On: Show HN: 2D PID controller simulation

[sjtrny]

It's a shame that the PID values aren't exposed directly in text boxes but you can affect how the ball behaves (react faster/slower or cause under or overshoot) by adjusting the kp, ki and kd values.

[nikital → sjtrny]

That's a good idea, I'll add them.

Edit: Done.

[icegreentea → nikital]

You should consider resetting your integral error when you accept new set points.

[nikital → icegreentea]

I agree, I pushed a new update.

Also, I adjusted the defaults for a snappier response with a bit of satisfying overshoot

[phkahler → icegreentea]

This is a special case. If there were gravity acting on the ball the integrator would be required to maintain altitude.

[icegreentea → phkahler]

Ya, but then you would also probably use a separate controller for each dimension.