weird lap job

I just tried lapping my cpu heatsink (stock amd2500) and reaplied as3 and mbm5 reports a gain in temperature and my external probe (mounted under cpu) reports a drop... wth.

Did you get the bottome of the heatsink really smooth?

Also, make sure you didn't sand at an angle so the bottom is slanted now.

How big are the gains and drops?

I'd give it a few days to let the AS3 cure. But, it's weird that it's reporting that kind of a difference. Keep an eye on it and keep us posted.

Maybe you lapped the surface so it became convex. Did it on plate glass going only one direction?

I had a wierd lap dance before.

Did you use too much AS3? If you lap a heatsink well, you need to put on less compound.

Ya what was your lapping technique? Amount of As3? Tryed re-seating it?

you said the probe reported a drop and the board reported an increase. thats really wierd, but I would trust the probe over the motherboard's sensors. What kind of motherboard is it?

weird lap dance?

me too, mine was with this BUFFF chick all cool at first and everything THEN she flexed her thighs or something and then you could see the MUSCLE....



anyways on topic now, the external temp probe reports temperatures relative to the contact it makes against the core, e.g. if its taped somewhere its not gonna be accurate...

but in general you should try relapping the h\sink and see how it goes, only going to 600 grit this time..

and besides, temperature is only relative to a good overclock, so dun worry too much about it

hehe,,, buff chick? lol,, yeah 600 is what is useed for water cooler blocks,,,, so more than that isnt always neccesary,,

Okay...

Sorry for the late reply. I was trying to get moved to a new apartment.


Procedure:

Glass door front from my entertainment center laid down on floor with various sheets of sandpaper tapped to it. Motions included lapping in 2 directions though alternating every 5-10 mins. Basically I'd lap back and forth till fairly satisfied then would rotate 90 degrees and repeat. Finally, I'd increase grit# and repeat procedure again, until I was through 1500 (though not long at that one).

AS3 app and Reseating:
I'd heard most ppl put compounds on too thick so I decided to try the thin route. Afterall, it is designed to just fill the nearly microscopic voids right? So I wrapped a finger in plastic wrap and essentially put it on a thin as possible without seeing the core. The gentlest brush against core would score it down to the core. Maybe to thin I suppose. Also I wiped the heatsink off with 91% Isopropyl afterwords, so it should be fairly clean in that sense.

MOBO:

ABIT NF7-S

p.s. gains were around 5C .

Hopefully that answers all. I'll try to get a sig up soon with box description. Thanks all!!

LxOxSxI

you're also supposed to put a very very small amount on the heatsink where the core would make contact, rub it in with plastic on your finger, then whip it off with a lint free cloth..

you prolly just dind't apply enough on the cpu core, give it another go...

One way I have found to be an effective way of measuring the contact between the core and heat sink base is to apply a very thin layer of as3 to the heat sink base only. Then attach the base, let it sit for a minute or two and remove it again. Look at the base. You should see a perfect imprint of the cpu core. Again, only a very thin layer of compound should be used, because if it's too thick, it won't reveal any gaps.