[TEST] GE72VR - Liquid Metal Grizzly Conductonaut

Hi all !
My laptop is the MSI GE72VR 6RF
I wanted better temps so I tried the famous grizzly conductonaut and here are the results.

First, some pictures :



















First Test with original thermal past (auto cooling profil):
(10min OCCT)
(The core 1 was always the higher so I only show you the results of this core.)





96°C after 6 minutes.
93,25°C average for the 4 cores.



Test with Grizzly Conductonaut (auto cooling profil) :







80°C (16°C Drop !)
79,25°C average for the 4 cores (14°C drop !)




And with the notepal u3 :






For me this is insane results. Big thank to this forum who give me big help.
(sorry if my english is approximative)


See you !

 

Isnt dangerous to use liquid metal termal paste? there are no risks?

 

I saw my favorite LM and was like "ooh must go read", then i saw this :


and I was like THANK HEAVEN he repasted, lolol

However, I would suggest using grizzly kryonot instead of LM...

the reason for it, is because LM splashed out when pressed between the two even surfaces of the die surface and the heatsink's copper surface.

I take pictures some times to show people, I think I have one recently , let me show you what I did in something relatively similar to avoid the splashing, though it was a shimming and not using LM to direct on die...


So if you look closely, those little drops of silvery goodness in the edges came out, but not right away, I made very even contact and already cleaned this up before covering the edges with tape to be safe.

If I didn't prepare for this eventuality those drops would have flown out when using the laptop or moving it in the future. They're conductive and could spell doom for your laptop. just adding this, not because I think your not careful, but incase anyone else would read and try this at home and maybe not take the precautions.

 

I want to point out , what I believe to be a small mistake that is hard to initially understand because it's not explained very well in any instructions :

1) this is too much LM
2) it's not "worked" in enough, and thats why it hasn't seeped into the surface on the molecular level yet (the molecules that make up or bind these sort of LM tims are really small and fall closer on an molecular level, physically even seeping inbetween particles.... like water over a strainer)
3) if you "scratch" the tim into the surface a bit more, working it, it will stick better and become smoother without the variance. Your temps will probably improve by 10C more if I had to speculate.
4) I believe if you closed this and ran it and moved it around, you should now be able to open it up and find those little silvery metal pellets all over the place. They go far... like an ice cube on an oil frying pan thats hot.

 

There are some serious risks.

I do not normally add my input on the MSI section of these forums, but I had to once I saw this, I will get the LHz team to add thier comments since this is exactly thier thing too and they should weigh in, I wouldn't be surprised if a hearty convo or discussion is already on - in the discord server they have.

 

@iunlock Hey care to weigh in, on your teams thoughts ? (maybe you have documentation and how-to video's or have done this laptop specifically..)

 

Right, i used normal Kryonaut termal paste on my gt72vre 7re, but there were folks saying to use liquid metal instead, i google it and read lots of risks using this.

So i think i did good

 

You did great !

I do repaste something pretty much everyday , and it's been steadily like this for awhile, I absolutely think you did the right thing by using Kryonaut instead. The temps will not be much different, while the margin of error is enormous using LM for this setup.

I would not be confident doing it myself and guaranteeing safe results. i just know the LM will splash out, no matter how little you use, and pretty much unless you make a complete container for it, it will go everywhere.

I've gotten motherboards and videocards sent to me, that were killed by this, so I always warn people or caution them.

the OP's temps, based on the OEM pasting I seen, should have been at the least 25-30C lower if done right. Kryonaut would have done at least 25C better then that stock silkscreened thermal paste job that was 1.5mm thick.

 

Thank you for your feedback.

 

Looks like the cpu fan is pretty dirty, might try cleaning it up with some compressed air?

 

Wow, I didn't expect get so many warning :O
Before doing this, I read a lot of stuff, one this forum and other, and even if there is a risk, no doubt about that, I have never read any people being so alarming !
First time I read something about "splash everywhere".

I know there is too much.
But I tried to press the heatsink and remove it to see how the LM spread, and it looked totally fine to me.

So right now Ill turn off my computer and open it again to see the LM, and Ill try to remove a bit to have a clean thin surface.

I really dont believe having same results with kryonaut.

I did

 

So you were right



It spread !
I cleaned everything, removed the tape, take off some LM, and put clean tape all around to protect.



New OCCT test :

Before/After



Core 0 : -4°
Core 1 : -6°
Core 2 : -6°
Core 3 : -5°

So now I have 23°C Drop from the original thermal past.

Whatever... I will not recommand this product. I think we must control the heatsink contact before try LM, and we also must use tape to protect.
For me I think it's now fine, because I don't move with my laptop.

 

please don;t hate me, but I must insist you still have too much thermal paste. Not by as much as before though. It really shouldn't make a mirror shine, it should be thin and evennly spread out, but just wet enough to make proper contact.

I think you should check to see how much contact this system makes with the CPU as well, sometimes, the heatsinks do not make enough contact.

 

BTW, your the first I've seen to admit it when they put too much. Mysteriously everyone else goes silent, and I find thier notebook on Kijiji broken or for parts a week later and never see them again on NBR or something.

SO, thank you for paying attention and not running away, and I'm happy you didn't break the laptop.

I always tell people exactly what you just said " I would not recommend this project" once they know what is what, risk outweighs the rewards and cost, when you can just get Gelid Extreme or Tuniq's TX-4 or Grizzly Kryonout

 

Liquid metal shines (no pun intended) more with [extreme overclocking]. But I agree: In most cases, the payoff is hardly worth the risk.

 

Of course I admit. First time I read your warning I was "wtf he is talking about, everything is perfect". But I supposed you know the product better than me so I decided to open and see.
It's hard to remove. Micro drop can EASILY go everywhere... I was scared.
I should open it again I guess... But Gelid Extreme or kryonaut will not give me the same insane results...

 

Liquid metal is dangerous stuff. While the results are impressive, I honestly would NOT suggest anyone who isn't rich mess around with this stuff. Some people can afford to buy a new laptop if things go south. Some can't. You literally need to as much work as if you were using sub-zero cooling on a 6 ghz desktop...

 

Have you tried though ? Apply grizzly or Gelid Extreme so thin is almost translucent. Should just be enough to fill in the invisible to the eye air gaps in the surface of the die and copper heatsink's surface. That's how it's meant to be used. Less= more. never do you want it to actually prevent the copper from making direct contact with the die itself, it's only meant to fill in the microscopic gaps or air that hold in place.
As for the LM, same thing, but it's viscosity is being interfered with by the surface tension of the air and moisture on the surfaces of the die and copper and that is why you have to scratch it in, the molecules are that light and small. Thats why it runs everywhere.

I know in this day and age of internet information, everyone can so easily dismiss anyone's comments or experience , the net is full of trolls or would be self taughts or considered experts on subjects all over the place, this is why I am happy you took a minute to consider the advice, and in admitting it, you might have helped someone else to do the same thing, step back and do a closer examination of thier project too, and maybe save a motherboard.

As a reseller, I do not want to have to do RMA's for my customers. No ones ever happy in that situation, so you know what goes around comes around.

But - in my experience, thermal pastes can be as good as LM, its not nearly as dangerous, but if you can get an ideal surface going, perfectly levelled, make great contact to the point it's as if the copper and die are one solid piece welded together, than its the best.

Even Liquid metal acts as a INSULATOR compared to direct copper contact.... so less is better still. I've been saying this for decades, hoping to make a difference. Back in the day, they used to ship this gummy white silicon paste , and they would show directions to apply it liberally and stick a heatsink on it.

Directions from like INTEL or IBM themselves. In testing, inschool, we noticed a quick smudge of peanut butter did better , it was like a big joke. We realised it was the fact the peanut butter we used was craft and mostly just oil that became really thin. One thing led to another, even silicon paste was okay, what we used to do was just squish the paste into the cracks on the fiberious copper blocks they used to use and apply nothing to the CPU ceramic (they used to have ceramic, lol...like 286 or 486DX etc..) just a clean surface after using a cheap 70% alcohol solution to clean it etc.. (BTW, Artic Clean is great and NOT expensive for how many years it will last and it does infact work much better than alcohol - and when whiping surfaces, use a coffee filter made of bamboo, which has no small fibers, and was never bleached -because bleach weakens materials and leaves behind invisible particles. never use kleenex or tissue for that reason)

 

Hello,
which screwdriver do you use to remove heatsink and fan screws? (the big copper thing)

 

So I'm one of the ones that fried a GPU using CLLU, it's a learnt experience now (not on an MSI notebook).

woodzstack, do you have the comparable temps of LM vs Kryonout or TX-4?

I'm tempted to have another go with LM on my notebook so would be intrigued at the values.

 

Oh NOES ! Friggin nightmares of that happening man, so sorry to hear you lost your GPU... the stuff is so much more tricky than any advertising would let on.
I always try and warn people....

Yeah I find all three of those to be the same in practice. Maybe within 1-2 C

But I have perfected my technique and done so many applications, when it's not perfect I've gotten so used to just opening things up again and redoing it.. until perfect.

Those are the best TIMS. Add in Gelid Extreme as well.

I soon will be testing Phobya's TIM and will be comparing it in a gauntlet of comparisons and tests to grizzly's Kryonout.

Can you go into some detail how you spilled the LM, maybe that would help the next guy, and maybe we can add some tips or come up with some preventative tips etc..

 

My experience has it's own thread MiSJAH's guide on how NOT to Liquid Metal repaste!!.

 

LOL oh no you didn't ! Ah man.. okay I go there, thank you kindly !

extra rep for being such a sport !