Sign-in / Sign-up
Your question
Solved

Use diamond powder in thermal compound?

Tags:
  • Diamond
  • Thermal Compound
  • Systems
Last response: in Systems
March 19, 2014 6:13:23 AM

How to improve existing thermal compound?

There are many commercially available thermal compounds and I use Arctic Ceramique or Arctic Silver 5. I am sure that there are some others that are marginally better.

After reading on the thermal conductivity of materials I got interested in diamond powder and bought some 0.5 micron (about 20 micro inch) diamond powder. My plan is to continue using the Ceramique compound and then sprinkle or dip the coated CPU and the heatsink base with the diamond powder; just a light sprinkle.

I looked at mixing the powder with glycerin, glycol, but I don't know enough about this.

My question is “What do you think of this?” Anything to add?

All replies are welcome!

More about : diamond powder thermal compound

March 19, 2014 7:09:37 AM

Adding material between your IHS/HSF is going to increase thermal resistance thus decreasing the heat transfer rate.
m
0
l
Related resources

Best solution

March 19, 2014 9:36:04 AM

The more particulates you add to thermal compound the less effective it actually is, because it creates a static layer!

All thermal compound is for, is to fill in the microscopic imperfections between the 2 contacting metal surfaces taking the place of the air that will occupy those areas.

The thicker the layer the more insulating properties it has, so the heat conductivity begins to get lost.

That's why the meltable type thermal compound that comes pre-applied on some heat sinks are not the best performers, as they leave a solid layer as they melt which is exactly what you do not want when it comes to cooling performance.

In all cases when those thermal layers are replaced with just the bare minimum TIM needed cooling performance seriously improves.

Warranty wise heat sink manufacturers want something on their shiny heat sinks out the manufacturing door, because they are fully aware to what finishing extents they went to in the manufacturing process.

In some cases they know they need something on them because the contacting surfaces are far from flat, and their gaps are more than microscopic!

So they use the pre-applied melting type as something is better than nothing in those cases.

To manufacturing machine perfectly flat to a mirror finish is very expensive and costly!

Heat sink manufacturing cost is usually the reason we don't see but the most expensive with mirror finishes, that's why some lap their heat sinks, or water blocks, and CPU heat spreaders.

Almost nothing is needed TIM wise to fill the microscopic imperfections on a good lapping job, and you surely wouldn't want a diamond layer in those circumstances. (Think about it.)

Anyway you look at it when it comes to thermal compound less is best and the diamond particulates will form a solid layer, impeding the very reason you're applying the TIM in the first place.

Artic MX4 is a thinner consistency TIM and it fills the imperfections very well, does not scratch the surface as diamond will, and will thin out to almost nothing as diamond won't, and is some of the best thermal compound I've used yet, as less is best, anyway you look at it.

However Ubrales good luck in your venture, I do hope it's successful!

Ryan
Share
March 19, 2014 9:55:12 AM

this looked interesting to me as i have lived by less is more. the first couple times i did my thermal paste was tricky ended up to much making some up and down idle temps.
now that i got it down its perfectly stable temps with the standard artic silver and even did my card too.

i just wanted to add a little more physics
diamonds carry frequency and in fact no two diamonds can carry the same frequency. so by touching 1 u are changing every single diamond in the universe.
u may find that you end up with strange frequency interference happening randomly
m
0
l
March 19, 2014 11:23:07 AM

Thanks and keep posting guys! I need as much input as possible. The conductivity of diamond is phenomenal and the powder I have is like a fine ground flour. Any particle size less than 0.5 micron will bind together and form an inseparable brick.

Please keep posting your ideas. This is for personal use; not commercial use.
m
0
l
March 20, 2014 2:23:41 AM

IC Diamond was originally invited to be a TIM participant in the Thermal Compound Roundup 2011, I email chatted with Mr. IC Diamond for days and he would not submit a test sample until I agreed to his terms.

They were that his product needed a guaranteed 50psi of clamping pressure and he would supply the laboratory testing paper to ensure I was clamping with that much pressure, before he would submit a test sample.

I finally agreed that I would use 50psi of clamping pressure for IC Diamond, but to be fair to the other candidates supplying test products, they also would be clamped under the same 50psi of pressure, he responded with No, NO, those other products do not need that much clamping pressure!

I told him that, the only way to be fair was to test on equal ground, but he knew that thinner consistency compounds under that much pressure would squish down to almost nothing, (forming the perfect heat conductor!), and far outperforming his product, so he refused to submit a testing sample.

Most heavy heat sinks clamp under approximately 25psi ~ 35psi, a guaranteed 50psi has actually been enough pressure to warp some water block bases, and there are complaints, or there were at the time of that happening, and also CPU warranty loss from the diamond scratching the CPU heat spreader.

Since this test is personal you won't be getting into any legal territory by CPU warranty loss, so that won't be a problem for you.

For the record I still have the email correspondence between us, but Mr. IC Diamond is so quick to file lawsuits, it will remain confidential unless I have to produce it in court!

At one point there was a thread here at Toms regarding legalities that were going on with the IC Diamond lawsuits, I don't know if it is still here or was closed and deleted?

I asked Mr. IC Diamond, why the paste consistency was so thick regarding his product, that if he thinned it out some maybe it would not take so much mounting pressure to seat?

He never answered me, and I'm sure it would have meant going back to the product drawing board, but to me that's what product improvement is all about?

Make it better!

Maybe you will! :) 

FYI: Note the pressure chart, he cannot legally require a higher pressure than 50psi, even though it is clearly insinuated with the display of the chart as his own tests report the best pressure for the product is actually 65psi, and that is definitely water block copper base damage pressure!
m
0
l
March 20, 2014 2:53:44 AM

it is possible to use a man made quirts crystal on a loop with circuitry to produce less power draw and achieve higher stability and is currently being used in some manufacturing. to see someone to modify a pc in this way would be... amazing
i think in real world fair compound testing over synthetic, silver, gold, diamonds would yield that gold in the proper manner would produce the best results as far as performance and temps are concerned
now as synthetics evolve we will most likely see graphene replace silicone and will not require such extreme measure or overclocking of any kind to any consumer of any level
m
0
l