#linuxcnc Logs
Dec 20 2022
#linuxcnc Calendar
12:03 AM CaptHindsight[m]: small watch gears and similar may be 3d printed now
12:14 AM XXCoder: sls type?
12:14 AM XXCoder: probably
12:15 AM solarwind: CaptHindsight[m] you never disappoint
02:47 AM -!- #linuxcnc mode set to +v by ChanServ
03:21 AM Guest4421 is now known as Guest4421b
03:49 AM Loetmichel: *brrr* China-quality. I just tried to put two 8mm holes into an old Harddisk (to destroy it) on the drill press. Second hole started to glow... 8mm "TiN" Drill bit came out 8mm shorter and molten into a glob at the tip. What stuff do they use to make that "HSS"? Lead?
03:58 AM jpa-: maybe it is a glass based platter?
04:01 AM Loetmichel: nope, was alu platter
04:02 AM Loetmichel: i checked
04:04 AM sensille: tin drill? high speed solder?
04:04 AM Loetmichel: sensille: certainly felt like it, yes
04:05 AM travis_farmer[m]: G'Morning
04:05 AM Loetmichel: on the other hand: it lit up bright red... Tin would probably be as liquid as water at that point, not a paste that can wrap around the rest of the drill :)
04:06 AM XXCoder: alum alloy probably
04:11 AM Deejay: moin
04:53 AM Tom_L: morning
05:40 AM JT-Cave: morning
08:37 AM travis_f[SHOP]: very odd issue. firstly, it is still quite cold in my shop, but i can't imagine that would affect things. my config that homed fine yesterday, now sends the Y axis homing in the wrong direction. i have changed nothing in that regard. been debugging for half an hour, and nothing i do fixes it. i can change the homing search direction, but then the home switch doesn't trigger. could the cold be affecting the mesa card? it is about 30F
08:41 AM travis_f[SHOP]: jogging is still the correct direction
08:47 AM travis_f[SHOP]: hmmm, no signal from the home switch... none at all...
08:54 AM travis_f[SHOP]: hmmm, me thinks the mechanical home on Y has failed... time to replace it with one of the inductive sensors i have, when i figure out how to mount it...
09:07 AM JT-Cave: prox switches work well if you have home to index...
09:12 AM travis_f[SHOP]: hmm, installed the ind-prox sensor, temporarily with some tape, and it sends a signal, but hal meter shows no change on the pin
09:37 AM unterhaus: stewmac wants $23 to ship a small package from central ohio to central Pa
09:38 AM unterhaus: at least it's not going to Canada
09:38 AM travis_f[SHOP]: changed pins on the mesa, and still nothing.
09:40 AM * travis_f[SHOP] heads inside, defeated :-(
09:44 AM unterhaus: I think I would find another mechanical switch to test with just for a sanity check
09:45 AM Bleepshop: unterhaus: He's on the internet. I would say sanity check failed. ;D
09:46 AM unterhaus: I might be crazy but I'm not stupid
09:49 AM roguish[m]: roycroft: hey, did you get a little rock-n-roll last night??? 6.3
09:51 AM travis_farmer[m]: unterhaus: the opto-isolator boards i use have LEDs on both the 24V (machine side) and 5V (mesa card side) sides. both LEDs responded to the ind-prox sensor, and i tested with a DMM to verify output. the sensor is sending a signal to the mesa card. i will play more after lunch. my mind is a tad burnt out right now...
09:52 AM unterhaus: it's a 7i96? Which input?
09:53 AM travis_farmer[m]: 7i80HD-16, ping 57
09:53 AM travis_farmer[m]: s/ping/pin/
10:01 AM unterhaus: ping!
10:02 AM travis_farmer[m]: ping what?
10:03 AM unterhaus: I just like calling pins ping. All my boards are going to be labeled that way going forward
10:04 AM unterhaus: I think I've spotted the problem, there is no pin 57
10:04 AM sensille: pings output, pongs input?
10:05 AM travis_farmer[m]: the 7i80HD-16 has no pin 57? that will be news to a lot of people...
10:06 AM unterhaus: is that the fpga pin?
10:06 AM travis_farmer[m]: it is the pin according to HAL
10:07 AM unterhaus: okay, so connector 3 pin 19?
10:08 AM unterhaus: io57
10:08 AM travis_farmer[m]: yes
10:09 AM travis_farmer[m]: hm2_7i80.0.gpio.058.in
10:10 AM travis_farmer[m]: *currently, i changed pins
10:10 AM unterhaus: is your opto board configured to pull down the pin?
10:11 AM travis_farmer[m]: yes
10:14 AM travis_farmer[m]: should be "hm2_7i80.0.gpio.058.in_not" actualy, but hal meter still shows no activity, regardless of which one i use
10:16 AM zephyr9900: zephyr9900 likes to make the gpio inputs ON when high. That way if a signal wire is shorted to ground it will not turn an input on. Short to ground is more likely because chassis and machine iron is grounded. Is there a downside to that way of configuring?
10:17 AM unterhaus: a lot of inputs are unreliable in that configuration
10:17 AM unterhaus: the fpga has pull up resistors and it's better to signal by pulling them down
10:18 AM unterhaus: in this case there is an IO board that probably doesn't care, and a short on the connection between that io board and the fpga is unlikely
10:18 AM zephyr9900: Oh yes if going directly to the FPGA but I use the opto isolated inputs
10:18 AM travis_farmer[m]: well, i gotta make me some lunch. back in a while..
10:20 AM zephyr9900: I think NFPA 70 requires start logic for motor start buttons so that a short to ground will not cause the motor to start.
10:20 AM unterhaus: I have always used a mesa input board and they are usually configured one way or the other
10:20 AM unterhaus: your i/o board would take care of that
10:24 AM zephyr9900: I am using 7i37 for gpio so you can drive an input either way.
10:26 AM unterhaus: right, so you can be fairly sure what the fpga is seeing
10:30 AM roycroft: wow
10:30 AM roycroft: big quake just off the coast of northern california - 6.0
10:31 AM roguish[m]: 6.4
10:31 AM zephyr9900: yup
10:31 AM roycroft: centered off of humboldt county, i think - the news just said that there was some damage in ferndale
10:31 AM roycroft: which is just south of eureka
10:33 AM roguish[m]: no tidal wave though. damn it.
10:41 AM unterhaus: I bought some magnetic v-pad knockoffs and the magnets were disappointing
10:41 AM unterhaus: so I got some of the strong hands ones and the magnets are also disappointing
10:42 AM roycroft: maybe your iron is what is disappointing :)
10:42 AM unterhaus: I have a negative aura
10:44 AM unterhaus: I was hoping the Chinese would come up with some better maganets
11:11 AM zephyr9900: iocontrol.0.emc-enable-in - Should be driven FALSE when an external estop condition exists. halui.estop.activate bit in pin for setting Estop (LinuxCNC internal) On It seems like these 2 inputs kind of do the same thing. How are they related? What is the difference? It seems like these are some areas where the documentation could be
11:11 AM zephyr9900: better and I would like to suggest improvements if I can figure out some of these details.
11:42 AM travis_f[SHOP]: back at machine. the signal from my Y home sensor is correct, all the way to the mesa card, yet the mesa card does not report change on that pin. i checked with a scope and DMM. voltage change is correct, as is polarity of signal. the only thing different from when it worked, until it did not work, is that it is simply a different day...
11:50 AM zephyr9900: travis_f Maybe add all of the gpio inputs to the watch window in HAL_configuration window from axis gui and see if anything works?
11:51 AM zephyr9900: and save as a watch list for latter use
11:54 AM travis_f[SHOP]: the other axis will home, so the 7i80 must be working. i will add a bunch of pins in the area to watch, and see...
11:55 AM jpa-: travis_f[SHOP]: could be burned IO pin due to overvoltage
11:57 AM travis_f[SHOP]: could be, jpa-
12:02 PM travis_f[SHOP]: i put the wire on a far different mesa pin, no responce. seeing a trend here. going to try a different pin on my opto-isolation board
12:07 PM roycroft: swap the cable from the working axis to the pins for the non-working axis
12:07 PM roycroft: you need to determine if the problem follows the io pins on the mesa card, the cable, or the axis hardware
12:08 PM travis_f[SHOP]: good point
12:23 PM travis_f[SHOP]: well... this is embarrassing...
12:23 PM travis_f[SHOP]: i forgot the pullup resistor on the 24V side of my opto-isolation board for that channel
12:24 PM travis_f[SHOP]: all is well in the kingdom again... (hangs head in shame)
12:25 PM roycroft: don't feel shame
12:25 PM roycroft: just be happy that it was a simple fix
12:25 PM zephyr9900: glad you found it to be simple.
12:25 PM roycroft: and a no-cost one
12:26 PM travis_f[SHOP]: true...
12:32 PM Guest46 is now known as smcollins
12:33 PM smcollins: anybody here familiar with C ?
12:34 PM sensille: [x]
12:40 PM roycroft: c
12:40 PM roycroft: c spot run
12:40 PM roycroft: run, spot, run
12:40 PM roycroft: that should take some of us back to when we were 3-5 years old :)
12:53 PM travis_farmer[m]: i didn't read those stories when i was a kid... i read the world book encyclopedias (true story, actualy)
12:54 PM roycroft: the dick, jane, and sally readers were used up until the 1970s
12:55 PM travis_farmer[m]: <-- 1979
12:55 PM roycroft: but i think everyone here who is 60 or older grew up with them
12:55 PM roycroft: they were the standard readers from the 1930s until the 1970s
12:55 PM roycroft: everyone here who grew up in the united states, i should say
12:59 PM travis_farmer[m]: my reading teacher didn't think i could read. then he asked my mother what i read at home... "World Book Encyclopedia"... sure shocked that teacher :-)
12:59 PM CaptHindsight[m]: tell me where you grew up without telling me where you grew up
01:00 PM travis_farmer[m]: lobster, paper-mills...
01:01 PM CaptHindsight[m]: travis_farmer: did they then beat you until you read only the designated books?
01:01 PM travis_farmer[m]: no, the teacher had to up his game on reading level
01:03 PM CaptHindsight[m]: I was once told to leave the section of the library where they kept the books on physics, chemistry, electronics etc etc by the librarian and to go to the children's section
01:03 PM travis_farmer[m]: dang, that is harsh
01:03 PM CaptHindsight[m]: we mostly had assholes for librarians and school teachers
01:05 PM travis_farmer[m]: being Autistic (undiagnosed at the time), most of my teachers didn't know how to handle me. but i had one teacher, he was very good (went on to become a state biologist)
01:07 PM CaptHindsight[m]: one year they wanted to fail me for science the same year I won the school and district science fair
01:07 PM CaptHindsight[m]: travis_farmer: I'm glad that you ran into someone decent back in your early school years
01:08 PM travis_farmer[m]: that was in my freshmen year of HS actualy ;)
01:09 PM CaptHindsight[m]: I think my kindergarten teacher was a decent human, i think i had some others in HS
01:10 PM travis_farmer[m]: my kindergarten teacher thought i was wild... just because i was climbing the shelves...
01:11 PM zephyr9900: OK I have a 3 speed gearbox with tally switches. I was using the hal component called spindle which accommodates different gears. That was working until I found out that it wouldn't work for reverse direction because it would not output negative speed (technically velocity) the hy_gt_vfd component does not have direction inputs but responds to
01:11 PM zephyr9900: negative speed. So Andy Pugh suggeted I use the lincurve component that he made and uses for his lathe. I have that hooked up and working Now the issue I have is getting spindle-at-speed to work. My commanded speeds are pretty close to the actual speed reported by the encoder.2.velocity-rpm. The hy_gt_vfd has an at-speed output but it is based on
01:11 PM zephyr9900: what the VFD thinks the speed is and it is off by too much. The hy_gt_vfd at-speed only goes true when it is within 2%. My actual speeds from the encoder are actually about 2.5 to 5% accurate but the VFD does not agree. So is there a component that can just make the desired speed to actual speed comparison and give me a TRUE output if within
01:11 PM zephyr9900: some tolerance?
01:12 PM CaptHindsight[m]: "My actual speeds from the encoder are actually about 2.5 to 5% accurate but the VFD does not agree. "
01:13 PM CaptHindsight[m]: why not?
01:13 PM CaptHindsight[m]: which do you wish to please?
01:14 PM zephyr9900: Not sure exactly why the VFD thinks the speed is different from the actual. Maybe because the rotor slip is less under light loads.
01:14 PM CaptHindsight[m]: how does the VFD measure its speed?
01:14 PM zephyr9900: I don't know for sure but I assume it estimates based on what I thinks the rotor slip is or should be?
01:14 PM zephyr9900: *it
01:15 PM CaptHindsight[m]: sounds like you might have to scale whatever the VFD gets for feedback
01:15 PM CaptHindsight[m]: IF you want to make the VFD happy
01:15 PM zephyr9900: I think the VFD is happy enough. I want LinuxCNC to be happy
01:17 PM zephyr9900: Other VFD components let you put an at speed tolerance in but hy_gt_vfd is fixed to 2% - that seems pretty tight.
01:18 PM zephyr9900: Im just running in V/Hz mode so of course there is an error from rotor slip.
01:19 PM CaptHindsight[m]: what is slipping?
01:20 PM zephyr9900: Induction motor rotor. They do not run at synchronous speed. They all have what is known as "slip"
01:21 PM CaptHindsight[m]: but you want it synchronized within 2% of commanded speed
01:22 PM zephyr9900: That is why the nameplate will give a rated speed that is not a multiple of the line frequency ie nameplate says 1725 but synchronous is 1800
01:22 PM zephyr9900: I compensate with the lincurve component.
01:22 PM CaptHindsight[m]: but you want it synchronized within 2% of commanded speed
01:22 PM jdh: There is a 'near' component that let's you specify nearness
01:22 PM zephyr9900: I don't need it to be that close
01:22 PM CaptHindsight[m]: yeah or change the rules
01:24 PM * zephyr9900 looks to add near to my hal file - thank you jdh !
01:25 PM CaptHindsight[m]: lie to the VFD
01:26 PM zephyr9900: CaptHindsight that is what I _am_ already doing! I think.
01:30 PM CaptHindsight[m]: zephyr9900: are you on the Milwaukee area?
01:30 PM CaptHindsight[m]: on/in
01:31 PM zephyr9900: yes
01:35 PM CaptHindsight[m]: nice makerspace up there
01:36 PM CaptHindsight[m]: with two locations
01:37 PM zephyr9900: Oh yes I was there once for a tour of it but that was maybe 10? years ago. I have my own makerspace :]
01:37 PM zephyr9900: So I haven't been to both locations
01:41 PM CaptHindsight[m]: I haven't seen the new space
01:42 PM CaptHindsight[m]: but I used to drive past the older one frequently
01:45 PM CaptHindsight[m]: and they have a sense of humor https://wiki.milwaukeemakerspace.org/mms-api
01:51 PM * zephyr9900 checked it out. Yup I was there a while back. Yes it is nice. but It is a bit of a drive for me.
01:53 PM zephyr9900: CaptHindsight and jdh At speed working now with "near" component. Thank you!
01:54 PM CaptHindsight[m]: 👍️
01:54 PM zephyr9900: I use 3 instances of lincurve - one for each gear and then use the tristate_float to mux them back to the VFD speed command
01:55 PM zephyr9900: Now I am almost ready to make something with the ClausingNC!
01:55 PM CaptHindsight[m]: zephyr9900: how is the system setup mechanically? Is there a gear change of some sort?
01:56 PM zephyr9900: Its a manual gear change leaver on the geared headstock. 3 speeds
01:57 PM zephyr9900: Inside there are limit switches to show the gear
01:58 PM zephyr9900: machine was made in 1972 and originally run by feeding a paper tape.
01:58 PM zephyr9900: I have serial number 003 I doubt there were many made and maybe this is the only one left?
01:59 PM zephyr9900: But because this machine was really a lot of work to use in its day it seems to have little mechanical wear. Its like new.
02:00 PM zephyr9900: I even have the Clausing Manual for the machine :-)
02:02 PM * zephyr9900 goes to get some (late) lunch then will add some tools to the turret for my first real job.
02:05 PM CaptHindsight[m]: zephyr9900: I picked up a manual Clausing a few years ago. It came with some original cans of spindle and way oil.
02:06 PM CaptHindsight[m]: https://i.imgur.com/xw5PV90.jpg
02:08 PM CaptHindsight[m]: I think it needs some new spindle bearings now
02:08 PM CaptHindsight[m]: the original hydraulic speed changer still works, just needed to be bled and refilled
02:08 PM CaptHindsight[m]: but it runs on 3-ph now
02:11 PM CaptHindsight[m]: https://i.imgur.com/OVgL40b.jpg I give it a 6/10 as a manual lathe
02:36 PM roycroft: 3 phase is so pedestrian
02:36 PM roycroft: why don't you mix it up a bit, give it a dave brubeck treatment, and run it on 5 phase?
02:37 PM CaptHindsight[m]: oohh run 3 phase from two sides! 6 phase
02:37 PM CaptHindsight[m]: but it is a pedestrian lathe
02:38 PM bjorkintosh: is there (other than machining cloud) an open database of tools?
02:38 PM CaptHindsight[m]: factory stock is single phase
02:40 PM CaptHindsight[m]: bjorkintosh: that spies on you and tries to manipulate you as much as possible?
02:40 PM bjorkintosh: haha. hopefully without the surveillance.
02:47 PM Centurion_Dan1 is now known as Centurion_Dan
02:51 PM CaptHindsight[m]: and they want you to pay for the service
02:51 PM roycroft: https://eugene.craigslist.org/hvo/7559077544.html
02:51 PM roycroft: i suppose there is some scrap value to that
02:52 PM roycroft: but the way it was removed from the lathe, i think it would be hard to index it to any other machine
02:53 PM CaptHindsight[m]: universal adapter, just needs welding
02:55 PM CaptHindsight[m]: https://i.imgur.com/BdJAScA.jpg scam or not?
02:55 PM CaptHindsight[m]: spray on fat reducer
02:56 PM XXCoder: 1000% scam
03:00 PM roycroft: if the spray is flammible, and you ignite it, the fat would burn off quickly
03:00 PM roycroft: i would be 100% effective, but with some negative side effects
03:00 PM XXCoder: leaving just feet in shows wit rest burnt off
03:01 PM XXCoder: you remember that in news in umm 90s?
03:01 PM roycroft: i do not
03:01 PM roycroft: but i found much of the '90s not particularly memorable
03:01 PM roycroft: and 'memberberries had not been invented yet
03:02 PM XXCoder: its too old to be easily found on internet
03:03 PM CaptHindsight[m]: the 90's were the birth of LCNC, grunge and Seinfeld
03:04 PM XXCoder: anyway basically there was few cases of person burning up completely leaving just feet in shoes
03:04 PM roycroft: spontaneous combustion?
03:05 PM XXCoder: ah yes thats the term. been too long
03:05 PM CaptHindsight[m]: and Bat Boy
03:06 PM XXCoder: https://en.wikipedia.org/wiki/Spontaneous_human_combustion
03:06 PM XXCoder: capt nah it was real, though "it magically starting fire" is myth part. theres ingition sources
03:11 PM CaptHindsight[m]: https://en.wikipedia.org/wiki/Ball_lightning
03:11 PM CaptHindsight[m]: and St Elmo
03:11 PM XXCoder: whats funny is it was thought to be a myth for long time
03:11 PM XXCoder: but you can make one in microwave now
03:12 PM XXCoder: https://youtu.be/ydSN2BhBCdk
03:13 PM CaptHindsight[m]: instant St. Elmo , Sriracha flavored
03:21 PM sensille: something changed from 2.8.3 to the current master, my config doesn't load anymore. something in mb2hal from float to s32?
03:22 PM CaptHindsight[m]: sensille: maybe
03:22 PM CaptHindsight[m]: lots of changes
03:23 PM CaptHindsight[m]: https://github.com/LinuxCNC/linuxcnc/commits/master/src/hal/user_comps/mb2hal/mb2hal_HOWTO.ini
03:23 PM roycroft: two deaths related to the ferndale quake are reported now
03:24 PM CaptHindsight[m]: sensille: well a few changes this year
03:26 PM sensille: strange. this version compatibility looks the wrong way round. but maybe i'm just tired
03:33 PM CaptHindsight[m]: roycroft: did you notice it?
03:35 PM roycroft: no
03:35 PM roycroft: folks on the southern coast of oregon did, but not up here
03:37 PM sensille: after a handfull of missing python packages: segmentation fault --> tomorrow
03:38 PM sensille: maybe 2.8.3 is a good version for now
03:49 PM JT-Cave: my daughter sent me some chicken seasoning https://www.amazon.com/Big-Cock-Chicken-Poultry-Seasoning/dp/B011Q21H1G
03:57 PM CaptHindsight[m]: what are the going rates for CNC time? $70/hr is what shops charge customers and they pay operators ~$30/hr?
03:58 PM zephyr9900: CaptHindsight I have the same Clausing Flame Hardened badge for mine. But is loose and I haven't figured out where it goes yet.
03:59 PM CaptHindsight[m]: 👀
03:59 PM CaptHindsight[m]: zephyr9900: have pics of the lathe?
03:59 PM zephyr9900: I wanted to drill a hole in the headstock to re-apply the NC badge but the casting is so hard I couldn't get a HSS twist drill to drill into it. I didn't know cast iron could be that hard
04:01 PM zephyr9900: https://photos.app.goo.gl/DokKB5WGWhqpSmuf9
04:01 PM CaptHindsight[m]: heh how I found out that some cheap Chinese "carbide" spot drills were not carbide
04:02 PM CaptHindsight[m]: I must have tried something on ebay
04:03 PM CaptHindsight[m]: zephyr9900: you model looks pretty rigid
04:03 PM zephyr9900: https://photos.app.goo.gl/DokKB5WGWhqpSmuf9
04:05 PM zephyr9900: yes its really kind of goofy though. the tool post is kind of in the center of the cross slide so you cant really reach anything that is bigger around than what goes over the cross slide. Just over 6" capacity over slide.
04:05 PM * JT-Woodshop never wastes money on cheap cutting tools just to throw them away
04:06 PM CaptHindsight[m]: not even on wood?
04:06 PM zephyr9900: https://photos.app.goo.gl/bCxwo8fdwGRWNup17
04:12 PM * JT[m]1 uploaded an image: (246KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/LxkRNnhSWOXcdXkMJjeqEGJQ/20221220_161031.jpg >
04:19 PM zephyr9900: wow thy guy that sold me the lathe put a new 3 jaw chuck on it or maybe the guy before him did. I just chucked a piece of .875" round bar and measured a TIR of 0.002" Im excited to have a decent 3 jaw chuck. I've never seen one that good for not being a set true type.
04:20 PM zephyr9900: Its rusty in the pics but I have since cleaned it and re-greased it
04:22 PM roycroft: that's the runout with the jaws at 0.875"
04:22 PM roycroft: try another size bar and see what it is
04:23 PM * JT-Woodshop never uses a 3 jaw chuck on the metal lathes
04:23 PM zephyr9900: yes I'm sure it will be different with a different size piece
04:23 PM zephyr9900: I have a 4 jaw but I need a D1-4 back for it
04:25 PM * JT-Woodshop ponders a safe way to cut the segments to 22.5°
04:29 PM roycroft: a chop saw would be good for that
04:29 PM unterhaus: I should build a sled for my current project making blocks
04:29 PM roycroft: or a good miter gage with an auxliiary fence on the table saw
04:30 PM roycroft: i built a sled for making kumiko frames - it is kind of interesting
04:30 PM unterhaus: kumiko frames have random angles?
04:31 PM roycroft: i used jonny tromboukis' design
04:31 PM roycroft: no, but it has a number of fixed angles
04:31 PM roycroft: https://www.youtube.com/watch?v=1JyUgvOnjHM
04:32 PM roycroft: that sled is really nice for cutting the half-laps and 1/3-laps (2/3 laps?) in the framework
04:43 PM CaptHindsight[m]: zephyr9900: the 6" limit is a feature, it keeps the operator from trying to use the lathe on parts that are too large for it
04:44 PM CaptHindsight[m]: not like the $900 HF lathes that have a 23" swing and 60" CC :)
04:45 PM unterhaus: that's an awfully big lathe to just swing 6"
04:45 PM CaptHindsight[m]: https://www.harborfreight.com/7-inch-x-10-inch-precision-mini-lathe-93212.html if you are turning cheese
04:46 PM unterhaus: work used to have a clausing cnc. I think it was a 15" lathe
04:51 PM roycroft: that 7x10 lathe can actually be improved a lot and can turn useful parts
04:53 PM CaptHindsight[m]: yes, HF tools are kits
04:53 PM CaptHindsight[m]: you just have to finish them
04:53 PM roycroft: yes, and that's true of most imported machines these days
04:54 PM CaptHindsight[m]: you mean Chinese or all imports even from Canada?
04:54 PM roycroft: and there's nothing inherently wrong with that as long as you set your expectations correctly
04:54 PM CaptHindsight[m]: so perception is reality?
04:54 PM roycroft: i thought canadia was a wholly-owned subsidiary of usa, inc.
04:55 PM CaptHindsight[m]: if that is your perception
04:56 PM CaptHindsight[m]: what exactly is a modern machine tool? say modern means made in the past 20 years
04:56 PM XXCoder: states 51 to 60
04:57 PM CaptHindsight[m]: solarwind: they are picking on your country again
04:57 PM XXCoder: capt some canadians already cite some usa laws as their own ;)
04:58 PM CaptHindsight[m]: like when to mooses come to an intersection at the same time?
05:00 PM CaptHindsight[m]: or is it two misse?
05:00 PM travis_farmer[m]: two moose
05:00 PM travis_farmer[m]: ;-)
05:04 PM travis_farmer[m]: time for supper, KFC tonight! and therefore, i sign off for the night. :-)
05:08 PM roycroft: it's tuesday where i live
05:09 PM roycroft: i believe that means i need to go find a taco cart for my dinner
05:09 PM roycroft: if only we had elected hillary in 2016 we'd have a taco cart on every corner
05:09 PM roycroft: i distinctly remember the republicans making that promise
05:10 PM roycroft: which is the main reason i'm sad she lost
05:11 PM CaptHindsight[m]: roycroft: that was a code word for cardboard box as a home
05:27 PM Tom_L: nothing like a pot of chili on a cold day
05:31 PM JT-Woodshop: yup
05:32 PM JT-Woodshop: crap I cut the slot a few thou too wide... must be time to quit
05:34 PM Tom_L: nothing a little paint won't fix
05:37 PM JT-Woodshop: actually it's exactly the width of 3x5 card
05:40 PM CaptHindsight[m]: JT-Woodshop: I use the same metric for quitting when it's late
05:41 PM CaptHindsight[m]: others are, did it break the skin and how far did it fly out of the machine
05:45 PM CaptHindsight[m]: might be of interest https://github.com/auto-mation-assist/LinuxCnc-OpenDDS-Work
05:55 PM JT-Woodshop: snafu fixed it's really time to quit now
05:59 PM zephyr9900: andypugh_ remember my post about how to cut fine splines? I am close to being ready to try it. I am using 4140 RC 30 steel and a tool like a threading tool on its side. I need to grind a tool. I have a carbide tool and it has 0 degrees back rake. Do you think that will work OK? its basically like a metal shaper kind of operation to cut each spline.
06:00 PM andypugh_: This was the blind spline up to a shoulder?
06:03 PM andypugh_: Maybe a tool like this? https://www.eccentricengineering.com.au/products/diamond-tool-holder/the-diamond-tool-holder
06:04 PM zephyr9900: yes but because I will use the ClausingNC I can make passes that back off at an angle near the shoulder
06:06 PM andypugh_: I think we talked about cutting the spline conventionally and then pressing a collar on?
06:06 PM andypugh_: (Which is what they often do inside a gearbox)
06:23 PM zephyr9900: These are external splines that I want so I want to try similar to the keyway cutting video but external. Hold a thread cutting tool sideways in the tool post with the point at the center height. A tool path like this https://drive.google.com/file/d/1JDe9ZSE6ZkbylSPcYsOPy1g5ZhTrIBre/view?usp=share_link
06:25 PM zephyr9900: spline detail https://drive.google.com/file/d/1Qg4ZtX2TORvTQnHhRY_OEbrm6pB5OLtS/view?usp=share_link
06:26 PM zephyr9900: so the spline does not go completely to the shoulder. the tool back out on each pass but in a different spot.
06:28 PM zephyr9900: So anyway I will grind and set up a tool to try it. I was wondering if you had any thoughts on the tool if it should have a back rake or not. Maybe I should just try it and see what happens.
06:58 PM CaptHindsight[m]: zephyr9900: can you tell how the factory makes them?
07:00 PM Tom_L: https://learnmech.com/shaper-tools-types-of-shaper-tools-classification-of-shaper-tools/
07:00 PM zephyr9900: andypugh_ thought maybe like this; https://youtu.be/7hVuH9XH948
07:00 PM zephyr9900: but that is too expensive for just one part
07:01 PM Tom_L: i used to use a thread roll similar to that spline roll
07:01 PM XXCoder: ice to see ice around here
07:02 PM Tom_L: zephyr9900, you're wanting to single point the spline?
07:02 PM XXCoder: seems snow just missed this region, and it rained then froze lol
07:02 PM zephyr9900: I have an index wheel mounted to the back of the chuck and I will just run the program to shap it for each tooth
07:02 PM zephyr9900: *shape
07:02 PM Tom_L: you would probably shape the tool like you would a shaper tool
07:02 PM Tom_L: link provided
07:02 PM zephyr9900: exactly
07:02 PM andypugh_: It might still need a run-out groove to break the chip. But see how it goes with just tapering out.
07:02 PM zephyr9900: it looks like a thread tool on its side in the tool post
07:03 PM Tom_L: use a thread tool?
07:03 PM Tom_L: what angle?
07:03 PM zephyr9900: hmm what do you mean by run out groove
07:03 PM zephyr9900: 60 deg
07:03 PM zephyr9900: it seems to match pretty close
07:04 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/thread_mills/Thread_mill2.jpg
07:04 PM Tom_L: run a thread mill down the side instead of 'shaper' down the top
07:04 PM Tom_L: live spindle
07:05 PM Tom_L: dunno what shape you need exactly
07:05 PM zephyr9900: there is not room against the shoulder for a spinning tool
07:05 PM Tom_L: where is the chip going when it reaches the shoulder?
07:06 PM Tom_L: you may want a slight undercut at the shoulder
07:06 PM zephyr9900: The tool will progressivly back out near the shoulder on each pass
07:06 PM zephyr9900: See tool path pic above
07:07 PM Tom_L: i did
07:08 PM zephyr9900: Plus I need to make it with the machines that a hobbyist has:-) I don't have an rotary table or indexer for milling machine.
07:08 PM Tom_L: hmm, i figured you had a rotary head
07:10 PM zephyr9900: someday maybe well actually I have a Niken 16 inch servo controlled rotary table but it weighs about 550 lbs and even if I get it working it is such a beast.
07:13 PM zephyr9900: Maybe tomorrow I will finish the setup and give it a try
07:13 PM andypugh_: You could potentially skive the spline. It’s not easier, but it would be cooler: https://youtu.be/VJOem40ggkI?t=1148
07:16 PM zephyr9900: andypugh_ yes that looks cool it reminds me of rotary broach - kind of. but then I need a live tool or a rotary table.
07:17 PM andypugh_: It is basically rotary broching. And if the pitch of the cutter is correct then it _ought_ to freewheel into synch, so the cutter coud just freewheel. In theory,
07:19 PM CaptHindsight[m]: https://youtu.be/VJOem40ggkI?t=1258 take a closer look at the teeth
07:19 PM * zephyr9900 would spend an inordinate amount of time making a tool that can do that
07:21 PM zephyr9900: I think the tool might need a slight helical tooth shape.
07:22 PM zephyr9900: In order to cut straight teeth like https://youtu.be/7hVuH9XH948 does the same think I think
07:22 PM zephyr9900: That is the link andypugh_ shared months ago
07:32 PM CaptHindsight[m]: https://www.youtube.com/watch?v=tuEqKKQWZJ8
07:35 PM CaptHindsight[m]: no mill or lathe required https://youtu.be/vsXGgYrHaFo?t=44
07:35 PM CaptHindsight[m]: brute force
07:35 PM Tom_L: i'm guessing that setup costs more than either a mill or lathe
07:36 PM Tom_L: less wear on the tooling
07:37 PM CaptHindsight[m]: some old cheap beams ground into shape and some decent bearings
07:44 PM zephyr9900: its nice to see how to do it in mass production.
07:54 PM CaptHindsight[m]: the Alro outlet near me has some i-beams for cheap, scrap price but cut to ~5ft lengths
07:55 PM CaptHindsight[m]: was thinks of making a few glass lathes out of them
07:55 PM XXCoder: interesting
07:58 PM CaptHindsight[m]: ~25lbs per linear ft
07:58 PM CaptHindsight[m]: so decent thickness
07:59 PM CaptHindsight[m]: oh H beams even better
08:01 PM CloudEvil: I wanna see a glass lathe.
08:01 PM CloudEvil: (made of glass)
08:01 PM CloudEvil: Do _not_ drop the chuck.
08:02 PM Plaid: Just a random report. I communicate with my VDF via RS232. I got a lot of noise on the connection when my VFD started. Added a cheap $20 opti-isolator from amazon. Now works with about 80% reliability. I lose a small number of messages.
08:03 PM CaptHindsight[m]: Plaid: what is your baud rate?
08:03 PM Plaid: 9600
08:03 PM CloudEvil: Add two optoisolators in series and you should get 96% reliability.
08:03 PM roycroft: that sounds like progress
08:03 PM CloudEvil: Do you own any form of oscilloscope?
08:03 PM Plaid: Nope
08:03 PM CloudEvil: At 9600, I guess even audio card scope may work
08:03 PM Plaid: The serial connection on the side of the VFD probably needs a ferrite core around it
08:04 PM CloudEvil: Optoisolator I'd have thought would wholly cure that
08:04 PM CloudEvil: I'd also want to check the optoisolator on a 9600 serial connection to see if it's reliable seperately.
08:04 PM CaptHindsight[m]: CloudEvil: a glass lathe made of glass would work, the are about the accuracy of a wood lathe
08:05 PM Plaid: I think the failure is now in the PC => VFD stage. The isolator fixed the VFD => PC side of things
08:05 PM CloudEvil: Do you have a laptop you could use to test - on battery means it's isolated.
08:05 PM CaptHindsight[m]: Plaid: i wonder how well they designed the PCB in the VFD
08:06 PM Plaid: CaptHindsight[m]: Yeah, I wonder. As long as I transmit the messages often enough, it should work...
08:06 PM CloudEvil: Unless you understand why it's 80%, one day it could be 0.
08:06 PM Plaid: Also, it's just a router, so changing the spindle speed isn't something I'm constnatly doing
08:07 PM CaptHindsight[m]: yeah but CE has a point
08:08 PM Plaid: I understand the wise advice. If this was a production machine, I'll agree with you. I'm going to be dumb, and ignore wise advice.
08:08 PM CaptHindsight[m]: it's also the lowest cost industrial solution until it dies
08:09 PM Plaid: CaptHindsight[m]: Ignore the problem and hope it goes away?
08:09 PM CaptHindsight[m]: of course
08:09 PM CaptHindsight[m]: and then blame it on someone else
08:09 PM Plaid: Ah, right. The dog bumped the cabinet and messed it up.
08:10 PM Plaid: Excuse scheduled.
08:10 PM CaptHindsight[m]: animals are great scapegoats
08:10 PM CaptHindsight[m]: also ghosts
08:11 PM Plaid: I'll try adding a ferite core on the VFD side, and see if it cleans up the signaling. I'm not seeing noise on the PC side.
08:21 PM CaptHindsight[m]: Plaid: which side is currently connected to the shield of the conductors?
08:24 PM Plaid: I think it's on the VFD side, but I don't the cable is fully shielded however. I don't want to change the "stock" wiring. I've got enough things to get working already :P
08:25 PM Plaid: The first foot from the VFD, I can't tell if it's shielded or not
08:30 PM roycroft: overlay new wiring that you know is shielded and see if that takes care of the problem
08:30 PM roycroft: if so, then it's time to replace the stock wiring
08:30 PM roycroft: if not, then you have a different problem to solve
08:31 PM Plaid: It requires a funky connector :(
08:31 PM Plaid: Doesn't have a screw block for the RS232
08:36 PM roycroft: that makes things a bit more challenging
08:37 PM Plaid: Yeah, funky connectors have wasted soo much of my money. Like I need to replace an encoder. Found a cheap encoder that cost $40. Finding a breakout cable for it... cost me $35
08:37 PM roycroft: i just hate taking stabs in the dark as a debugging methodology
08:37 PM Plaid: Still the cheapest solution for an encoder.
08:38 PM roycroft: i just spent $120 on a jumper cable to go from the batteries in the front of my pickup to a winch that will be mounted in the back
08:38 PM CaptHindsight[m]: sounds like a ghost reference
08:38 PM roycroft: i have to crimp new lugs on the jumper cable to connect it to the winch
08:39 PM roycroft: i got some nos crimp lugs on ebay for $13 for five
08:39 PM roycroft: that's not bad
08:39 PM roycroft: the crimp die for my hydraulic crimper is going to cost me $98
08:39 PM roycroft: for two crimps
08:39 PM roycroft: because it's one of the sizes i don't have
08:39 PM Plaid: Ooof
08:40 PM XXCoder: roy get snow or ice?
08:40 PM roycroft: i'm waiting for the lugs to arrive to be sure they fit the cable properly before i drop the $98 on the lug
08:40 PM roycroft: not yet, xxcoder
08:40 PM XXCoder: interesting. ice here
08:40 PM roycroft: that storm that dumped 400mm of snow on vancouver this morning is headed south
08:40 PM Plaid: If it's for one cable, you can order crimped cable.
08:40 PM roycroft: it is due to arrive her tomorrow night
08:40 PM roycroft: i haven't found a place that can make up what i need
08:41 PM roycroft: and if i did, it would probably cost me $100 for the special cable
08:41 PM roycroft: so i might as well buy the die
08:41 PM Plaid: roycroft: what do you need?
08:41 PM roycroft: the wire is 2/0 gauge
08:41 PM roycroft: but it's copper-plated thin strand aluminium wire
08:42 PM roycroft: i'm not 100% positive a standard 2/0 crimp lug will fit it properly
08:42 PM roycroft: anyway, on the winch end it needs to have 90 degree lugs
08:42 PM roycroft: with short necks
08:42 PM roycroft: and it has a quick-disconnect plug on the other end
08:43 PM Plaid: Oh, I thought you were needed a cable with a ring terminal on the end
08:43 PM roycroft: https://www.amazon.com/gp/product/B09R47HSSV/?th=1
08:43 PM roycroft: so that's the jumper cable
08:43 PM roycroft: it has ring terminals on the short end, that connects to the winch
08:43 PM roycroft: but i can't safely install them on the winch
08:44 PM roycroft: https://www.ebay.com/itm/165694299498
08:44 PM roycroft: those are the lugs i need
08:44 PM roycroft: and i hope they fit
08:44 PM Plaid: Okay, you can join me in the being angry at connectors club
08:44 PM roycroft: the cable isn't quite the same diameter as solid copper 2/0 wire
08:45 PM roycroft: so i ordered some 1/0, hoping they will fit that
08:45 PM roycroft: i've been working on telecom gear for decades - this is just part of the game
08:45 PM roycroft: my crimp tool cost me close to $2k
08:45 PM roycroft: again, that's just the price one pays to play the game
08:46 PM roycroft: i hardly ever use it these days, but i figured that when i'm working on wiring a winch that draws 350a, i should probably make sure i build the cable properly
08:46 PM CaptHindsight[m]: hmmm 2/0 Gauge 30 FT = 60 ft of 2/0
08:46 PM roycroft: and i aleady have the tool
08:46 PM roycroft: i just need the die
08:47 PM CaptHindsight[m]: 30 ft or red and 30 of black
08:47 PM roycroft: the only part about this that may be frustrating is that it's a temporary solution
08:47 PM roycroft: at some point i plan on installing hard wiring to both the front and rear for the winch
08:47 PM roycroft: and eventually i'll get a second winch
08:47 PM CaptHindsight[m]: maybe copper plated alu
08:48 PM Plaid: Setting this router up is a side-project to my main project. I'm building a 4 seat airplane... and I lost access to the router I'd been using right after I tested it out and verified it'd work.
08:48 PM roycroft: anyway, yes, i do share the connector problem with you
08:49 PM CaptHindsight[m]: who sells 60 ft of 2/0 copper for $85? Noone.
08:49 PM Plaid: They went from a nice custom made router to a cheap chinese one. It's better for wood cutting, which most people use it for, but it's not as percise as the old machine.
08:50 PM Plaid: That's probably what you'd sell that wire for as scrap....
08:50 PM CaptHindsight[m]: noone is responsible, case dismissed
08:59 PM roycroft: the cable i got is copper-plated aluminium
09:00 PM roycroft: it would be $400 if it were solid copper
09:04 PM CaptHindsight[m]: the cable from noone?
09:04 PM roycroft: yes
09:06 PM CaptHindsight[m]: why won't you tell us?
09:06 PM CaptHindsight[m]: fun with Chinese brands
09:10 PM XXCoder: noone company exists? thought you guys was being sarastic
09:10 PM roycroft: https://www.amazon.com/gp/product/B09R47HSSV/?th=1
09:11 PM roycroft: that's the actual brand name
09:11 PM XXCoder: fun
09:12 PM XXCoder: whats funny is I found a few noone companies
09:12 PM XXCoder: including school uniforms one
09:12 PM Tom_L: same company probably :)
09:13 PM roycroft: yes
09:21 PM redgreenlegion is now known as redlegion
11:17 PM mrec_: interesting, for generating some C code you have to correct chatgpt quite a lot
11:18 PM mrec_: and the answer is always ... yes you're correct and then it starts fixing the code
11:50 PM CaptHindsight[m]: XXCoder: they might pronounce it Noon-Eeee
11:50 PM XXCoder: yeah pronounces is meaningless to me basically lol
11:50 PM CaptHindsight[m]: but No-1 is more fun
11:50 PM XXCoder: lol I bet
11:51 PM CaptHindsight[m]: who makes the best jumper cables No one
11:52 PM CaptHindsight[m]: another "who is on first"
11:52 PM XXCoder: hi noone im dad