Possible Memory Error Dell XPS

[Texun]

I'm working on a 6 year old Dell XPS that had been running perfectly since day one. Last week the customer said it crashed. He tried to reboot and it crashed again either right after or during POST. He never saw Windows trying to load.

Specs;
Dell XPS
1x1g Samsung PC3200
1x512M Samsung PC3200
P4 @ 3.0
No HD
HD 4650 AGP\512M
Dell OEM PS

He pulled the HD for now so he could access his files with an external enclosure. I have the XPS now and I'm using a Hirens Boot CD to test it. It booted fine so the first test I ran was Memtest v2.11. When I returned, memtest had completed 3 full passes and showed 147 errors. I cleared the screen before writing down the errors. I took the short cut of skipping the errors since it seemed obvious at that moment that at least one of the two sticks was causing the problem. Of course it wasn't that simple.

Pulled the first Samsung stick and ran memtest again leaving only the one original Samsung stick. No errors at all.

Removed the last remaining stick and filled both DIMM slots with a matched pair of G-Skill 2x1g PC3200. Again, no errors after a full pass.

Removed the G-Skill and reinstalled a single Samsung stick, the first one I pulled because I suspected it might be bad, and again no errors.

Reinstalled the original Samsung PC3200's back in their original slots and ran memtest again, and again there were no friggin errors after one full pass.

So I've gone full circle with Memtest, pulled one, then pulled the other, filled both slots with G-Skill PC3200, reverted back to the originals and I am back where I started but with no errors at all.

Can someone give me a reasonable explanation as to why it crashed, and why memtest reported 147 errors on the first pass but now tests clean with the same memory back in their original slots?

The first thing that come to mind is that reseating the memory cleared the problem. It was just slightly dusty on the inside and the modules felt fully tight before I removed them. I would like to believe something that simple could have fixed the problem but I'm just not convinced. :\

Perhaps something is wrong with the MBR and that's why I can't duplicate it here, but he said it never made it that far and said he saw no error message after the crash.

Thanks!

 

[RebateMonger]

I had a new motherboard that got one memory connector damaged by a DIMM module that was on the thick side and had no chamfer on its edges. That tore up several of the spring connector inside the slot rub against the gold contacts on the memory module.

Under a magnifying glass, I managed to straighten the connectors, but it forever gave memory errors in that slot.

 

[Blain]

Can someone give me a reasonable explanation as to why it crashed, and why memtest reported 147 errors on the first pass but now tests clean with the same memory back in their original slots?

It may not have been the memory...

Troubleshooting Memory Errors

"Please be aware that not all errors reported by Memtest86 are due to bad memory. The test implicitly tests the CPU, L1 and L2 caches as well as the motherboard.
It is impossible for the test to determine what causes the failure to occur. However, most failures will be due to a problem with memory module.
When it is not, the only option is to replace parts until the failure is corrected."

 

[Texun]

It may not have been the memory...

Troubleshooting Memory Errors

"Please be aware that not all errors reported by Memtest86 are due to bad memory. The test implicitly tests the CPU, L1 and L2 caches as well as the motherboard.
It is impossible for the test to determine what causes the failure to occur. However, most failures will be due to a problem with memory module.
When it is not, the only option is to replace parts until the failure is corrected."

I thought about that too right after I posted this. Sure wish I had caught the errors. It ran all night and successfully completed numerous passes with zero errors.