I've borked it

[msknight]

Ignore this post... I should have thought of the maths before I opened my mouth

[msknight]

I'm looking at this, to replace the resistor pack. 2% tolerance, while I believe that the blue one (which is actually labelled 22x... so might have been cut from an array of 22?) is 5%.

BOURNS - 4609X-101-103LF - RESISTOR, NETWORK, 9 PIN, 10K, 2% - https://www.ebay.co.uk/itm/171324463458

[1024MAK]

A tolerance of 5% or better is fine.

The old pack, the 22X is most likely just part of the part number.

Mark

[msknight]

Got some ordered. They'll be a while.

I've still got no clue as to what happened. I was swapping ROM chips when it happened. I've got no firm indication as to what broke, so if this resistor pack doesn't solve it, I'm up the creek and having to de-solder another board in order to fit sockets and test chips.

[msknight]

The more I sit here and think, the more I doubt that a new resistor pack will solve this problem. Packs aren't known for failing, purely because of what they are and how they're constructed. Plus I tested it when removed and found no problem.

I'm at a dead end now and I don't know how to go forward in troubleshooting a very strange issue which is affecting all the data lines simultaneously.

[msknight]

I don't know why I didn't think of this earlier.

I plugged in a ZXpand and.... I get a K straight away.

That puts the fault somewhere with the rom/ram side of things.

In a few days I'll pick up an original 16K RAM pack, and if that works or doesn't work... it will help me isolate a RAM or ROM issue. I can take things further from there.

It does also vindicate the resistor pack.

As I was changing ROMs at the time the issue happened, then that does bring ROM as being the more likely of the two, to be the issue. But further testing will tell... and the ultimate question of why two different ROM chips would produce the same failure. Or it could be the same again, but with the RAM. I'll know for sure in three or four days.

[Moggy]

I'm assuming keyboard works etc with spand attached?

Now that you have the cursor present and if also a working k/board type the command CONFIG" X" (CONFIG is shifted G.) no spacing the spand does that for you.

If at that point you still have the cursor then I would say it points to your RAM mod' as the spand switches it out in the manner of a memory module and after using the CONFIG "X" command you are then running on the original internal ROM with the spand now acting as a glorified RAM pack.

[msknight]

I remember reading that there was a way to get the ZXpand to surrender to the ZX81 ROM, but you saved me a chunk of work trying to figure that out!

The result was that I retained a K and typed in a line of code and it retained it.

The RAM was also replaced. It had a 1K chip which was replaced with a socket and a modern 16K RAM chip, along with a capacitor to solve the speed issue. That was removed and replaced with a vintage RAM chip, and the fault persisted.

This means that I've got another problem somewhere. I did test the 5V into the chip and that comes up OK on the scope, so I've now got to figure out what could be wrong with the RAM that would cause all data lines to go to garbage simultaneously.

Getting there. Getting there. Many thanks for the help.

[msknight]

I'm also working from Ron Reuter's re-drawn schematic and I've just realised that it doesn't show a number of things including how the edge connector hooks up to the data lines.

Can anyone recommend a fuller schematic please?

[msknight]

This is the current RAM chip and wiring. This was only installed after the fault appeared and it replaced the MyRetroStore 16k module.