The best of both worlds

[derelict]

So it seems in the past few years water cooling hasn't really progressed much in terms of performance, but air on the other hand has. I was thinking about combining the 2.

Has anyone tried taking a heatpipe tower and running water through the fins? Would performance be better than just the usual water block? Any thoughts?

[derelict]

I've thought about all kinds of air heatsinks (not just heatpipe towers) and wondered how they'd perform if immersed in water, or just enclosed and converted into a water block.

[billbartuska]

Originally Posted by derelict

I've thought about all kinds of air heatsinks (not just heatpipe towers) and wondered how they'd perform if immersed in water, or just enclosed and converted into a water block.

How about this!!!!
Series or parallel?

[ben333]

Silly bill (j/k) He is saying turn the heatsink into a water block. I say it will preform much worse than a good water block. I believe thats been proven here before - anyway people realized water blocks don't need to be giant to be effective.

[derelict]

yeah I don't mean cool a water block with a heatpipe tower, I mean run water through the fins of a heatpipe tower, or enclose a heatsink so water can go between all the fins not air.

[HammerSandwich]

Every extra interface increases the C/W.

[UnaClocker]

Yeah, but the phase change properties of a heat pipe SHOULD be far better than just water. It seems like it'd work well to me.. Someone just needs to try it.

[ben333]

Leaky leaky. But it would give us the opportunity to test five inch diameter hose and fittings

[ben333]

[Anarking]

I want to drop a hint here about watercooling that will boost performance on it and get that C/W ratio to 0.00 ... compression fittings. 5 years ago I started using them and figured everyone else would catch on, but I still see people using hose clamps!!! We didn't ask dangerden to start making radiators with threaded holes so that people could just put hose barbs back in them, we asked for the radiators to be manufactured that way so we could insert threaded compression fittings, and have that perfect, immovable, screwed-on seal in the case so they could actually be shipped around the country. The compression fittings also increased the thermal transfer properties a ton because of the pressure of the water coming down on the waterblock through the fitting, and I always get room temperature temps now no matter what, and have for years (hell it's easier now that chips aren't 95w). straight 0.00 C/W

Compression Fittings in action with a MAZE4


btw, good to see ya UnaClocker, all my watercooling started with gettin more and more info and then even that 1GHz AVIA Athlon chip and waterblock from you back in 2000 or 2001 or whenever it was. Got that puppy to 1.7, 1.63 stable Good times. Just wanted to say Hi.

Peace and Chilled Cooling All.

[derelict]

Now you're onto something ben333, thats kind of what I had in mind (its all pretty impractical I know). However crazy ideas spurn new, more feasible ones.

For testing one could get a bucket of water, get a mobo with a heatpipe tower, then flip it upside down and dip it in the bucket. a simple hose could move water through it for sim purposes. I don't know if having the HSF upside down would significantly effect its cooling.

@ Ben333 - such an arrangement could at least make it on your ghetto fab competition for water cooling.

[Brians256]

Originally Posted by Anarking

I want to drop a hint here about watercooling that will boost performance on it and get that C/W ratio to 0.00

Huh? You cannot achieve 0.00 C/W with straight water cooling. Laws of thermodynamics aren't negotiable. Heat won't "flow" unless there is a differential, and a 0.00 C/W is a zero differential boundary.

To reach down to 0.00 C/W or lower, you would have to have active cooling and take the C/W of the whole rather than a particular boundary. Active means Peltier, phase change or other heat pump.

0.0 C/W means that the CPU is at ambient temperature.

Not that compression fittings are bad, mind you. They do reduce the restriction that you normally find at a joint.

[Anarking]

Yes. Ambient temperature. All the time. Multiple systems all at ambient temperature doing things this way. Maze 4 waterblock, Black Ice radiator (much better than the heater cores from pintos I had to use back in 2000), reservoir up high and 350gph pump with 1/2" tubing compression fittings all around. Ambient temperature readings of CPU temps from the motherboard, even after overclocking. 0.00 C/W is possible with watercooling without having to use a peltier. which i've also done plenty of, but there's not much need unless doing extreme overclocking with high vmods. anyways, just wanna say it is possible. the compression fittings mainly work so well because of the increased flow at the waterblock and the pressure with which it's hitting the waterblock right above the core which makes for better thermal transfer between the copper into the water. it's sucking the heat away as quick as it can be made and staying at ambient temps. or i could be crazy and have been doing everything wrong and should never have had dangerden start manufacturing their radiators with threaded holes (it cracks me up that people then just put hose barbs back on). so do i need to be more technical and say 0.000001 C/W?

[ben333]

ask pH I'm sure he has a shit load of radiators and pumps for the upside down cooler in bucket idea.

[Brians256]

Originally Posted by Anarking

so do i need to be more technical and say 0.000001 C/W?

Yes.

[derelict]

what do you think pH? care to put the idea to rest (or life)?