On Saturday 20 October 2018 01:12:02 Nicklas Karlsson wrote: > > > On 10/19/2018 2:12 PM, Nicklas Karlsson wrote: > > > > Probably not, I have done quite a lot of search on Internet, > > > > have a few books and have taken two courses specifically in the > > > > area but if you have readily available I could always take a > > > > look. > > > > > > > > I find a lot of documents about PID controllers and theory using > > > > bode, nyquist diagrams. > > > > > > > > I found state space feedback and bode, nyquist diagram are > > > > useful also for state feedback but then state space model is > > > > available I prefer method to calculate gains using the model. I > > > > could do state space feedback including integral action but get > > > > the feeling a chapter or so is missing, maybe I have to write it > > > > myself. > > > > Someone who really understands it needs to.. Whats out there tends > > to talk more theory than hardware. ... > > The hardware need to be properly modeled. There are plenty of > litterature available including for freely accesible on internet about > electric motors. Static friction may be a problem but I think a > properly lubricated machine have rather viscous friction which easy to > put in a mathematical model suitable for control of machine.
The std milling machine is a bigger problem WRG static friction than it first appears, particularly the X motions. because the table is so long, there's a huge disparity between the friction of the ways when the table is centered vs when its at the end of its range. I am limited to about 1/4 the table and motors speed just because I put the homeing switch near the end of its travel. And this I use as an indicator that its time to take some time to clean and relube the ways, it will head toward the switch, do its homeing thing at creep speeds where the motor is adequate, but once its homed, I also have to use quite a low speed just to get it back off the 50 lb imbalanced & high drag point back to center, otherwise the 470 motor hangs and home is lost. I could move the home switch to the middle of the travel, but that would take a looong bar to activate the switch so that once tripped, it would stay tripped and the home logic would know which way to move to find the switch. I had no clue that it knew which way to move to find the switch until I had put individual switches on the lathes, starting with TLM, using the bottom of the tapered Z gibs to activate miniature roller tipped microswitches, which never move close enough to the spindle to release the switch by going past it. The docs on homeing don't mention that and it was a pleasant surprise when I was adjusting the homeing constants in the .ini file, and left it in the switch tripped condition when I restarted LCNC to test and found its first move was to the right to get off the switch! I reused the g0704's original stops, round things that can be loosened and moved along the front of the table, and which are set to trip about 1/16" from crashing into the end of travel just because an inch away gives it a small chance of rideing on by and releasing the switch again. Same theory on the z, clear at the top of the post & maybe 30 thou from a crash. The X problem could be solved by making a bar of 1/8 alu so the switch remained closed once closed. I'll see about doing that when I can see out of both eyes again, I'm in between cataract operations ATM. I seem to have wandered off topic some but the static friction when the x table is way off center is a point to consider. Been tempted to experiment with a bar off the top of the post to attach screen door springs to the ends of the table to help support that unbalanced condition. But I need to make some more round tuits first. ;-) thanks Nicklas. -- Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
