Phase Change Cooling (newbie)

nhusby

Limp Gawd
Joined
May 1, 2006
Messages
303
Ok, to thwart any flaming: my research thus far is minimal, and I know only some principles behind phase change cooling. I am posting this thread to get input on my plans, and the answers can help to serve as a refrence to other newbie users like myself.

If I understand correctly, the basic components are:

1) compressor
2) condensor (high pressure side, aka heat exchanger)
3) condenser fan
4) regulator valve
5) evaporator core (low pressure side, aka heat exchanger)

did I miss anything?

and again, as I understand it, the compressor pumps a gas (freon I am assuming works best, although I have heard of CO2) from the low pressure side (heat exchanger) into the high pressure side (condenser aka radiator). here is where it might confuse someone, the act of compressing a gas actually heats the gas, but that is the desired effect. once the gas is compressed, it is then cooled to (as near as possible) room temperature. At this point the compressed gas is at near room temperature, so when it is released into the heat exchanger the temperature drops (because it is no longer compressed.) Because the high pressure gas is cooled to near room temperature, the low pressure gas is below room temperature.


I want to start off simple (and cheap) so I plan on building a phase cooler out of an old air conditioner.
basicly I intend to remove the evaporator core and fit the hoses with connectors for a water block. for a matter of space, I would remove the fan and replace it with something like an automotive electric 12v radiator fan. I could route the 12v power through the a/c fan control. I am also thinking of soldering the origional a/c thermistat to the water block.

alternatively, I was thinking of splicing the water block in between the evaporator and the compressor, and using the origional a/c function to cool the rest of the case & components (to avoid condensation)

any thoughts? good / bad idea? you are going to report me to the epa? :eek:
 

boshuter

Limp Gawd
Joined
Mar 25, 2003
Messages
379
Ok... at the risk of starting another flame thread :rolleyes: I'll try to help clear up some of the things you are a little off on.

The freon (in gas form) is compressed in the compressor which turns it into a liquid and adds heat, from there it goes to the condensor which cools the liquid to near room temp, from the condensor the liquid runs to a metering device (cap tube, tvx, cpev, etc.) then into the evaporator, the liquid referigerant boils inside the evaporator and turns back into a vapor , from there the vapor is returned to the compressor via the low side suction line. This is a VERY basic description of the loop ;)

As to your idea of removing the evaporator and just plumbing in a water block... not a good idea. The evaporator needs to be designed as an evaporator, a water block doesn not have enough restriction and the liquid referigerant will not be in the block long enough to boil off completely. You would run a very great risk of having liquid referigerant returning to the compressor which would most likely destroy your compressor.

A better idea is to use the existing evaporator and submerge it in a cooler to use as a water chiller. Or make or purchase an evap designed for direct die cpu/gpu cooling.
 

DFI Daishi

2[H]4U
Joined
Feb 14, 2005
Messages
2,839
on track and on target so far. unlike that previous thread this one has some specifics.

an additional consideration is that waterblocks do not have seals rated to handle the pressure differentials generated within a phase change loop.

an easier option is to mount the cool coils of the A-C unit into a water resevoir and run an otherwise normal water loop with condensation proofing and insulation from that. look up "water chillers" for more info. automotive antifreeze should be able to keep the coils of all but the most powerfull A-C units from icing up too badly, depending on the mix ratio you use.

for extreme low temperature opperation, i have been pointed to acetone in a closed loop because it keeps a viscosity similar to water down to about -30C, whereas water-antifreeze mixtures get pretty thick at those temps.

removing some of the fins from the cold coils migh help performance some, but you run the risk of damaging the evap tubes.

one thing to consider is that the hot coils are going to dump out a fair amount of heat. it might be good to keep them outside the house if at all possible, to save strain on your newer A-C or just to keep yourself cooler if you're going without.

and just a final thought regarding the EPA: for the love of god, if you're going to drain the system of existing gas, bottle the waste gas and take it in for proper disposal! you're supposed to take the system down to a near perfect vacuum in order to clean it out prior to re-filling it anyways.........
 

best [486]

Limp Gawd
Joined
Aug 28, 2005
Messages
476
that and there are a lot of kinds of freon, i used R-12 before i realized the can was $70 :mad: :eek: I wouldve used r-290 but if that leaks [a very small chance] it'll ignite, so then that leaves you with r134a available at walmart, just buy the cans that don't contain oil [mixing oils is bad {i think}] and the adapter kit [if you use the 1/4" flare type schreder valves]
and if you use a piercing valve and vaccumed recovery cylinder then it only has to get to 0 PSIG [14.7 PSIA]
 

nhusby

Limp Gawd
Joined
May 1, 2006
Messages
303
I live in wisconsin, so it would be illegal for me to purchase any freon without a license. however, my roomate is going to school for HVAC and my dad is a mechanic licensed to do A/C work (on cars) and he has a ready supply of r134. I know r12 works much better, but as you mentioned its more spendy. I do know a source that is selling bulk r12 (30lbs) for $550 if anyone is interested.

that being said, I actually dont want to mess with the freon at all, I was hoping to isolate the high pressure side (via regulator valve if possible) so venting would be minimal.

also, thanks for the advice on not removing the evap core.

I am now thinking that I would route the output of the evap core to a water block (really trying to skimp on money) and from there to the compressor. My thinking behind this is that the freon would be in a gasius state as to not hydrolock the compressor. I would also build a custom case using the evap core as a dehumidifyer (not that I can spell.) with air drawn in through the evap core's fins and exhausting through the condensor core. I would have to figure out an efficent drainage method for that too, but it should limit condensation in the computer, and help cool the other components.

thanks for your input so far guys.
 

nhusby

Limp Gawd
Joined
May 1, 2006
Messages
303
also, I forgot to mention, if you are converting from one type to another, 4 seasons sells an oil or additive that allows the mixing of the oils, but you should try not to mix them. (thats what dad told me)
 

cyks

[H]ard|Gawd
Joined
Dec 31, 2005
Messages
1,234
DFI Daishi said:
unlike that previous thread this one has some specifics.
This thread would not even be here if it was not for the previous thread :rolleyes: :

nhusby said:
I didnt even know thats what is called "Phase changing" until I read this thread.
"This" refers to the last one. K+ to all those that put forth the effort to teach this user. Good luck nhusby, post some pics when you finally finish:cool:
 

nhusby

Limp Gawd
Joined
May 1, 2006
Messages
303
actually I was waiting for some feedback on wether or not splicing in after the evap core would be advisable.

like I said, I grasp the general principle, but lack any practice or experience...
 

nhusby

Limp Gawd
Joined
May 1, 2006
Messages
303
I was wondering, how hot does the high pressure side typically get? and how much of a booster effect would it have to water cool the high pressure side?
 

cre8chaos

[H]ard|Gawd
Joined
Aug 29, 2005
Messages
1,629
nhusby said:
I was wondering, how hot does the high pressure side typically get? and how much of a booster effect would it have to water cool the high pressure side?

Water cooling (worthless) is starting to get to casscade cooling and that is way over your head for now.
Go over to http://www.xtremesystems.org/forums and read, read, read. Then after you have it all, read one more time and then ask questions.

GL
 

TheStigma

n00b
Joined
Jun 8, 2006
Messages
4
For water-chilling, you want a mix of water and methanol. its viscous, carries heat very well (not quite as well as pure water, but better than pretty much all other subzero mixes), is cheap and well... just best to use simply put...

negatives are:

- most alcohols including methanol tends to made rubber dry and brittle over time. Don't rely on rubber seals, or you could get a leak later on.

- Methanol is poisonus. if you drink, inhale (lots), or get into contact with skin (again, you need lots) it can potentially poison you. Large quanatities can even blind you, but you usually have to drink a large quantitiy for this. Still, you need to be aware of this fact and work in a well ventilated room when you dowork with it. and preferably with some kind of latex gloves or something to be 100% safe.

- Methanol is flamable. very flamable. Explosive in vapor form even. water/methanol mixes over meth30% water70% are potentially flamable. blwoe that they are considerd as safe as water.

Depending on the mix concentration you can chill down VERY low. more than any normal chiller can handle anyway.

You NEED to be aware that most normal water pumps cant handle liquid when below -20 to -30 celcius. They crack open and are destroyed. Below this you need more robust pumps (check ebay).

Finally... chiling liquid that then chills your electronics is never as eficient as direct-die chilling. No matter how you design it. You do however have much more "manouverability" and stuff you can do with chilled water than you can with direct-die.

-Stigma
 
Top