Alright gentlemen, I need some help here.

P

Pyrex238

Limp Gawd
Joined
Jul 12, 2006
Messages
274
Well, I've started having some issues with my rig lately. It all started when I switched to the latest 175 nvidia drivers. During games of company of heroes, my computer will just click off.

My system is...

E6600 @ 3.0GHz +.0875v over stock
Antec True Power 850W (~6 months old)
DFI Infinity P965 Dark motherboard
(4x1gb) 4GB Crucial Ballistix PC2-1000
Watercooled system (CPU, NB, SB, GPU)
eVGA 8800GTS

I've never had this issue before where the system just clicks off. I run very low voltages on most all my components because the system seemed to get unstable with higher voltages. My ram was running at 2.0v (2.3v stock) but I run my system on a low ram strap because I'm running 4 sticks, it gets unstable with the higher straps. my northbridge was running at stock voltage (1.25v) as is all my other components except the cpu which is +.0875v over stock. I went ahead and clicked everything up a notch and it didn't reset for a few hours of gameplay, however the next day it clicked off on me again. I've since upped voltages further but haven't really tested it.

From my experience, an instant shutdown is usually a power supply issue, however my system is most likely not utilizing even 50% of my PSU's total output (my waterpump is separate AC), so cap drain shouldn't really be an issue for me since the PSU is still fairly new.

The odd thing is all this started with the 175 drivers. I rolled back to the 174, still didn't fix it, rolled back to the older 169 (?) drivers, and that is where it's seemed to be most stable, but I seem to think this is just a coincidence.. I've run out of solutions and I need some of you experienced guys to chime in. Also, I might add that I also considered a heat issue, however the moment it resets itself I go right into my bios to check temps to see if something overheated, and the hottest component is my northbridge @ 45c... so I don't think that heat is the problem here.

My video card is overclocked normally @ 700MHz core / 1800MHz memory. It will run 720/915, but I keep it a little conservative. I never see artifacts... I've run the card with the clock at stock speeds and it seems to have no effect on the reset issue.
Please, let me know what you guys think, I'd really appreciate it.
 
It all started when I switched to the latest 175 nvidia drivers.
Click to expand...

Patient: Doctor ! Doctor ! It HURTs when I do this !
Doctor: Well, stop doing that.

revert to the previous drivers until they find the bug and investigate increasing the voltages with a rational testing method keeping notes etc. Offhand I would give the cpu and memory minor increases until stability returns. Run your video card at stock settings, OCs determined by tools like ATI tool will not mimic gameplay and make you think your card can do OCs that will crash immedately in games.
(be sure to fully uninstall and clean out the new ones before re-installing the old ones)
 
Hey Bill, I also said I went all teh way back tot he 169's with the same results though.
 
Yeh you got me there, I just edited to fix my reply. I agree its a tricky thing, several things are probally all working together to hurt stability. I doubt just one thing will fix it. Apologize for missing that, got in too big of a hurry. It is going to be a bit of work to find out exactly what is going on, 3dmark loops while adjusting system settings would probally be the best testing method and its going to kill some weekends tracking it down.
 
BillParrish said:
Yeh you got me there, I just edited to fix my reply. I agree its a tricky thing, several things are probally all working together to hurt stability. I doubt just one thing will fix it. Apologize for missing that, got in too big of a hurry. It is going to be a bit of work to find out exactly what is going on, 3dmark loops while adjusting system settings would probally be the best testing method and its going to kill some weekends tracking it down.
Click to expand...

Yeah I was running 3dmark Vantage for a few hours straight and it never reset... It's only COH it seems... but it might be that RTS's really flog the CPU and 3dmark doesn't put that much of a strain on it, I could try orthose too I guess, but it's just a damn shame it happens like 3 hours into a game and just clicks off..
 
Been thinking some more about it, the "just clicks off" is the clue but I dont like where that takes me. It implies the power supply. What I mean is that if the computer froze or crashed from heat or software etc. I could see that being a board or settings problem and while the computer would crash etc, power would still be on. But to just "click" and completely shut off, all power gone, I would think that would have to be something in the power supply or (not up on that power supply) something on the board signaling the power supply to shut off.

I dont see your components pulling enough current to cause an 800W supply to trip from overcurrent, I run roughly the same stuff on a 550W. It could be that one rail is overloaded, maybe, but that too would susprise me as the cabling should be arranged so that it is very difficult to attach too much on a rail but it might be something to check on if your manual or the Antec web site details what connectors are hooked up to which internal rail.

Any sensors or automatic shutdown settings in bios (again just throwing ideas out there, not up on that specific motherboard) should be looked at carefully and maybe the setting increased. Even if set at high values the sensor might be off and signaling a shutdown when everything is really ok. Kinda back to that "it is odd it just completely clicks off thing".

Wonder if any watercooling components, pump?, could be causing it as a last resort? I have my pumps rpm sensor read by bios with an alarm if the rpm falls too low for detecting pump failure. This seems unlikely but again a sensor failure could occur after things get warmed up either in the mb or the pump.


Hope this gives you some ideas.
 
it could also be a short in one of the wires. i just had this problem with a heatsink i bought (maybe i mounted incorrectly), but when i had that heatsink in, and I set OC settings in the BIOS, the system would POST Beep, and then shut off. i'd have to unplug it, and then turn it on again, and it'd turn on for 1 second, and then shut off.
also tried different power supplies, same issue

swapped the heatsink back to the orginal, and it hasn't done this since.

so i'm assuming the heatsink shorted something out... i don't know what though.

back to your situation, something seems to be shorting the computer out causing it to shut off :/ i would try chaning the PSU out for another one, maybe disconnect any fans and run it with your case open.
 
You must log in or register to reply here.
Back
Top