Question about homing

[MRX8024]

You mention the following in homing-squence:

My question is: how will turning off the AB motors change the corexy belt tension balance? In my opinion, after parking, when the beam has assumed a "square position", the belt tension difference will become even greater, and the head will shift to a certain position on both axes, because nothing holds it.

Thanks for the hard work!
-Maxim

[ChipCE]

@MRX8024 sorry, for the late reply!

In my opinion, after parking, when the beam has assumed a "square position", the belt tension difference will become even greater, and the head will shift to a certain position

Yes, we turn off the motor and make the head shift on purpose. The unbalance on the AB belt that make the head move when AB motor off, after the head move to its nature position, the AB belt tension will be more balance than before we turn off the motor.
After tuen off the AB motor, the toolhead lost its X homed position, so we have home it once again.

[MRX8024]

Hi, during this time I have rethought the principle of operation a little, and it has become more attractive to me. Indeed, corexy has this problem, unlike hybrid index kinematics. But you do not take into account one thing, the tmc drivers do not just turn on in the new head position, but restore their previous position between 2 fullsteps (turn on the current on two coils from the last position). Because of which, in any case, we will get a certain tension failure between A and B or AB and YY1, this is, it will probably because smaller, but it will still remain. I plan to try to deal with this problem a little later. This is also why it is important to use two limit switches to align the beam with respect to both the activation of the YY1 motors and the primary AB kinematics. But only if they are really exactly posted on the frame.

Perhaps you have some thoughts?
If you don't mind, I'd like to keep this issue open for now.
Thanks.

[ChipCE]

Thanks for the feedback!
In the worst case, A and B motors are in the middle of their full-step at neutral position, and will "snap" to the nearest full-step position in the opposite direction when re-enabled.
That will change the belt length by delta mm. We can calc the stretch by
delta < = (0.5 step * 2 )/motorStepCount * pulley teeth * belt pitch.
That will be 0.2mm with 20t pulley and 200 steps steppers.

In my test machine (about 600x600mm bed), when turned off the motors to release the belt tension, the toolhead will move about 2~5mm, much larger than the 0.2 mm by re-enable the motor.

As you said, the homing sequence helps, but not perfect. I would like to hear more about your ideal to deal with this problem.

Thanks!