#linuxcnc Logs
Aug 12 2018
#linuxcnc Calendar
12:00 AM fragalot: HI
12:04 AM Tom_L: the flip side is waking up..
12:13 AM fragalot: We've got all sort sof ants her
12:13 AM fragalot: black, red, HUGE (relatively speaking) ants, flying ants, etc..
12:13 AM XXCoder: heard of new invastive ants thats killing off fire ants
12:13 AM XXCoder: in usa
12:14 AM jesseg: I wonder what kind of aluminum sculpture the new kind of ants make
12:14 AM XXCoder: dunno
12:14 AM XXCoder: odd I cant access reddit
12:14 AM jesseg: maybe you already read it
12:14 AM fragalot: hehe
12:17 AM XXCoder: can access now. weird
12:26 AM Wolf__: ugh, you know work is getting big when you find yourself looking at cranes…
12:29 AM fragalot: lol
12:30 AM XXCoder: lol
12:30 AM Wolf__: 9k lbs/ 4k kg truck bodys
12:30 AM XXCoder: watch out when you need overhead cranes
12:30 AM XXCoder: one with rails and many tons lifting ability
12:31 AM XXCoder: you know when you have gone big by then lol
12:31 AM Wolf__: I need a actual yard crane, looking at 15ton pettibone
12:32 AM XXCoder: http://www.hkequipment.com/images/Pettibone_-_Extendo.jpg this?
12:33 AM Wolf__: https://www.easternsurplus.net/VehicleDetails/217/Pettibone-Model-30-Multi-Crane-with-Telescopic-Boom
12:36 AM XXCoder: youre definitely getting into bigger stuff
12:49 AM Wolf__: I wish I didn’t need to get something like that, but I think we are at around 10 bodies to install at 9,000lbs each, need to unload off a trailer then install on truck
12:52 AM XXCoder: lot og work
12:55 AM Wolf__: not really much to it, just big things
12:56 AM XXCoder: oh yeah its BULLET ant not gunshot ants
12:56 AM XXCoder: quite large ant
12:56 AM XXCoder: people is known to sucide after bites because of pain
12:57 AM fragalot: Wolf__: up next: http://www.verticaaltransport.nl/documents/nieuws-items/2013-09-20-bkv-neemt-nieuwe-750-tons-mobiele-kraan-in-gebruik/images/2013-09-20-bkv-neemt-nieuwe-750-tons-mobiele-kraan-in-gebruik.jpg
12:57 AM Wolf__: I think shipping on that would kill me
12:57 AM fragalot: :P
12:57 AM XXCoder: https://www.youtube.com/watch?v=1Ka3hNgR4Qo hes insane
01:00 AM XXCoder: Wolf__: not reallt
01:00 AM XXCoder: just drive it
01:00 AM XXCoder: no shipping nesscary
01:00 AM Wolf__: this is the type of bodies I have to install https://previews.123rf.com/images/robwilson39/robwilson390905/robwilson39090500002/4807066-white-garbage-truck-isolated-with-a-driver.jpg
01:00 AM XXCoder: its immune to traffic as bonus. just drive over
01:00 AM Wolf__: XXCoder: how am I going to drive that from europe to the usa?
01:01 AM XXCoder: floatation devices of course
01:01 AM XXCoder: https://www.youtube.com/watch?v=wsZI1fdf2RU
01:01 AM XXCoder: that guy is an asshole lol
01:03 AM XXCoder: finally found boat mode one lol http://pics.imcdb.org/0is220/ccbb21op3.2825.jpg
01:55 AM XXCoder: roycroft: https://www.youtube.com/watch?v=7FhifTGKtUQ
02:02 AM roycroft: yeah, that is the totally not way to do it
02:02 AM XXCoder: seemed to work lol whats wrong with it
02:03 AM roycroft: it used expensive, toxic chemicals where there was no justification for their use
02:03 AM XXCoder: you didnt watch all of it?
02:03 AM roycroft: no
02:03 AM roycroft: i gave up early on :)
02:04 AM XXCoder: watch then. I linked it for second part not dumb chemical first part of it
02:06 AM Wolf__: tl;dr plastic tote with water+dish soap + board with meat (chicken?) on it to catch yellow jackets, kinda neat to know
02:07 AM XXCoder: yeah its kinda surpising. seems wasps need couple inches to turn and fly
02:07 AM XXCoder: and soap water being too close means they cant fly
02:10 AM roycroft: ok, the second half was better
02:11 AM roycroft: i've never used chicken to attract wasps
02:11 AM XXCoder: apparently they LOVE em so hard
02:11 AM roycroft: but i'll keep it in mind as an option
02:11 AM XXCoder: you can see wasps already flying in as hes screwing it in
02:13 AM XXCoder: he was using whole ones. me? I probably would order KFC then use leftovers
03:03 AM Wolfmetalfab is now known as Wolf__
03:09 AM Deejay: moin
04:03 AM phipli: Morning
04:04 AM Deejay: hey
04:06 AM phipli: Deejay, do you guys have Velleman electronics kits?
04:12 AM Deejay: nope, nothing from velleman here
04:12 AM Deejay: why do you ask?
04:13 AM phipli: I dug out some of the kits I built when I was learning electronics (not that you ever stop learning)
04:13 AM phipli: They pre-date the resurgence in electronics - they were reasonably available through the 90s when most companies vanished
04:14 AM phipli: I owe them a lot... but I was wondering how common they were in other places
04:14 AM phipli: the company itself is Belgian
04:15 AM phipli: A lot of their kits are actually useful things for larger electrical projects
04:15 AM phipli: https://quasarelectronics.co.uk/Item/velleman-vm8090-8-channel-usb-relay-card-module
04:16 AM phipli: Educational, build it yourself, comes with functional instructions etc. etc.
04:16 AM phipli: my Nephew is getting old enough that I can start thinking about buying him their simple kits
04:28 AM xynashi is now known as Guest20960
04:39 AM jthornton: morning
05:02 AM jthornton: installing debian 9 again...
05:03 AM XXCoder: nothing like freshly installed debian to wake you up eh
05:04 AM jthornton: lol yea, found out that buildbot.linuxcnc.org and linuxcnc.org use different keys
05:04 AM jthornton: so instructions for installing 2.7 are different than master
05:06 AM XXCoder: interesting
05:06 AM jthornton: so it only took me 3 days to get that info
05:08 AM phipli: I've completely broken the nvidia drivers on my laptop
05:08 AM jthornton: ouch
05:08 AM phipli: I'm resigned that at some point I'll have to re-install after I've probably spent about 16 hours in total trying to get them to work on the existing install
05:09 AM phipli: issue is, I was trying to learn opencl / gpu computing stuff just before I did it
05:09 AM phipli: all I did was upgrade them :(
05:09 AM phipli: using opensource drivers now and none of the stuff I was trying to learn is compatible
05:10 AM jthornton: linux the upgrade dependency hell
05:10 AM phipli: nvidia hell
05:10 AM phipli: they've done something weird that stops you rolling back
05:10 AM phipli: and if you try /really/ hard to roll it back
05:10 AM phipli: you break something
05:11 AM phipli: I feel like they've put a dummy version of something in for the newer version, and it isn't been replaced when I install an older one...
05:12 AM phipli: but by this point I've tried so much stuff I have no idea what state the remains of nvidia drivers are in on this machine
05:12 AM phipli: probably a good point to move back to Mint
05:12 AM jthornton: I like linux mint much better than debian anything
05:13 AM jthornton: debian just makes everything a pia
05:14 AM phipli: I'm on Ubuntu atm - not really by choice
05:15 AM phipli: Usually use Mint on this laptop
05:15 AM jthornton: I liked Ubuntu up to 10 after that not so much
05:15 AM phipli: Never liked Unity
05:15 AM phipli: only tried it this time around because Unity has gone
05:15 AM jthornton: chickens are back on extra light now they get up 20 minutes before sunrise today
05:15 AM phipli: but then they started making gnome more like unity
05:16 AM phipli: jthornton, that time of year already?
05:17 AM jthornton: yea 13:41:29 of sunlight today
05:17 AM jthornton: from sunrise to sunset
05:21 AM phipli: We've had two eggs so far today
05:21 AM phipli: good strong eggs
05:21 AM phipli: don't think they were liking the hot weather
05:21 AM jthornton: some of mine have thin shells dunno why
05:21 AM phipli: do you give them grit with shells?
05:22 AM phipli: sometimes they just do it
05:22 AM phipli: othertimes it is the heat
05:22 AM phipli: sometimes they try to lay two eggs in a day and don't get a chance to make enough shell
05:22 AM jthornton: they have oyster shell for calcium along with crushed egg shells
05:22 AM phipli: then it is probably the weather
05:23 AM phipli: also, with time, the shells get thinner
05:23 AM phipli: in a few years the average shell thickness will go down
05:23 AM phipli: and I think egg size goes up
05:23 AM phipli: if I remember
05:23 AM phipli: atm, we can't shut a large eggbox on a random sample of our eggs
05:24 AM phipli: when I bring them into work for people I have to sort out all the "small" (read "only large") eggs that fit in the boxes properly
05:27 AM jthornton: yea egg size is still getting larger for my group, the thin shells are always a lighter shade of brown
05:33 AM jthornton: http://gnipsel.com/images/chickens/egg-size-chart-8-12-18.jpg
05:34 AM jthornton: december is when I got the first eggs
05:40 AM agris_ is now known as agris
06:09 AM * fragalot has caved
06:10 AM phipli: what have you bought fragalot
06:11 AM fragalot: I haven't hit "send yet"
06:11 AM fragalot: but it is to buy a shaublin 13 with a ton of accessoires
06:11 AM fragalot: anyone interested in a free RF45? pickup only :P
07:18 AM fragalot: hey
07:21 AM Tom_L: morning
07:47 AM JT-Shop: morning
07:47 AM fragalot: mornin' JT-Shop
07:51 AM JT-Shop: hey fragalot
09:45 AM JT-Shop: went to get the paper and there's a half dead armadillo in the driveway... yuck
09:46 AM fragalot_: dinner.
09:47 AM fragalot_: That's it. The deed is done. I am now poor. lol
09:48 AM Wolf__: lol
09:48 AM Wolf__: I know that feeling
09:49 AM fragalot_: :P
09:49 AM fragalot_: how is your hunt going
09:50 AM Wolf__: I need a 12,000lb fork lift or a small crane now =/
09:50 AM fragalot_: hefty forklift
09:51 AM Wolf__: http://refusetrucks.scrantonmfg.com/rear-loader/viper isnt light
09:51 AM fragalot_: lol, no
09:51 AM fragalot_: a crane would arguably be more useful
09:51 AM fragalot_: for that type of jobs
09:52 AM Wolf__: we have one on the ground right now, 9,000 lb truck body
09:52 AM Wolf__: not really sure how the hell we are going to get it on the truck
09:52 AM fragalot_: have you thought about putting the truck on the body instead & flipping it
09:53 AM Wolf__: would scratch up the paint
09:53 AM fragalot_: put a blanket under it
09:53 AM fragalot_: this is why manufacturing and assembly is going to china, Wolf__
09:54 AM Wolf__: lol
09:54 AM fragalot_: you're too focussed on quality to get the job done
09:56 AM Wolf__: or 100 guys to pick it up and put in on the truck
09:56 AM Wolf__: https://i.imgur.com/zswGLcl.jpg wouldnt roll over well either
09:57 AM fragalot_: dig a hole
09:57 AM Wolf__: hmm
09:57 AM Wolf__: we do have 2 excavators there
09:57 AM fragalot_: dig hole, drive truck in said hole until back is level with the ground
09:58 AM fragalot_: shimmy loader onto truck
09:58 AM fragalot_: :P
10:00 AM sync: an overhead crane is probably the easiest
10:01 AM Wolf__: I wish we had shop space big enough to work on this stuff in, with a big overhead gantry crane
10:04 AM fragalot_: it's surprising just how large cranes get when you go above 5t
10:04 AM fragalot_: overhead gantry, I mean
10:05 AM Wolf__: small rough terrain crane, 15t lift, 55,000 vehicle weight…
10:11 AM Wolf__: over built beats having the gantry falling on your head
10:11 AM fragalot_: certainly does
10:11 AM fragalot_: and the odd job that's JUST a bit too big will be less likely :P
10:12 AM Wolf__: I call that utilizing the safety factor
10:12 AM fragalot_: well it's going to be a dynamic load, so .. x9?
10:12 AM fragalot_: well, x4.5 since it's only dynamic in one direction
10:19 AM Wolf__: these looks cool http://www.metreel.com/products-and-parts/met-track-workstation-cranes/
10:20 AM Wolf__: least for light weight things
10:20 AM fragalot_: Aye
10:22 AM fragalot_: Wolf__: in your opinion, on a schaublin 13.. magnetic or optical DRO?
10:24 AM sync: I would always go for optical
10:24 AM Wolf__: usually I would say optical, but that Y axis looks like it puts the table close to the column
10:24 AM sync: more options to choose from
10:25 AM Wolf__: wait, Y axis is on the head?
10:25 AM fragalot_: yes
10:25 AM Wolf__: optical if the scale will fit
10:25 AM Wolf__: cheap and works =)
10:26 AM fragalot_: :)
10:26 AM fragalot_: i'm only worried about 2 things here
10:26 AM fragalot_: coolant & clearance
10:27 AM Wolf__: https://i.imgur.com/xPohnls.jpg https://i.imgur.com/RQJrIWW.jpg fun to install…
10:27 AM fragalot_: :D
10:27 AM fragalot_: I remember it being hell on my RF45 with the wonky castings
10:28 AM Wolf__: https://i.imgur.com/GyxdNkC.jpg https://i.imgur.com/PhftEBY.jpg I did my lathe and mill at the same time
10:28 AM fragalot_: also - LCD panel version? how luxurious
10:28 AM fragalot_: does the LCD actually make any real world difference in terms of usability?
10:29 AM Wolf__: yeah I kinda wish I had got it for both but was almost $200 more
10:29 AM * fragalot_ mulls this over
10:29 AM Wolf__: its nice on doing bolt circles and whatnot
10:29 AM Wolf__: plus its a color LCD so you can change the colors
10:31 AM fragalot_: found an LCD readout kit with magnetic 5µ for £660 or I can get the same optical 1µ set I got for my RF45 for €590..
10:31 AM fragalot_: (similar to the "standard" one you show in that picture)
10:31 AM Wolf__: I was around $750 USD for both DROs, LCD w/ 5um and the led segmented 3 axis with 2x @ 1um
10:32 AM fragalot_: that's not a bad price
10:32 AM fragalot_: at all
10:32 AM Wolf__: $731 shipped lol
10:33 AM fragalot_: ._.
10:33 AM Wolf__: https://www.aliexpress.com/item/New-Easson-ES-12B-3-axis-digital-readout-mill-lathe-3-axis-DRO-system-and-3/32789719488.html
10:33 AM Wolf__: and the cheap https://www.aliexpress.com/item/free-shipping-complete-set-milling-lathe-drill-machine-dro-digital-readout-with-3-pcs-linear-scales/1524501216.html
10:34 AM fragalot_: going to get railed in taxes on both
10:36 AM JT-Shop: import tariff?
10:37 AM fragalot_: and the administrational fees they add on
10:37 AM Wolf__: that schaublin looks odd, like a mill and a shaper had a child
10:37 AM fragalot_: I bought 2 bottles of canode dye from the USA and wount up paying about $250 in total for them
10:37 AM fragalot_: Wolf__: :D
10:37 AM Wolf__: ouch
10:37 AM fragalot_: you can actually get a shaper head for it (vertical slotting head)
10:39 AM Wolf__: kinda wish I had got the K&T #2 I was looking at, it had a vertical slotting head and the vertical mill head for it
10:40 AM fragalot_: that's a nice machine, but it can't do what your bridgeport can
10:40 AM fragalot_: (obviously, the bridgeport also can't do what the K&T can :P)
10:41 AM Wolf__: well, it was a whole basement deal, $7k for the k&t #2, a index mill (oversized BP clone), a #2 tool and cutter grinder and a crap atlas lathe
10:42 AM fragalot_: should be able to sell most of it off individually and come out with a pretty cheap K&T if you're into that sort of thing
10:43 AM Wolf__: its all gone, that was over a year ago now
10:43 AM Wolf__: or 2
10:45 AM fragalot_: I think I'm going to go for an LCD with cut-to-length magnetics..
10:46 AM fragalot_: mainly because I see a lot of space constaints on this mill, and coolant will be plentiful
10:50 AM fragalot_: actually no. an LCD would look out of place
10:50 AM fragalot_: I need a neon DRO
10:50 AM fragalot_: :D
10:50 AM Wolf__: LSD dro
10:51 AM fragalot_: I think this style looks like it matches the machine.. https://www.machine-dro.co.uk/3-axis-milling-digital-readout-display-console.html
10:53 AM Wolf__: looks cool
10:54 AM fragalot_: should be alright if the membrane lasts :)
10:55 AM robotustra: neon DRO https://www.youtube.com/watch?v=EErY75MXYXI
10:56 AM fragalot_: I had completely forgotten about that monstrosity
12:08 PM gambakufu: what is an acceptable f-error value I should aspire to running with a mesa card and steppers with velocity control mode? I'm getting ~500u mm when accelerating/decelrating in halscope. is this ok?
12:20 PM pcw_home: 500u = .5mm? thats horrible
12:22 PM pcw_home: 500 nm (0.5 u) would be OK
12:22 PM gambakufu: 0.0005mm, I think. assuming 1/div means 1mm per division in halscope
12:23 PM pcw_home: Yes
12:23 PM gambakufu: cool. thanks.
12:23 PM pcw_home: you can improve trackin durring accel/decell with FF2
12:24 PM pcw_home: FF2 compansates for the time between reading the position and writing the velocity command
12:25 PM gambakufu: I see. I'll play around with that value, then.
12:26 PM pcw_home: it shoud be the time between read and writes (roughly .00013 on Ethernet, .00006 on PCI)
12:27 PM pcw_home: very roughly...
12:28 PM pcw_home: any DPLL pre-read time adds to this
12:28 PM gambakufu: I have the 7i92 with 0.00011 (which i believe I copied from the sample mesa config), so I guess it's about right. also, this is a modified x-carve so I don't think i can gain any actual accuracy :)
12:29 PM pcw_home: no, if you actually measured the axis positions with linear encoders you might be a bit dismayed
12:30 PM pcw_home: but no reason to not have the pulse stream as accurate as possible
12:35 PM -!- #linuxcnc mode set to +v by ChanServ
01:53 PM Jin^eLD: hi, I am learning linuxcnc, trying to write a component for a mill, and somehow I don't get one thing..is there some sort of API on the input side that I have to implement? for instance, something like my component would ge a request to set rpm and I it figures out how to do that for this specific hardware
01:53 PM Jin^eLD: what does it get as input?
01:55 PM JT-Shop: http://linuxcnc.org/docs/2.7/html/hal/comp.html
01:55 PM JT-Shop: or http://linuxcnc.org/docs/2.7/html/ladder/classic-ladder.html
01:56 PM Jin^eLD: I am actually at comp.html, i manaed to write something simple, juts playing around with input and output pins and a connected pyvcp
01:56 PM Jin^eLD: let me have a look at the ladder...
01:59 PM Jin^eLD: i don't get it... how is the ladder thing related to my question? :) maybe it is and I just dont get it, but maybe I was not able to properly convey what I wanted to know when asking my question...
02:00 PM Jin^eLD: who or what in Linuxcnc would call my component?
02:04 PM Jin^eLD: maybe I got the whole concept wrong, I assumed that in order to support some machine I would have to implement a hal component following some API to receive input linuxcnc and produce output that would further control the machine
02:07 PM fragalot_: are you actually building a component (eg, a full custom controller circuit board?)
02:07 PM fragalot_: or simply driving a standard step/dir motor via either oen of the recommended boards, or a parallel port?
02:11 PM Jin^eLD: good question... so as a backstory, my buddy bought an MAHO mill which he is converting to linuxcnc, but he is not into programming and linux, so he asked me to help out; he threw some MAHO related documents at me, but I am totall new to linuxcnc, hence all the n00b questions ;) I think he has some I/O cards that he wants to use to control the machine (I see here MESA 7i84 I/O card mentioned in one of his documents)
02:12 PM Jin^eLD: he's on holidays now so I can't ask for detais, I figured I learn linuxcnc / how to write HAL components while he is gone
02:12 PM Jin^eLD: *details
02:12 PM pcw_home: This should not need any new components
02:12 PM pcw_home: mainly needs an understanding of hal/ini files
02:13 PM Jin^eLD: for instance he sent me a table how spindle rpm values match to pin outputs that control the gearbox
02:13 PM Jin^eLD: well he did sent me a component that someone else wrote for MH700 (he has an MH400e)
02:13 PM Jin^eLD: thats why I thought that a component was needed
02:14 PM Jin^eLD: but the guy sold his machine before finishing the component so what I have is something half finished and not working
02:14 PM Jin^eLD: I wanted to write a new one from scratch
02:14 PM Jin^eLD: but indeed good point
02:14 PM pcw_home: I would first read the hal documentation then go through the existing components
02:14 PM Jin^eLD: do I need one at all?
02:15 PM Jin^eLD: I've been digging through docs the whole last week, but I am still not totally sure how everything fits together
02:16 PM pcw_home: Possibly, but I would tackle that after you thoroughly understand hal
02:16 PM pcw_home: you may well be able to do all thats needed in hal
02:17 PM pcw_home: bbl
02:17 PM Jin^eLD: thats his thread btw https://forum.linuxcnc.org/12-milling/33035-retrofitting-a-1986-maho-mh400e
02:24 PM Jin^eLD: from what i read his problem was performing the actual gearchange, thats what the component was needed for
02:24 PM Jin^eLD: and I guess this is the one that he sent me https://forum.linuxcnc.org/12-milling/33035-retrofitting-a-1986-maho-mh400e?start=330#106497
02:29 PM Jin^eLD: at least from what I read that it seems that he needs an extra component if he wants to use the original MAHO gearbox (other option would be to use a VFD, but I understsand he is not going for it)
02:35 PM jthornton: there is a gearchange component
02:36 PM jthornton: if you want to learn ladder you could do the classic traffic light or garage door opener exercise with classicladder and pyvcp
02:37 PM fragalot_: does anyone know if I can use ER25 collets in a schaublin E25 high-speed head?
02:42 PM Jin^eLD: jthornton: well, I dont want to learn ladder if I dont have to :) I currently seem to fail to understand how this whole stuff is used, I looked at the gearchnge component you mentioned, but would use it and how? all those hal components define input and output pins the way the component author likes it, and this confuses me in the sense that I dont get what the API towards linuxcnc is.. i mean, OK, you can connect all the pins the way
02:42 PM Jin^eLD: you want them, but that means that the one who uses the pin connetions of the particular component defines the API?
02:43 PM jthornton: what's an API?
02:44 PM Jin^eLD: ehm, application programming interface
02:44 PM jthornton: is that a windows thing?
02:45 PM Jin^eLD: no, its a generic programming term
02:45 PM jthornton: ah ok
02:45 PM Jin^eLD: when you have two modues and you want them to talk to each other, then you need to define an interface
02:45 PM jthornton: HAL is the interface
02:46 PM Jin^eLD: HAL itself is an interface towards "further up"
02:46 PM jthornton: http://linuxcnc.org/docs/2.7/html/hal/basic-hal.html
02:46 PM Jin^eLD: I understand what the purpose of HAL is
02:46 PM jthornton: what do you mean by "futher up"
02:47 PM Jin^eLD: by the part of the software that uses HAL
02:47 PM Jin^eLD: i.e. by the part that does not care about hardware details, which are hidden by HAL
02:48 PM jthornton: sorry I don't understand what your looking for
02:50 PM jthornton: as an integrator HAL and the components are your main tools to configure the machine connection and classicladder
02:51 PM * jthornton wonders why debian 9 apt-get install python3-pyqt5 fails to find the package yet it is in the Synaptic Package Manager
02:51 PM Jin^eLD: but how do I know what pins in the hal component to implement? towards the machine its clear, for instance I know i have a card with 12 pins and I know which pins to set to 1 or to 0 if I want get a certain spindle speed
02:52 PM Jin^eLD: so here is me, alone with my knowledge, say I implemented that in the component... but who tells me to set the speed? and how is my component told that now someone wants me to set the speed?
02:53 PM Jin^eLD: (imagining that "I" am the component now), so from the components perspective - who tells me what to do?
02:53 PM Jin^eLD: I know how to do it, but who tells me "set speed xyz now"?
02:54 PM jthornton: well I imagine G code would be what tells the spindle what speed to run
02:54 PM Jin^eLD: right, but thats on a much higher level, abstracted by HAL
02:55 PM Jin^eLD: how does the command from G code reach my component? and here I would expect this interface i was talking about
02:55 PM Jin^eLD: like, "set_speed_command 123"
02:55 PM Jin^eLD: and i do my magic with the pins
02:56 PM Jin^eLD: maybe I look at this too much from a programmers point of view... usually when you write plugins you have an API that you have to implement, some callbacks or whatever
02:57 PM jthornton: motion.spindle-speed-out-rps-abs
02:58 PM jthornton: do you have linuxcnc installed on your pc?
02:59 PM Jin^eLD: I do, i compiled the latest release
02:59 PM Jin^eLD: reading up on the motion thing now
02:59 PM jthornton: run the axis simulator
02:59 PM Jin^eLD: ok one sec
02:59 PM jthornton: then open the Show HAL Configuration
03:00 PM jthornton: start the sim and home it
03:01 PM jthornton: brb checking the ribs
03:02 PM Tom_L: mmmmm ribs...
03:03 PM Jin^eLD: not sure how to run it, the play button is greyed out
03:03 PM jthornton: low and slow buddy
03:03 PM Jin^eLD: but I think this might indeed be what I was looking for, commands or pins that I need to implement in my component
03:03 PM jthornton: press the estop the red x
03:03 PM jthornton: then press the power button
03:03 PM jthornton: circle with a |
03:04 PM jthornton: then press home all
03:04 PM jthornton: there is no play button :)
03:04 PM Tom_L: awww
03:05 PM Jin^eLD: it complaints about something exceedingonthe z axis
03:05 PM jthornton: you went too far
03:05 PM jthornton: are you homed?
03:05 PM Tom_L: yeah you're out of bounds
03:05 PM Tom_L: fowl!!
03:05 PM Jin^eLD: well the "begin execution button" kinda looks like play :)
03:05 PM Tom_L: ball goes to the other team now..
03:06 PM jthornton: aye, are you homed?
03:06 PM Jin^eLD: I pressed "home all"
03:06 PM Jin^eLD: and it did move around weirdly
03:06 PM jthornton: open machine show hal configuration
03:06 PM Tom_L: you could have some switches reversed as well
03:06 PM Jin^eLD: I have it open
03:07 PM jthornton: click on the watch window and add the motion.spindle-speed-out-rps-abs pin
03:08 PM Jin^eLD: I see it, I cant run the sim though, but I expect this value to change
03:08 PM Jin^eLD: so basically this would be my api? http://www.linuxcnc.org/docs/html/man/man9/motion.9.html
03:08 PM jthornton: your trying to get ahead of me
03:08 PM Jin^eLD: I implement those pins in my component
03:08 PM jthornton: do you have that pin in the watch window?
03:08 PM Jin^eLD: and thats how I am being talked to? via those commands
03:09 PM Jin^eLD: I do yes
03:09 PM jthornton: in axis select the MDI tab
03:09 PM jthornton: then type S300 and press enter and nothing changes
03:10 PM Jin^eLD: done
03:10 PM jthornton: then type m3 and press enter and watch the value of motion.spindle-speed-out-rps-abs change
03:10 PM Jin^eLD: it does
03:11 PM jthornton: so you can connect that pin to your component and with logic figure out what output to turn on in your component
03:12 PM Jin^eLD: thank you :) thats the interface that I was looking for
03:12 PM Jin^eLD: list of pins to implement that define the api
03:12 PM jthornton: now add motion.spindle-speed-out-rps to the watch window
03:12 PM jthornton: then type m4 and enter in the mdi
03:14 PM jthornton: http://linuxcnc.org/docs/2.7/html/
03:14 PM Jin^eLD: both change
03:14 PM Jin^eLD: which makes sense
03:14 PM jthornton: yes but one is -5 and one is 5
03:14 PM Jin^eLD: for some reason both are -5 here
03:15 PM Jin^eLD: ah no, sorry
03:15 PM jthornton: if you scroll down to the bottom of that webpage you can see all the commands, userspace components and realtime componnets
03:15 PM Jin^eLD: I had the wrongone selected, not the abs
03:15 PM Jin^eLD: abs is 5 of course
03:15 PM jthornton: aye
03:16 PM Jin^eLD: I did look at the rtapi, which is also important of course
03:16 PM Jin^eLD: but I was failing to find this pin-api that you explained just now
03:16 PM jthornton: you can connect the output of your spindle speed component to some pyvcp leds while testing
03:17 PM jthornton: and use the mdi window to send G code
03:17 PM Jin^eLD: thats my plan, to rig this whole thing first
03:17 PM Jin^eLD: then once it works in the sim he'll connet some hardware with lamps, not yet the machine
03:17 PM jthornton: you can do this either in a component or in classicladder
03:18 PM Jin^eLD: component is probably easier for me
03:18 PM Jin^eLD: I know C quite well
03:18 PM jthornton: ah yes ladder would be a bit strange then it has some funny rules
03:18 PM Jin^eLD: :)
03:18 PM jthornton: the meanest is last one wins
03:19 PM jthornton: that gets so many plc programmers when they start out
03:19 PM * jthornton wanders off to check up on the chickens
03:20 PM Jin^eLD: :)
03:20 PM Jin^eLD: thank you! your explanations helped me a lot
03:20 PM Jin^eLD: at least I now think that I know how to continue
03:20 PM Jin^eLD: and the picture is much clearer now
03:21 PM fragalot_: Jin^eLD: would you mind keeping us in the loop?
03:21 PM fragalot_: I know of someone with a maho 700 that would like to know how this turns out
03:22 PM Jin^eLD: fragalot: i suggest to follow the thread I posted earlier, thats the thread of my buddy whom I am helping, he'll be posting about our progress
03:22 PM fragalot_: can you link it again? it's beyond my scroll buffer
03:22 PM Jin^eLD: but I will surely pester you guys here with questions too :)
03:22 PM Jin^eLD: https://forum.linuxcnc.org/12-milling/33035-retrofitting-a-1986-maho-mh400e
03:22 PM fragalot_: thx
03:22 PM Jin^eLD: np
03:39 PM JT-Shop: that's one complicated gear box lol
03:44 PM SpeedEvil: 'Well, my project is over, and on the plus side, only four people were killed when the gearbox setting went wrong and spun the part up to 48000RPM backwards'
03:51 PM JT-Shop: lol
04:03 PM Jin^eLD: lol ))
04:03 PM Jin^eLD: lets hope it does not end like that
04:03 PM Jin^eLD: I'll make sure not to be around when my code gets tested in real life ;)
04:17 PM XXCoder: wow they cleaned up the site! http://sevspace.com/
04:17 PM XXCoder: it used to be super awful
06:21 PM skunkworks: http://electronicsam.com/images/greenmachine/20180812_142249R.jpg
06:22 PM skunkworks: http://electronicsam.com/images/greenmachine/20180812_140623.jpg
07:19 PM Tom_L: skunkworks, makes more sense now
07:19 PM skunkworks: ?)
07:19 PM skunkworks: :)
07:20 PM skunkworks: I think I am going to make it 5"
07:20 PM skunkworks: I can always make it shorter..
07:20 PM Tom_L: i added 2" to mine last night now i gotta re'tram it all
07:20 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Final/Z-Riser_blocks.jpg
07:20 PM skunkworks: when I am all done - depending on how well things like up - I can always grind the surfaces
07:21 PM skunkworks: awesome!
07:21 PM Tom_L: for that little machine i think it will do fine
07:22 PM skunkworks: *mating surfaces..
07:22 PM skunkworks: http://electronicsam.com/images/greenmachine/20180812_141512.jpg
07:22 PM Tom_L: i may have to get a tiny bit thicker sacrificial block but it gives me more travel for the vise
07:22 PM skunkworks: it fit together awesome. Like I measured it or something.
07:23 PM Tom_L: hah
07:23 PM skunkworks: right - as of now - 4 inches of rise - almost any tool in the spindle will be touching the table.
07:23 PM Tom_L: it's awesome when you have machines to build machines...
07:23 PM skunkworks: so - 5 will make a vice less painful
07:23 PM XXCoder: machines all way in
07:25 PM Tom_L: may start checking it tomorrow or tuesday
07:25 PM skunkworks: yes - we really lucked out when dad pulled the K&T out of the junk yard...
07:26 PM skunkworks: also a great learning experience.
07:26 PM Tom_L: yeah
07:26 PM skunkworks: not to mention that for 68' The machine is really awesome
07:27 PM Tom_L: metal was real then
07:27 PM skunkworks: not much skimping
07:30 PM {HD}: Hey, anyone know about flatcam or pcb2gcode? I am having unusual results with both. Here are some pictures and a short description: https://github.com/a7hybnj2/KicadStuffs/tree/master/Issues%20I%20have%20been%20having
08:00 PM hazzy-lab1 is now known as hazzy-lab
08:05 PM skunkworks: hazzy-lab: http://electronicsam.com/images/greenmachine/20180812_142249R.jpg
08:09 PM hazzy-lab: skunkworks: Very nice!
08:09 PM XXCoder: lathe rail?
08:09 PM hazzy-lab: I like the way you have it go up to support the column better
08:09 PM hazzy-lab: XXCoder: mill column
08:09 PM Kevin`: i'm currently running linuxcnc on debian wheezy. is that still current? updated?
08:09 PM XXCoder: ahh nice
08:09 PM TurBoss: Kevin`: its current
08:09 PM XXCoder: curious if linuxcnc changes variable names with updates?
08:09 PM skunkworks: hazzy-lab: I had enough iron to do it.. :)
08:09 PM XXCoder: ie HAL works on older version and newer without changes
08:10 PM skunkworks: I keyed it into the column and base.. (well - the base part isn't done yet... - but it is mounted in the mill, trammed and probed)
08:10 PM skunkworks: http://electronicsam.com/images/greenmachine/20180812_141512.jpg
08:11 PM skunkworks: XXCoder: if there was - it would be in the change logs. I know there have been some conical name changes to make things more consistant..
08:12 PM XXCoder: not too bad
08:13 PM XXCoder: marlin seems to set variable names on wind
08:48 PM {HD}: Wolf__: added some better pictures of my laser issues. KicadStuffs/Issues I have been having/flatcam_result_1808122000.JPG
08:48 PM XXCoder: url dont work
08:49 PM {HD}: XXCoder: dang!
08:50 PM Wolf__: fail =P
08:50 PM {HD}: https://github.com/a7hybnj2/KicadStuffs/blob/master/Issues%20I%20have%20been%20having/flatcam_result_1808122000.JPG
08:50 PM {HD}: Thats what I get for using the github copy link button
08:51 PM XXCoder: looks like it dont quite turn off between engrave points
08:51 PM Wolf__: strange
08:52 PM {HD}: XXCoder: oh yeah, don’t mind that. I did that on purpose so I could see if the air was coming into the cut or going out of the cut.
08:53 PM XXCoder: ahh ok
08:53 PM {HD}: The issue is that last little leg isnt contacting the rest of the circle. There’s that little piece of black pie.
08:53 PM XXCoder: the circles at bottom left looks bit squat
08:54 PM XXCoder: I also see misalignment at circle near your pointer
08:54 PM XXCoder: similiar placement eveb
08:54 PM XXCoder: the one to up and left of arrow has same problem
08:54 PM XXCoder: all ones I can see have similiar issue
08:55 PM XXCoder: rest is too blurry
08:55 PM {HD}: Yes the issue is a parent with almost every circle cut. The circles at the very bottom are generated with different G code so those have different problems all together. Yes this is a macro shot and it’s difficult to get much more than I have in focus at once.
08:55 PM XXCoder: I suggest brighter light so lens open shorter time, meaning less blur
08:55 PM {HD}: Yes the issue is a parent with almost every circle cut. The circles at the very bottom are generated with different G code so those have different problems all together. Yes this is a macro shot and it’s difficult to get much more than I have in focus at once.
08:55 PM XXCoder: I HIGHLY recommand ikea led light. very bright and cheap
08:56 PM {HD}: For my photos or just in general?
08:56 PM XXCoder: both reallyt
08:56 PM XXCoder: but photography bighter light means less blur as general rule
08:59 PM {HD}: Yep. I could probably take a better photo pretty easily. To put it into perspective the ID of those circles is a scoosch less than an eighth.
08:59 PM Wolf__: moar light = quicker shutter speed
08:59 PM {HD}: “
08:59 PM Wolf__: I say its norias fault
09:00 PM norias: probably
09:00 PM {HD}: Moar light = tighter aperture = more focus.
09:00 PM {HD}: It was on a tripod anyway...
09:00 PM Wolf__: anyways =) can you paste bin the nc for that?
09:01 PM {HD}: Wolf__: step back and all the support files are there
09:01 PM XXCoder: for micro I use pretty bright light, flash isnt enough
09:01 PM Wolf__: shouldn’t be to hard to figure out if its a hardware or cam processor issue
09:01 PM XXCoder: I use my 500 lumen flashlight for it
09:03 PM Wolf__: wtf its all xy steps to do that?
09:05 PM {HD}: Wolf__: ¯\_(ツ)_/¯
09:05 PM Wolf__: what control on your laser?
09:06 PM XXCoder: wild guess, manual
09:06 PM XXCoder: just switch on.
09:06 PM {HD}: The G code was generated by flat cam or PCB to G code depending on which file you’re looking At.
09:06 PM Wolf__: I mean cnc controller flavor
09:06 PM XXCoder: my cnc router orginial "spindle" I had to turn on and off myself lol
09:06 PM {HD}: Its a elekslaser 5.5. I am running grbl 1.1 with laserweb
09:07 PM Wolf__: =/ I hate to admit that I haven’t looked at the gcode my laser runs lol
09:07 PM {HD}: XXCoder: I have a new 4 kilowatt spindle for my C&C machine I haven’t hooked in yet. I am looking forward to setting up the VFD.
09:08 PM XXCoder: nice! thats way more powerful than mine
09:08 PM XXCoder: mines just 400w
09:08 PM Wolf__: my laser is running smoothie w/ laser web
09:12 PM {HD}: Wolf__: give the G code a test burn. See if you have that same little leg. I am currently burning off flat black spray paint from a copper clad PCB.
09:13 PM XXCoder: hm didnt think of bad cam intertering
09:13 PM Wolf__: I can do that, tho no copper clad in the garage, its all at my other place
09:13 PM XXCoder: interpetering
09:13 PM XXCoder: Wolf__: honestly anything that leaves a mark would work lol
09:13 PM Wolf__: yeah, I was going to use cardboard
09:15 PM {HD}: You might not get the resolution necessary on cardboard. The error in question is .07 mm...ish
09:15 PM Wolf__: hmm
09:15 PM Wolf__: I’ll see what I have to use
09:16 PM Sabotend_ is now known as Sabotender
09:20 PM {HD}: All right this is as good as it’s going to get. At least for now.
09:21 PM {HD}: The new photo is on that page now
09:22 PM {HD}: That ruler is in inches btw
09:26 PM wolf_100w: ok lets see what I have to work with out here
09:27 PM {HD}: 100w apparently
09:27 PM wolf_100w: I wish
09:28 PM XXCoder: 100cw
09:34 PM {HD}: Latest pic if you didn’t see that msg: https://github.com/a7hybnj2/KicadStuffs/blob/master/Issues%20I%20have%20been%20having/flatcam_wruler.JPG
09:36 PM wolf_100w: ok looking
09:40 PM wolf_100w: is that a inch scale
09:40 PM {HD}: Wolf__: yep
09:40 PM {HD}: But each line is .01
09:41 PM wolf_100w: yeah 10th engineer scale
09:41 PM {HD}: Yep.
09:43 PM -!- #linuxcnc mode set to +v by ChanServ
09:44 PM {HD}: I need a metric version but I’ve never gotten around to acquiring one.
09:44 PM wolf_100w: https://i.imgur.com/a7pRvd0.jpg waiting on paint to dry on steel
09:47 PM {HD}: Hum hard to tell in that pic. The circles on the bottom appear to have some kind of issue.
09:48 PM wolf_100w: https://i.imgur.com/ImYAA3e.jpg
09:49 PM wolf_100w: my table isnt deep enough so my focus is off some
09:50 PM XXCoder: hard to tell, think I can see same gaps though
09:50 PM {HD}: https://i.imgur.com/PUTs0WE.jpg
09:51 PM {HD}: Looks like a similar issue. So I wonder if that means I’m leaning most towards the G code. Or I guess I could still be laser web since that’s what we’re both using.
09:51 PM XXCoder: wonder if laser focus size setting and actual focus burn area size isnt agreeing?
09:51 PM norias: _fuck_
09:52 PM {HD}: Gasp!
09:53 PM wolf_100w: that was the flat cam one, trying the p2 now
09:53 PM {HD}: XXCoder: how do you mean? You mean the burn width versus what I have it set to in code.
09:53 PM wolf_100w: then I'll try the nc direct on the smoothie
09:54 PM {HD}: Wolf__: direct on the smoothie? Neat.
09:54 PM XXCoder: yeah though I kinda doubt it, as it would be much more consisent
09:54 PM {HD}: XXCoder: right its only in the same spots around the circles.
09:55 PM wolf_100w: p2 came out ugly
09:55 PM {HD}: I kind of think it’s associated with acceleration. The last pass is the opposite direction of the four previous passes. Which means whatever slight error you get in circles based on acceleration would be opposite that for that last pass. Does that make sense?
09:56 PM {HD}: But then again all of your settings are probably different than my own. And the errors look almost identical. So…?
09:58 PM {HD}: I wish flat cam had an option to make all passes of isolation routing the same direction.
09:58 PM {HD}: I think that would solve The problem
10:00 PM wolf_100w: ok yeah, that p2 file just sucks lol
10:01 PM {HD}: Wolf__: lol. Picture! That one hasn’t been tweaked at all.
10:01 PM {HD}: Is it worse the the picture I have of the p2 file?
10:02 PM {HD}: Whoops I keep mentioning the wrong Wolf__
10:05 PM wolf_100w: same wolf, just too many instances
10:07 PM XXCoder: aw. toggle gif plugin is still not compitable. I miss that plugin so hard
10:08 PM wolf_100w: {HD}: https://i.imgur.com/G69Pcox.jpg both direct off smoothie sd card
10:09 PM XXCoder: yep I see same issue
10:09 PM XXCoder: burn width is larger so its hiding better but its still there
10:11 PM {HD}: Wolf__: looks better...? What do you think?
10:11 PM {HD}: Burn the metal
10:12 PM {HD}: I like that piece of wood the cardboard is on top of. What is that, Oak?
10:13 PM wolf_100w: yeah oak ply scrap I had laying around
10:13 PM wolf_100w: 550mm x 550mm table on this thing
10:13 PM {HD}: I had to get my extension tubes out for my macro photos.
10:14 PM {HD}: Thats big
10:14 PM wolf_100w: derp, 500x500 useable
10:15 PM wolf_100w: ok going back in the house now, its hot and humid af out here
10:17 PM {HD}: Wolf__: so whats the conclusion? Bad gcode?
10:17 PM Wolf__: yeah its in the cam
10:19 PM {HD}: Ugh. So, what do I do now? Ha...
10:19 PM Wolf__: run cam with a smaller dot size
10:20 PM Wolf__: looks more like its just missing part of the engraving area
10:20 PM Wolf__: that error might be visible in laser web when you load the NC file
10:20 PM Wolf__: but I didn’t think to zoom in and look
10:21 PM Wolf__: my test on steel didn’t work well the 2nd time, all I had for black paint here is textured rustolum
10:22 PM {HD}: Wolf__: I don’t see the error from laserweb gui
10:22 PM Wolf__: that missed spot, on the NC plot
10:24 PM {HD}: I have also tried very small tool size and more passes but I always have those little legs.
10:24 PM Wolf__: weird
10:25 PM Wolf__: I have been doing my stuff from DXF import in to laser web
10:25 PM {HD}: https://i.imgur.com/mrO8KEH.jpg
10:26 PM Wolf__: I have laserweb cam to gcode file, then send it to smoothie SD and then run it direct
10:26 PM {HD}: You can see here that the earlier versions were even worse.
10:26 PM {HD}: So you definitely get better results running a direct as opposed to sending it trickle from laser web?
10:26 PM Wolf__: usb sucks for live data
10:27 PM Wolf__: you know, it could be lens vibration
10:27 PM Wolf__: I wish I had better paint and copper clad to mess with
10:28 PM Wolf__: and AC where the laser is
10:30 PM Tom_L: i got a pile if you wanna come get it
10:31 PM Wolf__: I have a 4’x8’ sheet at my shop…
10:31 PM Wolf__: came with my house…
10:31 PM {HD}: Wow
10:31 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/Copper3.jpg
10:31 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/etching/Copper1.jpg
10:32 PM Wolf__: damn Tom_L wins
10:32 PM Tom_L: that's all i kept
10:32 PM {HD}: But why?
10:32 PM Tom_L: well i've got several of the big sheets
10:32 PM Tom_L: i had 2 pickup loads
10:32 PM {HD}: From where?
10:32 PM Tom_L: cause i didn't need it all
10:33 PM Tom_L: a closed boardhouse i got the stock from
10:33 PM Tom_L: and random cutoffs
10:34 PM Wolf__: I know who to bug for a flat rate game board box full when I run out =]
10:34 PM Tom_L: i'm about tired of tripping over it
10:35 PM Tom_L: the big sheet was full sheets but i cut them in half to store
10:38 PM {HD}: Wolf__: the testpcb files are in that data. And the exported gerber is in there too if you wanted to go crazy! I am going to try a different cam.
10:39 PM Wolf__: Tom_L: where are you at?
10:40 PM Wolf__: {HD}: I havent tried going from board design to laser, yet
10:43 PM {HD}: Just for fun I am trying “universal G code sender“
10:50 PM {HD}: Same error :-(
11:14 PM ziper: jesseg, im calling off the autoboat
11:44 PM jesseg: ziper, awww wha?
11:44 PM jesseg: as it giving up?
11:45 PM jesseg: probably not enough time to get in done in time this go around huh