Originally Posted by DaClownie
Smear method usually isn't deemed the most effective. Reason being is when you go to apply your cooler, it could leave some air trapped in between. The dot works great because the TIM starts in the middle of the chip and the pressure spreads it outwards, leaving no gaps. Someone post the link for him, the youtube video about TIM spreading. Its been posted on these forums a million times but I'm at work so I'm having trouble finding it.
I think those videos are bull****. I already know what you're talking about.
1. It looks like it's thick as hell, I never put that much on while using spread.
2. There are f'ing gaps and air pockets in it before he even presses down with that piece of glass/plexi or whatever so how is that even an accurate representation of what's going to happen when he starts out with air under it...
3. That clear piece he's using to put pressure down isn't even flat, look at the second video the sides don't even touch the thermal compound and when the light moves over it right when he takes it away you can see the uneven surface on the edge. How do we know it's not imperfect in the middle? Someone needs to redo this test with accurate equipment and not this plastic looking junk. (And might I also add with an accurate amount of paste too.)
4. How do we know that the (what looks like a plastic plate) isn't bending when he pushes on it from the sides.
My conclusion so far on my lapped processor and ghetto-lapped cooler the dot method is the most effective giving me the best temps, but my setup is proven to be uneven if you look on page 2 so of course the pea method will work better, you're starting out with a glob that spreads outwards. The only problem is you also have a thicker layer of paste compared to if you used spread, and a thicker layer of paste = hotter temps. From my understanding you can only use the spread method when you have a flat cpu and cooler. (Lapped systems only usually) otherwise it will not be effective as the dot method which is forgiving to uneven surfaces unlike the spread method.
When I get my NH-D14 I will make sure both are flat(cpu and cooler) and redo temp tests with MX-4 using pea and spread.Edited by Khaotik55 - 5/20/11 at 1:40pm