#linuxcnc-devel Logs

Oct 14 2019

03:56 PM skunkworks: pcw_mesa: ^
04:00 PM skunkworks: running the same config on a hp 8300 - following error is .000023 and hasn't budged
04:02 PM skunkworks: rpi4
04:02 PM skunkworks: I think there are still issues with it :)
04:03 PM cradek: is that the same one gene is fighting with?
04:03 PM skunkworks: I think? he has a 3..
04:03 PM cradek: bleh
04:03 PM skunkworks: I haven't really had any issues with it - other than it's realtime is not that great (in my experience)
04:04 PM skunkworks: it works as a debian computer...
04:04 PM skunkworks: it seems like it is () close to being a very small form factor solution...
04:05 PM skunkworks: the preview is decent (you can pan and zoom pretty large programs) and over all is pretty zippy
04:08 PM skunkworks: ah - I guess I haven't cleaned my glasses in a few weeks.. Much better
04:11 PM cradek: ewww
04:11 PM cradek: I can't function unless I wash them every morning with soap
04:12 PM skunkworks: windex works well in a pinch
04:12 PM skunkworks: heh
04:12 PM skunkworks: I think I know my problem...
04:13 PM skunkworks: I think an update stomped on the realtime kernel
04:13 PM skunkworks: Linux raspberrypi 4.19.75-v7l+ #1270 SMP Tue Sep 24 18:51:41 BST 2019 armv7l GNU/Linux
04:14 PM skunkworks: amazing it worked as well as it did?
04:14 PM cradek: oh! wow
04:19 PM pcw_mesa: Ahh that wont work too well, but you could try my VPE patch
04:22 PM pcw_mesa: (which masks bogus following errors caused by Ethernet read data timeouts)
04:38 PM skunkworks: pcw_mesa: what kind of dpll should I use for latencys that could aproach 500us?
04:38 PM skunkworks: I currently have it set to -100
04:40 PM skunkworks: well - having a realtime kernel running make it sound way better too... ;)
04:42 PM skunkworks: .00004 following error..
04:42 PM skunkworks: so far
04:42 PM skunkworks: by now it would have been into the .0005 or more
04:43 PM skunkworks: no triggers of long read times either..
04:45 PM skunkworks: I also am running 1750mhz trying to figure out the issue with it running like crap
04:45 PM skunkworks: up to .00006" following
04:45 PM skunkworks: peak
04:47 PM skunkworks: ugh - so stupid..
04:49 PM pcw_mesa: -100 usec should be pretty good (you can plot dpll-error-usecs to see haw bad the baseline drift is)
04:51 PM pcw_mesa: what happens if you have really bad latency for many cycles in a row is that the DPLL baseline shifts far enough that when it gets better it exceeds the pretrigger time
04:51 PM pcw_mesa: this has to be really bad to exceed 100 usec...
04:53 PM pcw_mesa: note that you need to add the pre-trigger time to the loops FF2 setting (which should = the time from position sample to velocity write)
04:58 PM skunkworks: pcw_mesa: I don't see dpll-error-us
04:59 PM skunkworks: I see phase-error-us
05:00 PM pcw_mesa: sorry, thats right
05:01 PM skunkworks: pcw_mesa: seems to flip between -.0014 and .0014
05:01 PM pcw_mesa: what does?
05:01 PM skunkworks: dpll.phase-erro-us
05:02 PM pcw_mesa: well thats not right...
05:03 PM skunkworks: heh - still not realtime enough.. the thing just thumped and I have a .0027" f-error
05:05 PM pcw_mesa: I get about 100/150 peak usec on my RPI3
05:05 PM skunkworks: wait
05:06 PM hazzy-m: sweet!
05:07 PM skunkworks: sorry - it was still showing backlash - duh even though it was labled error-us
05:08 PM skunkworks: yes - that is about what I get - spikes around 150us
05:10 PM pcw_mesa: you shoud be able to see the baseline wander, it not more than perhaps 30 usec on the RPI3
05:12 PM pcw_mesa: do you have the VPE hal file stuff? that should make occasional timed out packets harmless
05:12 PM skunkworks: base line seems to go between 0 and -30us
05:13 PM skunkworks: mostly
05:13 PM pcw_mesa: yeah so -100 should be fine (try starting,stopping videos)
05:15 PM pcw_mesa: You shoud be able to get rid of the thump with the PE handling and a reasonable timeout setting
05:15 PM pcw_mesa: (thump and FE)
05:16 PM pcw_mesa: remember the FE is bogus (stale data)
05:16 PM pcw_mesa: but a large bogus FE will result in a large (also bogus) correction
05:31 PM skunkworks: FE handling and reasonable time out?
05:33 PM skunkworks: where is the vpe hal file stuff
05:35 PM skunkworks: pcw_mesa: google isn't helping..
05:36 PM pcw_mesa: VPE = Vamp on Packet Error
05:36 PM skunkworks: it has run though the program again with following error staying under .00006"
05:37 PM skunkworks: oh - was that something jepler was playing with?
05:37 PM skunkworks: do you have an example?
05:37 PM pcw_mesa: Ill send you a hal file, basically what it does is feed the PID/Motion the commanded position as feedback if there is a packet error
05:37 PM skunkworks: https://forum.linuxcnc.org/27-driver-boards/35145-7i96-joint-following-error
05:38 PM pcw_mesa: yeah that should be it
05:39 PM pcw_mesa: its not perfect as it now will not notice if very many packets are lost
05:40 PM skunkworks: so - the dirivitive should not be used - right?
05:41 PM pcw_mesa: not after the PID loop was fixed so it actually used the pin...
05:42 PM pcw_mesa: (not sure why motions velocity is not right, maybe a 1 sample timing offset)
05:42 PM skunkworks: setp hm2_[HOSTMOT2](BOARD).0.packet-read-timeout 15
05:42 PM skunkworks: this is reasonable?
05:42 PM pcw_mesa: thats pretty aggressive, probably more like 60 for a RPI
05:42 PM skunkworks: ok
05:43 PM pcw_mesa: plot the read time to get an idea of the statitics
05:44 PM pcw_mesa: (15 means timeout in 150 usec on reads assumin 1 ms servo thread)
05:44 PM pcw_mesa: (its percent of the thread time)
05:49 PM skunkworks: I am running a 2ms thread
06:01 PM skunkworks: pcw_mesa: off and running... wasn't too painfull integrating my stuff in your hal.... Very similar anyway as I think I got the sample config from you originally
06:02 PM pcw_mesa: Missing from hm2_eth is a count of lost packets
06:05 PM skunkworks: I should try a full screen youtube video :)
06:07 PM pcw_mesa: I know thats impossible (or maybe just impossibly slow) on a RPI3
06:17 PM skunkworks: pcw_mesa: http://electronicsam.com/images/greenmachine/IMG_20191014_175826.jpg
06:24 PM skunkworks: following error is staying around .00006
06:32 PM skunkworks: hal is still pretty amazing...
06:32 PM skunkworks: running youtube..
06:33 PM skunkworks: yah.. don't
06:33 PM skunkworks: actually youtube runs ok.. Don't grab the browser and move it around... :)
06:35 PM skunkworks: actually - if you move any window around..
06:35 PM skunkworks: wow
07:02 PM linuxcnc-build: build #3556 of 4017.deb-wheezy-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-amd64/builds/3556 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:04 PM skunkworks: ok - if you enable the opengl driver - it is way better...
07:09 PM linuxcnc-build: build #3813 of 4009.deb-precise-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4009.deb-precise-rtai-i386/builds/3813 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:21 PM linuxcnc-build: build #4984 of 4008.deb-precise-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4008.deb-precise-amd64/builds/4984 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:23 PM linuxcnc-build: build #4984 of 4007.deb-precise-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4007.deb-precise-i386/builds/4984 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:39 PM linuxcnc-build: build #2113 of 4019.deb-jessie-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4019.deb-jessie-rtpreempt-i386/builds/2113 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:45 PM linuxcnc-build: build #3553 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/3553 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:47 PM linuxcnc-build: build #2111 of 4020.deb-jessie-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4020.deb-jessie-rtpreempt-amd64/builds/2111 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:52 PM linuxcnc-build: build #3243 of 4018.deb-wheezy-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4018.deb-wheezy-rtai-i386/builds/3243 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:55 PM linuxcnc-build: build #2222 of 4022.deb-jessie-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4022.deb-jessie-amd64/builds/2222 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:57 PM linuxcnc-build: build #2845 of 4015.deb-wheezy-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4015.deb-wheezy-rtpreempt-amd64/builds/2845 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
07:58 PM linuxcnc-build: build #2814 of 4014.deb-wheezy-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4014.deb-wheezy-rtpreempt-i386/builds/2814 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
08:01 PM linuxcnc-build: build #2225 of 4021.deb-jessie-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4021.deb-jessie-i386/builds/2225 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
08:03 PM linuxcnc-build: build #1009 of 4031.deb-stretch-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4031.deb-stretch-rtpreempt-amd64/builds/1009 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
08:04 PM linuxcnc-build: build #1007 of 4030.deb-stretch-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4030.deb-stretch-rtpreempt-i386/builds/1007 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
08:52 PM linuxcnc-build: build #3557 of 4017.deb-wheezy-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-amd64/builds/3557 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:00 PM linuxcnc-build: build #4985 of 4007.deb-precise-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4007.deb-precise-i386/builds/4985 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:00 PM linuxcnc-build: build #3814 of 4009.deb-precise-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4009.deb-precise-rtai-i386/builds/3814 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:07 PM linuxcnc-build: build #4985 of 4008.deb-precise-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4008.deb-precise-amd64/builds/4985 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:09 PM skunkworks: pcw_mesa: https://www.youtube.com/watch?v=qS7fy4qBC-g
09:09 PM linuxcnc-build: build #2846 of 4015.deb-wheezy-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4015.deb-wheezy-rtpreempt-amd64/builds/2846 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:24 PM linuxcnc-build: build #2815 of 4014.deb-wheezy-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4014.deb-wheezy-rtpreempt-i386/builds/2815 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:33 PM linuxcnc-build: build #1008 of 4030.deb-stretch-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4030.deb-stretch-rtpreempt-i386/builds/1008 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:34 PM linuxcnc-build: build #3554 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/3554 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:44 PM linuxcnc-build: build #3244 of 4018.deb-wheezy-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4018.deb-wheezy-rtai-i386/builds/3244 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:45 PM linuxcnc-build: build #2223 of 4022.deb-jessie-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4022.deb-jessie-amd64/builds/2223 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:47 PM linuxcnc-build: build #1010 of 4031.deb-stretch-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4031.deb-stretch-rtpreempt-amd64/builds/1010 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:47 PM linuxcnc-build: build #2226 of 4021.deb-jessie-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4021.deb-jessie-i386/builds/2226 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:49 PM linuxcnc-build: build #2112 of 4020.deb-jessie-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4020.deb-jessie-rtpreempt-amd64/builds/2112 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
09:49 PM linuxcnc-build: build #2114 of 4019.deb-jessie-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4019.deb-jessie-rtpreempt-i386/builds/2114 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>