These are chat archives for machinekit/machinekit

16th
Jun 2017
zhivko
@zhivko
Jun 16 2017 13:41
Oh there is similar kins there already - just need to adapt for my different coord system... https://github.com/LinuxCNC/linuxcnc/blob/master/src/emc/kinematics/rotatekins.c
I will see if it helps...
zhivko
@zhivko
Jun 16 2017 14:06
This is not working for me:
sudo halcompile --install [name of kins.c]

machinekit@beaglebone:~/git/machinekit$ cd mycomponents/
machinekit@beaglebone:~/git/machinekit/mycomponents$ sudo halcompile --install pipeCutKins.c
sudo: halcompile: command not found
zhivko
@zhivko
Jun 16 2017 14:14
this works:
comp --install ./pipeCutKins.c
zhivko
@zhivko
Jun 16 2017 14:42
try to load MK with my kins I get this:
16:36:35,455 [Thread-39] 1011138 INFO commands.MachineKitStart: cape-bone-iio overlay found
16:36:41,514 [Thread-39] 1017197 INFO commands.MachineKitStart: CRAMPS.hal:19: Program './setup.sh' finished
16:36:41,526 [Thread-39] 1017209 INFO commands.MachineKitStart: CRAMPS.hal:28: Realtime module 'pipeCutKins' loaded
16:36:41,534 [Thread-39] 1017217 INFO commands.MachineKitStart: CRAMPS.hal:33: Realtime module 'tp' loaded
16:36:41,770 [Thread-39] 1017453 INFO commands.MachineKitStart: CRAMPS.hal:34: insmod failed, returned -1:
16:36:41,770 [Thread-39] 1017453 INFO commands.MachineKitStart: rtapi_app_main(motmod): -1 Operation not permitted
16:36:41,770 [Thread-39] 1017453 INFO commands.MachineKitStart: 
16:36:41,770 [Thread-39] 1017453 INFO commands.MachineKitStart: See /var/log/linuxcnc.log for more information.
zhivko
@zhivko
Jun 16 2017 14:48
Oh it loaded now - I missed new kins in
loadrt [EMCMOT]EMCMOT servo_period_nsec=[EMCMOT]SERVO_PERIOD num_joints=[TRAJ]AXES tp=tp kins=pipecutkins
line of hal
Bas de Bruijn
@luminize
Jun 16 2017 14:56
Installing a comp can be done with instcomp. http://www.machinekit.io/docs/hal/instcomp/
zhivko
@zhivko
Jun 16 2017 15:00
Im now on new kin
Bas de Bruijn
@luminize
Jun 16 2017 15:01
Ok!
zhivko
@zhivko
Jun 16 2017 15:01
I will try to get new motion.curr-vel halscope image
Brb in 30min
zhivko
@zhivko
Jun 16 2017 16:13
I actually reconfigured hal to use new kins, but actually in my case I have 4 independent axis - I mean... all 4 axis X,Y,Z and A are not linked however...
So now I am not sure what is right way to calculate motion.req-velocity or why it is going to 0... I will try to zoom in grpah as you @luminize suggested.
zhivko
@zhivko
Jun 16 2017 16:35
image.png
"where do you generate your trajectory?" - I build gcode with some preprocessor... I use inverse time and from gcode I pasted last time I can see that F word is always >0 means I do not force feed rate to 0 anywhere...
zhivko
@zhivko
Jun 16 2017 16:41
I also now disabled thc.comp so I know I get this 0 oscillations of motion.req-velocity also without thc.comp
Bas de Bruijn
@luminize
Jun 16 2017 20:05
What do you mean "all axes not linked" it has to do with the nr of axes in the motion component. You still have 4 axes so it still stops at points of lines.
Yes, also what i said. THC component knows nothing of motion. Motion is the cause
Bas de Bruijn
@luminize
Jun 16 2017 20:10
There might be something else, cause some (blue line) go to zero velocity, while other parts not. You need to find a piece of gcode which you can reproduce this behaviour. I'm not sure if i can help you further. You might have more succes at the Google Group list