Receiver Theory

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Chapter 9

The Superheterodyne
Now that we understand the fundamentals of AM
and FM, it is time to look at the typical radio receiver.
Fig. 9-1 is the same crystal radio we introduced in
Chapter 7. We pointed out that it consists of just a few
parts:
An outside wire antenna, which captures the various radio signals coming our way
An antenna coil and variable capacitor, which do
two jobs. The antenna coil is wired as a transformer, coupling the antenna signal to the radio.
The secondary of the coil, along with the variable capacitor also form a tuned circuit which selects the station we want, while rejecting stations
we do not want.
A diode which rectifies the AM signal.
A 0.001 F capacitor which filters out the highfrequency carrier and sidebands, and keeps only
the envelope the audio signal.
A pair of headphones which convert the audio
signal to sound.
Variations of this radio circuit date back to the early
days of radio. The big advantage, of course, is that it is
simple, and requires no batteries. But that simplicity
carries a price the radio doesnt work too well. It has
poor sensitivity, and poor selectivity.

Sensitivity
Sensitivity describes the ability of a radio to pick up
weak signals. Our crystal radio has low sensitivity, because it can only pick up really strong stations.
Sensitivity has to be judged in relation to noise. Just
picking up a station is not enough, if the station is so
noisy that it is not pleasant to listen to. Spec sheets and
advertising literature usually specify receiver sensitivity
by measuring how much voltage from the antenna (usually measured in microvolts) is required to make the de-

sired signal (usually the sound out of the speaker) 10


times or 100 times stronger than the noise. This ratio of
signal to noise is then called the signal-to-noise ratio; a
decent radio might provide a 10-to-1 or 100-to-1 signalto-noise ratio with an antenna signal of under 1 microvolt.
This definition of sensitivity is useful for most radio
receivers, but not for a crystal radio. Typical receivers
have amplifiers which produce noise when tuned to a
weak or no station, so measuring the signal-to-noise
ratio is possible. With a crystal set, however, there is really no noise to be heard from the headphones, so measuring the ratio is tough. Still, you need several hundred
thousand microvolts of antenna signal to hear anything
at all, so sensitivity is clearly bad.
The signal-to-noise ratio is usually measured in
decibels as follows.
Suppose you turn on a receiver, disconnect the antenna (so there is no input signal), and adjust the volume
control to get some measurable amount of noise from
the speaker. Measure the speaker voltage and call that
the before voltage. (For our example, suppose it is 0.1
volts.)
Now connect the receivers antenna leads to a signal
generator, properly adjust its frequency so the radio can
receive it, and set the generator output until the
receivers speaker voltage is 1 volt (10 times more than
before.) Call this the after voltage; it is the voltage
needed to give a signal 10 times stronger than the noise.
Now insert the before and after values into the
formula for decibels:
after
1
db = 20 log
= 20 log
= 20 log 10 = 20 1
before
0.1
(since 10 to the 1st power is 10, log 10 is 1). And so the
SNR or signal-to-noise ratio is 20 db.
Now go back to measure the amount of signal coming from the generator into the receiver antenna connection; this value (in microvolts) is the sensitivity to
produce a 20 db signal-to-noise ratio in the receiver output.
Strictly speaking, this is not entirely correct, because
the after measurement is not just the signal, but also
includes a bit of noise. Hence many people will call our
20 db value the signal-plus-noise-to-noise ratio, rather
than just signal-to-noise ratio. They may also write it as
(S+N)/N, meaning that the signal-plus-noise is divided
by the noise output.

Selectivity
Selectivity describes the ability of a receiver to select
the station you want, and keep out other stations that
Fig. 9-1. A simple AM crystal radio

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-1

you dont want. Our crystal radio has poor selectivity,


because it has trouble separating nearby stations.
As weve discussed in previous chapters, a radio signal consists of a carrier and sidebands, and has a certain
bandwidth which depends on the type and amount of
modulation. Resonant circuits in the receiver tune in the
station you want; ideally, they should pass all the frequencies in the carrier and sidebands equally well, and
completely reject all other frequencies below and above
that. Practically, however, that is not possible.
Fig. 9-2 shows the actual frequency response of a
tuned circuit at the left, and the ideal response we would
like to have for a radio at the right. The ideal response
would be a rectangle, where all the signals within the
bandwidth of the radio signal (carrier and sidebands) get
through equally well (the response is 1 or 100%), while
nothing gets through above or below that range (the response is 0 or 0%.)
You can see there is a big difference between what
we want and what we get. The ideal rectangular response at the right has
1. A flat top; this lets the carrier and all sidebands
get through the tuned circuit equally well.
2. Steep skirts; the skirt is the vertical part at the left
and right. Steep skirts make sure that the response drops
very fast, so that no adjacent stations get through.
3. A definite bandwidth; ideally this should be just
as wide as the bandwidth of the signal we are trying to
receive no more, and no less.
The actual tuned circuit response on the left of Fig.
9-2 has none of these. The top isnt flat, so the carrier
can get through, but the farther out a sideband is, the
less of it gets through. The sides arent steep enough to
keep out adjacent stations, since even pretty far away
from the peak, the curve still has fairly high response.
And finally, there is no definite bandwidth to the circuit.
We can flatten out the top a bit by widening the
whole curve. The bandwidth of a tuned circuit determines the relative width of the curve (see Appendix A.)

The bandwidth in turn is determined by the Q or Quality


Factor of the circuit. The higher the Q, the narrower the
response is; the lower the Q, the wider it is. Unfortunately, there is a conflict here a lower Q would flatten out the top and thus provide more even transmission
of the desired signal, but it also widens the bandpass
and makes the skirts even less steep.
It now becomes obvious that a single tuned circuit
simply cannot provide the right selectivity for a radio,
even under ideal conditions. Even worse, the circuit of
Fig. 9-1 is very far from ideal. The antenna and the primary of the coil feed the signal into the tuned circuit, so
their resistance affects the Q. The headphones are the
load on the circuit, so their resistance also affects the Q.
And because all of these have fairly low resistance, the
Q is terrible! In a typical crystal radio, the response of
the tuned circuit is so wide, and the skirts so broad, that
it is almost impossible to separate stations from each
other. Unless you are lucky to live very close to one
radio station and far from all others, dont expect to get
very good performance from such a simple crystal
radio.

How do we improve sensitivity and


selectivity?
Improving sensitivity seems fairly simple just
add some amplifiers. Well ... its not quite that simple,
because you have to do it just right, as we shall see in a
moment, but it can be done.
Improving selectivity, on the other hand, is somewhat more complicated. There are some modern components, such as crystal or ceramic filters, which can
provide a fairly sharp bandpass. The more traditional
method, however, is to just add more tuned circuits. For
example, the dark curve in Fig. 9-3 shows how using
three resonant circuits, each tuned to a slightly different
frequency, and having different Qs, can improve the
overall response. There are several ways of getting this

Fig. 9-2. Actual and ideal resonant response

Copyright 2002 by Peter A. Stark - All Rights Reserved

Fig. 9-3. Improving the bandpass by combining tuned


circuits

9-2

The circuit started with an antenna, usually a long


wire strung outdoors. Then came two or more RF tuned
circuits, separated by RF amplifiers. These were called
RF because they all amplified the actual radio frequency
(RF) signal. Eventually came a detector, which was simFig. 9-4. This is one, not three tuned circuits
ply a rectifier diode and capacitor, that worked just like
those in the crystal radio of Fig. 9-1. This was followed
by an AF amplifier, called AF because it now amplified
same result, and obviously the circuit must be carefully
the audio frequency signal, not the radio frequency sigdesigned and set up, or else the resulting bandpass may
nal. The audio signal then went to a speaker. Youll note
be lopsided or have lumps in the top.
how RF amplifiers separated the tuned circuits, so they
The problem is that you cant just parallel a bunch of
would act separately instead of becoming one single
tuned circuits together as in Fig. 9-4. Even though this
tuned circuit, as in Fig. 9-4.
looks like three separate tuned circuits, if you parallel
The TRF receiver worked quite well for its time, but
the three inductors into one, and the three capacitors
it had some major flaws. One difficulty was that, each
into one, you see that there is really only one tuned cirtime you wanted to change stations, you had to retune
cuit here. To use more than one tuned circuit in the
all the tuned circuits. Although Fig. 9-5 shows only
radio, you must somehow separate them so they are not
three, some more expensive radios had four or even
all in parallel with each other. The secret is to separate
more. But as Fig. 9-3 shows, even three tuned circuits
them with amplifiers.
have to be carefully adjusted if you want to get an overall response with a fairly flat top and steep skirts. It was
almost impossible for the average owner to get it right.
The TRF or Tuned Radio Frequency
A second problem had to do with the actual physical
Receiver
construction of the radio. If two tuned circuits were too
The TRF, or Tuned Radio Frequency receiver, beclose to each other, the two inductors would act as a
came popular as soon as the electronics industry got to
transformer. Some of the amplified signal from one of
the point where it was possible to build amplifiers
the later stages would get back into an earlier stage,
cheaply enough. Fig. 9-5 shows the block diagram.
only to be amplified again and again this positive
feedback made the radio
into a perfect oscillator! And
the more tuned circuits there
were, or the more gain the
amplifiers had, the worse
the problem became. It was
very difficult to build a receiver that had both high
sensitivity and high selectivity.
Fig. 9-5. The Tuned Radio Frequency (TRF) receiver

Fig. 9-6. The superheterodyne receiver block diagraam

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-3

Superhet to the Rescue!


The solution was the superheterodyne (superhet for
short) receiver, invented around 1930. This idea revolutionized radio; all modern receivers use it.
Fig. 9-6 shows the superhets block diagram. The
amplification in the circuit is provided in three separate
sections: the RF section, which extends from the antenna to the mixer, the IF section which goes from the
mixer to the detector (and is the main feature of the
superhet), and the AF section, which extends from the
detector to the speaker.
As before, the RF section contains some RF tuned
circuits and amplifiers, which amplify the radio frequency signal. Similarly, the AF section contains some
audio amplifiers, and amplifies the audio signal. But between them is the IF or intermediate frequency section,
which further amplifies the signal, but also provides all
of the selectivity for the entire radio.
The superhet solves both problems of the TRF design: First, the IF section operates at a different frequency from the RF section; moreover, the IF section
stays tuned to the same frequency regardless of which
station we listen to. So the tuned circuits in the IF section can be properly aligned in the factory to give the
best bandpass curve, and they dont get retuned by the
user.
Further, because the overall radio gain is split into
three sections, each sections gain is smaller. With less
gain, feedback from the output back to the input is less
of a problem. And because each of the three sections
the RF, IF, and AF operates at a different frequency,
it doesnt matter if a signal from one section sneaks into
another, since it just gets rejected.
So the superhets significant feature is that the signal
in the IF portion of the radio stays at a constant frequency regardless of what station you tune to. This is
done by heterodyning or beating two signals.

Heterodyning is so important to radio that we have


to look at it some more.
Consider the circuit of Fig. 9-7. We have a box containing some circuitry, and two inputs into the box; one
is a 100 Hz sine wave, the other a 1000 Hz sine wave.
What comes out?

Fig. 9-7. Mixing two signals in some circuit

Assuming there is something inside the box (not just


empty air!), the two input signals will usually somehow
combine into the output. Electrical engineers will now
explain that there are two main possibilities:
If the circuitry in the box contains only resistors,
inductors, and capacitors, it is called a linear circuit. In linear circuits, the output is proportional
to the inputs; there is nothing in the output which
didnt come from the input. This is just a fancy
way of saying that, if 100 Hz and 1000 Hz go in,
then only 100 Hz and 1000 Hz can come out.
But if the circuitry in the box also contains some
diodes, transistors, tubes, or other non-linear
components, then this becomes a whole new
ballgame things can come out that didnt go
in.
For one thing, non-linear circuits can distort; they can
change the waveshape of the sine waves going in. As
we explained way back at the beginning, this introduces
harmonics. So the 100-Hz signal could now produce
harmonics of 200, 300, 400, or more Hz, while the
1000-Hz signal could now have harmonics at 2000,
3000, Hz. etc.
Much more important for us, though, is that the two
input signals can interact with each other. This process
is called heterodyning or beating. When two signals interact like this, they produce new signals whose frequencies are the sum and difference of the original two
signals. In our case, the sum would be 1100 Hz (1000
plus 100), and the difference would be 900 Hz (1000
minus 100). These new frequencies would be called heterodynes.
As usual, things are just a bit more complicated. The
distortion harmonics also produce sums and differences.
For example, the 200 Hz harmonic of the 100-Hz signal
could heterodyne with the 3000 Hz harmonic of the
1000-Hz signal to produce 2800 and 3200 Hz, and so
on. Fortunately, the harmonics are usually smaller than
the fundamentals, and so these heterodynes are also
smaller than the main ones at 1100 and 900 Hz.
At a first glance, you may think this heterodyning is
a terrible complication. But remember that, without heterodyning, the superheterodyne receiver would be impossible, and radio and TV reception would be a lot
worse today.

Lets now see how heterodyning is used in the


superhet. As an example, Fig. 9-8 shows a superhet AM
radio tuned to a radio station at 880 kHz. Coming in the
antenna is not just this stations signal, but also signals
from all sorts of other stations radio, TV, radar, etc.
The tuned circuits in the RF section remove most of the

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-4

Fig. 9-8. Superhet with a 455 kHz IF, tuned to 880 kHz

undesired signals, but not all, so that the signal coming


into the mixer is mostly 880 kHz, but still has many
other signals at nearby frequencies.
The mixer is a nonlinear circuit; it receives this combined signal, but it also gets a 1335 kHz signal from the
oscillator below it. Since it is nonlinear, it heterodynes
these signals. There are a lot of different signals going
in so it produces a lot of heterodynes, but the most important ones are the sum and difference of the desired
station at 880 kHz, and the oscillator signal at 1335
kHz. This gives us 2215 kHz, the sum, and 455 kHz, the
difference.
But note that the tuned circuits in the IF section are
all tuned to 455 kHz, so they keep the 455 kHz signal
and reject the others. By the time signal gets to the detector, the filtering has been pretty much completed, and
the signal is almost pure 455 kHz (plus any nearby sidebands .)
Now, suppose we retune the radio to a different station, say one at 770 kHz. We retune the RF tuned circuits, but these only do a rough job of removing
faraway signals; they arent the main tuned circuits in
the radio, so it isnt important to get them just right. But
and here is the important thing to note we also retune the oscillator to 1225 kHz. The difference between
1225 kHz and 770 kHz is again 455 kHz! And so the IF
section again amplifies the resulting signal, without having to be itself retuned.
So the trick when changing stations is to retune the
RF circuits (but a slight error here isnt critical), and
also retune the oscillator (and this is important!) so the
difference frequency between the station you want, and
the oscillator, stays at 455 kHz. Since the RF tuning adjustment isnt that critical, it is possible to use a single
knob to adjust all the tuned circuits at the same time,
without having to worry about whether all of them are
right on target.
In case you wonder why we chose 455 kHz for the
IF frequency ... well, other values are possible, but this

just happens to be a popular one in AM broadcast receivers. FM broadcast receivers usually use 10.7 MHz
IF, and other IF frequencies are also used in other kinds
of receivers.
If we let fstation be the frequency of the station we
want, and FIF be the IF frequency, then the oscillator
frequency fosc should be
fosc = fstation + fIF
But its also possible to let
fosc = fstation fIF
Either way, the difference between fstation and fosc is
equal to the IF frequency fIF, so either will work.

The Converter
Many radios combine the mixer and the oscillator
into one circuit called the converter. This is a popular
technique for lowering the radios cost, because several
components in the circuit do multiple jobs at the same
time. Fig. 9-9 shows the converter used in many popular

Copyright 2002 by Peter A. Stark - All Rights Reserved

Fig. 9-9. Typical converter in an AM receiver

9-5

AM broadcast radios; there are several useful techniques that are worth mentioning.
L1 and C1 are the RF tuned circuit, with C1 being
the tuning capacitor. But L1 does several different jobs.
The top part of the winding (above the ground connection) is the part that actually resonates with the capacitor; the bottom part (connecting to C2) acts as the
secondary of a transformer, to bring the signal from L1
to the transistor without loading down the tuned circuit
(which would reduce the Q.)
At the same time, L1 is also the antenna. As we have
seen, coils or loops of wire can act as antennas; in this
case, L1 is wound on a ferrite core (a ceramic core
which contains ferrous metal particles); the core helps to
pick up the energy from the radio signal, and concentrate it in the coil.
The transistor also does two jobs. First, it oscillates
at a frequency 455 kHz above the signal you want to
pick up. To do this, we need an amplifier with positive
feedback. The transistor is the amplifier, with its output
coming out of the collector, going through oscillator
coil T1, and back through C3 into the emitter of the transistor. Capacitor C4 resonates with the secondary of this
coil to control the oscillator frequency.
At the same time, however, the transistor amplifies
the RF signal coming from the antenna coil, and mixes
it with the oscillator signal. Because the transistor is
non-linear, it also produces the sum and difference heterodyne frequencies. The primary of IF transformer T2
and capacitor C5 resonate at 455 kHz, and send the 455
kHz difference frequency on to the IF amplifier.
Note how T1 and T2 both use taps on one winding
(the tap is a third connection part way into the winding.)
This reduces the loading on the resonant circuit, and
keeps the Q from being lowered.

Variations on a Theme
Although Fig. 9-6 showed one RF amplifier (with
two RF tuned circuits), and one IF amplifier (also with
two IF tuned circuits), there is nothing sacred about
these numbers. Many cheap radios use only one RF
tuned circuit, and no RF amplifier; the converter in Fig.
9-9 is a good example.
On the other hand, more expensive radios might
have more RF amplifiers and/or more IF amplifiers. In
fact, quite a few radios use double-conversion; in this
scheme, there are two mixers and oscillators (or two
converters), and two different IF amplifier sections.
Well see the reason for this in a moment.

Superheterodyne Sensitivity and Selectivity

By splitting the amplification into separate sections,


a superhet can provide more total gain without the danger of signals feeding back and causing oscillation. Further, because the IF amplifier does not need to be
retuned each time you change stations, it can be optimized, and carefully adjusted at the factory, to provide
the best possible bandpass characteristics steep skirts
and a flat top. But there is more to it than that.
Recall our definition of the Quality factor Q of a resonant circuit:
resonant frequency
Q=
3db bandwidth BW
The 3-db bandwidth doesnt really specify how well
the circuit will reject adjacent stations; in order to reject
such interference, the response of the tuned circuit has
to be 30, 40, or even more db down from the top of the
curve at the frequencies of any adjacent stations. Still, it
provides a useful yardstick for comparison.
We can rewrite the above equation as
resonant frequency
3db bandwidth BW =
Q
To get a small bandwidth, you have to either make
the resonant frequency small, or make the Q big. But in
most resonant circuits, there is a limit on how big Q can
get; it is affected by the resistance of the rest of the circuit, and is seldom more than 20 or 30. So making Q big
is not a feasible approach to making the bandwidth
small. So, to get a small bandwidth, it would help if you
could make the resonant frequency small.
But in a TRF receiver, you must tune the resonant
circuits to the frequency of the station you want to receive, so you really cant make the resonant frequency
small to get a good bandwidth. And the bandwidth will
change as you tune to different stations, further complicating the design.
In a superhet, on the other hand, all the selectivity is
obtained in the IF stages, and their frequency stays the
same for all stations. Moreover, the IF frequency is
lower than any of the stations you want to receive, so
you can get the same narrow bandwidth for every station you listen to. In theory, at least, you could get the
bandwidth as narrow as you want, simply be going to a
lower IF frequency. (Note how AM broadcast radios,
which need a lower bandwidth than FM broadcast radios, also have a lower IF frequency of 455 kHz instead
of 10.7 MHz.) But there is a fly in the ointment, as they
say the image.

The Image
Lets return to the radio in Fig. 9-8. It is tuned to 880
kHz, has a 455 kHz IF, and an oscillator frequency of
1335 kHz. Here we see that

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-6

1335 kHz 880 kHz = 455 kHz.


So far, so good. But suppose there was a station at
1790 kHz. Look at the following calculation:
1790 kHz 1335 kHz = 455 kHz
In other words, the difference between the new station at 1790 kHz and the 1335 kHz oscillator frequency
is also 455 kHz. This new radio station could also now
be heard, though not as well as the one at 880 kHz because the RF tuned circuits largely remove it. But if it
were strong enough, it would come through anyway.
The 1790 kHz frequency is called the image frequency.
Note how the image frequency is calculated:
Desired station
880 kHz
+ IF frequency
+455 kHz
Oscillator frequency
1335 kHz
+ IF frequency
+455 kHz
Image frequency
1790 kHz
That is, the image frequency fimage is
fimage = fdesired station 2 fIF
(We used the sign in the equation because in some radios the oscillator could also be below the desired station frequency; in that case, the image frequency would
be below the oscillator frequency, and we would need
the minus sign.)
This brings us to a problem just a few paragraphs
ago, we said In theory, at least, you could get the
bandwidth as narrow as you want, simply be going to a
lower IF frequency. But if you do that, then the image
frequency gets closer to the desired frequency, and then
the RF tuned circuits may not be able to get rid of it. So
you have two conflicting requirements:
To get better selectivity lower bandwidth
you want to lower the IF frequency.
To get better rejection of the image frequency,
you want to raise the IF frequency.

This is particularly a problem with high-frequency receivers intended to receive narrow-band signals. For example, consider an amateur FM receiver for 146.94
MHz. Since the bandwidth of FM signals on this frequency is typically only 10 or 15 kHz, a low IF frequency (such as 455 kHz or even less) would be ideal.
But then the image would be at
146.94 MHz + (2 455 kHz) = 147.85 MHz
which is not even 1% away from the desired frequency.
There is no way that a typical RF tuned circuit could
keep the image out youd need a tremendous Q to do
it.
Typical receivers solve the problem one of two
ways. A few use a much higher IF frequency (around 10
MHz), but with special crystal or ceramic filters which
can achieve the narrow bandwidth even at this higher IF
frequency.
But a much more common alternative is to use two
separate IF sections and double conversion. Fig. 9-10
shows the block diagram of a double-conversion superhet to receive 146.94 MHz. Since 10.7 MHz and 455
kHz IF transformers are fairly inexpensive (they are
manufactured by the zillions for use in standard AM and
FM broadcast receivers), many communications radios
use them as well, and we show them here.
To receive 146.94 MHz, the first oscillator runs at
146.94 minus 10.7 MHz, or 136.24 MHz (the oscillator
could be either 10.7 MHz above the desired signal, or
10.7 MHz below; in this case, we chose to use the lower
frequency.) The second oscillator and mixer converts
the 10.7 MHz first IF signal to 455 kHz by using an oscillator at 10.7 MHz + 0.455 MHz, or 11.155 MHz.
By using two IF frequencies, the double-conversion
receiver solves our two problems. The high first IF frequency does not provide much selectivity, but it helps to
eliminate the image. Since the image frequency is at
fimage = fdesired station 2 fIF

Fig. 9-10. A double-conversion superheterodyne

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-7

(note that we use a minus sign since the oscillator is


below the desired signal, so the image must be even farther below that), the image frequency is now
146.94 MHz (2 10.7 MHz) = 125.54 MHz
which is far enough away from 146.94 that the RF tuned
circuits can remove it (or at least significantly reduce it).
The second IF frequency of 455 kHz, on the other
hand, is low enough that even transformers with reasonable Q can provide a narrow bandwidth.
Incidentally, suppose we wanted to use a similar circuit to receive 145.015 MHz instead of 146.94 MHz.
This circuit would not do, and for an interesting reason:
The 11.155 MHz signal from the second oscillator goes
into the second mixer, and the mixer is intentionally
non-linear (to produce a heterodyne.) Hence it also generates harmonics of all the signals going in. It turns out
that the 13th harmonic of 11.155 MHz is exactly
145.015 MHz. Although this harmonic is weak, a slight
amount of it will still sneak back into the RF stage, and
fool the receiver into thinking there is a weak, unmodulated signal at that frequency. Unless your desired signal
is substantially stronger than this false signal (called a
birdie,) it will not be heard.
The solution in this case is to change the second oscillator frequency from 10.7 plus 455 kHz, to 10.7
minus 455 kHz, or 10.245 MHz. This new oscillator frequency has harmonics at different places; while this removes the birdie at 145.015 MHz, it introduces birdies
elsewhere, such as 143.43 (which is the 14th harmonic
of 10.245 MHz.) Designing wide-band receivers (receivers designed to receive a wide range of frequencies)
is thus always a problem; there are always some birdies
somewhere, and the designer has to carefully choose his
oscillator and IF frequencies to try to place the birdies at
places where they will not interfere with normal operation.

Summary
If the superheterodyne receiver had never been invented, communications as we know it would probably
not exist. The combination of features we have described allows radio receivers to have selectivity and
sensitivity which allows millions of transmitters around
the world to coexist with each other, yet allows us to select and listen to even extremely weak signals from far
away.
We have touched on some of the important concepts, yet have had to skip many others. In the next
chapter, we will try to cover some more concepts having to do with the transmitters and receivers which we
run into daily.

Copyright 2002 by Peter A. Stark - All Rights Reserved

9-8

You might also like