[apk919]

Originally Posted by xtn

The most efficient use of an external mister would be to get the smallest droplet size, and get the mist spread out over the largest possible surface area of the IC, and get those millions of tiny droplets to vaporize during contact with that surface.

Some of you seem to have a bit of a misconception about cooling the air first, and then having the cooling air flow through the IC. Spraying the perfect mist, by itself, does not cool the air. Put your hand in the flow a couple inches down-stream from the nozzle and you will feel cooler air. By the time the air reaches your hand, you will feel a bit of cooling due to evaporation, but the bulk of the cool effect you feel is just the fact that a cool mist is being sprayed on you. The real (most effective) cooling comes from the evaporation. The exact point of evaporation is what pulls the largest hunk of heat energy out of the surroundings.

So you want a fine mist, located just far enough up-stream to spread out and reach the whole width of the IC, but not so far up-stream that a lot of it has evaporated by the time it gets to the IC. So yes, you want the liquid water (albeit in tiny mist droplet form; a single squirt-gun stream aimed at the middle won't be nearly as effective) to reach the IC in the liquid state.

If the IC is not hot enough to turn it all into steam, then use less water, or just accept liquid blow-by. Not like you should complain about it. "Awww my IC is too cool to vaporize all the mist," is kind of a silly complaint, don't you think?

xtn

Hmm.... as long as the droplets evaporate completely, I'm not sure whether it matters whether that happens first in the air going through the IC, or on the surface of the IC itself. The latent heat of evaporation still has to be carried away by the air flowing through the IC. If there's not enough air to remove the water vapor, it will eventually become saturated and additional droplets on the IC will not evaporate. So I think bottom line, it's the air's ability to remove heat from the IC, either by starting at a lower temperature (misting the air), or by removing water vapor from the IC (spraying directly on the IC).

EDIT: Also, spraying directly on the IC would only get mist on the first few mm's of surface area from the front of the IC... theoretically you'd want the entire surface area of the IC to be more efficient (true evaporative chillers are designed to get a maximum surface area wet), and it seems the best way to do that would be to cool the air, which is going to contact the entire surface area.