11-16-2005, 01:12 PM | #1 (permalink) |
Insane
|
Power Problems
Short Description: When the computer is under load, it will shut down, seemingly because of the CPU trying to pull more than needed. What can I do to fix this?
Long Description: I have a 3.2Ghz Prescott overclocked to 3.6Ghz with an XP-120 and Panaflo High speed fan on it. Ambient case temp is about 28-30 deg C (which, given that in South Carolina and the highs are still in the 70s, isn't too bad). Idle, the CPU is about 44-45 deg C. Under load, it gets to about 60 deg C. Computer parts: Asus P4P800E-Deluxe (I believe the CPU volt drooping is the major problem here, as detailed below) 3.2E P4 6800GT 7 case fans 420W Thermaltake Power Supply (about 2 years old) Voltages (according to Asus Probe): +12 = 11.9 (idle) to 11.7 (load) +5 = 5.1 (idle/load) +3.3 = 3.39 (idle/load) VCore = 1.5V idle, drops to 1.40-1.44 under load About the CPU Vcore drooping. Basically, to do my overclocking, I have to set the VCore much higher in BIOS than it actually needs to be because when the CPU is under load, the voltage supplied by the board drops. Thus, to supply the ~1.4V needed for the CPU to function under load at 3.6Ghz, I have to set VCore to 1.5000V or 1.4875V in the BIOS to try and account for the drop. Is there anything anyone can suggest (other than lowering the clock, preferably) to combat this problem? I think the problem is that the Vcore drops too low for just a second, causing the computer to shut down. I believe this because I have no trouble with the computer if it is not left under load for a long time. For instance, I don't have trouble with games or short-term loads. |
11-16-2005, 08:23 PM | #4 (permalink) |
Knight of the Old Republic
Location: Winston-Salem, NC
|
The symptoms are a simple case of overclocking instability. Many processors will cause the system to shut down if clocked too high for stability. Mine does it, my brothers does it, and basically everyone that has tried OC'ing has done it. You have to find a stable clock speed before it starts shutting down under load. The shutting down is caused by one or many of a large number of things, ranging from the chipsets being too hot, to the PSU not applying enough voltage, and even to your CPU simply not being capable of stability at the OC'ed speed.
I really don't know what to tell you besides to clock your CPU back down. It is sort of odd that it doesn't do it in games though... This very problem is why I choose not to go with OC'ing. It's much less annoying to simply buy a better processor. *EDIT* May I ask why you're even OCing? Your CPU is more than enough for any game on the market. You won't notice a difference in any application when OCing that CPU. -Lasereth
__________________
"A Darwinian attacks his theory, seeking to find flaws. An ID believer defends his theory, seeking to conceal flaws." -Roger Ebert Last edited by Lasereth; 11-16-2005 at 08:25 PM.. |
11-16-2005, 08:31 PM | #5 (permalink) |
Insane
|
OCing for the sake of OCing, I suppose. It's more for fun (despite the hassles) than anything else. I guess for now I'll see if lower ambient temperatures (thanks to some finally colder temperatures) help maintain stability. The shutting off is not a constant thing, and it usually happens overnight when it does happen, so it's very hard to pin exactly what causes it.
Edit: I should mention that when I say "under load," I'm referring primarily to a java program I run to try and solve the RSA prime number security challenges. Actually, now that I check that...it's only 50% load. Very bad that with only 50% load, I still see that voltage droop to VCore. Last edited by wombatman; 11-16-2005 at 09:52 PM.. |
Tags |
power, problems |
|
|