nuclear
08-20-2004, 08:21 PM
Realizing it's nearly impossible to condensation proof my hardware in the humid environment I'm working in, I began looking at submersive cooling.
Glanced over the few articles I could find that ranged from using LN2 to mineral oil; and regardless of liquid used there were some common problems such as circulating the liquid to prevent hotspots and not freezing the chips and caps to the point of failure.
I was wondering if it would be a viable option to set up a chilled rig in the traditional way, and then submerse the hardware in mineral oil to combat condensation. I know, most people would just recomment better insulation...and I'm more than open to some suggestions that would actually work; "generous use of insulation" just wasn't enough...the room is probably near %100 humidity.
While keeping the blocks and hardware insulated in the normal way, this would allow some cold transfer to the oil...but hopefully not enough to freeze the components to the point of failure and eliminate the formation of condensation. This would leave only the problem of liquid getting into the caps. (I haven't seen this problem actually verified but have read where this can happen.) That problem should be solved by using sealstring or silicone to prevent this. I suppose another issue could be any non-cooled components submerged in oil could be a problem as they would no longer be aircooled.
Glanced over the few articles I could find that ranged from using LN2 to mineral oil; and regardless of liquid used there were some common problems such as circulating the liquid to prevent hotspots and not freezing the chips and caps to the point of failure.
I was wondering if it would be a viable option to set up a chilled rig in the traditional way, and then submerse the hardware in mineral oil to combat condensation. I know, most people would just recomment better insulation...and I'm more than open to some suggestions that would actually work; "generous use of insulation" just wasn't enough...the room is probably near %100 humidity.
While keeping the blocks and hardware insulated in the normal way, this would allow some cold transfer to the oil...but hopefully not enough to freeze the components to the point of failure and eliminate the formation of condensation. This would leave only the problem of liquid getting into the caps. (I haven't seen this problem actually verified but have read where this can happen.) That problem should be solved by using sealstring or silicone to prevent this. I suppose another issue could be any non-cooled components submerged in oil could be a problem as they would no longer be aircooled.