Hi Pieter:
Thanks for another most interesting article.
The background to the PC88 was particularly welcome, as I haven’t seen very much on this previously.
It would appear that all of the Philips European valved UHF tuners were of the three-gang type, with an aperiodic input to the RF amplifier. As a corollary, only the UK versions had the fourth gang for a tuned RF input, as necessitated by the significant increase in image rejection following the decision to move the IF from 38.9 to 39.5 MHz, whilst still retaining co-siting of n and (n+10) channel transmitters. I suspect that the noise factor suffered a little because of the tuned input. Here is the data for the Mullard AT6360-02:
Mullard AT6360-02.pdf
The AGC curves for the PCC189 and EF183 case are interesting. It looks as if the AGC bias for the PCC189 had around a 15-volt delay, with the slope then close to that of the EF183 bias curve. For the PCC84 and EF85 case, for system A, Mullard used a 3.5-volt delay for the PCC84 relative to the EF85, both AGC bias curves having the same slope. It also provided for AGC application to an EF80 2nd IF stage (which would have handled vision only), not delayed, but scaled down to be at -5 volts when the EF183 was at -13.5 volts, so having a lower slope. Optionally, for a PCC84 RF amplifier with EF80 1st and 2nd RF stages, there was a 2-volt delay for the PCC84, with the EF80 bias scaled to be at -5 volts when the PCC84 was at -10 volts. I think the latter would have applied in cases where the sound take-off was immediately after the tuner, and not after the 1st IF stage.
For the negative modulation case, with a PCC88 RF amplifier and EF80 IF amplifiers, Mullard showed the RF amplifier AGC with a 3-volt delay relative to the IF amplifier AGC, but a much steeper slope, so that both reached -5 volts at the same time.
I haven’t seen any information on how Mullard used the PCC89 and PCC189 in System A applications, but I’d guess that it was with sliding bias for the upper triode with a resultant reasonably wide grid base, similar to that of the PCC84 with sliding bias. Another unknown (to me) is the nature of the AGC bias (delay and slope) applied to the PCF86 and PCF801 pentode parts when used as UHF IF amplifiers.
In respect of IF choices for the Belgian and French multi-system receivers, it appears that French System L was accommodated with a 39.9 MHz picture IF, which allowed the sound to be at 33.4 MHz, the same as for Systems B, C and F. In some cases, System E picture IF stayed at 38.9 MHz, with sound at 27.75 MHz, but in others, it moved up to 39.9 MHz, the same as for System L, with sound thus moving up to 28.75 MHz. A post-facto rationalization of this change would be that the Nyquist slope over 39.9 MHz was gentler (±1.25 MHz for System L) than that at 38.9 MHz (±0.75 MHz) and so closer to the ±2.0 MHz that was ideal for System E. Also, assuming that lower edge bandstop events at and below 33.4 MHz were the same as previously, there was an extra 1 MHz of video bandwidth gained.
Cheers,