Jr. and mini all have fixed bed, the bed are fixed by three screws, and add soft washer of different thickness under the bed allow some flexibility to slightly adjust bed level with the screw. The adjustment is not as convenient as bed leveling screw other printer equipped, but it will be more stable once you get the bed leveled. combined with adjustment of z-offset, you should only need to adjust two screws to get the bed leveled.

I guess it might be easier if you can make miniMover to adjust the auto calibration data manually.

As for manual leveling, unfortunently the mini/nano printers have a fixed bed with no leveling screws.

One thing I noticed is the limit switches are just loosely fitted in and can move around a lot. There optical so in theory that is ok since nothing touches them but it seems all to easy for a bump or vibration to move the switch.

I may try a dab of hot melt glue to lock them in place hopefully it will be easy to peel that back up if needed. In any event having a loose limit switch can’t be good for repeatability.

Sometimes it is my fault that I didn't clean up all the debris from the print bed or reapply the glue evenly on glass bed, but the point of auto calibration is to save my time, not add more babysitting to the print bed. Other times, I can not figure out what's wrong with it, I power off the printer and power back on, it will work all the sudden.

It is a lot easier to level your print bed manually, and then it will stay leveled for a month or so, I just send the print file and go do whatever I want to do, it will print nicely and reliably. And with small adjustment, it is good for another month or even longer. With automatic calibration, I have to watch it and make sure first layer is good before I can let it do the rest of the printing.

I think other brand auto bed leveling works better than Da Vinci, but they also require good cleaning of print bed and nozzle(for FSR based sensor), manual bed leveling is still my first choice.

Anyway on my machine this cycle does not happen just once but keeps going for a very long time, eventually either erroring out or finally catching and then running the rest of the calibration routine.  The question is why?

I'm assuming either my z axis limit switch is buggy, or my detector switch is buggy, but it is possible that the motor driver is borked or the coupler on the z threaded rod is loose.  I was hoping someone else may have an idea of where to start looking.