Thermal transfer material squeezed out? Just the facts.

  • Thread starter Thread starter Matt
  • Start date Start date
M

Matt

We have a bunch of folks who have installed and removed heat sinks to
and from CPUs, along with the question of whether excess material is
squeezed out by the pressure of heatsink clamps. Maybe we can clear
things up.

Here I solicit observations related to the hypothesized squeezing
phenomenon.

To retain focus, I propose that for the first 48 hours of the thread we
keep reasoning, deduction, conclusions, and controversey to a minimum,
and stick to observations and observed patterns. The arguments if
necessary can start a couple days from now.

The emphasis is on the description of relevant details, especially
brands and characteristics of materials and components used.
 
Matt said:
The emphasis is on the description of relevant details, especially
brands and characteristics of materials and components used.
Click to expand...

I installed a retail-boxed XP 2500+ Barton with stock heatsink and fan
on a Biostar M7NCD Pro mobo maybe eight months back. The heatsink had a
gray patch of wax-like thermal transfer material maybe a half to a
fourth of a millimeter thick. Hooking the HS clamp onto the motherboard
with a screwdriver required considerable force. After running the
machine for a few days, I removed the heatsink from the CPU and found
IIRC that I could see the surface of the heatsink through the layer of
thermal material, at least in some spots. Definitely I could see a
ledge of material that had been surrounding the die: I could see the
imprint of the die in the material, and the imprinted area had a very
thin layer of material. Also I could see the purple surface of the die
through the remaining thin layer of thermal material. All of this is to
say the material appeared to have been almost all squeezed out.

I used 91% isopropyl alcohol and coffee filters to remove the used
thermal material (easily), and reinstalled the heatsink using Arctic
Silver 3 and following as well as I could the AS3 instructions. I seem
to recall removing it again to check whether I had used enough AS3,
found that I hadn't and reinstalled using more AS3. I did observe a
breaking-in period for the AS3 installation, which is to say that the
CPU temps did improve a bit (1 to 3 deg C? I forgot how much) after a
few weeks of use. Sorry I don't recall many details of the AS3
application, except that the recommended amount was not enough, meaning
that the die did not the first time make a complete AS3 print on the
heatsink.
 
Matt said:
We have a bunch of folks who have installed and removed heat sinks to
and from CPUs, along with the question of whether excess material is
squeezed out by the pressure of heatsink clamps.
Click to expand...

From what I've seen, the excess (using plain 'ol white HS silicone grease
aka radio shack type) get's mushed out the sides no problem. Maybe some
thicker (arctic -insert your favorite low temp sounding metal-) type
doesn't do this? I've never had a problem with a rice sized blob, smear it
around a little and clamp the HS down. It always has a very thin layer when
removed later and never had any problem removing a HS with this stuff
either.
 
Stacey said:
Matt wrote:




From what I've seen, the excess (using plain 'ol white HS silicone grease
aka radio shack type) get's mushed out the sides no problem. Maybe some
thicker (arctic -insert your favorite low temp sounding metal-) type
doesn't do this? I've never had a problem with a rice sized blob, smear it
around a little and clamp the HS down. It always has a very thin layer when
removed later and never had any problem removing a HS with this stuff
either.
Click to expand...
The stuff that oozes out can form a HEAT dam and hold IN the heat!

Use thermal grease very sparingly, only enough to fill the micro voids
in the two surfaces, and not enough to keep the heat sink seperated from
the top of the cpu.
 
Since the proposed cease fire and waiting period is now null and void....
The stuff that oozes out can form a HEAT dam and hold IN the heat!
Click to expand...

Then clean it off. Did you have problems with diapers or something when you
were a kid?
Use thermal grease very sparingly, only enough to fill the micro voids
in the two surfaces, and not enough to keep the heat sink seperated from
the top of the cpu.
Click to expand...

By definition, it's impossible for something as coarse as human fingers or
related tools to fill *only* "micro voids" in the 2 separate surfaces, so
don't bother trying.
 
Patrick said:
Since the proposed cease fire and waiting period is now null and
void....
Click to expand...
In
jeffc said:
Then clean it off. Did you have problems with diapers or something
when you were a kid?


By definition, it's impossible for something as coarse as human
fingers or related tools to fill *only* "micro voids" in the 2
separate surfaces, so don't bother trying.
Click to expand...

I've always applied Thermal Grease with a razor blade. (Read that somewhere)
In other words, just put a little dab on the CPU and use the edge of the
razor blade to kind of very gently scrape it on...that way you get a very
even fine coat with very little thermal grease used. Always has worked for
me. I've always heard the smaller amount the better, as long as it covers
the CPU, as your only trying to fill in microscopic little 'pits', not big
cracks.

....Allen
 
Patrick said:
The stuff that oozes out can form a HEAT dam and hold IN the heat!
Click to expand...

?? Since the heat is being transfered between the -top of the CPU- and the
-bottom of the heat sink-, that doesn't sound reasonable. How is that going
to stop heat from being transfered between those two surfaces? How much
heat do you think is being transfered out the side of the chip to the air
between the tiny gap between the sink and the CPU board?
Use thermal grease very sparingly, only enough to fill the micro voids
in the two surfaces, and not enough to keep the heat sink seperated from
the top of the cpu.
Click to expand...

If there was enough staying between the top of the CPU and the heat sink to
"seperate" them, it wouldn't have squished out from between them! Given how
tight most clamps hold the heat sink down, the posibility of that
seperation happening is nil. Now maybe with some sort of "ultra silver" HS
grease it's so thick this is a problem? I wouldn't know about that one..

I agree you don't need to use any more than needed to cover the CPU and if
you insist on using conducting material you have to be very careful, that's
why I don't use that stuff. I've used silicone grease for over a decade of
building computers, overclocked etc and never had the first cooling related
problem.
 
Back
Top