson, B. I., ‘‘Global Satellite Communications’’, Sci. Am., Feb., 1977.
Basawaptna, G. R. and Stancliff, R. B., ‘‘A Unified Approach to the Design of Wideband
Microwave Solid-State Oscillators’’, IEEE Trans. MTT-27(5), 379–385, 1979.
Edwards, T. C., Foundations for Microstrip Circuit Design, John Wiley & Sons, New York,
Ghandi, O. P., Microwave Engineering Applications, Pergamon Press, New York, 1981.
Gonzalez, G., Microwave Transistor Amplifiers: Analysis and Design, Prentice-Hall, Englewood Cliffs, NJ, 1984.
Ha, T. T., Solid-State Microwave Amplifier Design, John Wiley & Sons, New York, 1981.
Ishii, T. K., Microwave Engineering, Harcourt Brace Jovanovich, Publishers, San Diego CA,
Liao, S. Y., Microwave Circuit Analysis and Amplifier Design, Prentice-Hall, Englewood
Cliffs, NJ, 1987.
Midford, T. A. and Bernick, R. L., ‘‘Millimeter-wave CW-IMPATT Diodes and Oscillators’’,
IEEE, MTT-27(5), 483–492, 1979.
Sadiku, M. N. O., Elements of Electromagnetism, Saunders College Publishing, New York,
White, J., Microwave Semiconductor Engineering, Van Nostrand Reinhold, New York, 1982.
Describe how you would transmit a video signal of approximate bandwidth
1.5 MHz over a twisted-pair cable.
Derive an expression for the input impedance of the circuit shown in Figure
P12.1, assuming that the operational amplifier is ideal. How can this circuit be
used to improve the transmission characteristics of the twisted-pair or coaxial
Figure P12.1.
Describe two techniques which can be used for the compensation of gain loss
on a twisted-pair and=or coaxial cable. Discuss how these techniques can be
modified to compensate for the relative loss of gain as a function of
Discuss the merits of optical fiber transmission relative to satellite communication.
Discuss the role of the ionosphere in the free space propagation of very low
frequencies (VLF) and short-wave transmission between two points on the
Earth’s surface.
Terrestrial microwave radio transmission is subject to severe attenuation in the
atmosphere. What elements in the atmosphere are responsible for this; can
you suggest a simple explanation of this phenomenon?
Describe the processes that take place at a typical terrestrial microwave
analog relay station using a suitable block diagram. What are ‘‘upconversion’’
and ‘‘downconversion’’ and why are they necessary?
Using suitable block diagrams, discuss the major differences between analog
and digital microwave relay stations.
Compare and contrast the satellite transponder to the microwave relay station
paying particular attention to signal processes that occur in them.
A transformer is a device which is used for coupling a signal or power from a source
to a load in an efficient manner. It does this by changing the impedance of the source
to match that of the load. A transformer can also be used for phase inversion and dc
isolation. The transformer consists of two or more coils of wire wound on a common
core so that the magnetic field of one coil links with the other coils. When a
changing current is applied to one winding, it causes a changing magnetic field in
the common core. The changing magnetic field induces a voltage in all the other
coils and if a path exists, current will flow. The coils are said to be mutually coupled.
Consider the two coils connected as shown in Figure A.1. The application of
KVL to loop 1 gives
joL11 I1 þ joM21 I2 E ¼ 0
where the voltage joM21 I2 is induced in loop 1 as a result of the current I2 flowing in
loop 2 and it opposes the applied voltage E.
The application of KVL to loop 2 gives
joM12 I1 ðjoL22 þ ZL ÞI2 ¼ 0:
where the voltage joM12 I1 is induced in loop 2 as a result of the current I1 flowing in
loop 1.
It is evident that
M12 ¼ M21 ¼ M
where M is the mutual inductance between the two coils.
Figure A.1. An equivalent circuit of a two-winding transformer showing the applied voltages
and currents.
Two coils may be closely or loosely coupled, depending on how much of the
magnetic field generated by one links with the other. The coupling coefficient k is
related to the mutual inductance M and the self-inductances of the two coils by
k ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi :
L11 L22
A transformer may be considered to be ideal if it satisfies all the conditions below:
a¼ 1¼ 1¼ 2¼
n2 V2 I1
(2) The primary inductance L11 and the secondary inductance L22 both approach
infinity while the ratio L11 =L22 remains finite.
(3) The coupling coefficient k ¼ 1.
(4) The windings have no resistance and the magnetic core has no losses.
(5) Stray capacitances of the coils are zero and stray capacitance between the
coils is zero.
To investigate the impedance transformation property of the transformer, consider
the circuit shown in Figure A.2. Assuming that the resistance seen across the
primary of the transformer is Rin then power flowing into the transformer is
Pin ¼
Figure A.2. The equivalent circuit of the ideal transformer used for matching the source
resistance, RS to the load resistance RL .
and power flowing into the load RL is
P0 ¼
Since the transformer is ideal, Pin ¼ Pout . Therefore,
Rin ¼
RL :
Rin ¼ a2 RL :
The source then ‘‘sees’’ a load equal to a2 RL instead of RL if the load has been
connected directly to the source. Since the turns ratio, a is a variable, it can be
chosen so as to optimize the transfer of power to the load. This occurs when
a2 RL ¼ Rs :
The result can be generalized for the load impedance
ZL ¼ RL þ jXL
ZS ¼ RS þ jXS
and the source impedance
ZS* ¼ a2 ZL :
A transformer can be constructed in which the conditions in Equation (A.5) apply
within a reasonable approximation. However, it is impossible to make the selfinductances L11 and L22 infinite. The coupling coefficient k can be made approximately equal to unity by using bifilar windings on a high permeability core. Winding
resistance and core losses can be reduced considerably by a judicious choice of
materials. Stray capacitance can be minimized by increasing the thickness of
insulation between layers of the windings. By taking into account all the imperfections, an equivalent circuit of the practical transformer is as shown in Figure A.3,
R1 is the winding resistance of the primary,
R2 is the winding resistance of the secondary,
L1 is the leakage inductance of the primary,
L2 is the leakage inductance of the secondary,
C1 is the total stray capacitance of the primary,
C2 is the total stray capacitance of the secondary,
Rc represents an equivalent resistance in which the core losses are dissipated,
Lm is the primary inductance in which the magnetizing current flows,
C m is the total stray capacitance between primary and secondary.
Figure A.3. The equivalent circuit of a practical transformer. The ideal transformer is used in
conjunction with the non-ideal components which are shown outside the black-box containing
the ideal transformer.
A number of the components shown in Figure A.3 can be made negligible by taking
the following precautions:
(1) Using low resistance wire for the windings will make R1 and R2 negligibly
small compared to the source and load resistances.
(2) Using bifilar windings in which the primary and secondary coils are wound
side-by-side and turn-by-turn, on a highly permeable core, the leakage
inductances L1 and L2 can be considered to have been eliminated. It is
evident that the use of bifilar windings will increase the value of Cm .
(3) The use of a thicker insulation between layers of winding will reduce C1 and
C2 but will increase the leakage inductances.
(4) By laminating the core, and thus increasing the resistance of the path for eddy
currents to flow, the core losses can be reduced and therefore Rc can be left
out of the equivalent circuit.
It should be noted that some of the steps described above are contradictory and
suitable compromises have to be made to get the simplified equivalent circuit shown
in Figure A.4.
The ideal transformer and its load RL can be replaced by a resistance equal to
a2 RL as shown in Figure A.5. By applying Thevenin’s theorem to the left of X X 0
the circuit can be simplified as shown in Figure A.6
Vth ¼
a2 R L
V :
þ RS S
a2 RL
It can be seen that this is a high-pass filter with a slope of 6 dB per octave and a
corner frequency
oc ¼
a2 RL RS
Lm ða2 RL þ RS Þ
Figure A.4 The simplified equivalent circuit of the practical transformer when steps have been
taken to minimize the effects of the non-ideal components.
Figure A.5 The equivalent circuit for the analysis of the frequency characteristics of the
Figure A.6 The Thevenin equivalent circuit of the transformer showing that it is a high-pass
It is advantageous to make Lm (the primary inductance L11 ) as large as possible in
order to keep the corner frequency as low as possible.
The condition for maximum power transfer exists if
a2 RL ¼ RS
oc ¼
30 300 Hz
300 3 kHz
3 30 kHz
30 300 kHz
300 3 MHz
3 30 MHz
30 300 MHz
300 3000 MHz
3 30 GHz
30 300 GHz
Extremely Low Frequency ðELFÞ
Voice Frequency ðVFÞ
Very Low Frequency ðVLFÞ
Low Frequency ðLFÞ
Medium Wave Frequency ðMWÞ
Short Wave Frequency ðSWÞ
Very High Frequency ðVHFÞ
Ultra High Frequency ðUHFÞ
Super High Frequency ðSHFÞ
Extremely High Frequency ðEHFÞ
Frequency band (MHz)
Video carrier (MHz)
(Not normally used)
Commercial FM Broadcast Band (88–108 MHz)
Frequency band (MHz)
Video carrier (MHz)
537.25 etc.
The channels follow in order, each one being 6 MHz in bandwidth with the carrier at
1.25 MHz above the lower cut-off frequency. The total number of channels is 69 and
channel 69 has a bandwidth from 800–806 with the carrier at 801.25 MHz. It should
be noted that in every case the FM carrier for the audio is 4.5 MHz above the video
Figure C.1 is a simplified electromagnetic spectrum with various uses and bandwidths shown.
Figure C.1.
Reprinted with permission from McConnell, Bodson and Urban, Fax: Facsimile
Technology and Systems, 3rd Ed., Artech House, Boston, MA, 1999.
The Modified Huffman data compression code for facsimile machines is given in
Tables D.1 and D.2. Table D.1 covers black and white runs of 0–63 and Table D.2
covers run lengths from 64 to the maximum of 1728. Run lengths not given in Table
D.2 can be made up from both tables.
Run length
Code word (white)
Code word (black)
(continued )
TABLE D.1 (continued )
Run length
Code word (white)
Code word (black
Run length
Code word (white)
Code word (black)
*EOL ¼ End of Line.
Modern electronic systems would be inconceivable without the possibility to store
information. Information in the form of music has been stored on wax cylinders,
vinyl discs, magnetic tapes and compact discs (CD). As electronic systems become
increasingly digital, the need to store large quantities of data has grown. In digital
electronic storage systems every bit (1 or 0) is stored in a device or circuit which can
take only one of two states at any time. There has to be a mechanism for changing
the state of the device or circuit according to the nature of the bit to be stored. It must
be possible for the stored information to be retrieved when required. Magnetic tapes
and discs are the preferred media for the storage of electronic data where the data is
to be retained even after the system power is switched off. A second medium
commonly used for data storage is the flip-flop which can take one of two states
(high or low corresponding to 1 or 0). In this case when the system power is shut off,
the information stored is lost.
Figure E.1(a) shows the NOT gate (non-resettable) equivalent of the flip-flop, (b)
shows the resettable equivalent using NOR gates and (c) shows a second realization
using NAND gates.
In the flip-flop circuit shown in Figure E.1(b), the S and R terminals are the
are the outputs and they are complementary (that is when
control inputs. Q and Q
one of them is a 1 the other must be a 0). There are four different combinations of 1s
and 0s that can exist at the terminals S and R.
(1a) If we assume that S ¼ R ¼ 0 then, since we are dealing with two-input
NOR gates, it follows that the output, Q (and its complement), is determined
by the logic levels present at the input terminals. If we assume that Q
Figure E.1. (a) Two NOT gates connected to form a non-resettable flip-flop. (b) The connection of two two-input NOR gates to form an S-R flip-flop (resettable). (c) The connection of two
NAND gates to form an S-R flip-flop.
then the lower input terminal of NOR gate, x is a 1 and hence the output of
NOR gate x is a 0 (Q ¼ 0). It follows that the upper input terminal of NOR
¼ 1, as initially assumed. There is
gate y is a 0 and hence the output Q
consistency throughout.
(1b) We now repeat the process using the condition S ¼ R ¼ 0 and assuming this
¼ 0. A quick check shows that there is consistency throughout.
time that Q
This means that there are two possible logic states, one with Q ¼ 0 and
¼ 1 and the other with Q ¼ 1 and Q
¼ 0. The first state is referred to as
the reset or clear state and the second as the set or preset state. The reset
state (Q ¼ 0) is important because it usually represents the starting point of
the logic operation that is to follow. However, in terms of storing and
retrieving bits, the condition S ¼ R ¼ 0 is not a promising prospect because
the output is independent of the input.
(2) We now consider the situation when S ¼ 0 and R ¼ 1. Following the
¼ 1Þ is the only possibility
procedure used above, we find that Q ¼ 0 ðQ
(that is, the reset state). Consider the reset state (S ¼ 0 and R ¼ 1 and hence
Q ¼ 0). If we now go back to Figure E.1 and change R from 1 to 0, we find
that the flip-flop remains in the same (reset) state as before.
(3) Similarly, with the opposite initial condition, that is, S ¼ 1 and R ¼ 0, the
¼ 0) and a change in S from 1
flip-flop will be in the set state with Q ¼ 1 ðQ
to 0 so that S ¼ R ¼ 0 will not affect the state of the circuit.
Qðt þ tÞ
Not used
Note used
Condition in (1a)
The flip-flop goes from the reset to the set state
Condition in (2); the reset state is not affected
This can lead to an ambiguous state
This is the condition in (1b)
Condition (3); the set state is not affected
The flip-flop goes from the set to the reset state
This can lead to an ambiguous state
The final possibility is that S ¼ R ¼ 1. From Figure E.1(b) this is not
possible since both outputs have to be 0s, thus violating the condition for
them to be complementary to each other. The condition S ¼ R ¼ 1 is not
Table E.1 is a summary of the above.
The operation of the flip-flop circuit shown in Figure E.1(c) is left as an exercise
to the reader.
A flip-flop is normally a part of a much larger system of gates. Each gate takes a
finite amount of time to change its state so there is no guarantee that the two inputs
(S and R) of a flip-flop will arrive simultaneously or whether one will arrive before
the other. For example, because of delays in their respective paths, suppose S ¼ 0
and R ¼ 1 and they are both supposed to change states simultaneously, then if S
arrived a little ahead of R, we would momentarily have S ¼ 1 and R ¼ 1; an
unacceptable condition which could lead to ambiguity and hence errors. It is clearly
advantageous to wait for the proper values of S and R to be established before
allowing the state of the flip-flop to change. This can be achieved with a third signal,
namely, a clock pulse.
A clock pulse is used to enable a transition to occur in digital circuits. Because of
delays (variable from gate to gate) it is necessary to enable the transition at the
instant when the clock changes from enable-to-disable rather than from disable-toenable. This allows all, the variables to reach their proper values before the clock
pulse causes a transition to occur. It is therefore usual to trigger digital circuits on the
negative-going edge of the clock pulse. This is described as edge-triggered logic
(refer to Figure E.3).
Figure E.2(a) shows the NAND gate version of the S-R flip-flop with two twoinput NAND gates added. We have chosen the NAND gate version because in the
design of an integrated circuit one would normally choose an architecture that
exploits the repetition of the same cellular building block; in this case, the NAND
gate. The S-R flip-flop using NAND gates is shown in Figure E.1(c).
Figure E.2. (a) The NAND gate version of the S-R flip-flop [see Figure E.1(c)] modified to
change states only when a clock pulse is supplied. (b) A further modification which provides
terminals for ‘‘preset’’ and ‘‘clear’’.
In Figure E.2(a), NAND gate a will switch states only when both R and the clock
pulse, CP are present. This will then provide an input to the (original) S-R flip-flop.
Similarly, NAND gate b will switch states when both S and CP are present. It is now
possible to introduce the clock pulse when the inputs S and R have had time to reach
their proper binary values.
We revisit the basic characteristics of the S-R flip-flop by referring to the timing
diagram shown in Figure E.3. It is assumed here that the negative-going edge of the
clock pulse triggers the S-R flip-flop but whether it changes states or stays the same
is determined by the values of R and S when the clock pulse arrives. The values of S
and R are binary but arbitrary at the instants when the flip-flop is triggered. It can be
seen that, at t ¼ t0 , S ¼ 1, R ¼ 0 and Q ¼ 1. The arrival of the clock pulse does not
cause the flip-flop to change states. At t ¼ t1 , S ¼ 0 and R ¼ 0; the flip-flop does not
change states. At t ¼ t2 , S ¼ 1, R ¼ 0; the flip-flop does not change states for the
same reason as at t ¼ t0 . At t ¼ t3 , S ¼ 0, R ¼ 1, the arrival of the clock pulse
¼ 1. At t ¼ t4 , S ¼ 0,
causes the flip-flop to change states making Q ¼ 0 and Q
Figure E.3. The changes in the output(s) and the corresponding inputs on the arrival of a
number of clock pulses.
R ¼ 0; the flip-flop does not change states. At t ¼ t5 , S ¼ 1, R ¼ 0; the arrival of the
¼ 0. Finally, at
clock pulse causes the flip-flop to change states making Q ¼ 1 and Q
t ¼ t6 , S ¼ 0, R ¼ 1; the arrival of the clock pulse causes the flip-flop to change
¼ 1. These agrees with Table E.1.
states making Q ¼ 0 and Q
Before data can be stored in a flip-flop, it is necessary to ensure that it is in a known
state, usually the set (or clear) state (Q ¼ 0). Figure E.2(b) shows that NAND gates x
and y in Figure E.2(a) have been changed from two-input to three-input and two
NOT gates have been added. Note that when both inputs of a two-input NAND gate
are connected together, we have a NOT gate.
When it is necessary to clear (or preset) the flip-flop, a 1 is applied to the
appropriate terminal. Assume that the clock pulse is a 0 (it is undesirable to clear or
preset the flip-flop when the clock pulse is present) so that the outputs of both
NANDs a and b are 1s. If a 1 is applied to the clear terminal (0 to the preset
, will be a 1, regardless of the value of the
terminal), the output of NAND y, namely Q
signal on the third terminal. All three inputs to NAND x are 1s and Q ¼ 0. The flipflop is ready (cleared) to accept data. Similarly, a 1 applied to the preset terminal will
make Q ¼ 1.
Figure E.4. The structure of a four-stage shift register formed from S-R flip-flops.
A shift register constructed with S-R flip-flops is shown in Figure E.4. The NOT gate
connected to FF0 ensures that S0 is the complement of R0 . Thus FF0 will flip into the
state presented at the data terminal if it is not already in that state. The connection of
0 to R1 , etc., assures complementarity at the interfaces between flipQ0 to S1 and Q
flops and hence the acceptance of the data presented at the input of the next flip-flop.
A pulse can then be applied to the clear terminal to set Q0 ¼ Q1 ¼ Q2 ¼ Q3 ¼ 0.
Assume that the data to be stored in the shift register has the following four bits:
1101. The bits are presented at the data input and are clocked into the shift register
Figure E.5. The input data 1101 changes the states of the flip-flops on the arrival of each clock
pulse to store the input bits.
one bit for every clock pulse. The first to be clocked in is the 1 on the right-hand side
followed by the 0 and then the two 1s. Before the first clock pulse S0 ¼ 1, R0 ¼ 1
and Q0 ¼ 0. So after the first clock pulse, FF0 changes states and Q0 ¼ 1; all the
other FFs are not affected because their inputs were 0 during the first clock pulse.
The next bit presented at S0 ¼ 0 and S1 ¼ 1. So after the second clock pulse Q0 ¼ 0
and Q1 ¼ 1. Figure E.5 shows the waveforms of the clock, the input, and the
outputs. After the fourth clock pulse we have Q0 ¼ 1, Q1 ¼ 1, Q2 ¼ 0 and Q3 ¼ 1.
A fifth clock pulse will shift the data one position to the right, leading to the loss of
the first bit.
A memory is required to store information in the form of bits. Each bit is stored in a
flip-flop. There are various classifications of memories depending on the material
(paper tape, magnetic medium or semiconductor) used to hold the data and the mode
for storage and retrieval of the information.
When we store information in a memory, we say that it has been written into the
memory. When we retrieve data stored we say that the memory has been read.
The data can be stored magnetically on tape, disc or drum and since the magnet
retains its properties after the power is switched off, it is an example of a non-volatile
memory. Semiconductor memories which rely on the states of a set of flip-flops lose
the data stored in them when the power is switched off. They are examples of
volatile memories.
If the information is stored on tape (magnetic or paper) it is described as
sequential access memory because if we wish to retrieve data which is stored in a
different location than where we are currently, we have to wind=rewind the tape to
get there. Sequential access memories are, in general, cheap but the process of
storing and retrieving information from them is slow. When the data is stored on a
magnetic disc or drum, where it is possible for the read=write head to go directly to
the location of the information, we describe it as random access memory (RAM).
Accessing a RAM to read or write information into it is much faster than in a
sequential access memory. Semiconductor RAMs are used extensively in computers
to store large amounts of data.
Another form of memory, in which the information is only meant to be read, is the
read only memory (ROM). The information can usually be accessed randomly but
the data is placed there at the time of manufacture. A compact disc with recorded
music is an example of a ROM.
In Section E.5, we discussed a simple memory which was used to store the bits 1101.
It is advantageous to use this as a platform for the description of the semiconductor
random access memory.
The information to be stored is in the form of bits (1s and 0s). A group of bits
form a word so that 1101 is described as a four-bit word. It may represent a number
to be used in an arithmetic operation. Each bit has to be stored in a specific memory
cell. When a word is stored, the specific location or address must be known so that
the correct information can be retrieved when required.
Figure E.6 shows a single memory cell. It can be seen that there is a read=write
terminal. When the read=write terminal has the value 1 data can be written into the
cell and when it has the value 0, the output can be read. When the address line is a 1,
the memory cell is activated and it can be written into or read. The bit to be stored is
presented at the input. A suitable clock pulse (not shown) is used to drive the cell
from the current state to the new state.
(1) Write. The memory cell must be reset (Q ¼ 0) before the operation can
begin. To write data into the memory cell we set the address line to 1, the
read=write line also to 1. Now if the data to be written is presented at the
input as a 1 then a careful examination shows that the output of AND gate c
is 0 (R ¼ 0) and that of AND gate d is 1 (S ¼ 1). According to Table E.1, Q
goes from 0 to 1. At the input to AND e, the output of NOT a is 0, Q ¼ 1 and
the address line is 1. Therefore the output of AND e is 0.
(2) Read. To read the data, the address line has to be 1 and the read=write line
must be 0. The input can be 1 or 0 but this is of no interest since the
read=write line is 0 and the data cannot enter the cell. A careful examination
of the circuit shows that all three inputs to AND e are 1s and hence the output
of AND e is 1.
Note that when the address line is 0 (no address is specified), AND e is disabled and
its output is 0, independent of what happens in the rest of the circuit.
The read and write operations of the circuit in Figure E.6 when the input data is 0
are left as an exercise to the reader.
Figure E.7 shows a five-bit four-word memory. This means that the memory can
store four words, each consisting of five bits. All the read=write lines are connected
Figure E.6. A single memory cell in which the S-R flip-flop has been modified to provide a
‘‘read=write’’ and an ‘‘address’’ line.
Figure E.7. A memory capable of storing four words each of which has five bits. Note that the
input of the data is carried out serially but the output is presented in a parallel format. Circuits are
available to change the format of the data from serial to parallel and vice-versa (see Section
together and, depending on the binary signal applied, the memory can be read or data
can be stored in it. The address lines of all the cells in each row are connected
together so all the five bits which make up the word can be accessed together. The
input lines connect all the cells in each column.
In the write mode, suppose we apply a 1 to address line 2 and data is applied to
the input lines. Then it follows that the data will affect only the cells in row 2 and
they will take on the values of the bits presented at the inputs. In the read mode, the
read=write line is activated and if a 1 is placed on address line 3 then the information
at the outputs of all the cells in row 3 will appear on the output lines through the
respective OR gates. Note that this is an example of a parallel operation; the
retrieval of the five bits of the word happen simultaneously on separate lines, not one
bit at a time. Note also that the process of reading data from the memory does not
destroy the contents. The content remains until new data is written over it. This is an
example of a non-destructive reading of the contents of a memory.
A practical memory is usually much larger than the 4 5 example shown in
Figure E.7 and it is possible only in integrated circuit form. Clearly, as the size of the
memory increases, the sheer number of leads that have to be brought out of the
integrated circuit package becomes a very important consideration. Sophisticated
digital techniques are used to reduce the number of leads required for a large
memory. The interested reader is encouraged to refer to a suitable textbook on
The MC5400 and MC7400 series (see Figure F.1) belong to a family of circuits for
decoding binary coded decimal data to drive seven-segment liquid crystal displays
(LCD) and light emitting diodes (LED). Reproduced with permission from Motorola
Semiconductor Inc.
Figure F.1.
4-to-2 wire hybrid 251
2-to-4 wire hybrid 244, 298
90 phase shift circuit 125, 127
A-plane 33
A=D converter 339
Accelerator anode 165
Active LC filters 100
Active RC filters 100
Adder 126, 135
Address 412
Adjacent channel interference 334
Advanced Mobile Phone System (AMPS) 347
Alert generator 339
Alert transducer 339
Aliasing 273
All-zero state 360
Alternate mark inversion (AMI) 292
American Wire Gauge (AWG) 368
Amplifier 8
class-A 54, 55
class-B 54, 60
class-C 40
power 23, 319, 356
radio frequency 67, 143, 146
Amplitude limiter 117, 118, 148, 291
Amplitude modulation (AM) 18, 333
Amplitude modulation theory 18
Amplitude modulator 22, 163
Analog mixer 90
Analog multiplier 85
Analog radio 381
Analog-to-digital (A=D) converter 245, 248, 358
Angle modulation 333
Anode 202
ANSnet 15
Antenna 23, 71
Apogee 384
Aspect ratio 170, 376
Astable multivibrator 195, 280
Asymmetrical digital subscriber line (ADSL)
14, 298
Attack time 102
Audio frequency amplifier 22, 54
Auto-transformer 201
Automatic Frequency Control (AFC) 117,
122, 145
Automatic Gain Control (AGC) 83, 101
Automatic telephone exchange 8
Ayrton, William 9
B-H characteristics 295
Back porch 182
Baffle 106
Bain, Alexander 4
Bakewell, Frederick 5
Balanced mixer 333, 359, 361
Balanced modulator 126, 132, 269, 326
Baird, John 10
Band split filter 259
Barkhausen Criterion 28, 231
Baseband signal 187, 327, 384
Basic group 269
Basic mastergroup 270
Basic supergroup 270
Bell, Alexander Graham 6
Bell System T-1 PCM Carrier 286
Bessel functions 113
Bias point 60
Bifilar winding 394
Binary coded decimal 415
Binary digits 11
Binary Phase Shift Keying (BPSK) 333
Binary quantization 317
Bipolar junction transistor (BJT) 54, 354
Bistable Multivibrator 239, 294
Boron 312
Branly, Edouard 8
Buffer amplifier 37, 205
Bucket Brigade 312
Cable modem 14
Cable television 14
Camera tube 163
CAP code 338
Carbon microphone 216
Carbon transfer printer 308
Carey, George 9
Carrier, 18
Caselli, Giovanni 6, 9
Cathode 202
Cathode ray tube (CRT) 10, 187, 202
CCD image sensor 311
CDMA Cellular Telephone System 358
Cellular telephone 325, 347
Central office 7, 214, 243, 253
Central office tone receiver 259
Central power supply 220
Ceramic RF duplexers 346
Cesium beam atomic clock 290
Chain matrix 371
Charge-coupled devices (CCD) 30
Chat rooms 15
Chemical valency 311
Chip clock rate 361
Chips 329
Chrominance 181
Circular orbit 386
Circulator 226, 386
Classical feedback theory 26
Classical RC twin-tee filter 231
Clock pulse 407
Clock rectifier 292
Clocked S-R flip-flop 407
‘Clonning’ 357
Coaxial cable 11, 270, 375
Code Division Multiple Access (CDMA) 329
Codec 244, 286
Coherent demodulation 327
Coherent detection 327
Coherer 8
Color television subcarrier 182, 234, 299
Color television 180
Color triangle 180
Colpitts 30, 231
Common-emitter 58
Common-source amplifier 90
Commutator 273
Compact disc (CD) 11, 333, 405, 411
Compander 351
Comparator 248
Complementary metal oxide semiconductor
(CMOS) 234, 354
Complex frequency 25
Composite video signal 168, 187
Computer mouse 14
Concentration 257
Conductance 33
Consultative Committee of the International
Telegraph and Telephone 307
Control channels 348
Control grid 202
Conversion efficiency 40, 54
Convert enable 248
Cooke, William 2
Cordless telephone 325, 343
Correlation 331
Coupling coefficient 392
Cross-over distortion 61
Cross-talk 374
Crossbar switching 257
Crystal filters 100
Crystal microphone 225
Crystal-controlled oscillator 21, 22, 35, 117, 241
Cut-off 55
Damping 42
Darlington pair 321
Datalink 297
Datapac 297
Dataroute 297
De-multiplexed 182
Decelerator grid 165
Degenerate 2-port 372
DeForest, Lee 9
Delay lock 361
Demodulation 86, 103, 268
Depletion layer 312
Depth of modulation 19
Depth of notch 232
Detent 254
Diagonal clipping 103
Dielectric 375, 379
Differentiator 197, 291, 293
Digital filter 100
Digital circuit multiplication equipment (DCME)
Digital radio 384
Digital revolution 10
Digital telephone 243
Digital tone dial 234
Digital-to-analog (D=A) converter 234, 245
Diode 9
Diode modulator 47
Dipole 72, 146
Direct amplitude modulation 49
Direct radiation loudspeaker 105
Direct sequence spread spectrum technologies 329
Direct wave 379
Discriminator 117, 120
Dispersion 369
Dither technique 322
Divide-by-N 234
Doping 311
Dot-matrix displays 341
Double superheterodyne 107
Double-sideband suppressed carrier modulator
(DSB-SC) 124, 132
Down-conversion 334, 356, 361, 381
Dual-gate FET mixer 147
Dual-mode cellular telephone 358
Duplexer 343
Earth return 7
Earth-reflected wave 379
Echo 308, 370, 387
Eddy currents 395
Edison, Thomas 7
Edge-triggered logic 407
Edison effect 9
Eight-stage shift register 234
Electret microphone 225
Electric power grid 12
Electromagnetic spectrum 399
Electron beam control 191
Electron collector 165
Electron gun 164, 165, 202
Electron lens 202
Electron multiplier 165
Electron tube amplifiers 10
Electronic mail (e-mail) 15
Electronic scanning 10
Electronic serial number (ESN) 350
Electronic sidetone control 224
Electronic switching 261
Electronic telephone 224
Electrostatics 202
Embedding 24
Emitter follower 61, 319
Enable 407
Encryption 347, 357
Envelope detector 81, 103, 122, 145, 148, 300
Equalization 381
ERMES (European Radio Message System) 341
Error correction 319
Error signal 155
Error-detection codes 12
ETACS (European Total Access Communication
System) 357
European Postal and Telecommunications
Administrations (CEPT) 362
Exclusive OR (XOR) 360
Expander 351
Extremely high frequency (EHF) 397
Extremely low frequency (ELF) 397
Facsimile machine 4
Fading 347
Farnsworth, Philo 10
Federal Communications Commission (FCC) 15
Feedback theory 24
Fiber optics 11
Film bulk acoustic resonator (FBAR) 346
Finger wheel 223
Fleming, Sir John Ambrose 9
FLEX (Flexible wide-area paging protocol) 341
Flip-flop 241, 405
Flow diagram 262
Focus coil 165
Folded dipole 146
Forward or downlink 348
Foster-Seeley discriminator 148
Four-phase ring counter 273
Four-quadrant analog multiplier 51
Four-wire repeater 374
Fourier series 97, 113
Fractional detuning 152
Frame bit detector 292
Frame bit extractor 286
Free electrons 311
Free space propagation 17, 71, 297, 367, 377
Frequency changer 82, 83, 90, 147
Frequency discriminator 145, 148
Frequency division duplex (FDD) 343
Frequency division multiple access (FDMA) 328,
Frequency division multiplex (FDM) 20, 267, 327
Frequency equalizers 375
Frequency hopped multiple access (FHMA) 329
Frequency mixer 83
Frequency modulation (FM) 18, 111, 333, 354
Frequency multiplier 22, 37
Frequency-shift keying (FSK) 297,299, 300
Frequency-to-amplitude convertor 120, 145
Full-duplex 326
Gamma rays 367
Ganged capacitor 82
Gauss, Carl Friedrich 2
Geostationary satellites 384
Ground plane 73
Ground returns 307
Group 3 (G3) 307
Group special mobile (GSM) 362
GSM (Global system for mobile communications)
Gyrator 226
Gyroscopic effect 386
Half-wavelength dipole 72
Half-duplex 326
Halske, Johann 3
Handoff 347, 350
Handshake 308, 310
Harmonic telegraph 6
Hartley oscillator 30
Henry, Joseph 8
Hertz, Heinrich 8, 377
High-bit-rate digital subscriber line (HSDL) 9, 14
High-definition television (HDTV) 209
Histogram 318
Holes 312
Hook switch 221, 254
Horizontal Deflection System 200
Horizontal trace 164
Horn loudspeaker 105
Horn reflector 381
Hue 181
Hunt 257
Hybrid 222, 226
Hybrid-to-hybrid feedback 374
Hyper-abrupt junctions 116
Hysteresis loop 295
In-phase (I) signal 183, 334
Iconoscope 10, 163
Ideal transformer 392
Image dissector 10
Image Orthicon 165
Inductive loading 10, 267, 368
Infra red 312, 367
Input impedance 26, 58
Institute of Radio Engineers 307
Integrated Services Digital Network (ISDN) 14,
296, 297
Integration time 315
Integrator 126, 133
Intel 4004, 8080 13
Intel Pentium III 13
International telecommunication satellites
Inter-digit pause 224
Interlacing 167
Intermediate frequency (IF) 82, 98, 143, 147, 189
International Telecommunication Union (ITU) 307
Internet 11
Internet browsers 14
Internet service providers (ISP) 15
Inversion layer 312
Inverting input 246
Ionosphere 107, 379
Isolated dipole 72
J-K flip-flop 241
Jenkins, Charles 10
Jumbogroup 271, 381
Korn, Arthur 6
Ladder network 246
Large-scale-integration (LSI) 250
Laser 376
‘‘Last mile’’ 14, 298
Leakage inductance 394
Leaky surface acoustic wave (LSAW) 346
Least significant bit (LSB) 245
Light emitting diode (LED) 376, 415
Line-of-sight transmission 327, 380
Liquid crystal display (LCD) 340
Load line 55
Local oscillator 82, 143, 147, 268, 290, 381
Long wave 75
Loop-gain 28
Loop-gain phase shift 28
Loudspeaker 105, 187
Low frequency (LF) 397
Low-noise amplifier 354, 387
Luminance, 181
Magnetic flux density 106, 167
Magnetic tape 405
Magneto 7, 220
Magneto-motive force (MMF) 253
Magnetostatics 202
Majority carriers 312
Manual central office 253, 257
Marconi, Guglielmo 8
Marine telegraphic cable 3
Mark 297
Maxwell, James 8, 377
May, Joseph 9
Mechanical filters 100
Mechanical speed governor 6, 223
Medium wave 75
Medium wave frequency (MW) 397
Memory 261, 308, 318, 340
Message Switched Data Service (MSDS) 296
Metal-oxide semiconductor field-effect transistor
Metal-semiconductor field-effect transistor
(MESFET) 354
Microprocessor 308
Miller oscillator 37
Miller-effect capacitance 177
Minimum frequency reuse distance 349
Minority carriers 312
Mixer 82, 83, 90, 187, 381
Mixer gain 96
Mobile 331, 347
Mobile identification number (MIN) 350
Mobile radio 325
Mobile Telephone Switching Office (MTSO) 348
Modem 297, 298, 308
Modified Huffman (MH) Code 308, 318, 401
Modified modified READ (MMR) 308
Modified READ (MR) 308
Modified READ Code 318
Modified twin-tee filter 232
Modulation index 113
Modulation 18, 47, 381
Molniya 386
Monochromatic 181
Monopole 72
Morse, Samuel 2
Mosaic 164
Most significant bit (MSB) 245
Moving-iron telephone receiver 218
MPEG-2 209
Multipath fading 331
Multiple-hop signal 380
Multiplex system 181
Multiplier sampling gate 279
Mutual coupling coefficient 150, 391
Mutual inductance 391
N-coder 234
N-channel JFET 284
N-type metal-oxide semiconductor (NMOS) 354
N-type negative conductance 25
N-type silicon 312
NAND gate 234, 237, 407
Narrow band frequency modulation (NBFM) 124
Negative conductance oscillator 24, 115
Negative feedback 27
Negative resistance 24
Negative-impedance converter 369, 370
Newsgroups 15
Nipkow, Paul 10
Nipkow disc 10
Non-coherent demodulation 327
Non-volatile memory 411
Non-destructive reading of memory 414
NOR gate 234, 237
NOT-gate 234, 236
Notch frequency 232
NSFnet 15
Nyquist sampling rate 271
Operating point 55
Operational amplifier 130
Operational amplifier sampling gate 279
Optical fiber 367, 376
Oscillation 28
Output impedance 26
P-type silicon 312
Packet switching 12, 297
Paging system 325, 338
Pantelegraph 6
Parabolic dish antenna 381
Parallel operation 414
Parallel transmission 161
Parallel-to-serial (P=S) converter 245, 250
Passive LC filters 99
Pawl 254
Pel 305
Pendulum 4, 305
Pentodes 25, 115
Perigee 384
Perry, John 9
Phase angle 305
Phase equalization 369, 375
Phase modulation (PM) 18, 333
Phase shifter 291
Phase-locked loop 290
Phase-locked loop FM detector 155
Phasing 5
Phosphor 167
Phosphorus 311
Photo-transistors 312
Photocathode 165
Photoconductivity 167
Photogate 315
Photomultipliers 10
Photon 312
Photosensitive cells 10
Phototube 161
Picture tube 202
Pierce oscillator 37
Piezoelectric substrate 346
Pilot signal 138, 268
Pixel 305
Polarized light 340
Poles 174
Popov, Aleksandr 8
Portable 347
Positive feedback 27
Post Office Code Standardization Advisory Group
(POCSAG) 338
Pre-amplifier/equalizer 292
Pre-picture tube matrixing. 203
Printing telegraph 3
Propagation delay 376
Pseudo-noise (PN) generator 358, 359
Pseudodigital modulation 273
Pseudo-random sequence 329
Public Switched Telephone Network (PSTN) 343
Pulse code modulation encoder=multiplexer 286
Pulse Regenerator 294
Pulse shaping filter 334
Pulse-amplitude modulation (PAM) 273, 279
Pulse-code modulation (PCM) 244, 273, 286
Pulse-position modulation (PPM) 273
Pulse-width modulation (PWM) 273
‘‘Push-pull’’ 54
Quadrature detector 154
Quadrature multiplex 182
Quadrature phase shift keying (QPSK) 334
Quadrature-phase (Q-) signal 183, 334
Quartz crystal 21, 22
Quiescent 55, 59
R-2R ladder 247
Radio frequency choke (RFC) 148
Radio frequency duplexer 343
Ramp counter A=D converter 248
Random access memory (RAM) 411
Ratio detector 153
RC twin-tee circuit 231
Read memory 411
Read only memory (ROM) 411
Real power 32
Recording telegraph 4
Rectifier 230, 259
Refraction 379
Refractive index 377
Regenerative repeater 288, 291, 369, 376, 381, 384
Register 261
Relay 2
Release time 102
Reset 241
Reset or clear state 406
Resetable binary counter 234, 248
Resetable bistable multivibrator 234
Resettable flip-flop 318
Resistor transistor logic (RTL) 234
Resonance 22
Reverse or uplink 348
Reverse-emphasis filter 356
Ribbon microphone 225
Right-hand rule 167
Ring counter 359
Rotary dial 223
S-plane 174
S-R flip-flop 241, 405
S-type negative resistance 25
Sample-and-hold (S=H) 280, 283
Sampling 273
series gate 277
series-shunt gate 278
shunt gate 278
Sampling theorem 271
Satellite communication 11, 271, 367, 384
Saturation 55, 181
Scanner 162, 167, 305
Schmitt trigger 280, 300, 308
Search engines 14
Siemens, Werner 3
Selenium cell 6
Semafore 1
Sensors 312
Sequential access memory 411
Serial transmission 162
Serial-to-parallel converter 251, 245
Set or preset state 241, 406
Seven-segment display 341, 415
Shadow mask 207
Shift register 318, 410
Short dipole 72
Short wave frequency (SW) 76, 397
Short-circuit admittance 31
Sideband 20, 98
Sidetone 222, 229
Siemens Telecommunication Company 3
Signal plate 164
Silicon-controlled rectifier 25
Simplex 325
Single sideband (SSB) 179, 268
Single-hop distance 380
Single-sideband-suppressed carrier (SSB-SC) 326
Sinusoidal oscillator 28
Sky-reflected wave 379
Sliding correlator 361
Space-division multiplex (SDM) 261
Speech compressor 351
Speed-up capacitor 320
Spiral scanning 6
Spread spectrum techniques 329, 347
Square-law modulator 47
Stagger tuning 146
Station class mark (SCM) 350
Step-by-step switching 253
Stepper motor 308, 315
Stereophonic FM signal 137, 156
Stray capacitance 394
Strowger, Almon B. 8
Strowger switch 256
Submarine cable 376, 387
Sun spot cycle 380
Super High Frequency (SHF) 397
Superheterodyne 81, 82, 143, 187
Supervisory audio tone generator (SAT) 354
Surface acoustic wave (SAW) 100, 346
Susceptance 33
Switched capacitor filters 101
Switching-type mixer 96
Symmetrical clipper 119
Sync-pulse separator 193
Synchronism 5, 273, 305,329
Synchronous demodulation 158
Synchronous detection 205
Synchronous tuning 146
T-equivalent model 373
T1 PCM carrier 14
Tank circuit 40
Target 165
TAT-6 (SG) trans-Atlantic cable 376
TAT-8 (SG) trans-Atlantic cable 377
Telecom Canada Digital Network 290
Teletypewriter exchange (TWX) 296
Television front end or tuner 189
Terrestrial microwave network 11, 271, 380
Thermal printer 308, 321
Thevenin’s Theorem 395
Thin-film inter-digital electrodes 346
Three-phase clock 315
Threshold voltage 312
Time division multiple access (TDMA) 328
Time division multiplex (TDM) 261, 271
Timing information 273
Toll line 267
Tone dial 224, 230
Tone Ringer 224, 229
Touch-tone 338
Traffic channels 348
Transadmittance 228
Trans-Atlantic cable 3
Transconductance 227
Transfer gate 315
Transistor-transistor logic (TTL) 234
Transponder 386
Transverse electric (TE) 376
Transverse magnetic (TM) 376
Triple-mode cellular telephone 362
Troposphere-reflected wave 379
Tropospheric scatter propagation 379
Trunks 267
Tuned circuit 40
Tuning fork 305
Tunnel diode 25, 115
Turns ratio 393
Turnstile array 180
Tweeter 106
Twisted nematic 340
Twisted pair cable 270, 368
Two-port 31
Ultra high frequency (UHF) 397
Ultra violet 367
Uni-junction transistors 25
Universal asynchronous receiver/transmitter
(UART) 250
Up-converter 354, 381
Upper sideband 326
USDC (United States Digital Cellular) 362
Vail, Alfred 3
Varactor diode 116
Variable gain amplifier 101
Variable transconductance 51
Vertical deflection oscillator 195
Vertical sync-separator 193
Vertical trace 165
Very high frequency (VHF) 397
Very low frequency (VLF) 397
Vestigial sideband 163, 179, 190
Video 161
amplifier 163, 170, 191
image 161
Vidicon 167
Vinyl discs 405
Virtual ground 246
Visual telegraphic system 1
Voice coil 106
Voice frequency (VF) 397
Volatile memories 411
Voltage-controlled capacitor 116, 122
Voltage-controlled crystal oscillator (VCXO) 290
Voltage-controlled current source 227
Voltage-controlled oscillator (VCO) 155
Waveform clipping 60
Waveguide 376
Wax cylinders 405
Weber, Wilhelm 2
Western Union Telegraph Company 3
Wheatstone, Charles 2
Winding resistance 69, 87, 394
Wireless telegraphy 8
Wireline telephone 338
Woofer 106
Word 412
World Wide Web 13
Write into memory 411
xDSL 298
Xerographic printer 308
Yagi-Uda antenna 146
Yoke coil 199
Zener diode 248, 293
Zener diodes 120
Zworykin, Vladimir 10, 163
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF