Also, some devices show temperature hysteresis, giving them error
which is impossible to calibrate out (at least, unless you want to
bother tracking the temperature history and using it as a correction
factor).

Sean


On Wed, Dec 24, 2008 at 9:53 AM, Marcel Duchamp
<marcel.duchamp@sbcglobal.net> wrote:
> You want to build a temperature sensor. But everything is a temperature
> sensor.  All the parts you would use, the IC's, the resistors, the
> capacitors, the wire, the connectors, the solder, everything has a
> temperature coefficient.  For accuracies around 1% or greater, these can
> often be ignored. But not so at 0.05% accuracy.
>
> Every two or more "things" connected become a thermocouple, generating
> temperature dependent voltages.  Small but finite.  Just like your
> desired accuracy.
>
<snip>
>
> So yes, there is a reason that it's so hard...
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist