[Synchrodyne]

The Marconi H2900 may well have been a nodal point in the convergence of point-to-point ISB and general coverage receivers. And I have now managed to find some information on the H2900, namely the (brief) article in SWM 1970 April.

As Radio Wrangler has already said, it was clearly of the traditional “3-IF-channel” ISB form. According to a comment in Hawker’s HF receiver article in Wireless World 1970, it had PLL AM demodulation. Thus it might have been one of the first to use this approach for reconditioned carrier generation rather than the established quasi-synchronous technique using filtered and limited carrier.

Frequency coverage of the H2900 was said to be 1.5 to 30 MHz. This was consistent, even slightly better than established ISB receiver practice, examples from the valve era being the Marconi HR21 family at 3 to 27.5 MHz, and the Mullard GFR 552, at 4 to 30 MHz. The then-recent solid-state Marconi Hydrus covered 1.5 to 30 MHz.

The H2900 was not quite “general coverage”, the norm for which one might take as being 0.5 to 30 MHz. But on the other hand, there were earlier Marconi general-purpose HF receivers that did not tune below the top of the MF broadcast band. For example, the HR120 of 1960 covered 2.1 to 30 MHz. And the “Radio & Television Engineer’s Reference Book, 3rd Edition, chapter 22, includes a description of an unidentified HF receiver that was probably a Marconi model. The chapter was written by a Marconi staffer, and the valve line-up is very Marconi, with for example a Z77 (EF91) 1st RF stage, W77 (EF92) 2nd RF stage, X77 (6BE6) 1st & 2nd mixers, W77 IF stages, just like the HR24. This receiver covered 1.5 to 30 MHz, and was dual-conversion, with IFs of 1.2 MHz and 100 kHz. So the H2900 frequency coverage was aligned with long-established Marconi precedent for general-purpose HF receivers.

The Marconi H2900 delivered its multi-mission capability in just one box, albeit a very expensive one. It arrived on the scene about two years after the Hydrus solid-state ISB receiver. The trade press advertisements for the latter show it to occupy four rack bays in dual-diversity form, so presumably the basic non-diversity option would have taken three bays. And the Hydrus itself was much more compact than its valved predecessors such as the HR21 family.

On the question of frequency coverage, and pertinent to the convergence theory, is the fact that hitherto, marine main receivers had been somewhat separate from general coverage receivers because of their need to tune down to 15 kHz or thereabouts, as well as their need for certification against the pertinent GPO and other regulations. But this distinction had started fading by the late 1960s. The Plessey PR155 general coverage receiver of 1967 tuned down to 15 kHz, although I do not know whether it was ever qualified as a marine main receiver (or whether it had an ISB option). The Eddystone EC958 of 1968-69 covered 10 kHz to 30 MHz, and was designed with marine applications in mind. The Marconi Nebula marine receiver was a rebrand of the EC958/5; this might have provided a lower cost alternative to its own Apollo marine receiver.

The Apollo seems to have arrived on the scene at about the same time as the Hydrus ISB receiver, so at that time Marconi was still treating the marine and point-to-point cases as being separate. Whether it also had a contemporary solid-state general coverage receiver, perhaps as a successor to the HR120, I do not know. But the Apollo was also described as being a general-purpose receiver, so it might also have been sold into non-marine applications. (I think that might also have happened with its Atalanta valved predecessor, which apparently had a model number in the main Marconi series as well as in the marine series.)

So the H2900 did not quite connect all of the dots. That step appears to have come in 1974, with the Racal RA1772, which covered 15 kHz to 30 MHz, and for which a 3-IF-channel ISB variant was offered from the start. And at least judging by the Radio Bygones “Marconi Mishap” article, the RA1772 was about a quarter of the price of the H2900. The Eddystone EC958/12 also I think dated from 1974 or thereabouts. This was an EC958/7 (an update of the original) with “add-on” ISB equipment, so that it had four IF channels in all.

Subsequent general-purpose HF receivers, such as the Plessey PR2250, and the Redifon R551, appeared to include ISB options almost as a matter of routine.

An assumption here is that the general-purpose HF receivers with ISB options also came with a range of ISB filters according to customer requirements. The Marconi valve-era point-to-point receivers seemed to have offered both 100 Hz to 3.5 kHz and 100 Hz to 6 kHz passbands, the former for voice traffic (and whatever else used a voice channel) and the latter either for broadcast relays or to accommodate two frequency-multiplexed voice channels. The 100 Hz lower limit might well have been determined by broadcast relay needs, which would point to go as low as reasonably feasible whilst still adequately rejecting the carrier. I imagine that anything much above 100 Hz would cause a significant decline of “naturalness: for voices and be prejudicial to music. On the other and data for the Eddystone EC958/12 indicate that the baseline sideband filters were of the voice traffic type, and probably followed the normal voice bandwidth, 300 Hz to 3.4 kHz, or something close to it.

Cheers,