#linuxcnc Logs

Jun 14 2022

#linuxcnc Calendar

01:06 AM Deejay: moin
02:59 AM travis_farmer853: G'Morning ☕
03:56 AM pere: anyone here know how to talk to a Sanyo Denki RT 4XC 11?
05:06 AM Tom_L: morning
05:23 AM JT-Cave: morning
09:21 AM silopolis[m]: Hello friends
09:23 AM silopolis[m]: We're having a discussion in FreeCAD/Path channel about Dynamic Work Offsets and Tool Center Point Control, trying to get our head around these and also wondering what's their status in LinuxCNC ?
09:23 AM silopolis[m]: Any of you can shed some light on this for us please ?
10:15 AM ZincBoy[CAON]066: Linuxcnc does not have DWO or TCPC features. It also only has lookahead trajectory planning for 3 axis work. As soon as you add a rotary axis it falls back to the older single block lookahead. The 4 and 5 axis capabilities of linuxcnc are fairly basic. It works fine, just does not have the advanced features of a modern control.
10:16 AM JT-Shop: http://linuxcnc.org/docs/stable/html/lathe/lathe-user.html#_tool_path
10:16 AM JT-Shop: control point only makes sense for a lathe
10:17 AM ZincBoy[CAON]066: Or for 5 axis work.
10:47 AM JT-Shop: that's above my pay grade
11:51 AM roycroft: i checked out my cnc router project files, but sw is almost unusable with them on the old, slow display
11:51 AM roycroft: i guess i'll have to wait for my new one to come back from repair before i can work on the project again
11:51 AM roycroft: i hope they don't take too long to repair/replace it
11:51 AM JT-Shop: got the drawers done?
11:51 AM roycroft: no
11:52 AM roycroft: but they should be done by friday
11:52 AM JT-Shop: E-Bike build starts tomorrow
11:52 AM roycroft: i have 12 more to glue up and install, and i'll get 4 of them glued up tonight
11:52 AM roycroft: it takes a day after glue-up before they're ready to install, so i'll glue 4 up tonight, tomorrow night, and thursday night
11:53 AM roycroft: i still have the 8 systainer trays to make, but i can't start on that until i get the drawer fronts milled
11:53 AM * Tom_L wonders if roycroft has stock in titebond
11:54 AM roycroft: that gallon jug of it i had at the start isn't empty yet, and it was mostly empty when i started
11:54 AM roycroft: i've gone through a gallon of shellac, though
11:54 AM roycroft: and over a gallon of alcohol
11:55 AM Tom_L: industrial or consumer grade? :)
11:55 AM roycroft: denatured
11:55 AM JT-Shop: good Menards has 35 oscillating pedestal fans in stock in case my oscillator fix does not work
11:56 AM roycroft: the finish is sloppy enough without my consuming refreshing beverages while applying it
11:56 AM roycroft: actually, the insides of the drawers are finished nicely
11:56 AM JT-Shop: I think I'll get 2 if my glue fails
11:56 AM roycroft: i'm just not caring too much about the outsides, since they won't be seen, and since these are shop cabinets
11:57 AM roycroft: besides, it's shellac - if i ever decide that i do care about the finish, it's easy enough to fix it
11:57 AM * roycroft doubts he'll ever care
12:20 PM JT-Shop: the oscillating fan seems to work so I'll only get one for a spare
12:44 PM silopolis[m]: <ZincBoy[CAON]066> "Linuxcnc does not have DWO or..." <- Does that mean CAM system must know machine geometry/kinematics exactly ?
01:00 PM ZincBoy[CAON]066: Not exactly. The CAM system needs to have the same work offset and relationships between the axis (both linear and rotary). This may or may not be an exact kinematic model of the machine. Most CAM systems will use a 3d model of the machine with all axis represented for collision checking.
01:02 PM roycroft: oh, the insanity
01:02 PM roycroft: i just got more spam - "create stunning furniture with resin"
01:02 PM roycroft: for the love of god, will this resin fad never end?
01:03 PM ZincBoy[CAON]066: What, it doesn't resinate with you?
01:03 PM pcw--home: lol
01:04 PM XXCoder: lol
01:04 PM XXCoder: it looks like it'll be everywhere
01:04 PM roycroft: it's tied with our 45th president and the covid pandemic as the worst things that have happened in this century
01:05 PM * JT-Shop continues to search for his desktop
01:06 PM unterhaus: I'm pretty sure the boom in epoxy furniture has died down some, along with slab furniture
01:06 PM roycroft: it hasn't, sadly
01:06 PM roycroft: it should have, but it hasn't
01:07 PM roycroft: the economic crash may cause it to die down somewhat
01:07 PM roycroft: even rich people won't be able to afford $20k ugly tables
01:08 PM sliptonicBradCol: I installed Debian 12/bookworm on two computers and then installed lcnc using APT. One is an i386 based machine. The other is AMD. Launching lcnc on the AMD machine I'm getting an error about a missing simple. libTifftcl, if memory serves me right. I'm not in front of that machine at the moment. I assume there's a missing package. Anyone know what I need to install?
01:08 PM sliptonicBradCol: s/simple/symbol/
01:10 PM roycroft: that issue was recently mentioned on emc-developers, but not resolution has been posted yet
01:11 PM roycroft: but on buster
01:11 PM roycroft: so it seems it affects multiple debian distros
01:11 PM roycroft: oh, i'm sorry
01:11 PM sliptonicBradCol: Can you point me to a link so I can follow?
01:12 PM roguish[m]: _discord_522858624599719959[m]: launch linuxcnc via a cmd window prompt and see what exactly is going on.
01:12 PM roycroft: the issue is that there's a hard-coded version of that library that is the one that shipped with buster
01:12 PM roycroft: i just see a discussion on the mailing list
01:12 PM unterhaus: the problem is in a package that is not part of lcnc
01:13 PM roycroft: i don't know of a website where it's documented
01:13 PM unterhaus: it's hosted by sourceforge, do they have the emails?
01:13 PM roycroft: yeah, it's an issue with tkimg
01:14 PM sliptonicBradCol: Can I avoid this by building from source?
01:15 PM unterhaus: here is one message, does sourceforge thread? https://sourceforge.net/p/emc/mailman/message/37667381/
01:15 PM XXCoder: honestly id use iso file and install it from that
01:15 PM XXCoder: it has everything you need ready to gio
01:16 PM unterhaus: someone broke the threading. Here is today https://sourceforge.net/p/emc/mailman/emc-developers/?viewmonth=202206&viewday=14
01:16 PM silopolis0866[m]: most probably as it didn't show up earlier and I believe more people are building RIP than installing straight from Debian
01:18 PM roguish[m]: it's not always the best to be out on the bleeding edge. I second XXCoder's statement
01:18 PM srk- is now known as srk
01:18 PM unterhaus: I think I would install from debian. Glad I know it doesn't work right now
01:18 PM sliptonicBradCol: I thought installing from the distro apt package manager was staying AWAY from the bleeding edge!
01:19 PM silopolis0866[m]: It will soon be !
01:20 PM silopolis0866[m]: LinuxCNC was introduced in Debian not so long ago and there have been lots of problems with Python versions
01:20 PM roycroft: installing on debian 12 would seem to me to be fairly "bleeding edge"
01:21 PM silopolis0866[m]: It looks like it is stabilizing now and we discussed initiating 2.9 pre release very soon(c)
01:22 PM JT-Shop: I just clone the repo and build a deb
01:22 PM JT-Shop: then install that deb
01:23 PM silopolis0866[m]: Main target for 2.9 should be Bullseye (11), but Bookworm (12) will be supported too and Andy insisted that we don't abandon Buster users in 2.8...
01:24 PM silopolis0866[m]: I ould have bet so... But you're not exactly Jo Rooky 😉
01:25 PM roycroft: debian 12 is still a moving target
01:25 PM silopolis[m]: JT-Shop: BTW, we need you in next meeting to talk about MesaCT 🙂
01:25 PM roycroft: i would strongly recommend installing on debian 11 at the moment
01:27 PM sliptonicBradCol: If I need a stable machine, debian 11 is fine. I don't even mind building from source.
01:27 PM sliptonicBradCol: But I want the ability to spin up a machine fast for testing with minimum hassle.
01:28 PM JT-Shop: meeting?
01:30 PM JT-Shop: sliptonicBradCol, https://forum.linuxcnc.org/9-installing-linuxcnc/45793-install-2-9-on-debian-11
01:30 PM silopolis[m]: > <@_discord_522858624599719959:jauriarts.org> If I need a stable machine, debian 11 is fine. I don't even mind building from source.
01:30 PM silopolis[m]: > But I want the ability to spin up a machine fast for testing with minimum hassle.
01:30 PM silopolis[m]: Then maybe install from lcnc ISO or Buster and add builbot repository
01:30 PM silopolis[m]: http://buildbot.linuxcnc.org/
01:31 PM silopolis[m]: JT-Shop: The one we had yesterday, next should be in 3/6 weeks
01:31 PM silopolis[m]: we had announced it weeks ago on MLs and I also called for participation here
01:32 PM sliptonicBradCol: Generally what I'm testing is features in FreeCAD. So I need to be able to build freecad and the linuxcnc isos tend to lag. Sometimes a lot. That's why I was thrilled to see linuxcnc in bookworm
01:32 PM CaptHindsight[m]: i have about ten machines running LCNC 2.9 on Debian 11 Bullseye
01:33 PM roycroft: debian 12 is a different os every day
01:33 PM roycroft: and will be for some time to come
01:33 PM silopolis[m]: Using testing on my main PC for years and years, I'd say Bookworm is OK
01:33 PM sliptonicBradCol: Does linuxcnc have a Bullseye iso?
01:33 PM roycroft: it is lunch time
01:33 PM roycroft: so time to apply some more shellac
01:34 PM silopolis[m]: sliptonicBradCol: not before 2.9 release
01:34 PM roycroft: it is also not raining - i am going to head over to the country fair after work, and hopefully i'll be able to make it to my booth to assess what repairs need to be done
01:34 PM CaptHindsight[m]: silopolis: have that link handy for qtpyvcp debian 11 install?
01:35 PM CaptHindsight[m]: sliptonic (Brad Collette)#2339: I'll get you the link with the simple instructions
01:35 PM silopolis[m]: https://www.qtpyvcp.com/install/index.html
01:35 PM Rab_ is now known as Rab
01:36 PM CaptHindsight[m]: https://www.qtpyvcp.com/install/bullseye.html yeah that is it
01:36 PM CaptHindsight[m]: sliptonic (Brad Collette)#2339: ^^^
01:40 PM JT-Shop: silopolis[m], I missed it... the mailing list is so active I can't keep up. Did someone mention mesact?
01:40 PM sliptonicBradCol: Thanks. Not to be too critical but this has bugged me about the linuxcnc community for 15 years. The right answer is almost always out there but it's hidden among so many wrong answers that it's almost impossible to find. Between the wiki, the forum, the blogs, the mailing list, discord, etc etc.
01:40 PM sliptonicBradCol: Maybe that's just endemic to larger mature projects. But if someone would ever start (and maintain) a curated / vetted list of tutorials, I'd kiss 'em on the lips.
01:41 PM XXCoder: maybe thats why nobody does it, they dont want kisses
01:41 PM sliptonicBradCol: fair 'nuff
01:41 PM JT-Shop: https://gnipsel.com/linuxcnc/index.html
01:41 PM JT-Shop: and you can't kiss me I'm taken lol
01:41 PM CaptHindsight[m]: sliptonic (Brad Collette)#2339: well the old devs have some sort of mental block with the rest of the LCNC community
01:42 PM * JT-Shop has 1/3 of his desk clean... I might attempt another third
01:42 PM XXCoder: just tip table over
01:42 PM XXCoder: ;)
01:42 PM XXCoder: nice though
01:43 PM JT-Shop: pretty heavy table...
01:43 PM JT-Shop: old oak desk
01:43 PM CaptHindsight[m]: they don't engage on IRC/matric/discord the forums or the ml but for some reason still keep the servers and the git tree going
01:45 PM CaptHindsight[m]: so yeah it's a bit dysfunctional
01:46 PM sliptonicBradCol: @JT-Shop, I've used your tutorials extensively in the past. At a bare minumum I owe you a hug. Great stuff. You need to add a couple things: a 'recent updates' page and an index.
01:46 PM JT-Shop: good idea
01:56 PM travis_farmer853: slightly OT (3D printer question): the extruder keeps losing traction on my Ender 3. i have tried tightening the pressure. ... and as i am writing this, i recall the tip of the nozzle was skuffed from the previous owner... perhaps i need a new nozzle...
01:57 PM c101horse8332[m]: have a look at the bowden tube at the nozzle end, see if its still tight up against the nozzle
01:57 PM silopolis[m]: <JT-Shop> "silopolis, I missed it... the..." <- Not this time, but I believe it would be nice.
01:57 PM silopolis[m]: I hope TurBoss will talk a bit about QtPyVCP and work he's done on tool management there too
01:57 PM roycroft: travis_farmer853: does your printer have a plastic extruder?
01:57 PM travis_farmer853: metal tooth wheel
01:58 PM roycroft: but the extruder itself is plastic?
01:58 PM travis_farmer853: i think so
01:58 PM silopolis[m]: > <@_discord_522858624599719959:jauriarts.org> Thanks. Not to be too critical but this has bugged me about the linuxcnc community for 15 years. The right answer is almost always out there but it's hidden among so many wrong answers that it's almost impossible to find. Between the wiki, the forum, the blogs, the mailing list, discord, etc etc.
01:58 PM silopolis[m]: > Maybe that's just endemic to larger mature projects. But if someone would ever start (and maintain) a curated / vetted list of tutorials, I'd kiss 'em on the lips.
01:58 PM silopolis[m]: You'll see in the minutes of the meeting that that's one of our goals for docs
01:58 PM roycroft: you should get this:
01:58 PM silopolis[m]: Add and improve user orientated docs
01:58 PM roycroft: https://www.amazon.com/gp/product/B07B96QMN2/
01:58 PM roycroft: and you should get a new bowden tube
01:59 PM travis_farmer853: ok, thanks
01:59 PM roycroft: https://www.amazon.com//dp/B08BHVWNWK/
01:59 PM roycroft: get that, not just any bowden tube - you want capricorn, and that comes with a cutter so you can make a nice, square end
02:00 PM roycroft: after you get those, clean out your hot end really well
02:00 PM roycroft: remove the nozzle and examine it
02:00 PM roycroft: replace if it looks worn or damaged
02:00 PM roycroft: you should see a significant improvement in performance and print quality after you do those things
02:00 PM travis_farmer853: i have a Titan metal extruder, for a CR10 or something, that i was going to try on my big router/3d printer experiment, i wonder if i can just swap that in...
02:00 PM sliptonicBradCol: Improving user docs great (and hard). I wish documentation included a 'freshness' stamp. It's one thing to see, like on JT-Shops docs, that it was updated on a date. If that date is recent, it implies freshness. But if the date is not recent, it doesn't imply not-fresh. So docs should be routinely reviewed and marked 'reviewed on xxxx by xxxx'
02:01 PM travis_farmer853: https://www.ebay.com/itm/154647079845
02:01 PM CaptHindsight[m]: sliptonic (Brad Collette)#2339: the format foe the online docs also changed a few years back
02:02 PM CaptHindsight[m]: if you get a google search result it is often for an older version
02:02 PM roycroft: some folks like those direct extruders
02:02 PM CaptHindsight[m]: and if you just change the version number you get a dead link
02:02 PM roycroft: personally, i don't want to add that much weight to the z axis
02:02 PM c101horse8332[m]: yes, it can be swapped in, we made a custom mount to mount a hemera direct drive on the ender 3
02:03 PM CaptHindsight[m]: s/foe/for/
02:03 PM roycroft: with a good bowden tube and a good, well placed extruder, i haven't seen any drawbacks to having the extruder separate from the hot end
02:03 PM travis_farmer853: 🤔
02:03 PM travis_farmer853: good point...
02:03 PM roycroft: that is my personal opinion, and i have not done any testing to compare a direct extruder and a remote extruder
02:04 PM c101horse8332[m]: direct drive has a little less to worry about imho
02:04 PM sliptonicBradCol: Here's one for you. http://linuxcnc.org/docs/stable/html/code/building-linuxcnc.html#_non_realtime
02:04 PM Loetmichel_: Bowden is crap for TPU
02:04 PM travis_farmer853: i already have it, so i could try it. if it doesn't work, i go back separate
02:04 PM Loetmichel_: everything else its ok
02:04 PM sliptonicBradCol: The Quick start clone url is wrong. github doesn't support that protocol anymore.
02:05 PM sliptonicBradCol: should be git clone git@github.com:LinuxCNC/linuxcnc.git
02:05 PM CaptHindsight[m]: sliptonic (Brad Collette)#2339: there was a big effort to get LCNC into the debian repos and a new format for docs
02:05 PM roycroft: there are many "correct" ways to build a 3d printer
02:05 PM travis_farmer853: true
02:05 PM roycroft: don't look for the right way
02:05 PM Loetmichel_: roycroft: IS there a "correct" way?
02:05 PM c101horse8332[m]: i think that was the reason we went with the directdrive, but in the end never printed tpu🙃
02:05 PM roycroft: look for the way that's right for you
02:06 PM Loetmichel_: or just ways that excel in one or the other aspect?
02:06 PM roycroft: any way that produces the desired results at an acceptable cost in terms of time and money is a correct way
02:06 PM firephoto__ is now known as firephoto_
02:06 PM Loetmichel_: i meant: one excels in speed, the other in price, the next in quality... Pick two ;)
02:07 PM roycroft: pick what works for you
02:07 PM roycroft: and that is your correct way
02:08 PM silopolis[m]: <sliptonicBradCol> "Improving user docs great (and..." <- Timestamping is also on my TODO of improvements for the docs
02:09 PM Loetmichel_: yeah, I think what i meant to say was "there are many "correct" ways to make a 3d-printer, bit none that fits all/is perfect"
02:09 PM roycroft: that is one of the things i really like about RCS/CVS
02:09 PM CaptHindsight[m]: no link from http://linuxcnc.org/docs/stable/html/ to http://linuxcnc.org/docs/stable/html/code/ or http://linuxcnc.org/docs/stable/html/code/building-linuxcnc.html
02:09 PM Loetmichel_: and as i am a lazy bum i just tend to buy stuff like ender3s instead of making one myself ;)
02:10 PM roycroft: i just add a $ID:$ line in the footer, and whenver i check in the file, the version control system inserts a timestamp there
02:10 PM silopolis[m]: <CaptHindsight[m]> "sliptonic (Brad Collette)#2339..." <- For docs, the big 9 month work that has just been commited was to put in place a translation support infrastructure using po4a and weblate
02:11 PM CaptHindsight[m]: silopolis: ok, so how does that help the avg LCNC new user?
02:11 PM silopolis[m]: This needed to sanitize docs formating and sync structure of Fr and Es docs with En source
02:11 PM roycroft: i suppose a drawback of that is if you are just checking in an entire project, and did not review the docs, the updated timestamp on the doc file(s) is not necessarily an accurate indicator of when the file(s) are/were reviewed/changed
02:11 PM CaptHindsight[m]: a new user goes to the wiki or searches for info just like sliptonic and finds a mess
02:12 PM CaptHindsight[m]: silopolis: look at my example above, no connection between pages
02:13 PM roycroft: hmm, with git you can use a pre-commit hook to insert a timestamp at commit time
02:13 PM CaptHindsight[m]: if you already know where it is great but if you are new what then?
02:13 PM silopolis[m]: CaptHindsight: like I said, at this point, that was really infrastrutcure work... But from now on, docs can be easily translated on https://hosted.weblate.org/projects/linuxcnc/ which should in the end benefit loooots of users
02:14 PM silopolis[m]: This has also allowed us to gain a better knoledge of the docs, and what's need to be done
02:14 PM CaptHindsight[m]: silopolis: how do you fix http://linuxcnc.org/docs/stable/html/ if you don't have keys to the server?
02:15 PM silopolis[m]: Beware that stable is 2.8 !
02:15 PM silopolis[m]: Master is here https://linuxcnc.org/docs/devel/html/
02:16 PM CaptHindsight[m]: and how does one fix that??
02:17 PM silopolis[m]: In turn, the benefit of our work on docs is that there is now a single source: the english docs here https://github.com/LinuxCNC/linuxcnc/tree/master/docs/src
02:17 PM CaptHindsight[m]: where is the link to http://linuxcnc.org/docs/stable/html/code/building-linuxcnc.html
02:18 PM CaptHindsight[m]: how do you volunteer to fix something that you can't access?
02:18 PM silopolis[m]: This sources are now ingested by the po4a/weblate toolchain and translatable by users with a nice web UI
02:19 PM silopolis[m]: CaptHindsight[m]: Oh, ok... in 2.8 branch
02:19 PM silopolis[m]: https://github.com/LinuxCNC/linuxcnc/tree/2.8/docs/src
02:27 PM Connor1 is now known as Connor
02:27 PM CaptHindsight[m]: silopolis: OK and where is this explained to new users and devs?
02:27 PM silopolis[m]: Should be in the docs
02:27 PM CaptHindsight[m]: I am a terrible mind reader
02:27 PM CaptHindsight[m]: likely what keeps me sane -ish
02:27 PM silopolis[m]: https://github.com/LinuxCNC/linuxcnc/blob/master/docs/src/code/contributing-to-linuxcnc.adoc
02:27 PM rene-dev59 is now known as rene-dev5
02:27 PM andypugh_ is now known as andypugh
02:27 PM silopolis[m]: https://github.com/LinuxCNC/linuxcnc/blob/master/docs/README.adoc
02:27 PM silopolis[m]: But that's far from satisfying
02:40 PM * roycroft wonders if unterhaus' real name is red
02:41 PM unterhaus: I'm green
02:41 PM roycroft: yup, sounds like you
02:42 PM unterhaus: someone misspelled Andy's name, but I would say it's fitting
02:42 PM unterhaus: I wish I was that funny
02:42 PM CaptHindsight[m]: have front doors become like telephones?
02:42 PM XXCoder: red green, isnt that famous tv show :P
02:42 PM CaptHindsight[m]: people under 30 don't seem to know how to work them
02:42 PM unterhaus: obsolete?
02:43 PM CaptHindsight[m]: How to knock on a door
02:43 PM CaptHindsight[m]: How to ring a bell
02:43 PM unterhaus: I'm surprised they even approach the door
02:43 PM * Tom_L wonders if unterhaus is the hulk
02:43 PM Tom_L: bein green n all..
02:43 PM CaptHindsight[m]: How to leave a voice message
02:44 PM unterhaus: shorter Cap'n: get offen my lawn
02:44 PM Tom_L: you knock and run right?
02:45 PM roycroft: delivery people don't knock any more
02:45 PM roycroft: it's toss and run for them
02:45 PM Tom_L: some do some don't around here
02:45 PM Tom_L: typically amazon don't
02:45 PM roycroft: all but usps used to here
02:45 PM roycroft: but now nobody does
02:46 PM roycroft: if there was no signature required, they would knock and run in the past
02:46 PM roycroft: and that's fine
02:46 PM ZincBoy[CAON]066: I am glad when they don't. Scares the cats when they do 🙂
02:46 PM Tom_L: they been talkin to your neighbor kids
02:46 PM roycroft: but now they just run
02:46 PM Tom_L: maybe they're afraid
02:46 PM CaptHindsight[m]: Tom_L: yeah Amazon driver lightly tapped on the door
02:46 PM roycroft: oh, i did not give you the update on the neighbor kids
02:46 PM Tom_L: haha
02:46 PM Tom_L: can't wait..
02:46 PM roycroft: i was at the market the other day buying some groceries
02:46 PM roycroft: i saw the kids there, with someone i assume is their mom
02:47 PM roycroft: as they were leaving, one of them said to me "hello, fat old ugly hippie neighbor"
02:47 PM roycroft: and they all laughed, including the mome
02:47 PM Tom_L: kinda rude
02:47 PM roycroft: i still don't know what house they live in
02:47 PM roycroft: but i know that if they cause any more problems, going to the parents is not going to help
02:48 PM roycroft: i suspected that would be the case all along
02:48 PM Tom_L: tell em some day you'll be just like me only in prison
02:48 PM CaptHindsight[m]: sounds like my Fox News neighbors
02:48 PM roycroft: the kids' behaviors and the things they say were taught to them
02:48 PM roycroft: they're not old enough to come up with any of those ideas on their own
02:49 PM roycroft: anyway, they are least staying off my property
02:49 PM roycroft: that's all i care about
02:50 PM CaptHindsight[m]: good news
02:50 PM * roycroft notes as well that the mom definitely weighs much more than he does
02:50 PM Tom_L: heh figures
02:50 PM CaptHindsight[m]: if that changes maybe try a trench filled with pointed sticks
02:50 PM roycroft: or raspberries?
02:50 PM roycroft: or maybe a banana?
02:50 PM roycroft: i haven't been taught how to use pointed sticks yet
02:51 PM CaptHindsight[m]: and now for something completely different
02:51 PM CaptHindsight[m]: sometimes you need to choose your neighbors wisely
02:52 PM roycroft: i was here first
02:52 PM Tom_L: we get new ones from time to time and 'break them in' to the neighborhood if they look like a problem
02:52 PM Tom_L: i definitely was here first
02:52 PM roycroft: well i live in a very redneck neighborhood
02:52 PM roycroft: i'm the misfit here
02:52 PM roycroft: but most of my neighbors at least tolerate me
02:53 PM Tom_L: don't quite fit the mold ehh?
02:53 PM Tom_L: i think i broke it
02:53 PM unterhaus: shoulda bought more land
02:54 PM CaptHindsight[m]: sad and pathetic, they need to attempt to put down those around them to feel slightly better for a moment
02:54 PM roycroft: i probably would not fit in well in a hippie neighborhood either
02:54 PM roycroft: it's not that i'm intentionally non-conformist
02:55 PM roycroft: it's just that i do what i think is the right thing to do, and do not consider tribalism or conformity when making decisions
02:57 PM CaptHindsight[m]: i run into it everywhere, city, suburbs, country, corporate offices and the factory floors
03:15 PM CaptHindsight[m]: 97F (36C)
03:16 PM CaptHindsight[m]: heat index 113 today, tornadoes yesterday
03:17 PM CaptHindsight[m]: at one point i counted 4 tornadoes over Chicago on radar but none touched down
03:23 PM c101horse8332[m]: i am the redneck in the neighborhood, only neighbor that has problems with me is import
03:24 PM roycroft: i remember the tornados when i lived in chicago
03:25 PM roycroft: most folks don't realise that there is tornado alley there
03:25 PM roycroft: they think of kansas and oklahoma when they think of tornados
03:26 PM Tom_L: we have 2 of the biggest on record
03:31 PM roycroft: i'm not comparing the ones in chicago with yours
03:31 PM roycroft: just saying that tornados happen all over the place
03:31 PM Tom_L: i know
03:31 PM roycroft: there's a little tornado alley just south of portland
03:33 PM unterhaus: we shouldn't have tornadoes, but we do. Mostly only do damage in the mountains
03:34 PM unterhaus: knocks down about the same number of trees as a heavy rain, but more concentrated
03:35 PM roycroft: we just had two in the blue mountains last month
03:35 PM roycroft: which is in the northeastern part of the state, and is a very unusual place for tornados
03:37 PM unterhaus: i wonder if there is a bug report with the tcl library
03:38 PM unterhaus: tkimg
03:40 PM unterhaus: it's weird how many repositories there are for tkimg. Which one is debian
03:42 PM unterhaus: there is a dev at debian that is responsible for the package https://repology.org/project/tkimg/versions
03:43 PM CaptHindsight[m]: silopolis: how does weblate effect me as a LCNC user/dev?
03:43 PM CaptHindsight[m]: can i swim as usual? Will it effect my diet?
03:45 PM silopolis[m]: Weblate really is for translation management, so if you don't need translated docs or don't plan to translate docs, the main effect for you is hopefully dragging more people to read it, fix it, and improve it by making it accessable to non english speaking people
03:45 PM CaptHindsight[m]: ah ok
03:46 PM * JT-Shop waits for the guy with a pocket full of bengies to arrive
03:48 PM CaptHindsight[m]: silopolis: I did not know that the website docs had the doc github as its backend
03:52 PM silopolis[m]: As in most FOSS projects, almost everyting is kept in VCS, further, LinuxCNC uses a single repository for everything
03:53 PM * JT-Shop loves it that the mesact can download the deb for you and has a cool progress bar
03:54 PM silopolis[m]: Only major projects I know outside of main repo are JT-Shop 's MesaCT and hazzy 's QtPyVCP
03:56 PM silopolis[m]: Oh, website also has its own repo
03:57 PM JT-Shop: in the beginning I didn't want to step on Chris's toes by adding another configuration tool but I've surpassed pncconf with what my tool does
03:57 PM JT-Shop: yep WLO
03:58 PM JT-Shop: I also like that the tool can check for new releases
03:59 PM silopolis[m]: I believe it's a gret contribution that will prove to be a cornerstone
03:59 PM silopolis[m]: And I'm sure Chris has enough on its plate to leave you the Mesa configuration utility lol
04:00 PM silopolis[m]: At least I, now that MesaCT is there, prefer having him working on QtVCP or anything else 🙂
04:01 PM JT-Shop: I don't understand what you mean "leave you the Mesa configuration utility"
04:01 PM silopolis[m]: Worst with MesaCT is that now that it's there... There are no more boards to config ! 😜😅
04:02 PM JT-Shop: they are coming...
04:02 PM JT-Shop: now that the 7i97 has firmware for daughter cards I need to work on adding a daughter card to an all in one card
04:02 PM silopolis[m]: Yeah, was just joking 🙂
04:04 PM JT-Shop: the worst thing about pncconf is it's done in glade...
04:05 PM JT-Shop: looks like phil is working on pncconf to keep it up to qtplasmac
04:06 PM silopolis[m]: Why that ? what's special between pncconf and qtplasmac ?
04:07 PM CaptHindsight[m]: where is MesaCT hidden?
04:07 PM JT-Shop: they are friends and I guess phil works on it when needed
04:08 PM JT-Shop: it's hiding in plain sight on github.com/jethornton
04:08 PM JT-Shop: https://forum.linuxcnc.org/27-driver-boards/45994-mesa-configuration-tool-features
04:09 PM CaptHindsight[m]: JT-Shop: google must have some issues with your site, it comes up way after a bunch of non relevant links
04:09 PM CaptHindsight[m]: https://gnipsel.com/mesa/index.html
04:09 PM JT-Shop: that's the online docs
04:10 PM JT-Shop: https://github.com/jethornton/mesact
04:11 PM JT-Shop: 96°F 39%rh
04:11 PM JT-Shop: roycroft, would be able to shellac like a mad man in this weather
04:11 PM CaptHindsight[m]: https://gnipsel.com/mesa/mesact/index.html
04:12 PM JT-Shop: yep that's the online docs
04:12 PM JT-Shop: the tool also comes with a builtin pdf docs as well as the docs for all the mesa boards
04:13 PM CaptHindsight[m]: why come it's hotter here than in Missouri?
04:13 PM JT-Shop: your closer to the sun
04:13 PM silopolis[m]: Second week above 32°C, 34+ for 4 days and climbing for the week to come... That's insane ☹️
04:14 PM CaptHindsight[m]: oh yeah the sun is north in summer
04:14 PM JT-Shop: 32°C is easy to take..
04:14 PM JT-Shop: plus it's downhill to Missouri
04:14 PM silopolis[m]: Not for me, I'm melting above 20
04:17 PM JT-Shop: silopolis[m], you will have to remind me about the next meeting, I miss a lot on the mailing list
04:17 PM silopolis[m]: OK no problem 🙂
04:20 PM silopolis[m]: rene-dev3 is our next host and should announce it with a 2 weeks period for everyone to agree on date and time using something like framadate.org, then meeting happens on Debian's Jitsi server, directly in browser
04:24 PM JT-Shop: sounds better than the old irc meetings
04:28 PM Tom_L: what? no zoom meetings??
04:32 PM * roycroft hates zoom meetings
04:33 PM silopolis[m]: <JT-Shop> "sounds better than the old irc..." <- I'm enjoying them very much and we have hope it could help bringing a new breath to the project
04:34 PM silopolis[m]: > * <@roycroft:libera.chat> hates zoom meetings
04:34 PM silopolis[m]: Much more fun when it's for tchating with lcnc mates with a cold beer than for marketing briefs 😉
04:34 PM coralgoat: I am looking at switching out a 220V VFD and 220V spindle motor for a 120V VFD and 120V spindle motor.
04:35 PM coralgoat: When searching for a 120V VFD and spindle motor combination I ran across a 110V to 220V VFD.
04:35 PM coralgoat: https://www.amazon.com/dp/B09SLV71Y4
04:35 PM coralgoat: Could I plug this VFD into a 120V outlet and use it with a spindle motor that says this on it: WH80, 2.2kW, 220V, 8A, 400Hz, 2400RPM?
04:35 PM Tom_L: i doubt it
04:36 PM roycroft: no
04:36 PM Tom_L: 1.2kw is about all you're gonna get out of 120v
04:36 PM coralgoat: okay
04:36 PM Tom_L: on a normal circiut
04:36 PM roycroft: we've told you this several times
04:36 PM Tom_L: my first but ok
04:36 PM roycroft: i'm not sure how else to explain it so that it makes sense to you
04:37 PM coralgoat: I did find a couple of 110V VFD and 110V 2.2kW spindle motor combinations.
04:37 PM Rab: coralgoat, you would need to install a special 120V circuit, no less trouble than installing a new 240V circuit (and less ideal).
04:37 PM coralgoat: Interestingly they all came with water cooled spindles.
04:37 PM roycroft: to run a 2.2kw spindle on a 120vac circuit you need 30a to be safe
04:37 PM coralgoat: okay
04:37 PM coralgoat: What do people think of water cooled spindle motors?
04:37 PM Tom_L: go back to your googling..
04:37 PM roycroft: 20a would barely run the spindle, and given the inefficiencies of both the motor and the vfd, you would likely be popping breakers constantly
04:38 PM roycroft: i like water-cooled spindles
04:38 PM roycroft: the disadvantage is that you have to run two water lines to them, but those lines can be small, and are usually not difficult to route along with the power lines
04:38 PM roycroft: the advantage is that you don't have a noisy fan blowing dust all over the place
04:39 PM roycroft: but if you get a 2.2k2 water-cooled spindle, get a decent one, and make sure it has three bearings, not two
04:39 PM coralgoat: These are the three 110V VFD and 110V 2.2kW spindle motor combinations that I have found:
04:39 PM coralgoat: https://www.amazon.com/dp/B07BBHJMQB
04:39 PM coralgoat: https://www.amazon.com/dp/B07QD4DQ9D
04:39 PM coralgoat: https://www.amazon.ca/dp/B089SVKV1F
04:39 PM roycroft: none of which will work
04:40 PM roycroft: i don't even have to look at them :)
04:40 PM roycroft: for the reasons that myself and others have explained to you multiple times
04:40 PM roycroft: i promise you, finding a new link with a different product that has the same specifications is not going to make it work
04:40 PM Rab: Well, the first combo is set up for 800W.
04:40 PM roycroft: you need a 30a 120v circuit
04:41 PM roycroft: 800w is easy to do on 120v
04:41 PM roycroft: 2.2kw not worth the trouble, because running it on 240v would be cheaper
04:41 PM roycroft: 1.2kw, pretty iffy on a standard 20a circuit, but it will probably work most of the time
04:43 PM coralgoat: roycroft, so something less than or equal to 1.2kW for the motor spindle on 120V outlet?
04:43 PM roycroft: i would say "less than"
04:44 PM roycroft: because i'm rather conservative
04:44 PM coralgoat: okay
04:44 PM roycroft: you're being optimistic, so you probably want to go with "less than or equal to"
04:44 PM Rab: I'm a little skeptical of the nameplate rating on those chinese spindles. It would be interesting to profile them under load.
04:45 PM roycroft: just remember that (and i'm simplifying things) w = v * a
04:45 PM roycroft: as the voltage decreases, the amperage increases to obtain the same amount of power
04:46 PM roycroft: a 1.2kw spindle is going to require 10a of power @120vac
04:46 PM roycroft: and you might think that's fine, because that is 1/2 of a 20a circuit
04:46 PM roycroft: BUT
04:46 PM roycroft: 1. your vfd is not 100% efficient, so it may require 13a to deliver 10a
04:46 PM roycroft: or maybe 16a to deliver 10a
04:47 PM roycroft: 2. when the motor first starts up there will be a current surge
04:47 PM roycroft: you want a buffer in your circuit sizing to accomodate the surge
04:48 PM roycroft: if your vfd requires 15a to deliver 10a to the motor, it may well draw 22a or so at motor start-up
04:48 PM coralgoat: Thanks for explaining that.
04:49 PM roycroft: you'll also have some voltage drop on the circuit itself, depending on how far from the breaker box you are, although with a properly designed/installed circuit that is usually very minor
04:49 PM roycroft: now if you run a long extension cord from the receptacle to your vfd, and that cord is undersized, you'll see significant voltage drop there
04:50 PM coralgoat: I did find a couple 800W and 500W VFD and spindle motor combinations.
04:50 PM roycroft: which means even more current draw
04:50 PM roycroft: also, the extension cord will get warm
04:50 PM roycroft: you don't want warm extension cords
04:50 PM coralgoat: okay
04:51 PM roycroft: electrical wires are resistors
04:52 PM roycroft: the first thing you need to do is determine what size spindle motor you need for the work you want to do
04:52 PM roycroft: then you need to find a vfd that will safely and efficiently drive that motor
04:53 PM roycroft: then you need to get an electical circuit that will safely and efficiently feed the vfd
04:53 PM roycroft: you don't pick the circuit first and figure out what will work
04:53 PM roycroft: unless this is a strictly hobby machine and you don't care what you can and cannot do with it
04:53 PM coralgoat: I would like to be able to work with aluminum, plastic and wood.
04:54 PM roycroft: well with an 800w spindle you're going to be using very small cutters for aluminium, and taking fairly light cuts
04:55 PM roycroft: so make sure the spindle goes to 24,000rpm or higher
04:55 PM roycroft: before you do anything, get an estimate on pulling a 20a 240v circuit to the machine location
04:56 PM coralgoat: Yes I will do that.
04:56 PM roycroft: even if you don't end up doing that, it gives you a baseline cost for doing it properly
04:56 PM roycroft: and you can better evaluate the compromises you're thinking of making
04:57 PM coralgoat: Will the 2.2kW spindle motor will be better at aluminum than the 800W spindle motor?
04:57 PM roycroft: i would say get an estimate on a 30a 120v circuit at the same time, but i can tell you that will cost more and will just piss off the electrician whom you are asking for the estimate
04:57 PM roycroft: it will be able to take bigger cuts
04:57 PM Rab: Given that you already have the 240V hardware, I'd explore a new circuit first. (As I've mentioned.)
04:57 PM roycroft: you also said your breaker box is full
04:57 PM roycroft: two thoughts on that:
04:58 PM roycroft: 1. you can always add a subpanel, which would likely be rather expensive
04:58 PM roycroft: 2. there are breakers that can "double up" on circuits
04:59 PM Rab: It can be cheap, too. I ran my 240V mill for several years from a 12/3 extension cord I built that went to a clothes dryer outlet.
04:59 PM roycroft: so if you have two standard-size breakers directly touching each other, you can often replace them with what is in effect four half-size breakers, the middle two of which will be tied together and will provide a 240v circuit, and the outer two will replace the existing ones
04:59 PM roycroft: and the amperages of the various circuits do not have to be the same
05:00 PM roycroft: so you can match the amperage of the existing breakers, and add a 20a 240v circuit as well
05:00 PM roycroft: not all breaker types can handle that, and your breaker box may not be rated for it
05:00 PM coralgoat: okay
05:00 PM roycroft: you really need an electrician to look at the job and tell you what is possible
05:01 PM roycroft: i'm oversimplifying things to a great extent - i'm just trying to make the point that you likey do have that option, even if you think you might not
05:01 PM -!- #linuxcnc mode set to +v by ChanServ
05:01 PM coralgoat: That is good to know about.
05:02 PM XXCoder: roy hacky solution, I guess, would be plug to dryer
05:02 PM XXCoder: remove dryer plug and plug into that
05:02 PM roycroft: i would also posit that if you're going to build a cnc router, there's a good chance that you might want to install another 240v machine in the shop in the future
05:02 PM roycroft: i suggested that several days ago, xxcoder :)
05:02 PM XXCoder: not surpised
05:03 PM roycroft: the takeaways here are:
05:03 PM roycroft: 1. if you're going to be machining non-ferrous metals you probably want a minimum 2.2kw spindle
05:03 PM roycroft: 2. if you're running a 1.2kw spindle or bigger you should run it on 240v
05:03 PM roycroft: i think you know what to do now :)
05:05 PM roycroft: if this is a garage or basement, you can have the new circuit run through surface-mounted conduit - it doesn't have to be put inside the wall
05:05 PM roycroft: and if it's in the same room as the electrical panel, the closer you can locate the machine to the panel the less expensive the install
05:07 PM roycroft: and on that note, i need to head to the country fair and see if i can check out my booth
05:07 PM coralgoat: Thanks
05:08 PM JT-Shop: Spyder is gone
05:09 PM tcurdt: it's time to trying to get started with linuxcnc and setting up my 7i92 ... but I am still feeling lost on where to get started
05:10 PM tcurdt: I have flashed the bin to the mesa ... have mesact running ... but still don't see how to configure the pin to the functions
05:10 PM JT-Shop: I wonder if he will remember how to start it when he gets home?
05:10 PM JT-Shop: what daughter card?
05:10 PM XXCoder: jt nice
05:10 PM tcurdt: JT-Shop custom ... I want to assign all pins of P1/P2 manually
05:11 PM tcurdt: 5 step generators, some outputs, some inputs
05:12 PM JT-Shop: did you flash it with 7i92_5ABOBx2D.bit?
05:12 PM tcurdt: yes
05:13 PM JT-Shop: you can configure the axes on one connector by selecting P2 5ABOB
05:13 PM tcurdt: JT-Shop but why is P1 disabled then?
05:13 PM JT-Shop: and you will have the I/O for P2 in the tool
05:13 PM JT-Shop: I've not gotten that far with the tool
05:14 PM tcurdt: aaah ... ok ... bummer
05:14 PM JT-Shop: so many options make the code difficult
05:14 PM tcurdt: so it seems like I should write the hal and ini manually then I guess?
05:14 PM JT-Shop: you could start with what the tool makes for you then add the rest
05:15 PM JT-Shop: it _might_ work to build a config for P2 then for P1 and copy and paste but you would have to tell me if there is any issues
05:15 PM tcurdt: ok
05:16 PM JT-Shop: multiple daughter cards is in my brain but other things are more up front right now
05:17 PM tcurdt: the config UI reveals so many settings ... it's almost more scary than the raw files (to me) ... not sure I have all the answers :)
05:17 PM JT-Shop: if you look the GPIO tab will be for both p1 and p2 and you will be able to select what you want there but not done yet
05:18 PM tcurdt: is there some indication of what are required values?
05:18 PM JT-Shop: I/O will be pretty simple on the GPIO tab but axes get complicated
05:18 PM JT-Shop: values for?
05:19 PM tcurdt: let me quickly go through the tabs to gather them...
05:19 PM JT-Shop: ok
05:20 PM JT-Shop: I have about 40 minutes until chow time
05:20 PM tcurdt: ok, I'll work through faster :)
05:23 PM tcurdt: alright ... maybe let me go through one by one
05:24 PM JT-Shop: brb gotta go find a tree
05:24 PM tcurdt: display: not sure which GUI to pick yet, position offset and feedback is unclear to me
05:26 PM tcurdt: axes: PID Settings and FE I probably can just start with the defaults, wish there were some defaults for homing, too ... no idea what will make sense for the speeds etc yet
05:26 PM RoboticsNut[m]: aparently the servo drive on the Bridgeport is setup as a PDFF controller (Pseuro Derivative Feedback + FeedForward), any tips or experilence tuning this to operate with LinuxCNC?
05:27 PM tcurdt: spindle: I want to control via modbus. Is the tab for a PWM controlled output?
05:27 PM JT-Shop: fortunately I have a lot of trees and it didn't take long to find a suitable tree
05:27 PM JT-Shop: tcurdt, pick Axis, Relative, Commanded
05:28 PM tcurdt: tool changer type: is empty - I need manual too changes ... that's not supported yet (in the config?)
05:28 PM JT-Shop: Options Tab
05:28 PM JT-Shop: tcurdt, no defaults for homing as there is no way to predict them
05:29 PM tcurdt: JT-Shop not even speeds?
05:29 PM JT-Shop: speeds depend on your machine
05:29 PM JT-Shop: I see I need to move on screen tool change to the tool changer tab
05:29 PM JT-Shop: did you find it on the options tab?
05:30 PM tcurdt: yes
05:30 PM tcurdt: is there anything else in the options tab that is essential to get started?
05:30 PM RoboticsNut[m]: I've never heard of a PDFF Velocity controller before.
05:30 PM JT-Shop: it does make more sense to be on the tool changer tab
05:30 PM tcurdt: JT-Shop indeed
05:30 PM JT-Shop: no, press check config
05:31 PM coralgoat: For a 220V extension cord for a 220V CNC router would a 12/3 gauge be okay or would a lower gauge be better? I saw a welding machine extension cord with 8 AWG.
05:31 PM JT-Shop: depends on the amps...
05:31 PM tcurdt: JT-Shop saw the question about the spindle tab?
05:31 PM JT-Shop: reading back
05:32 PM tcurdt: spindle: I want to control via modbus. Is the tab for a PWM controlled output?
05:32 PM JT-Shop: that's for a mesa pwm output
05:32 PM coralgoat: 3500W/220V = 15.9A for the CNC router.
05:32 PM tcurdt: JT-Shop ok ... so if I use modbus that somewhere else
05:33 PM CaptHindsight[m]: illinois emissions testing
05:33 PM JT-Shop: hmm modbus as in a hal component?
05:33 PM CaptHindsight[m]: fun
05:33 PM tcurdt: I read there are two hal components for it somewhere ... to chose from
05:33 PM JT-Shop: coralgoat, without looking it up 12 awg should handle up to 20 amps
05:34 PM JT-Shop: tcurdt, I use the gs2 component because I have an automation direct gs2
05:34 PM coralgoat: JT-Shop, okay I will look up AWG amps.
05:34 PM tcurdt: JT-Shop what goes into "Coordinates" in the Machine tab? seems to be just a string field
05:35 PM JT-Shop: I would suggest creating a custom hal file on the options tab and put the hal in there
05:35 PM tcurdt: ok
05:35 PM JT-Shop: when you define an axis it gets added to the coordinates
05:35 PM tcurdt: ah! ok
05:36 PM tcurdt: what's a good ui to start with?
05:36 PM JT-Shop: the more questions the better the tool becomes and the documents as well
05:36 PM JT-Shop: Axis
05:36 PM JT-Shop: it's bullet proof
05:38 PM tcurdt: what's the "scale" on the axis?
05:38 PM tcurdt: just put 1 for now?
05:38 PM JT-Shop: for steppers it's pulses per user unit
05:38 PM coralgoat: <https://en.wikipedia.org/wiki/American_wire_gauge#Tables_of_AWG_wire_sizes> says 12 AWG has an ampacity of 20 at 60 degrees C.
05:38 PM tcurdt: ah
05:38 PM DigitalReaper: don't forget to divide that 15.9 amps by the power factor of your VFD. watts/volts = amps only applies to DC and purely resistive loads on AC
05:39 PM JT-Shop: so if it takes 1200 pulses to move one mm or one inch that is scale
05:39 PM tcurdt: ok
05:39 PM DigitalReaper: alternatively size the cable to handle the highest current your outlet can provide, and if it's a long cable a low enough voltage drop
05:39 PM JT-Shop: I will expand the documents due to our conversation
05:40 PM JT-Shop: tcurdt, did you open the mesact pdf?
05:40 PM tcurdt: max velocity and max acceleration I haven't figured out yet ... any probably OK values to start off with?
05:41 PM tcurdt: JT-Shop where is that pdf?
05:41 PM tcurdt: "Documents"?
05:41 PM JT-Shop: see the button that says Documents?
05:41 PM JT-Shop: yup
05:42 PM tcurdt: alright ... that wasn't obvious (to me) ... sorry, if that's all detailed in there :)
05:42 PM JT-Shop: tcurdt, on the info tab > manuals > main board manuals select 7i92 then press open manual
05:43 PM JT-Shop: all Mesa manuals are at your finger tips :)
05:44 PM RoboticsNut[m]: if the servo drive I'm interfacing with incorporates a Feedforward term, do I set FF to zero?
05:44 PM tcurdt: JT-Shop is the "Documents" pdf online somewhere?
05:45 PM JT-Shop: no, only the html version
05:45 PM tcurdt: or that!
05:45 PM _unreal_: back to milling more of the small parts for this new cnc machine.
05:45 PM JT-Shop: tcurdt, https://gnipsel.com/mesa/mesact/index.html
05:45 PM tcurdt: https://gnipsel.com/mesa/mesact/machine.html#menu?
05:45 PM tcurdt: ah .. ok
05:47 PM tcurdt: would you accept a PR for the docs? ... it feels still quite light
05:48 PM JT-Shop: of course I would love the help
05:49 PM tcurdt: best way to say thx for the help :)
05:50 PM JT-Shop: that would allow me to focus on the code
05:50 PM tcurdt: what's home and home offset?
05:50 PM JT-Shop: a problem when the programmer writes documents is he/she knows it already and does not have fresh eyes
05:50 PM tcurdt: I assume home offset is the distance of the sensor from the origin?
05:51 PM tcurdt: JT-Shop I know. I am a software dev myself :)
05:51 PM JT-Shop: home offset is how much to move after finding the home sensor, needed usually to move off the switch
05:52 PM JT-Shop: distance from the sensor TO the origin
05:52 PM tcurdt: ok ... what's "Home"?
05:52 PM tcurdt: and what is "Home Use Index"?
05:53 PM JT-Shop: home is usually 0 home use index is for encoder with index pulse
05:53 PM JT-Shop: on the axes tab press F1
05:53 PM tcurdt: ah!
05:54 PM JT-Shop: cool?
05:55 PM JT-Shop: https://github.com/jethornton/mesact/blob/master/mesact/src/libmesact/helptext.py
05:56 PM tcurdt: about the FE ... when using closed loop drivers ... that's just the difference between the machine and the mesa?
05:57 PM JT-Shop: closed loop steppers?
05:57 PM tcurdt: s/machine/computer/
05:58 PM JT-Shop: closed loop is usually servo/encoder
05:58 PM tcurdt: servos but with integrated drivers ... basically the same as closed loop steppers (from the linuxcnc perspective)
05:58 PM tcurdt: I don't have the encoder signals in linuxcnc
05:59 PM JT-Shop: with either if you can send the encoder to linuxcnc then it's closed with linuxccn
05:59 PM JT-Shop: linuxcnc
05:59 PM coralgoat: DigitalReaper, thanks I think I will size it to the highest current the outlet can provide.
06:00 PM tcurdt: JT-Shop I brought that up before and people said it's overkill for a small machine
06:00 PM tcurdt: I don't understand the "Home Sequence" help
06:01 PM JT-Shop: yep I don't see any need for stepper closed loop on a small machine or plasma
06:02 PM JT-Shop: if you have home sensors you can define the order each axis homes as well as designate two or more joints to be the same axis as in a gantry
06:03 PM tcurdt: that I get ... just not how to specify it ... it says for a gantry should be negative?
06:04 PM CaptHindsight[m]: Robotics Nut: what is the interface to LCNC? step+dir or analog?
06:04 PM JT-Shop: yep -1 for any joint when two or more joints have the same axis means it's a gantry
06:04 PM JT-Shop: or - any number in sequence
06:05 PM _unreal_: ok TAPPING done on an other location.
06:05 PM JT-Shop: tcurdt, call me anything but late for dinner :)
06:05 PM _unreal_: finally getting closer
06:05 PM JT-Shop: see you tomorrow
06:05 PM tcurdt: JT-Shop sorry!! :)
06:05 PM _unreal_: I'll just be glad when these parts are finally cut.
06:05 PM tcurdt: thanks .. and have a good dinner!
06:05 PM _unreal_: I'm tired of cutting aluminum on my mini cnc
06:06 PM _unreal_: so slow and weak.
06:08 PM _unreal_: I just picked up some tension pins.
06:08 PM _unreal_: When I get the arms to there final location I'll drill the other half of the final location holes
06:09 PM _unreal_: and toss in the pins.
06:09 PM _unreal_: if I ever have to take it apart for some reason. I can re-ALINE much faster.
06:13 PM -!- #linuxcnc mode set to +v by ChanServ
06:18 PM _unreal_: just mounted an other one of the HARD end stops for the Z
06:23 PM _unreal_: I have not done the math but I'm hoping 3 8-32 screws on an aluminum block will be enough to dead end the Z. I guess it depends on the max load the Z lead screw is capable of generating.
06:40 PM tcurdt: can't I configure the stepgens to any pin I want on a mesa? seems like I am not allowed to do this in pncconf
06:43 PM _unreal_: tcurdt, I could be wrong but I dont think so.
06:43 PM _unreal_: what card
06:44 PM tcurdt: 7i92
06:44 PM Tom_L: tcurdt, they need to be pins that the bit file has configured the firmware for
06:44 PM _unreal_: most cards have set pins that are defined as step gen in firmware
06:44 PM tcurdt: so if I had the right bin file it should work
06:45 PM Tom_L: generally yes
06:45 PM Tom_L: also depends on any daugheter card you plug into it
06:45 PM _unreal_: most cards also have pins that have external hardware that is designed FOR step gen
06:45 PM Tom_L: ^^
06:46 PM tcurdt: well, since I could not get (as in order) any daughter bords I designed some myself ... for easier routing I assigned pin as I saw fit
06:47 PM tcurdt: I guess that was not ideal
06:47 PM _unreal_: I dont know how universal the daughter card system is. I've never msessed with it.
06:47 PM tcurdt: or I have to create my own bin file
06:47 PM _unreal_: that would be a question for pcw--home
06:48 PM Tom_L: if you're new to it i wouldn't suggest making your own bit file
06:48 PM Tom_L: rather make your hardware fit an exhisting one
06:48 PM tcurdt: this is such a rabbit hole
06:48 PM Tom_L: you can use mesaflash to find what functions are associated with the pins
06:48 PM Tom_L: not really
06:49 PM _unreal_: tcurdt, yes it is if your DIY'ing it.
06:49 PM _unreal_: the mesa system has a strong comunity. but it also has a FIXED design scheme. you can change it is is designed to do it. but you had better be up to the challange.
06:49 PM Tom_L: there are bit files that work with alot of the generic BOB's out there
06:50 PM tcurdt: Tom_L well, re-routing isn't great either ... so also building my on bin makes it a rabbit hole
06:52 PM tcurdt: Tom_L how can mesaflash tell me the assigned functions? looking at --help
06:52 PM tcurdt: --readhmid?
06:52 PM Tom_L: yes
06:54 PM tcurdt: hm ... how do I know the device name? I only have the ip
06:54 PM Tom_L: 7i92
06:54 PM tcurdt: yes
06:54 PM tcurdt: or that's the "name" ?
06:54 PM Tom_L: --device 7i92
06:55 PM tcurdt: calling that a "name" is weird
06:55 PM tcurdt: but thx!
06:57 PM Tom_L: will list both headers
06:58 PM Tom_L: and if they're assigned a function and you don't use it, you can still use that pin for GPIO
07:00 PM tcurdt: *nod*
07:03 PM tcurdt: what's with the IOs that have "none" assigned?
07:04 PM tcurdt: I can also just override that function
07:05 PM tcurdt: for GPIO
07:05 PM tcurdt: I just cannot re-assign the QCount and StepGen without a different bin
07:11 PM tcurdt: I am confused ... according to the manual pin 14 should be GND ?!
07:19 PM tcurdt: so what pins is mesaflash printing? I would have expected HDR pins
07:19 PM tcurdt: but seems to be DB25 pins
07:19 PM tcurdt: even for HDR connectors
07:25 PM RoboticsNut[m]: Its +/- 10v Analog Differential signal to input the the control from a MESA 7i77.
07:27 PM RoboticsNut[m]: There is a gearing input that can take step+direction for control, I don't know how performant it is.
07:28 PM RoboticsNut[m]: I tested the positioning mode over the RS232 port on the controller, and its very performant, but the drive and that motor were tuned together, so all the parameters are in the config of the drive.
07:29 PM RoboticsNut[m]: its a Kollmorgen Servostar + AKM series motor.
07:30 PM RoboticsNut[m]: its specifically in the Velocity PDFF controller mode at startup, as it was previously controlled by the non open controller via the +/-10v inputs.
07:34 PM CaptHindsight[m]: "Velocity PDFF controller mode at startup" does that imply that it changes after startup?
07:35 PM CaptHindsight[m]: Robotics Nut: ^^
07:35 PM CaptHindsight[m]: was there any feedback to the non open controller from the drive?
07:57 PM RoboticsNut[m]: no, I can change the mode of the controller with commands from a serial command console.\
08:00 PM RoboticsNut[m]: It connects the A & B quadrature channels, and we can read position on LinuxCNC, in fact a control loop with P=50, FFW set to 1, or 0 works, I have been wrestling with the HALScope to graph the commanded position and actual position trajectories to see what the error is, but since this is my first time working with Motors and LinuxCNC, I am still trying to feel comfortable with it, and I'm not sure where it should be from a tuning
08:00 PM RoboticsNut[m]: perspective, or how everything should be configured.
08:01 PM RoboticsNut[m]: in PNConf, I can run the servo tune loop, though with FF1 is set to 1, which causes the motor to move.
08:01 PM RoboticsNut[m]: This document covers it:
08:02 PM * RoboticsNut[m] posted a file: (266KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/CxMCeBxNcSXBGAkLshyZfyvk/ServoStar_S_CD_Series_Velocity_Tuning.pdf >
08:03 PM RoboticsNut[m]: Its running COMPMODE=1
08:03 PM RoboticsNut[m]: which is the PDFF controller.
08:04 PM RoboticsNut[m]: The drive supports a position control loop PID type control internally, which I enabled to make sure the motor was working well, and it was working nicely, very very performant, I was able to get very high positional accuracy on the drive to commanded positions.
08:05 PM RoboticsNut[m]: The resolvers on the drive were set to interpolate to 65536 counts per revolution, and it was holding the position control loop to with in a few counts of 65536.
08:05 PM pcw--home: tcurdt: what pinout do you need?
08:06 PM RoboticsNut[m]: With regard to the servo control loop for LinuxCNC I Have the simulated quadrature output set to 1024 counts per revolution which produces 4096 counts per revolution with full quadrature decoding on the MESA interface cards.
08:10 PM tcurdt: pcw--home ideally just the dir/step on different pins ... I can send that through ... but could first just confirm the mesaflash output? it shows IO18 on pin 14 (which is a HDR connector) ... but according to the manual on 14 there is GND
08:10 PM tcurdt: does mesaflash always only show the DB25 pins?
08:11 PM tcurdt: even for HDR connectors?
08:11 PM pcw--home: Mesaflash 25 pin pinoust are DB25 pinouts (as all breakouts use DB25)
08:13 PM pcw--home: "none" in the readhmid listing means no secondary function (so GPIO only)
08:13 PM tcurdt: ok ... that clears things up :)
08:14 PM tcurdt: wish mesaflash would list it like it's in the manual ... or call it DB25 pin ... that would have caused less confusion
08:17 PM pcw--home: all in one boards show both:
08:17 PM pcw--home: IO Connections for P1 -> 7I96_2
08:17 PM pcw--home: Pin# I/O Pri. func Sec. func Chan Sec. Pin func Sec. Pin Dir
08:17 PM pcw--home: P1-01/DB25-01 34 IOPort QCount 1 Quad-A (In)
08:17 PM pcw--home: P1-02/DB25-14 35 IOPort None
08:17 PM pcw--home: P1-03/DB25-02 36 IOPort None
08:17 PM pcw--home: P1-04/DB25-15 37 IOPort None
08:17 PM pcw--home: P1-05/DB25-03 38 IOPort None
08:18 PM pcw--home: P1-06/DB25-16 39 IOPort None
08:18 PM pcw--home: P1-07/DB25-04 40 IOPort None
08:18 PM XXCoder: whoa
08:18 PM XXCoder: try to use dpaste when more than 3-4 lines
08:19 PM tcurdt: pcw--home I got the 7i92 ... I wasn't lucky enough to find a 7i96 :-/
08:19 PM tcurdt: it only shows a number
08:19 PM pcw--home: Yeah 7i96's are maybe 1 year out
08:19 PM XXCoder: roycroft: just finished 40mm version of the joinery lol
08:20 PM pcw--home: 7I96S's are available though
08:21 PM * tcurdt searches for the difference of the "S"
08:23 PM pcw--home: The 7I92 manual shows both DB25 and header pinouts
08:23 PM tcurdt: pcw--home yes, the manual is fine
08:24 PM tcurdt: but (my) mesaflash does not show both
08:26 PM * tcurdt is a little confused on the 7i96S details ... doesn't look like daughter card
08:28 PM pcw--home: the 7I96S is like a 7I96 (built with an available FPGA) but adds analog spindle output and better ESD protection for Plasma use
08:29 PM Tom_L: pcw, that board should arrive friday so it says
08:31 PM Tom_L: are both the 96's using the newer FPGAs?
08:32 PM pcw--home: Only the 7I96S
08:36 PM * RoboticsNut[m] uploaded a video: (16408KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/bWXTwgOlHsIrhjIvIBeDIysk/IMG_5818.MOV >
08:36 PM RoboticsNut[m]: This is the motor running from the Test/Tune in pncconf
08:36 PM RoboticsNut[m]: which looks like it doesn't play.
08:38 PM * RoboticsNut[m] uploaded a video: (1921KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/dODbyetmxfcxOnsLQGQhXNUx/IMG_5818.MOV >
08:38 PM RoboticsNut[m]: LinuxCNC Test/Tune pncconf
08:39 PM RoboticsNut[m]: P=50, all other terms are zero.
08:39 PM RoboticsNut[m]: controller is in the PDFF mode
08:40 PM pcw--home: for velocity mode drives most of the "tuning" is just setting FF1
08:42 PM pcw--home: FF1 should be 1.00 IF the analog output is scaled in machine units/second
08:43 PM RoboticsNut[m]: the PDFF mode of the controller has a FF term
08:44 PM RoboticsNut[m]: if FI is set to one it turns into a PI controller
08:45 PM pcw--home: I would expect the drive to be set into velocity mode
08:46 PM RoboticsNut[m]: yes it is in velocity mode as best I can tell.
08:47 PM RoboticsNut[m]: kvi=10000, kv=2500, and kvfr=10
08:48 PM tcurdt: pcw--home but why is the 7i96S listed as daughter board?
08:49 PM tcurdt: pcw--home I now got the 7i92 and it allows me to have 5 stepgens, 6 outputs, 18 inputs ... which is pretty great ... so I will probably stick to that
08:50 PM tcurdt: plus availability in the EU is another matter ... and shipping from the US isn't exactly cheap
08:52 PM pcw--home: Eusurplus should have some 7I96S's soon
08:52 PM pcw--home: If you send me the desired pinout, I'll make a 7I92 bitfile that matches
08:56 PM * RoboticsNut[m] uploaded an image: (6746KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/xJlSVdvFCwpKlSAjoImIyami/C6352181-606D-4B64-B1D0-C3D9B05DB49F.jpg >
08:56 PM tcurdt: pcw--home that would be great <3 ... just 5 stepgens with slightly different pins and the rest GPIO https://gist.github.com/tcurdt/f9949af939eb9e0a8ec812234021e526
08:56 PM RoboticsNut[m]: I'm graphing POS-cmd vs POS-fb
08:56 PM RoboticsNut[m]: but having trouble getting scaling.
09:07 PM * RoboticsNut[m] uploaded an image: (3833KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/WzDubFGIthcbbEmpKJnVsfXm/E091A713-6123-43F8-A033-141CD7E6036D.jpg >
09:07 PM RoboticsNut[m]: if I'm reading this right, I'm getting good performance.
09:08 PM RoboticsNut[m]: its so exact that I'm skeptical
09:10 PM pcw--home: tcurdt: Ill try and get this done by tomorrow
09:19 PM RoboticsNut[m]: not sure precisely which signals I should be looking at right now.
09:25 PM pcw--home: following error, commanded velocity
09:25 PM * RoboticsNut[m] uploaded an image: (4900KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/mPaTtdJqOapIjRqLzSTvZRdt/8B7C1814-D279-4AF9-ADEE-2CE172CADC87.jpg >
10:02 PM * RoboticsNut[m] uploaded an image: (5558KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/qbkUdGPmjOxUyQOJwgNJYoyP/09073368-240B-48ED-83DA-BB3F6227B02F.jpg >
10:02 PM RoboticsNut[m]: i get something like this
10:08 PM pcw--home: Error in the direction of velocity means FF1 is too low IICRC
10:17 PM RoboticsNut[m]: Thank you pcw--home, I will have to look at this again, to frustrated for now.
10:31 PM roycroft: so i went to check out our booth at the fair today, but in spite of that, i got 4 drawers installed, 4 drawers glued-up, as per usual
10:31 PM roycroft: but as well, 4 sets of drawer sides prepped for finish, and four of the systainer bottoms half finished
10:31 PM XXCoder: roy wish me luck. printing 40mm version of japan joinery lol
10:31 PM roycroft: i hope it works out well
10:31 PM roycroft: it's a really neat project
10:31 PM roycroft: in fact, i'm thinking of making that joint when i have a bit of spare time
10:32 PM XXCoder: nice :) I do have freecad model if you want to make, rather than design and make
10:32 PM roycroft: i'll make mine out of wood, though
10:32 PM XXCoder: i have 400mm and 40mm versions
10:32 PM roycroft: well, the dimensions are all published
10:32 PM XXCoder: yeah, though not very easy to read
10:32 PM roycroft: i have measuring devices and knives
10:33 PM XXCoder: and tip for you, design is draw on end that is cut already
10:33 PM XXCoder: ie sketch on slopes for female, while male I guess you need to do it differently
10:36 PM XXCoder: on male, I did a sketch on slopes, then did a loft
10:38 PM XXCoder: cant build up in real wood so dunno how it'd be designed lol. video would help I guess
10:38 PM roycroft: i'll need to study it a bit more
10:39 PM XXCoder: yeah. took me quite a while to get everything right
10:39 PM roycroft: it seems doable, although it's only been done once ever on something real
10:39 PM roycroft: i wonder if there's a reason for that
10:39 PM XXCoder: theres so many japanese joinery
10:39 PM XXCoder: book lists in least 20?>
10:40 PM roycroft: yes, but there are hundreds of thousands of japanese buildings done in traditional styles
10:40 PM roycroft: that's the only traditional joint i've ever seen that's only documented use is one time
10:41 PM XXCoder: oh really? one time?
10:41 PM XXCoder: so interesting
10:41 PM roycroft: yes
10:41 PM roycroft: that one has only been used once that anyone has found
10:41 PM roycroft: other than reproducing it
10:43 PM * xxcoder8265[m] uploaded an image: (81KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/BFOtUfjnYpPIrgfYXdkpnnaF/unknown.png >
11:20 PM XXCoder: roycroft: success lol
11:20 PM XXCoder: slides in easily
11:20 PM roycroft: awesome
11:20 PM XXCoder: one and ONLY way to put together and apart
11:22 PM XXCoder: I think clearance can be tightened some more, but as it is, it hardly rattles at all
11:22 PM XXCoder: just a tiny bit
11:23 PM roycroft: there's a lot of mating surface area