Quote:
Originally Posted by SteveCG
BTW I wonder whether Dolby B applied only to the stereo sub-carrier component might have assisted fringe area reception?
|
That I think would be debatable.
The partial compression of the subcarrier, along with the probable use of a different pre-emphasis constant (25 µs), would have reduced the extent to which the baseband and subcarrier interleaved. This would have required lower levels of maximum modulation for each in order to avoid exceeding the ±67.5 kHz maximum. Perhaps there could have been a trade-off, with a lesser reduction of modulation depth than nominally required accompanied by more severe peak limiting. But the effect of a modulation depth reduction would have been to reduce the mono service area for a given transmitter power output. Maybe, with partial compression of the subcarrier, the stereo service area would have been a higher fraction of the mono service area than for the basic Zenith-GE system, but with minimal, if any net increase in actual service area. Applying a greater maximum modulation depth reduction to the subcarrier than to the baseband might have helped, but at the expense of reasonable compatibility with non-Dolby equipped receivers that were not configured to handle the difference.
One could outpoint that the American Zenith-DBX TV stereo sound system did use a highly compressed subcarrier. In that case though, the subcarrier was allocated an aliquot of ±50 Hz deviation, additional to the ±25 kHz allocated to the baseband. In the TV sound case, there was plenty of bandwidth available to do this, whereas in the FM case, the ±75 kHz maximum deviation was a fairly hard limit.
Incidentally, with Dolby on the sub-carrier only, FDM decoding would be required, rather than TDM, which was the modal choice for the receivers and most decoding ICs. The Philips TDA1005A could do both TDM and FDM, but off-hand I don’t know whether it allowed access to the raw demodulated sub-carrier for separate processing. But the National LM1884, designed for American TV sound, did.
Given the ±75 kHz maximum deviation constraint, in turn derived from channel allocations in the FN band, and the resultant established receiver design, I don’t think that there was a “magic bullet” that would have solved the stereo “problems”. The reasons were well-explained by RW in post #11. Additionally, a graphic illustration was provided by Keller (*):
I have taken Keller’s graph (d) and roughly added in the Zenith-GE subcarrier, with pre- and de-emphasis also shown:
One may see that whatever kind of subcarrier was used, it was going to be in a much higher noise zone than the baseband. Having the subcarrier at a low a frequency as reasonably possible was clearly desirable. One may also see the attraction of using only the lower sideband of an AM or AMSC subcarrier. However, transmitting this way would mean losing interleaving, which was probably worth around 6 dB. Also, it would have been impracticable with the realizable consumer technology of the day. Using only the LSB (of a DSB subcarrier) at the receiving end would be a way of taking advantage of its lower noise without giving up interleaving at the transmitting end, and as mentioned previously was actually done.
Maybe therein is a challenge for the those interested in FM stereo DX - design and build a decoder that uses only the LSB of the subcarrier, and so ascertain what improvement may be obtained by this means. I imagine that using some of the AM stereo decoding techniques, including quadrature demodulation and AF phase-shifting for ISB sideband separation would have been applicable. As an example, The Hershberger synchronous detector (Popular Electronics 1982 April p.61ff), adapted for 38 kHz (rather than 455 kHz) operation and for an external carrier input (19 kHz) might be a starting point for circuitry that extracts the LSB from the subcarrier.
(*) V.H.F. Radio Manual; P.R. Keller; Newnes, 1957; see p.29.
Cheers,