Sparkly pixels, take two

[Lardo Boffin]

A while back I started a thread about a zeddy with 'sparkly pixels' (viewtopic.php?f=7&t=2201&hilit=Sparkly) and now I have another one!

This one does not do it as badly but it does do it with both the ZXPand and a Sinclair 16k RAM pack. As well as the pixels flashing on and off it also had the problem on the LCD where the image became unstable and turned a bit orange rather than being black and white. Interestingly this one didn't stay orange once it started having issues it comes and goes in a pretty stable pattern.
My CRT TV didn't have this issue so I assume it is less sensitive to input variations?

Having had the previous one repaired at MCG by chip swapping I tried the same. I changed the ROM and it now appears fine (soak tested for three hours now).

So that leaves me to wonder what it is in the ROM that can cause this problem? Or what is it in the rest of the zeddy that can cause the ROM to cause this issue!

Lardo

[PokeMon]

Yes - I think so. I had this issue with a ZXmore as well which could be solved while adding a 47pF capacitor to /OE signal. You could try between pin 20 and pin 12 of ROM (IC2) of a usual ZX81.

[Lardo Boffin]

Yes - I think so. I had this issue with a ZXmore as well which could be solved while adding a 47pF capacitor to /OE signal. You could try between pin 20 and pin 12 of ROM (IC2) of a usual ZX81.

Not being too technical I won't ask you to explain why that works but would be interested to see a photo of how you installed it!

Many thanks

Lardo

[PokeMon]

I haven't installed it in a ZX81 but used with ZXmore to avoid sparkling pixels.

It works as the capacitor extends the OE signal of ROMCS. I think it could be to fast switched off which let read sometimes wrong ROM contents during refresh, displayed as pixels. It is sparkling as it is quite on the edge. Sometimes showing wrong contents, sometimes correct, depending on address lines used as well. It is more or less a kind of random pattern.

Could work with lower capacities as well, maybe 33pF or 27pF could solve this pattern, too.
47pF I have many and is an acceptable load for most ICs.
You could also try if the pattern disappears when putting a finger in this area and cover GND too (pin 12 and pin 20).
This is also a typical body capacity.

[Lardo Boffin]

One thing I did notice after removing the dodgy ROM was that the legs are quite tarnished. The other chips present are not like this.

Could this have been causing problems? The lower part of the legs are not too bad now - presumably removing the ROM and reseating it elsewhere helped.

What is the best method to clean these, if it is possible?

IMG_6385.JPG

(1.1 MiB) Downloaded 286 times

[sirmorris]

Glass fibre pencil. You'll never want to be without one.
until then, kitchen scourer.

[Lardo Boffin]

Ordered thanks! There isn't a scourer in the house...

[Lardo Boffin]

And back to square one... After working fine for several hours yesterday the problems are back with a vengeance.

Hey ho - time for more chip swapping to see if that helps.

[PokeMon]

You can change components of course and it may help - but in my eyes it is a simple timing problem which can be probably solved with a simply added capacitor.

And it may appear or disappear depending on the weather (room temperature). Most of these effects are temperature sensitive. Cool air or open case could stop this effect temporarily but a capacitor is cheaper and easier to install than a cooler.

[Moggy]

@Lardo

Don't worry about the tarnish it's quite common on Mostek chips and I agree with Karl re timing as I'm sure Andy has had similar problems with certain roms when creating devices for himself.