I did not think linuxcnc had an issue. Maybe rose colored glasses.. ;) Part of the issue is you thought the current TP behaviour was a 'bug' while it just was performing as designed. The 1 segment look ahead doesn't work as well as you get into higher speeds/shorter segments. My eyes where opened when I compared Mach to linuxcnc a half a year ago.
Heck - in all this testing I have done - I have found a few constraint violations in the current linuxcnc. http://imagebin.org/295715 in certain rare-ish situations the acceleration would go over. in my testing the 30in/sec^2 would approach 35. I think rob has fixed them.. http://imagebin.org/295717 sam On 02/25/2014 02:19 AM, Steve Blackmore wrote: > On Mon, 24 Feb 2014 23:09:15 -0600, you wrote: > >> On 02/24/2014 10:03 PM, sam sokolik wrote: >>> I am surprised people still put up with me... I do have a lot of help >>> from the community. The compact 5 configs using the latch and dual >>> stepgens would not have been possible if not for Jeff and Chris's >>> smarts.. (I am the big picture guy usually...) >>> >>> As far as machs acceleration violations. I am pretty confident that >>> they are real. (again unless I am doing something wrong) If you want >>> to see some huge violations - run the tort.ngc program. It has lots of >>> diabolical motion. (granted - it is diabolical) I think though you >>> should be able to throw any gcode to a control and is should process it >>> correctly. >>> >>> >> I saw a number of downward spikes where the velocity >> suddenly dipped. >> I can imagine those were times when the Mach driver was >> unable to >> get the CPU. But, they appeared to be where the instantaneous >> velocity dipped by 25 - 33%, which is a HUGE velocity >> discontinuity >> for a motor to follow. > Despite what those figures appear to show, Mach's output behaves better > than LinuxCNC's. It sounds much much smoother. I posted many moons ago > Art's explanation on how it works and a Youtube video which clearly > shows LinuxCNC's problems in comparison to Mach. > > I no longer use either on my router, I use an external control so am > unable to test Rob's new TP. The annoying part about all this was the > denial there was a problem in the first place and the time taken to get > a fix. > > Well done Rob for taking this on and shame to those who buried their > head in the sand or were in denial. > > Steve Blackmore > -- > > ------------------------------------------------------------------------------ > Flow-based real-time traffic analytics software. Cisco certified tool. > Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer > Customize your own dashboards, set traffic alerts and generate reports. > Network behavioral analysis & security monitoring. All-in-one tool. > http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ Flow-based real-time traffic analytics software. Cisco certified tool. Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer Customize your own dashboards, set traffic alerts and generate reports. Network behavioral analysis & security monitoring. All-in-one tool. http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
