best compound for subzero?

[Ewok]

tried AS3, tried akasa shinetsu TIMs. the akasa is complete crap being about 7c worse than the AS3 and i cant see it turning that around after a couple of days to set.

[water_cooler 20]

I think i hread someone say as1 is good for subzero

[bowman1964]

i have used either as3 or antiseize compound.
but it would be nice if i found something better.

[Ewok]

i have a couple little tubes of alumina that came with the mach2 so ill whip this shinetsu TIM crap off in a while and see wot the alumina does.

thats assuming i can get the evap head off the cpu, its hard enuff pulling it off with just AS3 and seal string holding it down, god knows how im gonna get it off when its practically glued to the chip with shinetsu. gotta be careful i dont snap the tube off the evap i guess.

[Major]

Good question Ewok! I always use AS3 but but with lower temps I'm not sure what would be best. what do you guys think about cermique ?

[Ewok]

im having abit of trouble at the moment with this chip. 3x 2.8c at 1.7v and about 3700mhz and they were all about -7 under load with AS3.

get this 2.6, 1.70v, 3700mhz and it seemed to be running at 0c with AS3, and is now at +4c with the shinetsu TIM (and i actually saw +9c for a second)

i even reapplied the AS3 thinking i had done it wrong somehow but that seemed to make it even worse. only thing i can think of is that for some reason this chip is just hotter than any of the 3x 2.8s even though its same speed and voltage.

[berkut]

different stepping ?

[Ewok]

stepping on these chips are all the same stepping for the same speed as far as i can tell, all 3 of my 2.8s were SL6WJ even though they were weeks apart and every 2.8 i have seen has been a WJ, the 2.6 is SL6WH.

[Major]

The on die diode might just be reporting higher temps in the cpu. Not necessarily running hoter.

[Ewok]

i was under the impression nothing used the on die sensors and that it was a sensor in the socket on all these boards.

i just figured out why the shinetsu was SO bad, it was only touching maybe 20% of the cpu surface, took the evap off and most of the tim was untouched. which would also explain why since i put this chip in, temps were so bad.

before with the other chips i must have had better contact with the cpu, and also used a bit more AS3, so that probably also helped bridge any gaps from the poor contact.

all the contact was in the opper left of the cpu going along the top and left side with the rest being untouched. the screws were fully tightened though so i dont know how i can get proper contact on it

only thing i can think of is maybe i have too much seal string and thats keeping the evap head a bit above where it would normally be, but im scared of using too little and killing something from condensation.

i have stuck the alimina on it now and im now getting the same temps i was with the 3 2.8 chips and AS3, an average of -7c when under prime and superpi load and if anything its very slightly better.

i still dont know what the cpu contact is like but its obviously a lot better than ive yet managed with this chip. but i still dont know if its good contact or not.

[Teus]

AS Céramique all the way baby
it's the newer edition of AS Alumina. both keep working fine subzero

[Ewok]

well this alimina is doubt pretty much the same as my AS3 did, maybe very slightly worse at this point but we are talking about 1c.

how much better is ceramique than alimina? cos if it isnt 3c or more then its not really worth it with phase change as its unlikely to gain you anything.

[Teus]

difference between AS3 and Céramique is temp-related very small. go for Céramique anyway as it's better then Alumina and easier to clean of then AS3

[Ewok]

priority now is trying to get better evap contact because if when the screws are fully tightened it can still only be touching less than half the core, then even when its getting better contact its not exactly going to have decent pressure on the contact. so next time i take it off ill add the extra washers to the screws and maybe use less seal string so its not getting in the way as much.

i can probably get much better temps if i can somehow get this contact GOOD because at the moment its obviously dodgy.

probably wont bother with ceramique as i think AS3 will still be slightly better once ive managed to sort the contact properly.

[Holst]

Im using AS2 I got a bigish tub and its not run out yet

Ewk is correct that a good contact is much more important.

I dont quite understand how you can get a bad contact, ive remounted my system maybee 10-15 times on AMD CPUs and ive allways got a good contact.

Maybee your hose is twisted or your lower mount plate is stuck on an SMD ?

[Ewok]

i think the hose must have been slightly twisted or something, only the top left part of it was touching the cpu as there were no marks on the rest of the TIM.

i think this alumina might be slightly better than AS3 but theres 1 or 2c in it at most.

after some more testing i can confirm that alumina IS a bit better than AS3 for subzero temps, the warmest ive had this at 1.70v has been -8c average under mbm and AS3 was never that good, at best AS3 was -7c and under these very warm conditions i think it dropped to more like 5c.

so by my estimation alimina gives something between 2 and 3c better temps under load than AS3 (and it seems to be 2-3c better temps at idle as well).