High temperatures with Q6600 and TRUE120?

[er12345]

I have just installed a TRUE 120 with a scythe s-flex 1600rpm fan. Neither CPU or the HSF is lapped. Case is a P182 with all 3 fans running on low speed.

I was just wondering if my tempeatures were normal for Q6600 (G0) at stock speeds (2.4hz):

Core 1 and 2: 52-54 (both cores identical temp)
Core 3 and 4: 46-47 (both cores identical temp)

This is after 1 hr of Prime95 v25 toture test at 25C ambient temp.

Are these about right or should i reseat my heat sink (it takes ages... =[). I was originally aiming to OC to about 3.4GHz but with these temps i am not sure i can.


Thanks guys

 

[er12345]

oh yeah idle temps were 30-32-28-28 (at a 23 degree ambient)

 

[DerwenArtos12]

While not exactly out of line they are a little high. The inconsistancy between cores is telling me that the heatsink is convex and needs to be lapped, sorry.

 

[er12345]

kk, thanks. I will see what i can do.


Just for reference, what is the ballpark normal figures for Q6600 (stock speeds) and TRUE both idle and load?

The delta between cores 1/2, 3/4 may just be the way it is. With the intel stock HSF i noticed the same pattern, and as before core 2 was always the hottest one.

 

[DerwenArtos12]

Originally posted by: er12345
kk, thanks. I will see what i can do.


Just for reference, what is the ballpark normal figures for Q6600 (stock speeds) and TRUE both idle and load?

The delta between cores 1/2, 3/4 may just be the way it is. With the intel stock HSF i noticed the same pattern, and as before core 2 was always the hottest one.

It is absolutely 100% possible. Unfortunately unless someone has the same processor(not just model but one made in the same silicon batch), in the same case with the same heat sink with the same fans in the same climate, there is not a reliable temperature ballpark to give you.

 

[Thor86]

Temp variances between cores on Quads are normal. Never seen one with all temps the same or relatively (1-2 degrees). Most are 3-7 degrees in variance.

I would recommend you try turning your rear exhaust 120mm fan to medium. You probably won't notice much more noise, but definately drop 2-3 degrees more under full load.

And if you want to see your max cpu temps, try running Prime95 using LargeFFT tests for 2+ hours.

 

[aigomorla]

ER first welcome to anandtech.

Next, the TR has been sometimes shown to have a unflat base:
http://i125.photobucket.com/al...orla/Thermalright6.jpg


You need to do a razor test to see if your sink falls i this area.

The uneven coretemps lead me to think that your sink or ihs maybe a little warped and your not geting a good firm clamp on it. [its not bad tho. You have a delta of 8 from your lowest to highest. Usually 8 is the upper max, but its not drastically bad like a pushpin sink would do.]

However anything below 70C loaded is considered safe and okey. So your still in the safe mark.


Razor test:
Put the sharp side of the razor perpendicular to the sink base. Look at it though light from the side. If you see a lot of light bleed though the sharp part, its not flat, and you'll need to lap it.

You need to do a razor test on each 90 degree so you know how unflat it is.

 

[er12345]

ok.......

i tried a moderate overclock to 3.2GHZ. I have kept all voltages at stock and its been prime95 stable for the past 6 hrs so far. Temps are 66-58-58-66 for cores 1-4 respectively. Given these temps should i push ahead or call it a day?



crap BSOD in P95 on 3.4ghz. too ambitious i guess. oh well

 

[DerwenArtos12]

I hate to have to tell you but, I'm betting you're going to need to lap to go any further. If you pull your heatsink you should be able to see where you're getting the most contact in the TIM and that should give you a decent idea if you'll need it.

 

[er12345]

ur probably right there

i am stable now at 3.4ghz 1.35 vcore, but tempeatures are 72-63. a full delta of 10C!