Erik Christiansen wrote: > I'm thinking of building a chunky machine for wood & aluminium, so the > datasheet accuracy of quarter of a degree is OK, even though that 1/1440 > of a rev limit means that 2048 ppr is humbug, AFAICT. > > What causes wrinkles on my brain is the "time constant" of 0.4 ms. > If it takes 40% of an EMC2 sampling loop for current motion pulses to > work their way through the encoder electronics, then that'll look like > backlash when we change direction? > > What does this time constant actually mean? Is is a response to changes in velocity? That seems to be the only thing I can figure out it means. If EMC is sampling position at 1 ms intervals, then 400 us is 2.5 time constants, so the velocity error should be pretty small. > And when we're just accelerating, it's fibbing to the PID loop, but no > more than if we had a slower sampling loop in EMC2? > > It would be nice to be able to fit encoders to 3 axes on a woodworking machine > for the price of one axis with US Digital encoders, but what's the > stability like? > > I have put these encoders on the Keling motor, and gotten the servo loop to work, using my own servo amps. I DID notice the stability margin is less than with some other motor/encoder combinations. The Keling motors have VERY light rotors, and so I added just a little mass to the motor by mounting a small shaft extension, and it helped quite a bit. I would assume this is a lot less angular momentum than even a small leadscrew.
I also had a customer who was experimenting with a very high acceleration machine, and I believe he will eventually abandon the CUI encoders. He was accelerating at something like 10000 rad/sec^2, which is really high acceleration, though. Currently, he has cut the acceleration rates to prevent missing counts. Jon > Has anyone had problems with the index pulse on servomotors? > (Datasheet note 1 warns about steppers) > > Erik > (Hoping to learn from the brave souls who've taken the plunge.) > > ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
