21-02-2018, 07:01 PM
I suspect that the UK 625 TV vision IF of 39.5 MHz had not been established at the time that the ‘1961’ Pye receiver was designed. Furthermore, I speculate that that Pye receiver case may have been a catalyst for reconsideration of the UK IF.
The reasoning may have been that if the IF could be moved upwards enough, then a dual-Nyquist basic IF bandpass might have been feasible. Clearly, avoiding running into channel B1 was a fairly hard upper limit. But the image rejection requirements for the (n+10) channel probably determined the practical upper limit. The use of co-sited n and (n+10) channels was a given, from the ITU Stockholm 1961 (ST61) planning meeting, and in the UK case, this probably assumed a 38.9 MHz IF. As the IF moved closer to 40 MHz, the spacing of the (n+10) channel, the image rejection requirement increased steeply. Evidently 39.5 MHz was as high as it could be pushed without going beyond reasonably achievable image rejection requirements. It also had the side benefit that the lower adjacent channel vision came out at 41.5 MHz, the same as the channel B1 sound carrier, for which a rejection notch would be desirable anyway.
As you have observed, the 39.5 MHz IF allowed a dual-Nyquist allowed a 625 vision bandwidth of 4.85 MHz, not that good but probably enough for the less scrupulous of the UK setmakers. The consequent higher image rejection requirement typically required an extra RF tuned circuit in the UHF front end, in the valve era normally added ahead of the RF amplifier. Established European UHF tuner practice was to use an aperiodic input for lowest noise, with a bandpass interstage. (And Cyldon’s first UHF tuner version was three-gang), so the definitive UK tuners probably incurred a noise penalty from the tuned input. But by the time that they were required, the EC88/PC88 valve had arrived, offering an improvement of the EC86/PC86, which was originally used for both the RF and mixer functions. I should not be surprised if the 1960 and 1961 Pye receivers at interest had 3-gang UHF tuners.
In the early solid-state era, whilst the four-gang tuner with single-tuned input and bandpass interstage was probably the most common type, there were variations. The Thorn 2000 and 3000 models come to mind. The 2000 had an aperiodic input, and used the 4th gang for a tuned image rejector at the input. The 3000 also had an aperiodic input, with a triple bandpass interstage.
Returning to the UK IF, the earliest mention of the 39.5 MHz number that I have found is in WW 1962 October:
Perhaps I missed it, but I could not find any earlier mention in WW, nor any mention in Practical Television for the 1961-62 period. Yet I imagine that there was a BREMA missive issued once the decision was made.
Cheers,
Steve
The reasoning may have been that if the IF could be moved upwards enough, then a dual-Nyquist basic IF bandpass might have been feasible. Clearly, avoiding running into channel B1 was a fairly hard upper limit. But the image rejection requirements for the (n+10) channel probably determined the practical upper limit. The use of co-sited n and (n+10) channels was a given, from the ITU Stockholm 1961 (ST61) planning meeting, and in the UK case, this probably assumed a 38.9 MHz IF. As the IF moved closer to 40 MHz, the spacing of the (n+10) channel, the image rejection requirement increased steeply. Evidently 39.5 MHz was as high as it could be pushed without going beyond reasonably achievable image rejection requirements. It also had the side benefit that the lower adjacent channel vision came out at 41.5 MHz, the same as the channel B1 sound carrier, for which a rejection notch would be desirable anyway.
As you have observed, the 39.5 MHz IF allowed a dual-Nyquist allowed a 625 vision bandwidth of 4.85 MHz, not that good but probably enough for the less scrupulous of the UK setmakers. The consequent higher image rejection requirement typically required an extra RF tuned circuit in the UHF front end, in the valve era normally added ahead of the RF amplifier. Established European UHF tuner practice was to use an aperiodic input for lowest noise, with a bandpass interstage. (And Cyldon’s first UHF tuner version was three-gang), so the definitive UK tuners probably incurred a noise penalty from the tuned input. But by the time that they were required, the EC88/PC88 valve had arrived, offering an improvement of the EC86/PC86, which was originally used for both the RF and mixer functions. I should not be surprised if the 1960 and 1961 Pye receivers at interest had 3-gang UHF tuners.
In the early solid-state era, whilst the four-gang tuner with single-tuned input and bandpass interstage was probably the most common type, there were variations. The Thorn 2000 and 3000 models come to mind. The 2000 had an aperiodic input, and used the 4th gang for a tuned image rejector at the input. The 3000 also had an aperiodic input, with a triple bandpass interstage.
Returning to the UK IF, the earliest mention of the 39.5 MHz number that I have found is in WW 1962 October:
Perhaps I missed it, but I could not find any earlier mention in WW, nor any mention in Practical Television for the 1961-62 period. Yet I imagine that there was a BREMA missive issued once the decision was made.
Cheers,
Steve
