|
|
Xtreme Cooling LN2, Dry Ice, Peltiers, etc... All the usual suspects |
Thread Tools |
07-02-2003, 03:57 PM | #1 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
My Heatpipe / Radiator Design
Just 3 or 4 heatpipes made into a radiator with external and internal fins. Connects at 45 degrees with the cpu block so it works in both tower and desktop configurations. The fan also blows the air directly at the power supply, where the warmed air can exit.
Any feedback/suggestions welcome. |
07-02-2003, 04:04 PM | #2 |
Cooling Savant
Join Date: Apr 2003
Location: Blackburn / Dundee
Posts: 451
|
I thought the best thing about warer cooling was thr idea you could transfer thr heat AWAY from the CPU to cool it elsewhere. you might as well just put a big ass heatsink with a big ass fan on and you'll get the same effect.
|
07-02-2003, 04:12 PM | #3 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
Well heat pipe absorbs heat more effectively than a plain heatsink/fan. This allows a large quieter fan to be used.
|
07-02-2003, 04:25 PM | #4 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
Thats going to have to be one tiny ass heat pipe. 1" long? Thats how much room I have from my CPU to the bottom of my Power supply in my case. And your just going to be blowing hot air on the bottom of the power supply and probably recirculating it through the heat pipe.
|
07-02-2003, 04:27 PM | #5 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
It's not to scale... guess I should have mentioned. It will probably be a 120mm fan. Figure sizes from that. Also, the whole thing, finned part included is the heat pipe, the blue is just the liquid filled section.
Hmm, there should be a thermal break between the sink and rad. Maybe a piece of less conductive pipe joining them? Or a foam rubber gasketted pipe joint? |
07-02-2003, 05:33 PM | #6 | |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
Quote:
The grey in the modified pic below represents the power supply if it were in a case. |
|
07-02-2003, 05:34 PM | #7 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
...
|
07-02-2003, 08:32 PM | #8 |
Cooling Savant
Join Date: Jun 2003
Location: Miami, Florida US
Posts: 117
|
I think a very similar heatpipe design has already been done by AeroCool, take a look here
Now.. if i could get rid of those alu fins on the AeroCool Deep Impact, flaten the heat pipe somewhat, and then clamp two (2) brand new cascade water blocks, one on each side, I think I would get something really interesting.. hmm...
__________________
I stop for 1 C. "Those that give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -- Benjamin Franklin (1773) Last edited by iggiebee; 07-02-2003 at 08:57 PM. |
07-02-2003, 08:36 PM | #9 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
Do you know of any good scale diagrams of a standard atx chassis with mobo and power supply in it?
|
07-02-2003, 09:22 PM | #10 | |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
Quote:
Anyway here is a pic of a semi modern system: See were the water block is and were the power supply is. There is about 1"-2" of space between the CPU and the bottom of the power supply. Not enough for a 120mm fan. The heat pipe needs to come strait out like the one iggiebee posted above. |
|
07-02-2003, 10:28 PM | #11 |
Cooling Neophyte
Join Date: May 2003
Location: Montreal, QC
Posts: 14
|
Well, to ignore the limitations of ATX specs for a second, the design is sound. It's technically a thermosyphon, and not a heatpipe. A heatpipe by definition has a wicking material on the inside to pump the fluid, a thermosyphon only needs gravity.
Your thermosyphons don't need any fins on the inside, either, Condensation and Boiling are the most effective forms of heat transfer, and fins won't help. Also, thermosyphons and heat pipes need very little fluid to work. Basically you want to coat the entire inside of your pipe, and a few drops more. Trying to boil a whole pool of fluid as shown in your diagram reduces the efficiency. You'd need some sort of texture or coating over the cpu that would remain wet, and you'd direct the condensed fluid to drip down onto that to keep it wet as the fluid boils off of it. Something like a wire screen or even a piece of cloth. Well, at least for a vertical CPU. If the CPU is horizontal, you'd just need to make sure it's level. And since I've already said screw ATX, why don't we get rid of the fan entirely. Just make the Thermosyphons nice and long, with vertical fins, and you'll have enough surface area for Passive convection to cool the whole thing down. |
07-02-2003, 10:54 PM | #12 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
How long can an 1/2" ID passive heatpipe be? Thinking of running a heatpipe from cpu out the back or top of case. Is it possible to make separate evap and liquid return lines? Maybe a tiny pump that turns on when the fluid in a small tank reaches a certain level. A one way valve keeps the fluid moving from tank to sink. A 120mm power supply fan is modded in to serve two functions, reducing the number of fans needed. Would natural flow dynamics eliminate the need for a pump?
Ok, I admit it's complicated.. but i don't really know how to use the capillary return stuff.. Please, trash/flame away Last edited by Seyeklopz; 07-02-2003 at 11:01 PM. |
07-03-2003, 12:17 AM | #13 |
Cooling Savant
Join Date: Jun 2003
Location: Miami, Florida US
Posts: 117
|
Building a heatpipe is not a small undertaking, as I understand it for a heatpipe to work needs to be evacuated, that means all air has to be taken out and create a vacuum. This you will find rather difficult to achieve as you may need to freeze the entire assembly, in order to create the vacuum without having all the water boil off.
Also the pump is not necessary, and will interfere with the mechanics of the heatpipe. As I understand it the boiling and condensation process in a heatpipe can take place at supersonic speeds, that's why heatpipes are up to 1000 times more efficient transferring heat than any known metal. If you would like to see the experiments of someone who tried to build some heatpipes, you can do so here
__________________
I stop for 1 C. "Those that give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -- Benjamin Franklin (1773) Last edited by iggiebee; 07-03-2003 at 12:22 AM. |
07-03-2003, 10:34 AM | #14 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
Wow, thanks for the link. Great article!
Now to improve on his methods... Need some sort of sensor to detect liquid in the pipe, maybe a studfinder? A better fitting for adjusting pressure would help as well. I think I would use water. Put some in, then just freeze the water, vacuum the air, and seal it. Last edited by Seyeklopz; 07-03-2003 at 10:44 AM. |
07-03-2003, 12:34 PM | #15 |
Thermophile
Join Date: Jun 2001
Location: The deserts of Tucson, Az
Posts: 1,264
|
To get the air out ghetto style:
Boil the water inside as hard as you can. use a torch if possible. Then cap it. The gas inside will be 99% vapor anyway if you heat it hot enough. Sort of like a boiling tea kettle that blasts out a plumb of vapor. Also, why use a homemade sink? for testing I'd just use a heatercore. Cheap and easy. |
07-03-2003, 12:37 PM | #16 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
I just bought an all copper heatercore from a 1961 mercury comet on ebay. It's pretty
I'm working on a watercool system, with individual heat pipe waterblocks... |
07-03-2003, 01:09 PM | #17 |
Cooling Neophyte
Join Date: Jun 2003
Location: Saint John, New Brunswick, Canada
Posts: 94
|
Well, here's a small heatpipe water block. 27 individual small diameter heatpipes attached to a copper block. The upper ends are inside a waterblock with [3] 1/2" inputs and outputs. A final design would have some sort of manifold, or a round pipe to rectangular adapter. I'd like to streamline this design. Could this sort of heatpipe, waterblock combination be any more effective than a normal waterblock?
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|