The only thing I would add - set G61 (exact stop mode) to be sure emc is using the max acceleration in the ini. (Thanks ChrisR for pointing that out).
sam On 10/11/2011 9:52 AM, s...@highlab.com wrote: > Chris taught me to tune emc2 axes with motion provided by gcode. > > Set your following error sloppy enough to let your machine run untuned, but > tight enough that if it runs away emc will stop it. Run axis (or your gui of > choice), run halcmd (for interactively setting pid gains), run halscope (for > getting a detailed look at pid and joint behavior), and finally run this > program: > > g91 ; incremental motion > o10 while [1] > g0 x1 ; go an inch right > g4 p1 ; pause for a sec > g0 x-1 ; go an inch left > g4 p1 ; pause for a sec > o10 endwhile > m2 > > > > ----- Reply message ----- > From: "andy pugh"<bodge...@gmail.com> > Date: Tue, Oct 11, 2011 04:43 > Subject: [Emc-users] all those who thirst knowledge. > To: "Enhanced Machine Controller (EMC)"<emc-users@lists.sourceforge.net> > > On 11 October 2011 07:15, Frank Tkalcevic<fr...@franksworkshop.com.au> wrote: >> I opened the first pdf and it has an article on pid tuning. It suggests >> applying a square wave command and tuning from that. How hard is it to >> generate a square wave in emc? Will this require hal trickery or can it be >> done with GCode? > It depends on where you want to generate the square wave, but you > could use either siggen (or just a 50% pwmgen) or a G-code loop > driving an analogue output. > ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2d-oct _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
