Hi Mario I don't have a scope on the pulse train but I don't see any rt error messages or "hear" any long delays in the stepper motor with the setup I described a couple weeks ago. It also uses a Semperon processor but a PCI parport card.
I'm pretty certain that at the high pulse rates I'm getting I'd get step loss if a 1 ms pulse delay stuck itself in there. I'll be interested in the code you are planning for a high frequency pulse generator. Rayh On Sun, 2006-12-24 at 20:26 +0100, Mario. wrote: > I know it does not change - should not - so, it is either a hardware > bug, or a "software feature" - whatever it is, it results in 1 > milisecond updates, instead of 20 microseconds - my thought was that > it could be the sudden change by 5% in the feed rate in zero time (not > sure how this is applied in the code). If division by time would be > involved, we could get large numbers... not sure if the system tick is > set to 1ms or 0.5ms - if the former, my theory was it could be > something around that. > > One explanation is that moving the slider can cause more than just one > write to the feed override rate, thus blocking some calculations or > counters, depending on how is this done. > > I was looking now into HAL documentation, stepgen.0 does not have a > feed override connection, but the update_freq() has some d/dT > calculation from position-cmd. > > I need to learn how to use the probing utilities probably, right? If > one could hook position-cmd input to stepgen, start recording and > quickly jump on the override slider > > > Or it can be something simpler? > > code from stepgen.c > > /* this periodns stuff is a little convoluted because we need to > calculate some constants here in this relatively slow thread but the > constants are based on the period of the much faster 'make_pulses()' > thread. */ > > hmm, still not sure. > > On 12/24/06, Jeff Epler <[EMAIL PROTECTED]> wrote: > > Moving the "feed override" slider does not modify the 'period' of the > > HAL functions. 'period' is the approximate number of ns between > > invocations of a realtime thread, and does not change during an emc2 > > run. > > > > Jeff > > > > ------------------------------------------------------------------------- > > Take Surveys. Earn Cash. Influence the Future of IT > > Join SourceForge.net's Techsay panel and you'll get the chance to share your > > opinions on IT & business topics through brief surveys - and earn cash > > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > > _______________________________________________ > > Emc-developers mailing list > > Emc-developers@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-developers > > > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys - and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Emc-developers mailing list > Emc-developers@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-developers > ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
