Applying TIM on quad-cores

[krnmastersgt]

I've seen a lot of guides on how to mount coolers and apply TIM on quad cores but a lot of them differ, some say you should apply a dab in the middle and just press down with the cooler, others say dab and use a razor to spread it, more say apply a line etc.
Since I'm going to remount this weekend to kill my heat problem, I wanted to ask how everyone thinks I should apply any TIM I get, probably IC Diamond 7 but I have some AS5 leftover so I might use that instead.

 

[myocardia]

When I mounted my Scythe Ninja B on my Q6600, I decided to try doing it in a slightly different way. So, I researched where the cores were located on Kentsfields, then I put a 1/3 to 1/2 an uncooked grain of rice sized "spot" of TIM over each core. As far as I'm concerned, it worked out very well.

 

[DSF]

These are Arctic's own instructions for Arctic Silver 5:

http://www.arcticsilver.com/pd...s5_intel_quad_wcap.pdf

Personally, that's what I'd go with, but as long as you apply the goop properly, the manner in which you do it shouldn't make a whole lot of difference. Not more than couple degrees C I would think.

 

[krnmastersgt]

What'd you use to spread the TIM? Or did you just mount the cooler on it and let the TIM spread on its own?

 

[krnmastersgt]

Originally posted by: DSF
Personally, that's what I'd go with, but as long as you apply the goop properly, the manner in which you do it shouldn't make a whole lot of difference. Not more than couple degrees C I would think.

Those couple of degrees is what I'm trying to get, I know it shouldn't make too big a difference but with my TEC cooler I need the best possible reading from the cpu -> cooler, the more accurate the better since the higher the temp the more the TEC unit works.

 

[BonzaiDuck]

Dab in the middle, spread with a razor.

I put 2/3 of the diamond TIM I plan to use on the lapped IHS. With razor-blade, I spread to all four sides, scrape the blade with another blade, dab where its sparse, spread until there's a layer all around.

The remaining TIM I use to "tint" the heatsink base. Then I scrape the the blade with the other blade, put the residue in the center of the IHS, and apply the heatsink assembly.

 

[myocardia]

Originally posted by: krnmastersgt
What'd you use to spread the TIM? Or did you just mount the cooler on it and let the TIM spread on its own?

If this question was for me, I just let the heatsink spread it. If you're going to spread it yourself, put the TIM in the center, then spread it.

 

[imported_wired247]

I prefer the razor method for anything more than a single core in dead center

It's very difficult to judge "just the right amount" to dab in the middle so that you cover the 4 cores evenly without over-applying and getting goop all over the place.

Of course, for single core dead center, you really don't need TIM spread out to every corner of the cpu's integrated heat spreader... so a small dab in the middle is fine.

 

[jiffylube1024]

Dab in middle, spread evenly with razor. MX-2 makes spreading suprisingly easy (picked up a tube of it to try it out); puts Arctic Silver 5 to shame in terms of 'spreadability' (and performs at least as well, in part, I think, because it's so easy to spread).

 

[sgrinavi]

I just tried the diamond paste, the website instructions are for the blob & smash method, my temps on my q6600 are down a tad so I'll take it.

 

[BonzaiDuck]

sgrinavi--

"Ah been done sayin' dis for six months. But done nobody lissen. Ah been wantin' to sell dem Wolf-tickets cuz dey don' LIS-sen!!"

To really see it, you need to collect CoreTemp samples over an hour and process with Excel statistical functions. It's obvious to me. People keep talking about "MX-2" and "Shin-et-su" and of course, Arctic Silver 5.

For the $5, no degradation over life of the installation, optional "re-use-ability," no conductivity or capacitance, it's worth it. Just harder to spread.

Even a 2 to 3C improvement in temps over AS5 with the same bench configuration is worth the $5.