strange heatsink problem

Ok, this is strange. When I put a fresh coat of ASC on my sk-7, my temps are just fine. They stay in the 40-47 range, idle to full load respectively. Well, after a month of operation, my temps slowely started to go up and up. Eventually my system would start overheating. I had no idea what the problem could be so I pulled off my heatsink to investigate. What I found was that the thin layer of ASC I had applied to the die had been pushed out around its sides. Hardly any thermal compound was on the top of the die. So I cleaned off the cpu, put a fresh coat, and slapped the heatsink and fan back on. Everything was fine till now, a month later, my temps are getting way high again. They just topped out at 51.5c. Im pretty sure for some reason the thermal compound is getting squeezed out. Do I have a defective heatsink? Did I mess it up by lapping it(does this void any warranty)? I don't understand.....should I try to return it to newegg?

well...

lapping it wasnt neceesary, and from what it sounds like, you actually gave yourself an uneven surface.

It might be vibration from the HS and the pressure that is slowly forcing the ASC out from where it should be.

You think the surface is uneven? Then why would everything be fine for the first 3 weeks or so, and why would being uneven cause all the asc to be pushed out the sides of the die?

So should I just buy a new HS? Does lapping a HS void any warranty??

thermal compound does not transmit heat as well at metal->metal contact, but MUCH better than air, that's why people use it, to replace the air that would be trapped without it.

If it's getting pushed out by the pressure, then it's not getting replaced by air, so if anything, thermal conductivity should go up. (lower temps)

have you tried just blowing the dust out of your heatsink and seeing if that helps at all?

Quote:

---

Originally posted by Big_Mac
thermal compound does not transmit heat as well at metal->metal contact, but MUCH better than air, that's why people use it, to replace the air that would be trapped without it.

If it's getting pushed out by the pressure, then it's not getting replaced by air, so if anything, thermal conductivity should go up. (lower temps)

have you tried just blowing the dust out of your heatsink and seeing if that helps at all?

---

Big_Mac,

Need to check those Chem notes again ;)

Air is going to have a higher viscosity then AS3, which means its going to occupy any space the the AS3 gives up.

It sounds like you're getting the effects of thermal expansion and contraction. The thermal expansion of the board is putting pressure on the heatsink forcing the compound out. Then releasing during cooldown.

You might want to take a look at the motherboard screws. Make sure they are all snug, but not over tightened. Some may be too loose letting the board flex slightly during heating and cooling.

The thermal compound is just supposed to fill in the tiny gaps that could be on the surface of the chip and or heatsink. Do you think that the case temps could be too high ?

Quote:

---

Originally posted by Isezumi
Big_Mac,

Need to check those Chem notes again ;)

Air is going to have a higher viscosity then AS3, which means its going to occupy any space the the AS3 gives up.

---

What I meant is that AIR will not creat it's own space, it will only fill a gap that becomes created, and if the heatsink is pressing down hard enough to squeeze out the compound, then it's unlikely to leave a nice gap for air to rush in.

I'm going with the vibration theory here. A possible solution is to place rubber washers between the HS and the Fan, this would in theory lower the amount of vibration going to the heatsink.

Quote:

---

Originally posted by Big_Mac
What I meant is that AIR will not creat it's own space, it will only fill a gap that becomes created, and if the heatsink is pressing down hard enough to squeeze out the compound, then it's unlikely to leave a nice gap for air to rush in.

---

He has an SK-7 which means its a tab mount, which means he must Push down on the sink to get the tabs to connect...in fact for a short period of time he has to exert MORE force against the core then is necessary (in order to attach the clips over the tabs) that displaces the thermal compound around the edges. Once pressure(s) begins to normalize it could cause there to be gaps in the contact surface, gaps that gorw more volumous as compound cures. Having even remotely uneven surfaces form self lapping can make this problem worse.

Ok, well, I dunno if it is vibrations, expanding and contracting, or to much pressure from installation! lol I am thinking about just ordering a new HS and not worry about it anymore! The idea of having to keep pulling off my HS and fan bother me. I already put a screwdriver to the mobo once on accident. Thank goodness I didn't mess anything up. So should I go with the bolt on Thermalright SLK-947-U? I was looking at this bad boy :D Also, my SFII is loud and soon I am going to be using my pc for music production, so I need a good quite fan. What do you guys suggest, I heard panaflo (sp?) makes a good quite fan, anyone?