Why does this TIM product not exist?

Legend

Legend

Legend

Soldato
Joined
31 Dec 2006
Posts
7,224
I was thinking of a product, a wafer micron thin layer of TIM that could be quickly and easily applied to a CPU in a matter of seconds, ensuring even coverage and optimal cooling. It could be sandwiched between two layers of plastic, you peel off one side, place TIM side down on the CPU, remove top layer, job done.

Now, I know putting a rice grain blob on there as is the standard method is hardly rocket science, but people DO still get it wrong, and it's messier. Plus Thermal Grizzly recommend to spread theirs, so there is obviously something in that method. Just occurred to me that there could be an utterly foolproof solution, and it might even offer slightly better performance (not that there's more than a few degrees separating all TIMs anyway). Hardly Dragon's Den worthy, but just a thought. :)
 
After Eight mints are the perfect size and come in a multi-pack too.

In all serious though it is a good idea, but it would have to be wafer thin and could be more fiddly, break apart on application.
 
Legend said:
Plus Thermal Grizzly recommend to spread theirs, so there is obviously something in that method.
Click to expand...

There are different approaches (without there being one that is inherently superior in itself) to the make up of a thermal compound that require different application methods. Hand spreading is often a poor idea but some pastes have a consistency and/or viscosity that can compensate for that or requires it - it largely stems out of older thermal pastes being of a type that required it so people got in a habit of doing it.
 
Legend said:
Is that Coollaboratory one any good? I had read the liquid form needs to be applied VERY carefully, so presumably the pad is simpler?
Click to expand...

Seems to be a lot of reports it doesn't work well unless both surfaces are perfectly flat - if you have any concave/convex shape to the CPU, etc. it seems to move away from and/or doesn't fill where you need the interface material the most (somewhat ironically).

(This is a lot of the reason I still get great results with a well positioned line of AS5 - despite what some reviewers say who insist on hand spreading it :( )
 
Last edited:
Rroff said:
Seems to be a lot of reports it doesn't work well unless both surfaces are perfectly flat - if you have any concave/convex shape to the CPU, etc. it seems to move away from and/or doesn't fill where you need the interface material the most (somewhat ironically).

(This is a lot of the reason I still get great results with a well positioned line of AS5 - despite what some reviewers say who insist on hand spreading it :( )
Click to expand...

Makes sense, the pad would be the same thickness regardless of the smoothness, whereas the pea/rice blob would be able to leave a little more in the less smooth areas.
 
Legend said:
I was thinking of a product, a wafer micron thin layer of TIM that could be quickly and easily applied to a CPU in a matter of seconds, ensuring even coverage and optimal cooling. It could be sandwiched between two layers of plastic, you peel off one side, place TIM side down on the CPU, remove top layer, job done.

Now, I know putting a rice grain blob on there as is the standard method is hardly rocket science, but people DO still get it wrong, and it's messier. Plus Thermal Grizzly recommend to spread theirs, so there is obviously something in that method. Just occurred to me that there could be an utterly foolproof solution, and it might even offer slightly better performance (not that there's more than a few degrees separating all TIMs anyway). Hardly Dragon's Den worthy, but just a thought. :)
Click to expand...

If you spread you get pockets of air trapped, this air expands when heated resulting in poor heat transfer because you have air pockets where there should be TIM, I'd imagine you'd have the same issue with this thermal tape idea.

When using blob method, all of the air is pushed off to the outside and you get a much more better contact.
 
Last edited:
I have always spread my thermal paste and have never had any issues. Thermal Grizzly, Gelid GC Extreme and some others even come with spreaders.
 
I've always just plopped a blob the size of a large grain of rice in the middle, then with some clingfilm or a bit torn off a plastic carrier bag on the end of my finger, spread it evenly across the whole surface of the chip.. It's always seemed to work fine.

There should be just enough paste on the chip to very thinly evenly cover the surface so you can't see any of the chip any more.
 
Last edited:
pastymuncher said:
I have always spread my thermal paste and have never had any issues. Thermal Grizzly, Gelid GC Extreme and some others even come with spreaders.
Click to expand...

You aren't going to have like a 20C increase in temps from incorrect application in the context of how its spread - some like TG come with spreaders because they are designed to be hand spread - stuff like AS5 doesn't come with spreaders because its designed to work optimally when naturally spread.

mattyfez said:
I've always just plopped a blob the size of a large grain of rice in the middle, then with some clingfilm or a bit torn off a plastic carrier bag on the end of my finger, spread it evenly across the whole surface of the chip.. It's always seemed to work fine.
Click to expand...

If its stuff like AS5 you are just wasting your time and potentially slightly degrading its performance - blob or line and stick the heatsink on will have either identical results or slightly better results. For other pastes it can be a different story.

EDIT: I really don't care how people chose to do it themselves - just annoys me when I see reviewers, etc. write a paste down based on stubbornly incorrect application when I know from experience of apply pastes as per the application notes it can do better.
 
Last edited:
I'm not debating that just putting a blob in the middle and using the heatsink pressure to spread it is better or worse, but spreading has never caused me issues, undoubtedly if the chip surface or the cooler surface is not perfectly flat, you could end up with 'air pockets'.

It would be interesting to see some exhaustive group tests comparing both methods, maybe with 5 different types of popular tim, and 5 popular coolers..but I think in reality the results could still be debated as there could be manufacturing tolerances to consider, for chip, cooler and paste, youd have to do extensive testing with with maybe ten 'identical' chips, ten coolers each from each vendor, and the same with the paste..it would take days and days, and be expensive, and the results I imagine would be inconclusive.

Only my experience but ive always used AS5, apart from my current build, where I couldn't be bothered to dig out the tube, so I just used the paste that came with my cooler master 212 evo, and shock horror, it runs nice and cold.

It goes against my O.C.D, but unless you have a temp issue, I wouldn't pay too much heed to type of paste or method of application.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom