Zeddy displaying wrong characters on screen

[Lardo Boffin]

I recall, along time ago, that I had a similar problem and replacing the voltage regulator fixed the problem. I got this idea from a forum posting but I can't find it. Not being a hardware guy I have no idea why the voltage regulator would garble the display, but replacing it worked for me.

Regards from Hardware Challenged;

Greg

Annoyingly I recently used my last one so I will try that at some point thanks!

[Lardo Boffin]

In the spirit of quick checks I have plugged in my ChromaSCART.

With the onboard RAM switched off I get the display issue.
With the onboard RAM switched on I still get the display issue.

I was expecting it to work properly with the RAM on as this is in effect the same situation as having a RAM pack or Spand plugged in.

With the Memotech RAM pack plugged into the back of the ChromaSCART but its on board RAM switched off it works properly - no garbled screen!

[Lardo Boffin]

The ROM is really quite hot to the touch.

[Lardo Boffin]

Guess the program!

Running through the ChromaSCART.

[Lardo Boffin]

I put a different ULA in and it works perfectly. Put the old one in and back to Venusian.

Any ideas how a ULA could do this? Given that it pretty much works fine when a RAM pack / Spand is plugged in?

[msknight]

The only thing that comes to mind is timing signals. It might be worth seeing if that ULA works or fails in another system that you've replaced the voltage regulator, the capacitors, transistors, etc. of... just to see.

[msknight]

The ULA is, if I understand correctly, basically a collection of logic gates in one chip, so it's possible for it to partially fail without completely breaking down... if that makes sense.

[Moggy]

A cell from a de-capped 2c184 ULA, the second picture is a cell with the overlay from Sinclair.

I've had one or two that partially failed, membrane keyboard presses causing a crash but with a Memotech plugged in all ok.
Replacement of the ULA cured it so not a board fault.

[Moggy]

Guess the program!


IMG_5789.jpeg


Running through the ChromaSCART.

10 PRINT "HELLO WORLD"
20 GOTO 10

[Lardo Boffin]

The ULA is, if I understand correctly, basically a collection of logic gates in one chip, so it's possible for it to partially fail without completely breaking down... if that makes sense.

That does make sense - I have some ‘danger ULAs’ that are part failed that I keep for test purposes. It doesn’t matter if they go into a zeddy that may break them. For this one I used a ULA that doesn’t save but the keyboard interaction is fine.

What I don’t really get is why the part failed area works fine when a RAM pack is plugged in but does not work otherwise. Clearly there is nothing too badly wrong with the board otherwise a working ULA would not work.

The only thing I can assume is that the ULA is marginal and some aspects of the board are marginal. The presence of certain peripherals must be affecting this interaction?

One thing I have learnt from endless repair videos and similar is that although computers are considered digital technology they are anything but with tolerances and timing etc.