new logo
Wednesday – Saturday 10am - 5pm
Closed Sunday - Tuesday and holidays
Adults $8
Students $6

FM Multiplex Stereo
How It's Done

Until recently, there were only two kinds of radio:  AM and FM, which stand for Amplitude Modulation and Frequency Modulation, respectively.

In either case, the radio station’s transmitter has an oscillator that produces a sine wave at a particular frequency. 
This is the carrier wave, and this is what you tune your radio to.

The information – music, announcer’s voice, annoying political campaign ad, whatever – modulates that carrier wave
That’s essentially all radio is.

AM radio came first. 
In fact, it’s been around since December 24, 1906, when Reginald Fessenden used a telephone transmitter (microphone) to modulate a carrier wave produced by a high-frequency alternator.

In the case of AM, modulation is done by changing the amplitude (height) of the carrier wave.

diode detector Inside your radio is a detector or demodulator circuit that rectifies this modulated carrier wave – separating the audio information from the carrier and passing it on to the amplifier.

It’s a tried and true technology that's relatively easy to understand.

AM can cover great distances.

AM receivers are easy to build.

AM over modulationAM DISADVANTAGES:
AM has limited dynamic range; in other words, it's easily overmodulated, causing audible distortion. 
AM has limited frequency response. When AM channels are only 10 kHz apart, they can only reproduce audio up to 5 kHz.
AM is very susceptible to noise.  Remember, your AM radio’s detector is designed to respond to changes in amplitude.  Most noise – lightning, static, etc. – is amplitude-based, and your AM radio is sensitive to it.

In the early days of radio, the noise/static issue was seen as the biggest problem, and a number of engineers were working on ways to deal with it.

One of these people was Edwin Howard Armstrong.  Armstrong was a fascinating, and ultimately tragic, figure.  When he started looking at how the problem of noise and static, he already had two very important radio advances to his credit.

The first was regeneration, using feedback through a triode to increase its power.  Armstrong found that if you do this enough, the tube begins to oscillate, or create radio waves.  This made it possible to build radio transmitters around vacuum tubes instead of 20-ton alternators.

The other was the superheterodyne circuit, which is still the basic circuit design used in most modern radios, televisions, and radar sets.

Armstrong was given the task by RCA’s David Sarnoff of developing a “black box” that could be connected to or built into an AM radio to get rid of static.  Armstrong studied the problem, and the nature of radio static, and in 1933, came up with a solution that basically reinvented radio.  Instead of modulating the amplitude of the carrier wave, he would modulate its frequency.  This was the beginning of FM radio. 

In 1934 Armstrong conducted large scale field tests of the new system from the 85th floor of the Empire State Building and in 1937 financed the construction of the first FM radio station, W2XMN, in Alpine, New Jersey. He ordered 25 receivers from G.E. to demonstrate the new system. By 1938 both REL and GE were making FM receivers. By 1939 there were almost 40 experimental FM stations on the air, mostly on the east coast, and a few commercial stations by 1941.  We even had one in Minnesota, built by Charles Persons.  It eventually became WEBC-FM, and was claimed to be the first FM station west of Chicago.  The pre-war stations operated at 42 – 50 MHz and they were all monaural.

The primary difference between AM and FM is that with FM, the amplitude never changes but the frequency of the main carrier is constantly changing. With AM the frequency stays the same, but the amplitude of the wave is constantly changing.

FM radio circuits have limiter stages that keep amplitude constant and detectors that are not sensitive to changes in amplitude. 

FM has wide dynamic range creating the potential for much higher fidelity. 

FM has much better frequency response, typically from 50 Hz to 15 kHz.

The monaural FM signal that we’ve used since 1938 is capable of delivering static-free audio from 50 Hz to 15,000 Hz. Unlike the AM signal, which is easily overmodulated, FM has a wide dynamic range. Each station can deviate ± 75 kHz from its center frequency – in other words, if you tune in a station at 97.1 MHz, a 1 kHz tone at maximum volume would make that frequency vary by plus or minus 75 kHz.  So 1,000 times per second, the carrier wave would sweep up to 97.175 and down to 97.025. There’s also a 25 kHz guard band at each end, giving that station at 97.1 a 200 kHz spectrum of 97 to 97.2 MHz.

With so much bandwidth, it's easy to back off the modulation enough to accommodate subcarriers. On May 2, 1955, the FCC began issuing Subsidiary Communication Authorizations (SCA) for multiplexing via 41-67 kHz subcarriers to offer so-called “functional” or background music to subscribing clients. FM stations began leasing one or both SCAs to Muzak or the Storecast Corp. of America among others for use on a non-broadcast point-to-point basis and for some outlets, this income was a matter of survival.

Also in 1955, the FCC authorized experimental FM multiplex stereo broadcasts.  There were as many as 17 different systems tested by various companies over the next few years.  The FCC also mandated that whatever system was adopted had to be completely compatible with existing monaural receivers.

In the meantime, there was an option:  getting two-channel stereo by using both an FM and an AM signal.  This was called simulcast stereo, and since most FM stations were operated by companies that also operated sister AM stations, it was an easy method to adopt.  The right channel of the program was broadcast on the FM station, and the left on the AM.

What the listener needed then, was two radios, ideally two radios in one.  In the late 1950s, several hi-fi companies built special tuners and receivers to meet this demand.  They featured separate tuning dials for FM and AM.  A selector switch let you listen to just AM, just FM, or the FM-AM stereo simulcast. 

From its inception in 1938 until the mid 1960s FM was primarily used to broadcast classical music and educational programming. In the mid to late 1960s FM was adopted by the “Alternative Rock” (A.O.R.) format, and in 1978 listenership to FM exceeded that of AM.

Fisher MPX10As the race to true FM stereo escalated, it looked like Murray Crosby's "sum and difference" system, patented in 1953, would be the winner. By 1959, a reported 70 stations were using the Crosby system. Adaptors were being marketed by Fisher, Karg and Madison-Fielding among others.

The downfall of the Crosby system was not its commitment to full fidelity stereo. 
To the contrary, many listeners at the time insisted that it was better than the current system.  The failure of the Crosby system was it’s inability to help the handful of broadcasters who were invested in FM (in 1960) to find a way to pay for it.  The Zenith/GE system allowed the inclusion of subcarriers that could provide special programming without a noticeable degradation of main channel audio; the Crosby system did not.

On April 19,1961 the FCC announced their choice of the Zenith/GE system and the final standard effective midnight June 1, 1961.

Fisher component hi-fi system
top shelf: 1957 FM 90X FM mono tuner, Reko Kut turntable
bottom: matched pair 1952 Fisher 50A power amplifiers, 50C preamplifiers,
1961 MPX 100 multiplex adapter

How the Zenith/GE system works:

The key to the system is the "sum and difference" technique.
If "L" represents the signal corresponding to the left-channel program material, and "R" the right-channel material, then it is possible to add and subtract these two signals to form L+R and L-R. When the L+R and L-R signals reach the matrix of the multiplex decoder these two "sum and difference" signals are added and subtracted to produce the following results:

(L+R) + (L-R) = 2L
(L+R) - (L-R) = 2R

At the transmitter:

  • L+R signals are added to modulate the main carrier by 90% (of full 75 kHz modulation).
  • A 19 kHz pilot tone is generated and modulates the main carrier by 10%.
  • A 38 kHz subcarrier in phase with the 19 kHz pilot tone is amplitude modulated with the L-R signal.
  • The 38 kHz subcarrier is suppressed, but the L-R upper and lower sidebands, along with the L+R main channel audio and the 19 kHz pilot tone frequency modulate the main carrier.
  • When a 67 kHz subcarrier is added, the modulation of the main program material is "backed off" enough to accommodate the additional programming.

At the receiver:

  • In mono-mode, the main L+R signal, along with the19 kHz pilot tone and L-R sidebands, is passed to the demodulator stage (either a discriminator or a ratio detector). The 19kHz pilot and L-R sidebands are filtered out and the 50 Hz to 15 kHz (L+R) audio is sent out to the audio amplifier.

  • In stereo mode, the main L+R signal , along with the 19 kHz pilot tone and L-R sidebands, is passed to the demodulator stage (either a discriminator or a ratio detector). The 19 kHz pilot tone is not filtered out, but instead passes to the multiplex decoder along with the the L-R sidebands and the L+R main channel information.
  • The 19 kHz pilot tone is used to regenerate the 38 kHz signal that is re-inserted at the input of the balanced detector, thus restoring the proper phase relationship of the L-R sidebands at that point.
  • The L+R signal passes directly to the matrix network.
  • The L+R and L-R signals are added and subtracted at the matrix resulting in 2L and 2R at the respective left and right audio outputs.

  • The 19 kHz pilot tone is also used to trigger "stereo beacon" indicators and attenuators that hold the stereo audio signals to the same level as the monaural audio.

FM Stereo in Minnesota

from vol. 7, no. 3 (March - April 1996) of the
Pavek Museum of Broadcasting Newsletter

It Happened on the Air
by J.R. Lonto

The first attempt at stereo broadcasting in the Twin Cities came on November 30, 1957.  In a one-month joint venture between commercial station WLOL-FM (99.5) and the University of Minnesota’s KUOM-AM (770), the two stations agreed to simulcast an hour of symphonies every Saturday through the end of December.

One had to go to great lengths to enjoy the stereophonic sound.  The listener at home needed separate AM and FM receivers, placed six to eight feet apart, with the left tuned to 770 AM and the FM receiver on the right tuned to 99.5.  The radios had to face into the room, parallel to each other, with the volume just right.  It was suggested that the listener be ten to fifteen feet away from the speakers.

In 1959 KWFM 97.1 was one of only a handful of stations nationwide given permission to begin stereo testing using a multiplex system.  The experimental broadcasts aired in the wee hours over a frequency known as KA2XEDO.

KWFM (later known as KTCR-FM, then KTCZ-FM) obtained FCC permission and began regular stereophonic broadcasting in late December, 1961.  Those with multiplex adapters experienced the rich, three-dimensional sound of an entire symphony in their home without having to align two separate radios.

Soon, other stations followed, and by 1969, eight of the twelve FM stations in the Twin Cities, WCAL, KSJN, KQRS,WAYL, WLOL , WPBC, KEEY, and KTWN were broadcasting in stereo; ironically, KTCR, the former KWFM was not.       JRL

When we initially published this in 1996, we heard back from a few people. Some felt that WLOL came first, while others thought it was WAYL.  So, I called the oracle, Bruce Elving, longtime publisher of the FM Atlas.  He maintains that it was WAYL, which operated in the same Northwestern Electronics Institute building at 3800 Minnehaha Avenue in Minneapolis as KWFM.

One unnamed source reported that “the KWFM 97.1 and WAYL 93.7 antennae tower was right there and the transmitter was right in the studio.  Engineer Wayne Nelson lived under the stairs in the building.  He modified a GE console so that WAYL (or was it KWFM ?) could do the first real full-time FM-stereo broadcasts.