Heck, in Marlin 2.0 it has come so far and I, for one, don’t want to be detached from the firmware as I like lifting the hood and getting dirty. G0/G1 is coming I read somewhere and ARC is already in it (infant stages though).
No, I don’t but over on the old Prusa forum Josef mentioned it and to bring them back into alignment the old way (before he began to use E1 for Z2 and dual endstops) you simply moved the X carriage all the way to the top and let it grind (no harm was done doing this) until both sides slam against the top then the cycle would repeat. For my old I3 it would go out of sync and with a heated bed I had so much trouble.
Oh, I don’t think that is because of the steppers while moving, I think that is just because they can move independently when they are off. It is exactly the same as dual endstops on XY. You can correct that with two different endstops. But you need to be able to move them independently. The only way to do that is to have them on separate drivers.
Our original solution is just to hold them square before enabling them. I still do it this way and it works well for most.
Over time of a lot of printing (1-6 months) the old Prusa machines would lose tram due to each Z motor never ever turning precisely in sync (no mechanical open loop motor will) so what he decided to do is use two drivers (E1 and Z1) so each motor heads towards the endstop and once hit it stops and the other keeps going until it hits its endstop. Once done you are within the margin of error of the enstop in sync with each other. I would think if that were still being done today but using the series method the same outcome would happen of losing sync. I know this tuning the A4988 for Z was always less available than X/Y/E because you had two motors on a single driver in parallel which, as Josef once said, was done that way for cost savings at the time of design.
With my Prusa printers the Z motors still get out of sync but NOT when they are actively enabled. It happens when the motors are off and I am working with the extruder. Even loading or pulling out filament can cause enough pressure to make one of the z motors move a few steps. Usually the one on the left side because that is where it gets closer most often.
So I agree with Jeffeb3. I thought that they get out of sync only when idle.
Nope, and that was even according to Josef though it can happen whenever but even if you never touched it, and just sat there going up and down with Z, you will notice one side is off after awhile and the off gets worse over time. My latest machine is a CoreXY and I removed one of its dual Z motors and replaced it with a pulley and belt though I used 2mm per revolution acme rods so it takes 4800 steps for Z at 1/16 microstepping but no longer do I have sync issues.
Btw, the dual Z with a Y adapter is on the SKR V1.3 but now the 1.4 has been released and they did what Ramps did and have a connector on the board just mimicing the parallel method (no Y adapter needed but electrically the same).
If the motors are wired in series then the stators of both motors have the magnetic fields in the same configuration. Always. They cannot slip a tiny amount over time, they can only slip a by one whole magnetic pole and this is exactly what happens when steps are lost due to excessive torque for example. I think everyone here is familiar with skipped steps. If this does not happen and the motors stay energized then the rotors will stay in sync because the magnetic field configurations stay in sync.
If you can provide a link for Josef saying otherwise then I would consider it in context but otherwise I can’t see how it would be true.
I am not about to go back through the threads from 2014-2018 on his forum to find him telling people about it and how to correct it every so often and, as I said, his were not wired in series they were wired in parallel.
Did you have any questions left for me or have they all been satisfactorily been answered?
In summation, we have an easy way to do things, and a more precise way to do things that takes a bit more knowledge of exactly why you would want to do them that way.
This is what I have been understanding after reading documentation and trying to catch up on the forum. It is always good to have better granularity in understanding both the systems we are using and the users using them. Dual end stops are something that the MPCNC can do without, just like a Z probe. This fits with the design aesthetic that you have chosen and it makes great sense. Less is more and the basic model and construction is accessible to anyone.