Originally Posted by Frontman
Ok I see now. Well the TEC is going to get very hot, so the water will be absorbing a lot of heat. Unless the water is getting cooler than the ambient, like you said, the radiator will be cooling the water.
I'm not sure what the temps will be, so I will have to experiment and see what happens.
Hopefully, this is how it will work:
1. Water starts in reservoir
2. Water gets pumped through pump
3. a. Water gets pushed through waterblock
b. Water absorbs heat and rises above ambient temperature in the process
4. Water runs through radiator and gets cooled closer to ambient temperature
5. Water flows through cooling reservoir and gets cooled much below ambient temperature
6. Water flows back into starting reservoir.
If water is not rising above ambient temperature after flowing through the waterblock, I could get a valve that would allow me to skip the radiator and go directly into the cooling reservoir.
Do you know of any valves to switch paths in a water loop?
Ok, next set of problems.....
1) You're talking about using a mini-fridge compressor. The same arguments apply: ie, the compressor will run continuously until it burns out. That said, it really comes down to your use of the PC. If it's a few hrs a day for gaming, then the compressor may last, but you won't see the effect that you are thinking. If it's 24/7 folding/crunching, etc...then the compressor won't last.
2) #1 aside....the water is not going to rise more then ~1.5C going through the block. On that same note, not more than ~1.5C is going to be lost going through the radiator. Once the system settles to equilibrium those figures will drop to ~0.75C in and out. These measurements have already been determined. The same is basically true with the compressor, except it's a bit more variable because of size/power, refrigerant used, etc, etc, etc. However, it's still going to require "pulldown time", and that's simply because it does not work quite the way you are thinking: The water temp is not going to rise, say, 15C going through the block, then cool 15C (at equilibrium) going through the radiator, then drop another 20C going through the cooling res....it just does not work that way. Toss a quart of water on the stove and see how long it takes to reach the boiling point....it just takes time for things to heat up and/or cool down. Toss an ice-cube tray of water in the freezer and tell me how long it takes to freeze solid.Keep in mind that the water in the ice-cube tray will not be having any heat added....just heat removed, and it will still take a considerable amount of time considering the amount of water involved.
Here's the deal: if you just want to chill the TEC water to or slightly below ambient then you'll save yourself a lot of hassle and some $$$ by just using a bong or HVC. Beyond that, toss the TEC and just go with direct-die phase or a proper chiller.
My questions would be: Are you going to recover, rebuild and tune? Or, just try to incorporate the mini-fridge system as is? If recover, rebuild and tune...do you have the tools and supplies for that..or at least access to them?
You could use a Y or T and a couple of these: Alphacool Ball valve with 2 x G1/4" IG-connection