On Monday 13 September 2021 16:51:42 Alan Condit wrote: > John, Peter and Gene, > > I have the Cycletrol-150 Manual. It says that you can replace the 5k > ohm potentiometer with a 0-10v analog signal. The 7i76 manual says > that you can use it to replace a 5k potentiometer control (0-10v). So > that should work and indeed seems to be working. I have tried > adjusting the detectors to 90°. I have an old Tectronix scope and they > look pretty good there. It now has a solid “at speed indication” up to > about 250rpm. So now I am suspecting that I don’t have the PNCConf > configuration quite correct for the spindle. At the top of the range > the Speed Indicator fluctuates from about 400 to 600 rpm while the > motor sounds like it is holding dead steady. One of my problems is > that I don’t actually know what the rpm of the motor should be at > 90volts. It is a Pacific Scientific motor rated 1.5hp at 110v DC. > > Thanks, > Alan
If you are falling apart at 250 rpms, the first thing I would check is the speed of any opto stuff in the path from your encoder output back to LinuxCNC to the the encoder that shuld be labeled a decoder in the hardware at the computer end of the path. Get that old tech to look at the encoder inputs at the 5i25 or 7i76. I'd be willing to bet a small sum the good, square waves from your hardware are half amplitude saw teeth by the time they get to the mesa card(s) > > > I built a new controller for my 9x20 lathe using a 5i25/7i76 combo. > > I thought I had everything working but I still can’t get the spindle > > working correctly. I have a 1.5hp PMDC treadmill motor that I am > > controlling with a Cycletrol-150 DC controller. The am using the > > analog out from the 7i76 to replace the potentiometer to control the > > spindle speed. I have a 100 slot encoder with index (homemade) on > > the spindle. That would be 400 edges per turn. And since the motor doesn't seem to be affected, the feedback to the spindle's pid is also getting lost. > > > > The spindle starts under LCNC control but I am not getting an “at > > speed indication” at above about 100rpm. The motor sounds like is is > > running at a steady speed but the Spindle Speed indicator fluctuates > > wildly. I am suspicious of the spindle encoder. Before I was just > > running it as a 100 count counter with index and it worked fine. I > > can’t figure out in PNCConf how to set up a spindle counter with > > index for the 7i76. > > > >> From: "John Dammeyer" <jo...@autoartisans.com> > >> > >> Hi Alan, > >> > >> This is what I found for a Cycletrol DC motor controller. > >> http://www.grahammotorandcontrols.com/item_dc_motor_controller/176B > >>6002/58 > >> > >> > >> There is no guarantee that a PWM output will emulate a > >> potentiometer. That's the first place to look First put the pot > >> back and measure the voltage between the wiper and ground as you > >> control the speed from 0 to full RPM open loop. And depending on > >> how it's powered be careful that the pot inputs aren't indirectly > >> connected to ground. > >> > >> It's always possible the speed control won't work unless there is > >> also 5K Ohm between pin 9 and 11. > >> > >> It's also possible the pot changes the resistance (as a rheostat) > >> in a feedback circuit so input PWM may or may not work. > >> > >> Also PWM is exactly that pulse width. Unless you have a filter > >> capacitor across that with a bleeder resistor to ground (Pin 11) > >> the internal circuit many not integrate the PWM enough for it to > >> work. > >> > >> For example I bought a small Chinese PWM to 0-10V module that has > >> the rule that the PWM frequency can't be higher than 1kHz. Try it > >> with 10kHz and it won't create a lower voltage. So PWM frequency > >> is a parameter that must be correct. > >> > >> I also, way back, bought a step/dir to 0-10V controller that > >> produced 10V at 25kHz step rate. The problem was my ELS was > >> limited to 20kHz so I never could get 10V out of it. > >> > >> I don't believe your 100 PPR encoder is the issue. I'm running my > >> spindle with 60 PPR and can do power tapping on the mill. > >> > >> It may still be possible to find these UHU or HP_UHU kits > >> somewhere. I ended up using them because the GECKO servo drives I > >> bought were only good to 80V and my servo motors like more than > >> 90V. The HP_UHU kit was good to over 200VDC. > >> https://www.cnczone.com/forums/servo-motors-drives/342836-uhu-servo > >>-controller-build.html > >> > >> > >> Similarly there are still some STMBL kits out there. These can > >> actually take the smart serial input from the MESA boards or > >> step/dir. I'm not sure that 0-10V was very implemented. > >> https://www.pcbway.com/project/shareproject/stmbl_servo_motor_drive > >>.html > >> > >> John > >> > >> > >> > >> From: "Peter C. Wallace" <p...@mesanet.com> > >> > >> I would first check the A/B phasing and symmetry by monitoring the > >> A and B inputs with halscope when turning at a slow speed (say 100 > >> RPM or less) The pins to monitor would be: > >> > >> hm2_5i25.0.encoder.00.input-a > >> hm2_5i25.0.encoder.00.input-b > >> > >> Ideally both A and B should be square waves (50% duty cycle) > >> and have a 90 degree phase difference (A changes in the middle of B > >> high and B low times and vice versa) > >> > >> > >> Peter Wallace > >> Mesa Electronics > >> > >> > >> > >> From: Gene Heskett <ghesk...@shentel.net> > >> > >> On Sunday 12 September 2021 17:49:43 Alan Condit wrote: > >> > >> > >> getting proper quadrature out of homemade stuff is a bear with 4 > >> sore paws. I know, I done it about 3 times now. I would in your > >> shoes do this which has worked really well here. > >> > >> How many "gear shifts" do you have between the motor and the actual > >> spindlle? You could probably do more than 2, but thats all the > >> gears I have on my g0704. I have tally switches on that knob that > >> tell linuxcnc what, if any, gear it is in. I have a single screw > >> glued to the side of the drawbar cap that is read by an ATS-667 > >> hall effect device for the once per turn "index signal" to an > >> encoder. Prowl ebay and get a cheap 1000 line encoder for a $20 > >> bill, and spin it with the back of the motor shaft. It will be way > >> to fast to get thru the average BoB's opto stuff, so you'll have to > >> use one of the encoders in the 5i25, which can do that fast a > >> signal or hack the BoB to remove them. Feed the same encoder you > >> are feedingthe index into with the outputs of this encoder . > >> Not knowing the gear ratio in either gear, I hacked up a counter in > >> hal that starts counting pulses from an encoder on the back of the > >> spindle motor, counting them for 100 turns of the spindle as seen > >> by the ATS-667, and freezing the count so you can read it with a > >> hal-meter. Divide that count by 100 and you have the "SCALE" value > >> for that gear. Repeat for the other gear. then use the tally > >> switches and a mux to feed the correct SCALE into that input to > >> linuxcnc. For my g0704, that scale is a hair over 7,000 in high > >> gear, and a different hair over 14,000 in low gear. And with that > >> high a scale, quantization noise flat disappears, allowing very > >> high pid Pgain values, 20 or more for extremely stiff spindle speed > >> control. I use the PICO pwm-servo amp, and the only time I can tell > >> the spindle is loaded really heavily is when the iron in the motor > >> starts squeaking as that 9.7 amp motor is being throttled at around > >> 18 amps by the pwm-servo. Thats about 2 hp from a 1 hp rated > >> motor, but its had 6 years to complain and hasn't, not even asking > >> for fresh brushes. > >> > >> When you are ready, let me know and I'll pm you my hal files and > >> such so you can cherry pick and use what you want. > >> > >> > >> Take care now Alan. > >> > >> Cheers, Gene Heskett > >> -- > > _______________________________________________ > 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
