On Saturday 22 February 2014 14:59:52 John Kasunich did opine: > On Sat, Feb 22, 2014, at 02:20 PM, Mark Tucker wrote: > > John > > > > Thank you so much for the insight of how that work and it explains a > > lot. But as you stated it would be hunting back and forth.Which i > > assume it would have to pulse the step line? > > Lets take my example of a 100uS base thread and an > accumulator that makes a step every time the 5th > digit changes. > > Suppose the position command is 324,556, and you start > out somewhere below that, say 300,000. > > The velocity is positive, and a positive number is added > to the accumulator 10000 times a second. At some > point, the accumulator passes 320,000, and the final > step of the move is sent to the drive. There is still a > small position error (4556 counts, or 0.4556 steps). > The position loop will keep commanding a non-zero > velocity until the accumulator reaches 324,556. It > might overshoot a bit - maybe to 325,920. Then it > starts going back down. It hunts around and hopefully > settles down at 324,556. Maybe it never fully settles, > but is bouncing around from 324,550 to 324,561 or so. > But in any case, it never reaches 330,000, so it never > sends another forward step. And it never gets as low > as 320,000, so it doesn't send a reverse step. > > > And a number of people have scoped the outputs and only found the Dir > > line hunting back and forth,i wonder why it is not detected on the > > step line.? > > > > > > And if it is only the dir line,why would the motors make a noise at > > all? > > That is a very good question, and depends on the design of the drive.
I am with Peter on this, until a step is received by the driver, the state of the dir pin shouldn't mean squat, however if the fluttering sign is being propagated on to the output of the dir pin, it seems to me that the problem could be solved by not toggling the dir line until the NEXT loop WILL issue the step. The step timing should not be held back as that would confuse the TP, but it seems like we could detect the need for a step one pass through the thread hence, and use that to toggle the dir line. Maybe a hal calculated guard band value dependent of the instant speed might be needed to handle the transition from stopped to moving? How much more complex would that actually make the stepgen? 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> NOTICE: Will pay 100 USD for an HP-4815A defective but complete probe assembly. ------------------------------------------------------------------------------ Managing the Performance of Cloud-Based Applications Take advantage of what the Cloud has to offer - Avoid Common Pitfalls. Read the Whitepaper. http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
