Dan wrote:

larry wrote:

my 5K gold star window air:
bg5200er
limited schem on top of black plastic control box)
sensor - thermistor(non polar) in black epoxy glob:
oF ohms
212 = 1.2k
finger(98) = 6.6k
warm room(82) = 7.8k
32 = 26k

negative factor (hotter = lower resistance)

draw a linear chart, see if my numbers are on the line

use ohms law

have fun

-larry / dallas

BTW- if anyone trashes one of these, I would like to buy your circuit
board and keyboard/display. I have a GE I want to teach new tricks ;-)


Thanks Larry, very interesting info. If the sensor in mine also turns
out to be a thermistor (a likelihood) it really simplifies tweaking. In
the case of your figures, you could fool the thermostat into thinking it
was 98 when it was actually 82 by shunting the thermistor with 42,900
ohms. Since the range of actual thermistor values over real room
temperatures would be fairly narrow, something along these lines would
probably be safe (only down side is the temp readout would be wrong, but
since it is anyway (says 60 when it's 69) no loss there). Also very
easy/cheap to implement.

Dan

Hi Dan...

With all this great info, seems you can't go wrong!

Another suggestion might be to consider using perhaps a 100k
variable rather than a fixed resistor.

A bit of tweaking might let you make the readout "wrong" by
10 degrees F, so that for instance you could more understandably
have 65 degrees read 75, etc.

Take care.

Ken