> > I have been playing with it for a bit now and am wondering about a few > things... > > Could there be a e-offset error? (difference between where the axis is and > the o-ffset input) > > Can the e-offset ratio be disabled or changed from within linuxcnc? I > don't see an ini. hal pin for it. The reason is - for doing some of the > fancy spindle synced motion stuff I need to apply most of the acc/vel to > e-offset. This makes for very slow normal motion. In this situation - I > only need the high ratio during polygon motion > > Is this making sense? >
I was thinking the same when I did my tests. The problem I see with the error tracking is that there are two factors to consider. First there is the phase displacement error between the e-offset input and the actual offsetted axis position. And second there's an amplitude attenuation whenever the axis can't keep up with the desired e-offset. I guess the error should be splitted in to X and Y coordinates but I'm not sure. I mean this is not like a normal f-error like the one experimented with a normal axis where position magnitude is a considered a scalar (I mean the commanded and actual position of an axis). I don't know if there is an easy way to compute a near exact error but that's what I was trying to figure out to better tune my camshafts for lobe turning. In reality what I need is simple, I just need to detect the phase displacement as the spindle revs. I can't do anything about the amplitude loss, and since my test was to rough cut the cams I can live with a little bit of lift (amplitude) loss. For me, this is harder than what it looks like and the best way I found to solve this is to really overkill the speed and acceleration of the axis to be e-offsetted. Off course I'm far from being a LCNC/maths/programmer expert but that's roughly what I see. Thanks for sharing your progress Sam. El lun., 5 oct. 2020 a las 18:52, Rod Webster (<[email protected]>) escribió: > Sam, the pins are listed here > http://linuxcnc.org/docs/2.8/html/motion/external-offsets.html > If your offset motion is mutually exclusive to normal motion, you can trick > the system by doubling the axis and velocity and acceleration and use a > ratio of 0.5. This gives you full velocity and acceleration to eoffsets and > how we manage the Z axis with Plasma as normal Z axis motion is not used at > all when eoffsets are active. > > Rod Webster > *1300 896 832* > +61 435 765 611 > VMN® > www.vmn.com.au > > > > On Tue, 6 Oct 2020 at 03:08, Sam Sokolik <[email protected]> wrote: > > > I have been playing with it for a bit now and am wondering about a few > > things... > > > > Could there be a e-offset error? (difference between where the axis is > and > > the o-ffset input) > > > > Can the e-offset ratio be disabled or changed from within linuxcnc? I > > don't see an ini. hal pin for it. The reason is - for doing some of the > > fancy spindle synced motion stuff I need to apply most of the acc/vel to > > e-offset. This makes for very slow normal motion. In this situation - I > > only need the high ratio during polygon motion > > > > Is this making sense? > > > > thanks! > > sam > > > > _______________________________________________ > > Emc-developers mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-developers > > > > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers > _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
