h3rk

Thanks for sharing the empirical information, turbocad6. That is very useful. I would have surely fried some stuff.

What if the computer had a climate controller that never allowed the cooling air to get too cold (by using its own heater/temp control door)? It can use RH sensor & temp sensors to select, from mapped data, a temperature that is closer to ambient; close enough to preclude condensation. Then the metal components of the computer will not 'chill'. I know there is an abundance of dew-point and related data that would be helpful. This system would benefit from being sealed when not being cooled by its fancy climate control (but that would be tough). And timing the puter to remain on longer than the cooling system, long enough for temps to rise above ambient (or until triggered by sensors/ setpoints), would negate ill effects of the chilled components with warm damp air (to that of at least more normal arrrangements).