#linuxcnc Logs
Sep 04 2024
#linuxcnc Calendar
01:06 AM Deejay: moin
01:23 AM lcnc-relay: <meisterdippel> moin
01:30 AM lcnc-relay: <TurBoss> good morning
01:39 AM lcnc-relay: <meisterdippel> good morning
01:39 AM lcnc-relay: <meisterdippel> setting new name and avatar
01:42 AM lcnc-relay: <turboss> 🙂
02:07 AM lcnc-relay: <captainhindsight_.> Repairing the really big iron is more like working on ocean freighters
02:07 AM lcnc-relay: <captainhindsight_.> Not at all like an F1 or F16 tech
02:19 AM -!- #linuxcnc mode set to +v by ChanServ
02:19 AM lcnc-relay: <captainhindsight_.> It is impressive to test 60hp spindles and see <1um of runout
02:25 AM -!- #linuxcnc mode set to +v by ChanServ
02:42 AM lcnc-relay: <meisterdippel> 60hp, wow
04:41 AM Tom_L: morning
04:46 AM Tom_L:
04:46 AM Tom_L: Spindle RPM
04:46 AM Tom_L: 33,000 rpm - 130 kW
04:46 AM Tom_L: meisterdipple now imagine that
04:46 AM Tom_L: https://www.makino.com/en-us/machine-technology/machines/horizontal-5-axis/mag3-ex
04:55 AM lcnc-relay: <meisterdippel> impressive, unfortunately my apartment and my wallet are too small for this 🥺
05:08 AM JT-Cave: morning
05:36 AM lcnc-relay: <captainhindsight_.> Pretty sure they offer payment plans
06:54 AM lcnc-relay: <meisterdippel> yes, i sent them an offer and now i receive a photo of the machine every month
06:58 AM rdtsc-wk: Nice. Guess that's like watching a Tormach YT video, then the next week's worth of ads being Tormachs.
07:33 AM lcnc-relay: <meisterdippel> my bad english, sorry, should be a joke 🙂 that's all i get for my budget
07:39 AM {HD}_ is now known as {HD}
07:59 AM rdtsc-wk: English is fine; certainly understood as a joke. The world would be a much better place if everyone chuckled more and scorned less.
08:55 AM rdtsc-wk: Interesting RFID tag design: https://www.reddit.com/r/microchip/comments/1f80mwj/question/#lightbox
09:02 AM Ev: Eveready
09:47 AM Deejay: re
12:09 PM -!- #linuxcnc mode set to +v by ChanServ
12:13 PM Rab: rdtsc-wk, the wiggles are for the required antenna length. Somebody in the thread asked why no sharp right angles--could be that design is more effective or reliable when it's conformed to a curved surface.
01:55 PM lcnc-relay: <festlv> Oh, howdy there
01:56 PM lcnc-relay: <skunkworks8841> my ham radio uncle talked about rf and sharp corners causing problems..
01:56 PM lcnc-relay: <skunkworks8841> but - I think he was talking about kw power
01:58 PM xxcoder: those tags run under extremely low power, they probably wanted very low losses?
01:58 PM lcnc-relay: <festlv> Boards and parts shipped today, probably will have something more to report in two weeks or so
02:14 PM Rab: I have investigated the question of PCB trace angles in the past, mainly to defend my love of curvy traces. The consensus seems to be that there is no meangful difference below 3GHz, which is way higher than those tags operate.
02:16 PM xxcoder: interesting. thanks
02:16 PM Rab: There is a benefit to beveling the corners of right-angle traces in RF power applications in a very specific way (everything is very specific in RF power). I don't remember why...could be skin effect or theoretical perfect impedance match or something.
02:17 PM roycroft: i think it helps avoid micro-arcing around the corner
02:18 PM roycroft: something in the back of my mind recalls something like that from an ee class taken decades ago
02:19 PM Rab: Well, it was just the outside corner. There might be additional issues beyond the range I was concerned with. ;)
02:20 PM Rab: Etchant traps are also a historical concern with right-angle traces, but I think that's less of an issue with modern fab processes.
02:21 PM roycroft: yeah, that should be easy to avoid
02:21 PM Rab: I do not happen to like the look of 90-degree angles, but apparently they are totally electrically valid up to 3Hz.
02:21 PM Rab: uh GHz
02:22 PM * roycroft wonders how a person can eat 80 hot dogs in 10 minutes
02:22 PM roycroft: i'm not a hot dog eater, but given the size of one, even a small one, i could possibly eat 2 in 10 minutes
02:22 PM roycroft: i don't know where a third one would go
02:23 PM Rab: 7.5 sec per hot dog.
02:23 PM Rab: Buns included?
02:24 PM Tom_L: it's just a weiner without the bun
02:24 PM roycroft: yes
02:24 PM roycroft: and he actually ate 83
02:25 PM roycroft: no, it includes the bun
02:25 PM Rab: Condiments?
02:25 PM roycroft: probably not
02:25 PM roycroft: would condiments really make much of a difference, though?
02:26 PM Rab: 83 tsp of mustard has got to add up to something.
02:26 PM roycroft: i won't even get into a discussion of *why* a person would want to do that
02:26 PM Rab: 1.73 cups!
02:26 PM roycroft: just that someone actually did it is mind-blowing
02:26 PM roycroft: and the guy is not like sumo wrestler huge
02:27 PM roycroft: he's a tall guy with a full build, but he doesn't appear obese
02:27 PM Rab: It's well over 10 lbs of meat.
02:27 PM roycroft: well
02:27 PM Tom_L: most of those winners are slender
02:27 PM roycroft: 10 lbs of processed pig anus bits and other miscellaneous animal parts
02:27 PM -!- #linuxcnc mode set to +v by ChanServ
02:27 PM roycroft: to call them "meat" would, i should think, be an insult to actual meat
02:28 PM Tom_L: i'm amazed the topics that show up here
02:28 PM roycroft: i just stumbled across the "news" story on a website
02:28 PM roycroft: it felt worthy of comment
02:28 PM Tom_L: condiment
02:28 PM roycroft: and is not as controversial as many stories mentioned here
02:29 PM roycroft: provably worth of condiment too, yes :)
02:29 PM Tom_L: bless you
02:29 PM roycroft: again, not a hot dog eater
02:29 PM roycroft: but i grew up in chicago
02:29 PM roycroft: and i *know* that there is exactly one condiment that is appropriate for a hot dog - mustard
02:29 PM roycroft: one could make an argument for onions and/or pickle relish
02:30 PM roycroft: but catsup is right out
02:38 PM rdtsc-wk: testing a 32-year-old Parker Apex6152 drive... firmware date 1992
02:39 PM Tom_L: things were made to work and last back then
02:39 PM roycroft: when cleaning out my old office i ran across a pile of 5-1/4" full height drives, with capacities varying from 100MB to 300MB
02:40 PM rdtsc-wk: strange failure... the bus caps leaked through their positive terminal studs
02:40 PM roycroft: several of them had an esdi interface
02:40 PM Rab: Back when "Compumotor" sounded hardcore.
02:40 PM rdtsc-wk: had to relearn DRIVE1, GO, and !S over serial terminal to control it :)
02:40 PM Tom_L: back when apps were programs
02:40 PM roycroft: i used to have some fujitsu eagles - 8" 3 phase drives with a 100MB capacity
02:42 PM Tom_L: you've been spending alot of time moving... did you retire?
02:44 PM roycroft: no
02:45 PM roycroft: not yet
02:46 PM roycroft: i took a week of vacation last week, though
03:12 PM Rab: rdtsc-wk, I have been investigating a box of small Logosol servo drives which were last checked around the turn of the millennium...no idea how old they actually are. Only one was bad (no response to software).
03:13 PM roycroft: most of those older drives will last (and store) indefinitely, unless they're already dead
03:13 PM roycroft: it's the new-fangled millennial stuff that is designed to die microseconds after the warranty expires
03:13 PM Rab: Logosol still has all of their old software publicly available for download, which is a real class act IMO.
03:14 PM Rab: Parker does that too IIRC (maybe I had to register first).
03:17 PM Rab: Runs great in a Windows XP VM with USB passthrough to a commercial USB-to-RS485 converter. Now I am trying to build a full-duplex RS-485 interface to drive these things from a microcontroller, but I haven't quite figured it out.
03:18 PM rdtsc-wk: Try a SEW RS485 communications dongle - it's full-duplex, use it for several other things (not SEW) ;)
03:18 PM Rab: Specifically full-duplex RS-485 doesn't seem to have a lot of exposure among hobbyists, few designs to steal.
03:20 PM Rab: That looks legit. The one I got was an Advantech. It shows up as an FTDI UART, but I don't know what else is inside the dongle.
03:20 PM rdtsc-wk: one of these - comes up as a serial port on the PC but is full-duplex (unlike most others) https://www.ebay.com/p/663214394
03:21 PM rdtsc-wk: have used it to talk to several different things; don't think there is much else inside - seems to just work. :)
03:22 PM rdtsc-wk: downside is the output is an RJ11 cable - need to break that out into four pins
03:23 PM Rab: I dunno if full-duplex RS-485 as implemented by these drives is just standard serial RX/TX through a MAX488 or similar line-driver, but that approach isn't working so far.
03:24 PM Rab: Half-duplex needs some signaling trickery to pick a transmitter.
03:24 PM rdtsc-wk: I've tried to roll-my-own and had mixed results. There are lots of little gotchas that creep into a seemingly easy and benign goal.
03:26 PM Rab: I got to the point of triggering and decoding the data from the dongle on my DSO, I think...need to hook up the homemade driver and see what's different.
03:28 PM rdtsc-wk: Sounds like a plan for success. I'd wanted to make an isolated 485 / profibus driver, even drafted a whole schematic for it, but alas, too many other projects
03:33 PM Rab: I bought this among other things. The ISO35T looks like a nice product. Unfortunately the PIC on the board has to be instructed over I2C to enable passthrough serial communication, it's kinda lame. I jumpered straight to the ISO35T but didn't get a result from the drive. https://www.ebay.com/itm/174428102937
03:35 PM Rab: I think if I compare the USB dongle with the homebrew driver I will find something very elementary, like wrong UART parity bits (although probably not that).
03:37 PM Tom_L: "Full duplex RS-485" (a misnomer) is really called RS-422.
03:37 PM Tom_L: https://electronics.stackexchange.com/questions/482406/use-half-duplex-or-full-duplex-rs-485
03:37 PM Tom_L: maybe search for that
03:40 PM Tom_L: https://www.netburner.com/learn/introduction-to-rs485-and-rs422-standards/
03:40 PM Tom_L: stuff you probably already know
03:43 PM JT-Shop: https://www.youtube.com/watch?v=3jW2eOU0ZNo&t=1s
03:46 PM Tom_L: https://www.analog.com/en/resources/app-notes/an-960.html
03:49 PM Tom_L: JT-Shop, get your parts sorted out?
03:50 PM JT-Shop: yup, he's sending me new bolts and I got some better lock washers from McMaster Carr
03:50 PM JT-Shop: https://www.corvetteforum.com/forums/c3-tech-performance/4859669-caliper-fasteners.html#post1608141795
03:51 PM Rab: Tom_L, I was having difficulty disambiguating fdx RS-485 from RS-422, maybe that's why. Thanks!
05:16 PM JT-Shop: Tom_L, decided to replace the master cylinder as well so waiting on it
05:31 PM Tom_L: pretty soon you'll decide to replumb it all too :)
05:31 PM Tom_L: i noticed they have tube & fastener kits
05:32 PM JT-Shop: well I replaced the hoses and lines aft of the trailing arm but I "think" the lines to the mc are ok
05:33 PM Tom_L: no reason they shouldn't be unless the ends are screwed up
05:33 PM JT-Shop: yup I got the fastener kit but replaced the lock washers with high collar or shcs lock washers
05:33 PM Tom_L: the lines are SS aren't they?
05:33 PM JT-Shop: no
05:33 PM Tom_L: hmm
05:36 PM JT-Shop: you can get SS lines
05:36 PM lcnc-relay: <roguish> well, heck, it's only 98 here now.
05:36 PM JT-Shop: but they are harder to install
05:37 PM lcnc-relay: <roguish> JT-Shop: how's Flex coming along?
05:37 PM Tom_L: flex is one version closer to 1.0
05:37 PM JT-Shop: version is 0.9.8...
05:37 PM lcnc-relay: <roguish> is there probing in the latest deb ?
05:37 PM JT-Shop: so it's pretty solid and stable
05:38 PM JT-Shop: you're the probe master lol
05:38 PM lcnc-relay: <roguish> yeah, i can try and test,
05:39 PM JT-Shop: probing is just subroutine calls to chris radec's probe routines
05:39 PM lcnc-relay: <roguish> is that what's in it now?
05:40 PM Tom_L: which is easy to attach a button to
05:40 PM JT-Shop: yup
05:40 PM lcnc-relay: <roguish> ok, i just finished my $$$ machining, so I might be able to get to it
05:40 PM Tom_L: the nickel job?
05:40 PM * JT-Shop calls it a night
05:42 PM lcnc-relay: <roguish> Tom_L: yeah. only broke 3 cutters...... but got it done
05:43 PM Tom_L: the only thing cutters cut when you break them is your profit
05:43 PM lcnc-relay: <roguish> Tom_L: this is Probe Screen, by verser. It's what i am used to and like. morley copied and modified it for qtdragon
05:43 PM lcnc-relay: <roguish> https://jauriarts.org/_matrix/media/v3/download/matrix.org/NQYZTrOObJIPnkHESsYcISjO/image.png
05:44 PM lcnc-relay: <roguish> each big button runs a macro, some call other macros. but I'm pretty sure it's all gcode. morley's is lots of python
05:45 PM lcnc-relay: <roguish> https://github.com/verser-git/probe_screen_v2
05:46 PM lcnc-relay: <roguish> Tom_L: if you buys can do something similar, it'd be great
05:46 PM Tom_L: i've no idea how to put a graphic on a button
05:46 PM lcnc-relay: <roguish> I don't think verser would mind getting plagurized
05:46 PM lcnc-relay: <roguish> Tom_L: that's for the big guy....
05:47 PM lcnc-relay: <roguish> Tom_L: so to cut the 2mm nickel, and not leave nasty burrs, i had to sandwich it between 2 alum pieces.
05:48 PM Tom_L: i think i saw mention of that
05:48 PM lcnc-relay: <roguish> no too hard for the flat plate, but difficult for the 2.2 ID tube.
05:49 PM lcnc-relay: <roguish> made a plug to fit inside the tube, and rolled a thin sheet around the outside.
05:50 PM lcnc-relay: <roguish> 3 flute, carbide, 1/4" end mill.
05:50 PM Tom_L: coated?
05:50 PM lcnc-relay: <roguish> no
05:51 PM Tom_L: not sure what coating nickel would need
05:52 PM Tom_L: dunno if TiN would work or not
05:52 PM lcnc-relay: <roguish> https://jauriarts.org/_matrix/media/v3/download/matrix.org/IygnaQeZhiXiNCpvAXxzAMSw/20240902_094825.jpg
05:52 PM lcnc-relay: <roguish> https://jauriarts.org/_matrix/media/v3/download/matrix.org/JGprantJmBpzKjRJqdruiudF/20240902_095003.jpg
05:52 PM lcnc-relay: <roguish> 1.5 inches long, 1/2" wide
05:53 PM lcnc-relay: <roguish> see the weld in the top picture? in the middle of the dogbone
05:53 PM Tom_L: yeah
05:53 PM lcnc-relay: <roguish> the curved ones have a weld in the middle too. the instron test is for the welding
05:54 PM Tom_L: these are test specimens?
05:54 PM lcnc-relay: <roguish> yes
05:55 PM Tom_L: i've made a few of those
05:55 PM Tom_L: not like that, more plate and roundbars
05:55 PM lcnc-relay: <roguish> so, check with JT about images on buttons.....
05:56 PM Tom_L: put in a feature request :)
05:56 PM Tom_L: right now i gotta figure out wtf is going on with this debian install
07:08 PM lcnc-relay: <skunkworks8841> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/gwXrJSxYFouHhqGSZWeCjJkG/PXL_20240905_000032877.jpg
07:09 PM lcnc-relay: <skunkworks8841> Finally
07:09 PM Tom_L: it works
07:11 PM lcnc-relay: <skunkworks8841> last problem was I missed a jumper - so one of the diodes wasn't grounded as it should have been
07:15 PM rdtsc: Nice. Which Debian Tom, the old machine?
07:30 PM Tom_L: 12
07:50 PM xxcoder: ugh next 2 days going to be real hot