|Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting
Departments | Features | ADs | Equipment | Music/Recordings | History
by LEONARD FELDMAN
By the time you read this, you may be able to turn on the radio in your car or home and listen to a stereo broadcast--on the AM radio band. Or, you may be able to listen to two different AM stereo broadcast systems--or three-or four, or possibly even five! That's because the Federal Communications Commission, unable to decide which of five proposed systems for AM stereo was "best," took the easy way out. Instead of providing the country with a national standard for this new service (one of the purposes for which the FCC was ostensibly created), that august body of seven commissioners and assorted technical staff, after spending more than five years on this subject, issuing a tentative decision in favor of one of the systems (Magnavox) more than two years ago, and then, a few months later, retracting its tentative decision for the purpose of "taking another look at the data," has brought forth what amounts to a colossal non-decision.
The FCC has asked the American public to decide which system (if any) should dominate. In other words, U.S. AM radio stations are now free to go on the air, using any one of the five systems they wish, and transmit stereo programming. What the American public is supposed to do to receive these broadcasts is left to the manufacturers of home and car stereo equipment.
The makers of such equipment can "bet" on a single system at the outset and produce receivers which can detect and decode that system. Or they can build receivers containing multiple decoder circuits and a selector switch-at considerable additional cost to the purchaser. They can probably even devise receivers which will automatically respond to the type of transmission being received and switch over to appropriate decoder circuitry without the listener having to do a thing--again, at a considerable increase in cost of the receiver. Or, as may very well happen, they can sit back in a classical chicken and egg dilemma, waiting for broadcasters to make up their minds, while the broadcasters wait for the receiver manufacturers to make up theirs! It would be one thing, of course, if all five systems were nearly identical in operation. Under those circumstances, a single circuit with minor switched-in modifications might be possible without adding too much cost to the receiver. But, as a quick look at the sidebars will show, the five proposed AM stereo systems are quite different from each other. To be sure, National Semiconductor, for example, has developed a single chip (LM1981) which, though originally designed for the "about to be approved" Magnavox system, can, according to its developers, be modified (by changes in external componentry) for use with at least two other proposed systems (Motorola and Belar), and, possibly, with the addition or substitution of yet additional components, for the other two proposed systems (Harris and Kahn/Hazeltine) as well. Still, time alone will tell whether the FCC non-decision leads to the creation of a new type of radio service in the United States or results in abandonment of the entire idea of AM stereo.
At stake are enormous sums of money. Car audio is particularly ripe for AM stereo. In many places, such as the Bay Area of California, automobile commuters cannot listen to a single FM station throughout the drive to and from work. Indeed, at least six leaders in car audio already have built prototype receivers.
While AM stereo will not automatically make existing gear obsolete, the industry estimates that five million AM stereo receivers could be sold in 1983, if a single system is established. But by leaving the decision to the marketplace, where price, advertising and other considerations may be as significant as quality, this becomes a very large IF. Nevertheless, some interested parties have already picked their horse in the AM stereo derby. National Semi conductor, for example, one of the world's leading suppliers of integrated circuits, entered the AM stereo field in 1976. By the time the FCC gave its initial approval of the Magnavox system in 1980, National Semiconductor was ready to go to market with its first AM stereo decoder. It was specifically designed for the Magnavox system, and although the chips can be made to suit two other systems, National openly prefers Magnavox.
Dan Shockey, an executive of the company, has said, "Momentum in the receiver industry towards the Magnavox system, based on the 1980 decision, is the only clear momentum anywhere." He has also claimed that it was the only "existing, practical solution to AM stereo at the crucial receiver level ... the only one system ready for production at major receiver manufacturers." Shockey warns that failure to establish an AM stereo system within the next six months will confuse the public, resulting in the kind of reaction that has all but destroyed quad and slowed the growth of videodiscs.
Pioneer has offered cautious support for the Magnavox system. "After extensive research on the proposed system of implementing AM stereo," said company President Jack Doyle, "Pioneer has determined the Magnavox format to be one of the best and most effective." However, Pioneer has yet to make any AM stereo receivers.
Sansui has built a prototype tuner capable of automatically detecting and decoding the Magnavox, Harris or Kahn systems. But Sansui Vice President Tom Yoda says, "We are not supporting any specific AM stereo system at this time." Ford Motor Company, whose subsidiary Ford Aerospace and Communications Corp. manufactures car receivers, notified the FCC that it favored either the Magnavox or Belar systems.
Ford reported that its engineers found that the costs for Motorola, Harris and Kahn systems would be roughly double that of Belar or Magnavox "without perceivable improvement." On behalf of its candidate, the Harris Corp. intends to enter the market aggressively, using the company's unique leverage. Of all the manufacturers, Harris is the only one that also produces transmitters and in that role supplied more than half of all AM and FM transmitters installed by U.S. broadcasters last year. The company claims to have 150 "contingent" orders from AM stations for stereo systems and will immediately seek to con vet/ these to firm commitments.
Meanwhile, Sanyo, Matsushita, Trio-Kenwood, Aiwa, Sony, Sharp, JVC, Nippon Columbia, Akai, et al. appear to be waiting for more information before making any decisions on a system. Their hesitation is not based solely on the uncertainty in the U.S. market, because the Japanese government has also failed to choose a system for AM stereo in that country.
While there are some similarities between the five systems now "approved" by the FCC, by no stretch of the imagination could all five systems be said to be compatible with each other. No one can predict at this time which system (if any) will gain supremacy. What is certain is that even if one system ultimately wins out, there is little guarantee that it will be the "best" system, from a technical point of view. At least three commissioners at the FCC agree with that unfortunate conclusion.
I'd like to quote, verbatim, from a statement of FCC Commissioner James H. Quello who, while concurring in the result of the FCC deliberations, said, in part: "The marketplace has very little competence to determine the relative merits of one technical standard versus another over the short term since its decisions are generally influenced by marketing efforts more than by technical superiority. To expect the American public to select a nationally compatible AM stereo system in a reasonable period of time from among five systems now before this Commission is sheer folly. In the first instance, the decision as to what system to select falls upon the broadcaster. He must guess which system will gain I enough public acceptance over time to survive. It is at that point that the public must cast its votes with each broadcaster, principally with a view toward avoiding, if possible, early obsolescence of receivers. Should two or more incompatible systems develop within a listening area, competition for listeners effectively takes place only until stereo receivers are purchased. From then on, the purchaser is "locked" into the system selected insofar as stereo is concerned.
"I am appalled that it has taken this Commission five years to decide that it cannot decide this issue ...." So are we, Commissioner Quello, so are we!
The Magnavox System
Almost gaining approval of the FCC, the Magnavox system uses a combination of amplitude modulation and linear phase modulation. As is true in FM stereo broadcasting (and with all of the proposed AM stereo techniques), stereophonic program material is first matrixed into (L+R) and (L-R) signals. The (L+R) signal amplitude-modulates the r.f. carrier, while the (L-R) signal phase-modulates the carrier. This system also includes a 5-Hz FM signal on the carrier which can be used for automatic identification of stereo broadcasts and, perhaps, automatic switching of receiver circuitry from mono to stereo.
A three-step process generates the required composite signal for the Magnavox system; the elements are shown in the block diagram of Fig. 1. First, a 3.69 MHz oscillator is frequency-modulated with a 5-Hz identification tone. Phase deviation is limited to 4 radians. Next, a phase modulator adds the (L-R) audio component as phase deviation to the output of a tunable frequency synthesizer. Tuning range is from 4 to 6 MHz, and the maximum peak deviation contributed by the (L-R) component is limited to a maximum of one radian. The two modulated signals are down-converted by a heterodyning process to the desired broadcast band frequency of the station, applied to a standard AM transmitter's r.f. input, and amplified to full station power. This amplified carrier is also amplitude-modulated by the (L+R) component using the modulation circuitry already found in the standard AM transmitter. A delay network equalizes the time delays which the (L+R) and (L-R) signals encounter prior to transmission.
A block diagram of a possible decoder circuit for the Magnavox system is shown in Fig. 2. All r.f. and i.f circuitry is the same as it would be in a monophonic AM receiver. At the i.f. amplifier output, the signal splits into two directions. A typical envelope detector may be used to extract the amplitude modulated material (L+R) and to detect the r.f. carrier level for an a.g.c. function. At the same time, the i.f. amplifier output is limited to remove amplitude modulation, and a phase modulation (PM) detector extracts the (L-R) audio signal and the 5-Hz identification tone. Next, AM and PM channel outputs are applied to the matrix block. If a 5 Hz tone is present (indicating that stereo is being transmitted), the stereo/mono switch is placed in the stereo mode. In the absence of this identification tone, the mono/stereo switch reverts to the mono mode, and the (L+R) signal appears at both outputs-a monophonic signal from both loudspeakers.
The Belar System
Belar Electronics Laboratory proposed a system derived from an original proposal by RCA (which was not an active proponent in the FCC AM Stereo Docket). The Belar transmission method is illustrated in the block diagram of Fig. 3. As in the Magnavox system, matrixed (L+ R) and (L-R) signals are used. The (LR) component is first pre-emphasized using a 400-mS time constant. It is then routed through a variable-delay line to an FM modulator input, where it frequency-modulates the r.f. signal used to drive the transmitter. At the same time, the (L+R) signal amplitude-modulates the carrier in the conventional way. Peak low-frequency deviation of the carrier by the (L-R) component is limited to ±1.25 kHz. A typical receiver decoding circuit for the Belar system is shown in block diagram form in Fig. 4.
Again, the r.f. and i.f. stages would be the same as in an ordinary monophonic AM receiver. The output of the i.f. amplifier is split into two separate detection paths. One path applies the signal to a limiter amplifier which re moves any AM modulation, leaving an FM square wave to be de-modulated by an FM detector such as a discriminator or ratio detector. The other path applies the signal to a conventional envelope detector where the (L+R) information is recovered. After detection, the recovered (L + R) signal is matrixed with the recovered and de-emphasized (L R) signal to yield separate "L" and, "R" output signals. Belar proposes to use a 10-Hz pilot signal.
The Harris System
Harris Corp. calls its system a compatible phase-multiplex system (CPM). As shown in Fig. 5, in this system, the left-channel signal amplitude-modulates a carrier which lags the transmitted resultant by 15°; the right channel modulates a carrier which leads the transmitted resultant carrier by 15°. When these two signals are added together, they form coherent (L + R) sidebands which are identical to those of a conventional monophonic signal, as well as (L-R) quadrature side bands which are reduced in amplitude with respect to the (L+R) sidebands because of the use of the 30° angle. The CPM system might therefore be described as a modified quadrature system. It uses a technique which is very similar to that used in color TV for transmitting two chroma (color) signals on a single carrier. In Fig. 5, the composite signal, V(t), created by the addition of the leading and lagging r.f. signals, must be separated into its envelope components, Ve(t), and phase modulated carrier components, V(c)t, before it can be transmitted by a conventional AM transmitter.
A clipper circuit recovers the phase modulation, while an envelope AM detector recovers the amplitude modulation. These two components, Vc(t) and Ve(t), are then applied, respectively, to the r.f. and audio inputs of the transmitter.
One possible form of receiving decoder circuitry for the Harris system is illustrated in the block diagram of Fig. 6. An ordinary AM receiving circuit is used up to the point of i.f. detection. There, a phase-locked loop circuit regenerates an unmodulated i f. signal from the incoming i.f. signal. The regenerated i.f. signal from the voltage-controlled oscillator of the PLL circuit is 90° out of phase with respect to the incoming modulated i.f. signal, so that mixer M2 demodulates the quadrature (L R) information directly. A 90° phase-shift network produces umodulated i.f. that is in phase with the incoming modulated i.f. signals. This signal is combined with the modulated i.f. signal In a second mixer, M1, to demodulate the in-phase (L+R) information.
After appropriate gain adjustments, the two recovered signals are matrixed to produce separate "L" and "R" signals. Though not shown in the diagrams of Figs. 5 and 6, Harris has also suggested that a 20 to 25 Hz signal, at 9% modulation of the (L-R) channel, be transmitted as a stereo-indicating or switching signal.
The Motorola System
This system is known as C-QUAM. One method of transmitting two signals on one carrier is to modulate two carriers of the same frequency which are 90° apart in phase.
The, problem with ordinary quadrature amplitude modulation (QUAM), when applied to the broadcast band for the transmission of stereo, is mono compatibility. When large amounts of stereo information (L-R) are present in the composite signal, the audio signal recovered by a monophonic receiver's envelope detector becomes not the linear sum of left-plus-right, as it should be, but contains high orders of intermodulation distortion. To counter this problem, Motorola developed compatible QUAM or C-QUAM, and a block diagram of their transmitter arrangement is shown in Fig. 7. The monophonic-equivalent (L+R) signal follows two paths. One goes directly to the transmitter's AM modulator, while the other impresses a crystal-controlled carrier, Fe, with (L + R) modulation through a balanced modulator. At the same time, the (L-R) audio signal is passed through a different balanced modulator which generates the required pure-quadrature modulation. The (L+R) amplitude-modulated carrier and the quadrature-modulated carrier are added together and filtered to form an ordinary QUAM r.f. signal. This QUAM signal is limited (incompatible AM sidebands are removed), amplified, filtered, and delivered as a phase-modulated carrier to the transmitter's r.f. in put. The standard AM transmitter amplitude-modulating this phase modulated carrier with an (L+R) signal, is what actually generates the C-QUAM signal. For stereo identification purposes, the Motorola system uses a 25-Hz tone.
To receive and decode Motorola C-QUAM AM stereo, a de coder such as the one in Fig. 8 would be used. Here, the output of the i.f. amplifier is applied to a carrier level modulator and limiter. The voltage-controlled oscillator (VCO), which is locked in phase with the i.f. carrier, is used with the limiter output to provide input signals to a phase-detector circuit. The phase detector and low-pass filter provide the control signal which maintains the VCO locked in quadrature phase relationship with the i.f. carrier signal The VCO output is shifted 90° to provide a signal in phase with the i.f. carrier. When the phase-shifted VCO signal is used together with a signal from the limiter to supply the phase detector, a signal proportional to cos 0 is derived. The cos A signal is used to supply the carrier level modulator which restores QUAM signals at its outputs. Left and right audio signals can be demodulated by synchronous detectors (balanced modulators), supplied with signals at cos (wct ± π/4) to derive "L" and "R" signals directly, since the signals are QUAM at their inputs.
The Kahn/Hazeltine System
Leonard Kahn, an inventor and engineering consultant based on Long Island, New York, has been involved in AM stereo since the late 1950s. In 1977, he gained the support and endorsement of the well-known Hazeltine Corporation, also of Long Island. Kahn calls his system ISB for Independent Sideband. A block diagram of the transmission side is shown in Fig. 9. Left and right audio signals are again matrixed together and fed through a constant-phase-difference (45°), constant-amplitude network to form an (L+R) signal. At the same time, an (L-R) signal is fed through another phase-shift network (+45°) so that the (L + R) and (L-R) signals are now 90° apart in phase. The (L-R) signal now feeds a summing circuit and a level-squaring circuit. A variable time-delay network adjusts the time delay of the (L-R) path so that it equals the time delay of the (L + R) path. Outputs from the time-delay circuit and the transmitter's crystal oscillator are sent to a phase-modulator/frequency multiplier chain which finally provides a phase-modulated carrier wave, at the correct frequency, to the standard AM transmitter's r.f. input. What may not be immediately apparent from the above description is that in the Kahn/ Hazeltine system, phase modulation of the carrier with (L-R) information occurs in such a way that amplitude modulation of the carrier will place most of the left channel stereo information in the lower sideband of the AM transmission, and most of the right channel stereo information will appear in the upper sideband. To this basic system, Kahn/Hazeltine have added a pilot (identification) tone at 15 Hz which angle-modulates the r.f. carrier by approximately 0.1 radian.
As a result of the distribution of left and right signal energy in the upper and lower sidebands, an interesting side benefit of the Kahn/Hazeltine system is that two ordinary radios, placed apart from each other in the listening room, can be used to receive AM stereo broadcasts transmitted using this system. The left radio is simply tuned slightly below the center station frequency, while the right radio is tuned slightly above center frequency. Tuned in this manner, the two radios will directly demodulate the upper and lower AM sidebands and produce "L" and "R" audio channels. Of course, a more sophisticated approach in the form of a single receiver specifically designed for the system is also possible and is illustrated in the block diagram of Fig. 10.
(Source: Audio magazine, Jul. 1982)
= = = =
Prev. | Next