#linuxcnc Logs

Sep 01 2024

01:40 AM Deejay: moin
02:21 AM lcnc-relay: <turboss> morning
04:27 AM JT-Cave: morning
04:29 AM Tom_L: morning
04:41 AM pere: Anyone here interested in <URL: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1080191 >, svg2gcode.
06:46 AM JT-Cave: hmm even though the docs say increments is comma separated axis does not complain if they are space separated
06:57 AM JT-Cave: so should I reject an increments string that is not comma separated?
07:08 AM pere: I suspect it is better to do what the documentation claim should happen...
07:17 AM JT-Cave: yup
07:28 AM JT-Cave: and the docs say INCREMENTS = 1 inch, 1 mil, 1 cm, 1 mm, 1 µm is a valid entry.
07:28 AM JT-Cave: however µm is not valid as Axis only checks for um
07:29 AM JT-Cave: elif jogincr.endswith("um"):
07:29 AM JT-Cave: scale = from_internal_linear_unit(.001/25.4)
08:31 AM lcnc-relay: <Travis Farmer> had some real maple syrup i had put on my sausage for breakfast, and it tasted... odd, this time. checked the syrup and it was all cloudy with mold or something... 😦
08:36 AM Tom_L: JT-Cave, i don't see that it say it has to be comma separated. the example shows that
08:37 AM Tom_L: went back to 2.5 and the docs haven't changed
08:39 AM Tom_L: prior to that all i see ar fr docs
08:41 AM Tom_L: err no, it shows both
08:42 AM JT-Cave: it doesn't say but the examples show comma delimited
08:42 AM Tom_L: right
08:42 AM Tom_L: but it didn't say it has to be
08:42 AM JT-Cave: and how would one parse 1 inch 2 mm, 5in
08:43 AM JT-Cave: there has to be a consistent separator of some kind
08:43 AM Tom_L: you figure that out lemme know :)
08:43 AM Tom_L: i'd go with comma separated too
08:44 AM Tom_L: that documentation was introduced in 2.2
08:45 AM JT-Cave: yep I'm sticking with comma separation
08:45 AM JT-Cave: increments?
08:45 AM Tom_L: don't see mention of it prior to that
08:45 AM Tom_L: yes
08:45 AM Tom_L: http://linuxcnc.org/docs/2.2/html/config_ini_config.html#sub:%5BDISPLAY%5D-Section
08:46 AM Tom_L: 1/16 is also valid
08:48 AM JT-Cave: yup
08:48 AM Tom_L: i'm not sure that option was well though out
08:48 AM Tom_L: thought
08:52 AM JT-Cave: actually 1/16 is not hard to deal with I just use .contains('/') and then do the math
08:53 AM Tom_L: but it could be in mm cm fathom etc
08:53 AM JT-Cave: the combo box has two thing per row the text and the value
08:53 AM JT-Cave: ah no fathom is not allowed
08:53 AM JT-Cave: lol
08:53 AM Tom_L: just checking
08:54 AM rdtsc: AUs, parsecs? ;)
08:56 AM JT-Cave: I'm checking for invalid units and going to complain :)
08:57 AM JT-Cave: INCREMENTS = 1 inch, 1 mil, 1 cm, 1 mm, 1 um
08:57 AM JT-Cave: so 2.2 is correct with the um unit
08:58 AM Tom_L: better check for 'um' and 'µm'
09:00 AM JT-Cave: as is 2.3, 2.4, 2.5, 2.6, 2.7, 2.8 however 2.9 is wrong and uses µm
09:00 AM JT-Cave: not going to check for µm, going to treat it as an error
09:00 AM Tom_L: somebody just finally figured out the symbol
09:01 AM Tom_L: alt 230
09:01 AM JT-Cave: well emc does not use µm so it's wrong
09:01 AM JT-Cave: break fast time
09:02 AM lcnc-relay: <Travis Farmer> why not just do a replacement of "µm" with "um"
09:02 AM Tom_L: because
09:03 AM lcnc-relay: <skunkworks8841> lol - dad had a container of 103 (.01uf) capasitors.. I am still using them today.. I have been using them since I was 8 atleast
09:03 AM Tom_L: ceramic?
09:04 AM lcnc-relay: <skunkworks8841> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/FSvPnkYhfUSjgsrYxjYKUyNd/PXL_20240901_135716205.jpg
09:04 AM Tom_L: no doubt they're still good
09:05 AM lcnc-relay: <Travis Farmer> i bought a bag of 1000... still using them
09:05 AM lcnc-relay: <skunkworks8841> so far..
09:05 AM Tom_L: i just recapped a board for a guy with dry electrolytics that had swelled up
09:05 AM lcnc-relay: <skunkworks8841> yeck
09:05 AM rdtsc: heat is the enemy of electrolytics
09:05 AM lcnc-relay: <skunkworks8841> I think at work we are past those computers..
09:05 AM lcnc-relay: <skunkworks8841> knock on wood
09:06 AM Tom_L: only change is size
09:07 AM Tom_L: you can get those today the size of a breadcrum
09:08 AM lcnc-relay: <skunkworks8841> right.. I want to have someone make boards for this - but I still probably want them to be non-surface mount.. But - I don't know yet. I do have a re-work station at work - so I probably could do it.
09:08 AM Tom_L: how many do you need?
09:09 AM lcnc-relay: <skunkworks8841> not many.. maybe 10 at the most
09:09 AM rdtsc: Today's computer boards often use hybrid or solid caps, so seldom fail.
09:09 AM Tom_L: https://oshpark.com/
09:09 AM Tom_L: that's gonna be your best bet
09:10 AM Tom_L: you used to get 3 per layout back
09:10 AM lcnc-relay: <Travis Farmer> i wonder if my 1980-90s security alarm system needs to be recapped yet...
09:10 AM lcnc-relay: <skunkworks8841> sure - pcbway reached out to me.. I might call in their favor if I get that far
09:10 AM Tom_L: purple mask
09:10 AM rdtsc: About to send a design to OshPark today :)
09:10 AM Tom_L: i remember when ian first got started
09:12 AM rdtsc: think pcbway and jlcpcb can do assembly too... although I'd prefer OshPark and self-assembly to keep more of the economics local.
09:13 AM rdtsc: wans't there when Ian got started, but shortly after
09:17 AM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/boards/atmega32u4/atmega32U4_3.jpg
09:17 AM Tom_L: that was the early 'purple'
09:17 AM Tom_L: he was still 'perfecting' it
09:18 AM Tom_L: oh and if somebody wanted gold plating, everybody on that order got it
09:21 AM rdtsc: nice. that design looks like one I started awhile ago with a LT 16-channel ADC. aahh, the things I'd like to get back to, but need the bench finished first, which is waiting on the mill...
09:26 AM lcnc-relay: <skunkworks8841> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/vYhSxXvwaVgiafpjMjhLvjTa/PXL_20240901_141901990.jpg
09:26 AM lcnc-relay: <skunkworks8841> lol - my current soldering station
09:29 AM rdtsc: hey if it gets the job done :)
09:30 AM lcnc-relay: <skunkworks8841> mostly.. 🙂
09:30 AM lcnc-relay: <skunkworks8841> still cleaning - the 42' is parked down by the polebuilding - starting to fill it up with steel that has been laying around outside.
09:31 AM lcnc-relay: <skunkworks8841> need to decide if I am taking apart the fork lifts and scrapping them
09:31 AM JT-Cave: travis where would replacing incorrect units end?
09:32 AM lcnc-relay: <skunkworks8841> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/YdAqhYtXntUeWETAbYSbxfYK/PXL_20240901_142517935.jpg
09:39 AM Tom_L: skunkworks, when they see you coming down the road in that do you get extra space?
09:39 AM Tom_L: :)
09:44 AM lcnc-relay: <skunkworks8841> Lol.. usually just double takes
10:02 AM lcnc-relay: <JT> https://jauriarts.org/_matrix/media/v3/download/matrix.org/HeUYStUGLJEiilhQzTMPxaoQ/1000002310.jpg
10:02 AM lcnc-relay: <JT> Caliper mounting bolts, new
10:03 AM JT-Shop: I may not have a stripped out threads after all, just a bad lockwasher or bolt head
10:03 AM Tom_L: nice quality control
10:07 AM JT-Shop: when the bolt tightened up I got the torque wrench out then it got loose, I'm thinking the threads stripped as I can't see the bolt head it's on the back side
10:07 AM JT-Shop: both bolts and lockwashers are new and I didn't see anything wrong with either
10:08 AM Tom_L: hmm
10:12 AM lcnc-relay: <skunkworks8841> not flat were the bolt tightens?
10:16 AM JT-Cave: yup very flat but a 7/16 bolt with a 5/8 head is a bit small for a split lockwasher IMHO
10:16 AM JT-Cave: I looked at the old ones and only one had a lockwasher
10:18 AM JT-Cave: but that seems to be the correct head size according to mcmaster
10:18 AM lcnc-relay: <Travis Farmer> good point, JT, on the units... i think of logical solutions, not idiot-proof ones 😉
10:27 AM lcnc-relay: <JT> https://jauriarts.org/_matrix/media/v3/download/matrix.org/pCqLPSyMTfnufcDzOZkwCfCJ/1000002311.jpg
10:29 AM lcnc-relay: <JT> https://jauriarts.org/_matrix/media/v3/download/matrix.org/BCJamtBPkjlBwvhdSIDufXPt/1000002313.jpg
10:35 AM lcnc-relay: <skunkworks8841> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/npiqNnMzZiDLBMYBiNWTvgqC/PXL_20240901_152250707.jpg
10:35 AM lcnc-relay: <skunkworks8841> 2 working drives for the monarch
10:36 AM JT-Shop: no point in looking for a 7/16" lockwasher the caliper has issues
10:36 AM lcnc-relay: <skunkworks8841> What are you working on?
10:37 AM JT-Shop: `82 C3
10:37 AM lcnc-relay: <skunkworks8841> Neat! Did you get rid of the new vet?
10:38 AM lcnc-relay: <JT> https://jauriarts.org/_matrix/media/v3/download/matrix.org/nJHvCXteBUPMGdpfFLQygwxn/1000002259.jpg
10:38 AM JT-Shop: oh no the C8 is my traveling vehicle
10:38 AM JT-Shop: the C3 is something to work on it seems lol
10:38 AM lcnc-relay: <skunkworks8841> Lol.. cool!
10:47 AM Tom_L: got some bad parts?
10:48 AM JT-Shop: looks that way
10:48 AM * JT-Shop heads to town
11:24 AM rdtsc: Tom, is there an easy way to also have an "Exit and shutdown" menu item in Flex?
11:26 AM Tom_L: there is 'Quit'
11:27 AM rdtsc: suppose could autostart linuxcnc and include a shudown afterwards
11:29 AM lcnc-relay: <JT> You can have a shutdown Hal file
11:32 AM lcnc-relay: <JT> If that's what you're asking
11:32 AM Tom_L: SHUTDOWN = shutdown.hal - Execute the file shutdown.hal when LinuxCNC is exiting. Depending on the hardware drivers used, this may make it possible to set outputs to defined values when LinuxCNC is exited normally. However, because there is no guarantee this file will be executed (for instance, in the case of a computer crash), it is not a replacement for a proper physical e-stop chain or other protections against software failure.
11:33 AM lcnc-relay: <JT> Yup
11:33 AM Tom_L: i fixed your log issue i think as well as index
11:34 AM Tom_L: didn't know if you saw that
11:34 AM lcnc-relay: <JT> Cool
11:34 AM Tom_L: try index
11:34 AM lcnc-relay: <JT> Ondex
11:36 AM Tom_L: lunch done... upcoming events: nap
11:52 AM rdtsc: Tom when you get a minute, could you check over a small PCB for powering this beast? Perhaps a novel idea, perhaps a fool's endeavor.
11:59 AM Tom_L: for 5v i'd get an SMPS off ebay or such
12:00 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/misc_stuff/5v_switcher.jpg
12:00 PM Tom_L: i made a few up but the parts cost more than the ebay ones
12:01 PM Tom_L: LM2596
12:01 PM rdtsc: oh the 7C80 powers up fine (using a 12V PSU), it's just that there's no way to shut it off, except for pulling the plug - that's what my board does
12:02 PM rdtsc: or the concept anyways
12:03 PM rdtsc: Machine power supply wiring: https://pasteboard.co/MAMca34109qM.png
12:03 PM rdtsc: Power adapter schematic: https://pasteboard.co/iBrYjTvjH1IV.png
12:04 PM rdtsc: Power adapter board: https://pasteboard.co/DmpWUgIMO5iB.png
12:04 PM rdtsc: Render Front: https://pasteboard.co/cPjBsYoJZHiu.png
12:04 PM rdtsc: Render Rear: https://pasteboard.co/XuT42NkTD2nY.png
12:07 PM lcnc-relay: <zincboy_ca_on> What is the problem you are trying to solve?
12:09 PM rdtsc: no way to power on/off 7C80 automatically
12:09 PM JT-Shop: could shutdown.hal power it off?
12:10 PM lcnc-relay: <zincboy_ca_on> Just leave it on when the machine is on? I don't see any issue with that.
12:11 PM lcnc-relay: <big_kevin420> yeah are you trying to power off the 7c80 or the rpi?
12:11 PM rdtsc: it's on whenever the power cord is plugged in... if I forgot about it for a week, etc.
12:11 PM lcnc-relay: <big_kevin420> the main power? then you need some kind of contactopr
12:12 PM lcnc-relay: <zincboy_ca_on> So? No real power consumption if the system is not running.
12:13 PM lcnc-relay: <big_kevin420> i think it was easier for me to control the power opn the 7c81 becasue i could have control of the 5v power suoprce from the pi
12:13 PM lcnc-relay: <big_kevin420> but if you are powering everything from the c80, im not sure how you expect to power it off
12:13 PM rdtsc: The Rpi is powered from the 7C80, along with a fairly noisy cooling fan. The only way to get these off is by cutting power to the 7C80, but only want to do that when the Rpi is actually shut down (else corrupt SD card.)
12:15 PM rdtsc: I'm getting the impression that not many others have a 7C81 lol
12:15 PM rdtsc: err 7c80
12:15 PM lcnc-relay: <big_kevin420> well, he sold all 50
12:15 PM lcnc-relay: <big_kevin420> so at least 50
12:16 PM lcnc-relay: <big_kevin420> it becomes a rabbit hole for the 7c81
12:16 PM rdtsc: oh good, I'm not alone :)
12:16 PM lcnc-relay: <big_kevin420> the 7c80 seems even more challenge though
12:16 PM rdtsc: I think my board will work, if I haven't forgotten anything
12:16 PM lcnc-relay: <zincboy_ca_on> I think you are overthinking things. I would just shut down the pi and then turn the system off.
12:17 PM lcnc-relay: <big_kevin420> he wants a nice thing, cant blame him, i spend a long time tryign to deal with the same on/off issue
12:17 PM lcnc-relay: <zincboy_ca_on> The fuses on your power supplies are likely undersized. Especially using fast blow.
12:17 PM lcnc-relay: <big_kevin420> press button, turns shit off / press button turns it on
12:17 PM lcnc-relay: <zincboy_ca_on> Oh, I understand and do the same thing myself 🙂
12:18 PM lcnc-relay: <big_kevin420> it was hard thought, and i never found a good option, except for making some kinda of pcb to do it
12:18 PM rdtsc: yeah fuse sizes require a measurment; guesstimate at the moment
12:19 PM rdtsc: A 1A fast fuse will actually pass tens of amps for one cycle; just not for two :)
12:19 PM lcnc-relay: <big_kevin420> turn off the 24v using the RPI io, so when rpi turns off the 24v power turns off
12:19 PM rdtsc: right on, that's the idea anyways.
12:22 PM rdtsc: I'm just concerned that I'd use a contactor, hit shutdown and throw the switch before Linux was fully shut down. Then the next time I boot it, fsck errors and a lengthy repair
12:22 PM lcnc-relay: <Travis Farmer> use USB power... when USB power shuts off...
12:23 PM lcnc-relay: <Travis Farmer> have it trigger a SSR
12:36 PM JT-Cave: nap time
01:13 PM rdtsc: Think the 10k gate pullup should be reduced to a 330R - at initial power-on, that gate is going to be at zero volts and take some time to rise to 5V, potentially letting a brief pulse through the SSR. Simulating, yes, 10k = 5us unintended pulse. With 330R = 200ns
01:26 PM lcnc-relay: <zincboy_ca_on> Look at something like the LTC2954
02:08 PM xxcoder: nice frank howarth video today
02:13 PM lcnc-relay: <vibram> https://youtu.be/-G4Xcev4-tc?si=lVJFZ89Wt0_iayp0
02:14 PM lcnc-relay: <vibram> This is my spindle stability test. Do you think I can do better?
03:47 PM lcnc-relay: <skunkworks8841> The debian system that was hardlocking - nothing in the logs.. I updated the bios and it has been up for a week so far
03:50 PM lcnc-relay: <skunkworks8841> asrock motherboard
03:57 PM Tom_L: cool
03:57 PM lcnc-relay: <skunkworks8841> don't know why the locking up happened when the ssd failed.. Go figure
04:32 PM Tom_L: i've had crazy things happen when swapping ssd around on mine
04:32 PM Tom_L: takes a couple bios resets to clear it up sometimes
04:33 PM Tom_L: and i found out it doesn't like 3 ssd attached at once
05:51 PM * JT-Shop calls it a night
06:00 PM _unreal_: https://photos.app.goo.gl/ggAPhV6uwbvr4jLs8 what cha think?
06:11 PM lcnc-relay: <captainhindsight_.> I bet that exhaust tube really sucks
07:40 PM roycroft: yay
07:40 PM roycroft: the last of the heavy things are moved out of the old house
07:41 PM roycroft: and my back is still semi-intact
07:41 PM roycroft: i do have three heavy things in an outbuilding still, but they don't have to move for a bit
07:43 PM _unreal_: will they move to the in building?
07:49 PM roycroft: they will move to the storage container for now
07:49 PM roycroft: eventually to the shop, once the shop extension is constructed
07:50 PM * roycroft is hoping that happens next spring/summer
07:50 PM * roycroft , btw, is in with the in crowd
07:53 PM Rhine_Labs: I really enjoy backing up firmware.. and always make me happen when chip is not secured. LOL https://youtu.be/2F0kaJtwcGM
07:54 PM Rhine_Labs: Happen = Happy
07:59 PM roycroft: if you're happy and you know it crack a chip
08:00 PM rdtsc: omg, so lucky... 've tried dozens but they were all locked
08:06 PM t4nk_fn: well, .. _unreal_ ... I'm thinking.. buy some proper leadscrews and nuts :b
08:07 PM t4nk_fn: and I'm also thinking.. I bet a 14-foot drill bit would have been even better suited for that skeleton :b
08:07 PM _unreal_: 6" of working z
08:07 PM Rhine_Labs: Not a common MCU
08:09 PM t4nk_fn: what size are those rails?
08:12 PM t4nk_fn: also, I think that this spindle-thingy oughtta be positioned way more to the center of the clamp
08:13 PM t4nk_fn: I'm far from an expert on all this, as you know, far from a decent user at all even.. but I think that this all contributes to an enormous amount of instability/play/feebleness
08:14 PM t4nk_fn: I couldn't really imagine that cutting any serious metal methinks, right?
08:22 PM _unreal_: Its not for metal
08:22 PM _unreal_: its for wood, mdf, PVC foamboard, aluminum a bit
08:23 PM _unreal_: I've already done a few tests and its fine.
08:23 PM _unreal_: I do have some defelection in the X axis but I discovered I had missed installing a few screws in the cars. I need to get the right length m3's
08:24 PM _unreal_: and I'm prob going to need a cross wire pully setup for steel cable to sure up any rotation in the X axis
08:24 PM _unreal_: Z and Y are rock solid
08:25 PM _unreal_: its like 5k steps per mm for resolution :)
08:25 PM _unreal_: its NOT fast but its very powerful
08:29 PM t4nk_fn: I've been doing a whole bunch of projects to the side myself lately
08:31 PM t4nk_fn: as I was a bit disappointed with the z-axis on my new build
08:32 PM t4nk_fn: didn't seem to perform as I'd hoped
08:33 PM t4nk_fn: heh, it's basically largely the same as it always has been.. but for some reason it's not the same anymore
08:33 PM t4nk_fn: I've had some good results before
08:33 PM _unreal_: I used my little machine to build most of the parts the aluminum parts to build the new machiune
08:33 PM _unreal_: machine
08:49 PM t4nk_fn: yeah, that's nice if you could make some accurate alu parts
09:01 PM _unreal_: and 1.5kw spindle
09:01 PM _unreal_: my little cnc machine is accurate with the aluminum just not fast. and only having a 12k rpm spindle
09:02 PM _unreal_: took a while
09:03 PM _unreal_: at this moment I think my little cnc mahcine is a bit more rigid then the big machine. BUT then again it has a spindle that spins 2x faster and is 5x more powerful
09:03 PM _unreal_: the small spindle on the small machine is a .3kw
09:03 PM _unreal_: lol
09:04 PM _unreal_: so i'm sure I can at least do ruff cuts and then do a finish pass when it comes to aluminum
09:22 PM _unreal_: t4nk_fn, so what does the current build you have look like now?
10:31 PM lcnc-relay: <captainhindsight_.> oo0ff LinuxCNC with a Fanuc GUI
11:47 PM lcnc-relay: <onceloved.cn> Where is the Fanuc GUI?