I'm sorry but I'm not following you or I'm confused or both. If you could post the complete component that might help.
Just because I managed to hack this component out by trial and error doesn't mean I feel like I own it in any way shape or form and would welcome anyone with more programming smarts to make changes to it to make it better. I don't feel comfortable making changes I don't understand. I really wish someone like Jeff would give it a once over and make any changes they see fit. JT On 6/27/2014 4:07 PM, Marius Liebenberg wrote: > On 2014-06-27 22:53, John Thornton wrote: >> Marius, >> >> Are you dropping the correction_velocity parameter? I'm confused. > Not dropping it but making it rate dependant > > #### This is the corrected line that was there > min_velocity = requested_vel - ( requested_vel > -(requested_vel*(velocity_tol/100))); > #### and this is what I added. It applies the velocity change over a > period of time to simulate acceleration > correction_vel= ((requested_vel *(speed / 100)) * fperiod)/10; > > This way the correction velocity becomes a percentage of the requested > velocity. It is applied over a period of time in order to prevent the > stepper to slip if the change it to big all at once. This might not be > an issue with servo systems but definitely with steppers. It is a > similar implementation approach that was followed in Mach. > > Hope this make sense. It did implement this on several systems with > great results. > >> JT >> >> On 6/27/2014 12:53 PM, Marius Liebenberg wrote: >>> Seb >>> That fixes some of the problem. If you can ask John to look at the extra >>> line of code I added to make the velocity change at a rate that is a >>> percentage of the requested velocity instead of banging the velocity to >>> the next value. The fix is attached to the comment in the ticket >>> manager. If he can look at that and comment please. >>> >>> >>> On 2014-06-27 19:08, Sebastian Kuzminsky wrote: >>>> On 6/27/14 05:22 , Marius Liebenberg wrote: >>>>> I just found that the issue is still open. The originator never closed >>>>> the damn issue. Can you please close it and somehow get the fixed >>>>> component into the release. >>>>> The ticket number is #348 under 2.6. >>>>> >>>>> >>>>> >>>>> >>>>> On 2014-06-27 13:09, Marius Liebenberg wrote: >>>>>> Hi Sebastian >>>>>> I am busy helping someone with a plasma setup and found that the changes >>>>>> that was made prior to the release of 2.6 does not reflect in master. So >>>>>> I would expect that it is not in 2.6 either. >>>>>> I am referring to a change that I made to the THCUD component. Someone >>>>>> else logged the issue but I fixed it and submitted fixed code on the >>>>>> issue tracker. >>>>>> >>>> Hi Marius, I spoke with John Thornton about the thcud velocity issue. >>>> He'd fixed the problem in thc.comp but not in thcud.comp, but he fixed >>>> it just now: >>>> >>>> http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2394f70 >>>> >>>> The fix is in 2.6 (v2.6.0-pre4-10-g2394f70) and master >>>> (v2.7.0-pre0-576-g79fd496). >>>> >>>> Does this fix the problem? If so i'll close #348. >>>> >>>> Thanks for reminding me that this was only half fixed.... >>>> >>>> >> ------------------------------------------------------------------------------ >> Open source business process management suite built on Java and Eclipse >> Turn processes into business applications with Bonita BPM Community Edition >> Quickly connect people, data, and systems into organized workflows >> Winner of BOSSIE, CODIE, OW2 and Gartner awards >> http://p.sf.net/sfu/Bonitasoft >> _______________________________________________ >> Emc-developers mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/emc-developers >> ------------------------------------------------------------------------------ Open source business process management suite built on Java and Eclipse Turn processes into business applications with Bonita BPM Community Edition Quickly connect people, data, and systems into organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner awards http://p.sf.net/sfu/Bonitasoft _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
