![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
I thought I'd post my chiller here, maybe I can bribe some help out of you by posting some pics! I've posted it elsewhere, but maybe I can get some additional ideas here. Also, this forum could use a few projects.
Here is my old my case. I've had it since my old thunderbird 1400. Kinda ugly, and no space to work.
I wanted to do an in-case chiller. I've done a direct die project before, but it was large and clunky. I wanted a compact chiller. Here's the parts:
I needed to make a compartment at the top of the case for the chiller to sit in. I used some aluminum plate:
Here is the plate installed:
I drilled a 120 mm fan hole in the plate so I could put a fan under the compressor to help cool it. Sorry, no picture on that.
I wanted to do a fancy evaporator/reservoir. I designed it to have a double shell... The inner one to hold the evaporator and water, and the outer to contain the inner reservoir and some insulation. Unfortunatly, this made the inner reservoir very small. I had a hard time making an evaporator coil with a tight enough radius to fit in the 2.5 inch inner reservoir. Here's what I finally came up
with. It's 5/16 tubing. You can see the 3/8 one I tried to make. Just couldn't get the 3/8 tubing to bend properly in that tight of a radius.
Because the reservoir will be mounted vertically, I created a tube that would route the incoming hot water to the top of the reservoir, while cold water would be sucked out the bottom. You can just the the tube in the middle of the res in this picture:
Here is the evaporator inside the reservoir:
Here is the top and bottom of the inner reservoir. Notice the fill port and the line for the temp sensor for the automatic temp relay.
Another picture of the evaporator with cap tube in the reservoir
Here is the entire reservoir assembled. The inner reservoir has about a 1/2 of that spray foam between it and the outer shell.
Here is the entire thing assembled in the case for leak testing;
With no load and no circulation, I was able to get the fluid temp down to -40 C. Once I installed the computer parts and loaded the chiller up, it had no capacity. Here's a shot of the assembled machine:
This project uses a AE9415E tecumseh. It's a 1/5 hp compressor I believe. I'm using R-22 and a small amount of propane. The capillary tube came from chilly1, as did the condensor. Lots of the other stuff came from gclg2000. I can't tell you the diameter of the capillary tube. I told chilly1 I wanted some cap tube for a 300 watt chiller using a 1/5 hp compressor and R-22 and he sent me what I have. His suggestion was to start with 7 feet. After some trimming, I'm down to 6 feet. It will hold the fluid at -5C at idle. Load jumps to 0C. I have to really charge the system to get it to hold temps. Right now, low side pressure is 20 psi, high side is 225 psi.
It forms some condensation at these temps:
I'll insulate the water lines when I get everything squared away.
I'd really like to hit -10c fluid temp under load at least. Of course, -20C would be better.
I'm using an older D0 stepping 550 prescott at 1.6 v and 4.15 GHz. A Radeon X800 XT is also in the loop. Should I continue to trim the capillary tube length to try and get better temps? Anyone think I should move to an expansion valve?
Here is my old my case. I've had it since my old thunderbird 1400. Kinda ugly, and no space to work.
I wanted to do an in-case chiller. I've done a direct die project before, but it was large and clunky. I wanted a compact chiller. Here's the parts:
I needed to make a compartment at the top of the case for the chiller to sit in. I used some aluminum plate:
Here is the plate installed:
I drilled a 120 mm fan hole in the plate so I could put a fan under the compressor to help cool it. Sorry, no picture on that.
I wanted to do a fancy evaporator/reservoir. I designed it to have a double shell... The inner one to hold the evaporator and water, and the outer to contain the inner reservoir and some insulation. Unfortunatly, this made the inner reservoir very small. I had a hard time making an evaporator coil with a tight enough radius to fit in the 2.5 inch inner reservoir. Here's what I finally came up
with. It's 5/16 tubing. You can see the 3/8 one I tried to make. Just couldn't get the 3/8 tubing to bend properly in that tight of a radius.
Because the reservoir will be mounted vertically, I created a tube that would route the incoming hot water to the top of the reservoir, while cold water would be sucked out the bottom. You can just the the tube in the middle of the res in this picture:
Here is the evaporator inside the reservoir:
Here is the top and bottom of the inner reservoir. Notice the fill port and the line for the temp sensor for the automatic temp relay.
Another picture of the evaporator with cap tube in the reservoir
Here is the entire reservoir assembled. The inner reservoir has about a 1/2 of that spray foam between it and the outer shell.
Here is the entire thing assembled in the case for leak testing;
With no load and no circulation, I was able to get the fluid temp down to -40 C. Once I installed the computer parts and loaded the chiller up, it had no capacity. Here's a shot of the assembled machine:
This project uses a AE9415E tecumseh. It's a 1/5 hp compressor I believe. I'm using R-22 and a small amount of propane. The capillary tube came from chilly1, as did the condensor. Lots of the other stuff came from gclg2000. I can't tell you the diameter of the capillary tube. I told chilly1 I wanted some cap tube for a 300 watt chiller using a 1/5 hp compressor and R-22 and he sent me what I have. His suggestion was to start with 7 feet. After some trimming, I'm down to 6 feet. It will hold the fluid at -5C at idle. Load jumps to 0C. I have to really charge the system to get it to hold temps. Right now, low side pressure is 20 psi, high side is 225 psi.
It forms some condensation at these temps:
I'll insulate the water lines when I get everything squared away.
I'd really like to hit -10c fluid temp under load at least. Of course, -20C would be better.
I'm using an older D0 stepping 550 prescott at 1.6 v and 4.15 GHz. A Radeon X800 XT is also in the loop. Should I continue to trim the capillary tube length to try and get better temps? Anyone think I should move to an expansion valve?