Quote:
Originally Posted by Synchrodyne
Another related idea of the late 1940s was that of Norgaard of GE, for the generation of ISB signals by wideband AF phase-shifting and quadrature modulation. The proposed receiver was of the locked oscillator type, with I and Q demodulation, audio phase-shifting and dematrixing, recognizable as the basic technique used in more recent HF receivers with PLL fully synchronous selectable sideband demodulation. A brief description of the Norgaard system was provided in Communications magazine for 1948 April, available here: http://www.americanradiohistory.com/...ns-1948-04.pdf, see page 12. |
Quite recently I happened across confirmation that the Norgaard system was actually used in its early days in a commercial product, namely the General Electric (GE) YRS-1 Single Sideband Selector. It was covered by an article in Wireless World for 1948 July, p.244ff, “Single Sideband Selector - Unit for Attachment to Communications Receivers”, which may be found here:
http://www.americanradiohistory.com/...ld-1948-07.pdf.
Further searching then found a similar article in Radio News for 1948 August, p.53ff, “A Single Sideband Selector for Ham Use”:
http://www.americanradiohistory.com/...-1948-08-R.pdf. Interesting is that it was aimed primarily at the amateur market rather than the commercial market, and was considered to be simpler than commercial approaches that included the use of filters. I have attached the page from that article which shows the block schematic. Both articles include the full schematic. Also attached is a GE advertisement for the YRS-1.
Its primary function appears to have been to enable the selection of either sideband of an AM transmission. It also provided exalted carrier demodulation of DSB AM signals, and demodulation of reduced carrier SSB signals. All of these were done fully synchronously using a locked oscillator. It could demodulate suppressed carrier SSB and CW signals, in which case its oscillator served as CIO and BFO respectively. Sideband selection was done by the phase-shifting and matrixing method, not by filters, and was reasonably wideband, 70 to 7000 Hz. Oscillator locking was done via an AFC loop, using the Q demodulator DC output and a reactance valve. Its input was 455 kHz (or thereabouts) IF from the associated communications receiver. All of this required 14 valves, including a power supply with voltage stabilization. Overall, it looks to have been quite an impressive product for the late 1940s, particularly as it was aimed at the consumer market.
I hasten to add that the GE YRS-1 used fully synchronous demodulation, not the quasi-synchronous type. So strictly speaking it is off-centre in this thread, topic-wise, but in practice it is difficult to discuss one without some mention of the other. And perhaps the distinction can be a bit blurred. TMC has been mentioned previously. Its TDRS triple-diversity receiving system used Hammarlund SP600X HF receivers modified inter alia to incorporate Crosby ECC exalted carrier adaptors. The Crosby exalted carrier demodulation system was quasi-synchronous; the carrier was extracted in a narrow filter and then limited and used to drive the demodulators. In the Crosby adaptor as used by TMC, the incoming receiver IF was down converted to a 200 kHz IF for exalted carrier processing. This was done by a local oscillator that had reactance-valve AFC controlled by a discriminator acting on the filtered and limited carrier signal. The primary purpose of this AFC system was to keep the 200 kHz IF closely centred even if the incoming IF was wandering somewhat. It seems to me that the presence of this “long” loop did move the Crosby demodulation system some way along a line drawn between quasi-synchronous and fully synchronous. Not only that, but sometimes both fully-synchronous and quasi-synchronous demodulation were used in the same unit, or even in the same IC. For example the Motorola MC13020 CQUAM AM stereo decoding IC – and its successors – used PLL fully synchronous demodulation to retrieve the I and Q signals, but (wideband) quasi-synchronous demodulation to retrieve the envelope signal.
The GE YRS-1 was a “surprise” find, and reminded me of the maxim that when doing casual and scattered lightweight “research”, one should not be surprised by surprises. This was purely a happenstance find, though, as I was looking at WW 1948 for other reasons and nearly didn’t even look at the GE YRS-1 article. Effectively this unit presaged the solid-state outboard synchronous demodulators for (consumer and semi-professional) HF receivers that were offered in the 1980s. Some of the latter also included quasi-synchronous as well as PLL full synchronous demodulators.
Cheers,