"mike" wrote in message
...
On 6/23/2012 1:16 PM, Winston wrote:
On Sat, 23 Jun 2012 07:13:01 -0400, Jim Wilkins wrote:

...If I can find a non-destructive way to do it
I'll add a temperature sensor to the cold side to see if it can
serve as
a dew point indicator.

That's not gonna be very accurate.
The peltier is a pump. For a given input temp, output temp
and power applied, you're gonna get a fixed amount of BTU
transferred.
Part of that is sensible heat, part latent heat. It's all affected
by
air flow, humidity, surface effects, ambient temp and probably other
stuff.

First I'll have to determine if the Peltier has enough excess cooling
power to further chill the liquid water after condensing it from
vapor. It might stabilize at or slightly below the dewpoint.

If not I could have an Arduino ramp it up and down and look for the
flat on the cooling slope, or reduce DC power until it does stabilize
at the dew point, which I can measure independently with a lab-quality
psychrometer.

jsw