[Synchrodyne]

Re the issue of RF amplifiers (or not) for UHF TV tuners during the valve era, was there much net advantage for the European approach as compared with the standard American approach?

The standard American approach was predicated on the use of a bandpass tuned input circuit followed by a passive diode mixer followed by a low-noise IF amplifier, the last-mentioned usually being the RF amplifier of the VHF tuner, and so equipped with (delayed) AGC.

The European (German) approach was based upon the use of a triode self-oscillating mixer. Since this would have been a lot noisier than a diode mixer, it would have needed to have been preceded by a low-noise grounded-grid RF amplifier just to match performance of the American configuration. Whether it was any better in terms of noise factor I don’t know. But presumably it was better in terms of blocking the oscillator signal from getting back to the aerial input.

With a noisy mixer, there was no point in feeding the UHF IF into the (low-noise) VHF RF amplifier, so instead it went into the VHF mixer which then operated as an IF amplifier. And the overall gain at UHF provided by the RF amplifier and active mixer meant that the additional gain that would have been provided by the VHF RF stage was less needed. Perhaps though the deletion of an AGC’d stage was the reason for the introduction (in Europe) of triode-pentode VHF mixers with remote cutoff pentodes that could be connected to the RF AGC line.

I think that German practice was to use a broadband aerial input (presumably for lowest noise) to the RF amplifier, with a bandpass interstage. For UK application, an additional tuned RF stage was necessary, so that a single-tuned rather than aperiodic input to the RF amplifier was used. This was evidently necessitated by the BREMA decision to move the IF upwards from 38.9 to 39.5 MHz, whilst retaining the same European channel matrix as previously agreed at the ITU Stockholm 1961 meeting, including the co-siting of channels n and (n+10) transmitters. Thus, the required image rejection became 53 dB, whereas it would have been 29 dB with the standard 38.9 MHz IF (which was quite suitable for System I, and for example was used in South Africa).

The American standard valved UHF tuner might not have been a product of necessity, though. Quite early on the US tubemakers had developed a range of valves that could be used for the various UHF tuner functions, and were designed with the consumer application in mind, rather than being industrial products pressed into consumer service. By mid-1952 GE was offering the 6AJ4 as a grounded grid RF amplifier, the 6AM4 as a grounded-grid mixer, and the 6AF4 as an oscillator. (I think that the 6AF4 was originally developed by RCA, from the 6F4 acorn valve.) And by the end of 1952 Sylvania was offering the 6AN4 for use as an RF amplifier and as a mixer, and the 6T4 for use as a local oscillator. RCA added the 6BC4 RF amplifier by mid-1954. So, one might infer that it was the tuner makers and setmakers who opted for the simpler form of UHF tuner, used in tandem with VHF tuners. Evidently the benefits that accrued from using RF amplifiers and/or active mixers did not justify the extra costs. The 6AF4, later succeeded by the 6AF4A, became the most common American UHF oscillator valve, but the others did not achieve any prominence.

The 6AJ4 became the EC84 in Europe, and I understand that from this was developed the EC86/PC86 for European UHF tuner use. Originally the EC86/PC86 was intended to cover both the RF amplifier and mixer applications, but the EC88/PC88 was then developed for better RF amplifier performance. The European choice of self-oscillating UHF mixer does not seem to have had an American precedent. It might have been done for cost reasons, to avoid the need for a third UHF valve; in general, self-oscillating mixers were seen as being not as good as the externally excited type.

The simple American UHF tuner design was evidently viewed as being satisfactory, as it was retained until the early 1970s. In the mid-1960s, transistor oscillators replaced the UHF triode oscillators, but the nature of the UHF signal path remained basically the same. (In the USA, the move to solid-state VHF tuners was slowish until suitable dual-gate mosfets became available, given that circuits based upon bipolar devices had rather poor cross-modulation performance.)

It would appear that the advent of varactor tuning in the early 1970s was the driver of changes in US UHF tuner design. The early varactors were lossy as compared with their variable capacitor predecessors, and the use of low-noise RF amplifiers became desirable to offset these losses. Typically, a grounded base amplifier, with single-tuned input and agc was added ahead of the bandpass interstage and crystal diode mixer. The extra tuned circuit might have been desirable to offset the possibly lower Q of early varactors. From circa 1975 suitable UHF-capable dual-gate mosfet RF amplifiers became available, and these offered a step improvement in cross-modulation performance.

At the same time, the advent of varactor tuning in VHF tuners made it less convenient to route the UHF IF signal via the RF amplifier. Rather it would go direct to the VHF mixer transistor, bypassing the VHF varactor circuitry. This change required extra gain in the UHF signal path, typically achieved with a grounded base stage following the UHF mixer and included in the UHF tuner, although sometimes I think within the VHF tuner (which then allowed said VHF tuner to be used with a conventional UHF tuner if desired). An alternative approach, used by Sylvania at least, was to use an active (transistor) mixer, still fed by a separate transistor mixer. That the agc’d VHF RF amplifier was no longer in the UHF signal path was probably the reason why the “new” UHF RF amplifiers were agc’d, in order to retain the full agc range.

Early European solid-state UHF tuners appeared to follow the same basic pattern as their valved predecessors, although sometimes with the addition of agc for the RF amplifier. So, they had grounded base RF amplifiers followed by grounded base self-oscillating mixers. In the UK case, there were some variations in the distribution of the RF selectivity, for example an aperiodic input with the fourth gang used for a tuned image rejector, and as another example, an aperiodic input with a triple-bandpass interstage.


Cheers,