Help solving seemingly random crashes...

Socks

Socks

Socks

Associate
Joined
2 Jan 2004
Posts
14
Alright, well, my computer first started BSODing about a month ago. At first it only happened sporadically, probably once a week or so, mebbe twice if particularly unlucky. I was concerned at this point, but due to how infrequent the problem was I had no particular drive to fix it. Anyway, about 2 or so weeks after the initial crashing began, It started to happen about once every 2 days, sometimes once a day. At this point it's obviously a serious problem and had to be diagnosed, which i've attempted to do with very little luck so far.

First, a bit about the BSOD error messages. There is NO pattern as far as I can tell. They vary completely from the error name to what caused the error. I'd post some dump logs but i've done a reformat recently (more on that later) and therefore i've only got the crash logs since then.

Okay, so, since the BSOD errors were so completely random and had no discernable connection, I took that as a clue in itself, possibly meaning that either the PSU or memory were failing and therefore giving the illusion that other components/software were at fault (the dump reports had all sorts of reasons for what could have caused the error).

I then replaced my PSU without hestitation, mainly because I've had a few issues with it in the past (while gaming) and I've just been looking for an excuse to get a new one. So I ordered a nice Corsair 520w jobby and it's been sitting in my computer ever since. Unfortunately the BSOD problem remains...however my gaming power issues have dissapeared.

So yea, I moved onto the memory from here. I ran the latest version of Memtest for a good 10 hours with absolutely no faults at all, both sticks. I'm aware that memory can still be faulty even if it passes Memtest successfully, but I felt that with the frequency of the crashes by this point (sometimes up to 7 in a day), Memtest would damn well find something if memory was the culprit. So yea, clean memtest, still no luck, crashing remains :/

I'm generally outta ideas at this point, especially in concerns to hardware, so I thought about the software side of things. I've been running this computer for about 2+ years, possibly 3+, without reformatting once. So suffice to say my boot disk is pretty messy and i've done all sorts of various installations and uninstallations over the years. I figure a reformat can't hurt because even if the problem isn't solved, my computer still desperately needs one. Also, the BSOD dump reports did hint at driver faults fairly frequently, so i figured a clean Windows install with fresh drivers might do the trick.

Alas, no luck, which brings us to exactly where I am now. I've got a brand new shiny PSU, clean memory (according to Memtest anyway) and a fresh installation of Windows...but the problem remains.

Overall, I'm really all out of GOOD ideas. I do have a few ideas left like flashing the motherboard and removing/turning off any PCI devices one by one to see if the problem stops, but those are more wishful thinking solutions than anything else.

So yea, I'm hoping you guys will have some decent ideas/suggestions. If you need any more info from me, just ask.

PC Spec:

3.0 Ghz P4
Abit IC7-G Mobo (onboard sound turned off)
2x 512mb Corsair Twin sticks
Corsair 520w PSU
ATi x800 XT
3x Seagate HDDs (120, 120, 500)
Audigy 2 sound card

OS: Windows XP Pro SP2 (all recent critical updates)

(all on default voltages, apart from memory which I needed to increase upon initial PC build due to memtest errors)

Minidump logs (since reformat):

Date: Wed Oct 24 11:06
Error: IRQL_NOT_LESS_OR_EQUAL (a)
Process_name: msnmsgr.exe

Date: Wed Oct 24 11:31
Error: PAGE_FAULT_IN_NONPAGED_AREA (50)
Process_name: msnmsgr.exe

Date: Wed Oct 24 11:41
Error: IRQL_NOT_LESS_OR_EQUAL (a)
Process_name: utorrent.exe

Date: Wed Oct 24 12:43
Error: IRQL_NOT_LESS_OR_EQUAL (a)
Process_name: Idle

The dump logs are pretty big, so I wasn't sure exactly what info you need from them, but just ask and I can post more info. Btw, i know 3 of those errors are the same type, and certain ones have definitely occured more than others, but as a whole they really are random. If i posted all the errors from before the reformat you wouldn't find any pattern.

Btw, i'm really sorry this post ended up so huge, but i wanted to describe my situation in full.
 
The errors your getting suggest overheating or dodgy ram, what temp is your cpu running at? As the system is a few years old have you checked the heatsink isn't clogged with dust.

You could also try running the system with just one stick of memory installed and see if you still get the errors then.
 
Here's my current temps, according to Everest.

Mobo - 30 C
CPU - 45 C
PWM (PSU?) - 32 C
GPU - 38 C
GPU Ambient - 35 C

HDDS, all 3, between 30 - 32 C

As for the dust suggestion, It probably wouldn't hurt to take off the HSF and give it a clean, but it's a Zalman artic cooler that I got less than a year ago, so i doubt it's that dusty yet. Also, When I installed my new PSU a few weeks ago, I gave the whole inside of the case a fairly decent clean.

Edit: Zalman Artic? Sorry, Zalman Flower :P
 
Last edited:
Or HDD? Can you run the PC with only the boot HDD for a while and see if it's one of the other two? If not it'd rule them out of the problem at any rate.

Also, what settings have you got you memory dialled into? If it's Auto SPD it might need manually dialling in.
 
After reformatting, i actually did run the system for a good few hours (in which it crashed) without the SATA drivers, which are required for my other 2 HDDs, not the boot disk. So they were plugged in, but Windows couldn't see them and what have you, so I dunno if that counts.

As for memory, yes, the motherboard has it on Auto SPD. I originally bought a Corsair TwinX pack with guaranteed SPD settings of 2-2-2-5 (or is it 2-2-5-2?). I've checked the BIOs and that is what they're set at by Auto SPD. Keep in mind though, i've had them on Auto SPD for years, so i'm a little confused why having them on auto would only be a problem now? Unless you're suggesting the settings need to be loosened?
 
Okay, scrub disks then - unless it's your primary drive which I'm guessing is gonna be a pain to test with. (Linux live CD maybe?)

I'd be less inclined to think it was memory if it was working OK. However, I'd still dial it all in properly though just to rule it out - make sure the voltage is spot on too. Every Corsair module I've used need to be explicitly dialled in unless I was just unlucky.
 
I had 3 sticks of Corsair running fine for 2-3 years and suddenly errors.
I thought hard disc, but it turned out to be a bad stick of ram.

Just shows it can die.

I have the feeling we might need to see the full dumps here.....they might reveal something.
 
Randal, consider it done in regards to manual memory timings. I'll see if that makes any difference.

I'll happily post the full memory dumps, but first is it possible to attach files to a post? Otherwise i'll have to paste gargantuan amounts of text...

Edit: Crash dumps*
 
Last edited:
Sorry i don't :(

I searched file upload on google and found a site called Mediafire, so i've uploaded all 4 crash logs (thus far) to there, here's some links:

First Crash:
http://www.mediafire.com/?3x23jxtdxte

Second Crash:
http://www.mediafire.com/?am5jmdmlojz

Third Crash:
http://www.mediafire.com/?bezx2fiboxm

Fourth Crash:
http://www.mediafire.com/?ewm9xc1mx0c

Hopefully those links will work. Anyway, i gotta go back to work for a bit now. I'll give those Seagate tools a try when i can too.

Thanks for all the help and suggestions thus far guys. I honestly didn't expect such a rapid response.
 
Alright, sorry for not updating, but i've been busy and made little progress with the problem up until now.

I took out both of my memory sticks and borrowed a spare memory stick from a mate. I determined this was the only way I could truely 100% rule out memory as the cause of the crashes, since I wasn't entirely sure whether Memtest86+, or the Windows based Memtest, were telling me the whole truth. Anyway, turns out that it does seem memory related, as my computer has been running flawlessly on this spare stick for a while now.

So today I've been trying various things to further pinpoint where the exact problem is. I had a good look at both my memory sticks and noticed one of them had siginificant dust build up on one side, which I believe was sitting in slot DIMM1, nearest the CPU. Unfortunately I wasted a large chunk of my day attempting to remove the heatspreader from that particular side to give it a real good clean. I gave up after a while convinced that it wasn't gonna budge and therefore i've cleaned it best I can with it attached. I did a fairly decent job and then proceeded to put the memory back in, cept this time in slots DIMM2 and DIMM4. Anyway, comp booted, I went about playing on my comp like I do before it crashed aproximately 2 hours later.

So now I decided to test both sticks individually in DIMM1. I tried the formerly dusty one first and my computer was a disaster...Windows just couldn't load. The computer kept crashing during the Windows loading screen (at different points during the load) and rebooting back to square one. So i opened the PC back up again and tried the other stick, voila, Windows booted first time and it's seeming stable thus far.

So yea, my plan is to purchase some new memory unless I get any replies with better ideas. I also wouldn't mind if someone can give me an explanation as to why either of the Memtests i tried picked up NOTHING. Seriously, I just don't get it, I mean Windows couldn't even load with that one particular stick, and I personally consider that a sign of rather severely botched memory...yet Memtest picks up nothing?!? How is that possible?!?!

I just wouldn't mind hearing any theories or explanations you guys could possibly provide.
 
You must log in or register to reply here.
Back
Top Bottom