I’ve been trying to figure a way to go subzero without prepping the MB, the only viable way to do it would be to lower the temperature of the air around the board so I came up with this idea based on a dice slush pot. YES it’s over the top and not that practical, but I like making things along these lines.

Air tight construction for recirculation. Fans are conceptual, twenty four 120x38s could get a little expensive. Overall size is 4x8 feet +/-, about 19 cubic feet minus the hardware, will add an area to hold some silica gel as a chemical air dryer. Hopefully see air temps below -25c with the mb loop just above. Fans and dice side pump will probably all be supplied with standard house current.
15-Inch oval covered aluminum roaster as the main heat exchanger, acetone and dice. Flow channels either soldered or brazed to the bottom, constructed from .250 x 2.5 inch bar aluminum. Place the roaster in the ice chest, pump 60/40 mix of antifreeze water mix past the roaster then to a condenser coil to act as an air dryer, (coldest component for ice buildup not as restrictive to air flow as a radiator), through some heater cores, then a Koolance heat exchanger for the computer side cooling loop. 70/30 mix alcohol/water to reduce viscosity, computer side.
Hard drives and power supply well be mounted outside with something like a Frontx 5.25 bay adaptor for the video, ps2 sata and usb connections. Still working on how to install connections for mb and video card volt mod adjustments and probes.
At this point concerns, questions and thoughts.
1: Going to take a lot of dice, but should level out once to temp, everything will be insulated.
2: The roaster isn’t large enough
3: Roaster is to large, can be controlled by the amount of dice.
4: Flow channels, size to small freezing up, to large not enough transfer.
5: How much water is present in 19 cubic feet of air at 20% humidity, will it overly restrict air flow through the condenser coil and heater cores once frozen?
6: Pump selection dice side, my thoughts (Pan World 100PX Magnetic Water Pump). A lot of head and fluid movement.
7: Need an idea on how to make a circuit board that allows the changing of VRs and will the leads hinder things if overly long?
8: Will it work to expected temps of -25c or better?
Probably build a smaller prototype first out of 6 inch HAVC ducting components, a lot of duct tape and a couple of MDF boxes. If not doable will have some parts for something else.

Air tight construction for recirculation. Fans are conceptual, twenty four 120x38s could get a little expensive. Overall size is 4x8 feet +/-, about 19 cubic feet minus the hardware, will add an area to hold some silica gel as a chemical air dryer. Hopefully see air temps below -25c with the mb loop just above. Fans and dice side pump will probably all be supplied with standard house current.
15-Inch oval covered aluminum roaster as the main heat exchanger, acetone and dice. Flow channels either soldered or brazed to the bottom, constructed from .250 x 2.5 inch bar aluminum. Place the roaster in the ice chest, pump 60/40 mix of antifreeze water mix past the roaster then to a condenser coil to act as an air dryer, (coldest component for ice buildup not as restrictive to air flow as a radiator), through some heater cores, then a Koolance heat exchanger for the computer side cooling loop. 70/30 mix alcohol/water to reduce viscosity, computer side.
Hard drives and power supply well be mounted outside with something like a Frontx 5.25 bay adaptor for the video, ps2 sata and usb connections. Still working on how to install connections for mb and video card volt mod adjustments and probes.
At this point concerns, questions and thoughts.
1: Going to take a lot of dice, but should level out once to temp, everything will be insulated.
2: The roaster isn’t large enough
3: Roaster is to large, can be controlled by the amount of dice.
4: Flow channels, size to small freezing up, to large not enough transfer.
5: How much water is present in 19 cubic feet of air at 20% humidity, will it overly restrict air flow through the condenser coil and heater cores once frozen?
6: Pump selection dice side, my thoughts (Pan World 100PX Magnetic Water Pump). A lot of head and fluid movement.
7: Need an idea on how to make a circuit board that allows the changing of VRs and will the leads hinder things if overly long?
8: Will it work to expected temps of -25c or better?
Probably build a smaller prototype first out of 6 inch HAVC ducting components, a lot of duct tape and a couple of MDF boxes. If not doable will have some parts for something else.













