-----Original Message-----
        From:   Spehro Pefhany [SMTP:speff@INTERLOG.COM]
        Sent:   Wednesday, August 21, 2002 4:06 PM
        To:     PICLIST@MITVMA.MIT.EDU
        Subject:        Re: [PIC]: First pic project

        At 09:31 AM 8/21/02 -0400, you wrote:


> Over a halfway wide temperature range, polynomials are quite ill-suited
> for
> this purpose. I've tried it up to about 9th order. It may well be possible
> to do a couple of operations first to make the equation more suitable for
> polynomial approximation.
>
> Your suggestion of low-order polynomials would probably work if the range
> was split up into segments, along the lines of classic spline
> interpolation
> algorithms.
>
>
Forming a potential divider with a resistor value equal to the thermistor
resistance in the middle of the desired range has a linearising effect.
Combined with a table and simple first order interpolation this can yield
surprisngly good results, but again over a limited temperature range.
Towards the temperature extremes the non-linearity becomes progresively
worse, and the slope reduces to the point that you start losing a lot of
resolution.  Then again, thermistors are generaly only used for relatively
narrow temperature ranges or for cheap and cheerfull non-precision
applications.

Regards

Mike

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.