#linuxcnc Logs

Sep 15 2018

12:58 AM justan0theruser is now known as justanotheruser
01:07 AM veek: has anyone tried making powder metallurgy parts using an array of 5T car jacks ? is anything else needed
01:09 AM veek: oO https://www.researchgate.net/post/i_want_to_fabricate_a_die_for_powder_metallurgy_what_should_be_the_design
01:28 AM fragalot: hi
01:39 AM CaptHindsight: veek: pressure is pressure, you just need enough in the right direction
01:40 AM CaptHindsight: veek: are you trying to press metal powders and then sinter?
02:06 AM veek: CaptHindsight, not actually trying.. as such.. just curious.. i'd like to make parts for angle grinders/tools if they break and such
02:06 AM Deejay: moin
02:17 AM fragalot: miss0r|office: face mill arrived. I think the 80mm one I bought is actually 120mm
02:18 AM fragalot: miss0r|office: it looks.. actually kinda alright when mounted in the spindle, but awfully out of proportion on the taper >.>
03:48 AM gloops: done, with 45 minutes to spare haha https://ibb.co/jDbtez
03:48 AM XXCoder: hh your webpage scales it waaay up
03:48 AM XXCoder: looks great!
03:50 AM MarcelineVQ: It does, probably even better head on as well, it's hard to read the cleary at that angle
03:52 AM gloops: well i just took a couple of snaps in what look a good light, mrs wanted to show someone, shes taking it to a wedding in a bit lol
03:53 AM gloops: needed a bit of sanding worked where i touched second tool off slightly out but too late to mess about
03:59 AM XXCoder: you dont use paper to touch off?
04:01 AM gloops: on this occasion i simply felt for resistance while turning the tool, not much out but a few little deeper marks here and there, old oak doesnt sand quickly by hand, so it was just a quick rub over, should have been done 2 week ago but i didnt get round to it lol
04:02 AM gloops: nothing noticeable
04:06 AM XXCoder: cool
04:59 AM gloops: any cutting with the router yet Blumax?
05:00 AM Blumax: Hello gloops
05:00 AM gloops: hello
05:00 AM Blumax: I'm looking for how to check the perpenducularity
05:01 AM Blumax: I use endstop effect hall, I can adjust the moment of detection accurately but before I must be sure that the machine is well perpendicular
05:02 AM XXCoder: theres few ways to do hat, Blumax
05:02 AM gloops: well, tramming is one way
05:02 AM XXCoder: do you have dial test indictor or bigger dial indictor
05:04 AM Blumax: XXCoder I do not understand
05:04 AM XXCoder: example of dial test indictor data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAAAQABAAD/2wCEAAkGBxMSEhUQEhMVFRMXGBgVGRUVFRYXGhcgFxgXFhoVFhkaICggGBsnHxYYITEhJSktLi4uGB83ODMsNyktLisBCgoKDg0OGhAQGy4mHyUtKy0tLS8wLSstLS0tKy0tLS0vLystLS0tLS0tLS0tKy01LS0tLS0tLS0vKy4tLSstLf/AABEIAOUA3AMBIgACEQEDEQH/xAAcAAEAAgMBAQEAAAAAAAAAAAAABQcDBAYBCAL/xAA/EAABAwIEBAQCBwcDBAMAAAABAAIRAyEEEjFBBQZRYRMicYEHkTJCUqGxwfAUI2KSotHxgrLhQ3LC0iREU//EABoBAQADAQEBAAAAAAAAAAA
05:04 AM XXCoder: AAAABAgMEBgX/xAArEQEAAgIBBAECBAcAAAAAAAAAAQIDESEEBRJBMSJRE2Gh8RVScZHB4fD/2gAMAwEAAhEDEQA/ALxREQEREBERAREQEWHE4plNpfUe1jRq5zg0D1J0UXiObMDTOV+LoAxMeK38jr2QTSLnm888OMf/ADKPm0l8fzT9H3hbmG5kwdQSzFUHXy2qs16aoJVF4CvUBERAREQEREBERAREQEREBERAREQEREBePcAJJAA3NlHce43SwlM1ap9GgiT6SRYakqmOZ+b6+OdkBIpZi5jQBPQFvQR9d3UwBKCwOYviXh8P5aQNZ8kEXbpuLEu+Ud1XvE+esfiZAreAzNMUpDh0BLLgdi5aHD+CPqiGNlpMOdJDQf436vP3dgut4Zycxo/eEuLYlrPI2OojzH57FcGfuGLFxvlpXHayv8Z
05:04 AM XXCoder: RlxfVNSo91y6pUgk9bSXH3X7wfDs8RRBB6U6rvvzXVuUeB0qQDmUmDK6ZDRJBsZOpgOP8oUpSwgEdnF3zm33r5l+8/wAsNY6f7yo7HcLqDTDmJi1CptPQ9lF1qeWc9IN2u17PxlfQr8KLdnF3zzW/q+5atLBAtbIEZnPIP8RcY/q+5Z/xq0TzX9VvwIUXw3itegSaFarSkXNOoYMaAwQT8l3XAfiljaZis1uJYBFvI8dHOIBPzaulxvKOGqCX0WA3c5zRlPpLYJ1/pXMcR5Ec0g0XSfpZX2LBtDxodvWb2K7MHd8d+LcM7YJj4WbyzzphcaAKb8tUiTTdqI1Ado72vGy6NfO1TBlroqtc1/2wAHnuR9Gs31v0cV2PKHP76Rbh8SfEZGVjmi5jpMGQNWO83Tovr0yVvG4YzEx8rYRY6FZr2h7SHNN
05:04 AM Blumax: LinuxCNC is correctly configure it works as I wish.
05:04 AM XXCoder: wQZBWRXQIiICIiAiIgIiICIiAiIgLQ45xanhaLq9UkNECwkknQBb6oj4wcyur4k4VhcKdGWuE2c76zoHrlv0PVBBcd5grcQrl7yCIAsIbDZiRu0SYB1Oq6bgPL4Iz1AY3H1yftVO38P5WEVybwWG+M5ocdQ2YMjQ+g/H0CsThNVgvUlrh9sRr1OnpovP9x6+3NMbpxYvctjC8OLQCA2dMos1w9NipHDVGHytIMd57fkb9iskrDRwjWuzCfrQJsMxBdG9yBbRfAi3l8uhnKBfivWDdT7blaVbFGCSckaaXIuZMyfQW7lbYumvknj4UvkrVn4lhzUYWh0ayNnSCMrtwL6i+ix4LClkyReDAEAHcx1O5EC2gWzSqZmyRB3Fx+O269Cwy+VZmktKzExshY/AF5vmMmd+3pFo/uVmX4rVA0ZjOwt1JAA+
05:06 AM Blumax: Hello SVG URL spammeur :p
05:06 AM XXCoder: oops! sorry had to close client to stop the spam
05:06 AM XXCoder: http://www.kaefer-messuhren.de/files/kaefer-messuhren/Englische%20Version/01_Products/Dial_Test_Indicators_big.jpg
05:06 AM XXCoder: thats test version
05:07 AM Blumax: Yes I have this :)
05:07 AM XXCoder: you could use it to check stuff
05:07 AM Blumax: But I'm not a reference surface
05:07 AM XXCoder: its for trimming your machine
05:08 AM XXCoder: basically you attach it to spindle somehow with offset as large as you can make it
05:08 AM gloops: https://www.youtube.com/watch?v=D6pfIoyhgTc
05:08 AM gloops: also Blumax - surface the spoil board
05:08 AM XXCoder: then you rotate it around and adjust spindle till its nicely trimmed
05:09 AM XXCoder: interesting gloops lol
05:10 AM gloops: well, simple idea but will get you pretty close
05:10 AM XXCoder: yeah depends on your tols range
05:10 AM Blumax: Oh sorry, i want settle two motor Y granty
05:10 AM XXCoder: that wont work for .001" +- tol parts
05:10 AM gloops: ahh you mean square Blumax?
05:11 AM gloops: square the gantry, X > Y
05:11 AM Blumax: yes square :)
05:12 AM gloops: nobody really knows how to do that lol
05:12 AM XXCoder: measuring and adjusting is so much fun
05:12 AM XXCoder: measure corner to opposite corner
05:12 AM XXCoder: then other set
05:13 AM XXCoder: see if equal
05:13 AM Blumax: With a precision square and a feeler
05:13 AM Blumax: But the big square is very expensive :(
05:13 AM gloops: i started with a kitchen door - i knew had been machined reasonable square, then get x and y travel passing along each side just brushing it
05:13 AM gloops: you can do pythagoras 345 measurements etc
05:13 AM Blumax: I have a big glass table.
05:14 AM gloops: cut squares and try them with engineers square, draw lines on the board with a pencil and test that
05:15 AM gloops: it comes down to nudging the home switches a touch this way and that way until youre happy, i couldnt devise any scientific method to do it
05:16 AM Blumax: Ok, shame, I'll do as I can with the glass table ^^
05:16 AM XXCoder: well heres silly thing you can do
05:16 AM XXCoder: tape a paper to your table
05:16 AM XXCoder: attach a pen
05:16 AM XXCoder: program it to draw a recangle or square
05:16 AM XXCoder: it would show if your machine is out of square
05:16 AM XXCoder: circle is good though bit harder to check
05:17 AM XXCoder: in fact draw a diamond inside circle inside square
05:17 AM XXCoder: it combines all motion ponental issues so it could expose issues
05:18 AM XXCoder: square is single axis, diamond is equally two axis, circle hits ALL ratios
05:21 AM XXCoder: Hmm I'd suggest perment marker as it has softer tip so you dont break pen or pencil on paper lol
05:21 AM Blumax: The machine is 838 x 1425mm
05:22 AM gloops: yeah that circle in a square cut is a good guideline
05:22 AM XXCoder: diamond circle square yeah
05:23 AM MarcelineVQ: XXCoder: you'd want a mid-quality bic ball-point pen, low quality would clog up, other writing implements ends would wear during use
05:23 AM MarcelineVQ: Just to min-max this task :>
05:23 AM XXCoder: lol ok
05:23 AM MarcelineVQ: What we need is good card stock!
05:26 AM gloops: if you could set up 2 laser measures at one end of X, pointing diagonally across each other to the other end of X, you
05:26 AM gloops: could get a perfect square
05:27 AM XXCoder: hold on a second
05:27 AM gloops: but then you gotta get both angles perfect etc, its not worth it for a router
05:27 AM XXCoder: whoa picpaste is down?
05:27 AM gloops: imagbb
05:28 AM XXCoder: Blumax: https://imgur.com/a/5Y8c0KG
05:28 AM XXCoder: I eyeballed the diamond a little but you get the idea lol
05:28 AM gloops: i had a vid of mine doing that test but must have deleted it
05:28 AM XXCoder: too bad
05:29 AM gloops: if the circle is a true circle, the square should fit nicely in it
05:31 AM Blumax: So if I understand correctly, I make a circle with the hand and I ask the CNC to do the square?
05:31 AM XXCoder: nah it does all the drawing
05:32 AM XXCoder: if machine has errors there is bunch different symtoms
05:32 AM XXCoder: squashed a little, skew, circle not properly placed inside square (or other way if gloops version)
05:33 AM XXCoder: circle is hardest because there is bunch of ponental issues
05:33 AM XXCoder: the transition where it goes to, stops and reverses on one axis while other is going max speed can cause issues if theres hardware issues
05:34 AM gloops: i could make you some code to run, but have to dash out right now for 10 minutes
05:35 AM Blumax: I know how to use the GCode and LinuxCNC, I have already machined quite a lot but I found problems of perpendicularity on certain piece. That's why I changed my machine.
05:36 AM XXCoder: heh I havent really worked on my machine for quite a while, but then I run big machines at work :P
05:36 AM XXCoder: biggest being a61 makino lol
05:36 AM XXCoder: whole cell
05:40 AM Blumax: Beautiful machine
05:41 AM XXCoder: it runs quite well, most times
05:41 AM XXCoder: couple flaws is chips tend to build up near tool break detect laser at far end
05:41 AM Blumax: Okay, I'll do with my glass table and a comparator. With pytagore I will check if the table is perpendicular.
05:42 AM XXCoder: today I ran tiny robodrill though lol
05:42 AM XXCoder: engraving very hard metal parts :P
05:42 AM XXCoder: its so hard tool gets useless after 30 parts
05:42 AM XXCoder: compare it to similiar engrave of stainless steel it can run 1,000 parts or more before tool gives up lol
05:47 AM diverdude: Hi, I am trying to run this code: https://www.brainy-bits.com/setting-stepper-motors-home-position-using-accelstepper/ to home my stepper motor...but when the code has been uploaded the motor is just giving a buzzing sound and is not moving. What am I doing wrong?
05:47 AM XXCoder: buzzing sound hm have you configured microstepping and such?
05:50 AM miss0r: XXCoder: This is how far I got with the control cabinet last night: https://imgur.com/a/13lPAwY
05:50 AM diverdude: XXCoder: hmm yeah... i can move my motor if i just do something like this: https://paste.ubuntu.com/p/6Tg7rT9ZqM/ so the motor is working
05:50 AM diverdude: XXCoder: its just the homing using accelstepper which is not working for some reason
05:51 AM XXCoder: I cnt directly answer that, its been a while but there may be troubleshooting stuff I can suggest
05:51 AM XXCoder: your machine may be trying to accerate too hard
05:51 AM XXCoder: given its configuration
05:51 AM miss0r: diverdude: Does the stepper give off sound?
05:52 AM XXCoder: miss0r: high quality cardboard!
05:52 AM diverdude: miss0r: yeah its like bzzzzzzzzzzzzz
05:52 AM miss0r: XXCoder: indeed hehe
05:52 AM XXCoder: no pizza leftover stains on it
05:52 AM miss0r: diverdude: Alright. That sounds like you are either a) accelerating too hard or b) running it intirely too fast
05:52 AM XXCoder: or microstepping is too high
05:52 AM XXCoder: it loses torque with that
05:52 AM diverdude: miss0r: aha so if i increase the delay it might help
05:53 AM miss0r: what delay?
05:53 AM miss0r: line no.
05:53 AM miss0r: the delay(1000) ?
05:54 AM diverdude: miss0r: here is a delay: https://www.brainy-bits.com/setting-stepper-motors-home-position-using-accelstepper/ delay(5);
05:54 AM miss0r: no
05:54 AM miss0r: that delay is during the setup phase. That does not effect your loop time.
05:54 AM miss0r: And using accelstepper you do not need to think about the loop time.
05:55 AM miss0r: you need to ajust: stepperX.setMaxSpeed(100.0); & stepperX.setAcceleration(100.0);
05:55 AM miss0r: although, those both seem alright
05:55 AM diverdude: miss0r: hmm ok. the thing is that i set stepperX.setMaxSpeed(100.0); stepperX.setAcceleration(100.0); so those values are definitely not to big
05:55 AM miss0r: (if those are, infact, the values you use
05:55 AM miss0r: so, what is it you have working, exactly?
05:56 AM diverdude: because in my working accelstepper code i set it to stepperY.setAcceleration(5000.0); stepperY.setMaxSpeed(10000.0);
05:56 AM miss0r: ahh...
05:57 AM diverdude: this very simple code is working;: https://paste.ubuntu.com/p/3X5vQNQbFx/
05:57 AM miss0r: hehe. in fact, this might be so slow that it does not seem to move / accelerate at all.
05:57 AM miss0r: Try to increase those values
05:58 AM miss0r: 2 secs, on the phone
05:58 AM gloops: give it some stick diverdude
05:59 AM diverdude: but if i set it to stepperX.setMaxSpeed(10000.0); stepperX.setAcceleration(5000.0); (in the homing code) no sound is even coming from the motors
06:02 AM gloops: im not familiar with stepperaccel whatever it is, but it sounds like the motors are being asked to perform extremes they cannot achieve
06:03 AM miss0r: back
06:04 AM XXCoder: wb
06:04 AM miss0r: diverdude: Theres something off about the code that is not working.-
06:04 AM miss0r: The way you are calling the accelstepper
06:05 AM XXCoder: yo jt
06:05 AM miss0r: diverdude: Please make me two pastebins; one of the code you want to have working, like you have tested it and failed. and one of the working code you have
06:06 AM jthornton: morning
06:06 AM miss0r: mornin' jt
06:06 AM jthornton: morning miss0r
06:07 AM MarcelineVQ: in case it's important to the task at hand: according to the docs setMaxSpeed is in steps/sec and setAcceleration is steps per sec per sec
06:08 AM MarcelineVQ: Which you guys probably guessed but I was curious about what units were being used so there it is just in case
06:10 AM miss0r: MarcelineVQ: No guessing :) We read manuals
06:10 AM miss0r: This is not ikea furniture. The manuals are worth something
06:10 AM XXCoder: yeah understanding machine is key
06:11 AM jthornton: thank you I spent a long time working on the manuals
06:11 AM XXCoder: hguy at work who badly hurt hand didn't understand that chip conveyer at back could run seperately and does, contrast it with most machines its linked
06:11 AM miss0r: pop, theres goes the warranty on my new computer PSY
06:14 AM jthornton: I need a 17" laptop that I can install linux for the road using a VM sucks
06:14 AM diverdude: miss0r: this is the code which i am trying to get to work (the homing procedure) https://paste.ubuntu.com/p/5vs4zPrRF9/ AND here is the code which is working - which is just accelstepper moving forward and backward https://paste.ubuntu.com/p/3X5vQNQbFx/
06:14 AM Loetmichel: *hmpf... you think my wife will ever learn NOT to send the husband out for grocerys on an empty stomach? Wife has written up foods for 60 eur... i came back with foods and snacks for 217 eur... :-)
06:15 AM XXCoder: lol
06:15 AM jthornton: opps there goes the diet
06:17 AM diverdude: Loetmichel: haha nice :D and thats how we learn!!!
06:19 AM miss0r: diverdude: and you have sent the information it needs via the serial connection?
06:19 AM gloops: money spent on food is never wasted
06:19 AM diverdude: miss0r: mmm the setup procedure should just begin to home directly without getting input
06:19 AM diverdude: miss0r: its only in the loop procedure its asking for input
06:19 AM Loetmichel: gloops: not wasted unless it spoils before you can eat it
06:20 AM miss0r: diverdude: sure. But the 'setup' part does not actualy move the stepper
06:20 AM gloops: better get eating Loetmichel lol
06:20 AM miss0r: well, okay, it does
06:20 AM diverdude: miss0r: i think it does - its doing: initial_homing-=1000; // Decrease by 1 for next move if needed stepperX.run(); // Start moving the stepper
06:21 AM miss0r: Theres a bad error there, I would say
06:21 AM diverdude: miss0r: ok, which one is that
06:21 AM miss0r: The while loop, that does the homing oyu speak of there
06:22 AM miss0r: it calls stepperx.run(); each cycle, but in the end of that loop, theres a delay of 500ms, which means, the homing cannot move faster than two steps per second
06:22 AM miss0r: I would try to decrease that delay to perhaps 5-10ms
06:23 AM jthornton: is this on a raspberri pi or something?
06:23 AM miss0r: jthornton: Looks like arduino to me
06:23 AM diverdude: its arduino uno
06:24 AM jthornton: ah yea that's what I meant I just woke up lol
06:24 AM miss0r: :]
06:24 AM Loetmichel: gloops: already at it... got my first coffee (latte macciato made by the wife) for today, sitting in front of 6 bread rolls layerst with cheese and cold meats... there is a reason i weight 19 and a half stone ;)
06:24 AM diverdude: miss0r: originally that was just delay(5); and it also did not work
06:25 AM Loetmichel: s/layerst with/covered in
06:25 AM MarcelineVQ: is a stone 20lbs?
06:25 AM miss0r: diverdude: Well, you can be very sure it does not work with 500 ;)
06:25 AM miss0r: well, it would move. but be ridiculously slow
06:25 AM miss0r: to the point you'd be unaware that it is even moving
06:25 AM diverdude: miss0r: yes that i also found out :)
06:26 AM Loetmichel: MarcelineVQ: i weight 124kg. a stone is 14lbs
06:26 AM MarcelineVQ: Er no that'd be too high wouldn't it, you work too hard for that number to work out
06:26 AM MarcelineVQ: gosh you weight more than I do, you must be tall
06:27 AM diverdude: miss0r: ok i tried completely removing the delay and then i can see that its actually moving
06:27 AM diverdude: miss0r: but its just moving very slowly
06:28 AM miss0r: is your homing switch activated at this point? (during your test now) ?
06:29 AM diverdude: miss0r: yes
06:30 AM miss0r: okay :) then move to the next loop - there you also have a delay, remove that as well
06:30 AM diverdude: miss0r: yeah already did that
06:30 AM diverdude: but still very slow
06:30 AM miss0r: Darn it. that would've made sense
06:30 AM Loetmichel: MarcelineVQ: 180cm
06:31 AM Loetmichel: <. fat old bastard ;)
06:32 AM miss0r: diverdude: comment out all serial communication
06:32 AM miss0r: and run it again (that could very well be slowing it down alot)
06:32 AM diverdude: miss0r: already done....does not help
06:32 AM miss0r: meh. Send me an updated version of your code, please
06:33 AM MarcelineVQ: Loetmichel: well, you're not short, that is a little higher weight-wide for that height though hehe
06:33 AM MarcelineVQ: *weight-wise
06:34 AM Loetmichel: MarcelineVQ: i SHOULD be around 90kg according to my doctor ;)
06:35 AM miss0r: ladies ladies... I'm ~120kg 187cm :)
06:35 AM MarcelineVQ: that sounds about right, though if you've got some muscle that could be low
06:36 AM miss0r: It mostly is, but I have around 5-6kg of stomach fat I would like to loose at some point
06:36 AM MarcelineVQ: Just get really buff, then if you're also fat you can blame the weight on your muscle density.
06:36 AM miss0r: but I just like burgers and soda too much, I guess :)
06:36 AM * jthornton should be 12.5 stone but has been slacking and is 13 stone
06:37 AM MarcelineVQ: all y'all talkin' in stones and cm, what a crowd
06:38 AM XXCoder: $wa 13 stone to pounds
06:38 AM XXCoder: %wa 13 stone to pounds
06:38 AM MarcelineVQ: "I only weight one stone, it's that one over there holding up the bridge."
06:38 AM diverdude: miss0r: the only thing which seems to make it run faster is to increase the stepsize....but that only makes it run so far...its still quite slow
06:38 AM andypugh: MarcelineVQ: Yes, very odd. cm are right out. I think in stones and mm
06:38 AM diverdude: miss0r: moment i will hsow you updated code
06:39 AM diverdude: miss0r: this is updated code: https://paste.ubuntu.com/p/2NjgwZkTj2/
06:39 AM miss0r: diverdude: So the speed does not change *at all* when you increase/decrease the MaxSpeed or MaxAcceleration?
06:40 AM diverdude: miss0r: motor is at least moving now but its still quite slow and i cannot see how to increase speed
06:42 AM miss0r: so editing the MaxSpeed or acceleration does not change anything?
06:42 AM diverdude: miss0r: well no and it makes sense why also since its moving many small steps so i guess it will accellerate and deaccell always
06:42 AM miss0r: Increase the length it has to move then
06:42 AM miss0r: If that does not change it either, I would have to say theres something slowing down your cycle time
06:44 AM diverdude: miss0r: no it does not help really
06:44 AM miss0r: hmm
06:44 AM * miss0r is studying the text
06:44 AM diverdude: miss0r: maybe accelstepper is just bad for making homing procedure?
06:44 AM miss0r: how 'slow' it is, would you guess, in RPM?
06:45 AM miss0r: Could be. Write your own, if in doubt
06:45 AM diverdude: miss0r: oh thats a difficult question.
06:45 AM miss0r: ballpark answer
06:46 AM diverdude: i would say its moving maybe 1/3 cm per second
06:46 AM diverdude: in RPM i have no idea
06:47 AM diverdude: i dont even know how to see that
06:48 AM miss0r: hmm
06:48 AM miss0r: hang on, reading like a mad man here
06:48 AM miss0r: It seems you still have the serial in working order?
06:49 AM miss0r: never mind. its not 'inside' the loop
06:49 AM miss0r: so it shouldn't matter
06:52 AM miss0r: hmm
06:52 AM miss0r: Can you move the axis a good way away from the home switch and run it again?
06:53 AM diverdude: miss0r: yeah for sure
06:55 AM miss0r: Have you tried this with a good long run before? (After removing the delays in the loops)?
06:56 AM diverdude: yeah
06:56 AM diverdude: its moving at constant speed
06:56 AM miss0r: and it is creeping up on the homing switch very slowly?
06:57 AM diverdude: yes
06:57 AM miss0r: alright. and with the working example you had it running quite fast, with the same electrical setup?
06:58 AM miss0r: i.e. no change to microstepping on the driver?
06:58 AM diverdude: yeah much faster
06:58 AM diverdude: no change on driver
07:01 AM miss0r: https://pastebin.com/gpjMzqaW
07:01 AM miss0r: try that, and let me know how it does
07:05 AM diverdude: miss0r: its moving a bit faster now...but its stopping before reaching the switch
07:05 AM miss0r: yeah. now it is only moving the amount of steps defined in the home_step
07:05 AM miss0r: faster, but not as fast as the working code?
07:06 AM diverdude: miss0r: i would say same speed as working code actually
07:06 AM miss0r: yeah
07:06 AM miss0r: Good. I think it was just using some stupid math in there, that made the loop take too long to cycle.
07:07 AM diverdude: miss0r: right...but i think its not homing now...its just moving to an absolute position
07:07 AM miss0r: I would make two constants: home_stepin and home_stepout
07:08 AM miss0r: One for each direction. And make sure the home_stepin(the one replacing home_step) is large enough to reach the home switch from anywhere on the travel
07:08 AM miss0r: thereby eliminating the math in the loop
07:08 AM miss0r: and the home_stepout should only be a distance large enough for it to release the homing switch
07:08 AM diverdude: miss0r: but is it possible to forexample ask accelstepper to run 10000000 pulses and then stop it when homeswitch is reached?
07:09 AM miss0r: diverdude: That is basically what you are doing right now
07:09 AM diverdude: miss0r: i thought that was the whole reason to move it small steps
07:09 AM diverdude: miss0r: yeah but right now its not even reaching the home switch. it stops before
07:09 AM miss0r: the loop that calls stepperX.run() exits when the homeswitch is acticated. afterwards you set the distance it needs to go to zero
07:10 AM miss0r: basically you just need to have a number large enough
07:10 AM diverdude: miss0r: ok you are actually right
07:11 AM miss0r: what a chocker :P
07:11 AM diverdude: miss0r: very interesting :D
07:11 AM diverdude: haha nono i did not mean it that way
07:11 AM miss0r: Yeah :)
07:12 AM miss0r: Would you mind humouring me with a picture of your setup?
07:12 AM diverdude: I just was surprised why this article then jumps so many hoops for a stupid slow scheme when i can just move the damn motor in a single move
07:13 AM miss0r: They are trying to make an example of using a constant speed method, which is flawed, in my opinion
07:14 AM Loetmichel: miss0r: i always use G64P0.05 on my programs. i can live with the 0.05mm deviation and it makes the run so much smoother, especially if you have a lot of corners/small arcs in the code
07:14 AM diverdude: miss0r: so now comes next question....how can i make the motor move faster....its now running at same speed as bespoke working code....but would be awesome if it was working faster
07:15 AM miss0r: Increase the maxSpeed
07:17 AM miss0r: just take note of one thing: when doing a movement like this.. or rather stopping it like you do - sending it no more pulses when the homing switch is reached, you do not have a ramp down. it stops abruptly. It might not be wise to move too fast
07:17 AM diverdude: miss0r: hmm no it does not really increase speed
07:17 AM miss0r: test with your other program, if you can increase the speed with that
07:17 AM miss0r: if you can, it still comes down to the cycle time of the arduino
07:17 AM diverdude: miss0r: yeah ok...but for homing i think its ok
07:17 AM miss0r: Indeed
07:18 AM diverdude: miss0r: yeah i am actually testing with other program already
07:18 AM diverdude: but it also does not really increase speed
07:18 AM miss0r: ha... :)
07:18 AM diverdude: what i remember though is that i once used grbl on this machine and it was able to move it faster than this
07:19 AM miss0r: https://www.airspayce.com/mikem/arduino/AccelStepper/classAccelStepper.html#ae79c49ad69d5ccc9da0ee691fa4ca235
07:20 AM miss0r: perhaps one should start looking at maxSpeed as just that. A sort of failsafe :)
07:20 AM miss0r: 2 secs, I will have a quick glanse at the last stepper controller I programmed
07:20 AM diverdude: ahh yes ofc
07:21 AM diverdude: setspeed would be the one to use
07:22 AM miss0r: not sure about that, though
07:22 AM miss0r: :)
07:22 AM miss0r: I did not use it. I can upload you some code if you like to take a look at it? sadly all the comments are in danish, so I don't know what use you could have looking at it
07:23 AM diverdude: miss0r: well then its quite lucky that i am fluent in danish ;)
07:23 AM miss0r: That helps ;)
07:24 AM miss0r: https://pastebin.com/R2KPAvf4
07:24 AM diverdude: anyway...setspeed did not seem to make much difference either hmmm
07:24 AM miss0r: yeah, I didn't think it would.
07:24 AM miss0r: How does that acceleration feel? does it do a nice ramp when starting?
07:25 AM diverdude: hmm
07:25 AM diverdude: actually not really
07:26 AM diverdude: why in denmark are you based?
07:26 AM diverdude: why=where
07:27 AM miss0r: A small city called �lstykke. north of copenhagen(ish)
07:27 AM diverdude: oh cool. im in copenhagen hehe :D
07:28 AM miss0r: Sweet. I did not know there were any "locals" in here :)
07:29 AM miss0r: On another node: I'm not sure you can completely compare grbl and arduino when it comes to speed. I'm not sure how the internals of the grbl run when it comes to cycle time. it might be able to do it faster
07:29 AM miss0r: Try reducing the ramp time, and see if that makes any difference
07:29 AM diverdude: but grbl was running on this very arduino
07:29 AM diverdude: its just code which i downloaded to arduino
07:30 AM miss0r: oh, so a grbl shield?
07:30 AM miss0r: not the actual grbl software on the arduino
07:31 AM miss0r: well, there you have it: the grbl shield does not use microstepping, if I recall correctly.
07:31 AM miss0r: Are you doing so with your current controller?
07:31 AM miss0r: controller=driver
07:31 AM diverdude: hmm yes i actually am doing microstepping now
07:32 AM diverdude: is that the reason for the slow speed you think?
07:32 AM diverdude: moment i will show you
07:32 AM Blumax: XXCoder and gloops, i have solution.
07:33 AM gloops: elaborate Blumax
07:33 AM miss0r: The arduinos cycle time can only produce pulses at a given maximum speed. If your maxspeed is set to something higher/faster than the cycle time of the arduino it sorta' just caps off doing it as fast as it can.
07:34 AM miss0r: When you were using the grbl, one step pulse was translated into one full step on the stepper. i.e. what we shall call 'full speed'. Now that you are running microstepping
07:34 AM miss0r: one pulse translates to 1/16(? or what microstepping setup you have) which would make the maximum output of the arduino move 1/16 the speed as before
07:35 AM miss0r: so, yeah
07:36 AM miss0r: if I were you, i'd reduce the maxSpeed to a point where you can see it moves slower(the steps can be performed within the cycle time of the arduino)
07:36 AM miss0r: Not doing so could cause issues later on
07:39 AM miss0r: diverdude: If I have answered what questions you have for now, I'll move to the mill to do some brackets for my new controller.
07:39 AM miss0r: :) You are welcome to write me here, but don't expect a fast answer in the forseable future
07:39 AM diverdude: ah ok sorry...i was trying to capture and image of the driver settings for you. Its in an inaccessible place hehe
07:39 AM Blumax: XXCoder, gloops :
07:39 AM Blumax: https://prntscr.com/kupe0b
07:40 AM Blumax: If i draw triangle and that I measure it is 3 sides, we can calculate all other
07:40 AM miss0r: diverdude: I'm building this at the moment: https://imgur.com/a/13lPAwY
07:40 AM Blumax: In the red rectangle, one has the value adjusted.
07:41 AM Blumax: Precise measurements must be made, but it should work well.
07:41 AM diverdude: miss0r: its here anyway: https://imgur.com/a/nN1332Y
07:41 AM Blumax: The end stops of my machine are positioned at the center of the guide rails. They are modeled on the drawing by the red dots.
07:42 AM diverdude: miss0r: cool!! what is that
07:42 AM miss0r: diverdude: Nice to see a fellow builder using wire ferrules (ledningstyller)
07:43 AM miss0r: diverdude: its the replacement controller for my maho 500c. Currently it is sporting a 1986 philips 432
07:43 AM diverdude: oh yeah...that makes it nice and tidy
07:44 AM gloops: yes geometry helps Blumax, however, we have to make exact physical measurements of these lines and angles, at this scale it is not easy with a laymans tools
07:46 AM diverdude: miss0r: so right now i have microstepping set to 12000. You are saying that to increase speed i should set it to maybe 200 instead?
07:46 AM Blumax: gloops, yes, with this method there is no need to measure angle, "just" length.
07:47 AM miss0r: diverdude: well 200 would land you the same place the grbl does
07:47 AM miss0r: But I assume you've choosen this to be able to microstep(make the motion less jerky). So perhaps you could settle for a slower speed, that is more smooth.
07:47 AM miss0r: That is something you have to try out to see what you're most happy about
07:48 AM Blumax: With a little rigor we should have a good result.
07:48 AM miss0r: bah.. Time got away from me. I'll smoke a cigarette, then I have to get back into the house'n make pancakes for the kid & wife
07:49 AM gloops: a^2 + b^2 = c^2
07:49 AM miss0r: diverdude: Where in copenhagen do you live, if you don't mind me asking?
07:50 AM diverdude: miss0r: well this is actually in my office. its in østerbro
07:50 AM miss0r: Great; what do you do for a living, that requires writing arduino software for steppers? :)
07:51 AM diverdude: miss0r: its kind of a hobby thing hehe. But I am building a machine for controlling a microscope
07:51 AM diverdude: so i can move in 3 dimensions over a large area
07:51 AM miss0r: Hehe, my shop started as a hobby as well :)
07:52 AM diverdude: what do you do in your shop?
07:52 AM miss0r: A little bit of everything, basically. Machine work, new machine builds/addons. But primarily I drive out and fix machinery, like mills'n lathes ect
07:53 AM miss0r: Alright. I realy need to get gone, or the wife will strangle me in my sleep :] I will be back later to see if can make some headway on the controller build
07:53 AM miss0r: talk to you later
07:54 AM diverdude: miss0r: ohhh cool. So you know a lot about stepper motors and drivers etc. and what is on the market and stuff?
07:54 AM miss0r: sure :) We can talk later.
07:54 AM * miss0r is out
07:58 AM Loetmichel: ah, btw: i am REALLY astonished how resilient those "single use" 1.5 PET coke bottles are... just bought a six-pack of them this morning... which broke just at my flats door... so the six bottles bounced down a half-flight of stairs and then out on the driveway... not ONE developed a leat. I refuse to open them until this evening though ;)
08:00 AM diverdude: miss0r: cool, let me know when you are back :)
08:20 AM andypugh: Loetmichel: Try putting 25mm of liquid nitrogen in an empty one then putting the lid on.
08:21 AM andypugh: Stand well back.
08:21 AM andypugh: You typically can’t find _anything_ left.
08:25 AM sync: yes
08:25 AM sync: dry ice also works
08:25 AM Loetmichel: hihi. a bit of water and a few chunks of dry ice works just as well
08:26 AM Loetmichel: indeed
08:30 AM Loetmichel: i was just surprised that they survive half a dozen bounces on exposed aggregate concrete under full pressure and at full weight
10:24 AM Jin|away is now known as Jin^eLD
10:24 AM Jin^eLD: hi
12:03 PM Rab: Loetmichel, interestingly, those bottles shrink if you hit them with a heat gun: https://hackaday.com/2017/03/23/soda-bottles-used-as-heat-shrink-for-wood-joinery/
12:03 PM Rab: (Although that project tends more toward hideous art than beautiful practicality IMO)
12:03 PM diverdude: I have a 3-phase stepper motor which runs at 300 steps per revolution and a driver which pulses 200 steps per revolution. I made arduino program like this: https://paste.ubuntu.com/p/XmZmnbGsH7/ which runs the motor 300 steps at 300 steps/sec at very high accelleration. In my understanding this should make the motor drive 1 second in 1 direction and then 1 second in the opposite direction. For some reason it runs 2 seconds in each
12:04 PM Spida: Loetmichel: oh, those "single use" ones have the advantage the plastic is fairly fresh. try this with the multi use ones....
12:05 PM Spida: Loetmichel: I dropped one from like 20cm on concrete. the bottom gave way, and the bottle did that rocket-start thing....
12:06 PM Spida: (which ended quite spectacular, because the concrete roof of the underground parking garage was just 2m away - bottle billard)
12:07 PM miss0r: diverdude: I'm back :)
12:07 PM miss0r: looking at your question now
12:08 PM miss0r: diverdude: Theres something wrong with your math alltogether
12:09 PM diverdude: miss0r: ohh i see
12:09 PM diverdude: maybe i confused some things
12:09 PM miss0r: diverdude: What exactly do you see? :)
12:09 PM miss0r: Are you quite positive your stepper does not have 200 steps/rev? i.e. the driver is doing a 1:1
12:09 PM diverdude: i see the machine is taking 2 seconds to move in each direction
12:10 PM diverdude: miss0r: well my driver has 200 steps/rev and my motor has 300 steps/rev
12:10 PM diverdude: miss0r: which makes it diffucult for me to understand
12:10 PM miss0r: is it a geared stepper?
12:10 PM diverdude: yes i believe so
12:10 PM miss0r: do you have a part number?
12:11 PM diverdude: i do
12:11 PM miss0r: you see, with steppers the most common ones are 200steps/rev. There are some 400steps/rev. but with these smaller ones 200 is the standard ish
12:11 PM miss0r: let me see :)
12:11 PM diverdude: this is my driver: https://www.lightobject.com/3MD560-Leadshine-Compatible-3-Phase-84A-1-axis-Stepping-Motor-Driver-P836.aspx and this is my motor: https://www.lightobject.com/-NEMA-23-3-phase-573S09-35A-127in-oz-Stepper-motor-P914.aspx
12:12 PM diverdude: miss0r: yeah i can adjust this driver to different settings...now its at 200 steps/rev
12:13 PM miss0r: Indeed. You've found one of the odd ones :D
12:13 PM diverdude: yeah... unfortunately....if you know of a better motor/driver combination than this i am all ears
12:13 PM diverdude: :)
12:14 PM miss0r: I know of others, I'm not sure about better :)
12:14 PM diverdude: miss0r: ack...i have to go for 30-60 minutes....will you be here in about 1 hour?
12:15 PM miss0r: Sure.. I'm about to fire up the cnc and make the parts I didn't manage to get done earlier. Just write me, and I will answer as soon as I see it
12:15 PM diverdude: ahh awesome :) thnx man
12:16 PM Rab: miss0r, did you mill or punch the dsub cutouts in that enclosure you posted yesterday?
12:16 PM miss0r: rab: I milled those
12:17 PM miss0r: unfortunatly I don't own a punch'n die set for d-subs
12:17 PM miss0r: and I didn't feel like making three different ones for this small project ;)
12:17 PM Rab: It's a big financial outlay.
12:18 PM Rab: What was your source for the dsub dimensions?
12:18 PM miss0r: rab: This quite excellent one_ https://www.norcomp.net/storage/wysiwyg/panel-cutout_3.jpg
12:18 PM Rab: miss0r, sweet, thanks! I think that's what I've used in the past.
12:19 PM miss0r: and you lost it? :)
12:19 PM Rab: Just wondering if there's a canonical source. I usually hit up datasheets from a big player like Norcomp or Amphenol.
12:20 PM miss0r: hehe. That there was basically one of the first three hits on google image search, when searching for "d-sub cutout template"
12:26 PM miss0r: rab: About the financial aspect of it; This is a project I'm making for myself.. so time is not realy a factor in this. I just couldn't be bothered :)
12:27 PM miss0r: rab: also, should you be curious; this is what it looks like bend into shape & mounted in the control panel: https://imgur.com/a/13lPAwY
12:29 PM Rab: miss0r, looks great! Nice latching dsubs.
12:29 PM miss0r: rab: not so nice, actualy. I'm having issues with it
12:29 PM Jin^eLD: miss0r: what thickness did you chose?
12:29 PM miss0r: Jin^eLD: its 1mm plate
12:30 PM miss0r: rab: the latches I bought are designed to be mounted directly on the d-sub, so it does not pull the connector close enough. I'm searching for a solution
12:31 PM Loetmichel: Rab: we always do them symetrically
12:31 PM Loetmichel: sou you can fit the dSub in both orientations
12:31 PM miss0r: and the only reason I'm *stuck* with the latches, is because the controller I'm replacing is using them.
12:31 PM Loetmichel: http://www.cyrom.org/palbum/main.php?g2_itemId=15002&g2_imageViewsIndex=1
12:31 PM Loetmichel: like that
12:32 PM miss0r: Loetmichel: Do you have any experience with the latch locks?
12:32 PM Loetmichel: miss0r: like on the old thicknet connectors?
12:32 PM Loetmichel: nope
12:32 PM Loetmichel: i dont use them because they dont work well
12:32 PM miss0r: Me neither, I like the screw better. But as I am doing a plug'n play replacement of what is on my mill currently, I sorta' need to use them
12:33 PM Loetmichel: cant you remove them from the controller?
12:33 PM Loetmichel: and replace them with screws/bolts?
12:33 PM miss0r: sadly, no. They are factory made in one piece with the d-sub
12:34 PM miss0r: like riveted
12:34 PM * miss0r did think of that
12:34 PM Loetmichel: rivets can be drilled open...
12:34 PM Loetmichel: but i see your predicament
12:34 PM miss0r: sure, but then I would have no way of attaching anything new. The idea is, that I can reinstall the old controller within a few hours, should the new one fail
12:35 PM Loetmichel: i am sorry to say: you are fubared...
12:35 PM miss0r: The problem is, I'm using this machine in my daily production, so I can't realy be too intrusive to it
12:35 PM miss0r: I had that feeling :]
12:35 PM Loetmichel: those clips didnt work well in the 70s and 80s... they dont work any better now ;)
12:36 PM miss0r: I mean, other than making a craptacular solution, where I add a male->female block bolted to the d-subs in my plate, and have the latchs on the back side of those. But that just seems exsessive
12:36 PM miss0r: bulky.. and another point of faliure ext
12:36 PM miss0r: ect*
12:37 PM Rab: Panel thickness can mess up any retention mechanism if the connectors are rear-mounted. I have that problem with threaded jack nuts. Are the cutouts big enough to front-mount the connectors?
12:38 PM miss0r: rab: sadly no. And even if they were, I would have issues with the two 50 pole d-subs, as they are with a screw terminal breakout board on the back, that I would not be able to get through the hole
12:38 PM Rab: oof
12:39 PM miss0r: I just want a solution that does not look like crap & is relieable
12:39 PM miss0r: Short of double sided tape, I'm out'a ideas :]
12:41 PM miss0r: gawd, it will be sweet with the new controller.. I'm transfering a 52kb g-code to the mill at the moment. I'm nearing completion, and I have been "uploading" for 16minuts
12:41 PM Loetmichel: miss0r: hrhr
12:41 PM Loetmichel: i once did a 35MB gcode
12:41 PM miss0r: Thats what you get for emulating a punch tape terminal, through a 2400baud serial connection
12:41 PM miss0r: Nice :D
12:42 PM miss0r: I only have 64k of memory on this one, anyway :D
12:42 PM Loetmichel: https://www.youtube.com/watch?v=cEG7-6zwpt4 <- took quite a while... about 30 hours IIRC. (caution loud)
12:43 PM miss0r: Yeah, those 3D parts are a bitch :D That is something I will be able to do once the new controller is in place.
12:43 PM miss0r: You can imagine, that is very limited with 64k memory
12:43 PM * miss0r off to start the mill
12:47 PM fragalot: hi
12:48 PM miss0r: 'ello
12:49 PM miss0r: fragalot: The new controller is comming together: https://imgur.com/a/13lPAwY
12:51 PM fragalot: cardboard enclosure?
12:51 PM fragalot: I expected better from you
12:51 PM miss0r: I just don't want to make a 'noise box'
12:51 PM fragalot: xD
12:52 PM miss0r: you know - with all the electrical noice bouncing around inside making everything worse :D
12:52 PM fragalot: I like the obviously salvaged igbt
12:52 PM miss0r: hehe...
12:52 PM miss0r: IGBT is actualy a transistor.. This was just the 15volt unit to drive them
12:52 PM gloops: yes miss0r is right imo
12:52 PM gloops: re - noise box
12:53 PM fragalot: gloops: yea.. that's not how electricity works
12:53 PM miss0r: gloops: that was actualy a remark, mocking what you said earlier ;)
12:53 PM gloops: i know haha
12:53 PM miss0r: hehe
12:54 PM fragalot: sorted through the schaublin accessoires today
12:54 PM fragalot: it came with a LOT of arbors
12:54 PM miss0r: you didn't happen to find a tapping head in there as well? :)
12:54 PM miss0r: how many are we talking?
12:54 PM fragalot: you say that
12:54 PM fragalot: and you'd be right
12:55 PM gloops: i fried a whole pack of bacon today
12:55 PM gloops: throughout the day
12:55 PM fragalot: we're talking an entire drawer (400mm wide) full
12:55 PM fragalot: and another drawer of spacers
12:55 PM miss0r: fragalot: Pictures or....
12:55 PM fragalot: I think the previous owner was a tool maker
12:55 PM fragalot: or at least very creative with making specialty cutters
12:56 PM miss0r: Nice. Its good to have bunches of that stuff :D
12:56 PM fragalot: miss0r: didn't take pics - forgot to take my phone with me and cba to go back now :P
12:56 PM fragalot: will do tomorrow
12:56 PM fragalot: tried the high speed head too
12:56 PM fragalot: that thing has ZERO play in it
12:57 PM * miss0r drules
12:57 PM fragalot: backlash, I mean
12:57 PM gloops: how much play in it?
12:57 PM fragalot: gloops: absolutely nothing :P
12:57 PM fragalot: well, barely enough for a thin oil film on the gear, but no more
12:58 PM gloops: like my router then
12:58 PM fragalot: i'm not sure if there are any gears in those routers
12:59 PM gloops: not in the spindle, no
01:00 PM miss0r: that is pretty sweet. What speed was it able to do?
01:00 PM fragalot: miss0r: don't have a tachometer
01:01 PM miss0r: fragalot: In about 30 mins I think, a guy called diverdude(apparently a fellow dain) will be back here asking questions about grbl shield on arduino. He is having some issues. I think your insight will do him good
01:01 PM fragalot: you mean my superior documentation reading skills?
01:02 PM miss0r: I helped him earlier to the best of my knowledge, but theres just something off about driving a three phase 300steps/rev with a driver that says 200 steps/rev in the lowest microstepping setting... Something does not add up
01:02 PM miss0r: Sure.. call it that ;)
01:03 PM miss0r: (19:10:42)<diverdude>this is my driver: https://www.lightobject.com/3MD560-Leadshine-Compatible-3-Phase-84A-1-axis-Stepping-Motor-Driver-P836.aspx and this is my motor: https://www.lightobject.com/-NEMA-23-3-phase-573S09-35A-127in-oz-Stepper-motor-P914.aspx
01:04 PM fragalot: there's always going to be something off with that motor and driver combination
01:04 PM fragalot: thankfully the driver itself doesn't give a damn what the motor will actually do
01:05 PM miss0r: Indeed.
01:05 PM fragalot: it probably assumes 1:1 step is 200 steps/rev
01:05 PM miss0r: yeah. What I was thinking exactly
01:05 PM fragalot: and 400 steps/rev is 1:2
01:05 PM miss0r: He just can't use the readout for other than assuming 200 is 1:1, and so forth
01:05 PM fragalot: yea
01:05 PM fragalot: easy. :)
01:05 PM miss0r: yep
01:05 PM miss0r: Always nice to have a second opinion.
01:06 PM miss0r: So, wensday, I will call GLS and claim insurance
01:06 PM fragalot: alright
01:06 PM miss0r: I'm pretty sure you can start looking for another tapping head.. Sorry man
01:06 PM fragalot: at least I have the collets & nut
01:07 PM miss0r: Always good to have :]
01:07 PM fragalot: I'll just make my own tapping head, with blackjack;. and hookers.
01:07 PM miss0r: Indeed :D
01:07 PM miss0r: you know, my closest friends and I use that same refrence pretty damn often.. we are sad people
01:08 PM fragalot: does your wife know?
01:08 PM miss0r: But I'm sure they will be more than ready to pay up, as I have been calling them daily since I started :D
01:08 PM miss0r: sure.. she "looks to the sky" every time
01:08 PM fragalot: different department
01:08 PM fragalot: you've been calling customer service
01:08 PM fragalot: those people exist to make you feel they care
01:08 PM fragalot: without them actually having to do so
01:09 PM miss0r: sure. Insurance is another matter entirely. I know. Lets just hope for the best
01:11 PM miss0r: fragalot: Do you have experience with the d-sub latch locks?
01:12 PM fragalot: nope
01:12 PM miss0r: great :D
01:40 PM Tecan: https://imagebin.ca/v/4FoXaDrA37DD distributed computing
01:54 PM fragalot: miss0r: https://www.borenexpress.nl/a-nieuw-geplaatst/1269 <=== maybe an alternative to the ER32 hex blocks? :P
01:55 PM miss0r: yes. that IS sexy...
01:55 PM miss0r: but with my luck ordering / shipping stuff lately, I do not dare
01:55 PM fragalot: xD
01:56 PM fragalot: he doesn't ship that online
01:56 PM fragalot: pick-up only
01:56 PM miss0r: hehe
01:59 PM Jin^eLD: hmm, so, my buddy is now trying to compile the comp on his linuxcnc installation, so RT mode as a kernel module; I as developing it in userspace mode
01:59 PM Jin^eLD: and for him it all fails because of a different compiler
01:59 PM Jin^eLD: I was assuming that halcompile would enfornce the same flags everywhere?
01:59 PM Jin^eLD: is that not the case?
02:02 PM Jin^eLD: ok turns out he is on 2.7.11, not sure if that could have such an influence
02:02 PM fragalot: it might, depends on why it fails
02:02 PM Jin^eLD: from what I see it tries to use ISO C90 mode
02:02 PM Jin^eLD: while I was surely using C99 stuff during development, because it "just worked" here
02:08 PM Jin^eLD: does linuxcnc support "distro updates" via apt-get? so that he would not lose his setup/settings but just update linuxcnc?
02:10 PM Roguish: Jin^eLD: no distro updates. remember it's based on realtime kernel.
02:11 PM Jin^eLD: I thought linuxcnc rolls its own debian based distro?
02:12 PM Jin^eLD: so what would you suggest, whats the easiest way to get 2.7.14 without losing configuration? I think he starte with a 2.7.11 live cd and installed to disk from it (that was a while ago)
02:12 PM Roguish: yes it does that. sorry, probably misread your question.
02:13 PM Roguish: I run a master version, and when it gets a new build, it will update. without blowing the config.
02:14 PM Roguish: it's when, say you want to go from debain 8 to debian 9 that things don't work.
02:15 PM Jin^eLD: ok, thats what I meant... and is linuxcnc 2.7.11 to 2.7.14 like going from a distro to distro, or would a "regular" update do it? I am really not that familiar with the linuxcnc distro itself
02:16 PM Roguish: regular update should work fine, given the proper repositories are spec'd.
02:17 PM Jin^eLD: and any ideas about the kernel flags? should I rewrite the comp in C90?
02:17 PM Jin^eLD: or will 2.7.14 use C99?
02:18 PM Roguish: sorry, way above my pay grade.
02:18 PM Jin^eLD: I should probably ask in -devel
02:18 PM Jin^eLD: thank you for the update hints
02:18 PM Roguish: yes
02:18 PM Roguish: no problem.
02:18 PM Jin^eLD: I'll tell him to try a normal update first and lets see how far we get
02:19 PM Roguish: sounds like a good call.
02:27 PM Jin^eLD: did not work, he ran apt-get update, but he's still on 2.7.11
02:28 PM CaptHindsight: Jin^eLD: what kernel flags are being used?
02:29 PM CaptHindsight: my question has little to do with the upgrade not upgrading
02:30 PM Jin^eLD: CaptHindsight: I do not see the exact flags in his paste, but I see messages like:
02:30 PM Jin^eLD: warning: ISO C90 forbids mixed declarations and code
02:30 PM Jin^eLD: or
02:30 PM Jin^eLD: error: �
02:30 PM Jin^eLD: crap
02:30 PM Jin^eLD: loop initial declarations are only allowed in C99 mode
02:30 PM Jin^eLD: stuff like that, so it seems to use C90 mode
02:31 PM andypugh: Jin^eLD: Might be easiest to update via Synaptic
02:32 PM Jin^eLD: i am currently looking at http://linuxcnc.org/docs/2.7/pdf/LinuxCNC_Getting_Started.pdf chapter 4.1
02:32 PM andypugh: If the repository is set up right (and an ISO install will be) then it’s automatic.
02:32 PM CaptHindsight: what is being compiled? maybe I need to read more of the backlog
02:32 PM Jin^eLD: I am not a debian/ubuntu guy, do you have a link for me or do I just google for synaptic update?
02:33 PM Jin^eLD: CaptHindsight: a C .comp via halcompile, it worked fine in userspace mode where I was developing it using halcompile from 2.7.14
02:33 PM Jin^eLD: but now my buddy is trying to build it on the actual PC that is connected to his mill, and turned out he has an older linuxcnc version there and the comp won't compile due to the above compiler errors
02:34 PM diverdude: miss0r: hi, are you still here?
02:34 PM andypugh: Can you halcompile a very simple (and pre-existing) comp?
02:35 PM CaptHindsight: Jin^eLD: if apt-get update didn't work then he needs to setup his repos
02:35 PM andypugh: halcompile should just compile for the system it is running on.
02:35 PM andypugh: 2.7.11 to 2.7.14 should have no effect at all on halcompile. It’s a set of minor bugfixes
02:35 PM Jin^eLD: andypugh: his halcompile works as such, and the compiler errors indicate that I was simply using too advanced C :) so I think it works "for his system", which we now found out is Debian Wheezy
02:36 PM andypugh: You have to be careful what you use in ,comp files. You should only really use basic C, no libraries.
02:37 PM Jin^eLD: andypugh: I am not using any libraries or whatever, its really stupid stuff like variable declarations in for loops and thingsd like that, C90 vs C99
02:37 PM andypugh: (and you can automatically use the rtapi_ math (etc) libraries.
02:38 PM Jin^eLD: andypugh: I would expect that halcompile uses the same restrictions in userspace mode, i.e. if it can only do C90 for the kernel, then I'd expect it to use C90 in userspace as well
02:38 PM Jin^eLD: is that not the case?
02:38 PM Jin^eLD: or is it a 2.7.11 vs 2.7.14 thing? thats what I am trying to find out...
02:38 PM Jin^eLD: if its the same on 2.7.14 then I should rewrite my comp
02:38 PM Jin^eLD: userpace did not give me any warnings whatsoever
02:40 PM Jin^eLD: would anyone who has 2.7.14 in rt mode be willing to quickly try and compile my comp and tell me if it works? that'd be a great help in determining how to approach the issue
02:41 PM andypugh: Nothing sugnificant as far as infrastructure changed between 2.7.11 and 2.7.14
02:41 PM Jin^eLD: so both would still be on wheezy?
02:42 PM andypugh: Both can be on a large range of OS-es.
02:42 PM andypugh: There is no link between any version of LinuxCNC and a specific OS.
02:42 PM andypugh: (years ago someone even managed to run it on MacOS)
02:42 PM Jin^eLD: well, he started with a linuxcnc live cd download, dunno how its being packaged
02:43 PM Jin^eLD: I thought that linuxcnc kind of maintains a distro spin based on debian/ubuntu
02:43 PM Jin^eLD: but I may be worng here
02:43 PM andypugh: The liveCD installs a Linux version, but the package can run on others. (and that version can be updated without changing the OS)
02:44 PM Jin^eLD: that I understood, the most trickery why the live cd is helpful is all the RT stuff
02:44 PM andypugh: Yes, you get Linux + RT kernel + LinuxCNC all working for little effort.
02:44 PM Jin^eLD: and of course I wrote the comp on my regular notebook, just compiled myself 2.7.14 linuxcnc from git and ran it in userspace mode
02:45 PM andypugh: But you can choose Linux distribution pretty freely as long as you can make on of the RT kernels work on it.
02:45 PM andypugh: (that’s RTAI, PREEMP-RT or Xenomai)
02:45 PM Jin^eLD: that'd be too advanced for him - not an IT vampire :)
02:45 PM andypugh: Where is your comp?
02:46 PM Jin^eLD: https://github.com/jin-eld/mh400e-linuxcnc
02:47 PM diverdude: I have a 3-phase stepper motor which runs at 300 steps per revolution and a driver which pulses 200 steps per revolution. I made arduino program like this: https://paste.ubuntu.com/p/XmZmnbGsH7/ which runs the motor 300 steps at 300 steps/sec at very high accelleration. In my understanding this should make the motor drive 1 second in 1 direction and then 1 second in the opposite direction. For some reason it runs 2 seconds in each
02:47 PM Jin^eLD: if you have the environment set up and halcompile is available, then just "make gearbox" will do it
02:48 PM fragalot: diverdude: the driver likely assumes you have a 200 step motor, and will act as 1:1 step for your 300 step motor (eg, 300 pulses will be 1 revolution)
02:49 PM fragalot: diverdude: as for your understanding of the time: not really. it's not a linear acceleration
02:50 PM diverdude: fragalot: ok i see. What is then a good way to figure out what the maximum speed of my system is?
02:50 PM fragalot: you keep increasing the speed (with a load on it) until it stops being reliable
02:50 PM fragalot: then back down a bit
02:51 PM fragalot: there are calculators out there that should get you close but I can't remember the link
02:53 PM andypugh: Jin^eLD: That’s a heck of comp. I haven’t seen one with separate headers before.
02:54 PM diverdude: fragalot: with a load on it?
02:55 PM Jin^eLD: andypugh: that's "fake" because I #include the sources too in the end, but I still wanted to have it separate for a better overview, didnt want to dump everything in one file (I let the preprocessor do that)
02:57 PM andypugh: Anyway, it isn’t wanting to compile on my RTAI LinuxCNC 2.8.0 system.
02:57 PM Jin^eLD: andypugh: same kind of C90 / C99 errors?
02:58 PM andypugh: https://pastebin.ubuntu.com/p/6mCdk8vyfv/
03:00 PM Jin^eLD: k, C90 as well, thank you!
03:01 PM Jin^eLD: btw he managed to update using synaptic, but says the splash screen still shows 2.7.11
03:17 PM diverdude: fragalot: so if i set my driver to 200 it will really be 300, if i set it to 400 it will raelly be 600, if i set it to 10000 its really 15000...the driver does that mapping automatically. correct?
03:20 PM Jin^eLD: ok, hacked up the linuxcnc makefiles to enforce -std=c90, I think I should be able to fix my comp in a userspace env without a "real" system
03:20 PM Jin^eLD: andypugh: thanks once more
03:25 PM andypugh: Jin^eLD: The splash screen should show 2.7.14 if he really did update.
03:25 PM andypugh: What does Synaptic itself think the version is?
03:25 PM XXCoder: yo andy
03:25 PM andypugh: Hai
03:25 PM XXCoder: how goes yiur spindlke project
03:25 PM andypugh: Which one?
03:26 PM XXCoder: one you had video of
03:26 PM andypugh: Finished
03:26 PM XXCoder: cool :)
03:26 PM andypugh: Have you seen the video? (do autocaptions work at all)
03:26 PM Jin^eLD: andypugh: he reran it and rebooted again and seems now it worked, with 2.7.14 in the splash, so seems
03:26 PM Jin^eLD: OK
03:27 PM XXCoder: autocraptons does work but cuts out sometimes and once a while weird results when high inference
03:27 PM XXCoder: otherwise no issues I can see. you speak clearly apparently
03:37 PM andypugh: It’s doing very well.
03:38 PM andypugh: When I did my first video in 2010 it was utterly, laughably, hopeless.
03:38 PM andypugh: I don’t think that they had used any Yorkshire acents to train it with.
03:38 PM XXCoder: nice
03:39 PM andypugh: It made absolutely no sense at all. I think “ER32 collet” came out as “wrinkled popes” for example.
03:39 PM XXCoder: heh thats why I call it autocraptons
03:39 PM XXCoder: its definitely a lot better, but not always
03:40 PM XXCoder: its still iffy on specialized words but good
03:58 PM miss0r: XXCoder: Getting a little closer to completion: https://imgur.com/a/t7OcojW
03:58 PM XXCoder: that looks good. :)
03:58 PM miss0r: It seems I have the most of the day tomorrow only for completing this. I think by evening tomorrow, I will be programming this thing. or atleast realy close
03:58 PM XXCoder: I really need to work on mine
03:59 PM miss0r: wait. don't tell me its nice yet. Just wait till I start pitchforking wires in there :D
04:00 PM Deejay: gn8
04:00 PM miss0r: nite nite
04:00 PM XXCoder: metal spaghetti
04:00 PM miss0r: Yeah :D
04:01 PM miss0r: You need to be extra carefull to get all the chips out of the cabinet, now that theres a motherboard in there as well
04:01 PM XXCoder: getting those very puffy filters for air vents?
04:01 PM XXCoder: dunno what its called, we use lot of em at all machines at work
04:02 PM miss0r: Theres those filters in both the intake and the output vents on this one
04:02 PM XXCoder: nice
04:02 PM XXCoder: I have no idea where to buy em
04:02 PM miss0r: I don't know that they have a specific name. I just buy some complete units from schneider electric.
04:02 PM miss0r: I can get you the EAN number, if you are interrested?
04:03 PM XXCoder: while I dont think I will need for while sure I can find a link with number so I can look up later
04:03 PM miss0r: 7512404045
04:04 PM miss0r: just google that, you'll find a seller near you
04:04 PM XXCoder: interesting thats little different
04:04 PM miss0r: I used to use them all of the time on electrical cabinets when I build those
04:05 PM XXCoder: one at ork comes in large roll sheet maintance guy cuts em into rectangles
04:05 PM miss0r: sure. I just buy the ready made ones that fit in these. its probally 100x as expensive.. but at a few $ a piece, who cares when you only need two :D
04:06 PM miss0r: I'll call it a night. See you around
04:06 PM * miss0r is out
04:06 PM XXCoder: night
04:08 PM _unreal_: Got my safty switches today :)
04:09 PM _unreal_: There so cute :)
04:09 PM XXCoder: salty switches? ;)
04:09 PM _unreal_: little smaller then I expected but oh well :)
04:43 PM Jin^eLD: nite
04:44 PM Jin^eLD is now known as Jin|away
04:44 PM diverdude is now known as Guest8301
04:45 PM diverdude: How can I see what the Voltage (V) of this stepper motor is? https://www.lightobject.com/-NEMA-23-3-phase-573S09-35A-127in-oz-Stepper-motor-P914.aspx
04:46 PM XXCoder: wow that is awesome video https://twitter.com/weatherchannel/status/1040305773664980993/video/1
05:03 PM andypugh: diverdude: You don’t need to know the voltage
05:04 PM _unreal_: Looking for a key switch like this that can handle 20amp at 115VAC http://img.weiku.com/waterpicture/2011/10/24/0/electrical_key_switch_634590921947212130_1.jpg
05:06 PM andypugh: _unreal_: You could have a light-duty switch operate a relay
05:06 PM _unreal_: I'm aware
05:07 PM andypugh: They do exist, at least:
05:07 PM andypugh: https://uk.rs-online.com/web/c/switches/key-switches-accessories/key-switches/?redirect-relevancy-data=636F3D3126696E3D4931384E53656172636847656E65726963266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E2E2A2426706F3D31333326736E3D592673723D52656469726563742673743D43415443485F414C4C5F44454641554C542673633D592677633D4E4F4E45267573743D6B657920737769746368267374613D6B65792073776974636826&r=f&applied-dimension
05:07 PM andypugh: 94569406
05:08 PM andypugh: They make 100A battery isolatros, but they are only rated 24V
05:10 PM andypugh: Shorter URL, 440VAC 20A. Not cheap: https://uk.rs-online.com/web/p/key-switches/1245262/
05:11 PM andypugh: And here is how to pick one ;-_ https://www.youtube.com/watch?v=ZBk9w4bc_b4
05:12 PM _unreal_: ya thats a bit much
05:17 PM andypugh: I would asssume that you can build what you want from modular industrial switches?
05:18 PM andypugh: eBay to the rescue. 20A 660V https://www.ebay.com/itm/20A-Panel-Mount-3-Position-Universal-Rotary-Cam-Changeover-Switch-With-Key/112921794396?hash=item1a4aa9d75c:g:6KIAAOSwgd1ayNCX
05:21 PM _unreal_: to large :/
05:21 PM _unreal_: wont fit
05:25 PM andypugh: Relay then :-)
05:29 PM andypugh: diverdude: Steppers are all OK with 100V or so. It’s what the drivers can handle that matters. (and you want to run as high a voltage as you can while still having a little margin)
05:31 PM andypugh: So I was given a brand-new Moog G400 servo yeterday. 29Nm 3kW. What should I use it for :-)
05:53 PM * jthornton gets back to how to cleanly sum each float in a tuple and subtract that from a third tuple without using a bunch of for loops in python of course
08:13 PM CaptHindsight: andypugh: precision potato masher, get creative with the servo
08:14 PM XXCoder: nahh
08:14 PM XXCoder: make supser powerful blender
08:14 PM XXCoder: make your machine can blend "can you blend this?" blender
08:16 PM andypugh: It’s a 4000 rpm motor, that might work.
08:17 PM XXCoder: if you make lots sawdust you can make pellet maker with that
08:17 PM CaptHindsight: high shear mixer blender
08:37 PM skunkworks: https://www.youtube.com/watch?v=mudZ739aWnQ
08:40 PM XXCoder: thats fast
08:41 PM CaptHindsight: \0/
08:45 PM skunkworks: 2 of the 3 steppers had the thrust bearings in backwards..
08:47 PM XXCoder: how do you find that out
08:47 PM norias: ask them
08:52 PM skunkworks: took them apart after they never seemed to tighten up without binding
08:54 PM Tom_L: skunkworks is that stepper or servo?
08:54 PM Tom_L: 1:1 ratio?
09:08 PM skunkworks: Tom_L: stepper 1:1
09:09 PM Tom_L: that's pretty quick
09:09 PM Tom_L: what screw ratio?
09:09 PM Tom_L: pitch
09:21 PM skunkworks: uhh
09:30 PM skunkworks: Tom_L: 4mm
09:30 PM skunkworks: *5mm
09:30 PM Tom_L: that's pretty quick
09:31 PM Tom_L: makes me think i should work on mine some as far as tuning
09:40 PM skunkworks: don't know if it will be run that fast...
09:40 PM Tom_L: yeah but it's fun do demo that way
09:41 PM Tom_L: i turned mine down a bit to 'practical' but i don't have the drives current up all the way either
09:42 PM Tom_L: i can't cut anything that fast anyway but it would help on the rapids
10:07 PM _unreal_: finally getting my wires cut and crimped
11:16 PM ziper: _unreal_, is there any reason I couldnt use a wood router on a FRP mold that had fiberglass sticking up around the edges?
11:16 PM ziper: so that the top surface is totally flat?
11:19 PM XXCoder: man this is crazy!! https://www.youtube.com/watch?v=OP30okjpCko
11:19 PM XXCoder: pumpler cant stop water
11:19 PM XXCoder: why didnt he cut water off
11:23 PM XXCoder: read the discription. no wonder
11:24 PM XXCoder: owner is way too chea[
11:53 PM _unreal_: ?
11:53 PM _unreal_: define wood router?
11:53 PM _unreal_: are ou talking about a hand held router? or a CNC machine router
11:53 PM _unreal_: as to standard ROUTER bits. what kind of a router bit are you talking about?
11:55 PM _unreal_: As long as the router BIT is a CARBIDE tipped it will work ok BUT I assure you. it will DULL the router bit and you will likely get smoke when cutting wood. I use standard router bits on fiberglass all the time. they DULL reall fast but they DO keep cutting. just remember DONT force the router. it will break the bit when cutting fiberglass.
11:56 PM _unreal_: I would also strongly suggest having dust collection and a fan/exhaust system to get any FRP smoke fumes out.
11:57 PM _unreal_: lool