Blakkout47

Hi, sorry if this is the wrong forum but I need help. Every now and then, my computer will randomly reboot. It turns off as if I've held the power button, and then turns back on. The first couple times it would reboot. One time it sat at my motherboard's welcome screen and just sat there, and now it doesn't boot at all. No video, now keyboard power. If it doesn't boot, I have to leave it off for about 10 minutes before I can boot it again.

I have ruled out overheating. It is pretty damn cold in here, I cleaned out the dust last night, and it rebooted with a temp checker open, all temps were below 40 C.

It seems to happen at random times, although the first few times were all when I was playing WoW. It has only happened once out of a game, and I was just sitting in windows with IRC and ventrilo open.

I've considered it to be a possible power supply issue, but I am not 100% convinced. I ran a stresstest (prime 95) and it did it fine, did not reboot, so being under load and not rebooting made me less convinced.

However I've noticed that some of my rails are very skewed. My 3.3 and my 5V rails are within range, however my + and - 12V rails are very very off. my +12V rail reads between +2.3V and +8V (8V was achieved under the stress test). My -12V rail is constantly at -16.97V. During a virus scan, my -5V went to -8.58V. I know that it is possible for speedfan to be wrong, but I cant' find my multi meter so I'll have to wait till later to check them myself.

Also, I've noticed my PSU shows that I have 4 +12V rails and what they're designated to, however speedfan and my bios only indicate one 12V rail. The plugs for my PSU have the same indicators that are on the label, such as pci-e 1 and pci-e 2. So I'm curious if those correlate to the rails or if that is just for ease of installation. Also, my graphics card has a 36 amperage requirement, and my 12V rails are rated at 18A, but since it's getting 2 it's ok? but one is possibly shared so it's bad? but it's worked for a year?
Image of this label can be found on newegg here: http://images17.newegg.com/is/image/newegg/17-341-002-S06?$S640W$

Specs:
GIGABYTE GA-P35-DS3L LGA 775 Intel P35 Mobo
Intel Core 2 Duo E6850 Conroe 3.0GHz LGA 775 65W Dual-Core BX80557E6850
OCZ GameXStream OCZ700GXSSLI 700W ATX12V SLI Certified CrossFire Ready Active PFC PSU
G.SKILL 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 1066 (PC2 8500) RAM
Western Digital Caviar Blue WD6400AAKS 640GB 7200 RPM SATA x2
SONY DRU840A and DRU710A DVD burners
EVGA nvidia GTX260

any help is greatly appreciated

dai

where are you getting those readings from we prefer what the bios lists

for accurate readings you need to check with a multi meter

http://www.driverheaven.net/guides/testingPSU/

Blakkout47

I'm getting them from speedfan, and once it becomes a better time of day I'll be able to borrow a multimeter, will post results when I do.

Tyree

Speedfan can give flaky readings. Check the Voltages in the BIOS.

Blakkout47

so my results with my multimeter, all my 12V rails were at about 10-10.5V, which is under the 12V rail tolerance rating. Unfortunately I had a multi meter with a needle, so it's not the most accurate, but I never broke 11V on any of them. I checked them while running prime95 so that it would be under full load. While doing this stress test, it did not reboot, nor did it on the last stress test with prime95.

all my other rails were within tolerance.

Tumbleweed36

If that +12 is accurate, you have found your problem most likely. YOu need a decent power supply in there.

Blakkout47

even though it's working fine? even under full load there were no issues

Tumbleweed36

If in fact the bios states that it is less than the 5% +/- deviation from the +12 standard and that seems to be a fact, then you run the chance of harming other components on your computer in the long run. When you look at the cost of a decent power supply vs. replacing burned up ram or a motherboard, that is something to consider.

If it is randomly turning off as you reported, that is another indicator that something serious is wrong with the rig. I have no way of knowing (since I am not there to see it) if that is the total issue, but I would not run a rig with that low of reading on the +12 rail unless I was planning on buying a new power supply, plus the other things it takes with it from damage that is done from lack of power. That is my opinion.

Blakkout47

the bios had them within range, well, it only lists one, but my voltages were

Core: 1.332V
DDR: 1.984V - specs from manufac. site say 2-2.1V necessary
3.3V - 3.344V
12V - 12.556V

But my multimeter tests gave different results. It's possible my multimeter was wrong? it is fairly old.

dai

check the meter on your car battery and see what it gives you for voltage

Blakkout47

i tested it on a new AA battery and it read 1.5V

so my computer has been working all day, although it's just been web browsing, watching videos, doing visual basic homework. once i was done with my homework, i was like, gonna play some wow i guess, and as soon as the login screen loads my computer restarts

once i get it back on i'm gonna try playing cod4 and see what happens

cod 4 also died, at the choose a server page.

i've noticed that after it dies and it fails to reboot, that there's less noise coming from the case, and the CPU fan isn't spinning, but all today while it was running fine the fan was spinning. I know that it has autodetection it usually doesn't spin on boot until it gets warm, so that may be why, and it won't ever spin on a failed boot cuz it doesn't even get to the bios which is controlling the fan.

so i was looking into things, and i realized that my case (Antec 900) has no internal speaker. So that's definitely proabbly why i have no clue what's going on when ijust have a black screen. gonna try to rig the speaker from another case into this one and see what happens.

rigged it up, then magically it booted up fine. side panel was off so wasn't sure if i had totally debunked overheating. it ran fine for a while, i decided to shut it off myself to undo my speaker rig and now it won't boot back up. i'll let it sit for 30 mins again, that usually seems to get it to boot.

sorry when i say won't boot, case fans start up, no keyboard power, no video, no post, no cpu fan.. waited a while, it boots, too afraid to play any games tho lol, seems to be the most common cause

also ordered a new PSU. the sagging voltages cause me to worry, even if not the cause now, could be later.

ordered http://www.newegg.com/Product/Produc...82E16817139009

dai

good choice

see how you go with the new psu when it arrives,it sounds like that is what is the problem

the batteries in the meter may be down or it's time for a new one

Tumbleweed36

Please be sure to let us know if the new supply (great choice) solves the issue. OCZ supplies of the past have had some rail stability problems.

Blakkout47

seems to be working, managed a whole day of gaming and no crashes. speedfan is still not reading my rails right, so i give up on that, it'll be good for temperatures at least.

dai

glad you have it sorted