#linuxcnc Logs
May 27 2025
#linuxcnc Calendar
12:10 AM lcnc-relay: <meisterdippel@> good morning
12:29 AM lcnc-relay: <TurBoss> good morning
02:11 AM lcnc-relay: <meisterdippel@> I am currently constructing my own absolute encoder that outputs the data via RS485.
02:12 AM lcnc-relay: <TurBoss> nice!
02:13 AM lcnc-relay: <meisterdippel@> AS5600 -> ESP32-C3 -> RS485
02:13 AM lcnc-relay: <meisterdippel@> but search for a faster mag-encoder
02:22 AM lcnc-relay: <TurBoss> the AS5048A has 0.5º accuracy but the AS5600 0.4º but is more expensive 😲 it cames with spi and 14 bit resolution
02:22 AM lcnc-relay: what is the accuracy type good more or less ( looks less ) 😐️
02:24 AM lcnc-relay: <TurBoss> both EOL in 2031 😄
02:47 AM lcnc-relay: <meisterdippel@> i think i need one with spi
02:53 AM lcnc-relay: <TurBoss> 👍️
03:09 AM lcnc-relay: <meisterdippel@> WTF: AS5048A mouser: 6-8Euro
03:21 AM lcnc-relay: <meisterdippel@> i need the encoders for my BLDC-FOC stuff
03:22 AM lcnc-relay: <meisterdippel@> i need a cheap solution to retrofit old robots 🙂
03:51 AM lcnc-relay: <rs> https://docs.broadcom.com/doc/AS20-M42M-Series-DS
03:51 AM lcnc-relay: <rs> cheap-ish €90
04:07 AM lcnc-relay: <meisterdippel@> sweet
04:09 AM lcnc-relay: <meisterdippel@> https://www.youtube.com/shorts/ougZnA4ItKw
04:14 AM lcnc-relay: <rs> https://www.tme.eu/at/katalog/encoder_112582/
04:22 AM Tom_L: morning
04:24 AM lcnc-relay: <meisterdippel@> rs: but all incremental encoders, i need absolute
04:24 AM lcnc-relay: <meisterdippel@> and a little bit out of my price-range
04:30 AM Tom_L: https://www.ebay.com/itm/365476174475
04:30 AM Tom_L: that one may not be absolute
04:31 AM Tom_L: https://www.amazon.com/Rugged-Absolute-Encoder-E6C3-AG5B-E6C3-AG5C/dp/B0DF36Z9DP?th=1
04:31 AM Tom_L: based on the price
04:37 AM lcnc-relay: <meisterdippel@> ok ok, but i talk about chips, not complete encoders, i pay <5euro for a complete assembled board from jlcpcb
04:42 AM lcnc-relay: <meisterdippel@> https://github.com/multigcs/riocore/blob/dev/riocore/plugins/rioencoder/rioencoder/src/main.ino
04:51 AM lcnc-relay: <meisterdippel@> very simple code, read encoder, inc/decr revs , send via serial
04:54 AM lcnc-relay: <meisterdippel@> and if anyone is interested in how to read this back into the fpga: https://github.com/multigcs/riocore/blob/dev/riocore/plugins/rioencoder/rioencoder.v
05:14 AM mrec: can anyone help me to remove a shaft coupling?
05:14 AM mrec: https://www.cnczone.com/forums/mitsubishi-controls/479100-cnc.html#post2615470
05:15 AM mrec: behind the plate with the screws there should be 2 thrust bearings, they don't matter, but I cannot get the black shaft coupling off
05:18 AM mrec: is this some kind of heat shrink coupling?
06:13 AM lcnc-relay: <lackdaz@> has anyone tracked the rock solid mill project? looks like its grown legs and aspires to become a linuxcnc https://www.printables.com/model/1237272-rock-solid-milling-machine-v09/comments
06:20 AM lcnc-relay: <meisterdippel@> i like the '3d printing and concrete' mix idea
06:29 AM Deejay: hi
07:07 AM rdtsc-w: meister, what about a ring magnet, like is used on small fans, and two hall sensors placed 90 degrees apart? sin / cos resolver output.
07:37 AM lcnc-relay: <meisterdippel@> I am satisfied with the normal mag encoder ic's , very easy to install and easy to handle
07:47 AM rdtsc-w: some have surprisingly fast output too; good for high speeds
07:50 AM lcnc-relay: <meisterdippel@> but i need to read this data and sin/cos is not so simple and needs adc's
08:07 AM rdtsc-w: sin/cos is old tech, so is well-supported: https://duckduckgo.com/?q=sin+cos+resolver+ic
08:19 AM lcnc-relay: <skunkworks8841@> andy has a sketch that runs on an arduino - it works ok - but only has 1000? counts per rev - iirc.
08:20 AM lcnc-relay: <skunkworks8841@> (resolver)
08:20 AM lcnc-relay: <skunkworks8841@> does 3 axis
08:21 AM lcnc-relay: <skunkworks8841@> 0-1023
08:21 AM lcnc-relay: <skunkworks8841@> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?ResolverToQuadratureConverter
09:41 AM rdtsc-w: something like the https://www.analog.com/en/products/ad2s1210.html can take in sin/cos and emulate an encoder
09:47 AM lcnc-relay: <meisterdippel@> it's ok now 🙂 i have all components i need, thanks to all
09:57 AM lcnc-relay: <skunkworks8841@> I think jon elson makes resolver -> quadrature boards using the above chip.
09:57 AM lcnc-relay: <skunkworks8841@> mesa also makes a resolver interface for their boards
09:58 AM lcnc-relay: <skunkworks8841@> https://pico-systems.com/resolver.html
10:11 AM roycroft: hi folks
10:17 AM lcnc-relay: <meisterdippel@> skunkworks8841@: the AD2S1210 is cool, yes, and i can use it as an additional rio interface/plugin, only problem is the price ~30Euro 😭
10:18 AM lcnc-relay: <meisterdippel@> I hope that one day I will be able to read out a sin/cos via normal ADCs and calculate it in the FPGA
10:22 AM lcnc-relay: <rs> i found some DOGE optimzed documentation in linuxcnc. https://linuxcnc.org/docs/html/gcode/g-code.html#gcode:g52
10:22 AM lcnc-relay: <rs> or maybe that is a deliberate AI trap?
10:23 AM lcnc-relay: <rs> the link should probably point to https://linuxcnc.org/docs/html/gcode/coordinates.html#sec:g52-and-g92-offsets
10:43 AM roycroft: ev sales are up 28% in the eu this year so far
10:43 AM roycroft: tesla sales are down 49% in the eu this year so far
10:43 AM lcnc-relay: <travis_farmer@> LMAO
10:44 AM roycroft: tesla are saving a lot of money by not having to buy and ship as may cars as they use to
10:44 AM roycroft: er, build and ship
11:02 AM lcnc-relay: <roguish> roycroft: how's the back coming along ????
11:09 AM roycroft: i'm feeling much better today, having taken the day off yesterday from moving
11:10 AM roycroft: saturday we moved the last of my big, heavy stuff
11:11 AM roycroft: it's been almost 2 months since the surgery, so i should have full bone strength very soon - i have lifting restrictions because it takes time for the bones to grow around the screws that hold the rod in place
11:11 AM lcnc-relay: <roguish> yeah, go easy. ya don't wanna mess up the fix....
11:11 AM roycroft: it will be months before i'm fully healed, but i'm like 80+% functional now, as long as i take it easy
11:15 AM * roycroft is figuring out where to put limit switches on the cnc router while waiting for controller parts to arrive
11:17 AM * djdelorie always wondered if it was worth the effort to install "near the limit" switches, to speed up homing...
11:42 AM rdtsc-w: https://linuxcnc.org/docs/html/config/ini-homing.html should be able to approach very rapidly, overshoot to a stop, backup, and reapproach slowly for accuracy
11:43 AM roycroft: i do not like the homing switches i have, and the ones that xxcoder provided are almost identical
11:44 AM roycroft: they are too large to conveniently mount on the mini-router
11:45 AM roycroft: they are mechanical switches, and i see that some folks prefer inductive limit switches
11:45 AM roycroft: any thoughts on one vs. the other?
11:45 AM roycroft: it seems to me that mechanical switches would be more reliable
11:46 AM rdtsc-w: ditto roycroft... I am using these: https://www.digikey.com/en/products/detail/omron-electronics-inc-emc-div/D2VW-01L1B-1HS/368540
11:46 AM roycroft: but if inductive switches are 4 nines reliable vs 5 nines reliable for mechanical (arbitrary numbers) the effective difference may be insignificant
11:51 AM mrec: https://i.snipboard.io/6vlGTI.jpg does anyone know how to get that coupling off safely without destroying it? Did anyone do something like that before?
11:52 AM mrec: well without damaging it, rather than destroying it.
11:53 AM rdtsc-w: if no rubber (hard to tell), try heat - loctite probably used
11:53 AM mrec: https://i.snipboard.io/HjIxCa.jpg
11:57 AM roycroft: odd - the inductive limit switches seem to be almost universally normally open
11:58 AM roycroft: normally closed would be the safer circuit
12:01 PM rdtsc-w: mrec I bet JT-Shop knows and has done it - bet he is out playing with his car lift at the moment
12:07 PM xxcoder: roy wonder if you could use current invertor after sensor, so in least wire failures wont be an issue, but sensor failure would still be.
12:09 PM roycroft: that's the problem with a normally open circuit
12:09 PM roycroft: if a component in the circuit fails, or the wiring fails, there is no way to detect it
12:09 PM xxcoder: yep. I always prefer NC
12:10 PM roycroft: i think i've figured out the y axis with the mechanical switches
12:10 PM roycroft: so i guess i'm 1/3 of the way done
12:11 PM roycroft: x should not be too hard, but there's not much room for limit switches on the z axis
12:11 PM xxcoder: I was thinking 2 designs for z
12:11 PM lcnc-relay: <skunkworks8841@> https://www.youtube.com/watch?v=vnIabBllhmg
12:11 PM xxcoder: one is adding tiny hole to holder and put in switching in such way that moving up pushes switch in correct spot
12:12 PM xxcoder: other is adding flange to spindle holder in such way that it presses switch on way up, also at correct spot trigger
12:12 PM rdtsc-w: spring-load the switch, such that it is NC and use it in reverse
12:13 PM roycroft: i've found some nc inductive limit switches
12:13 PM roycroft: i'm just surprised taht 90+% of them are no
12:16 PM xxcoder: er I meant rod in first type I thought up
12:16 PM xxcoder: honestly I was leaning towards second design idea
12:18 PM roycroft: i'm not sure a limit switch on the positive end of the z axis makes sense, as the tooling stickout will vary
12:18 PM roycroft: but the negative end, certainly
12:19 PM xxcoder: I was sued to z zero to be top
12:19 PM xxcoder: used
12:19 PM roycroft: yes
12:19 PM roycroft: and it goes positive as it goes down
12:20 PM roycroft: so what i'm saying is a limit switch at the top and not at the bottom
12:20 PM xxcoder: yah both of my ideas is at top
12:21 PM roycroft: x and y can have limit switches at both ends
12:26 PM djdelorie: rdtsc-w: wrt homing, I have a shared home/limit setup. If I approached the "home" sensor at full speed, "overshoot" would rip the end of the machine off
12:28 PM djdelorie: I would need a second sensor a couple inches away that tells it to start slowing down. I guess that would make it a separate home/limit setup, but I wonder if it would be faster
12:28 PM roycroft: some folks do that, or set soft limits in the controller software
12:28 PM djdelorie: soft limits won't help before homing
12:28 PM roycroft: with this being such a tiny machine i'm not too concerned about things getting ripped off
12:28 PM roycroft: no, of course not
12:29 PM roycroft: maybe not use rapid when homing?
12:29 PM roycroft: rapids
12:29 PM djdelorie: if I forget to park the axes before powering off, homing takes forever
12:29 PM roycroft: but again, on a tiny machine like mine that would work out well
12:29 PM JT-Cave: you set the search and latch speeds for homing
12:29 PM roycroft: with a big machine it could take a long time to home
12:29 PM djdelorie: ten feet of homing, worst case
12:30 PM JT-Cave: there is a jog function...
12:30 PM roycroft: yes, i was going to say
12:30 PM djdelorie: JT-Cave: yes, but again, if I overshoot the combined home/limit sensor by more than 1/8" it hits the hard stop
12:30 PM djdelorie: and you can't jog a dual-Y before homing
12:30 PM roycroft: power on, jog at high speed to get close, then run the homing sequence at low speed
12:30 PM djdelorie: I end up powering *off*, pushing the gantry to the end, and powering back up
12:31 PM JT-Cave: what I do on my plasma is I have a button that sends the joints to the home position so on startup homing only take a few seconds
12:31 PM roycroft: or remember to park :)
12:31 PM roycroft: but if the power goes off it doesn't matter
12:31 PM djdelorie: I have a script to park everything, yes. o<sleep>call
12:31 PM roycroft: you're still stuck somewhere weird
12:32 PM djdelorie: if I'm doing maintenance on the machine it's not always in the park position at powerup
12:32 PM JT-Cave: sleep that's a good idea here shortly
12:33 PM JT-Cave: found out my lg fridge has wifi and is still in warranty
12:33 PM JT-Cave: the fridge thinks it's ok but I know better
12:34 PM lcnc-relay: <skunkworks8841@> nothing saying the home sensor has to be at the end of the travel
12:34 PM lcnc-relay: <skunkworks8841@> put it in the middle.
12:34 PM djdelorie: if it's in the middle, won't it need to signal for the whole half of travel?
12:34 PM lcnc-relay: <skunkworks8841@> yep
12:35 PM djdelorie: I'd thought of mounting a 5 foot steel bar under there, for this purpose...
12:35 PM lcnc-relay: <skunkworks8841@> or - over the last 4 inches..
12:35 PM djdelorie: or that
12:35 PM lcnc-relay: <skunkworks8841@> whatever works for your accelleration
12:35 PM JT-Cave: so I had to modify one of the plugins for wp to remove the enhancement and now there's another update
12:35 PM djdelorie: then just tell the config that the home sensor is at the 6 inch mark?
12:36 PM lcnc-relay: <skunkworks8841@> yes - if you watch the video I posted - the home on some of the axis is on the ramp.
12:36 PM lcnc-relay: <skunkworks8841@> when you re-home -it backs off and finds it again.
12:37 PM JT-Cave: if your on the wrong side of the home switch it will not be found...
12:37 PM lcnc-relay: <skunkworks8841@> sure - but with that - you can't be on the wrong side. (it can't over-senter)
12:37 PM lcnc-relay: <skunkworks8841@> center
12:39 PM xxcoder: roy honestly router is so tiny it took me only a minute to manually "home" it. I power it on, move it to near home position, turn steppers off, manually set position, then power it and set home opn all. :) but with switches I would just use certain speed (not rapid) and double tap the switch, second tap at slower speed.
12:39 PM xxcoder: even rapid it wouldnt be too fast I think but been years
12:40 PM djdelorie: I wish linuxcnc allowed a dual-Y gantry to rapid before homing...
12:41 PM roycroft: manually homing is no big deal
12:42 PM roycroft: the switches are also there to prevent runaway movements from bad gcode
12:42 PM roycroft: and again, with such a tiny machine the consequencs are not so dire
12:43 PM roycroft: but this tiny machine is also a place for me to experiment/learn for my build of the bigger machine
12:43 PM xxcoder: indeed. had impact of side once while testing. essentallt nothing happened other than having to retighten coupler.
12:49 PM djdelorie: on my "little" cnc, I had a broken limit switch once. Drove the spindle right through the side mount.
12:50 PM xxcoder: ouch
12:51 PM roycroft: i view safety devices such as limit switches as insurance, not guarantees
12:51 PM djdelorie: with a combined home/limit you need both
12:52 PM roycroft: there really are no guarantees
12:52 PM roycroft: just better or worse insurance
12:52 PM djdelorie: that's why you keep an estop switch handy :-)
12:52 PM roycroft: which is also just good insurance
12:52 PM roycroft: because it could also fail
12:53 PM djdelorie: and you hope it fails safe
12:53 PM roycroft: yes, but one never knows
12:53 PM djdelorie: yup, and I've broken my machine in those cases
12:54 PM djdelorie: "what's the worst thing that could happen?" er, that?
12:54 PM * roycroft always considers edge cases, no matter how improbable
12:55 PM roycroft: with the mini-router, the worst that could happen, assuming i keep my various body parts away from it while it's operating, is that it will destroy itself and/or the parts being made, make a big mess, and possibly catch on fire
12:55 PM roycroft: none of which is likely
12:56 PM xxcoder: yep
12:56 PM djdelorie: it's always better to let it fail than to stick your fragile soft fingers in there and fail along with it
12:57 PM xxcoder: thankfully cnc machines dont need blood to work. unlike IT repairs
12:57 PM roycroft: with a big cnc machine one has to start considering things like structural damage to the building as possibilities as well
12:57 PM lcnc-relay: <skunkworks8841@> use a safety relay then..
12:57 PM roycroft: what if the safety relay fails?
12:58 PM lcnc-relay: <skunkworks8841@> it is designed to fail safe.
12:58 PM roycroft: what if it does not fail safe?
12:58 PM lcnc-relay: <skunkworks8841@> that is why they are usually required in the real world
12:58 PM xxcoder: roy yep ran machines big enough that they had to reforce floor under it, deeper concerete etc.
12:58 PM roycroft: the real world often requires redundant safety systems
12:59 PM xxcoder: otherwise it could break floor on rapids that dont go well
12:59 PM roycroft: which even then are not absolute guarantees
12:59 PM xxcoder: even with that, I could still feel machine doing a short rapid, few feet away from me
01:00 PM xxcoder: 800 pounds moving 1 inch or so, so fast I couldnt see it move. I sure could feel it tho
01:00 PM roycroft: it's almost impossible to design a machine that is 100% safe under all circumstances
01:00 PM roycroft: you want to design to be 5 nines or 6 nines or 7 nines safe
01:00 PM roycroft: triple redundancy of critical safety systems, etc.
01:01 PM djdelorie: and be on the other side of the planet, or better, on another planet
01:01 PM roycroft: but those machines are in the universe
01:01 PM roycroft: and in the univese, entropy happens
01:02 PM roycroft: on my mini-router, 3 nines safe is probably good enough, but i'll try to build it better than that
01:03 PM xxcoder: yeah. its not exactly dangerous to anyone but quite careless people
01:07 PM rdtsc-w: I'll have to add some striped caution tape to my X axis, as the Z plate creates two fairly nasty pinch-points against the Y gantry uprights
01:09 PM rdtsc-w: could leave an inch or so, but dangit I paid a lot of money to get this far, want to use every square inch
01:10 PM lcnc-relay: <skunkworks8841@> lol - cough42 is converting a mill to cnc and the home switch for the y axis is a crazy pinch point. like shear you finger off.
01:10 PM rdtsc-w: at that point, I'd put a pause switch in there :)
01:11 PM lcnc-relay: <skunkworks8841@> anti tie down..
01:17 PM roycroft: i think i've figured out good places to mount the mechanical limit switches
01:17 PM roycroft: now for the fun part - metal fabrication without a lathe or milling machine
01:18 PM roycroft: fortunately it's very simple fabrication
01:20 PM rdtsc-w: files, hammers, grinders, taps all existed long before cnc
01:22 PM roycroft: yes, and i have all of that
01:23 PM roycroft: i also have my surface plate and precision layout tools
01:24 PM roycroft: all my small metal stock is still at the old shop, though
01:24 PM roycroft: but i'll be getting that by this coming weekend at the latest
02:48 PM lcnc-relay: <skunkworks8841@> starship to launch in 4 hours
02:49 PM lcnc-relay: <skunkworks8841@> re-using the booster
03:20 PM -!- #linuxcnc mode set to +v by ChanServ
03:33 PM roycroft: if it's either the bezos rocket or the musk rocket i hope there are no people in it and i hope it crashes somewhere remote and i hope i don't hear anything more about it
03:52 PM -!- #linuxcnc mode set to +v by ChanServ
03:52 PM -!- #linuxcnc mode set to +v by ChanServ
03:56 PM lcnc-relay: <skunkworks8841@> lol - yes. I could understand that. I have to separate the man from the company. They are doing amazing engineering
04:02 PM roycroft: i usually do, but i don't have to and in this case i certainly won't
04:55 PM xxcoder: ltest tech connection video is about captions and yeah I have experenced that issue, in least till I simply stopped using tv
05:42 PM roycroft: so the shaper benchpilot demo will be at 4pm pdt this thursday, in case anyone is interested in seeing how it works
05:43 PM roycroft: and since it's a shaper sessions episode, it will almost certainly feature jake and russ
05:43 PM roycroft: so there will be some inadvertent comic relief
06:49 PM xxcoder: WOLFBOX 150000 RPM Jet Dry Blower Portable Turbo Jet Air Blower Electric Jet Fan
06:50 PM xxcoder: crazy. it blows almost like air pressure, but weaker of course. but no hose and no compressor
06:50 PM xxcoder: battery life is fairly short, but plenty to clear floor or something
06:54 PM Tom_L: just use a leaf blower
06:54 PM roycroft: this trend towards wimpy, fragile battery tools where you have to swap batteries every 3 minutes is hopefully just a fad
06:54 PM xxcoder: dont have one, and cant use it to clean inside house lol
06:55 PM xxcoder: wolfbox is usb chargable so no batteries to swap
06:55 PM roycroft: battery weed whacker, battery lawn mower, battery table saw, etc.
06:55 PM roycroft: they always seem underpowered and cheaply built
06:55 PM xxcoder: yeah those definitely dont replace actual plug tools
06:56 PM roycroft: which is odd, because smaller tools like a battery sawzall or impact driver can be more powerful than their corded equivalents
06:56 PM xxcoder: isnt it because it can suck more amps out of batteries than plug?
06:57 PM roycroft: maybe it's the dc motor
06:57 PM xxcoder: hmm maybe
06:58 PM roycroft: some middle tier, in terms of power consumption, tools such as a circular saw or sawzall are kind of a wash when it comes to battery vs. corded
06:58 PM xxcoder: anyway yeah it have power settings, max power may be short, but I probably wouldnt use it to clear out dust from pc for example
06:58 PM roycroft: and i have battery versions of some of those tools for convenience
06:59 PM xxcoder: indeed
06:59 PM xxcoder: it all comes down to usercase
07:00 PM roycroft: i know part of the popularity of some of the big battery tools is that they are allegedly "greener" than their ice counterparts
07:00 PM roycroft: but that just depends on where you get your electricity
07:00 PM xxcoder: this blower probably wont be used in huge shop trying to clean floor, but maybe for cnc machine interior for example
07:00 PM roycroft: with an ice lawn mower you make the pollution in person in real time
07:00 PM xxcoder: 1 mower 1 hour = hmm 100? cars hour drive pollution
07:01 PM roycroft: with a battery lawn mower a large percentage of people are just buying energy that was made by polluting, but farther away
07:01 PM xxcoder: mower is far less regulated after all
07:01 PM xxcoder: yeah even that case, its more effecient
07:01 PM xxcoder: and less polluting
07:02 PM roycroft: i just got back from hauling things to the storage containers
07:03 PM xxcoder: oh its 40 cars, not 100
07:04 PM JT-Shop: yup unless your electric device is powered by solar, wind, or water it's a loss of energy
07:05 PM xxcoder: yeah. but then coal or gas plantions is far far more regulated than mower
07:05 PM xxcoder: so ultimately still quite a bit less pollution
07:05 PM xxcoder: well for now
07:05 PM roycroft: small ices are hard to make clean
07:06 PM JT-Shop: the problem is the conversion loss
07:06 PM roycroft: i had a propane weedeater for a while
07:06 PM roycroft: that was nice
07:06 PM roycroft: it burns a lot cleaner than gasoline
07:06 PM roycroft: and still has power
07:08 PM roycroft: and yes, jt-shop, turning chemical energy into electrical energy and then into mechanical energy, or mechanical energy into electrical energy back into mechanical energy usually is not nearly as efficient as skipping the electrical conversion in the midde
07:08 PM roycroft: middle
07:09 PM * JT-Shop calls it a night
07:10 PM JT-Shop: https://gnipsel.com/images/wildfire-lift/wf-24.jpg
07:10 PM * roycroft calls it a late afternoon
07:11 PM xxcoder: still lots to go eh jt
07:11 PM roycroft: nice
07:11 PM JT-Shop: yup, lots of work left
07:11 PM JT-Shop: but getting close
07:13 PM roycroft: the older i get the less interested i am in working on vehicles
07:14 PM JT-Shop: I find it rewarding to work on older vehicles
07:17 PM roycroft: yeah, i can see that
07:17 PM roycroft: i just find vehicles messy to work on, for the most part
07:18 PM roycroft: working on a well-maintained vehicle that doesn't have mud and oil an stuff caked on it all over the place would be ok to work on
07:18 PM JT-Shop: yup
07:18 PM roycroft: i also never had a garage to work on vehicles in
07:18 PM roycroft: i was always a driveway mechanic
07:19 PM JT-Shop: oh yeah I remember working on vehicles in the snow... what fun
07:19 PM Tom_L: pretty good progress there
07:19 PM roycroft: so my disinterest in working on them is more due to circumstances than anything else
07:19 PM JT-Shop: understand that for sure
07:19 PM Tom_L: out of necessity instead of hobby
07:20 PM roycroft: exactly, tom_l
07:20 PM roycroft: i worked on vehicles because i could not afford to pay someone to work on them
07:20 PM JT-Shop: same here
07:20 PM Tom_L: that and finding one you can trust
07:20 PM roycroft: abom is doing a tour of the starrett museum in atholl mass this week
07:21 PM roycroft: it's interesting seeing the old, mostly prototype tools and machines
07:22 PM roycroft: i would like to start work on the limit switch mounts for the mini-router, but with all my small metal stock still at the old shop that will have to wait :(
07:23 PM roycroft: for wiring them up, i'm thinking i can wire them all in series
07:23 PM mrec: JT-Shop: do you know about heatshrunk couplings?
07:24 PM JT-Shop: no
07:24 PM Tom_L: mrec, you figure out that was heated on the shaft?
07:24 PM roycroft: and program the homing sequence so that every step reads the same input, and assumes that when the circuit is broken it's reached the limit on the axis it's currently homing
07:24 PM roycroft: is that a common way to do it?
07:24 PM mrec: Tom_L: I just think so
07:24 PM Tom_L: roycroft, single input for them all?
07:25 PM roycroft: yes
07:25 PM Tom_L: no i used an io for each one
07:25 PM roycroft: i have plenty of io pins
07:25 PM roycroft: i just want to conserve on wire
07:25 PM Tom_L: not saying it won't work but you don't know what axis faulted
07:25 PM xxcoder: that was why I decided against single wire for all 3
07:26 PM roycroft: well if you're homing x, y and z should not be moving, so when it faults it's good to assume it's x
07:26 PM roycroft: but conversely, if done that way, each axis has to be homed individually, in sequence
07:26 PM Tom_L: i home x and y at the same time
07:26 PM Tom_L: Z first
07:26 PM roycroft: and that would take a lot longer
07:26 PM Tom_L: then x & y
07:26 PM JT-Shop: yup
07:26 PM Tom_L: the homing sequence is in the ini
07:28 PM roycroft: on an axis with limit switches at both ends it would be fine to wire those serially, though
07:28 PM Tom_L: and on a gantry i would assume you'd have 2 switches. one per stepper on a 2 stepper gantry
07:28 PM roycroft: on this little machine there's one motor for each axis
07:29 PM Tom_L: i do share home and limit switches but have a +- wire per axis
07:29 PM JT-Shop: say goodnight gracie
07:29 PM roycroft: ciaoder
07:31 PM flyback: https://github.com/FlyGoat/csmwrap
07:42 PM roycroft: it's kind of ironic that jt is installing a car lift right now, as skunkworks is busy digging out the car lift in the shop he's cleaning up
07:51 PM * _unreal_ scans the channel O,o .... o.O..... O,O
08:19 PM xxcoder: whoa https://youtu.be/pxxvCUDeS1w
08:19 PM xxcoder: truck explodes and essentally disappears
08:24 PM roycroft: sounds like it blowed up real good
08:25 PM xxcoder: sounds like I dont know, but man its like truck is there then it arent
08:44 PM lcnc-relay: <big_kevin420@> justr saying it exploaded isnt a good enouight answer for me
08:45 PM lcnc-relay: <big_kevin420@> how does somethign like that just explode
08:49 PM xxcoder: propone leak apparently
09:09 PM roycroft: my wish was fulfilled
09:09 PM roycroft: the spacex rocket had no humans in it, it crashed, and in a remote area
09:55 PM lcnc-relay: <plaidrabbit@> I'm trying to come up with a good idea on how to fabricate a part. I have a 4'x8' router. I need to take a 3/4"x3/4"x1/8"x20' piece of aluminum L, and drill holes in it. The tricky part is that I need to drill the holes after it's been gently bent a bit, with about 1/3 of it bent around a roughly r3000mm curve. I need to place the holes within about +/- .2mm or so. Not too crazy tight tolerances. My plan is to take a piece of...
09:55 PM lcnc-relay: ... MDF, and carve a 1/8" channel that is the exact curve I need into it, form the part to the curve. I now have a set of soft-jaws for holding the part, and that'll let me CNC drill holes from the top. But the holes from the side are a bit of a challenge.
09:56 PM lcnc-relay: <plaidrabbit@> My backup plan is I'll hand-drill those parts, but I'm trying to come up with a good primary plan.
09:57 PM xxcoder: can't drill it pre-bend?
09:58 PM lcnc-relay: <plaidrabbit@> Bending it changes the exact length in some of the sections, and I'm worried about accumulated error.
09:58 PM * roycroft was wondering why it can't be pre-bent
09:58 PM roycroft: er, pre-drilled
09:59 PM roycroft: take a piece of scrap, pre-drill, bend, and measure
09:59 PM lcnc-relay: <plaidrabbit@> But... you've got me thinking, I can do at least half the holes pre-bending, cause it's a long straight section
09:59 PM roycroft: so you can pre-drill the easy holes :)
09:59 PM xxcoder: the curved, I suppose you could add wood block and have it cnc it into fixture for drilling other holes
09:59 PM lcnc-relay: <plaidrabbit@> Well, fewer spots to screw up as well
10:00 PM xxcoder: depends on how many of those, it might be worth it
10:06 PM lcnc-relay: <plaidrabbit@> Humm, can I not paste images in here or something? Or am I failing to operate discord?
10:06 PM roycroft: no idea, i'm not on discord
10:06 PM roycroft: so i would not see your images anyway
10:07 PM xxcoder: you can paste into off topic then get url for image there (use view, not displayed)
10:07 PM xxcoder: then paste url
10:07 PM xxcoder: or use some image site url
10:07 PM roycroft: if you need semi-precision then the thing to do might be to machine some drill guides on the cnc router and use those to index a hand drilling operation
10:10 PM lcnc-relay: <plaidrabbit@> https://cdn.discordapp.com/attachments/970073761770512384/1377119050571710554/image.png?ex=6837cd8f&is=68367c0f&hm=2cb4de77a8c716cfc026d4e93b7d1af4ced3e30e103b2f3f76cf2a2b21726b14&
10:10 PM xxcoder: hmm from what I see, theres 2 probably easy to do sections
10:10 PM lcnc-relay: <plaidrabbit@> I'm basically hanging as sheet on the top and outside of the L's.
10:11 PM xxcoder: so 2 fixtures, first for just unbent stright, then second one after bend, the latter probably hand drill the curvy part
10:13 PM roycroft: you could machine a little roller wheel with spikes spaced $holeinterval apart, sharpen the spikes, harden the wheel, line up the first hole on the part, then roll the wheel along it, thus marking the hole locations
10:13 PM lcnc-relay: <plaidrabbit@> Im making a big-ass tapered box pretty much.
10:13 PM roycroft: kind of like a stitch marking wheel that leathercrafters use
10:15 PM lcnc-relay: <plaidrabbit@> Well, the outer skin will be pre-drilled, and that's why the tolerance. It just has to line up with the skin.
10:15 PM lcnc-relay: <plaidrabbit@> And I can use the skin as a template
10:15 PM xxcoder: yeah can use tape to measure it also
10:15 PM xxcoder: tape it over 2 or 3 times, punch the holes
10:15 PM xxcoder: remove and measure
10:15 PM roycroft:
10:16 PM roycroft: a fiberglass tape would not stretch
10:16 PM xxcoder: ah yeah. forgot that thing lol
10:16 PM lcnc-relay: <plaidrabbit@> As long as it mates to the skin, that's where the tolerance comes from
10:17 PM roycroft: what we're getting at is that sometimes automation is not the solution
10:17 PM Tom_L: it isn't?
10:17 PM roycroft: especially in a semi-complex part that is only made once
10:17 PM xxcoder: yeah. if only 1 or 2 manual probably is better
10:17 PM lcnc-relay: <plaidrabbit@> Well, you have me thinking about drilling the middle straight section
10:17 PM roycroft: or twice, in this case, as it appears you're making mirrored parts
10:17 PM lcnc-relay: <plaidrabbit@> Yep, I just need one LH and one RH part
10:18 PM roycroft: you *could* devise a fixture for it and do the calcs and cnc it
10:18 PM roycroft: but by the time you did all that work you probably could have done it manually and still had time for a refreshing beverage while you admire your work
10:18 PM lcnc-relay: <plaidrabbit@> You've got me thinking the right way. I'll CNC the long straight section, and that'll make it easy to do the curved section
10:18 PM lcnc-relay: <plaidrabbit@> +by hand
10:19 PM roycroft: hybrid machining is pretty normal
10:20 PM Tom_L: oddly similar :) https://hips.hearstapps.com/pop.h-cdn.co/assets/17/24/1497379529-vsv.jpg?resize=980:*
10:20 PM lcnc-relay: <plaidrabbit@> Yeah, I'm just trying to make it less hybrid. I was just going to CNC the top. Now I'll CNC the top and 1/3 of the side
10:21 PM roycroft: still just as hybrid, but less on the manual side and more on the automation side
10:22 PM lcnc-relay: <plaidrabbit@> Plus in reality, it makes it much easier, because it's being held together/lined up
10:23 PM lcnc-relay: <plaidrabbit@> https://cdn.discordapp.com/attachments/970073761770512384/1377122326218543134/image.png?ex=6837d09c&is=68367f1c&hm=6fb393c137a4aaab9abf8ddc252d545d572b8107fe03259dcb1948e37c12929c&
10:23 PM lcnc-relay: <plaidrabbit@> That's it with a lot more parts
10:23 PM Tom_L: homebuild aircraft?
10:23 PM xxcoder: interesting
10:24 PM lcnc-relay: <plaidrabbit@> yep
10:24 PM roycroft: so you'll be holding the parts in with clecos as you do the manual hole locations/drilling
10:25 PM lcnc-relay: <plaidrabbit@> now I can just rivet the middle 1/3 in place
10:25 PM lcnc-relay: <plaidrabbit@> or that
10:25 PM roycroft: that makes the manual operations even easier
10:25 PM lcnc-relay: <plaidrabbit@> Yeah, cause the skin will be pre-drilled
10:25 PM * roycroft doesn't understand why the aircraft industry is the only one that uses clecos
10:26 PM roycroft: they are so useful for so many things
10:26 PM lcnc-relay: <plaidrabbit@> cause aircraft industry is the only ones doing this much manual sheet metal work, with the need to partially do it, and take it apart.
10:27 PM xxcoder: oh thats the thing wesley treat always uses
10:27 PM xxcoder: he makes large signs using sheets also so it makes sense.
10:27 PM roycroft: it makes sense for a lot of things, but nobody uses them except the aircraft folks
10:28 PM xxcoder: mr. treat doesnt make plants but basically most non-airplane people doesnt I guess
10:28 PM lcnc-relay: <plaidrabbit@> well, it only makes sense if you need to make it, and pull it apart after drilling too do stuff like finishing/deburring
10:28 PM xxcoder: *planes
10:28 PM roycroft: which is not unique to aircraft construction
10:29 PM roycroft: i'll concede that the scale of doing those kinds of operations is much greater on an aircraft than most other things
10:30 PM roycroft: making them pretty much essential for an aircraft, while only very useful for most other applications
10:30 PM roycroft: anyway, i did not mean to get us sidetracked on that
10:30 PM lcnc-relay: <plaidrabbit@> I'd asssume more fields would just rivet the parts together, and finish the parts as one. I don't think most fields care about burrs between the materials reducing the strength.
10:31 PM roycroft: only people who care about appearance, if strength is not an issue :)
10:31 PM lcnc-relay: <plaidrabbit@> then the rivets will cover up any burrs.
10:31 PM roycroft: btw, i used to use them for automotive sheet metal applications that were not riveted
10:32 PM roycroft: i would fit the part with clecos, then pull the clecos one at a time and spot weld through the hole
10:32 PM lcnc-relay: <plaidrabbit@> Okay, why/how? It only makes sense to me to hold the riveted surfaces together? Are they going to be bolted in the final product?
10:32 PM lcnc-relay: <plaidrabbit@> ah
10:32 PM roycroft: works a charm
10:32 PM roycroft: you get strong, welded fitments with no weld bulge
10:33 PM lcnc-relay: <plaidrabbit@> this design is hilariously simple. It's all pop-rivets. (a very specific brand, but not an expensive one)
10:33 PM lcnc-relay: <plaidrabbit@> * 99%
10:54 PM roycroft: it looks like a fun fabrication project, plaidrabbit
10:54 PM roycroft: are you an experienced aircraft builder?
11:13 PM lcnc-relay: <plaidrabbit@> Nope. First project!
11:15 PM lcnc-relay: <plaidrabbit@> I wanted to build something, but the 640 was the only one that fit my general mission. I wanted something with a higher weight limit than most other aircraft. Most kits are only 2 seaters with a tight weight limit.
11:16 PM roycroft: it will be a lot of fun - i'm sure it already is
11:16 PM roycroft: and you'll develop heaps of new skills
11:17 PM lcnc-relay: <plaidrabbit@> Already am
11:17 PM lcnc-relay: <plaidrabbit@> I'm several years into this project...
11:17 PM roycroft: so you're nearing the partway point :)
11:18 PM lcnc-relay: <plaidrabbit@> Yeah, I've got the rudder and stab mostly fabricated.
11:18 PM xxcoder: certainly long project. nice
11:19 PM lcnc-relay: <plaidrabbit@> Yep. I've gotten back onto the project after a bit of a break. I was stuck on one part for a while, but I got some help, and I'm almost unstuck on that part
11:21 PM lcnc-relay: <plaidrabbit@> There's a spot where a few parts of the plans just... disagree with each other. So I paid a guy to actually spend some time just designing that small section.
11:25 PM xxcoder: interesting. I wonder why design error happened
11:26 PM lcnc-relay: <plaidrabbit@> a) not a lot of them have been built. b) it's normally done by hand, so small errors are ignored. c) it's a small-ish error. d) this part is mentioned as being a PITA to do, probably because of the design error.
11:27 PM lcnc-relay: <plaidrabbit@> It's off by a couple inches in design. And I've realized there's 3 conflicting statements, and only two of them can be true.
11:27 PM xxcoder: while deaf pilots do exist, I am not remotely close to being able to fly one, let alone build one lol
11:27 PM lcnc-relay: <plaidrabbit@> Yeah, you just have to fly out of untowered airports.
11:28 PM lcnc-relay: <plaidrabbit@> What region are you in, if you don't mind me asking? I can tell you how realistic it is.
11:28 PM xxcoder: oh im at tacoma, wa but yeah I dont plan to ever fly a plane lol
11:29 PM lcnc-relay: <plaidrabbit@> If you want, you could totally get in a few lessons if you want. It's not that expensive, if you save up.
11:29 PM xxcoder: sadly also poor retiree, forced into retirement by health issues. (not deafness, always been deaf)
11:30 PM lcnc-relay: <plaidrabbit@> if I had a plane, and I was up in the washington area, I'd help you out
11:31 PM xxcoder: lol riding a small plane have been my minor wish
11:31 PM lcnc-relay: <plaidrabbit@> You can knock that out for about $150-200
11:32 PM xxcoder: interesting. maybe someday :) thought it was more
11:32 PM roycroft: i've had friends who were private pilots
11:32 PM lcnc-relay: <plaidrabbit@> For that price point, you'll get about a 1h flight in a bit of a junker aircraft.
11:32 PM roycroft: i've flown in lots of small planes
11:33 PM lcnc-relay: <plaidrabbit@> a lot of places have what are known as "discovery flights" Basically discounted first flights to hopefully get you in the door.
11:33 PM roycroft: does anyone know if those cheap z-touch plates are tolerably repeatable?
11:34 PM roycroft: i'm interested not because they're cheap, but because they're small
11:34 PM roycroft: kind of like that first line of cocaine that the dealer gives you for free, plaidrabbit? :)
11:34 PM lcnc-relay: <plaidrabbit@> Pretty much. Just with worse profit margins for the dealer.
11:35 PM * roycroft thinks fire season will be arriving soon
11:37 PM lcnc-relay: <plaidrabbit@> Pretty much the bare minimum actual cost to run a c-172 is about $75-100/hr. Normally billed out at 150-200/hr.
11:37 PM lcnc-relay: <plaidrabbit@> +marginal
11:41 PM lcnc-relay: <plaidrabbit@> xxcoder@ If you want to dream a little harder, I can point you at a few things you can read up on.
11:49 PM xxcoder: sorry was in call. well still is but yeah
11:50 PM lcnc-relay: <plaidrabbit@> want links, or maybe some other time?
11:50 PM xxcoder: nah its fine :)
11:50 PM xxcoder: thanks for offer tho!
11:50 PM lcnc-relay: <plaidrabbit@> sure