exxos's Atari, Amiga & retro forum

My 1040 STf suddenly died yesterday :shock:

Motherboard model: C070523-001 REV.D
There are no modifications applied to this ST, everything stock.

This is what happened:

• ST didn't want to run colour monitor (SC1224, works great on my STfm) anymore (looks like 70 hz signal going out, screen running vertically, no distinguishable graphics)
• Mono works fine still (SM124)
• Started basic from the language disk
• After a minute the ST froze, no mouse pointer visible but screen still visible
• Power off, power on: nothing happens. Screen stays black (although a single frame can be seen with like 5% of non-blackness, looking distinctively like analogue distortion, nothing digital). The drive light doesn't light up. No digital activity of some sorts can be perceived.

What I've tried:

• I always smell first (general life advice) but nothing out of the ordinary.
• Disconnect the disk drive
• Look for anything that could cause a short
• Tried 3 different PSUs which all show decent Vpp readings on the scope when sitting in this ST
• Pushed shifter firmly in its socket
• Reseated glue/mmu chips after cleaning with cleaning alcohol

AND SUDDENLY IT WORKS!

The last steps were clearly the solution in my case, I already suspected the glue or mmu chips being the culprit here because apparently ram wasn't even accessed.

So while I fixed the problem while writing this post, I still wanted to share it because someone with the same symptoms might find it useful.

:bravo:

Glad we could help :lol:

:lol:

It's interesting though, how the ST would not switch to 50 Hz and keep sending 70 Hz to the colour monitor. That's probably the glue chip going out of its mind, right? I don't know too much about how everything works together but as far as I know the glue chip keeps the shifter constantly under control, including timings based on whether the ST runs in mono/colour. Right?

Glad its working again and nothing serious :D

Is there some science behind chips coming loose suddenly almost 30 years later?
I've always wondered, there's no moving parts to wear out and its not like your waving the thing about all the time.

Thermal expansion/cycling??? tiny moves adding up over time to create a bad connection?

Yeah, I've heard this is design error on Atari's part. Thermal effects slowly causing the chip to creep out of its socket, regardless of the retainer bar.

It's weird, I have a test mainboard that I'm running a PAK accelerator in, but also had a memory upgrade that I had to remove, including removing the Shifter adapter, but I forgot to push the Shifter back into its socket, but it ran fine just sitting in the socket.

rubber_jonnie wrote: 26 Aug 2018 14:24 It's weird, I have a test mainboard that I'm running a PAK accelerator in, but also had a memory upgrade that I had to remove, including removing the Shifter adapter, but I forgot to push the Shifter back into its socket, but it ran fine just sitting in the socket.

I don't think pushing the shifter fixed my problem here, I'm pretty sure it was either the glue or the mmu. I did it just to be sure though.

I have run into issues with the PLCC socketed chips too, I think also one of the problems is that the contacts are not gold plated on either the chip or the socket, so there is a chance for both the chip pins and the socket pins to have oxidisation occur over a long period, which leads to poor connectivity, and that push can be enough to give the contacts a quick scrape and clean them sufficiently to restore contact.

I wouldn't rule out chip creep either though.

Today the ST started acting up again. Somehow it seems to be stuck in 70 Hz for colour resolutions.

The screen flickers exactly like it does when holding the reset button. I can see the desktop and everything but the picture is running like crazy.

Sounds like a glue chip leg needs more cleaning...