Television Magazine Testcard Generator

[FRANK.C]

The glitch just below the colour bars looks like a line sync timing error.
Is there any subroutine called in this part of the code? I found that subroutines can introduce such errors. Especially if a subroutine is called within a subroutine.

I don't think that type of error can be seen on an analogue scope but a decent digital scope will show it.

Frank

[See_Mos]

I solved the glitch, a couple of clock cycles had crept into some of the blocks of line code, it's surprising how much 125nS can affect the display. If every line was extended by the same amount I think I might have got away with it.

[FRANK.C]

Well done on finding it. It can take a bit of counting to find but where it appears on screen is a good clue.
Even if out by a single cycle can be noticed on screen.

Frank

[Philips210]

Hi.

A further update on the progress of this project.

I finally completed the breadboard assembly of the logic board for the test card generator. The IC positions are essentially the same as on the original PCB but I worked entirely from the circuit diagram which I corrected from the originally published circuit. I made a point of ensuring that the +5V supply and ground connections to each IC were correct before proceding any further.

I also wanted to use the original TBA540/TBA520 PAL colour encoder for coupling to the logic board. I built the encoder on matrix pad board. The reference oscillator quadrature coil should have been from a GEC 2110 series colour TV but I used the one as in the Philips G11 chassis kindly supplied by Glyn (Welsh Anorak).

I coupled the logic board to the PAL encoder, gave the whole circuit a good check over, set the current limiter on the bench power supply and tentatively powed up the circuit. Fortunately there were no signs of distress or overheating.
As the circuit is very complicated I wasn't expecting any initial success but was pleasantly surprised to see a form of test card. It turns out there were about several problems.

The first pic shows the first test. The odd colour fault was due to an error connecting the burst gate signal to TR1 on the encoder board. I also had an intermittent disturbance effect on the encoder due to a poor solder joint. The remaining problems obviously being on the logic circuit layout.

The fault with the 1.25MHz frequency gratings extending from top to bottom of the display was actually due to IC42 (7410) A replacement put that right.

The remaining problems listed as follows:

(1) Wrong colour bars ie cyan, magenta and blue missing due to a wrong connection IC26 pin should go to IC27 pin 9. I connected it to pin 10. Correcting that gave correct colour bars. (see pic 2)

(2) IC21 pin 9 not connected resulted in a double height colour fit pattern (see pic 3)

(3) IC26 pins 12 and 13 weren't connected which affected the rendition of the letter box pattern (see pic 4)

(4) The right hand side of the pattern was considerably brighter than the left. I left off a link between IC34 pin 3 and IC41 pins 2, 5 and 10 (see pic 5)

I still have some issues with the PAL encoder. There's a green cast present at the top of the screen and I am trying to see if it is associated with the different quadrature coil. The video output is quite low and I haven't been able to get the modulator to work probably due to a low amplitude video signal.

Anyway, I am quite happy with the progress so far. More to follow shortly...

Regards,
Symon.

[Philips210]

Hi

The PAL encoder using the original circuit is shown in pic1. It has an additional stage for composite video out as per the Television magazine (August 1979) circuit recommendation in the article "Simple Sync Pulse Generator" by M Burrell. I've not included a UHF modulator but intend to have a separate board for that in due course.

Earlier in the year I built a TEA1002 PAL encoder as recommended by member SeeMos. I initially tested it with a colour bar signal which worked well after overcoming problems with the PAL switching and colour burst flag (CBF) signals.
I've adapted this board for use with the test card generator. (see pic 2).

Also Trevor (member SeeMos) kindly provided me with a couple of his own designed TEA1002 PAL encoder PCBs. I built one and also tested it and it performs well. (see pic 3).

I am also in the process of assessing the Mullard TEA2000 PAL encoder and will update on this in due course.

In pic 4 the testcard display with the TEA1002 encoders.
I am quite pleased with the TEA1002 circuit. More to follow.

Regards,
Symon

[Philips210]

On an old Philips televideo set, the display with the TEA1002 encoder with chroma band limiting (pic1) and without (pic2).

I intend to experiment with other patterns in addition to the test card. I'll then have an idea what the final circuit will look like and then progress to a PCB.

It's been good fun building the prototypes on this retro project.

Regards,
Symon

EDIT: On my TEA1002 PAL encoder, I used the chroma band limiting filter coil together with the 270ns luminance delay line. On a flat panel LCD display there was negligible improvement but on an old analogue colour TV the dot crawl effect was much reduced

[Chris55000]

Could you tell me where I can obtain a suitable Chroma Band Pass Coil and 270 ns Delay Line please?

I don't want to buy a huge great CRT TV just to take two or three bits out of it!

I don't mind buying a panel if anyone knows a chassis that has suitable parts!

Chris Williams

[See_Mos]

Took bandpass coils are available on eBay. Search 5 Toko 166NNF-10264AG.

For the delay line I was lucky and a new one was available from the same source.

I bought them for a MC1377 encoder which still has to be completed.

[Chris55000]

Pack of Toko coils now puchased, just need a 270 ns Luma Delay Line now – if anybody can identify a TV chassis with one or has one please let me know!

HD44007A and 17.704 MHz crystal now on its way so I'll start the SPG PCB 1 when I can get the "Easy PC" up again – it's just thrown a wobbly on me!

Chris Williams

[Philips210]

Chris.

For the TEA1002 encoder I'm using the 270ns luminance delay line and chroma band limiting filter coil pulled from an old Ayr Teletext add on unit. I don't think you'll find a 270ns delay line in any colour TVs, they're generally 330ns. I've also been trying to find some spare 270ns delay lines for some time.

I'm also working on a TEA2000 encoder. That also uses a 270ns delay line but is of the green rectangular box construction as opposed to the cylindrical unit on the TEA1002 encoder. I originally purchased the delay line together with the TOKO chroma band limiting coil from Maplin over thirty years ago. Finding these parts now isn't so easy.
I seem to recall reading somewhere that the chroma band limiting coil needs to have a fairly high Q to be effective.

I have been recently working on the Test card project and have made further progress. I've been experimenting with the border castellations and trying to produce a similar arrangement to Test Card F. On the left there are the red and blue castellations for decoder checks. To the right, white a yellow ones for sync checks. The top of the card has cyan castellations and the bottom has green ones. I have been using timing diagrams for the line and field waveforms to plan the final position of the castellations on the test card. I write the Boolean algebra expressions by referring to the waveforms then simplifying them to minimise the number of gates required.

A combination of AND and OR gates are used to obtain the intended size and position on the card. Further gating is required to add the extra waveforms to the existing colour sections of the card.
I may also add a green square to the letter box for LF response checks as on Test Card J. Extra frequency gratings may also be added. Further improvements are also being considered.

It's proving to be a very enjoyable project and I will add some further pics of the progress in due course.

Regards,
Symon

[Philips210]

Hello again.

Just a small update on the progress of the test card generator.

I've been experimenting with adding colour castellations to the card along the lines of Test Card F and ETP1 patterns.
On the left of the card are broad and narrow red and blue castellations.
To the right are broad yellow and narrow white castellations.
On the top are six cyan castellations and on the bottom six green castellations. There are also four white corner castellations.

By sketching out a rough plan on a printed test card, I referred to the line and field timing waveforms from the original circuit and devised the gating logic to fulfil the idea. Additional OR gating was included to insert these castellations together with the existing colour parts of the card. It's taken a further 42 gates (13 ICs) to provide the colour castellations.

Please see attached pics. In the first pic, a Toshiba LCD TV is displaying the latest version of the card. In the second pic, a Philips CRT televideo combi. The third pic is of the breadboard layouts.

It's easy to see why a mask programmed IC makes life easier but I wanted the challenge to do it with bog standard TTL and it's surprising what can be done.

I'm considering increasing the number of multiburst sections (frequency gratings). Also, a green square in the black letter box for LF checks. It's an enjoyable project, when I can find the time that is...

Regards,
Symon

[1980s_john]

Originally Posted by Philips210

but I am always interested in pattern generator circuits and enjoy experimenting with them.

Symon,
I hope you have seen the BlackStar Orion pattern generator, circuit diagram is in the service manual:

http://w.ko4bb.com/getsimple/index.p...ment/BlackStar

Regards,
John

[Chris55000]

Any chance of a circuit sketch of your Coloured Castellation Logic Symon?

Chris Williams

[Philips210]

Hello Chris.

Sorry for the delay in replying. I hadn't checked this thread recently and only just noticed your post.

The colour pattern generator had been put on hold this last month due to other projects but I hope to return to it soon.
The circuits for the colour castellations are only hand drawn at present. I haven't yet finalised the extra logic for the colour castellations and may need to make some alterations to the version previously shown but I will gladly send you my circuits when finalised. Also, I hope to add a green square to the the black letter box for luma/chroma timing. Extra multiburst frequency gratings are also being looked into. I will update the circuits soon.

Regards,
Symon

[Philips210]

Originally Posted by 1980s_john

Symon,
I hope you have seen the BlackStar Orion pattern generator, circuit diagram is in the service manual:

http://w.ko4bb.com/getsimple/index.p...ment/BlackStar

Regards,
John

Hello John.

Thank you for the link, that's very good. I'll have a good look at the circuit. What a great website, plenty to look through!. Sorry for delay in replying.

Regards,
Symon

[Chris55000]

For anyone wanting 270ns Delay Lines for TEA2000
PAL Encoder circuits can I point Members in the direction of :–

https://www.ebay.co.uk/itm/392205500...mis&media=COPY

I've just bought one for my forthcoming TEA2000 board!

Chris Williams

[Philips210]

Thank you Chris, that's useful to know.

Re. the circuit diagram for my colour castellation modifications, I haven't forgotten about it and will return to the project before too long. I've had a lot on recently and have also been catching up with outside maintenance and gardening as we're having some decent sunny weather for a change. Once I'm happy with the mods, I'll attach the completed circuit.

Regards
Symon

[Chris55000]

Morning Symon!

Thanks for the update on the Colour Castellations Project, I look forward to seeing the diagrams when completed to your satisfaction!

I'll be building my T.C.G. in Modular Format, with (provisionally) the Sync Generator, Audio & Modulator on the first board, crosshatch/vert/horizontal/crosses, etc., on board two, multiburst and circle add–on on board three, grey scale on board four, composite pattern/colour gating on board five, board six being Trevor Wilson's PAL Encoder, one of which I obtained from him.

Am I likely to spoil the fidelity/quality of the completed T.C. by splitting up the circuits in this way, or are the switching speeds low enough for me to get away with it? The Philips PM5518 is modularized!

I understand you developed one, and your Colour Castellations Project on pluggable breadboards by interlocking a large number together to make one huge "mega–breadboard", am I correct?

I don't trust myself to try and design board D062 in one go without making an atrocious pigs–ear of it, but I am happy to untangle all the separate pattern circuits from the full diagram and do it a bit at a time!

How would Malcolm Burrell have worked out which TTL to put where?

Chris Williams

PS!

How are white/red rasters done? – are they simply peak–white voltage levels fed to the colour gating logic? – The Manor Supplies Mk 5 circuits are too furry to read & the ones in Philips's manuals make my head go round the way they draw 'em!

[Philips210]

Hello Chris,

It's funny you should mention building the Test card generator in a modular way. I've come to the conclusion this will be advantageous, especially with all the extra features compared to the original large logic board. I would still keep the interconnections between boards short. I thought the boards could be arranged in a stack but still accessible for service.
Interesting question how Malcolm decided on the layout of the logic board. I don't really know. The Sync/Blanking IC, monostables and counters seem to follow a reasonable layout. It's all the combinational logic gates that look daunting to lay out. I guess it was a lot of trial and error. I reckon there are a few redundant gates here and there in the line/field logic but haven't entirely proved this.

My current set up is entirely on pluggable breadboard with the exception of the PAL colour encoder. That is currently on matrix pad board. I also have Trevor's PCB version of the PAL encoder that works well. The original TBA520/TBA540 encoder works quite well but something I've noticed is the blue and yellow colour bars look a little light. This may be due to my using the quadrature coil from a Philips G11 rather than from a GEC 2110 chassis. I'd like to find the correct coil to prove the point either way. I think it would probably be better to use the TEA1002 encoder which does give good results. The later TEA2000 encoder is also a good circuit but I haven't been proving the circuits recently. There should be no problems using the TEA2000 as an alternative. In fact it ought to be easier than the TEA1002. THE CBF and PAL switching is done within the IC.

Re the red and white rasters, I haven't sorted that out at this stage but it should be fairly easy to apply the relevant logic to the RGB inputs to the encoder. It will all be down to switching as more and more separate patterns are added plus the complete test card itself. Have a look at M Burrell's Simple Colour Pattern Generator, TELEVISION, February 1990 but the PAL encoder won't work without modification as mentioned earlier in this thread.

For pattern switching I thought about doing this electronically as opposed to switches for better reliability. Doing the switching electronically also lends itself to adding remote control as well as a local keypad. I reckon the old Mullard 5000 system would fit the bill. The SAA5000A encoder/transmitter and the SAA5012 receiver/decoder would work well.

I've also proved several chequerboard patterns are easily done with the existing line/field waveforms with a small amount of additional gating. I followed the approach Malcolm Burrell used in his Diagnostic Pattern Generator, TELEVISION, August 1978. The positive and negative streak patterns also worked well as did the teletext interference test for slow field blanking.

I haven't done any work on the circle pattern but looking at the Philips PM5519 it basically starts by using a parabola waveform to build the circle. I have a Körting 82515 pattern generator that also produces a circle pattern and will take a look at this possible route. Have you any ideas what circuits look straightforward for generating the circle?

I have also been having ideas of including additional colour patterns as in the Philips PM5508 such as the Delay, Matrix and Phase but that would be quite complicated as you'd need access to the three colour difference signals to do this. It would need a different PAL encoder.

The colour castellations seem to be working very well but wondered if I should change the position of these. They're a sort of cross between Test card F and the IBA ETP 1 castellations. Any criticisms or suggestions welcome.
Regarding ETP 1 pattern, I think it would make a good project to replicate this pattern using TTL or CMOS. I don't think it would be too difficult to do.

Sorry for the long post!

Regards,
Symon

[Wendymott]

When I designed my PIC TV signal generator, there was a glitch centre screen at the end of line 255. I had no idea how to go about de bugging it, so I designed and made a Line selector delay box. This had "brightup" with totally variable timing. Again a PIC was used, and the glitch was found very quickly, added a "NOP" and that was it. The Delay generator was never used again, but it really was invaluable at the time. If anyone requires details I still have the SCH's and PIC software.