The latest member of the family

[rubber_jonnie]

Maybe resolder the on-board ram chips/data path, and reseat the Agnus..

I am thinking of doing that, Agnus is in a new socket so I think it will be fine.

[rubber_jonnie]

OK, so the Amiga is still playing up. I get a green screen on boot and if I leave it to stand long enough it will eventually boot, either to the SD card or to the kickstart ROM. That suggests to me a thermal issue, but I haven't been able to pick up anything specific with the thermal camera.

AFAIK the flashing green screen means it can't find at least 512MB RAM onboard.

I have reflowed the data path IC's (U10, U11, U12 & U13) as per the suggestion from @Maximilian but it's still the same. I also reflowed U34 and u35 and all of the 4 onboard RAM IC's and that made no difference either.

I've tested with a different Agnus, Gary, Paula and Denise and different CIAs, however I don't believe it's any of those as whatever is installed the problem remains the same. A green flashing screen that when left long enough, the machine will actually boot.

To put the spectre of the Amiga 8MB plus IDE expansion to bed, I have also tested the machine with its original CPU and it behaves the same.

I did try piggybacking new ICs on top of the data path IC's (U10, U11, U12 & U13), but that didn't work either. I have also tested with the trapdoor RAM removed and it's the same. I haven't tried with Amiga Diagrom, however it wouldn't even boot to that ROM when I first noticed this problem.

So what's next? Well I have some new 74F244's coming and I am going to socket the the data path IC's and try replacing rather than piggybacking to see if that helps.

It's a rather annoying fault, that's for sure.

[Maximilian]

DiagRom should work fine, maybe the chip is dead?

Maybe reflow the Agnus socket?

[rubber_jonnie]

DiagRom should work fine, maybe the chip is dead?

Maybe reflow the Agnus socket?

It has a new Agnus socket. I presume DiagROM requires RAM which is why it doesn't start.

[Maximilian]

It has a new Agnus socket. I presume DiagROM requires RAM which is why it doesn't start.

Reflowing the socket could help if the solder didn't flow through completely while changing the socket.
Maybe the through hole plating was damaged, so a reflow could help.

I'm not completely sure, but I think Diagrom should give flashes/colors and serial output even if there is no available ram.

[rubber_jonnie]

It has a new Agnus socket. I presume DiagROM requires RAM which is why it doesn't start.

Reflowing the socket could help if the solder didn't flow through completely while changing the socket.
Maybe the through hole plating was damaged, so a reflow could help.

I'm not completely sure, but I think Diagrom should give flashes/colors and serial output even if there is no available ram.

I will reflow it when I fit sockets for the data path ICs and I have some replacement 74F244s to try as well.

I'm not convinced it's the Agnus socket from the point of view that it had these issues before I changed the socket.

I had previously cleared the holes for the rest of the onboard RAM and never fitted it, so I refilled the holes just in case there was a cracked via.

[rubber_jonnie]

OK, so today I reflowed the Gary, Agnus and CPU sockets. I also removed the data path IC's, U10-13, and fitted sockets and brand new tested ICs there. The old ones tested fine, but I'll feel better with new.

I also replaced U34 and U35 with new tested IC's. It turned out U34 was actually bad, so now both are new and both are socketed.

Incidentally I have the original ROM and CPU installed, so no fancy mods other than RGB2HDMI, and that is working.

However, I still have the same green screen. It did used to boot, but now I just get the green screen. Which as I still understand it, is that the machine couldn't find at least 256KB of onboard RAM.

So I got out my thermal camera and all 4 onboard RAM IC's are cold. By that I mean only ambient, like it isn't even powered up, though it seems to have 5v. If I plug in the trapdoor RAM in, that does get warmer than ambient.

So it looks like there is a RAM issue, so next up I'll break out the oscilloscope and look at what's going on, and I should probably check the address and data lines for continuity.

I suppose with just 4 RAM ICs I could just socket them all, but I would like to know what this is as it appears to have been a slow fail.

[HigashiJun]

It reminds me of my first ever retro computer rapair.

It was also an AMIGA 500 board and I had a headache with a RAM issue.

It ended on a local auction site for parts...

Anyway, good luck !

[rubber_jonnie]

I suspect that something was failing and has now died.

I know it's none of the custom chips as I have spares for all of them and it doesn't work with any.

I'll get it on the scope later, see what the RAM tells me.

[stephen_usher]

Sod's law states that it will be the last component left to replace after you've done all the others, chasing red herrings.