Re: [CAD_CAM_EDM_DRO] Re: EMC Contouring Settings/ downloading BDI

Art Fenerty wrote:

> Jon:
>
> Yes, I was under the impression that a move like G01X.1A180 should ramp
> the 180 as it is the longest move. But in fact the system computes the move
> of .1 in the X and "fits" the 180 degree change into that movement, which
> causes A axis to lose pulses as it goes too fast. What if A is a linear
> axis?

Then it is trivial, just computing the actual feed rate of any combination of
linear
axes. But, when a rotary axis is added, then the radius of the tool to the
rotary
axis's center is added to the mix, and the G-code DOESN'T define that in
any standard way that the control knows about, so there is NO WAY to
compute the actual feedrate in linear units/time anymore. This is a rough
spot in the design of the RS-274D language. I don't really know what
other commercial controls do about this. I suppose you could force the
definition, and demand that if the A axis is parallel to X, then Y=0 and Z=0
must put the tip of the tool on the A axis. Then, the control has all the info

it needs to properly compute the feedrate at the center of the tool.

This only works for these indexing head on the cartesian machine table
4-axis setups, not for the machines with the swivels on the spindle head.
On those, the radius is fixed (for each A, B, C axis) by the gimbal mechanism,
and the computations get grotesque, but the machine builder can at
least specify the radius and be done with it. (Radius actually equals the
gimbal to spindle reference plane + tool length (these machines USUALLY
don't have quills on them)). Tool length then has to be specified anyway
to have any idea where the tool tip really is when the ABC axes are not at
zero.

Jon