• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Applying thermal paste on a lapped CPU?

K

kkant

n00b
Joined
May 23, 2007
Messages
7
Check out this article:

http://arstechnica.com/news.ars/pos...ilities-with-simple-manufacturing-change.html

Seems to imply that maybe a slightly ridged surface is better when applying paste using the AS5-recommended method (apply a blob of paste in the middle, and let the heatsink squash it down).

For our lapped mirror-finish CPUs and heatsinks, which method is better? The AS5 method, or the "spread it by hand with saran-wrapped fingers" method?
 
for my lapped & polished CPU I just use a razorblade to spread the AS5 on ultra thin.
 
I usually do the razor blade method as well. The only thing I do, and this may just be for superstitious reasons is put a small, less than half of the rice grain dot recommended, on the center over the die to fill in any unevenness there might be from the lapping or the razor blade.
 
put it on any way you want but if both surfaces are properly lapped flat much less than normal is needed.
 
Thanks for the replies folks. Has any one experimented with both methods on lapped surfaces, to see if there's a difference in temps?

If not, I'll probably go with a tiny stripe of Shin-Etsu X23 paste, and the AS5 "squash with heatsink" method. Just because it's easier.
 
kkant said:
Thanks for the replies folks. Has any one experimented with both methods on lapped surfaces, to see if there's a difference in temps?

If not, I'll probably go with a tiny stripe of Shin-Etsu X23 paste, and the AS5 "squash with heatsink" method. Just because it's easier.
Click to expand...

I'd suggest Shin-Etsu X23 and then razorblade method myself.

the only reason you might shy away from the squash method, is you apply it dead center, and when you squash, you're almost certainly going to get a high spot in the center. you have no idea how thin you can apply compound.. it's crazy. Best method that I can recommend is apply with a razorblade, then set the HSF/waterblock on the core and twist and apply pressure to get all air pockets out. make sure you dont clamp one side and then drop down the other side.. it will force more compound to one side of the sink. It's important to realize 2 properly polished faces are literally gapped by microns.

I've had good results with mine, I run a lapped E6600 on water, my idle temp @ 75*F ambient is 24*C - exact room temperature and max load of 37*C.
 
Cool, thanks for the info cerebrex. I think I'll try out both methods, and see how much the temps change.
 
.
.
why kinda of alcohol?

Rubbing? Isopropyl? Everclear?


I've always wondered why, when you buy thermal paste, they give so much...and you need so little.

Why not sell 75% less, for 50% the cost?
 
squatpuke said:
.
.
why kinda of alcohol?

Rubbing? Isopropyl? Everclear?


I've always wondered why, when you buy thermal paste, they give so much...and you need so little.

Why not sell 75% less, for 50% the cost?
Click to expand...

Isopropyl - 90% or better.
They sell 91% alcohol at just about any local store for a few bucks.
It's recommended to use on all surfaces that will be getting new TIM.

Personally, I apply thermal paste like this:

*Use one end of a Q-Tip dipped in 91% alcohol to clean surface of HSF and CPU.
*Use other end of Q-Tip to dry and ensure that the area is completely clean and no moisture is left.
*Take a sandwich baggy / zip-lock bag and place it over your hand.
*Apply a rice-grain size of TIM to both the HSF and the CPU.
*Use your finger inside the baggy to rub the paste completely over the entire surface of both items. Rub thoroughly until the entire surface is coated.
*Then take a paper towel abd wipe both surfaces completely clean.
(The idea is to force TIM into whatever microscopic nooks and crannies are on both surfaces.)
*After that, place either a rice grain sized little dab of TIM on the CPU or if you feel more comfortable, a small strip directly over where both cores sit inside the CPU.
*Mount HSF, make sure it is tightly sealed, and boot 'er up.

It should take a few days for the TIM to completely burn in and spread itself perfectly over the hottest part of the processor.
By day 3, you'll know if the application was perfect because your temps should be noticeably lower.

If you have higher temps after day 3 than they were before to reapplied the TIM - take it out and start over.

My method has worked perfectly for me since I started using it and I recommend to anyone as a basic guide in TIM application.

Good luck. ;)
 
Great post man...thanks...

I hadn't heard of this method before, but it makes sense. I'll try it, when my time comes.


Last time I used TIM was with my AMD TBird 800 mghz...7 or 8 years ago. Used the razor blade method...thing is still working...(never OC'd though).
 
So here is a shot of my q6600 installed in the MB with AS5 right before I added the HS. It shows the right amount in my opinion given a lapped HS and CPU (which is a thicker line than I used before); the red triangle I drew shows where that tag is on the CPU, remember that on quad core chips, the dies are placed in a different located relative to a dual core, see the instructions on AS5's website for more):

line800qe9.jpg
 
I'd be happy to have anyone correct me if I'm wrong, but that looks like it will do the trick perfectly, graysky.

Some people block off the section where the 2 cores sit and perfectly flatten a smooth strip on the CPU before adding the TIM....but I think that's usually only with more "pasty" types of goop.

With AS5, it will spread a little on it's own with heat, so I think you're more than covered and both cores should be nicely protected now.

Good luck and keep us posted.
I have Ceramique arriving tomorrow or Tuesday and I'll try to snap a few pics when I get around to reapplying my thermal.
 
Yeah, I've done the Evoic method also, it's what I'll probably do this weekend, work in the AS5, wipe it off, then apply the strip in the middle.
 
Is this method applicable to any kind of HSF mounting or can it only be applied to lapped CPU's?
 
Here is a GREAT guide for applying some of the thicker, more industrial, type of TIM....such as Shin-Etsu that's being used in the guide.

Shin-Etsu Application guide

I would recommend my method of coating the CPU, then removing, and finally reapplying for more solvent types of TIM.

For the paste, I'm thinking that the guide above is probably the best possible method out there.
 
I go for the razor blade/credit card method myself. I base that on my understanding of how thermal paste is supposed to work.
Thermal paste is a reasonable conductor of heat. It is a LOT better than air, but nowhere near as good as copper. Therefore it is NOT a good idea to have a layer of thermal paste sitting between the contact surfaces of a heatsink and CPU, as it will actually act as an insulator between the two metal surfaces. Rather the thermal paste should only be filling in the microscopic gaps and irregularities in the HS and cpu surfaces (as these would otherwise be filled with air, which is a crap conductor), with as much metal-metal contact as possible. If heatsinks and cpus were perfectly machined then we wouldn't use thermal paste at all as you'd get a perfect metal-to-metal contact area but as they aren't giving the surface a skim with a razor/credit card ensires that paste is forced into any smalll gaps and that all excess paste is skimmed off, leaving only the thinnest of films.
 
You must log in or register to reply here.
Back
Top