Linearizing a Thermistor


Thread Starter

Steve Tomporowski

When you use a resistor in parallel with a 91k thermistor, you get fairly large linear region and then at low temps and high temps the curve levels off. In these areas you can't get enough counts per degree for any kind of accuracy. I know that adding a couple of amps and diodes can stretch these areas out, but before I 'reinvent the wheel', has anyone done this before? Basically I'm looking not for total linearity, but enough counts per degree (10 bit A/D) to do 0 to 450 F. I should be able to get at least 2 counts/degree....

lee eng lock

yellow springs instruments of ohio makes what u are looking for and they call them thermilinears. they have some nice application notes and u can
find ysi on the web.