I'm sure this is a basic question but I am unable to find an answer anywhere online that explains it in a simple way.

I am researching NTC sensors and figuring out how to use them.

I want to know that the advantages or disadvantages of changing the NTC resistance will be ?

So far I have the assumption that a larger resistance ( 50K ) could meausre a greater range of temperatures ? ( than say a 10K )

But if both are 1% tolerance the larger resistance would not be as accurate because the range is bigger ?

Am I on the correct track here?

Also I take it that in my voltage divider circuit I should get a better tolerance resistor if I want the reading to be more accurate ? ( 1% instead of a 5% )

I am looking to measure temperatures of around 90 - 130 degrees C, with an accuracy of around 1 or 2 degrees, if that makes a difference.


1 Answer 1


Typically the usable range will be the same regardless of the resistance at 25°C. The datasheet will tell you what range it is rated for and the accuracy at various temperatures (typically it will be most accurate at 25°C and less accurate the further you are from 25°C).

The dynamic range is an issue with thermistors. At high temperatures the resistance will be much lower than at 25°C so it may be more convenient (or require less expensive circuitry) to use a higher resistance thermistor, and if you have long wires to the thermistor their resistance may start to have an effect on the measurement.

Another factor is self-heating. You can calculate that from the power dissipated in the thermistor. Note that thermistor data sheets typically will show self-heating (and response times) in very favorable conditions such as water moving at 1m/s. In still air the performance will be much worse.

Typically NTC sensors change around 4% per degree C, so a 1% resistor is called for if you are looking for 1-2°C accuracy. Also an accurate thermistor- not only the tolerance at 25°C but the tolerance on \$\beta\$.

  • \$\begingroup\$ I am planning on using a 100K NTC having 1% tolerance and 1% beta accuracy. Would a 1Meg NTC be even better? Besides self-heating, why shouldn't I choose the highest resistance? And why does 10K seem to be the de-facto standard? Or should this be a bona fide question? Thanks, Spehro! \$\endgroup\$ Commented Jul 10, 2021 at 23:45
  • 1
    \$\begingroup\$ It has been a while since I worked on this project but I think you don't want to use a 1M Ohm because then your Voltage divider equation will skew to one side and you wont have much of a drop across your other resistor, I can't explain this perfectly but I am pretty sure that its important. For my project I used a 100K NTC and a 4.7K with 5V and it worked pretty well, hope this helps you. \$\endgroup\$
    – Spider999
    Commented Jul 11, 2021 at 6:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.