Re: [CAD_CAM_EDM_DRO] Threading on a lathe with EMC

> > Is the problem that the hobby machines do not have a proper spindle encoder?
>
>
> No, as hobbyists we should all be capable of fitting one. The problem
> is that the architecture of EMC will need some adaptation to be able
> to coordinate other axis moves to an encoder on a spindle that cannot
> be servoed in position. So your choices are servo the spindle with a
> positioning-class drive or modify the innermost workings of EMC.

Perhaps I should further clarify: the part of EMC that plans
coordinated movements operates open loop in the sense that when it
tells an axis to move it expects that this command is followed
exactly. There is no mechanism for feedback to this level compensate
for an axis that does not respond quite as expected, short of the
extreme case of a fatal following error which will shut down the
machine.

The part of EMC that runs the servo loops does know where the axis
actually are, and operates in a closed-loop mode that compensates for
deviation from the commanded position. But this part of EMC only
takes position commands (which are points on move profiles caluclated
by the motion planner) it does not understand feedrates or
acceleration profiles or tool advance/retract or anything like that.
You could trick it into maintaining a relationship between the Z
position and spindle, but you would have no way to nicely transition
in and out of this mode.

The result is that it's not clear where to implement a hard
feed-per-revolution capability.

It is something I'd like to see work, and would be willing to work on,
as I think it is a gating issue to being able to design an inxepensive
home-shop cnc/manual lathe that does not need an expensive set of
threading gears.

Chris