I recently talked my bro (ardawg) into watercooling (not that it took much talking - halfway into a 12 pack and we were good to go) so we started looking at parts and planning out the system.
I already had a spare full tower that I had used for H2O cooling (left over after I switched one of my systems to phase change) and quite a few spare parts but there was an interesting and fortunate twist to the layout of his house. The room his computer is located in shares a wall with his enclosed gargage....running copper pipe thru the wall was obviously easy enough and there was plenty of room in the garage for me to go as crazy as possible with this project. also, while the garage is enclosed, there is no heat out there in the winter (other than that which would be provided by some heavy H2O gear) so we had a temp differential working on our side for winter -- a waterchiller would resolve any problems with temps in summer. sold.
so, I started the search for a second enclosure for the H2O gear that would go in the garage. amazingly enough, I found an empty Sun Mass Unit cabinet for dirt cheap ($60). after almost destroying every door in his house...but doing no damage to the sun cabinet as it is built like a fricken tank (prolly weighs 150lbs), we got the cabinet in place...to put the size into perspective, it is about 5 feet tall, 3 feet deep and 3 feet wide.
pic of course:
^that is with the side panels removed. the white thing in the bottom is a 5 gallon res....it will be replaced with an insulated model for chiller use later. see the copper pipes coming thru the wall??
for winter use, we thought that a big heatercore would be plenty so I modded out a '77 bonneville core, added a shroud with my usual divided-chamber-venturi design packing dual SanAce 120s (about 103cfm each):
for testing, we set a PSU on the res and left the wires everywhere as in the pics.
temp differential between rooms paid off - the OCed 2400 mobile barton in his system showed 2C below case temps in the computer room on the first run hitting 2.5 ghz easy. that is with a maze 2-1 until I get enough time to build another block like the one in my main H2O system [which is a jet impingement block that is brazed together with 15% silver brazing rods (good to 50,000psi)]
next, straightening up the wiring, adding some insulation, dropping in a garage-built waterchiller and adding ball valves so the water can be routed thru the heatercore or the chiller depending on the season.
more to follow...
I already had a spare full tower that I had used for H2O cooling (left over after I switched one of my systems to phase change) and quite a few spare parts but there was an interesting and fortunate twist to the layout of his house. The room his computer is located in shares a wall with his enclosed gargage....running copper pipe thru the wall was obviously easy enough and there was plenty of room in the garage for me to go as crazy as possible with this project. also, while the garage is enclosed, there is no heat out there in the winter (other than that which would be provided by some heavy H2O gear) so we had a temp differential working on our side for winter -- a waterchiller would resolve any problems with temps in summer. sold.
so, I started the search for a second enclosure for the H2O gear that would go in the garage. amazingly enough, I found an empty Sun Mass Unit cabinet for dirt cheap ($60). after almost destroying every door in his house...but doing no damage to the sun cabinet as it is built like a fricken tank (prolly weighs 150lbs), we got the cabinet in place...to put the size into perspective, it is about 5 feet tall, 3 feet deep and 3 feet wide.
pic of course:
^that is with the side panels removed. the white thing in the bottom is a 5 gallon res....it will be replaced with an insulated model for chiller use later. see the copper pipes coming thru the wall??
for winter use, we thought that a big heatercore would be plenty so I modded out a '77 bonneville core, added a shroud with my usual divided-chamber-venturi design packing dual SanAce 120s (about 103cfm each):
for testing, we set a PSU on the res and left the wires everywhere as in the pics.
temp differential between rooms paid off - the OCed 2400 mobile barton in his system showed 2C below case temps in the computer room on the first run hitting 2.5 ghz easy. that is with a maze 2-1 until I get enough time to build another block like the one in my main H2O system [which is a jet impingement block that is brazed together with 15% silver brazing rods (good to 50,000psi)]
next, straightening up the wiring, adding some insulation, dropping in a garage-built waterchiller and adding ball valves so the water can be routed thru the heatercore or the chiller depending on the season.
more to follow...