July 1961 Electronics World
Wax nostalgic about and learn from the history of early electronics. See articles
Electronics World, published May 1959
- December 1971. All copyrights hereby acknowledged.
FM radio has been in the news fairly frequently in the last
couple years as phone manufacturers and the
National Association of Broadcasters lobby the FCC and politicians
to mandate the inclusion of FM radio capability into every phone
manufactured. In a ploy to exploit the gullibility and egos
of said bureaucrats and pols, their primary argument that FM
radio is a 'first informer in times of crisis,' assuming of
course that people will miss news of 'the big one' when and
if it occurs. To my knowledge, successful reception of FM radio
on a cellphone requires the listener wear a set of wired ear
buds since the wire from the phone to the ear buds functions
as the antenna. What percentage of cellphone users would bother
to carry a set of ear buds? I, of course, am a huge proponent
of the continuance of over-the-air broadcasting, including that
of television. However, I also resent the continual manipulation
of citizens and implementation of regulations that impose unnecessary
financial and inconvenient burdens.
Stereophonic FM Multiplex System
By Daniel R. von Recklinghausen/ Chief Research Engineer,
H. H. Scott, Inc.
Authoritative details on the newly approved method of broadcasting
stereo programs over a single FM channel.
On April 19th, the Federal Communications Commission issued
specifications for the standard of modulation of FM multiplex
broadcasts. This decision came after long, careful deliberation
and study of many possible multiplex systems.
At least 17 different systems were proposed, some of them
dating back a good many years. The National Stereophonic Radio
Committee, with the help of engineers, members of the Electronic
Industries Association, representatives of the Institute of
High Fidelity Manufacturers, Inc., as well as other agencies
including the FCC, studied all of these systems and narrowed
the field to eight systems of six basically different groups.
In May 1960, the FCC issued Docket 13506 calling for technical
comments and field tests of these systems from the same station.
Six of these systems underwent field tests in July 1960, using
the facilities of KDKA in Pittsburgh - with receiving facilities
at the Uniontown Motel in Uniontown, Pa. Many hours of test
recordings were made on tape, along with the measurement of
many pertinent parameters. This data, plus an analysis of all
aspects of the systems tested, was then submitted to the Federal
Communications Commission in the form of a report - as bulky
as two New York City telephone directories.
The system adopted by the Federal Communications Commission
is along the lines of the methods proposed by General Electric
Company and Zenith Radio Corporation, with certain aspects of
the system developed by the FCC. These are the essential specifications
of the FM stereo broadcast transmission:
1. The modulating signal for the main channel is the sum
of the left and right audio signals.
2. A pilot carrier at 19,000 cycles, ± 2 cycles, modulates
the main carrier between 8 and 10%.
3. The sub carrier is the second harmonic of the pilot carrier
and crosses the time axis with a positive slope simultaneous
to each crossing of the time axis by the pilot carrier.
4. The sub carrier is amplitude-modulated and the subcarrier
itself, but not its sidebands, is suppressed to less than 1%
5. The modulating signal for the sub carrier is the difference
of the left and right signals with a frequency response of at
least 50 to 15,000 cycles with 75-microsecond pre-emphasis.
The main channel modulation, as in present monophonic service,
has the same requirement of frequency response and pre-emphasis.
6. The sum of the sidebands resulting from amplitude modulation
of the sub-carrier causes a peak deviation of the main carrier
of 45% when only the left or right signals are present. The
individual maximum modulation capabilities of the main carrier
and the subcarrier are 90% since the former reaches a maximum
when the latter is zero, and vice versa.
7. The frequency response of the sub-channel is identical
to the main channel (including pre-emphasis) to within 0.3 db
over the whole frequency range and the phase response must match
to ± 3 degrees. This results in a minimum separation of 29.7
db, maintained from 50 to 15,000 cps.
8. The distortion requirements of the subchannel are the
same as for the main channel and also for monophonic service.
The distortion of the audio equipment in the studio, the transmitter,
and its monitor operating together with proper pre-emphasis
and de-emphasis networks cannot be higher than 3.5% between
50 and 100 cycles, 2.5% between 100 and 7500 cycles, and 3%
between 7500 and 15,000 cycles.
9. Background music operation (SCA) is permitted. However,
the background music carrier may not modulate the transmitter
more than 10% and the crosstalk from the background music channel
into either stereo channel must be at least 60 db down from
Spectrum & Waveforms
Fig. 1 shows the spectrum which can be found at the output
of the FM detector. In normal monophonic operation of the station,
only the main channel and perhaps the additional subchannel
for background music is used. For stereo operation, the maximum
modulation of the main channel is then reduced from 100% to
90%, causing a volume change of less than 1 db which is inaudible
for all practical purposes. In addition to that, the pilot carrier
and the subchannel with stereo information is activated. It
should be noted that the main carrier of the station is modulated
with FM and never with AM. Therefore, all advantages of FM are
Fig. 1 - The spectrum of signals that is
found at the output of the FM detector.
The FCC system of stereophonic broadcasting had its inception
in a time-division multiplex switching system. Here, the input
of a transmitter is switched rapidly between the left and the
right stereophonic program channels, switching at a 38-kc. rate.
The system was analyzed mathematically and with test equipment.
It was discovered that this system was basically a "sum and
difference" system. The sum of the left and right stereophonic
channels appeared as audio modulation of the main carrier, whereas
the difference between the left and right stereophonic channels
appeared as suppressed-carrier amplitude modulation of a series
of odd harmonics of the switching rate. Since FCC regulations
do not permit the radiation of any signals in excess of 75,000
cycles, filtering of the higher harmonics of the switching rate
resulted in the adopted system.
Fig. 2 shows the waveforms which can be found on the output
of the FM detector of a high-quality FM tuner. In this illustration
the presence of the pilot carrier has been omitted for the sake
of clarity. For example, only the left channel (channel A) is
modulated with a sine wave. The sine wave itself will be found
if the output of the detector is examined with an oscilloscope
and a low-pass filter. The use of a high-pass filter will then
show the suppressed-carrier amplitude-modulated subchannel.
The elimination of all filters will then show the composite
modulation which is the sum of the audio modulation and the
subchannel modulation. Here it can be seen that channel A is
turned on several times during the complete audio cycle and
turned off just as many times. The right channel (channel B)
is not modulated.
Fig. 2 - Waveforms at the output of the FM
detector with a sine-wave modulation on the left channel. Pilot
carrier has been omitted.
The left and right channels can carry completely separate
and uncorrelated information, which is often true in stereophonic
recordings. In Fig. 3, channel A carries the high-frequency
audio modulation, whereas channel B carries a low-frequency
audio modulation. The summing of both of these modulations results
in the main channel modulation which shows the high-frequency
sine wave superimposed on a low-frequency sine wave. The subchannel
shows the difference of the two modulating wave-forms as the
carrier envelope of the suppressed carrier signal. Both the
main channel and the subchannel modulations form the composite
modulation which now clearly shows the presence of both waveforms.
The shaded areas shown in the last two portions of the figure
are actually the high-frequency sidebands of the subcarrier.
Fig. 3 - Output waveforms with high-frequency
audio modulation on channel A, and low-frequency audio modulation
on channel B.
The new FCC regulations concerning stereo broadcasting became
effective June 1st. Because broadcasting equipment is still
in short supply and only the major cities are likely to feature
multiplex stereo broadcasting, the waveforms described probably
cannot be observed immediately. However, since stations presently
engaged in AM-FM simulcasting are the ones most likely to carry
stereo multiplex broadcasts, a phone call to the station should
elicit information as to when stereo broadcasting is likely
Receiving Stereo Broadcasts
To receive these stereo broadcasts, a multiplex stereo tuner
or an FM tuner with a multiplex adapter will be required. Fig.
4 shows a block diagram of a stereo tuner. The signal from the
antenna passes through the r.f. amplifier, mixer, and oscillator
and is then amplified further in the i.f. amplifier at an intermediate
frequency of 10.7 mc. and detected as in monophonic service.
For normal monophonic service, the output of the detector is
then passed through a de-emphasis network and amplified still
further in an audio amplifier. At the output of the FM detector,
the composite stereo signal is available. It is then split into
two paths, one path containing the 19-kc. pilot or synchronizing
carrier. A filter selects this pilot carrier and this block
also contains means to produce a 38-kc. carrier frequency to
be fed to the stereo demodulator, which also receives the composite
signal. After de-emphasis of the left-and right-input signals,
further audio amplification is required to give a left and right
audio output of sufficient amplitude.
Fig. 4 - Simplified block diagram of a stereo
FM tuner that will produce signals for application to a stereo
Fig. 5 is a block diagram of a multiplex stereo adapter,
shown in greater detail. Since the stereophonic system that
has been described is a "sum and difference" system, the stereophonic
subcarrier may be demodulated by a synchronous detector to recover
the difference modulation and the main channel sum modulation
may be matrixed at audio frequencies with the conventional circuitry
to derive the left and right signal outputs. Here, the composite
output signal of the FM detector must first be amplified to
overcome the insertion loss of the matrix network and to operate
the AM detector at its proper level. The carrier of the AM detector
may be provided by selecting the output of this amplifier with
a 19-kc. filter, amplifying this signal, and using the second
harmonic (38 kc.) as the inserted carrier to the AM detector.
The output of the amplifier will also pass through a 23-53 kc.
bandpass filter into the AM detector. In this case, the synchronous
detector would be a double diode for recovering both polarities
of the A - B audio signal which is then matrixed with the A
+ B audio signal from the amplifier. Both the A - B and A +
B signals may contain a de-emphasis network. However, better
separation results if the de-emphasis network is placed after
the matrix network in the circuit. The matrix network would
be nothing but a set of resistors. Further audio amplification
is required to give proper audio output.
Fig. 5 - Block diagram of a multiplex stereo
adapter showing the circuits required.
The synchronous detector could be driven by an electron-coupled
19-kc. oscillator with frequency doubling in the plate circuit
so that the 38-kc. reference carrier may be correctly re-inserted
in the modulated subcarrier signal. The oscillator synchronization
would be accomplished by injecting the 19-kc. pilot carrier
into the oscillator tank circuit. It is important that the proper
phase relationship between the inserted carrier and the A -
B double-sideband signal be maintained. A phase error of 12
degrees can result in separation being reduced to 20 db. An
180-degree phase shift would result in an interchange of the
left and right audio outputs.
The A and B output signals may also be derived directly in
one operation on the composite modulating signal appearing at
the composite signal amplifier output. Such a detector would
make use of a tube similar to the 6AR8 beam deflection tube.
This tube has one control grid, two plates, and two deflecting
plates. The reference carrier of 38 kc., derived as above, is
applied to the deflecting plates and the composite stereophonic
signal is inserted at the control grid. The two plates will
have as outputs the product of the reference carrier and the
composite signal which is, of course, the left and right signal.
The adapters described here are of a rather simple nature.
Such adapters would be used where cost is of prime importance
and the various aspects of performance are perhaps of a secondary
nature, especially as far as ultimate quality performance is
concerned. The quality of the FM tuner employed will also affect
the various performance aspects rather severely. For example,
if it is desired to maintain a 30-db channel separation, then
the frequency response of the tuner with the adapter amplifier
may not vary more than 3/10 of 1 db from 50 to 53,000 cps, and
the over-all phase response from antenna input to stereo detector
input may vary no more than ± 3 degrees from a linear phase
line over the same frequency range. This performance aspect
is considerably more stringent than that required for monophonic
The internal impedance of the multiplex output of the tuner
and the input impedance of the adapter can affect frequency
response severely and, to a greater extent, the phase response
if an adapter of one make is used with a tuner of another make.
For this reason, the prospective buyer of a multiplex adapter
should check with the manufacturer of his tuner as to the type
of multiplex adapter to buy and also as to the extent of the
modifications required of the tuner to be able to obtain maximum
As the block diagram of Fig. 4 indicates, full advantages
of FM are maintained since the limiter or limiters in the tuner
are in normal operation as in standard monophonic service. The
amplitude modulation of the subcarrier is strictly a subsidiary
modulation and the output voltage of the subchannel detector
will be proportional to the subchannel output voltage of the
FM detector in the tuner. Since the main channel output voltage
at the tuner may vary with signal strength, the subchannel output
voltage will vary as well. Because of the proportionality of
output voltages, separation of left and right channels will
be maintained irrespective of signal strength and, therefore,
no front-panel separation adjustment is required. Such a control
may actually be harmful since this control is very likely to
be misadjusted by uninformed friends, neighbors, or children
and the correct setting can be found only when the test signal
is transmitted by the station equipped to broadcast multiplex
A separation control may still be found as a screwdriver
adjustment on certain adapters to compensate for tolerances
in the matrix network or the subchannel frequency response of
the tuner. For the latter deficiency, this control will be quite
beneficial in adjusting for good separation at low- and mid-audio
frequencies. However, a good high-frequency separation will
be obtained only if proper equalizing networks for the tuner's
frequency response are used. The characteristics of these networks
will vary among different types of tuners. This is one example
where the quality of the tuner, along with the quality of the
adapter, and the design of both will have an important effect
on stereo performance.
Background Music Operation
As the FCC specifications indicate and as illustrated in
Fig. 1, background music operation (SCA) of FM radio stations
is still permitted. However, these types of signals will have
to be broadcast on a fairly high-frequency subcarrier (such
as 67 kc.) when the station is engaged in multiplex stereo operation.
The careful design of the stereo multiplex adapter will have
a great influence on the amount of interference which will be
experienced when such a background music channel is in operation.
The transmitter's design and adjustment will also affect this
type of interference. However, the maximum transmitted interference
(-60 db) is specified by the FCC.
First photo received by the editors of a
multiplex adapter designed in accordance with the recently announced
FCC-approved system. The unit is the H. H. Scott Type 335, a
wide-band self-powered adapter, which is being sold for just
A station not engaged in stereo multiplex may operate its
background music carrier or carriers in the band between 20
and 75 kc. It is therefore possible that this type of background
music modulation may be picked up by a multiplex stereo tuner.
The type of interference which will then be heard can be described
as a "swishing whistle," the swishing controlled by the rate
at which the background music is performed. Such a noise would
be an indication that that particular station has a background
music carrier in operation and is not broadcasting stereo. These
types of noises are not the fault of the tuner or the adapter.
The stereo multiplex system adopted by the FCC, as well as
any other multiplex system that has ever been proposed, is subject
to interference to a greater degree than the monophonic main
channel. Ignition noise may perhaps sound louder; stray radiation
from FM tuners, TV sets, and other transmitters may cause whistle
tones to appear in the background; and the tube hiss of the
first stage in the FM tuner will degrade the signal-to-noise
When listening to signals of marginal signal strength, the
signal-to-noise ratio of the left- and right-channel stereo
output signals will be poorer than the monophonic main channel
by approximately 20 db. Measurements agree with this theoretical
prediction to within a few decibels. The ultimate signal-to-noise
ratio of the left- and right-channel outputs will be well in
excess of 60 db. Because of the reduced signal-to-noise ratio
when listening to stereo on weak stations, it will be desirable
to have a directional outdoor antenna available to provide an
increased amount of signal for the tuner.
The distortion which will result in a high-quality stereo
multiplex adapter connected to a high-quality FM tuner will
be on the order of 1/2% to perhaps 1%. This is considerably
lower than the distortion in subcarrier detection employing
other types of modulation. However, in listening to a station
it is possible that considerable distortion may be experienced
when listening stereophonically, yet relatively little distortion
be present when listening monophonically. This increased distortion
may be due to incorrect adjustment of the tuning control of
the tuner or more likely is caused by a reflection of the station's
transmitted signal from nearby buildings. This is an interfering
signal. For this problem, correct adjustment of the tuning control,
reorientation of the antenna, perhaps a reversal of the antenna
leads, will reduce this type of interference.
Tape recorder users will recognize that multiplex stereo
broadcasting provides an opportunity to obtain good quality
off-the-air recordings. However, during multiplex field tests
some of the tape recordings which were made indicated that there
could be definite interference problems. This is true since
the bias frequency of many tape recorders is not far removed
from the re-inserted carrier frequency or its harmonics. In
well-designed adapters, these potential problems have been eliminated.
Multiplex can be a greater boon for the listener than was
the development of the stereo record. The fact that certain
problems have been mentioned in this article shouldn't cause
too much consternation. You can't get something for nothing.
However, with proper thought given to the selection of equipment
and care taken in installation, the results should make the
effort completely worthwhile.
Posted September 21, 2015