On Tuesday 19 January 2021 12:48:32 Jon Elson wrote: > On 01/18/2021 05:47 PM, John Dammeyer wrote: > >>> So what does LinuxCNC do? Is the thread mucked up if spindle > >>> speed is changed during a feed hold and then start? > >> > >> Feed hold has nothing to do with it John, you can't change the > >> spindle speed in mid thread. Full stop. Period. I think it could do > >> what is needed. But its like Yogi Berra once said about theory and > >> practice. Which in this case means machine balistics are hard to > >> do. > > > > Gene, > > Are you saying that for threading on a lathe LinuxCNC _must_ have > > control over the spindle speed? Or can I have a manual knob and > > adjust the speed as it's threading. Or as it's returning back to > > the start? > > No, not at all. On desktop machines, the spindle always > slows down when the cutting starts. > LinuxCNC adapts the Z feed to the speed of the spindle, as > sensed by the encoder. You can literally shut the motor off > and turn it by hand, even back up and then move forward again. > > > Or are you saying that if LinuxCNC is cutting a thread keep your > > fingers off the speed control? > > When in a threading operation, at least most manual controls > and overrides are disabled, other than E-stop. Feedrate > override is certainly disabled, and abort (esc key) and > pause are disabled until the cutter is out of the > workpiece. I have no idea if spindle override is disabled, > I have not tried it. > Seems like maybe that would be OK. > > Jon > It wasn't the last time I tested it Jon, and most of the recent conversations here are ignoreing the fact that every new stroke, starting from its resting position, regardless of whether its a G76 or you are "pecking" a G33.1 for rigid tapping, is a NEW sync action, and its here that diddling the the spindle speed affects both the accel time to get Z up to sync speed, so in addition to the spindle turning further at a higher speed, it also takes Z longer to get up to that higher speed and become phase locked. This combined change in the delay from 0 to locked is changing the phasing between the spindle and the tool at about 2x the change in spindle speed. And screwing up your thread.
The ONLY way I can see to get away from that would be to set the allowable error in the PID way high at the index pulse that starts the stroke and let the PID establish a 1 to 1 relationship at the index pulse, letting the PID do all the work of jerking Z up to speed. But that means everybody has to use a PID to drive Z, configureing the allowable error back to its default at say 10 turns after the index pulse that triggered the current stroke and train themselves to allow more space between the resting point, and the tool entering the work as the PID, dealing with that large an initial error, may well ring like the cracked liberty bell for several turns of the spindle before its tracking good enough to cut a useable thread. If, 10 turns into the stroke it triggers a following error as the default is restored, you really need to tune the PID better. Or slow the spindle to give it time to stabilize. This might even be done better if each stroke starts from a dead spindle, because the PID would spared much of the huge initial error while accellerating both at the same time. Diddling the PID'a error inputs is something hal can do well. You just have to look at the problem, and sometimes think a bit outside the box to fix it. That was the major reason I changed out that 1600oz/in on the G0704 driving Z, 27 IPM was simply too slow for any reasonable threading speed as it couldn't get up to speed to cut a 11 tpi USS thread. A shorter, much lower inductance 940oz/in motor with an AC powered driver got it up to over 100 IPM, and all my following errors went away. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis Genes Web page <http://geneslinuxbox.net:6309/gene> _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users