IC Diamond 7 Carat Thermal Compound Giveaway

[IC Diamond 7]

Originally Posted by screwballl

you have ghas? was wondering what that smell was


------------

and aziani and garet.... I know I am by no means and expert but in my limited testing, if you take the grain of rice size and stretch it from one end of the IHS to the other, you end up with almost the same results. the problem with going all the way from one side to the other is you almost always get some that spills over off the edge making it that much less that makes contact with the heatsink and slightly reducing the effectiveness.
Also with my mention of pea size for the IC7, I have one or two SocketA's that I will be testing it with and a PEA will cover the entire core and spill over onto everything else...

You are dealing with two components here.

1. Bond Line thickness

2. Bulk thermal conductivity

A "wetter" grease with less bulk and more liquid will spread and cover a larger area and give you a thin BLT but the thermal conductivity may be low. That kind of paste gets it's performance due to its relatively thin thickness

A high bulk paste can have twice the BLT, so with say 2x conductivity both perform the same.

The best of all worlds then is great conductivity and a thin bond line which is what we tried to do here with ICD7.

A quality paste like a couple of the ShinEtsu's are high bulk loaded and are intended to last years. Some of the retail grades which are more spreadable or "wetter" Give you a kind of instant gratification in performance but quickly degrade, some in just a couple of months due to pump out or bake out.

We tried thinner and thicker mixes. The thin ones were lower performance as we lost a couple of degrees and were subject to pump out. The really thick ones boosted the conductivity from 4.5Wmk to 10Wmk the advantage was only .5 C @ 100 W. some people think this stuff was thick, that mix took an unbelievable amount of dedication to use.

Use common sense. Put some on, half mount the sink, take it off check to see if the coverage is good and if it is mount the sink.

In a higher heat flux density with a bare die vs IHS the spread or result should double over what you would see on an IHS.

Again enough pressure to hit the proper BLT is critical to optimizing performance.

The test result you get, good, bad or indifferent is what it is and is fine with us. What we are looking for is a picture of the overall group so don't agonize too much just follow the general guidelines and post ambient, before and after result so we have enough to work with.

[screwballl]

thank you for the update, we will do some digging and when it comes available, we will post the results with as much before and after info as possible. My first tests will be on a SocketA and then a Core2Duo setup so there will be at least 2 setups... and if i can find a heatsink and fan for my FX5200 video card, I will use it for that as well. Although in my case I will not have a before and after result for the FX5200 since the current heatsink fan is dead.

[RangerXLT8]

Originally Posted by IC Diamond 7

You are dealing with two components here.

1. Bond Line thickness

2. Bulk thermal conductivity

A "wetter" grease with less bulk and more liquid will spread and cover a larger area and give you a thin BLT but the thermal conductivity may be low. That kind of paste gets it's performance due to its relatively thin thickness

A high bulk paste can have twice the BLT, so with say 2x conductivity both perform the same.

The best of all worlds then is great conductivity and a thin bond line which is what we tried to do here with ICD7.

A quality paste like a couple of the ShinEtsu's are high bulk loaded and are intended to last years. Some of the retail grades which are more spreadable or "wetter" Give you a kind of instant gratification in performance but quickly degrade, some in just a couple of months due to pump out or bake out.

We tried thinner and thicker mixes. The thin ones were lower performance as we lost a couple of degrees and were subject to pump out. The really thick ones boosted the conductivity from 4.5Wmk to 10Wmk the advantage was only .5 C @ 100 W. some people think this stuff was thick, that mix took an unbelievable amount of dedication to use.

Use common sense. Put some on, half mount the sink, take it off check to see if the coverage is good and if it is mount the sink.

In a higher heat flux density with a bare die vs IHS the spread or result should double over what you would see on an IHS.

Again enough pressure to hit the proper BLT is critical to optimizing performance.

The test result you get, good, bad or indifferent is what it is and is fine with us. What we are looking for is a picture of the overall group so don't agonize too much just follow the general guidelines and post ambient, before and after result so we have enough to work with.

I think what happened with that AS5 I was talking about is "bake out".
When I get my tube I am testing it on my Xeon quad core, 2900XT and my dads E6550. I have my DMM with the temp probe all ready to go.

[polobunny]

A question for IC Diamond 7;

Just how much pressure is necessary and how should one calculate how much pressure he should put to get the best results?

It's fine to talk about how the B.L.T. raises as you put more pressure, but with no way (that I know) for common people to calculate how much pressure there is by tightening their mounting mechanism, how do we know we're doing it right?

Thanks.

[theburan]

Hmm, I'm going to have to go hunting for some temp monitoring software that runs with ubuntu... other wise I'm goign to have to rely on just the bios monitoring for testing my socket a computer... or install xp on that computer.

[IC Diamond 7]

Originally Posted by RangerXLT8

I think what happened with that AS5 I was talking about is "bake out".
When I get my tube I am testing it on my Xeon quad core, 2900XT and my dads E6550. I have my DMM with the temp probe all ready to go.

A couple of the more exaggerated improvements that we saw on the Overclockers.com promotion, I think were in some part due to comparing an old install vs. a new app of ICD7 where some degradation had occurred .I have an ICD7 installation in the office that has been in place for 14 months with no degradation. We designed for stability as well as performance

Electronics-cooling.com I think has some pictures of bake out you might find if you do a search

[Beetle Bailey]

I think I need to get out of the woodwork and get back to posting again. I would like to get my hands on this for my 6000+. But I still need to get a better HSF.

[IC Diamond 7]

Originally Posted by polobunny

A question for IC Diamond 7;

Just how much pressure is necessary and how should one calculate how much pressure he should put to get the best results?

It's fine to talk about how the B.L.T. raises as you put more pressure, but with no way (that I know) for common people to calculate how much pressure there is by tightening their mounting mechanism, how do we know we're doing it right?

Thanks.

Great question, and I do not have a great answer. It is impossible for me to troubleshoot installations in this format so I can just throw out a couple of things to be mindful of, getting enough paste on and watch your mounting pressure.

There is a wide variety of spring mounting hardware out there some good, some marginal. The marginal ones will affect the performance of both TIM's differently. So it is in your general interest and not ICD7 specific for optimization of performance. Might make a good article for some motivated reviewer to buy some mix of springs 5, 10, 20 30lb springs to determine optimum pressures. Might even be able to make a business out of selling replacement hardware.

I would look for a snug fit that won't move on mount. Otherwise I leave it up to your motivation. Most geeks that I have run into on the forums are generally a bright lot to begin with, Just pointing it out will double check their mount.

It's your call If you think it's OK, then run with and post whatever result you get. If you think the mount is a little light but does not make a difference to you that's OK. If it really bothers you and you want to fix it go ahead and buy or scavenge some heavier springs, shim it and see whether it makes a difference. Up to you

One option as a simple test with a tower sink would be to lay it (case) on it's side so the sink is in a vertical position to prevent the weight lifting one side.

This is just a product promotion, maybe I am throwing to much detail but there are some differences in product I like to point out initially but other than that just post whatever result you get and do not worry, just be ready for the final exam at the end of this exercise

[garetjax]

If nothing else, this thread should be stickied for those people who are looking for more information about TIM and the application of it in general. IC Diamond 7 is a wealth of information and I can see many of his posts in this thread answering many of the more technical questions that posters may have when it comes to TIM in the future.

I sincerely hope, as all the staff and members of Hardwarelogic do, that IC Diamond 7 sticks around HL long after this promotion ends.

[theburan]

Yeah, it's been very informative... although I have to go google what the "magic cross" method is, lol.

*Edit: Ahh, I see now... although I wish my first google search had been more specific... there are some magic crosses out that that I wish I didnt know about.... *shudders