|
Television Standards Converters, Modulators etc Standards converters, modulators anything else for providing signals to vintage televisions. |
|
Thread Tools |
27th Jun 2010, 6:18 pm | #21 |
Nonode
Join Date: Aug 2007
Location: South Bradford, West Yorkshire, UK.
Posts: 2,571
|
Re: Test Card / Teletext Generator - Phase 1
About 30 years ago I designed a 625 line sync pulse generator using LSTTL. This used a 12MHz clock and a synchronous counter to generate the line and field sync pulses and all the line and field drive and blanking pulses. It used 25 LS devices on a 4" x 6" double sided PCB. The timings were within the tolerances permitted by the appropriate specification.
It also generated burst gating pulses although it was only ever used for monochrome. I did work out a method of generating the colour subcarrier that would have been locked to the line sync, or should it have been the other way round, but never got round to fully building it. I just got as far as getting a circuit breadboarded that would generate the 25Hz offset present in PAL. If you are interested I can dig out the circuits, scan them and post them. Also many of the processors used for digital TV receivers have the capability to generate a PAL signal and include teletext but they are incredibly complex to program. Keith |
27th Jun 2010, 11:15 pm | #22 | |
Retired Dormant Member
Join Date: Sep 2008
Location: London, UK.
Posts: 117
|
Re: Test Card / Teletext Generator - Phase 1
If it was only 30 years ago it will still probably be more modern than what I've knocked up
Quote:
Wouldn't they then be designed for digital text decoding, rather than analogue encoding? Bit of a rhetorical question really because, as you mention, I wouldn't want to get into the programming aspects. Basically, I don't intend the hardware to develop much past a solid state data stream player, as that should be all that's required, thus avoiding any hideously complicated chippery Parabola |
|
28th Jun 2010, 8:07 am | #23 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Test Card / Teletext Generator - Phase 1
While it would be interesting to see Keith's design it's not very relevant to Parabola's ideas. His basic aim is effectively an arbitrary waveform generator, based on a large memory. This is how the Dinosaur and Richard Russell TCgens worked.
I think I still have an SPG that I designed when I was 18, about 35 years ago. When I look back, the methods I used were crude. I certanly didn't understand synchronous design back then. Think of a counter, decoded by gates to give pulse start and stop times and feeding cross coupled gates acting as flipflops. Hardly surprising it needed a few caps to suppress glitches. If I had used dual JK FFs instead of cross coupled gates the whole thing would have been a much better design. |
28th Jun 2010, 12:54 pm | #24 |
Nonode
Join Date: Aug 2007
Location: South Bradford, West Yorkshire, UK.
Posts: 2,571
|
Re: Test Card / Teletext Generator - Phase 1
Actually posting my old design may be a bit difficult as I can't find the circuits etc. I've got the board up in the loft so the circuit may be up there.
From what I remember Jeffrey has described the design more or less exactly. I used LS 163s as counters, LS138s as decoders and LS73s set and reset to provide the pulse outputs plus a bit of glue logic to combine them all to give the required outputs. All counters and JKs were clocked by the 12MHz clock. I seem to remember it worked but put out trememdous RFI and virtually wiped out most of the 2m band in a radius of a few hundred yards. I built it when I was young and enthusiastic but I had no intention of generating a "proper" test card as I had no means of programming EPROMS let alone working out what to put in them. In many respects it was built because I had the facilities to design and build it with a proper PCB and it seemed like a good idea at the time. These days using programmable logic it would use fewer ICs and be much smaller and consume much less power with much of the design effort transferred from the hardware to software. I'll follow this thread with interest as potentially this design, once perfected, could be a useful piece of test kit. Keith |
28th Jun 2010, 1:45 pm | #25 | |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Test Card / Teletext Generator - Phase 1
Quote:
|
|
28th Jun 2010, 9:02 pm | #26 |
Retired Dormant Member
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
|
Re: Test Card / Teletext Generator - Phase 1
I bought an FPGA starter kit. It has a VGA connector. But no video logic AT ALL.
The HS, VS connect to FPGA pins The R G and B pins connect via resistors as a simple DAC to FPGA pins. You can do any video signal up to SVGA, "just" by Verilog or VHDL, create gate design and download via USB to a Flash chip. It's Spartan 3E Digilent Kit with also CLPD, ethernet phy, 2 x RS232 (simply connecting to FPGA), rotatory encoder, buttons switches and 2 x 16 LCD. You could certainly simply combine the VGA connector pins and some other FPGA pins to have any video standard from 405 line to 525/625 PAL/NTSC/SECAM to VGA/SVGA via LCD menu.. Link of video mods http://www.fpgaarcade.com/displaytest.htm Includes code to make Spartan board do SVGA (800x 600) Colour Bars That guy is emulating classic arcade machines! It cost me about €120 from Germany. I was going to make a DSP radio with it but so far have only designed some Hilbert filters for 300Hz to 3kHz flat -45 and +45 phase shift. |
29th Jun 2010, 2:31 pm | #27 |
Retired Dormant Member
Join Date: Sep 2008
Location: London, UK.
Posts: 117
|
Re: Test Card / Teletext Generator - Phase 1
Those FPGAs are so clever you could probably just wire one straight to a round CRT and emulate a TV22 - EHT may flash over to an adjoining pin though if you don't get a nice round blob of solder on it Seems a shame to replace all that those lovely boards full of TTL (Don't scowl Jeffrey) with an FPGA - that's what 70's games machines were all about. But we digress!
FPGAs - work of the devil they are. Having said that, I finally cracked under extreme pressure and a nice man in a brown van has just delivered a Xilinx Coolrunner II CPLD starter kit to me. It's actually a very useful bit of kit. At the moment it's acting as a paper weight to stop my TTL data sheets blowing away. One day I may even open the box In the meantime, for you 405 fans out there (if there are any ), attached is a picture of the same hardware playing out a different PROM programmed with Test Card D (for a change). The Aurora is only playing a small supporting role in modulator only mode (sorry Darryl). Any patterning you see on the screen is a combination of digital camera and file downsizing to match forum image criteria. It all looks good when you're really there. The second image is of a purposely rolling frame in order to show the VITS that Mr Tazman asked about (don't say I never listen to you). Think that's it for now. Onward towards colour..... Parabola |
29th Jun 2010, 3:01 pm | #28 | |
Hexode
Join Date: Jul 2004
Location: Michigan USA
Posts: 325
|
Re: Test Card / Teletext Generator - Phase 1
Quote:
Darryl |
|
29th Jun 2010, 5:41 pm | #29 |
Nonode
Join Date: May 2006
Location: St Albans, Herts, UK.
Posts: 2,193
|
Re: Test Card / Teletext Generator - Phase 1
Very authentic. Looks good sir
__________________
All the very best, Tas |
2nd Jul 2010, 1:08 pm | #30 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Test Card / Teletext Generator - Phase 1
A lesson in the use of non-synchronous and marginally spec'ed logic.
My faithful old (1977 vintage) home built TV line selector uses a wild and woolly selection of async logic, monostables etc. Basically the design published in Television Feb 77. I was using it to strobe out a single line from a frame as part of some development work I am doing. I have always trusted this box to do its job but today there was an obvious malfunction - looked like it was displaying both fields when it should only show a line from 1 field. It was also disagreeing with my Tek 1755A vector/waveform monitor as to which line was which. The poor thing was overheating, ironically at least in part from being stood on top on the 1755A. I've moved it, put a fan to blow on it and a few minutes later it's much happier. The both fields/one field error is almost certainly down to a monostable. Hideous things but almost unavoidable here. The line count error was probably due to some of the async logic. May this be a lesson to all of us who design digital systems. |
2nd Jul 2010, 3:44 pm | #31 |
Nonode
Join Date: Aug 2007
Location: South Bradford, West Yorkshire, UK.
Posts: 2,571
|
Re: Test Card / Teletext Generator - Phase 1
At one company I worked for we were not allowed to use monostables. They are also frowned upon where I work now due to too much variation between ICs.
Best to design them out if possible. When I designed ny SPG all those years ago, it was inspired by a BATC design which used monos, but I didn't like the idea of all those ajustments so decided to do it digitally which turned out to be a much more reliable method with no adjustments needed. Keith |
2nd Jul 2010, 3:52 pm | #32 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Test Card / Teletext Generator - Phase 1
The original 74121 monostables aren't too bad (as monostables go!) and the 74LS221 is OK too. The reall horrors are the CMOS ones. 4098: Ugh. 74HCT221: double ugh. The only 74HCT221 I found to be at all consistent was the Philips one. All the others had timings all over the place and drifted like crazy.
Actually Philips HCMOS was the best and fastest of the lot. I used a lot of their HC4046 PLL chips which would work at 27MHz. None of the others would. Just occasionally you need a mono but these days with plentiful logic in FPGAs and CPLDs you can usually find a solution with a counter etc. Parabola: If I catch you using a monostable in your TCgen you will be excommunicated |
3rd Jul 2010, 1:55 pm | #33 |
Retired Dormant Member
Join Date: Sep 2008
Location: London, UK.
Posts: 117
|
Re: Test Card / Teletext Generator - Phase 1
What, again? Anyway, there's no chance of that - you catching me, I mean
Unfortunately, a bit of a snag has come along in the quest for colour. All centred around timing issues with extended bus lengths (70ns PROMs may be a tad* too slow). So a bit more work required yet before a Phase 2 thread appears. Parabola * a quantity that, in this case, requires someone better at maths than me to define correctly. |