Making 405 line NTSC - UK Vintage Radio Repair and Restoration Discussion Forum

tubesrule · 20th Mar 2007, 4:21 pm

[Split from https://www.vintage-radio.net/forum/...ad.php?t=15601]
Jeff and I have been discusing the possibility of adding this to the low cost converter, and it does look possible with minimal changes, the biggest being the need for a larger fpga. I could develop the code under 525 NTSC so I could tst the output, and then switch to 405 NTSC.

This brings up the need for a spec on the 405 NTSC system. I know the subcarrier frequency, but not the specifics of the burst and such. I can use the modern NTSC specs and this would probably work, but it would be nice to know if there where differences.

Darryl

Ian - G4JQT · 20th Mar 2007, 4:34 pm

Darryl,

Have a look here: http://www.bbc.co.uk/rd/pubs/archive/index.shtml

If you click on "List of BBC Monographs (numbers 1 to 80)" the link takes you to a list. Monographs 18 and 32 (colour TV tests) may give some clues, but I've not read them...

Ian

Dave Moll · 20th Mar 2007, 4:37 pm

Darryl, I shall check in my copy of [405-line] Colour Television The N.T.S.C. System Principles & Practice by Carnt & Townsend. I will contact you by PM.

ppppenguin · 20th Mar 2007, 4:47 pm

On p109 of Carnt and Townsend there's a diagram that shows:

Sync width: 9us
Breezeway: 1.5us
Overall H blanking: 16.5us
Burst length: 9 +/-1 cycles of subcarrier
Burst amplitude: Same as sync amplitude

I haven't managed to find a spec showing on which lines the burst is blanked in the vertical interval. It just say that the burst is not transmitted during field sync pulses. I can't imagine this is too critical, just adapt the 525 NTSC rules.

Subcarrier frequency is 525/2 * fh = 2.6578125MHz

PS: Just found the luminance notch spec on p419. They recommend 6dB attenuation at subcarier with +/-200kHz width at the 3dB points. This was a suggestion only, not a firm spec, but it corresponds pretty well with subsquent industry practice in PAL and NTSC.

tubesrule · 20th Mar 2007, 4:55 pm

Thanks Ian, this is a wonderful resource I was not aware of! Publiction #18 "The BBC Colour Television Tests: An Appraisal of Results" is very informative from a user experience point of view, and #32 "A New Survey of the BBC Experimental Colour Transmissions" has the technical information I was looking for. On brief inspection, it discusses the cross-color notch filter that was employed in these tests. Jeff had mentioned this, and from this reading, it sounds like the benefit was only marginal, so if I implement this, I will make it an option so the user can decide to insert it or not.

Darryl

tubesrule · 20th Mar 2007, 4:58 pm

Thanks Jeff. I think I have enough info to start now. The first thing will be building a modified unit, so it may be a few days before I'm making sparks!

Darryl

ppppenguin · 20th Mar 2007, 5:10 pm

The diagrams on p26 of BBC monograph #32 refer to a 5% lift on the Y signal. I have never properly understood the history of lift (aka setup) in the 405 system. 625 systems never used it while 525 systems have 7.5IRE lift excpet for the Japanese variant which has none.

The 405 system seems to have acquired and discarded lift at least once over the years. I reckon we can live without it.

The only reason I can see for having lift is to help the H flyback blanking of early sets which is either poor or non-existent. I don't beleive it really helped that much and it certainly wasted transmitter power and/or reduced signal to noise ratio.

ppppenguin · 20th Mar 2007, 5:14 pm

I've started a new thread about making 405 line NTSC signals. https://www.vintage-radio.net/forum/...842#post107842

Dave Moll · 20th Mar 2007, 5:36 pm

Further to the posts mentioned above, I have sent scans of page 109 and appendix 1 of Carnt & Townsend to Darryl.

tubesrule · 20th Mar 2007, 5:56 pm

Thanks for starting up a new thread Jeff.

It looks like the burst was on every line in the vertical blanking area with the obvious exception of the vertical sync.

You mentioned lift (setup) as described in the spec. Is this something that should be included? I tend to agree that I see no benefit in this, and it does decrease dynamic range in the output.

Darryl

ppppenguin · 20th Mar 2007, 6:37 pm

Forget about lift (setup). It really isn't worth having. I don't know why NTSC (US version) persisted with it. If it had been abolished in the 1970s it wouldn't have caused any trouble.

tubesrule · 24th Mar 2007, 5:50 pm

I have finished the preliminary work on the color converter and am making great 525 NTSC color pictures. I have looked at the 405 NTSC output, and it looks fine on a monochrome television and on the scope, but I have no way to verify the color portion, although it should be correct since the 525 works, and it is only a chroma subcarrier frequency difference. The cross color filter per the BBC spec is working nicely, and can be switched on and off to compare the difference.

The last thing is deciding what to do with the default image flash. Since there is not enough memory to store a full resolution color image, I am debating wether to leave it as is, storing a full res monochrome image, or change it over to store a half res color image. If I change to half res color, it would be better to cut the horizontal resolution in half, but this is fairly complicated. I could just cut the vertical resolution in half, but this would not look as good.

Darryl

ppppenguin · 24th Mar 2007, 10:21 pm

Darryl is making great progress with the design. One snag is that it's not been possible to use the existing Xilinx programmable gate array 'cos it just ain't big enough. This means that we can't just reprogram existing units to do 405 NTSC. It's actually worse than that. The device in the existing design is a XC3S100E in VQ100 package. This is actually the smallest device in the Spartan 3E series. The data sheet says that the XC3S250E is also available in the VQ100 package but it's proving to be elusive. It's certainly not worth re-doing the entire PCB to use a different package so Darryl and I are hoping to find a few more of these almost mythical devices. He and I both have good relations with Xilinx distributors in our respective countries but if anyone knows something that we don't I'm sure Darryl would appreciate the information.

tubesrule · 26th Mar 2007, 1:13 pm

The 405 NTSC converter is finished. I have updated the manual to include this model, and have added it to the list of available converters on the website, although it will remain as a special order item until I can sort out the availability of the larger fpga required.

Although I have not been able to visually verify the output on an actual 405 NTSC set, I did verify it on a standard 525 NTSC set when programmed for it's timing. I further looked at the 405 output with a standard color bar input on a scope, and the frequency, phase and amplitude of the chroma appeared to be correct.

The unit has a selectable cross color filter that adds a notch in the luma channel at the chroma subcarrier frequency. The filter characteristics where designed to match the original BBC spec. It also has a selectable color/monochrome mode which allows it to act like a standard 405 monochrome converter if the chroma is not desired. This can be useful when using this converter on a B&W 405 set that does not have a chroma subcarrier trap resulting in interference in the picture.

Darryl

chipp1968 · 28th Mar 2007, 8:57 am

Very clever . Now we need to see the working set ! how many exactly are there ? 1 1/2?

FERNSEH · 28th Mar 2007, 10:45 am

Quote:

Originally Posted by chipp1968

Very clever . Now we need to see the working set ! (

And you will soon.

DFWB.

channel405 · 30th Mar 2007, 11:31 am

This is all great news!!

Now that colour 405 signals is about to be obtainable for the masses, the next question is a suitable receiver. As most of us won't have access to the very few original 405 colour sets, can any one recommend a suitable 625 set for conversion? I had in mind a Decca Bradford and by passing the IF stage. Would this be as simple as changing the xtal, removing the PAL delay line in the decoder and lowering line frequency? I know changing line frequency will have horrible effects on the convergence but other than having a dual standard colour set are there better solutions?

murphyv310 · 31st Mar 2007, 7:32 pm

Quote:

Originally Posted by channel405

This is all great news!!

I had in mind a Decca Bradford and by passing the IF stage. Would this be as simple as changing the xtal, removing the PAL delay line in the decoder and lowering line frequency? I know changing line frequency will have horrible effects on the convergence but other than having a dual standard colour set are there better solutions?

Not that simple I am afraid!
The easiest set to modify would be dual standard sets, 625 line stages do not like to run as slow as 10125 khz they produce enormous amount of EHT at those low frequencies so the simple way of doing this is a dual standard set. You then have a good time base and suitable IF strip.
It would be nice to just change the crystal and whip out the delay line, sadly no, although PAL is loosely based on NTSC there are large enough differences to warrant a redesigned decoder.
I have ordered a 405 line NTSC converter from darryl hock and I hope to track down the 405 line colour set that was in Kilmarnock Tech 30 + years ago and also convert a dual standard colour set in the near future but not a job for the faint hearted.
Trevor

Panrock · 31st Mar 2007, 7:58 pm

What about getting an old American colour set on Ebay, changing the crystal and modifying the line/output and frame timebase ?

Steve

ppppenguin · 31st Mar 2007, 8:03 pm

Modifying the line output would be just as hard as on a 625 UK set. 525 line scan is 15734Hz.

A dual standard set is the way to go. I can see no showstopping problems in modifying a decoder. Remove the PAL switch and delay line. Get a subcarrier Xtal cut, retune the various tuned circuits and possibly lengthen the Y delay line. This last item is the hardest but it probably wouldn't be necessary. No chips in sets of this generation so you canget at everything. A Thorn 2000 with its pluggable decoder strikes me as ideal.

20th Mar 2007, 4:34 pm	#2
Ian - G4JQT Octode Join Date: Jun 2004 Location: Reading/Fakenham, UK. Posts: 1,323	Re: Pye 405 line colour TV Darryl, Have a look here: http://www.bbc.co.uk/rd/pubs/archive/index.shtml If you click on "List of BBC Monographs (numbers 1 to 80)" the link takes you to a list. Monographs 18 and 32 (colour TV tests) may give some clues, but I've not read them... Ian

20th Mar 2007, 4:47 pm	#4
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Pye 405 line colour TV On p109 of Carnt and Townsend there's a diagram that shows: Sync width: 9us Breezeway: 1.5us Overall H blanking: 16.5us Burst length: 9 +/-1 cycles of subcarrier Burst amplitude: Same as sync amplitude I haven't managed to find a spec showing on which lines the burst is blanked in the vertical interval. It just say that the burst is not transmitted during field sync pulses. I can't imagine this is too critical, just adapt the 525 NTSC rules. Subcarrier frequency is 525/2 * fh = 2.6578125MHz PS: Just found the luminance notch spec on p419. They recommend 6dB attenuation at subcarier with +/-200kHz width at the 3dB points. This was a suggestion only, not a firm spec, but it corresponds pretty well with subsquent industry practice in PAL and NTSC.

20th Mar 2007, 4:55 pm	#5
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Re: Pye 405 line colour TV Thanks Ian, this is a wonderful resource I was not aware of! Publiction #18 "The BBC Colour Television Tests: An Appraisal of Results" is very informative from a user experience point of view, and #32 "A New Survey of the BBC Experimental Colour Transmissions" has the technical information I was looking for. On brief inspection, it discusses the cross-color notch filter that was employed in these tests. Jeff had mentioned this, and from this reading, it sounds like the benefit was only marginal, so if I implement this, I will make it an option so the user can decide to insert it or not. Darryl

20th Mar 2007, 4:58 pm	#6
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Re: Pye 405 line colour TV Thanks Jeff. I think I have enough info to start now. The first thing will be building a modified unit, so it may be a few days before I'm making sparks! Darryl

20th Mar 2007, 5:10 pm	#7
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Pye 405 line colour TV The diagrams on p26 of BBC monograph #32 refer to a 5% lift on the Y signal. I have never properly understood the history of lift (aka setup) in the 405 system. 625 systems never used it while 525 systems have 7.5IRE lift excpet for the Japanese variant which has none. The 405 system seems to have acquired and discarded lift at least once over the years. I reckon we can live without it. The only reason I can see for having lift is to help the H flyback blanking of early sets which is either poor or non-existent. I don't beleive it really helped that much and it certainly wasted transmitter power and/or reduced signal to noise ratio.

20th Mar 2007, 4:21 pm	#1
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Making 405 line NTSC [Split from https://www.vintage-radio.net/forum/...ad.php?t=15601] Jeff and I have been discusing the possibility of adding this to the low cost converter, and it does look possible with minimal changes, the biggest being the need for a larger fpga. I could develop the code under 525 NTSC so I could tst the output, and then switch to 405 NTSC. This brings up the need for a spec on the 405 NTSC system. I know the subcarrier frequency, but not the specifics of the burst and such. I can use the modern NTSC specs and this would probably work, but it would be nice to know if there where differences. Darryl Last edited by Duke_Nukem; 21st Mar 2007 at 3:03 pm. Reason: Split thread

20th Mar 2007, 4:37 pm	#3
Dave Moll Dekatron Join Date: Feb 2005 Location: West Cumbria (CA13), UK Posts: 6,127	Re: Pye 405 line colour TV Darryl, I shall check in my copy of [405-line] Colour Television The N.T.S.C. System Principles & Practice by Carnt & Townsend. I will contact you by PM. __________________ Mending is better than Ending (cf Brave New World by Aldous Huxley)

20th Mar 2007, 5:14 pm	#8
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Pye 405 line colour TV I've started a new thread about making 405 line NTSC signals. https://www.vintage-radio.net/forum/...842#post107842

20th Mar 2007, 5:36 pm	#9
Dave Moll Dekatron Join Date: Feb 2005 Location: West Cumbria (CA13), UK Posts: 6,127	Re: Making 405 line NTSC Further to the posts mentioned above, I have sent scans of page 109 and appendix 1 of Carnt & Townsend to Darryl. __________________ Mending is better than Ending (cf Brave New World by Aldous Huxley)

20th Mar 2007, 5:56 pm	#10
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Re: Making 405 line NTSC Thanks for starting up a new thread Jeff. It looks like the burst was on every line in the vertical blanking area with the obvious exception of the vertical sync. You mentioned lift (setup) as described in the spec. Is this something that should be included? I tend to agree that I see no benefit in this, and it does decrease dynamic range in the output. Darryl

20th Mar 2007, 6:37 pm	#11
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Making 405 line NTSC Forget about lift (setup). It really isn't worth having. I don't know why NTSC (US version) persisted with it. If it had been abolished in the 1970s it wouldn't have caused any trouble.

24th Mar 2007, 5:50 pm	#12
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Re: Making 405 line NTSC I have finished the preliminary work on the color converter and am making great 525 NTSC color pictures. I have looked at the 405 NTSC output, and it looks fine on a monochrome television and on the scope, but I have no way to verify the color portion, although it should be correct since the 525 works, and it is only a chroma subcarrier frequency difference. The cross color filter per the BBC spec is working nicely, and can be switched on and off to compare the difference. The last thing is deciding what to do with the default image flash. Since there is not enough memory to store a full resolution color image, I am debating wether to leave it as is, storing a full res monochrome image, or change it over to store a half res color image. If I change to half res color, it would be better to cut the horizontal resolution in half, but this is fairly complicated. I could just cut the vertical resolution in half, but this would not look as good. Darryl

24th Mar 2007, 10:21 pm	#13
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Making 405 line NTSC Darryl is making great progress with the design. One snag is that it's not been possible to use the existing Xilinx programmable gate array 'cos it just ain't big enough. This means that we can't just reprogram existing units to do 405 NTSC. It's actually worse than that. The device in the existing design is a XC3S100E in VQ100 package. This is actually the smallest device in the Spartan 3E series. The data sheet says that the XC3S250E is also available in the VQ100 package but it's proving to be elusive. It's certainly not worth re-doing the entire PCB to use a different package so Darryl and I are hoping to find a few more of these almost mythical devices. He and I both have good relations with Xilinx distributors in our respective countries but if anyone knows something that we don't I'm sure Darryl would appreciate the information.

26th Mar 2007, 1:13 pm	#14
tubesrule Hexode Join Date: Jul 2004 Location: Michigan USA Posts: 325	Re: Making 405 line NTSC The 405 NTSC converter is finished. I have updated the manual to include this model, and have added it to the list of available converters on the website, although it will remain as a special order item until I can sort out the availability of the larger fpga required. Although I have not been able to visually verify the output on an actual 405 NTSC set, I did verify it on a standard 525 NTSC set when programmed for it's timing. I further looked at the 405 output with a standard color bar input on a scope, and the frequency, phase and amplitude of the chroma appeared to be correct. The unit has a selectable cross color filter that adds a notch in the luma channel at the chroma subcarrier frequency. The filter characteristics where designed to match the original BBC spec. It also has a selectable color/monochrome mode which allows it to act like a standard 405 monochrome converter if the chroma is not desired. This can be useful when using this converter on a B&W 405 set that does not have a chroma subcarrier trap resulting in interference in the picture. Darryl

28th Mar 2007, 8:57 am	#15
chipp1968 Rest in Peace Join Date: Jul 2006 Location: Somerset, UK. Posts: 2,356	Re: Making 405 line NTSC Very clever . Now we need to see the working set ! how many exactly are there ? 1 1/2?

30th Mar 2007, 11:31 am	#17
channel405 Tetrode Join Date: Jun 2004 Location: CH3, Blaen Plwyf Transmitter, Aberystwyth, WALES Posts: 72	Re: Making 405 line NTSC This is all great news!! Now that colour 405 signals is about to be obtainable for the masses, the next question is a suitable receiver. As most of us won't have access to the very few original 405 colour sets, can any one recommend a suitable 625 set for conversion? I had in mind a Decca Bradford and by passing the IF stage. Would this be as simple as changing the xtal, removing the PAL delay line in the decoder and lowering line frequency? I know changing line frequency will have horrible effects on the convergence but other than having a dual standard colour set are there better solutions? __________________ 405 line TV, 1936- 1985 2007.........to the present day

31st Mar 2007, 7:58 pm	#19
Panrock Nonode Join Date: Feb 2005 Location: Worcestershire, UK. Posts: 2,533	Re: Making 405 line NTSC What about getting an old American colour set on Ebay, changing the crystal and modifying the line/output and frame timebase ? Steve __________________ https://www.radiocraft.co.uk

31st Mar 2007, 8:03 pm	#20
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Making 405 line NTSC Modifying the line output would be just as hard as on a 625 UK set. 525 line scan is 15734Hz. A dual standard set is the way to go. I can see no showstopping problems in modifying a decoder. Remove the PAL switch and delay line. Get a subcarrier Xtal cut, retune the various tuned circuits and possibly lengthen the Y delay line. This last item is the hardest but it probably wouldn't be necessary. No chips in sets of this generation so you canget at everything. A Thorn 2000 with its pluggable decoder strikes me as ideal.