|
Looking for input here, related to the IJK radius
error, even when perfectly co-incindent with commanded values. i.e. whole number
values.
Bits of reply/fix provided by Max Heise a month or
so ago...
>
>I had the same problem a while ago. >Look at src/rs274ngc_new/rs274ngc.hh at the beginning an decrease these >values. > >/* numerical constants */ >#define TOLERANCE_INCH 0.0002 >#define TOLERANCE_MM 0.002 > >Once you have changed them => recompile. At inch tolerance .0002 I still get the "radius is different" errors, so I
think I need to change the mm values. (0.002mm = .0000787")
My theory on it is this: it isn't so much the
accuracy of EMC, but of the encoder connected. My encoder resolution is
0.0000488 per inch, which doesn't divide evenly into 0.0000787 per inch, hence
every so often I get the IJK interpolation error. The linear allowances not
kicking up errors because their tolerances are set through the .ini. Tolerance
values being applied to the commands outgoing to the trajectory planner.
(?)
I'm going to try setting it to a
multiple of my encoder's resolution (in mm). And see what happens here.
Comments? - Keith Rumley
0.0000488" comes from 1024 line encoder on 5:1
ballscrew, with 4 edges counted per 'line'. 1024 lines * 5 rev per
inch = 5120 lines counted per inch. 5120 * 4 transitions =
20480 transition counts, 1" / 20480 = 0.0000488"
0.0002 encompasses one line (4 transitions),
0.002 only covers 1/4 line (1 transition)
|