Sign in with
Sign up | Sign in
Your question

Applying Thermal Compound

Last response: in Overclocking
Share
a c 226 K Overclocking
September 26, 2012 11:51:52 AM

One frequently reoccurring question is regarding the application of thermal compound, which application method is best to use, my way is better than your way, Blah, Blah, Blah!

Personally I'm not a gambler and never have been, I'm not going to apply any heat sink wondering if the thermal application is correct and I have the coverage I need, because I decided to use one method over another because someone said it was OK.

Most in this overclocking section are after the maximum cooling performance they can get from the cooling solution they've chosen.

The effectiveness of the thermal compound application has everything to do with your heat sink or water block cooling performance for the best overclocking temperature edge possible.

First off I want to explain what thermal compound is for, and what it does.

No matter how good the machining process of your heat sink is your heat sink base has imperfections, even one taken to a mirror finish has microscopic imperfections, a CPU heat spreader has imperfections also, it may be cupped in, it may be bowed out, and it definitely has microscopic imperfections.

When both surfaces come together there is air space between the two contacting surfaces, the air space becomes an insulator between the two surfaces and an insulator is bad for the conduction of the heat from one surface to the other.

Thus the purpose of thermal compounds or TIM (Thermal Insulating Material), now the application of TIM is very important to the overall cooling performance of any heat sink cooler, the worst TIM on the planet is the pre-applied TIM that comes on some heat sinks that has to melt in place, because it always forms a solid layer.

No matter which application method you use your ultimate goal for the best performance is the thinnest full coverage application possible, you are only filling the microscopic imperfections between the two surfaces.

If you use too much TIM it acts as an insulator between the two surfaces instead of a conductor of the heat, still much better than air but it becomes a maximum performance temperature conduction insulator.

So to be sure, you need to pull the heat sink and inspect the thermal footprint left on the heat spreader and the heat sink base, it will reveal if you're using too much or not enough.

It will also reveal if you have a problem with the heat sink base and heat spreader, for this thread regarding applying thermal compound, the resolution to problems is simple either more or less TIM.

Some of you that think you've done everything right but still have higher than others load temperatures, you need to pull your heat sink and see what's going on underneath it.

It's far better to know for sure than wonder, and if you had enough mental ability to install that heat sink in the first place, you have the same ability to pull it, inspect, and re-install it with a fresh application of TIM.

The different application methods, drop in center, X across, solid line, dual solid lines, spreading a thin even layer, are just an application methods that have been used over and over, the thermal footprint is the most important thing, what happens once your chosen method is clamped in place under pressure?

Pull that heat sink, and Find Out!

Every TIM is a different consistency, some flow much easier than others, the thinner consistencies require less application and less clamping pressure as they spread under pressure more easily, the thicker consistencies are usually more difficult to work with and require a higher clamping pressure.

If you're using a TIM that requires a high clamping pressure you need to be aware of that or it will not seat properly, there are TIMs that require as high as 50psi clamping pressure, and standard after market heat sinks do not apply that much clamping pressure, most clamp around the 35psi range.

Worst case scenario.

Unfortunately some problems require going against what's best regarding TIM application, for example if your CPU heat spreader is cupped inward, upon inspection the center of the CPU area will show a solid TIM layer and that's unfortunately bad, because the only true cooling performance resolution to that problem is to lap the heat spreader, instantly voiding the warranty of the CPU.

No one wants to void a brand new CPU warranty so a lesser cooling performance for that specific cupped in situation will just have to be lived with, using as little TIM as possible to fill the problem area because you do not want air in that gap! TIM is still better than air even if you're stuck with a cupped inward CPU heat spreader and are forced to have a solid layer over the CPU die area just make sure there are no air bubbles trapped inside the layer.

Ryan




a c 226 K Overclocking
September 26, 2012 11:52:33 AM

Something interesting to share with you all regarding heat sinks refined down to a mirror finish.

The picture below is to show you just how polished and reflective the finish of the heat sink base is.

It is so reflective and literally mirror perfect that the picture is of the reflection of the business end of the microscope.

The microscope itself took that picture.



The picture below is beginning to adjust the focus down to the actual surface area itself, note you still see the microscopes shadow.



The picture below is the actual heat sink surface coming into sharpness and you still see a slight microscope shadow.



The picture below is the actual heat sink raw surface at approximately 100X magnification.



Those microscopic scratches and imperfections is all you should be filling with your thermal compound!

Actually almost all thermal compounds today are too thick for 2 contacting mirror finished surfaces, if you have lapped your heat spreader on your cpu and have a mirror finished heat sink base the amount you actually need between the 2 contacting surfaces is very small amount.

This amount of course is determined by if the 2 contacting surfaces are actually perfectly flat which is why I suggest the trial fit, removal and inspection method, so you know what the surfaces are like when they contact each other.

If the surfaces are perfectly flat, all the thermal compound you would actually need with mirrored surfaces is to take a sandwich bag to cover your finger tip and smear a film over both surfaces with no build up at all!
a c 226 K Overclocking
September 26, 2012 11:53:05 AM

Reserved 2
Related resources
a c 226 K Overclocking
September 26, 2012 12:05:33 PM

Reserved 3
a b K Overclocking
September 26, 2012 12:16:14 PM

I always tell people to follow the instructions on the package instead of listening to some random posters advice which is probably wrong anyways.

All thermal compounds are different, and made of different ingredients so it's best to follow the instructions on the package.

a c 226 K Overclocking
March 18, 2013 9:55:50 AM

New information added above.
a c 226 K Overclocking
March 19, 2013 4:43:22 AM

geekapproved said:
I always tell people to follow the instructions on the package instead of listening to some random posters advice which is probably wrong anyways.

All thermal compounds are different, and made of different ingredients so it's best to follow the instructions on the package.



That of course is, if the TIM has packaging instructions.

a c 226 K Overclocking
April 2, 2013 3:20:12 AM

Updated information above.
!