|
|
Xtreme Cooling LN2, Dry Ice, Peltiers, etc... All the usual suspects |
Thread Tools |
01-06-2003, 10:39 PM | #26 |
Cooling Savant
Join Date: Dec 2002
Location: Dallas
Posts: 339
|
I dont know I kind of like the bluecooling design
|
01-12-2003, 05:42 AM | #27 |
Cooling Savant
Join Date: Feb 2002
Location: Oxford University, UK
Posts: 452
|
I'm sorry, but this thing is just badly designed, plain and simple.
In an ideal set up, you want a coldplate in contact with the cpu which is then clamped to the block with a peltier sandwiched in between. In order for the peltier-block setup to be efficient, you need two things. You need a cold plate which is completely insulated save from contact with the cpu. That way, the peltier is only having to transfer heat from the cpu to the water block portion. Secondly, you need a water block which can remove the heat from the peltier and pass it into the water. Ok, to look at the cold plate. This is the bit that scares me. Firstly, look at the picture linked below http://www.hardocp.com/image.html?im...ZfMV8yX2wuanBn The portion which is in contact with Kyles hand is the cold plate. In order from the device to only be extracting heat from the cpu, this whole finned section will need to be insulated. I can understand why they have built this into the design. It's the part about cooling system failure. They are trying to market it in such a way that, if the peltier fails, then you will have this heatsink portion clamped to the cpu which should protect the processor. However, you have to remember that heatsinks work both ways. They are designed to maximise the area between the heatsink and the air flowing through it in order to increase heat transfer. However, if this heatsink is on the cold side of a peltier, the heat is gonna be flowing from the air into the heatsink. Also, it will be so cold, that the moisture in the air will condense onto the fins releasing large quantities of latent heat. So the peltier is now having to move the heat from the processor, the heat fro the surrounding air as well as the latent heat from all of the moisture condensing and freezing onto the fins. I'm sure you all know the way peltiers work where they provide a maximum temperature difference for zero heat moved and maximum heat moved for zero temperature difference. Therefore, because the peltier is having to move so much heat, it will be doing so with a low delta T. This is not going to be keeping your processor cool. Then there is the matter of all this ice and water building up on the fins. What's gonna happen to this when you switch the computer off or you start really using your processor. Bye Bye video card, and anything else it drips on. Secondly, the water block. Companies like swiftech have spent a lot of time perfecting there waterblock designs to efficiently remove heat from a processor, and it is no different for removing heat from a peltier. So why should you need such a different block design. I'll let you think about this. 8-ball |
01-12-2003, 04:11 PM | #28 | ||
Cooling Savant
Join Date: Nov 2002
Location: in my chair
Posts: 574
|
Wow... all that and never tested it. Your amazing!
Quote:
As per your words.. it must be an ideal setup.. because every thing you just said.. it has (A), except for the extended fins. It also have B which is the unit above the coldplate which rests on a peltier when it is installed in the system. The unit you are referring to was not designed to work without one. Running this unit without a TEC would be like running a car without gas. It wasn't designed to work without it. Per the technical designer and tester of BC the manufacturer. They have never had one freeze up when mounted and used within designed specs, and he referred me to thispicture. He pointed out that the unit has no mounting screws holding it down to the CPU on the [H] article. And the system has no power on the board (far right of picture). Both or one of these will always cause condensation when working with a TEC, any TEC of capable power. In order to have had that much condensation, you would have had to run the unit 10 min (w/85w pelt) in a very humid environment with the cpu not even powered on, or the block left unscrewed to the powered cpu. This apparently was the case. I will test his conclusions when I purchase one when they are for sale. The unit comes with an 85 watt pelt. The only way to get one different is to buy it seperately and add it yourself. Quote:
__________________
-winewood- |
||
01-12-2003, 05:26 PM | #29 |
Cooling Savant
Join Date: Dec 2002
Location: Dallas
Posts: 339
|
This unit is a cross between a heat sink/water block / peltier cooler. Putting all these three together to come with a new configuration makes this unit a hybrid of some type. So is it fair to keep calling it a water block any one else know of anyone doing this? what are these type of units called if in fact they are out there?. Since there may not be other units using all three of these concepts then the static principles to each specific type may not apply.
|
01-12-2003, 06:57 PM | #30 |
Cooling Savant
Join Date: Feb 2002
Location: Oxford University, UK
Posts: 452
|
Winewood, I have to agree, after close examination, that the atx power molex is not plugged into the header on the board. It is possible that this has lead to the icing shown in the picture.
That was the last nail in the [H] coffin as far as I'm concerned!!! However, I am still skeptical of the design. If you do get one of these, then post the results. Maybe then I will reconsider. 8-ball |
01-12-2003, 07:06 PM | #31 |
Cooling Savant
Join Date: Nov 2002
Location: in my chair
Posts: 574
|
You have good reason to be hesitant 8-ball. Im waiting for them to open. I will post you some data. The only thing im actually scared of is not being able to provide the technical specs in a valid way for these gurus in the room. I appreciate your evaluation.
__________________
-winewood- |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|