#linuxcnc Logs
May 21 2024
#linuxcnc Calendar
12:59 AM Deejay: moin
03:57 AM lcnc-relay: <travis_farmer> Morning
04:08 AM Tom_L: solarwind, should have gotten a General 78
06:17 AM -!- #linuxcnc mode set to +v by ChanServ
06:28 AM JT-Mobile: ok so PyOpenGL is part of debian so I'm ok using that
09:18 AM lcnc-relay: <skunkworks8841> For gcode preview?
09:36 AM solarwind: Tom_L any advantages?
09:36 AM solarwind: JT-Mobile doesn't matter if it's in the debian repository or not, you can just install via pip
09:38 AM solarwind: roycroft I know, that's not the part that's unreliable. If you look at the animation on the wikipedia page, it's the bottom spring that's tapered, the one putting force on the blue middle pin
09:39 AM solarwind: part of its job is to bias the pin off-axis so it naturally prefers to bite onto the ledge of the hammer
11:14 AM Rab_: I've never had good luck with automatic center punches. I have a smaller Starrett 18AA and it's OK. Mostly they fail to reset and I have to thwack them on a benchtop.
11:17 AM lcnc-relay: <skunkworks8841> I have never had one of those style center punches..
11:18 AM Rab_: Tom_L, Central Texas has two overlapping toll systems, TxTag and KTAG (administered by KTA). KTAG is vastly superior. TxTag is basically incompetent, bills spurious charges with no redress which quickly rack up exorbitant late fees.
11:20 AM Rab_: I suspect careful application of a hammer punch will give better accuracy than an automatic punch.
11:21 AM Rab_: Generally I find the automatic punch hops off, and I re-punch at an angle to get the divot centered.
11:21 AM Rab_: Always wanted to try one of those optical center punches.
11:26 AM Rab_: Center punches aren't appropriate for accurate setups anyway. For those cases I'll use the mill, edge-find then dial in the offsets and use a spotting or stub-length drill to get started.
11:45 AM faheus8 is now known as faheus
11:55 AM -!- #linuxcnc mode set to +v by ChanServ
12:14 PM CaptHindsight: center punches might be accurate enough if you have robot hands
01:07 PM Unterhaus_ is now known as Unterhausen
01:36 PM XXCoder: https://youtu.be/OGOb_rZ8dWU
01:36 PM XXCoder: pretty cool
01:36 PM XXCoder: step up from pocketnc
01:44 PM bjorkintosh: $80,000 base price.
01:44 PM * Tom_L orders half dozen
01:44 PM XXCoder: order one for me
01:48 PM Tom_L: bjorkintosh, so $90-100k by the time you get a fair amount of tooling
01:48 PM bjorkintosh: approximately. with probe.
01:57 PM Rab_: "Rolling through your man door"
02:04 PM Rab_: Safety Birkenstocks detected: https://imgur.com/a/F44FEHj
02:14 PM bjorkintosh: heh. they're also using safety squints.
02:15 PM JT-Mobile: takes 10 to buy a tractor
02:23 PM XXCoder: think we spotted roy. (right one)
02:23 PM XXCoder: kidding
02:25 PM XXCoder: lol couple days I considered removing engineering guy channel because its been quiet for a while, and suddenly 2 videos today
02:28 PM XXCoder: https://youtu.be/E-F2QQuZZGk prety good (engineering guy)
03:23 PM JT-Mobile: Tom_L, https://www.youtube.com/watch?v=hPhcCh8aBx4
03:55 PM Tom_L: JT-Mobile, aren't you gettin fancy
03:56 PM JT-Mobile: lol, just blundering about really
03:56 PM Tom_L: did that site yesterday help any?
03:56 PM JT-Mobile: yes it did
03:57 PM JT-Mobile: there's still way to many questions and not enough answers like converting the gcode to relative position
03:58 PM JT-Mobile: the gcode module should do that but I can't find any info on it at all
03:58 PM Tom_L: that guy seemed to have a site with tutorials maybe
03:58 PM JT-Mobile: just the one I found
03:58 PM Tom_L: you mean what turboss referred to?
03:58 PM JT-Mobile: import gcode then gcode.parse() but I don't know the arguments lol
03:59 PM JT-Mobile: I forgot what turboss referred to
03:59 PM Tom_L: the cannon?
03:59 PM Tom_L: i dunno without finding it again
04:02 PM JT-Mobile: there seems to be two parts of the puzzle, opengl and parsing the gcode file
04:02 PM JT-Mobile: I've managed to do a little of both
04:03 PM JT-Mobile: Help on built-in function parse in module gcode:
04:03 PM JT-Mobile: parse(...)
04:03 PM JT-Mobile: Parse a G-Code file
04:03 PM JT-Mobile: help doesn't tell me much
04:04 PM JT-Mobile: down to 3 hens now
04:05 PM JT-Mobile: Spaetzle bit the dust
04:05 PM Tom_L: that's sad
04:05 PM lcnc-relay: <travis_farmer> what a colossal screw-up... bought tires for my bike... thought sure they were 29in as it is a XL frame... the bike is 26in... so tomorrow i gotta take them back and swap them...
04:05 PM JT-Mobile: yup she was doing ok then just went downhil fast
04:06 PM JT-Mobile: frame size doesn't matter to wheel size
04:06 PM lcnc-relay: <travis_farmer> apparently...
04:09 PM Tom_L: i'm not finding the links he posted yet
04:10 PM Tom_L: http://paste.debian.net/1317688/
04:10 PM Tom_L: there
04:12 PM Tom_L: i'll post those so they're in the log.. the log missed them that day
04:13 PM JT-Mobile: I'll have to check them out later fixing to head to the graduation
04:13 PM Tom_L: have fun
04:13 PM Tom_L: as graduations go...
04:14 PM Tom_L: <turboss> JT-Cave, to plot the gcode you don't need to parse the file, just subclass canon and implement the GL plot line there
04:14 PM Tom_L: <turboss> https://github.com/kcjengr/qtpyvcp/blob/main/src/qtpyvcp/plugins/gcode_properties.py#L408
04:14 PM Tom_L: <turboss> https://github.com/kcjengr/qtpyvcp/blob/main/src/qtpyvcp/widgets/display_widgets/vtk_backplot/base_canon.py
04:14 PM Tom_L: <TurBoss> https://github.com/kcjengr/qtpyvcp/blob/main/src/qtpyvcp/plugins/gcode_properties.py#L508
04:15 PM Tom_L: at least theyre logged now
04:16 PM lcnc-relay: <travis_farmer> logged twice 😉
05:21 PM Unterhausen: Travis, I could have told you they were 26", but I did tell you not to buy them
06:10 PM lcnc-relay: <travis_farmer> Huh, when did you tell me not to buy them?
06:17 PM lcnc-relay: <travis_farmer> Unterhausen, what are you talking about?
09:15 PM JT-Mobile: graduation over with...
09:16 PM XXCoder: yay. one of your kids or?
09:16 PM JT-Mobile: granddaughter
09:16 PM XXCoder: nice
09:18 PM JT-Mobile: <turboss> JT-Cave, to plot the gcode you don't need to parse the file, just subclass canon and implement the GL plot line there
09:18 PM JT-Mobile: I wish I knew what that meant
09:22 PM XXCoder: I guess theres some library object that can parse them for you
09:22 PM XXCoder: otherwise dunno
09:23 PM lcnc-relay: <JT> https://jauriarts.org/_matrix/media/v3/download/matrix.org/kQvTVKMxAlevVzTleDVZTPsP/20240521_194602.jpg
09:31 PM XXCoder: nice picture
09:34 PM Leeloo_ is now known as Leeloo
09:34 PM JT-Mobile: mom, daughter and son
09:35 PM Leeloo: Hmm, "sudo ./mesaflash --device 5i25 --reload" and get "ERROR: Active firmware too old to support --reload" - I just wrote 5i25_7i76_7i85.bin onto it... from 5i25.zip
09:35 PM Leeloo: Is there some other firmware I am to use?
09:36 PM JT-Mobile: you might use a fresh copy of the firmware or use Mesa CT to flash
09:37 PM Leeloo: What is Mesa CT?
09:37 PM JT-Mobile: my Mesa configuration tool
09:37 PM JT-Mobile: https://gnipsel.com/linuxcnc/mesact/index.html
09:42 PM Leeloo: Hmm, those links for .deb packages give me 404
09:43 PM JT-Mobile: on the install page?
09:44 PM Leeloo: Yes
09:44 PM JT-Mobile: one moment
09:44 PM JT-Mobile: https://github.com/jethornton/mesact/releases
09:44 PM Tom_L: Leeloo, how old is mesaflash?
09:45 PM Tom_L: wow jt's up way past his bedtime!
09:45 PM JT-Mobile: and he is very tired
09:45 PM Leeloo: I jsut downloaded it - --version returns 3.5.8
09:45 PM Leeloo: JT-Mobile, Thank you
09:46 PM Tom_L: https://github.com/linuxCNC/mesaflash
09:46 PM JT-Mobile: no problem
09:46 PM JT-Mobile: crap I'm not setup on this pc to clone
09:46 PM Tom_L: although you won't need the lastest mesaflash for those cards
09:46 PM Leeloo: I think that's where I got it from.
09:47 PM Leeloo: I restarted PC and my 7i76 is seen, but not 7i85.
09:47 PM Tom_L: the just sudo make install after the dependencies are met
09:47 PM Tom_L: for mesaflash that is
09:47 PM Tom_L: 7i85 a sserial card?
09:47 PM JT-Mobile: it said the firmware is too old...
09:47 PM JT-Mobile: no
09:47 PM Tom_L: oh
09:47 PM JT-Mobile: db25
09:48 PM Leeloo: It's an encoder rs-485 card
09:48 PM JT-Mobile: 7i85 and 7i85s
09:48 PM JT-Mobile: and maybe I'm very tired
09:48 PM JT-Mobile: https://mesaus.com/product/7i85/
09:49 PM JT-Mobile: Tom_L,
09:50 PM JT-Mobile: remind me when I get home to copy my ssh keys to this laptop...
09:50 PM Tom_L: :)
09:50 PM Tom_L: i keep them on a thumbdrive
09:51 PM Tom_L: you still have to chmod the files & dir
09:51 PM JT-Mobile: I keep them on the NAS but it's a bit out of reach here in Tomball
09:51 PM Leeloo: It seems that it writes the new firmware fine: https://imgur.com/EyVfMSR
09:52 PM Tom_L: did you specify an address?
09:52 PM Tom_L: --addr
09:52 PM JT-Mobile: 5i25 is pci
09:52 PM Tom_L: ok
09:53 PM JT-Mobile: it requires sudo
09:53 PM Leeloo: Yes, I used it afterwards.
09:55 PM Tom_L: JT-Mobile, you headin back in the morning?
09:56 PM JT-Mobile: no, heading back Thursday and stopping in Texarkana for the night
09:56 PM JT-Mobile: be home Friday evening
09:56 PM Tom_L: k
09:56 PM Tom_L: god willin and the creek don't rise
09:57 PM JT-Mobile: yup
09:57 PM JT-Mobile: I'm betting on the creek not rising
09:57 PM Tom_L: i wonder how hard galveston was hit
09:58 PM JT-Mobile: I think mostly Houston and some a bit north of there
09:58 PM JT-Mobile: I don't know why my link to the deb in mesact is off...
09:59 PM * JT-Mobile calls it a night
09:59 PM Tom_L: later
10:01 PM Tom_L: Leeloo, you can get the mesact deb here: https://github.com/jethornton/mesact/releases/tag/2.1.3
10:02 PM Leeloo: Yes, that downloads
10:02 PM Tom_L: he bumped versions and forgot to update his webpage
10:03 PM Leeloo: When I do this command: https://imgur.com/OMRfTBF
10:03 PM Leeloo: I should see 7i85 on port 1, right?
10:04 PM Tom_L: i'm not sure it reports sserial. it may on newer mesaflash versions
10:05 PM Tom_L: i haven't looked at addressing sserial cards that much so i'm not too familiar with the command for that
10:05 PM Tom_L: see what the help says
10:06 PM Tom_L: it may not list the sserial pins
10:10 PM Leeloo: I changed number of encoders form 1 to 5: loadrt hm2_pci config="num_encoders=5 num_pwmgens=0 num_stepgens=5 sserial_port_0=00xxxxxx"
10:10 PM Leeloo: And they are listed by hal show, but I have no idea if that 7i85 is there or not :)
10:11 PM Tom_L: if hal show sees them they should be there
10:12 PM Leeloo: I was kind of thinking that 7i85 will show in the tree like 7i76 - but it's not there... is that supposed to be this way?
10:12 PM Tom_L: like i said, on sserial i'm not sure they show
10:12 PM Tom_L: it's been a while since i messed with a sserial card and i don't really remember
10:14 PM Tom_L: bedtime here
10:38 PM Leeloo: Interesting - went with command line halrun: halrun / loadrt hostmot2 / loadrt hm2_pci / show pin and indeed, I see 5 encoders.
10:38 PM Leeloo: They are all under 5i25.
11:53 PM pcw-home: Yes, the 7I85 is a passive card, so there are no "7i85s" in pin names
11:55 PM pcw-home: It's just an interface card for FPGA card hardware
11:57 PM pcw-home: "sudo mesaflash --device 5i25 --dbname1 7i76 --dbname2 7i85 --readhmid" will show the pinout