PC randomly shuts down
The last week or so, I 've been having a problem where my PC will suddenly shut off very abrubtly, like flicking a circuit breaker... It does not try to reboot automatically, but always reboots without an issue following this random shutdown.
I don't get any message from winxp asking if I want to restore from last known config, etc... it will just do a normal boot once I hit the power switch after failure....
My fist inclination is a possible PSU failure... Running an OCZ psu... the rails appear to be ok when I peek at the voltages within the bios screen.....
Any other ideas as to what can be causing this ? Power Switch, Virus, short on board, etc.????
Hail to the Victors
First thing I'd do is run a memory and hard drive diagnostic.
Next I would suspect a PSU.
Speedfan will give you temps and voltages when there is a load. You should check for bad caps also on all of your components. If you suspect the case switch, then disconnect it from the mainboard and use a ballpoint pen to power on, its only a momentary switch anyways.
Overheating and failing psu's are the two mail culprits.
Thanks for all the good input...
1) I did suspect heat.. but my temps appeared to be fine, ... .in any case i did clean up the heatsink fins and reapplied some thermal past, etc on both the video and processor heat sinks....
2) I wouldn't think a bad hard drive would cause the machine to just to a hard stop/switch off ? good suggestion on the PSU switch.. I'll give that a try next...
3) the voltages looked to be within range when monitoring them via the bios as well as via coretemp .....
I just swapped in another PSU from another PC, so we'll let that run for a day or so to see if that's the culprit.. My wife said the problem happens almost hourly if she is using the machine.. and it seems to do it during the evening randomly if not used at all....
Thanks again for the suggestions.. hopefully its the PSU..if not, then I'll try the suggestions on bypassing the power switch as well as the memtest
hmmm .. my wife just told me the pc shut down again, even with the swapped PSU, while using photoshop...
would a virus or spyware cause it to abruptly shut down like this? I would have thought not.. wouldn't it try to reboot back up? In this case its just shuts off/black screen with no power.. and it doesn't try to reboot.......
Last edited by mlc; 11-03-2009 at 02:14 PM.
How old is the mobo?
You need to look for buldging caps..
i did check for that.. .its a socket 939 board (A8N32-SLI DLX).. pretty good quality board by most standards... and I didn't notice any bulging caps...
Next step.. bypass the purge/power strip .. and plug directly into the wall to rule out the power strip....
And for grins:
When Running windows and it crashes you will get a blue screen and it will automatically restart, ofter it will restart too fast for you to see the error message. You could check the error log in this case but that is too easy. We are going to disable auto restart on system failure.
1. Go to Start -> Control Panel -> System (Windows+Pause works, too)
2. Go to Advanced
3. Under the Startup and Recovery section, click Settings...
4. Under System Failure un-check "Automatically restart"
Checking the power strip is a good idea.
You'll need to check temps and voltages when under load, it makes a big difference after Windows loads and especially when running PS. Checking in the bios wont cut it, use Speedfan.
FWIW, old systems can suffer bad caps, chip creep, cold solder joints, not to mention loss of thermal contact under heatsinks. You can check the north and southbridge chips for over heating and the vga cards gpu if you have a vga card, if they have heatsinks then reapply thermal compound after removing the old thermal pad.
Its not likely malware is turning the system off, but they can load the system up so badly it overheats , hence a shutdown. But, you would notice the pc running very slow among other problems if that were the case.
thanks Rocketmesh... the power strip idea didn't work...
I do agree with your assessment on both the malware and thermal load.. The ASUS PROBE II program I typically use is not very accurate, and I notice it consistently reports temps 5-6C lower than what the CoreTEMP program shows. I ran Orthos to stress both cores, along with SPYBOT, and noticed the temp under CORETEMP approached 62C. I hadn't realized that I had still had this rig setup with a slight overclock.. I'm running an OPTERON 170 overclocked from 2.00 to 2.30 using stock voltage. Supposedly 65C is the thermal limit they recommend you stay below for this chip.. so that could very well be the problem.
In any case, I pulled off the HSF and cleaned it up nicely and reapplied thermal paste.. and I've reset the bios to run w/stock settings, where appropriate, and the reran the Orthos test for an hour or so, and it now seems to be in the mid 50C range under stress, so I'm hoping that does the trick, but I should know by this evening once my wife starts up with her PhotoShop and Illustrator work, etc.....
I'll keep my fingers crossed, but thank for all your patience....
well... I thought I was in the clear... I put the original PSU back in, since that was ruled out as an issue.. thinking it was the thermal overload..
I just booted back up.. and as soon as windows was done loading .. it abruptly shutdown again.. and the temperature of the 2 cores were only in the high 30's centigrade... So I think its safe to say.. thermal overload is out....
Remaining variables in my mind are as follows:
1) Case Switch
2) Error on Hard Drive ? Could this cause an abrupt shutdown?
3) Memory? would this cause an abrupt shutdown ?
I will run memtest now and a checkdisk to rule out those two..
Additional thoughts/ideas are welcome!
Stark Raving MOD
chkdsk is not enough. Run the hd manufacturer's diags like Seatools.
Its sounding more and more like a short or bad solder connection on the mobo. Look again for a bad cap. They can bulge slightly on top or on the bottom.
Bad memory can give you hard shutdowns.