Hesitation between two consecutive G33 codes

The main part of the thread must have a taper of 3/4" per foot. I think that G76 will only taper at the end of a straight cut. I could be mistaken. When I first started looking at writing code to thread pipe, I wanted to use G76, since I already had written code to do threading. I read posts here and other places that said G76 could not be used.

You used to be able to create a continuous , variable pitch , variable diameter thread:

wiki.linuxcnc.org/cgi-bin/wiki.pl?JMKsFusee

Seems a setting is wrong or a possible bug (which would be easy to not exercise).
What version of linuxcnc are you using?

Chris M

I'm using 2.7.8

I wonder if this is a feature of the new trajectory planner?

Try ARC_BLEND_ENABLE=0

linuxcnc.org/docs/2.7/html/config/ini-config.html#_traj_section

I added to my .INI file:

I threaded a pipe at 300 RPM. No hesitation at all between G33 codes! It's working great.

Is this likely to affect anything else I'm doing on the lathe? Or is it mostly for milling?

I'll post the code to " O Codes (subroutines) and NGCGUI - Lathe Library" later today.

It changes the Trajectory planner back to what it was before 2.7, with one line look ahead and parabolic blending. It can cause a series of short moves to be slower than they have to be, by limiting the maximum feed to what what the acceleration settings can bring to a controlled stop by the end of the next line. It is less likely to be a limiting factor on a lathe than a router.

I've posted the pipe threading routine at:

It's not the same issue as this forum.linuxcnc.org/20-g-code/29685-g76-p...aper-linux-cnc-2-7-0 is it? I'm not sure if this has been corrected yet in 2.7.x

If you are set up to build from source, try this branch here:

github.com/robEllenberg/linuxcnc-mirror/...-overhaul-2.7-rebase

Among other things, that branch fixes some bugs I found in the blend calculations for spindle-sync motions.