#linuxcnc Logs
Oct 07 2024
#linuxcnc Calendar
12:19 AM lcnc-relay: <meisterdippel> moin
02:15 AM -!- #linuxcnc mode set to +v by ChanServ
04:40 AM Tom_L: morning
05:13 AM JT-Cave: morning
05:40 AM pere: Only 5 strings left to complete the spanish translation on <URL: https://hosted.weblate.org/projects/linuxcnc/linuxcnc/ >. Only 20 left to complete Italian. Anyone here up to the task?
06:11 AM lcnc-relay: <captainhindsight_.> No comprendo
06:26 AM lcnc-relay: <vibram> can't we use something like chatgpt?
06:53 AM lcnc-relay: <meisterdippel> it is sometimes difficult to guess the context in order to find a suitable text/translation
07:13 AM lcnc-relay: <vibram> i need to work on the french one but this is quite time consuming
07:28 AM Tom_L: JT-Cave, PR #117
07:29 AM JT-Cave: thanks
07:33 AM Tom_L: subs should be done unless i add a retract var to the angle subs
07:34 AM Tom_L: and a tool setter height
07:40 AM * JT-Cave starts his chicken day
09:20 AM lcnc-relay: <Travis Farmer> <-- heads out to verify a simulation
10:21 AM rdtsc-wk: apparently it is "G-code": duckduckgo.com/?q=gcode wonder if G is capitalized due to Gerber being a name
10:23 AM Tom_L: geometry code
10:25 AM jpa-: could be a backronym, considering gerber was there first and there are clear similarities in the standards
10:25 AM Tom_L: RS-274, was published in 1963
10:27 AM jpa-: hmm, actually yes
10:28 AM lcnc-relay: <Travis Farmer> JT: sim posted on the issue
10:38 AM lcnc-relay: <Travis Farmer> <-- waits semi-patiently, in the hopes the sim shows the issue... 🤔
10:38 AM Tom_L: is the #xxxx address for G53 known or is it just accessible thru #<_abs_x> etc?
10:40 AM Tom_L: also should Z for a tool setter be defined in absolute coordinates or local (default G54) coordinates?
10:41 AM Tom_L: some could be below the G54 Z0 and some above. case: a machine vise would likely make the setter appear below G54 Z0
10:42 AM Tom_L: a tabletop gantry most likely it would appear above it
10:43 AM Tom_L: seems to be the last piece of my puzzle
10:43 AM lcnc-relay: <zincboy_ca_on> I would always have the Z for a tool setter defined in absolute coordinates. It won't move, but G54 may be reset due to touch off or offesets.
10:43 AM Tom_L: kinda what i was thinking
10:44 AM Tom_L: it locates it in xy in absolute
10:44 AM jpa-: yeah, i agree that if the toolsetter is mounted to the table at fixed position, it should be in abs coordinates
10:44 AM jpa-: if you had some weird per-fixture toolsetter then it would make sense in G54
10:44 AM Tom_L: the tool set relationship appears to be working good
10:44 AM Tom_L: updateing the tool table
10:45 AM Tom_L: i've simulated a tool setter with a 123 block and my probe
10:46 AM Tom_L: it's roughly .500 below G54 Z0
10:46 AM Tom_L: i plan to test different heights too
10:47 AM lcnc-relay: <Travis Farmer> i use G59.3 as my tool setter "playground", and just zero it in code
10:47 AM Tom_L: what i don't quite understand yet is i'm used to setting one tool to 0 and using that to locate G54 Z0 and reference the rest of the tools from that tool
10:48 AM Tom_L: using a setter, all the tools end up with a non zero number
10:48 AM Tom_L: so i need to establish a proper reference point
10:48 AM lcnc-relay: <Travis Farmer> set your 3D probe as a tool, with zero offsets, this is your reference tool
10:49 AM lcnc-relay: <Travis Farmer> then calculate the difference between the 3D probe and the measured tool, and set that as your offset in tool db
10:50 AM jpa-: i use the longest tool as zero length, and set rest of the offsets as negative.. that way if i mess up and deactivate tool offset, nothing bad happens
10:50 AM lcnc-relay: <Travis Farmer> to get the difference, probe your 3D probe to the tool setter
10:51 AM Tom_L: jpa- that's always a good practice
10:51 AM lcnc-relay: <Travis Farmer> my method works, but do what you want...
10:54 AM JT-Shop: unfortunately /me is busy working
10:54 AM Tom_L: werkin fer da man
10:54 AM Tom_L: you :)
10:55 AM Tom_L: i had a bit of time this morning and ran a few tests
10:55 AM lcnc-relay: <Travis Farmer> ok, fair enough, test wen you can 🙂
10:56 AM Tom_L: i fully understand how you do it
10:56 AM Tom_L: wanted to get a feel for other methods ppl may use as well
10:57 AM lcnc-relay: <Travis Farmer> ok
11:00 AM lcnc-relay: <Travis Farmer> Andy Pugh helped me out here: https://forum.linuxcnc.org/40-subroutines-and-ngcgui/50139-my-calculations-are-off-aparently-need-help-with-spacial-math thats where i got my method from
11:31 AM * JT-Cave takes a lunch minute
11:35 AM Tom_L: took longer to heat it than to eat it
11:35 AM Tom_L: short nap then out the door
11:36 AM Tom_L: what andy describes sounds very similar to what zincboy said
11:53 AM * JT-Cave thinks a short nap is near
11:53 AM rdtsc-wk: zincboy has put my CRS in check at least twice now
12:07 PM lcnc-relay: <Travis Farmer> Lunch complete, helped JT solve an issue, watched part of a movie... i feel as though i have accomplished a little something today 🙂
12:12 PM lcnc-relay: <Travis Farmer> Tom_L: "what andy describes sounds very similar to what zincboy said" perhaps i am wrong to use G59.3 as my tool setter playground, but my routines work, and i have yet to need G59.3 for work offsets... yet. i "could" change my routines to get rid of the G59.3, and it would be simpler... but that would require re-proving new code routines, and i just got things working just right... for now. so i personally will stick with my...
12:12 PM lcnc-relay: ... "playground" method. 😉
12:19 PM Tom_L: world coordinates (absolute) are constant
12:20 PM lcnc-relay: <zincboy_ca_on> In industry standard machines you use a calibration tool with known length from the holder taper gauge line to calibrate the Z location of the tool setter. This way the offset table is just the length of the tool and will match a tool presetter. The table surface is Z0.
12:20 PM lcnc-relay: <Travis Farmer> well fine... i will change my routines then... see if i can break another probe... 😉
12:20 PM Tom_L: if it's working for you why break it?
12:21 PM Tom_L: i'm not fixing this for me
12:21 PM Tom_L: zincboy, right
12:21 PM lcnc-relay: <Travis Farmer> "The table surface is Z0." my Z0 is up in the air
12:21 PM Tom_L: my kid's work they have a guy that's all he does is work in the tool crib setting tools
12:23 PM lcnc-relay: <Travis Farmer> i don't see how i can home Z0 on the table, without crashing something, hence why i home at the top of Z
12:24 PM Tom_L: like i said if it works why break it?
12:24 PM lcnc-relay: <Travis Farmer> ... true
12:24 PM Tom_L: afk
12:43 PM jpa-: you don't have to have Z=0 at the point where you home it
12:44 PM jpa-: table at Z=0 and tool length offset is nice in that if you have it set correctly, you won't accidentally run your tool into the table :)
12:44 PM jpa-: but still plenty of ways to make mistakes
12:44 PM lcnc-relay: <zincboy_ca_on> My Z home is a +18".
12:47 PM lcnc-relay: <Travis Farmer> how would i do that? set HOME=<blah> and HOME_OFFSET=<blah>, both the same number, when i measure the table distance?
12:55 PM lcnc-relay: <skunkworks8841> same - all my home locations are x,y,z 38",36",24"
12:55 PM lcnc-relay: <skunkworks8841> Lol - I would have to look at the config..
12:58 PM lcnc-relay: <zincboy_ca_on> In my case, the home is just where the limit switch is and sets the G53 abs coordinates. The tool setter Z/table relationship is calibrated using a fixed length tool against a gauge block. This offset is stored in the static parameters and used in the tool probe routine.
12:58 PM lcnc-relay: <skunkworks8841> HOME = 24
12:58 PM lcnc-relay: HOME_OFFSET = 23.95
12:58 PM lcnc-relay: HOME_SEARCH_VEL = .5
12:58 PM lcnc-relay: HOME_LATCH_VEL = .16
12:58 PM lcnc-relay: HOME_USE_INDEX = YES
12:58 PM lcnc-relay: HOME_IGNORE_LIMITS = YES
12:58 PM lcnc-relay: HOME_SEQUENCE = 0
12:58 PM lcnc-relay: <skunkworks8841> so - my home is slightly off from where the encoder index fires
01:13 PM lcnc-relay: <Travis Farmer> ok, i will create a copy config so i can set my machine up right, yet still be able to use the current config until the copy is proven.
01:27 PM lcnc-relay: <roguish> Travis Farmer: the docs are pretty good about Homing..... http://linuxcnc.org/docs/devel/html/config/ini-homing.html#cha:homing-configuration
01:29 PM lcnc-relay: <Travis Farmer> i still need to find a fixed-length tool for a BT30/ISO30 spindle..., but i have the config in place to work on when i find one i can afford...
01:32 PM lcnc-relay: <roguish> what do ya mean 'fixed lenth' ??
01:33 PM lcnc-relay: <Travis Farmer> the tool setter calibration tool
01:33 PM lcnc-relay: <roguish> make one
01:35 PM lcnc-relay: <Travis Farmer> with a woodworking router table?
01:36 PM lcnc-relay: <roguish> oh, thought you have a mill
01:37 PM lcnc-relay: <Travis Farmer> i wish
01:42 PM lcnc-relay: <Travis Farmer> wow! $199 for one from tormach... way beyond my budget! i will stick with my current method for now 🙂
01:53 PM Unterhaus_ is now known as Unterhausen
01:54 PM lcnc-relay: <skunkworks8841> my probe is my fixed lenght tool 😉
01:54 PM Unterhausen: I would just get a precision dowel and dedicate a collet holder for it.
01:55 PM Unterhausen: but I went on an ebay buying frenzy for toolholders for my machine so I have extras
01:56 PM * JT-Shop opens the garage door in the machine shop for the first time this year
01:57 PM lcnc-relay: <Travis Farmer> that warm in there JT?
01:57 PM lcnc-relay: <Travis Farmer> 51F here
02:01 PM JT-Shop: 70°F 41% humidity
02:01 PM JT-Shop: feels nicer outside so letting the outside in
02:02 PM lcnc-relay: <Travis Farmer> @70F, i understand, for working
03:38 PM Tom_L: skunkworks, you don't use a tool setter then?
03:40 PM lcnc-relay: <skunkworks8841> ni
03:40 PM lcnc-relay: <skunkworks8841> no
03:40 PM lcnc-relay: <skunkworks8841> actually - I don't ususally use a tool file either... I set each tool with a 1/2 dowel.. ;0
03:41 PM lcnc-relay: <skunkworks8841> I do a lot of one offs and just grab whatever tool
03:42 PM Tom_L: i set all mine off a gage block
03:48 PM * roycroft has decided that microsd cards have one major flaw - they are tiny and hard to find when misplaced, especially when it's an old person looking for them
03:48 PM lcnc-relay: <Travis Farmer> very true, roycroft
03:51 PM roycroft: i think it's intentional, as i'm about ready to go buy yet another one because i can't find the one i know i have
03:52 PM roycroft: the only reason i haven't done so already is because best buy (the nearest store that carries them) is almost an hour away from here
03:53 PM lcnc-relay: <Travis Farmer> you need something like this roycroft https://www.amazon.com/dp/B00R52FIH6/
03:55 PM roycroft: that would be awkward for me to manage
03:55 PM roycroft: i have a nice wooden box that holds my usb thumb drives
03:55 PM lcnc-relay: <Travis Farmer> but you could find them
03:55 PM roycroft: i should just make a habit of storing my microsd cards in it as well
03:55 PM rdtsc-wk: have seen a credit-card-sized holder for 9 or so micro-sd cards (bright red)
03:56 PM lcnc-relay: <Travis Farmer> stick your SD cards to a magnet so you don't lose them 😉
03:57 PM rdtsc-wk: https://www.amazon.com/BANDC-Micro-SDHC-Storage-holder/dp/B0196PR0H0
03:57 PM roycroft: better yet, i can take my hot glue gun and glue them to the wall
03:58 PM lcnc-relay: <Travis Farmer> thats cool, rdtsc, added to my amazon wish list
03:59 PM roycroft: yeah, that is actualy useful
04:00 PM roycroft: and instead of storing a standard sd card on it, one can store an sd to microsd adapter
04:00 PM lcnc-relay: <Travis Farmer> oooh, god point roycroft
04:00 PM lcnc-relay: <Travis Farmer> *good point
04:02 PM lcnc-relay: <roguish> roycroft: I keep those adapter things around too, just in case I gotta store a micro card
04:04 PM Rhine_Labs: Not running linux cnc but it is a CNC machine. https://www.youtube.com/watch?v=hd5c6RGNlXg&
04:21 PM Tom_L: JT-Shop, PR #118
04:26 PM Rhine_Labs: So far this machine works there were some chips floating around used them to test programmers. only thing I found was a cracked line in the Pneumatic system prob why it was decommissioned. New comparable machines start at about 90K
04:28 PM JT-Shop: thanks
04:38 PM Tom_L: nice out today ehh
04:39 PM Tom_L: gettin organized with the project?
04:41 PM rdtsc-wk: Good luck with the programming RL :) (Now if only we could forcibly read all those proprietary chips...)
05:32 PM * roycroft heads out to finish the plumbing for his air compressor
05:56 PM lcnc-relay: <roguish> well. only 98 right now.....probably go up a few more
06:10 PM Tom_L: i'll keep what i got at 76F
06:34 PM lcnc-relay: <captainhindsight_.> Rhine_Labs : easy enough to convert to LCNC
06:34 PM lcnc-relay: <captainhindsight_.> I use LCNC for things like this all the time, or well i did
06:37 PM JT-Cave: hmm I need to get the menu improvements from the laptop
06:37 PM JT-Cave: flex gui is great
06:37 PM * JT-Cave loves flex gui
06:38 PM Tom_L: JT-Cave, anything to report on a button for 'recents'?
06:39 PM JT-Cave: I've not worked on that, is that something that might be useful?
06:40 PM Tom_L: maybe not quite as much as i initially thought
06:40 PM Tom_L: but on a full touch system it might
06:41 PM JT-Cave: yup, I agree for a touch screen it could be useful
06:41 PM * JT-Cave goes to get the windblows 10 laptop
06:42 PM Tom_L: i spent 3 days working on a 10 laptop and now there is no boot sector
06:42 PM Tom_L: the usual windows 'fix' process
06:42 PM lcnc-relay: <captainhindsight_.> with great power comes great responsibility
06:43 PM Tom_L: it likely has a virus
06:43 PM Tom_L: no menus no icon response etc
06:43 PM Tom_L: takes 30 min between mouse clicks
06:45 PM JT-Cave: chow time here, see you in the morning
06:45 PM Tom_L: later
07:46 PM lcnc-relay: <big_kevin420> shit i cant remember how i set up the doosan at work
07:46 PM lcnc-relay: <big_kevin420> i wanted the bottom of the vise to match up with g54 z0, so it does
07:47 PM lcnc-relay: <big_kevin420> makes it easy to adjust t offsets manually to set z height
07:47 PM lcnc-relay: <big_kevin420> hmm wait i have no idea what i did now
07:48 PM lcnc-relay: <big_kevin420> i set something prob offset, then probed the vise, and adjusted......something fanuc coordinates
07:48 PM lcnc-relay: <big_kevin420> then .... kept chjanging offsets until the bottom of the vise was always z....somewhere...
07:49 PM lcnc-relay: <big_kevin420> i need to install the qtvcp offset table on my axis gui
07:51 PM lcnc-relay: <big_kevin420> so bottom of the vise is 0 somewhere machine, maybe g53.... then i probed all the tools again to the toolsetter, and it works great.... but how did i do it and how do i do it on lcnc...
07:55 PM lcnc-relay: <big_kevin420> https://jauriarts.org/_matrix/media/v3/download/jauriarts.org/daJezqdYAzMMRmfAxGjjpwjq/offsets.png
07:56 PM -!- #linuxcnc mode set to +v by ChanServ
08:25 PM rdtsc: Insert your choice of explitives here... https://www.youtube.com/watch?v=LJ5yLBKj0Rk
08:29 PM roycroft: i almost used an expletive earlier today
08:29 PM roycroft: i had almost finished plumbing in the air compressor - i was on the last fitting
08:30 PM roycroft: which was the extended drain line
08:30 PM roycroft: i cut and threaded the pipe, put the sill cock on, and started to install it on the elbow coming out of the bottom of the tank
08:31 PM roycroft: while i was still hand tightening it into the street elbow coming out of the tank, the elbow broke off flush with the tank
08:31 PM xxcoder: ow
08:32 PM roycroft: now i have to disconnect the power and the plumbing, remove the motor, remove the compressor pump, remove the tank from its mounting board, extract the broken piece, and then reassemble everything
08:32 PM xxcoder: too bad we cant access 4d space. it'd make that much easier
08:33 PM roycroft: there's a tiny chance i can remove the broken bit with a screw extractor
08:34 PM Tom_L: there's alot of things i'd try before going thru all that
08:34 PM roycroft: i'm certainly going to give that a go before i spend the better part of a day disassembling and reassembling
08:34 PM roycroft: but i don't have high expectations that it will work
08:34 PM roycroft: the clearance at the bottom of the tank is very minimal
08:34 PM Tom_L: just the threads left in the tank?
08:34 PM roycroft: yes
08:35 PM roycroft: and the street elbow is brass
08:35 PM roycroft: its replacement will be galvanized
08:35 PM roycroft: i'm pretty sure i have enough room to get the screw extractor in the hole
08:35 PM Tom_L: galv will be a bit longer than brass.. hope you have room
08:35 PM roycroft: i don't know that there are many other options
08:35 PM roycroft: yeah, it might not fit well
08:36 PM rdtsc: a big ez-out?
08:36 PM Tom_L: i have mine sitting on a riser
08:36 PM Tom_L: for that reason
08:36 PM roycroft: i really don't now how/why it broke
08:36 PM Tom_L: rust?
08:36 PM roycroft: brass doesn't rus
08:36 PM roycroft: t
08:36 PM Tom_L: overtightened once
08:36 PM roycroft: at the old shop there would be room for a riser
08:36 PM roycroft: but not here
08:36 PM xxcoder: metal fatique?
08:37 PM roycroft: i installed the current drain about 5-10 years ago
08:37 PM rdtsc: brass does work-harden, right?
08:37 PM Tom_L: only ~3" rise
08:37 PM roycroft: the elbow was installed once, and never moved until today
08:37 PM roycroft: it could be metal fatigue
08:37 PM roycroft: but it seems unlikely to me
08:37 PM rdtsc: it could be plain-old corrosion
08:38 PM Tom_L: just about everything work hardens to a point
08:38 PM roycroft: i should look on the bright side
08:38 PM roycroft: if it broke that easily, it would probably have failed on its own soon after i put the compressor back in service
08:39 PM roycroft: and murphy has taught me that the time it would fail would be the time i forgot to turn the compressor off and then headed off for the day
08:39 PM roycroft: this whole drain redo was my cunning plan to make the compressor last longer
08:40 PM roycroft: i figured if i had the drain running to a sill cock outside the shed i'd be more likely to turn the compressor off and drain it every day when i'm done using it
08:40 PM roycroft: with the drain on the inside and needing a hose hook-up to drain, i was honestly only draining it a few times/year
08:41 PM roycroft: once i let it go so long that it started cycling so frequently my vacuum pump (venturi design) would not be able to pull a good vacuum
08:42 PM roycroft: i did some troubleshooting, and discovered that the compressor tank was about 3/4 full of water
08:42 PM roycroft: this is a 250l tank, so that's a lot of water
08:42 PM xxcoder: free water
08:43 PM xxcoder: lol that is a lot yeah. I bet it was very heavy then
08:43 PM roycroft: i'm sure it was
08:43 PM roycroft: i did not pick it up or move it at the time
08:43 PM roycroft: i just drained it, and that took a long long time
08:44 PM roycroft: i've gotten better about periodically draining it, but it's still a hassle and i still don't do it as often as i should
08:44 PM roycroft: the ideal thing would be to do as i described - turn it off at the end of the day and drain it, every single time
08:45 PM Tom_L: maybe once a week/month
08:45 PM xxcoder: I wonder if there can be system to autodrain it
08:45 PM roycroft: yes, there are autodrain systems
08:45 PM roycroft: the good ones are expensive
08:45 PM Tom_L: not sure i'd trust a cheap one.. rust acumulates at the bottom
08:45 PM Tom_L: and gunk
08:45 PM roycroft: right
08:46 PM Tom_L: i bet when it was full of water it cycled alot more often
08:46 PM roycroft: yes, that's how i figured out it was full of water
08:47 PM roycroft: and i could not pull a good vacuum with it cycling that often
08:48 PM roycroft: it wasn't able to deliver enough air to the venturi
08:48 PM xxcoder: venturi is funny stuff. blow enough air and it makes vacuum
08:49 PM roycroft: i don't know how long it had been since i drained it, but it was in the spring when it happened, so the compressor had just gone through the rainy season, where the humidity is often 80% or higher
10:09 PM lcnc-relay: <big_kevin420> damn i had my toolsetter math correct the whole time
10:12 PM Tom_L: good to hear
10:21 PM lcnc-relay: <big_kevin420> and this toolsetter is more repeatable than its spec'd out for
10:22 PM lcnc-relay: <big_kevin420> https://us.misumi-ec.com/vona2/detail/110302460870/