Phenom II 940 temp question's

[andrewvgils]

First off, thanks to everyone who helps!

Ok, with stock air cooling on my Phenom II, I am getting 58 C. When I run benchmarks with AMD Overdrive, I was at 94 C (same temps with coretemp) before I stopped the benchmarks! I am starting to believe there has to be something wrong with the temp readings. The thing is, I have my MB (Gigabyte UD4H) set to turn off or show warning screens when I get above 70 C. I've read tons of threads, even some on this forum, but these are by far the highest temps I've seen posted. I have some AS5 that I can apply, but would it really make that much of a difference. At last resort, which I know most of you are going to recommend, I can buy a new HSF, preferably one where I would not have to remove the MB, but I just feel these temp readings are insane.

Thanks for all the help!

 

[pat]

And if you carefully touch the heatsink with your finger, does it looks like hot? normal human body temp is 37C, water boil at 100C.

 

[AdioKIP]

With stock cooling my Phenom 2 940 was idling around 54. Stress tests pushed it into the 60's before I would shut them down. There is no way your processor should be hitting 94 c, failsafes should have cut the computer off way before that. As far as cooling your processor, I would get an aftermarket cooler. Xigmatek HD ST1283. Brought my temps down to 35 C idle, maxing around 45 C. Gaming and average use is around 38 C.

 

[smithereen]

That's not a sure-fire solution. If the thermal compound is bad, the heatsink won't get hot. I've seen this effect in person - my cousin ran a e8400, and when he bumped the heatsink, sure enough, 100C temp readings. After reapplying thermal compound - right back to the low 40s

 

[andrewvgils]

Sorry I haven't answered guys, been out for awhile. I'm going to apply AS5 and see if that does the trick. As i'm typing this, my temps are 58-61, which is pretty ridiculous. I forgot to mention I have pretty good airflow in the case, so it really doesn't make sense. Any tips for applying AS5?

 

[keithlm]

Sorry I haven't answered guys, been out for awhile. I'm going to apply AS5 and see if that does the trick. As i'm typing this, my temps are 58-61, which is pretty ridiculous. I forgot to mention I have pretty good airflow in the case, so it really doesn't make sense. Any tips for applying AS5?

Use less than you think you should. In your head think: "very thin layer". You only need a dab about the size of a BB in the center of the CPU.

Using too much won't help and can actually hurt.

 

[andrewvgils]

I put about half the size of a pea of AS5 on my CPU. As I'm typing, my temps are at about 50 C. I'm guessing either the AS5 needs to set in, or something is up with whatever is used to measure my CPU temps. My temps should only get lower from here, not higher right? Also, I ran stress tests for about 10 min and at 100% CPU usage, my temps were at 59-61, which from what I read is right below the AMD 62 C safe zone.

 

[smithereen]

That's a little warm, but the compound does need to set. Run Prime95, and if your temps go over low sixties, re-apply.

 

[andrewvgils]

Just an update guys...

I believe I read somewhere on this forum that an owner of a PH II 940 was conducting overclocking tests. My stock voltages where at 1.35 v, but the poster was overclocking to 3.2 on 1.30 v. I decided to lower my voltages to 1.30, which brought my temps down to 46-50 idle 57-58 during stress tests (for about 30 min). I even did another 45 min stress test at 3.2 GHz and was performing with no problems at the lower voltage. I ran these tests using AMD overdrive, which is showing the same temps as Coretemp. Thanks for all your help!

 

[keefygeorge]

Just a note.
Sometimes if sensors are not connected they can give high temp readings.
Check that all your system fans and cpu fan are plugged into the right power sockets.

 

[keefygeorge]

Also your CPU heatspreader may have a large concave on the heatspreader.
If this is the case AS5 wont really help. Return it under warranty or consider lapping your cpu heatspreader.