On Sunday 07 March 2021 14:37:12 John Dammeyer wrote: > > -----Original Message----- > > From: Mark Johnsen [mailto:[email protected]] > > Sent: March-07-21 10:22 AM > > To: Enhanced Machine Controller (EMC) > > Subject: Re: [Emc-users] isolcpus? > > > > On Sun, Mar 7, 2021 at 7:25 AM Gene Heskett <[email protected]> wrote: > > > On Sunday 07 March 2021 06:58:01 Gene Heskett wrote: > > > > > It takes at least 5 seconds for it to get up to search speed, > > > > > and another 5 seconds to get back to a stop after the home > > > > > switch has been found, coasting 20-25 degrees getting stopped, > > > > > then it goes into latch speed and takes another minute to get > > > > > back to the switch edge. And theres a random offset being > > > > > applied to the home position in the linuxcnc dro that does not > > > > > exist > > > > I don't know the inner workings of the LinuxCNC home switch > > routine as my machine doesn't have a home switch (someday > > hopefully). However, it sounds like an acceleration/deceleration > > setting for the homing routine is too low. At least that's what I > > would look at with an off-the-shelf motion system. > > > > This isn't related in terms of code to LinuxCNC, but in general > > motion control practice: > > For an old Parker compumotor 6k2 motion controller, the homing > > routines had their own Accel, decel, home to index (was it John D w/ > > the 2 mechanical switches), and other routines such as approach from > > CW or CCW, go past and slew back slowly, etc... I would look at > > that rather than the linux setup (isolcpus, latency, etc). But, I'm > > not an expert on this.. > > > > Mark > > I share the home and limit switch for each axis. The other end of > each axis has a limit switch in series with the other two axis. Total > inputs used is 4 then. The Z always homes first by itself to clear > tools. Then XY happen at the same time. > > I've also got soft limits set up so loading a G-Code program tells me > if I'm outside my work envelope regardless of G54, G55 etc. > > I think I used this document to determine what to set and how and it's > pretty comprehensive. > http://linuxcnc.org/docs/2.8/html/config/ini-homing.html
I don't think there is a diff, I'm on master with all my machines, playing the canary in a coal mine, good explanations in there. > I don't have a home or limit switch on my rotary A axis. And I just found out that touch off does do something to A but 0.0 is not 0.0 ack the dro. And the offset changed when I rebooted. So something is definitely fubar someplace. > This is Z and notice the sign of the OFFSET verses final velocity. > The knee moves away from the spindle towards the home/limit switch. > The Z value gets smaller (more negative) as we move closer to the > spindle just like on a lathe. > HOME_OFFSET = 0.2 > HOME_SEARCH_VEL = 0.8 > HOME_LATCH_VEL = 0.01 > HOME_FINAL_VEL = -0.2 > HOME_IGNORE_LIMITS = YES > HOME_USE_INDEX = 0 > HOME_SEQUENCE = 0 Yes, I'm doing that sequence too, only with faster speeds since my scale is per degree. > Compare that with Y axis which moves the table towards the column and > home/limit switch. That puts the Y=0 location on the table just like > an XY graph where the Y axis value increases as it we move up the > axis. HOME_OFFSET = -0.1 > HOME_SEARCH_VEL = -0.8 > HOME_LATCH_VEL = 0.01 > HOME_FINAL_VEL = 0.2 > HOME_IGNORE_LIMITS = YES > HOME_VOLATILE = 1 > HOME_SEQUENCE = 1 > > If you get those signs wrong it behaves very funny. > > John Dammeyer We'll see if I get a different scale this time, as I'm re-running that test now with the "lights out". See my reply to Phil T, for a description of the method. I let hal do the heavy lifting for that stuff. Thanks John D. Take care and stay well. 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) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis Genes Web page <http://geneslinuxbox.net:6309/gene> _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
