On Tue, 22 Jul 1997 21:49:15 +0000 Gerald Lewis <k2mfh@WORLDNET.ATT.NET>
writes:
>At 10:00 AM 7/22/97 GMT+0200, you wrote:

>
>Whats wrong with a 25 cent thermistor and a series 1 % resistor to the
>+5
>supply and possibly a shunt resistor to control range.  The curve of
>the
>thermistor can be linearized by a small look up table in memory or by
>scaling according to the magnitude of the output of the 16C71 or
>equiv.
>on-board A/D converter.


        Good point!  I wonder how interchangable thermistors are.  Can
you set up just one table and use it for production?  Years and years ago
I discovered that the nonlinearity of therimistor resistance versus
temperature was just about the opposite of the nonlinearity of output
voltage of a voltage divider versus the resistance of the shunt resistor.
 Thus, if you choose the pull-up resistor right, you can get a pretty
linear output directly out of the thermistor.  It can, of course, always
be improved by a look-up table!
        another interesting therimistor to use is an automotive water
temperature sensor.  Pretty cheap and nice package fore measuring water
temperature in a pipe.

Harold