Nov. / Dec. 1941 Radio-Craft
People old and young enjoy waxing nostalgic about and learning some of the history of early electronics.
Radio-Craft was published from 1929 through 1953. All copyrights are hereby acknowledged. See all articles
When you look at the circuit board and/or chassis of a radio set
- new or old - you see a lot of components including resistors,
semiconductors (and/or vacuum tubes),
inductors, capacitors, transformers, switches , potentiometers,
shielded cables, shielded compartments, displays, indicator lights,
connectors, etc. With the possible exception of some semiconductors
(ICs and discretes), the function
of just about every component can be discerned by most people who
are at all familiar with radio electronics by its location in the
circuit, with the exception being inductors and transformers
(other than those in the power supply).
Inductors and transformers tend to be the least understood and therefor
the most mysterious. They are the least likely to bear any identifying
marking unless they happen to be encapsulated like a resistor or
capacitor. Articles like this one help remove some of unknowns about
Coil Coupling Problems
L. V. Sorensen
circuit diagrams given to the right are explained in the text and
the important role played by the coupling coil in radio receiving
circuits especially are discussed at length by the author. The loop
antennas used so widely on portable sets as required some fine engineering,
to properly solve the problem of the best type of coil for the purpose.
The author explains in an easily understood manner, some of the
problems met with in radio design, so far as the coils and coupling
methods are concerned. Every student of radio technique will find
many valuable pointers in this article. Some of the topics are -
Loop Antennas, Reducing Hum Modulation, Effect of High Impedance
Primaries, Short-Wave Coils and R.F. Amplifiers.
Every radio set employs resonant circuits to select the desired
signals and to discriminate against the undesired signals, but the
methods by which these resonant circuits are coupled into the preceding
and following tubes or circuits vary widely as conditions dictate.
It is the purpose of this article to discuss the general methods
used to couple tuned circuits into radio receivers, outline the
general characteristics of each method, and to point out the limitations
that influence the type and amount of coupling.
A loop antenna is the outstanding example of a simple method
of coupling a resonant circuit to a tube. The entire tuned circuit
is directly connected between the grid and the cathode return of
the first tube. Signal voltage is introduced into the loop by direct
induction from the signal field existing in space.
When the signal from the desired station is too weak for satisfactory
reception on a loop, it is necessary to couple an outside antenna
to the receiver. The most common method of coupling is to connect
the antenna to a few turns of wire wound around the loop. When the
loop is of relatively large diameter the coupling winding is frequently
only a single turn, When the loop is small in diameter, and consequently
requires many turns of wire to make up the required inductance,
the coupling winding frequently has several turns in place of the
single turn used on the larger diameter loops. This type of coupling
is known as "low-impedance" coupling. Its characteristics are: low
gain at the low-frequency end of the tuning range, very much higher
gain (often 5 to 8 times) at the high-frequency end of the tuning
range, poor image-ratio, and considerable sensitivity (at the high-frequency
end of the tuning range) to variations in antenna capacity.
The principal limitation in the amount of this type of coupling
that can be employed is the amount of capacity "reflected" from
the antenna across the tuned circuit. When this "reflected" capacity
becomes large enough to throw the circuit badly out of alignment
with the other circuits in the receiver, a net loss in sensitivity
is caused by further increases in coupling. Under such a condition
of misalignment the image-ratio suffers very badly. In some sets
an isolating condenser of limited capacity, say 50- to 100-mmf.,
may be connected in series with the coupling winding to limit the
maximum capacity that may be reflected across the tuned circuit
by antennas of unusually high capacity.
A fairly common variation of the coupling winding just described
is to make the coupling conductive rather than inductive. In other
words, instead of using a separate winding for the antenna coupling,
a part of the secondary is used for the dual function of coupling
winding and for part of the tuned circuit. In construction, this
is accomplished by putting a tap on the loop winding one or more
turns from the low-potential end of the coil, and connecting this
tap to the antenna, either directly or through an isolating condenser.
The tapped loop may present some manufacturing advantages over the
separate coupling winding, but it has the distinct performance disadvantage
of sometimes introducing "hum-modulation" into the received signal.
Following the circuit in Figure 1, it can easily be seen how this
method of coupling the antenna to the loop introduces hum modulation
in AC-DC sets.
Reducing "Hum" Modulation: When the line plug is 50 connected that
side "A" is connected to the "ungrounded" side of the power line,
the A.C. line voltage is divided across C1, C2, and C3 in series.
Since only C1 has both terminals in the grid circuit this is the
only condenser across which an A.C. voltage of power-line frequency
can introduce hum into the grid circuit. A moment's consideration
of these three condensers in series will show that increasing the
value of C1, decreasing the value of C2 or reducing the capacity
of C3 (the capacity of antenna to ground) will reduce hum-modulation.
Of course, the quickest and easiest method of reducing hum-modulation
in the above case would be to reverse the line plug. Reversing the
line plug, however, will have little effect on hum voltages that
the antenna may pick up from high-voltage power lines that may be
in the neighborhood of the antenna. A.C. sets with the same type
of coupling from antenna to loop may experience similar trouble
but probably to a lesser degree, because the power line is not connected
to the chassis, but is merely bypassed to it with a relatively small
If hum-modulation only when the antenna is connected is a complaint
on any receiver using the type of antenna coupling shown in Figure
1, it is suggested that the antenna connection be removed from the
tap on the loop and that a completely separate coupling winding
be wound on the loop, right over the turns formerly included in
the antenna circuit.
This new winding should have the same number of turns as there
were between the tap and the AVC end of the loop. The AVC end of
the new coupling winding should be connected to chassis on an A.C.
set, and to "B" minus on an A.C.-D.C. set. In the latter case, an
isolating condenser should be connected between the antenna and
the high-potential end of the coupling coil while in the former
case no isolating condenser is necessary. It is very important that
the coupling winding be located at the AVC end of the loop, where
the capacity between the coupling winding and the loop adds virtually
nothing to the circuit capacity on the loop. If the coupling winding
is located at the grid end of the . loop, the winding will add so
much fixed capacity to the grid circuit, that it may be impossible
to "trim" the loop to resonance with the oscillator at the proper
Another scheme for coupling an antenna to a loop is to connect
the antenna directly to the grid through a very small capacity -
something around 2 mmf. Such a scheme is extremely simple but results
in a poor image ratio. To prevent serious mistracking at the high-frequency
end of the band, it is essential that the wiring from the low-capacity
coupling condenser to the antenna terminal or lead of the set have
a relatively high capacity to chassis, say 10 to 20 mmf. With such
a capacity coupling from the low-capacity condenser to ground or
chassis, little change is made in the effective grid-circuit capacity
when the antenna is connected to the antenna post, but if the low-capacity
coupling condenser is so placed that its lead to the antenna terminal
has negligible capacity to ground, the maximum mistracking will
result when an antenna is connected.
In broadcast radio sets employing conventional antenna coils,
the low-impedance primary was abandoned years ago in favor of high-impedance
coupling to provide better image ratio, flatter gain characteristics
from one end of the band to the other, and greater freedom from
the detuning effects of antennas of different capacities. The same
idea holds true in the better grades of sets employing loop antennas.
These sets are employing high-impedance coupling in either of two
forms: (1) a high-impedance primary of large diameter (perhaps 4
to 6 inches) coupled directly to the loop, or (2) a conventional
high-impedance primary of the usual diameter coupled rather tightly
to a small coil that is connected in series with the loop.
In either case the net result is the same as a conventional high-impedance
antenna coil having the same inductances, Q's and coupling capacities.
There is no question but what such a method of coupling an antenna
to a loop gives better performance than the low-impedance couplings
previously described, but many designers have used the poorer coupling
circuit because of its economy, and because of the fact that relatively
few loop sets ever have an external antenna connected to them.
Large-Loop Capacity Problem:
In some sets that cover a rather large tuning range on each band,
the distributed capacity of a large-diameter loop, comprising the
entire inductance of the tuned circuit, is so high that the tuning
range is seriously restricted. In such cases, an effective reduction
in distributed capacity can be obtained by separating the inductance
of the resonant circuit into two parts, putting approximately half
of the required inductance in the loop and the remainder in a small
coil of low distributed capacity connected in series with the high
side of the loop. If this inductance is mistakenly connected in
the low side of the loop, virtually no advantage has been gained
because the capacity of the loop to ground is still connected to
the grid. When this capacity is connected to ground, however, with
the small coil connected to the grid, the effective circuit capacity
is much lower and consequently the tuning ratio is greater.
Sets of the better grade employing the above method of stretching
the tuning ratio of the loop circuit, almost always have a high-impedance
primary coupled loosely to the above mentioned small coil. Thus,
when an outside antenna is connected to the receiver the loop circuit
partakes of the desirable characteristics of an antenna coil with
high-impedance primary. rather than the less desirable characteristics
of the low-impedance coupling scheme that employs one or more coupling
turns wound around the loop.
In general, all broadcast-band antenna coils of modern receivers
employ high-impedance coupling for the same reasons as given in
favor of such coupling in the better grade loop sets. The limiting
factor in the amount of coupling that can be used is the amount
of mistracking caused at the low-frequency end of the tuning range
when antennas of different capacities are used, and by the amount
of change in the apparent inductance of the tuned secondary as the
low-frequency end of the tuning range is approached. This directly
affects the tracking of an antenna coil with an R.F. or oscillator
Effect of High Impedance Primary:
Mathematically, it can be shown that actually a high-impedance
primary (with antenna connected) reflects a large capacity in series
with the tuned secondary, giving much the same effect as a padding
condenser in that circuit, and that the value of that condenser
becomes more disturbing to the normal tuning curve of the circuit
as the resonant frequency of the primary circuit (primary inductance
with antenna capacity) approaches nearer to the low-frequency end
of the tuning range, and as the degree of coupling between the circuits
becomes greater. A reasonably good general design specification
for coupling in a high-impedance circuit is 15% magnetic coupling
and a primary resonant frequency of about two-thirds of the lowest
frequency to which the secondary will tune.
In some instances, a little capacity coupling is added to the
high-impedance magnetic coupling for the purpose of changing the
gain characteristics of the coil. Since the polarity of the magnetic
coupling between two coils can be reversed by reversing either winding,
there are two possible polarities of connections.
With one of these polarities, the magnetic coupling reinforces
or aids the coupling provided by the coupling capacity while, with
the other polarity, the magnetic coupling counteracts or opposes
the coupling produced by the coupling condenser.
When the magnetic coupling aids the capacity coupling, the general
effect is to raise the gain all over the band, but most effectively
at the high-frequency end. When the magnetic coupling opposes the
capacity coupling there is a reduction of gain over the entire band
but most effectively at some one point, which may accidentally be
in the band in a poorly designed set (or one in which the coil has
been accidentally hooked up incorrectly).
This point of minimum response may be made to fall out ide of
the band at the "image" of some important frequency in the band.
When the latter is done the image ratio is improved most at the
one point where cancellation of coupling is greatest but the beneficial
effects of this opposition of couplings extend for a considerable
range on either side of the cancellation frequency. This opposed
coupling is most likely to be found in two-gang super-heterodynes
where image ratio is seldom as good as may be desired.
Short wave coils almost universally employ low-impedance coupling
for several good reasons: (1) A high-impedance winding for the range
just higher than the broadcast band resonates in the broadcast band.
If this resonance should accidentally fall at the frequency of a
local station, the station, in all probability, would cause "cross-talk"
on any station tuned in on the short wave band in question. (2)
Low-impedance coupling makes a much better impedance match between
a doublet antenna and the first tuned circuit.
The limiting factor in the amount of low-impedance coupling that
can be used on an antenna coil is the amount of mistracking caused
by different antennas at the high-frequency end of the tuning range,
and by the broadness of resonance of the antenna circuit when aligning
with the standard 400-ohm dummy antenna. This is especially important
in sets having only a two-gang condenser. If too many primary turns
are used, the resistance of the dummy antenna is reflected into
the tuned circuit in such an amount that the circuit becomes so
broad that it is difficult to distinguish between the desired signal
and the image, when the set is working on frequencies such as 16
to 18 mc. with a 456-kc. intermediate frequency and virtually impossible
to pad at the low-frequency end.
The types of coupling employed in R.F. amplifier stages probably
vary more than the coupling in any other kind of a radio-frequency
circuit. First, the two general classes of R.F. amplifiers, tuned
and untuned, determine whether coils and coil couplings will be
employed or not.
The simplest coupling is resistance coupling, in much the same
manner as a resistance-coupled audio amplifier with the exception
that the values of the coupling resistances and condenser are usually
considerably smaller than in audio circuits, and that a trap circuit
(tuned to the intermediate frequency) is usually employed to keep
out of the converter tube the noises arising in the R.F. tube at
the intermediate frequency. If these noises are not suppressed by
means of a wave-trap, the "signal-to-noise" ratio suffers. Such
a circuit is shown in Figure 2.
Untuned R.F. amplifiers of more complicated design may be used
to cover a greater band of frequencies, as in some All-Wave receivers.
The most commonly used tuned R.F. coupling circuits are high-impedance
coupling on the broadcast or long-wave bands and low-impedance coupling
on the shortwave bands.
The high-impedance circuit for R.F. coupling has several versions:
straight magnetic coupling, magnetic plus capacity coupling, and
capacity coupling (more familiarly known as choke coupling). These
coupling circuits are shown respectively in Figure 3, together with
representative examples of coils employing the coupling methods
The characteristic of the above type of coupling is a fairly
flat curve of gain vs. frequency, especially in the circuit employing
combined magnetic and capacity coupling. All of these circuits yield
better image ratios than low-impedance coupling, but to of these
three the circuit employing only magnetic coupling yields better
image ratios than either the combined magnetic and capacitive coupling
or the straight capacitive coupling (except in the case where the
magnetic coupling opposes the capacitive coupling for the specific
purpose of improving the image ratio).
In the design of such circuits, particular care should be exercised
to see that the primary resonant frequency does not coincide with
the intermediate frequency of the receiver because poor I.F. rejection
results when this occurs. If the intermediate frequency is 456 kc.
the primary resonance may be between the intermediate frequency
and the low-frequency end of the broadcast band, but such a primary
resonant frequency requires very close control of coil constants
and wiring capacities to keep the resonance at the desired frequency.
More uniform and more comfortable production will be experienced
if the primary resonance is placed well below the intermediate frequency,
where a reasonable variation of resonant frequency will have virtually
no effect on the uniformity of the finished sets.
The coupling employed in a few broadcast R.F. circuits of low
gain is of the low-impedance type, for reasons of economy and convenience.
It is true that the Image ratio obtained is not as good as that
obtained from a high-impedance circuit, but the image ratio of a
complete set employing a stage of such low-impedance amplification
is so much better than that of a set without a tuned R.F. stage
that some designers use such low-impedance coupling for the economy
that results. If they use high-impedance coupling they would unquestionably
obtain a still better image ratio, but at the price of a primary
of many turns of fine wire, and perhaps a mica bypass condenser
from the plate of the R.F. tube to plus "B" to cut the gain of the
stage instead of a low-impedance primary of a relatively few turns
(10 to 20) of fairly heavy wire.
The coupling employed on short-wave R.F. coils is almost universally
of the low-impedance type, in many cases with the primary turns
wound between the secondary turns, in an effort to obtain maximum
coupling and maximum gain.
The characteristics of this type of coupling on short-wave R.F.
coils are: maximum gain at all frequencies, greatest gain at the
high-frequency en d of the tuning range and image ratio inferior
to high-impedance coupling. This type of coupling is used however,
in spite of this limitation, in order to achieve high gain.
The limiting factor in the amount of coupling so employed is
single-stage oscillation in the R.F. tube and/or the amount of capacity
reflected from the R.F. plate circuit across the following grid
circuit. In order to see how single-stage oscillation is the limiting
factor and how to adjust coupling to avoid this difficulty, first
consider the circuit in Figure 4.
Inspection of this circuit shows it to have a tuned circuit directly
in its grid and plate circuits. If the frequency is anywhere in
the range from 2 to 18 mc. the gang condenser is of normal size
(365 to 410 mmf.), the coils are of normal "Q" design, and an A.C.
tube operating at normal voltages is employed, this stage will oscillate
as a tuned-grid-tuned-plate oscillator in exactly the same manner
as the old tuned-grid-tuned-plate transmitters operated.
Even with coupling between grid and plate circuit wiring reduced
to zero by careful placement of parts, the capacity inside of the
tube is enough to cause oscillation when the circuits are properly
tracked. It is necessary to reduce the effective impedance in either
the grid or the plate circuit of the R.F. tube to stop oscillation.
This could be readily done by connecting the plate to some point
part way down on the tuned circuit and, as a matter of fact, such
an arrangement is sometimes used. The far more common method, however,
is to keep the tuned circuit out of the high-voltage D.C. circuit,
and merely to couple the plate to this tuned circuit by a separate
winding, such as shown in Figure 5.
Since the impedance of any tuned circuit is maximum for the highest
inductance, it is obvious that the highest stage gain results on
the lowest-frequency band of a multiband short-wave receiver, unless
steps are taken to prevent this inequality. The steps usually taken
are to use a lower percentage of coupling on the lower-frequency
coil. For example, in a certain set the 5.5- to 18-mc. R.F. coil
had three-fourths as many turns on the primary as on the secondary,
while on the 1.7- to 5.6-mc: band the R.F. primary had only one-third
as many turns as the secondary.
This article prepared from data supplied by Meissner Mfg. Co.
(To be concluded)
Posted October 22, 2014