Use slide potentiometer for axis positioning

[Kofferkulli]

@lyndwyrm666 came up with the idea to use a slide potentiometer on every gripper axis. Ideally this will make the end stops and calibrating obsolete as we can read the resistance value and map the position of the axis onto it. Kind of like a servo motor. Alternatively we could use an absolute value encoder, which is more costly. Or an incremental encoder, with which we would need to keep the end stops for referencing.

We would gain more speed as we do not have to reset the gripper and reliability as we would always know whether the stepper motors missed steps or not.

[nlimpert]

To me this sounds rather risky as from my experience reading raw analogue values from potentiometers tends to be quite noisy.

Also I do not see why we would have to give away the accuracy from counting steps just because we're saving time during initialization of the gripper (or re calibration in general). Did we ever really lose steps?

IMO this would only add one source of errors for no real benefit. I doubt that reading a linear potentiometer can be as accurate (and reliable) as using the current resolution that we have by counting the steps.

But maybe @lyndwyrm666 can give a more elaborate explanation on this.

[lyndwyrm666]

You can get high precision potentiometers. I only suggested Potentiometers as an easy to use alternative to optical distance sensors. As far as I understood the discussion we would like to get rid of the calibration because it is unnecessary movement. (Also I would like to get rid (or use it at different times) of it because we lost some workpieces during calibration in the test games.)

[vmatare]

They do tend to get scratchy, especially when they gather dust due to not being used for a longer time...

[Kofferkulli]

I will check if we still miss steps, in future test games. Until then I will put this Issue on hold.