Potmeter feedback on linear actuator - Low pass filter

I am in design of a custom shield for an Arduino Due, to control the linear actuators of a small scale stewart platform to simulate offshore vessel movement. This PCB will sit on top of the Due and will include three dual bridge motor drivers (TB6612FNG) together with a power supply circuit (12V) and current sensing circuit.

The motor drivers is to PWM/DIR control the six linear actuators from Actuonix (P16, 100mm stroke, 22:1 & 46mm/s). These actuators has built in potentiometers as position feedback. The potentiometers are listed with 11k ohms +-50% and less than 2.00% feedback linearity. Picture from datasheet is attached in post.

I do not have the actuators yet so i cannot do any readings/measurements... The potentiometers will be wired to a 3.3V ref signal from the arduino and wiper will be connected to one of the analog inputs. What i want is a simple passive low pass filter on the analog input signal with a resistance and a capacitor. To prevent the likelihood of high frequency noise from the brushed DC motor inside the actuator as well as other noise sources. I am aware of filtering in software with moving average readings and similar to smooth out the signal. This will probably also be done, but i want the "worst" noise in my readings to be filtered away before it reaches the arduino's ADC.

Is there someone who can help me with how i can find the right cutoff frequency that filters out as much as possible, but not the actual readings. What parameters do i need to consider in order to find the feedback signals working frequency? If someone has any other tips or considerations to this problem, it is much appreciated.

I use a 100 nF capacitor from the wiper to GND to reduce noise, but you might want to use shielded wire from the position pots to the Arduino as well to eliminate radiative pickup.

There is probably no need for an additional resistor.

If a pot is used for position feedback, you have to be careful not to filter out too many
frequencies or you could reduce the stability of the control loop by adding delay within
the feedback path.

Screened cable for all the pots is better starting point I think.

You'll need to have a high sample rate for the same reason.

This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.