Use a higher impedance thermistor.  I typically use thermistors that are
470K at room temperature.  This allows a very small, stable polyester cap
(like you) and very short charging times (milliseconds).  Make the duty
cycle of measuring the thermistor very low (1%)  and keep everything at zero
volts between cycles (critical to get the capacitor REALLY discharged, no
soakage.  )

I've used this method a lot, with really good luck, in oven controllers.

-- Lawrence LIle



-----Original Message-----
From: M. Adam Davis <adavis@UBASICS.COM>
To: PICLIST@MITVMA.MIT.EDU <PICLIST@MITVMA.MIT.EDU>
Date: Tuesday, February 22, 2000 2:26 PM
Subject: Thermister heating itself, how to resolve?


>I am using the POT command which is shown on Dontronics Web site to read
the
>resistance of a thermister. (Pin to thermister, thermister to cap, cap to
>ground)  The sequence is as follows:
>
>0. Bring pin high, wait 5 * R * C for the capacitor to become fully
charged.
>1. Bring pin low for an instant
>2. Turn pin into input
>3. If pin is high, increment counter, and goto step 1
>4. Return counter value
>
>I notice that the faster I call this routine, the higher the temperature
the
>thermister returns.  I'm assuming that the initial charging (step 0) is
warming
>the thermister up.
>
>I'm wondering if there is another method that is commonly used which
wouldn't
>heat the thermister up as much.  I'm using a polyester cap as well, do they
leak
>much?
>
>I'm thinking about leaving the pin low all the time, which should only
induce
>current if the cap has a charge, and counting the time it takes to charge
it up
>to the point at which it trips the pin.
>
>I would appreciate your comments and experiences!
>
>-Adam