New system, isolating the problem...

[ry]

So, I just put together a new system. Pentium D 940, ASUS P5N32-SLI MB, OCZ 2 GB DDR2 667 Ram, EVGA 7600GT, Seagate Barracuda 7200.9 300GB HD, 430W Antec Power supply in the Antec P150 Case. My CPU is running about 70 F and the motherboard is hitting around 95 or so.

Have Windows XP Pro installed. Updated drivers for everything. Now, I will be playing a game and then all the sudden get a blue screen, reboots before I can really see what is going on. Check the system event viewer and it seems to be giving me HD warnings when it crashes. Then after that happened a few times, it started not booting about 1 out of 3 times without installing any new hardware. Any ideas on a place to start to isolate the problem? Just don't know what to start sending back to get replaced...

Thanks.

 

[doolittle]

My CPU is running about 70 F and the motherboard is hitting around 95 or so.

Hmm, maybe your sensors are flimm-flammed that is way hot for mobo temps.

I would feel the northbridge heatsink and the voltage regulator temps (near the cpu socket) if it is uncomfortable, try leaving the lid off and getting some extra cooling (like a standalone room fan blowing in) in there to see if that helps solve the problem.

the system event viewer and it seems to be giving me HD warnings when it crashes.

Could be heat related, feel the HDDs also. The fan should help in this case too...
Get the ultimate bootcd everyone should have this. It has every hdd diagnostic give it a run, if you get errors do you have a different cable to try? That could be your problems if you're lucky, if you have to rma the hdd then that sucks... unless you have a spare handy...

 

[sailer]

A couple guesses. Your power supply might not be giving enough juice to the hard drive, or a connecting wire might be loose. Try reseating the wires just to make sure. Another possibility is the hardrive overheating. Make sure that its clear of obstructions and the air flow is good. A fan on the HD itself might help.

This does sound more like a power supply problem to me than anything else, whether the power supply is bad or just loose connection, I don't know. If after making sure the connections are good, there are still problems, try a spare power supply if you can. If the problem then goes away, then your power supply is bad.

 

[shadowduck]

My CPU is running about 70 F and the motherboard is hitting around 95 or so.

Hmm, maybe your sensors are flimm-flammed that is way hot for mobo temps.

I would feel the northbridge heatsink and the voltage regulator temps (near the cpu socket) if it is uncomfortable, try leaving the lid off and getting some extra cooling (like a standalone room fan blowing in) in there to see if that helps solve the problem.

70 and 95 F are not that hot for motherboard temps. Thats like 21C to 35C. Quite normal really.

 

[ry]

Thanks for the ideas. Put some fans on the passive MB heat sinks, didn't change the temp that much. Seems to be a bit more stable the last couple days. Who knows why that is...

Thanks for the ideas, if it keeps dying, I'll probably try the HD first, since it seemed to be the thing that was having trouble. I had already tried reseating the cables, but I haven't tried a new one in it yet. Has to be an adventure doesn't it?

 

[The_Prophecy]

My CPU is running about 70 F and the motherboard is hitting around 95 or so.

Hmm, maybe your sensors are flimm-flammed that is way hot for mobo temps.

I would feel the northbridge heatsink and the voltage regulator temps (near the cpu socket) if it is uncomfortable, try leaving the lid off and getting some extra cooling (like a standalone room fan blowing in) in there to see if that helps solve the problem.

70 and 95 F are not that hot for motherboard temps. Thats like 21C to 35C. Quite normal really.

shadowduck is right, those temps are fine....

But if you're really concerned about the temps that much, grab some Arctic Silver 5 and replace the thermal paste on the CPU, North Bridge, and Video card.

 

[edwuave]

is it only happen during the time u play games? if yes, try to check your graphic card. run the PCMARK2005 and see which test it fails, make sure u observe the test all time. the particular test(CPU,memory,hard disk, graphics) is corresponding to the particular hardware it test. so u may isolate the problem from there.

 

[ry]

Well, I hadn't gone with the stock heatsink/fan, I suppose I should have stated that too. I got a Scythe Ninja cooler with a fan blowing across the fins to the case exhaust fan. I had the Arctic Silver on the CPU as well. So, I was trying to be smart, but maybe outsmarted myself again. This is the first computer I've had in quite awhile that doesn't sound like a jet taking off. It's really nice. (As long as it stays stable.) Thanks for all of the replys. I'll try a couple more of the suggestions and see if I unearth anything new.

 

[ry]

It's not just crashing when I play games.

I ran the memtest86 program and it crashed the first time. I set the timings to what my mem is supposed to be 4-4-4-12 and it made it through the test without locking up, but showed that I had errors in test 5, so that could be a problem.

I'm getting blue screens with IRQL_NOT_LESS_OR_EQUAL. These are happening at random times. Also, when running 3DMARK06, comp crashed with a nv4_dsp error on the second test.

So, it looks like at the very least I may have some bad memory, not sure if the vid card should go back with it though, maybe try one thing at a time.

Not sure what exactly the IRQL_NOT_LESS_OR_EQUAL is, or what is causing it.

 

[The_Prophecy]

video driver problem.... uninstall the video driver, reboot and get XP to use the generic driver. Then go to your card manufacturer's website (not the nvidia website) and get the latest driver provided by them. Make sure the driver you're getting is not a beta version either... you want the latest stable release.

 

[doolittle]

I ran the memtest86 program and it crashed the first time. I set the timings to what my mem is supposed to be 4-4-4-12 and it made it through the test without locking up, but showed that I had errors in test 5, so that could be a problem.

I would try to get it stable (zero errors) to run at least 8 hours - I am wondering if you got the 1.8v or the 1.9v OCZ and what your current VDIMM setting is... Have you tried only a single DIMM of the pair in memtest, and does only one fail and the other is good?

Also try Driver Cleaner for a fresh start w/ those video drivers...

 

[ry]

I would try to get it stable (zero errors) to run at least 8 hours - I am wondering if you got the 1.8v or the 1.9v OCZ and what your current VDIMM setting is... Have you tried only a single DIMM of the pair in memtest, and does only one fail and the other is good?

Also try Driver Cleaner for a fresh start w/ those video drivers...

I have the 1.8 V memory. The first pass I got two errors in test 5 at 1.906 and 1.924 GB. The second time I got 3 errors on test 5, but am not remembering exactly where they were. My current setting for voltage was on Auto, I had tried to set it manually, but it only went down to 1.85 V. I had it like that for a bit, but didn't memtest it at that V.

video driver problem.... uninstall the video driver, reboot and get XP to use the generic driver. Then go to your card manufacturer's website (not the nvidia website) and get the latest driver provided by them. Make sure the driver you're getting is not a beta version either... you want the latest stable release.

I had installed the first version of the drivers since the download was so slow from EVGA. Then I bit the bullet and spent the time to get the updated driver. I uninstalled the driver, then reinstalled the new one. I'll try the driver cleaner program and then reinstall the driver.

I'm going to send back the memory. It totally failed when on auto timings, and when the timings were entered manually, it still had errors. They came as a set, so I have to return them as a set. So, hopefully, in a couple of days I'll have new memory and will be able to reenter the fray. rawr.

Thanks again for the comments.