Thermalright 120 Extreme Idle Temps High??

C

Concept1

n00b
Joined
Oct 5, 2007
Messages
35
I got a E6850 and the above HSF, but for some reason my temps seem a bit high at idle?
They usually sit at 35-38... I looked around a bit and found a few places (Anandtech, THW) usually have a idle on a E6800 of 28ish... so I thought with my E6850 it would be reasonable to say it would sit at 29-30ish. When I applied the thermal paste I put a small layer on both the HSF and the heat spreader on the chip. I am assuming that is too much??

I also installed the HSF so the fan is blowing N/S, ive heard that some people get better results if it is blowing E/W?

Any thoughts?
 
You aren't supposed to apply thermal paste on both the CPU and the HS, just a small layer on the CPU will suffice.

What's the room temperature?
 
Off the top of my head I really could not tell you. But it is in the basement and its definately not that warm down there.

Well that is probably why I am running hot then, too much thermal paste. Man its going to be a pain in the ass to take that off.
 
Do you have your fans running on high? medium? low? I find better idle temps with C1E and speedstep on. The power is there only when I need it.

CPU temps (not core) idle at 25c (according to speedfan) and with room temp probably in the low 70s (F). If the reviewers could indicate what their BIOS settings are, especially C1E and speedstep, it might help unravel the mystery of low review temps.
 
I turned off C1E and Speedstep and the temps did go up a bit, but that is kind of expected as I went from 2000mhz at idle to 3000mhz. It was still in the 30's even with both of those settings on though.

:S
 
Concept1 said:
I turned off C1E and Speedstep and the temps did go up a bit, but that is kind of expected as I went from 2000mhz at idle to 3000mhz. It was still in the 30's even with both of those settings on though.

:S
Click to expand...

I assume you are referring to your cpu temps?

I orient N/S and tried E/W; did not make a difference. Tried laying the case down horizontally; hardly a difference. Using an Antec 900 case btw.
 
MooMooCow said:
You aren't supposed to apply thermal paste on both the CPU and the HS, just a small layer on the CPU will suffice.

What's the room temperature?
Click to expand...

don't say stupid shit, i applied it on both the cpu IHS and heatsink bottom surface. just gotta know how to work your paste, kid.

abiteq.jpg
 
How much should I be putting on then? I thought I didnt put *that* much on both of them, but it looks like I did. I was just using the stuff that came with the 120 Extreme... maybe I need some AS5 or whatever it is called?
 
Concept1 said:
I got a E6850 and the above HSF, but for some reason my temps seem a bit high at idle?
They usually sit at 35-38... I looked around a bit and found a few places (Anandtech, THW) usually have a idle on a E6800 of 28ish... so I thought with my E6850 it would be reasonable to say it would sit at 29-30ish. When I applied the thermal paste I put a small layer on both the HSF and the heat spreader on the chip. I am assuming that is too much??

I also installed the HSF so the fan is blowing N/S, ive heard that some people get better results if it is blowing E/W?

Any thoughts?
Click to expand...

Your temps are what I am getting with my qx6700 overclocked to 3200. It sounds like it could be paste or contact. I lowered my temps by putting a washer between the cooler and the mount to tighten things up. A penny will work also. The base of your sink might need to be lapped also.
 
Concept1 said:
How much should I be putting on then? I thought I didnt put *that* much on both of them, but it looks like I did. I was just using the stuff that came with the 120 Extreme... maybe I need some AS5 or whatever it is called?
Click to expand...

Just a thin line on the cpu will work. I used shin etsu x-23 on mine but as-5 willwork also.
 
Get yourself a ziplock bag, stick your index finger in it, now with your other hand, add on about 2 1/2 rice size dots of thermal paste onto the surface of the IHS, dot the surface evenly, now with your finger still in the ziplock plastic bag, start spreading the paste on the IHS to an even layer, making sure it has covered the IHS, if you think you put on too much, you probably did, now on the heatsink bottom surface, put a rice size dot with thermalpaste in the dead center where you think it makes contact with your cpu, then spread it, making sure it's scarcely.
 
RangerSVT said:
Get yourself a ziplock bag, stick your index finger in it, now with your other hand, add on about 2 1/2 rice size dots of thermal paste onto the surface of the IHS, dot the surface evenly, now with your finger still in the ziplock plastic bag, start spreading the paste on the IHS to an even layer, making sure it has covered the IHS, if you think you put on too much, you probably did, now on the heatsink bottom surface, put a rice size dot with thermalpaste in the dead center where you think it makes contact with your cpu, then spread it, making sure it's scarcely.
Click to expand...

Hmm, that's interesting. I only spread the paste on the heatsink rather than on both. How much of a difference will i see if I spread paste on both?
 
ScourggeFX said:
Hmm, that's interesting. I only spread the paste on the heatsink rather than on both. How much of a difference will i see if I spread paste on both?
Click to expand...

In my situations, it yields same results, maybe 1c different if i do it with alittle paste spread scarecely on the bottom of the heatsink surface
 
MooMooCow said:
You have a E6550 running at 3.4 GHZ with air cooling at 17 celsius?
Click to expand...

Now I'm being questioned, dang if it ain't in counter-strike source getting banned for "hacking", now it's being called out on hardforums...people don't believe anything nowadays huh. guess the streets influenced the internet fairly quick. haha.

well, buddy what do you want me to provide? proof? i just took a dang snapshot now you want more?
 
MooMooCow said:
You have a E6550 running at 3.4 GHZ with air cooling at 17 celsius?
Click to expand...

It's probably that Tjunction problem with some intel chips where some chips have Tj of 100 and some have a Tj of 85 so some chips read 15C low, it's probably really 32C which would jive more with his other temperatures he has listed. Unless his room temperature is around 50F/12C.
 
-Dragon- said:
It's probably that Tjunction problem with some intel chips where some chips have Tj of 100 and some have a Tj of 85 so some chips read 15C low, it's probably really 32C which would jive more with his other temperatures he has listed. Unless his room temperature is around 50F/12C.
Click to expand...

That's what I figured.
 
Ok so I re-applied the thermal paste to the IHS and the HSF
These are the readings I am getting from Speedfan:
Temp 1: 34c
Temp 2: 21C
Core 0 : 19C
Core 1: 19c
On Core Temp I get:
Core #0: 34c
Core #1: 34C
And the Tjunction is: 100c

Does that still seem a bit off or?
 
MooMooCow said:
That's what I figured.
Click to expand...

The 15 degree difference is only for quads in speedfan. Duos are reported correctly. Since Ranger and the op have duos I don't see how that could be it. Sorry kids I know you thought you were on to something there.
 
Well I'll be damned, didn't think anyone could obtain a 17c temp with such an overclock while living in CA.

Ranger, you have my apologies.
 
The ultra 120 extreme is an air cooler like no other. If you lap it properly and use good quality paste like as5 or shin etsu it can be almost as good as watercooling.
 
First I lap both cpu and heatsink.

The method I use is to first dot the heatsink and cpu and wipe it down with a paper towel so you can see the metal "change color" indicating there is minute thermal paste filling in the cracks.

I then apply a little larger than a grain of rice on the cpu and place it flat and mount the heatsink.

I had cases on some of my older boards (pre 775) that I would rip the cpu out of the zif socket with the heatsink. That's how strong a thin even layer with no trapped air of thermal grease can get.

Even with socket 775 I have to slide the heatsink off, I've had it strong enough to start to move the pc first rather than come off. Luckily I didn't damage the socket in the process.

Temp wise, since I have a quad, I'm hovering at 37C idle. This is being clocked at 3.6. At load I peak around 62C if my room ambient is around 28C, not exactly a comfortable room at that point. If I can keep the room ambient at 22-24C I can see peak load at 51C roughly. All closed case P160, looks great but sucks in airflow compared to today's cases.
 
I just bought one - the base wasn't flat - visibly convex. Obviously that will have a major effect on how well it cools. This is a common problem so I'd check yours - if it's not flat then lap it flat or send it back.
 
Do you all just use Core Temp to get your idle/load temps or what?
I wanna compare apples to apples here, I didnt get a response to figure out if my temps were accurate for Coretemp/Speedfan.
 
Concept1 said:
Do you all just use Core Temp to get your idle/load temps or what?
I wanna compare apples to apples here, I didnt get a response to figure out if my temps were accurate for Coretemp/Speedfan.
Click to expand...

I use the BIOS and Abit EQ for my readings. Don't trust any other software out there.
 
Concept1 said:
Do you all just use Core Temp to get your idle/load temps or what?
I wanna compare apples to apples here, I didnt get a response to figure out if my temps were accurate for Coretemp/Speedfan.
Click to expand...

I use Speedfan. It has both cpu and core temps.
 
Concept1 said:
Do you all just use Core Temp to get your idle/load temps or what?
I wanna compare apples to apples here, I didnt get a response to figure out if my temps were accurate for Coretemp/Speedfan.
Click to expand...

I use everest ultimate for mine.
 
I was just thinking... maybe I put the fan on wrong? I also bought some AS5 tonight.
I got the Scythe S-flex 120mm Fan. Does the Logo on the center of the fan face the HS or does it face me? As it stands I have it facing the HS.

I think when I take off the HS tomorrow I am gonna lap the HS, not the CPU and use the AS5, and potentially remount the fan if its facing the wrong way.

Thanks for all the responses guys, I really appreciate it.
 
If you want to improve temps, put one fan blowing into the heatsink, and have another sucking that air out, and into your exhaust case fan. i plan on doing this later this weekend
 
Ill cross that bridge when I think its necessary on my behalf, but I just need to make sure I got the fan on right in the first place. I guess the dummy proof way to figure out is use my hand and seeing where the air feels like it is coming from.

Sorry I just got some wisdom teeth + cysts removed today, I am a little slow and groggy. Tomorrow will be a good day to get stuff figured out!
 
There should be arrows on the side of the fan somewhere indicating airflow direction and blade direction, generally air flows from the spinning side to the static hub side, over and around the dome covering the motor
 
Just took the xp-120 off again and tried to see how it was sitting, seems it is a bit convex. I can see under every side and only the middle is touching.

So I am going to assume that I will have to lap it a bit to make it sit properly. Does anyone know of a good how-to for this? Ive never tried this before. And I will NOT be lapping my CPU.

Thanks.
 
actompkins said:
The 15 degree difference is only for quads in speedfan. Duos are reported correctly. Since Ranger and the op have duos I don't see how that could be it. Sorry kids I know you thought you were on to something there.
Click to expand...

I don't believe this to be true. So you're telling me, my e6850 runs at 14C idle with an ambient of 72F? I find it much easier to believe that adding 15C onto that, making it 29C is the accurate temp.

This is using speedfan.
 
Last question(I hope): On speedfan 2 of my fans are running at 1400rpm or so but my processor fan is only at 500rpm, but is good up to 1600 rpm, its the Scythe S-flex. How can I max that to see how it is?

My mobo is a Gigabyte DS3R rev 1.0

Thanks

Edit:
I am using Speedfan and Core Temp to monitor my temps... I loaded Prime95 and ran the Small FFT's.
Now I am getting 2 different core readings from both of them and I really dont know which one to believe...
SPEEDFAN: 37c average load on the cores
CORE TEMP: 52c average load on the cores.

Which one do I believe as correct? I am losing my mind trying to figure out which one is accurate!
 
Ok nevermind, I figured out the Fan Issue, got it running full tilt now.


Still have core temp reading: 34c average
Still have Speedfan reading :19c average

Is that normal for most people w/ a E6850 out there?

Im done bugging!
 
actompkins said:
The ultra 120 extreme is an air cooler like no other. If you lap it properly and use good quality paste like as5 or shin etsu it can be almost as good as watercooling.
Click to expand...

what about Coollaboratory Liquid Pro? and if anyone has heard about the Coollaboratory Liquid Ultra? does it actually work all that better compared to as5 as claimed?
 
Maverick986 said:
what about Coollaboratory Liquid Pro? and if anyone has heard about the Coollaboratory Liquid Ultra? does it actually work all that better compared to as5 as claimed?
Click to expand...

I have not heard of that. I used the shin-etsu x23 on my 120 ultra extreme.
 
You must log in or register to reply here.
Back
Top