Suggestion to make the graphene sheet thermal solution work

Abhrant · July 20, 2020

Hi guys, I am new here and I just watched the ltt video about the graphene thermal pad ( Too good to be true was the video title). I have a suggestion about how it can be made to work. We can take a large piece of that sheet and stick it to the cpu. We can then fold the remaining sheet and stick it to the cooler. In this way, we can use the high surface conductivity to transfer the heat .

spartaman64 · July 20, 2020

the thing is graphene is one layer of atoms and the point of a thermal interface is to fill the valleys and peaks between the two surfaces so one sheet of graphene probably wont do the trick. and as you add layers you are basically making graphite and im not sure if graphite is a really good thermal interface

IkeaGnome · July 20, 2020

It's a good one apparently, just not for this. For this application it wouldn't be that great. Graphite has a thermal conductivity on 10W/m-K as a through plane TIM with 150W/m-K in an in plane configuration. This would be through plane. For reference, Grizzly Conductonaut is 73W/m-K through plane and MX-4 is 8.5W/m-K.

I like when a discussion on here ends up with Googling "useless" information

Falkentyne · July 21, 2020

Just fyi, Grizzly conductonaut is only about 30 w/mk at absolute best, and probably lower than that.

You can't have an alloy that has a HIGHER w/mk than its metal component compounds. That would violate basic chemistry and metallurgy laws. And this stuff is a science.

Gallium, the main component in LM, is 40 w/mk. The entire eutectic alloy can **NOT** be higher than 40 w/mk and MUST be at least a few lower, if not much more lower.