Re: [Emc-users] Slow G code (sam sokolik)
some more random paths running the circular arc blend rc3... Playing around with http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Oword#Sample_3_loops_math_example_spirograph P64Q.005 Current TP 6min 33s http://imagebin.org/300518 Plot of velocity. (peaks at about 180ipm) http://imagebin.org/300515 New TP 2min 9s http://imagebin.org/300517 Plot of velocity... (hits machine limit for most of the motion) http://imagebin.org/300516 Still having fun.. :) sam On 12/18/2013 8:56 PM, sam sokolik wrote: running it as fast as it could go - feedrate set to way more than the config would allow (config is 500ipm 30in/sec^2) While the penguin - (metric so .1 is about .004) parabolic blend (old)P.1=2:21 Strait G64=4:12 circular blend (new) P.1=1:41 Strait G64=2:27 this was from a while ago... there may have been improvements.. from my playing - around twice as fast for this configs. (for programs with lots of line segments) sam On 12/18/2013 05:41 PM, Greg Bentzinger wrote: Hey Sam; How does it the improved look ahead compare on the chips 3d NGC sample file? Greg -- Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349831iu=/4140/ostg.clktrk ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users -- Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349831iu=/4140/ostg.clktrk ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
On 03/20/2014 01:28 PM, sam sokolik wrote: some more random paths running the circular arc blend rc3... I just tested trochoidal milling using the example I created for gcmc: - image: http://www.vagrearg.org/gcmc/example-trochoidal-large.png - source: https://gitorious.org/gcmc/gcmc/source/example/trochoidal.gcmc Target: axis_mm simulator The runtime on 2.5.3: - trochoidal milling: 386s - total time: 482s Runtime on 2.6.0~pre, branch circular-blend-arc-rc3: - trochoidal milling: 110s - total time: 205s LinuxCNC reports the runtime at 2.7min (162s). It is clear that the old LinuxCNC version cannot keep the speed high and wanders between 500..600mm/min. The code sets the speed for the trochoidal path at 3000mm/min. The effective speed is still not entirely up to the requested speed and seems to hover arround 1900mm/min. That may be due to the simulator setup (maybe acceleration limits?) or the arc-blending not looking ahead far enough. Still, a very impressive improvement. -- Greetings Bertho (disclaimers are disclaimed) -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
Could you post that gcode somewhere? (I don't have time right now to play with your cool scripting language..) It could be the centripetal accel limits of the spirals.. sam On 3/20/2014 8:29 AM, Bertho Stultiens wrote: On 03/20/2014 01:28 PM, sam sokolik wrote: some more random paths running the circular arc blend rc3... I just tested trochoidal milling using the example I created for gcmc: - image: http://www.vagrearg.org/gcmc/example-trochoidal-large.png - source: https://gitorious.org/gcmc/gcmc/source/example/trochoidal.gcmc Target: axis_mm simulator The runtime on 2.5.3: - trochoidal milling: 386s - total time: 482s Runtime on 2.6.0~pre, branch circular-blend-arc-rc3: - trochoidal milling: 110s - total time: 205s LinuxCNC reports the runtime at 2.7min (162s). It is clear that the old LinuxCNC version cannot keep the speed high and wanders between 500..600mm/min. The code sets the speed for the trochoidal path at 3000mm/min. The effective speed is still not entirely up to the requested speed and seems to hover arround 1900mm/min. That may be due to the simulator setup (maybe acceleration limits?) or the arc-blending not looking ahead far enough. Still, a very impressive improvement. -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
On 03/20/2014 02:38 PM, sam sokolik wrote: Could you post that gcode somewhere? (I don't have time right now to play with your cool scripting language..) It could be the centripetal accel limits of the spirals.. Yes, I also suspect the centripetal accel limits may be a problem here. You can get the gcode here: http://www.vagrearg.org/gcmc/trochoidal.ngc.gz -- Greetings Bertho (disclaimers are disclaimed) -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
with my setup of 30in/s^2 - I get 3000mm/m (cool) the current TP peaks at about 600-700mm/s http://imagebin.org/300539 sam On 3/20/2014 8:46 AM, Bertho Stultiens wrote: On 03/20/2014 02:38 PM, sam sokolik wrote: Could you post that gcode somewhere? (I don't have time right now to play with your cool scripting language..) It could be the centripetal accel limits of the spirals.. Yes, I also suspect the centripetal accel limits may be a problem here. You can get the gcode here: http://www.vagrearg.org/gcmc/trochoidal.ngc.gz -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
On 03/20/2014 02:46 PM, Bertho Stultiens wrote: On 03/20/2014 02:38 PM, sam sokolik wrote: Could you post that gcode somewhere? (I don't have time right now to play with your cool scripting language..) It could be the centripetal accel limits of the spirals.. Yes, I also suspect the centripetal accel limits may be a problem here. You can get the gcode here: http://www.vagrearg.org/gcmc/trochoidal.ngc.gz I just set the axis accelerations in the INI file from 508 to 1000, but the result is the same. Actual result is 109s vs. 110s, which is within the tolerance of my eye-on-the-clock timing. This is where my expertise end ;-) -- Greetings Bertho (disclaimers are disclaimed) -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
On 03/20/2014 03:19 PM, sam sokolik wrote: the limit of sim is 1800mm/m - so if you change all the axis velocities to 58mm/s (3500mm/m) it peaks at 2900mm/min. Upping the acceleration to 1000mm/s^2 makes the program run at 3000mm/m steady. Indeed, with that setup the trochoidal path will be finished in ~73s and the total time ~168s. Also, LinuxCNC's runtime calculation of 2.7min will finally be correct for all practical purposes and intents. with my setup of 30in/s^2 - I get 3000mm/m (cool) the current TP peaks at about 600-700mm/s It works :-) However, if you are going to do trochoidal milling at those speeds, then you better make sure that you have one of those machines that weighs too much and is bolted securely or it will start dancing in the room. Great work! keep it up. -- Greetings Bertho (disclaimers are disclaimed) -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
like... http://electronicsam.com/images/KandT/oldkandt.JPG well - it could do the velocity.. but the acceleration is set pretty safe (not too many spare parts.) iirc 10in/sec^2 or something like that. sam On 3/20/2014 9:44 AM, Bertho Stultiens wrote: On 03/20/2014 03:19 PM, sam sokolik wrote: the limit of sim is 1800mm/m - so if you change all the axis velocities to 58mm/s (3500mm/m) it peaks at 2900mm/min. Upping the acceleration to 1000mm/s^2 makes the program run at 3000mm/m steady. Indeed, with that setup the trochoidal path will be finished in ~73s and the total time ~168s. Also, LinuxCNC's runtime calculation of 2.7min will finally be correct for all practical purposes and intents. with my setup of 30in/s^2 - I get 3000mm/m (cool) the current TP peaks at about 600-700mm/s It works :-) However, if you are going to do trochoidal milling at those speeds, then you better make sure that you have one of those machines that weighs too much and is bolted securely or it will start dancing in the room. Great work! keep it up. -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Slow G code (sam sokolik)
For anyone who wants to experiment with settings for trochoidal milling, there is now an adapted version of Bertho's program for git master for use with [py]ngcgui: nc_files/gcmc_lib/trochoid-path.gcmc The routine is included by default as a tab in several sim configurations for gcmc usage. An independent install of gcmc (v=1.5.1) is required. Debs are (or will be soon) available from the buildbot. Thanks Bertho! -- Dewey Garrett -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users