#linuxcnc-devel | Logs for 2015-03-16

[08:26:57] <skunkworks> pcw_home, ran over the weekend without and realtime errors. (2khz)
[08:27:20] <alex_joni> any*
[08:34:02] <skunkworks> exactly!
[08:34:10] <skunkworks> Hi alex_joni
[08:39:06] <alex_joni> hi sam
[08:40:22] <skunkworks> how is it going? Number 2 at the end of april.. Don't know if it is a boy or girl.
[10:41:39] <skunkworks> pcw_home, I don't know if I have ever seen marius's problem.. (although it usually takes me 3 tries to get all the network config correct)
[10:43:34] <pcw_home> Yeah, he also has a slow machine and may have all hardware enabled
[10:43:35] <pcw_home> (a blank config line does this) so that may be an issue also
[10:45:53] <pcw_home> And I have seen that error when changing network setup
[11:26:32] <alex_joni> skunkworks: congrats
[11:26:51] <skunkworks> Thanks.. Shirley is feeling it...
[11:26:57] <alex_joni> I can imagine :)
[11:27:10] <alex_joni> we're starting school in september
[11:27:26] <skunkworks> wow.. Stella is just over 2.5 years.
[11:27:48] <alex_joni> victor is 5.5, he'll be 6 by autumn
[11:28:06] <alex_joni> over here they introduced a preschool class (0th grade)
[11:28:20] <alex_joni> but it gets organized and teached in schools not kindergardens
[11:28:53] <alex_joni> I see they call it 'reception' in the UK
[11:29:14] <skunkworks> pre-k, pre-school..
[11:29:28] <skunkworks> 4-k
[11:30:49] <alex_joni> I think there's a small difference, as here it's also compulsory
[11:31:02] <alex_joni> or mostly compulsory :)
[11:31:34] <skunkworks> interesting.
[11:31:42] <alex_joni> anyways.. he just had his first exam on saturday
[11:31:48] <skunkworks> wow
[11:32:11] <alex_joni> but we should probably take this elsewhere :D
[11:32:25] <skunkworks> heh
[12:24:26] <skunkworks> jepler, Hi!
[12:26:44] <seb_kuzminsky> jepler!
[12:29:55] <skunkworks> seb_kuzminsky, Don't spook him!
[13:11:27] <KimK_laptop> skunkworks: Did you see the piston and connecting rod animation?
[13:12:03] <skunkworks> I did - very cool
[13:12:27] <KimK_laptop> OK, great! Hope all is well there.
[13:13:50] <skunkworks> things are well - are you still at stuarts?
[13:17:03] <KimK_laptop> Ha, that's a long story, more later. For now I'll just say yes.
[13:17:54] <skunkworks> heh - interesting.. :)
[13:18:49] <KimK_laptop> Maybe there'll be a fest somewhere.
[13:19:13] <KimK_laptop> (That we can both attend, I mean.)
[13:19:37] <skunkworks> Yes - dad is hoping to go to one this year
[13:20:41] <KimK_laptop> I haven't heard anything, but it's still early. Was the last one the Houston one, or was there another Ann Arbor one?
[13:21:18] <KimK_laptop> I thought the Ann Arbor folks tried a restart? Maybe I'm mistaken.
[13:21:20] <skunkworks> I think there is going to be one (more like the ann - arbor one) near detroit
[13:22:37] <KimK_laptop> Oh, OK, maybe that's the one I'm thinking of. This summer? If so, that will make it hard for anyone to hold a competing fest (maybe that was the idea?)
[13:23:02] <skunkworks> who knows...
[13:23:18] <skunkworks> It certainly isn't a linuxcnc only fest
[13:26:08] <KimK_laptop> I've got to shutdown and relocate. Good chatting with you, maybe more later. Say hi to everyone for me. Oh, wait, a survey: what's your PCB app, Eagle, gEDA, Kicad, other?
[13:26:38] <jepler> hi all
[13:26:50] <skunkworks> I used eagle.. Once you learn it - it is pretty darn powerful..
[13:26:57] <KimK_laptop> Hi jepler
[13:27:02] <skunkworks> Hi jepler - good to see you!
[13:27:20] <jepler> skunkworks: likewise
[13:27:35] <KimK_laptop> Yes, welcome back(?)
[13:27:46] <skunkworks> how was the vacation? :)
[13:27:59] <jepler> skunkworks: relaxing
[13:28:04] <skunkworks> awesome
[13:28:37] <jepler> I haven't decided how much time I'll be spending on linuxcnc but I decided it would be OK to hang out on IRC some
[13:29:31] <KimK_laptop> You guys can chat, I've got to go. More later, maybe. Bye!
[13:32:35] <skunkworks> jepler, I have been using uspace quite often - no issues.. damn impressive.
[14:19:10] <seb_kuzminsky> jepler: i'm working on a robot arm, also using uspace, it's working great :-)
[14:26:19] <jepler> cool both of you
[14:34:47] <cradek> yuck, something in SF's list handler still escapes ^From_
[16:02:05] <jepler> has anyone tried building 3.18.x-rt yet?
[16:02:42] <jepler> unfortunately, it looks as if the debian 8 kernel, 3.16, will never get a rt patch
[16:03:05] <jepler> hi PCW
[16:03:51] <PCW> Hi
[16:03:54] <cradek> is there a particular pci parport card that can do epp that anyone can recommend?
[16:04:58] <PCW> Only thng I know is avoid NetMOS
[16:05:08] <PCW> =MOSChip
[16:05:21] <cradek> yeah I was hoping for a positive, not negative, recommendation
[16:05:29] <cradek> the wiki has several pages with pictures of cards
[16:05:59] <cradek> I looked in the mesa store - I once had a fever dream that you resold one that was known to work
[16:06:05] <jepler> I used Oxford Semiconductor Ltd Device c110 (pci express) with mesa 7i43 and 7i90 -- that's the one I offered you, but mine was pci express
[16:06:22] <jepler> but I haven't tried it with pluto, which iirc is what you're trying to get running
[16:06:28] <cradek> yeah that's the rub
[16:06:41] <seb_kuzminsky> jepler: i think pcw tried 3.18-rt and said it works better than 3.2-rt
[16:07:43] <seb_kuzminsky> i'm working with some gsoc students to build new kernel debs for us, rtai is the #1 goal, newer rt-preempt is #2
[16:07:52] <jepler> seb_kuzminsky: cool beans
[16:08:13] <jepler> (are beans, in fact, cool? I have my doubts)
[16:08:43] <PCW> 3.18 certainly works better than 3.2 on newer hardware
[16:09:33] <PCW> (works well on my laptop that 3.2 is hopeless on)
[16:10:59] <cradek> aha here's one with SUNIX SUN1989
[16:11:06] <seb_kuzminsky> coffee beans are
[16:12:15] <cradek> since the SUN1888 works, the SUN1989 sure might too
[16:19:11] <PCW> I think i tested one of the earlier ones and it was OK
[16:23:45] <cradek> huh, I haven't ordered from newegg for years and years
[16:23:55] <cradek> wonder if they pissed me off last time
[16:24:07] <cradek> PCW: I'll report back, fwiw
[16:25:19] <PCW> yeah would be nice to know which current ones work (almost all MB ports seem to work even new ones)
[21:05:24] <kwallace> I have a routine that does a fast rough probe, backs off a little, then a slow fine probe. It looks like the fine probe stops right at the trigger point and cycles the signal and sometimes produces and error. http://wallacecompany.com/tmp/Screenshot-probe_bounce-1a.png
[21:07:17] <kwallace> I'm thinking a hal one-shot could be used to hold the signal stable long enough to process the trigger on the first edge, then go on to the retract move.
[21:08:02] <kwallace> Has anyone tried to solve this problem already?
[21:09:08] <kwallace> By the way, the above screenshot did not produce an error, but some do.
[21:15:45] <kwallace> Another thing that comes to mind, it might be nice to have a parameter to adjust how far the probe travels after a trigger event to ensure the new state is stable.
[21:20:58] <cradek> kwallace: my real renishaw has a time constant of several ms and the opto22 input module may add a couple more. definitely the problem is solved before - but in hardware
[21:22:36] <cradek> kwallace: you know the probe is read every ms (or whatever your servo cycle is), so seems like you "just" need to know how far your probe has to move to be sure it's triggered to select the appropriate slow feed rate
[21:23:23] <cradek> if the probe has .004" of noise in it, you should be moving .010/ms and not .001/ms
[21:34:52] <cradek> yeah renishaw MP7 "enhanced trigger circuit" nominal 7 milliseconds
[23:14:36] <memfrob> Ah i was able to narrow down the kernel crashes to rtai
[23:15:04] <memfrob> not caused by how linuxcnc handles the new ipipe code which is nice
[23:15:47] <memfrob> jepler: nice to see you again!
[23:16:22] <memfrob> i've been absent from rtai and linuxcnc for months and just a few days ago I started getting back into it again