TechPowerUp! Official IC Diamond Test

[sneekypeet]

All I have is what you have for the original image. As for the "now" image, I could change its position and lighting to make it look similar to the original.

Damage is not as visible from this angle...

 

[IC Diamond]

First after pic is angled so highlights sink imperfections and hides the prominent stains seen here.

I would keep in mind that machining the base is an intermediate step in the whole process, probably clamped for the soldering, in and out of production bins, probably several cleaning steps on and off assy and packing tables....you are going to get small micro dings out of the gate.

Pure copper has an MOHS of 3 - For comparison On the Mohs scale, graphite (a principal constituent of pencil "lead") has a hardness of 1.5; a fingernail, 2.2–2.5; a pocketknife 5.1

Copper is soft, barely harder than your fingernail add a little weight move it around, stack it etc. on any uneven surface and you will get micro dings especially in fragile raised areas like with machine tool cuts.

As noted generally they are camouflaged being small and same color as the background. now use the sink and get a micron or two of oxide or or dirt in the micro depression and now you see something that you did not see before with the new contrast.

a better factory fresh picture with better resolution would provide more points of comparison with any luck I can line up 10 to 20 with the current sample.

 

[sneekypeet]

I already told you I only have the one resolution. Explain away all you want, but your TIM did the damage to this base. If I read your statement correctly, you say that the diamond is too hard for copper bases and will most certainly damage any copper base it is used on.

Like I said man, I'm just presenting the facts, you can do with them what you will, but these "scripted" explanations are holding no weight with me at the moment.

I know the base was fine before applying your product...fact.
I tested the product...fact
I removed the cooler and cleaned off the TIM to find pits and damage to my base....fact
Watching you scramble with technical babble about how you try to just pass it off....priceless

At this point I am having more fun watching you squirm than actually bothering to think I would get some logical support from my findings with your product

 

[DOM]

So wuts up with the smell ? And it also left a stain on my copper wb and made the letters on the ihs lighter....

Some said about the smell and didn't see any response...

 

[IC Diamond]

Happy New Year

sneekypeet - Thanks for the pics.

The smoking gun is here. You can clearly see the pits in the tool marks in the left upper quadrant in the contact area of the CPU.

Outside the contact area you find the same type of pits on the tool marks less contrasted without the oxide or grease in the groove between the cuts clean copper scoring against copper would be hard to spot with the naked eye but there they are.

Now as this scoring is outside the contact area and not subject to pressure there could be no crunching as the say to produce identical size type marks so looks to be a case of mistaken identity as to the source

As further confirmation you can send me your sink on my UPS account and I will clean it up and put it under the scope to provide clearer pictures and measure the size and depth of the scoring.

Also send me the sample tube of paste as we can wash it to a powder and put that under the scope and measure the particle size although after looking at this picture I know that finding particles 20X larger than spec is extremely unlikely considering the result or alternatively I could send it to a materials lab for independent testing there is one right up the road from me, either way not a big deal they will just wash it and scope it also.

As noted Pure copper has an MOHS of 3 - For comparison On the Mohs scale, a fingernail, 2.2–2.5

Machining the base is an intermediate step as clamping the base after full assembly is to awkward, sink was clamped for soldering, I see screws so there was one assembly step at least, a cleaning step to remove machine oil, another to remove flux from soldering, packaging & a lot of physical handling went on between steps, in and out of production bin or carts and unto tables.

This was a mass production item in China not fine jewelry from Tiffany's New York.

So against a back drop of mass production you have a base with tool marks I'll estimate at .001 in height and width made out of soft copper it would take minimal effort to score a base tool mark, plenty of grit in a production environment and plenty of opportunities in movement from station to station. This is consistent as the scoring is uniform across the base and same/type as the contact area

In any event I will PM my UPS account number, send me the sink and sample tube and I will run the confirmation tests

 

[AphexDreamer]

So how many people are saying their heat sinks took damage from the compound? Is it only certain heatsinks? I've yet to check mine.

 

[TRWOV]

I didn't see damage on the heatsink other than some stains (got rid of them with lemon juice) but my CPU showed a pit and light scratches:

Before the test:

The scratches I can understand but I don't know how the pit formed.

I was using IC7 before the sample arrived and after that I tested with Cooler Master's standard compount, MX-4 and then IC24. IC says that's some sort of contamination and maybe it was because of all the TIM shuffling but I'm still baffled on how that pit formed

 

[IC Diamond]

I didn't see damage on the heatsink other than some stains (got rid of them with lemon juice)

lol, I will file that advice and use it

If you are concerned you can send me your sample and we can wash it and measure the particles

 

[RejZoR]

I'm thinking of ordering some of this stuff if it's really up ti 7°C better. I'll just have to check Amazon or eBay or somewhere else as no one is selling this in my country But then i'm a bit afraid that i'll get the actual thing (at least for eBay)...

As for the damage, if you add diamond particles between two metals, it's logical that it will scratch it one way or another, even if they are all at specified 40 microns size. If there are so big particles to make dents half a milimeter big, then there has to be massive chunks of diamonds which you'd see during application of thermal compound imo. So it had to be something else.

 

[TRWOV]

It really gives results, no doubt. In my case it got about 1-2°C less than MX-4 but needs several hot-cold cycles to get to its optimal performance. Idle temps are nothing to write home about, load temps are stellar.

 

[sneekypeet]

I'm thinking of ordering some of this stuff if it's really up ti 7°C better. I'll just have to check Amazon or eBay or somewhere else as no one is selling this in my country But then i'm a bit afraid that i'll get the actual thing (at least for eBay)...

As for the damage, if you add diamond particles between two metals, it's logical that it will scratch it one way or another, even if they are all at specified 40 microns size. If there are so big particles to make dents half a milimeter big, then there has to be massive chunks of diamonds which you'd see during application of thermal compound imo. So it had to be something else.

If you are supposed to apply a blob in the center of the cpu, explain to me how I would see chunks again?

 

[TRWOV]

There wasn't a chunk of diamond in the pit if that's what you're wondering. I think it's some kind of corrosion but it'll be nice if IC could replicate it and tell us how to avoid. Maybe offer a cleaning compound along with the TIM?

 

[IC Diamond]

I'm thinking of ordering some of this stuff if it's really up ti 7°C better. I'll just have to check Amazon or eBay or somewhere else as no one is selling this in my country But then i'm a bit afraid that i'll get the actual thing (at least for eBay)...

As for the damage, if you add diamond particles between two metals, it's logical that it will scratch it one way or another, even if they are all at specified 40 microns size. If there are so big particles to make dents half a milimeter big, then there has to be massive chunks of diamonds which you'd see during application of thermal compound imo. So it had to be something else.

Plenty of sample tubes left -

 

[RejZoR]

So, i only have to submit current temperature vs ICD temperature in order to apply for the sample?

Which version are you sending? ICD7 or ICD24 ? Also what about postage costs if any?

 

[ALMOSTunseen]

So, i only have to submit current temperature vs ICD temperature in order to apply for the sample?

Which version are you sending? ICD7 or ICD24 ? Also what about postage costs if any?

ICD24. Shipping is free.

 

[RejZoR]

I'll contact IC Diamond representative for the sample. I wonder how much i'll gain with my ANTEC H2O 920...

 

[Random Murderer]

I didn't see damage on the heatsink other than some stains (got rid of them with lemon juice)

Good to know. Just be careful with it, lemon juice is highly conductive.



Still no word on what causes the rancid smell? IC Diamond, I'm looking at you >.>

 

[IC Diamond]

Good to know. Just be careful with it, lemon juice is highly conductive.



Still no word on what causes the rancid smell? IC Diamond, I'm looking at you >.>

subjective to the individual, I can barely smell anything, much less than any household cleaning product- 92% diamond couple % polymeric binder and carbon black and a perhaps a drop of synthetic oil per application so the only thing would be the oil.

altogether pretty inert stuff

 

[cadaveca]

subjective to the individual, I can barely smell anything, much less than any household cleaning product- 92% diamond couple % polymeric binder and carbon black and a perhaps a drop of synthetic oil per application so the only thing would be the oil.

altogether pretty inert stuff

I don't think it's "large particles" of diamond causing the pitting...It's a chemical reaction.


Perhaps a batch of rancid oil used? This might explain the smell, and why some users have issues?


Honestly, to me, the pitting found by some users seem like some chemical etching. Is it possible some left-overs from past paste applications is causing a chemical reaction?

If you look at Sneekypeet's picture, sure, there is a bit of minor pitting there, but where the base of the cooler made contact with the chip, there is far greater pitting. Personally, I'd think that the constant heating and cooling cycles is causing the oil used to become rancid, accelerated possibly by some sort of contamination.

 

[fullinfusion]

I've been using IC Diamond for a few years now and cant say I have ever seen any scratches or have smelled any thing that others are complaining about.

I did ask for a tube of this test compound but never got anything and it's been months now.

From what I see everyone asked for a test tube and now some are they're bitching?!
I don't think anyone held a gun to there heads to use this stuff. It was FOR TESTING and like anything else in the world nothing is perfect,,, It's called SHIT HAPPENS so suck it up butter cup!

God it just blows my mind reading some of the posts here and seeing all the bitching and complaining on things... So ungrateful are many ppl.

I pulled off my water block which is a EK-Supreme HF EN Full Nickel block, I seen a slight I mean only at a certain angle in the sun light can you see a stain but just like any other metal it WILL stain! The cpu has more of a stain but who cares? It's not hurting anything.
Metal on metal and different metal on metal will cause some discoloration, its a fact! Be it water blocks, car engines you name it!

I have a simple solution to all the moaning in here. Just don't use the product. Plain and simple! Thank the rep for the sample and move on. Just be grateful your alive and not 6 ft under...

 

[theonedub]

Well I thought I posted my results here, but I may not have. I did get a chance to pull my CPU and look for damage, etc and did not see any of the scratching people have mentioned. The paste did severely stain my block (which will need to be sanded) and the paste did smell really bad.

Basically the paste performed about 2C cooler than what I had been using before.

Spoiler: Results

 

[fullinfusion]

Well I thought I posted my results here, but I may not have. I did get a chance to pull my CPU and look for damage, etc and did not see any of the scratching people have mentioned. The paste did severely stain my block (which will need to be sanded) and the paste did smell really bad.

Basically the paste performed about 2C cooler than what I had been using before.

Spoiler: Results

http://img.techpowerup.org/130102/ICDiamond_43_LoadHigh.jpg

Why would you need to sand the block? Your using an H100 correct? It has the worst smoothest surface in a block I've ever seen.

You sand that block and your going to loose cooling effect... What is the stain going to do?
The smoothest surface still has micro pores that you cant see unless under a microscope... Thats what you see as a stain and where it is coming from... If I were you Id leave well enough alone

Thermal paste is made to fill in micro imperfections and pretty much that's it.

 

[IC Diamond]

I don't think it's "large particles" of diamond causing the pitting...It's a chemical reaction.


Perhaps a batch of rancid oil used? This might explain the smell, and why some users have issues?


Honestly, to me, the pitting found by some users seem like some chemical etching. Is it possible some left-overs from past paste applications is causing a chemical reaction?

If you look at Sneekypeet's picture, sure, there is a bit of minor pitting there, but where the base of the cooler made contact with the chip, there is far greater pitting. Personally, I'd think that the constant heating and cooling cycles is causing the oil used to become rancid, accelerated possibly by some sort of contamination.

You only can get rancid out of an organic that contains fats and lipids, this is synthetic and contains none. This has essentially the same base as a motor oil no difference and is pretty inert just not much it is going to react to.

As to acid you might see that in a car engine but it is a by product gas combustion and accumulates over time and can separated or filtered out but does cause the oil itself to be acidic.

Having just talked to a chemist at the company that produces the oil he tells me that there is no known combination of chemicals that could be introduced that would turn the base oil acidic

Actually you can find the damage in the fresh from factory pic posted, I did not post it as the connection was tenuous but for me provided a clue as to source to look for.

I can only say that in my experience running on test setups with pure copper test dies and sinks from 30W to 200W in time frames from several hours to several months I have never experienced any pitting or scratching on mirror finish contact surfaces.

I enlarged the original and changed brightness and contrast to highlight imperfections

To test whether they were not dust on the camera lens or an artifact from the computer I then used the measuring tool in Gimp on the similarly sized pics I located one point then located the others at the appox. same angle and distance on the second picture from the original selected point and found 12 or 13 out of 18 that were on the original factory fresh sink. I am not counting the ones marked red and counting the one blue.

Better resolution on the factory fresh & I might have pulled out more but as it was I assumed due to the resolution only the grossest of imperfections would show up and so could be found on the second pic. There is some offset due to perspective angles of approx 20 px on the horizontal and some on the vertical but some good matches in the group.

But as I mentioned this exercise was more of a clue search and the other higher resolution has enough to indicate likely source and we are continuing to look into it.

 

[Random Murderer]

I've been using IC Diamond for a few years now and cant say I have ever seen any scratches or have smelled any thing that others are complaining about.

I did ask for a tube of this test compound but never got anything and it's been months now.

From what I see everyone asked for a test tube and now some are they're bitching?!
I don't think anyone held a gun to there heads to use this stuff. It was FOR TESTING and like anything else in the world nothing is perfect,,, It's called SHIT HAPPENS so suck it up butter cup!

God it just blows my mind reading some of the posts here and seeing all the bitching and complaining on things... So ungrateful are many ppl.

I pulled off my water block which is a EK-Supreme HF EN Full Nickel block, I seen a slight I mean only at a certain angle in the sun light can you see a stain but just like any other metal it WILL stain! The cpu has more of a stain but who cares? It's not hurting anything.
Metal on metal and different metal on metal will cause some discoloration, its a fact! Be it water blocks, car engines you name it!

I have a simple solution to all the moaning in here. Just don't use the product. Plain and simple! Thank the rep for the sample and move on. Just be grateful your alive and not 6 ft under...

There's a big difference between bitching about a free product and being angry about damage from a test product. It's obvious some of these tubes should not have passed QC.
And as far as the rep on here trying to say the damage is caused by us contaminating the TIM, that's absolute bullshit, and no, I won't be nice about it. Awfully strange that the only compounds anybody has ever seen this type of damage from are ICD7 and ICD24. Sure, Shihatsu or whatever it's called caused a chemical reaction and caused pitting, but that was actually eating away the metal. This is completely different from that and is actual physical damage, as if something ground into the copper.
Staining? I can deal with that. Dead fish smell when removing the heatsink? I can deal with that, too. Very fine scratches/polishing? Not only can I deal with that, it's to be expected from tiny diamonds and I actually wish all compounds did that, lol. Pitting/gouging from oversized particles and/or manufacturing oversights? COMPLETELY UNACCEPTABLE, and the fact that the company representative is blaming us compounds the issue.
That being said, I'm not bitching in the least. For me, the product has worked exactly as expected and done a damn good job. In fact, I recently swapped from an air cooler to this H100i. Guess which TIM I'm using? ICD24, and I will continue to use it until I run out or have an issue.

 

[cadaveca]

~snip~

What's curious to me is that the "new" pitting shows the copper of the base through quite obviously, while the "pre-pasting" pic does not.

Maybe plating is being removed somehow?


As mentioned before, I did not get a recent test sample, so my interest in this is purely as an intellectual exercise. It's quite curious, to me.

I have a Noctua cooler here, have used it for a couple of years now nearly in my review testing, which means multiple mounts per week, every week. I've used every paste on the market at this point, since actual cooling means little to me testing, and is more for checking fitment issues, but that said, my heatsink has a similar base to the original picture posted by Sneekypeet. I have now sent this cooler to another staff member here. He has a sample tube from you, I think, and I'll ask him to try the paste out and see what happens. Maybe it's some issue with the plating.



I just want to know what's causing the issue..specifically...the idea that it's contamination of some sort, doesn't help prevent the same from being repeated by other users in the future. Properly identifying exactly what causes the problem will go a long way to build consumer confidence, IMHO.