David Szakovits wrote: > hi, > i use gecko drives with 10 "micro steps" and i don't know if i ever lost > a step or not. i would like to add encoders to the back of my double > shafted stepper motors. i would like emc to stop all motors and report > any lost steps. i definitely don't want emc to try to make up for the > detected mistake. just want it to stop. Why? If the mistake is due to a stalled motor, then you DO need to stop, this is hard to fix while still moving. If the error is due to electrical noise or mixing up the change in direction and step pulses, then the error is only a microstep or two, and can easily be corrected. Since the motors can be up to 20 microsteps off commanded position without permanent loss of alignment, you can't set the limits too tight or you will get nuisance trips constantly. > > these are my questions: > is this practical? > does emc support this? > what should i look for when purchasing encoders for this application? > how many positions/pulses per revolution will work best? Something like 2000 counts/rev (500 lines) sounds about right, to match the motor's microstep resolution. Maybe Mariss at Gecko can shed some light on this.
My Universal Stepper Controller handles all the I/O for this. The USC puts out step pulses to the drives, and can read signals from the encoders. EMC2 has provisions to set minimum and sliding-scale following error limits, so more position error can be tolerated during faster moves. Don't think that an exact 1:1 correspondence between motor position and step pulses can be maintained at all times, what you need to trap on is accumulating errors. Due to manufacturing tolerances in the motors, the encoders and the logic in the Gecko microstepping drive, plus friction and inertia, there are always small discrepancies between commanded and actual position. These may actually be quite small compared to leadscrew thread errors, backlash, slop in the ways, etc. in the machine tool. So, quibbling over one microstep is not an issue. An accumulation of many missed microsteps is what you want to avoid, as well as outright stalls. Of course, you will KNOW if this ever happens! So, yes, EMC definitely handles it, it is done on a number of systems, and it does work. Whether it will actually benefit you in any way is not clear. If you have never seen any positioning error, why are you worrying? For more info on my USC board, see http://jelinux.pico-systems.com/univstep.html Jon ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
