#linuxcnc Logs

Apr 02 2023

01:31 AM Deejay: moin
05:19 AM JT-Cave: morning
05:20 AM travis_farmer[m]: G'Morning ☕️
05:51 AM Bing[m]: morning fellow humans
06:10 AM Tom_L: morning
06:36 AM Sariah[m]: Marning
09:50 AM JT-Woodshop: yo
09:53 AM TurBoss: hao 🖖
10:33 AM Sariah[m]: added my second axis in, and finally ran my first EMC2 AXIS Gcode 😛
12:11 PM Tom_L: 70°F & sunny
12:12 PM Tom_L: oh and don't forget the wind..
12:33 PM roycroft: we have three more days of winter
12:33 PM roycroft: including a chance of snow
12:34 PM roycroft: but wednesday it will turn into spring
01:14 PM JT[m]: Quiet in here
01:14 PM skunkworks[m]: non-circular turning tool geometry is a pain
01:14 PM skunkworks[m]: JT: thought the same thing
01:14 PM skunkworks[m]: I thought I had enough relief - but no...
01:16 PM Tom_L: well, you don't pick the easiest projects...
01:17 PM skunkworks[m]: True.
02:42 PM CaptHindsight[m]: if anyone is looking for a manual Bridgeport J-head or a manual 12x32 Clousing lathe in the Chicago area, ping me
03:03 PM spray[m]: hello,
03:03 PM spray[m]: does anybody have a working hal & ini config for open loop motors? I'm still getting Joint 0 following error. I'm using default configs from Scott's Remora NVEM github.
03:03 PM spray[m]: I have read manual regarding following error, but Im still lost
03:06 PM rmu: spray[m]: what do you mean with open loop motors?
03:07 PM spray[m]: like no feedback (steppers for example)
03:07 PM spray[m]: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Following_Error
03:08 PM skunkworks[m]: that is pretty old info..
03:09 PM rmu: spray[m]: just make a config with stepconf
03:10 PM skunkworks[m]: I don't know how the nvem is setup for steppers - but 'usually' it is caused by the stepgen in the hardware not having enough acceleration headroom compared to the axis acceleration setting.
03:10 PM skunkworks[m]: (if you are not getting any realtime errors)
03:14 PM rmu: http://linuxcnc.org/docs/html/config/stepconf.html#sub:finding-maximum-acceleration
03:14 PM spray[m]: the hardware is capable of 460kHz
03:15 PM CaptHindsight[m]: spray: "Scott's Remora NVEM github" ask on the LCNC forums for this project
03:15 PM CaptHindsight[m]: spray: nobody on IRC is using Remora, so it will just be guessing
03:16 PM CaptHindsight[m]: the are pretty responsive on the forums for the Remora project
03:17 PM CaptHindsight[m]: spray: https://forum.linuxcnc.org/18-computer/44828-remora-ethernet-nvem-cnc-board
03:17 PM spray[m]: The hardware is working fine I think, so I was thinking its the general linuxcnc config now
03:18 PM CaptHindsight[m]: Remora has its own quirks
03:18 PM CaptHindsight[m]: be sure to be using whatever version they say is stable
03:27 PM CaptHindsight[m]: spray: somebody, maybe you has already posted the same question there, so just hold tight for responses
03:27 PM spray[m]: my inquiry in that thread is the last
03:27 PM CaptHindsight[m]: yeah it's 6:19 AM at Scotts home now
03:46 PM pcw--home: You could also widen the following error limits, at least to gather more test data
03:47 PM spray[m]: MAX_ACCELERATION = 2000.0
03:47 PM spray[m]: STEPGEN_MAXACCEL = 2200.0
03:47 PM spray[m]: FERROR = 1.0
03:47 PM spray[m]: MIN_FERROR = 0.5
03:47 PM rmu: spray[m]: what are you doing when the following error occurs?
03:48 PM spray[m]: just jogging left right
03:50 PM rmu: open halscope and watch signals j0pos-cmd and j0pos-fb
03:53 PM rmu: are you sure your networking setup is working and stable?
03:55 PM spray[m]: yes, pings can go for hours
03:57 PM rmu: looking at the code it doesn't look like remora hal component will trigger estop in case of excessive packet loss
04:00 PM rmu: halscope, that will reveal network problems, that should show as a kink / horizontal line in position feedback
04:04 PM pcw--home: plotting commanded velocity may help diagnosis (for insufficient velocity/acceleration overhead issues)
04:36 PM skunkworks[m]: does it also use velocity headroom too?
04:39 PM pcw--home: I would guess so
08:50 PM iggw[m]: How should I set up my spindle servo motor on the 7i96s in step/direction mode? It is not working for me so far
09:38 PM CaptHindsight[m]: the Clausing was sold sight unseen in <24 hrs
10:17 PM pcw--home: iggw[m]: normally you would set the spindle stepgen into velocity mode and feed the stepgen velocity command from spindle.0.speed-cmd-rps
10:17 PM pcw--home: (and step scale to steps per turn)
10:37 PM pcw--home: also you would probably drive the stepgen enable from motion.motion-enabled (so spindle off ramp-down to 0 RPM is done at a controlled rate)
10:54 PM XXCoder: mill enclosure https://youtu.be/iQk7Hap6MgQ
11:30 PM XXCoder: design reminds me of hurco
11:37 PM ZincBoy[CAON][m]: Nice but no seals or overlap on the doors or panels. It will leak coolant from everywhere. I have built a few enclosures now and if there is a way for coolant to get through it will. And with that nice white paint, it will show up beautifully 🙂
11:37 PM XXCoder: yeah if I made it, I would add rubber lips to many spots in back
11:38 PM XXCoder: bottom edge would have rubber barely overlap so you can open and close it, but coolant spray will miss the edge
11:39 PM XXCoder: or alternatively, thin steel bent in such way its not on way of closing or opening but coolant on walls will just flow down without touching edge
11:39 PM ZincBoy[CAON][m]: You really need to overlap the panels/joints in such a way that the coolant drips/flows back into the enclosure.
11:40 PM ZincBoy[CAON][m]: I found that seals are not that reliable unless you have a very accurate enclosure (i.e. not welded from a bunch of parts).
11:41 PM XXCoder: https://imgur.com/aOsHCSz.png
11:42 PM XXCoder: its design I saw on hurco and some other machines. works pretty effective
11:42 PM XXCoder: the top bolted on part needs sealant but honestly even if its not perfect what little left that gets in would would drop down and not go into outside
11:43 PM XXCoder: of course that drawing is not perfectly accurate but its enough to show concept
11:44 PM XXCoder: lol. at end he shows machine running
11:44 PM XXCoder: and I feel like its hurco or haas lol
11:50 PM XXCoder: zinc one im not sure if his enclosure doors edges is low enough
11:50 PM XXCoder: ideally its well below the line of table so coolant dont go stright though it