Notices

Overclockers Forums > Hardware > Cooling > Extreme Cooling
Extreme Cooling Below ambient cooling is extreme. The crazies discuss chillers, phase change, dry ice, & liquid nitrogen in here...
Forum Jump

Need help on extreme project

Post Reply New Thread Subscribe Search this Thread
 
 
Thread Tools
Old 10-07-02, 11:29 PM Thread Starter   #1
Caffinehog
Übercaffinated Member

 
Caffinehog's Avatar 

Join Date: Dec 2001
Location: In the lab

 
Need help on extreme project


Ok, here it is: I'm planning to build a hybrid cooler. I plan to use two 220W peltiers (or maybe more, smaller peltiers,) to chill a reservoir containing alcohol. This will be pumped to the CPU in a typical watercooling fasion. I need help with the following:

1. How do I control the temperature so that it does not get cold enough to freeze the alcohol?

2. How will I cool the peltiers without making enough noise to wake the entire block?

Why am I doing this? It should actually be better than Vapochill at a fraction of the cost.
Caffinehog is offline   QUOTE Thanks
Old 10-07-02, 11:41 PM   #2
LittlePiggie
Member

 
LittlePiggie's Avatar 

Join Date: Sep 2002
Location: Oregon

 
I don't think you should have problems with freezing your alchohol if you are circulating the water through the waterblock.

A suggestion for cooling the water: Use a reseviour(Spell?) and attach heatsinks to the cold sides of two 226w pelts. Put the heasinks in the side(or top if you can). Just a suggestion.
LittlePiggie is offline   QUOTE Thanks
Old 10-08-02, 12:00 AM Thread Starter   #3
Caffinehog
Übercaffinated Member

 
Caffinehog's Avatar 

Join Date: Dec 2001
Location: In the lab

 
Uh, I am not planning on using water! I am planning on pumping the methanol directly throught the waterblock!
Caffinehog is offline   QUOTE Thanks
Old 10-08-02, 02:55 AM Thread Starter   #4
Caffinehog
Übercaffinated Member

 
Caffinehog's Avatar 

Join Date: Dec 2001
Location: In the lab

 
The thing that interests me about alcohol cooling is this:
Methanol is a liquid down to -97C. Ethanol to -117C. These substances have about half the heat capacity of water, making them far better at carrying away heat than most substances. Phase change, while capable of reaching lower temperatures under light load, has a pathetic capacity to carry away heat.
(Here is an example of heat capacity: One calorie, or 4.19joules, of heat will raise the temperature of one gram of water by 1C. The same amount of heat will raise the temperature of one gram of copper by 11C. Or 1g ethanol by 1.7C.)
Add to this the fact that a phase-change system is only passing maybe 10 grams per second, while a pump going at 150gph is passing in excess of 115 grams per second.
When all is considered, the ability of such a system to cool a processor is limited only by the wattage and temperature differential of the peltier units.
Caffinehog is offline   QUOTE Thanks
Old 10-08-02, 03:03 AM   #5
beezee
Member



Join Date: Aug 2002

 
use antifreeze solution?
beezee is offline   QUOTE Thanks
Old 10-08-02, 05:49 AM   #6
strokeside
Member

 
strokeside's Avatar 

Join Date: Jan 2002
Location: Dublin, Ireland

 
Instead of spending the money to buy the pelts, the power supplies and blocks to place in the resevoir, why not buy two 120mm fans and an old freezer, use the fans to cool the condensor of the freezer (to allow you to run it continuously) and place the cold plate of the freezer into the resevoir. Depending on the power of freezer, you could get yourself some pretty good resevoir temps.

Also, using pelts as chillers only works good if you have many pelts and lots of surface area for the liquid to come in contact with.
strokeside is offline   QUOTE Thanks
Old 10-08-02, 09:05 PM   #7
LittlePiggie
Member

 
LittlePiggie's Avatar 

Join Date: Sep 2002
Location: Oregon

 
Quote:
Originally posted by LittlePiggie
I don't think you should have problems with freezing your alchohol if you are circulating the water through the waterblock.

A suggestion for cooling the water: Use a reseviour(Spell?) and attach heatsinks to the cold sides of two 226w pelts. Put the heasinks in the side(or top if you can). Just a suggestion.
I know, I meant if you circulate the alchohol, you shouldn't have probs cause the heat will warm it up enough to keep it from freezin.
LittlePiggie is offline   QUOTE Thanks
Old 10-08-02, 09:22 PM   #8
TLKfatman
Member

 
TLKfatman's Avatar 

Join Date: Sep 2002
Location: Concord NC

 
what is the boiling point of Methanol and Ethanol

if at the cpu it reaches this temp u are gunna be in big truble when it expands 750 x its size and over presures your cooler

other than that i think it would be cool
TLKfatman is offline   QUOTE Thanks
Old 10-08-02, 09:57 PM   #9
JFettig
Hey! I showered! Senior

 
JFettig's Avatar 

Join Date: Jan 2002
Location: MN

 
i suggest you make 2 blocks the size of a hdd block, 4x6inches, then run 6 80watt peltiers, and watercool the hot side, just run it thru that and the heatercore with a decent sized pump

__________________
||http://www.wc101.com - Watercooling101||
||http://jfettigmachines.com CNC kits||
JFettig is offline   QUOTE Thanks
Old 10-09-02, 12:47 AM   #10
Since87
Member



Join Date: Jul 2002
Location: Indiana

 
Quote:
Originally posted by maskedgeek
i suggest you make 2 blocks the size of a hdd block, 4x6inches, then run 6 80watt peltiers, and watercool the hot side, just run it thru that and the heatercore with a decent sized pump
I think it would make a lot more sense to use four 226 Watt pelts powered at 6V than 6 80 Watt pelts powered at 12V.

(You'd wire pairs of 226 Watt pelts in series across a 12V supply to get 6V to each pelt.)

For six 80 Watt pelts (12V across each pelt):
Total current draw from 12V rail = 33.5 Amps
Power dissipation in the pelts = 402.4 Watts
Cooling (Q) = 335.4 Watts
Price (Danger Den) = $144
Operating close to max operating point.

For four 226 Watt pelts (6V across each pelt):
Total current draw from 12V rail = 17.1 Amps
Power dissipation in the pelts = 205.2 Watts
Cooling (Q) = 322 Watts
Price (Danger Den) = $152
Plenty of room to increase cooling by increasing pelt voltage.
Since87 is offline   QUOTE Thanks
Old 10-09-02, 01:51 AM   #11
aenigma
Member



Join Date: Jul 2002

 
Quote:
Originally posted by Caffinehog
The thing that interests me about alcohol cooling is this:
Methanol is a liquid down to -97C. Ethanol to -117C. These substances have about half the heat capacity of water, making them far better at carrying away heat than most substances. Phase change, while capable of reaching lower temperatures under light load, has a pathetic capacity to carry away heat.
(Here is an example of heat capacity: One calorie, or 4.19joules, of heat will raise the temperature of one gram of water by 1C. The same amount of heat will raise the temperature of one gram of copper by 11C. Or 1g ethanol by 1.7C.)
Add to this the fact that a phase-change system is only passing maybe 10 grams per second, while a pump going at 150gph is passing in excess of 115 grams per second.
When all is considered, the ability of such a system to cool a processor is limited only by the wattage and temperature differential of the peltier units.
I suggest you learn about phase change....
For instance, how is 13,500BTU a light load?Air conditioners dont handle "light loads" and they dont have a pathetic capacity either.

Pelt chillers are not worth the cost, they are very unneffeciant.

And I would rule out the methanol/ethanol boiling anytime soon
aenigma is offline   QUOTE Thanks
Old 10-09-02, 03:34 AM Thread Starter   #12
Caffinehog
Übercaffinated Member

 
Caffinehog's Avatar 

Join Date: Dec 2001
Location: In the lab

 
Quote:
Originally posted by aenigma


I suggest you learn about phase change....
For instance, how is 13,500BTU a light load?Air conditioners dont handle "light loads" and they dont have a pathetic capacity either.

Pelt chillers are not worth the cost, they are very unneffeciant.

And I would rule out the methanol/ethanol boiling anytime soon
OK, that's a lot of joules! But...

It's notable that vapochill units are only capable of moving 130W or 160W depending on the model. While substantial, it pales in comparison to what pelt-chilled alcohol can achieve. I'm just wondering if I can achieve similar results with a compressor and how much a capable compressor would cost. Notably, walk-in freezers are only capable of moving about 700-1000W of heat at their normal operating range. Air conditioners are capable of far more efficent transfer of heat because the difference between the outside and inside air temperature is small. While the same holds true for peltiers, this can be overcome by adding more to spread the load. Adding another compressor is not a cheap or easy proposition.
Caffinehog is offline   QUOTE Thanks
Old 10-09-02, 02:57 PM   #13
JFettig
Hey! I showered! Senior

 
JFettig's Avatar 

Join Date: Jan 2002
Location: MN

 
this absolutely makes no sence, it would be the same as having 2 226 watt tecs wich would handle 452watts, of heat, and put off like 800watts total, your calculations make no sence to me, explain
and there is no way you can use less wattage than what is being handled not possible

Quote:
Originally posted by Since87


I think it would make a lot more sense to use four 226 Watt pelts powered at 6V than 6 80 Watt pelts powered at 12V.

(You'd wire pairs of 226 Watt pelts in series across a 12V supply to get 6V to each pelt.)

For six 80 Watt pelts (12V across each pelt):
Total current draw from 12V rail = 33.5 Amps
Power dissipation in the pelts = 402.4 Watts
Cooling (Q) = 335.4 Watts
Price (Danger Den) = $144
Operating close to max operating point.

For four 226 Watt pelts (6V across each pelt):
Total current draw from 12V rail = 17.1 Amps
Power dissipation in the pelts = 205.2 Watts
Cooling (Q) = 322 Watts
Price (Danger Den) = $152
Plenty of room to increase cooling by increasing pelt voltage.

__________________
||http://www.wc101.com - Watercooling101||
||http://jfettigmachines.com CNC kits||
JFettig is offline   QUOTE Thanks
Old 10-09-02, 03:48 PM   #14
aenigma
Member



Join Date: Jul 2002

 
Quote:
Originally posted by Caffinehog


OK, that's a lot of joules! But...

It's notable that vapochill units are only capable of moving 130W or 160W depending on the model. While substantial, it pales in comparison to what pelt-chilled alcohol can achieve. I'm just wondering if I can achieve similar results with a compressor and how much a capable compressor would cost. Notably, walk-in freezers are only capable of moving about 700-1000W of heat at their normal operating range. Air conditioners are capable of far more efficent transfer of heat because the difference between the outside and inside air temperature is small. While the same holds true for peltiers, this can be overcome by adding more to spread the load. Adding another compressor is not a cheap or easy proposition.
That is because vapochill units are absolute garbage.They are a disgrace.Puny little 12v compressors.Might as well use a friggin pelt!

But pelts are terribly uneffeciant.
Getting a compressor that would own a vapochil and/or a pelt is VERY easy.Just use a refrigerator/freezer/air conditioner etc. compressor.(free if you know where to look)
My current system has a 1hp low stage and a 3/4hp high stage, both a/c compressor (13,500btu and 8000btu).
aenigma is offline   QUOTE Thanks
Old 10-09-02, 03:52 PM   #15
JFettig
Hey! I showered! Senior

 
JFettig's Avatar 

Join Date: Jan 2002
Location: MN

 
ok heres what my thoughts are over all

phaze change systems are huge
peltier systems can be huge if you go cheap and use a bunch of atx power supplies
peltier systems can be small
peltier systems can cost a ton
phaze change systems can cost a ton
peltier systems probably use more energy because of efficiency

__________________
||http://www.wc101.com - Watercooling101||
||http://jfettigmachines.com CNC kits||
JFettig is offline   QUOTE Thanks
Old 10-09-02, 08:03 PM   #16
LittlePiggie
Member

 
LittlePiggie's Avatar 

Join Date: Sep 2002
Location: Oregon

 
I heard that the more undervolted pelts are, the more efficient they get. Like WAY efficient, so a bunch of 80w pelts at like 5v would be better than four 226pelts.
LittlePiggie is offline   QUOTE Thanks
Old 10-09-02, 10:32 PM   #17
Since87
Member



Join Date: Jul 2002
Location: Indiana

 
Quote:
Originally posted by maskedgeek
this absolutely makes no sence, it would be the same as having 2 226 watt tecs wich would handle 452watts, of heat, and put off like 800watts total, your calculations make no sence to me, explain
What would be like having 2 226 Watt TEC's? Having 4 226 Watt TEC's? I think you are the one who's not making sense.

Suppose you have four 1 Ohm resistors. You connect two in series to get a 2 Ohm resistor. Then you connect the other two in series to get another 2 Ohm resistor. Then you connect the two 2 Ohm resistors you've built in parallel to get a total resistance for the network of 1 Ohm. Suppose you connect this 1 Ohm network to 12V. What is the voltage across each of the individual 1 Ohm resistors?

Discussing the cooling effectiveness of pelts, without mention of their operating voltage, is meaningless. The cooling capacity of pelts is very nonlinear with respect to voltage or current.

My calculation are based on this paper: "Universal Thermoelectric Design Curves" by Richard J. Buist.

In this thread I have posted Excel spreadsheets based on the curves shown in that paper. I'm not going to go through all the math here. That's why I created the spreadsheet, so that I wouldn't have to work through all the math. If you like you can check the math in the spreadsheet versus the paper by Richard J. Buist.

Quote:
Originally posted by maskedgeek

and there is no way you can use less wattage than what is being handled not possible
Why would you think so? A pelt is a heat moving device. A fan is also a heat moving device. Fans are frequently used to continuously move 100 Watts of heat while only consuming a couple Watts. Why exactly is a pelt moving 332 Watts of heat while consuming 205 Watts impossible?
Since87 is offline   QUOTE Thanks
Old 10-09-02, 10:37 PM   #18
JFettig
Hey! I showered! Senior

 
JFettig's Avatar 

Join Date: Jan 2002
Location: MN

 
ok its making some sence..... i guess i havent been doing my studying... why dont they just make some 24watt 24amp 50x50mm peltiers and just run them at 12v and have them really efficient? or cant they make them that small?

__________________
||http://www.wc101.com - Watercooling101||
||http://jfettigmachines.com CNC kits||
JFettig is offline   QUOTE Thanks
Old 10-10-02, 12:01 AM   #19
Since87
Member



Join Date: Jul 2002
Location: Indiana

 
Quote:
Originally posted by maskedgeek
ok its making some sence..... i guess i havent been doing my studying... why dont they just make some 24watt 24amp 50x50mm peltiers and just run them at 12v and have them really efficient? or cant they make them that small?
I'm not sure what you are asking? Is it something like:

Why don't they make 24 Volt, 12 Amp pelts that you could run at 12V and get high efficiency?

If that is the question, then the answer is, "They do." The "Drift 0.8" pelt that DD sells has specs similar to this although it is 40mmx40mm. At 12V its Q (heat capacity) would only be 76 Watts though, which is not enough for a CPU.

You can only take advantage of the increased efficiency when the heat load is about 1/4 of the pelt's Qmax. Or in the case of a chiller, when you can increase the number of pelts to get the Q up while keeping the voltage low.
Since87 is offline   QUOTE Thanks
Old 10-10-02, 12:26 AM   #20
Since87
Member



Join Date: Jul 2002
Location: Indiana

 
Re: Need help on extreme project


Quote:
Originally posted by Caffinehog

1. How do I control the temperature so that it does not get cold enough to freeze the alcohol?
You can get a temperature controller like this.

Or you can design and build your own. (Not a simple project.)


Quote:
Originally posted by Caffinehog

2. How will I cool the peltiers without making enough noise to wake the entire block?
Attach the hot side of the pelts to a really long heatsink and then duct air down the length of the heatsink, or watercool the hot side of the pelts and cool the water with a big radiator, and big low RPM fan.
Since87 is offline   QUOTE Thanks

Post Reply New Thread Subscribe


Overclockers Forums > Hardware > Cooling > Extreme Cooling
Extreme Cooling Below ambient cooling is extreme. The crazies discuss chillers, phase change, dry ice, & liquid nitrogen in here...
Forum Jump

Thread Tools Search this Thread
Search this Thread:

Advanced Search


Mobile Skin
All times are GMT -5. The time now is 09:18 AM.
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2013, vBulletin Solutions, Inc.
You can add these icons by updating your profile information to include your Heatware ID, Benching Profile ID or your Folding/SETI profile ID. Edit your profile!
X

Welcome to Overclockers.com

Create your username to jump into the discussion!

New members like you have made this the best community on the Internet since 1998!


(4 digit year)

Why Join Us?

  • Share experience
  • Max out your hardware
  • Best forum members anywhere
  • Customized forum experience

Already a member?