Machine gantry is misaligned and locks up when a program is run
Have you tried manually realigning it? I can come by between 5 and 6 to try that.
Yah I tried but no luck so far
1 Like
i just had a thought- it is very possible that the new controller also has the gantry channel set with a posedge/negedge mismatch.
the two motors each have their own drive but that is commanded by a shared pair of step/dir (which really should never be done that way but, well, they did it anyways)
you’d think that would guarantee they’d step in sync, even if the net movement was wrong. but if it’s switched on the wrong edge, it’s a race condition. it’s very possible one drive could read a step as fwd while the other registers it as a reverse. it depends on how the signals are implemented but usually the ambiguity only occurs on the one cycle where direction changes and that is like 0.01% of the steps so the error is subtle and builds up over time.
the two motors can get further and further out of sync and, given how powerful the gearing is, the racking can physically bend and damage the gantry. this is wholly consistent with what i saw with it a few nights ago. i didn’t have an easy way to measure it but the gantry was showing clear signs of having been bent out of shape even with the controller unpowered and the motors freewheeling. this was confusing how that could be, but an edge mismatch would explain how that happened
and you probably wouldn’t see anything wrong on an initial inspection. it will seem ok and could take time, even hours of running continuously to build up a destructive amount of sync error. when powered down the motors freewheel and release most of the built up sync error
the interpretation of posedge/negedge is actually not entirely the device config. the drives are opto input with two ungrounded pins and you can easily wire them the wrong way and inadvertently reverse the edge. since the controller was just replaced, all those wiring connections are new. this strongly suggests that’s the source of the problem
also, if the gantry was indeed physically bent as it seems to be, it can be unbent in the same way- with the motors, by intentionally desyncing the drives and forcing the exact amount of counter-bend, then resync them. it’s a tricky process but sounds like the best way to do it
The red tag had been removed from the Swift. The repairs were completed tonight, and the testing went well.
As with any machine upgrade, it would be beneficial to keep a close eye on it for the first few runs.
2 Likes
what did you fix?
I corrected the wiring.
1 Like
is there an updated unstruction sheet for the new controller ? It looks great - but currently guessing at how to make things work. There are new sheets - but they are not working either
There is a video on the desktop, and new workflows in the pocket of the machine.
1 Like
I used the CNC today. Worked flawlessly. I like the new controller. I am a newly though. Brian was also great. He stopped by to see if I had any questions. He was a great help.
I am sure I am missing something simple but i keep having the same issue ;
initial start up and setting xy good .
Once I have completed a run it won’t allow me to reset the xy to a new position and has the same error on restart .
The error on the controller i get is "in machine coordinate state - please switch to workpiece state
That happens occasionally. On the old controller, Menu + High/Low switched back to the normal state, but can’t say for sure about the new controller.
User may have to press Home + Number (usually 1) to switch to workpiece coordinates.
I encountered this yesterday with a user. I’m not sure what the scenario is that causes the controller to switch coordinate systems.
got it ! Thanks all !