Thread Starter


Hi everybody.
This is my first time in the list.

Can I connect a thermistor directly to a PLC input?

Thanks everybody

Martin Rimmele

Curt Wuollet

Yes and no;

With a current source, or even just a series resistor, you can derive a voltage that you can measure that is a function of temperature. That function would require a look up table to linearize or a set of ( I believe)

three linear equations in three variables. The matrix solution could perhaps be coded on a PLC, but the lookup would be easier. I would expect that many companies offer signal conditioners to do this but a small program to crunch out the table values isn't too difficult either. I might have one someplace if you get stuck. The Omega site might have the equation up.


Curt Wuollet, Owner
Wide Open Technologies

You could take the signal, amplify it and put it through a V-I converter (or leave it at V if short run). Burr-Brown have such a chip for approx $11. This approach requires the need to construct the circuit and board but good app notes are general available for this.
An alternative is to source a signal conditioning puck that has a similar input range to the thermistor (anything around 3-4 volts). The
signal generated is nonlinear but it is a simple matter of taking a number of point from the graph and generating a polynomial. The order required will depend on accuracy (and response time required to a lesser extent). I think you would find a 4th order adequate. If you do not have the software to generate the poly I could assist if you provide the graph of the signal output.

Allan Dow
Embedded Systems & Solutions
More on Thermistors:
Volume 26 Omega 1998 Section F is what you need to read first. They also offer linear devices.
Look closely at the linearisating equation they propose (F-9). In contradiction to previous answers, this type of device is not of polynomial format at all. Any attempt in that direction is waste of time, unless you would know other avenue approaching polynomial format.
Watch the interchangeability of your thermistor.
In case you are still unhappy after your Omega education, let me know. Being specialised in numerical approximation of functions, I would give a rational approximation. I only need pairs of coordinates spanning your device.
Why have you not used type T thermocouple.
That one I love, it is well established from absolute zero to nearly 400 deg C. I can calibrate
and personalise an installation c/w the extension line down to approximately .01 deg C.
In that situation, each Thermocouple has its own individual linearising set of coefficients.
dear sir I am PG student of control system in my dissertation work i have developed one software for transduer lineaisation. Which linearises thenonlinear char. of sensors like Thermistor. Now I have to prove the advantages of linearisation over nonlinear curve. Please suggest me the solution How i can prove it?