water cooled heatpipe cpu cooler

A heat pipe works well because it STAYS warm and the fins dissipate enough heat to cool while still being above a certain temperature to allow the properties to work effec5ively. If the top 2/3 of the heatpipe remain below that it takes away some of the efficiency of it.

(In my opinion. I could make an argument from both sides though I think, lol)

I could care less if it actually works at this point, pop an led or two in it and it would look cool in the case!!!!

I'm curious to see if this actually works....though, I fear it won't fare well when compared to a good CPU block under heavy, overclocked loads.

It will be interesting to see if the UK mag takes OP up on his plan to get it tested.

The acrylic looks nice, doesn't it? If the Water lavaging the heatpipes keeps them too cool to act as heatpipes, they will warm up their working fluid until they do act like heatpipes. Or if the water still keeps them cool, soon enough the water will warm up. How do you keep it cool? With a rad, I think.

This is something I would like to test out. Perhaps when the UK mag is done?

ehume said:

It will be interesting to see if the UK mag takes OP up on his plan to get it tested.

The acrylic looks nice, doesn't it? If the Water lavaging the heatpipes keeps them too cool to act as heatpipes, they will warm up their working fluid until they do act like heatpipes. Or if the water still keeps them cool, soon enough the water will warm up. How do you keep it cool? With a rad, I think.

This is something I would like to test out. Perhaps when the UK mag is done?

Click to expand...

You make a good point ed... for some reason I forgot the water would heat up (say with a 2x120 rad) so maybe it would still work...well.

I'm with velly though, I think it wont do as well as other products with higher loads.

This gives me a chance to flex my nerd muscles. Leaving the fins on is what thermaltake did with the rx2 ufo. I think it's a bad idea because that would require a larger chamber to house them. This means the flow would reduce quite dramatically, possibly allowing a thermal barrier layer to develope, insulating the cooling area. The water heated by the pipes wouldn't be carried away as quick meaning they would be surrounded by warmer water. Heat would then not be transferred to water as quickly. In my design the pipes are each housed in individual tubes not much wider than the pipes themselves ( pipe d:6mm, tube d:7.5mm ).
As long as it doesn't go below zero ( water freezes at zero even in the near vacuum inside a heatpipe ), the colder the condensor is the quicker it will condense the water inside the heatpipe and return it to the evaporator.
The intention is to put it in a regular loop with a radiator, reservoir and pump.

Two new videos:
Hope you like em.

Hey Rob,

I like the upgraded fittings makes it look much cleaner. As for load testing try running Prime 95 small fft or Intel Burn Test, either one will really push the cpu to it's max heat generation and give you a better idea of how the cooler is performing.

So... temps? Have you tested it with a load? Something considerable as in not an old dual core amd?

What's the deal with testing at the place you may?

It's time for some substance.

idle temps are fine and all but don't really show anything about the cooling ability of the heatsink, please show load temps

Don't really know much about benchmarking n I could really do with some detailed instructions n the software I need.
Anyway with my limited knowledge I've tried to do a test:
I'd be grateful if you guys could tell me what the numbers mean.

Well I hate to be "that guy" but I don't really see a scenario where this would be beneficial. You guys need to remember, heatpipes may be a good way to get heat from point A to point B, but in this scenario it is an extra step. We've discovered long ago in the WC world that you get the best performance out of relatively thin WC block bases. Big fins, pins, thick bases, "cold plates" (copper slabs / shims / heat spreaders) have all been tested and don't offer any benefit.

The goal here is to get the CPU heat from the CPU, into the water. As fast as possible. And the most efficient way to do that is through a couple MM of copper. In this case, you still need to go through a few millimeters of metal, but there's more going on too. It'd be kind of like thinking that if you were to drive a generator with an electric motor, you'd end up with more electricity in the end than what you started with. You can't get anything for free, and that includes heat loss. We already have very efficent WB designs.

The acrylic work is impressive though! And I'm still interested in results on how this works. I don't think it'll be ground breaking though, my guess is weaker performance than any WB on the market. I hate to be rude, but that is simply where my best judgement points me. I definitely appreciate the ideas and craftsmanship that went into this thing. I'd like to see those efforts applied into the construction of a custom built AIR cooler, where there may be some room for improvement.

I'm still very curious. I see what you mean ben, but at the same time It may work better than one might think because it has a larger surface area to make contact with the water. All those heat pipes touch water.

I'd love to see this on an FX 6300 or something! Rob, did you ever hear back from that magazine about testing it at all??
I'm sure we can do an in house test if they don't want to

Rob,

I watched your video and thanks for the update. I have included a link below that talks about different programs used to stress a cpu. You may want to give it a read to get a better handle on how to stress the cpu temps.

http://www.overclockers.com/forums/...Testing-for-Beginners-Which-tool-should-I-use


Another thing you may want to look at is using a different utility to monitor temps. Realtemp can give faulty readings and based on the 22c it is reporting I bet it is not accurate. As for the ASRock utility I'm not real sure how accurate it is. Either way try using hwinfo which I included a link for below. When you start it, a pop up will appear and tell it to run sensors only and it should give you a little more accurate info, if nothing else it will at least confirm the temps the ASRock utility is reporting.


http://www.hwinfo.com/download.php

Realtemp is for Intel anyway I thought? He is using an old dual core amd. Coretemp was the thing back then. HW info should be good too.

So did that website not want to test for you or....


I would imagine the offer from our guy is still in the table?

EarthDog said:

Realtemp is for Intel anyway I thought? He is using an old dual core amd. Coretemp was the thing back then. HW info should be good too.

So did that website not want to test for you or....


I would imagine the offer from our guy is still in the table?

Click to expand...

You got me to thinking and I went back to watch the video again and he is using core temp and not realtemp. Either way if it is reporting 22c it is probably wrong.

It was idle though..., no?

Edit.. it wasn't I see the new video and p95. Coretemp is reading wrong. The software is showing 30c. But it's water so it should be tested for at least an hour to allow the loop to warm up. We also don't know what the loop consists of rad wise, do we?

Well in his earlier video he had it hooked into a 240 rad so I assume that is what it is hooked to. He did run the test for an entire hour and got a peak temp of 35 according to the ASRock utility. I'm wondering if he ran small fft our blend. He said he left it on the default but I cannot remember what it defaults to.

I wonder if he just ran the application and not the stress tests....

It says stress test started in the prime95 window.

Those dual core AIIs ran pretty cool, much cooler than any modern chip. You should be able to put a Phenom II processor in that board. What's it's max TDP rating?