BRAND NEW computer randomly turns off

[dalearyous]

check my sig for my computer specs. i just built it 4 days ago. ive been playing games and doing heavy amounts of work on the machine and i haven't noticed any problems at all. then i started playing warcraft 3 and things started to go wrong. war3 is not a very demanding game so it doesn't make sense to me, ive been playing call of duty 2 etc... a ton and no problems. ive played 2 games of war3 today and my computer randomly turns off around 45 mins into the game. i checked all the temperatures and cpu and mobo is definately cool enough. however, upon restart i saw that the ati CCC showed my temp of vid card was 56 C degrees. i have nothing overclocked. im going to run standard diags on cpu, HD, and memory but i have feeling thats not the problem.

 

[mechBgon]

1) your rig link needs to be changed to this in order for it to work for everyone else too: http://www.anandtech.com/mysystemrig.aspx?rigid=32218

2) you have a heatpipe-cooled motherboard in a case that puts it upside down from its intended orientation. Put the system into a normally-oriented case and see if the problem persists.

 

[dalearyous]

there changed it. umm about your suggestion. im going to say no, thats a ton of work and why would that change anything. i have great airflow and my cpu and mobo idle at 32/42 C. is there anything else i can try. that would be last thing i tried.

 

[mechBgon]

Do you understand how heatpipes work? The coolant runs to the bottom. Because of your upside-down case design, your motherboard's heatpipe has the radiator at the bottom and the hot bits at the top. Just what we need, keep the radiator cool and send the heat up to the hot parts to make them hotter yet

If you don't want to go to the trouble of trying a different case yet, then try this: Lay the case on its side so the motherboard is slots-up, then try again. It could be you've got some other issue, such as an overheating power supply, but I would not pick that case for a heatpipe-cooled motherboard (or any other motherboard, actually, after hearing from an owner of this model about its heat issues).

Good luck with the testing Incidentally, your mobo temperature does not indicate the temperature of the parts cooled by the heatpipe.

 

[dalearyous]

hmm i c now. i thought the heatpipe just transfered the heat, didn't know there was coolant in it. crap i'd hate to have to rma the board and get different one. could be DFI for me afterall.

btw, memtest passed

 

[mechBgon]

Bah, just get a normal case Check out this one, if you want a very classy black one: http://www.endpcnoise.com/cgi-bin/e/p180.html

Full review If I had $180 to blow on a case, that's the one I'd get. If not, the Antec SLK3000-B is pretty sweet for the price, and superior to my own Lian-Li PC-86.

 

[dalearyous]

i dunno, this case is amazing. also its going to be cheaper to send the mobo back than the case.

 

[lopri]

I've been suggesting this very simple mod to everyone who owns the BTX-style case.

Here
http://img498.imageshack.us/my.php?image=dsc001121yz.jpg

Result
http://img498.imageshack.us/my.php?image=temp10wa.gif

 

[dalearyous]

what is that? just a fan that mounts above the chipset?

 

[Kaeishiwaza]

Just as a note, though, although the heatpipe is affected by orientation, having it upside down won't disable it. The coolant returns to the hot core via capillary action, not gravity. However, if you have gravity fighting for you instead of against you, all the better. But the difference in temperatures you should see would be in terms of a few degrees, not in terms of failure to cool at all.

 

[dalearyous]

ok so what were saying is:

keep the board and figure out a way to mount a fan above or close to the chipset?

 

[Kaeishiwaza]

I don't think it was ever really established that cooling was the problem. Your temps don't seem that far out of line, esp if you're not overclocking. Randon restarts can also be attributed to power supply problems. Also, did you get a POST error from windows? You can check your logs.

 

[dalearyous]

its not random restarts though. the thing just shuts off. the only way i can make it happen is by playing war3...nothing else makes it turn off. i checked my logs and no errors could be found anywhere near time it happened.

i played call of duty 2 for like 3 hours and nothing

i ran cpu test for forever nothing happened

ive left it on all day and done various things nothing

i guess your right, only thing left is possible faulty PSU. just don't see how psu would work fine 99 percent and then decide to turn off. i have a UPC battery back up so power fluctuations can be ruled out.

 

[Kaeishiwaza]

I had overheating problems before. Every time the system rebooted due to heat it was because either the RAM or the video card fouled up and gave a HALT command to windows causing it to shut down and reboot. The 'Blue Screen of Death' if you will, but windows doesn't show you that screen any more. You will get the 'Windows has recovered from a serious error' popup when the reboot is finished.

If you're not seeing that, then your hardware didn't generate an error, so something pretty drastic happened. Could be a short, or could be a power problem. If you can, swap your PS out for another one and see if the problem continues. I had a bad Seasonic on my first build and I had a few random crashes before it dies altogether.

 

[imported_nocturne]

Well... before you do anything too drastic, make sure your computer is set to not Automatically Restart in case of BSOD. If it's a driver crash or windows crash, the error code might point you more in the right direction. In most cases it will point you to a non-sp2 driver or bad memory settings in the bios.

Also, check the driver release notes for your gfx card, as there are sometimes noted bugs that could save you a lot of time diagnosing (although the fact it works for 45min. point's to heat).

 

[dalearyous]

woah ok i made my computer do it again playing war3. afterwards i immediately opened my case and i could smell bad burning electronic smell in there (and i have a cold right now) i touched the heatblock on the chipset and its seriously hot i couldn't leave my finger on it at all.

here is error i received from event log:
The COM+ Event System detected a bad return code during its internal processing. HRESULT was 8007041F from line 44 of d:\qxp_slp\com\com1x\src\events\tier1\eventsystemobj.cpp. Please contact Microsoft Product Support Services to report this error.

its got to be heat, specifically on chipset because under good load my cpu usually stays around 37C. still wierd war3 is the only game to make it happen.

the only thing i can think of is:
the start of my vid card air intake is right next to the chipset. call of duty 2 etc... makes my vid card fan run a lot and prolly pulls some heat off the chipset keeping it cool enough to run. however, war3 doesn't require lots of graphics so my vid card fan stays relatively slow, causing chipset to overheat. well, unless they make a heatsink/fan that mounts to that block i need to swap the board.

 

[dalearyous]

going to rig up some extra fans near the heatsink for now while i wait for new mobo

wow 2 fans pointed at the chipset and after hour of war3 my MB shows to be at 28C

 

[dalearyous]

crap at 3am this morning my computer restarted...i was in bed by 1 which means it might not be due to heat now...im so confused now.