LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Hardware (https://www.linuxquestions.org/questions/linux-hardware-18/)
-   -   What would cause a computer to shut off? (https://www.linuxquestions.org/questions/linux-hardware-18/what-would-cause-a-computer-to-shut-off-562831/)

davidguygc 06-18-2007 11:04 PM

What would cause a computer to shut off?
 
Hi, earlier today while I was taking a nap my computer all of a sudden shut off for no apparent reason. When I tried to turn it back on, it wouldn't, the lights on the mobo weren't on. I tried unplugging and plugging it back in, still no luck. I tried doing this for about 5 mins in different sockets, then all of a sudden it started working fine again. I don't think the power went out, my alarm clock didn't reset. Any ideas?

tredegar 06-19-2007 01:00 AM

Overheating. So it shut down to protect itself. When it had cooled down, you could restart it.
Clean out the case. Check your cooling fans.

j-ray 06-19-2007 01:01 AM

a defective contact in your power pack maybe?
but tredegar's explanation is more likely.

slakmagik 06-19-2007 01:03 AM

yeah, bad psu was going to be my guess. I've had both the fans/overheating situation and the psu situation.

davidguygc 06-21-2007 09:05 PM

OK, that's what I figured, thanks.

davidguygc 06-24-2007 02:15 AM

OK, well it happened again, I think. How can I tell if it overheated? It seems to happen only when I'm not paying attention to it.

jimmy512 06-24-2007 09:51 AM

Go into your BIOS and run the computer for a while. Monitor the temperatures on the cpu and motherboard. I had a similar situation where the fans were clogged with dust, and this was causing the cpu to get hot. There should be software for linux too that can monitor the temperature of your cpu (I think conky can), but I'm not sure how to use this.

slackhack 06-24-2007 09:54 AM

monitor lm_sensors while running something that taxes your cpu. if it shuts off when it gets to a certain temperature, you'll know that's the problem. otherwise, i would guess bad power supply along with the others.

IsaacKuo 06-24-2007 01:10 PM

I'd suspect a bad motherboard or bad PSU before I'd suspect overheating. But then, I work a lot with older hardware, and the motherboard is often the first thing to go bad.

slakmagik 06-24-2007 02:29 PM

Well, I agree with you IsaacKuo - when the machine that overheated overheated, it came right back up. The problems you had with restarting your system, davidguygc, makes me think PSU. If your cooling fans are operational and it's got good airflow around it, it really shouldn't be overheating.

davidguygc 06-24-2007 04:09 PM

Well I have a brand new PSU, and it finally happened while I was on the computer, I was just on imdb.com, not even listening to music or anything. Just Firefox. The PSU is Thermaltake TR2 W0070RUC ATX 430W Power Supply 115/230 V I decided to get a slightly better than average power supply. My MoBo is MSI P6N SLI-FI LGA 775 NVIDIA nForce 650i SLI ATX Intel Motherboard I have been running Windows past day or so, and it has happened on it too. So I got SpeedFan, and it has yet to set off any alarms I set. The fans are completely unblocked, are clean and about 2 mos old. The entire system is about 2 mos old. The computer's main temp has been running around 32 C, and the CPU's between 35-45. The limit is around 81. I don't have any real feel for what those temps are, I'm used to Fahrenheit.

Thanks for the input

P.S. I don't think it is OS related at all, since I cannot turn it back on. The routine to get it back going is to unplug it, plug back in. I hit the power button, it powers on for about 1/2 sec, then shuts back off, repeat until it works. I ran "backed up" a DVD and the CPU Temps ran between 40-45 C So I really don't see it hitting the 80 C ceiling

slackhack 06-24-2007 04:50 PM

well at least it doesn't seem to be overheating. those temperatures are fine. it could still be the PSU, or as IsaacKuo said the motherboard could be bad.

the next time it dies, you might try pressing in the power button for a few seconds *with the power cable unplugged* to "clear" any charge in the motherboard before plugging it in again and trying to start it. that might at least let you power back on sooner until you can find out the problem.

i wonder if kernel.log or messages or anything like that would log a problem if it were something to do with the motherboard. there might be a hint at least. or maybe your bios needs to be flashed, is another idea. could be a lot of things, sorry.

davidguygc 06-24-2007 07:06 PM

I know what ya mean, million of things, not a lot to go on with my info. First time something like this has happened, I don't know what info I need to give yall. Well, I'll look into the BIOS, and I have the PSU that came w/ the case, if it keeps happening I'll switch it out w/ that and see if it that fixes it. If so, then yay, I have it pinned down.

If yall have any suggestions keep em coming please, they all help immensely!

Thanks,
David

Crito 06-24-2007 07:37 PM

Quote:

Originally Posted by davidguygc
The limit is around 81. I don't have any real feel for what those temps are, I'm used to Fahrenheit.

You observed a peak temp of 81C? My BIOS auto-shutdown doesn't even go higher than 85C. I currently have it set to 70C on my dual core 4400+. In other words, that's too hot. So IMHO the original diagnosis was correct.

I'd remove the heatsink and replace the pad/compound with some artic silver. I'd also look into the case ventilation, especially with a power hungry nVidia card cooking your capacitors. ;)
http://www.aavidthermalloy.com/techn...apacitor.shtml

Allen614 06-24-2007 07:54 PM

It's a bad power supply as long as your heatsink is not laying in the bottom of the case. (I actually have seen that.) Overheating CPU problems manifest themselves in poor performance long before they shutdown.


All times are GMT -5. The time now is 02:21 AM.