Question Different SMC3 Ki values for each direction of the motor ( G-seat )

Hello everyone,

I'm tuning my G-seat, and am running into a problem with the motor settings. I imagine that most people who have used motors rather than servos have faced this problem.

As you may know, when the bottom flaps of the G-seat go up, there is a lot of weight on it, but when it goes down, there is no resistance (or negative resistance, rather).

If I increase the power of the motor so that the bottom flap (yellow) will reach the ideal line (blue) in SMC3, the motor has no problem pushing my body up. However, if I do this, the motor arm oscillates when the flap goes down because too much power is applied when there is no resistance. If I decrease the power of the motor so that it doesn't oscillate when going down, the flap doesn't reach the ideal line when it goes up, and I lose detail and force.

I have tried adjusting the Kp, Ki, Ks, and PWMmin, but I don't think the problem is with tuning these settings, because there is an obvious physics problem.

I'm wondering if it's possible to have different Ki value for each direction of the motor, as that is a solution to the problem. For example, can the Ki be set to "10" only when the motor arm goes up (and "0" when going down)? I'm not a coder, so I don't know if this is even possible.

Or is the only option to change something physically so that the motor's negative value does not affect movement?

What has been other's solutions to this problem?

It's interesting because I switched to wiper motors/ worm-gear motors from servos because I wanted to have enough power, but I am facing a problem that servos don't have.

Thank you for your help!

Its possible but it should be done with the proportional value since it has the most influence and then use integral only if needed. If "P" and your weight is enough for a flap to keep up with the set point as it drops then the" I" term will stay near zero anyway and won't really do much.

On worm gears motors vs RC servos the obvious difference is gearing. Unfortunately while high friction work gear motors are nice for their self-locking properties they can cause oscillate when power eventually overcomes friction and the worm breaks loose and lunges. Because of this your best option is probably to attempt to counter balance first.
But it just so happens I'm currently developing a new Pid controller to supersede the aging SMC3 that does allow separate up and down PID settings when needed.

Hey, thank you for the response! I don't full understand the proportional vs. integral -- but it's great that you are developing a new Pid controller! The problem I have applies to both the bottom flaps and the back flaps of my G-seat, and I imagine there are builds where balancing the 2dof might be too difficult. Eagerly awaiting your program!

Has there been any update on this? I’m also interested in getting the new app that supports parameters per direction.

Counter-balancing might also mean it will only work for one person or at least people with the same weight. I guess the solution might be to use a more powerful motor such that your weight becomes negligible.

I noticed that SMC3 doesn’t really try to control the descend so giving it its own Kp parameter will still not slow down the descent unless a new parameter is added to indicate the target velocity of the movement. In which case SMC3 will try to counter gravity during downward movement.