On Sat, 10 Dec 2022 at 23:14, Dewey Garrett <dgarr...@panix.com> wrote:
> a) not reproducible if all joints set immediate homing
> as there are no residual position errors
>
Thanks for that, the original problem description was very garbled but
seemed to be indicating that there was no movement during homing on his
machine, and so I had deliberately set immediate hiking for most of my
tests.
The underlying issue seems to be that homing doesn't actually get all the
way to the defined home position. In fact it seems to be 2.5 * CART_FUZZ
away from the defined home position.
Interestingly g2 x0 y0 z-1 j4 f1000 gives the error but g2 z-1 j4 f1000
does not, because that uses the actual not-quite-zero starting point.
I don't know how important _this_ bug is as it would be very unusual to
machine an arc immediately after homing.
A kludgy workaround is to just widen the tolerance distance at which a full
circle is assumed. CART_FUZZ is 1e-8 units (appears to be the units of the
config)
if (d < 10 * CART_FUZZ){
circle->angle = PM_2_PI;
}
But I think it would be better to set the axis position _exactly_ to the
home position at the last point in the homing sequence.
--
atp
"A motorcycle is a bicycle with a pandemonium attachment and is designed
for the especial use of mechanical geniuses, daredevils and lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1912
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers
