So, I have to switch to 2.8-pre for JA? Danny
---- Charles Steinkuehler <char...@steinkuehler.net> wrote: > On 8/25/2016 11:38 AM, dan...@austin.rr.com wrote: > > > > So I guess it does do that. Now if one home was physically > > installed where it trips 0.53" before physical end-of-travel, if > > this were NOT the gantry axis I'd just give its final machine coord > > as 0.53" and its machine coord is correct (0=end-of-travel). But > > in this one, say one gantry switch is mounted to trip at 0.5" but > > the other trips at 0.65". If homing acts like non-gantry joints, > > it would physically leave it at 0.5" and 0.65" and leave joint mode > > with it physically out of sync like that. Which would mean the > > joints are racked by 0.15" and will forever be locked like that > > because future moves are in axis mode, not joint mode. > > > > Does it have the ability to physically move the joints into > > alignment based on .ini parameters saying one switch is 0.15" off, > > or do I just need to keep physically remounting one switch until > > its trip point is "close enough" to the other? > > No. On the machines I wrote the gantry component for, typically there > is a small screw used to adjust the tripping point for each homing switch. > > As Andy mentioned, you may want to just use a version of LinuxCNC that > supports JA. When I wrote the gantry component that wasn't an option, > and the behavior of LinuxCNC with any non-trivial kinematics (even > something as simple as a gantry) was very painful from a user > perspective (or at least from *THIS* user's perspective). I haven't > messed with JA, but it's supposedly *MUCH* better at handling these > sorts of machines. > > -- > Charles Steinkuehler > char...@steinkuehler.net > ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users