Anyone used used Liquid metal on P650HS-G?

Considering using LM on my machine, but before hand want to find out if anyone else has used it before.

 

You can use liquid metal on any CPU or GPU layout. Just make sure you make a barrier to protect the liquid metal from slipping onto you circuits. See my post on this link for some reference of how the foam barrier is placed:

http://forum.notebookreview.com/thr...876-owners-lounge-phoenix-3-0.800081/page-218

And have a look at MrFox videos of liquid metal on the GPU's and CPU's here:

 

I plan to put LM on my RS (a few weeks or a month)

 

I put some on my P650SG, 10-15°C drop overall.

 

Having a barrier is a good idea, just pay close attention to the material used in the video.

 

Cheers for the Info! I will be honest. I had already tried using LM, and the results were amazing. But it leaked out. So I had to do the emergency clean and go back to kyronaut. >_<

 

How did u know it leaked out? Or was it pure luck?
Also, you should add isolation tape under the foam for extra security, then u're all set i'd say

 

Usually people open the machine and start finding it places it was not before

 

THIS... After they can't boot up their machines anymore

 

Yea it didn't boot up at first, then I had to clean it all out. Once I was fully confident I gave it another go and it worked as intended again lol, the level of sadness I had! But I did really like the temps, now when I play xcom 2, yea it barely goes above 80, but its obnoxiously loud , starting to wish I got the alienware 13 its so annoying.

 

Is the difference between the LM and Kryonaut big enough to run the risk with the LM again?

 

Hmm this is a good question. I will say that when I had it the CPU stuck at a steady 60c-69. With the kyronaut its peaks at 79c and it is much louder. Stress testing it hovers around 60c though with kyronaut, its odd.

I most probably will be doing it again with the concessions people have recommended and getting it insured. I am really not a fan of the fans.

 

I haven't tried it but i intend to use a non conductive thermal paste at the perimeter of the die to prevent this. My concern is that putting foam might actually insulate the chip even further though there isn't much airflow going on there already. I experienced a leak once before in one of my desktop gpus when i blew on it with a compressor, at first it didn't boot then i jsut did a repaste. It's very important to apply the LM on both the die and the heat sink or in about 3 months it's going to migrate to one section. Applying it on both sides mitigates this and i can say that it lasts longer than applying it on just one side. I haven't fully tested until when but it's 7 months and counting.

 

So what you suggest is to apply a bit of lm on both the die and heatsink to avoid it "sticking" to one side?

 

Always both sides.

 

I read mixed things about the benefits over non-conductive/non-metal TIMs. Would be happy to read about your findings if you compare the two again.

 

I did. here are some results:

there are thing you need to know about this model and LM though before you do it!


since its in Russian: it runs XTU stress test together with Valley with 4k DSR at ultra settings to get close to 99% load on both GPU and CPU

7700hq+1070

EDIT: oh, that's automatic fan profile

 

it is. and the risk is minimal when done properly

 

Hi Danishblunt,

Yes and by doing that you can apply it very thin. The good thing about LM compounds is that you don't have to clean them off every time you take them apart, simply spread them again using the included brushes. You'll notice that over time if you don't apply it on both sides and remove it, some will stick to the die and some will stick to the heatsink, but usually on the heatsink it only appears to be half of the square not entirely matching the die on the opposite side, which means it didn't really stick to it. What i also do is i clean the copper heat sink and die with metal polish first and it gets really clean and shiny removing much of the oxidation and a sticky grey coating that initially from on the Die (probably due to the thermal compound i used before). LM is a bit difficult to apply and i would recommend to start small, and just add as needed, if you put too much it's more difficult to remove and will most likely turn to waste unless you can suck it back in the syringe. Too much LM also makes it easier for it to "run" outside the die and get into the contacts.

 

Thats the first video i seen of a p65x machine with LM running a stress test. Thanks! What do you get in normal usage conditions, like gaming?

 

around 72C, peaking at 75 after a very long gaming session.

to apply LM here you need to force the the heatsinks springs, will make a photo next time I will disassemble one

 

Hello what do you mean by force the heatsinks springs? I applied conductonaut I'm getting around 78c-80c on gpu at 28c room temperature, but the cpu is in the late 80s (one of the cores is higher than the others)

 

Due to terrible heatsink design the pressure on the CPU is rather meh, usually it's recommended to apply a bit more force on them. So that the fit is usable:


It's basicially the same issue on the P1xx notebooks and their GPU heatsink design.

 

IIRC, only the GPU heatsink has spring-loaded screws.

@Tmash You can try tightening the CPU screws more, but don't overdo it (you'll risk stripping them). Have you checked the spread of your Conductonaut?

 

I did try tightening using a genuine philips screw, the spread seems ok, the gpu wasnt (south was tiny bit higher than north) but getting very good temps, probably since the die is large?
I'm waiting for more days before taking into conclusion that the thermal pads near the cpu are way too thick, with time hopefully the pads will compress a bit and see lower temps.

 

Wow amazing trick, thanks will try! Ill be careful not cracking something though, ill tighten the screw under stress test and adjust level that way.

 

Could you post your CPU and ambient temps when running prime95 Small FFTs test only?

 

I tighten all the way down, made around 20 laptops like this, never broke anything.

 

I don't have it anymore, the last photo is from a Clevo VR machine that has the same MOBO

 

Shame. I get 83*C max with 20*C ambient with Gelid Extreme and was wondering if LM was worth it.

 

I have around -18c decrease with LM on gpu, cpu got worse due to the leveling issue, and LM is not thick enough.

 

And this is my main concern. Too many things that can go wrong, especially in a laptop without proper (when speaking about LM) cooling. I rarely go above 80*C in normal use with fans on auto.

 

On CPU with OCCT? If not can you kindly test it? Oddly p95 triggers tdp limit so it does not run as hot as OCCT, OCCT from my test seems to load the CPU the same as Adobe Premier export function.

 

Ambient ~12*C.

 

I assume ambient is 22 or 21c?
At that temp you don't need thermal paste haha
Edit: Your heatsink also appears to be uneven even though its less than mine as the temperature difference between your cores are 4c, and Intel describes DTS accuracy in +-5c. Mine is around 8-10c, two cores are 79-80, others are 91-92 on 28c ambient (its hot here in KSA ). Guess I should try CLU as it seems slightly thicker than conductonaut, or just remove the bottom thermal pads as they seem unnecessary for 7700hq imo (measured with probe + older clevos didn't use heatsink on them: chokes, inductors, vrms), affecting the heatsink balance on non centered CPU die

 

Nope, it's 12. 13 ATM. It's possible the paste is uneven, cause I've done some messy job. Gelid has gluey consistency and I didn't bathe it in hot water. Without it, you can forget to spread it with a razor blade. Even the rice grain resembles more of a turd than a grain. :/
Only thing I've done was heatsink polishing and bending those flat metal bars away from MB, so they apply more pressure when screwed back in. Didn't wat to go for toothpick nor paper clip tricks yet.

 

Should try this with conductonaut http://www.coollaboratory.com/product/coollaboratory-liquid-metalpad/

 

What do you mean "with conductonaut"? Make a die-LM-PAD-LM-heatsink sandwich? Isn't it for standalone use? Can't find thermal conductivity value for it anywhere.

 

Ahh yeah, not sure if the thermal pad is compressible. I've seen a video 4 years ago using clu with 2 sheets of these sandwiched, still trying to find it. I think LM would help filling heatsink imperfections as the die is usually very flat and perfect (non porous)

 

The pad you've linked has a melting point of 59*C. In other words, it looks like a LM for noobs - place it and let it melt under heat. Maybe they were using normal copper sheets, which seems to be more reasonable. I was looking for a solder with low enough melting point and high enough thermal conductivity and simply solder the heatsink do die just like AMD with their IHS.

 

Those have been around for a little bit, they don't tend to work to well with laptops with the lower pressure heatsinks.