3
\$\begingroup\$

I'm developing a data acquisition board for 8 thermistors hooked up through 3 feet wires. The board is gonna be installed inside a package unit air conditioning. In there, we have compressors, fan motors and contactor generating EMI. I was wondering how to protect those analog inputs without degradation the thermistor resistance the microcontroller's ADC is measuring. I saw here and in other forums some alternatives based in 100-ohm resistor and TVS diodes. Also I took a look at the ADG5412xx, but it is for voltages higher than 3.3 v which is the MCU operation voltage. I would appreciate very much some help with that.

\$\endgroup\$
1
  • \$\begingroup\$ Take a lot of readings and filter. Temperature does not change quickly. \$\endgroup\$
    – DKNguyen
    Commented Apr 23, 2019 at 4:41

2 Answers 2

2
\$\begingroup\$

Do this on the inputs. The variable resistor is the remote thermistor.

How concerned are you, about voltage drop across the 10,000 ohm resistors?

Notice we use the 1uF to heavily filter the trash injected.

The assumption is that the injected electric-fields and magnetic-fields have ZERO average current, and thus heavy filtering ---- by a capacitor or in software ---- is all you need to do.

Thus I'm having the 1uF capacitor (3 of them) replace the TVS.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
1
\$\begingroup\$

If the major concern is AC-coupled interfering signals, a simple approach is to sample the thermistor value often, then sum the samples over a tenth of a second. Either 50 Hz, or 60 Hz, AC may be superimposed, but the average over several full cycles will only capture the DC component. Temperature variations over short times are also rejected, of course.

Other measures (twisted-pair wiring, shielding, common-mode chokes) are possible, but add hardware complexity. Filtering the signal before conversion is also possible, but that depends on the excitation of the thermistor being constant (and causes some measurement offset due to self-heating of the sensor).

\$\endgroup\$

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.