On Wednesday 21 July 2021 17:19:05 Chris Albertson wrote: > If you feed those averages to the PID controller, that PID is reacting > to old data and you get either very sluggish control or oscillation. > > The best way is to predict the future and the have the control > programmed to make that predicted future the future you want. That > hockey quote comes to mind, "I don't skate to the puck, I skate to > where the puck will be." PID algorithms using averaged data are even > worse and "skate to where the puck used to be." > > A PLL is actually a predictive model of the spindle that tells you > where the index will be at any point in time, past or future. It > works in as much as spindles don't teleport randomly to different > angles and RPMs but exhibit predictable behavior. For very short > periods of time, (one servo cycle) the prediction is very good. > This spindle could, it is being controlled by one of jon's PWM-Servo amps, which is a full 4 quadrant controller. A 1hp rated motor at 9.7 amps and 90 volts, but the power supply is about 127 volts with enough microfarads that a cold turnon trips a 30 amp breaker instantly.
And my home made optical disk encoder had so much quantization noise I actually caught it stopping at 50 rpms occasionally, and the nylon gears were being slapped around like all the bearings had square balls in them. Pgain was limited to the 2.5-3 range. I installed a 1024 ppr encoder on the back of the motor and wrote some hal to calibrate it again. Pgain can now go as high as 40 but its plenty stiff enough at 20, and below 500 revs its so quiet I have to look at it to see if its turning. That PSU gets soft started by having a 50 ohm, 200 watt resistor in series with it for the first 4 seconds, and now it lives on a 20 amp breaker even with the iron chirping because the servo limit is about 17.5 amps. I thought that would be hard on brushes, but its been running that way for 5 years now. On the same brushes. G0704's are rated for 2250 fwd, 1100 in reverse when grizzy ships, but with that PSU, it does 3 grand in both directions. Whats not to like? :) I had to run the change direction thru some hal to slow it, otherwise the Z can't keep up and errors off, but it still does that bottom of the stroke reverse in a g33.1 in 400 milliseconds at 1500 res. Using that same idea in the sheldon, with has a vfd, reverse overshoot at 100 revs is done in .25 of a turn with an 8" chuck north of 30 lbs mounted. 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