UPDATE! 2/11/06 2pm
If I can achieve a cpu temp of maybe -10F or -15F running a wc system with a refrigerated chiller would you guys consider that in the realm of these phase change systems?
my other question is can I use Vasaline to slobe over the cpu ,in the socket and such for a condensation barrier?
or do I NEED to buy the stuff thats made for insulating everything?
If I build the chiller system like I have some plans for my wc lines will frost.
I wouldn't mind my lines or cpu block frosting and building ice (it would look cool)
but I want to protect the cpu and mobo.
I'm thinking Superman's fortress of solitude looks like ice..........
I don't want to go strictly refrigeration or "phase change"
cause with wc I can valve in an emergency backup radiator in case the refrig. pukes on me.
any suggestions?

The 1st Question I can Answer with a Yes, the rest a pro will answer.

No, you will not get those temps with a chiller. Maybe in the 20's with a real good setup. For more than that you will need to go full phase change.

Do not use Vaseline since it is conductive. Get the dielectric grease for the socket pin holes, and use closed cell foam for the CPU. Neoprene works well, I got a 5 foot roll off of eBay for $12. Don't forget to insulate the back of the board as well.

This is what it'll look like when your done.
I build little foam boxes around the blocks with removable tops instead of neoprene.
also you need to seal the mobo with conformant coating or such to resist any stray condensation .
you MUST insulate the fluid lines or when you turn it off the frost melts it'll drip inside your case.

Quote: Originally Posted by BigAkita No, you will not get those temps with a chiller. Maybe in the 20's with a real good setup. For more than that you will need to go full phase change. Do not use Vaseline since it is conductive. Get the dielectric grease for the socket pin holes, and use closed cell foam for the CPU. Neoprene works well, I got a 5 foot roll off of eBay for $12. Don't forget to insulate the back of the board as well.

What about Styrofoam? Ive heard people can use that, is it true?

Styrofoam is not recommended since it can hold quite a bit of an electrical charge. Enough to zap your components. Neoprene works great, but any block of closed cell foam works fine.

EDIT: RB makes a great point about the lines, I went through 2 vid cards before I figured out I had an insulated line that wasn't insulated that great. A couple drops from condensation is all it took.

Isn't extreme OC'ing fun?

Quote: Originally Posted by BigAkita Styrofoam is not recommended since it can hold quite a bit of an electrical charge. Enough to zap your components. Neoprene works great, but any block of closed cell foam works fine. EDIT: RB makes a great point about the lines, I went through 2 vid cards before I figured out I had an insulated line that wasn't insulated that great. A couple drops from condensation is all it took. Isn't extreme OC'ing fun?

Armaflex is one of the best closed cell foams for insulation. It is good up to -200C. No nerdz that would not be considered a phase change that would be a water chiller. As far as temps. I have seen a water chiller made out of an AC unit that went down to -30C over at extremeoverclocking. You will need to insulate everything really. I dont think a pump could withstand those kinds of temps as well. How long do you plan on running this system? If its a couple hours I wouldnt be concerned with using vaseline or die electric grease. I ran my dry ice setup with no reall insulation except for around the tube for a couple hours.

I've been thinking about this alot.
Altho it would probably be more effiecient to directly cool the cpu with a refrigeration system (I'm no longer gonna call it phase change since it's a marketing gimmic)
there is definite benifit's with a chiller unit. There are as follows.

1. with a chiller you can have a decent sized reservior then you can use a thermostat and shut down the refrig. system instead of it running the full time the pc is running. (saves elect. and helps keep a more consistent temp.)

2.In the unnavoidable event when your refrig. cpu cooler fails you still have the fluid in the res. thats cold circulating. (see #3)

3. a backup rediator can be installed to either manually or electrically open valves to allow for a typical wc setup cooling and you can keep using your computer till the refrig. system is repaired or replaced.

4. And sorry BA but if Automotive antifreeze and food safe Glycol can protect from freezing to -20F
then cooling a wc setup in a chiller type system is definitly possible.
the biggest question IS are wc waterblocks able to withstand those temps.? AND are they effeicient enough to give you a -15F temp on your cpu?

Since the words "you will not" were used in reference to achieving those temps with a chiller setup.
I will start a "side project" to either prove my point OR prove I'm wrong.
I'm going to continue to make the chiller setup with a beer fridge compressor,but once I get my hands on a suitable freezer condensing unit then I'll go with it.
To get a chiller to -20F isn't hard (for me being in refrigeration) but to find the stuff to do it will take time and patience.
I can easily design a system that will reach -45F in a walk in box roughly 8' square. (I worked on one)
hitting -15F is easy. Doing it within my budget is the problem.
Have patience..... within a year I'll hit AT LEAST -10F with a chiller setup on mbm5 running F@H.
pics will obviousley be included.
So...... BA you say I can't??? I say I WILL!
(realize this is all meant in a friendly and respectfull manner...with some humor)
I will freeze BA's bunns solid!!!

Btw. I've been running 107F or 41C on my cpu for 3 hours now with F@H going.
THATS out of my beer fridge still keeping my beer cold aswell.

If you look around any construction sight,they always throw away left over neoprene insulation.

Heres some blocks that will take sub zero temps
D D,Brass maze 4 Gpu block http://www.dangerdenstore.com/produc...&cat=48&page=1

D D RBX, With optional brass top (2 outlets so you can run lines to your gpu and chipset) http://www.dangerdenstore.com/produc...&cat=19&page=1

You really need to find a pump with a metal impeller housing otherwise the plastic ones become quite fragile at sub zero temps.
If you use a plastic housed pump make sure to completely support the pump and lines individually ,or any strain on the housing will cause it to crack.
Use a lower flow pump (300gph or less) ,high flow pumps arent neccesary and can actualy impede the thermal exchange between the evaporator and cooling medium.
Keep in mind that the coolents viscosity will increase substantially as temps get lower.

Use a Low temp Thermal Expansion valve instead of a cap tube, it'll allow for easier tuning and adaptability to changing loads.
dont bother with R12 or 134a go for R22 Or R290 with a Polyol Ester lubricant ,I prefer "Solest LT-32"
It's got decent miscibility at low temps.

P.M Me if you need help sizing the TEV/TVX or captube when you get a compressor

You sound like a fellow Refrigeration Tech.
I'd try direct refrig. cooling the cpu,(phase change) but I don't want to have to depend on the refrig system.
I think a good pump would be more dependable than a compressor. (I've seen too much money lost on spoiled food from a refrigeration failure)
Also I can have a back up reservior piped in but valved off for backup.
As far as Refrigeration I'm thinking around a 1/4 ton txv with r404a to cool a 1 gallon reservior.
I may make it a smaller reservior with a big chunk of aluminum to help mediate the res. temp.
That way the compressor could cycle off ( an empty freezer shortcycles)
As far as coils and compressor and that I'll have to see what I can come up with from scrapped units at work. I'll just have to make sure I have enough load on the refrig. unit to prevent floodback to the compressor.
I may need some help on designing this tho. And thanks

You can not pump frozen water!OH CRAP,inserted entire foot and shoe in own mouth!!!

Quote: Originally Posted by jofultner You can not pump frozen water!

Yea, because Its Frozen!

However, if you insulate the pipes ( shoot me if Im wrong) It shouldnt freeze up (well as long as you dont go below 32c..) But If you have AntiFreeze, it shouldnt freeze (well it should lower the freezing temp..). I do know for a fact theres Alcohol in Windsheild wiper fluid and thats why it doesnt freeze on your wind sheild in the winter.

Automotive antifreeze is good down to around -20F.
I know I need a different cpu block cause the wc fluid temp is 28F and my cpu is at 68F.
It's a crappy one that came with my Aquiraius 2 setup.
Also the cpu block is 1/4" and everything else is 3/8" id.
(Iknow I said 1/2" earlier, I didn't realize the tubing is measured OD. I also had a six pack prior to putting it together )
I wonder if Beer is a good wc fluid???? ahh nevermind I'll never have any in the system cause I'd drink it all
I do have a serious question... What is a decent temperature difference between the fluid temp. and cpu temp? ( tells me the cpu block effeciency)
I use mbm5 and verify accuracy with bios.

A further thought on your responses, water has the same viscosity with a glycol additive at 100F room temp. or at the threshhold of the antifreeze's ability (around -20F) water is unlike any other liquid known to man. It's viscosity never changes regardless of the temperature.
When water reaches freezing temperature it "slushes" meaning ice crystals are suspended in the unfrozen water.
Water alone will never get "thick" like cold oil.
The glycol additive may.
I'm not a scientist but to my knowledge water is the only liquid known to man that goes directly from liquid to solid without "gelling".
Slush is suspended solids. If you go buy a slushie at 711 you can pour off the liquid and be left with a cup of wet ice crystals.
Water is very unique It to my knowedge is the only fluid that can be an insulator and a conductor to electricity but always retains it's thermal transfer properties.

If I'm wrong please feel free to educate me....But I know for a fact I can use glycol to reach at least - 20F
The glycol will then gell and seperate from the water and the water will then chrystalize and slush.

Pure methonal wont freeze till below -85f,unfortunately its toxic is easily absorbed through skin,and produces noxious fumes,it also it tends to degrade the tubing and any plastic.

ethonal freezes around -80 and isnt toxic, its also easier on the tubing and plastic.

in certain controlled situations water will hit the triple point phase.
that is it can contain liquid,suspended crystals and solids together .very bad for pump

I mix 25% ethonal,25% ethylene glycol,and 50%distilled water for my chillers.
Ive never had any problems with crystalization above -45f with this mix.

Im sorry to say your incorrect about the coolent viscosity ,it Does get much thicker when you approch
the lowest temps.
In my impromtu tests ive noted that my danner 5 pump(500Gph)only flows around 300 gph at -40f with the above coolent blend.

Running a smaller reseviour will keep the coolent mixed better and help prevent crystalization.
Run 2 outlets into the res to produce a swirling motion,you'll like the results.

And about liquid refrigerant entering the compressor you have several options.
I incorperate all 3 listed below into my designs.

1. buy a low temp copeland hermetic compressor,thayre designed to occasionally slurp liquid refrigerant without damage(large housing with more internal passageways to heat the remaining liquid to its gas phase).
2.use a TEV.
3.provide several loops after the evaperator to allow any excess liquid to turn to gas before the comp.
youll need a drip pan for option 3

this is what ive come up with,now that I have a new camera I'll finish my worklog
http://forums.pcapex.com/project_wor...d_frost_2.html

Attached Thumbnails

 

no offense guys but a 50/50 mix used in automotive cooling systems can go below -20F if it didn't there would be a lot of pissed off people in Idaho and montana when temps drop below -30F every winter and if that were to happen wouldn't the water/anti-freese freeze? and due to the laws of nature crack teh engine block? and before i get a angry flame about how cast iron is stronger than water so it wouldn't there are thing very thin parts of the engine where water passes that would crack under minimal internal pressure... so thats my .02

oh yeah last winter when temps hit -35F my lil aluminum 4 banger didn't crack or anything so that would be some credibility to my claim...oh yeah before i forget if it did freeze when ithit -35 when my engine started (Yeah i was up that early) why did my waterpump pump if it would have been frozen it should have seriously damaged that part....

Thanks for the feed back! I almost forgot about this thread With work and tryin to finish my case.
I got some minor things left with the case but I need to decide what to do with the chiller end of things.
I've really thought about this and I'll try to use an Igloo cooler of suitable size as my chiller reservior/exchanger.
The most efficient design I've seen is used at Burger King, which is a "MultiPlex" system.
It's basically a 20 gallon tank with a refrigeration unit that develops an ice bank inside by a simple refrigeration loop of copper (about 20 loops no fins) inside the water bath.
I'm going to try making a scaled down version and use Glycol to prevent freezing.
I've scored a freebie... An LV1A (refrig. guys know what I mean) It's basically a packaged condensing unit that pumps all different types of refrigerant out of a system for repair or disposal.
I'm hoping to use the LV1A and the Igloo cooler setup to atleast get below 32F (liquid temp)
I'm planning on a TXV system and will incorperate a thermostat and other refrig. controls.
the idea is to have a reservior/chiller tank large enough that the refrig. can cycle off like your home refrigerator to mantain a set temperature. NATURALY I'm going to see "How Low Can I Go?!!"
This may take some time, but if all goes as planned I'd like to do this in such a way that I can write a how to guide that will shed alot of light on the refrigeration end of things.
Hopefully a step by step progress.. I'd like to cover everything from determining if the freebie fridge or A/C can handle the job, to actual brazing copper and charging the refrigeration system.

I should be able to achieve respectable temps. That would help me prove that a $1000 dollar "Phase Change" system is a rediculously overpriced product. and a wc setup can compete with those temps PLUS is more reliable and efficient.

I know thats alot to prove but after 15 years of repairing and installing refrigeration units I think I can get some good results. After all the surface to be cooled is less than 1 square inch. And the capacity of the cooling setup will be more than 3 square feet.
The key to success is the best possible heat exchange at the processor.


Stay tuned! (it could get interesting, If I'm wrong then atleast I tried)

well i don't know if this makes sense but you know those little mini fridges with the freezers? couldn't you just pipe the radiator into there have fans setup to force air over radiator? that in my mind should almost be capable of freezing the CPU to the lowest temp the fridge can take, not to mention it would give you a mtn dew cooler next to your PC but in all honesty i think takin apart a full size freezer and modding it to fit into a cooler would work the best and just have a couple of 120mm fans in there blowing over a radiator would work wonders, ( as long as the fans continually move they shouldn't freeze up, our walk in freezer has fans that circulate the air and it maintains temps at -20) just an idea because then if your doing a setup like that you could probably have like 2xdual 120mm rads in there which would probably kill any heat you could build.