exxos's Atari, Amiga & retro forum

And the Electron is complete:


Time to figure out the Plus 1 next.

:girldance:

OK, so the Plus 1 is putting up a bit of a fight...

I've managed to find a bunch of dead ICs and now the Electron will boot with the Plus 1 attached, which it didn't before, it just hung the machine.

The main issue now is that the ROM isn't being seen, so either the ROM is bad (I can read it in my TL866) or there is something else going on but it seems that the ROM is just directly connected to the address/data buses so I'm going to check the /CE and /OE lines are doing what's expected as well, and I'll probably also continuity check the address/data lines from the ROM socket to the edge connector.

Yeah, checking the /CE or /OE lines on the ROM and then tracing back if they don't appear is a good idea.

I believe that there's a schematic on-line for the Plus 1.

P.S. Found one.

OK, the ROM /OE goes through a block of links... Looks like it can be enabled directly by A11 (LK4) or via a train of TTL logic (LK2) for address decode. Similarly A11 on the ROM can be A11 or the address decode logic.

stephen_usher wrote: 20 Aug 2024 12:27 Yeah, checking the /CE or /OE lines on the ROM and then tracing back if they don't appear is a good idea.

I believe that there's a schematic on-line for the Plus 1.

P.S. Found one.

OK, the ROM /OE goes through a block of links... Looks like it can be enabled directly by A11 (LK4) or via a train of TTL logic (LK2) for address decode. Similarly A11 on the ROM can be A11 or the address decode logic.

I do have the schematics (2 lots to be fair) and the links are set with 1 & 2 connected, which should be correct. /CE is connected to ground, and that's fine, and I do see a signal on /OE but it looks kind of off.

I need to check the signals at the links I think, see what's what there.

OK, so I'm looking at the signals at the ROM, specifically pins 22 & 23 which are fed by links 1&2 as best I can tell.


Link 1 -> Pin 23 (A11) This is the blue trace and it looks OK to me.

Lin 2 -> Pin 22 (/OE) is the yellow trace and it looks bad.

R/W from the edge connector, which feeds into a gate in IC2 related to /OE looks just fine on the scope, but the output on pin 3 looks like the yellow trace.

So /OE looks bad.

Tracing back it looks like IC6 is feeding out garbage up to the IC's between it and link 2, so I will pull it out and test it.

Hopefully I have a spare in stock...

And I still get the same garbage with IC6 removed, plus it tests just fine...

It is clearly an /OE issue here, the signal looks just awful. Not much left that can be wrong...

More digging and IC5 is dead. So dead you couldn't even use any spare gates on it. Ho hum.

On a positive note, IC1 is all good. TBH I only have 2 IC's left that I can pull and test, IC10 & IC12, so I'm just going to do it.

It is unfortunately a shotgun approach, however the IC's I've removed so far have had a lot of residue under them from the water that seems to have got into this device, so it's not a bad thing as it means I can clean all the crud from underneath and at least verify the ICs as good or bad with them out.

EDIT - IC10 & IC12 pulled and I've cleaned under where they were. Turns out IC10 is also dead as a dead thing that's really dead...

I only have 74S04's, but given that there are a number of other bad IC's awaiting replacement, I'll just order some LS's while I'm at it.

rubber_jonnie wrote: 21 Aug 2024 14:14 More digging and IC5 is dead. So dead you couldn't even use any spare gates on it. Ho hum.

On a positive note, IC1 is all good. TBH I only have 2 IC's left that I can pull and test, IC10 & IC12, so I'm just going to do it.

It is unfortunately a shotgun approach, however the IC's I've removed so far have had a lot of residue under them from the water that seems to have got into this device, so it's not a bad thing as it means I can clean all the crud from underneath and at least verify the ICs as good or bad with them out.

That's one of the things I dislike the most about fault finding, especially 8 bit era machines as the chips are so rarely socketed. I've had to remove no end of logic chips and 4116s lately, just to see if they have died. At least they are then socketed which makes it easier for the next person.

PhilC wrote: 21 Aug 2024 14:45

rubber_jonnie wrote: 21 Aug 2024 14:14 More digging and IC5 is dead. So dead you couldn't even use any spare gates on it. Ho hum.

On a positive note, IC1 is all good. TBH I only have 2 IC's left that I can pull and test, IC10 & IC12, so I'm just going to do it.

It is unfortunately a shotgun approach, however the IC's I've removed so far have had a lot of residue under them from the water that seems to have got into this device, so it's not a bad thing as it means I can clean all the crud from underneath and at least verify the ICs as good or bad with them out.

That's one of the things I dislike the most about fault finding, especially 8 bit era machines as the chips are so rarely socketed. I've had to remove no end of logic chips and 4116s lately, just to see if they have died. At least they are then socketed which makes it easier for the next person.

Indeed. There were some sockets, but I swapped them out too as they were single wipe and corroded too.

I hate that I have removed and socketed each and every unsocketed IC on this Plus 1, but by the same token, I have been able to verify all the ICs as good or bad and I have been able to clean under the chips where there was what looked like scale from a kettle under all of the ICs.

TBH IC's 5 and 10 look fairly important in the grand scheme of things, and the fact they are proven dead is at least something.

The whole damn thing is socketed now, and once I get my replacement ICs I can see what's what. If it still doesn't work after that it's going to be down to tracing the PCB to look for missing connections...