Ultra 120 Extreme Temperature Problems - Overclock.net

STN71190 · 06-29-07

Well, I just got this cooler today, installed, then removed it and put in a washer between the mounting bracket due to its loose mounting pressure, however after all this I am loading at 50°C in a 22°C ambient @3.3Ghz with 1.2875 vCore, help me.

afzsom · 06-29-07

Try checking the heatsink's base with a razor edge and a flashlight, that should give you some indication of un-even-ness and whether or not you should lap the base. Other thing to check is the amount and method of your TIM application. The weight of the Ultra 120 eXtreme may also be leaning it away from the mobo and "off" the chip slightly, so it may need a brace/mounting mod to keep it firmly planted.

**grunion** · 06-29-07

Quote:

Originally Posted by STN71190

Well, I just got this cooler today, installed, then removed it and put in a washer between the mounting bracket due to its loose mounting pressure, however after all this I am loading at 50°C in a 22°C ambient @3.3Ghz with 1.2875 vCore, help me.

Hmm..Did it come with the rubber washer already mounted?

STN71190 · 06-29-07

I checked the base with a razor and it is in fact very convex, but I've heard they were designed this way, but after taking the cooler off for a third time I noticed the TIM is not being spread correctly compared to my other lapped heatsinks I used to use, guess I have to lap this one too. Also mine didn't come with a washer so I just put my own in.

**grunion** · 06-29-07

Quote:

Originally Posted by STN71190

I checked the base with a razor and it is in fact very convex, but I've heard they were designed this way, but after taking the cooler off for a third time I noticed the TIM is not being spread correctly compared to my other lapped heatsinks I used to use, guess I have to lap this one too. Also mine didn't come with a washer so I just put my own in.

Is your CPU IHS concave or convex?

STN71190 · 06-29-07

Neither, its actually pretty flat and mates up well with my lapped heatsinks.

**grunion** · 06-29-07

Quote:

Originally Posted by STN71190

Neither, its actually pretty flat and mates up well with my lapped heatsinks.

IMO don't lap the Extreme, I didn't and have excellent temps.
What fan are you using and how did you apply the TIM?

txtmstrjoe · 06-29-07

Quote:

Originally Posted by STN71190

I checked the base with a razor and it is in fact very convex, but I've heard they were designed this way, but after taking the cooler off for a third time I noticed the TIM is not being spread correctly compared to my other lapped heatsinks I used to use, guess I have to lap this one too. Also mine didn't come with a washer so I just put my own in.

Last night I did some quick back-to-back testing between a lapped Ultra 120 Extreme and a fresh, out-of-the-box one on my naked Opteron 170. I had the notion that, on a naked die, lapping a cooler's base wouldn't be much of an advantage. But to be absolutely sure, I just had to test.

(The new one is meant for a client, whose permission I asked for some quick tests.)

Just to be sure that TIM application was not a factor, I did two separate installations for each Ultra 120 Extreme (lapped and unlapped).

There's no doubt about it: With my system, lapping the Ultra 120 Extreme results in a reduction of around 6 degrees C at maximum load temperatures. Bear in mind that I used the "2 cent mod" technique during my testing as well. I did all I could to ensure that conditions between the two HSFs under test were as close to identical as possible.

Conclusion: Out of the box, Thermalright's Ultra 120 Extreme is nowhere close to optimized. Lapping the base will gain at least a few degrees C headroom and improve thermal dissipation.

**grunion** · 06-29-07

Quote:

Originally Posted by txtmstrjoe

Last night I did some quick back-to-back testing between a lapped Ultra 120 Extreme and a fresh, out-of-the-box one on my naked Opteron 170. I had the notion that, on a naked die, lapping a cooler's base wouldn't be much of an advantage. But to be absolutely sure, I just had to test.

(The new one is meant for a client, whose permission I asked for some quick tests.)

Just to be sure that TIM application was not a factor, I did two separate installations for each Ultra 120 Extreme (lapped and unlapped).

There's no doubt about it: With my system, lapping the Ultra 120 Extreme results in a reduction of around 6 degrees C at maximum load temperatures. Bear in mind that I used the "2 cent mod" technique during my testing as well. I did all I could to ensure that conditions between the two HSFs under test were as close to identical as possible.

Conclusion: Out of the box, Thermalright's Ultra 120 Extreme is nowhere close to optimized. Lapping the base will gain at least a few degrees C headroom and improve thermal dissipation.

I wonder how it would apply to a non naked C2D