What is the best way to apply Prolimatech PK-1?

[Stealth3si]

that video was in my link except it wasn't embedded...

 

anyway should i 'prime' the surface of the heatsink and/or cpu first?

 

you see the AS instructions on how to apply paste says to apply a small mound on the HS, spread an even layer throughout the surface with a credit card and then use a cloth to wipe it all away leaving a clear haze on the heatsink. this allegedly fills in the microscopic holes.

 

at this point i imagine i would do also do this on the cpu? and finally apply the dot or x method on the cpu?

 

i think that sounds good.

[vandreadstriker]

You can follow the instructions that it come with, it's really up to you which one you're gonna use.

Also, if you spread it onto the HS and the CPU, I'm concerned that it'll build up a thick layer of the TIM which could in turn result as an insulator.

IMO it's better to just use the dot method right away. But remember to squeeze some amount of it so that it spreads more (since the AS5 doesn't spread very well)

[Stealth3si]

You can follow the instructions that it come with, it's really up to you which one you're gonna use.

Also, if you spread it onto the HS and the CPU, I'm concerned that it'll build up a thick layer of the TIM which could in turn result as an insulator.

IMO it's better to just use the dot method right away. But remember to squeeze some amount of it so that it spreads more (since the AS5 doesn't spread very well)

I'm using the Prolimatech PK-1.

 

This is waht I'm going to do:

 

spread a thin even layer on both HS and CPU. Take a lint-free cloth and basically remove the entire layer from the surfaces. This way I don't need to worry about having any layer on it since there will be only a transparent haze remaining on the surface.

 

And from the video, I noticed that the x-method seems do cover more of the surface but that isn't really necessary as long as the main center area of the CPU is covered (by the pea method), am I right?

[Drdeath]

As much as i am all over the pea sized method, the temp differences are only a few degrees any way you slice it.

[Stealth3si]

Sweet.

 

Drdeath....is that name happen to be from Arachnaphobia?

[Stealth3si]

I don't disagree with any of ya'll. I think I might try both ways, somewhat...

 

Because i emailed Prolimatech Tech support and they said:

 

Dear sir,

 

After checking with our engineers about the spread application method, our engineers suggest to apply PK-1 to evenly a layer which is not thin (vaguely seen).

We've modified our website's FAQ (from a thin layer to a layer).

Since PK-1's thermal conductivity is very good (http://www.prolimatech.com/en/products/detail.asp?id=163&subid=1528#showtab), our engineers suggest to apply a layer on both CPU and heatsink which will have better performance.

If you like, you may apply a thick layer; however, a normal vaguely layer is enough. (thick layer means waste of the thermal compound)

 

About your question of spread application method may introduce the possibility of air bubbles, our engineers' reply as follows:

We've tried both ways (a dot in the center and spread it on both CPU and Heatsink) and found both ways have similar performance.

 

Best regards,

 

 

**************************************

Selena Lee

Prolimatech Co.,LTD.

E-mail:*************

Visit our Website at: http://www.prolimatech.com

Facebook: http://www.facebook.com/pages/Prolimatech-Co-LTD/143936609425

Twitter: https://twitter.com/#!/Prolimatech

 

****************************************

 

Now I suspect that they prefer the spread application maybe because they want us to buy more of them in the future?? I don't know.

 

I guess I would try spreading a thin layer on the heatsink and next removing the entire layer but leave a translucent haze that covers the microscopic holes and then applying the dot method on the CPU (or X method to cover more area if I have enough paste remaining.)

[Drdeath]

Just do it both ways and record the results. My opiinion would be use the method that gets the loweest results. This is not brain surgery. It takes lesst than a half hour.