TechPowerup IC Diamond Results Thread

[IC Diamond]

PLEASE READ THOROUGHLY BEFORE TESTING/POSTING YOUR IC DIAMOND OR PERIHELION TEST RESULTS!!!



PLEASE ONLY POST WITH RESULTS WHEN YOU HAVE FULL RESULTS, OTHERWISE YOUR POST MAY NOT BE SEEN BY ALL IF YOU RESERVE THE POST FOR LATER RESULTS WITH IC7. PEOPLE TEND TO GO TO THE NEWEST POST, SO IF YOU WANT IT TO BE SEEN BY ALL, POST YOUR RESULTS ONCE YOU HAVE THEM ALL. ALL OTHER DISCUSSION/SUPPORT POSTS, GO RIGHT AHEAD!!

PLEASE BE SURE TO BENCHMARK YOUR CPU TEMPS WITH YOUR CURRENT TIM AND RECORD YOUR RESULTS BEFORE USING IC7 FOR TESTING. WE NEED BEFORE AND AFTER PLEASE!

ALSO, PLEASE MAKE SURE YOU DON'T CHANGE ANY VARIABLE OTHER THAN THE THERMAL COMPOUND USED FOR BEFORE/AFTER (I.E. SAME VOLTS/CLOCKS, COOLER, AND OTHER HARDWARE)


APPLICATION METHOD


For all those participating in the ICD giveaway, please follow this link to the Innovation Cooling website for more information on the products:

Innovation Cooling

Proper application is the key to this product. The nature and consistency of this product is likely quite different from what most of you are used to in a Thermal Compound. We don't recommend use of traditional methods of applying the TIM like razor blades, lines, Rice size, etc. ARE ALL Big BIG NO NO's. Through extensive end user testing, the best overall application method has been determined to be a pea sized amount directly in the center of the CPU. The TIM should be adequately spread via the down force of the CPU cooler (a sufficient amount of down force will provide best results; minimum 50psi of down force should be the goal for optimum results. please see the Innovation Cooling website for more info). Also, for coolers/surfaces that have many peaks and valleys, it may be beneficial to wet the surface(s) with a small amount of ICD to initially fill the voids in the mounting surfaces...then proceed with recommended application as usual. Please click on the link below for application methodology and shoot me a PM if you have any questions.
Keep in mind, ICD has been formulated with the specific goals of providing superior thermal conductivity with short set times, but also for long term stability and performance without pump/bake out through extended thermal cycling or usage. This stuff is very stable over long periods of time.



Please do feel free to navigate through the whole Innovation Cooling website as there is tonnes of information on there that might prove beneficial.


HOW TO POST RESULTS


THE FORMAT IS AS FOLLOWS (ALL TEMPS IN C):

OPTIONAL INFO: PROCESSOR, VOLTS/CLOCKS, HEAT SINK/COOLING USED, ETC.

Before Compound:
Ambient Temp:
Idle Temp:
Load Temp:

ICD
Ambient Temp:
Idle Temp:
Load Temp:

Cure Time- Cure Time averages 2 hours in most situations for sinks with adequate pressure loading with maybe .5C more overnight. Lightly pressure loaded sinks may take a day or more depending on the amount of pressure


If you have any problems or need support, please feel free to post here for now. Remember, proper application and good even heat sink pressure (goal should be minimum of 50psi or more) are crucial for peak performance. Too much or too little compound will impact your results so please do visit the Innovation Cooling website for Application Instructions with images and explanations. For those with direct touch heatpipe coolers, you might want to try wetting the contact surface of the cooler with a tiny amount of ICD to fill any small voids before proceeding with the normal application method.

If anyone has any questions, please feel free to post up. .


Contact/ Pressure Paper Instructions

The C/P is sourced from a company called Sensor Products and is the Ultra Low (LLLW) with a rage of
28 - 85 PSI (2 - 6 kg/cm²).

To apply (do not have paste installed, this done with a clean sink and IHS) Take one "A" & one "B" from the key envelopes and place shiny (uncoated) sides to metal - one to the sink shiny side and one to the IHS shiny sides to metal.

Mount your sink as you normally would then dismount and remove the paper - one will have a pink impression showing the areas of contact and the amount of pressure indicated by the density of the colour.

I will post a colour chart as soon as I find it on their site so you can estimate within about 10% your total pressure.

http://www.sensorprod.com/prescale/p...ructions-2.pdf

[OneMoar]

AC5 12H cure
g AC5 12H

IC Diamond No cure Center dot
IC Dimond No cure package-sensor
I had a fair bit of difficultly with the application(it has a thick consistency that I would compair to JB-weld) I finally had to resort to the single center dot method not my preferred method that said even with my less then stellar application temps are down 3c on Avg call me impressed considering it has not had ANY cure-time

[VulkanBros]

And the benchmark/testing software we should use? OCCT? LinX? or?

[crazyeyesreaper]

Testing using LinX and Wprime, with the Former being worst case senario temps and the latter being closure to real world load temps

3770K MX-4 STOCK

Ambient: 24c
Idle: 35c
WPrime Load:57c
Linpack Load:65c

Spoiler

MX-4 Overclocked @ 4.6GHz

Ambient: 24c
Idle:36c
WPrime Load:74c
Linpack Load:90c

Spoiler

IC Diamond STOCK

Ambient: 24c
Idle:33c
WPrime Load:53c
Linpack Load:61c

Spoiler

IC Diamond Overclocked @ 4.6 GHz

Ambient: 24c
Idle:32c
WPrime Load:69c
Linpack Load:82c

Spoiler

Pressure Test

Spoiler

MX-4 vs IC Diamond
STOCK
Idle: 35c vs 33c = Temp Drop of 2c
Wprime Load: 57c vs 53c = Temp Drop of 4c
IBT Load: 65c vs 61c = Temp Drop of 4c

MX-4 vs IC Diamond
OVERCLOCKED
Idle: 36c vs 32c = Temp Drop of 4c
Wprime Load: 74c vs 69c = Temp Drop of 5c
IBT Load: 90c vs 82c = Temp Drop of 8c

So all told depending on the situation and test the IC Diamond Thermal paste dropped temps by 2-4'c at stock and by a massive 4-8c at load

[IC Diamond]

Note C/P instructions added to the intro.

Contact/ Pressure Paper Instructions

The C/P is sourced from a company called Sensor Products and is the Ultra Low (LLLW) with a rage of
28 - 85 PSI (2 - 6 kg/cm²).

To apply (do not have paste installed, this done with a clean sink and IHS) Take one "A" & one "B" from the key envelopes and place shiny (uncoated) sides to metal - one to the sink shiny side and one to the IHS shiny sides to metal.

Mount your sink as you normally would then dismount and remove the paper - one will have a pink impression showing the areas of contact and the amount of pressure indicated by the density of the colour.

I will post a colour chart as soon as I find it on their site so you can estimate within about 10% your total pressure.

[hat]

Hmm, unfortunately my results don't look so good. I followed the application method listed in the OP (in fact that's the same method I use all the time anyway). Thermometer in my room read 78f during both tests. I ran my i7 920 at 3.8GHz 1.175v for both tests.

First, the MX-2:




And now the new IC Diamond paste:




Looks like the IC paste actually raised my temps by 2-3*c. Not sure why this is. I thought maybe I used too much paste the first time, especially considering I lapped my processor, so I took the h70 block off, cleaned it up with 91% rubbing alcohol and a coffee filter like I always do, and re-applied paste, this time using a little less. The paste seemed really thick so I thought I'd leave it run overnight and give it some time to spread out and settle in.

[IC Diamond]

You Might try the C/P paper for trouble shooting

I thought this was an Interesting breakout of the numbers Delta' between compounds on the Water Cooling shrink considerably about 2C on the MX2 and MX3 and 2.75C on the MX4.

The only one cooler that bucked the trend was the Corsair and perhaps the EK block with better mounting schemes.




Below is a data sort of the test results The blue bars are air cooling all others are water cooling.

The Corsair results are marked in green and follow the usual distribution

[OneMoar]

I had better luck warming the tube and thinning the paste with rubbing alcohol gonna give it some time to cure and then ill post more screenys
if be blut stuff is great but you really need to get it thinner nobody's heatsink ever goes on with even pressure especially the big heavy aftermarket beasts they tend to wanna twist and pull when you install them stuff is basically plastic at room temp I had to heat it to 100f to get proper air-free coverage that said the performance is impressive once you get it applied ....

[hat]

Care to elaborateness on that process onemoar?

[OneMoar]

I just put the tube in some hotwater for 5-10min and then mixed a few drops of rubbing alcohol into about a pea-sized blot and then placed 5 short-grain rice dots on the cpu and applyed the head-sink work quickly or it will stiffen right up

[stinger608]

@OneMoar: I don't believe that "thinning" the ICD24 is going to give you correct readings. If the TIM was suppose to be thinned out, it would have come thinned out.

[OneMoar]

Originally Posted by stinger608

@OneMoar: I don't believe that "thinning" the ICD24 is going to give you correct readings. If the TIM was suppose to be thinned out, it would have come thinned out.

alcohol evaporates and any moisture left will also evaportes within a matter of seconds after system power on leaving nothing but paste I have used this method before with other-pastes has no impact on the performance of the paste

[IC Diamond]

Originally Posted by OneMoar

I just put the tube in some hotwater for 5-10min and then mixed a few drops of rubbing alcohol into about a pea-sized blot and then placed 5 short-grain rice dots on the cpu and applyed the head-sink work quickly or it will stiffen right up

On the larger IHS's over 30 X 30 mm people have been reporting better results with the 5 dot method - A 5.5mm bead on center with four dots 1/3 towards the corner edges. We are in process of updating application procedures with a video.

With adequate pressure of about 50PSI (which encompasses 80% of those sampled in the forums0 a 5.5mm bead on centre should work best.

If you are unsure about the 5.5mm bead you can peel back the label and use the gradations on the syringe which works out to .15- .2 ML

On the version we sell to the OEM's we add an acetate to thin for stencilling, silkscreen or ink roller application so ISP is fine just make sure you give it enough time to dry (10 min) or you will get EXPLOSIVE VOIDS for fast evaporation of the solvent.

Application Notes -Amount of Applied Compound
*
*
There is always some debate on amount of applied compound to use, our approach is based on a best to worst case scenario with a one size fits all so the user can experience best odds of a successful mount out of the gate without having to redo his mount.
*
*
Say that Intel has a spec of Flat and parallel // to .002. - At the extreme end you get something like the 2 images of the candidate below, to fill that gap/volume you need .05 ml volume of compound for a 30mmk X30mm area about the size of a grain of rice.
*

*
Our difference with the rice method comes from the other side of the equation [B]The heat sink base, while many HS bases are good some can be off as much as the IHS and to cover the contingency of irregular contour on both sink & IHS and as few people lap to correct the problem (5%?)and fewer lap to perfection so we add another .05 ml for a total of 0.1ml.
*
Another factor associated with irregular contours is another mil or two layer of compound is added on the high spots depending on the pressure applied. To fill that gap we add another .05 ml for a total recommended amount of 1.5 ml which approx. equals our 5-5.5mm size bead on center.
*
Thermal compound is a volume application to fill gaps, not one of weight, liquid spread-ability etc. and

[IC Diamond]

Try the Supplied Contact and Pressure Paper

[OneMoar]

yea ill need to try the contact paper when I get my new psu for my 2500k rig this Socket Am2+ board is beat and probably a little warped and yea I should note you need to give it a good 30 seconds for the ISP to evaporate if you are brave you can power the machine on for a few seconds to speed up teh process

[hat]

Hmm, it seems like my temps may have actually went up after lapping the h70. The contact paper showed high pressure contact around the edges of the cpu ihs and virtually no contact in the center. After lapping there's more contact coming in from the edges, but still hardly any contact in the center... but there's still more contact so temps should be better. Maybe too much paste, especially since both surfaces should be flat now. Maybe it needs curing time as well.

Don't understand how the contact paper would be showing no contact in the center. There were scratches all across the ihs base and now there's only fine marks from the sandpaper (up to 1000 grit) on the base. The CPU has got to be flat since all the nikel is gone... nothing but copper all across the ihs. Hell I even drew an X across the ihs and it was all gone save the very edges were the ihs tapers off within not even a minute of sanding.

[chevy350]

I've re-ran Prime95 for 20 minutes with ambient room temp of 24C. Idle temps on cpu are around 25-27C and after 20 minutes of prime 71C was highest core temp. CPU is running 4.5GHz@1.36v with an Antec Kuhler 620 and 2 Scythe Kaze 3000RPM 120x38 fan's in push-pull


Heading to the garage now to tear down, check contact and apply the sample. Will update results once I get back up and running and then again in a week or so if anything changes

Here's the results after applying the sample, left the dust (little that there was) so it would be the same conditions as the before test run......so here's what temps look like now, Idle is 27-28C (same) Load at 20 mins of prime highest core was 66C ambient room temp is 24.5C. I will do another run in a week or so and see if it gets any better or stays the same....hopefully better



Getting ready to do the 6950's.....ran the burn-in 1080 test with highs at 74c and 67c

[IC Diamond]

Originally Posted by hat

Hmm, it seems like my temps may have actually went up after lapping the h70. The contact paper showed high pressure contact around the edges of the cpu ihs and virtually no contact in the center. After lapping there's more contact coming in from the edges, but still hardly any contact in the center... but there's still more contact so temps should be better. Maybe too much paste, especially since both surfaces should be flat now. Maybe it needs curing time as well.

Don't understand how the contact paper would be showing no contact in the center. There were scratches all across the ihs base and now there's only fine marks from the sandpaper (up to 1000 grit) on the base. The CPU has got to be flat since all the nikel is gone... nothing but copper all across the ihs. Hell I even drew an X across the ihs and it was all gone save the very edges were the ihs tapers off within not even a minute of sanding.

For those that have the knack for lapping you can get great results, for people like me who lack the art of it I was surprised how bad my efforts were when I started checking it with the C/P paper. I now resort to my trusty belt sander for flatness, the edges end up a little rounded but at least flat within .001.

I do a final dressing with a set of Diamond Stones (DMT)

Check out the contact images below, 3 of them are lapped results guess which ones they are? Hint, 1A was a water block from our 2nd giveaway and was not lapped. I have only seen 4 or 5 lapping exercises that were in the near perfect category out of 50 examples.

Note - You can not put to much compound on, we have tested this with a synthetic copper IHS applied with IC Diamond and clamped with binder clips (60PSI) between 2 glass slides.

We then mic'ed the assembly at one min. intervals at room temperature (20C). The baseline for comparison is the average particle size. After 5 min the bond line thickness (BLT) to within 95% of the average particle size, overnight to 98%. Excess just squishes out.

Better too much than too little - too little is more of a waste as you end up redoing the mount and run into the problem of laplacian growth entrapping air into the final joint

[OneMoar]

tried the contract paper mine looks closer to sample 32 but with more pressure in the center

[Steevo]

I tried a few different compounds and the issues I ran into with the water loop was it never got warm enough to allow the phase change to complete occur and I always had too thick of a layer left. Trying AS5 and some WD40 mixed to thin it out gave me some good temps but it ran out and never set, so I just started using ceramique and the 3C temp difference was not enough to worry about. On the GPU it actually got warm enough once I pinched the tubing to set completely, but I still ended up on the last cleanout running ceramique on it too.

What is the phase change temp of this thermal compound?

[hat]

My contact looks like 19a

Can't see why though... I know my cpu ihs has to be flat because I drew an X on it and sanded it off in under a minute as I mentioned before. The block has got to be flat too as I got all the milling marks out and now it's just marks from the sandpaper, up to 1000 grit. Maybe it's because of the weird mounting the h70 uses?

[popswala]

Just wanted to throw this out there, That is some stiff tim. I tried to work with it a lilttle just to see and man its like old gum lol.

System Used:
AMD Phenom X4 945 3.1GHz
Corsair H50 Cooler
Gigabyte 990FXA-UD7
Thermaltake Chaser MK-1

Heres all my results:

Test 1
TIM - AS5





Test 2
TIM - IC Diamond




Pressure Test:


TIM Spread:



As you can see with this mount for the H50 is still kinda tricky for me to work with and to use the pressure paper. I took everything out of my way to try and accomplish that lol.

I hope all this is thorough enough. If there's any other test or anything else to do/try. Feel free to just say.

I wanted o give as much info to help aid in the process/data of the project.

Thanks IC Diamond for this opportunity.

[hat]

? Almost a 10c increase? Did you clean the old paste off first? Looks like there's some old gunk left on that chip.

[popswala]

I did a complete clean up of both (spu and block). The TIM spread I did was just after I dismounted it, cleaned old AS5 TIM off and did the pea size bead and reseated it, Then poped it off to see how the spread was. Re-cleaned it up and put fresh TIM back on and re-seated block to do test. I fully mounted it and all 4 screws tightened. It was completely clean. Its just has a stain to it. That chips been around lol.

[hat]

Hm, so I'm not the only one seeing an increase...