Random crashes, PLZ HELP!

[CowMan]

Ok, here's the problem. At least once a day I get a random reboot or crash. Usually it's a total restart without any warning but sometimes I get the BSOD and the message: DRIVER_IRQL_LESS_OR_EQUAL
STOP 0x000000D1 (0x00000000, 0x00000002, 0x00000000, 0x00000000)
BEGINING DUMP OF PHYSICAL MEMORY

My system specs are in my sig, so that's what I'm working with. My temps are around 40C at all times so heat isn't the problem. The power supply never wavers, so it's not that, and both sticks of ram (after lots of testing) appear to be fine. I thought it might be the RAM at first, so I tested it in my old computer which houses a Duron 850. I ran MEMTEST86+ and Prime95 for a couple hours for each setting I could think of. It ran flawlessly with the RAM at 100 and never crashed. Then I put my 2500+ in it and ran it at 100 and then 133 (this particular mobo doesn't support above 133 fsb) and it ran flawlessly as well. So now I've put everything back in my normal system. Here's what I've tried doing so far:

RUN MEMTEST86+
1) CPU and RAM at 166FSB -> Crashes within 5-30 minutes
2) CPU at 133, RAM at 166 -> Stable for at least 3 passes (stopped after that)
3) CPU at 133, RAM at 200 -> Stable for 8 hours (never crashed)

Here's the real weird thing. After running the CPU at 133 and the RAM at 200 for 8 hours without fail, I drop the RAM back to 133 to match the FSB, go into Windows, and BAM! I get the dreaded BSOD within 10 minutes. Honestly now, wtf?!? I don't have access to a spare mobo, 2500+, or more RAM without buying it. How can I tell whether it's the RAM, CPU, or motherboard now?! Please help!

||||||||||||||||||||||||||||||||
Barton 2500+ (stock)
Leadtek K7NCR18-D nforce2
2x256mb Geil pc-3200
Ultra XConnect 500W PSU
WD 80gb SE
BFG GeForce 6800GT OC

 

[4Ryan6]

RUN MEMTEST86+
1) CPU and RAM at 166FSB -> Crashes within 5-30 minutes
2) CPU at 133, RAM at 166 -> Stable for at least 3 passes (stopped after that)
3) CPU at 133, RAM at 200 -> Stable for 8 hours (never crashed)

I've lost count at the times I've run Memtest, I've not yet had it crash a machine? Memtest runs independent of any drivers loaded at all, so you're literally dealing with CPU, Chipset, Memory, interaction. Report errors Yes! But crash a machine No!

Have you previously OC'd this machine?

Is your RAM running at default settings or do you have the settings tweaked?

Have you tried clearing the CMOS and resetting all the values to default?

You should always run Memtest to diagnose problems at the settings your CPU's stock FSB is which for yours is 333mhz with a 166 setting, and not be constantly changing the settings, at the right settings you're reporting Crashes within 5-30 minutes.

No errors at all before the crash???????????

What doesn't make sense at all is the time variance of 5-30min, that would indicate an intermittent failure of some kind, sometimes heat related sometimes just a defective product, you say your temp is around 40c, I assume thats the CPU DIE reading, what is the actual M/B chipset/system temp?

If the problem is heat related and it crashes in 5 min, theres just no way it will run for 30 min, unless you've done something to change the environment like opening the side of the case indicating poor case airflow if the time to crash extends or putting a fan blowing into the case as some do, to change the heat parameters.

What I'm saying is usually a heat related crash with no memory errors, will occur around the same time interval maybe a couple minutes difference but not 30min.

If I was to guess at your problem I would say after all the Memtests you've run, if you never got even 1 error on the Memory, that you probably have a bad M/B chipset and I'd replace the M/B.






<A HREF="http://forums.btvillarin.com/index.php?act=ST&f=41&t=2541&st=0#entry20385" target="_new">My Puter</A>

 

[CowMan]

Ok, glad someone's replied. Here's what's more interesting. I've had several ocassions where after booting into Windows, the computer will crash after about a minute. One day, after turning it on in the morning, it crashed twice within 5 minutes. Then, later, when I wanted to try playing a game, it was stable for about 2 hours. The 40c temperature is the CPU die, and airflow wasn't considered an issue because I have a very clean (just cleared out the dust yesterday) Thermaltake SilentBoost fan on the core, and right above it rests my 120mm side intake fan. The northbridge chipset is cooled by a Thermaltake crystal orb and it hasn't had any troubles as far as I can tell.

When I ran Memtest and it crashed, I had it on stock speeds. CPU and ram at 166fsb, so a standard 333Mhz bus. Like I said, it wasn't always instantaneous but it would eventually crash. This is another reason why I don't think it's a heat issue. I turned it on today to test it at stock speeds and it crashed after 10 minutes. Then I rebooted it and tried it again and it made it through all the tests once and then crashed about 10% of the way through the second go-round. The progress bars and walltime clock will freeze, I won't be able to access any of the commands, and the really weird this is that the "+" on 86+ keeps flashing like it's doing something.

The computer is currently stable at 133 FSB (in fact, I'm typing on it right now) but the 20% performance loss is noticable in windows and sure as hell will be in games.

||||||||||||||||||||||||||||||||
Barton 2500+ (stock)
Leadtek K7NCR18-D nforce2
2x256mb Geil pc-3200
Ultra XConnect 500W PSU
WD 80gb SE
BFG GeForce 6800GT OC

 

[sbchemist]

I've been having the same problems as you CowMan. I have no clue what the cause is. My comp would crash a few times with the same BSOD and then I would get hours of stability.

 

[pat]

bad capacitors on motherboard anybody?

<font color=red>Sig space for rent. make your offer.</font color=red>