UK Vintage Radio Repair and Restoration Discussion Forum - View Single Post - Tuners in Philips G6-G11 CTV chassis

Pieter H · 7th Jul 2018, 3:58 pm

Hi Synchrodyne,

our discussion on where the UK IF choices, or standard-related IF choices in general, came from kept haunting me. Putting it all together I've come to the following observations.

It seems that in the ideal case one would like to position the LO frequency in an "empty" frequency space between channels.
This is most consistently done with the French system-E 819-line standard, where most (Philips) receivers had the LO exactly on the band edge ow the active pair/impair band, while additionally it co-incided with the sound carrier of the opposite band.
This is exactly what is being done for UK CCIR-I. The need to change from the CCIR-G/H 38,9MHz IF was that the sound carrier of N+4 was in comparison 0,5MHz higher. For reasons as yet unclear to me there was apparently a preference to put the LO nearer the PC than the SC.
So for CCIR-I the centre between N+4 SC and N+5 PC would have been 39,0MHz, but it was likely moved to 39,5 to have the N+1 IF sound trap be on the channel A1 sound carrier at 41,5MHz, as discussed earlier.
The interesting thing is that the G/H 38,9MHz is similarly perfect for the standard with 5,5MHz picture-sound distance AT UHF! At the time the 38,9MHz was standardized when only VHF off-air was being transmitted. However, this suggests the 38,9MHz choice was mainly determined by the future UHF use. I don't have the 1954 article by Holm and Werner on this, so can't check whether this assumption is true.
I also looked at the location of the image frequency given the IF choice (so 2*IF above the wanted PC). Here again we see that the G/H and I IF's are perfectly in between a SC and PC. Even stronger, looking at the numbers one could argue that the location of the image frequency EXACTLY between the N+9 SC and N+10 PC was the main design criterium, with the LO frequency and IF then automatically being half that number. That would then be the reason of the 39,5 for CCIR-I: image exactly between the 78 and 80MHz of these two carriers. And similarly for the 38,9MHz for G/H!
The "older" VHF off-air standards from the early 1950s (B and D) probably had their IF choice either determined by UHF (B) or by other practical reasons. Given the much lower number of channels, the above arguments of the LO positioning only applied for the lowest VHF-III channels, in all other cases the LO was outside the received band and the choice thus less critical.
This changed, however, with the introduction of cable S-channels, when all of a sudden the VHF-III band was substantially extended, this could have become an issue. I'm still checking if and how this translated into additional tuner specs.
As to the Italian (Philips) IF of 45,9, it is noteworthy that this is exactly one channel width (7MHz) up from 38,9, and would thus from an interference rejection perspective give the same performance as the standard 38,9. Might have been the reason, but just a theory.

Below is a table I made to summarize all this.
Green fields indicate the LO and image frequencies exactly between a SC and PC. Dark orange fields indicate the LO in the middle of the video band, while light orange are cases where the LO is just between the upper video band edge and the SC. The dark yellow fields in the image column are only an issue in case of cable channels.

I don't pretend this the full or only story, but it is too nice and consistent to be irrelevant for the IF choice. Looking forward to your view on this.

Cheers, Pieter

7th Jul 2018, 3:58 pm	#14
Pieter H Tetrode Join Date: Sep 2017 Location: Waalre, Netherlands Posts: 67	Re: Tuners in Philips G6-G11 CTV chassis Hi Synchrodyne, our discussion on where the UK IF choices, or standard-related IF choices in general, came from kept haunting me. Putting it all together I've come to the following observations. It seems that in the ideal case one would like to position the LO frequency in an "empty" frequency space between channels. This is most consistently done with the French system-E 819-line standard, where most (Philips) receivers had the LO exactly on the band edge ow the active pair/impair band, while additionally it co-incided with the sound carrier of the opposite band. This is exactly what is being done for UK CCIR-I. The need to change from the CCIR-G/H 38,9MHz IF was that the sound carrier of N+4 was in comparison 0,5MHz higher. For reasons as yet unclear to me there was apparently a preference to put the LO nearer the PC than the SC. So for CCIR-I the centre between N+4 SC and N+5 PC would have been 39,0MHz, but it was likely moved to 39,5 to have the N+1 IF sound trap be on the channel A1 sound carrier at 41,5MHz, as discussed earlier. The interesting thing is that the G/H 38,9MHz is similarly perfect for the standard with 5,5MHz picture-sound distance AT UHF! At the time the 38,9MHz was standardized when only VHF off-air was being transmitted. However, this suggests the 38,9MHz choice was mainly determined by the future UHF use. I don't have the 1954 article by Holm and Werner on this, so can't check whether this assumption is true. I also looked at the location of the image frequency given the IF choice (so 2*IF above the wanted PC). Here again we see that the G/H and I IF's are perfectly in between a SC and PC. Even stronger, looking at the numbers one could argue that the location of the image frequency EXACTLY between the N+9 SC and N+10 PC was the main design criterium, with the LO frequency and IF then automatically being half that number. That would then be the reason of the 39,5 for CCIR-I: image exactly between the 78 and 80MHz of these two carriers. And similarly for the 38,9MHz for G/H! The "older" VHF off-air standards from the early 1950s (B and D) probably had their IF choice either determined by UHF (B) or by other practical reasons. Given the much lower number of channels, the above arguments of the LO positioning only applied for the lowest VHF-III channels, in all other cases the LO was outside the received band and the choice thus less critical. This changed, however, with the introduction of cable S-channels, when all of a sudden the VHF-III band was substantially extended, this could have become an issue. I'm still checking if and how this translated into additional tuner specs. As to the Italian (Philips) IF of 45,9, it is noteworthy that this is exactly one channel width (7MHz) up from 38,9, and would thus from an interference rejection perspective give the same performance as the standard 38,9. Might have been the reason, but just a theory. Below is a table I made to summarize all this. Green fields indicate the LO and image frequencies exactly between a SC and PC. Dark orange fields indicate the LO in the middle of the video band, while light orange are cases where the LO is just between the upper video band edge and the SC. The dark yellow fields in the image column are only an issue in case of cable channels. I don't pretend this the full or only story, but it is too nice and consistent to be irrelevant for the IF choice. Looking forward to your view on this. Cheers, Pieter Attached Thumbnails