(and here's the whisper quiet version):
Let me answer some of your questions:
First the block was insulated except the bottom of the cold plate. Yes the side of the block stayed close to -27C though most of the test. Although it did start to rise 5 minutes into the load at first,
because I didn’t have enough Freon in the system so I added a little bit more and started to stay at -27C.
Now the pelt was rated at 156 watts at 13.8 volts and I ran it with my 40 amp power supply, pelt draws close to 15 amp at load. Now the hot side of the pelt was against the bottom of the block and the block and pelt are the same size 50MM x 50MM.i machined a groove in the bottom of the block to put a temp probe close to the cpu core. So the probe sat against the hot side of the pelt and the block that’s how I did that.
Now the compressor return (suction) line does have frost to the compressor with no load, but when running a load, no frost.
The copper is alloy 110 I believe. The block has 4 mazes connected
.200 x.300 x 5 1/2 feet long grooves. That gives the Freon long enough
time to transfer the heat.
Cost is about $200.00.but that includes the price of my proto type unit.
Now I have dropped my temps even more by tweaking some more.
May meter won’t read past -58F and it will go past that now no load.
Now before everyone jumps up and says r22 boils at -40F so it can’t
get that cold. Look at a chart and my compressor runs in a vacuum
on the suction side of about 12 inches of vacuum. At that vacuum r22
boils about -60F.
Now how come I get the temps I get, well......
I designed the system from scratch with the most efficient use of components I could.
Compressor is 1/4HP found anything smaller doesn’t have enough suction to make a strong vacuum to get below -40F. The condenser is mounted horizontal with no tubes going up and down. This keeps the oil and Freon from collecting in a tube, also keeps a steady flow of Freon to the capillary tube.
The capillary tube is run inside of the suction line from the compressor to inside the evap block. This makes the return gas super cool the incoming liquid Freon, making it colder and denser.
The suction line is bent in a way that is almost impossible for liquid
Freon to get into the compressor.
the evaporator it self is my design with a injector nozzle right over the cpu die spraying super cooled Freon straight on the bottom plate of the evaporator and returning though a maze of almost 5FT 6"long. I used flexible automotive ac hose to allow easy flexing and mounting of the block to the board prototype unit I used copper and found it is very hard to bend even soft copper time and time again between changing cpu's. This is major failure of the units you buy online.
|