Which would cool a vrm better

[pqwoerituytruei]

1. Mosfet -> thermal adhesive -> small aluminum fin stack
2. Mosfet -> 0.5mm pad (claimed 6.0 W/mk) -> massive copper heat spreader w/ small aluminum fin stacks attached via thermal adhesive everywhere they will fit, this would have probably around 2 to 2.5x the surface area

both would have active cooling

 

[Zerileous]

VRM on what? And for what purpose? It may be a moot point with active cooling. Thermal adhesive is going to transfer better than the pad but is permanent.

 

[pqwoerituytruei]

On this card:

https://imgur.com/a/DvthwXb

1x High side mosfet: http://www.aosmd.com/res/data_sheets/AON6594.pdf
2x Low side mosfet: http://www.aosmd.com/res/data_sheets/AON6508.pdf

I was thinking of using a L shaped piece of copper with a thin thermal pad and covering it with heat syncs, the layout would favor the high side mosfet over the low side as that is the weakest link

I was just wondering if it would be better by brute force

I already going to be using adhesive elsewhere on the card, so permanent is not a issue, but i think it may be easier to do using the pad

 

[ShrimpBrime]

Sometimes the best way to cool a VRM is with nothing but a fan. No heatsinks.
If looking to go big time with hard v-mods, you don't have a reason to worry much cooling the VRMs so long as they have air across them.

 

[pqwoerituytruei]

Been getting everything ready to start on this this morning, about to start cleaning and gluing in the next hour

 

[pqwoerituytruei]

Well i got the cooler installed, it complained about uefi support so i booted with csm, guess it is time to mod the bios
tested heaven a few minutes and found a problem, the black plate now that it sinks heat so well it overheated my M.2 (that idles at stupid high temps anyway)
i setup a jank fan to make it work till i get a chance to work on it