ZX81 with mismatched ROM, jumper cable and flat cursor

[nicad]

HI! I've just picked up my first ZX81 from ebay and unfortunately can't get a readable prompt.
The board has already been modified with a composite board, but it also seems to have a couple of other mods:

• It's an issue 1 board with the ROM fix standoff board, but the ROM chip appears to be a later 649 variant.
• There's a wire on the bottom of the board from ULA pin 14 to a point just the other side of the ULA (see pic below - red line added for clarity).

I did also have to remove a couple of solder blobs bridging pins on the ULA and RAM chip, so I'm hoping that hasn't caused any damage.

So that gives rise to a couple of questions:

• Will the ROM fix cause problems for the later ROM installed?
• Any idea what the yellow jumper wire would be for? Removing it seems to make the output worse - I get just a plain white screen rather than the one shown below with it.

But the main issue:
I can get a picture via composite, but the cursor appears badly skewed/flattened. Though I'm not that is the cursor as pressing anything on the keyboard doesn't change the display - though I don't know if the keyboard is working yet!

Any help or suggestions welcomed, I would love to get this operational again.

[1024MAK]

Bizarre!

Anyway, some answers:

With any ROM other than the version with the maths bug, the ROM fixer board (the daughter board above the processor) should be removed. Firstly because it's not needed and secondly, because it may cause problems.

The ULA looks like it did not originally have a socket. It may or may not be the original. If it is the original unless the pins were cleaned and straightened, it's possible one or more pins may not be making good contact with socket or the socket contact may have been damaged.

For at least some of the early production run, the ULA was soldered to the board. The soldering looks different for the ULA socket compared to the soldering elsewhere in the board. The "wire on the bottom of the board from ULA pin 14" is likely replacing a damaged/broken PCB pad/track. ULA pin 14 is the clock output from the ULA that goes to R6, which eventually feeds the clock input to the processor.

There may be other damaged/broken PCB pads/tracks.

I can still see a solder blob linking two pins on the RAM chip. Is this the blob that you have removed? Or is this another one?

The fact that you get anything on the display shows that the processor is trying to run.

See what happens if you disconnect and remove the ROM fixer board (the daughter board above the processor).

Mark

[nicad]

Thanks for the info!

I've carefully removed the ROM fixer board, and plugged the Z80 back in, but the screen looks exactly (pixel by pixel) the same.

So based on what you said about broken tracks, I removed the ULA too, however when I pulled the chip out, the metal contact from the socket for pin 31 (KBD1) also came out, suggesting it wasn't soldered to anything. I've put the contact back in the socket, but there's no trace on the back side to solder the pin to - so I popped a solder blob on it to hold it in for now. Looks like I'll need to add another jumper wire from the solder blob to the keyboard connector to reconnect it?

If that sounds like a good idea, do you know which point it needs connecting to? I tried tracing the wire, but can't see under the connector, and continuity testing both sides of the board was problematic!

As for the solder blob on the RAM, that has been taken care of, ta.

Thanks, Nik

[nicad]

After a bit more of an inspection, I found my board has a section of missing links which probably won't help:

So I'll work on figuring out where the ends should be and add them back in and check the connectivity of the rest...

[1024MAK]

There is a good schematic here. Just keep in mind that it only shows the French video circuitry between the ULA and the modulator. Everything else is correct (to the best of my knowledge).

Mark

[Moggy]

Picture of one of my old motherboards just for a visual reference with the tracks painted in. The black "X"s indicate the track goes to a resistor pad before exiting.

The red track goes to an empty pad and the white track I added, again as a reference point.

Click to magnify.

[nicad]

Fantastic, thanks for both of those, much better than the reference images I found. Now down to work...

[1024MAK]

This site has some pictures of an unpopulated issue one board that may help

Mark

[nicad]

Success!

And this is what it looks like now!

So on to the next challenge, the keyboard doesn't do anything. I couldn't immediately find any, but are there any guides on diagnosing keyboard issues? The connector cables look like they have been bent quite a bit in the past, so I'm not totally surprised it's not working, but could do with knowing if the keyboard is at fault, or if there could be another connection issue on the board.

Thanks again for your help so far!

[1024MAK]

Notice that one keyboard connector has eight contacts and the other five. The membrane is a matrix of 8 x 5 = 40 keys. So using a piece of suitable insulated wire (0.6mm to 0.9mm solid core tinned copper is what I recommend) with the ends stripped, you can touch one end on each contact in the 5 way keyboard connector, then in turn, touch the other end on each contact in the 8 way keyboard connector. Each different combination should produce a different result apart from the shift key (which on its own does nothing).

The most common problem is that the plastic that the membrane is made of goes brittle, and cracks form where tails are bent, causing breaks in the thin metal tracks. Sometimes if the break is near the end, you can use cut the bad part off and then very carefully push the new end into the PCB connector.

Otherwise you need a new keyboard membrane. This is what I use for replacements.

Mark