Advanced Search

I was upbraided earlier by a friend from the Motor Club about this, as he watched your video and thought that I should have defaulted to a 55 degree angle as anything other than Whitworth is an abomination.

Adding the compound angle wouldn't be too hard, but I doubt that cutting buttress threads is that common, and you can always MDI the G76 directly if you need to.
I think that 29.5 is "not terrible" for Whitworth, but I would need to sketch a diagram to be sure. It's probably not ideal for BA threads (47.5 degree) or Panzergewinde (80 degree)

I also suspect that the macro takes many more passes than strictly necessary,

As a first step, get the encoder counting so that you can see a "counts" pin incrementing when you turn the encoder.
There are a few ways to do this mentioned above, and there is always the possibility of using the software encoder counter too, connected to GPIO pins,

Once you have the counts HAL pin working, that then needs connecting to all the axis (and joint) jog-counts pins. You then wire the switches to set the jog-enable pins of the axes/joints as required,

For the actual pin names see:
linuxcnc.org/docs/stable/html/man/man9/m...n.9.html#AXIS%20PINS

linuxcnc.org/docs/stable/html/man/man9/m....9.html#JOINT%20PINS

And possibly linuxcnc.org/docs/stable/html/man/man1/h...html#Joint%20jogging

But prefer the pins in "motion" to those in "halui" as the former run in realtime and so are more dependable.

I think that this is a consequence of the system trying to get into the same state after a run-from-line request.
But it does rather assume a toolchanger and/or repeatable length tools.

Rather than re-post the toolpath, you should be able to to comment-out the preceding tool changes in the G-code file. (Just don't forget to un-comment them, or save the modified file under a new name)

Which LinuxCNC version are you running? I think that there have been various attempts to improve run-from-line over the years.

There is discussion, but no conclusion, here: github.com/LinuxCNC/linuxcnc/issues/246#issuecomment-423049528