IC Graphite Thermal Pad Available for Test and Review

No in my personal opinion the issue a lot of folks had was that the sample size given by IC was too small and they had to double up to completely cover the chip which is something Panasonic says not to do ie fold, scratch or stack.

What I bought should work great. Pano says it is a TIM replacement and so I am taking them at their word. The application it was designed for was circuitry, power modules and semiconductors. So again good application useage.

The issue is that a lot of people didn't read the white paper. PGS tape is not a thermal interface material. It is too stiff to fill voids. The only one that is...IS soft PGS (EYG-S) which suspends the graphite layers in a pyrolytic silicon substrate making it both flexible and soft. Anyone who ordered PGS and their product code does not start with EYG-S is not going to have good results.

.2mm is plenty as a single sheet of paper is .05 thick and most CPU thermal pads (the wax kind) are also .2mm thick.

What I bought (and I believe I'm the only one who has) is the soft TIM variant. I plan on testing it first on a compromised thermally throttling and physically damaged 13r3 that is already going back to Dell

If that goes well I'll use it on my new 15"

 

Has anyone ever tried soldering their heat sink to the CPU die? It's a stupid idea - it would make servicing the laptop nigh on impossible, but it would give better results than liquid metal.

 

No not a stupid idea, the issue is finding a soldier that will whet to glass

 

This is directed, in part, at me with the 0.025mm tape I mentioned, while ignoring the mirroring work I referenced, the pressure film to verify even pressure, and testing that would be needed in order to verify tolerance on flatness, which I mentioned there would be lots of work I would need to do to actually use that one. Good attention to detail there.

Edit: My apologies, I suppose it wouldn't be me, as I did reference EYGS with this product sku: http://static6.arrow.com/aropdfconversion/8f2ff4ca6aae88065df9b0a98b2ce90c20fd1f8e/3aya0000c16.pdf

 

Couldn't you use the stuff AMD uses to solder their Ryzen dies to the IHS?

 

Sorry... typed that post on my phone on which grabbing image links is a 5min each job... and forgot to edit it later with pics.

The bridge is the 30mm x 80mm copper surface on the upper left

it screws into this corresponding structure that is located on top of the CPU heatsink grill; below you can see the stock 30x80mmx1mm grey thermal pad in place.

 

@Sentential - my apologies. Upon re-reading, you directly referenced tape vs "s", which S is the type without any other layers or laminates on the sheet. I read too quickly and owe you an apology, which I also posted on my response from earlier in an edit. Once again, sorry.

 

No worries! My reference were other posters including bloodhawk who appear to have purchased the non soft PGS which may or may not work well based on what I read on the whitepapers.

The person who used the egg-crate anaolgy is trying to split hairs about how much or how little airgap it covers.

IMO the same arguement can be made about improper paste storage, mixing, application etc. I mean that's the whole point of using pads is to simplify the application and prevent user error which is FAR more likely than an irregular surface

 

So you guys
@Papusan @Sentential @jaybee83 @bennyg @Mr. Fox @ajc9988

what exactly is wrong with THIS stuff?

700 W/MK ?
https://www.digikey.com/product-det...ronic-components/EYG-S091210/P12726-ND/678309

Am I missing something here?

 

Far as I can tell? Nothing. All soft PGS starts with EYG-S, it's just the .1mm version of the same material I bought. It can be purchased in multiple thicknesses and I chose the thickest at .2mm instead since it appeared to be the preferred TIM thickness since they have precut versions for specific power module brands.

Is it possible mine is too thick, sure is the opposite true, yea only one way to find out!

Go for it!

For anyone reading I literally went into the site searched EYG-S found the cheapest strip I could find since I'm going to cut it anyway who cares about its shape. I think what I bought was a 90mm by 30mm strip for like $7

 

nothing wrong with these graphene pads at all. this is actually the first time ever that ive seen ANY thermal pads even coming CLOSE to the real life performance of actual thermal paste!

even when you compare the most expensive and super high end thermal pads weve known so far, the fujipoly extreme pads, with something like kryonaut, you wouldnt even be able to turn on your machine when applied to CPU and/or GPU, let alone put any stress whatsoever onto the system. it would just instantly overheat.

so even if these graphene pads are worse by a few degrees C than LM or kryonaut, this is already lightyears ahead to what any other thermal pads have to offer.

i agree with previous posts here that the difference in real life temps (even though theoretical heat transfer rates look to be amazing on the graphene pads and higher than in any conventional thermal paste and even most LMs) is most likely to be found in the applied heatsink pressure, heatsink warping and resulting air gaps in the interface. no pad can compete in this regard with any kind of paste because of its material nature.

so from that perspective, these pads are quite amazing to say the least!

now...whos gonna be the first to properly isolate his gpu's SMDs and apply those suckers instead of regular pads?

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Nope, thats indium solder, the only way to solder copper with the chip die and its a very high precision job. You can read more about it here https://overclocking.guide/the-truth-about-cpu-soldering/

 

Looks like that's going to be me. I thought for sure blookhawk would be but my PGS sample arrives tommorow and I'm going to try it first on a half dead return-in-progress 13r3

On mine highest reported temps were 95C normally 91/88/76/71. Going to start with a small strip on the CPU and see how it does. Then once my 15" gets here I'm doing a full setup assuming it isn't terrible.

I've also purchased normal thermal pads for the RAM but wouldn't be opposed to using on RAM.

Ull report here once I'm done and if all goes well start a thread in the AW section

 

With a 20c core temp differential i highly suggest to repad your heatsink to balance it. I think no solution will work in your case as long if that is not fixed to be honest.

 

Let me know how this goes since you are using the 0.2mm variant and I'm considering either the 0.05mm or 0.025 variant with lapping to a mirror. I'm very interested in the results. Since there is no layer, make sure if using this anywhere other than IHS to heatsink to be careful that this does not touch a solder joint or other component. Since there is no laminate, it will carry voltage and could short components. But you read up on it with the whitepapers, so you are aware, I'm sure. Also, look up K-core, which is another novel use of graphite ( https://www.thermacore.com/products/kcore.aspx).

Actually, this could help potentially with that, and that is what makes this such a good test. One of the potential benefits of PGS is that the A-B plane has the extraordinarily high W/mK. This means that it should be able to reduce hot spots, such as a single core running 20C higher than the other cores. How much is an open question, but that makes this a great test (and potential advertisement for using a similar product by OEMs, especially considering it overcoming issues in QC in heatsinks for products and reducing the number of RMAs and servicing). That is another reason to be so interested in the results of this test!

Edit: What can cut against it is the thickness of the pad and the tolerances necessary to implement the thinner variants, which is why most are trying the tape variant which has the thicker sizes of 1.5mm+....

 

If you guys have good results with the 28 w/mk soft PGS, please order the 700 w/mk and see how well that does.

https://www.digikey.com/product-det...ronic-components/EYG-S091210/P12726-ND/678309

They even have 1000 w/mk version..........

 

They have higher variants than that: http://static6.arrow.com/aropdfconversion/8f2ff4ca6aae88065df9b0a98b2ce90c20fd1f8e/3aya0000c16.pdf

 

yikes.....so are you guys going to order those?

There HAS to be a catch somewhere......

 

Novel idea, not doing it only because of cost. This 13r3 has heavy physical damage to it in ways that I am not going to keep it and already have an exchange in place I am waiting for. This laptop is disposable therefor is a perfect canidate...FOR SCIENCE!

Yea will do. I'll post results either late Friday or early Saturday since I need to get this 13" back in the mail ASAP as I expect my replacement to arrive ant day now. The way I plan on installing this is using a postage stamp sized sheet (since I assume one side has adhesive) and I am going to insulate the PCB around it.

Also you get my point, using a turdbook with known issues will give a solid baseline on how well or not well this stuff works. If it can turn around a worst case scenerio using the .2mm pad then perhaps you are correct that using a .1 or an .05/.025 on a solid setup will show better results. Either way I'm base lining worst case scenario to see whether or not this stuff works as advertised in a low pressure mount condition.

I'm starting with mine first purely because it was designed at this thickness specifically to act as a high grade TIM alternative. I'll let other folks try the more exotic stuff since it takes a week for anything to arrive from these vendors. If it doesn't work then the odds the others will are slim (which is why I also ordered a set of wax based thermal pads) I'd go liquid metal but the oxidization concerns me. Had a test setup and I was seeing serious losses using conductonaut within a 1 year span and I want a permanent config

 

There is, those are very thin, and the high thermal conductivity is on the X and Y axes(between sides of the processor or heatsink). That may be nice, but I think what's needed is high Z axis thermal conductivity(from processor to heat sink), which can be provided by:
https://industrial.panasonic.com/ww/pgs2/graphite-pad
or
https://industrial.panasonic.com/ww/pgs2/soft-pgs

 

What I linked **IS** PGS2 (Soft PGS).
The sheets are EYG-S. That's exactly what you linked.

So......again....
what's the catch?

These should be as good as LM.

Someone should test the 700 w/mk (0.100mm) or the 1000 w/mk (these are even thinner).
I suggested the 700 because 0.100 might be required for laptop heatsinks and going too low may not have good contact.

 

A week?
Mine took 2 days (arrives today, ordered it Tuesday).

If you can try the 700 w/mk version. It's 0.100 mm.
I didnt mention the 1000 w/mk because the thickness may be too thin for laptop use.

 

So, you are assuming an adhesive. Look at the product spec sheet, the S variant is graphite sheet w/o the other laminates. http://static6.arrow.com/aropdfconversion/8f2ff4ca6aae88065df9b0a98b2ce90c20fd1f8e/3aya0000c16.pdf

The catch is the "Z" transfer being less and the thickness being so low that the tolerances are tighter, which can effect the applications as stricter tolerances usually increase costs for machining and costs of QA/QC to a degree.

 

100% truth. But all we need is for this 700 w/mk pad to equal liquid metal in temps, not exceed it. That's all.
We get that and its an auto win because no degradation or oxidation will occur.

So 700 w/mk being actually 35 w/mk in laptops is still a win win.

 

That is the idea. Also, this is why Linus brought up how tight the core temps were, which is the A-B plane removing the hot spots of a single core, which is an issue if you have uneven mounting or a bad application of TIM. If you do the 0.05mm sheet, that increases the AB plane to 1300W/mK while reducing the thermal resistance, which may help with the C plane, which is what we need. Now, I need to repaste anyways and service my loop soon, so that is why I'm considering getting what is needed and mirroring the surfaces like I did with my laptop (much easier now compared to doing it with the heatsink with heatpipes already attached). Wet-dry sandpaper and polishing paper going from 600grit up to 2500 grit (on the polishing paper) with a flat pane of glass will get you a very good result. You need to watch the edges so you don't round them, constantly change direction, etc. If you have an ultrasonic plate, you can attach sand paper to it, stick the IHS or cold plate on it, turn it on, walk away, and make sure to clear the sanded material which can log the sand paper or polishing paper over time with removed nickel or copper from the parts being sanded. If you have access to a machine shop with a CNC, you can also get it very tight on the tolerances. So how you do it is a matter off what you have access to and how you want to proceed, as well as what tolerances you want to achieve. Pressure film helps to find spots on the plane that have weaker contact, which can be the results of not meeting the tolerances on the planes. I could go on, but take your pick, from using the techniques of older gens with a glass plate to modern techniques.

 

https://www.amazon.com/Innovation-Cooling-Graphite-Thermal-Pad/dp/B07CKVW18G/

Back on amazon, might get some to test while I try to stream/game on these up coming summer days.

 

What's the consensus on this one? A no go for laptops due to low pressure mounting? I am interested in the other pads, hope others can test soon

 

Yeah i would wait before wasting money on the overpriced IC Pads.

 

I hope it doesn't turn out to be $24 flushed down the toilet. Hopefully, it will work well.

https://www.digikey.com/product-detail/en/EYG-S091210/P12726-ND/678309

 

Worst guess is it will be at least equal to Nanogrease Extreme (on a desktop).
Best case, a few C worse than Conductonaut.

 

Keep us posted! I doubt it will rival liquid metal compounds, but if it does, then we have our solution!

 

https://www.aliexpress.com/item/Gra...Products-Thermal-Dissipation/32832215417.html

probably will work as well, maybe, possibly, could be

 

For you science dorks out there like me:
https://industrial.panasonic.com/cdbs/www-data/pdf/AYA0000/AYA0000C27.pdf

 

didn't read the previous pages, which one is sold as ic graphite?

 

None that I'm aware of, I'm just picking this out for the ones that interested in reading up on the properties of the pads @Mr. Fox has just purchased for testing, specced to 700 w/mk apparently (EYGS091210)

 

ah

will wait

 

@Mr. Fox If you find a way to solder the outer edge of the pad so that it stays flush with the bare die which might provide some greater degree of reduction in temps.

 

I am probably not going to test it between the die and IHS unless it does a superior job between the IHS and heat sink. If it does, then I might try it there. I don't have a problem using liquid metal to achieve the results that I want. However, my main attraction to this, (but only if it works well enough,) is primarily ease of use and re-usability, convenience, etc. Would be great to be able to remove and replace parts a few times, with no cleanup or special handling, and without having to spend money on one-shot consumables such liquid metal or thermal paste, and cleaning supplies.

 

If its cheaper than LM then its worth the investment and time for LM like thermal dissipation.

 

I agree. It could even be a little more expensive if it works as well and provides the other benefits mentioned. Re-usability could make it less expensive in the long run.

 

It's not cheaper than LM. Unless assuming the 700 W/MK or 1000 W/MK is equal to LM, that it is reusable enough on different builds. Minus the LM cleaning supplies as well.

Here is enough LM to paste 500 computers.

https://www.ebay.com/itm/323213005662?ViewItem=&item=323213005662

 

Shady?
Anyone can make galinstan alloy.
I bought some of it and it works fine. Maybe 1C worse than conductonaut.

Now if you want to pay $15 for 1 gram of conductonaut, that's on you, man.

 

was more of a joke

 

Fair enough @Mobius 1 +Rep

 

Back on topic.

I'm debating whether or not I should do a comparison of the sample ic graphite pads vs kryonaut on the P71.

Currently I have one with a P5000M and P4000M, but the disassembly process is ass.

 

No, the 28 w/mk one isn't good enough.
If you're going to do a comparison, get the 700 w/mk version and test that (or even the 1000 w/mk version).
Might as well make it count.

 

I doubt this stuff will be any better than the IC Graphite pad. The IC graphite pad has a higher thermal conductivity in the xy plane (800 W/mK vs 700 W/mK) and it therefore follows that this pad has worse thermal conductivity in the z direction as well. Is the pad that @Mr. Fox bought compressible? If it isn't, it will do poorly in laptops that have a low heatsink contact pressure.

 

I so want to get my hands on that. Couldnt find anyone shipping it in the EU.

 

So, going to try to clear up some confusion on the different lines from Panasonic from a little research:

EYG-S types are JUST graphene. If you want SOFT PGS, then you need to look at the "ZL" later in the sku for the type of sheet. https://industrial.panasonic.com/cdbs/www-data/pdf/AYA0000/AYA0000C50.pdf
If you want the sheets "optimized" for Z/C planes (not A-B plane of the material), this is the EYG-T line. https://na.industrial.panasonic.com...les/graphite-pad-eyg-t-series-intro-paper.pdf
The EYGS line that does not have the "ZL" in it is the normal PGS sheet. https://na.industrial.panasonic.com...es/pyrolytic-graphite-sheet-brochure-2016.pdf. This compresses, but not nearly as much as the Soft PGS, while also being thinner than the soft PGS line.

How well each of these works is up for testing in our use case, but they each have properties that are beneficial and each have properties that are detriments for our use cases. It seems that IC is using the Z plane optimized variant. They may have tested the other variants mentioned here or not. I do not know. IC could tell us more on that, though. Either way, looking forward to seeing people's results, especially if someone wants to do a pre and post-lapping comparison with these sheets and with a high quality thermal paste (excluding LM). If it exceeds a high quality thermal paste after lapping, then maybe a comparison with LM would be something to do.