Re: mdi g0

[Fred Proctor <proctor-at-cme.nist.gov>]

Ray Henry wrote:

> I'm currently using the 18-Oct 2.0.36 release with steppers.  When I mdi a
> g0x8y4z1 all axis arrive at the same time.  It looks like the motion
> process is still doing a linear interpolation rather than sending each axis
> to its commanded position at rapid feed rate.

That's right. Rapid moves are done exactly like linear moves, except
that they run at the max feed rate set by [TRAJ] MAX_VELOCITY. This is
the speed of the tool tip, so each axis will run more slowly than this
if more than one axis is specified.

It is relatively easy for me to change things so that at least one axis
moves at the max rate. The axes would still finish at the same time, but
the overall time for the move would be the same as if all axes moved at
the max rate.

What you describe, in general, is joint motion instead of Cartesian
motion. These are difficult to separate conceptually in a Cartesian
machine, but for a hexapod joint motion means moving the struts
independently. This leads to all sorts of problems with struts binding
in their sockets. However, it is a very useful mode, particular with
machines like robots. Ideally you could switch between the two motion
modes on the fly, so that you could move along a line, switch to joint
mode and continue motion through a singularity, and pick up along the
same line at a different arm configuration without stopping.

I'll look at the simple fix, for now, since that will enable the machine
to run as fast as possible cycle-time wise.

--Fred