TIM comparison at 120°F Ambient temps

[I.M.O.G.]

Joe Citarella did some good testing about paste application patterns. Didn't matter much, except you shouldn't attempt to spread the paste using a razor blade or credit card - its more likely to result in uneven coverage resulting in air gaps.

Center is best, or a line following the alignment of modern die. I use a peasize amount in the center, then let pressure spread it out.

When designing your hold down, don't worry too much about pressure - apply a lot, but not enough to flex the PCB or crush the socket. Preferably, you want to ensure the pressure is even around all sides. The easiest ways to go wrong are to not apply enough pressure, or to flex the mobo PCB which could cause traces to fail eventually which would cause errors or mobo death.

From experience strapping really heavy copper pots onto processors, its hard to apply too much pressure, especially if you use a backplate that reinforces the mobo.

And welcome to OCF!

 

[hxtasy]

My understanding is the chips make the heat and size of cover is sized to match the PCB area needed for the 1000+ pins connecting CPU to motherboard. Chips are in center of CPU or in a line. I have never seen them reach the edge. The rest of cover is there as a base for cooler to set on. The PugetSystems pea test was not centered so possible the bottom edge of chips wasn't covered by TIM. The line test appears to be running the wrong direction. I think the chip line is up and down not side to side.

Yes, like this for example:

The IHS is exactly just that, it gets the heat away from the core quickly and spreads it outward, from there it is up to the heat sink to wick it away.

a lot of i7's or BGA style processors dont have the IHS, so you are just mounting the top of the core dies right to the heat sink. or people that de-lid and then don't use an IHS.

I WISH i could do this in production environment. i would kill to get a 10C drop right now!

you can also see the inconsistencies across the different cores, and on a lot of de-lid test you can see the core temps more stable against eachother.

 

[Conumdrum]

Too much TIM test:

Processor heat sink clearance dry: 0.051 mm (about two thousandths of an inch)

Processor heat sink clearance excessive, almost sarcastic amount of TIM: 0.051 mm


temperature results are exactly the same as my last coolermaster graph, no need to graph it.

heat sink had shoulder screws and torque does not change clearance. this cannot be said for clamp style or other style heat sinks.

Great pics! Now we can see more. Your testing methods are superior unless your heatsink doesn't apply enough pressure to squeeze the material, which is used for engine building for crankshafts? The torque value or lbs per inch or 'squeeze' used on crankshaft bolts must be 10's of times greater. A heatsink unless a bolted on no springs method probably isn't near enough force to get the 'squeeze' you see typically with 10mm bolts of special steel with grade 10+ bolts. If your not in the range of what the measuring green device is designed to work with, then it ain't valid.

So not sure if that way works the best. Do they make a softer grade stuff for use for much less torque loads?

Now we are talking science here. Now it's fun.

Finally, you sure your heatsink and CPU are perfectly flat? You make no mention of the cooler your using or your mounting method.

Lets keep it going.........

 

[I.M.O.G.]

If you want to test to see how much mounting pressure you are getting and how even it is, order a free sample of fujifilm prescale ultra low: http://www.sensorprod.com/fuji-prescale.php

Ultra low is the one that does 30-85 psi... This the range your mount should be in. Ideally on the higher end of 50-80psi.

80psi is the ideal mostly, though some pastes may say different if they state anything. If you read any science focused or industrial focused thermal interface stuff, there will be discussion of °C-in²/W which relates to thermal impedance, and is only relevant when stated alongside a contact pressure in PSI. This is the same stuff Joe used to talk about also on the front page with his heatsink and TIM testing.

I don't know what PSI typically starts to bend the PCB of the mobo, but that probably varies too depending on the mobo model/layers and the use of a backplate or not.

 

[hxtasy]

Great pics! Now we can see more. Your testing methods are superior unless your heatsink doesn't apply enough pressure to squeeze the material, which is used for engine building for crankshafts? The torque value or lbs per inch or 'squeeze' used on crankshaft bolts must be 10's of times greater. A heatsink unless a bolted on no springs method probably isn't near enough force to get the 'squeeze' you see typically with 10mm bolts of special steel with grade 10+ bolts. If your not in the range of what the measuring green device is designed to work with, then it ain't valid.

So not sure if that way works the best. Do they make a softer grade stuff for use for much less torque loads?

Now we are talking science here. Now it's fun.

Finally, you sure your heatsink and CPU are perfectly flat? You make no mention of the cooler your using or your mounting method.

Lets keep it going.........

Hey Conumdrum, the plastigauge was just to show that the clearance did not change with application of the paste. Of course it is going to change, but in a manner so minute it wouldnt be measureable (not in my lab).

the torque on your girdle or connecting rod caps that hold your bearings into place is a lot, but this is just for preload of the surfaces being clamped together. since it's two machined surfaces being clamped together (eg con rod and con rod end cap) the clearance between them should not change.

it is designed to have a certain amount of clearance since your connecting rod is spinning on top of this metal to metal bearing, a very thin layer of oil is the only thing keeping them metal floating.

So since i am just measuring clearance, the plastigauge should be sufficient, if it was out of this range it would compress more or less.


As I.M.O.G. mentioned, measuring actual pressure and deformation you will need a different device.


thanks I.M.O.G., ive always wanted to use the fuji film but stayed away from it because of the price. This will help with my homemade heat sink torque settings (ordered a sample

as for the heat sink in my previous post, its a 1U system, so its all passive heat sinks.