My intent is to use the R only to set origin(s). After that the most useful bit of information would be the difference between the stepper-servo position and the position being reported by the glass scale. The stepper-servos have about the same torque as the SEM servo motors I was using so we’ll see. The parts I make are relatively small so G0 speeds are not so critical. I may have to swap out the ball screw on the Y, hope not. On going project keeps me off the streets and out of the bars. ;-)
> On Feb 23, 2025, at 3:19 PM, gene heskett <ghesk...@shentel.net> wrote: > > > On 2/23/25 10:39, Dave Engvall wrote: >> This is all Gene’s fault. :-) His evangelism of steppers-servos has >> penetrated my thick skull. Side note: tinyG has signal pickoff before the >> final stage so it is possible to drive stepper servos off a tinyG.I’ve not >> gotten there yet but headed that way. Of course “the difference between >> theory and practice….” Be patient nothing I do these days is fast or >> efficient. Stay tuned! >> >> D >> >>> On Feb 22, 2025, at 5:41 PM, gene heskett <ghesk...@shentel.net> wrote: >>> >>> On 2/22/25 18:29, Dave Engvall wrote: >>>> There is a method to my madness, to wit I really want to know how good, or >>>> bad the tinyG is. Older version of the emc interp but a much faster cycle >>>> time. Glass scales are problematic for velocity control but rather good >>>> for position. Servos on that machine with good encoders on the ball screw >>>> were much better at velocity control, roll that in with a glass scale for >>>> position and you have the best of both worlds. >>> I've had very good results on the one machine I have converted to closed >>> loop stepper/servo's. Totally doing away with the PID's. Feedback is from >>> the TP that drives that axis, the stepper/servo's have their own PID's. My >>> sheldon rapids are 2x faster, and quite a bit more accurate. The TP outputs >>> good code. These motors stay w/in a count of error And it makes beautiful >>> metal. Well within the thousandths I wrote the code to do. No ticklish PID >>> tuning ever, no PID's to function as rubber bands, so what the TP spits out >>> is exactly what the motor does. > > To explain that further, there are /PID's in the controller and an encoder in > the motor, but that encoder feedback is not given to linuxcnc. That feedback > to the controller serves to allow the controller to do a couple things. The > controller then knows where the motor is at a considerably higher bandwidth > than linuxcnc's 1 or 2 kilohertz servo thread. If the motor is doing this w/o > any great error, the motors excitation current is minimal so there is no > great heat. If the motor is behind, the controller turns up the current until > the motor is where its supposed to be. By the same mechanism if the motor has > overshot, the controller will turn up the current AND step the motor > backwards until the motor is in position. Applied to a 3d printer, the motor > can be 5 to 10x faster w/o a layer shift, which occurs if the motor has lost > steps and is no longer HOMED. This relative lack of burn your hand heating, > and the active motor control averages quite a bit less wasted power you can > see in how fast the power meter spins and because the controller can use the > full power supply output for maybe 20 microseconds, makes the machine 5 to 10 > times faster. Hanpose LC42 controllers can tolerate 90 or more psu volts to > make a nema-17 motor do exactly what the TP told it to do. The LC57 > controller for Nema-23's is rated for about 110 volts. I am using 72 volt > psu's on one converted printer, not the usual 24 volts. > / > > /I hope this is a better explanation. And helps to explain my enthusiasm./ > > On Feb 20, 2025, at 1:41 AM, Todd Zuercher <tzuercher1...@gmail.com> wrote: > >>>>> If they are equidistant wouldn't simply use them the same as an ordinary >>>>> encoder index signed? >>>>> >>>>> On Thu, Feb 20, 2025, 4:09 AM Andy Pugh <bodge...@gmail.com> wrote: >>>>> >>>>>>> On 19 Feb 2025, at 17:15, Dave Engvall <dengv...@charter.net> wrote: >>>>>>> >>>>>>> I think both have R marks at 50 mm intervals. >>>>>>> Seems logical? >>>>>> There are systems where the index marks are at varying distances from >>>>>> each >>>>>> other such that you know exactly where you are after seeing two marks. >>>>>> There was discussion of this on the forum, which resulted in the homing >>>>>> process being modularised, (HOMEMOD) >>>>>> I don’t recall if we saw a suitable HOMEMOD module for this style of >>>>>> scale >>>>>> in that thread. >>>>>> >>>>>> Emc-users mailing list >>>>>> Emc-users@lists.sourceforge.net >>>>>> https://lists.sourceforge.net/lists/listinfo/emc-users > > Cheers, Gene Heskett, CET. > -- > "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 > > > > _______________________________________________ > 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
