On Friday 25 April 2014 11:08:55 Florian Rist did opine: > Hi, > thanks for all your comments so fare. > > >> 0. move to home position (0آ°) > >> 1. wait for low frequency trigger, a hardware or a software signal > >> 2. within say 10آ° accelerate and synchronises to the 50Hz trigger > >> > >> so that the 10آ° position is reached 20ms after the last trigger > >> and at a given speed (specified in degree per trigger pulse, > >> not in seconds as the pulse rate of the trigger may change > >> slightly) > >> > >> 3. keep on moving at the give speed (in sync with the trigger) for > >> 100آ° 4. decelerate an move to a safe position > > > > This sounds like fun. It might "just work" with a G33 or G33.1 > > synchronised move, or it might need an elaborate system of PIDs and > > PLLs. > > I've been doing some research on rigid taping and I think is should > work. > > As fare as I understood it rigid taping takes the information from the > spindle encoder and moves Z accordingly. > > So in my case I take the ca. 50 Hz signal from the scanner and treat it > like a spindle encoder signal (if necessary I can create A/B and an > index pulse somehow, eg. using an external micro controller) and my > rotary axis as a linear Z axis. OK. > > Now, how is acceleration and deceleration handled? > > I move to a specific starting location (G1 Z0 F100), start the > synchronised motion (G33 Z90 K0.1). This should give me a total scan > angel of 90آ° and 900 scan lines (not considering acceleration), total > scan time is 18 seconds (+ acceleration/deceleration time). > > The LinuxCNC G-Code reference provides some information on acceleration > [1] but I'm not sure if I really understand the paragraph. > > Lets say I takes about 2 seconds for the 'Z' axis to reach the necessary > speed, this is equivalent to about 100 'spindle encoder' ticks (a=50 > ticks/s). Where does my synchronised motion start? I guess at a 'Z' > position of about 10 (z=1/2*t^2*a=100 ticks=10آ°). So I have to drop my > first 100 scan lines and start collecting them at Z=10, or start at -10. > > What about the end of the synchronised motion? Using G33 the spindle is > not revered and there will no return move. Where does Z stop? At the > given Z value (90) or at the given Z value + 10 for deceleration? > > See you > Flo
And you have just stumbled over the conundrum that prevents one from changing the spindle speed in the middle of a G33.1 peck cycle you've just carved the gcode for. To do so ignores the accel delay in the z, and it will actually become synchronized at some point determined by the accel settings, but many tens of degrees after the index has passed. I am hoping that at some point that these "synchronized" motions will be made smart enough to spend an extra turn of the spindle while it calculates this delay based on those settings, and then issues the start synch command far enough ahead of the spindle index that it will be truly in sync at the index pulse. I have not experimented, but for Z in that case it might be helpful, since the Z speed when threading is not generally anywhere near as fast as a rapid move, to allow a larger much larger Ferror, and set MAX_VEL & MAX_ACCEL that is much more brutal, based on the theory that it could play catch up and have lots less actual lag in terms of the rotational delay. If the start point for the sync is a couple turns of the spindle before the tool contacts, giving it time to truly get in step with minimal error, it might just work. If I get to the shop today, its set in to rain now, I might try playing with that. As long as it restores to the same starting position, we haven't lost any counts. This would apply to both the G33 stuffs and equally to the G76. It might also beat a 1" travel dial indicator into junk too. If nothing else, I can afford to throw away some rapid speed since my lathe can do at least 60 ipm for z, and close to 100 ipm for x. Way more than it needs. > [1] http://linuxcnc.org/docs/html/gcode/gcode.html#sec:G33-Spindle-Sync > > ------------------------------------------------------------------------ > ------ Start Your Social Network Today - Download eXo Platform > Build your Enterprise Intranet with eXo Platform Software > Java Based Open Source Intranet - Social, Extensible, Cloud Ready > Get Started Now And Turn Your Intranet Into A Collaboration Platform > http://p.sf.net/sfu/ExoPlatform > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users Cheers, Gene -- "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> US V Castleman, SCOTUS, Mar 2014 is grounds for Impeaching SCOTUS ------------------------------------------------------------------------------ Start Your Social Network Today - Download eXo Platform Build your Enterprise Intranet with eXo Platform Software Java Based Open Source Intranet - Social, Extensible, Cloud Ready Get Started Now And Turn Your Intranet Into A Collaboration Platform http://p.sf.net/sfu/ExoPlatform _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
