#linuxcnc Logs
Jun 23 2025
#linuxcnc Calendar
01:16 AM Deejay: moin
02:47 AM lcnc-relay: <sodakaustik@> beautiful fan inlet roycroft
04:44 AM Tom_L: morning
05:21 AM rifraf: mornin
08:49 AM lcnc-relay: <skunkworks8841@> already 80f at 8:30am
08:50 AM Tom_L: same here
09:49 AM lcnc-relay: <skunkworks8841@> I think I left the green machine running...
09:49 AM lcnc-relay: <skunkworks8841@> find out when I get home
09:53 AM rdtsc-wk: impromptu extended test :)
09:57 AM lcnc-relay: <skunkworks8841@> yep
10:39 AM * rdtsc-wk secretly hacks into the greenmachine and changes the system locale to Klingon
10:44 AM lcnc-relay: <skunkworks8841@> I remember that being more of a problem a few years ago.. Like the wifi would not connect until you set the locales correctly
10:44 AM lcnc-relay: <skunkworks8841@> the latest iso's seem a bit easier for inital setup.
10:45 AM lcnc-relay: <skunkworks8841@> although - it took me 2 tries to get the interfaces file setup correctly for mesa
10:53 AM rdtsc-wk: the rpi5 6.12 .iso comes with the location set to Australia and no locale set - those are a must-set right away
10:54 AM rdtsc-wk: well at least the locale
10:55 AM roycroft: or you could just move and not have to set it
10:55 AM roycroft: reset the location, that is
11:12 AM lcnc-relay: <pavle_dlltech@> hello, do i need any hut for raspberry pi 3b or can i directly connect steper motor drivers to gpio pins and configure them in linuxcnc?
11:29 AM roycroft: first, i'm not sure that an rpi 3b would work well with linuxcnc - most folks are using at least an rpi4
11:29 AM roycroft: but that said, you will almost certainly experience way too much latency to be able to do anything approaching real time motion control using just the rpi3b
11:31 AM roycroft: you would need a smart motion control board such as a mesa 7i96 to reasonably control a machine with any rpi
11:39 AM lcnc-relay: <skunkworks8841@> agreed - I don't think the pi3 is supported..
11:43 AM lcnc-relay: <skunkworks8841@> hmm - looks like it might be doable. I thought the 3 was way underpowered.
11:43 AM roycroft: i find my rpi 4b with a 7i96s barely works
11:44 AM roycroft: that setup uses wifi for the lan interface, though, and that can cause a lot of needless latency
11:44 AM roycroft: i'm considering adding a second ethernet port to the rpi and doing a hard-wired connection to the lan
11:44 AM lcnc-relay: <skunkworks8841@> To answer you question... From my experience the rpi4 can barley output fast enough pulse train to drive steppers - In my experience it is less than 10khz.
11:45 AM lcnc-relay: <skunkworks8841@> (software step generation)
11:45 AM roycroft: i haven't even tried - so many folks have said that it won't work that i did not bother
11:46 AM lcnc-relay: <skunkworks8841@> I have had good luck with mesa and the rpi4.. I only allow 1 core for the os and it is snappy enough for me.
11:46 AM lcnc-relay: <skunkworks8841@> The rpi5 is way better
11:46 AM roycroft: in my initial testing it seems to work fine
11:47 AM roycroft: if, when i finally get the controller done and the machine actually doing work, i find it too sluggish i'll upgrade to an rpi 5
11:47 AM rdtsc-w: the rpi has come a long ways from 3 to 5 - suggest getting a 5; very versatile, more compatibility, way faster... also for a dedicated milling machine, consider a small x86-64 board instead of a pi. amd64 is better-supported than arm64 in many ways.
11:51 AM Tom_L: 3 doesn't have enough computing power to be useful as a cnc control
11:51 AM Tom_L: 4 does but barely
11:52 AM Tom_L: 5 so far looks promising as an inexpensive cnc platform with some sort of mesa card on it
11:52 AM Tom_L: and compared to the 4 it's FAF :)
11:53 AM Tom_L: you might be able to do more things that just watch grass grow or paint dry
11:53 AM Tom_L: than
11:54 AM lcnc-relay: <mikemader@> I started out with a 7i96 on a PI4 and it wasn't stable. Though it was on a USB SSD, so that may have contributed to the instability.
11:54 AM Tom_L: you can tweak the config some to make a few improvements
11:55 AM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/Rpi4/Rpi4_histogram294..png
11:55 AM Tom_L: i ran that yesterday on mine with a stock image from the download page before any tweaks
11:55 AM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/Rpi5/Rpi5_histogram2.png
11:56 AM Tom_L: rpi5 in comparison
11:56 AM lcnc-relay: <mikemader@> I switched to an HP Elitedesk 800 g100 mini, and it has been solid.
11:57 AM lcnc-relay: <mikemader@> That has been running for a few years now
11:57 AM Tom_L: i'm still running a D525 on the mill
11:57 AM Tom_L: intel board
11:58 AM Tom_L: started with wheezy and now running bookworm with 2.9.4
11:58 AM Tom_L: opens the door for alternate gui too
12:06 PM lcnc-relay: <skunkworks8841@> I have been running a 4 on the green machine since I got it running.. It has been stable and good.
12:06 PM rdtsc-w: axis gui?
12:06 PM lcnc-relay: <skunkworks8841@> (and I do some wierd stuff with it)
12:06 PM lcnc-relay: <skunkworks8841@> yes
12:06 PM rdtsc-w: 'cause it would have struggled with QtDragon :)
12:06 PM lcnc-relay: <skunkworks8841@> https://photos.app.goo.gl/k8heMAQ7S2vB268Y7
12:09 PM lcnc-relay: <mikemader@> Like I said, it could have been because it was on an USB SSD rather than an SD card in my case. But I would get jitter isses and random disconnects from the 7i96, usually part way through a job.
12:09 PM roycroft: i just happen to have several of the rpi 4b and would like to use them if i can
12:10 PM roycroft: but if i need an rpi 5 i'll get one
12:10 PM roycroft: i won't know until i start making parts
12:10 PM lcnc-relay: <mikemader@> I don't plan on changing out the mini PC I have anytime soon, but should I build another machine, I may try an rpi 5 with an NVME drive.
12:11 PM * roycroft has decided to relabel the computer interface panel for his controller after all, as that should take about 5 minutes of sanding and then however long it takes on the laser (unattended, so the time does not matter)
12:11 PM roycroft: it just bugs me too much
12:11 PM lcnc-relay: <mikemader@> Though I have 3 other mini PCs laying around right now, so one of those may get used.
12:12 PM roycroft: way more than the damage to the opening from the cutter slipping
12:12 PM roycroft: i'll probably sand it down at lunch time and get it back on the laser
12:12 PM rdtsc-w: unattended laser... wasn't SW just saying how one caught fire
12:12 PM roycroft: i have a web cam for monitoring
12:13 PM roycroft: and i'm engraving brass, not cutting something flammible
12:14 PM rdtsc-w: surprised wood doesn't just burn from a laser... but surprisingly it doesn't... most of the time.
12:14 PM Tom_L: yeah one guy here set his house on fire with one
12:14 PM roycroft: it's all about power and speeds
12:15 PM roycroft: take several fast/low power passes vs fewer slow/high power passes and it's a lot less likely to burst into flame
12:15 PM lcnc-relay: <mikemader@> I have two lasers (both CO2, one not functional right now). I never leave them running unattended. Way too much of a risk.
12:15 PM rdtsc-w: salvaged a 20W 1064nm module; gonna be tinkering with that soon (od7+ goggles on order)
12:16 PM roycroft: again, i have a web cam for monitoring
12:16 PM roycroft: i suppose one could consider remote monitoring (from within the same building) "attended"
12:16 PM roycroft: i'm using a 2w 1064nm laser head for the brass engraving
12:17 PM roycroft: which is why it takes so bloody long
12:17 PM rdtsc-w: results are nice though
12:17 PM roycroft: but it gets the job done
12:17 PM roycroft: yes
12:17 PM roycroft: i'm pleased with the results, other than my screwup on reindexing that one plate
12:17 PM roycroft: which i still don't understand
12:17 PM * roycroft thought he was really careful
12:18 PM lcnc-relay: <skunkworks8841@> the only problem we have had with fire and lasers at work here is caused by the bits and buildup of plastic on the table or ducting
12:18 PM xxcoder: if you bumped and moved the head it might look like plate misplacement
12:18 PM roycroft: i rehomed and then moved to my set offset
12:18 PM xxcoder: hmm not sure then
12:18 PM roycroft: and the offset was at an angle
12:19 PM roycroft: if it had been that the head moved the offset would be square to the original
12:19 PM rdtsc-w: wanted to see if brake (friction) material could be laser-cut... long story short, no; contains glass/ceramic fibers, melts and prevents beam from cutting further
12:19 PM xxcoder: true dunno then
12:19 PM roycroft: https://roycroft.us/MiniRouter/ComputerInterfacePanelLabeled.jpeg
12:19 PM roycroft: that's not from a bumped head
12:20 PM roycroft: but my indexing plates are attached to the honeycomb
12:20 PM roycroft: so i'm thinking the honeycomb moved somehow
12:20 PM xxcoder: might be it
12:20 PM roycroft: or the laser frame moved in relation to the honeycomb
12:20 PM lcnc-relay: <mikemader@> If that were a CO2 laser, I would be thinking that the laser was hitting the cone and reflecting off it making the shadow
12:20 PM lcnc-relay: <skunkworks8841@> I do machining like this with the rpi4...
12:20 PM lcnc-relay: <skunkworks8841@> https://www.youtube.com/shorts/E1-oQAhWiXI
12:21 PM lcnc-relay: <skunkworks8841@> which requires syncing the spindle with the axis
12:21 PM roycroft: either could be the case, and if it is one of those two things, when i get the enclosure finished that issue will go away, as the laser frame will be fixed in relation to the base, and the honeycomb will be indexable on the base
12:21 PM lcnc-relay: <mikemader@> I had that issue with my big 60W, and I had to adjust one of the mirrors to get it centered.
12:22 PM roycroft: mikemader: i did a single pass, removed the part to make sure the text was where i wanted it to be, then reindexed (i thought) the part and did the rest of the passes
12:22 PM roycroft: the first pass was so light i did not notice the offset until after the job was done
12:23 PM roycroft: it's much more apparent in the pic than when viewed in person, but i can still see it in person, even from a distance
12:23 PM lcnc-relay: <mikemader@> Gotcha, yeah, that would do it too.
12:23 PM xxcoder: mayb you can sand it so "shadow" fades away, as its more shallow
12:23 PM roycroft: https://roycroft.us/MiniRouter/MotionControlPanelLabeled.jpeg
12:24 PM roycroft: that one came out perfect
12:24 PM xxcoder: might fade rest of engrave too much though
12:24 PM roycroft: the engraving is not very deep
12:24 PM lcnc-relay: <mikemader@> Yeah, that could sand away
12:24 PM roycroft: my intent is to sand it off completely, both the original pass and the second set of passes
12:24 PM roycroft: it should take about 5 minutes to sand it away
12:25 PM roycroft: then i'll re-engrave
12:25 PM rdtsc-w: E pluribus carborundum
12:25 PM lcnc-relay: <mikemader@> I did something similar that you are doing. I used the laser to make the plug panel for my CNC controller, though I used cast acryllic for it.
12:25 PM lcnc-relay: <skunkworks8841@> roycroft: what laser is that?
12:25 PM roycroft: https://roycroft.us/MiniRouter/BackPanelSecondFitment.jpeg
12:25 PM lcnc-relay: <skunkworks8841@> and do you like it?
12:25 PM roycroft: that's the panel it's going into
12:26 PM roycroft: it's an xtool d1 pro
12:26 PM roycroft: and yeah, it was cheap but it works
12:26 PM roycroft: they don't make it any more - xtool have shied away from open frame lasers
12:26 PM roycroft: i guess too many people put their eye out :)
12:26 PM roycroft: i should say that the initial cost was cheap
12:27 PM roycroft: but then i added a 20w diode laser for way more than i paid for the original unit, and then i added the 2w 1064nm laser for even more than the cost of the 20w diode laser
12:27 PM roycroft: but i have a decent setup for my use - i do not intend to do production work on it
12:28 PM roycroft: it's for shop/household projects, and for engraving my brand into products
12:28 PM xxcoder: when I looked at xool site, there was a some sort of gambling thing pop up, for getting savings for buying things there. ugh
12:28 PM roycroft: i would never make parts in production with it
12:29 PM roycroft: i did get the rotary axis, though, and there's a chance i might make products for sale with that
12:30 PM roycroft: i got the xtool for the same reasons i got the ender 5 3d printer - i wanted something turnkey to get my feet wet with the process, to learn more about the process from a practical use point of view, and to be able to build on the machine to make improvements
12:30 PM roycroft: with the plan being to build my own from scratch in the future
12:31 PM roycroft: and i'm very pleased with both the xtool and the ender 5 - they have fulfilled my criteria capably
12:34 PM roycroft: and yeah, that spinning wheel thing on the xtool website is annoying
12:35 PM roycroft: so centurylink never showed up on friday to remove that illegal drop
12:36 PM roycroft: i just got off the phone with them and they were trying to set a new due date of july 4
12:36 PM roycroft: when i pointed out to them that july 4 is a holiday they tried to push it back to july 28
12:36 PM roycroft: i reminded them that they committed to having it gone by the end of june, and they said i can't expect them to do things that quickly
12:37 PM roycroft: they thought my ticket from last week was my first contact with them about this
12:37 PM JT-Cave: sounds like the warranty appliance repair service... my fridge has been broken for 6-8 weeks
12:37 PM roycroft: i gave them all the other ticket numbers, and they finally got a supervisor to reschedule for june 26
12:38 PM xxcoder: oh man
12:38 PM roycroft: so we'll see if the truck shows up on thursday
12:38 PM JT-Cave: 2 weeks ago they canceled the service call due to who the heck knows
12:38 PM roycroft: they were really pushing for late july, and i told them that by then the line would be cut and dangling from both the pole and my neighbor's house
12:38 PM JT-Cave: that's the same day my fridge is supposed to be fixed
12:38 PM roycroft: because nobody needs a fridge, jt-cave
12:38 PM roycroft: it's a luxury item
12:39 PM roycroft: so they can deal with it when they get around to it
12:39 PM JT-Cave: it's sears home appliance repair and if I'm paying they can be here the next day
12:39 PM roycroft: i try to be a little patient, as everthing is a long truck roll from here
12:40 PM roycroft: but my patience with centurylink has grown thin
12:40 PM roycroft: i know you don't live in a big city, but my sense is that you're not as far removed from civilization as i am
12:41 PM roycroft: so an expectation of quick service is not unreasonable in your case
12:42 PM lcnc-relay: <skunkworks8841@> My mom is cooler than your mom...
12:42 PM roycroft: that's 'cuz her fridge works
12:43 PM lcnc-relay: <skunkworks8841@> (sorry - local bumper sticker joke when they found this guys mom in the freezer because he wanted to keep getting her ss checks)
12:43 PM JT-Cave: holy crap
12:44 PM lcnc-relay: <skunkworks8841@> (she did die of natural causes.. but still...)
12:45 PM * rdtsc-w thinks the crickets are even quiet...
12:47 PM xxcoder: yeah heard of similiar happenings once a long while. I remember one where someone just covered their mother with bunch of blankets and left room closed
12:47 PM xxcoder: got caught as usual.
12:47 PM xxcoder: its impossible to pull off such thing for long at all
01:04 PM JT-Shop: nap time
01:16 PM roycroft: i don't know, xxcoder
01:16 PM roycroft: i've heard that there are millions of people over 150 years old still collecting ss
01:16 PM xxcoder: lol
01:41 PM Tom_L: i would like to live a nice long productive life but maybe not that long
02:05 PM lcnc-relay_ is now known as lcnc-relay