Acer Aspire 8730G, MB: Wistron 08223-2, Restarts with RAM in both slots

Brand: Acer
Model: Aspire 8730G
Motherboard: Wistron Big Bear 2, 08223-2, 48.4AV01.021
***************************************************************
Hello!

About two years ago I was working on this "old guy" with a problem it randomly shut down while in use. It was a standard issue - bad "NEC/TOKIN" capacitor under the CPU (on another side of the board). I replaced this capacitor with a six pieces of 220µF/6.3V tantalum capacitors.

Now this laptop has returned with a similar, but different problem - it shuts down and restarts automatically while in use. It turns out this only happens when RAM memories are inserted into both slots. In the case that only one RAM is inserted into either of the two slots, the laptop works flawlessly.
The issue only occurs when the laptop is booted into OS. If I leave the laptop turned on the BIOS screen the issue does not occur.

I did the following:
- Freshly installed OS on another disk with only one of the two RAMs inserted, otherwise the laptop shut down and restart during installation.
- Both RAM memories were replaced with known good ones.
- Original CPU Intel Pentium T4200 (SLGJN) were replaced with Core 2 Duo P8400 (SLB3R).
- Checked all six replacement tantalum capacitors at the place of "NEC/TOKIN".

None of the above measures changed anything.
I'm out of ideas, please help.

Does anyone have any ideas?

tekort napisał(a):I would try resoldering the RAM slots, maybe some pins have loose contact.

I pushed each pin of both RAM slots aside with a needle under a microscope and all pins are solid.

tekort napisał(a):Also I would test it with a memory tester to see if all signals between CPU > RAM are good on both slots.

I did some more tests with a different RAM sticks and found the following:
- With two identical RAM sticks (2x 1GB sticks or 2x 2GB sticks) it usually shuts down and restarts automatically after a few minutes in the system. Sometimes this already happens while the system is loading.
- With two different RAM sticks (1GB + 2GB sticks) the issue only appears after several hours of operation.

During operation, I pressed and slightly twisted the RAM sticks in different directions, and this does not affect the operation.

Any other ideas?

Have you tried replacing those "new" capacitors under CPU socket? If so, I'm 99.5% sure that the CPU socket is the culprit here. Had a very similar issue with those MLBs while repairing for Acer long time ago, most ended up with CPU socket replacement as a successful repair. You may try to reflow it first - I could not do that on any component due to Acer's strict "fail-replace" policy.

Vogelek23 napisał(a):Have you tried replacing those "new" capacitors under CPU socket?

I measured the capacitance on the VCC_CORE power rail in reference to ground (with the CPU removed of course), and the meter shows a capacitance of 1330µF.
I also measured the voltage on this power rail to ground with an oscilloscope (with the CPU connected) and I got a ripple of about 50mV p-p, excluding the voltage ramping generated by the power supply controller chip, which is also clearly visible.

Vogelek23 napisał(a):If so, I'm 99.5% sure that the CPU socket is the culprit here.
You may try to reflow it first

I reflowed the CPU socket (with a nudge test), and the laptop worked flawlessly for about two days, then the initial problem reappeared .
But I can say that it appears a little less frequently than originally.

tekort napisał(a):I would try resoldering the RAM slots, maybe some pins have loose contact.

Just in case, I also resoldered (drag the solder all over the pins) both RAM slots, but nothing changed.

Do you think I need to replace the CPU socket, or is there anything else I can do first?

matic napisał(a):Do you think I need to replace the CPU socket

Yes I think so.