Thermoelectric cooling for portables

[Shank]

As I have been looking for different kinds of cooling I have found out about thermoelectric cooling. The concept seems interesting, and inexpensive. Putting a pad in-between a processor and a fanned heatsink sounds like a great idea.

However, I have never seen someone put one in a portable before.

Would it be beneficial to include one? Are they a viable cooling option for portables?

If no, what's the catch? Is there a drawback?

If yes, why hasn't anyone done it yet?

 

[Blargaman91]

I personally don't pioneer new ideas, I just build off what I see going around. That is probably the main reason it hasn't really been done. Still, maybe there are drawbacks, I don't know.

 

[legofan623]

It adds thickness, maybe only 4mm but that could be crucial for some people. If I knew about it before I made a heatsink I probably would have done it.

 

[Ampz]

I think it's a good idea, I doesn't seem like it'll hurt to try?

 

[grossaffe]

if you're talking about using a peltier-effect type thing, I don't think it'll do what you want. Yes, one side of it will be cool, but the other side will be hot creating more total heat to dissipate. They can be useful sometimes in a liquid-cooling setup where you can dissipate heat nicely and the cooling is bottlenecked by how fast you can transfer heat from the chip to the heatsink rather than by how fast you can dissipate heat.

 

[thatguythere]

As I have been looking for different kinds of cooling I have found out about thermoelectric cooling. The concept seems interesting, and inexpensive. Putting a pad in-between a processor and a fanned heatsink sounds like a great idea.

However, I have never seen someone put one in a portable before.

Would it be beneficial to include one? Are they a viable cooling option for portables?

If no, what's the catch? Is there a drawback?

If yes, why hasn't anyone done it yet?

Like someone else had said, the main drawback I could see is it adding thickness but if you have the space to spare I don't see how it could be a bad idea? The only way new things become trends is to test it out. You dont know if things work unless you try it! I think it be pretty cool if it got implemented in some way.

 

[ttsgeb]

Thickness, power draw, and everything that groose said.

 

[vskid3]

Size is really the least bad thing about using a thermoelectric cooler in this application. To expand on what groose said, thermoelectric coolers move heat like an AC or fridge and can make the cold side cooler than ambient (vs just a heatsink that can only cool to about ambient). Unless you're doing extreme overclocking, most chips will run fine at 100-150F with a properly sized heatsink/fan. With a themoelectric cooler, the hot side heatsink not only has to dissipate the heat moved from the cold side, it also has to dissipate the heat generated by the process of moving the heat, which looks like its even higher than what it moves (which also means your energy consumption would be more than doubled). So you'll use twice as much power with a larger heatsink and probably about the same temp at the chip. Much better to cut out the middle man and just use a bigger/better heatsink/fan.

Themoelectric coolers have their place, portables aren't one of those places.

 

[Shank]

Thanks vskid3
That's exactly the kind of answer I was looking for.