"the ultimate waterblock"

[Etacovda]

http://forums.bit-tech.net/showthread.php?t=62162

Or rather, 'lets use big pelts'!

I'd be intrigued to know what you fellas think of this creation....

[firtol88]

An Italian copy of the MCW5002-PT with a controller card.

[Etacovda]

To the point

[HAL-9000]

You guys are right about the MCW-5000PT similarity. The baseplate looks almost like they just bought one from Swiftech.

I like the idea of the controller card though. Peltiers have such markedly different charectristics depending on the Dt and Dc temps, and the DT between them, that optimization is difficult. If you buy a 226W QMAX TEC, and even if you run it at perfect voltage and current, the QMAX number only applies if the DT is identical to what the manufacturer tested for QMAX. So being able to tweak those things based on temps on the fly should really help with the pelt's biggest weak spot, which is the efficiency of the thing.

[qbert95]

I also like the controller card idea. As much as i would like to go with a pelt the whole condesation thing is a big turn off for me and this looks like a great solution.

[SlaterSpeed]

I like there 'lets see where the bubbles go' way of testing flowrate through the differant channles. Im not convinced

The 2 guys that were posting that stuff sounded like they were just sales people to me. I mean the only reason they posted the thing is to advertise it! They didnt seme to like answering my questions much. Would like to hear from the attual designer. Theres several things about it that just made me think 'why the hell have you done it like that!?'

slater..

[AngryAlpaca]

So it won't go below (or much below) ambient... Anyway, that block is less symmetrical with the barbs at the same side, is it not? Also, in those bubbles pictures, much more water was moving through the close ones than the far ones, hence the more bubbles at the far end... Nearing the end the visual goes away (as the air eventually disappears) but the flow remains the same.

[SlaterSpeed]

Originally Posted by AngryAlpaca

So it won't go below (or much below) ambient... Anyway, that block is less symmetrical with the barbs at the same side, is it not? Also, in those bubbles pictures, much more water was moving through the close ones than the far ones, hence the more bubbles at the far end... Nearing the end the visual goes away (as the air eventually disappears) but the flow remains the same.

Well thats exactly what i thought when i first saw it. I cant see any logical reason to have it like that :shrug:

[BillA]

lack of resolution, poor procedures
if you never 'see' it, you'll never 'miss' it

a 226W TEC is so big that it hides the inefficiencies,
needs a better wb to shine

a greater question is the need for such a controller
the enthusiast will run flat out and deal with the potential for condensation

I'm thinking the whole idea may be flawed,
it is the cold plate which can become the coldest therefore that is where the temp should be sensed/controlled
by my calcs that's ~ 10° above ambient for the cpu, better than WCing - but worth the effort ?

[Etacovda]

Hmm....
theoretically, with this product, if the whole unit was designed better, with the controller card, could you drop the pelt to very low power levels, with little ineffects?

Its a nice way to actually USE a pelt, rather than having a huge power bill... If it could be optimized for temperature/cpu load, stop condensation, and keep the system very very cool only when its needed.

Something that uses basically no power at idle, and ups the power quickly when under load (w/o instability) and only uses the power when it needs could be a very good product... I mean, its no harder than installing a waterblock, and you dont have to deal with condensation etc. The idea is good, I think, perhaps the implementation is not so.

Perhaps im missing something vital...

[BillA]

would be tricky
a TEC responds FAR slower than a CPU
Intel does NOT recommend cooling devices 'controlled' by the CPU heat for this reason

[Etacovda]

Yes, this is very true; obviously a pelt cant take a 5-10 degree change in less than a second.

[jaydee]

Hummm, that's not "ultimate" anything. Disappointing actually. Better to use some kind of phase change for quicker temp changes for better themral control I would think. IMO TEC's should be run flat out or not at all but that's just me.

[jlrii]

If your concerned about efficiency at all TECs should NOT be run flat out. But then again it all depends on what you are trying to achieve. If you want remps that are not way below ambient and want to reuse your current WC gear minus block the option is not so bad. If you want all out, balls to the wall, frost on your MB cooling it would be wise to consider phase change as it gets real expensive real fast with TECs (especially considering running cost and added heat) to achieve that sort of cooling.

[SlaterSpeed]

I think TECs have had there day. If you attualy take into account all the costs (including bigger rads etc.)not to mention the powerbills you may aswell buy a low end vapochill.

Add this controler card into the equation and you must be looking at it attualy costing more than a vapochill ?

[qbert95]

I wish vapochill will sell a lower end phase change for us lpoor folk

[AngryAlpaca]

TEC's did have their day, when a 40W one would get you to -10... You could run that off of a cheap PSU or maybe even your computer PSU... When Dothan gets out or when VIA gets their shit together ($200 case?!), then, maybe, TEC's will have their time again.