I've had good luck with the CUI encoders (Compared to US Digital) on DC Servos and the HP_UHU drives. I found I lost position over time with either missing or extra pulses. I was using differential signalling. The CUI dispensed with that problem.
I was never able to get satisfactory motion from the STMBL driving the same DC Servo with the US Digital. I didn't go back to and try the STMBL with the CUI encoders. John > -----Original Message----- > From: Sam Sokolik [mailto:samco...@gmail.com] > Sent: May-05-20 7:41 PM > To: Enhanced Machine Controller (EMC) > Subject: Re: [Emc-users] Fluctuating RPM using CUI ATM 10 encoder > > I don't know if this is related - but I have had bad luck with > capacitive encoders like the CUI ATM 10 > in it's spec sheet it is only accurate to 0.25 degrees.. when I tried to > use it for a closed loop servo - I got a lot of noise.. > > https://www.youtube.com/watch?v=gHbemDWXudQ > > https://www.youtube.com/watch?v=i6rJsCKzUK0 > > On Tue, May 5, 2020 at 9:13 PM Gene Heskett <ghesk...@shentel.net> wrote: > > > On Tuesday 05 May 2020 21:24:50 Dan Henderson wrote: > > > > > I just reverted to a LinuxCNC configuration that isn�t setup to use > > > the encoder. The spindle speed is much more consistent without the > > > feedback from the encoder?? > > > > > And I assume that its steady if the pid is bypassed. That does point a > > finger at the encoder. Can you post a link to it as I'm not familiar > > with that one. > > > > > On Tue, May 5, 2020 at 8:19 PM Dan Henderson <luvtof...@gmail.com> > > wrote: > > > > I have LinuxCNC fully operational. In halscope I�m able to view the > > > > signals for both the spindle index, phase a, and I now have phase b > > > > hooked up. I�m conducting my testing with the LinuxCNC MDI. The > > > > number on the opto chips is EL817C807. > > > > The BOB PN is DB25-1205. > > > > useing the halscope, look at the a and b signals, you should see each > > with a 50% duty, and a time overlap in each state of 50%. + or - not > > more than 5% maximum, when its running steady. But you'll have to stop > > the hunting before you can read that very accurately. You won't see > > amplitude variations there unless something is poorly connected, only > > time variations. > > > > > > On Tue, May 5, 2020 at 8:08 PM Gene Heskett <ghesk...@shentel.net> > > wrote: > > > >> On Tuesday 05 May 2020 20:17:45 Dan Henderson wrote: > > > >> > Slowing things down had marginal effect. Okay sounds like I need > > > >> > to by pass the optocouplers. Would you recommend I bypass all the > > > >> > encoder inputs as well as the PWM output to the MC2100 > > > >> > controller? > > > >> > > > >> Only if it has opto's for isolation, they can effect to duration of > > > >> the pwm playing hob with the speed linearity. If the motor is > > > >> "hunting", something else is going on. The m2100 ISTR is the > > > >> controller that came in the treadmill and its a very serious case > > > >> of built by lowest bidder, I was not able to use it in anything > > > >> I've tried. Control was extremely non-linear until I put a 47k > > > >> resistor between linuxcnc and its summing point, That helped but > > > >> it was still flaky. Not to mention that all of its control > > > >> circuitry is on the hot side of the power line and quite dangerous. > > > >> > > > >> I am not a machinist, but a long retired broadcast engineer and a > > > >> CET so much of the high power stuff comes out of my junk box. The > > > >> psu for instance, makes about 120 volts DC at a surge current of > > > >> 20 amps, 10 forever, way more that the 1hp motor I'd put on the > > > >> back of a 7x12 lathe is rated for. Capable of getting 2hp out of > > > >> that 90 volt motor. But it kept breaking drive parts once I had > > > >> obtained one of the pico systems pwm-servo drivers. So I've gone > > > >> into the hal file and slowed the motors acceleration to something > > > >> that doesn't break stuff any more. But that might explain why my > > > >> troubleshooting is geared to the electronics, and possibly is > > > >> different than what a real machinist might do to solve an identical > > > >> problem. > > > >> > > > >> Other than that it doesn't sound as if the opto's speed is your > > > >> problem. > > > >> > > > >> So 2 other questions: > > > >> > > > >> What to you have in the way of meters and oscilloscopes for > > > >> troubleshooting? > > > >> > > > >> Do you actually have LinuxCNC running well enough to try to control > > > >> your motor? > > > >> > > > >> 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 > > > > > > _______________________________________________ > > > 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) > > 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 > > > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
