#emc-devel | Logs for 2009-12-24

Back
[00:10:54] <micges> cradek: volatile home and halui pins both doesn't work in estop mode
[00:11:37] <micges> (I understand now why volatile home code isn't fully enabled in task)
[00:22:53] <micges> proposed patch to make volatile home works: http://www.pastebin.ca/1725472
[00:25:41] <micges> (this patch also makes possible to unhome in estop mode via halui unhome pins if they were added)
[01:00:12] <cradek> micges: if it works, that seems fine to me
[01:01:23] <micges> thanks
[01:08:06] <jepler> likewise
[01:08:59] <cradek> yes
[01:15:39] <jepler> (I also didn't test it)
[01:28:17] <CIA-62> EMC: 03micges 07master * r0a7f2d34fcad 10/src/emc/ (motion/command.c task/emctask.cc): Enable volatile home support in task
[01:31:02] <cradek> cool, I guess it was only 98% done, not 99% like I thought
[01:31:06] <cradek> thanks for finishing and testing it
[01:37:44] <cradek> wow, a probe is sure nice for measuring a hole circle
[01:43:14] <micges> cradek: do you have machine with index homing architecture?
[01:44:25] <cradek> yes two
[01:45:02] <micges> you have some latches for index pulses?
[01:45:48] <cradek> I don't understand that
[01:46:18] <micges> I wonder how can be done if encoder is high resolution and there is chance to miss a index
[01:47:56] <cradek> that is the responsibility of the encoder hardware
[01:48:23] <cradek> emc doesn't read the index directly
[01:48:47] <cradek> the hardware sees index and either resets the count or latches the count, and then at the next servo cycle emc notices
[01:50:04] <micges> yes that is on full servo
[01:50:25] <cradek> yes, what is your question
[01:50:27] <cradek> ?
[01:50:39] <micges> my option is that I have closed loop stepper system (electronics)
[01:50:58] <cradek> oh ok
[01:50:58] <micges> and I have pulled only index signal to mesa port
[01:51:34] <cradek> it would be nice if our (various) stepgens handled index homing
[01:58:40] <jepler> hm2 stepgen doesn't have reset-on-index functionality.
[01:59:26] <cradek> does it have latch?
[01:59:51] <jepler> I don't think it has that either
[02:00:08] <jepler> there are no input pins associated with stepgen instances
[02:00:14] <cradek> hm
[02:00:32] <cradek> so it's a big project, and involving pcw
[02:01:26] <cradek> software stepgen could easily (?) read index at the base cycle and latch or reset
[02:01:52] <cradek> but if hm2 (and ppmc?) can't do it, what's the point
[02:04:21] <cradek> micges: you could fake it by going slow enough to not miss the index, and use it in hal to mask the real home switch at the appropriate time (based on the home-state hal pin)
[02:05:11] <cradek> but only do that if you have a limit switch past it... I bet you could still miss it.
[02:08:03] <micges> thanks I'll try make some tests
[02:08:12] <micges> good night all
[02:08:19] <cradek> goodnight
[09:51:48] <alex_joni> micges: good morning
[09:52:16] <micges> good morning
[09:52:40] <alex_joni> Wesolych Se
[09:52:46] <alex_joni> Wesolych Swiat
[09:55:49] <micges> thanks
[09:56:06] <micges> Wesolych swiat to You too ;)
[11:26:23] <CIA-62> EMC: 03micges 07master * r88ba8a07d9e1 10/src/emc/usr_intf/halui.cc: Add joint unhome pins to halui, update comments
[17:50:00] <Guest358> Guest358 is now known as skunkworks_