P4 temps

[Hitman]

It seems to me my cpu is running too hot. I checked my cmos setup after running my computer overnight. The SYSTEMP was 35ºc and CPUTEMP was 44ºc.

 

When I play Call of Duty sometimes my computer will just shut down, and I was thinking that might be because my cpu is overheating. Any thoughts on that?

 

After reading this I thought my suspicion could be correct.

 

Specs:

-P4 2.8c (Northwood) not overclocked

-Stock HSF, using the craptastic thermal pad or whatever came on it

-2 80mm intake fans in the front (bottom one doesn't run at full speed ever since I had to hook it up with a 3-to-4 pin power adapter, top one blows between my HDDs)

-1 80mm outtake fan in the back (plus whatever my PSU adds)

-Cable management is decent, cables are folded and on the side, extra power cables are tucked between my DVD and CD writers.

 

Would investing in some arctic silver help me substantially? Or perhaps a new HSF? (Can anyone recommend a good one?)

[All Kill3r]

cpu temp was 44c idle or load ? i hit 50c loaded and have no stability issues, idle is around 42-44c. could be the prob but might not be, is your room hot or does it have air conditioner ?

 

AS5 would help 2-3C and a better heatsink another 3-4c with more heat removal capabilities. Download and run Prime95 and see what your temps do and if you crash, that will let ya know if its the cpu overheating. SP-97 or SLK-947U are the best P4 heatsinks, Vantec Aeroflow is good also and it uses the stock heatsink assembly.

 

Whats your motherboard ? If its an Asus board and your at 44c then yea your prob overheating. If its Abit then your prob ok. If its something diff then refer to Prime95

 

Whats your PSU and room temp.

[All Kill3r]

Run torture test with small tft's, my room is 80F right now and my cpu just hit 53C with no stability issues Hitman. System temp is 33C.

[Hitman]

Thanks for the response killer.

 

cpu temp was 44c idle or load ? i hit 50c loaded and have no stability issues, idle is around 42-44c. could be the prob but might not be, is your room hot or does it have air conditioner ?

I guess idle, I had been downloading stuff all night, but low cpu activity. Room is the same one I sleep in so it better have AC.

 

Download and run Prime95 and see what your temps do and if you crash, that will let ya know if its the cpu overheating.

OK I'm running it now, does it stop eventually or should I turn it off after a certain period of time?

 

Whats your motherboard ?

Oh yeah, forgot to mention that. It's an AOpen AX4SPE-UN.

 

Whats your PSU and room temp.

 

Room temp should be 78ºf and I have a couple fans to keep it a bit cooler. How would I go about getting my PSU temp? I have a 500 watt Powmax. Also, how can I get my cpu/system temps from outside the cmos setup. I tried MoBo Monitor 5 (recommended by Dweezil in another thread) but that reported rediculous numbers. I have SiSoft Sandra installed but I can only find "Board Temperature" in the "Mainboard Information" section, not cpu temp.

 

Thanks for the help!

[All Kill3r]

heh, wasnt asking for your PSU temp, just what psu as that may have been the prob but you have plenty of powa.

 

I would let prime run for an hour atleast to see if it crashes, if not then temps prob arent what was causing the crashes.

 

I monitor temps with a program that came with my motherboard. MBM5 would be my recommendation also, dont know why it doesnt work for you i think ya should try again

 

if it doesnt work or is wacky, uninstall MBM5, delete it's program directory from program files, then reinstall, make sure your logged on as admin ?

 

-theres also a program called "speedfan" to monitor temps but ive not tried it, just heard of it...

Edited June 10, 2004 by All Kill3r

[Hitman]

heh, wasnt asking for your PSU temp, just what psu as that may have been the prob but you have plenty of powa.

LOL oops.

 

I monitor temps with a program that came with my motherboard. MBM5 would be my recommendation also, dont know why it doesnt work for you  i think ya should try again 

 

if it doesnt work or is wacky, uninstall MBM5, delete it's program directory from program files, then reinstall, make sure your logged on as admin ?

Alright I'll try it again. Thanks again.

[All Kill3r]

np mang hope ya get everything sorted.

[[Mmmm]Homer]

The two fans "in" aren't a problem, but the heat getting out might need more help. The one 80mm fan "out" and the PSU fan are doing what they can, but do you have room for another one blowing "out" on the back. I'd experiment with one of the ones on the front blowing "in" by swapping it to the back blowing "out" if you do have the room.

 

Make sure you cables and wires are tucked up clean. Make sure your fans are clean, blow them off with a can of air if necessary.

 

The Arctic silver might help your CPU a little as Killer suggested, but I'm more inclined to think that your systme temperature is the culprit. If it's going much higher under load you might get some shutdowns - chipset or RAM could cause this. If your chipset has a crappy fan or heatsink it might need to be replaced. Sometimes mobo manufs don't put any thermal paste betweeen the HS/HSF on a chipset and adding this could help, but will require pulling the mobo.

[Hitman]

Rgr that Homer. I wish I had another spot on the back for more outtake but alas, I don't. However the back runs full speed and the front two don't so I think it's ok.

 

I ran killer's stress tests for a good 2 1/2 hours with absolutely no problems so I'm pretty convinced my temps are good. Just have to figure out what is wrong with CoD.

[[Mmmm]Homer]

It could be your video card overheating then, if a mem/cpu test didn't crash.

[All Kill3r]

could still be the mem, the torture test i had him run was cpu intensive only, wouldnt make sense to have him run something that stressed more then the component that was in question. That would put us back where we started if it crashed.. not knowing if it was the cpu or the mem. There is a blend test on Prime that stresses both but i find Memtest better for RAM stability testing... its probably not a agp related prob though... (a guess) Few hours of 3DMark01 looping nonstop for agp stability testing... that way you know where your problem is.

 

Also, your system temps arent northbridge temps, the temp sensor on intel boards is somewhere near the dimm's. Its pretty useless to use that temp for anything, the mosfets around the chip are probably scorching hot (burn you if you touch them kinda hot) but they are designed for the heat and are ok if their blazin.

 

What im saying, when you stress test, test one thing at a time, so if it crashes you know what crashed. Since theres no overclocking in this case and prime didnt fail (as i didnt think it would) i dont think your hardware was the culprit.