One more question, regarding optimal TIM application.

[daehxxiD]

Hey guys...

I have yet another little question about TIM application. As I'm striving to reach best possible contact between cooling surfaces on all my electronical devices (lapping=done ) I've been all over it with TIM, from the Liquid Metal Pads (let's not use those anymore), to AS5, MX-2, IC Diamond 7 and lately, the best until now, MX-3.

Those last few TIMs I mentioned all have one thing in common: They advice to put a tiny dab (in case of multicore and AS5 a line) right onto the HS and let the Heatsink spread it... Doing some testing by myself in these last years, I've found that not only does that save a whole lot of time, but it is easier to do and more efficient than a manual full-spread and usually even will cover the whole HS over time...

And this is where my question starts. So I've observed the TIM thinning out with Thermal cycles and stuff going on, is it adviced to refasten the Heatsink every once in a while in the first few weeks of a mount?
The reason I'm asking is mostly because of my PS3. I've applied MX-3 to it (it's a good ol' 90nm 60GB) because I've noticed quite an improvement over MX-2 when I did the swap in my laptop (seriously, I thought it was dead for a moment there, for more info check the Cat 9.9 Thread... Guess it was "just" a bad-solderpoint; heatgun and added pressure FTW!)... It really blew my mind how much better it worked right off the bat then the AS5 I had applied on the PS3 (which had already spread full surface and cured perfectly fine, having been in that thang for over 1 year), but I've noticed that the screws, even though thightened very firmly on first install get a very tiny bit (like 1/20 of a turn) looser... Does this impact on pressure and thus cooling efficiency or is it basically not worth thinking about?

If it is of any importance, on the ps3 you have these brackets:

that hold down the Heatsink, they're basically being pressed down by the screws which in turn are screwed into the HS. This basically "pulls" the HS right onto the Chips.

Thanks in advance for your time and help people

 

[Onderon]

Well depends on the mounting sistems, there are some that have little springs on the screws and that apply tension and thus preventing them to move, it could cause trouble only IF they unscrew as munch as loosening the heatsink and creating such a gap that allows air, but if its that small and the heatsink is still in the same place theres no problem and no noticeable diference.

Remember or well in my experience, that thermal pastes (not liquid but dense pastes) like mx-2-3 or artic silver 5 when the heatsink applies pressure it creates a little suction betwen the two parts, that why when we remove them it doesnt come right away and you have to use minimal force, and that allows that even if it loosses a tiny bit it will still transfer the same amount of heat.

On a third tought by my experience the only time that I have had the temps go up was only when my AC5 dried up on an athlon 64 that had over 2 years witouth moving the heatsink or changing the paste.

Edit: Remember that with each thermal cicle the thermal pastes thinnes so it can reach it perfect spread coverage, with the pressure applied. it's normal, the all reach a moment that they wont get thinner and that moment if you apply the right amount of tim you will see the whole proc covered and if you apply less you will see that its not covered completely even if you apply more force or go throught more cycles