CNS Basic Train ing- Analog Co mmunication ECATC

Ethiopian Civil Aviation Training Center

CNS Engineering Personnel (CNSEP)
Basic Training

Analog Communications

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Ethiopia Civil Aviation Training Center(ECATC)

Analog Communication

Prepared By

AlemayehuYeshitila

AndualemLegesse

Ayalkbet Lemma

BushraNurhussein

KibreabKibebew

Thomas Asres

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Table of Contents

Module 1- Communication Concepts

1 Introduction
2 The dB in Communications
3 Noise
4 Noise Designation and Calculation
5 Noise Measurement
6 Information and Bandwidth
Module2 – Amplitude Modulation
1 Introduction
2 Amplitude Modulation Fundamentals
3 Percentage Modulation
4 AM Analysis
5 Circuits for AM Generation
6 AM Transmitter Systems
7 Transmitter Measurements
8 Receiver Characteristics
9AM Detection
10 Superheterodyne Receivers
11 Superheterodyne Tuning
12SuperheterodyneAnalysis
13Automatic Gain Control

Module 3- Single Side Band

1 Single-Sideband Characteristics
2 Sideband Generation: The Balanced
Modulator
3 SSB Filters
4 SSB Transmitters
5 SSB Demodulation
6 SSB Receivers

Frequency Modulation
1Angle Modulation
2 FM Communication
3FM Transmission
4FM Reception

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Module 1

Communication Concepts

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Communication Concepts

OBJECTIVES

After completing this Module, the student will be able to:

• Describe a basic communication system and explainthe concept of modulation

• Develop an understanding of the use of the decibel(dB) in communications systems
• Define electrical noise and explain its effect at thefirst stages of a receiver
• Calculate the thermal noise generated by a resistor
• Calculate the signal- to-noise ratio and noise figure foran amplifier

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

1. Communications concepts
1.1. Introduction

We will provide an introduction to all relevant aspects of communications systems.These

systems had their beginning with the discovery of various electrical, magnetic,and electrostatic
phenomena prior to the twentieth century. Starting withSamuel Morse's invention of the
telegraph in 1837, a truly remarkable rate ofprogress has occurred. The telephone, thanks to
Alexander Graham Bell, camealong in 1876. The first complete system of wireless
communication was providedby Guglielmo Marconi in 1894. Lee DeForest's invention of the
triode vacuum tubein 1908 allowed the first form of practical electronic amplification and really
opened the door to wireless communication. In 1948 another major discovery in thehistory of
electronics occurred with the development of the transistor by Shockley,Brattain, and Bardeen.
The more recent developments, such as integrated circuits,very large-scale integration, and
computers on a single silicon chip, are probablyfamiliar to you.The rapid transfer of these
developments into practical communicationssystems linking the entire globe (and now into outer
space) has stimulated abursting growth of complex social and economic activities. This growth
has subsequentlyhad a snowballing effect on the growth of the communication industrywith no
end in sight for the foreseeable future. Some people refer to this as theage of communications.
The function of a communication system is to transfer information from onepoint to another via
some communication link. The very first form of "information"electrically transferred was the
human voice in the form of a code (i.e., theMorse code), which was then converted back to
words at the receiving site. Peoplehad a natural desire and need to communicate rapidly between
distant points onthe earth, and that was the major concern of these developments. As that
goalbecame a reality, and with the evolution of new technology following the inventionof the
triode vacuum tube, new and less basic applications were also realized,such as entertainment
(radio and television), radar, and telemetry. The field of communicationsis still a highly d ynamic
one, with advancing technology constantlymak ing new equipment possible or allowing
improvement of the old systems.Communications was the basic origin of the electronics field,

1
CNS Basic Train ing- Analog Co mmunication ECATC

and no other majorbranch of electronics developed until the transistor made modern digital
computersa reality.

Modulation

Basic to the field of communications is the concept of modulation. Modulation isthe

process of putting information onto a high- frequency carrier for transmission. Inessence, then,
the transmission takes place at the high frequency (the carrier) whichhas been modified to
"carry" the lower- frequency information. The low- frequencyinformation is often called the
intelligence signal or, simply, the intelligence. It followsthat once this information is received,
the intelligence must be removed fromthe high- frequency carrier-a process known as
demodulation. At this point youmay be thinking, why bother to go through this
modulation/demodulation process?Why not just transmit the information directly? The problem
is that the frequency ofthe human voice ranges from about 20 to 3000 Hz. If everyone
transmitted those frequenciesdirectly as radio waves, interference would cause them all to be
ineffective.Another limitation of equal importance is the virtual impossibility of transmittingsuch
low frequencies since the required antennas for efficient propagation would bemiles in length.
The solution is modulation, which allows propagation of the low- frequencyintelligence with a
high- frequency carrier. The high- frequency carriers arechosen such that only one transmitter in
an area operates at the same frequencyto minimize interference, and that frequency is high
enough so that efficientantenna sizes are manageable. There are three basic methods of putting
lowfrequencyinformation onto a higher frequency. Equation (1) is the
mathematicalrepresentation of a sine wave, which we shall assume to be the high-
frequencycarrier.
𝑣 = 𝑉𝑝 𝑠𝑖𝑛 𝜔𝑡 + 𝜙 (1)

Where v=instantaneous value

Vp =peak value

𝜔=angular velocity=2π f

𝜙 =phase angle

1
CNS Basic Train ing- Analog Co mmunication ECATC

Any one of the last three terms could be varied in accordance with the low- frequencyinformation
signal to produce a modulated signal that contains the intelligence. Ifthe amplitude term, Vp , is
the parameter varied, it is called amplitude modulation(AM). If the frequency is varied, it is
frequency modulation (FM). Varying thephase angle, 𝜙,results in phase modulation (PM). In
subsequent chapters we shallstudy these systems in detail.

Communications systems

Communications systems are often categorized by the frequency of the carrier.Table l provides
the names for various frequency ranges in the radio spectrum.The extra- high- frequency range
begins at the starting point of infrared frequencies,but the infrareds extend considerably beyond
300 GHz (300 X 109 Hz).After the infrareds in the electromagnetic spectrum (of which the radio
waves area very small portion) come light waves, ultraviolet rays, X rays, gamma rays,
andcosmic rays.

Figure 1 represents a simple communication system in block diagram form.

Notice that the modulated stage accepts two inputs, the carrier and the information(intelligence)
signal. It produces the modulated signal, which is subsequently amplifiedbefore transmission.
Transmission of the modulated signal can take place by any one offour means: antennas,
waveguides, optical fibers, or transmission lines. These fourmodes of propagation will be studied
in subsequent chapters. The receiving unit of thesystem picks up the transmitted signal but must

1
CNS Basic Train ing- Analog Co mmunication ECATC

reamplify it to compensate for attenuationthat occurred during transmission. Oncesuitably

amplified, it is fed to thedemodulator (often referred to as the detector), where the information
signal is extractedfrom the high- frequency carrier. The demodulated signal (intelligence) is then
fed to theamplifier and raised to a level enabling it to chive a speaker or any other output
transducer.
Many of the performance measurements in communication systems are specifiedin dB
(decibels). The next section introduces the use of this very important conceptin communication
systems. This is followed by two basic limitations on the performanceof a communications
system: (1) electrical noise and (2) the bandwidthof frequencies allocated for the transmitted
signal. Sections 3 to 6 are devoted tothese topics because of their extreme importance.

Fig. 1: A communication system block diagram .

*A transducer is a device that converts energy from one form to another.

1
CNS Basic Train ing- Analog Co mmunication ECATC

1.2.The dB in communications

Decibels (dBs) are used to specify measured and calculated values in noise analysis,audio
systems, microwave system gain calculations, satellite system link-budgetanalysis, antenna
power gain, light-budget calculations, and many other communicationssystem measurements. In
each case, the dB value is calculated with respect to astandard or specified reference.The dB
value is calculated by taking the log of the ratio of the measured orcalculated power (P2 ) with
respect to a reference power (P1 ) level. This result isthen multiplied by 10 to obtain the value in
dB. The formula for calculating the dBvalue of two ratios is shown in Equation (2). Equation (2)
is commonly referred toas the power ratio form for dB.

dB = 10 log10 P2 /P1 (2)

By using the power relationship P = V2 /R, the relationship shown in Equation (3) isobtained:

dB = 10 log 10 (v2 2 /R2 )

Let R1 = R2 : v 1 2 /R1

dB = 10 log 10 v2 2 /v1 2 (3)

Note that we have assumed that the resistances (R1 and R2 ) are equivalent; therefore,these terms
can be ignored in the dB power equation. This is a reasonableassumption in most communication
systems since maximum power transfer (adesirable characteristic) is obtained when the input and
output impedances arematched. Equation (3) can be modified (using a prope1ty of logarithms) to
providea relationship for decibels in terms of the voltage ratios instead of power ratios. Thisis
called the voltage gain equation and is shown in Equation (4).

dB = 20 log 10 (V2 /V1 ) (4)

Applying thedb value

The dB unit is often used in specifying input- and output-signal- level requirementsfor
many communication systems. When making dB measurements, a referencelevel is specified or
implied for that particular application. An example is found inaudio consoles in broadcast
systems, where a 0-dBm input level is usually specifiedas the required input- and output-audio

1
CNS Basic Train ing- Analog Co mmunication ECATC

level for 100% modulation. Notice that alowercase m has been attached to the dB unit. This
indicates that the specified dBlevel is relative to a 1- mW reference.
Jn standard audio systems 0 dBmis defined as 0.001 W measured withrespect to a load
termination of 600 Ω. A 600-Ω balanced audio line is thestandard for professional audio,
broadcast, and telecommunications systems.However, 0 dBm is not exclusive to a 600-Ω
impedance.

Example 1
Show that when making a dBm measurement, a measured value of l mWwillresult in a 0 dBm
power level.

Solution
dB = 10 log 10 P2 /P1 = 10 log 10 1mW/1 mW = 0 dB (2)

This result, 0 dB, is expressed as 0 dBm to indicate that the result was obtained relativeto a 1-m
W reference.
It can be shown that the voltage measured across a 600-Ω load for a 0-dBmlevel is 0.775
V. This value can be obtained by first modifying Equation (2) byinserting the 1 mW value for P
1 , as shown.
dB = 10 Log 10 (P2 /P1 )
Where P2 = V2 2 / 600
P1 = 0.001 W

*dB (decibel)is a relative unit ofmeasurement usedfrequently in

electroniccommunications todescribe power gain or loss
*dBmis dB level using a 1-mWreference
*0 dBmis1 mW measured relativeto a 1-mW reference
*dBm(600)is decibel measurementusing a 1-mW reference with respect to a 600-Ω load

1
CNS Basic Train ing- Analog Co mmunication ECATC

Since 1 mW is the specified reference for dBm, the voltage reference for 0 dBmcanbe
developed as follows:

The voltage value 0.77459 (0.775 V) is the reference for 0 dB with respect to a 600-
Dload when a voltage measurement is used to calculate the dBm(600) value. ThedBm(600) term
indicates that this measurement or calculation is made using al- mW reference with respect to a
600-D load.
dBm(600) = 20 log 10 (V2 /0.775) (5)

Example 2 demonstrates how to solve for the voltage value (V2) if a +8-dBm levelis specified.

Example 2

A microwave system requires a +8-dBm audio level to provide 100% modulation.Determine the
voltage level required to produce a +8-dBm level. Assume a 600-Daudio system.

Solution

Since this is a 600- fl system, use the 0.775-V reference shown in Equation (5).

dBm(600) = 20 log 10 (V2 /0.775) (5)

+8 dBm = 20 log V2 /0.775)
0.4 = log v2 /0.775
Log -1 (0.4) = V2 /0.775
V2 = 1.947 V

1
CNS Basic Train ing- Analog Co mmunication ECATC

Thus, to verify that a +8-dBm level is being provided to the input of themicrowave
transmitter, approximately 1.95 V must be measured across the 600Ωinput.ThetermdBmalso
applies to communication systems that have a standardtermination impedance other than 600 Ω.
For example, many communicationsystems are terminated with 75 Ω. The 0-dBm value is still
defined as 1 mW, butit is measured with respect to a 75Ωtermination instead of 600 Ω.
Therefore, thevoltage reference for a 0-dBm system with respect to 75 Ωis obtained by solving
ForV in the expression P = V2 /R as shown
V= 𝑃𝑅 = 0.001 (75) = 0.274 V

To calculate the voltage gain or loss with respect to a 75- Ω load, useEquation (6). This value is
specified as dBm(75)to indicate that this measure wasmade or calculated using a 1- mW
reference relative to a 75- Ω load.
dBm(75) = 20 log 10 V/0.274 (6)

Fifty-ohm systems are usually used in radio conununications. The dBm voltage referencefor a
50-Ω system is
V= 𝑃𝑅 = 0.001 (50) = 0.2236 V

To calculate the voltage gain or loss expressed in dB for a 50-Ω system[dBm(50)], use Equation
(4) with V1 = 0.2236. This relationship is shown inEquation (7).

dBm(50) = 20 log 10 V/0.2236 (7)

It is common for power to be expressed in watts instead of milliwatts. In this casethe dB
unit is obtained with respect to 1 W and the dB values are expressed as dBW.

dBm(50) = 20 log 10 V/0.2236 (7)

*dBm(75)is a measurement madeusing a 1-mW referencewith respect to a 7 5- Ωload

*dBm(50)is a measurement madeusing a 1-mW referencewith respect to a 50- Ω.load
*dBWisa measurement madeusing a 1-W reference.
*dBμVisa measurement madeusing a 1-μ,V reference.
*0 dBW is defined as 1 W measured with respect to a 1-W reference.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Remember, dB is a relative measurement. As shown by Equations (1) and (3), bothpower

and voltage gains can be expressed in dB relative to a reference value. In thecase of dBW, the
reference is 1 W; therefore, Equation (1) is written with 1 Wreplacing the reference P 1. This
gives Equation (8).
dBW = 10 log 10 P2 /1W (8)

In some applications, it may be necessary to convert from one reference dBto another.

Example 3demonstrates how to convert from dBm to dBW.

Example 3

A laser diode outputs + 10 dBm. Convert this value to

(a) watts.
(b) dBW

Solution

(a) Convert +l 0 dBm to watts. Substitute and solve for P2:

+10dBm = 10 log P2 /0.001

Log -1 (1) = P2 /0.001 => 10 = P2 /0.001
P2 = 0.01 W

(b) Convert +10 dBm to dBW.

dBW = 10 log 0.01W / 1W = -20 dBW

It is common with communication receivers to express voltage measurements

in terms of dBμV, dB- microvolts. For voltage gain calculations involving dBμV,use Equation
(4) and specify 1 μ, V as the reference (V1 ) in the calculations, as shown in Equation (9).
dBμV = 20 log 10 V2 /1 μV

1
CNS Basic Train ing- Analog Co mmunication ECATC

Conversiontableforcommon dBmvalues

There are many applications using decibels in calculations involving relativevalues. The
important thing to remember is that a relative reference is typicallyspecified or understood when
calculating or measuring a decibel value. Table 2 isa conversion table for many common dBm
values. A conversion table is providedfor dBm, voltage, and watts for 600-Ω, 75- Ω, and 50- Ω
systems. Additionally, a listof common decibel terms is provided in Table 3.

1.3. Noise

Electrical noise may be defined as any undesired voltages or currents that ultimately end up
appearing in the receiver output. To the listener this electrical noise often manifests itself as
static. It may only be annoying, such as an occasional burst of static, or continuous and of such
amplitude that the desired information is obliterated.
Noise signals at their point of origin are generally very small, for example, at the
microvolt level. You may be wondering, therefore, why they create so much trouble. Well, a
communications receiver is a very sensitive instrument that is given a very small signal at its
input that must be greatly amplified before it can possibly drive a speaker. Consider the receiver
block diagram shown in Figure l to be representative of a standard FM radio (receiver). The first
amplifier block, which formsthe "front end" of the radio, is required to amplify a received signal
from the radio's antenna that is often less than 10 μ.,V. It does not take a very large dose of
undesired signal (noise) to ruin reception. This is true even though the transmitter output may be
many thousands of watts because, when received, it is severely attenuated.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Therefore, if the desired signal received is of the same order of magnitude as the undesired
noisesignal, it will probably be unintelligible. This situation is made even worse because the
receiver itself introduces additional noise.
The noise present in a received radio signal that has been introduced in the transmitting
medium is termed external noise. The noise introduced by the receiver is termed internal noise.
The important implications of noise considerations in the study of communications systems
cannot be overemphasized.

External noise

Human made noise

The most troublesome form of external noise is usually the human- made variety. It is
often produced by spark-producing mechanisms such as engine ignition systems, fluorescent
lights, and commutators in electric motors. This noise is actually "radiated" or transmitted from
its generating sources through the atmosphere in the same fashion that a transmitting antenna
radiates desirable electrical signals to a receiving antenna. This process is called wave
propagation. If the human- made noise exists in the vicinity of the transmitted radio signal and is
within its frequency range, these two signals will "add" together. This is obviously an
undesirable phenomenon. Human- made noise occurs randomly at frequencies up to around 500
MHz.
Another common source of human-made noise is contained in the power lines that supply
the energy for most electronic systems. Jn this context the ac ripple in the de power supply
output of a receiver can be classified as noise (an unwanted electrical signal) and must be
minimized in receivers that are accepting extremely small intelligence signals. Additionally, ac
power lines contain surges of voltagecaused by the switching on and off of highly inductive
loads such as electrical motors. It is certainly ill- advised to operate sensitive electrical equipment
in close proximity to an elevator! Human-made noise is weakest in sparsely populated areas,
which explains the location of extremely sensitive communications equipment, such as satellite
tracking stations, in desert-type locations.

Atmospheric noise

Atmospheric noise is caused by naturally occurring disturbancesin the earth's atmosphere,

with lightning discharges being the most prominentcontributors. The frequency content is spread
over the entire radio spectrum, but itsintensity is inversely related to frequency. It is therefore
most troublesome at thelower frequencies. It manifests itself in the static noise that you hear on
standard AMradio receivers. Its amplitude is greatest from a storm near the receiver, but
theadditive effect of distant disturbances is also a factor. This is often apparent whenlistening to
a distant station at night on an AM receiver. It is not a significant factorfor frequencies exceeding
about 20 MHz.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Space noise

The other form of external noise arrives from outer space and is calledspace noise. It is
pretty evenly divided in origin between the sun and all the otherstars. That originating from our
star (the sun) is tern1ed solar noise. Solar noise iscyclical and reaches very annoying peaks about
every eleven years.All the other stars also generate this space noise, and their contribution
istermed cosmic noise. Since they are much farther away than the sun, their individualeffects are
small, but they make up for this by their countless numbers and theiradditive effects. Space noise
occurs at frequencies from about 8 MHz up to 1.5 GHz(1.5 X 109 Hz). While it contains energy
at less than 8 MHz, these components areabsorbed by the earth's ionosphere before they can
reach the atmosphere. Theionosphere is a region above the atmosphere where free ions and
electrons exist insufficient quantity to have an appreciable effect on wave travel.It includes the
area from about sixty to several hundred miles above the earth.

Inte rnal noise

As stated previously, internal noise is introduced by the receiver itself. Thus, thenoise
already present at the receiving antenna (external noise) has another componentadded to it before
it reaches the output. The receiver's major noise contributionoccurs in its very first stage of
amplification, where the desired signal is at its lowestlevel, and noise injected at that point will
be at its largest value in proportion to theintelligence signal. A glance at Figure 2 should help
clarify this point. Even thoughall following stages also introduce noise, their effect is usually
negligible withrespect to the very first stage because of their much higher signal level. Note
thatthenoise injected between amplifiers 1 and 2 has not appreciably increased the noise on

Fig. 2: Noise effect on a receiver's first and second amplifier stages.

the desired signal, even though it is of the same magnitude as the noise injected into amplifier 1.
For this reason, the first receiver stage must be carefully designed to have low noise

1
CNS Basic Train ing- Analog Co mmunication ECATC

characteristics, with the following stages being decreasingly important as the desired signal gets
larger and larger.

Thermal Noise

There are two basic types of noise generated by electronic circuits.The first one to
consider is due to the1mal interaction between the free electrons andvibrating ions in a
conductor. It causes the rate of arrival of electrons at either end ofa resistor to vary randomly,
and thereby varies the resistor's potential difference.Resistors and the resistance within all
electronic devices are constantly producing anoise voltage. This form of noise was first
thoroughly studied by J. B. Johnson in1928 and is often termed Johnson noise. Since it is
dependent on temperature, it isalso referred to as thermal noise. Its frequency content is spread
equally throughoutthe usable spectrum, which leads to a third designator: white noise (from
optics,where white light contains a11 frequencies or colors). The terms Johnson, thermal,and
white noise may be used interchangeably. Johnson was able to show that thepower of this
generated noise is given by
Pn = kT/∆f (10)
- 23
Where k = Boltzmann's constant (l.38 X 10 J/K)
T = resistor temperature in kelvin (K)
∆f= frequency bandwidth of the system being considered
Since this noise power is directly proportional to the bandwidth involved, it isadvisable to
limit a receiver to the smallest bandwidth possible. You may be wonderinghow the bandwidth
figures into this. The noise is an ac voltage that has randominstantaneous amplitude but a
predictable rms value. The frequency of thisnoise voltage is just as random as the voltage peaks.

Transistor Noise

In Example 5, the noise introduced by the transistor, other than its thermal noise, was not
considered. The major contributor of transistor noise is called shot noise. It is due to the discrete-
particle nature of the current carriers in all forms of semiconductors. These current carriers, even
under de conditions, arenot moving in an exactly steady continuous flow since the distance they
travel varies due to random paths of motion.
The name shot noise is derived from the fact that when amplified into a speaker, it sounds
like a shower of lead shot falling on a metallic surface . Shot noise and thermal noise are additive.

* Johnson Noise:another name for thermalnoise, first studied byJ. 8. Johnson

*Thermal Noise:internal noise caused bythermal interaction betweenfree electrons and
vibratingions in a conductor
*White Noise:another name for thermalnoise because its frequencycontent is uniform
acrossthe spectrum.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Unfortunately, there is no val id formula to ca1culate its value for a complete transistor
where the sources of shot noise are the currents within the emitter-base and collector-base
diodes. Hence, the device user must refer to the manufacturer's data sheet for an indication of
shot noise characteristics. The methods of dealing with these data are covered in Section 5. Shot
noise generally increases proportionally with de bias currents except in MOSFETs, where shot
noise seems to be relatively independent of de current levels.

Frequency noise effects

Two little- understood forms of device noise occur at the opposite extremes of frequency.
The low-frequency effect is called excess noise and occurs at freq uencies below about 1 kHz. It
is inversely proportional to frequency and directly proportional to temperature and dc current
levels. It is thought to be caused by crystal surface defects in semiconductors that vary at an
inverse rate with frequency. Excess noise is often refered to as flicker noise, pink noise, or 1/f
noise. lt is present in both bipolar junction transistors (BJTs) and field-effect transistors (FETs).
At high frequencies, device noise starts to increase rapidly in the vicinity of the device's
high- frequency cutoff. When the transit time of caniers crossing a junction is comparable to the
signal's period (i.e., high frequencies), some of thecarriers may diffuse back to the source or
emitter. This effect is termed transit-time noise. These high- and low- frequency effects are
relatively unimportant in the design of receivers since the critical stages (the front end) will
usually be working well above 1 kHz and hopefully below the device's high- frequency cutoff
area. The low-frequency effects are important, however, to the design of lowlevel, low- frequency
amplifiers encountered in certain instrument and biomedical applications.
The overall noise intensity versus frequency curves for semiconductor devices (and
tubes) have a bathtub shape, as represented in Figure 3. At low frequencies the excess noise is
dominant, while in the midrange shot noise and thermal noise predominate, and above that the
high- frequency effects take over. Of course, tubes are now seldom used and fortunately their
semiconductor replacements off er better noise characteristics. Since semiconductors possess
inherent resistances, they generatethermal noise in addition to shot noise, as indicated in Figure
3. The noise characteristics provided in manufacturers' data sheets take into account both the shot
and thermal effects. At the device's high- frequency cutoff, f hc ,thehighfrequency effects take
over, and the noise increases rapidly.

Fig 3: Device noise versus frequency.

1
CNS Basic Train ing- Analog Co mmunication ECATC

1.4. Noise designation and calculation

Signal to Noise Ratio

We have thus far dealt with different types of noise without showing how to dealwith noise in a
practical way. The most fundamental relationship used is known asthe signal-to-noise ratio (S/N
ratio), which is a relative measure of the desired signalpower to the noise power. The S/N ratio is
often designated simply as S/N and canbe expressed mathematically as

(12)
at any particular point in an amplifier. It is onen expressed in decibel form as

(13)
For example, the output of the amplifier in Example 5 was 1 m V rms and the noise
Was0.514 m Vrms, and thus (remembering that P = E2 /R)

Noise Figure

S/N successfully identifies the noise content at a specific point but is not useful inrelating
how much additional noise a particular transistor has injected into a signalgoing from input to
output. The tem1 noise figure (NF) is usually used to specifyexactly how noisy a device is. lt is
defined as fo11ows:

(14)

WhereSi/Ni is the signal-to-noise power ratio at the device's input and S0 /N0 is thesignal- to-noise
power ratio at its output. The term (Si /Ni)/ (S0 / N0 )is called the noiseratio (NR). If the device
under consideration were ideal (injected no additionalnoise), then Si/Niand S0 /N0 would be equal,
the NR would equal 1, and NF= 10log 1 = 10 X 0 = 0 dB. Of course, this result cannot be
obtained in practice

1
CNS Basic Train ing- Analog Co mmunication ECATC

Discussion

• Describe a basic communication system and explain the concept of modulation.

• What are the uses of the decibel (dB) in communication system.
• Define electrical noise and explain its effect at thefirst stages of a receiver.
• Explain the main factors responsible for the thermal noise generated by a resistor based
on the noise formula.
• Explain the main factors that affect signal-to- noise ratio and noise figure for an amplifier
based on the S/N formula.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Module 2
Amplitude Modulation

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Amplitude modulation

OBJECTIVES

After completing this Module, the student should be able to:

• Describe the process of modulation.

• Sketch an AM waveform with various modulation indexes
• Explain the difference between a sideband and side frequency.
• Analyze various power, voltage, and current calculations in AM systems
• Understand circuits used to generate AM.
• Determine high- and low- level modulation systems from schematics and block diagrams
• Perform AM transmitter measurements using meters, oscilloscopes, and spectrum
analyzers.
• Define the sensitivity and selectivity of a radio receiver
• Describe the operation of a diode detector in an AM receiver
• Sketch block diagrams for TRF and superheterodyne receivers
• Understand the generation of image frequencies and describe how to suppress them
• Recognize and analyze RF and IF amplifiers
• Describe the need for automatic gain control and show how it can be implemented
• Analyze the operation of a complete AM receiver system
• Perform a test analysis on the power levels (dBm) at each stage of an AM receiver system

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

2. Amplitude modulation (AM) transmission

2.1.Introduction

The reasons that modulation is used in electronic communications is explained as:

l .Direct transmission of intelligible signals would result in catastrophic interferenceproblems
because the resulting radio waves would be at approximatelythe same frequency.
2. Most intelligible signals occur at relatively low frequencies. Efficienttransmissionand
reception of radio waves at low frequencies is not practicaldueto the large antennas required.
The process of impressing a low-frequency inte11igence signal onto a
higherfrequency"carrier" signal may be defined as modulation. The higher- frequency"carrier"
signal will hereafter be referred to as simply the carrier. It is also termedthe radio-frequency (RF)
signal because it is at a high-enough frequency to be transmittedthrough free space as a radio
wave. The low-frequency intelligence signalwill subsequently be termed the "inte11igence." It
may also be identified by termssuch as modulating signal, information signal, audio signal, or
modulating wave.
Three different characteristics of a carrier can be modified to allow it to"carry" intelligence.
Either the amplitude, frequency, or phase of a carrier arealtered by the intelligence signal.
Varying the carrier's amplitude to accomplish thisgoal is the sub ject of this chapter.

2.2. Amplitude Modulation Fundame ntals

Combining two widely different sine-wave frequencies such as a carrier and intelligencein a
linear fashion results in their simple algebraic addition, as shown in Figure 4. A circuit that
would perform this function is shown in Figure 4(a)-

Fig. 4: Linear addition of two sine waves.

1
CNS Basic Train ing- Analog Co mmunication ECATC

the two signals combined in a linear device such as a resistor. Unfortunately, the result[Figure 4
(d)] is not suitable for transmission as an AM waveform. If it were transmitted, the receiving
antenna would be detecting just the carrier signal because the lowfrequencyintelligence
component cannot be propagated efficiently as a radio wave.

The method utilized to produce a usable AM signal is to combine the carrierand intelligence
through a nonlinear device. It can be mathematically proven thatthe combination of any two
sine waves through a nonlinear device produces thefollowing frequency components:
1. Adc level
2. Components at each of the two original frequencies
3. Components at the sum and difference frequencies of the two originalfrequencies
4. Harmonics of the two original frequencies

Figure 5 shows this process pictoria1ly with the two sine waves, labeled f cand f ito
represent the carrier and intelligence. If all but the f c-f i,fc, and fc+f icomponentsare removed
(perhaps with a bandpass filter), the three components left forman AM waveform. They are
referred to as:
1. The lower-side frequency (fc-fi)
2. The carrier frequency (fc)
3. The upper-side frequency (fc+ fi)
Mathematical analysis of this process is provided in Section 4.

Fig. 5: Nonlinear mixing.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Am waveforms

Figure 3 shows the actual AM waveform under varying conditions of the intelligencesignal. Note
in Figure 3(a) that the resultant AM waveform is basically asignal at the carrier frequency whose
amplitude is changing at the same rate as theintelligence frequency. As the intelligence
amplitude reaches a maximum positivevalue, the AM waveform has a maximum amplitude. The
AM waveform reaches aminimum value when the intelligence amplitude is at a maximum
negative value.In Figure 3(b), the intelligence frequency remains the same, but its amplitude
hasbeen increased. The resulting AM waveform reacts by reaching a larger maximumvalue and
smaller minimum value. In Figure 3(c), the intelligence amplitude isreduced and its frequency
has gone up. The resulting AM waveform, therefore, hasreduced maximums and minimums, and
the rate at which it swings between theseextremes has increased to the same frequency as the
intelligence signal.

Fig 6:AM waveform under varying intelligence signal (e i) conditions.

1
CNS Basic Train ing- Analog Co mmunication ECATC

It may now be correctly concluded that both the top and bottom envelopes ofan AM
waveform are replicas of the frequency and amplitude of the intelligence(notice the 180° phase
shift). However, the AM waveform does not include anycomponent at the intelligence frequency.
The equation for the AM waveform(envelope) is provided in Equation (1).

e = (Ec + 𝐸𝑖 sin 𝜔𝑖𝑡) sin ωct

WhereEc= the peak amplitude of the carrier signal

Ei= the peak amplitude of the intelligence signal
wit= the radian frequency of the intelligence signal
wct = the radian frequency of the carrier signal
w = 2𝜋f

This equation indicates that an AM waveform will contain the carrier frequencyplus the
products of the sine waves defining the carrier and intelligence signals.Based on the
trigonometric identity,
(sin x)(siny) = 0.5 cos(x - y) - 0.5 cos(x + y)

Wherex is the carrier frequency and y is the intelligence frequency.

The product ofthe carrier and intelligence sine waves will produce the sum and differences of
theAmplitude Modulation: Transmissiontwo frequencies. If a 1-MHz carrier were modulated by
a 5-kHz intelligence signal,the AM waveform would include the following components:
1 MHz + 5 kHz = 1,005,000 Hz (upper-side frequency)
1 MHz = 1,000,000 Hz (carrier frequency)
1 MHz - 5 kHz = 995,000 Hz (lower-side frequency)
This process is shown in Figure 4. Thus, even though the AM waveform hasenvelopes
that are replicas of the intelligence signal, it does not contain a frequencycomponent at the
intelligence frequency.
The intelligence envelope is shown in the resultant waveform and results fromconnecting
a line from each RF peak value to the next one for both the top and bottomhalves of the AM
waveform. The drawn- in envelope is not really a component of thewaveform and would not be
seen on an oscilloscope display. In addition, the top andbottom envelopes are not the upper- and

1
CNS Basic Train ing- Analog Co mmunication ECATC

lower-side frequencies, respectively. Theenvelopes result from the nonlinear combination of a

carrier with two lower-amplitudesignals spaced in frequency equal amounts above and below the
carrier frequency. Theincrease and decrease in the AM waveform's amplitude is caused by the
frequencydifference in the side frequencies, which allows them alternately to add to and
subtractfrom the carrier amplitude, depending on their instantaneous phase relationships.
The AM waveform in Figure 7(d) does not show the relative frequenciesto scale. The
ratio of f cto the envelope frequency (which is also f i) is 1 MHz to 5 kHz,or 200: 1.

Fig. 7:Carrier and side- frequency components resultin AM waveform.

*Upper Sidebandis a band of frequenciesproduced in a modulatorfrom the creation of

sumfrequencies between thecarrier and informationsignals.
*Lower Sidebandis a band of frequenciesproduced in a modulatorfrom the creation
ofdifference frequenciesbetween the carrier andinformation signals.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 8:Modulation by a band of intelligence frequencies.

Thus, the fluctuating RF should show 200 cycles for every cycle of envelopevariation. To
do that in a sketch is not possible, and an oscilloscope display ofthis example, and most practical
AM waveforms, results in a well-defined envelopebut with so many RF variations that they
appear as a blur, as shown in Figure 4(e).
Modulation of a carrier with a pure sine-wave intelligence signal has thus farbeen shown.
However, in most systems the inte11igence is a rather complex waveformthat contains many
frequency components. For example, the human voice containscomponents from ro ughly 200 Hz
to 3 kHz and has a very erratic shape. If it wereused to modulate the carrier, a whole band of side
frequencies would be generated.The band of frequencies thus generated above the carrier is
termed the upper sideband,while those below the carrier are called the lower sideband. This
situation is illustrated in Figure 5 for a 1-MHz carrier modulated by a whole band of
frequencies,which range from 200 Hz up to 3 kHz. The upper sideband is from 1,000,200
to1,003,000 Hz, and the lower sideband ranges from 997,000 to 999,800 Hz.

Example 1

A 1.4-MHz carrier is modulated by a music signal that has frequency componentsfrom 20 Hz to

10 kHz. Determine the range of frequencies generated for theupper and lower sidebands.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Solution

The upper sideband is equal to the sum of carrier and intelligence frequencies. Therefore,the
upper sideband (usb) will include the frequencies from
1,400,000 Hz+ 20 Hz= 1,400,020 Hz
to
1 ,400,000 Hz + I 0,000 Hz = 1,410,000 Hz

The lower sideband (lsb) will include the frequencies from

1,400,000 Hz - 10,000 Hz = 1,390,000 Hz

to
1,400,000 Hz - 20 Hz = 1,399,980 Hz

This result is shown in Figure 6 with a frequency spectrum of the AM modulator's output.

Fig. 9: Solution for Example 1.

*Percentage Modulationis a measure of the extent towhich a carrier voltage isvaried by

the intelligencefor AM systems.
*Modulation Indexis another name forpercentage modulation;represented as a
decimalquantity between 0 and 1for AM transmitters.
*Modulation Factoris another name formodulation index.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 10:AM representation using vector addition of phasers.

2.3. Percentage Modulation

In Section 2 it was determined that an increase in intelligence amplitude resulted inan AM

signal with larger maximums and smaller minimums. It is helpful to have amathematical
relationship between the relative amplitude of the carrier and intelligencesignals. The percentage
modulation provides this, and it is a measure of theextent to which a carrier voltage is varied by
the intelligence. The percentage modulationis also ref erred to as modulation index or
modulation factor, and they aresymbol ized by m.
Figure 11 illustrates the two most common methods for determining the
percentagemodulation when modulating with sine waves. Notice that when the intelligencesignal
is zero, the carrier is unmodulated and has a peak amplitude labeledas Ee. When the intelligence
reaches its first peak value (point w), the AM signalreaches a peak value labeled Ei(the increase
from Ee). Percentage modulation isthen given as
%m = Ei/ Ec x 100%(3)
or expressed simply by a ratio:
m = Ei / Ec(4)

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 11: Percentage modulation determination.

The same result can be obtained by utilizing the maximum peak-to-peak value ofthe AM
waveform (point w), which is shown as B,and the minimum peak-to-peakvalue (point x), which
is A in the following equation:

𝐵−𝐴
% m = 𝐵+𝐴 X 100 % (5)

This method is usually more convenient in graphical (oscilloscope) solutions.

Overmodulation

If the AM waveform's minimum value Afalls to zero as a result of an increase in

theinte11igence amplitude, the percentage modulation becomes

𝐵−𝐴 𝐵−0
%m = X 100% = X 100% = 100%
𝐵+𝐴 𝐵+0

1
CNS Basic Train ing- Analog Co mmunication ECATC

This is the maximum possible degree of modulation. In this situation the carrier isbeing
varied between zero and double its unmodulated value. Any further increase inthe intelligence
an1plitude will cause a condition known as overmodulation to occur.If this does occur, the
modulated carrier wi11 go to more than double its unmodulatedvalue but will fall to zero for an
interval of time, as shown in Figure 12. This "gap"produces distortion termed sideband splatter,
which results in the transmission offrequencies outside a station's normal allocated range. This is
an unacceptablecondition because it causes severe interference to other stations and causes a
loudsplattering sound to be heard at the receiver.

Fig. 12 :Overmodulation.

*Over modulation is a state when an excessive intelligence signal overdrives an AM

modulator producing percentage modulation exceeding 100 percent.
*Sideband Splatter is a distortion resulting in an over modulated AM transmission
creating excessive bandwidths.

1
CNS Basic Train ing- Analog Co mmunication ECATC

2.4.AM analysis

The instantaneous value of the AM waveform can be developed as follows. Theequation for
the amplitude of an AM waveform can be written as the carrier peakamplitude, Ee, plus the
intelligence signal, ei. Thus, the amplitude Eis
E = Ec + ei
butei= Eisinwit, so that
E = Ec + Ei𝑠𝑖𝑛 𝑤it
From Equation (2), Ei= mEc, so that

The instantaneous value of the AM wave is the amplitude term E just developed
times sin wcf. Thus,

Notice that the AM wave (e) is the result of the product of two sine waves. Asdefined by
Equation 2, this product can be expanded with the help of the trigonometricrelation

sinx sin y = Hcos (x - y) - cos (x + y) ]. Therefore,

The preceding equation proves that the AM wave contains the three terms previouslylisted: the
carrier(1), the upper sideband at f c+ f i(3)and the lower sidebandat f c- f i(2) It also proves that the

1
CNS Basic Train ing- Analog Co mmunication ECATC

instantaneous amplitude of the side frequenciesismEc/2. It shows conclusively that the bandwidth
required for AM transmissionis twice the highest intelligence frequency.
In the case where a carrier is modulated by a pure sine wave, it can be shownthat at 100
percent modulation, the upper- and lower-side frequencies are one- halfthe amplitude of the
carrier. In general, as just developed,
ESF = mEc/2
(6)
WhereESF= side- frequency amplitude
m = modulation index
Ec= carrier amplitude

In an AM transmission, the carrier amplitude and frequency always remain

constant,while the sidebands are usually changing in amplitude and frequency. The
carriercontains no information since it never changes. However, it does contain the mostpower
since its amplitude is always at least double (when m = 100%) the sideband'samplitude. lt is the
sidebands that contain the information.

Example 3

Determine the maximum sideband power if the carrier output is I kW andcalculate the total
maximum transmitted power.

Solution

Since
ESF = mEc/2(6)

it is obvious that the maximum sideband power occurs when m = 1 or 100 percent. At
thatpercentage modulation, each side frequency is 1/2the carrier amplitude. Since power
isproportional to the square of voltage, each sideband has 1/4 of the carrier power or 1/4 X 1
kW,or 250 W. Therefore, the total sideband power is 250 W X 2 = 500 W and the total
transmitted power is 1 kW + 500 W, or 1.5 kW.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Importance of high percentage modulation

It is important to use as high a percentage modulation as possible while ensuring

thatovermodulation does not occur. The sidebands contain the information and have
maximumpower at 100 percent modulation. For example, if 50 percent modulation wereused in
Example 3, the sideband amplitudes are1/4 the carrier amplitude, and sincepower is proportional
to E2 , we have (1/4)2 , or 1/16 the carrier power. Thus, total sidebandpower is now 1/16x 1 kW x
2, or 125 W. The actual transmitted intelligence is thusonly ¼ of the 500 W sideband power
transmitted at full l00 percent modulation. Theseresults are summarized in Table 1. Even though
the total transmitted power has onlyfallen from 1.5 kW to 1.125 kW, the effective transmission
has only! the strength at50 percent modulation as compared to 100 percent. Because of these
considerations,most AM transmitters attempt to maintain between 90 and 95 percent modulation
as acompromise between efficiency and the chance of drifting into ovem1odulation.
A valuable relationship for many AM calculations is

Pt= Pc(1+m2 /2)(7)

WherePt= total transmitted power (sidebands and carrier)

Pc = carrier power
m = modulation index
Equation (7) can be manipulated to utilize cmrent instead of power. This is a usefulrelationship
since current is often the most easily measured parameter of a transmitter'soutput to the antenna.

(8)

whereIt= total transmitted current

Ic= carrier current
m = modulation index

Equation (8) can also be used with E substituted for I (Er= Ec l + m2/2).

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 4

A 500-W carrier is to be modulated to a 90 percent level. Determine the totaltransmitted power:

Solution
Pt = Pc (1+m2 /2)
Pt = 500 w(1+0.92 /2) = 702.5w(7)

Example 5

An AM broadcast station operates at its maximum allowed total output of 50 kWand at 95

percent modulation. How much of its transmitted power is intelligence(sidebands)?

Solution

Pt = Pc (1+m2 /2)
50 kw = Pc (1+ 0.952 /2)
Pc = 50 kw/ 1+(0.952 /2) = 34.5kw

Therefore, the total intelligence signal is

Pi= Pt- Pc= 50kW - 34.5 kW = 15.5 kW

1
CNS Basic Train ing- Analog Co mmunication ECATC

2.5. Circuits for AM generation

Amplitude modulation is generated by combining carrier and intelligence frequenciesthrough a

nonlinear device. Diodes have nonlinear areas, but they are not oftenused because, being passive
devices, they offer no gain. Transistors offer nonlinearoperation (if properly biased) and provide
amplification, thus making them ideal forthis application. Figure l3 (a) shows an input/output
relationship for a typical bipolarjunction transistor (BJT). Notice that at both low and high values
of current, nonlinearareas exist. Between these two extremes is the linear area thatshould be used
fornormal amplification. One of the nonlinear areas must be used to generate AM.
Figure 13 (b) shows a very simple transistor modulator. It operates with no basebias and
thus depends on the positive peaks of ecand eito bias it into the firstnonlinear area shown in
Figure l3 (a). Proper adjustment of the levels of ecand ei is necessary for good operation. Their
levels must be low to stay in the first nonlineararea, and the intelligence power must be one-half
the carrier power (or less)for 100 percent modulation (or less). In the collector a parallel resonant
circuit, tunedto the carrier frequency, is used to tune into the three desired frequencies-the
upperand lower sidebands and the carrier. The resonant circuit presents a high impedance tothe
carrier (and any other close frequencies such as the sidebands) and thus allows ahigh output to
those components, but its very low impedance to all other frequencieseffectively shorts them out
Recall that the mixing of two frequencies through a nonlinear device generates more than just the
desired AM components, as illustrated in Figure 2. The tuned circuit then "sorts" out the three
desired AM components and serves to provide good sinusoidal components by the flywheel
effect.

* Base Modulationis a modulation system inwhich the intelligence isinjected into the base
of atransistor.
*High-Level Modulationissuperimposition of the intelligence on the carrier at the
lastpoint before the antenna in an AM transmitter.
*Low-Level Modulation is superimposition of the intelligence on the carrierbefore
reaching the antenna then themodulated waveform is amplified in an AM transmitter.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 13: Simple transistor modulator.

In practice, amplitude modulation can be obtained in several ways. Fordescriptive

purposes, the point of intelligence injection is utilized. For example, in 13(b) the intelligence is
injected into the base, hence it is termed base modulation.Collector and emitter modulation are
also used. In previous years, whenvacuum tubes were widely used, the most common form was
plate modulation, butgrid, cathode, and (for pentodes) suppressor-grid and screen-grid
modulationschemes were also utilized.

High and low level modulation

Another common designator for modulators involves whether or not the intelligence isinjected at
the last possible place or not. For example, the plate- modulated circuit shown in Figure 14 has
the intelligence added at the last possible point before thetransmitting antenna and is termed a
high- level modulation scheme. If the intelligencewas injected at any previous point, such as at a
base, emitter, grid, or cathode, or evenat a previous stage, it would be termed low- level
modulation. The designer's choicebetween high- and low- level systems is made largely on the
basis of the requiredpower output. For high-power applications such as standard radio
broadcasting, whereoutput~ are measured in terms of kilowatts instead of watts, high- level
modulation isthe most economical approach. Vacuum tubes are sti11 the best choice for many
highfrequency,high-power transmitter outputs. Recall that class C bias (device conductionfor less

1
CNS Basic Train ing- Analog Co mmunication ECATC

than 180°) allows for the highest possible efficiency. It realistically provides 70to 80 percent
efficiency as compared to about 50 to 60 percent for the next best configuration,a class B (linear)
amplifier. However, class C amplification cannot be usedfor reproduction of the complete AM
signal, and hence large amounts of inte11igencepower must be injected at the final output to
provide a high-percentage modulation.

Fig. l4: Plate- modulated class C amplifier.

Fig.15 :(a) High- and (b) low- level modulation.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The modulation process is accomplished in a nonlinear device, but a11 circuitrythat

follows must be linear. This is required to provide reproduction of the AM signalwithout
distortion. The class C amplifier is not linear but can reproduce (and amplify)the single
frequency earner. However, it would distort the carrier and sidebands combinationof the AM
signal. This is due to their changing amplitude that would be distortedby the flywheel effect in
the class C tank circuit. Block diagrams for typicalhigh- and low- level modulator systems are
shown in Figure 15(a) and (b), respectively.Note that in the high- level modulation system Figure
15(a)] the majority of poweramplification takes place in the highly efficient class C amplifier.
The low-level modulationscheme has its power amplification take place in the much less
efficient linearfinal amplifier.
In summary, then, high- level modulation requires larger intelligence power toproduce
modulation but allows extremely efficient amplification of the higher-poweredearner. Low- level
schemes allow low-powered intelligence signals to be used, but allsubsequent output stages must
use less efficient linear (not class C) configurations.Low- level systems usually offer the most
economical approach for low-powertransmitters.

Neutralization

One of the last remaining applications where tubes offer advantages over solid-stateddevices is
in radio transmitters, where kilowatts of output power are required athigh frequencies. Thus, the
general configuration shown in Figure 14 is still beingutilized. Note the variable capacitor, CN,
connected from the plate tank circuit back

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 16: Collector modulator.

to the grid. It is termed the neutralizing capacitor. It provides a path for the returnof a signal that
is 180° out of phase with the signal returned from plate to grid viathe internal interelectrode
capacitance (CINT ) of the tube. CN is adjusted to cancelthe internally fed-back signal to reduce the
tendency of self-oscillation. The transformerin the plate is made to introduce a 180° phase shift
by appropriate wiring.
Self-osci11ation is a problem for all RF amplifiers (both linear and class C).Notice the
neutralization capacitor (CN ) shown in the transistor amplifier in Figure 16. The self-oscillation
can be at the tuned frequency or at a higher frequency.The higher- frequency self-oscillations are
called parasitic oscillations. Inany event these oscillations are undesirable. At the tuned
frequency they preventamplification from taking place. The parasitic oscillations introduce
distortion andreduce desired amplification.

*Neutralizing Capacitor: is a capacitor that cancels fed-back signals to suppress self-

oscillation
*Parasitic Oscillations :higher-frequency self oscillations in RF amplifiers

1
CNS Basic Train ing- Analog Co mmunication ECATC

Transistor high level modulator

Figure 16 shows a transistorized class C, high- level modulation scheme. Class Coperation
provides an abrupt nonlinearity when the device switches on and off,which allows for the
generation of the sum and difference frequencies. This is incontrast to the use of the gradual
nonlinearities offered by a transistor at high andlow levels of class A bias, as previously shown
in Figure 13 (a). Generally, theoperating point is established to allow half the maximum ac
output voltage to be suppliedat the collector when the intelligence signal is zero. The Vbb supply
provides areverse bias for Q 1 so that it conducts on only the positive peak of the input
carriersignal. This, by definition, is class C bias because Q1 conducts for less than 180° percycle.
The tank circuit in Q 1 's collector is tuned to resonate at fc, and thus the full carrier sine wave is
reconstructed there by the flywheel effect at the extremely highefficiency afforded by class C
operation.

Fig 17: Collector modulator waveforms.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The intelligence (modulating) signal for the collector modulator of Figure 16 is added directly in
series with the co11ector supply voltage. The net effect of theintelligence signal is to vary the
energy available to the tank circuit each time Q1conducts on the positive peaks of carrier input.
This causes the output to reach amaximum value when the intelligence is at its peak positive
value and a minimumvalue when the intelligence is at its peak negative value. Since the circuit is
biasedto provide one-half of the maximum possible carrier output when the intelligenceis zero,
theoretically an intelligence signal level exists where the carrier will swingbetween twice its
static value and zero. This is a fully modulated (100 percentmodulation) AM waveform. In
practice, however, the collector modulator cannotachieve 100 percent modulation because the
transistor 's knee in its characteristiccurve changes at the inte11igence frequency rate. This limits
the region over whichthe collector voltage can vary, and slight collector modulation of the
precedingstage is necessary to allow the high modulation indexes that are usually desirable.This
is sometimes not a necessary measure in the tube-type high- level modulators.
Figure 17(a) shows an intelligence signal for a collector modulator, and Figure 17(b) shows its
effect on the co11ector supply voltage. In Figure 17(c), theresulting collector current variations
that are in step with the available supplyvoltages are shown. Figure 17(d) shows the collector
voltage produced by the flywheeleffect of the tank circuit as a result of the varying current peaks
that areflowing through the tank.

1
CNS Basic Train ing- Analog Co mmunication ECATC

2.6. AM trans mitter systems

Section 5 dealt with specific circuits to generate AM. Those circuits are only oneelement of a
transmitting system. It is important to obtain a good understanding of acomplete transmitting
unit, and that is the goal of this section.
Figure 18 provides block diagrams of simple high- and low-level AMtransmitters. The
oscillator that generates the carrier signal will invariably becrystal-controlled to maintain the
high accuracy required by the FederalCommunications Commission (FCC). The FCC regulates
radio and telephone communications in the United States. JnCanada the Canadian Radio-
Television andTelecommunications Commission performs the same function.
The oscillator is followed by the buffer amplifier, which provides a highimpedanceload
for the oscillator to minimize drift. It also provides enough gain todrive the modulated amplifier
sufficiently. Thus, the buffer amplifier could be a singlestage, or however many stages are
necessary to drive the following stage, the modulated amplifier.

Fig 18: Simple AM transmitter block diagram.

The intelligence amplifier receives its signal from the input transducer (oftena microphone) and
contains whatever stages of intelligence amplification arerequired except for the last one. The
last stage of intelligence amplification is calledthe modulator, and its output is mixed in the

1
CNS Basic Train ing- Analog Co mmunication ECATC

following stage with the carrier togenerate the AM signal. The stage that generates this signal is
termed the modulatedamplifier. This is also the output stage for high- level systems, but low- level
systemshave whatever number (one or more) of linear power amplifier stages required.Recall
that these stages are now amplifying the AM signal and must, therefore, belinear (class A or B),
as opposed to the more efficient but nonlinear class C amplifierthat can be used as an output
stage in high- level schemes.

2.7.Transmitte r measure ments

Trapezoid patterns

Several techniques are available to make operational checks on a transmitter'sperformance. A

standard oscilloscope display of the transmitted AM signal willindicate any gross deficiencies.
This technique is all the better if a dual-trace scope isused to allow the intelligence signal to be
superimposed on the AM signal, asillustrated in Figure 16. An improvement to this method is
known as the trapezoidalpattern.
It is illustrated in Figure 23. The AM signal is connected to the vertical inputand the
intelligence signal is applied to the horizontal input with the scope's internalsweep disconnected.
The intelligence signal usually must be applied through anadjustable RC phase-shifting network,
as shown, to ensure that it is exactly in phase with the modulation envelope of the AM
wavefo1m. Figure 23(b) shows the resultingscope display with improper phase relationships, and
Figure 23(c) shows the properin-phase trapezoidal pattern for a typical AM signal.
It easily allows percentage modulation calculations by applying the B and A dimensions to the
fo1mula presented previously,

(5)

*Keying :is ensuring that an oscillatorstarts by turning the dc onand off

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 19: Citizen's band transmitter PC board layout and complete assemblypictorial.

In Figure 23(d), the effect of 0 percent modulation (just the carrier) is indicated.The trapezoidal
pattern is simply a vertical line because there is no intelligencesignal to provide horizontal
deflection.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 20: Trapezoidal pattern connection scheme and displays.

Figures 23(e) and (f) show two more trapezoidal displays indicative of somecommon problems.
In both cases the trapezoid's sides are not straight (linear). Theconcave curvature at (e) indicates
poor linearity in the modulation stage, which isoften caused by improper neutralization or by
stray coupling in a previous stage.The convex curvature at (t) is usually caused by improper bias
or low can-iersignalpower (on en termed low excitation).

Meter measure ment

It is possible to make some meaningful transmitter checks with a de ammeter in thecollector (or
plate) of the modulated stage. If the operation is correct, this currentshould not change as the
intelligence signal is varied between zero and the pointwhere full modulation is attained. This is
true because the increase in current duringthe crest of the modulated wave should be exactly
offset by the drop during thetrough. A distorted AM signal will usually cause a cha nge in de
current flow. In thecase of overmodulation, the current will increase further during the crest but
cannotdecrease below zero at the trough, and a net increase in de current will occur. It isalso
common for this current to decrease as modulation is applied. This malfunctionis termed
downward modulation and is usually the result of insufficient excitation.The current increase

1
CNS Basic Train ing- Analog Co mmunication ECATC

during the modulation envelope crest is minimized, but thedecrease during the trough is nearly
normal.

Spectrum analyzers

The use of spectrum analyzers has become widespread in all fields of electronics, butespecially
in the communications industry. A spectrum analyzer visually displays(on a CRT) the amplitude
of the components of a wave as a function offrequency.This can be contrasted with an
oscilloscope display, which shows the amplitude of thetotal wave (all components) versus time.
Thus, an oscilloscope shows us the time domain while the spectrum analyzer shows the
frequency domain. In Figure 24(a) the frequency domain fo r a 1-MHz carrier modulated by a 5-
kHz intelligence signal is shown. Proper operation is indicated since only the carrier and upper-
and lower-side frequencies are present. During malfunctions, and to a lesser extent even under
normal conditions, transmitters will often generate spurious frequencies as shown in Figure
24(b), where components other than just the three desired are present. These spurious undesired
components are usually called spurs, and their amplitude is tightly controlled by FCC regulation
to minimize interference on adjacent channels.The coupling stage between the transmitter and its
antenna is designed to attenuate the spurs, but the transmitter's output stage must also be
carefully designed to keep the spurs to a minimum level. The spectrum analyzer is obviously a
very handy tool for use in evaluating a transmitter's performance.
The spectrum analyzer is, in effect, an automatic frequency-selective voltmeter that provides
both frequency and voltage on its CRT display. It can be thought of as a radio receiver with
broad frequency-range coverage and sharp sweep tuning, narrow-bandwidth circuits

*Low Excitationis an improper bias or lowcarrier signal power in anAM modulator

*Downward Modulationis the decrease in de outputcurrent in an AMmodulator usually
causedby low excitation
*Spectrum Analyzeris an instrument used to measurethe harmonic content of asignal by
displaying a plotof amplitude versusfrequency.
*Spurious Frequenciesis extra frequency componentsthat appear in the spectral display
of a signal,signifying distortion.
*Spursisundesired frequencycomponents of a signal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 21: Spectrum analysis of AM waveforms.

. The more sophisticated units are calibrated to readsignals in dB or dBm. This provides better
resolution between low- level sidebandsignals and the carrier and of course allows direct reading
of power levels withoutresorting to calculation from voltage levels. Most recent spectrum
analyzers utilizemicroprocessor principles (software programming) for ease of operation.
Fig. 25: shows a typical spectrum analyzer. Available software packageslink the waveform
analyzer and a computer that enables simplification andautomation of complex operations and
measurements. A typical display on the computer'sCRT is shown in Figure 25. It shows an AM
carrier that has four spuriousoutputs. Notice the noise between the spurs and the carrier. This is
commonly referred to as the noise floor of the system under test.

Harmonic distortion measurements

Harmonic distortion measurements can be made easily by applying a spectrally puresignal source
to the device under test (DUT). The quality of the measurement is dependent on the harmonic
distortion of both the signal source and spectrum analyzer

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 22: Spectrum analyzer and typical display.

.The source provides a signal to the OUT and the spectrum analyzer is used tomonitor the output.
Figure 26 shows the results of a typical harmonic distortion measurement.The distortion can be
specified by expressing the fundamental with respect to the largest harmonic in dB. This is
termed the relative harmonic distortion.

*Noise Floor isthe baseline on a spectrumanalyzer display.

*Relative HarmonicDistortion: expression specifying thefundamental
frequencycomponent of a signalwith respect to its largestharmonic in dB.
*Total HarmonicDistortion is a measure of distortionthat takes all significantharmonics
into account.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 23: Relativeharmonic distortion.

If the fundamental in Figure 23 is 1 V and the harmonic at 3 kHz (the largestdistortion

component) is 0.05 V, the relative harmonic distortion is

20 log 1V/0.05V = 26 dB

A somewhat more descriptive distortion specification is total harmonic distortion(THD). THO

takes into account the power in all the significant harmonics:

where V1 is the nns voltage of the fundamental and V2, V3, . . . are the rms voltages ofthe
harmonics. An infinite number of harmonics are theoretically created, but in practicethe
amplitude falls off for the higher harmonics. Virtually no error is introduced ifthe calculation
does not include harmonics less than one-tenth of the largest harmonic.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 9

Determine the THD if the spectrum analyzer display in Figure 23 has V1 = 1 V,V2 = 0.03 V, V3 =
0.05 V, V4 = 0.02 V, and V5 = 0.04 V.

Solution

THD calculations are somewhat tedious when a large number of significantharmonics exist.
Some spectrum analyzers include an automatic THD function thatdoes all the work and prints
out the THD percentage.

Special RFsignal measurement precautions

The frequency-domain measurements of the spectrum analyzer provide a morethorough reading

of RF frequency signals than does the time-domain oscilloscope.The high cost and additional
setup time of the spectrum analyzer dictates the continueduse of more standard measurement
techniques- mainly voltmeter and oscilloscopeusage. Whatever the means of measurement,
certain effects must beunderstood when testing RF signals as compared to the audio frequencies
withwhich you are probably more familiar. These effects are the loading of high-Q parallel-
resonant circuits by a relatively low impedance instrument and the frequencyresponse shift that
can be caused by test lead and instrument input capacitance.The consequence of connecting a 50-
fi signal generator into an RF-tuned circuitthat has a Zpin the kilohm region would be a
drastically reduced Q and increased bandwidth.The same loading effect would result if a low-
impedance detector were used tomake RF impedance measurements. This loading is minimized
by using resistors,capacitors, or transformers in conjunction with the measurement
instrument.The consequence of test lead or instrument capacitance is to shift the
circuit'sfrequency response. If you were looking at a 10-MHz AM signal that had its
resonantcircuit shifted to 9.8 MHz by the measurement capacitance, an obvious problemhas
resulted. Besides some simple attenuation, the equal amplitude relationshipbetween the usb and
lsb would be destroyed and wave form distortion results. Thiseffect is minimized by using low-

1
CNS Basic Train ing- Analog Co mmunication ECATC

valued series-connected coupling capacitors, orcanceled with small series-connected inductors. If

more precise readings withinconsequential loading are necessary, specially designed resonant
matchingnetworks are required. They can be used as an add-on with the measuring instrumentor
built into the RF system at convenient test locations.When testing a transmitter, the use of a
dummy antenna is often a necessity.A dummy antenna is a resistive load used in place of the
regular antenna. It is usedto prevent undesired transmissions that may otherwise occur. The
dummy antenna(also called a dummy load) also prevents damage to the output circuits that
mayoccur under unloaded conditions.

1
CNS Basic Train ing- Analog Co mmunication ECATC

AM Reception

2.8.Receiver Characteristics

If you were to envision a block diagram for a radio receiver, you would probably gothrough the
following logical thought process:
1. The signal from the antenna is usually very small- therefore, amplification isnecessary. This
amplifier should have low-noise characteristics and should betuned to accept only the desired
can- ier and sideband frequencies to avoidinterference from other stations and to minimize the
received noise. Reca11 thatnoise is proportional to bandwidth.
2. After sufficient amplification, a circuit to detect the intelligence from the radiofrequency is
required.
3. Following the detection of the intelligence, further amplification is necessaryto give it
sufficient power to drive a loudspeaker.

This logical train of thought leads to the block diagram shown in Figure 24. Itconsists of an RF
amplifier, detector, and audio amplifier. The first radio receiversfor broadcast AM took this form
and are called tuned radio frequency or, moresimply, TRF receivers. These receivers generally
had three stages of RF amplification,with each stage preceded by a separate variable-tuned
circuit. You can imaginethe frustration experienced by the user when tuning to a new station.
The threetuned circuits were all adjusted by separate variable capac itor controls. To receivea
station required proper adjustment of a11 three, and a good deal of time and practicewas
necessary.

Fig. 24:Simple radio receiver block diagram.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Sensitivity and Selectivity

Two major characteristics of any receiver are its sensitivity and selectivity. Areceiver's
sensitivity may be defined as its ability to drive the output transducer(e.g., speaker) to an
acceptable level. A more technical definition is the minimuminput signal (usually expressed as a
voltage) required to produce a specified outputsignal or sometimes just to provide a discernible
output. The range of sensitivitiesfor communication receivers varies from the millivolt region for
low-cost AMreceivers down to the nanovolt region for ultrasophisticated units for more
exactingapplications. In essence, a receiver's sensitivity is determined by the amount ofgain
provided and, more important, its noise characteristics. In general, the inputsignal must be
somewhat greater than the noise at the receiver's input. This inputnoise is termed the noise floor
of the receiver. It is not difficult to insert more gainin a radio, but getting noise figures below a
certain level presents a more difficultchallenge.
Selectivity may be defined as the extent to which a receiver is capable of differentiating
between the desired signal and other frequencies (unwanted radio signalsand noise). A receiver
can also be overly selective. For instance, on commercialbroadcast AM, we have seen that the
transmitted signal can include intelligence signalsup to about a maximum of 15 kHz, which
subsequently generates upper andlower sidebands extending 15 kHz above and below the carrier
frequency. The totalsignal has a 30-kHz bandwidth. Optimum receiver selectivity is thus 30 kHz,
but ifa 5-kHz bandwidth were selected, the upper and lower sidebands would extend only2.5 kHz
above and below the carrier.

*Tuned Radio FrequencyReceiveris the most elementaryreceiver design, consistingof RF

amplifier stages, adetector, and audio amplifier stages
*Sensitivity is the minimum input RFsignal to a receiverrequired to produce aspecified
audio signal at output
*Noise Floor is the baseline on aspectrum analyzer display,representing input noise ofthe
system under test RFamplification Detector

1
CNS Basic Train ing- Analog Co mmunication ECATC

The radio's output would suffer from a lack ofthe full possible fidelity because the output would
include intelligence up to a maximumof 2.5 kHz. On the other hand, an excessive selectivity of
50 kHz results inthe reception of unwanted adjacent radio signals and the additional external
noisethat is directly proportional to the bandwidth selected. Unfortunately, TRF receiversdid
suffer from selectivity problems, which led to their replacement by the superheterodynereceiver.
As has been stated, broadcast AM can extend to about 30-kHz bandwidth. Asa practical matter,
however, many stations and receivers u ea more limited bandwidth.The lost fidelity is often not
detrimental because of the talk-show format of many AM rations. For instance, a 10-kHz
bandwidth receiver provides audio output up to5 kHz, which more than handles the human voice
range. Musical reproduction witha 5-kHz maximum frequency is not high fidelity but is certainly
adequate for casuallistening.

TRFselectivity

Consider a standard AM broadcast band receiver that spans the frequency rangefrom 550 to 1550
kHz. If the approximate center of I 000 kHz is considered, we canuse the following equations to
find that, for a desired 10-kHz BW, a Q of 100 isrequired.
Q = f r/BW (1)
=1000kHz/10kHz
= 100
Now, ince the Q of a tuned circuit remains fairly constant as its capacitance is varied,a change to
1550 kHz will change the BW to 15.5 kHz.

Q =f r/ BW(2)
Therefore,
BW =f r/Q
=1550 kHz / 100
= 15.5 kHz

1
CNS Basic Train ing- Analog Co mmunication ECATC

Selectivityis the extent to which areceiver can differentiatebetween the desired signaland other
signals

The receiver 's BW is now too large, and it will suffer from increased noise. On theother hand,
the opposite problem is encountered at the lower end of the frequencyrange. At 550 kHz, the BW
is 5.5 kHz.
BW = f r/Q
=550kHz/100
= 5.5 kHz

The fidelity of reception is now impaired. The maximum intelligence frequency possibleis 5.5
kHzJ2, or 2.75 kHz, instead of the full 5 kHz transmitted. This se lectivityproblem led to the
general use of the superheterodyne receiver (see Section 3) inplace of TRF designs.

1
CNS Basic Train ing- Analog Co mmunication ECATC

2.9. AM Detection

The process of detecting the intelligence out of the carrier and sidebands (the AMsignal) has thus
far been mentioned but not explained. In fact, the detection processcan be easily accomplished.
Recall our discussions about generating AM. We said iftwo different frequencies were passed
through a nonlinear device, sum and differencecomponents would be generated. The carrier and
sidebands of the AM signalare separated in frequency by an amount equal to the intelligence
frequency. If theAM signal is passed through a nonlinear device, difference frequencies between
thecarrier and sidebands will be generated and these frequencies are, in fact, the intelligence.It
follows that passing the AM signal through a nonlinear device will providedetection, just as
passing the carrier and intelligence through a nonlinear deviceenables AM generation.Detection
of amplitude- modulated signals requires a nonlinear electrical network.
An ideal nonlinear curve for this is one that affects the positive half-cycle ofthe modulated wave
differently than the negative half-cycles. This distorts anapplied voltage wave of zero average
value so that the average resultant currentvaries with the intelligence signal amplitude. The curve
shown in Figure 25(a) iscalled an ideal curve because it is linear on each side of the operating
point P anddoes not introduce harmonic frequencies.When the input to an ideal nonlinear device
is a carrier and its sidebands, theoutput contains the following frequencies:
1. The carrier frequency
2. The upper sideband
3. The lower sideband
} Original

4. A de component
5. A frequency equal to the earner minus the lower sideband and the upper sidebandminus the
carrier, which is the original signal frequencyThe detector reproduces the signal frequency by
producing a distortion of adesirable kind in its output. When the output of the detector is
impressed upon alow-pass filter, the radio frequencies are suppressed and only the low-
frequencyintelligence signal and de components are left. This is shown as the dashed average
current curve in Figure 25(a)
.In some practical detector circuits, the nearest approach to the ideal curve is thesquare- law curve
shown in Figure 25(b). The output of a device using this curve contains,in addition to all the

1
CNS Basic Train ing- Analog Co mmunication ECATC

frequencies that were listed, the harmonics of each of thesefrequencies. The harmonics of radio
frequencies can be filtered out, but the harmonics

Fig 25: Nonlinear device used as a detector.

of the sum and difference frequencies, even though they produce an undesirabledistortion, may
have to be tolerated because they can be in the audio- frequency range.

Diode detector

One of the simplest and most effective types of detectors, and one with nearly an idealnonlinear
resistance characteristic, is the diode detector circuit shown in Figure 26(a).Notice the 1-V curve
in Figure 26(b). This is the type of curve on which the diodedetector at (a) operates. The curved
part of its response is the region of low currentand indicates that for small signals the output of
the detector will follow the squarelaw. For input signals with large amplitudes, however, the
output is essentially linear.
Therefore, harmonic outputs are limited. The abrupt nonlinearity occurs for the negativehalf-
cycle as shown in Figure 26(b).The modulated carrier is introduced into the tuned circuit made
up of LC1 in Figure 26(a). The waveshape of the input to the diode is shown in Figure 26(c).
Sincethe diode conducts only during half-cycles, this circuit removes all the negativehalf-cycles
and gives the result shown in Figure 26(d). The average output is shownat (e). Although the

1
CNS Basic Train ing- Analog Co mmunication ECATC

average input voltage is zero, the average output voltage acrossR always varies above zero.The
low-pass filter, made up of capacitor C 2 and resistor R, removes the RF (carrierfrequency),
which, so far as the rest of the receiver is concerned, serves no usefulpurpose. Capacitor C2
charges rapidly to the peak voltage through the small resistanceof the conducting diode, but
discharges slowly through the high resistance of R. Thesizes of Rand C 2 normally form a rather
short time constant at the intelligence (audio)frequency and a very long time constant at the radio
frequencies. The resultant outputwith C 2 in the circuit is a varying voltage that follows the peak
variation of the modulated

Fig 26 : Diode Detector

carrier [see Figure 26(f)]. For this reason it is often termed an envelope detector circuit.The dc
component produced by the detector circuit is removed by capacitor C 3 , producingthe ac voltage
wave shape in Figure 26(g). In communications receivers, the dc component is often used for
providing automatic volume (gain) control.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The advantages of diode detectors are as follows:

1. They can handle relatively high power signals. There is no practical limit tothe amplitude of
the input signal.
2. Distortion levels are acceptable for most AM applications. Distortio ndecreases as the
amplitude increases.
3. They are highly efficient. When properly designed, 90 percent efficiency isobtainable.
4. They develop a readily usable de voltage for the automatic gain control circuits.
The disadvantages of the diode detectors are:
1. Power is absorbed from the tuned circuit by the diode circuit. This reduces theQand selectivity
of the tuned input circuit.
2. No amplification occurs in a diode detector circuit.

Detector diode types

In noncritical applications, the standard pnjunction diode can be used. It is usuallyadequate for
the LF, HF, and low VHF bands. Low-cost silicon switching diodes suchas the 1N914 and
1N4148 are frequently used. For higher frequencies, point contactdiodes are often used. These
diodes have the pnjunction on the surface of the substrateand make contact to the p-type material
via a small wire. The junction is very small,yielding very low capacitance that makes them
useful for microwave operation up to40 GHz. The commonly used point contact varieties are the
1 N21, 1 N23, and 1 N34.An important specification for detector diodes is voltage sensitivity.
This is ameasure of detector output per unit of RF input power. It is specified as V/mWor m
V/μW at some specified de bias current.

DiagonalClipping

Careful selection of component parts is necessary for obtaining optimum efficiency in diode
detector circuits. One very important fact to consider is the value of the time constant RC2,
particularly in the case of pulse modulation.
*Envelope Detectoris another name for diodedetector.
*Product Detector:oscillator, mixer, and lowpass filter stage used toobtain the
intelligencefrom an AM signal.
*Heterodyne Detectoris another name for synchronous detector orproduct detector.
1
CNS Basic Train ing- Analog Co mmunication ECATC

When a carrier modulatedby a square pulse Figure 27(b)J is applied to an ideal diode detector,
the wave shape shown in Figure 27(c) is produced. Notice that for clarity, the amplitude of
thewave at (c) is exaggerated in comparison to the high frequency carrier shown at (b).
If the time constant of RC 2 is too long compared to the period of the RFwave, several
cycles are required to charge C 2 , and the leading edge of the outputpulse is sloped as shown in
Figure 27(d). After the pulse passes by, the capacitor dischargesslowly and the trailing edge is
exponential rather than square, as desired.This phenomenon is often referred to as diagonal
clipping. The diagonal clippingeffect from a sine-wave intelligence signal is shown at (e). Notice
that the detectedsine wave at (e) is distorted. The excessive RC time constant did not allow
thecapacitor voltage to follow the full changes of the sine wave. On the other hand, ifthe time
constant is too short, both the leading and trailing edges can be easilyreproduced. However, the
capacitor may discharge considerably between carriercycles. This reduces the average amplitude
of the pulse, leaving a sizable componentof the carrier frequency in the output, as shown in
Figure 27(f).
For these reasons the selection of the time constant is a compromise. The loadresistor R
must be large because the total input voltage is divided across R and theinternal resistance of the
diode when it is conducting. A large value of load resistanceensures that the greater part of this
voltage is in the output, where desired. On theother hand, the load resistance must not be so high
that capacitor C 2 becomes smallenough to approximate the size of Ci [Figure 27(a)], the internal
junction capacitance ofthe diode. When this occurs, capacitor C 2 will try to discharge through Ci
during thenonconducting periods, which would reduce the amplitude of the detector output.

Synchronous detection

Diode detectors are used in the vast majority of AM detection schemes. Since highfidelity is
usually not an important aspect in AM, the distortion levels of several percentagepoints or more
from a diode detector can be tolerated easily.In applications demandinggreater performance, the
use of a synchronous detector offers the followingadvantages:
1. Low distortion-well under 1 percent
2. Greater ability to follow fast- modulation waveforms, as in pulse-modulationor high- fidelity
applications

1
CNS Basic Train ing- Analog Co mmunication ECATC

3. The ability to provide gain instead of attenuation, as in diode detectors

Synchronous detectors are also called product or heterodyne detectors. Thep1inciple of
operation involves mixing in a nonlinear fashion just as in AM generation.

Fig. 27:Diode detector component considerations.

Dashed lines indicateaverage voltage during pulse.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Imagine receiving a transmission at 900 kHz. If it contained a 1-kHz tone, the receptionconsists
of three components:
1. The carrier at 900 kHz
2. The usb at 901 kHz
3. The lsb at 899 kHz
If this AM wavefo1m were mixed with an internally generated 900-kHz sine wavethrough a
nonlinear device, a resulting difference frequency is 1 kHz-the desiredoutput intelligence. Of
course, a number of much higher sum frequencies are alsogenerated, but they are easily filtered
out by a low-pass filter. Detection has beenachieved in a completely different fashion from that
for the envelope detector(diode detection) discussed previously. A circuit commonly used for
product detectionis the balanced modulator. It is widely used in single-sideband (SSB) systems.

2.10. Superheterodyne receivers

The basic variable-selectivity problem in TRF systems led to the development andgeneral usage
of the superheterodyne receivers in the early 1930s. This basic receiverconfiguration is still
dominant after an these years, an indication of its utility. A blockdiagram for a
superheterodynereceiver is provided in Figure 28. The first stage is astandard RF amplifier that
may or may not be required, depending on factors to beAudio Antenna and power :[J

Fig 28: Superheterodyne receiver block diagram.

1
CNS Basic Train ing- Analog Co mmunication ECATC

discussed later. The next stage is the mixer, which accepts two inputs, the output ofthe RF
amplifier (or antenna input when an RF amplifier is omitted) and a steady sinewave from the
local oscillator (LO). The mixer is yet another nonlinear device utilizedin AM. Its function is to
mix the AM signal with a sine wave to generate a newset of sum and difference frequencies. Its
output, as will be shown, is an AM signal with a constant carrier frequency regardless of the
transmitter's frequency. The nextstage is the intermediate-frequency (IF) amplifier, which
provides the bulk of radiofrequencysignal amplification at a fixed frequency. This allows for a
constant BWover the entire band of the receiver and is the key to the superior selectivity of the
superheterodyne receiver. Additionally, since the IF frequency is usually lower thanthe RF,
voltage gain of the signal is more easily attained at the IF frequency. Followingthe IF amplifiers
is the detector, which extracts the intelligence from the radio signal.It is subseque ntly an1plified
by the audio amplifiers into the speaker. A de levelproportional to the received signal's strength
is extracted from the detector stage andfed back to the lF amplifiers and sometimes to the mixer
and/or the RF amplifier. Thisis the automatic gain control (AGC) level, which allows relatively
constant receiveroutput for widely variable received signals. Detail on AGC is provided in
Section 6.

Frequency Conve rsion

It has been stated that the mixer pe1forms a frequency conversion process. Considerthe situation
shown in Figure 29. The AM signal into the mixer is a 1000-kHz carrierthat has been modulated
by a 1-kHz sine wave, thus producing side frequencies at999 kHz and 1001 kHz. The LO input is
a 1455-kHz sine wave. The mixer, being anonlinear device, will generate the following
components:
l. Frequencies at all of the original inputs: 999 kHz, 1000 kHz, 1001 kHz, and1455 kHz.
2. Sum and difference components of all the original inputs: 1455 kHz ± (999 kHz,1000 kHz,
and 1001 kHz). This means outputs at 2454 kHz, 2455 kHz,2456 kHz, 454 kHz, 455 kHz, and
456 kHz.
3. Harmonics of all the frequency components listed in 1 and 2 and a de component.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 29: Frequency conversion process.

Fig 30: Frequency conversion.

The IF amplifier has a tuned circuit that accepts components only near 455 kHz,in this case 454
kHz, 455 kHz, and 456 kHz. Since the mixer maintains thesame amplitude proportion that
existed with the original AM signal input at999 kHz, 1000 kHz, and 1001 kHz, the signal now
passing through the IFamplifiers is a replica of the original AM signal. The only difference is
that nowits carrier frequency is 455 kHz. Its envelope is identical to that of the originalAM
signal. A frequency conversion has occurred that has translated the carrierfrom 1000 kHz to 455
kHz-a frequency intermediate to the original carrier andintelligence frequencies-which led to the

1
CNS Basic Train ing- Analog Co mmunication ECATC

terminology intermediate- frequencyamplifier, or IF amplifier. Since the mixer and detector both
have nonlinearcharacteristics, the mixer is often referred to as the first detector.

Tunedcircuit adjustment

Now consider the effect of changing the tuned circuit at the front end of the mixer toaccept a
station at 1600 kHz. This means a reduction in either its inductance or capacitance(usually the
latter) to change its center frequency from 1000 kHz to 1600 kHz.If the capacitance in the local
oscillator's tuned circuit were simultaneously reduced sothat its frequency of oscillation went up
by 600 kHz, the situation shown in Figure 8would now exist. The mixer's output still contains a
component at 455 kHz (amongothers), as in the previous case when we were tuned to a 1000-
kHz station. Ofcourse, the other frequency components at the output of the mixer are not
acceptedby the selective circuits in the IF amplifiers.Thus, the key to superheterodyne operation
is to make the LO frequencytrack with the circuit or circuits that are tuning the incoming radio
signal so thattheir difference is a constant frequency (the IF). For a 455-kHz IF frequency,
themost common case for broadcast AM receivers, this means the LO should alwaysbe at a
frequency 455 kHz above the incoming carrier frequency. The receiver'sfront-end tuned circuits
are usually made to track together by mechanically linking(ganging) the capacitors in these
circuits on a common variable rotor assembly,as shown in Figure 9. Note that this ganged
capacitor has three separatecapacitor elements.

*First Detector is the mixer stage in asuperheterodyne receiverthat mixes the RF

signalwith a local oscillatorsignal to form theintermediate frequencysignal.
*Trimmeris a small variable capacitance in parallel with eachsection of a gangedcapacitor
*PadderCapacitoris a small variable capacitor inseries with each gangedtuning capacitor
in asuperheterodyne receiverto provide near-perfecttracking at the low end ofthe tuning
rangeTrimmers

1
 CNS Basic Train ing- Analog Co mmunication ECATC

Fig 31: Variable ganged capacitor.

2.11. Superheterodyne tuning

Tracking

It is not possible to make a receiver track perfectly over an entire wide range of frequencies.The
perfect situation occurs when the RF amplifier and mixer tuned circuitsare exactly together and
the LO is above these two by an amount exactly equal to the IFfrequency. To obtain a practical
degree of tracking, the following steps are employed:
1. A small variable capacitance in parallel with each section of the ganged capacitor,
called the trimmer, is adjusted for proper operation at the highest frequency.The trimmer
capacitors are shown in Figure 9. The highest frequency requires the main capacitor to be at its
minimum value (i.e., the plates all theway open). The trimmers are then adjusted to balance out
the remaining straycapacitances to provide perfect tracking at the highest frequency.
2. At the lowest frequency, when the ganged capacitors are fully meshed (maximumvalue), a
small variable capacitor known as the padder capacitor is putin series with the tank inductor. The
padders are adjusted to provide tracking atthe low frequency in the band.
3. The final adjustment is made at midfrequency by slight adjustment of theinductance in each
tank.
The curve in Figure 10(a) shows that performing the steps above, and thenrechecking them once
again to allow for interaction effects, provides nearly pe1fecttracking at three points. The minor

1
CNS Basic Train ing- Analog Co mmunication ECATC

imperfections between these points are generallyof an acceptable nature. Figure lO(b) shows the
circuit for each tank circuit andsummarizes the adjustment procedure.

Electronic tuning

The bulk and cost of ganged capacitors have led to their gradual replacement by a
techniqueloosely called electronic tuning. The majority of new designs use electronic
frequencysynthesis. Another electronic method relies on the capacitance offered by areverse-
biased diode.

Fig 32: Tracking considerations.

Since this capacitance varies with the amount of reverse bias, apotentiometer can be used to
provide the variable capacitance required for tuning.Diodes that have been specifically fabricated
to enhance this variable capacitance versusreverse bias characteristic are referred to as varactor
diodes, varicap diodes, orVVC diodes. Figure 11 shows the two generally used symbols for these
diodes and atypical capacitance versus reverse bias characteristic.
The amount of capacitance exhibited by a reverse-biased silicon diode, Cd,can be approximated
as

where C0 = diode capacitance at zero bias

VR= diode reverse bias voltage

1
CNS Basic Train ing- Analog Co mmunication ECATC

Trimmer capacitor.

Fig 33:Varactor diode symbols and C/V characteristic.

*VaractorDiodesis a diodehaving small internalcapacitance that varies asa function of the

irreversebias voltage.
*VaricapDiodes:another name for varactordiodes.
*vvcDiodes :another name for varactororvaricap diodes.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The use of Equation (1) allows the designer to determine accurately the amount ofreverse bias
needed to provide the necessary tuning range. The varactor diode canalso be used to generate
FM.
Figure 12 shows the front end of a broadcast-band receiver. It does not incorporatean RF
amplifier, and Q1 performs the dual function of mixer and localoscillator. The varactor diode D 1
provides the variable capacitance necessary to tunethe radio signal from the antenna while D 2
allows for the variable LO .

RF inductors.

Fig. 34: Broadcast-band AM receiver front end with electronic tuning.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The - 1- to - 12-V supply comes from the tuning potentiometer and provides thenecessary
variable reverse voltage for both varactor diodes. The matched diode characteristicsrequired for
good tracking often lead to the use of varactor diodes fabricatedon a common silicon chip and
provided in a single package.

2.12. Superheterodyne Analysis

Image frequency

The superheterodyne receiver has been shown to have that one great advantage overthe TRF --
constant selectivity over a wide range of received frequencies. This wasshown to be true since
the bulk of the amplification in a superheterodynereceiveroccurs in the IF amplifiers at a fixed
frequency, and this allows for relatively simpleand yet highly effective frequency select ive
circuits. A disadvantage does exist, however,other than the obvious increase in complexity. The
frequency conversion processperformed by the mixer-oscillator combination sometimes will
allow a station otherthan the desired one to be fed into the IF. Co nsider a receiver tuned to
receive a20-MHz station that uses a 1-MHz IF. The LO would, in this case, be at 21 MHz to
generate a 1-MHz frequency component at the mixer output. This situation is illustratedin Figure
13. If an undesired station at 22 MHz were also on the air, it is possiblefor it also to get into the
mixer. Even though the tuned circuit at the mixer's front endis"selecting" a center frequency of
20 MHz, a look at its response curve in Figure 13shows that it will not fully attenuate the
undesired station at 22 MHz. As soon as the22-MHz signal is fed into the mixer, we have a
problem. It mixes with the 21-MHz LO

1
CNS Basic Train ing- Analog Co mmunication ECATC

Figure 35: Image frequency illustration.

signal and one of the components produced is 22 MHz - 21 MHz = 1 MHz-the IFfrequency!
Thus, we now have a desired 20-MHz station and an undesired 22-MHzstation, which both look
correct to the IF. Depending on the strength of the undesiredstation, it can either interfere with or
even completely override the desired station.

Example 2

Determine the image frequency for a standard broadcast hand receiver using a455-kHz IF and
tuned to a station at 620 kHz.

Solution

The first step is to detem1ine the frequency of the LO. The LO frequency minus the
desired station's frequency of 620 kHz should equal the IF of 455 kHz. Hence,
LO - 620 kHz = 455 kHz
LO = 620 kHz + 455 kHz
= 1075 kHz

*Image Frequencyis an undesired input frequency in a superheterodyne receiver that

produces the same intermediate frequency as the desired input signal

1
CNS Basic Train ing- Analog Co mmunication ECATC

Now determine what other frequency, when mixed with 1075 kHz, yields an output component
at 455 kHz.
X - 1075 kHz = 455 kHz
X = 1075 kHz + 455 kHz
= 1530 kHz

Thus, 1530 kHz is the image frequency in this situation.

In the preceding discussion, the undesired received signal is called the imagefrequency.
Designing superheterodyne receivers with a high degree of image frequencyrejection is
obviously an important consideration.Image frequency rejection on the standard broadcast band
is not a major prob1em. A g1ance at Figure 14 serves to il1ustrate this point. This tuned circuit at
themixer's input comes fairly close to fully attenuating the image frequency, in thiscase, since
1530 kHz is so far away from the tuned circuit's center frequency of 620kHz. Unfortunately, this
situation is not so easy to attain at the higher frequencies

Fig 36: Image frequency not a problem.

used by many communication receivers. In these cases, a technique known as doubleconversion

is employed to solve the image frequency problems.The use of an RF amplifier with its own
input tuned circuit also helps to minimizethis problem. Now the image frequency must pass
through two tuned circuits(tuned to the desired frequency) before it is mixed. These tuned
circuits at the inputof the RF and mixer stages obviously serve to attenuate the image frequency
to agreater extent than can the single tuned circuit in receivers without an RF stage.

1
CNS Basic Train ing- Analog Co mmunication ECATC

RF Amplifiers

The use of RF amplifiers in superheterodyne receivers varies from none in

undemandingapplications up to three or even four stages in sophisticated
conununicationreceivers. Even inexpensive AM broadcast receivers not having an RF amplifier
docontain an RF section- the tuned circuit at the mixer's input. The major benefits ofusing RF
amplification are the following:
1. Improved image frequency rejection
2. More gain and thus better sensitivity
3. Improved noise characteristics
The first two are self-explanatory at this point, but the last advantage requires furtherelaboration.
Mixer stages require devices to be operated in a nonlinear area to generatethe required difference
frequency at their output. This process is inherently morenoisy (i .e., higher NF) than nor mal
class A linear bias. The use of RF amplificationstages to bring the signal up to appreciable levels
minimizes the effect of mixer noise.The RF amplifier usually employs an FET as its active
component. WhileBJTs certainly can be utilized, the following advantages of FETs have Jed to
theirgeneral usage in RF amplifiers.
1. Their high input impedance does not "load" down the Q of the tuned circuitpreceding the FET.
It thus serves to keep the selectivity at the highest possiblelevel.
2. The availability of dual- gate FETs provides an isolated injection point for theAGC signal.
3. Their input/output square-law relationship allows for lower distortion levels.

*Double Conversion: superheterodyne receiver design that has two separate mixers, local
oscillators, and intermediate frequencies to avoid image frequency problems.
*Cross-Modulation distortion: that results from undesired mixer outputs.
*Converters: another name for mixers.
*First Detectors: another name for mixers.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 37: Dual- gate MOSFET RF amplifier.

The distortion referred to in the last item is called cross- modulation.A typical MOSFET RF
amplifier stage is shown in Figure 15. It is a dualgateunit, with the AGC level applied to gate 2
to provide for automatically variablegain. The received antenna signal is fed via a tuned coupling
circuit to gate
1. The gate 1 and output drain connections are tapped down on their respectivecouplingnetworks,
which keeps the device from self-oscillation without the needfor a neutralizing capacitor. Notice
the built- in transient protection shown withinthe symbol for the 40673 MOSFET. Zener diodes
between the gates andsource/substrate connections provide protection from up to 10-Y p-p
transientvoltages. This is a valuable safeguard because of the extreme fra gility of theMOSFET
gate/channel junction.

Mixer/LO

The frequency conversion accomplished by the mixer/LO combination can beaccomplished in a

number of ways. The circuits shown in Figure 16 illustrate someof the possibilities. They all
make use of a device's nonlinearity to generate sum anddifference frequencies between the RF
and local oscillator signals. This process generatesoutput components at the IF frequency. Mixers
are also referred to asconverters and first detectors.The most basic mixer involves using a single

1
CNS Basic Train ing- Analog Co mmunication ECATC

diode to provide the required nonlinearityfor generating sum and difference frequencies. This is
shown in Figure 16(a).

Fig. 38: Typical mixer circuits.

* SchottkyDiode: specially fabricated majority carrier device formed from a metal

semiconductor interface, with extremely low junction capacitance.
* Self-Excited Mixer: single stage in a superheterodyne receiver that creates the LO signaland
mixes it with theapplied RF signal to formthe IF signal.
* Autodyne Mixer: another name for self excited mixer.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The LC circuit at the output is tuned to the desired IF and attenuates all the othergenerated
frequencies. Diode mixers are especially useful at the higher frequencyranges above 500 MHz.
Although they don't offer the gain of transistorized mixers,operation at the se frequencies is
economically possible by using Schottky diodes. TheSchottky diode is a specially fabricated
device fonned from a metal-semiconductorinterface. Because the diode is not made from a
pnjunction, it is a majority carrierdevice and offers an e xtremely low junction capacitance. This
makes it useful up tofrequencies of 100 GHz!The circuit shown in Figure 16(b) is a self-excited
mixer because a singledevice does the mixing and generates the LO frequency. Self-excited
mixers aresometimes referred to as autodyne mixers. The oscillator-tuned circuit of C4 andL4
provides a positive feedback signal to maintain oscillation via coil L3, which ismagnetically
coupled to L4. The oscillator signal is injected into Q1 's emitter viaC3 and the RF signal into its
base via L1- Li. transformer action. The "mixed"outputatQ1 's collector is fed to the C5- L5 tank
circuit, which tunes in the desiredfrequency for the IF amplifiers. Reca11 that mixing signals
through a nonlineardevice generates many frequency components; the tuned circuit is used to
select thedesired ones for the IF amplifiers. Further detail on this circuit is provided in
thetroubleshooting section of this chapter (Section 8).
A widely used IC mixer is shown in Figure 16(c). The Philips SA602A(NE602) IC
contains a transistorized mixer and an npntransistor that generates thelocal oscillator signal based
on the frequency-selective components connectedbetween pins 6 and 7. The LI> C1 combination
is tuned to the desired LO frequencybut a crystal could be substituted if better precision is
desired. The C2, C3 combinationis used to form a Colpitts oscillator with the internal oscillator
transistor. Theoscillator frequency can be set up to 200 MHz. Notice the output at pin 5 that is
fed to a ceramic bandpass filter. This type of filter, which is similar to crystals usedas filters, is a
commonly used alternative to LC filters. This versatile IC mixer canbe used at RF input
frequencies up to 500 MHz.

1
CNS Basic Train ing- Analog Co mmunication ECATC

IF amplifiers

The IF amplifiers provide the bulk of a receiver's gain (and thus are a major influenceon its
sensitivity) and selectivity characteristics. An IF amplifier is not awhole lot different from an RF
stage except it operates at a fixed frequency. Thisallows the use of fixed double-tuned
inductively coupled circuits, which in turna11ow for the sharply defined bandpass response
characteristic of superheterodynereceivers.The number of IF stages in any given receiver varies,
but from two to four istypical. Some typical IF amplifiers are shown in Figure 17.The circuit at
(a) usesthe 40673 dual- gate MOSFET while the other two use LICs specially made for
IFamplifier applications. Notice the double-tuned LC circuits at the input and outputof all three
circuits. They are shown within dashed lines to indicate they are onecomplete assembly. They
can be economically purchased for all common IFfrequencies and have a variable slug in the
transformer core for fine tuning theircenter frequency. All three of the circuits have provision for
the AGC level. Not allreceivers utilize AGC to control the gain of mixer and/or RF stages, but
theyinvariably do control the gain of IF stages.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 39: Typical IF amplifiers.

1
CNS Basic Train ing- Analog Co mmunication ECATC

2.13. Automatic gain control

The purpose of automatic gain control (AOC) has already been explained. Withoutthis function,
a receiver's usefu1ness is seriously impaired. The following list givessome of the problems that
would be encountered in a receiver without this provision:
I. Tuning the receiver would be a nightmare. To avoid missing the weak stations,you would have
the volume contro1 (in the non-AGC set) turned way up. Asyou tuned to a strong station, you
would probably blow out your speaker,whereas a weak station might not be audible.
2. The received signal from any given station is constantly changing as a result ofchanging
weather and ionospheric conditions. The AOC allows you to listentoa station without constantly
monitoring the volume control.
3. Many radio receivers are utilized under mobile conditions. For instance, astandard broadcast
AM car radio would be almost unusable without a goodAGC to compensate for the signal
variation in different locations.

Obtaining the AGC level

Most AGC systems obtain the AGC level just following the detector. Recall that following
the detector diode, an RC filter removes the high frequency but hopefullyleaves the low-
frequency envelope intact. By simply increasing that RC time constant,a slowly varying de level
is obtained. The de level changes with variations inthe strength of the overall received signal.
Figure 18(a) shows the output from a diode detector with no filtering. In thiscase, the output is
simply the AM waveform with the positive portion rectified outfor two different levels of
received signal into the diode. At (b), the addition of a filterhas provided the two different
envelope levels while filtering out the highfrequencycontent. These signals correspond to an
undesired change in volume oftwo different received stations. At (c), a much longer time
constant filter has actually

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 40 : Development of AGC voltage.

filtered the output into a de level. Notice that the de level changes, however, withthe two
different levels of input signal. This is a typical AGC level that is subsequentlyfed back to
control the gain of IF stages and/or the mixer and RF stages.In this case, the larger negative de
level at C2 would cause the receiver's gainto be decreased so that the ultimate speaker output is
roughly the same for eitherthe weak or strong station. It is important that the AGC time constant
be longenough so that desired radio signal level changes that constantly occur do not causea
change in receiver gain. The AGC should respond only to average signal strengthchanges, and as
such usually has a time constant of about a full second.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Controlling the gain of a transistor

Figure 19 illustrates a method whereby the variable de AGC level can be used tocontrol the gain
of a common emitter (CE) transistor amplifier stage. In the case of astrong received station, the
AGC voltage developed across the AGC filter capacitor(C Aod is a large negative value that
subsequently lowers the forward bias on Q1• Itcauses more de current to be drawn through R2,
and hence less is available for thebase of QI>since RI>which supplies current for both, can
supply only a relativelyconstant amount. The voltage gain of a CE stage with an emitter bypass
capacitor(CE) is nearly directly proportional to de bias current, and therefore the strong station
reduces the gain of Q1. The reception of very weak stations would reduce the gain ofQ1 very
slightly, if at all. The introduction of AGC in the 1920s marked the firstmajor use of an
electronic feedback control system. The AGC feedback path is calledthe AGC bus because in a
full receiver it is usually "bused" back into a number ofstages to obtain a large amount of gain
control. Some receivers require more elaborateAOC schemes.

Fig. 41: AGC circuit il lustration.

1
CNS Basic Train ing- Analog Co mmunication ECATC

IF/AGC Amplifier

The IF/ AGC amplifier shown in Figure 20 operates over an extremely wide input(Jl) range of 82
dB. It uses two low-cost transistors (2N3904 and 2N3906) as peakdetectors. Q2 functions as a
temperature-dependent current source and Q 1 asahalfwave detector. Q2 is biased for a collector
current of 300 μAat 27°C with a1 μA/°C temperature coefficient.The current into capacitor CA v
is the difference in the Q 1, Q2 collector currents,which is proportional to the output signal at 12.
The AGC voltage (V AGc) isthe time integral of this difference current. To ensure that V AGC is
not sensitive tothe short-term output signal changes, the rectified current in Ql must, on
average,balance the current in Q2. If the output of A2 is too small, V AGC will increase,thereby
increasing the gain of Al and A2. This will cause Ql to conduct furtheruntil the cmTent through
Ql balances the current through Q2.The gain of ICs Al and A2 is set at 41 dB maximum for a
total possible82-dB gain. They operate sequentia11y because the gain of A1 goes from
minimumto maximum first and then A2's does the same as dictated by the AGC level. Thefull
range of gain occurs from V AOC = 5 V (0 dB) to V AOC = 7 V (82 dB). Thisis approximately a
linear relationship so that V AGC = 6 V would cause a gain ofabout 41 dB [(6 - 5)/(7 - 5) x 82 =
41 ].The bandwidth exceeds 40 MHz and thereby a11ows operation at standard Ifssuch as 455
kHz, 10.7 MHz, or 21.4 MHz. At 10.7 MHz the AGC threshold is 100μ,Vrms (-67 dBm) and the
output is 1.4 V rrns (3.9 V p-p). This corresponds toa gain of 83 dB (20 log 1.4 V/100 μ,V). The
output holds steady at 1.4 V rmsforinputs from -67 dBm to as high as + 15 dBm, giving an 83-
dB AGC range. Inputsignals above 15 dBm overdrive the device. The undesired harmonic
outputs aretypically at least 34 dB down from the fundamental.

2.14. AM Receiver systems

We have thus far examined the various sections of AM receivers. It is now time to putit all
together and look at the complete system. Figure 21 shows the schematic of awidely used circuit
for a low-cost AM receiver. In the schematic shown in Figure 21,the push-pull audio power amp,
which requires two more transistors, has been omitted.The L1-Li, inductor combination is wound

1
CNS Basic Train ing- Analog Co mmunication ECATC

on a powdered- iron (ferrite) coreand functions as an antenna as well as an input coupling stage.
Ferrite-core loopstickantennas offer extremely good signal pickup, considering their sma11 size,
andare adequate for the strong signal strengths found in urban areas. The RF signal isthen fed
into Qi. which functions as the mixer and local oscillator (self-excited).The ganged tuning
capacitor, Ci. tunes to the desired incoming station (the B section)and adjusts the LO (the D
section) to its appropriate frequency. The output ofQ1 contains the IF components, which are
tuned and coupled to Q2 by the T1 package.The IF amplification of Q2 is coupled via the T2 IF
"can" to the second IFstage, Q3, whose output is subsequently coupled via T3 to the diode
detector £ 2.Of course, T1, T2, and T3 are all providing the very good
superheterodyneselectivitycharacteristics at the standard 455-kHz IF frequency. The £2 detector
diode'soutput is filtered by C11 so that just the inte11igence envelope is fed via the R12volume
control potentiometer into the Q4 audio amplifier. The AGC filter, C4, thenallows for a fed-back
control level into the base of Q2.

Fig. 42: Wide-range IF/AGC amplifier.

*Auxiliary AGC Diode : reduces receiver gain for very large signals.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 43 :AM broadcast superheterodyne receiver.

This receiver also illustrates the use of an auxiliary AGC diode (£1 ).Under normal signal
conditions, E1 is reverse biased and has no effect on theoperation. At some predetermined high
signal level, the regular AGC actioncauses the de 1evel at £ 1 's cathode to decrease to the point
where £ 1 starts toconduct (forward bias), and it loads down the T1 tank circuit, thus reducing
thesignal coupled into Q2 . The auxiliary AGC diode thus furnishes additional gaincontrol for
strong signals and enhances the range of signals that can be compensatedfor by the receiver.

LI C AM Receiver

The complete function of a superheterodyne AM receiver can be accomplished withLICs. The

only hitch is that the tuned circuits must be added on externally. SeveralAM chips are available
from the various IC manufacturers. The Philips semiconductorTDA1572T is a typical unit and is
shown in Figure 22. Notice that the deviceuses electronic tuning with variable capacitance
diodes.Even though the use of the LIC greatly reduces component count, the physicalsize and
cost are not appreciably affected because they are mainly determinedby the frequency selective

1
CNS Basic Train ing- Analog Co mmunication ECATC

circuits. Thus, LIC AM radios are not widely used forlow-cost applications but do find their way
into higher-quality AM receivers, wherecertain performance and feature advantages can be
realized.The limiting factor of tuned circuits is the only roadblock to having completereceivers
on a chip except for the station se1ection and volume controls.Alternatives to LC-tuned circuits,
such as ceramic filters, may be integrable in thefuture. Another possibility is the use of phase-
locked-loop (PLL) technology inproviding a nonsuperheterodyne type of receiver. Using this
approach, it is theoreticallypossible to fabricate a functional AM broadcast-band receiver using
justthe chip and two externa1 potentiometers (for volume control and station selection)and the
antenna.

AM Stereo

It is known that the reproduction of music with two separate channels can enrich andadd to its
realism. Broadcast AM has stru.ted to move into stereo broadcasts since severalschemes were
advanced in the late 1970s. Unfortunately, the FCC decided to letthe marketplace decide on the
best system. This led to confusion and no clear favorite.At this juncture we find that the
Motorola system has become the de facto standard. Itisno wonder that AM stereo has not beco me
a favorite mode of broadcast as has FMradio, where a single approved system led to essentially
total market coverage.
The Motorola C-Quam stereo signal is developed as shown in Figure 23. Thecarrier is phase-
shifted so that essentially two carrier signals are developed. The twoaudio signals (left and right
channels) are used to modulate the two cm.Tiers individually.
Note that a reference 25-Hz signal also modulates one of the carrier signals.
When the receiver detects the 25-Hz tone, it lights up an indicator to indicatestereo reception.
The two AM signals ru.·e summed out of the modulator for finalamplification and transmission.
Regular receivers simply detect the left-plus-rightsignals for normal monaural reception. A
specially equipped stereo receiver can differentiatebetween the two out-of-phase carriers and
thereby develop the two separateaudio signals.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 44: TDA1572T AM receiver.

Figure45: M stereo block diagram.

Due to the phase-shifting of the carrier, two sets of sidebands are generated90° out of phase.
Figure 24 provides a pictorial representation of this condition.This is an example of combining

1
CNS Basic Train ing- Analog Co mmunication ECATC

two separate signals (left and right channels) into one frequency band. Additional information
on this concept is provided in subsequentchapters.
A block diagram of a C-Quam AM stereo receiver is shown in Figure 25. AnMC13024 IC is the
basis of this system and the required "external" components arealso shown. This circuit provides
the complete receiver function requiring only astereo audio power amplifier for the left and right
channel outputs at pins 23 and 20.

Fig 46: Phase relationships in AM stereo.

Fig. 47 :C-Quam receiver system. (Courtesy of Motorola, Inc.)

As you study this block diagram, you may not understand some of the "blocks."Forinstance,
instead of a local oscillator input to the mixer, a voltage-controlled localoscillator (VCLO) is

1
CNS Basic Train ing- Analog Co mmunication ECATC

provided. It is controlled by an automatic frequency control(AFC) signal at pin 7 that is the result
of a PLL.

Receiver Analysis

lt is convenient to consider power gain or attenuation of various receiver stages interms of

decibels related to a reference power level. The most often used referencesare with respect to 1
mW (dBm) and 1 W (dBW). In equation form,

dBm = 10 log10 P/ 1mW (6)

dBW= 10 log10 P/ l W(7)

A dBm or dBW is an actual amount of power, whereas a dB represents a ratio ofpower. When
dealing with a system that has several stages, the effect of dB anddBm can be dealt with easily.
The following example shows this process.

Example 3

Consider the radio receiver shown in Figure 26. The antenna receives an 8-μ, V signal into its
50-D input impedance. Calculate the input power in watts, dBm, and dBW Calculate the power
driven into the speaker.

Fig. 48:Receiver block diagram.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Solution

Notice that dBm and dBW are separated by 30 dB-this is always the case because 30 dB
represents a I 000: I power ratio. To determine the power driven into the speaker, simply add
the gains and subtract the losses (in dB) all the way through the system. The -89 dBm at
the input is added to the 8-dB gain of the RF stage to give -81 dBm. Notice that the 8-dB
gain is simply added to the dBm input to give -81 dBm. This, and all subsequent stages, is
shown in Figure 26. Thus,

Pout(dBm) = -89 dBm + 8 dB + 3 dB + 24 dB + 26 dB

+ 26 dB - 2 dB + 34 dB
= 30 dBm into speaker

Therefore,
30dBm = 10log10 Pout /1mW
3 = log10Pout/1mW
1000 = Pout /1mW
Pout = l W

Example 3 assumed that the receiver's AGC was operating at some fixed levelbased on the input
signal's strength. As previously explained, the AGC system willattempt to maintain that same
output level over some range of input signal.Dynamicrangeis the decibel difference between the

1
CNS Basic Train ing- Analog Co mmunication ECATC

largest tolerable receiver input signal(without causing audible distortion in the output) and its
sensitivity (usua11y theminimum discernible signal). Dynamic ranges of up to about 100 dB
represent currentstate-of-the-art receiver performance.

Discussion

• Describe the process of modulation

• Sketch an AM waveform with various modulation indexes
• Explain the difference between a sideband and side frequency
• Analyze various power, voltage, and current calculations in AM systems
• Understand circuits used to generate AM
• Determine high- and low- level modulation systems from schematics and block diagrams
• Perform AM transmitter measurements using meters, oscilloscopes, and spectrum
analyzers
• Define the sensitivity and selectivity of a radio receiver
• Describe the operation of a diode detector in an AM receiver
• Sketch block diagrams for TRF and superheterodyne receivers
• Understand the generation of image frequencies and describe how to suppress them
• Recognize and analyze RF and IF amplifiers
• Describe the need for automatic gain control and show how it can be implemented
• Analyze the operation of a complete AM receiver system
• Perform a test analysis on the power levels (dBm) at each stage of an AM receiver system

1
CNS Basic Train ing- Analog Co mmunication ECATC

Module 3
Single Side Band

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Single Side Band

OBJECTIVE
After completing this Module, the student will be able to:
• Describe how an AM generator could be modified to provide SSB.
• Discuss the various types of SSB and explain their advantages compared to AM.
• Explain circuits that are used to generate SSB in the filter method and describe the filters
that can be used.
• Analyze the phase-shift method of SSB generation and give its advantages.
• Describe several methods used to demodulate SSB systems.
• Provide a complete block diagram for an SSB transmitter/receiver.
• Determine the frequencies at all points in an SSB receiver when receiving a single audio
tone.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

3. SINGLESIDEBANDCOMMUNICATIONS

3.1. Single Side Band Characteristics

The basic concept of single-sideband (SSB) communications was understood asearly as 1914. It
was first realized through mathematical analysis of an amplitudemodulatedRF carrier. Recall that
when a carrier is amplitude modulated by a singlesine wave, it generates three different
frequencies: (1) the original carrier withamplitude unchanged; (2) a frequency equal to the
difference between the carrierand the modulating frequencies, with an amplitude up to one-half
(at 100% modulation)the modulating signal; and (3) a frequency equa l to the sum of the
carrierand the modulating frequencies, with an amplitude also equal to a maximum ofone-half
that of the modulating signal. The two new frequencies, of course, are theside frequencies.
Upon recognition of the fact that sidebands existed, further investigationshowed that after
the carrier and one of the sidebands were eliminated, the othersideband could be used to transmit
the intelligence. Since its amplitude and frequencynever change, there is no information
contained in the carrier. Further experimentsproved that both sidebands could be transmitted,
each containing different intelligence,with a suppressed or completely eliminated carrier.
By 1923, the first patent for this system had been granted, and a successfulSSB
communications system was established between the United States and England.Today, SSB
communications play a vital role in radio communications because oftheir many advantages over
standard AM systems. The Federal CommunicationsCommission (FCC) recognizing these
advantages, further increased their use byrequiring most transmissions in the overcrowded 2- to
30-MHz range to be SSBstarting in 1977.

Powe r distribution

You should recall that in AM all the intelligence (information) is contained in thesidebands, but
two-thirds (or more) of the total power is in the carrier. It wouldappear that a great amount of
power is wasted during transmission. The basic principleof single-sideband transmission is to
eliminate or greatly suppress the highenergyRF carrier. This can be accomplished, but accurate
tuning is not possiblewithout a carrier and it does affect the fidelity of the music and sounds.
However,voice reception is still tolerable.
If a means of suppressing or completely eliminating the carrier is devised, thepower that
was used for the carrier can be converted into useful power to transmit theintelligence in the

1
CNS Basic Train ing- Analog Co mmunication ECATC

sidebands. Since both upper and lower sidebands contain the sameintelligence, one of these
could also be eliminated, thereby cutting the bandwidthrequired for transmissio n in half.
The total power output of a conventional AM transmitter is equal to thecarrier power plus
the sideband power. Conventional AM transmitters are rated incarrier power output. Consider a
low-power AM system operating at 100 percent
modulation. The carrier is 4 W and therefore each sideband is 1 W. The totaltransmitted power at
100 percent modulation is 6 W ( 4 W + 1 W + 1 W), butthe AM transmitter is rated as a 4 W
(just the carrier power) transmitter. If thissystem were converted to SSB, just one s ideband at 1
W would be transmitted.This, of course, assumes a sine-wave intelligence signal. SSB systems
areSingle-Sideband Communicationsmost often used for voice communications, which certainly
do not generate asinusoidal waveform.
SSB transmitters (and linear power amplifiers in general) are usually rated interms of
peak envelope power (PEP). To calculate PEP, multiply the maximum(peak) envelope voltage
by 0.707, square the result, and divide by the load resistance.For instance, an SSB signal with a
maximum level (over time) of 150 V p-p driveninto a 50-D antenna results in a PEP rating of
(150/2 x 0.707)2 + 50 n = 56.2 W.This is the same power rating that would be given to the 150-
V p-p sine wave, butthere is a difference. The 150-V p-p level in the SSB voice transmission
may occuronly occasionally, while for the sine wave it occurs every cycle. These calculations
are valid no matter what type of wavefonn the transmitter is providing. This couldrange from a
series of short spikes with low average power (perhaps 5 W out of thePEP of 56.2 W) to a sine
wave that would yield 56.2 W of average power. With anormal voice signal an SSB transmitter
develops an average power of only one- fourthto one-third its PEP rating. Most transmitters
cannot deliver an average power outputequal to their peak envelope power capability. This is
because their power suppliesand/or components in the output stage are designed for a lower
average power(voice operation) and cannot continuously operate at higher power levels.

Types of sideband trans mission

A number of single-sideband systems have been developed. The major types includethe
following:
l. In the standard single sideband, or simply SSB, system the Cai.Tier and one ofthe sidebands
are completely eliminated at the transmitter; only one sidebandis transmitted. This is quite

1
CNS Basic Train ing- Analog Co mmunication ECATC

popular with amateur radio operators. The chiefadvantages of this system are maximum
transmitted signal range with minimumtransmitter power and the elimination of carrier
interference.
2. Another system eliminates one sideband and suppresses the Carrier to a desiredlevel. The
suppressed carrier can then be used at the receiver for a reference,AGC, automatic frequency
control (AFC), and, in some cases, demodulationof the intelligence-bearing sideband. This is
called a single-sideband suppressedcan- ier (SSBSC). The suppressed cru.Tier is sometimes
called a pilotcarrier. This system retains fidelity of the received signal and minimizes
carrierinterference.
3. The type of system often used in militru.- y communications is referred to astwin-sideband
suppressed carrier, or independent sideband (ISB) transmission.This system involves the
transmission of two independent sidebands,each containing different intelligence, with the carrier
suppressed to a desiredlevel.
4. Vestigial sideband is used for television video transmissions. In it, a vestige(trace) of the
unwanted sideband and the carrier are included with one fullsideband.
5. A more recently developed system is called ru.nplitude-compandored singlesideband
(ACSSB). It is actually a type of SSBSC because a pilot carrier isusually included. In ACSSB
the amplitude of the speech signal is compressedat the transmitter and expanded at the receiver.
Performance gains of ACSSBsystems over SSB are explained in Section 4.

* Peak Envelope Powe r:method used to rate theoutput power of an SSBtransmitter.

* Pilot Carrier:the suppressed carrier inSSB; the carrier isreduced to a lower levelbut not
removedcompletely.
* Twin-SidebandSuppressed Carrier:the transmission of
twoindependentsidebands,containingdifferentintelligence, with thecarrier suppressed to adesired
level.
* Independent SidebandTrans mission:another name fortwin-sideband suppressedcarrier
transmission.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Advantages of SSB

The most important advantage of SSB systems is a more effective utilization of theavailable
frequency spectrum. The bandwidth required for the transmission of oneconventional AM signal
contains two equivalent SSB transmissions. This type ofcommunications is especially adaptable,
therefore, to the already overcrowded highfrequencyspectrnm.
A second advantage of this system is that it is less subject to the effects ofselective fading. In the
propagation of conventional AM transmissions, if theupper-sideband frequency strikes the
ionosphere and is refracted back to earth ata different phase angle from that of the carrier and
lower-sideband frequencies,distortion is introduced at the receiver. Under extremely bad
conditions, completesignal cancellation may result. The two sidebands should be ident ical in
phase withrespect to the carrier so that when passed through a nonlinear device (i.e., a
diodedetector), the difference between the sidebands and carrier is identical. That difference is
the inte11igence and wi11 be distorted in AM systems if the two sidebandshave a phase
difference.
Another major advantage of SSB is the power saved by not transmitting thecarrier and
one sideband. The resultant lower power requirements and weightreduction are especially
important in mobile communication systems.
The SSB system has a noise advantage over AM due to the bandwidthreduction (one-half).
Taking into account the selective fading improvement, noisereduction, and power savings, SSB
offers about a 10- to 12-dB advantage over AM.

Obtaining a 12-dB Advantage

Given a 50-kW carrier, the peak power is
4 X 50kW = 200kW

1
CNS Basic Train ing- Analog Co mmunication ECATC

An SSB transmitter with peak power equal to one AM sideband would transmit
12.5 kW.
10 log( 200/12.5)= 12 dB
This means that to have the same overall effectiveness, an AM system must transmit10 to 12 dB
more power than SSB. Some controversy exists on this issue becauseof the many variables that
affect the savings. Suffice it to say that a 10-W SSBtransmission is at least equivalent to the 100-
W AM transmission (10-dB difference).

3.2. Sideband Generation:

The balanced modulator

The purpose of a balanced modulator is to suppress (cancel) the carrier, leavingonly the two
sidebands. Such a signal is called a DSBSC (double-sideandsuppressedcarrier) signal. A very
common balanced modulator is shown in Figure l. Itis sometimes called a balanced ring
modulator or simply a ring modulator andsometimes a lattice modulator. Consider the carrier
with the instantaneous conventionalcun-ent flow as indicated by the arrows. The cun-ent flow
through both halvesofL5 is equal but opposite, and thus the carrier is canceled in the output. This
is alsotrue on the carrier's other half-cycle, only now diodes Band C conduct instead
ofAandD.Considering just the modulating signal, cun-ent flow occurs from winding Li.through
diodes C and DorAand B but not through L5. Thus, there is no output ofthe modulating signal
either. Now with both signals applied, but with the carrieramplitude much greater than the
modulating signal, the conduction is detemlinedby the polarity of the carrier. The modulating
signal either aids or opposes thisconduction. When the modulating signal is applied, current will
flow from Li. AnddiodeD will conduct more than A, and the current balance in winding L5 is
upset.This causes outputs of the desired sidebands but continued suppression of theearlier. This
modulator is capable of 60 dB carrier suppression when carefullymatched diodes are utilized. It
relies on the nonlinearity of the diodes to generatethe sum and difference sideband signals.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 49: Balanced ring modulator.

LIC Balanced modulator

A balanced modulator of the type previously explained requires extremely wellmatched

components to provide good suppression of the carrier (40 or 50 dBsuppression is usually
adequate). This suggests the use of LICs because of the superiorcomponent- matching
characteristics obtainable when devices are fabricated onthe same silicon chip. A number of
devices specially formulated for balanced modulatorapplications are available. A data sheet for
the AD 630 is provided in Figure 2.

*Balanced Modulator :modulator stage thatmixes intelligence with thecarrier to produce

bothsidebands with the carriereliminated
*Double-Sideband :Suppressed Carrieroutput signal of abalanced modulator
*Balanced RingModulator :balanced modulator design that connects fourmatched diodes
in a ring configuration
*Ring Modulator :another name for balancedring modulator.
*Lattice Modulator :another name for balancedring modulator.

1
CNS Basic Train ing- Analog Co mmunication ECATC

3.3.SSB Filte rs

Once the carrier has been eliminated, it is necessary to cancel one of the sidebandswithout
affecting the other one. This requires a sharply defined filter, as Figure 3helps illustrate. Voice
transmission requires audio frequencies from about 100 Hz to3 kHz. Therefore, the upper and
lower sidebands generated by the balanced modulatorare separated by 200 Hz, as shown in
Figure 3.
The required Q depends on the center or carrier frequency, fc; the separationbetween the two
sidebands, !l.f; and the desired attenuation level of the unwantedsideband. It can be calculated
from

where dB is the suppression of the unwanted sideband.

Example 1

Calculate the required Q for the situation depicted in Figure 3 for

(a) A I-MHz carrier and 80-dB sideband suppression.
(b) A 100-kHz carrier and 80-dB sideband suppression.

Fig. 50: Sideband suppression.

1
CNS Basic Train ing- Analog Co mmunication ECATC

A practical consequence of the preceding example is that the SSB signalwould be generated
around the lower 100-kHz carrier in conjunction with a crystalfilter. Then, after removing one
sideband, an additional frequency translation isusually employed to get the sideband up to the
desired frequency range. This isaccomplished with a mixer circuit.
Both SSB transmitters and receivers require selective bandpass filters in theregion of 100 to 500
kHz. In receivers a high order of adjacent channel rejection isrequired if channels are to be
closely spaced to conserve spectrum space. The filterused, therefore, must have very steep
skittcharactetistics (fast roll-off) and a flatbandpass characteristic to pass all frequencies in the
band equally well. These filterrequirements are met by crystal filters, ceramic filters, and
mechanical filters. A fourthtype of high-Q filter of more recent popularity is the surface acoustic
wave (SAW)filter. It is often used in TV and radar applications. It is most applicable to higher
frequencies than are typically used in SSB systems.

CrystalFilters

The crystal filter is commonly used in single-sideband systems to attenuate theunwanted

sideband. Because of its very high Q, the crystal filter passes a muchnan·ower band of
frequencies than the best LC filter. Crystals with a Q up to about50,000 are available.
The equivalent circuit of the crystal and crystal holder is illustrated inFigure 4(a). The
components Ls, Cs, and Rsrepresent the series resonant circuit of thecrystal itself.Cprepresents
the parallel capacitance of the crystal holder. The crystaloffers a very low-impedance path to the
frequency to which it is resonant and ahigh- impedance path to other frequencies. However, the

1
CNS Basic Train ing- Analog Co mmunication ECATC

crystal holder capacitance,Cp, shunts the crystal and offers a path to other frequencies. For the
crystal to operateas a bandpass filter, some means must be provided to counteract the shunting
effect of the crystal holder. This is accomplished by placing an external variable
capacitor in the circuit [C1 in Figure 4(b)].
In Figure 4(b), a simple bandpass crystal filter is shown. The variable capacitorC1 , called the
phasing capacitor, counteracts holder capacitance CP. C 1 can beadjusted so that its capacitance
equals the capacitance of CP. Then both CP and C1 pass undesired frequencies equally well.
Because of the circuit arrangement, thevoltages across CP and C1 due to undesired frequencies
are equal and 180° out of

Fig. 51: Crystal equivalent circuit (a) and filter (b).

Therefore, undesirable frequencies are canceled and do not appear in theoutput. This cancellation
effect is called the rejection notch.For circuit operation, assume that a lower sideband with a
maximum frequency of 99.9 kHz and an upper sideband with a minimum frequency of 100.l kHz
are applied to the input of the crystal filter in Figure 4(b).

* Surface AcousticWaveFilter:an extremely high-Q filteroften used in TV and

Radar applications.
* Phasing Capacitor:cancels the effect ofanother capacitance by a180° phase difference.
* Rejection Notch:a narrow range offrequenciesattenuatedby a filter, which can be tuned to
minimize interference phase.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Assume that the upper sideband is the unwanted sideband. By selecting a crystal that wi11
provide a lowimpedancepath (series resonance) at about 99.9 kHz, the lower-sideband
frequencywi11 appear in the output. The upper sideband, as well as all other frequencies,
willhave been attenuated by the crystal filter. Improved performance is possible whentwo or
more crystals are combined in a single filter circuit.

Ceramic Filters

Ceramic filters utilize the piezoelectric effect just as crystals do. However, they arenormally
constructed from lead zirconate-titanate. While ceramic filters do not offerQs as high as a crystal,
they do outperform LC filters in that regard. A Q of up to2000 is practical with ceramic filters.
They are lower in cost, more rugged, andsmallerin size than crystal filters. They are used not
only as sideband filters but alsoas replacements for the tuned IF transfor mers for
superheterodynereceivers.

Ceramic filters.

Fig 52: Ceramic filterand response curve.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The circuit symbol for a ceramic filter is shown in Figure 5(a) and a typicalattenuation response
curve is shown in Figure 5(b ). Note that the bandwidths at 60 dBand at 6 dB are shown. The
ratio of these two bandwidths (8 kHz/6.8 kHz = 1.18) isdefinedas the shape factor. The shape
factor (60-dB BW divided by a 6-dB BW)provides an indication of the filter's selectivity. The
ideal value of 1 would indicate averticalslope at both frequency extremes. The ideal filter would
have a horizontal slope within the pass band with zero attenuation. The practical case is shown in
Figure 5(b ),where a variation is illustrated. This variation is termed the peak-to-valley ratio or
Rippleamplitude. The shape factor and ripple amplitude characteristics also apply to
the mechanical filters discussed next.

Mechanical Filters

Mechanical filters have been used in single-sideband equipment since the 1950s.Some of the
advantages of mechanical filters are their excellent rejection characteristics,extreme ruggedness,
size small enough to be compatible with the miniaturizationof equipment, and a Q in the order of
10,000, which is about 50 times thatobtainable with LC filters.
The mechanical filter is a device that is mechanically resonant; it receiveselectrical energy,
converts it to mechanical vibration, then converts this mechanicalenergy back into electrical
energy as the output. Figure 6 shows a cutaway view ofa typical unit. There are four elements
constituting a mechanical filter: (1) an inputtransducer that converts the electrical energy at the
input into mechanical vibrations,(2) metal disks that are manufactured to be mechanically
resonant at the desiredfrequency, (3) rods that couple the metal disks, and (4) an output
transducer thatconverts the mechanical vibrations back into electrical energy.
Not all the disks are shown in the illustration. The shields around the transducercoils have
been cut away to show the coil and magnetostrictive driving rods. As youcan see by its
symmetrical construction, either end of the filter may be used as theinput.

*Shape Factor: ratio of the 60-dB and6-dB bandwidths of ahigh-Q bandpass filter.
*Peak-to-Valley Ratio: another name for rippleamplitude.
*Ripple Amplitude: variation in attenuation ofa sharp band pass fi lterwithin its 6-
dBbandwidths.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 53:Mechanical filter.

Figure 7 is the electrical equivalent of the mechanical filter. The disks of themechanical filter are
represented by the series resonant circuits L1 C1 while C2represents the coupling rods. The
resistance R in both the input and output representsthe matching mechanical loads. Phase shift of
the input signal is introduced by theL and C components of the mechanical filter. For digital
applications, a phase shiftcan affect the quality of the digital pulse. This can lead to an increase
in data errorsor bit errors. In analog systems, the voice transmission is not affected as
muchbecause the ear is very forgiving of distortion.
Let us assume that the mechanical filter of Figure 6 has disks tuned to passthe frequencies of the
desired sideband. The input to the filter contains bothsidebands, and the transducer driving rod
applies both sidebands to the first disk.The vibration of the disk will be greater at a frequency to
which it is tuned (resonantfrequency), which is the desired sideband, than at the undesired
sideband frequency.
The mechanical vibration of the first disk is transferred to the second disk, but asmaller
percentage of the unwanted sideband frequency is transferred. Each ti methe vibrations are
transferred from one disk to the next, there is a smaller amountof the unwanted sideband. At the
end of the filter there is practically none of theundesired sideband left. The desired sideband
frequencies are taken off the transducercoil at the output end of the filter.
Varying the size of C2 in the electrical equivalent circuit in Figure 7 variesthe bandwidth
of the filter. Similarly, by varying the mechanical coupling betweenthe disks (Figure 6), that is,
by making the coupling rods either larger or smaller,

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 54: Electrical analogy of a mechanical filter.

the bandwidth of the mechanical filter is varied. Because the bandwidth varies
approximately as the total cross-sectional area of the coupling rods, the bandwidth
of the mechanical filter can be increased by using either larger coupling rods or
more coupling rods. Mechanical filters with bandwidths as narrow as 500 Hz and
as wide as 35 kHz are practical in the range 100 to 500 kHz.

3.4. SSB Trans mitters

Filter Method

Figure 8 is a block diagram of a modem single-sideband transmitter using a balancedmodulator

to generate DSB and the filter method of eliminating one of thesidebands. For illustrative
purposes, a single-tone 2000-Hz intelligence signal isused, but it is normally a complex
intelligence signal, such as that produced by thehuman voice.
A 9-MHz crystal frequency is used because of the excellent operatingcharacteristics of
monolithic filters at that frequency. The 2-kHz signal is amplifiedand mixed with a 9-MHz
carrier (conversion frequency) in the balanced modulator.
Remember, neither the carrier nor audio frequencies appear in the output of thebalanced
modulator; the sum and difference frequencies (9 MHz ::!::: 2 kHz) are itsoutput. As illustrated
in Figure 8, the two sidebands from the balanced modulatorare applied to the filter. Only the
desired upper sideband is passed. The dashed linesshow that the carrier and lower sideband have
been removed.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The output of the first balanced modulator is filtered and mixed again with anew conversion
frequency to adjust the output to the desired transmitter frequency.

Fig 55: SSB transmitter block diagram.

After mixing the two inputs to get two new sidebands, the balanced modulator
removes the new 3-MHz carrier and applies the two new sidebands (3102 kHz and
2898 kHz) to a tunable linear power amplifier.

Example 2

For the transmitter system shown in Figure 8, determine the filter Q required in
the linear power amplifier.

*Conversion Frequency: another name for thecarrier in a balancedmodulator.

*Crystal-Lattice Filter: filter containing at leasttwo but usually fourcrystals.
*Continuous Wave: undamped sinusoidal waveform produced by an oscillator in a radio.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Solution

The second balanced modulator created another DSB signal from the SSB signal of the preceding
high-Q filter. However, the frequency translation of the second balanced modulator
means that a low-quality filter can be used once again to create SSB. The new DSB signal
is at about 2.9 MHz and 3. 1 MHz. The required filter Q is

3MHz / (3.1 MHz - 2.9 MHz) = 3MHz / 0.2MHz = 15

The input and output circuits of the linear power amplifier are tuned to rejectone sideband and
pass the other to the antenna for transmission. A standard LC filteris now adequate to remove
one of the two new sidebands. The new sidebands areabout 200 kHz apart ( = 3100 kHz - 2900
kHz), so the required Q is quite low.
(See Example 2 for further illustration.) The high- frequency oscillator is variable sothat the
transmitter output frequency can be varied over a range of transmittingfrequencies. Since both
the carrier and one sideband have been eliminated, all thetransmitted energy is in the single
sideband.

Phase Method

The phase method of SSB generation offers the following advantages over the filter
method:
1. There is greater ease in switching from one sideband to the other.
2. SSB can be generated directly at the desired transmitting frequency, whichmeans that
inte1mediate balanced modulators are not necessary.
3. Lower intelligence frequencies can be economically used because a high-Qfilter is not
necessary.
Despite these advantages, the filter method is rather firmly entrenched formany systems because
of adequate performance and the complexity of the phasemethod. The increased availability of
special LICs in the past has increased SSBdesigns using the phase method.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The phase method of SSB generation relies on the fact that the upper andlower sidebands of an
AM signal differ in the sign of their phase angles. This meansthat phase discrimination may be
used to cancel one sideband of the DSB signal.
Consider a modulating signal f(t) to be a pure cosine wave. A resultingbalanced modulator
output (DSB) can then be written as

(2)
wherecosw;tis the intelligence signal and coswetthe carrier. The term cosAcosB
is equal to t[cos (A + B) + cos(A - B)] by trigonometric identity, and therefore
Equation (2) can be rewritten as

(3)
If another signal,

(4)
were added to Equation (3), the upper sideband would be canceled, leaving j ust the
lower sideband,

.
Since the signal in Equation (4) is equal to

by trigonometric identity, it can be generated by shifting the phase of the carrier and
intelligence signal by exactly 90° and then feeding them into a balanced modulator.
Recall that sine and cosine waves are identical except for a 90° pha~e difference.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 56 : Phase-shift SSB generator.

A block diagram for the system just described is shown in Figure 10. The upperbalanced
modulator receives the canier and intelligence signals directly, while thelower balanced
modulator receives both of them shifted in phase by 90°. Thus,comb ining the outputs of both
balanced modulators in the adder results in an SSBoutput that is subsequently amplified and then
driven into the transmitting antenna.
A major disadvantage of this system is the 90° phase-shifting networkrequired for the
intelligence signal. The carrier 90° phase shift is easily accomplishedbecause of its single-
frequency nature, but the audio signal covers a widerange of frequencies. To obtain exactly 90°
of phase shift for a complete rangeof frequencies is difficult. The system is critical inasmuch as
an 88° phase shift(2° enor) for a given audio frequency results in about 30 dB of unwanted
sidebandsuppression instead of the desired complete suppression obtained at 90° phaseshift. The
difficulty in obtaining adequate performance o f the intelligence phaseshiftingnetwork is
becoming less of a problem with the newer LICs designed toaddress this situation.

1
CNS Basic Train ing- Analog Co mmunication ECATC

ACSSB Systems

Amplitude compandoring (compression-expandor) single-sideband (ACSSB)systems are now

allowing narrowband voice communications with the performanceof FM systems for the land-
mobile communications industry. This equivalentperformance is provided with less than one-
third the bandwidth of thecomparable FM systems. The basis of ACSSB is to compress the audio
beforemodulation and to expand it following demodulation at the receiver. A commonly
used method to achieve this is use of the SA571 compandor LIC. It is shown inFigure 11
connected as an expandor. This IC has a unity gain for a 0-dBm input.
When used as a compressor, all negative dBm power levels are increased and positive
dBm powers are decreased. For example, - 40 dBm becomes -20 dBm, + 15

Fig 57: Amplitude expandor circuit.

dBm becomes + 7.5 dBm, and so on. The expandor reverses the process to restorethe signa l's
original dynamic range. Thus, a -20 dBm to +7.5 dBm signalbecomes - 40 dBm to + 15 dBm at
the expandor's output. The only signals notchanged by the 571 IC are those at 0 dBm.

*Compandor: compress/expand; to provide better noise performance, a variable gain circuit at

thetransmitter increases its gain for low- level signals; a complementary circuit in the receiver
reverses the process to restore the original signal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

This system significantly cuts down the dynamic range that must be dealtwith. It allows
the lower- level signals to be transmitted with greater power whileremaining within the PEP
ratings of the transmitter power amplifier for the highestlevelsignals. Thus, the SIN ratio is
significantly improved at the lower end, whilethe somewhat increased noise for the louder
passages (due to their reducedamplitude) is not a problem. At the receiver, the expandor restores
the demodulatedoutput to its original dynamic range.
These ACSSB systems also include a pilot carrier signal as illustrated inFigure 12. It is
shown added to the audio signal sufficiently separated so that thereceiver can ultimately
distinguish between the two. The audio passband for voicetransmission is fully attenuated by 3
kHz, and a pilot tone at 3.1 kHz above the eliminatedcarrier is the norm. It is suppressed by 10
dB from the maximum PEP as shownin Figure 12. Thus, the transmitter will have output power
of about to (-10 dB) ofthe maximum when there is no voice modulation. At the receiver, the pilot
tone isusually compared to a reference oscillator in a phase- locked- loop (PLL) circuit. The PLL
difference voltage is used to shift the receiver oscillator until error is eliminated. The pilot tone
can also be used for AGC and squelch circuits at the receiver.

Fig 58: ACSSB signal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Transmitte r Li near Power Amplifier

Once an SSB signal has been generated, a linear power amplifier is necessary to
obtainsignificant power 1eve1s for transmission. The circuit in Figure 13 provides 140 W of PEP
nominalout"put power from 2 to 30 MHz when supplied with about 3 W of signal input.Fairly
linear outputs up to 200 WPEP are possible with increased input drive. Its simplicityand use of
low-cost components makes it an attractive design for mobile transmitters.It operates on the
standard 13.6 V de available from automotive electrical systems.
The amplifier is a class AB push-pull design. The quiescent current for eachMRF454
power transistor (Q3 and Q4) is about 500 mA. This amount of bias is

1
CNS Basic Train ing- Analog Co mmunication ECATC

C1-33 pF dipped mica R2-10 k.Q, 0.25 W

Cz-18 pF dipped mica R3- 10 kQ, 0.25 W
C3-10 μF, 35 V de for AM operation R4-33 .Q, 5 W wirewound
100 μF, 35 V de for SSB operation R5, R6-10 Q , 0.5 W
Cc--0.1 μFErie R7-100 .Q, 0.25 W
C5-10 μF, 35 V de electrolytic RFC1-9 ferroxcube beads on No. 18 A WG
Co-1 μFtantalum wire
~. 001 pF Erie disk D1- 1N4001
Cg, C9-330 pF dipped mica D2-IN4997
CI0-24 pF dipped mica Q t. Q2-2N4401
C11-910 pF dipped mica Q3, Q4-MRF454
C12-1100 pF dipped mica T 1, T 2- 16: 1 transformers
C13-500 pF, 3 V de electrolytic K1- Potter & Brumfield KTl IA 12 V de
R1- 100 kQ, 0.25 W relay or equivalent

Fig. 59: Linear power amplifier.

1
CNS Basic Train ing- Analog Co mmunication ECATC

3.5. SSB Demodulation

One of the major advantages of SSB for voice transmission has been shown to be theelimination
of the transmitted carrier. However, this advantage does not apply tomusic. We have shown that
this allows an increase in effective radiated power (erp)because the sidebands contain the
information and the never-changing carrier isredundant. Unfortunately, even though the carrier is
redundant (contains no information),it is needed at the receiver! Recall that the intelligence in an
AM system isequal in frequency to the difference of the sideband and carrier frequencies.

Waveforms

Figure 14(a) shows three different sine-wave intelligence signals; in Figure 14(b), theresulting
AM waveforms are shown, and Figure 14(c) shows the DSB (no carrier)waveform. Notice that
the DSB envelope (drawn in for illustrative purposes) lookslike a full-wave rectification of the
corresponding AM waveform's envelope. It isdouble the frequency of the AM envelope. In
Figure 14(d), the SSB waveforms aresimply pure sine waves. This is precisely what is
transmitted in the case of a sinewavemodulating signal. These waveforms are either at the carrier
plus the intelligencefrequency (usb) or carrier minus intelligence frequency (lsb). An SSB
receiverwould have to somehow "reinsert the carrier" to enable detection of the original
audio or intelligence signal. A simple way to form an SSB detector is to use a mixerstage
identical to a standard AM receiver mixer. The mixer is a nonlinear device, andthe local
oscillator input should be equivalent to the desired carrier frequency.

Mixer SSB Demodulator

Figure 15 shows this situation pictorially. Consider a 500-kHz carrier frequency that hasbeen
modulated by a I-kHz sine wave. If the upper sideband were transmitted, thereceiver's
demodulator would see a 501-kHz sine wave at it~ input. Therefore, a500-kHz oscillator input
will result in a mixer output frequency component of I kHz,which is the desired result. If the
500-kHz oscillator is not exactly 500 kHz, the recoveredintelligence will not be exactly 1 kHz. If

1
CNS Basic Train ing- Analog Co mmunication ECATC

the receiver is to be used on severalspecific channels, a crystal for each channel will provide the
necessary stability. If the
receiver is to be used over a complete band of frequencies, the variable frequency
oscillator(VFO), often called the beat frequency oscillator (BFO), must have some s01t
ofautomatic frequency control (AFC) to provide adequate quality reception. This can
beaccomplished by including a pilot carrier signal with the transmitted SSB signal. The

Fig. 60 :AM, DSB, and SSB waves from sinusoidal modulating signals.

*Product Detector: using a balancedmodulator to recover theintelligence in an SSBsignal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 61: Mixer used as SSB demodulator.

pilotcanier can then be used to calibrate the receiver's oscillator at periodic intervals.Another
approach is to utilize rather elaborate AFC circuit~ completely at the receiver,and the third
possibility is the use of frequency synthesizers.

BFO Drift Effect

In any event, even minor drifts in BFO frequency can cause serious problems inSSB reception. If
the oscillator drifts ± 100 Hz, a 1-kHz intelligence signalwould be detected either as 1100 Hz or
900 Hz. Speech transmission requires lessthan a ±100-Hz shift or the talker starts sounding like
Donald Duck and becomescompletely unintelligible. Obtaining good-quality SSB reception of
music anddigital signals requires a carrier.

Example 3

At one instant of time, an SSB music transmission consists of a 256-Hz sine wave andits second
andfourth harmonics, 512 Hz and 1024 Hz. If the receiver's demodulatoroscillator has drifted 5
Hz, determine the resulting speaker output frequencies.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Solution

The 5-Hz oscillator drift means that the detected audio will be 5 Hz in error, either upor down,
depending on whether it is a usb or lsb transmission and on the direction of theoscillator's drift.
Thus, the output would be either 251, 507, and 1019 Hz or 261, 517, andl 029 Hz. The speaker's
output is no longer harmonic (exact frequency multiples), andeven though it is just slightly off,
the human ear would be offended by the new "music."

Product Detector

As we have discussed, to recover the intelligence in an SSB (or DSB) signal, youneed to reinsert
the carrier. The balanced modulators used to create DSB can also beused to recover the
intelligence in an SSB signal. When a balanced modulator is usedin this fashion, it is usually
called a product detector. This is the most commonmethod of detecting an SSB signal.
Figure 16 shows another IC balanced modulator being used as a productdetector. It is the
Plessey Semiconductor SL640C. The capacitor connected to outputpin 5 forms the low-pass
filter to allow just the audio (low)-frequency component toappear in the output. The simplicity of
this demodulator makes its desirability clear.

Fig. 62 :SL640C SSB detector.

3.6. SSB Receivers

To see the relationship of the parts in a single-sideband receiver, observe the block
diagram in Figure 17. Basically, the receiver is similar to an ordinary AM superheterodyne

1
CNS Basic Train ing- Analog Co mmunication ECATC

receiver; that is, it has RF and rF amplifiers, a mixer, a detector, and

audio amplifiers. To permit satisfactory SSB reception, however, an additional mixer
(demodulator) and oscillator must replace the conventional diode detector.

Example 4

The SSB receiver in Figure 17 has outputs at 1 kHz and 3 kHz. The carrier used and suppressed
at the transmitter was 2 MHz, and the upper sideband was utilized. Determine the exact
frequencies at all stages for a 455-kHz IF frequency.

As shown before, the carrier frequency was suppressed at the transmitter; thus,for proper
intelligence detection, a carrier must be inserted by the receiver. The receiverillustrated in Figure
17 inserts a carrier frequency into the detector, although the carrierfrequency may be inserted at
any point in the receiver before demodulation.
When the SSB signal is received at the antenna, it is amplified by the RFamplifier and
applied to the first mixer. By mixing the output of the local

*Butterworth Filter :a constant-k type ofLC f ilter.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 63: SSB receiver block diagram.

oscillator with the input signal (heterodyning), a difference frequency, or IF, isobtained. The IF
is then amplified by one or more stages. Of course, this isdependent upon the type of receiver.
Up to this point it is identical to an AMsuperheterodyne receiver. The IF output is applied to the
second mixer (detector).The detector output is applied to the audio amplifier and then on to the
output
speaker.
Tuning the sideband receiver is somewhat more difficult than in a regular AMreceiver.
The carrier injection oscillator must be adjusted precisely to simulate thecarrier frequency at all
times. As previously explained, any tendency to dtift withinthe oscillator will cause the output
intelligence to be distorted.

Basic SSB Receiver

A basic SSB receiver is shown schematically in Figure 18. This superhet designfunctions well
without an RF amplifier. The input signal comes in to a fixed Butterworthfront-end filter (FL1)
that passes 3.75 to 4.0 MHz without tuning. AButterworth filter exhibits a very flat response in
the passband and approaches a6-dB slope per octave. Individual channels in this amateur radio
band are tuned byvarying the local oscillator (Q3) frequency. Notice that this stage is labeled
with VFO,or variable.frequency oscillator, and has a range from 4.253 to 4.453 MHz. This
signaland the received signal are applied to the two gates of the mixer (Q1). This

1
CNS Basic Train ing- Analog Co mmunication ECATC

3N211MOSFET provides high gain (gm= 30,000 μ.,S) and its output is applied to a
mechanicalfilter, F~. The specified filter has a 2.2-kHz bandwidth at the 3-dB points and hasa
5.5-kHz bandwidth at -60 dB.
The mechanical filter output is applied to the IF amplifier, Q2 , which isanother 3N211 MOSFET.
Its gain, and that of the audio amplifier U"are manuallycontrolled by ganged potentiometers R1A
and R18. The bias at gate 2 of Q 2 is variedby RIA · To obtain a wide range of control, it is
necessary to have gate 2 a voltor two less than gate 1. This is done by "bootstrapping" this stage
with an LED, Di.that conducts at about 1.5 V. Thus, when R1 A has its arm at ground, gate 2
iseffectively at -1.5 V and minimum gain for Q2 occurs.
Another 3N211 device is used as the LO (or YFO if you prefer). Gates 1 and 2 of Q3 are
tied together. The oscillator signal is applied from the gate of Q3 to gate 2of the mixer, Q1 • A
pure 3-V p-p sine wave is thereby available. D2 is used as aswitching diode to offset the VFO
frequency when changing from usb to lsb.
The product detector stage (Q 4 ) is fed from the IF amp into the source of Q4 .The beat
frequency oscillator (Q 5 ) is a switchable crystal oscillator. SIB selects eithercrystal Y1 or Y2 for
lsb or usb, respectively. The product detector output (at Q4's drain)is applied to a 741 op amp
audio amplifier that offers up to 40-dB gain. Its output issufficient to drive headphones, or an IC
power amp could be added if a speaker isneeded.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Discussion

• Describe how an AM generator could be modified to provide SSB.

• What are types of SSB and explain their advantages compared to AM.
• Explain circuits that are used to generate SSB in the filter method and describe the filters that
can be used.
• What is phase-shift method of SSB generation and what are its advantages.
• List out methods used to demodulate SSB systems.
• Draw a complete block diagram for an SSB transmitter/receiver.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Module 4
Frequency Modulation

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank page

1
CNS Basic Train ing- Analog Co mmunication ECATC

Frequency Modulation

OBJECTIVE

After completing this Module, the student will be able to:

 Define angle modulation and describe the two categories.
 Explain a basic capacitor microphone FM generator and the effects of voice amplitude
and frequency.
 Analyze an FM signal with respect to modulation index, sidebands, and power.
 Describe the noise suppression capabilities of FM and how they relate to the capture
effect and preemphasis.
 Provide various schemes and circuits used to generate FM.
 Explain how a PLL can be used to generate FM.
 Describe the multiplexing technique used to add stereo to the standard FM broadcast
systems.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Blank Page

1
CNS Basic Train ing- Analog Co mmunication ECATC

4. Frequency Modulation

4.1.Angle Modulation

There are three parameters of a sine-wave carrier that can be varied to allow it tocarry a
low- frequency intelligence signal. They are its amplitude, frequency, andphase. The latter
two, frequency and phase, are actually interrelated, as one cannotbe changed without
changing the other. They both fall under the general category ofangle modulation. Angle
modulation is defined as modulation where the angle of asine-wave carrier is varied from
its reference value. Angle modulation has twosubcategories, phase modulation and
frequency modulation, with the followingdefinitions:
Phase modulation (PM):angle modulation where the phase angle of a carrieris
caused to depart from its reference value by an amount proportional tothe modulating
signal amplitude.
Frequency modulation (FM):angle modulation where the instantaneous
frequencyof a carrier is caused to vary by an amount proportional to the modulatingsignal
amplitude.
The key difference between these two similar forms of modulation is that inPM
the amount of phase change is proportional to intelligence amplitude, while inFM it is the
frequency change that is proportional to intelligence amplitude. As itturns out, PM is not
directly used as the transmitted signal in communications systemsbut does have
importance because it is often used to help generate FM, and aknowledge of PM helps us
to understand the superior noise characteristics of FMas compared to AM systems. In
recent years, it has become fairly common practiceto denote angle modulation simply as
FM instead of specifically referring to FMand PM.

Radio Emission Classifications

Table l gives the codes used to indicate the various types of radio signals. The first letter
is A, F, or P to indicate AM, FM, or PM. The next code symbol is one of the numbers 0
through 9 used to indicate the type of transmission. The last code symbol is a subscript. If

1
CNS Basic Train ing- Analog Co mmunication ECATC

there is no subscript, it means double-sideband, full carrier. Here are some examples of
emission codes:

Table 1
Codes Emission Types
A3a SSB, reduced carrier
A3i SSB, no carrier
F3 FM, double-sideband, full carrier
A 7i SSB, no carrier, multiple sidebands with different messages
10A3 AM, double-sideband, full carrier, 10-kHz bandwidth

Notice the last example. If an emission code is preceded by a number, that number is the
bandwidth of the signal in kHz.

*Angle Modulation:superimposing theintelligence signal on ahigh-frequency

carrier sothat its phase angle orfrequency is altered as afunction of theintelligence
amplitude.
*Phase Modulation:superimposing theintelligence signal on ahigh-frequency
carrier sothat the carrier's phaseangle departs from itsreference value by anamount
proportional to theintelligence amplitude
*Frequency Modulation:superimposing theintelligence signal on ahigh-frequency
carrier sothat the carrier's frequencydeparts from its referencevalue by an
amountproportional to theintelligence amplitude

1
CNS Basic Train ing- Analog Co mmunication ECATC

4.2. A Simple FM Generator

To gain an intuitive understanding of FM, consider the system illustrated in Figure l.This
is actua11y a very simple, yet highly instructive, FM transmitting system. Itconsistsof an
LC tank circuit, which, in conjunction with an oscillator circuit, generatesa sine-wave
output. The capacitance section of the LC tank is not a standard capacitorbut is a
capacitor microphone. This popular type of microphone is often referredto as a condenser
mike and is, in fact, a variable capacitor. When no sound wavesreach its plates, it
presents a constant value of capacitance at its two output terminals.When sound waves
reach the mike, however, they alternately cause its plates tomove in and out. This ca uses
its capacitance to go up and down around its centervalue. The rate of this capacitance
change is equal to the frequency of the soundwaves striking the mike, and the amount of
capacitance change is proportional to theamplitude of the sound waves.
Because this capacitance value has a direct effect on the oscillator's frequency,the
fo11owing two important conclusions can be made concerning the system'soutput
frequency:
l. The frequency of impinging sound waves determines the rate of frequencychange.
2. The amplitude of impinging sound waves determines the amount of frequencychange.

Fig. 64: Capacitor microphone FM generator.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 65: FM representation.

Consider the case of the sinusoidal sound wave (the intelligence signal)shown in Figure
2(a). Up until time T1 the oscillator's waveform in Figure 2(b) isa constant frequency with
constant amplitude. This corresponds to the carrier frequency(f c) or rest frequency in FM
systems. At T1 the sound wave in Figure 2(a)starts increasing sinusoidally and reaches a

1
CNS Basic Train ing- Analog Co mmunication ECATC

maximum positive value at T2.Duringthis period, the oscillator frequency is gradually

increasing and reaches its highestfrequency when the sound wave has maximum
amplitude at time T2.From time T2 to T4 the sound wave goes from maximum positive to
maximum negative and theresulting oscillator frequency goes from a maximum
frequency above the rest valueto a maximum value below the rest frequency. At time T3
the sound wave is passingthrough zero, and therefore the oscillator output is
instantaneously equal to thecarrier frequency.
The relationship for the FM signal generated by the capacitor microphone canbe
written as shown in Equation (1).
f out =fc+ kei(1)

wheref out= instantaneous output frequency

f c= output carrier frequency
k = deviation constant [kHz/V]
ei= modulating (intelligence) input
Equation 1 shows that the output can·ier frequency (fc) depends on the amplitude and
thefrequency of the modulating signal (ei) and also on the frequency deviation generated
bythe microphone (k) for a given input level. The unit k is called a deviation constant.
Thisdefines how much the can- ier frequency will deviate (change) for a given input
voltagelevel. The deviation constant (k) is defined in units of kHz/V. For example, if k =
1kHz /10 m V, then a modulating signal input level of 20 m V will cause a 2-kHz
frequencyshift. A modulating signal input level of -20 mV will cause a -2-kHz frequency
shift.Assuming that the 20-mV level is from a sinusoid (a sinusoid contains both a
positiveand negative peak), then the rate at which the frequency changes (deviates above
andbelow the original earner frequency) depends on the frequency of the modulating
signal.For example, if a 20- mV, 500-Hz input signal is applied to the microphone, then
the carrierfrequency will deviate ±2 kHz at a rate of 500 Hz.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 1

A 25-m V sinusoid at a frequency of 400 Hz is applied to a capacitor microphoneFM

generator. If the deviation constant for the capacitor microphone FM generator is 750
Hz/JO mV, determine
(a) The frequency deviation generated by an input level of 25 m V.
(b) The rate at which the carrier frequency is being deviated.

Solution

(a) positive frequency deviation= 25 mV x 750 Hz/10 mV = 1875 Hz or 1.875 kHz

negative frequency deviation = -25 m V x750 Hz/10mV= 1875 Hz or - 1.875 kHz
The total deviation is written as ±2.25 kHz for the given input signal level.
(b) The input frequency (f i) is 400 Hz; therefore, by Equation (1)
f out= fc+ kei (1)
The carrier will deviate+ 1.875 kHz at a rate of 400 Hz.

The two major concepts

The amount of oscillator frequency increase and decrease around Jcis called thefrequency
deviation, o. This deviation is shown in Figure 2(c) as a function of time.Notice that this
is a graph of frequency versus time-not the usual voltage versustime. It is ideally shown
as a sine-wave replica of the original intelligence signal. Itshows that theoscillator output
is indeed an FM waveform. Recall that FM isdefined as a sine-wave carrier that changes
in frequency by an amount proportionalto the instantaneous value of the intelligence
wave and at a rate equal to the intelligencefrequency.Figure 2(d) shows the AM wave
resulting from the intelligence signal shown in Figure 2(a). This should help you to see
the difference between an AM and FM signal.
*Deviation Constant (k) :how much the carrierfrequency will deviate fora given
modulating inputvoltage level
*Frequency Deviation :amount of carrier frequencyincrease or decrease aroundits
center reference value

1
CNS Basic Train ing- Analog Co mmunication ECATC

In the case of AM, the carrier's amplitude is varied (by its sidebands) in stepwith
the intelligence, while in FM, the carrier's frequency is varied in step with theintelligence.
The capacitor microphone FM generation system is seldom used in practicalapplications;
its importance is derived from its relative ease of providing anunderstanding of FM
basics. If the sound-wave intelligence striking the microphonewere do ubled in frequency
from 1 kHz to 2 kHz with constant amplitude,the rate at which the FM output swings
above and below the center frequency(fc) would change from 1 kHz to 2 kHz. Because
the intelligence amplitude wasnot changed, however, the amount of frequency deviation
(o) above and belowfcwill remain the same. On the other hand, if the 1-kHz intelligence
frequencywere kept the same but its amplitude were doubled, the rate of deviation above
and below fcwould remain at 1 kHz, but the amount of frequency deviationwould double.
As you continue through your study of FM, whenever you start getting
boggeddown on basic theory, it will often be helpful to review the capacitor mike FM
generator.Remember:
1. The intelligence amplitude determines the amount of carrier frequencydeviation.
2. The intelligence frequency (fi) determines the rate of carrier frequencydeviation.

Example 2

An FM signal has a center frequency of 100 MHz but is swinging between100.001 MHz
and 99.999 MHz at a rate of 100 times per second. Determine:
(a) The intelligence frequency fi.
( b) The intelligence amplitude.
( c) What happened to the intelligence amplitude if the frequency deviationchanged to
between 100.002 and 99.998 MHz.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Solution

(a) Because the FM signal is changing frequency at a 100-Hz rate, .fi= 100 Hz.
(b) There is no way of determining the actual amplitude of the intelligence signal. Every
FMsystem has a different proportionality constant between the intelligence amplitude
andthe amount of deviation it causes.
(c) The frequency deviation has now been doubled, which means that the
intelligenceamplitude is now double whatever it originally was.

4.3. FM Analysis

The complete mathematical analysis of angle modulation requires the use of high-
levelmathematics. For our purposes, it will suffice simply to give the solutions and discussthem.
For phase modulation (PM), the equation for the instantaneous voltage is

e = A sin(ωct + mp sin ωit) (2)

wheree= instantaneous voltage

A = peak value of original carrier wave
ωc= carrier angular velocity (21T/J
m p = maximum phase shift caused by the intelligence signal (radians)
ωi= modulating (intelligence) signal angular velocity (2Πf i)

The maximum phase shift caused by the intelligence signal, mp, is defined as themodulation
index for PM.
The following equation provides the equivalent formula for FM:

e = A sin(ωct + mf sin ωit) (3)

All the terms in Equation (3) are defined as they were for Equation (2), with theexception of the
new term, m1. [n fact, the two equations are identical except forthat tem1. It is defined as the
modulation index for FM, mr- It is equal to
m f = FM modulation index = ᵟ /fi (4)
where
ᵟ= maximum frequency shift caused by the intelligence signal (deviation)
f i= frequency of the intelligence (modulating) signal

1
CNS Basic Train ing- Analog Co mmunication ECATC

Comparison of Equations (2) and (3) points out the only difference between PMand FM.
The equation for PM shows that the phase of the carrier varies with themodulating signal
amplitude (because mp is determined by this), and in FM thecarrier phase is detennined by the
ratio of intelligence signal amplitude (whichdetermines o) to the intelligence frequency (jj).
Thus, FM is not sensitive to themodulating signal frequency but PM is. The difference between
them is subtle- infact, if the intelligence signal is integrated and then allowed to phase- modulate
thecarrier, an FM signal is created. This is the method used in the Armstrong indirectFM system,
as will be explained in Section 6. In FM the amount of deviation producedis not dependent on
the intelligence frequency, as it is for PM. The amountof deviation is proportional to the
intelligence signal amplitude for both PM andFM.These conditions are shown in Figure 3.

Ze ro-Carrie r Amplitude

Figure 4 shows the FM frequency spectrum for various levels of modulation whilekeeping the
modulation frequency constant. The relative amplitude of all componentsis obtained from Table
2. Notice from the table that between m f= 2 and

Fig. 66: Frequency spectrum for FM (constant modulating frequency, variabledeviation).

*Guard Bands :25-kHz bands at each endof a broadcast FM channelto help minimize
interference with adjacentstations

1
CNS Basic Train ing- Analog Co mmunication ECATC

1nJ = 2.5, the carrier goes from a plus to a minus value. The minus sign simply indicatesa phase
reversal, but when m1 = 2.4, the carrier component has zero amplitudeand all the energy is
contained in the side frequencies. This also occurs whenm1 = 5.5, 8.65, and between 10 and 12,
and 12 and 15.
The zero-carrier condition suggests a convenient means of determining thedeviation
produced in an FM modulator. A carrier is modulated by a single sinewave at a known
frequency. The modulating signal's amplitude is varied whileobserving the generated FM on a
spectrum analyzer. At the point where the carrieramplitude goes to zero, the modulation index,
m1, is determined based on the numberof sidebands displayed. If four or five sidebands appear
on both sides of thenulled carrier, you can assume that m1 = 2.4. The deviation, o, is then equal
to2.4 X A The modulating signal could be increased in amplitude, and the next carriernull should
be at 1nJ = 5 .5. A check on modulator linearity is thereby possiblebecause the frequency
deviation should be directly proportional to the modulatingsignal's amplitude.

Broadcast FM

Standard broadcast FM uses a 200-kHz bandwidth for each station. This is a verylarge allocation
when one considers that one FM station has a bandwidth that couldcontain many standard AM
stations. Broadcast FM, however, allows for a true highfidelitymodulating signal up to 15 kHz
and offers superior noise performance (seeSection 4).
Figure 5 shows the FCC allocation for commercial FM stations. The maximumallowed
deviation around the carrier is ±75 kHz, and 25-kHz guard bandsat the upper and lower ends are
also provided. The carrier is required to maintain a±2-kHz stability. Recall that an infinite
number of side frequencies are generatedduring frequency modulation, but their amplitude
gradually decreases as you moveaway from the carrier.

Fig. 67: Commercial FM bandwidth allocations for two adjacent stations.

In other words, the significant side frequencies exist up to±75 kHz around the carrier, and the
guard bands ensure that adjacent channelinterference will not be a problem.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Since full deviation (o) is 75 kHz, that is 100 percent modulation. By definition,100
percent modulation in FM is when the deviation is the full permissibleamount. Recall that the
modulation index, m1, is

(4)
so that the actual modulation index at 100 percent modulation varies inverselywith the
intelligence frequency, Ji. This is in contrast with AM, where full or100 percent modulation
means a modulation index of 1 regardless of intelligencefrequency.
Another way to describe the modulation index is by deviation ratio (DR).Deviation ratio
equals the result of dividing the maximum possible frequencydeviation by the maximum input
frequency, as shown in Equation (8).

(8)
Deviation ratio is a commonly used term in both television and FM broadcasting.For example,
broadcast FM radio permits a maximum carrier frequency deviation,!dev(max)• of ±75 kHz and
a maximum audio input frequency, fi(max) of 15 kHz.
Therefore, for broadcast FM radio, the deviation ratio (DR) is

and for broadcast television (NTSC format), the maximum frequency deviationof the aural
carrier, !dev(max) is ±25 kHz with a maximum audio input frequency,fi(max)• of 15 kHz.
Therefore, for broadcast TV (NTSC format), the deviationratio (DR) is

FM systems that have a deviation ratio greater than or equal to 1 (DR ;:::; 1)
areconsidered to be wideband systems, whereas FM systems that have a deviation ratioless than
l (DR < 1) are considered to be narrowband FM systems.

*Deviation Ratio (DR):maximum possiblefrequency deviation overthe maximum input

Frequency.
*Wideband FM:a system where the deviation ratio is > 1.
*Narrowband FM:FM signals used for voicetransmissions such aspublic
servicecommunication systems.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Narrowband FM

Frequency modulation is also widely used in communication (i.e., not toentertain) systems such
as those used by police, aircraft, taxicabs, weather service,and private industry networks. These
applications are often voice transmissions,which means that intelligence frequency maximums of
3 kHz are the norm.These are narrowband FM systems because Federal Communications
Commission(FCC) bandwidth allocations of 10 to 30 kHz are provided. Narrowband
FM(NBFM) systems operate with a modulation index of 0.5 to t .0. A glance at theBessel
functions in Table 2 shows that at these index values, only the first set (J1) of side frequencies
has a significant amplitude; the second (J2) and third (J3)lose amplitude quickly. Thus, we see
that NBFM has a bandwidth no wider thanan AM signal.

Example 6

(a) Determine the permissible range in maximum modulation index for commercialFM that has
30-Hz to J 5-kHz modulating frequencies.
( b) Repeat for a narrowband system that allows a maximum deviation of I-kHz and 100-Hz to 2-
kHz modulating frequencies.
(c) Determine the deviation ratio for the system in part (b).

Solution

(a) The maximum deviation in broadcast FM is 75 kHz.

mf=g/f 1

= 75 kHz/30Hz = 2500

For f i = 15 kHz
m f= 75 kHz/15kHz =5

(b) For f i = 2kHz:

m f= g/f 1 = 1 kHz/100 Hz = 0.5

(d) DR = f dev(max)/f i(max) = 1kHz/ 2 kHz = 0.5

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 7

Determine the relative total power of the carrier and side frequencies when m f= 0.25 for a 10-kW
FM transmitter:

Solution

For mf= 0.25, the carrier is equal to 0.98 times its unmodulated amplitude and the only
significant sideband is Ji, with a relative amplitude of 0.12 (from Table 2). Therefore,
because power is proportional to the voltage squared, the carrier power is

(0.98)2 X 10 kW = 9.604 kW

and the power of each sideband is

(0.12)2 X 10 kW = 144 W

The total power is

9604 W + 144 W + 144 W = 9.892 kW
= 10 kW

The result of Example 7 is predictable. In FM, the transmitted waveformnever varies in

amplitude, just frequency. Therefore, the total transmitted powermust remain constant regardless
of the level of modulation. It is thus seen that whateverenergy is conta ined in the side
frequencies has been obtained from the carrier.No additional energy is added during the
modulation process. The carrier in FM isnot redundant as in AM because its (the carrier's)
amplitude is dependent on the intelligence signal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The final power stage of a 20-kW amplifier uses a tube as the active element.

4.4. Noise Suppression

The most important advantage of FM over AM is the superior noisecharacteristics. You are
probably aware that static noise is rarely heard on FM,although it is quite common in AM
reception. You may be able to guess a reasonfor this improvement. The addition of noise to a
received signal causes a change inits amplitude. Since the amplitude changes in AM contain the
intelligence, anyattempt to get rid of the noise adversely affects the received signal. However, in
FM, the intelligence is not carried by amplitude changes but instead by frequencychanges. The
spikes of external noise picked up during transmission are clippedoff by a limiter circuit and/or
through the use of detector circuits that are insensitiveto amplitude changes.
Figure 6(a) shows the noise removal action of an FM limiter circuit, while inFigure 6(b) the
noise spike feeds right through to the speaker in an AM system. Theadvantage for FM is clearly
evident; in fact, you may think that the limiter removesall the effects of this noise spike. While it
is possible to clip the noise spike off, itstill causes an undesired phase shift and thus frequency
shift of the FM signal, andthis frequency shift cannot be removed.

*Limiter : stage in an FM receiver that removes any amplitude variations of the received
FM signal before it reaches the discriminator.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The noise signal frequency will be close to the frequency of the desired FMsignal due to
the selective effect of the tuned circuits in a receiver. In other words,if you are tuned to an FM
station at 96 MHz, the receiver's selectivity provides gainonly for frequencies near 96 MHz. The
noise that will affect this reception must,therefore, also be around 96 MHz because all other
frequencies will be greatlyattenuated. The effect of adding the desired and noise signals will give
a resultantsignal with a different phase angle than the desired FM signal alone. Therefore,
thenoise signal, even though it is clipped off in amplitude, will cause phase modulation(PM),
which indirectly causes undesired FM. The amount of frequency deviation(FM) caused by PM is

Fig. 68: FM, AM noise comparison.

Fig.69:Phase shift (</>) as a result of noise.

1
CNS Basic Train ing- Analog Co mmunication ECATC

δ =</J x f i (9)

where δ= frequency deviation

= phase shift (radians)
f i= frequency of intelligence signal

FM Noise Analysis

The phase shift caused by the noise signal results in a frequency deviation that ispredicted by
Equation (9). Consider the situation illustrated in Figure 7. Here thenoise signal is one- half the
desired signal amplitude, which provides a voltage S/Nratio of 2: 1. This is an intolerable
situation in AM but, as the following analysis willshow, is not so bad in FM.
Because the noise (N) and desired signal (S) are at different frequencies (butin the same
range, as dictated by a receiver's tuned circuits), the noise is shown asa rotating vector using the
S signal as a reference. The phase shift of the resultant(R) is maximum when R and N are at right
angles to one another. At this worst-casecondition

<P= sin–1 N/S = sin-1 1/2

= 30°

or 30°/(57.3° per radian) = 0.52 rad, or about t rad.If the intelligence frequency, f i, were known,
then the deviation (8) causedby this severe noise condition could now be calculated using
Equation (9). Since8 = <PX f i, the worst-case deviation occurs for the maximum intelligence
frequency.Assuming an f imaximum of 15 kHz, the absolute worst case 8 due to thissevere noise
signal is

8=δX f i = 0.5 X 15 kHz = 7 .5 kHz

In standard broadcast FM, the maximum modulating frequency is 15 kHz and themaximum
allowed deviation is 75 kHz above and below the carrier. Thus, a 75-kHzdeviation corresponds
to maximum modulating signal amplitude and full volume atthe receiver's output. The 7.5-kHz
worst-case deviation output due to the S/N= 2condition is

7.5 kHz/7 kHz = 1/10

and, therefore, the 2: 1 signal-to-noise ratio results in an output signal-to- noise ratioof 10:1. This
result assumes that the receiver's internal noise is negligible. Thus, FMis seen to exhibit a very
strong capability to nullify the effects of noise! In AM, a 2: 1signal-to-noise ratio at the input
essentially results in the same ratio at the output.
Thus, FM is seen to have an inherent noise reduction capability not possible with AM.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example

Determine the worst-case output SIN for a broadcast FM program that has amaximum
intelligence frequency of 5 kHz. The input SIN is 2.

Solution

The input S/N = 2 means that the worst-case deviation is about t rad (see the preceding
paragraphs). Therefore.
o = <f>Xf;
= 0.5 x 5 kHz = 2.5 kHz(9)

Because full volume in broadcast FM corresponds to a 75-kHz deviation, this 2.5-kHz worstcase
noise deviation means that the output S/N is

75 kHz /2.5 kHz=30

Example 8 shows that the inherent noise red uction capability of FM isimproved when the
maximum intelligence (modulating) frequency is reduced. A littlethought shows that this
capability can also be improved by increasing the maximumallowed frequency deviation from
the standard 75-k:Hz value. An increase inallowed deviation means that increased bandwidths
for each station would be necessary,however. In fact, many FM systems utilized as
communication links operatewith decreased bandwidths-narrowband FM systems. [t is typical
for them tooperate with a 10-kHz maximum deviation. The inherent noise reduction of
thesesystems is reduced by the lower allowed 8 but is somewhat offset by the lower
maximummodulating frequency of 3 kHz usually used for voice transmissions.

Example

Determine the worst-case output SIN for a narrowband FM receiver withOmax= JO kHz and a
maximum intelligence frequency of 3 kHz. The SIN inputis 3:1.

Solution

The worst-case phase shift (</>) due to the noise occurs when </>= sin -l(N/S).
1 </>= sin-1 1/3 = 19.5°, or 0.34 rad
and
o=</ > x fi
= 0.34 x 3 kHz = 1 kHz(9)

1
CNS Basic Train ing- Analog Co mmunication ECATC

The S/N output will be

10 kHz / 1 kHz = 10

and thus the input signal-to- noise ratio of 3 is transformed to IO or higher at the output.

Capture Effect

This inherent ability of FM to minimize the effect of undesired signals (noise in thepreceding
paragraphs) also applies to the reception of an undesired station operatingat the same or nearly
the same frequency as the desired station. This is known as thecapture effect. You may have
noticed when riding in a car that an FM station is suddenlyreplaced by a different one. You may
even find that the receiver alternatesabruptly back and forth between the two. This occurs
because the two stations arepresenting a variable signal as you drive. The capture effect causes
the receiver tolock on the stronger signal by suppressing the weaker but can fluctuate back and
forth when the two are nearly equal. When they are not nearly equal, however, theinherent FM
noise suppression action is very effective in preventing the interferenceof an unwanted (but
weaker) station. The weaker station is suppressed just as noisewas in the preceding noise
discussion. FM receivers typically have a capture ratioof 1 dB-this means suppression of a 1-dB
(or more) weaker station is accomplished.
In AM, it is not uncommon to hear two separate broadcasts at the same time,but this is
ce1tainly a rare occurrence with FM.The capture effect can also be illustrated by Figure 8.
Notice that the S/Nbefore and after demodulation for SSB and AM is linear. Assuming noiseless
demodulation schemes, SSB (and DSB) has the same S/N at the detector's input andoutput. The
degradation shown for AM is due to so much of the signal's powerbeing wasted in the redundant
carrier. FM systems with m1 greater than 1 show anactual improvement in S/N, as illustrated in
Examples 8 and 9. For example, considerm1 = 5 in Figure 8. When S/N before demodulation is
20, the SIN after demodulationis about 38-a significant improvement.Insight into the capture
effect is provided by consideration of the inflectionpoint (often termed threshold) shown in
Figure 8. Notice that a rapid degradation.

*Capture Effect:an FM receiverphenomenon that involves locking onto the stronger

of two received signals ofthe same frequency andsuppressing the weakersignal.
*Capture Ratio:the necessary difference(in dB) of signal strengthto allow suppression of
aweaker signal from astronger one.
*Threshold:in FM, the point whereS/N in the output rapidlydegrades as SIN ofreceived
signal isdegrading.
*Preemphasis: process in an FM transmitter that amplifieshigh-frequency more than
low-frequency audiosignals to reduce theeffect of noise.
*Deemphasis:process in an FM receiverthat reduces theamplitudes of highfrequency
audio signalsdown to their original values to counteract theeffect of the preemphasis
network in the transmitter.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig.70: S/N for basic modulation schemes.

inS/Nafter demodulation results when the noise approaches the same level as thedesired signal.
This threshold situation is noticeable when driving in a large city.The fluttering noise often heard
is caused when the FM signal is reflecting off variousstructures. The signal strength fluctuates
widely due to the additive or subtractiveeffects on the total received signal. The effect can cause
the output to blankout totally and resume at a rapid rate as the SIN before demodulation moves
backand forth through the threshold level.

Pre emphasis

The noise suppression ability of FM has been shown to decrease with higher
intelligencefrequencies. This is unfortunate because the higher intelligence frequenciestend to be
of lower amplitude than the low frequencies. Thus, a high-pitched violinnote that the human ear
may perceive as the same "sound" level as the crash of a bassdrum may have only half the
electrical amplitude as the low-frequency drum signal.In FM, half the amplitude means half the
deviation and, subsequently, half the noisereduction capab ility. To counteract this effect, almost
all FM transmissions providean artificial boost to the electrical amplitude of the higher
frequencies. This process is termed pre emphasis.
By definition, pre emphasis involves increasing the relative strength of thehigh-
frequency components of the audio signal before it is fed to the modulator. Thus,the relationship
between the high- frequency intelligence components and the noise isaltered. While the noise
remains the same, the desired signal strength is increased.A potential disadvantage, however, is
that the natural balance between highandlow- frequency tones at the receiver would be altered. A

1
CNS Basic Train ing- Analog Co mmunication ECATC

de emphasis circuit in the receiver, however, corrects this defect by reducing the high- frequency
audio by thesame amount as the pre emphasis circuit increased it, thus regaining the originaltonal
balance. In addition, the de emphasis network operates on both the highfrequencysignal and the
high- frequency noise; therefore, there is no change in theimproved signal-to- noise ratio. The
main reason for the pre emphasis network, then is to prevent the high- frequency components of
the transmitted intelligence frombeing degraded by noise that would otherwise have more effect
on the higher thanon the lower intelligence frequencies.
The de emphasis network is normally inserted between the detector and theaudio
amplifier in the receiver. This ensures that the audio frequencies are returnedto their original
relative level before amplification. The pre emphasis characteristic curve is flat up to 500 Hz, as
shown in Figure 9. From 500 to 15,000 Hz, there isa sharp increase in gain up to approximately
17 dB. The gain at these frequenciesis necessary to maintain the signal- to-noise ratio at high
audio frequencies. The frequency characteristic of the de emphasis network is directly opposite
to that of the pre emphasis network. The high- frequency response decreases in proportion to its
increase in the pre emphasis network. The characteristic curve of the de emphasis circuit should
be a mirror image of the pre emphasis characteristic curve. Figure 9shows the pre- and de
emphasis curves as used by standard FM broadcasts in theUnited States. As shown, the 3-dB
points occur at 2120 Hz, as predicted by the RCtime constant (7) of 75 μ,s used to generate them.

Fig. 71: Emphasis curves (T = 75 μ,s).

Figure 10(a) shows a typical pre emphasis circuit. The impedance to the audiovoltage is
mainly that of the parallel circuit of C and R1 because the effect of R2 issmall in comparison to
that of either C or R1 • Since capacitive reactance is inverselyproportional to frequency, audio
frequency increases cause the reactance of C todecrease. This decrease of Xc provides an easier
path for high frequencies as comparedto R.Thus, with an increase of audio frequency, there is an
increase in signalvoltage. The result is a larger voltage drop across R (the amplifier's input) atthe
higher frequencies and thus greater output.Figure 10(b) depicts a typical deemphasis network.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Note the physical positionof R and C in relation to the base of the transistor. As the frequency of
the audiosignal increases, the reactance of capacitor C decreases. The voltage division

Fig.72: Emphasis circuits.

BetweenRand C now provides a smaller drop across C. The audio voltage appliedto the base
decreases; therefore, a reverse of the preemphasis circuit is accomplished. For the signal to be
exactly the same as before preemphasis and deemphasis,the time constants of the two circuits
must be equal to each other.

Dolby System

The FCC has ruled that FM broadcast stations can, if desired, use a 25-μ.,stime constant.They
then must use the Dolby system, which works like preemphasis but in adynamic fashion. The
amount of preemphasis (and subsequent deemphasis in a Dolbyreceiver) varies depending on the
loudness level at any instant. Maximum noise reductionof this system is realized by listeners in
fringe areas where noise effects are mostdetrimental. Those people with regular 75-μ.,s, non-
Dolby receivers do not seem to beadversely affected but will usually turn down their treble
control somewhat.
The Dolby system used by FM transmitters is not concerned primarily withnoise reduction,
however. Standard 75-μ.,s pre emphasis provides the same amountof boost to both weak and
strong high- frequency signals. This created no real problemyears ago because few, if any, high
frequencies were strong anyway. The betterqualityrecordings of today do offer so me fairly
strong high frequencies. Theirstrength, combined with 75-μ.,spreemphasis, now causes excessive
bandwidth ofthe transmitted FM signal, and thus a station is forced to lower the strength of
allsignals.

*Dolby System :advanced noise reductionsystem in FM systems inwhich the preemphasis

Anddeemphasisnetworkswork in a dynamic manner.

1
CNS Basic Train ing- Analog Co mmunication ECATC

To counteract this effect, the transmitting station will artificially reduce thestrength of high-
frequency signals, resulting in a less-than-true signal reproduction(and less dynamic range) at the
receiver.
The Dolby solution to this problem is to provide varying degrees of highfrequencyboost, as
depicted in Figure 11. The Dolby receiver must reverse thesecharacteristics and thus requires
some relatively complex circuitry. Notice how theweak high frequencies are given a significantly
greater boost than the stronger ones.
The overall result of this system is a stronger received FM signal with greaterdynamic range as
compared to a broadcast signal, which unnaturally attenuates thehigh frequencies. The Dolby
system is also highly effective in minimizing highfrequency

Fig.73: Dolby dynamic preemphasis.

4.5. Direct FM Generation

The capacitance microphone system explained in Section 2 can be used to generateFM

directly. Recall that the capacitance of the microphone varies in step with thesound wave
striking it. Figure 1 showed that if the amplitude of sound striking it isincreased, the
amount of deviation of the oscillator's frequency is increased. If thefrequency of sound
waves is increased, the rate of the oscillator's deviation isincreased. This system is useful
in explaining the principles of an FM signal but isnot normally used to generate FM in
practical systems. It is not able to produceenough deviation as required in actual
applications.

1
CNS Basic Train ing- Analog Co mmunication ECATC

VaractorDiode

A varactor diode may be used to generate FM directly. All reverse-biased diodesexhibit a

junction capacitance that varies inversely with the amount of reverse bias.A diode that is
physically constructed to enhance this characteristic is termed avaractor diode. Figure 12
shows a schematic of a varactor diode modulator. Withno intelligence signal (Ei) applied,
the parallel combination of C1, L1, and D1 'scapacitance forms the resonant carrier
frequency. The diode D1 is effectively in parallelwith L1 and C1 because the - V cc
supply appears as a short circuit to the acsignal. The coupling capacitor, Cc, isolates the
de levels and intelligence signalwhile looking like a short to the high- frequency carrier.
When the intelligence

Fig. 74:Varactor diode modulator.

*VaractorDiode :diode with a small internalcapacitance that varies asa function of its
reversebias voltage.
*Reactance Modulator :amp I ifier designed so thatits input impedance has areactance
that varies as afunction of the amplitudeof the applied inputvoltage.
*Voltage-ControlledOscillator :designed so that its outputvoltage varies as afunction of
the amplitudeof the applied inputvoltageModulatingsignalinput.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig.75: Reactance modulator.

signal, Ei, is applied to the varactor diode, its reverse bias is varied, which causesthe
diode's junction capacitance to vary in step with Ei· The oscillator frequency
issubsequently varied as required for FM, and the FM signal is available at Q1 's
collector.
For simplicity, the de bias and oscillator feedback circuitry is not shown inFigure 12.
While the varactor diode modulator can be called a reactance modulator,the term is
usually applied to those in which an active device is made to look like avariable
reactance. The reactance modulator (see Figure 13) is a very popularmeans of FM
generation. The troubleshooting section at the end of this chapter providesdetails on the
operation and repair of a reactance modulator.

LIC VCO FM Generation

A voltage-controlled oscillator (VCO) produces an output frequency that is

directlyproportional to a control voltage level. The circuitry necessary to produce such
anoscillator with a high degree of linearity between control voltage and frequency
wasformerly prohibitive on a discrete component basis, but now that low-costmonolithic
LIC VCOs are available, they make FM generation extremely simple.
Figure 14 provides the specifications for the Philips Semicond uctors NE/SE566YCO.
The circuit shown at the end of the specifications provides a high-quality FMgenerator
with the modulating voltage applied to C2. The FM output can be taken atpin 4 (triangle

1
CNS Basic Train ing- Analog Co mmunication ECATC

wave) or pin 3 (square wave). Feeding either of these two outputs intoan LC tank circuit
resonant at the VCO center frequency (i.e., carrier) will subsequentlyprovide a standard
sinusoidal FM signal by the flywheel effect.

CrosbyModulator

Now that three practical methods of FM generation have been shown-varactordiode,

reactance modulator, and the VCO-it is time to consider the weakness ofthese methods.
Notice that in no case was a crystal oscillator used as the basicreference or carrier
frequency. The stability of the carrier frequency is slightlyco ntro11ed by the FCC, and
that stability is not attained by any of the methodsdescribed thus far. Because of the high
Q of crystal oscillators, it is not possible tofrequency- modulate them directly-their
frequency cannot be made to deviate sufficientlyto provide workable wideband FM
systems. It is possible to modulatedirectly a crystal oscillator in some narrowband
applications. If a crystal is modulatedto a deviation of :± 50 Hz around a 5-MHz center
frequency and both are multiplied by 100, a narrowband system with a 500-MHz carrier
± 5-kHz deviation
results. One method of circumventing this dilemma for wideband systems is to provide
some means of automatic frequency control (AFC) to correct any carrier driftby
comparing it to a reference crystal oscillator.
FM systems utilizing direct generation with AFC are caJled Crosby systems.A Crosby
direct FM transmitter for a standard broadcast station at 90 MHz isshown in Figure 15.
Notice that the reactance modulator starts at an initial centerfrequency of 5 MHz and has
a maximum deviation of ± 4.167 kHz. This is a typicalsituation because reactance
modulators cannot provide deviations exceedingabout ±5 kHz and still offer a high
degree of linearity (i.e., !:lf directly proportionalto the modulating voltage amplitude).
Consequently, frequency multipliers are utilizedto provide a Xl 8 multiplication up to a
carrier frequency of 90 MHz (18 X 5 MHz)with a ±75-kHz (18 X 4. 167 kHz) deviation.
Notice that both the carrier and deviationare multiplied by the multiplier.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Frequency multiplication is normally obtained in steps of X2 or X3 (doublersor triplers).

The principle involved is to feed a frequency rich in harmonic distortion(i.e., from a class
C amplifier) into an LC tank circuit tuned to two or three

Fig.76: Crosby direct FM transmitter.

Fig. 77: Frequency multiplication (doubler).

times the input frequency. The harmonic is then the only significant output, asillustrated
in Figure 16.

1
CNS Basic Train ing- Analog Co mmunication ECATC

After the Xl8 multiplication (3 X 2 X 3) shown in Figure 15, the FMexciter function is
complete. The term exciter is often used to denote the circuitrythat generates the
modulated signal. The excited output goes to the poweramplifiers for transmission and to
the frequency stabilization system. The purposeof this system is to provide a control
voltage to the reactance modulatorwhenever it drifts from its desired 5-MHz value. The
control (AFC) voltage thenvaries the reactance of the primary 5-MHz oscillator slightly
to bring it back onfrequency.
The mixer in Figure 15 has the 90-MHz carrier and 88-MHz crystal oscillatorsignal as
inputs. The mixer output accepts only the difference component of 2 MHz, which is fed
to the discriminator. A discriminator is the opposite of aVCO, because it provides a de
level output based on the frequency input. The discriminatoroutput in Figure 15 wi11 be
zero if it has an input of exactly 2 MHz,which occurs when the transmitter is at precisely
90 MHz. Any carrier drift up ordown causes the discriminator output to go positive or
negative, resulting in theappropriate primary oscillator readjustment.

4.6. Indirect FM Generation

If the phase of a crystal oscillator's output is varied, phase modulation (PM) will result.
As discussed previously, changing the phase of a signal indirectly causes its frequency to
be changed. We thus find that direct modulation of a crystal is possible via PM, which
indirectly creates FM. This indirect method of FM generation is usually referred to as the
Armstrong type, after its originator, E. H. Armstrong. It permits modulation of a stable
crystal oscillator without the need for the cumbersome AFC circuitry and also provides
carrier accuracies identical to the crystal accuracy,as opposed to the slight degradation of
the Crosby system's accuracy.
A simple Armstrong modulator is depicted in Figure 17. The JFET is biased in the ohmic
region by keeping VDS low. In that way it presents a resistance from drain to source that
is made variable by the gate voltage (the modulating signal). In the ohmic region, the
drain-to-source resistance for a JFET transistor behaves like a voltage-controlled
resistance (a variable resistor). The resistance value is con-trolled by the gate voltage (V
cs), where a change in the gate voltage will create a

*Exciter :stages necessary in a transmitter to create the modulated signal before

subsequent amplification .
*Discriminator :stage in an FM receiver that creates an output de level that varies
as a funct ion of its input frequency
*Armstrong Modulator :FM transmitter that uses a phase modulator that feeds the
intelligencesignal through a low-pass filter integrator network to convert PM to FM

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 78:Indirect FM via PM (Armstrong modulator).

change in the drain-to-source resistance. Notice that the modulating signal is first given
the standard preemphasis and then applied to a frequency-conecting network. This
network is a low-pass RC circuit (an integrator) that makes the audio output amplitude
inversely proportional to its frequency. This is necessary because in phase modulation,
the frequency deviation created is not only proportional to modulating signal amplitude
(as desired for FM) but also to the modulating signal frequency (undesired for FM). Thus,
in PM if a 1-V, 1-k.Hz modulating signal caused a 100-Hz deviation, a 1-V, 2-kHz signal
would cause a 200-Hz deviation instead of the same deviation of 100 Hz if that signal
were applied to the l/f network. In summary, the Armstrong modulator of Figure 17
indirectly generates FM by changing the phase of a crystal oscillator's output. That phase
change is accomplished by varying the phase angle of an RC network (C1 and the JFET's
resistance), in step with the frequency-corrected modulating signal.
The indirectly created FM is not capable of much frequency deviation. A typical
deviation is 50 Hz out of 1 MHz (50 ppm). Thus, even with a X 90 frequency multi-
plication, a 90-MHz station would have a deviation of 90 X 50 Hz = 4.5 kHz. This may
be adequate for narrowband communication FM but falls far short of the 75-kHz
deviation required for broadcast FM. A complete Armstrong FM system providing a 75-
kHz deviation is shown in Figure 18. It uses a balanced modulator and 90° phase shifter
to phase- modulate a crystal oscillator. Sufficient deviation is obtained by a combination
of multipliers and mixing. The X 81 multipliers (3 X 3 X 3 X 3) raise the initial 400-kHz
± 14.47-Hz signal to 32.4 MHz ± 1172 Hz. The carrier and deviation are multiplied by 81
. Applying this signal to the mixer, which also has a crystal oscillator signal input of
33.81 MHz, provides an output component (among others) of 33.81 MHz - (32.4 MHz ±
1172 Hz), or 1.41 MHz ± 1172 Hz.
Notice that the mixer output changes the center frequency without changing the

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 79: Wideband Armstrong FM.

Fig. 80: The pump chain for the wideband Armstrong FM system.

deviation. Following the mixer, the X 64 multipliers accept only the mixer difference
output component of 1.41 MHz ± 1172 Hz and raise that to (64 X 1.41 MHz) ± (64 X
1172 Hz), or the desired 90.2 MHz ± 75 kHz. The electronic circuitry used to increase the
operating frequency of a transmitter up to a specified value is called the pump chain. A
block diagram of the pump chain for the wideba nd Armstrong FM system is shown in
Figure 19.

1
CNS Basic Train ing- Analog Co mmunication ECATC

4.7. PHASE LOCKED-LOOP

FM TRANSMITTER

Block Diagram

The block diagram shown in Figure 20 provides a very practical way to fabricate an FM
transmitter. The amplified audio signal is used to frequency- modulate a crystal oscillator.
The crystal frequency is pulled slightly by the variable capacitance exhibited by the
varactor diode. The approximate ±200-Hz deviation possible in this fashion is adequate
for narrowband systems. The FM output from the crysta l oscillator is then divided by 2
and applied as one of the inputs to the phase detector of a phase- locked- loop (PLL)
system. As indicated in Figure 20, the other input to the phase detector is the same, and
its output is therefore (in this case) the original audio signal.

Fig. 81: PLL FM transmitter block diagram.

*Pump Chain: electronic circuitry used to increase the operating frequency up to a

specified value

1
CNS Basic Train ing- Analog Co mmunication ECATC

The input control signal to the VCO is therefore the same audio signal, and its
output will be its free-running value of 125 MHz ±5 kHz, which is set up to be
exactly 50 times the 2.5-MHz value of the divided-by-2 crystal frequency of 5 MHz. The
FM output signal from the VCO is given power amplification and then driven into the
transmitting antenna. This output is also sampled by a +50 network, which provides the
other FM signal input to the phase detector. The PLL system effectively provides the
required X50 multiplication but, more important, provides the transmitter's required
frequency stability. Any drift in the VCO center frequency causes an input to the phase
detector (input 2 in Figure 20) that is slightly differentfrom the exact 2.5-MHz crystal
reference value. The phase detector output therefore develops an error signal that
c01Tects the VCO center frequency output back to exactly 125 MHz. This dynamic
action of the phase detectorNCO and feedback path is the basis of a PLL.

1
CNS Basic Train ing- Analog Co mmunication ECATC

4.8. Stereo Fm

The advent of stereo records and tapes and the associated high- fidelity playback
equipment in the 1950s led to the development of stereo FM transmissions as authorized
by the FCC in 1961. Stereo systems involve generating two separate signals, as from the
left and right sides of a concert hall performance. When p layed back on leftand right
speakers, the listener gains greater spatial dimension or directivity.
A stereo radio broadcast requires that two separate 30-Hz to 15-kHz signals be
used to modulate the carrier so that the receiver can extract the left and right channel
information and an1plify them separately into their respective speakers. In essence, then,
the amount of information to be transmitted is doubled in a stereo broadcast. Hartley's
law tells us that either the bandwidth or time of transmission must therefore be doubled,
but this is not practical. The problem was solved by making more efficient use of the
available bandwidth (200 kHz) by frequency multiplexing the two required modulating
signals. Multiplex operation is the simultaneous transmission of two or more signals on
one earlier.

Modulating Signal

The system approved by the FCC is compatible because a stereo broadcast received by a
normal FM receiver will provide an output equal to the sum of the left plus right channels
(L + R), while a stereo receiver can provide separate left and right channel signals. The
stereo transmitter has a modulating signal, as shown in Figure 23. Notice that the sum of
the L + R modulating signal extends from 30 Hz to 15 kHz as does the full audio signal
used to modulate the carrier in standard FM broadcasts. However, a s ignal co1responding
to the left channel minus right channel (L - R) extends from 23 to 53 kHz. In addition, a
19-kHz pilot subcarrier is included in the composite stereo modulating signal.

*Frequency Multiplexing: process of combining signals that are at slightly

different frequencies to allow transmission over a single medium .
*Multiplex Operation: simultaneous transmission of two or more signals in a
single medium .
*Frequency-Division M ultiplexing: simultaneous transmission of two or more signals
onone carrier, each on its own separate frequency range, also called frequency
multiplexing
*Matrix Network :adds and/or subtracts and/or inverts electrical signals

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 82: Composite modulating signals.

The reasons for the peculiar arrangement of the stereo modulating signal are
beyond the scope of this chapter. Suffice it to say that two different signals (L + R and L -
R) are used to modulate the carrier. The signal is an example of frequency division
multiplexing because two different signals are multiplexed together by having them exist
in two different frequency ranges.

FM STEREO GENERATION

The block diagram in Figure 24 shows the method whereby the composite modulating
signal is generated and applied to the FM modulator for subsequent transmission. The left
and right channels are picked up by their respective microphones and individually
preemphasized. They are then applied to a matrix network that inverts the right channel,
giving a - R signal, and then combines (adds) Land R to provide an (L + R) signal and
also combines Land -R to provide the (L - R) signal. The two outputs are still 30-Hz to
15-kHz audio signals at this point. The (L - R) signal and a 38-kHz carrier signal are then
applied to a balanced modulator that suppresses the carrier but provides a double-
sideband (DSB) signal at its output. The upper and lower sidebands extend from 30 Hz to
15 kHz above and below the suppressed 38-kHz carrier and therefore range from 23 kHz
(38 kHz - 15 kHz) up to 53 kHz (38 kHz + 15 kHz). Thus, the (L - R) signal has been
translated from audio up to a higher frequency to keep it separate from the 30-Hz to 15-
kHz (L + R) signal. The (L + R) signal is given a slight delay so that both signals are
applied to the FM modulator in time phase due to the slight delay encountered by the (L -
R) signal in the balanced modulator. The 19-kHz master oscillator in Figure 24 is applied
directly to the FM modulator and also doubled in frequency, to 38 kHz, for the balanced
modulator carrier input.
Stereo FM is more prone to noise than are monophonic broadca<;ts. The (L - R)signal is
weaker than the (L + R) signal, as shown in Figure 23. The (L - R) signal isalso at a
higher modulating frequency (23 to 53 kHz), and both of these effects cause

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 83: Stereo FM transmitter.

poorer noise performance. The net result to the receiver is an SIN of about 20 dB less
than the monophonic signal. Because of this, some receivers have a mono/stereo switch
so that a noisy (weak) stereo signal can be changed to monophonic for improved
reception. A stereo signal received by a monophonic receiver is only about 1 dB worse
(SIN) than an equivalent monophonic broadcast because of the presence of the 19-kHz
pilot carrier.

4.9. FM Transmissions

FM is used in five major categories:

l .Noncommercial broadcast at 88 to 90 MHz
2. Commercial broadcast with 200-kHz channel bandwidths from 90 to 108 MHz
3. Television audio signals with 50-kHz channel bandwidths at 54 to 88 MHz, 174 to 216
MHz, and 470 to 806 MHz
4. Narrowband public service channels from 108 to 174 MHz and in excess of 806MHz
5. Narrowband amateur radio channels at 29.6 MHz, 52 to 53 MHz, 144 to 147.99 MHz,
440 to 450 MHz, and in excess of 902 MHz

The output powers range from milliwatt levels for the amateurs up to 100 kW for
broadcast FM. Note that FM is not used at frequencies below about 30 MHz because of
the phase distortion introduced to FM signals by the earth's ionosphere at these
frequencies. Frequencies above 30 MHz are transmitted line-of-sight and are not
significantly affected by the ionosphere.

1
CNS Basic Train ing- Analog Co mmunication ECATC

(a) The 86 l OOC digital communications analyzer with jitter analysis offers
breakthrough speed, accuracy, and
affordability. (b) The MT88208 radio communications analyzer was designed to support
the test needs of the manufacturing, R&D, and maintenance markets.

The limited range (n01mally 70 to 80 mi) for FM transmission is due to the earth's
curvature.
Another advantage that FM has over SSB and AM, other than superior noise
performance, is the fact that low- level modulation can be used with subsequent highly
efficient class C power amplifiers. Since the FM waveform does not vary in amplitude,
the intelligence is not lost by class C power amplification as it is for AM and SSB. Recall
that a class C amplifier tends to provide a constant output amplitude due to the LC tank
circuit flywheel effect. Thus, there is no need for high-power audio amplifiers in an FM
transmitter and, more important, all the power amplification takes place at about 90
percent efficiency (class C), as compared to a maximum of about 70 percent for linear
power amplifiers.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fm Demodulation

4.10. Block Diagram

The basic FM receiver uses the superheterodyne principle. In block diagram form, ithas
many similarities to the receivers you may have covered in previous chapters. InFigure l ,
the only apparent differences are the use of the word discriminator in placeof detector,
the addition of a deemphasis network, and the fact that AGC may or maynot be used as
indicated by the dashed lines.The discriminator extracts the intelligence from the high-
frequency carrierand can also be called the detector, as in AM receivers. By definition,
however, adiscriminator is a device in which amplitude variations are derived in response
tofrequency or phase variations, and it is the prefeITed term for describing an
FMdemodulator.
The deemphasis network following demodulation is required to bring thehigh- frequency
intelligence back to the proper amplitude relationship with the lowerfrequencies. Recall
that the high frequencies were preemphasized at the transmitterto provide them with
greater noise immunity.
The fact that AGC is optional in an FM receiver may be surprising to you.From your
understanding of AM receivers, you know that AGC is essential to theirsatisfactory
operation. However, the use of limiters in FM receivers essentially providesan AGC
function, as will be explained in Section 3. Many older FM receiversalso included an
automatic frequency control (AFC) function. This is a circuit thatprovides a slight
automatic control over the local oscillator circuit. Jt compensatesfor drift in LO
frequency that would otherwise cause a station to become detuned.It was necessary
because it had not yet been figured out how to make an economicalLC oscillator at 100
MHz with sufficient frequency stability. The AFC systemis not needed in new
designs.The mixer, local oscillator, and IF amplifiers are basically similar to
thosediscussed for AM receivers and do not require further elaboration.It should benoted
that higher frequencies are usually involved, however, because of the fact thatFM
systems generally function at higher frequencies.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 84: FMreceiverblock diagram.

The universally standard IFfrequency for FM is 10.7 MHz, as opposed to 455 kHz for
AM. Because of significantdifferences in all the other portions of the block diagram
shown in Figure 1,they are discussed in the following sections.

4.11. RF Amplifiers

Broadcast AM receivers normally operate quite satisfactorily without any RFamplifier.

This is rarely the case with FM receivers, however, except for frequenciesin excess of
1000 MHz (1 GHz), when it becomes preferable to omit it. The essenceof the problem is
that FM receivers can function with weaker received signals thanAM or SSB receivers
because of their inherent noise reduction capability. Thismeans that FM receivers can
function with a lower sensitivity, and are called upon todeal with input signals of 1 μ, V
or less as compared with perhaps a 30-μ, V minimuminput for AM. If a 1-μ, V signal is
fed directly into a mi xer, the inherently high noisefactor of an active mixer stage
destroys the intelligibility of the 1-μ, V signal. Therefore,it is necessary to amplify the 1-
μ, V level in an RF stage to get the signal up to atleast 10 to 20 μ, V before mixing
occurs. The FM system can tolerate 1 μ, V of noisefrom a mixer on a 20-μ, V signal but
obviously cannot cope with 1 μ, V of noise witha 1-μ, V signal.

1
CNS Basic Train ing- Analog Co mmunication ECATC

This reasoning also explains the abandonment of RF stages for the everincreasingFM
systems at the 1-GHz-and-above region. At these frequencies, transistornoise is
increasing while gain is decreasing. The frequency is reached whereit is advantageous to
feed the incoming FM signal directly into a diode mixer tostep it down immediately to a
lower frequency for subsequent amplification. Diode(passive) mixers are less noisy than
active mixers.
Of course, the use of an RF amplifier reduces the image frequency problem.Another
benefit is the reduction in local oscillator reradiation effects. Without anRF amp, the local
oscillator signal can get coupled back more easily into thereceiving antenna and transmit
interference.

FET RF Amplifiers

Almost all RF amps used in quality FM receivers utilize FETs as the active element.You
may think that this is done because of their high input impedance, but this is notthe
reason. In fact, their input impedance at the high frequency of FM signals isgreatly
reduced because of their input capacitance. The fact that FETs do not offerany significant
impedance advantage over other devices at high frequencies is not adeterrent, however,
because the impedance that an RF stage works from (theantenna) is only several hundred
ohms or less anyway.The major advantage is that FETs have an input/output square- law
relationshipwhile vacuum tubes have a ~-power relationship and BJTs have a diode-
typeexponential characteristic. A square- law device has an output signal at the
inputfrequency and a smaller distortion component at two times the input
frequency,whereas the other devices mentioned have many more distortion components,
with some of them occurring at frequencies close to the desired signal.
*Local OscillatorReradiation:undesired radiationof thelocal oscillator
signalthrough a receiver'santenna.
*Cross-Modulation:form of distortion resultingin overdriven mixer stages.
*IntermodulationDistortion:undesired mixing of twosignals in a receiverresulting
in an outputfrequency componentequal to that of thedesired signal.
Dynamic Range :decibel difference betweenthe largest tolerablereceiver input level
and itssensitivity.

1
CNS Basic Train ing- Analog Co mmunication ECATC

The use of an FET at the critical small signal level in a receiver means that the
devicedistortion components are filtered out easily by its tuned circuits because theclosest
distortion component is two times the frequency of the desired signal.This becomes an
extreme factor when you tune to a weak station that has a verystrong adjacent signal. If
the high- level adjacent signal gets through the inputtuned circuit, even though greatly
attenuated, it would probably generate distortioncomponents at the desired signal
frequency by a nonsquare- law device, andthe result is audible noise in the speaker output.
This form of receiver noise iscalled cross- modulation. This is similar to intermodulation
distortion, whichis characterized by the mixing of two undesired signals, resulting in an
outputcomponent that is equal to the desired signal's frequency. The possibility
ofintermodulation distortion is also greatly minimized by the use of FET RF
amplifiers.

MOSFET RF Amplifie rs

A dual- gate, common-source MOSFET RF amplifier is shown in Figure 2. Theuse of a

dual- gate device allows a convenient isolated input for an AOC level tocontrol device
gain. The MOSFETs also offer the advantage of increaseddynamic range over JFETs.
That is, a wider range of input signal can be toleratedby the MOSFET while still offering
the desired square- law input/output relationship.A similar arrangement is often utilized in
mixers because the extra gate al lowsfor a convenient injection point for the local
oscillator signal. Theaccompanying chart in Figure 2 provides component values for
operation at 100-MHz and 400-MHz center frequencies. The antenna input signal is
coupled into

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig.85: MOSFET RF amplifier

gate 1 via the coupling/tuning network comprised of C1, Li, and C2 . The outputsignal is
taken at the drain, which is coupled to the next stage by the L-i_-CrC4combination. The
bypass capacitor C8 next to !Jiand the radio- frequency choke(RFC) ensure that the signal
frequency is not applied to the de power supply. TheRFC acts as an open to the signal
while appearing as a short to de, and the bypasscapacitor acts in the inverse fashion.
These precautions are necessary to RF frequenciesbecause while power supply
impedance is very low at low frequenciesand de, it looks like a high impedance to RF and
can cause appreciable signalpower loss. The bypass capacitor from gate 2 to ground
provides a short to anyhigh- frequency signal that may get to that point. It is necessary to
maintain thebias stability set up by R1 and R2. The MFE 3007 MOSFET used in this
circuitprovides a minimum power gain of 18 dB at 200 MHz.

4.12. Limiters

A limiter is a circuit whose output is a constant amplitude for all inputs above acritical
value. Its function in an FM receiver is to remove any residual (unwanted)amplitude
modulation and the amplitude variations due to noise. Both of these variationswould have
an undesirable effect if carried through to the speaker. In addition,the limiting function
also provides AGC action because signals from the critical minimum value up to some
maximum value provide a constant input level to the detector.By definitio n, the
discriminator (detector) ideally would not respond toamplitude variations anyway

1
CNS Basic Train ing- Analog Co mmunication ECATC

because the information is contained in the amount offrequency deviation and the rate at
which it deviates back and forth around its centerfrequency.
A transistor limiter is shown in Figure 3. Notice the dropping resistor, Re,which
limits the de collector supply voltage. This provides a low de collectorvoltage, which
makes this stage very easily overdriven. This is the desired result.

Fig.86: Transistor limiting circuit.

Fig.87: Limiter input/output and flywheel effects.

*Sensitivity :minimum input RF signalto a receiver that isrequiredto produce

aspecified audio signal atits output.
*Quieting Voltage :the minimum FM receiverinput signal that beginsthe limiting
process.
*Threshold Voltage :another term for quietingvoltage.
*Limiting Knee Voltage :another term for quietingvoltage.

1
CNS Basic Train ing- Analog Co mmunication ECATC

As soon as the input is large enough to cause clipping at both extremes of

collectorcurrent, the critical limiting voltage has been attained and limiting action
hasstarted.The input/output characteristic for the limiter is shown in Figure 4, and itshows
the desired clipping action and the effects of feeding the limited (clipped)signal into an
LC tank circuit tuned to the signal's center frequency. The natural flywheeleffect of the
tank removes all frequencies not near the center frequency andthus provides a sinusoidal
output signal as shown. The omission of an LC circuit atthe limiter output is desirable for
some demodulator circuits. The quadrature detector(Section 4) uses the square-wave- like
waveform that results.

Limiting and sensitivity

A limiter, such as the one shown in Figure 3, requires about l V of signal to

beginlimiting. Much amplification of the received signal is therefore needed prior
tolimiting, which explains its position following the IF stages. When enough
signalatTives at the receiver to start limiting action, the set quiets, which means
thatbackground noise disappears. The sensitivity of an FM receiver is defined interms of
how much input signal is required to produce a specific level of quieting,normally 30
dB.This means that a good-quality receiver with a rated 1.5-μ, V sensitivitywill have
background noise 30 dB down from the desired input signal thathas a 1.5-μ, V level.
The minimum required voltage for limiting is called the quieting, threshold,or limiting
knee voltage. The limiter then provides a constant-amplitude output upto some maximum
value that prescribes the limiting range. Going above the maximumvalue results either in
a reduced and/or a distorted output. It is possible that asingle-stage limiter will not allow
for adequate range, thereby requiring a doublelimiter or the development of AGC control
on the RF and IF amplifiers to minimizethe possible limiter input range.It is most
common for today's FM receivers to use IC IF amplification. Inthese cases, the ICs ha ve
a built- in limiting action of very high quality (i.e., widedynamic range). Section 7
provides an example of these ICs.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 1

A certain FM receiver provides a voltage gain of 200,000 (106 dB) prior to itslimiter. The
limiter's quieting voltage is 200 mV. Determine the receiver's
sensitivity.

Solution

To reach quieting, the input must be

200mV/200,000= lμV
The receiver's sensitivity is therefore 1μ, V.

4.13. Discriminators

The FM discriminator (detector) extracts the intelligence that has been modulatedonto the
carrier via frequency variations. It shou1d provide an inte1ligence signa1whose
amplitude is dependent on instantaneous can- ier frequency deviation andwhose frequency
is dependent on the carrier's rate of frequency deviation. A desiredoutput amplitude
versus input frequency characteristic for a broadcast FM discriminatoris provided in
Figure 5. Notice that the response is linear in the allowed area offrequency deviation and
that the output amp1itude is directly proportiona1 to carrierfrequency deviation. Keep in
mind, however, that FM detection takes place followingthe IF amplifiers, which means
that the ±75-kHz deviation is intact but that carrierfrequency translation (usually to 10.7
MHz) has occurred.

Fig. 88: FM discriminator characteristic.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Slope Detector

The easiest FM discriminator to understand is the slope detector in Figure 6. The LCtank
circuit that follows the IF amplifiers and limiter is detuned from the carrier frequencyso
that fcfalls in the middle of the most linear region of the response curve.

Fig. 89: Slope detection.

When the FM signal rises in frequency above Jc, the output amplitude increases
whiledeviations be1ow fccause a smaller output. The slope detector thereby changes
FMinto AM, and a simple diode detector then recovers the intelligence contained in
theAM waveform's envelope. In an emergency, an AM receiver can be used to
receiveFM by detuning the tank circuit feeding the diode detector. Slope detection is
notwidely used in FM receivers because the slope characteristic of a tank circuit is
notvery linear, especiaJJyfor the large- frequency deviations of wideband FM.

1
CNS Basic Train ing- Analog Co mmunication ECATC

FOSTER-SEELY Discriminator

The two classical means of FM detection are the Foster-Seely discriminator and theratio
detector. While their once widespread use is now diminishing because of newtechniques
afforded by ICs, they remain a popular means of discrimination using aminimum of
circuitry. A typica1 Foster-Seelydiscriminator circuit is shown inFigure 7. In it, the two
tank circuits [L1 C1 and Cl.ti+ L:3)C2] are tuned exactly to thecarrier frequency.
Capacitors Cc, C4, and C5 are shorts to the carrier frequency. Thefollowing analysis
applies to an unmodulatedcarrier input:1. The carrier voltage e1 appears directly across
L4 because Cc and C4 are shortsto the carrier frequency.

Fig. 90: Foster- See ly discriminator.

2. The voltage esacross the transformer secondary (Li_ in series with L3) is 180°out of
phase with e1 by transformer action, as shown in Figure 8(a). The circulating0i,L3C2
tank current, i.50 is in phase with esbecause the tank is resonant.
3. The current is, flowing through inductance LiLJ, produces a voltage drop thatlags i.~by
90°. The individual components of this voltage, e2 and e3, are thusdisplaced by 90° from
is, as shown in Figure 8(a), and are 180° out of phasewith each other because they are the
voltage from the ends of a center-tappedwinding.
4. The voltage e4 applied to the diode D1, C3, and R1 network will be the vectorsum of
e1 and e2 [Figure 8(a)]. Similarly, the voltage esis the sum of e1 and e3•The magnitude
of e6 is proportional to e4 while e7 is proportional to e5.
5. The output voltage, e8, is equal to the sum of e6 and e7 and is zero because

1
CNS Basic Train ing- Analog Co mmunication ECATC

the diodes D1 and D2 will be conducting current equally (because e4 = es)but in opposite
directions through the R 1 C3 and R2C4 networks.The discriminator output is zero with
no modulation (zero frequency deviation),as is desired. The following discussion now
considers circuit action at someinstant when the input signal e1 is above the carrier
frequency. The phasor diagramof Figure 8(b) is used to illustrate this condition:

Fig. 91: Discriminator phase relations.

l. Voltages e1 and esare as before, but es now sees an inductive reactance becausethe tank
circuit is above resonance. Therefore, the circulating tankcurrent, is, lags es.
2. The voltages e2 and e3 must remain 90° out of phase with is, as shown inFigure 8(b).
The new vector sums of e2 + e1 and e3 + e1 are no longer equaJ,so e4 causes a heavier
conduction of D1 than exists for D2.
3. The output, e8, which is the sum of e6 and e7, will go positive because thecurrent
down through R1 C3 is greater than the current up through R2C4 (e4 isgreater than e5) .
The output for frequencies above resonance Uc) is therefore positive, while the
phaserdiagram in Figure 8( c) shows that at frequencies below resonance the outputgoes
negative. The amount of output is determined by the amount of frequencydeviation, while
the frequency of the output is determined by the rate at which theFM input signal varies
around its carrier or center value.

Ratio Detector

While the Foster-Seely discriminator just described offers excellent linear responseto
wideband FM signaJs, it also responds to any undesired input amplitude variations.The

1
CNS Basic Train ing- Analog Co mmunication ECATC

ratio detector does not respond to amplitude variations and thereby minimizesthe
required limiting before detection.
The ratio detector, shown in Figure 9, is a circuit designed to respond only tofrequency
changes of the input signal. Amplitude changes in the input have noeffect upon the
output. The input circuit of the ratio detector is identical to that ofthe Foster-
Seelydiscriminator circuit. The most immediately obvious difference isthe reversal of one
of the diodes.
The ratio detector circuit operation is similar to the Foster- Seely. A
detailedanalysis will therefore not be given. Notice the large electrolytic capacitor,
C5,across the R1- R2 combination. This maintains a constant voltage that is equal to
thepeak voltage across the diode input. This feature eliminates variations in the FMsignal,
thus providing amplitude limiting. The sudden changes in the input signa l'samplitude are
suppressed by the large capacitor. The Foster- Seely discriminatordoes not provide
amplitude limiting. The voltage Esis
Es = e1 + e2

Fig. 92: Ratio detector.

When f i = f c, e1 = e2 and hence the desired zero output occurs. When f in >fc,e1 >e2 ,
andwhen fin <ft., e1 < e2. The desired frequency-dependent output characteristicresults.
The component values shown in Figure 9 are typical for a 10.7-MHz IF FMinput signal.
The output level of the ratio detector is one- half that for the Foster-Seelycircuit.
Quadrature Detector

The Foster-Seely and ratio detector circuits do not lend themselves to integration ona
single chip due to the transformer required. This has led to increased usage of

1
CNS Basic Train ing- Analog Co mmunication ECATC

thequadrature detector and phase- locked loop (PLL). The PLL is introduced in the
nextsection.
Quadrature detectors derive their name from use of the FM signal in phaseand 90°
out of phase. The two signals are said to be in quadrature-at a90° angle. The circuit in
Figure 10 shows an FM quadrature detector using anexclusive-OR gate. The limited IF
output is applied directly to one input and thephase-shifted signal to the other. Notice that
this circuit uses the limjtedsignalthat has not been changed back to a sine wave. The L, C,
and R values used atthe circuit's input are chosen to provide a 90° phase shift at the
carrier frequencyto the signal 2 input. The signal 2 input is a sine wave due to the LC
circuiteffects. The upward and downward frequency deviation of the FM signal resultsin
a corresponding higher or lower phase shift. With one input to the gate shifted,the gate
output will be a series of pulses with a width proportional to the phasedifference. The
low-pass RC filter at the gate output sums the output, giving anaverage value that is the
intelligence signal. The gate output for three differentphase conditions is shown at Figure
93(b). The RC circuit output level for eachcase is shown with dashed lines. This
corresponds to the inte11igence level atthose particular conditions.
An analog quadrature detector is possible using a differential
amplifierconfiguration, as shown in Figure 11. A limited FM signal switches the
transistorcurrent source (Q1) of the differential pair Q2 + Q3. L1 and C2 should be
resonantat the IF frequency. The L1-C2-C1 combination causes the desired
frequencydependentphase shift between the two signals applied to Q2 and Q1.
Theconduction through Q3 depends on the coincident phase relationships of thesetwo
signals. The pulses generated at Q3's collector are summed by theR1-C3 low-pass filter,
and the resulting intelligence signal is taken at Q4'semitter. R2 is adjusted to yield the
desired zero- volt output when an undeviatingFM carrier is the circuit's input signal. The
popular 3089 LIC shown in Section 7 uses the analog quadratur e detectiontechnique. It
provides an excellent total harmonic distortion (THD) specificationof 0.1 percent
(typically) for a 10.7-MHz IF and ±75-kHz deviation.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig 93: Quadrature detection.

Fig 94: Analog quadrature detector.

*Quadrature:signalsat a 90° angle.

1
CNS Basic Train ing- Analog Co mmunication ECATC

4.14. Phase-Locked Loop

The phase- locked loop (PLL) has become increasingly popular as a means of
FMdemodulation in recent years. It eliminates the need for the intricate coil adjustments
of thepreviously discussed discriminators and has many other uses in the field of
electronics. Ltis an example of an old idea, originated in 1932, that was given a new life
by integratedcircuit technology. Prior to its availability in a single IC package in 1970, its
complexity indiscrete circuitry form made it economically unfeasible for most
applications.
The PLL is an electronic feedback control system as represented by the blockdiagram in
Figure 12. This input is to the phase comparator, or phase detector asit is also called. The
VCO within the PLL generates the other signal applied to thecomparator.
The comparator compares the input signal and the output of the VCO anddevelops an
error signal proportional to the difference between the two. This errorsignal drives the
VCO to change frequency so that the error is reduced to zero. If theVCO frequency
equals the input frequency, the PLL has achieved lock and the controlvoltage will be
constant for as long as the PLL input frequency remains constant.

*Phase-Locked Loop:closed- loop control systemthat uses negativefeedback to

maintainconstant output frequency.
*Phase Comparator:circuit that provides anoutput proportional to thephase difference
of twoinputs.
*Phase Detector:another term for phasecomparator.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 95: PLL block diagram.

PLL Capture and Lock

If the VCO starts to change frequency, it is in the capture state. It then continues tochange
frequency until its output is the same frequency as the input. At that point,the PLL is
locked; the VCO frequency now equals that of the input s ignal. The PLLhas three
possible states of operation:
1. Free-running
2. Capture
3. Locked or tracking
If the input and VCO frequency are too far apart, the PLL free-runs at the
nominalYCO frequency, which is determined by an external timing capacitor. This is not
anormally used mode of operation. If the VCO and input frequency are close enough,the
capture process begins and continues until the locked condition is reached. Oncetracking
(lock) begins, the VCO can remain locked over a wider input-frequencyrangevariation
than was necessary to achieve capture. The tracking and captureranges are a function of
external resistors and/or capacitors selected by the user.

*Capture State:when the phasecomparator of a PLLgenerates a signal thatforces the

VCO to equalthe input frequency.
*Locked:a PLL in the capture state.
*Free-RunningFrequency:frequency at which thePLL runs with the inputsignal
removed.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Example 2

A PLL is set up so th.at its VCO free-runs at JO MHz. The VCO does not
changefrequency until the input is within 50 kHz of JO MHz.After that condition,
theVCO follows the input to ±200 kHz of JO MHz before the VCO starts to free-
runagain. Determine the lock and capture ranges of the PLL.

Solution

The capture occurred at 50 kHz from the free-running VCO frequency. Assume
symmetrical
operation, which implies a capture range of 50 kHz X 2 = 100 kHz. Once captured, the
VCO follows the input to a 200-kHz deviation, implying a lock range of 200 kHz X 2 =
400 kHz.

PLL FM Demodulator

If the PLL input is an FM signal, the low-pass filter output (error voltage) is
thedemodulated signal. The modulated FM carrier changes frequency according to
themodu1ating signa1. The function of the phase-1ocked loop is to hold the VCO
frequencyin step with this changing carrier. If the carrier frequency increases,
forexample, the error voltage developed by the phase comparator and the low-pass
filterrises to make the VCO frequency rise. Let the carrier frequency fall and the
errorvoltage output drops to decrease the VCO frequency. Thus, we see that the
errorvoltage matches the modulating signal back at the transmitter; the error signal is
thedemodulated output.
The VCO input control signal (demodulated FM) causes the VCO output tomatch the FM
signal applied to the PLL (comparator input). If the FM carrier (center)frequency
dtiftsbecause of local oscillator drift, the PLL readjusts itself and norealignment is
necessary. In a conventional FM discriminator, any shift in the FMcarrier frequency
results in a distorted output because the LC detector circuits arethen untuned. The PLL

1
CNS Basic Train ing- Analog Co mmunication ECATC

FM discriminator requires no tuned circuits nor their associatedadjustments and adjusts

itself to any carrier frequency drifts caused by LO ortransmitted carrier drift. In addition,
the PLL normally has large amounts of internalamp1ification, which allows the input
signal to vary from the microvolt region up toseveral volts. Since the phase comparator
responds only to phase changes and not toamplitudes, the PLL is seen to provide a
limiting function of extremely wide range.The use of PLL FM detectors is widespread in
current designs

4.15. Stereo Demodulation

FM stereo receivers are identical to standard receivers up to the discriminator output.At

this point, however, the discriminator output contains the 30-Hz to 15-kHz (L + R)signal
and the 19-kHz subcarrier and the 23- to 53-kHz (L - R) signal. If a nonstereo(monaural)
receiver is tuned to a stereo station, its discriminator output maycontain the additional
frequencies, but even the 19-kHz subcarrier is above the normalaudible range, and its
audio amplifiers and speaker would probably not pass itanyway. Thus, the nonstereo
receiver reproduces the 30-Hz to 15-kHz (L + R) signal(a full monophonic broadcast)
and is not affected by the other frequencies. Thiseffect is illustrated in Figure 15.
The stereo receiver block diagram becomes more complex after the
discriminator.At this point, the three signals are separated by filtering action. The (L + R)
signalis obtained through a low-pass filter and given a delay so that it reaches the mattix

1
CNS Basic Train ing- Analog Co mmunication ECATC

Fig. 96:Monophonic and stereo receivers.

Fig. 97: Stereo signal processing.

network in step with the (L - R) signal. A 23- to 53-kHz bandpass filter selects the(L - R) double
sideband signal. A 19-kHz bandpass filter takes the pilot carrier andis multiplied by 2 to 38 kHz,
which is the precise carrier frequency of the DSB suppressedcarrier 23- to 53-kHz (L - R) signal.
Combining the 38-kHz and (L - R)signals through the nonlinear device of an AM detector
generates sum and differenceoutputs of which the 30-Hz to 15-kHz (L - R)components are
selected by alow-pass filter. The (L - R) signal is thereby retranslated back down to the
audiorange and it and the (L + R) signal are applied to the matrix network for furtherprocessmg.

1
CNS Basic Train ing- Analog Co mmunication ECATC

Figure 16 illustrates the matrix function and completes the stereo receiver blockdiagram of
Figure 15. The (L + R) and (L - R) signals are combined in an adder thatcancels R because (L +
R) + (L - R) = 2L. The (L - R) signal is also applied to aninverter, providing - (L - R ) = ( - L +
R), which is subsequently applied to anotheradder along with (L + R), which produces ( - L + R)
+ (L + R) = 2R. The twoindividual signals for the right and left channels are then
deempha~izedand individuallyamplified to their own speaker. The process of FM stereo is
ingenious in its relativesimplicity and effectiveness in providing complete compatibility and

4.16. FM Receivers

A modern approach to FM stereo reception is shown in Figure 20, which is a circuitdiagram for
the Philips Semiconductors TEA5767/68 single-chip FM stereo radio.The architecture of this IC
has dramatically reduced the number of external components.Compare this circ uit to the older
style of FM receiver shown in Figure 21 . InFigure 20, the RF signal connects to pins 35 and 37
on the TEA5767 /68. An RFAGC circuit prevents overloading and limits intermodulation
problems created bystrong adjacent channels.The circuit in Figure 20 has separate digital (pin 6)
and analog (pin 33)grounds and requires a typical analog (pin 34) and digital (pin 7) supply
voltage of3.0 V. The stereo audio appears at pins 22 and 23. The circuit uses a crystal
referencefrequency to facilitate PLL tuning. External clock frequencies are 32.768 kHzor 13
MHz (pin 16) or 6.5 MHz (pin 17). The crystal oscillator frequencies are usedfor reference by
the
• Frequency divider for the synthesizer PLL.
• IF counter timing.
• Stereo decoder free-running frequency adjustment.
• Center frequency adjustment of the IF filters.
Full microprocessor control of the TEA5767/68 including channel searching andselection audio
control, and power on reset are facilitated via the chip's bus connections(pins 8, 9, 11, 12, 13, 14,
and 15).
A typical older-style FM receiver involves use of discrete MOSFET RF andmixer stages with a
separately excited bipolar transistor local oscillator, as shown inFigure 21. The antenna input
signal is applied through the tuning circuit L 1, C1A tothe gate of the 40822 MOSFET RF
amplifier. Its output at the drain is coupled to thelower gate of the 40823 mixer MOSFET
through the C18-Li, tuned circuit. The 40244BIT oscillator signal is applied to the upper gate of
the mixer stage. The local oscillatortuned circuit that includes C1c uses a tapped inductor
indicating a Hartleyoscillator configuration. The tuning capacitor, Ci, has three separate ganged
capacitorsthat vary the tuning range of the RF amp and mixer tuned circuits from 88 to 108MHz
while varying the local oscillator frequency from 98.7 to 118.7 MHz to generatea 10.7-MHz IF
signal at the output of the mixer. The mixer output is applied tothe commercially available 10.7-
MHz double-tuned circuit T1•

1
CNS Basic Train ing- Analog Co mmunication ECATC

Discusssion
• Define angle modulation and describe the two categories.
• Explain a basic capacitor microphone FM generator and the effects of voice
amplitudeand frequency
• Explain an FM signal in terms of modulation index, sidebands, and power.
• Describe the noise suppression capabilities of FM and how they relate to the capture
effect and preemphasis
• Describe various schemes and circuits used to generate FM
• Explain how a PLL can be used to generate FM.
• Describe the multiplexing technique used to add stereo to the standard FM broadcast
systems