#linuxcnc Logs

Nov 11 2018

#linuxcnc Calendar

12:02 AM cnomad: evilroot: not too sure how to do this. it seems like i would need a pretty big reservoir cap to smoothen this to ~3.3v DC
12:06 AM cnomad: one sec let me grab a pic
12:06 AM cnomad: evilroot: https://i.imgur.com/jn01syP.png <-- 'tis what i get unfiltered
12:06 AM cnomad: so I guess I'd need an amp on the other end?
12:06 AM cnomad: this really doesn't seem right, something else might be wrong?
12:07 AM cnomad: this is 60 hz. suspicious
01:33 AM r0n0x: heyo, anyone have experience with taig mills?
01:48 AM XXCoder: lol https://www.aliexpress.com/item/Heavy-Duty-3D-Wood-CNC-Machine-4x8-Feet-CNC-Milling-Machine-With-Rotary-4-Axis-Milling/32917223303.html
01:48 AM XXCoder: just $5k and no shipping
02:20 AM Deejay: moin
04:27 AM r0n0x: anyone use or see a taig cnc mill in action?
04:28 AM XXCoder: not me
04:35 AM Tom_L: morning
04:36 AM XXCoder: hey tom
04:39 AM Tom_L: https://www.youtube.com/watch?v=TkEOMQ6rQ6s
04:39 AM Tom_L: suck it #reprap
04:39 AM XXCoder: yeah heard of that thing
04:39 AM XXCoder: though more finished
04:44 AM MarcelineVQ: are the plans and software for that thing free?
04:44 AM XXCoder: i dont know, because slicer for that thing would be very complex
04:44 AM Tom_L: that would be doubtful
04:45 AM MarcelineVQ: suck it reprap might not be the right exclaimation in that case :o
04:48 AM Tom_L: https://www.mdpi.com/2227-7080/5/2/36/htm
04:52 AM Tom_L: https://www.engineering.com/3DPrinting/3DPrintingArticles/ArticleID/14149/Autodesk-Project-Escher-Finds-First-Commercial-Release-in-Cronus.aspx
04:52 AM MarcelineVQ: impressives section 1, will read more later
04:57 AM Tom_L: https://www.autodesk.com/products/netfabb/overview
05:04 AM jthornton: morning
05:32 AM XXCoder: hey jthornton
05:32 AM XXCoder: I found something interesting, apparently you can buy a camera POV mount for less than dollar
05:32 AM XXCoder: I suppose could make a live camera view for your chicken coop or something
05:33 AM jthornton: mount it on a chicken?
05:33 AM XXCoder: that would be interesting but no
05:35 AM XXCoder: https://www.aliexpress.com/item/Hobbywing-LED-Program-Card-For-FlyFun-SkyWalker-Brushless-ESC-Airplane-Heli-Free-Shipping/32508992281.html
05:35 AM XXCoder: that one costs bit more because has servos
05:36 AM XXCoder: its designed for plane but had couple interesting ideas lol
05:37 AM jthornton: I could put a small camera like a USB camera on there hmm how to move the servos?
05:37 AM XXCoder: you know more about such requirements than I do
05:40 AM XXCoder: how goes your buttons project
05:49 AM jthornton: this seems to work well https://paste.ubuntu.com/p/h2Rc9Kt4hv/
09:12 AM phipli: anyone got any experience with Node-RED?
09:22 AM Tom_L: jthornton, that code looks solid now
09:23 AM phipli: Tom_L, you scared him off!
09:27 AM Tom_L: JT-Shop, is that a python thing to do += instead of ++ ?
09:28 AM Tom_L: or just swamp style?
09:29 AM tiwake: += is a standard programming thing
09:29 AM Tom_L: so is ++
09:30 AM tiwake: x=x+y or x+=y
09:30 AM Tom_L: it probably compiles out the same in the end
09:31 AM Tom_L: deb24 += 1 or ++deb24
09:32 AM phipli: I guess he's got used to ++ not being available if he's using a mix of languages
09:32 AM Tom_L: i don't do python was why i wondered
09:32 AM tiwake: Tom_L: ++x is different from x++
09:33 AM MarcelineVQ: notably ++deb24 is different than deb24++ as well, languages aren't so directly comparable sometimes
09:33 AM Tom_L: tiwake i realize that
09:33 AM phipli: Tom_L, yeah, no i++ in python
09:33 AM tiwake: wait you cant do x++ in python?
09:33 AM phipli: nope
09:33 AM tiwake: fail
09:33 AM Tom_L: i don't do python was why i wondered
09:33 AM phipli: it doesn't work in hardly anything other than the C family
09:34 AM MarcelineVQ: fwiw += is quite common in imperative languages like tiwake mentions
09:35 AM Tom_L: i just wondered why he chose it over ++ and it seems it's due to the lanuage barrier
09:35 AM tiwake: I generally like to do x=x+y because its a little more clear typing.. easier to read... but it does not really matter much
09:36 AM tiwake: language designers love their shorthands
09:36 AM Tom_L: really i just thought it was too quiet in here
09:36 AM phipli: Tom_L, do you add many of your own shorthands as a library?
09:36 AM phipli: (out of curiosity)
09:36 AM Tom_L: not much anymore
09:37 AM Tom_L: i don't program that much now
09:37 AM phipli: :'(
09:37 AM phipli: Job change?
09:37 AM Tom_L: i never did it at a job
09:37 AM MarcelineVQ: > succ deb24
09:38 AM phipli: Nor me - it has always been a tool to avoid repetitive tasks at work
09:39 AM phipli: but... I hide the quality of my code from the real programmers due to shame. (Except when I want them to re-implement something to make it client safe)
09:39 AM Tom_L: i've seen some real programmer's code too and some isn't all that pretty either
09:40 AM Tom_L: some just like to make it confusing to throw you off
09:40 AM MarcelineVQ: phipli: this is wise
09:41 AM MarcelineVQ: Tom_L: obfuscation is a pass-time, as is golfing, doing it to throw people off without them knowing is mean though
09:41 AM phipli: MarcelineVQ, my girlfriend used to work with them so they'd only make fun of me :)
09:51 AM JT-Shop: hi phipli how's it going?
09:52 AM phipli: Hey JT-Shop - not bad
09:52 AM phipli: had a week off work
09:52 AM phipli: pigeon died week before though
09:52 AM JT-Shop: I had to get rid of one hen, she was an egg eater and made a mess in the nest box
09:52 AM phipli: yourself?
09:53 AM JT-Shop: neighbor took her in, she got a boyfriend and roosts on the big dog roost next to the rooster so everyone is happy now :)
09:53 AM JT-Shop: doing ok going to take off next weekend for a short holiday
09:54 AM JT-Shop: trying to get some wind blocks put in today on the run
09:54 AM phipli: Going somewhere on the bike?
09:55 AM JT-Shop: naw, too cold to ride the bike we are taking the Honda car though
09:56 AM JT-Shop: got another hen that has decided to start laying in the roost box...
10:00 AM phipli: Ours are only laying about one egg per day at the moment
10:07 AM jthornton: trying to come up with a bullet proof debounce and a timeout for the chicken door if you have any ideas?
10:17 AM phipli: where are you getting the bounce?
10:20 AM phipli: if you're worried about false triggering, stick two switches on it? Or three?
10:22 AM phipli: If you're talking about in a manual control "up and down" button, I'd just test a few times over 10ms in a for loop to avoid false triggers, and then have a 500ms cool down period. Plus use edge detection (change of state) rather than "is it on".
10:27 AM JT-Shop: I tried edge detect but the debounce does not work as it should... the main thing I'm trying to do is have a timeout for the door lol
10:27 AM JT-Shop: https://paste.ubuntu.com/p/h2Rc9Kt4hv/ this sorta works but is clunky lol
10:28 AM phipli: is that detecting light levels?
10:29 AM phipli: no...
10:30 AM phipli: so every 10th second you are testing if the buttons are pressed or not
10:31 AM phipli: and counting up to 3 if they are each time, or down if they're not each time - within the range 0 to 3
10:32 AM phipli: but if you get false triggers, there is not "return to centre" type feature
10:32 AM phipli: no, sorry, it is the same gpio so that doesn't matter
10:34 AM phipli: I'd consider moving the "lights" check into the if statement above : if deb24 == 3 and not lights:
10:40 AM phipli: JT-Shop, simple solution - react to the first state change, then ignore everything for 100ms
10:40 AM phipli: if it was a spurious trigger, then you will know after that 100ms and it wont really matter. If you're getting constant spurious triggers, that is a wiring issue, not a debounce issue! :)
10:50 AM merkatorix: My latency running it for around 18 hours looks like this: https://pasteboard.co/HMGAajW.png
10:50 AM JT-Shop: yea it's just switch bounce and trying to do something one time
10:51 AM JT-Shop: phipli: how would you time the 100ms?
10:51 AM merkatorix: I still get this error: https://pasteboard.co/HMGB945.png
10:52 AM merkatorix: RTAPI ERROR: Unexpected realtime delay on taks 1
10:53 AM JT-Shop: maybe your base thread is set too fast or something on startup causes it???
10:53 AM phipli: JT-Shop, record the event time + 100ms as FutureTime and then only run the button tests if now>FutureTime
10:53 AM * JT-Shop goes back to the other shop
10:53 AM phipli: JT-Shop has too many Shops. Perhaps he should buy a single big computer and remote into it from all other rooms
10:53 AM JT-Shop: phipli: ok I'll give that try in a a bit
10:54 AM merkatorix: JT-Shop, So you mean the settings file says it should be even faster than whan it displays (which is below 21000ns)
10:54 AM JT-Shop: lol, yea I have a single big computer as a server that does nothing lol
10:54 AM phipli: :)
10:54 AM JT-Shop: merkatorix: not sure what your saying
10:54 AM phipli: you know it is serious when I go and start my desktop... I'm too lazy to do it most of the time
11:03 AM merkatorix: I tried to ask if "base thread is set to fast" is referring to the settings I went through in the stepconf wizzard.
11:03 AM merkatorix: I thought I turned everything way down, but I'll check again :)
11:22 AM JT-Shop: well turning down the base thread time is the wrong way... you need to increase it (make it take longer)
11:30 AM jthornton: phipli: the modified manual light test code https://paste.ubuntu.com/p/TMZ9hBNcwR/
11:30 AM * jthornton goes back to the shop
11:36 AM phipli: JT-Shop, that looks ok to me - I'm not good enough at python to spot issues though... I leave that to the error reporting :)
11:36 AM JT-Shop: lol
11:37 AM phipli: JT-Shop, I guess it is a toggle switch?
11:37 AM JT-Shop: aye
11:38 AM gloops: wonder what these are off https://www.ebay.co.uk/itm/Cnc-Automatic-Project-Stepper-motors/202497111127?
11:38 AM phipli: do you have a pull up or pull down resistor on it?
11:41 AM JT-Shop: I'm using the internal pull down resistors
11:43 AM phipli: so you are, on line 16
02:24 PM Jin|away is now known as Jin^eLD
03:03 PM JT-Shop2: test
03:11 PM cnomad: so, finally found a wiring diagram for a _related_ stage: https://www.global-optosigma.com/en/page_pdf/OSMS-ZF.pdf?v=1499623066
03:11 PM Tom_L: pass
03:11 PM cnomad: which confirms that +VS is 5-24 VDC
03:11 PM cnomad: also tells me the photo sensors in use
03:12 PM cnomad: doesn't suggest that there's any current being sourced from the motor leads, so oit might not be telling the whole story; but otherwise it's difficult for me to imagine why the sensors won't output _anything_ until i switch the driver on
03:13 PM cnomad: still, though... getting these 60Hz pulses through the limit sensor output
03:27 PM JT-Shop: Tom_L: you get any snow?
03:29 PM Tom_L: not yet but we're supposed to later on
03:30 PM Tom_L: 2-3"
03:59 PM cnomad: evilroot: issue resolved! it turned out to be an open collector in the sensor, so I had to add a pull-up between the input and the LIM output
04:00 PM cnomad: my assumption that it was sourcing current from the motor leads was false. it's just that it was really sensitive to coil noise when there isn't a voltage source on the other end of the collector
04:00 PM cnomad: yay!
04:08 PM Deejay: gn8
05:03 PM Jin^eLD is now known as Jin|away
06:41 PM XXCoder: JT-Shop: saw your new code. nice
06:43 PM XXCoder: nice and pretty tight debounce code
06:44 PM XXCoder: since it is pretty tight loop, maybe l.ook into trying to use else if to cut some branching checks
06:44 PM XXCoder: for example if deb24 == 3 blah blah else if deb24 == 0 blah blah
06:45 PM XXCoder: not too sure what python version would be
06:46 PM XXCoder: I see 2 locations where "else" can be added
09:31 PM skunkworks: https://youtu.be/mAZei3wpBe0
09:32 PM XXCoder: looking
09:34 PM XXCoder: nice
09:34 PM XXCoder: wow he can punch cut a pipe lol
09:36 PM skunkworks: he isz entertaining
09:37 PM skunkworks: he was given a free mach 4 industrial package...
09:37 PM XXCoder: dang
09:39 PM skunkworks: i like how he is showing off css.. (because it didnt work in mach 3
09:40 PM skunkworks: sorry.. irssi doesnt have spell check
09:42 PM XXCoder: I love his rotatary weld station
09:44 PM skunkworks: he has a lot of great videos
09:44 PM XXCoder: he does. thats why I subbed his channel quite a while ago
09:45 PM skunkworks: yes
09:47 PM XXCoder: "pro tip, if fitted bearing still turns its good sign" lol
09:48 PM XXCoder: man wish I had even 10% of his skills.
09:53 PM Rab: Did he just press the spindle shaft into the bearing without supporting the inner race? (14:20)
09:53 PM XXCoder: i dont know what inner race is
09:53 PM Tom_L: bearing ID
09:53 PM SpeedEvil: The inner cylinder of a ball bearing
09:54 PM Tom_L: i suppose 4 jaw would be best on something like that
09:55 PM XXCoder: so far he havent solved slight off center issue
09:55 PM XXCoder: maybe its just off angle or maybe its chuck itself being off center
09:55 PM Tom_L: not critical on a 4 jaw
09:55 PM XXCoder: yah thats what he said, just want mor etrue setup for easier time
09:55 PM Tom_L: no sound here
09:56 PM XXCoder: same here unfortunately, at both ends lol
09:56 PM Tom_L: was that a stepper or servo?
09:57 PM XXCoder: not sure, wiring connection style looks like servo but dunno
09:57 PM XXCoder: or maybe its encoded stepper
09:57 PM SpeedEvil: It's a complex encoded stepper/servo thingy that he's only using part of the functionality on.
09:57 PM SpeedEvil: At least it was last video of his
09:58 PM XXCoder: oh yeah!
09:58 PM XXCoder: he was programming it isnt he?
10:00 PM XXCoder: little backlash there
10:08 PM CaptHindsight: bit broke
10:09 PM XXCoder: on what mkachine maching what thing?
10:10 PM CaptHindsight: in the vid skunkworks posted
10:10 PM CaptHindsight: ~40 min ago ouch
10:10 PM XXCoder: oh, yeah surpised it lasted that long due to fairly bad backlash
10:11 PM CaptHindsight: https://youtu.be/mAZei3wpBe0?t=113
10:11 PM XXCoder: oh drill broke
10:11 PM XXCoder: no idea how that happened lol