Strange PID symptoms (Motion Magic)

I’ve been tuning the PID on our robot’s arm to make the presets more accurate, but have been noticing some strange things happening. The arm is accurate in reaching the setpoint but does so in a strange manner. It moves upward with reasonable speed about 1/2 - 3/4 of the way, but then stops and starts moving up very slowly afterward. Thanks!

Have you plotted the closed-loop error? That should reveal a lot about where you are vs. where you want to be.
Raising your F and P terms may help.

You mixed up parentheses and square brackets when making a link
Here’s the video for those who are wondering.

This. And also use Tuner to see what the active trajectory velocity and position are compared to actual position and velocity. I’m guessing you have a very aggressive F term and a lower P term, but the plot will reveal what’s going on.

1 Like

From what I’m seeing, this is what happens when you have not reached your expected setpoint, and have steady state error. And after a point of time integral is kicking in to correct the collected error.

However, if you are not using an I term, it is certainly weird. I would recommend using a feedforward that is the cosine or sine (based on how you calculate angle of your arm) of the power required to keep the arm in space horizontally.

What type of feedback device are you using and what firmware is installed on the SRX? If it a low resolution feedback like a potentiometer and firmware <4.22 (*NOT 4.20 - NEVER 4.20) this could explain what your see.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.