Advanced Search

There are no
N75 M3 $0 S1
in the gcode after the first piece, and no M5 at the end of the cut except for the first cut.
Those must be added by the post processor.

 
 remap=m6 modalgroup=6 prolog=change_prolog....

maybe I solved the problem, I removed one line of code from ini.file, I don't know if it was a good solution but for now everything works

Hello All, I'm using this post processor (PlasmaCv2.1) in Fusion 360. My issue is that my first round hole in the gcode cuts no problem, machine come downs and probes, makes a perfect cut and then the Z axis rises 1-2inches and continues along the path in the gcode but doesn't come back down to start the plasma arc, it just runs the whole program without coming back down to initiate arc and cut the material. I don't get an "arc okay" error which is odd .I went back into Fusion360 and set my retract height and clearance height to 0 to see if that would help and maybe the THC via THCAD through a MESA 7i76 board would work but it just gave me an error. "Safe Traverse Height has been reduced to 0.23in"


Images attached of my cut parameters ,material properties , gcode from Fusion 360 and some misc info. 

The machine cuts simple parts using the conversational in QTplasmaC no problem, and my THC works so I've isolated it to the Fusion 360 Gcode and unsure how to fix my issue. Possible bug maybe?

I feel like I'm doing something really dumb and overlooking something really simple, in an effort to not waste another couple days going in circles, could anyone point me in the right direction, i greatly appreciate it!
 

I'm trying to set the coordinate system based on the user's selection of a combobox from the on_changed handler. From the documentation only found the methods ACTION.SET_MDI_MODE() and ACTION.CALL_MDI() listed but no examples of their use. I've tried the following:

Gcode = 'G55'

ACTION.SET_MDI_MODE()
ACTION.CALL_MDI(Gcode)

Which gladevcp complains about when starting up as follows:

Traceback (most recent call last):
  File "/home/scsmith1/linuxcnc/configs/mmMill/gvcp_handler3.py", line 53, in on_changed
    ACTION.SET_MDI_MODE()
  File "/usr/lib/python3/dist-packages/gladevcp/gtk_action.py", line 143, in SET_MDI_MODE
    self.ensure_mode(linuxcnc.MODE_MDI)
  File "/usr/lib/python3/dist-packages/gladevcp/gtk_action.py", line 675, in ensure_mode
    truth, premode = STATUS.check_for_modes(modes)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3/dist-packages/hal_glib.py", line 954, in check_for_modes
    self.stat.poll()
SystemError: <method 'poll' of 'linuxcnc.stat' objects> returned NULL without setting an exception

Does anyone have code examples of how to use these two methods in a handler?

I know documenting is not anyone's favorite activity and it take a huge amount of time to write, but it would be really helpful for newbies like myself if the ACTION, STAT and INFO library functions included code snipets to help new users implement the libraries in their projects.

Regards

If you use the hal_table widget, you must set some pins active, for instance:
 setp test2.button-box-active 1

 

You likely need to condition the carousel control pins with
motion.motion-enabled or similar. 

Posting your hal file or at least the carousel portion would
allow a more specific suggestion.
 

Update: limit switch working. As pointed out by scotta, it was defined as home switch not limit switch.

But still no sign of energy in steppers. Always joint following error.
Serial terminal don't show anything

Do you mean 7i77, not 7177?

I can't find any MESA 7177, only a MESA 7i77.

The capital i (I) looks too much like a 1.

Sorta like how a lowercase L (l) (ell) looks like a capiltal i (I), or a 1.

Typography is annoying sometimes.

It is important to mention that while the 7i84 can drive inductive loads, you still need flyback diodes to protect the driver from the reverse current spike when the solenoid switches off.

Thank you for the quick reply!
It's good that I can build a separate version without interfering with my base installation.

I'm having a hard time following the path of the position command through the code because of all the shared memory.
Could you give me advice on which module would be the best candidate to inject the extra pins into? Ideally it would be whatever the trajectory planner outputs its cartesian coordinates to. So instead of taking values from the trajectory planner it would take values from HAL pins.

When run with gladevcp, the widgets don't become active. How do I provide the signals to activate them?

I debounce the signals... with a schmitttrigger like from the original controller.

Still issue with state 9.

It works.
I use a second channel to re invert the signal.
Now no noise on the signal.

Start here:
linuxcnc.org/docs/html/code/building-linuxcnc.html

Hi, I'd like to control my robot arm directly using linuxcnc.
So far I've been using MDI and external offset pins to move the robot but I've had some issues with latency (mdi) and absolute positioning (external offsets). (When the robot starts, the motor's target position is not set to 0 but some other value close to 0 like 18 or -23 for some reason)

I'd like to pretty much hack the motion module so I can expose the pos-cmd pins that the trajectory planner sets so i can just connect to hal and set my cartesian position as I wish and not go in all these roundabout ways to get the robot to move.
The mini PC running the robot has a debian image with linuxcnc preinstalled.

1. What would be the easiest way to modify the source to achieve this?
2. How would I build the module and use it in my existing linuxcnc installation?

I'm not at all wary of digging into the C code, but I've never build linuxcnc myself.
Thanks for any tips