Worse thermals with LM and new pads ...

My RB17 Pro is hitting 90 spinning up normal tasks like FireFox for eg ... have I done something wrong here?

I suspect perhaps 2mm pads were too thick and there isn't enough contact being made?

 

Please take a look if the DIE made contact after the repaste with LM. You can see this easily when a good contact between HS and DIE was made. If there is no imprint visible the pads are maybe not attached correctly or you have used not enough LM.

 

You manually spread the liquid metal right? It does not self spread like regular paste.

 

Yep, it's spread out thinly.

I replaced the pads with 1mm'ers today and the Liquid Metal is still doing nothing - 95 degree CPU when I run Firefox ...

Anybody else tried to Liquid Metal their RB17 Pro?

 

Make sure the heatsink has LM spread on it too.

 

It does - it was old LM from a few months back, I guess this is the only other factor? Does it "dry up" in the tube?

 

Some compound liquid metals can separate and harden. LM will react to and be absorbed by copper over time. Old LM on a heasink will sometimes need to be sanded off and reapplied for good wetting.

It sounds like you flat out lack contact. I would test a bit without any thermal pads to see if that is knocking your heatsink uneven. Just keep hard stress tests under say 30s, and give the board a few minutes to cool between.

 

The previous paste job was stock and it had indeed hardened and spread all over the place, I gave it a good rub down with the alcohol wipes provided in the LM package tho.

I put all the thermal pads in the same place as the stock ones tho - that's why I can't understand the lack of contact.

 

I assume you checked the LM application when you swapped pads?

Did it look like it’s making proper contact with the heatsink? If not, add a tiny bit more. (Either on the die, or right onto the heatsink)

Did you use tape on top of the conformal coating? If so, might wanna check the thickness and application. I personally use Kapton tape as it’s much thinner and I didn’t want to insulate the components too much.

If all checks out, only thing left is to replace the LM with a new batch.

 

It looks like it's drying up and separating under the heat of this now molten lava Razer Blade.

 

Ooof, no bueno.

I'd remove and replace with thermal paste if you have some handy.. at least until you get a new batch of LM.

 

New batch on the way! Lets hope it's the old LM otherwise this things an email reader now LOL.

 

My laptop has bad contact. I did many attempts when doing LM, kept spreading it thinly following instructions/guides/videos but thermals weren't great (always 1-2 hot cores). Eventually figured it was uneven contact and adding a little extra LM did the trick to get much better thermals. The triangle screw mounting mechanism for the cpu heatsink commonly used in laptops isn't the best at applying even pressure either.

Some other articles/videos/forums say with poor pressure/contact the LM may be exposed to more air which can speed up the "drying" process. It isn't actually drying but reacting and absorbing into the copper which is why various places mention re-applying something like once a year or when thermals go bad. As time progresses, eventually enough reaction would occurr at the copper surface that it no longer reacts as quickly with LM and you can go for longer periods whilst maintaining good thermal performance.

 

I had a similar problem in my M18x R2 when replacing thermal pads on the CPU phases. The problem? The thermal pads were too thick. Step down to 0.5mm and see if that improves. I had 1.0mm on my M18x and had to step down to 0.5mm to prevent this from happening again. Been good for 6 years now. Air pockets can significantly accelerate the degradation of galinstan in the presence of high heat I’ve found.

 

Thickness of the pads was it indeed!

 

So what do your temperatures look like now?