#linuxcnc Logs
Sep 15 2022
#linuxcnc Calendar
12:00 AM XXCoder: theres small classes of shapes thats impossible outside printing but yeah
12:09 AM XXCoder: really depends on what want to do
12:09 AM CaptHindsight[m]: yeah and how fast and how many and what size
12:09 AM XXCoder: yeah
12:09 AM XXCoder: it could be protoyping too
12:09 AM XXCoder: see if shape works before doing cast
12:09 AM CaptHindsight[m]: but clay is dirt cheap :)
12:09 AM CaptHindsight[m]: I can't cast glass pretty quick
12:09 AM XXCoder: but molds isnt that cheap though not like its made from gold
12:09 AM CaptHindsight[m]: heh I can
12:09 AM CaptHindsight[m]: getting tired
12:09 AM CaptHindsight[m]: I can print the molds really fast
12:09 AM CaptHindsight[m]: i have a synthetic sand that melts well above glass and steel temps
12:09 AM XXCoder: nice
12:09 AM XXCoder: sls correct?
12:09 AM CaptHindsight[m]: no, more like a resin printer
12:10 AM XXCoder: interesting
12:10 AM CaptHindsight[m]: ceramics also get into all the glases
12:11 AM CaptHindsight[m]: and they are actually types of glasses
12:11 AM CaptHindsight[m]: ceramic with glass coating
12:11 AM roycroft: it's just fancy sand
12:11 AM CaptHindsight[m]: maybe you have seen ceramics with a dull finish after firing
12:12 AM CaptHindsight[m]: then they coat with a glaze
12:12 AM CaptHindsight[m]: the glaze, sorry for earlier spelling
12:12 AM CaptHindsight[m]: is made of glass powder
12:13 AM CaptHindsight[m]: this get fired and melts a layer of glass over the ceramic giving it that clear hard shiny finish
12:14 AM CaptHindsight[m]: yeah fancy sand :)
12:15 AM XXCoder: we are using essentally thinky sand to talk with each other
12:15 AM XXCoder: *using
12:16 AM Jeremy: has there been any progress on s-curve/jerk limiting motion planning ?
12:18 AM CaptHindsight[m]: Jeremy: there a long thread on the forums
12:18 AM CaptHindsight[m]: Jeremy: and since what point in time?
12:20 AM Jeremy: I guess I was hoping for a tl;dr of it's status as the thread is quite long. I'll dig through it
12:21 AM Jeremy: its*
12:33 AM CaptHindsight[m]: Jeremy: in the past few years the trajectory planner got better and deeper
12:33 AM CaptHindsight[m]: there is a Swiss fork of LCNC that focused on jerk limiting
12:34 AM CaptHindsight[m]: but after a long discussion on the maillist I'm not sure what thye really did mattered
12:34 AM CaptHindsight[m]: but it works for them
12:35 AM CaptHindsight[m]: Jeremy: https://forum.linuxcnc.org/show-your-stuff/37855-announcement-new-fork-opencn
12:36 AM Jeremy: very interesting
12:36 AM CaptHindsight[m]: Jeremy: if you want to know more I'd ask on the mail list
12:36 AM CaptHindsight[m]: and too long forum discussion
12:36 AM CaptHindsight[m]: :)
12:38 AM Jeremy: it seems like it's still active, last push 23 hrs ago
12:58 AM CaptHindsight[m]: Stuart had a long discussion with the Swiss group
12:59 AM CaptHindsight[m]: he did one of the if not the largest machine port to LCNC with his 5 axis Cincinnati
12:59 AM CaptHindsight[m]: he never had any jerk issues
12:59 AM XXCoder: 2019 and few rplies later 2021
01:01 AM CaptHindsight[m]: I handle my 5-6-7 axis jerk issues with my accel decall rates in config
01:01 AM * roycroft chooses his friends carefully and avoids jerks
01:06 AM XXCoder: wow didnt know you could be friends with 4th derivative
01:06 AM jpa-: i assume that the more rigid machine, the less jerk matters? because it is change of acceleration = change of force
01:09 AM CaptHindsight[m]: the machine jerk issues I usually see are just related to the design of the machine
01:09 AM CaptHindsight[m]: toys vs machines
01:09 AM CaptHindsight[m]: routers trying to be mills
01:10 AM XXCoder: all machines, from toy mills to giants, have limitions. its up to programmer to figure what is good to use
01:11 AM CaptHindsight[m]: personal responsibilty except when things go bad :)
01:11 AM XXCoder: success has many parents. failure is a widow
01:11 AM XXCoder: errm ophian
01:12 AM CaptHindsight[m]: then it's not my fault, like a 5 year old
01:12 AM XXCoder: orphan
01:12 AM CaptHindsight[m]: the current trend
01:24 AM CaptHindsight[m]: https://www.mmsonline.com/articles/understanding-jerk-control
01:26 AM CaptHindsight[m]: https://www.linearmotiontips.com/how-to-reduce-jerk-in-linear-motion-systems/
01:26 AM Jeratan[m]: i think we're making some oversimplifications here. unfortunately we live in a universe where instant acceleration doesn't occur, hence why even Kerns etc have jerk control. sure, you can lower accelerations but ultimately that's a band-aid
01:28 AM CaptHindsight[m]: cnc routers do it by using undepowered steppers
01:28 AM jpa-: jerk is always limited by hardware due to physics (it takes time for motor to change torque and also time for force to propagate through components), so the question is when it is beneficial to further limit it by software
01:29 AM jpa-: which would be either when the machine will resonate if driven with the jerk levels the motors are able to generate, or when the multi-axis path accuracy is limited by the jerk inaccuracies in acceleration
01:33 AM jpa-: first one could be solved by per-axis filtering with e.g. a HAL component; i think klipper does such acceleration compensations
01:33 AM jpa-: the second one definitely requires changes to path planning
01:34 AM Jeratan[m]: jerk is always limited by physics, regardless of the hardware. the only thing you can change is whether your motion planning is aware of it or not
01:36 AM Jeratan[m]: ultimately it just comes down to what the required tolerances are. for probably 99.99% of cases having 4th derivative control or not isn't going to make or break it
01:41 AM CaptHindsight[m]: PC's have more than enough computational speed to per form S-curve accel as well as other non-linear curves
01:42 AM CaptHindsight[m]: I'd have to dig back into the original source and follow the evolution to see why they only did what they did
01:42 AM randy: morning
01:43 AM CaptHindsight[m]: when EMC was open sourced it was a work in progress
01:44 AM XXCoder: dont think we have enough devs now?
01:44 AM XXCoder: not too sure
01:44 AM CaptHindsight[m]: no we don't
01:45 AM XXCoder: no paid devs for lcnc?
01:45 AM CaptHindsight[m]: plus PLL/servo has always been on the edge of black magic when it comes to adding non-linear filtering to the loop
01:46 AM CaptHindsight[m]: XXCoder: Tormach has paid devs
01:46 AM XXCoder: they do upstream commits correct?
01:46 AM CaptHindsight[m]: and they have hired outside devs as well
01:47 AM CaptHindsight[m]: it'd all open like path pilot but I don't know how much gets pushed back to LCNC
01:47 AM CaptHindsight[m]: the whole machinekit fork was a big waste of resources
01:48 AM CaptHindsight[m]: but that's open source too often, lack of cooperation
01:48 AM XXCoder: do people still work on that or dead?
01:49 AM CaptHindsight[m]: https://github.com/machinekit/machinekit-hal
01:49 AM Deejay: moin
01:49 AM XXCoder: 7 months latest
01:49 AM CaptHindsight[m]: https://github.com/zultron/hal_ros_control/
01:50 AM CaptHindsight[m]: ROS- LCNC bridge
01:50 AM CaptHindsight[m]: https://github.com/tormach
01:50 AM XXCoder: year old since update
01:51 AM CaptHindsight[m]: didn't Tormach fund the latest trajectory planner?
01:52 AM CaptHindsight[m]: they developed that and it went back into LCNC as well
04:40 AM Tom_L: morning
04:52 AM JT-Cave: morning
05:14 AM -!- #linuxcnc mode set to +v by ChanServ
05:23 AM JT-Cave: damn I installed a driver for my realtek nic and the pcie one stopped working
05:24 AM JT-Cave: john@cave:~$ nmcli device status
05:24 AM JT-Cave: DEVICE TYPE STATE CONNECTION
05:24 AM JT-Cave: enp9s0 ethernet connected Wired connection 1
05:24 AM JT-Cave: lo loopback unmanaged --
05:24 AM JT-Cave: 07:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)
05:24 AM JT-Cave: 09:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8125 2.5GbE Controller (rev 04)
05:24 AM JT-Cave: lspci shows both nics...
05:31 AM -!- #linuxcnc mode set to +v by ChanServ
09:48 AM -!- #linuxcnc mode set to +v by ChanServ
09:48 AM -!- #linuxcnc mode set to +v by ChanServ
10:18 AM JT-Shop: https://www.woodpeck.com/indexable-combination-square-system-components.html
10:28 AM roycroft: i like the tab on the end of the square head, but i would not find the indexing feature useful
10:30 AM roycroft: i have a bunch of incra rules, that index to 1/64" or 0.25mm, and those are great
10:34 AM JT-Shop: the incra t rule looks pretty wide but can't tell from the photo
10:35 AM roycroft: there are several different sizes, with differing lengths and widths
10:36 AM roycroft: and i think the us customary ones only index to 1/32", not 1/64"
10:36 AM roycroft: i don't use a combination square for the kind of indexing the woodpeckers one is designed to do
10:36 AM JT-Shop: looks like only the wide ones can be used with the T
10:36 AM roycroft: i use incra rules for that
10:37 AM roycroft: i use a combination square when i need to scribe an exact line
10:37 AM roycroft: i.e. i index the end of the rule off whatever is the distance i need to scribe
10:37 AM roycroft: and scribe with the end
10:38 AM roycroft: i'm not saying the woodpeckers one is bad
10:38 AM roycroft: i just don't see how it would make sense in my workflow
10:39 AM roycroft: i have a few woodpeckers tools, and i like the ones i have a lot
10:44 AM perry_j1987: mornin
10:45 AM JT-Shop: the incra t rule looks like something I can use and much more cost effective than the woodpecker
10:58 AM roycroft: they work great
11:01 AM -!- #linuxcnc mode set to +v by ChanServ
11:14 AM Tom_L: JT-Shop, anytime
11:35 AM JT-Shop: you dev or L
11:35 AM Tom_dev: over here
12:02 PM -!- #linuxcnc mode set to +v by ChanServ
12:53 PM -!- #linuxcnc mode set to +v by ChanServ
12:55 PM perry_j1987: hows everyone
01:06 PM JT-Shop: I was doing great till the bp stopped working
01:11 PM perry_j1987: what happened
01:13 PM XXCoder: nonworking driver
01:20 PM perry_j1987: ah
01:20 PM perry_j1987: ok heading to shop bbiab
01:25 PM Alicrow-phone: I got a few of the Incra rules for Christmas last year. They're great.
01:31 PM roycroft: jt-shop: i also have some mechanical pencils whose tip flexes/swivels to prevent breaking the lead, and they are great for layout work, especially with the incra rules
01:32 PM roycroft: https://www.amazon.com/gp/product/B00NN3YZ1G
01:33 PM XXCoder: roy I remember that one where it rotates each time its lifted
01:33 PM XXCoder: so sharp point never happens
01:33 PM XXCoder: well unless you constantly write for a bit without lifting but yeah
01:38 PM roycroft: no, not that one
01:38 PM roycroft: that's another style
01:38 PM roycroft: many of the japanese pencils do that
01:38 PM roycroft: this one just has a swivel tip
01:39 PM XXCoder: I know its different. just saying it reminds me of that fancy one lol
01:39 PM roycroft: and since i rarely use 0.5mm pencils for writing, i don't need the rotational tip ones
01:39 PM roycroft: i mostly hold the pencil vertical or close to vertical
01:39 PM XXCoder: but yeah not bad. yeah rotating one is more effective on larger diameter ones
01:39 PM roycroft: well
01:40 PM roycroft: perpendicular to the part being marked
01:40 PM roycroft: which is not always horizontal :)
01:40 PM XXCoder: yeah angle of surface to pen
01:40 PM roycroft: 0.5mm leads break very easily if you apply the slightest side pressure to the pencil
01:40 PM roycroft: and the ones i linked above avoid that
01:41 PM roycroft: it's actually hard to break the lead with those
01:41 PM XXCoder: my bro preffered .9mm if I remember right. I like smaller one but forgot size
01:41 PM roycroft: well the incra rules only work with 0.5mm pencils
01:41 PM XXCoder: checked my old college pencil. its .5mm
01:42 PM XXCoder: its NOS lol over 10 years old. I cant find ones I acually used though college
01:43 PM JT-Shop: hal show says the 7i77 output is on but I get 0v on the output
01:45 PM roguish[m]: JT-Shop: just a little notice. some of your debian download links in you install instructions are broken. the 11.4 links.
01:46 PM JT-Shop: aye rod told me
01:46 PM JT-Shop: I need to put the page instead of the deb
01:46 PM JT-Shop: hmm field power is ok...
01:46 PM roguish[m]: I'm doing a VM for 12. starting with 11.5
01:50 PM JT-Shop: hmm no outputs seem to work
01:51 PM XXCoder: anything else changed when you installed driver?
01:53 PM JT-Shop: pcw-home, any ideas on what to check if no outputs work on the 7i77 and field power is 24vdc?
01:56 PM JT-Shop: XXCoder, who are you asking?
01:56 PM XXCoder: you
01:56 PM JT-Shop: the realtek driver?
01:57 PM XXCoder: yeah from what I understand, it stopped working after that?
01:58 PM JT-Shop: the pcie nic shows still there with nmcli device status
01:58 PM JT-Shop: but if address or something like that it does not show up
01:59 PM JT-Shop: no that's backward lspci shows both nicks but nmcli device status it's not there
01:59 PM JT-Shop: I reinstalled the driver from debian but no change
01:59 PM XXCoder: weird. I doubt its as simple as dropping old card
02:00 PM XXCoder: (driver support of models)
02:01 PM JT-Shop: going to pull the card and see if it's seen on another pc in a bit
02:03 PM XXCoder: ok
02:05 PM roguish[m]: didn't see how to download a 12 .iso
02:05 PM TurBoss: JT-Shop: Hello, does `ip a` show the interface?
02:06 PM JT-Shop: roguish[m], there is no 12 iso
02:06 PM JT-Shop: TurBoss, let me see
02:08 PM roguish[m]: no wonder I couldn'g find it
02:09 PM JT-Cave: TurBoss, ip a just shows the one on the motherboard
02:09 PM TurBoss: ok no luck
02:12 PM -!- #linuxcnc mode set to +v by ChanServ
02:17 PM -!- #linuxcnc mode set to +v by ChanServ
02:19 PM JT-Shop: TurBoss, the pcie nic is seen on this pc with ip a
02:19 PM JT-Shop: so it's not broken, just my dev pc :(
02:20 PM JT-Shop: and amazing enough I pulled the plug on the bp and just powered it up and it works dunno what happened
02:25 PM XXCoder: weird
03:09 PM Tom_L: JT-Shop, both should be there now
03:22 PM JT-Shop: did you see I pushed a while ago?
03:22 PM JT-Shop: https://github.com/jethornton/mesact/commit/8433b10e3517cd8128911174db89adac23aff07a
03:39 PM XXCoder: CaptHindsight[m]: https://www.reddit.com/r/3Dprinting/comments/xf61gk/rate_my_ceramic_glass_benchy_fresh_out_the_kiln/
03:39 PM XXCoder: ceramic glass benchy
03:40 PM roguish[m]: JT-Shop: hey, i'm in my debian 12 VM. finished the upgrade and all. now trying to install linuxcnc............. when I try to install it errors with 'trying to remove the kernel you're running' linux-image-5.19.0-1-rt-amd64-unsigned (5.19.6-1)
03:41 PM roguish[m]: what gives?
03:41 PM roguish[m]: errors whether using synaptic or apt install
03:45 PM JT-Cave: I'm running 5.18.0-3-rt-amd64
03:45 PM JT-Cave: maybe lcnc don't know about that kernel and that's why I install the rt kernel first
03:46 PM JT-Cave: too bad 5 doesn't work with mesa ethernet cards
03:53 PM CaptHindsight[m]: XXCoder: they used Tethon resin https://tethon3d.com/
03:53 PM XXCoder: wow pricy
03:53 PM XXCoder: cool though
03:54 PM CaptHindsight[m]: they have a patent that cover a resin that is not a resin but works as a resin when you add ceramic particles
03:54 PM CaptHindsight[m]: so you are still free to make and sell ceramic filled resin
03:55 PM CaptHindsight[m]: just don't sell a resin without ceramic particles that you add ceramic particles to later :)
03:55 PM XXCoder: easy to get around that patent
03:55 PM XXCoder: sell mixer and jar of particles
03:56 PM CaptHindsight[m]: I talked to them a few years ago at a trade show before their patent was approved
03:56 PM CaptHindsight[m]: I guess they couldn't get a patent on a resin that also has ceramic partciles
03:58 PM CaptHindsight[m]: https://www.youtube.com/watch?v=b65lfCUjpZ0
03:59 PM CaptHindsight[m]: the comments are funny
04:00 PM CaptHindsight[m]: What's the electrical conductivity / resistance like for these samples?
04:00 PM CaptHindsight[m]: This resin is still in beta and we have not done these tests as of yet.
04:00 PM CaptHindsight[m]: I'll loan you my multi-meter for a few seconds
04:01 PM CaptHindsight[m]: just from watching the video i can see it's hardly conductive at all
04:02 PM CaptHindsight[m]: they said they can sinter it so it must shrink a bit since there is so much resin
04:10 PM XXCoder: strunkage is a must anyway from wat I understand. even if resin is a lot less
04:10 PM XXCoder: I wonder if that plastic injectable glass process will become popular
04:12 PM XXCoder: https://youtu.be/4jzWPcU80Ig seems theres video of it now
04:48 PM roguish[m]: hey, have any of you guys done any 3D scanning???
04:49 PM XXCoder: not me sadly
04:50 PM roguish[m]: i just picked up a couple of MS Kinects. a V1 and a V2. gonna give it a try.
04:51 PM XXCoder: nice
04:51 PM roguish[m]: hopefully help a friend make some prosthetic parts.
04:54 PM roguish[m]: after I get the scanning figured out, then it's on to the 3d printing. or as it's called now 'additive manufacture' gotta have funky name for everything nowadaze
04:59 PM CaptHindsight[m]: was always called additive manufacturing, 3d printing was hijacked by FDM/FFF printing, it also used to just mean additive manufacturing of 3d parts
05:00 PM CaptHindsight[m]: roguish: there are a bunch of open source tools for scanning
05:01 PM CaptHindsight[m]: https://openkinect.org/wiki/Main_Page
05:02 PM roguish[m]: yeah, all over them.....
05:02 PM roguish[m]: but haven't seen that one. thanks.
05:07 PM JT-Shop2: I guess I should rip the 1x8x10 now... all my blades will need cleaning after cutting all this PT
05:10 PM Tom_L: JT-Shop2, no that must have been during naptime
05:15 PM JT-Shop2: I pushed an updated build date a couple of hours ago
05:15 PM Tom_L: i just grabbed it
05:16 PM JT-Shop2: about 4 hours ago I fixed an indention error
05:16 PM Tom_L: figuring out how to sync my local copy with my fork
05:16 PM JT-Shop2: ok, got that sorted out?
05:16 PM Tom_L: i think so
05:16 PM JT-Shop2: local copy on the rpi?
05:17 PM Tom_L: right
05:17 PM JT-Shop2: did you clone your repo on the rpi?
05:17 PM Tom_L: yes
05:17 PM Tom_L: i think i got it
05:17 PM JT-Shop2: git pull --rebase
05:17 PM Tom_L: need to branch reset first
05:18 PM Tom_L: i get uncommited changes otherwise
05:18 PM JT-Shop2: gotta commit your changes first
05:18 PM JT-Shop2: or stash them
05:18 PM Tom_L: i updated my fork on git on the main pc
05:19 PM Tom_L: so i had to reset the rpi one
05:19 PM JT-Shop2: I still remember the days when jepler1088[m] would have fun trying to teach me git :)
05:19 PM Tom_L: then git pull --rebase
05:19 PM Tom_L: sound ok?
05:19 PM JT-Shop2: I've never had to reset a clone
05:20 PM roycroft: i wrote a one page git cheat sheet for my boss to do the very basic things
05:20 PM roycroft: he could not comprehend it at ll
05:20 PM roycroft: all
05:20 PM Tom_L: you've probably never had 2 copies going at once either
05:20 PM JT-Shop2: if you commit on local clone then you can pull
05:20 PM JT-Shop2: I have a git-secrets.txt file :)
05:20 PM roycroft: which i don't undestand - the basics of git are very straightforward
05:20 PM Tom_L: i'm starting one
05:20 PM JT-Shop2: easy when you know how...
05:21 PM roycroft: git clone, git checkout, git status, git add, git commmit, git push
05:21 PM Tom_L: like everything else
05:21 PM roycroft: those are the basics
05:21 PM JT-Shop2: yup
05:21 PM JT-Shop2: git gui is easier
05:21 PM roycroft: i'm fairly late at using git
05:21 PM roycroft: i used sccs back in the '80s, and then moved to rcs when it came out
05:22 PM roycroft: and later cvs
05:22 PM roycroft: which is built on rcs
05:22 PM roycroft: i used rcs/cvs for a very long time, and i still prefer rcs for maintaining server configuration files
05:22 PM roycroft: but the world has moved on to git, and git does have some nice features that are lacking in rcs/cvs
05:22 PM JT-Shop2: I think the problem is because i do 'force pushes'
05:22 PM JT-Shop2: and you use 'git pull' and that is a bad match.
05:22 PM JT-Shop2: Everyone learns different ways to use git but i have learned to
05:22 PM JT-Shop2: never,never,never use 'git pull' and only use 'git merge' on
05:22 PM JT-Shop2: rare occasions.
05:22 PM JT-Shop2: comment from dewey
05:22 PM roycroft: i skipped subversion
05:23 PM roycroft: i don't think i missed much by doing so
05:23 PM JT-Shop2: resolving a conflict
05:23 PM JT-Shop2: <cradek> jthornton: I use "git config --global merge.conflictstyle diff3" which I think makes it much easier to understand conflicts
05:24 PM JT-Shop2: I never understood how to resolve conflicts lol
05:24 PM Tom_L: most ppl in life don't
05:24 PM JT-Shop2: <seb_kuzminsky> a while ago i convinced git to use meld in diff3 mode, works great if you're into that whole "gooey" thing the kids do nowdays
05:25 PM JT-Shop2: so 274 lines of help in my git-secrets.txt file lol
05:25 PM JT-Shop2: In other situations, for example when a commit was made on a newer branch
05:25 PM JT-Shop2: (let's say 2.8) but really belongs on an older branch (2.7), cherry-pick is the
05:25 PM JT-Shop2: correct answer. The work flow in this situation would be to cherry-pick the
05:25 PM JT-Shop2: commit from 2.8 to 2.7, then merge 2.7 back into 2.8 to address any merge
05:25 PM JT-Shop2: conflicts that might arise.
05:25 PM * roycroft finds that, while he for the most part has nothing against what the kids do nowadays, he also for the most part isn't too interested in much of what the kids do nowadays :)
05:26 PM JT-Shop2: that's for sure
05:26 PM Tom_L: wonder what our parents/grandparents said about us
05:27 PM roycroft: the first thing i'm not interested in is listing my pronouns and comprehensively listing what i identify as before speaking with a random stranger
05:27 PM roycroft: the same thing, tom_l
05:27 PM Tom_L: with that said, i've talked to quite a few ppl about 'this' generation and we all agree they're doomed
05:27 PM roycroft: i don't think they're doomed
05:27 PM roycroft: they have a lot going for them
05:27 PM Tom_L: or dumbed
05:28 PM XXCoder: roy someone else posted picture of gst3d filament. it was 1.93mm
05:28 PM XXCoder: .2 higher than spec.
05:28 PM * xxcoder[m] uploaded an image: (10848KiB) < https://libera.ems.host/_matrix/media/r0/download/jauriarts.org/NydHsoxEvjxadMEaoLHFKbFo/unknown.png >
05:30 PM JT-Shop2: I assume the grandparents I knew thought I was a good boy and a hard worker, every summer we would go to NC and I'd work the whole time on the farm
05:30 PM JT-Shop2: Clarence's brother thought I was shit for kicking loose hay down from the barn
05:31 PM roycroft: i should have the final decision on my disputes within a week or so
05:31 PM JT-Shop2: from your CC company
05:31 PM JT-Shop2: or PP
05:31 PM roycroft: from my cc issuer
05:32 PM * roycroft opened the disputes on july 1st
05:33 PM XXCoder: i checked gst fb page. last posts is july 28
05:33 PM XXCoder: silient since. wonder if company is collipising
05:33 PM roycroft: possibly
05:34 PM roycroft: that would be good for me, i think, as it means i would get a default judgement if they don't respond again
05:34 PM XXCoder: seems illdmax and gst3d split off and former got good stuff and latter just cheapened too much
05:34 PM XXCoder: hope so
05:34 PM roycroft: i'm about 99+% confident i'll get my money back for the order they never shipped
05:34 PM roycroft: and about 60% confident i'll get my money back for the defective order
05:35 PM roycroft: but i'm prepated to eat both of the charges, if it comes to that
05:35 PM roycroft: i'll be grumpy about it, but prepared
05:35 PM XXCoder: can always try print em though measure if its reasonable range of size unlike picture above wow
05:35 PM XXCoder: sadly my current rolls is last good rolls from them from what I understand on timing
05:35 PM roycroft: i need more grey filament
05:35 PM roycroft: not too much - a couple kg or so
05:36 PM roycroft: i still have 9kg of the gst3d stuff
05:36 PM roycroft: if i'm stuck with it i may try printing one box at a time
05:36 PM roycroft: it would be trash anyway
05:36 PM Tom_L: JT-Shop2, any easy way to spot your change when running the code?
05:36 PM roycroft: but i'm hoping i just get my money back
05:36 PM XXCoder: indeed I would hope so also
05:36 PM roycroft: in which case it goes straight to the bin
05:36 PM XXCoder: eh I would try print ones that conform on diameter
05:36 PM JT-Shop2: the last fix?
05:37 PM Tom_L: JT-Shop2, woopsie...
05:38 PM roycroft: yeah, i would probably open each reel and measure it before binning it
05:38 PM roycroft: but that picture i posted is pretty horrible
05:38 PM XXCoder: what was filament diameter on that horrible roll
05:38 PM roycroft: if any of the other reels look like that i think it's a total waste of time to try printing them
05:38 PM roycroft: iirc it varied from 1.70 to 1.75mm
05:39 PM XXCoder: varance is just too great there
05:39 PM XXCoder: should be .02+-
05:39 PM roycroft: yes
05:40 PM roycroft: i've gone through about 15kg of atomic filament since then, and have not had a single problem
05:40 PM XXCoder: atomic definitely good stuff
05:40 PM -!- #linuxcnc mode set to +v by ChanServ
05:41 PM JT-Shop2: that's all I use
05:42 PM roycroft: i can't afford the cheap stuff, as i found out
05:42 PM JT-Shop2: lol
05:42 PM roycroft: besides, gst3d have never sent me even a single gummi bear
05:42 PM XXCoder: that's a crime, not sending candy
05:42 PM JT-Shop2: there's your sign
05:43 PM CaptHindsight[m]: https://www.youtube.com/watch?v=O4cWHPjlRgU
05:44 PM CaptHindsight[m]: no longer need to be concerned with filament diameter
05:44 PM roycroft: i don't think filament diameter is the problem with what i got
05:44 PM roycroft: i think it was just crappy filament
05:45 PM roycroft: the varying diameter is just a symptom
05:45 PM CaptHindsight[m]: a gear pump embedded into the heat block of the hotend, the hotend independently supplies the pressure required to extrude molten plastic.
05:45 PM XXCoder: capt interesting but meh also
05:46 PM roycroft: https://roycroft.us/gst3d.jpeg
05:46 PM roycroft: that's what they sent me
05:46 PM JT-Shop2: you need a extruder screw not a gear pump
05:46 PM JT-Shop2: a gear pump won't pump solids
05:46 PM roycroft: i need a chipper and a hopper for that crap
05:46 PM roycroft: and it proably still would not work well
05:47 PM XXCoder: im not too sure also because whole gear area need to be heated, and stepper would get heat also
05:48 PM CaptHindsight[m]: the gear pump needs to heat the filament to above its melting point so that it flows
05:48 PM * roycroft thinks just purchasing from a realiable vendor is the easiest thing to do
05:56 PM CaptHindsight[m]: https://www.youtube.com/watch?v=a6XRs-g6ngw
06:03 PM roguish[m]: JT-Shop2: got the VM going. lcnc installed and runs in sim.. installed 5.18.2.1
06:04 PM JT-Shop2: good
06:04 PM * JT-Shop2 listens to the end of king crimson and wanders inside for the night
06:04 PM roguish[m]: had to add a bullseye repository. no big deal.
06:06 PM JT-Shop2: did you use my instructions to install lcnc on 12?
06:06 PM roguish[m]: JT-Shop2: try some new Bluegrass. Molley Tuttle she's really young, from over in Palo Alto, next to Stanford University, like 30 miles south of SF
06:06 PM roguish[m]: yes, been bouncing around in there.......as needed.
06:07 PM JT-Shop2: no e in Molly
06:08 PM roguish[m]: you're right
06:09 PM roguish[m]: Molly T is good background music for desk work.
06:10 PM JT-Shop2: she is pretty good I like Alison Krauss
06:10 PM roguish[m]: they're close...
06:10 PM JT-Shop2: yup
06:10 PM * JT-Shop2 wanders inside
06:10 PM roguish[m]: your instructions are very handy. thanks.
06:14 PM roycroft: tonight is hamilton night
06:14 PM * roycroft has been waiting a long time to see hamilton
06:17 PM roguish[m]: What, no more Queen funerals?
06:21 PM roycroft: freddy mercury passed years ago
06:21 PM roycroft: i think those are over for a while
06:22 PM roycroft: unless you know something about brian may that the rest of us don't
06:24 PM roguish[m]: funny
06:25 PM JT-Cave: Tom_L, just pushed a fix for lcnc version check
06:48 PM Alicrow-bridget: Seems it takes about 0.01s from applying PWM to actually reading a position change from the linear scales. Anyone know whether this is typical, or what might be causing it?
06:55 PM CaptHindsight[m]: what hardware?
06:57 PM Alicrow-bridget: 7i80HD with 7i40 going to an old servo motor
07:03 PM CaptHindsight[m]: I have to go but I have more questions , such as how are you measuring the latency? applying PWM from where to where, what is the aceel set to, etc etc
07:13 PM Alicrow-bridget: I'm measuring the latency by comparing signal times in HALscope. PWM is from the 7i80HD to the 7i40 and then to the motor. And I've done this with the PNCconf open loop test, so no PID at all, just applying a decent amount of PWM immediately.
07:18 PM Tom_L: JT-Cave, building
07:28 PM Alicrow-phone: Probably should get around to replacing that bad RAM stick in the CNC computer. Might be why it froze twice tonight
07:45 PM -!- #linuxcnc mode set to +v by ChanServ
07:50 PM -!- #linuxcnc mode set to +v by ChanServ
07:59 PM -!- #linuxcnc mode set to +v by ChanServ
08:06 PM -!- #linuxcnc mode set to +v by ChanServ
08:07 PM -!- #linuxcnc mode set to +v by ChanServ
08:09 PM pcw-home: JT-cave re: 7I77 issue, do the inputs work?
08:11 PM Tom_L: sry for all the reconnects.. testing
08:11 PM pcw-home: Alicrow-bridget: Electrical Delays should be 1 PWM cycle max, but motor inductance and mechanical inertia will add much more
08:12 PM Alicrow-bridget: Yeah, that's my theory, too.
08:15 PM Alicrow-bridget: So should I be concerned? Is there a way to compensate/account for this?
09:23 PM mrec: finally I was able to set the encoder to 25ppr, now I'm hobbing gears at 1k rpm via parallel port, even more seems to be okay
09:23 PM mrec: 25ppr (x4, 100 steps per revolution)
09:24 PM mrec: 1k ppr (encoder factory default) was unreliable at 400-500rpm
09:24 PM Tom_L: 36deg per step
09:26 PM mrec: I need to make my own encoder board with a microcontroller and rs422
09:27 PM mrec: 3.6 deg per step not 36deg per step
09:27 PM mrec: that's good enough for m0.5
09:27 PM Tom_L: https://www.youtube.com/watch?v=o8lezpJhfAo
09:27 PM Tom_L: quad
09:28 PM mrec: I'm also using x4
09:28 PM Tom_L: 2k quad iirc
09:28 PM mrec: 25*4=100
09:29 PM mrec: I received my efinix board yesterday
09:29 PM Tom_L: i've got quadrature code for avr as well
09:29 PM Tom_L: you try their software yet?
09:29 PM mrec: I only started it and looked around ... looks okay
09:30 PM mrec: no simulator it seems
09:30 PM Tom_L: similar to webpack?
09:30 PM mrec: in one of their videos they used gtkwave
09:30 PM mrec: something between icecube and altera stuff
09:31 PM mrec: not super exciting but it seems like it can do its job
09:33 PM mrec: I'm just happy that this poor single-ended encoder works for hobbing... it should be rs422 or something like that.
09:33 PM mrec: if the magnet is out of range the AS5047P will go crazy... the chip is poorly designed
09:33 PM mrec: ABI will be flooded with pulses
09:34 PM mrec: since it's synced with the A axis the A axis is going crazy
09:34 PM Tom_L: think of all the emf spikes :)
09:34 PM mrec: for some other applications this can be super dangerous
09:35 PM mrec: I'd rather go back use an encoder with a disc for critical applications for sure not AMS parts.
09:35 PM Tom_L: get one
09:35 PM mrec: I have a few encoders here
09:36 PM Tom_L: i have a handfull of small ones
09:36 PM Tom_L: from medical equipment
09:36 PM Tom_L: but for my little mill they're great
09:37 PM mrec: gear hobbing is quick I like it