I am trying to make a cooling system for my high end PC, but instead of the traditional air and water coolers, I wanted to go with sub-zero cooling. I do not want to use LN2 as it is expensive and not a relatively long term solution.

I was thinking, I should but a Mini-Split Ductless AC unit and instead of connecting the outdoor compressor-condenser unit to the evaporator unit that would be placed in the room that needed to be cooled, I would instead connect the copper tubes from the outdoor compressor unit to a metal CPU block. I have the Heat-killer IV Pro CPU block which has a full metal construction and was thinking it should be able to withstand the immense pressures of the system. I would solder connect the copper tubing to the CPU block and the other end of the tubing would be connected to the compressor unit and would be sealed.

This is the my basic view of the path that the refrigerant would take:
Compressor --> Condenser --> Expansion Valve --> CPU Block (instead of evaporator) --> Compressor

The reason i need help is because I know that the compressor unit comes pre-charged with the amount of refrigerant needed for the AC to run normally using compressor unit and the evaporator unit that comes with the mini-split system. I wanted to know what i needed to change so that it works with the cpu block instead of with a evaporator unit because the evaporator unit has a bigger volume for the refrigerant than the CPU block and i needed to know what i would need to adjust. Also I need to know how the refrigerant would behave inside the CPU block because in think that inside the evaporator unit, the refrigerant enters as a cold liquid and evaporates carrying the heat away as a gas. When it is replaced with a CPU block i needed to know if the heat of the CPU is enough for it to turn that liquid into a gas and what would happen if not.

I am comfortable modifying the pressures of the refrigerants, voltages for the compressor unit, lengths of tubing, and other things that don't involve replacing parts of the mini-split AC system.

ANY HELP IS APPRECIATED

 

... one necessary criterion of expansion cooling systems is that all the liquid must become gas before it re-enters the suction side of the compressor. Liquid that has been allowed to get to the compressor typically will cause mechanical damage.

 

... one necessary criterion of expansion cooling systems is that all the liquid must become gas before it re-enters the suction side of the compressor. Liquid that has been allowed to get to the compressor typically will cause mechanical damage.

Thank you, i see, which would mean that AC systems work in a way that all liquid is guaranteed to turn back to gas before entering back into the compressor reguardless of the temperature inside the room which typically ranges from 18 C (after the max cooling is done) and above. So how can i modify the system so that the heat of the Cpu block is guaranteed to turn the liquid into a gas is the question (Cpu outputs anywhere from around 10 watts on idle to 150 watts on load i believe). The mini-split Ac system i plan to use is 12000 BTU (3500watt). From those numbers alone I think its safe to assume that the Ac is more than enough to cool the CPU which is exactly where the problem lies. I need to make sure that the liquid needs to turn to gas before going back into the compressor.

 

... one possibility is to install a reservoir, essentially a copper container, having an inlet and outlet at the top. Any liquid would fall to the bottom at the entrance tube. Only gas would reach the exit, flowing into the compressor suction port. If all the liquid is successfully converted to gas, the reservoir would not be cool to the touch. If cooling is detected at the reservoir, that would indicate that some liquid is still being converted from liquid to gas.

 

Thats something I havent thought of, thanks. Some questions i have are,

Is it ok for the refrigerant to stay as a liquid inside the reservoir?

How would that effect the rest of the loop if some of the refrigerant is held back into the reservoir?(i heard that the compressor motor is cooled by the refrigerant also)

Also where would the reservoir be placed, after the cpu block or before? (i am thinking before so that the cold liquid runs through the cpublock first and then cools the cpu before the excess is stores inside the reservoir.)

Lastly would i need to change any pressures of the refrigerant when i do this idea since the volume of the original evaporator is probably greater than the the cpu block and the reservoir?

 

... The reservoir would be placed next to the compressor, after the CPU cooler. A possible difficulty is that the CPU cooling device is not physically large enough that there is enough heat transfer surface area to accomplish the desired degree of cooling.
Heat Transfer between chilled water, a liquid, and a Freon vapor refrigerant are two different things. The better design approach would be to produce chilled water and use that as a cooling medium in your CPU cooler. If you want temps below 32 degree F, just use a different liquid ... anti-freeze instead of water maybe.

The volume of refrigerant that is installed in the system will have upper and lower limits. The best approach to filling it would be to vacuum it out, and allow short increments of vapor refrigerant to be added gradually, allowing sufficient time for thermal temperatures and heat flows to stabilize.

 

But isn't the freon vapor refridgerant liquid when the heat is being transferred between CPU and the refridgerant? And since when the liquid is coldest after it leaves the expansion valve, I figured that liquid will be coldest. Also wouldn't using water or antifreeze as a median make it less effective. Also doing that would mean I need to have a temperature probe stop once the median reches a certain temperature so that the AC doesn't work on too low of a temp.(that's bad for the system I think, right?) Also this way I want to run it continuously as long as my PC is on using a relay.

If the anifreeze way is better, I can do that but I would need help with using a different evaporator coil and the setup for how it would work. I'm not sure exactly what you have in mind when you said use something as a median. But I'm assuming that what you meant was, instead of the evaporator part cooling the CPU directly, it will cool the antifreeze which in turn cools the CPU. I don't understand the point of that other than adding another thing between the CPU and evaporator which I would assume is a bad thing because now the heat has to go though more things to cool the CPU.

 

... there is fluid refrigerant, which is the result of removing heat in the condenser coil, usually aided by a fan of some sort.
... then there is refrigerant vapor, which is the result of the fluid passing through the expansion valve, and starts off at a cool temperature, determined by the low pressure in the evaporator. As the refrigerant vapor progresses through the evaporator, it absorbs heat and changes from a mostly liquid vapor mix to a mostly gas vapor mix ... (the technical term for the vapor gas/liquid ratio is quality) ... eventually exiting the evaporator as a complete gas (ideally), having absorbed heat from whatever is outside the evaporator, achieving a cooling effect, and proceeds to the suction side of the compressor intake. The compressor both heats and compresses the refrigerant gas, which allows it to go back to a liquid state in the condenser ... The condenser fan removing heat to the atmosphere.

 

... Yes ... you would have to have a separate coil, surrounded by the water/antifreeze mix in order to have chilled fluid for the CPU cooler. However, that could simply be a coil of copper tubing immersed in a container of the cooling fluid mix. A quick trip to an auto salvage yard would probably yield a working expansion valve.
... some soldering and assembly required, but not impossible.

The temperature that the evaporator cools down to is fixed by a so called boiling point of whatever refrigerant that you are using.

 

Ok I see now, thank you
How do I determine the boiling point of the refrigerant, because wouldn't it be different at a different pressure such as in the AC system?
Also how would I limit the temperature so that it doesn't get near the boiling point, because wouldn't that be a bad thing if it reaches that point?

 

... correct .... say you are you using R-134, the standard automotive refrigerant since about 1996, and you measure a low side pressure of 5 Psia, then the R-134 table indicates that the lowest temperature in the evaporator would be -50 degrees F. However, by the time that the R-134 vapor reaches the evaporator outlet, it will be completely gas and will start to rise in temperature. It should remain cold enough to chill any sort of water-antifreeze mix for extreme CPU cooling. You will have to experiment with the actual low side pressure ... The lower the pressure that is obtained, the lower the temperature that will be achieved.

 

I am going to be using R410a. Also how did you get the boiling temp of the r134 at 5psia. Is there some claculator I can use or a chart to determine the boiling temp? Also would I need to use the reservoir idea that you gave me when doing this to make sure that no liquid gets into the compressor?

 

Here is a table showing the relevant temperature and pressure values for R410a in psig and degrees Farenheit.:
https://www.google.com/url?sa=t&sou...FjASegQICBAB&usg=AOvVaw0b5DIQ9FNkRNBUzXMpnzwI

... regarding the use of a reservoir mentioned above, it would be a good idea, particularly because you do not know exactly the quantity of refrigerant that your custom built system will operate with.

 

Oh I see thank you for that. Also do I need to adjust anything else such as voltage or anything for the input for the compressor motor. Because under lower pressure wouldn't the compressor heat up faster since it would run with a lower load since there isn't much pressure?

 

... not really informed concerning compressor designs, particularly newer innovations. Generally speaking, there would not be any special arrangements due to refrigerant pressure, either high or low.
... If your compressor comes pre-charged with refrigerant, then it should have any necessary lubricating oil included. You should construct a reasonably dimensioned evaporator, with an expansion valve between the condenser outlet and the top end of the evaporator coil, and install a a low side charging/gauge port at the evaporator tail-end, not too far from the compressor suction inlet ... This part will be available from a refrigeration supply store. The store clerk may also be able to suggest the correct expansion valve to use.
... You will have to figure out some way to vacuum out any air from the evaporator, prior to releasing the freon into it ... maybe there will be some type of valve on the compressor art the condenser outlet and the suction inlet .. that would seem to be necessary ... requiring only solder connections between the compressor and the expansion valve/evaporator assembly. If you install a reservoir of some sort before the suction inlet, just make it about the size of a Freon can, with inlet and outlet connections at the top ... that should be sufficient.
... If there is any installation information concerning the distance between the compressor and the evaporator coil, that might be useful to know.

 

... If a separate charging port is installed on the evaporator tail end, then you will be able to find a vacuum pump someplace and remove any air, assuming you have a set of refrigeration gauges. ... Then maybe the only other parts that you will have to buy is an expansion valve and a length of copper tubing. See if you can purchase the Refrigeration Grade copper tubing ... it will be annealed, bendable, and thermally conductive, as long as you don't bend it too much
... after vacuuming out the air from the evaporator, turn off the vacuum pump and observe the gauge for 24 hours or so, to confirm that the vacuum holds, that nothing is leaking ... those port gauge connections are similar to tire valves, and have small O-ring seals, which should be checked to see if they are sealing.

 

Ok I see now. Thanks for the help, I will keep updated if any questions come up.

 

... hope that you get it running.

 

While you are designing this, you should consider draining the condensation so you don't get puddles on your mother board.
Tapping into a small refrigerator system would be more economical.

 

Yes I had taken that into account already. I am planning on using copper tubes so condensation definitely will be a problem. I had planned to fully cover copper tubes in insulation, and then use some nichrome wire to wind around the outside if the insulation and pass just enough power to make it warm enough to prevent any condensation from forming