Lathe Taper OD

Hey John,

Nice work, thanks for sharing, I will give it a try when I get to work.

Rick

I've recently got NGCGUI going on my lathe, using John's subs as a starting point (thanks John).

The behaviour of the original taper OD sub was not what I wanted for 2 reasons:
1) Cutter rad comp is a step beyond me at the moment
2) My machine has backlash, so needs some run in and out of the taper cut

I set something up to behave as I wanted (see attached screenshot). You specify a "run-in/out" value to start the cut beyond the end of the workpiece (Z), and the cut continues until clearing the diameter by the same run in/out value.

The problem I have is the angle. This screenshot was programmed as a 45 degree chamfer, and you can see that's not what is output (closer to 35 degrees). The rest of the program behaves as expected. I suspect the problem is something to do with the TAN function (and my maths). Before I post up the code (my first foray into subroutine programming), is there anything special I should know about the trig functions in LinuxCNC that could give this type of error? Degrees/Radians, etc?

I'm planning to add cutter comp back in when I understand what is happening. The angle shouldn't change.

I've attached the GCode, with all my maths. The basic concept is:

Start of cut: Z is always <face-z> + <runinout>
End of cut: X is always <diameter> + <runinout>

The tricky part (for me at least - I've spent hours debugging this) is calculating X for the start of the cut, and Z for the end of the cut.
I'll upload a sketch in a minute to show how I arrived at the trig...

Tom

Hi John,

Gotcha about the cutter compensation, thanks for that. I'm going to add it back in when I need accurate parts, and when I'm confident I know that everything else is behaving normally.

My problem is the angle is not as expected. Programmed 45 degrees, but actually less.
I've attached a sheet of workings, and looking at them it's perhaps not a surprise it's not working.
Polar co-ords is the way to do this, but they don't appear to be available outside of the XY plane.

I'm going to sleep on this a few nights and come back to it. My head hurts!

Hi John,

I had one last look at this tonight.
My code is a little more involved because of the "run in/out" move, but not much.
In fact, I made a spreadsheet as a sanity-check, which shows that my code (corrected from the version I posted earlier) should give the correct taper angle. It's quite confusing.

I've got to do my working week now, but when I'm back I will try pre-calculating the co-ordinates into variables, and then passing the results to the G00 and G01 lines.

Tom