VRM Thermals; Thermal Mass vs. Surface Area?

[suzerain_00]

I'm thinking about picking up a MSI X370 Gaming Pro Carbon local from a guy who is selling it for $120. In the Tweaktown article I linked, they describe the VRMs on the board as "not that amazing, but the VRM gets the job done.".

 

What I want to know is that THEORETICALLY (ignoring airflow because air is ultimately the dissipative bottleneck), is it more important to value thermal mass of a VRM heatsink or surface area?

 

I'm thinking about experimenting with the VRM thermal solution, to see if mounting aftermarket heatsinks something like this are worth it from a VRM thermals standpoint. Does anyone have any sort of relevant knowledge on the topic? I'd love to hear some thoughts on the concept!

 

Thanks a bunch!

 

[Jurrunio]

if you know the mass, you can. They are mostly aluminium anyway, plenty of numbers for a block of aluminium.

[nick name]

I guess it depends on how much heat the VRM put outs.  Overall I would assume surface area is ideal, but may not be needed if the heatsink isn't saturated.  

[dgsddfgdfhgs]

heat is all about rate of heat energy transferred.

surface area , temp diff, heat condction from heat source to surface area etc

[suzerain_00]

7 minutes ago, Jurrunio said:

if you know the mass, you can. They are mostly aluminium anyway, plenty of numbers for a block of aluminium.

I'll have to weight the heatsinks If I go through with this then. Thank you!

[suzerain_00]

5 minutes ago, nick name said:

I guess it depends on how much heat the VRM put outs.  Overall I would assume surface area is ideal, but may not be needed if the heatsink isn't saturated.  

The heatsink on the board looks impressive on first look, but I'd imagine that cross cut sinks would offer more surface area - my thinking is that this would improve thermals. So to your point I guess that it would have to be something I have to test

[suzerain_00]

3 minutes ago, dgsddfgdfhgs said:

heat is all about rate of heat energy transferred.

surface area , temp diff, heat condction from heat source to surface area etc

It's good that you mention that. Regular thermal pads seem to be around 6.0 w/mk, but the adhesive kind seem to be considerably worse - and thicker at that....

[LukeLinusFanFic]

There's obviously a balance you need to strike, but I definitely would value surface area for dissaption.

I'm sure there a formula you could plug that into and find out. Wanna wait for a bit until I'm done with intro to thermodynamics?

[TrigrH]

All you need to do is have direct airflow over the VRM and you will be fine for all your overclocking needs. Inwin Mars + AIO is a pretty easy way to do it.

[dgsddfgdfhgs]

1 hour ago, ExodusR said:

I'll have to weight the heatsinks If I go through with this then. Thank you!

mass really has minimal effect on cooling, you actually want as little mass as possible.

Energy = mass x heat capacity x temp diff

 

So if the design of Al block isnt too thin which affects conduction , it shall be fine.

[suzerain_00]

13 hours ago, TrigrH said:

All you need to do is have direct airflow over the VRM and you will be fine for all your overclocking needs. Inwin Mars + AIO is a pretty easy way to do it.

Yeah, ultimately airflow is the limiting factor. Sticking a fan directly in line with the heatsink should pretty much guarantee good thermals.

[suzerain_00]

13 hours ago, dgsddfgdfhgs said:

mass really has minimal effect on cooling, you actually want as little mass as possible.

Energy = mass x heat capacity x temp diff

 

So if the design of Al block isnt too thin which affects conduction , it shall be fine.

Good to know! I'll keep this formula in mind.