On Fri, Oct 23, 2015 at 03:57:38PM +0800, reise tiet wrote:
> Thank you,Rob,
>
> It's really the genserkins.c 's bug, not the Motion's , I made a mistake in
> the data checking
>
> But,I still want to know the representation of ABC's value, who can tell
> me?..
LinuxCNC does not enforce one single definition of rotational axes. It
is up to the kinematics to select whatever model is desired.
In reading the genserkins source code I see this block in
kinematicsInverse:
// FIXME-AJ: rpy or zyx ?
rpy.y = world->c * PM_PI / 180;
rpy.p = world->b * PM_PI / 180;
rpy.r = world->a * PM_PI / 180;
go_rpy_quat_convert(&rpy, &haldata->pos->rot);
suggesting that the values are interpreted as roll-pitch-yaw by
genserkins.
If you prefer a different representation, then you could change the use
of go_rpy_quat_convert/go_quat_rpy_convert to something else.
Jeff
------------------------------------------------------------------------------
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers
