On Thursday 01 November 2018 04:09:36 Chris Albertson wrote: > You don't want to block the hole in the spindle so you are looking for > an encoder with a large hole though it? Is that right? > > If the spindle is belt or gear driven you could place the sensor on > the gear or pulley that drives the spindle. You could also place a > timing belt puley on the spindle then drive the encoder with a 1:1 > belt drive. 1:1 drive would allow using a 3 channel encoder > > > > You could use a 1000 line optical encoder and > > On Wed, Oct 31, 2018 at 8:32 PM David Berndt <ber...@uberwin.com> wrote: > > Thanks Andy, > > > > That doesn't sound particularly hopeful for me and my plans. I guess > > maybe > > a more traditional/proven approach of a through-beam sensor with > > interrupting disk and much much lower resolution encoder ring with > > drilled/milled holes might be the way to go. > > > > But to make sure this AEDR thing is really dead. Any reason to > > believe the > > AEDR-831X single channel units, which are the only way the 36tpi > > units seem to be available, have the same actual reoslution > > restriction? If there is only one channel then are they really > > looking for a fixed line width? There would be no offset second > > sensor at a fixed pitch to require that. I have the space to mount > > two single channel units, build in a little adjustability in their > > mounting to get them "clocked" correctly to get a quadrature signal. > > > > Anyone have any idea what an AEDR-8311 is? The data sheets I can > > find only > > detail 8xx0, no idea what the last 1 is for. > > > > Any other options I should be looking at? I didn't see anything > > super compelling when I spent 3 hours of my life reading through the > > datasheets from the digikey optical sensors - reflective section. > > I'm working with a > > ~2" ID mounting for my encoder ring, so options for adapting an off > > the shelf encoder ring seem low as not many are that large. > > > > Dave > > > > On Wed, 31 Oct 2018 16:48:46 -0400, andy pugh <bodge...@gmail.com> wrote: > > > On Wed, 31 Oct 2018 at 19:00, David Berndt <ber...@uberwin.com> wrote: > > >> if so was it with one of the one channel output units > > >> instead of 2 channels or how did you handle creating such a fine > > >> pattern? > > >> Any other details about how it was done and how well or not well > > >> it works > > > > > > I used a 2 channel. I ended up having a target laser-cut by a > > > company who specialise in PCB stencils. > > > > > > This might currently be pretty inexpensive from China, for > > > example: https://www.seeedstudio.com/stencil.html > > > > > > It never worked all that well, though. I rather struggled with > > > soldering the devices, as at the time I had no hot-air or reflow > > > oven.
When I changed the encoder on the mill, I made a small brass shaft extension and mounted it to the rear of the motor shaft. But I did not use the z from the new encoder, staying with the z on the encoder I had made for the mills spindle, so I still had a zero reference from the spindle. The motor shaft turns much faster of course, so I wrote some hal to measure how many A pulses there were in 100 revs of the spindle in order to obtain the SCALE value for each gear since the head is a two speed head. In high gear it was a bit over 7 thousand. In low gear it was a bit over 14 thousand but was not exactly a 2/1 ratio, so I determined the ratio and used the single SCALE, switching its multiplier by hand with a pyvcp button. But I have since made a switch roller notch in the rim of the gear shift knob for a half moon shaped pair of 16ga alu pieces with two switches sandwitched between them and mounted on the side of the head so the notch, when the switch arm settles into it activates the switches one at a time. Then it occurred to me that a mux4 could be used to feed a very slow speed command to the motor if neither switch was closed but it was running. This creep speed allows the gears to mesh well when being shifted. So I can have the spindle turning 1500 rpms, wide open in low gear, and the motor is down to 20 revs after the knob has begun to turn, the teeth of the other gear mesh silently and without any shock load on the plastic teeth, all w/o grabbing the spindle and searching for tooth mesh by hand, it just works, and when the other switch closes indicating the gear shift is complete, the spindle returns to 1500 rpms in a few milliseconds. Spindle reversals from 3000 fwd to 3000 in rev are about 400 ms. And thats profiled by a limit3 to keep the pwm-servo from chirping. It still chirps if the spindle is loaded with lots of swinging weight. The huge differential between my original home made 268 edge, 67 slot optical encoder and the very high effective count of the Omron encoder since its a 1000 line being driven by the motor, has allowed the motor PID's Pgain to be upped from around 2.5 to something in the 20 to 40 range, and spindle speed control is so stiff that I am not aware the motor is working way harder than its rated 1HP, probably around 2HP by the time the pico pwm-servo starts chirping as it regulates the maximum current at about 17 amps. Motor nameplate says 9. At 90 volts, but my motor psu is outputting about 126 volts. Those "overloads" have yet to noticeably heat the motor because they are relative short duration events, like rigid tapping a 10mm or 3/8" hole. In steel. Best modification to a G0704 since I put the cnc kit in it. A lot of work to do it, but worth every minute of the time spent to do it. Sweet even. :) > _______________________________________________ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > 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> _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
