0

So I've been having BSOD problems for the last two years or so, and not been able to track down the problem. Among other things, I ran MemTest86 a half-dozen times for as many as 48 hours, and it never revealed a single fault. I recently upgraded my mobo and RAM, and started experiencing the same problems again. Naturally, I assume it's the same bad driver, that I just can't find.

However, being the circumspect troubleshooter I am, I admitted to myself that this new RAM might actually be bad. I decided to run MemTest86 and Windows Memory Diagnostic again. This time, WMD showed a fault of some kind.

Having never discovered what the other problem might have been, I'm reticent to blame all my troubles on this one fault.

The only hardware I haven't replaced or completely ruled out are the Power Supply and the CPU. So, my question is is it possible for something other than a bad module to create a fault in WMD and/or MemTest86? Could the power supply be doing something weird that would make the RAM seem bad when it really isn't? Could the CPU have a non-critical fault that occasionally corrupts memory, but doesn't cause any other problems?

How much should I trust the memory diagnostic?

8
  • ...I provided context in an effort to preempt "Why wouldn't you just RMA the modules?" ... well, because it might not be a bad module, if MemTest can give false positives.
    – mHurley
    Nov 11, 2015 at 20:13
  • yes there are claims that memtest86 can have false error , it is rare and specific. To trust any diagnostic would be to have a "control" to know how it would act normally on your machine, then to repeat that when something is changed or going bad. Can power to the MB cause ram issues, yes, can the motheboards powering regulation stuff cause ram issues , yes. Is it likely that your BSOD is caused by the ram you spend hours to test? What is the BSOD? The cpu does not usually up and fail, but a very common fail for ram can be the cpu socket connection (now that the ram control is on the cpu)
    – Psycogeek
    Nov 12, 2015 at 6:51
  • One last note, rams can also continue to heat when using a hard ram testing, and fail when they reach a very high temp like 80*C+ or way more. Windows memory diagnostics can up and miss a problem that memtest86 could get a fail on, memtest 86 can be way more abusive to the ram. When WMD failed out of the blue, that is weird, wall power surge around then? What the Windows said about the fail, just like the Bsod your trying to track are both ultimately important to Edit into your question.
    – Psycogeek
    Nov 12, 2015 at 7:01
  • Thanks for the information. That's pretty much what I was looking to find out. Basically, I still don't know what the problem is ;-) Oh well. I took out two of my 4 modules last night and tested them all night with no faults. I'll try the other two tonight and see if I get any faults. If I DON'T get any faults, that will be frustrating, but will at least give me a better idea of where to look. Thanks again.
    – mHurley
    Nov 12, 2015 at 13:52
  • Just ran a test on all four modules, two at a time, and none of them failed. sigh
    – mHurley
    Nov 13, 2015 at 14:35

0

You must log in to answer this question.

Browse other questions tagged .