diamond thermal grease?

[Tech Geek Deluxe]

I appreciate that ... I have 2 (ea) 120mm Intakes and 2 (ea) 120mm Exhausts ... one of those is the PS ... so I have positive pressure in the case and filtered air in it to keep the dust down.

I tested again today and ... sure enough ... I got 1 degree better cooling ...

That with upped core voltage and moving my Ram timings from 5-5-5-15 to 4-4-4-12

I will try another method with this PC ... one more time ... I'll use a mix of carbon black and diamond and polysynthetics ...

I suspect that the grease with the diamonds is just too lumpy ... partice size is fine but the diamonds must be lumping together and there isn't sufficient fluid to allow for a thin transfer media.

I figure that carbon black and polysynthetics will make for a much thinner application.

The diamond content will drop from 50% to 20% ... and carbon black will be in it at about 20% ... this hopefully will give me a thinner media and still have fast and efficient heat transfer.

[Tech Geek Deluxe]

Is there a particular way to upload images on this site? Must they be "hosted" elsewhere or can I simply install a pic within my article.

[Elysium]

Originally Posted by Tech Geek Deluxe

Is there a particular way to upload images on this site? Must they be "hosted" elsewhere or can I simply install a pic within my article.

you gotta host them on a different site like PhotoBucket or Imageshack. Or you can use the HL Gallery!

[Tech Geek Deluxe]

Well ... I could use some assessments. Are these temps pretty run of the mill for this setup ... or blah ... too high ... or slight eyebrow raised ... hmmm ... pretty good.

Here are pics of my Idle Temps and Orthos full load temps on my Q6600 at 3Ghz. [I took pictures of the highest combination of core temps I could get it to read. In their fluctuation the highest would drop some 5 degrees lower ... and the lowest would drop 3 degrees lower ... on full load ... idle was 3/2 respectively.]

Core 2 duo quad Q6600 temps at Idle speeds
Core 2 duo quad Q6600 at full load


Fans at idle are whisper quiet ... at full load they are at max capacity ... which is still very quiet since I set it up with excellent air moving fans that are super low decibel.

Runs rock solid stable.

Abit IP 35 Pro
Gemini2 HS
Very thick diamond thermal grease blend of at least 50% diamond sub micron particles.
2 Yate loon 120mm Fans on the HS
1x 120mm fan front intake [filtered]
1x 120mm fan side intake [filtered]
1X 120mm fan exhaust
1x 120mm fan power supply exhaust

I am considering another run ... another re-install with another formulation of diamond thermal grease ... I think it might just be better ... I wasn't happy with the consistency, clumping and thickness of the thermal grease ... though I was very happy with how the diamond skin turned out.

[Elysium]

I'd say those are pretty good results on air! Nicely done! But now lets see some pics of the diamond thermal grease...

[Tech Geek Deluxe]

Corsair ... new favorite ram ... you can't believe how close the Heatsink is to this ram when mounted ... they actually touch!

Abit IP35 Pro Motherboard with some sweet CORSAIR Ram


Yes ... that is mostly diamond in a carbon base baked directly to the pipes. It actually doesn't look as amateurish as the pics might lead one to believe.

DIAMOND SKIN TEST


Nother pic ... please note ... I wasn't going for a lapped surface ... just 1200 Grit smooth.

DIAMOND SKIN 2

[Tech Geek Deluxe]

I didn't do pics of the grease itself ... it looks like super thick arctic silver 5 ...

It is too thick ... I think I can shave off 2-3 degrees if I can thin it out and add another heat transfer media. If not this CPU ... I have a Pentium D that runs too hot ... If I can get 1-2 degrees shaved off this one ... I'm sure I can get 2-3 degrees off the Pentium D.

That would leave me with a Core 2 Duo for a raw comparison test with a Lapped CPU and Heatsink and a perfect mount with Arctic silver 5 ... If I can score 2-3 degrees shaved off of that ... I think I might have the Holy Grail of thermal grease compound.

[Elysium]

Wow, great pics! Really helped me understand what you were doing.

[Tech Geek Deluxe]

Yeah ... I'm visual too ...

The one test I wanted to get definable results on was "feeling the heat" immediately after PC on ... hoping the heat would travel along the "diamond skin" faster and be felt immediately.

It may be doing that ... and transfering heat over a larger surface area to the copper ... Can't say.

The idea is theoretically sound. If you had a perfect 1" square rod of diamond that was perfectly insulated against heat loss and directly attached to the CPU Die with perfect heat transfer media ... and the rod was 500 feet long .... you would feel almost the exact heat of the cpu at the end of the rod almost immediately when the cpu was turned on.

Heat transfers through diamond differently than other media ... it transfers it like sound and at the speed of sound.

So .... I figured a High content media would provide "improved" results at a faster frequency and wider bandwidth.

Heat transfer moving away faster ... and by a wider bandwidth or surface area simply relieves the headrooom that the electron friction causes in CPU heat bottlenecks.

In reality ... if you remove heat from something faster than its source production and/or through a broader surface than it can produce ... you freeze it or extinguish its "fire."

[Tech Geek Deluxe]

Well ... I did another test ...

Put POS thermal grease on that did not "connect" directly to the "diamond skin" ... bad results

I only did an idle test and this is the quad idling ... ouch we are too hot

Ran hot ... I started a full load test and temps were starting to pass 65 degrees C [a full 10 degrees and rising over previous tests] ... so I shut her down ...

The very very good news is ... I re-did the diamond thermal grease making sure the "skin" contact and application was one of perfection ... and my full load temps dropped 10 degrees over four cores ... or an average of 2.5 Degrees C per core.

Here Quad 100% Load 2

Idle was pretty much the same.

I'm thinking that this is a win for diamond thermal grease ... I won't tell you the formula however ... not yet ... I do believe that the test for the "diamond skin" is the real winner tho.