"Project Redneck" Phase change system

[GunnerMan]

Well I have decided to further my overclocking endevours so I am building a direct die phase change system. I will be using an store bought A/c because this is my first unit I stick to "KISS." I will be using Propane gas as a refigerant. So my A/c is 5,050 btu, overpowered but it works. I am making a crap homemade block out of store bought fittings. So I am speculating to see temps of -30c or so idle and with proper tweaking head down to -40C. So yesterday I ordered my manifold gauge, vacuum pump, .031 capillary line, some schrader valves, and a filter/dryer. I will be getting some 1/4" tube and tee's today for use on the low side of the unit going to the evaporator or block. So I am taking a very bold and frowned upon move in the HVAC world and sweating(soldering) instead of brazing for a few reasons: Its quicker, easyer, and cheaper. I am using Propane so it is not harmful to have a leak so if something fails, no big deal really. So today I made my evap(Based off of a guy over at XS) all it is is a 1" endcap soldered to a 1" to 1/2" reducer soldered to a small piece of 1/2" pipe soldered to a 1/2" to 1/4" reducer. Then filled with copper wool. Here is a pic:


This is the unit I will be working with:

As you see the large radiator lookin thing is the condenser where high pressure gas is cooled and turned to liquid, the smaller one is the evaporator where the liqid boils off in a low pressure(vaccum) environment. So the evaporator will be cut off and replaced with the evap shown above. Schrader valves will be added to the low and high ends of the unit for charging and evacuating, a filter/dryer will also be pressent to eliminate and trace amounts of water my .75 micron vacuum pump can not pull out. I am still trying to figure out how I am going to mount my block(I will eventually buy a mass produced phase block that is more user friendly) but I am thinking using a metal plate that screws over it on to the mounting brackets. Then insulate the hell out of it.

So tomorrow I will take it to get it evacuated, cut off the stock evap, find a metal plate for my evap, start bending and fitting hose, and figuring out some basic design questions and where to put my valves at. If all goes well I will get some soldering done. The farthest I can go is running my 1/4" from my new evap to the compressor and from compressor to the condenser plus add the tee's. Then I wait for my other parts to arrive so I can complete the system and begin to vacuum and charge it. So as you see it is a VERY basic setup with a lot of basic things but I am hopin' it will work out as a starter project ad open the doors of refrigerated computing to me! 8-)

[GunnerMan]

Well I got my system evacuated but did not do much else to it. I sanded my evap to far so the end cap was to thin so I am replacing the endcap  :-[ I did find a solution to mount my evap to cpu. I got a 2" pvc plug. A 1" hole will be drilled in the center of it so the evap base can fit through. I will then silicone the hole and fill the plug with low expansion spray foam. Put a piece of plexiglass over top of the plug with holes to mount it to the mobo. Seal it on with some silicone type adhesive. I found that PVC is very weak at cold temperatures it is also a decent insulator so it should work fine with the foam.

[GunnerMan]

Well today I fianlly got some work done!. I have gotten my filter dryer ijnstalled to the low end of the condensor and a line from compressor to the condesnor complete with schrader valve. Tomorrow I am going to pick up more 1/4" since I am nearly out(need another 10 feet) and am praying my guages and vacuum pump get here. Tomorrow if all goes well I will finish connecting my suction and cap tube to my evap and I can leak/pressure test. If all goes well then it't time to fill her up with Propane and make something cold! So tomorrow I should get some pics posted because now my camera is dead. So for all/any of you watching my completely primitive phase project you can watch for some pics tomorrow afternoon of what I have so far and maybe a complete setup!  :D

[GunnerMan]

Well I put on my last schrader and put my evap on my suction line(I don't actually have a true suction line, just a copper tube:P) Then it got to dark to continue work. My vacuum pump came today, my manifold will be here tomorrow so then I can leak test at least. Still need to figure out how to get propane from said can to system

[Groundbound1]

Maybe it's just 'cause I'm new, but at -30, won't condensation be an issue? :-?

[GunnerMan]

Yes, condensation is a big issue. it is combated by housing the evap in a extra insulated, air tight housing, coating the motherboard socket area in dielectric grease, and lots of sheet and pipe insulation. (Ill get pics of all that later.)

So today,going behind schedule as usual anyway as I did get the evap and cap tupe put on my suction line, modded a propane torch to fit my system. All I need to do is install the suction/cap tube to the unit and we are set to pull a vacuum. I can not do that untill my manifold guage gets here, they say it will be here today but its getting late and still not here so I am loosing faith

Here are some pics of what I have done so far:

This is the vacuum pump. It is louse, noisy, and slow but it should get the job done:


This is the unit partialy assembled, as you see it is missing the low side and the fan:


This is my modded propane torch. The torch pipe had a 7/16" ID tube and my schrader had a 1/4' od so I used copper epoxy to bond them because it could fill the gap really easy and fast:


Last one, my evaporater and capilary tube. The refrigerant is pushed through the small cap tube in a managable stream where it boils off in the low pressure environment and gets sucked back to the compressor. This evap is very primitive and I will soon try and make a proper evap.


More to come!

[JBaymore]

And people think us wacko simpit builders are crazy!

In your spare time maybe you can convert it into a still and make some whiskey  ;).

Careful with that propane gas.  It is not much.... but it is very flamable.  The gaseous state is heavier than air and it will pool if it leaks.  Find a spark ignition source and you'll get a nice little mess to deal with.  ( I work with propane a lot doing kiln combustion engineering work. )

Good luck.

best,

......................john

[GunnerMan]

Hehe, yeah I have thought about the ignition risk, thats why I will be testing it outdoors  ;D I am not to worried about a leak catching on fire though, as even if it all were to pool under my desk there is only about 2 oz of propane used. Although 2 oz could cause a decent size fireball the possibility of all 2oz staying in a pool under my desk is quite slim especially with the airflow in here. Also if gas were to leak I would see a dramatic temp change after just a small amount of gas is lost.  Ignition within the system is impossible because of the absense of o2.

Anyway I still have not gotten my mani guages yet, FedEx fked up big time and picked em up from the seller 2 days late so it's Monday or Tuesday. Other than that everythings ready to go. The world record for an Opteron 175 is about 3460 Mhz, I have a Opteron 180, lets see if I can get anywhere near that heh. Anything after 2.8Ghz on these chips required a large Vcore increase to get them to scale higher. With this SS I should be able to go higher volts and and better clock per volt ratio

[Groundbound1]

Besides that, won't pumping propane through the compressor heat it up enough for it to combust? I'm reminded of a trick I've seen, where someone replaced the air in a friends trailer tire with propane, and when he hit the highway, the heat from the friction of the tires on the road was enough to light it and BOOM! Just seems like an awful lot of effort for a few extra Mhz. :-?

[JBaymore]

Besides that, won't pumping propane through the compressor heat it up enough for it to combust?

As long as he has evacuated all the air first....... there will be no oxygen to support any combustion wiithin the system.  No O2.....no possible chemical reaction.  

Not sure what the gas permeability to O2 of Cu pipe is though  ;).

best,

..................john

[GunnerMan]

Well I don't know either John but I am not a Pioneer in using R-290, it has worked for many others before me, so im gonna give it a shot.

It is a lot of effort but it is fun and cool, learn a lot from it as well. Sometimes you might get a few extra Mhz sometimes you might get a few extra Ghz, it just depends on the CPU and the temp. People use phase and run the Core 2 Duos at 5 Ghz 24/7 easily.

The compressors are know to reach 100c, very hot it's probably not enough to make it spontaneously combust, even under pressure. The tire thing was probably ignited by enough heat, pressure, and some other source such as static electricity.

[JBaymore]

The tire thing was probably ignited by enough heat, pressure, and some other source such as static electricity.

Most likely the tire was not evacuated of all the air in it before the propane was pumped in.  Enough to allow the gas to start to combust, which likely raised the pressure in the tire enough to rupture it ... allowing more oxygen to enter the picture and there you have it.

best,

...................john

[GunnerMan]

Well since it was a joke, im sure evacuation procedures were not taken heh, im still pondering if Propane is volatile enough to just ignite with pressure and some heat alone.

[Groundbound1]

I didn't pay much attention to pressure or anything. Frankly, I didn't want that much involvement. But however they did it, it worked! Ah the simple pleasures of youth!

[JBaymore]

im still pondering if Propane is volatile enough to just ignite with pressure and some heat alone.

Propane is compressed to the liquid state for storage and transport.