I'm trying to switch four analog thermistors to one main thermistor reader I've tried to use a CD4066BE to switch them but then I've just found out that it doesn't work as they don't let through the resistance.

How can I go about switching four different analog thermistors to one main reader?

  • \$\begingroup\$ The 4066 has a Ron of ~125 ohms. What are the resistances of your thermistors? \$\endgroup\$
    – vir
    Oct 26, 2022 at 20:53
  • \$\begingroup\$ If you excite the thermistor directly, rather than through the mux, you eliminate a lot of the mux errors (like on resistance). \$\endgroup\$
    – SteveSh
    Oct 26, 2022 at 21:12
  • 1
    \$\begingroup\$ Can you sketch your circuit for us or provide the schematic? Otherwise, ten of us are going to come up with ten different schematics based on your verbal description. \$\endgroup\$
    – SteveSh
    Oct 26, 2022 at 21:14
  • \$\begingroup\$ The minimum resistance of your thermistor (at the maximum temperature for an NTC thermistor) and your tolerance for error at that temperature will determine the maximum switch resistance for a naive multiplexing arrangement. There may be other better ways depending on the circuit used. 4066 switches have a relatively high resistance, especially at low supply voltage, and it varies significantly with input voltage. They're cheap though. \$\endgroup\$ Oct 26, 2022 at 21:26

2 Answers 2


Like many cheap MOS switches, the CD4066 has a relatively high on resistance. That's not the worst of it, because this resistance is very variable, with supply voltage, switch voltage, temperature, and specific IC. If you're trying to measure resistance, this variability shows up as a direct error in your measurement.

It's far better to use the MOS switch as a voltage multiplexer. This way, the variable resistance of the switch does not affect the reading significantly. Give each thermistor its own divider resistor. Buffer the output of the switch if the input to your ADC is very low impedance.


simulate this circuit – Schematic created using CircuitLab

If you really want to measure thermistor resistance through switches, then choose thermistors and switches that have at least 100:1 difference in their nominal values. You can get MOS switches with resistance down in single ohms. You can get thermistors with resistance in the 10s or 100s of kohms.

  • \$\begingroup\$ Yes. I've used this exact scheme with good success on several projects. Like @Neil_UK said, this takes the on resistance of the mux out of the picture. \$\endgroup\$
    – SteveSh
    Oct 27, 2022 at 11:55

enter image description here

I'm surprised you had trouble with using a CD4066. I have done exactly what you are describing, also with other 4000 series and 7400 series analog multiplexers. The only thing to keep in mind is calibrating out the error the analog switches introduce. I would try it again and characterize the system methodically. That being said, mux selection may be tricky depending on our accuracy and linearity requirements.

And as Steve said in the comments, you may be better off with a common current source exciting the thermistors instead of the voltage source I have depicted below.


Your Answer

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

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