Vapor Chamber Water Block Question

[Eastcoasthandle]

Is there such a thing as a vapor chamber water block? If so could it offer better cooling then the standard ones?

[subtec]

Is there such a thing as a vapor chamber water block?

Very much doubt it. Not in the PC-enthusiast water-cooling realm anyway.

If so could it offer better cooling then the standard ones?

Possibly. Vapor chambers are basically flat heat pipes and have a very high thermal conductivity. Conceivably a VC could be used to "spread out" the heat from the concentrated area of the die, allowing for a potentially much larger area for the copper-water heat transfer to occur. Imagine instead of a 30x30mm micro fin array a 80x80mm array on top of the VC. That's over a 7x increase.

[Eastcoasthandle]

Thanks for the reply. I would think it would be better then conventional means. But to actually see one just to see if it does may or may not happen any time soon .

[defect9]

i thought the sapphire vapor-x ATI cards used vapor chambers, as well as the GTX580. think flat heatpipe.

[subtec]

i thought the sapphire vapor-x ATI cards used vapor chambers, as well as the GTX580. think flat heatpipe.

Yes, vapor chambers have been used in video card heat sinks for a while now, but that's not what the OP is asking about:

vapor chamber water block?

[defect9]

ah, brainfart. for some reason was thinking of the liquid inside of the block, which is often very pure water.

now, of course, i'd love to take a GTX580 heatsink and turn it into a waterblock for another GPU.

[m0r7if3r]

I can't think of any reason that using a vapor chamber under your waterblock would hurt performance (other than the additional TIM joint). The one problem I do see is mounting, you could only extend as far as the mounting holes on the platform it was designed for (and if it's multiplatform...). Alternately, you could have one of those back-side mounting systems with studs appropriately placed on the vapor chamber to line up with mounting holes, but I've never been impressed with the pressure on those sort of monuts.

[Vapor]

I can't think of any reason that using a vapor chamber under your waterblock would hurt performance (other than the additional TIM joint). The one problem I do see is mounting, you could only extend as far as the mounting holes on the platform it was designed for (and if it's multiplatform...). Alternately, you could have one of those back-side mounting systems with studs appropriately placed on the vapor chamber to line up with mounting holes, but I've never been impressed with the pressure on those sort of monuts.

If that TIM joint were solder...not a huge bump in thermal restriction (relative to 'perfect' and it would be better than paste-based stuff) and mounting is solved

A VC would allow for block makers to not have to focus cooling at the center (where only so much surface area can be had before manufacturing costs and restriction limit it). Would be interesting and be able to utilize different internals, not sure if it would be better though.

[Gimmpy224]

Now this is a curious concept.

[Eastcoasthandle]

If I were to guess, I would think it's something not hard to implement. I wonder why we've not seen such a thing yet?

[MagisD]

Might be simple, a replace ment base for an existing block.

Sent from my X10a using Tapatalk

[subtec]

If I were to guess, I would think it's something not hard to implement. I wonder why we've not seen such a thing yet?

Maybe no one at the various WC companies has thought of it before... although it seems like someone would've. I'd given it some thought in the past, which is why I had a quick answer for you the other day - already thought through it a bit.

I think one of the barriers is that vapor chambers need to be custom designed and manufactured for a specific solution, and maybe CPU blocks don't have the volume necessary for the economics to work.

It's also possible the gain over a standard high-performance waterblock isn't significant enough to justify the added cost.

[Eastcoasthandle]

Maybe no one at the various WC companies has thought of it before... although it seems like someone would've. I'd given it some thought in the past, which is why I had a quick answer for you the other day - already thought through it a bit.

I think one of the barriers is that vapor chambers need to be custom designed and manufactured for a specific solution, and maybe CPU blocks don't have the volume necessary for the economics to work.

It's also possible the gain over a standard high-performance waterblock isn't significant enough to justify the added cost.

Interesting response that shows you've been thinking about it. But could you imagine the marketing potential if it does lower temps? Even if it's by 1/2 a degree...

[eXa]

My guess.
I think water beats the internal workings of a vapor chamber. And the vapor chamber wont have any more contact area than a normal water block(to the cpu).

Edit: If i had plenty of cash i would buy a true, cut the top of the heatpipes and run water through them, just to se what happend.

[NaeKuh]

guys here is my take on things.

First off IF the Vapor chamber would be better, then blocks should not be bowed.

A bow concentrates cooling in the center, while warping the block so it in sense clamps the IHS.

The vapor chamber would be adding 1 more thermal material OVER the heat source which already has a thermal material over.. IE.. the IHS.

Since the waterblock itself is already showing limited values, and only gets better as you focus the cooling directly at the heat source... i dont think a vapor chamber would help at all.

Your spread the heat out, when its already shown spreading it out doesnt help a water block.

So on a air sink with many many fins, and a cooling area which is dependant on spreading out... vs water channels which moves heat... i dont think vapor channels would work well in our models...

Unless it was a big vapor chamber... or series of vapor chambers.. and u had 1 tiny block to cool the entire plate.
OR... the IHS itself was a vapor channel.

[m0r7if3r]

NaeKuh, I agree with most of your post, I do think that a vapor chamber has a chance to improve performance, however. I don't feel like the IHS is an effective heatspreader (in relation to a vapor chamber) and if you had a vapor chamber with a bow in it (so it was taking the heat from the hottest part of the ihs) to a large water block you might be able to see gains (depending on the design etc). I think at the very least it'd be cool to see it tested.

[meanmoe]

I'm not sure if I see the benefit over a thin copper base. Essentially, we're talking about the equivalent of a well designed heat pipe air cooler with water passing through the fins instead of air. right?

if so, someone build an acrylic resevoir around a tuniq tower or something like that (I know jack about current aircoolers) and tell us how it works.

[Vapor]

I'm not sure if I see the benefit over a thin copper base. Essentially, we're talking about the equivalent of a well designed heat pipe air cooler with water passing through the fins instead of air. right?

if so, someone build an acrylic resevoir around a tuniq tower or something like that (I know jack about current aircoolers) and tell us how it works.

I think the thought behind it is that current blocks focus all of their cooling at such a small spot and increasing surface area at that spot is getting tougher and tougher--this could help make the target area larger. Just having an IHS made from something better than plain copper would go a long way to improving it as well...but that's Intel's fault

I don't think current block designs with a VC would be ideal, however. Maybe an Apogee XT though...hmmmm, maybe I'll play around with this some.

[m0r7if3r]

I'm not sure if I see the benefit over a thin copper base. Essentially, we're talking about the equivalent of a well designed heat pipe air cooler with water passing through the fins instead of air. right?

if so, someone build an acrylic resevoir around a tuniq tower or something like that (I know jack about current aircoolers) and tell us how it works.

I was envisioning something more like this

[nvmc]

Swiftech MCW6500-T with the TEC replaced by a vapor chamber.

[subtec]

Interesting article about this here.

[Church]

I'm not too familiar with TECs, but if one doesn't hunt for that extremely low temperatures they can give, will TECs still use that much extra power even if i want to get temps only moderately lower? Imho if it can be made in some more reasonable setting, like just extra 100W for 10-15 degrees less then conventional waterblock can do, then it might be worth it? Or TEC always should be more powerful then what is cooled by it?

[subtec]

I'm not too familiar with TECs, but if one doesn't hunt for that extremely low temperatures they can give, will TECs still use that much extra power even if i want to get temps only moderately lower? Imho if it can be made in some more reasonable setting, like just extra 100W for 10-15 degrees less then conventional waterblock can do, then it might be worth it? Or TEC always should be more powerful then what is cooled by it?

This is off topic, but google "tec calculator" and play around with it. It should give you a good idea of what TECs can do. Best coefficient of performance is around 30% of a TEC's rating. You need excess radiator capacity to make them even remotely worthwhile; i.e., if you can add more rad and still get lower CPU temps, TECs won't help you.

Personally, I don't think they are worth the trouble these days.

[BeepBeep2]

I'm not too familiar with TECs, but if one doesn't hunt for that extremely low temperatures they can give, will TECs still use that much extra power even if i want to get temps only moderately lower? Imho if it can be made in some more reasonable setting, like just extra 100W for 10-15 degrees less then conventional waterblock can do, then it might be worth it? Or TEC always should be more powerful then what is cooled by it?

Once TEC's are under a load heavier than what they can handle things start to heat up pretty quickly.

This was the problem with the CoolIT freezones, once you passed a certain load, the use of TEC cooling was worthless. Not only this, but to cool a ~200w load you would need a rather large 350-450w rated TEC (or two), and you would need to cool the backside of the TEC (upwards of 400w-500w of heat) as well with another cooling device.

Phase change iirc is more efficient than pelts...

[meanmoe]

Interesting article about this here.

Very nice. I love it when someone backs up an idea with some analysis... Note conclusion...

This article shows that while a vapor chamber presents exciting technology, some calculations should be made to justify its use. As shown above, in some situations a solid copper block might provide better thermal performance than a VC. ...snip...

Assuming their analysis, gains appear to be minimal.