MPC 3K LE randomly crashes - reboots OK the crashes again

Hello all,

I have an MPC 3000 LE which shows a strange problem, making it quite unusable. I'm looking for some ideas about what it could be. Any helps would be super appreciated.
Please find below a list of details about the situation:

1. When I switch it ON, it boots OK, I can use it normally, until it randomly crashes after 3h... or 10min,
2. Then if I try to switch it again, it sometimes boots OK, until it crashes, or doesn't boot immediately,
3. Crash can be a freeze (with info on the screen), or nothing on the screen
4. The model is a MPC 3000 LE (Edition #0824), 32Mo, I had it for at least 10 years,
5. Only mod is Mansell Labs OS Vailixi 3.50 eproms, no gotek,
6. It has an SCSI ZIP 100 drive attached, but It crashes also without the drive connected (tested with no SCSI cable),
7. Now the fun part: I gave it to a local service shop, they couldn't replicate the fault. It worked flawlessly at their place for a day.
8. I took it back yesterday, and... crashes within 10min ON.

At this point, I am kind of lost, any insight would be appreciated. I remember this could be related to some fuse, but if so, some guidance would be helpful.

I wish you a good day!

MH

It‘s rather unlikely that a fuse is the source of the trouble, except a half burned one on the SCSI termination line, but that‘s a different story.

If crashes are memory based, the Simm might not sit properly, which includes corroded contacts.
The memory itself may be unreliable. Replacements may be difficult to source, but they do exist.

The usual suspects in such cases are electrolytic capacitors (power instability/memory instability). Inspect the PSU for bulky or leaking big ones, in particular the 5V rail.
There‘s a 12000, a 2200 and 100yF plus a 10yF tantalum cap (C807).
Check the latter for damage like a small hole, possibly with „smokey“ surrounding.
(if it‘s a drop-like shape)

Btw if you switch on the MPC directly after a crash, memory state may still be buffered by the big 12000yF cap.
To discharge the cap, unplug the power cable from the wall outlet, switch on/off a couple of times, and then connect the cable again and power on again.
I‘d expect the time until the next crash happens to be longer than before.

I would start inspecting the power supply or even better change for a new modern one. The old one is around 30 years old. Change to a new one and see if it still occurs. Contact me for a power supply in Germany/Europe with 100V-240V which is usable worldwide without any "setting" or use of an adapter.

Hm, always tricky to pinpoint these kind of errors correctly but a good check-up of the power supply would be a good starting point. The second thing that comes to mind is temperature? How did the local shop test the unit and is this a bit of the same as when u are working on it? Or did they tested it whit the engine hood opens up? Would make a bit of sense reading your times between errors .

Yours
Danny

Hello all and thank you for the insights!

The second thing that comes to mind is temperature? How did the local shop test the unit and is this a bit of the same as when u are working on it?

I had this in mind, as the first signs of failure happened last summer, during a warm day. Yet, it also happened this winter (after a long hiatus when I didn't switched it on). As for the shop, they let it on with the lid closed during 1 day with no failure on their side to their say. Still, they mentioned that they checked the power supply in details.

Contact me for a power supply in Germany/Europe with 100V-240V which is usable worldwide without any "setting" or use of an adapter

I'll sure contact you if needed. Thanks for the suggestion.

I'll try to have one of the guys of this shop come at my place so that I don't move the MPC anymore, and that he can check all connections. I'll post when/if any progress.

G'd day.