#linuxcnc Logs

Nov 15 2023

01:21 AM Deejay: moin
06:39 AM JT-Cave1: rooster just crowed
07:58 AM JT-Cave1: Tom_L, added jog increments and fixed some jog items in the ini
07:59 AM JT-Cave1: didn't get to fixing the exponent for rotary axes so atm all number allow exponents
09:04 AM Getty2 is now known as Getty
09:32 AM JT-Cave1: does anyone want to fix the tab order in mesact? it's fun and easy to do
10:14 AM Gothia: JT-Cave1:
10:14 AM Gothia: i really wish i could help... by now i am making my firsts steps
10:15 AM Gothia: :)
10:27 AM -!- #linuxcnc mode set to +v by ChanServ
11:06 AM Tom_L: ...9 commits, that's like a new record!
11:10 AM -!- #linuxcnc mode set to +v by ChanServ
11:16 AM JT-Cave: nap time
11:16 AM Tom_L: soon as i finish lunch yeah
12:22 PM Unterhaus_ is now known as unterhausen
12:24 PM unterhausen: someone on fb complaining that his 5th axis slows down motion. My memory is failing me as far as what could be the problem
12:28 PM bjorkintosh: G-Code is as moody as vi.
12:28 PM bjorkintosh: maybe he's in a feed?
12:29 PM unterhausen: I vaguely remember people having problems with a rotary axis, I think he's got a spindle that rotates
12:45 PM Gothia: HI there everybody
12:49 PM Gothia: http://www.filefactory.com/file/3cc7pxlq98bk/Comparative%20Voltages%20parallel%20vs%207i92t.pdf
01:04 PM unterhausen: that site is too annoying for me to complete the d/l
01:05 PM Gothia: https://forum.linuxcnc.org/27-driver-boards/50530-7i92tm-pings-but-not-recognised-by-mesaflash?start=20#285533
01:05 PM Gothia: unterhausen: sorry...
01:05 PM Gothia: just posted the case int he forum
01:06 PM unterhausen: okey
01:06 PM Gothia: the question to me now is if the behaiviour of the mesa card requires a reset
01:06 PM Gothia: or if i set up any jumper or any pncconf uncorrectly
01:07 PM unterhausen: I don't think it requires a reset. Probably just not driving the i/o electronics properly
01:08 PM Gothia: then i should only focus to change the gpio on the pncconf?
01:09 PM Gothia: i initially though there was any issue related to a floating signal, due to a wrong or missing gnd connection in the mesa card...
01:10 PM Gothia: but I think this voltage situation would discart that. Am I right?
01:10 PM Gothia: does it sound logical ?
01:16 PM unterhausen: I am fairly sure it's not a configuration problem. People usually have problems with a floating ground, so they aren't actually setting up the required voltages
01:41 PM pcw-home: JT-Cave1 ping
01:46 PM unterhausen: JT reported it was nap time
01:47 PM unterhausen: pcw-home do you have an app note about driving I/O?
01:52 PM CaptHindsight: " someone on fb complaining that his 5th axis slows down motion" under what circumstances?
01:52 PM Tom_L: unterhausen, that would depend on the IO and what is driving it
01:52 PM Tom_L: CaptHindsight, probably huge numbers of tiny linear moves
01:53 PM Tom_L: just a guess
01:53 PM CaptHindsight: you hope it's logical
01:53 PM unterhausen: I suspect it has to do with everything slowing down to match the rotary.
01:53 PM Tom_L: could be that
01:54 PM unterhausen: maybe the rotary has tolerances too small?
01:54 PM Tom_L: until i had mine configured right it was slow on some moves
01:54 PM CaptHindsight: do you know the cost of the 5th axis option for the VF2's?
01:54 PM unterhausen: you said it was 50k
01:54 PM CaptHindsight: what I heard, I haven't verified
01:55 PM CaptHindsight: few $k for the 4th, but the 5th is supposedly much much more
01:56 PM unterhausen: I lost track of where updating the new trajectory planner for more axis got to
01:57 PM CaptHindsight: somewhere in the path pilot source?
01:57 PM unterhausen: I saw that someone had that
01:57 PM CaptHindsight: sounds a a forum question
01:57 PM unterhausen: but it's been 2 years since I heard it was going to be ported to lcnc
01:57 PM CaptHindsight: that way the answer will stick
01:58 PM unterhausen: there is a forum subject about it, I would have to go look
01:58 PM unterhausen: my watch decided to update its firmware without asking
01:59 PM CaptHindsight: https://forum.linuxcnc.org/38-general-linuxcnc-questions/50268-trajectory-planner-using-ruckig-lib a new thread about a TP
02:02 PM CaptHindsight: https://ruckig.com/
02:04 PM CaptHindsight: now i want to read the drama that they deleted :(
02:06 PM CaptHindsight: a closed source motion generator
02:06 PM CaptHindsight: https://roboticsconference.org/2021/program/papers/015/index.html
02:07 PM CaptHindsight: https://www.youtube.com/watch?v=b6y_3JvGFJo&t=6s RSS 2021, Spotlight Talk 24: Jerk-limited Real-time Trajectory Generation with Arbitrary Target...
02:08 PM CaptHindsight: sounds German so it must be better
02:11 PM roycroft: ruckig is german for "jerky"
02:13 PM roycroft: so it must operate like this: https://www.youtube.com/watch?v=HQu_NLRvULM
02:15 PM CaptHindsight: NOT FUNNY!
02:16 PM roycroft: if not, it's only because germans have no sense of humor
02:17 PM CaptHindsight: we used to tease our family there about that
02:19 PM CaptHindsight: I'd like to hear an uhmerican try to pronounce Karlsruhe
02:22 PM CaptHindsight: https://github.com/pantor/ruckig
02:22 PM CaptHindsight: the community version
02:28 PM * roycroft can pronounce karlsruhe like a native german, but not one from karlsruhe
02:29 PM roycroft: i lerned my german pronunciation from a northerner, from hannover
02:29 PM roycroft: and that's stuck with me ever since
06:14 PM JT-Cave: pcw-home, just got home
06:18 PM * JT-Cave calls it a night
07:39 PM pcw-home: JT-Cave: I don't know if you have this fixed in MesaCT: https://github.com/LinuxCNC/linuxcnc/issues/2748
08:53 PM Tom_L: his spindle stepgen isn't completed yet
08:57 PM Tom_L: his normal one maps to spindle on though
08:57 PM Tom_L: maybe it should be moved to machine on too
09:24 PM Tom_L: pcw-home, what signal would the machine-is-on map to then? i'm not finding a specific pin for that right off
09:24 PM Tom_L: looking in motion. and iocontrol
09:25 PM Tom_L: iocontrol being userspace and motion realtime
09:25 PM pcw-home: At least in pncconf its: net machine-is-on halui.machine.is-on
09:26 PM Tom_L: should come somewhere in realtime space
09:26 PM Tom_L: ahh, halui
09:26 PM Tom_L: :)
09:26 PM pcw-home: Yes probabal a motion pin
09:26 PM Tom_L: halui is userspace
09:27 PM Tom_L: halui.machine.is-on bit out
09:27 PM Tom_L: there it is
09:27 PM Tom_L: but that's uspace
09:28 PM pcw-home: maybe motion.motion-enabled
09:28 PM pcw-home: so its instantly cleared by a fault
09:29 PM Tom_L: so far jt's maps to spindle.on on the other one not using stepgen
09:29 PM Tom_L: and his stepgen code for spindle isn't done yet
09:30 PM pcw-home: thats more OK, With a a stepgen and a position mode servo, setting the spindle speed to 0 is rather shocking...
09:31 PM Tom_L: the only time i set S to 0 is if i'm locating stock with a dowel pin
09:31 PM Tom_L: using that as a tool
09:32 PM pcw-home: also just turning off the spindle
09:32 PM pcw-home: (m5)
09:32 PM Tom_L: yeah
09:34 PM pcw-home: There's a forum thread with this and it looks like every stepgen spindle +mesa user trips over this
09:35 PM Tom_L: iirc i'm using pwmgen? with the analog out
09:35 PM Tom_L: does it need to be changed as well?
09:35 PM Tom_L: mine maps to spindle.on
09:36 PM pcw-home: In that case, it likely the VFD is programmed for reasonable accel/decel values
09:36 PM Tom_L: it is
09:36 PM Tom_L: it was actually too mushy when i got it
09:37 PM pcw-home: the stepgen issue comes up because of people using servos for the spindle
09:37 PM Tom_L: ahh
09:39 PM Tom_L: motion.motion-enabled doesn't really have a description in the docs... what i'm wondering is a case where spindle is turned on before motion starts (to get it up to speed)
09:39 PM Tom_L: before plunging into material
09:40 PM Tom_L: if the accel ramp is too high you could start cutting before reaching speed
09:40 PM Tom_L: that's why i set mine on ahead of moves
09:42 PM Tom_L: the docs don't really indicate what triggers motion.motion-enabled true
09:42 PM pcw-home: I think its just true when LinuxCNC is ready , not really related to motion
09:43 PM Tom_L: i haven't used it but would be easy enough to put in a watch window
09:44 PM pcw-home: Added card listing to mesaflash (if you give it a bad card name it prints them all)
09:44 PM Tom_L: the machine-is-on or such i was thinking of turns out to be an input used by estop
09:44 PM Tom_L: nice
09:44 PM Tom_L: i remember you said you were gonna add that
09:45 PM pcw-home: there may be a better pin, spindle on is fine for PWM though
09:45 PM Tom_L: i was just trying to find an alternative to spindle.on
09:46 PM pcw-home: Yeah, people would use --device 7i92tm and mesaflash would respond "no 7i92tm found"
09:46 PM pcw-home: and no way to see the supported card names
09:47 PM Tom_L: i recall that on the 7i80 i think when you added HD16 or whatever it was
09:47 PM Tom_L: i wound up looking in the code :)
09:48 PM pcw-home: well now just use --device foonman...
09:51 PM pcw-home: also need to add sserialb support (mesaflash can't see it currently and even the LinuxCNC hm2 driver doesn't print the version )
09:51 PM Tom_L: 3.5.4
09:53 PM Tom_L: you might check the 7i80.. i think it requires HD16 or such
09:54 PM Tom_L: you list 7i80 & 7i80HDT
09:57 PM pcw-home: Yeah 7i80 should just work
09:58 PM -!- #linuxcnc mode set to +v by ChanServ
09:59 PM Tom_dev: eth_boards.c: } else if (strncmp(buff, "7I80HD-16", 16) == 0)
10:00 PM Tom_L: i had to add that after ver 3.4.3
10:01 PM Tom_L: prior to that it was just 7i80
10:01 PM pcw-home: That's the part that compares the actual card name from the card to setup the structure
10:02 PM Tom_L: i just pulled that example from code... not sure what it was doing there but my notes told me to add HD-16
10:02 PM Tom_L: on the cmd line
10:03 PM Tom_L: don't have one nearby or i'd try it
10:03 PM pcw-home: I dont think that is true anymore
10:03 PM Tom_L: one sec..
10:05 PM Tom_dev: mesaflash --device 7i80 --addr 10.10.10.10 readhmid
10:05 PM Tom_dev: No 7I80 board found
10:05 PM Tom_dev: mesaflash --device 7i80HD-16 --addr 10.10.10.10 readhmid
10:05 PM Tom_dev: ERROR: Unsupported device 7I80HD-16
10:06 PM Tom_L: lemme try on the rpi on the older ver
10:09 PM Tom_L: funny, jt's tool finds it but mesaflash doesn't
10:10 PM * Tom_L checks the cmd line...
10:12 PM pcw-home: try 3.5.5
10:12 PM Tom_dev: something is fishy
10:13 PM Tom_dev: it is found using his tool but the cmd line doesn't find it
10:13 PM Tom_dev: mesaflash --device 7i80 --addr 10.10.10.10 --readhmid
10:13 PM Tom_dev: should work?
10:14 PM pcw-home: 7I80HD-16 should work with 3.5.5
10:15 PM pcw-home: I removed 7I80 and added all the variants
10:16 PM Tom_L: that works
10:17 PM pcw-home: that makes more sense since the long names are the names you can extract with --ether
10:17 PM Tom_dev: mesaflash --device 7i80 --addr 10.10.10.10 --readhmid
10:17 PM Tom_dev: doesn't work but doesn't give a list either
10:17 PM Tom_dev: mesaflash --device 7i80
10:17 PM Tom_dev: You must supply an IP address for Ethernet cards
10:18 PM Tom_dev: mesaflash --device foo does
10:20 PM Tom_dev: mesaflash --device 7i
10:20 PM Tom_dev: You must supply an IP address for Ethernet cards
10:20 PM Tom_dev: but
10:20 PM Tom_dev: mesaflash --device 7ixx
10:20 PM Tom_dev: gives a list
10:21 PM pcw-home: I'll take a look at that, probably using the entered string length for the compare but in thsi case it should use the lenth of the list strings
10:21 PM Tom_L: ok, time for bed
10:21 PM Tom_L: thanks
10:22 PM pcw-home: Thank you!
10:25 PM Tom_L: couldn't be entered string length because 'foo' works
10:33 PM pcw-home: I think I've fixed it, a previous name fix was wrong
10:33 PM pcw-home: didnt bump the version but I think its fixed
10:34 PM pcw-home: so --device 7i80 now complains
10:35 PM Tom_L: i'll check real quick then i'm out
10:37 PM Tom_L: everything i threw at it works ok now
10:37 PM pcw-home: thanks for the help
10:37 PM Tom_L: glad to do it
10:38 PM Tom_L: what is a 'LITEHM2' device?
10:40 PM pcw-home: Thats "sensilles?" port of hm2 to Litex on some inexpensive Chinese FPGA board used for signage
10:40 PM Tom_L: ahh yeah i recall that
10:41 PM pcw-home: There are a bunch of cheap Ethernet FPGA cards ised for the giant LED billboards
10:41 PM Tom_L: repurpose
10:42 PM Tom_L: ok, later...
10:42 PM pcw-home: Yeah, Thanks again, I'm getting sleepy myself
10:42 PM pcw-home: bbl