On Friday, December 17, 2021 7:11:42 PM EST John Dammeyer wrote: > And now with the 40T and 41T labels on each gear it was possible to spin > them on the mount so they were all in the same orientation and I then > constrained one mount hole on each. Then repositioned the teeth and the > two rings. Now they all match back and front. > > Have not yet figured out how to set up constraints so the gears turn with > each other. Of course the 3D prints take care of that exercise. > > Now as an aside would it be more prudent to work an encoder into the green > output gear for feed back into LCNC or for that matter the motor driver? > (With the appropriate division ratio). > > John > Great progress John.
As for the encoder, it depends on if using a servo or a stepper. In either case a quadrature scheme gets you an errorless bit of data, the difference is in where you put the encoder. If using one of the new 3 phase motors, driving it with a stepper, you still have two choices. I've not used #1, but have used #2 twice. 1. Breaking into the cable and borrowing the motor encoders output while still feeding that to the driver, but also feeding an encoder in linuxcnc so you know to the count exactly where the motor is. 2. Let the driver do its thing and use the stepgen fb signal as feedback. Just as if you were driving an old 2 phase motor. That 3 phase driver is pretty darned smart. Scale calibration in any event should involve a switch to set home position so linuxcnc knows where it is. I glue something on the back of a chuck to trip a microswitch as it goes by. Since it will normally be used in the angular mode, I have some hal code that gives a scale calibration accurate to 2 decimal points to the right of an angle in degrees, like g1 f30 A195.00, and it should settle at 195.00 degrees. Its normally commented out of the hal once its been used and I could cull it out to quote it if I have to. Just holler. What it does is fire it up to run at about half speed, let it go a full turn, freeze the encoder count ( use a couple mux2's as sample holds) on the 2nd full cycle of the home switch, incrementing a counter every time the home switch goes true. leaving it run until the counter hits 102 at which point it freezes the count in a 2nd mux2. Then it stops the motor, subtracts the first encoder count from the second one and displays that count with a hal meter. Shift the decimal point two places left and put that value into the A_SCALE in the .ini file. Thats the value for a full turn of 360 degrees, accurate to 2 decimal places right of the point. If that's not accurate enough, setp the counter comp to 1000 turns for 3 places. As for printing the spider disks, you might leave a rectangular hole to let you see the zero marks printed into the top of the pinions and I'd print then 2 high, switching to 41 teeth for the 2nd half of the print. Piece of cake in openscad. You could also put a half a mm spacer between them, and/or taper the ends of the teeth where they meet to prevent miss-cues if the rings move vertically. But I'd not leave them much room in the housing to do that. With lube, .2mm is plenty of room. Fedex must be busy, no parts for my Prusa yet. So today I've not done that much. Take care and stay well all. 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, 1940) 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
