On Friday 27 July 2018 21:15:03 Kurt Jacobson wrote: > On Fri, Jul 27, 2018 at 8:46 PM, andy pugh <[email protected]> wrote: > > Good point. How about... > > > > G21 G91 > > G38.2 z-20 F80 > > G0 X#5061 Y#5062 Z#5063 > > G10 L20 P1 Z0 > > That would probably crash the probe, but a variant of that would work. > > G21 G91 > G38.2 z-20 F80 > G0 Z[#5063 + .100] (move up a know amount from where the probe > tripped) G10 L20 P1 Z.100 (set out current location to the distance > we moved) > > But that gets a little dangerous, because the signs have to be right, > and if you are probing inside a hole or slot, then what do you do .... > I probe all 4 directions to get the center of the hole. Has worked well for me for almost a decade, accuracy well under a thou.
> > FWIW I rather agree that probing in absolute coords would seem > > useful G38,6, 38.7, 38.8 and 38.9 are free > > So, one person on my side ;) > It seems like it might be clearer to be able to do G53 G38.x, but > adding more gcodes would work too! > > This it a little off topic, but currently any probe trip during a non > probe moves causes an estop (good!). But, some times when backing off > after a probe with G0 or G1 the probe bounces and causes the machine > to estop, even though the probe is perfectly safe. So I have though > about adding a probe backoff gcode, that is basically a G1 (or G0 > maybe) move that ignores probe input. Tormach uses a G38.6 for this > ... > > So, I pondered a little more while cutting the grass today, what about > having a probe code that probes the part, and then automatically > retracts to the retract plain R? Who has ever probed with out > retracting, right? SO might as well have this build in. It would also > allow for ignoring probe trip on the retract move, which would be a > really neat solution IMO. > Agreed. I record the co-ord at the 2nd slow moving contact, doing a 15 thou retract with a g0 at first contact, then repeat for the real measurement at F.1, measure that and record it, back away with a g0 about 15 thou to make sure the contact is broken, then start looking for the other side of the hole, calculate the center of that and goto it to look for the readings at 90 degrees to the first set, but I would never consider doing it at F80, in fact I generally do it at F5, retract 15 thou, and then do the final measured probe at F.1 or .2. You can't expect real accuracy if moving at an F80 on an inch machine. That makes the granularity of the measurement pretty bad. F80/servo-thread is .080". Are you happy with that much of a +- to your measurements in inches? Maybe close enough in mm's, but surely not in inches. And my probe contact is a wire with a lead wrapped around it to the probe input, mounted in the center of a 3/4 teflon rod, and the spindle is running at about 1.5k revs, so the wobbling tip of the probe is essentially a perfect circle, centered on the spindles axis. I have a .1uf cap across the probe input so the first contact discharges it to a logic 0, with the cap being slow enough to recharge so as to still be a logic 0 when the next board servo-thread read takes place even if probe contact has been lost at that spindle position. My repeatability is typically .0004" thou. Generally thats "close enough for the girls I go with"... > > What happened to G38.1? > > Good question, if I try it I get "Unknown g code used", wonder why it > was skipped, or maybe it was superseded?? > > I am going to see if I can find where the probe stuff is, and try and > wrap my head around it .. -- 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) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
