New "miracle" TIM, but does it handle phase?...

[Stigma]

Hey all,

I'm curious about this new TIM:
http://www.heatsinkfactory.com/IC-Diamond-7-Carat-Thermal-Compound-15-gram-p-16605.html

Its been shown in reviews to actually outperform artic silver by a relatively large margin (as far as TIMs go that is), but I can find no info on wether this TIM is suitable for a phasecooler setup. I know all TIMs are not...

here are two reviews for others who are interrested:
http://xtreview.com/addcomment-id-2884-view-IC-diamond-7-vs-arctic-silver-5-review.html
http://www.amdzone.com/modules.php?op=modload&name=Sections&file=index&req=viewarticle&artid=307&page=1

So basicly, does anyone know how this reacts to phase, or better yet, have tested it with phase?

-Stigma

[Ssilencer]

Hey all,

I'm curious about this new TIM:
http://www.heatsinkfactory.com/IC-Diamond-7-Carat-Thermal-Compound-15-gram-p-16605.html

Its been shown in reviews to actually outperform artic silver by a relatively large margin (as far as TIMs go that is), but I can find no info on wether this TIM is suitable for a phasecooler setup. I know all TIMs are not...

here are two reviews for others who are interrested:
http://xtreview.com/addcomment-id-2884-view-IC-diamond-7-vs-arctic-silver-5-review.html
http://www.amdzone.com/modules.php?op=modload&name=Sections&file=index&req=viewarticle&artid=307&page=1

So basicly, does anyone know how this reacts to phase, or better yet, have tested it with phase?

-Stigma

No, but I can test it for you for free if you send me some :p:
For just 5 dolars I think you can try it withut a great lost.:)

EDIT: It is the same review in both places, same proc, same temps...
At least it is non conductive

[Stigma]

Well, the cost of it for testing is negible, but the time and effort it takes to make a halfway reliable scientific test between this and AC ceramique for example is.

Heck, it would be worth 5 bucks for me to know if the data was reliable ;)

-Stigma

[[XC] gomeler]

Interesting stuff, I wonder how it handles sub-zero temps though. $5 + shipping, should ask for a review sample.. actually I think I will and mention my website *grin*.

[Stigma]

That would be sweet. Considering the relatively insignificant pricertag, I'm sure that if anyone with a little "weight" on the forums or something requests a review it would not be hard to get a sample.

In fact, one of the people who runs overclockers.com was supposedly heavily involved in the design of this thing. Perhaps it would pay off to send them a mail too.

I can't remember the exact name, but im pretty sure it was mentioned in one of the reviews I linked to.

-Stigma

[[XC] gomeler]

Later tonight I'll don my PR hat and pull out the charm and see what I can whip up. Otherwise I'll just drop the cash on it. Had a few pertinent things to take care of first that'll facilitate a review :)

[Shalamay]

How do you go about a test like this? What's your idea?

[Stigma]

How? well thats rather simple...

Get this new TIM, and atleast 1 well known brand (like AS ceramique, since thats most used for phase), then simply measure the max load and idle temps with one of them, clean and reapply the other and then re-measure. Obviously give both enough time to settle to attain their maximum efficiency.

Its about as simple as it gets really. The differences in idle and load will tell us if this new TIM keeps its performance advantage even under extreme low temps, or if the curve is thrown off.

There isn't really anything technical involved, but a fair amount of patience and perfectionism (for applying the TIM and getting comparable results) is obviously required. The difference between a good and a bad TIM application can be just as big as the difference between a low-grade TIM and a high-grade one.

-Stigma

[[XC] gomeler]

Here's my idea, should have enough in each tube to do this.
Order 1 tube of Ceramique
Order 1 tube of miracle whip

Using Ceramique mount one of my single-stages to a Q6600 and 100% load it for 3 days time. Measure temps at precise times during the day. Something like 6am, noon, 6pm, midnight to account for differing ambient temps affecting the single-stage itself. Temps shouldn't flucuate from day to day but they will flucuate as the temps outside affect the temps inside my house.

Clean evap and processor, apply proper amount of miracle whip and repeat the same processs. At the end of the 3 days compare the temperatures and look for patterns. Check and see if there's a period of time during which the effectiveness of the TIM increases or if there's no increase in performance with time. Look for other patterns.

Next if I can muster the balls to run my cascade for 6 days (going to cost a pretty penny in electricity..) I'll test how effective both are under cascade temps but that might have to be put off till later. I should really then repeat each test twice and eliminate the low and high values from the data to get an average but that'd constitute nearly a month of testing and frankly I'm not eager to burn that much power during the prime of summer. Anyone have better ideas for tests? I was thinking of using speedfan and it's log function to make it as controlled as possible.

[jimmyz]

as long as any charts show both evap and cpu temps it should be a good indication of heat transfer. that is what i look at when i remount. a larger than 5 degree diff. @ 1.6v means poor contact on my rig.

[Stigma]

Yep, a lot of data can be extrapolated from the difference temperatures between the CPU and the evap.

It doesn't even matter that much if the temps are 100% correct (which can be damn hard to do, especially for the CPU sensor), as long as they are consistent. After all we just want to compare it to another product, so the hard numbers aren't nearly as important as the delta of the numbers between TIM1 and TIM2.

I think 3 days is a way overkill. As long as it gets some time to cure (neither this new one, nor caramique requires a lot of time), then something like 3 hours (or even 1,5) would be more than enough to warm/cool the system to the equilibrium point. The only more accuracy you would get from doing it over 3 days would be the extra time it had to cure, and you can accomplish that part without running the phase all the time.

If you wanna be thorough I'd rather you do some curing cycles each TIM before the actual test. Ie. make it run through hot/col/hot cycles to help the curing process. Basicly just run 5 minutes on, 5 minutes off, 5 minutes on maybye 5-10 times. Then at the end wait 24hours. That will easily fill the setting "requirements" for AS ceramique:


Due to the unique shapes and sizes of the particles in Céramique, it will take a minimum of 25 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop slightly over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.

As for this new TIM, it says it only needs as much as 2 hrs, so thats even less of an issue.



Curing Time IC Diamond requires minimal time to attain peak performance; in most cases, IC Diamond will reach peak performance after two hours of use.


Running your phase for 6 days straight is definitely NOT needed. In fact you will probably get more accurate results from not doing that since making the temperature cycle in the TIM helps it cure, at least for the AS ceramique. I don't think it would do anything to your phasecooler, but your right about the electricity bill, which is unnecessary ;)

-Stigma

[Stigma]

Oh, and there is one other TIM I'd like to mention as a candidate to compare against, one that I have had my eye on for a long time: Shin etsu.

I know it gives great results on normal setups, generally better than AC5, but I have no idea on how shin etsu works with phasecooling either. There is just way too few tests done on TIM under phase cooling scenarios.

EDIT: Oh and the name of the guy from overclockers.com who was involved in making this IC diamond stuff is Joe Citerella, if that helps any :)

-Stigma

[Shalamay]

so, who's testing what? This will be interesting.

[Stigma]

Here is an update folks:

Last night I contacted the guys from innovation Cooling who make the IC Diamond TIM to ask about how it would react to phasecooling.

The answer was very positive, and although they have not done any real-life testing at those temperatures, they were very interested in the results of a test like this, and have offered to send some samples as long as its in-US. Sadly that disqualifies me as a testing candiate, but you get a shot at it [XC] gomeler ;)

I've sent you a copy of the correspondence in a PM Gomeler, so you can let them know your shipping information to make this happen. The only other condition for this was that we make the results public on the forum, but that was the plan all along, so that shouldn't be an issue ;)

-Stigma

[Planet]

Why not ln2 or dry ice? I would like to see how it holds up vs ceramique. Most of the time the tim has like 15-20 minutes to set.

[[XC] gomeler]

I don't bench with LN2 or Dry Ice but I should be able to throw Dry Ice temperatures at it through an evap. First will be single-stage testing though as that's simpler, then I'll dip a bit lower into the DI temps.

[Stigma]

Personally I think the phase-cooler range of temps is most important as you can actually run a system 24/7 on that, and thus the small differences a good TIM can get you become more relevant.

Dry-Ice and LN2 by their nature are only suitable to short-time benchmarking, and also tend to have a less stable temperature through testing.

Do you mean to say you have both a single-stage and a cascade available gomeler?

[[XC] gomeler]

Yes, I've always got atleast an experimental cascade/autocascade on the table that I use as stress relief from commercially building the singlestages. Currently charged with R290/R744 but depending on NoL's results with a new Ethylene supplier it might be R507/R1150. If that works out then I'll be able to test a very wide range of evap temperatures and loads. Otherwise I'm limited to a maximum of -75C or so unloaded, load temps down around -6xC though on the cascade.
edit: waiting for a reply back from the manufacturer, will keep you guys posted. Will be a chance to play with my new load-tester also with all it's nice little ports for temp readings.

[Shalamay]

Awsome. I really want to see the results. Keep us updated on your stats please gomeler.

[Speederlander]

Wow, guys. You need to get to the aircooling forum more often. :p:

This stuff doesn't live up to the hype. It's been tried by people here and on other forums. Existing products beat or match it.

[Stigma]

Wow, guys. You need to get to the aircooling forum more often. :p:

This stuff doesn't live up to the hype. It's been tried by people here and on other forums. Existing products beat or match it.

I'm assuming those are non-phase tests, but regardless I would like to see some links to these tests you mention. The only ones I have seen so far are very promising ones.

-Stigma

[[XC] gomeler]

Aircooling? As in aircooling my condenser? Oh wait, people still aircool processors? :p: Just got an email back from the manufacturer's of IC Diamond 7, guessing I'll have the stuff here in a little bit. Will keep you guys posted, if anyone wants to see any other tests on it I"ll do what I can or I can mail the tube to anyone within the CONUS, I shouldn't need too much of it.

[[XC] gomeler]

Got a tracking number already, this guy is fast *gasp*.

[Stigma]

gomeler;2362671']Aircooling? As in aircooling my condenser? Oh wait, people still aircool processors? :p: Just got an email back from the manufacturer's of IC Diamond 7, guessing I'll have the stuff here in a little bit. Will keep you guys posted, if anyone wants to see any other tests on it I"ll do what I can or I can mail the tube to anyone within the CONUS, I shouldn't need too much of it.

What? people still aircool their condensers? ;)

-Stigma

[Xeon th MG Pony]

What? people still aircool their condensers? ;)

-Stigma

My goat! Commeners I say!

Water cooled all the way baby!

[[XC] gomeler]

Touche. I should have seen that one coming *doh*

[Stigma]

Waiting on package still, or got any updates for us? :)

-Stigma

[[XC] gomeler]

Don't think Fedex delivers on Sunday plus it'll take a few days to do all the first round of testing under the singlestage. Right now it's sitting about 30 miles from my house, should be delivered tomorrow afternoon. Probably have the first round of testing on Thursday or Friday. I need to do a little bit of modification on my dummyload to properly do these tests, need to resurface the face and then create a groove so I can put a temp probe right in the center of the evap.

[Stigma]

As long as we get plenty of updates, im a happy camper ;)

*jumps up and down all excited like a 7-year old on a sugar-high*

EDIT: One thing I'm wondering though, if you plan to use a dummy-load instead of a real CPU, how do you intend to get accurate temp readings on the "CPU" side? A temp probe on the evap will provide nice additional data, but it won't let us gauge the difference in efficiency between the TIMs. Or... did I just misunderstand you?

-Stigma

[[XC] gomeler]

I'll be doing both dummyload and processor tests. The dummyload itself when I'm done with it will have a temp probe on the evap/dummyload face, a probe about 7mm into the coldplate centered right in the center, then a probe in against the resistors. On the processor testing I'll have a probe on the backside of the evaporator, on the edge of the evaporator, and then the CPU temps from speedfan. Will be using either a Xeon 3040 or a Q6600 depending on how my bank account is looking towards the end of next week. I think I can swing enough to get a Q6600 in time for this testing, that'd be ideal.

[Shalamay]

Sounds like it should be productive. I'm intrested to see how things turn out.

[lloydsmart]

Can't wait for the results of this - something I've been wondering about for quite a while, now.

[[XC] gomeler]

Received the box from the manufacturer, got enough so if I screw up a dozen tests I can start over :p: If my schedule doesn't change I'll be starting tests on Thursday or Friday afternoon, will keep you guys updated.

[Shalamay]

Thanks gomeler. Have fun.

[Stigma]

Good luck Gomeler. Looking forward to the results :)

-Stigma

[Stigma]

Gomeler, any updates for us? Even any preliminary data would be interresting :)

-Stigma

[[XC] gomeler]

Worked fine on a Q6600 under air-cooling. Sadly I just had Rush on campus start up for Greeklife so I've been busy from noon to midnight since last Thursday. Tomorrow I have a day off and no classes asides from one so I should have a chance to test it. I'll keep you guys updated. Just from the look though it doesn't look anything different than the generic crap I have, pasty gray with a unique smell :p:

[Shalamay]

Anything new? Thanks.

[[XC] gomeler]

Bump, results? Oh wait, that's me :p: Waiting on a tube of AS Ceramique to arrive at my place, I ordered the wrong crap, ended up with a tube of AS5, perhaps I should read before hitting buy. However in about an hour I'm going to do some load testing with the testing compound, just need to finish up some tuning on the system I'm using to test.

[killermiller]

Gomeler, 10 days between updates is unacceptable...:)

[billdavis]

Gromeler, 10 days between updates is unacceptable...:)

ditto

i think he is trying to put the dimonds back togather to give to his girl

[[XC] gomeler]

Feel like such a slacker just now getting to test the compound. Just finished up the first round of testing and the compound did seem to be rather effective compared to my control. Going to do these tests once more once the AS ceramique arrives, just got a tracking number and it should be here on Friday. Here's some basic numbers, test was done with my dummy-load, temp probe measured the temperature of the dummy-load on the edge of the evap where the tester and evap surfaces met. Ambient was a steady 31-32C through the entire test.
load generic ICD7
100w -29.2 -29.2
150w -25.1 -25.1
200w -21.3 -21.6
250w -18.8 -19.5
300w -16.9 -16.0

I applied the ICD7 to the evap face, pressed the evap to the load-tester to flatten out the compound as it had a very thick consistency, then separated the evap and tester for the suggested 10 minutes while the solvent evaporated off. The generic TIM was applied to the evap face and immediately mounted to the load tester. I tested the generic compound first and then tested the ICD7 during which the ambient temperatures actually increased roughly 1C. The evap probe made contact in the same location during both tests so for a preliminary result I'd have to say the ICD7 is better than the generic $3 tube of TIM I purchased to get some results out. I'll be doing these exact tests again when the ASC arrives to see if there is a significant deviation or if atleast these preliminary results are mirrored and the compounds show a consistency between mountings. I imagine ambient temperatures will once again be in the 31-32C range and the charge on this system won't change.
Load tester
6 wirebound resistors sandwiched between 2 2"x2"x.75" copper plates with the open face sanded smooth. The other 5 faces are insulated with 1/2" of armaflex and the entire assembly is screwed to the mount for the evap. Load is adjusted with a 6amp variac and a kill-a-watt.
Test system
NF11FX, Cryostar evap, Chilly1 Blue condenser, SUPCO SUD111 Filter/Drier, 8'4" of 0.031" capillary tubing that is wrapped around the suction line. Under 300w load the system is running 8psi suction, 275psi discharge (hot ambients don't help at all) and has 16k of superheat measured 6" from the compressor. System is charged with R22, static charge is around 65psi I believe.
After playing with the ICD7 for a few hours with these single-stage temperatures I have to say it's decent but I don't know how it'll compare to ASC. From what I can tell it held up rather well but I imagine -30C to -15C temperatures aren't pushing the physical properties of the binding agents in the compound. If I get a chance to throw together a proper R1150 cascade I'll be testing this under those extremes, otherwise if someone with an R1150 cascade would like to test this compound please let me know, I've got a plenty of tubes.
Will be back on Thursday or Friday with some ASC/ICD7 results if Fedex arrives before the weekend. Sorry about the massive delays but juggling school, phase orders, and life is nuts.

[TopherTony]

nice job so far. keep it up. very interested

[killermiller]

Is there a period of time that the TIM should cure, other then the inital 10 min period?

[[XC] gomeler]

The manuf only told me to give it 10 minutes for the solvent to evaporate, otherwise there wasn't any sort of specification. Still waiting on the ASC to arrive, ordered from ewiz :shrug:

[[XC] gomeler]

ASC arrived this afternoon but I'll be busy all the way up till Sunday but I'll do another spat of dummyload tests and possibly a processor test on an E6600.

[oublie]

guys,

I designed a product identical to this a couple of years ago (can anyone one say copyright theft?) Although i found the results of the diamond paste to be slightly better on air than AS5 the Artic Silver brand name is so strong it would have been almost impossible to market. Have you ever tried to get your bank manager to give you a business loan for overclockers paste?

Heres the link to the forum stuff on it.

http://www.xtremesystems.org/forums/showthread.php?t=57280&highlight=slurry

[killermiller]

Well investors want to invest in something they know will return. Your best bet would have been to market and sell in forums and ebay and such keeping records and once you established some clients and profits, then pitch it to investors. Why did your thread stop so abrubtly?

[oublie]

Results went well, the paste i developed worked slightly better than as5 but once i found out that i wasn't going to be taken seriously by my bank when it came to running this as a business and because at the time i was unemployed it made more sense to get a job and put the idea on the back burner until my situation improved. I went off to work and the idea is on a shelf somewhere. But like many things in the XS community we lead and others follow in our wake :)

[kiwi]

If you want the best for air/water get liquid metal paste, it is at least 5C better (on C2D setup) than AS5 :)

[gosmeyer]

^ agree^