exxos's Atari, Amiga & retro forum

stephen_usher wrote: 17 Apr 2025 18:38 /OEB and /OEL are for CPU reads from RAM. If they're never happening, and Gary is tested then it suggests that Gary isn't recognising that it's a RAM read and hence isn't enabling the transfer. That could be an address line on Gary being bad.

Working through it all still, but seem to have some bad data lines from, IIRC, U11 to Agnus.

stephen_usher wrote: 17 Apr 2025 18:38 /OEB and /OEL are for CPU reads from RAM. If they're never happening, and Gary is tested then it suggests that Gary isn't recognising that it's a RAM read and hence isn't enabling the transfer. That could be an address line on Gary being bad.

I'm currently buzzing everything out. On the CPU side, all of the data lines connected to U10-U13 are good. I'm just about to print out a pic of Agnus so I can check that side of U10-U13 and I'll have a look at Gary as well.

OK, so to speed things up a little I made a brush probe for my multimeter. I'd seen one on Adrians Digital Basement and it makes it a lot quicker to find general traces than a standard probe, which can also be a bit harsh on the traces:


And it's worked a treat :)

So what I found it that U10, U12 & U13 are all connected to Agnus as expected. U11 on the other hand has no connection from pin 3 or 4 over to Agnus. They should as far as I can tell from the schematic, be connected to pins 14 & 13 of Agnus respectively.

So I'm going to trace back from the Agnus to U11 to cross check my findings and then see if I can find where the break is, and I'm thinking it's a bas via somewhere.

I'll report back later.

The broken data lines are fixed, but still no joy.

I have buzzed out pretty much everything in the Data path circuit and it is all good from CPU across to U10, U11, U12 & U13 and across the other side to Agnus.

I've done the same for everything coming into U34 & U35 and everything going from them to the RAM, so address line and cas/ras/we.

I've tried a different Agnus, Gary, CPU and swapped out U10, U11, U12 & U13 for new tested IC's, and I've also swapped out U34 & U35 for new tested IC's as well. The Agnus socket is new.

Still I get 3 cold RAM ICs and 1 hot one, they have power and GND OK (I will recheck for the sake of my sanity) plus I have tried 3 sets of RAM and the result is the same. I get a green screen of death every time.

I suppose the GSOD does at least mean the CPU is executing code from the ROM and is telling me 'No RAM found.

Also, DiagROM is trying to run, but it's crashing.

I don't think it can be anything other then RAM, but I am pulling what little hair I have out right now.

What I haven't done yet is to scope the RAM with DiagROM in there, so I'll try that next.

Have you checked the address lines to Gary yet? Gary is the gatekeeper after all.

stephen_usher wrote: 18 Apr 2025 17:36 Have you checked the address lines to Gary yet? Gary is the gatekeeper after all.

Indeed he is, and it's next on my list. Given what I found with U11, which was two vias with tiny fractures that were almost invisible to the traces, I can't help but think there are more somewhere.

I think it's a case of a fragile mainboard that hasn't responded well to all the manual handling of adding mods etc.

And phew!! It's running again and seeing all the RAM, plus booting from HDD :)


And no, it wasn't Gary it would seem but a combination of issues.

Firstly the 2 data lines were bad at U11, and with such a small fracture at the via it was hard to see even with a microscope.

Secondly, at the time I tested the data line repairs I had my spare RAM installed. Turned out there was a bad chip or two. This morning, before I carried on checking I put the original RAM back in to see if anything had changed, and it booted just fine :)

So I don't have any way to test 514256 RAM, and with a working Amiga I decided to test one IC at a time and found a couple bad, one out of the test set I was using. Damn.

Anyway, I found one more, the rest being good so at least now I know that I have good spare RAM.

It's annoying enough to have 2 bad vias (I have a Mega ST that suffered the same), but to trip myself up with bad RAM, that is super annoying :(

Glad you sorted it in the end @rubber_jonnie
Very good to see my old system back in action with those upgrades

:thumbup:

JezC wrote: 19 Apr 2025 13:04 Glad you sorted it in the end @rubber_jonnie
Very good to see my old system back in action with those upgrades

:thumbup:

I'm very pleased too and just running Amiga Test Kit memory diags.

I'm kicking myself around my spare test RAM, but very happy to have at least got it back up and running and tested all my spares now.

It's great having the upgrades working too, the RGB2HDMI looks fantastic, and the A500 8MB +IDE is just spot on.

Once again though, a warning to all that these old boards don't always like being manhandled. The damaged traces were there before I removed U11, then I didn't see them when I installed the socket. They can be a sod to find for sure.

Good work!!
Happy you found it!!