Another recently encountered older article on synchronous demodulation was “Second Detectors in TV”, in the journal RCA Engineer for 1958 October-November, p.39ff, available at:
http://www.americanradiohistory.com/..._Issue_Key.htm.
It provided a concise and clear treatment of the distortions produced by envelope (diode) demodulation of vestigial sideband signals. Then the article went on to show the improvements that could be obtained by demodulating the in-phase component rather than the envelope. Methods suggested for obtaining in-phase component (IPC) demodulation were enhanced carrier and sampling using a product demodulator. The concluding statement was:
“There is probably little risk in predicting that in the near future we shall see more of enhanced-carrier techniques using envelope detectors to obtain approximate IPC detection. Eventually, with greatly improved frequency memories becoming available, sampling detectors may well become important in home television receiver design.” I have attached a copy of the last page of the article, which also includes some useful diagrams.
In the event, and to the best of my knowledge, enhanced carrier techniques as described in the article were not widely used in domestic TV receivers in the valve era. But “approximate IPC detection” suitable for general use in TV receivers did arrive at the end of the 1960s, albeit in the form of the IC-based quasi-synchronous demodulator, for example the Motorola MC1330. As previously recorded, Motorola’s motivation was not solely improved demodulation. For its vision IF IC set it needed a low-level demodulator, and found this in the recent work of Macario – who was in search of better AM demodulation, and of Sprague, who had outpointed that its ULN2111A FM limiter/demodulator IC could also be used as a low-distortion AM demodulator.
The “afc-locked product detector”, to use terminology from the article, appeared in consumer IC form in the mid-1970s. The RCA CA3136 was the first PLL vision demodulator that I am aware of, although it seems to have been quite obscure. The National LM1821/2/3 of the early 1980s probably did more to put the type “on the map”. Nevertheless, discrete solid-state circuitry for PLL vision demodulation had earlier been used in professional equipment, such as the BBC RC5M/502 TV receiver. Overall the article was quite prescient, although I suspect that the timeline was longer than the author expected. The critical step, for consumer TV receiver applications where cost was so important, was the availability of suitable integrated circuits.
That aspect was emphasized in another article in RCA Engineer for 1958 October-November, namely “High Performance Television Receiver Experiment”, p.80ff. The article may be summarized by this quote from the introductory section:
“This paper describes an experiment that has been performed indicating an approach to an improved, high performance, color-TV receiver which is completely compatible with broadcast standards. The original work was conducted some years ago and served at that time to prove the validity of the principles involved: however, the cost/performance tradeoffs at this early period discouraged development in the consumer TV receiver product environment. The interim developments of integrated circuit technology and the growing availability of lower cost signal processing components may make the development of such a receiver both desirable and practical. This background, coupled with the growing interest in cable TV and high performance color television systems, make a review of the fundamentals worthwhile.”
Quite when the original work was done was not stated in the article, but it was evidently back in the valve era. One of the improvements was the use of synchronous demodulation, amongst others to eliminate cross-modulation in the demodulation process. It was said:
“Any properly operating synchronous detector, such as the integrated circuit units including bridge-circuit balanced diodes with an associated AFC loop, will meet the requirements. However. the particular unit available for test consisted of 6J6, injection locked, product detector operating at the RF level on Channel 4 frequency. Fig. 3 is a diagram of the RF and detector portion of the complete receiver.”
I have attached a copy of the ‘Fig. 3’ referred to. The 6J6-based circuit looks surprisingly simple. The valve choice – the erstwhile 6J6 from the summer of 1942, and one of the very early heater-cathode miniatures - may have been quite specific, as elsewhere in the circuit the newer 12AT7 was used. The experimental receiver was a channel A4 TRF. Two other key features of the experimental receiver were the use of a delay line-based comb filter and sound carrier cancellation rather than notching out. In respect of the latter, it somewhat presaged the principle of the pilot-cancelling technique used in mid-1970s FM stereo decoder ICs from Hitachi and Toko.
Cheers,