On Sunday 13 November 2016 12:43:26 Jon Elson wrote: > On 11/13/2016 03:34 AM, Gene Heskett wrote: > > On Sunday 13 November 2016 00:21:24 Jon Elson wrote: > >> On 11/12/2016 08:29 PM, Gene Heskett wrote: > >>> Back again; > >>> > >>> I stuck the 2nd one back in, right way to, but I guess it did not > >>> survive the reversed polarity, it can only pull down about half a > >>> milliamp. I have 7 more of them so I will use new ones for the > >>> last 2, A/B. > >>> > >>> But watching the one in the index position, I note that they stay > >>> in the last state, but does that affect the index adversely? > >> > >> Yes, they have a flip-flop that is driven by a comparator > >> that checks the difference between the two sensors > >> in each unit. So, they DO latch, that is the behavior as > >> per the data sheet. > >> > >> Now, with LinuxCNC, all you need is consistent operation in > >> ONE direction only. For instance, in a rigid tapping > >> operation, the spindle encoder is synched to the index at > >> the beginning of the operation and then the encoder is > >> allowed to count up and down as the tap progresses in and > >> then backs out. The spindle encoder is NOT resynched during > >> the whole operation. > >> > >> It is pretty clear that the index pulse would pick up the > >> opposite side of the hole or notch when running the opposite > >> direction, but that should not be a problem as mentioned above. > >> > >> It also seems that the A-B quadrature might slip one > >> quadrature count when the spindle reverses, but would then > >> slip the other way when reversing again. And, I'm not sure > >> that would actually occur, either. Anyway, a tiny slip in > >> the spindle count should not be much of a problem in most cases. > >> > >> I've rigid tapped thousands of 4-40 holes with combo > >> drill-taps without any problem, using the ATS667 sensors. > >> > >> Jon > > > > Not knowing the internal workings as you do, I was somewhat > > concerned that there might be enough of a phase angle difference > > between directions to loosen a thread, particularly when running the > > g33.1 in a peck loop, > > I don't know the peck cycle. Does it re-synch the spindle > each peck, or just stay synched for the entire cycle? I'd > kind of guess it just stays synched for the entire cycle, in > which case it should not matter.
No, it resyncs at the top of each peck. Its just a gcode loop that passes an ever deeping z to the g33.1 command until it has reached the target depth. Not a gcode command in LCNC, but a short paragraph from my own fingers. > > > I think some of my > > loose threads are because the tap isn't straight. > > Well, there is also lag in the trajectory planner (in old > versions you could run the TP at 1/5th the servo rate) and > backlash and other slack in the machine. > > > And I keep forgetting > > to order a couple 3mm from some other than dewalt/irwin because both > > of the $30 ea tap & matching drill kits I've bought have crooked > > taps, by about a 20 to 30 thou wobble. > > Ugh! You are being polite Jon... > I buy my taps from MSC or KBC, usually. I use > drill-taps for thin stock and do the spot-drill-tap in one > cycle. I use spiral-flute taps for other stuff in aluminum. > > > Z backlash setting accuracy should count > > for that too. > > Oh, i do NOT use backlash compensation. It makes a jerk > move on every reversal, and that can cause problems. I would think that not using it when its say 5 thou, would be a bigger problem. Currently using .0016" in the Z axis. > > Jon > > ---------------------------------------------------------------------- >-------- Developer Access Program for Intel Xeon Phi Processors > Access to Intel Xeon Phi processor-based developer platforms. > With one year of Intel Parallel Studio XE. > Training and support from Colfax. > Order your platform today. http://sdm.link/xeonphi > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users 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> ------------------------------------------------------------------------------ Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
