Re: [CAD_CAM_EDM_DRO] Re: re:OK - you convinced me - was "Automated "Manual CNC"

Brian Collier wrote:

> Thank you once again for the informative answer!
>
> I am a little confused when you talk about backlash being a problem
> in my proposed set up. In open-loop or closed-loop with position
> feedback from the motor, I understand about backlash being a problem.
> The PC has no way of knowing the true position of the workpiece due
> to the backlash. However, with closed-loop with feedback from the
> table, couldn't the CNC program just work around it?

If the backlash is small, around .001", it can do fairly well. If the
backlash is
large (my old Acme screws were up to .050" on the Y axis) it won't work
at all. When following smooth curves, like a circle, it requires a very
smooth
slow down to zero speed, then reverse and slowly accelerate, without any

jump or jerk. If there is much backlash, the motor will have to
suddenly jump
to the other side of the backlash, and it inevitably will cause the
table to jerk.

The other problem with backlash is you can't climb mill. I do almost
all
my work climb milling, because it gives better surface finish. If you
have
much backlash, the cutting forces can pull the work into the cutter,
causing
tool breakage, or workpieces being pulled out of the vise and thrown
across
the room.

> My encoders probably will be digital caliper heads running on
> lengthened scales. Output would be through the SPC port.

The SPC port gives 1 to 2 outputs per second, and is of no value at all
for
servo control. My EMC system reads the encoder position 1000 times /
sec.

Jon