Source of Vista64 BSOD? RAM, mobo, or PS?

1302 Views 5 Replies 2 Participants Last post by Orumph, Sep 23, 2008

ryanclemson

Discussion Starter · Sep 21, 2008

Hey all, I need some help determining the cause of a BSOD I am getting. First, specs:

Windows Vista Business 64-bit
Gigabyte GA-EP35-DS3L
4GB Crucial Ballistix (4 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400)
Intel Core 2 Duo E6550 Conroe 2.33GHz

I built the system in May and have had zero problems until about a month ago. Oddly enough, the problems started occuring after a storm. I suspect the storm may have affected my system because lightning struck the building, and knocked out part of the buzzer system. (My PC is only hooked up to a suge protector, no UPS).

The system began BSODing at irregular intervals. Sometimes, I could go hours without a BSOD, othertimes only minutes. However, I did notice that it would BSOD frequently after returning from Sleep mode.

At first, I ran Memtest, which produced a bunch of errors. Thinking the RAM was immeadiately the problem, I opened the case to remove a couple of sticks to isolate the problem. While the case was open, I noticed one of the sticks was improperly seated. Both clips were down on one of the sticks, but one wasn't fully locked. Thinking this is the source of the problem. I make sure all of the RAM is fully seated and locked, reconnect the power to the power supply and run the memory test again. After a couple of hours, I receive no errors, so I think I have found the problem.

However, BSODs continue at random. Today I finally found the time to isolate this issue, and beginning to suspect at virus or other software issue, I reformat the hard drive and reinstall Vista 64. Still, BSODs continue, so I decided i need to start isolating the problem again. I run another memtest, and find 200-some errors in eleven minutes. I disconnect the power supply from the power cord, reboot and run memtest again - no errors this time. At this point I feel like I can't be sure if the RAM or the mobo is causing the error, but I suspect it is one of the two (and maybe the power supply). CPU temps seem normal, and I usually recieve stop error code of 1a, but today also received a 50, a 24, a 3b, and a de.

I am currently installing the tools to analyze my minidumps, so I might have more information soon, but I believe this is a hardware problem. Any suggestions from anyone would be greatly appreciated.

See less See more

Save

Status: Not open for further replies.

1 - 6 of 6 Posts

Orumph

· #2 · Sep 21, 2008

First, I would run the memtest again one stick at a time in one slot at a time.
Meaning,,,,

Take one stick in one slot, if it errors, move it to the next slot, or try a diffrent stick, if it errors again, it is possibly a bad slot or stick.
then move to the next slot and do the same, if the memory errors again in the second slot try a different stick, if no errors then try that stick in the first slot, if that errors, then it's a bad slot, but you could also have more than one bad stick. So keep checking till you narrow it down.

Also, yes, the Power Supply could be the problem, so I would get a known good PSU to test with, in a different wall power socket and eliminate the power strip. Then try the mem testing. Lightening strikes can cause some very weird issues and the problems can be time consuming to narrow down.

If you replace the PSU and run memtests and get no errors, you could assum the PSU is the problem, otherwise, it could be bad memory or mobo

Save

ryanclemson

Discussion Starter · #3 · Sep 22, 2008

okay, i isolated the sticks on ram and still no go.

first, i tested all 4 sticks in their original slots for about 20 minutes. nothing.

i then tested stick 4 in slot 1, then slot 2, then slot 3 each for about 15 minutes. nothing.

this should mean that all four slots and all four sticks are good right? just for kicks i'm now going to test stick 4 in slot 1, stick 3 in slot 2, stick 2 in slot 3, and stick 1 in slot 4 for at least an hour, although i doubt i will get any errors - othertimes, when i would do ram tests, i would get errors after 30 seconds or so. i ran a memory test the other night for two hours after a bsod for two hours on all four sticks without errors.

anyone know what could possibly cause erradic bsods/memory tests?

Save

Orumph

· #4 · Sep 22, 2008

i ran a memory test the other night for two hours after a bsod for two hours on all four sticks without errors.

well,, 15min or even an hour, and I should have said this and didn't, apologies.
for complete memtests, they should be run 4 to 6 hours on all of them. But for quick tests, atleast one or 2 full passes should be done for each stick and slot.This is in no way considered absolutly good memory, but may give a good indication to some degree that the slot might be bad.
If nothing fails on that, then run them for the full 4 to 6 hours and see fi you get any errors.
If you do, you need to check the memory itself in another system. If it passes there, then you definitely have bad slots therefore a bad mobo.

Second,,,, I should have also recommended running Manufacturers Hard Drive Diags.

It could be a bad Power Supply. You need to try replacing this.

If all this passes and the PSU doesn't fix the problem and you ahve already reformated and reinstalled the OS, then you have a bad Mobo.

Save

ryanclemson

Discussion Starter · #5 · Sep 23, 2008

thanks for the reply. Overnight, I ran a memtest on all four sticks. It ran for 10 hours with zero errors so I am starting to think that the memory is okay.

As for the mobo, I have reinstalled Vista64 and have only had 1 BSOD, which occured the first time I tried downloading a file in IE. Since, I have been installing antivirus, windows update, etc. without any issues at all.

However, I still suspect the mobo, and possibly the PS. I know of a friend who has a spare I can try, but hopefully the reformat has fixed whatever issue was causing the bsods.

I think I'm going to run [email protected] overnight unless someone out there knows of a better, free stress test program for the x64 platform.

Save

Orumph

· #6 · Sep 23, 2008

I have never heard of [email protected], But I would check passmark burnin test for a commercial solution,,

Yes, if the ram test passed after 10 hours, it is safe to assume the slots and ram are fine.

Yes, the mobo could still be bad.

Still check the PSU, but also run manufacturers hard drive diags.

if [email protected] is anything like Prime95 then it's not really going to stress much more than the ram and CPU, maybe other parts of the board, but Prime95 is mainly used to stress test stability of overclocking a system, not for stress test trouble shooting or burn-in.

Save

1 - 6 of 6 Posts

Status: Not open for further replies.

Top Contributors this Month