If I got it correct from other discusssion trajectory is pluggable TPCOMP nanual https://linuxcnc.org/docs/devel/html/man/man9/tpcomp.9.html there it is mentioned custom module is loaded with [TRAJ]TPMOD=tpcomp and source code https://github.com/LinuxCNC/linuxcnc/blob/master/src/emc/tp/tpmod.c
Have professionally done trajectory planner with maximum limit on jerk for other purposes. Some math is needed particularly to get the distance then it is time to apply maximum jerk to stop at exact correct position with maximum value for jerk and velocity. Had to write down equations and spend a few weeks on it. Smoother will require more math. Looking at equation fo electric motors it is possible to change jerk instantly by changing voltage so maximum limit on voltage is probably a good choice. Back-emf will however limit what voltage is available for acceleration particularly to increase speed at high velocity. The resistance will also add some voltage drop but should most certainly be a lot smaller for most electric motors. Nicklas Karlsson sön 2024-12-08 klockan 23:52 +0200 skrev Viesturs Lācis: > Viesturs > > sestd., 2024. g. 7. dec., plkst. 00:28 — lietotājs gene heskett > (<ghesk...@shentel.net>) rakstīja: > > > > On 12/6/24 16:48, Viesturs Lācis wrote: > > > > > > I am fine with workarounds as well. I did find discussion from 3 years > > > ago here on mailing list with a suggestion of putting limit3 component > > > before pid component. That machine has 8i20 servo drives, if that > > > changes anything. pid.n. command is connected to joint.n.motor-pos-cmd > > > at the moment. > > > > > I'm far from an expert on that but a limit3 has done what I needed > > several times. And IIRC it has always been applied prior to the pid its > > controlling,, preferably in such a position as to shape it all in one > > module. > > > > I would like to ask for more details on how exactly to implement this. > Since now the pid input is position, adding limit3 does not make sense > as I would not control the jerk. I need to change the pid input to > velocity. Does it mean that it will change from closing position loop > to closing velocity loop? And I will have to redo pid tuning? > > Or is there anything experimental that I could volunteer to test on > actual machine? > I found this attempt from Chris Morley: > https://forum.linuxcnc.org/24-hal-components/40152-jerk-limited-trajectory-planner-hal-component#184119 > That post is 4 years old. Has there been any update to that code? Is > it worth testing on actual machine? > > Aaand then there is this thread: > https://forum.linuxcnc.org/38-general-linuxcnc-questions/50268-trajectory-planner-using-ruckig-lib?start=0 > > Is there anything else I have missed that is worth a test? > > Viesturs > > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
