"mike" wrote in message
...

I think the theory is sound. But the devil is in the details.
ALL of the input power comes out the hot side. Add to that the
much smaller pumped
power that's a nonlinear function of almost everything in the
environment. Once you start modulating the input power, things get
messy.
I think you're gonna find that the error terms are gonna swamp
the measurement.

I regulated the temperature of the Peltier cooler quite precisely on
that laser diode project. It's not difficult.

First place I looked had humidity sensors for under 5 bucks.
Whole modules under ten. Probably do much better on ebay??
No fuss, no muss.

I have several. They all disagree.


And what are you gonna do with the information?
My motto is, don't measure anything if the result won't change
the future.

I'd like a simpler, possibly automated system to tell when to open
windows and when to run the A/C, based on indoor and outdoor temps and
dew points. It isn't completely clear when the outdoor temp is between
65 and 70 and humidity is high, even if the house is 80F inside.

For example right now outdoors reads 64.4F and 97%, indoors shows
78.4F and 56%, on 6-month-old Oregon Scientific sensors. Earlier when
it was warmer they both read almost identically. By my 3x rule of
thumb a 14F temperature difference is equivalent to a 42% humidity
difference, 56+42=98.

But I haven't completely related those readings to the temperature and
humidity levels I'm comfortable with, which will change as I acclimate
to summer heat.

There's probably no way to sell it as a product, people's tolerances
vary too much and it requires user intervention to open and close
windows.

jsw