#linuxcnc Logs

Nov 02 2025

01:43 AM lcnc-relay: <redboro_@> roycroft: This is a fun rabbit hole. I’ve got quite a few different wax batches for various wood turning & metal polishing. Some have a pine rosin & turpentine base, others beeswax & walnut (or olive) oil. Some of them I’ve mixed in very fine pumice.
02:03 AM lcnc-relay: I stopped using any of the colorful polishing waxes off the shelf because they’re so difficult to get out of the tiny nooks & cranny’s of intricate pieces. My beeswax, oil, and pumice blend works just as good or better, and is really easy to clean off. Smells nicer too!
02:03 AM lcnc-relay: I’ll be interested to see your progress as you go along!
02:03 AM lcnc-relay: <redboro_@> Possible to use linear encoder position info for homing & limits instead of switches?
02:07 AM jpa-: redboro_@: only if it is absolute encoder or has Z output
02:35 AM lcnc-relay: <meisterdippel@> moin
05:26 AM Tom_L: morning
07:58 AM lcnc-relay: <lesliet@> moin
09:12 AM Tom_L: quiet this morning
09:19 AM lcnc-relay: <allaboutjake@> Gonna take another run at Probe Basic again today.
09:19 AM Tom_L: ever have a look at flex?
09:27 AM lcnc-relay: <allaboutjake@> Tom_L: No. Better alternative?
09:28 AM lcnc-relay: <allaboutjake@> I know it’s under more active development
09:28 AM Tom_L: check the forum
09:28 AM Tom_L: several are using it now
09:29 AM Tom_L: https://forum.linuxcnc.org/41-guis
09:35 AM Tom_L: you can build your interface any way you like without having to code anything
09:35 AM Tom_L: set a few parameters on some things and it just works
09:37 AM Tom_L: if you want something special it also has python hooks
10:16 AM lcnc-relay: <lesliet@> cold today
10:17 AM lcnc-relay: <lesliet@> hmm, maybe time to clean up my downstairs work area, so I can do some electronics projects
10:34 AM _unreal_: today I'm not feeling so crappy. going to see if I cant clear stuff out of the garuage so I can get stuff put away orginized and make more space. so I can start using some of my tools finally
10:34 AM _unreal_: so tired of having no access
10:34 AM fluffywolf: I wish I had a garage.
10:34 AM _unreal_: I wish I had a dbl wide
10:34 AM _unreal_: all my problems would be gone
10:35 AM _unreal_: Its my number one hickup
10:35 AM _unreal_: also wish I could find slide square. I have like 3 of them and they are all miss placed as of this moment
10:36 AM lcnc-relay: <skunkworks8841@> oof - were the trottle butterfly valve assembly attaches to the carberator - the surface is so not flat. Like almost 1/16 out of flat
10:37 AM lcnc-relay: <skunkworks8841@> that would have been one hell of a vaccum leak
10:37 AM _unreal_: just need to suck harder
10:38 AM lcnc-relay: <skunkworks8841@> yep - that is one of the reasons why it only ran with the choke almost all the way on
10:38 AM rdtsc: J-B weld?
10:39 AM rdtsc: or find another throttlebody :)
10:45 AM rdtsc: Once had a Honda CB360 where the timing chain was loose and wore through the front of the block; fabbed a rounded SS insert and JB welded it in place - worked great
10:50 AM lcnc-relay: <skunkworks8841@> I can flaten it - it is pot metal - the tabs got bent up it looks like
10:51 AM lcnc-relay: <lesliet@> was gonna say
10:51 AM lcnc-relay: <lesliet@> pretty typical, people overtighten them and the flange warps
10:52 AM lcnc-relay: <skunkworks8841@> it probably wasn't as bad as I thought - there is a venturi that gets sandwiched in there.. so - maybe not as big of vacuum leak as I thought
10:52 AM lcnc-relay: <skunkworks8841@> no - it was probably pretty bad.. the one side isn't a tight fit
10:53 AM lcnc-relay: <lesliet@> going to try and shorten some cabling on the electronic leadscrew thing on my lathe. Getting an occasional glitch where it doesn't read the rpm on the encoder properly
10:54 AM lcnc-relay: <lesliet@> have tried 2 encoders, and 2 mainboards
10:54 AM lcnc-relay: <lesliet@> signals look ok with my janky old oscilloscope, as best as I can see
10:55 AM lcnc-relay: <lesliet@> but maybe picking up noise or something
11:08 AM lcnc-relay: <skunkworks8841@> wow - I think I am going to clamp it down on one of the mills and deck it off. it is a good 15+ thousands different
11:08 AM lcnc-relay: <skunkworks8841@> even with pot metal - that is a lot of filing
11:14 AM lcnc-relay: <lesliet@> verified that my encoder actually has an optical disk, so not an obvious counterfeit
11:21 AM lcnc-relay: <roguish> lesliet@: check the encoder cable shielding and grounding
11:21 AM lcnc-relay: <lesliet@> will check again, have been through it a couple of times though
11:22 AM lcnc-relay: <roguish> no ground loops
11:23 AM rdtsc: clough42's electronic leadscrew project had some noise due to long (unshielded) runs
11:24 AM lcnc-relay: <lesliet@> digging through a similar thread on the github forum
11:25 AM lcnc-relay: <lesliet@> rdtsc: that's a thought, maybe one of the cables other than the encoder cable is picking up noise?
11:25 AM lcnc-relay: <lesliet@> trying to remember if I used shielded cable, will have to check that
11:26 AM lcnc-relay: <skunkworks8841@> lol - the 2 halfs of the carb seem to have about .010 gap in spots..
11:26 AM lcnc-relay: <skunkworks8841@> oof - whoever worked on this really torqued the screws down
11:27 AM rdtsc: the noise can be in the 20-200MHz range, hard for most hobbyist equipment to see
11:27 AM Deejay: dont worry, thats outside your hearing range :-P
11:27 AM fluffywolf: carbs are usually cast pot metal and not all that precise to start with.
11:27 AM rdtsc: can clip a ferrite suppressor on it, might help to verify :)
11:28 AM rdtsc: it sure would be nice if we could hear electrical noise :)
11:31 AM rdtsc: https://duckduckgo.com/?q=ferrite+noise+suppressor
11:31 AM lcnc-relay: <lesliet@> in the forum, a lot of people had the same issue, and it seems to be due to the encoder that was linked on Clough42's information, from ebay. Folks on the forum found one and linked it, on amazon, that apparently solves the issue
11:32 AM lcnc-relay: <lesliet@> Pretty sure even if my "omron" encoder from ebay has the optical disc and not a hall sensor, that at the price I paid, it is likely counterfeit
11:33 AM lcnc-relay: <skunkworks8841@> is this encoer single ended?
11:33 AM lcnc-relay: <lesliet@> what does single ended mean in this context?
11:34 AM lcnc-relay: <skunkworks8841@> is it just a,b and z - or is it a/-a b/-b and z/-z
11:34 AM rdtsc: once had seven servo motors come into the shop for repair - their Heidenhain encoders were all bad - in a very weird way - all the signals were ever-so-slightly too low. Come to find out those encoders were recalled due to manufacturing defects
11:34 AM lcnc-relay: <lesliet@> I think just a,b, and z, let me look
11:35 AM lcnc-relay: <lesliet@> yes, just that
11:35 AM lcnc-relay: <lesliet@> rdtsc: oof
11:35 AM lcnc-relay: <lesliet@> somebody got a "deal"
11:36 AM lcnc-relay: <lesliet@> ok, ordered the recommended encoder, we'll see if that works. Would love to have this setup be reliably working
11:36 AM lcnc-relay: <lesliet@> so much quieter than change gears, and a lot less effort
11:36 AM rdtsc: single-ended means one ground, A, B, and Z, where differential means A+/A-. B+/B-, Z+/Z- (two wires for each signal, opposite polarities, so that any noise picked up is done so on both signals - which is rejected by the receiver)
11:37 AM lcnc-relay: <lesliet@> rdtsc: and so, today has not been a complete waste, for I have learned something new
11:37 AM lcnc-relay: <lesliet@> makes sense with the noise rejection
11:38 AM lcnc-relay: <skunkworks8841@> single ended will have way more issues with noise. they are also limited in distance..
11:39 AM lcnc-relay: <skunkworks8841@> I think around 10 ft is really pushing it
11:39 AM lcnc-relay: <lesliet@> I only need about a meter
11:39 AM lcnc-relay: <skunkworks8841@> what are you using for cable?
11:39 AM lcnc-relay: <lesliet@> the encoders I got came with their own cables, I can see that at least the signal wires are shielded
11:40 AM lcnc-relay: <skunkworks8841@> ok
11:41 AM lcnc-relay: <lesliet@> lol, I know I have a ferrite suppressor somewhere around here, but my electronics piles are pretty disorganised
11:41 AM rdtsc: everything industrial uses differenial signalling (RS485, PROFIBUS, CAN, ARINC (aviation comms) etc)
11:41 AM lcnc-relay: <lesliet@> a lot of music gear does, too
11:42 AM lcnc-relay: <lesliet@> but I don't think Clough42's design incorporates that for the encoder, unfortunately.
11:42 AM rdtsc: he's otherwise a pretty smart guy :)
11:43 AM lcnc-relay: <lesliet@> oh, I'm definitely not dogging on him
11:45 AM lcnc-relay: <lesliet@> ok, time to go assemble more mill parts, and paint a few more
11:45 AM lcnc-relay: <lesliet@> thanks for the help and ideas!
11:45 AM rdtsc: have a good day
11:45 AM * rdtsc has his 7/16 nuts stripped of zinc (vinegar bath overnight) so is setting up to tack them in place on the articulating arm project
11:48 AM lcnc-relay: <skunkworks8841@> well - with some stretegic filing - I got the gap down to around .002
11:49 AM lcnc-relay: <skunkworks8841@> that should be good enough for the gasket
12:02 PM lcnc-relay: <skunkworks8841@> lol - also make sure your file you are using doesn't have a twist
12:20 PM rdtsc: oof, a twisted file ism't good for much
12:21 PM rdtsc: really like diamond files, but a bit pricey
12:21 PM lcnc-relay: <skunkworks8841@> we have so many random files.. but I think the twisted one is certainly newer - but probably cheap
12:21 PM rdtsc: use a little diamond disc on the dremel for dressing TIG tungsten, works great
12:22 PM lcnc-relay: <skunkworks8841@> I don't have a tig.. some day. I would really like to learn
12:23 PM rdtsc: noobie here too, fumbling my way through. envious of those with any skill at it, certainly an artform
12:24 PM lcnc-relay: <skunkworks8841@> stick and mig here only. And I have't stick welded in ages
12:25 PM lcnc-relay: <skunkworks8841@> getting windy here
12:26 PM rdtsc: decided to go with TIG over others due to the wide range of materials supported - can do silicon steel, copper, brass, aluminum, almost anything (with the right filler) maybe even pot metal :)
12:26 PM lcnc-relay: <skunkworks8841@> I think I got the carb flat enough in places it needs to be..
12:26 PM lcnc-relay: <skunkworks8841@> it has to work a 100X better than it was
12:27 PM rdtsc: shine a light through the gap to judge
12:27 PM lcnc-relay: <skunkworks8841@> still have to make up the accellerator pump linkage
12:27 PM lcnc-relay: <skunkworks8841@> I am down to a couple thousands..
12:27 PM lcnc-relay: <skunkworks8841@> gasket should make up the rest
12:28 PM * rdtsc puts it on his bucket list to see if pot metal can be TIG welded :)
12:28 PM lcnc-relay: <skunkworks8841@> I think I have seen others do it.. maybe. on YT
12:29 PM xxcoder: anyone know of a heater that is set to fixed temperate, and are impossible for it to overheat? I know cermic heaters exist, but those tend to be tuned to temps way above what I want.
12:30 PM xxcoder: its for entirely unattended warmer. even if its set to 40f or something its perfectly fine
12:32 PM rdtsc: The Apollo 13 mission was doomed due to a ceramic heater which did not shut off (and got red hot)
12:32 PM xxcoder: thats when cermic heater is higher than needed temperate, and need controls on it
12:33 PM xxcoder: cermic heater cannot heat above its tuned temperate heat
12:34 PM lcnc-relay: <skunkworks8841@> I really should get the 'new' air comressor hooked up. this old one is getting sketchy. Dad fixed it atleast 40 years ago - had a bad bearing.
12:34 PM lcnc-relay: <skunkworks8841@> dad got one from an auction that doesn't seem to have had many hours..
12:34 PM xxcoder: https://www.amazon.com/100x30x6mm-Degrees-Thermostat-Aluminum-Preservation/dp/B0FD3LXJR7?th=1 whats selected is tuned to 50c
12:34 PM xxcoder: still too hot
12:35 PM rdtsc: that'll work, but consider using two thermostats in series, such that if one fails, there is a backup
12:36 PM xxcoder: it doesnt have any thermostat
12:37 PM xxcoder: it simply lowers power usage as it heats, reaching zero at target
12:37 PM rdtsc: can use a specific-temperature thermostat, like one of these https://www.digikey.com/en/products/filter/thermostats-mechanical/516
12:38 PM rdtsc: select desired switching temp
12:39 PM xxcoder: thats cool though the heater would be hot as it rises to switch temp so dunno
12:40 PM rdtsc: it would be nice if an intrinsic 40c heater existed, but I haven't seen such - maybe one does exist
12:41 PM xxcoder: I suppose can always put sensor right on cermic heater?
12:41 PM rdtsc: definitely
12:41 PM xxcoder: it might switch too often though
12:41 PM rdtsc: the thermal mass will prevent it from switching super-fast
12:42 PM xxcoder: ok. something to think about :)
12:42 PM rdtsc: bolt the heaters to a block of aluminum and that can greatly slow it down also
12:42 PM xxcoder: I want to heat a space to above freezing at all times, but entirely unattended
12:43 PM rdtsc: an aluminum heatsink (with cooling fins) can be a good heater for a small space
12:44 PM xxcoder: maybe. I would want entirely no smarts that might fail
12:44 PM xxcoder: best would be simply plug it in, and it wont get on fire even if it falls to floor
12:44 PM xxcoder: or on wall or whatever
12:46 PM xxcoder: having good smart device is great when its attended, I have PID controlled dumb heater in restroom and living room, and it works wonderfully well
12:47 PM rdtsc: failure-proof is kinda challenging really; we all know that "smart" devices can still be rather dumb at times
12:47 PM xxcoder: yeah. best would be purely power. no pid, no switches, no fan
12:48 PM xxcoder: when its going to be entirely unseen and unattended most of time
12:49 PM rdtsc: could take a (long) length of nichrome resistance wire, coil it into a spiral, and plug that directly into the wall outlet - if designed properly, would not get over some temperature. But that is very "dumb" and has no way to control the temperature of the space - may be 50c or 10c
12:54 PM * rdtsc is shutting down systems to avoid welding-induced resets... bbl
12:58 PM lcnc-relay: <allaboutjake@> Tom_L: Yeah, I think I'll do that because I'm hitting the same issue with ProbeBasic again: "VTKBackPlot has no attribute named machine_ext_scale"
12:59 PM Tom_L: jt & i are glad to help
01:01 PM Tom_L: touch screen or kbd & rodent?
01:02 PM lcnc-relay: <allaboutjake@> For reference on the ProbeBasic problem in case anyone wants to look: https://pastebin.com/hGpMJJVd
01:02 PM lcnc-relay: <allaboutjake@> Touchscreen. I've got Flex running.
01:02 PM lcnc-relay: <allaboutjake@> Or at lest the sims now.
01:02 PM Tom_L: i'm working on a touch version of touch-probe3 currently
01:02 PM Tom_L: i get the monitor tuesday
01:03 PM lcnc-relay: <allaboutjake@> Are there instructions for moving over the hal/ini config from an Axis setup into touch-probe3? The docs seem sparse on tht after getting the examples to run
01:03 PM Tom_L: absolutely
01:03 PM Tom_L: it's rather simple
01:03 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/rpi5/
01:03 PM Tom_L: some screens i've been working on
01:04 PM Tom_L: https://www.gnipsel.com/linuxcnc/flexgui/ini.html
01:04 PM Tom_L: create a [FLEXGUI] section and put most of it there
01:05 PM Tom_L: for the ui settings anyway
01:06 PM lcnc-relay: <allaboutjake@> ahh, my bad. I didn't go that far into the docs. I was looking for a page similar to (this page in the probe basic docs)[https://kcjengr.github.io/probe_basic/machine_config.html] and thought the ini page in the Flex docs was more of a reference type page
01:07 PM Tom_L: you're on discord?
01:07 PM lcnc-relay: <allaboutjake@> I did notice that keyboard jogging in the flex simulator seems like it can mis a key-up event and jogging just keeps going.
01:07 PM lcnc-relay: <allaboutjake@> yeah
01:07 PM Tom_L: there is also #flex there
01:07 PM Tom_L: i think jt fixed the kbd
01:08 PM Tom_L: make sure you have the latest deb
01:37 PM Unterhaus__ is now known as Unterhausen
01:39 PM -!- #linuxcnc mode set to +v by copper.libera.chat
02:12 PM -!- #linuxcnc mode set to +v by ChanServ
02:40 PM lcnc-relay: <skunkworks8841@> these pigion holes haven't seen the light of day in 30+ years
02:40 PM lcnc-relay: <skunkworks8841@> https://photos.app.goo.gl/QV4YD2LN97q58nZZ7
02:42 PM xxcoder: great. another unknown item
02:43 PM lcnc-relay: <skunkworks8841@> I did find a zenith carb in there.. lol - just not the same model
03:11 PM lcnc-relay: <lesliet@> wow, took all the sheet metal off the mill motor for painting, and discovered that the steel box on the side with a 120mm fan in it was not original. Someone added it, presumably to replace the missing fan on the end of the motor. :picard_facepalm:
03:11 PM lcnc-relay: <lesliet@> hand fabricated box... lotta work to do it poorly
03:12 PM lcnc-relay: <lesliet@> and they cut a big hole in the shroud for the motor
03:13 PM lfluffywof is now known as fluffywolf
03:15 PM lcnc-relay: <lesliet@> think I need to set up the vfd and do a run test, make sure the motor is viable before I source a fan for it and patch the hole
03:16 PM lcnc-relay: <lucid.nonsense@> skunkworks8841@: You should sell tickets
03:31 PM rdtsc-wtf is now known as rdtsc
04:25 PM _unreal_: ok delt with a ton of stuff in the garage
04:25 PM _unreal_: tossed a lot as well
04:45 PM _unreal_: ok dinner is cooking. I have a 4tb harddrive but is only showing up in gparted as 1.8tb?
04:45 PM _unreal_: am I missing something or WTF? I've no real experience with drives above 1tb also
04:45 PM lcnc-relay: <captainhindsight_.@> https://www.reddit.com/r/SweatyPalms/comments/1omf29j/a_huge_metal_product_being_machined_on_a_cnc/
04:46 PM _unreal_: I just want to format it and have it be a dump drive
05:09 PM roycroft: _unreal_: a mbr partitioning scheme can only address 2TB - you have to use GPT the full capacity of larger disks
05:09 PM roycroft: as well, some filesystems only support 2TB partitions
05:09 PM roycroft: make sure you use one that is 64 bit
05:51 PM rdtsc: oww, cap'n!
06:07 PM lcnc-relay: <lesliet@> captainhindsight_.@: oh, shiznit
06:23 PM xxcoder: IOS lol. seems beds that was always connected was found to use 16 gb data a day
06:24 PM xxcoder: and on people limited data access, it causes data to run out and bed cant confirm a sub so it rises one end so people cant sleep on it
06:25 PM xxcoder: errm *per month but still bad
06:45 PM lcnc-relay: <lesliet@> bed as a service?
06:45 PM lcnc-relay: <lesliet@> > .<
06:46 PM xxcoder: apparently.
06:46 PM xxcoder: service that could be done by local ibm xt
06:51 PM lcnc-relay: <lesliet@> does this bed serve a medical function?
06:52 PM lcnc-relay: <lesliet@> _and my brain flashes to the motel beds you stick quarters in the box_
06:52 PM lcnc-relay: <lesliet@> what if a nation state DDOs's my bed?
06:53 PM lcnc-relay: <lesliet@> (sorry, but lack of context is leaving me in silly mode)
06:56 PM lcnc-relay: <xxcoder@> the image
06:56 PM lcnc-relay: <xxcoder@> https://jauriarts.org/_heisenbridge/media/jauriarts.org/CaeaQWZIjTOhocWlFWZPjDuM/FcB6MompixQ/image.png
06:58 PM lcnc-relay: <lesliet@> oh dear gods
06:58 PM lcnc-relay: <lesliet@> what bed is this?
07:00 PM lcnc-relay: <lesliet@> nm, found the thread
07:04 PM lcnc-relay: <lesliet@> and now I can't stop laughing
07:22 PM _unreal_: welcome to the you will pay continuously for everything you already own environment.
07:23 PM lcnc-relay: <lesliet@> and like it
07:46 PM lcnc-relay: <skunkworks8841@> I forgot how well welding stick work for linkages
08:13 PM rdtsc: stick welding is suprisingly effective
08:18 PM _unreal_: I cant wait to get all my welding tools back on line
08:19 PM _unreal_: I just got one of these bldc's the other day b630-30c0-ms7
08:20 PM _unreal_: trying to find a pinout for it
08:20 PM _unreal_: asswholes cut the wire
08:27 PM _unreal_: got one of these as well rib-40-1000z0a
08:44 PM rdtsc: not seeing any datasheets for it
08:45 PM lcnc-relay: <skunkworks8841@> https://photos.app.goo.gl/jQHMX1CSNeboz4BZ6
08:51 PM lcnc-relay: <skunkworks8841@> might play with it a bit - I could get about .2 more movement out of the acellerator pump by cutting the new plunger shorter. (it has 3 places to put the keeper..
08:53 PM _unreal_: found the data sheet for the encoder but not for the motor. rather annoyed. they cut the cord rather shot. not so short that I cant recover but shot enough to make it not fun.
08:53 PM _unreal_: and SOB. looks like I have to reboot. my cpu freq is back to 1.6ghz
08:53 PM _unreal_: ? bios isnt holding settings
08:57 PM lcnc-relay: <skunkworks8841@> nope - that is all the movement you have.
08:57 PM lcnc-relay: <skunkworks8841@> ok.
08:58 PM lcnc-relay: <skunkworks8841@> still waiting on one vent/valve that mikes carberator had - so then I can put it back together
08:58 PM _unreal_: running rock solid at 2.7ghz now
10:23 PM lcnc-relay: <plaidrabbit@> I'm helping inoita11@ setup his machine (we have the same phsyical machine) He has a 7i97 card. the config tool leaves him a PWM generator section using board.0.pwmgen.00. stuff. Mine (on a diff version of linuxcnc, with a diff carrd setup) uses board.0.1.analogout0 the actual signal needs to be -10/+10 for a servo amp
10:24 PM lcnc-relay: <plaidrabbit@> Where do I find the pin config for the cards?
10:24 PM lcnc-relay: <plaidrabbit@> so I know how the 7i97 is setup, and if that's the correct pin to use