On Sunday 28 April 2019 10:53:49 Peter C. Wallace wrote: > On Sun, 28 Apr 2019, Gene Heskett wrote: > > Date: Sun, 28 Apr 2019 10:47:13 -0400 > > From: Gene Heskett <[email protected]> > > Reply-To: "Enhanced Machine Controller (EMC)" > > <[email protected]> > > To: [email protected] > > Subject: [Emc-users] stretch testing install Q > > > > Greetings; > > > > As I wrote yesterday except for two things. > > > > 1. It won't let me run synaptic from anyplace but its own keyboard. > > 2. Very consistent and poor latency in the 53 microsecond range. > > > > Something in pam is stopping synaptic. > > > > And is there anything I can put in the kernel command line that will > > improve the latency? > > > > To PCW, is that latency enough to bother a 5i25|7i76 combo where > > there is no base thread? > > > > It seems to be quite consistent, you can clear the stats and its > > back above 50 u-s by the time you've unclicked the mouse. And I just > > noted that removing the base-thread drops the servo to around 30 > > u-secs. > > 100 usec is usually fine for Mesa PCI hardware and even 500 usec would > be OK if the DPLL is used to retime things like encoder count and step > count sampling
ISTR seeing that in the 5i25 docs someplace. I think I was using it in a previous config. I think I am using that in the pi<spi>7i90 on the Sheldon as the pi's latency is known to be over 100 u-secs regularly. yes, in the .hal file for a 7i90HD: setp hm2_[HOSTMOT2](BOARD).0.dpll.01.timer-us -150 is that the right syntax for a 5i25? Must be, it just accepted a -50 according to the startup log on the G0704. The other 2 machines are running on 5i25's and D525MW mainboards with 6 u-secs max latency. Latency-test measurements on the pi are probably confused. in real life as theres a 2nd, very slow thread also running on the pi, handling the joggin and such functions. I have a pair of the MPJA 100 ppa dials mounted on its apron to replace the old manual cranks, and a 100 to 200 millisecond thread handles all that extra stuff nicely with no noticeable lag between my hand and the machines reaction. Saves a lot of cpu cycles by not running that stuff at servo-thread rate. ISTR we discussed that before. I do need to change the jog mode though, I can wind it up way too easy and crash it, breaking tooling. Theres a mode where it stops when I stop and I need to convert to that. The DRO's track whatever the machine does in either event if I understand it right. From man motion: axis.L.jog-vel-mode IN BIT When FALSE (the default), the jogwheel operates in position mode. The axis will move exactly jog-scale units for each count, regardless of how long that might take. When TRUE, the wheel operates in velocity mode - motion stops when the wheel stops, even if that means the commanded motion is not completed. The latter would be better I think. But what does that do to the size of the jog per click that I went to quite some length to setup? .0002" to 20 thou a click. ??? The theory was that I could jog right close to position, then reduce the jog size to almost zip and fine tune it to the correct size. I'll give it a shot anyway. :) Thanks Peter and all. > > > > 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 > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users > > Peter Wallace > Mesa Electronics > > (\__/) > (='.'=) This is Bunny. Copy and paste bunny into your > (")_(") signature to help him gain world domination. > > > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users 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 [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
