Some further thoughts. On reviewing the available pin outs using the G540 with four axis I have left available 4 input lines and 2 output lines. The VFD PWM output is already dedicated, as are the 4 axis step and direction lines. I want to reserve 3 of the inputs for limit/home switches, so this leaves 1 for spindle input, thus I probably should dedicate this to an index pulse. I would then use one output as a direction signal and the other as power on. This gives me the option to try stabilizing the spindle speed to that commanded and also use the index as a sync pulse. Since the direction is commanded by the computer I already have it available and use software to determine on which edge to sync. At this time this limitation applies to both machines. Longer term plans are for me to change to the Mesa 7I43 I/O card which I believe will let me connect to the G540 and add feadback from all planned 5 Axis plus the spindle. (G540 = 17 pins, 6 Quadrature + index channels = 18pins for 35 of the available 48). I still have to learn how much of this can be done with hostmot2 and whether I need to add additional functions to the hostmot2 configurations to determine which size of FPGA to use. I think it will also allow me to use separate home switches and possibly some control for a tool changer (dreaming). Please warn me of pitfalls with this scheme.
Hubert Hubert Bahr wrote: > I am actually working on two machines in parallel. An X2 and an X3. > The X2 already has LMS addon pulley set for spindle power. For it I am > thinking of just Drilling single hole on one side of the larger pulley > and sensing that as a index pulse and starting along the lines suggested > here > http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Closed_Loop_Spindle_Speed_Control > > Tapping may not be desired on this machine. > > For the X3 I am planning to build a belt drive mod for the machine that > will give me three speed ranges, I am looking at cutting interrupter > notches into one edge of the spindle pulley or at least attaching a disk > containing the notches. The quadrature encoder/sensors I am familiar > with were designed to be attached to a servo shaft, not a tool holding > spindle. I am guessing if I provide slots of uniform presence and > absence of material around the periphery of a disk this will give my the > 50% duty cycle. Then if I set two detectors in a relationship where one > is offset from the other half a slot away this provides a method of > detecting 4 transitions per slot instead of just two. And then the > relationship of the highs and lows determines direction. A third > detector senses a single slot/hole elsewhere that rotates in a constant > relationship with the other slots. I think I am approaching the limits > of the number of pins available from a parallel port for a 4 Axis mill. > The number of slots to use is a trade off between desired/necessary > resolution. What resolution is required for tapping? Isn't the > starting edge of an index pulse sufficient, or do they need to track > loading on the spindle? > > Hubert > > Gene Heskett wrote: > >> snip >> If threading is to be done, then emc must know the position of the spindle >> pretty accurately. The usual output format of a quadrature encoder, where >> the A and B signals are close to 50% duty cycle, and the switching edges of >> one phase are taking place pretty close to the centers of the stable state >> of >> the other, will give emc a pretty good idea where the spindle is. With the >> addition of an index pulse once per rev to tell emc when the spindle is at >> zero degrees, then it can do rigid threading. To drive the index pulse so >> it >> is sensitive to only one edge, one of the things discussed in that thread, >> one can develop a direction signal from the quadrature signals that would >> control the phase of the edge detector so the same physical edge of the >> signal is used as the index regardless of the direction its turning. Either >> edge of the hole could be used. >> >> As for the 6000 rpms, there was some math presented in the previous thread >> that would show what the limits were, given a perfectly symetrical set of >> signals that were also phased at 90 degrees. >> >> I currently have a wheel installed that I cut with the code I posted a day >> or >> so back, this is the code from the emc wiki page, and I note after I posted >> it, that I didn't adjust the comments although that code itself has been >> adjusted for fewer holes, on smaller circles. Cutting with a 1/16" bit, the >> holes are nearly round, and the index hole is actually more like a notch in >> the edge. There isn't enough room to get any fancier than that in a 7x12's >> head. It is installed between the two nuts that adjust the bearing >> pre-load, >> and I'll need smaller opto assemblies to pull it off, the LMC piece, since >> I'd need 3 of them, would be difficult to fit into the space available, so >> I'm still in the thinking out loud stage. >> >> >> >>> --------------------------------------------------------------------------- >>> --- Come build with us! The BlackBerry(R) Developer Conference in SF, CA is >>> the only developer event you need to attend this year. Jumpstart your >>> developing skills, take BlackBerry mobile applications to market and stay >>> ahead of the curve. Join us from November 9 - 12, 2009. Register now! >>> http://p.sf.net/sfu/devconference >>> _______________________________________________ >>> Emc-users mailing list >>> Emc-users@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/emc-users >>> >>> >>> >> >> > > > ------------------------------------------------------------------------------ > Come build with us! The BlackBerry(R) Developer Conference in SF, CA > is the only developer event you need to attend this year. Jumpstart your > developing skills, take BlackBerry mobile applications to market and stay > ahead of the curve. Join us from November 9 - 12, 2009. Register now! > http://p.sf.net/sfu/devconference > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ Come build with us! The BlackBerry(R) Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9 - 12, 2009. Register now! http://p.sf.net/sfu/devconference _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
