#linuxcnc Logs

Sep 29 2021

#linuxcnc Calendar

01:51 AM Deejay: moin
05:01 AM JT-Cave: morning
07:00 AM JT-Cave: first Italian copy lathe conversion done... 3 more to go and I need to finish the second spoke lathe feeder
07:01 AM JT-Cave: and I have to help build an assembly machine... I should be retired lol
07:02 AM CloudEvil: :)
07:04 AM JT-Cave: I'm taking 1/2 a day off today because it's going to rain the next few days
11:14 AM roguish[m]: JT-Cave: sounds like you should work an extra long day.......
02:39 PM roycroft: i have a question about limit switches and linuxcnc
02:39 PM roycroft: two things are clear to me:
02:39 PM roycroft: 1. i should separate the physical limit switches from the home switches
02:40 PM roycroft: and 2. the physical limit switches can be wired in series, and it does not matter which axis exceeds the limit
02:40 PM roycroft: the thing i'm not clear about is if it's ok to wire the home limit switches in series
02:41 PM roycroft: i think i can - i think that i set the home limits one axis at a time, and once set, linuxcnc maintains soft limits for homing
02:41 PM roycroft: in which case all axes can monitor the same circuit
02:42 PM roycroft: is that correct?
02:43 PM roycroft: and if linuxcnc does home by using the limit switches, does it do so one axis at a time, or does it try to home all axes simultaneously?
02:43 PM roycroft: i'm trying to conserve i/o pins here, of course
03:27 PM rs[m]: roycroft: i would wire "positive" and "negative" physical limit switches separately, else you can not drive free one a limit trips (the idea is the "positive" limit only blocks movement in "positive" direction and vice versa)
03:27 PM rs[m]: s/one/once/
03:32 PM roycroft: well the physical limit switches should never, under any circumstances, be reached
03:33 PM roycroft: and if they are they'll triggre an estop
03:33 PM roycroft: so i'm not sure how that would apply there
03:33 PM roycroft: and the homing switches will be one per axis
03:33 PM roycroft: that's what i'm asking about
03:35 PM roycroft: if i reach a physical limit switch i'm not sure how to handle that once power is restored
03:35 PM roycroft: i figured i would have to somehow manually move the axis where it was reached, once i diagnose and fix the problem that caused it to be reached in the first place
03:36 PM roycroft: or am i fundamentally missing something here?
03:37 PM roycroft: my understanding is that i have "physical" or "extreme" limit switches on both extremes of each axis, and i should never travel to them
03:37 PM roycroft: then i'll have soft limit switches inside the physical limit switches, and that is as far as i should ever travel
03:38 PM roycroft: then i would have home limit switches that i would use for homing the machine
03:38 PM XXCoder: honestly you can use homing as that for one end of all axis
03:38 PM XXCoder: because homing switches should never be touched after homing
03:38 PM roycroft: those could be either in the same place as one of the soft limits for each axis, or could be inside the soft limits on the axes
03:39 PM XXCoder: you should set position where homing is triggered as -0.1" or mm whatever and set it to have soft limit at 0
03:39 PM XXCoder: so if it hits homing switch again, something went wrong
03:39 PM roycroft: so from left to right, on the x axis, for example, i could have hard limit-soft limit/home limit-work area-soft limit-hard limit
03:41 PM roycroft: that may be an option, xxcoder
03:41 PM roycroft: but my original question has still not been answered
03:41 PM roycroft: can the home limit switches be wired in series?
03:41 PM XXCoder: homing not sure, but other end limits sure
03:42 PM roycroft: and this should be obvious, but if it's not, i'll wire all my limit switches as normally open
03:42 PM roycroft: yes, i know that the other end limits can be in series
03:42 PM XXCoder: might be possible by only homing one axis a time, though if something hits both axis switches at once it might be an issue
03:42 PM XXCoder: but then its already issue should shouldnt hit em at all anyway
03:42 PM roycroft: and if linuxcnc homes one axis at a time when pressing the home button, then the home limit switches may be able to be wired in series
03:43 PM roycroft: if they can't, that's fine
03:43 PM roycroft: i just need to know
03:43 PM roycroft: if they can i want to conserve the pins
03:43 PM XXCoder: yeah. 2 vs 4
03:43 PM roycroft: well, this machine will have 3 axes and 4 joints
03:43 PM roycroft: so 2 vs 5, i think
03:44 PM roycroft: and one of the pins is common, and doesn't really count
03:44 PM roycroft: so 1 vs 4
03:44 PM roycroft: i can use the same common pins for all the 24v circuits
03:44 PM roycroft: common pin
03:45 PM roycroft: and i misspoke above - i'll wire them normally closed
03:45 PM roycroft: i don't know why i said no when i meant nc
03:46 PM XXCoder: lol well nc definitely better idea
03:46 PM XXCoder: you want it to fail triggered not untriggered *smash*
03:46 PM roycroft: if there's a loose wire then it will fail immediately
03:46 PM roycroft: if it's nc
03:46 PM XXCoder: yep
03:46 PM roycroft: if it's no it won't fail ever
03:50 PM roycroft: reading the homing page again, i'm thinking i can wire the homing switches in series
03:50 PM roycroft: it suggests that the soft limits should be inside the homing switches
03:52 PM roycroft: and that when i want to go to the machine origin i should use g53, which goes to the soft limits
03:53 PM roycroft: that means i only need to set the home limits once, or once per machine setup
03:53 PM roycroft: and i do that on a per-axis basis
03:53 PM XXCoder: yeah home setup is only done once per power on
03:53 PM roycroft: so as long as i can configure linuxcnc to have all axes monitor the same pin for the home limit switch, i should be fine
03:54 PM XXCoder: hopefully though test this :)
03:54 PM roycroft: of course
03:54 PM roycroft: my mesa board should arrive tomorrow or friday
03:54 PM roycroft: i can start testing the motion control system as soon as i get it
03:54 PM roycroft: and i have not started the machine build at all
03:55 PM roycroft: my intent is to bench test the whole system before i attach anything to the machine
03:56 PM roycroft: i'm still curious about what rs[m] was saying about separating the positive and negative hard limits
03:56 PM roycroft: i don't see how that would be useful, but he would not have mentioned it if it were not useful, i think
03:57 PM XXCoder: dont know about that either
03:57 PM roycroft: if i hit a hard limit i'm going to automatically kill power to the machine
03:57 PM roycroft: and i'll know at a glance what limit i hit
03:58 PM roycroft: i may not know why
03:58 PM roycroft: and i hope i never have to figure out why
03:59 PM rs[m]: roycroft: on my servo drives, physical limit switches only stop the drive from moving further in that direction, but allow movement off the limit switch
03:59 PM roycroft: oh
03:59 PM rs[m]: so you have a chance to move the machine off the limit, would be hard to do by hand
04:01 PM rs[m]: and it is not wired to trigger e-stop. linuxcnc will error out with following error if you try to move beyond limit switch because drive will not go there
04:01 PM roycroft: so i would want to set the machine up to not trigger estop immediately when a hard limit is reached at all times
04:01 PM roycroft: hmm
04:01 PM rs[m]: i guess it depends on the machine what you want/need to do
04:02 PM roycroft: i'm not sure i want to set it up to not auto estop if it reaches a hard limit switch
04:02 PM roycroft: i don't want to have to trust linuxcnc to stop it for me
04:02 PM rs[m]: if its open loop stepper motor i would estop
04:02 PM roycroft: it is
04:03 PM rs[m]: on a machine with position feedback you just need to inhibit motion of servo drive
04:07 PM roycroft: i was planning on having a latching contactor control the 24v and 48b buses, and have any estop buttons and the limit switches wired in series through a pin on the mesa board to control the latching part of the circuit
04:07 PM roycroft: so hitting the mushroom button or triggering any of the limit switches would kill 24v and 48v power
04:08 PM roycroft: the spindle vfd would also be fed by that latching contactor
04:08 PM roycroft: so everything that moves would be killed
04:17 PM rs[m]: disconnecting DC bus to motor drives would be unhealthy
04:19 PM XXCoder: wouldnt you want it to lock rather than just depower
04:19 PM XXCoder: if i recall right, for steppers, you could just feed power to both pairs so stepper just locks up and refuse to move
04:19 PM XXCoder: dunno if theres something like that for servo
04:19 PM Tom_L: i just use the enables
04:20 PM rs[m]: if your stepper driver have enable inputs, use that. disconnecting the dc bus while motor is in motion riscs destroying your stepper drives
04:21 PM Tom_L: enables are there for a purpose
04:22 PM roycroft: they are gecko drives, with a disable input
04:22 PM Tom_L: same thing
04:22 PM roycroft: yes, inversely
04:28 PM rs[m]: in principle disable is the wrong way round, usually you want your drive to be safe in case a cable gets cut or comes loose
04:28 PM roycroft: i'll have to study this some more
04:29 PM roycroft: but right now i have to fix a website that my boss just broke, and all he will tell me is that he "did nothing wrong"
04:29 PM Tom_L: they never do
04:29 PM Tom_L: just ask em
04:30 PM roycroft: after 30 years of working for this guy, i still cannot fathom why everything comes down to blame with him
04:30 PM roycroft: i was not blaming him for anything
04:30 PM roycroft: i just need to know what he was doing so that i know how to troubleshoot
04:30 PM Tom_L: maybe he felt guilty
04:30 PM roycroft: but when i asked 'what were you doing' he responded "i didn't do anything wrong"
04:31 PM roycroft: then he has been feeling guilty for 30 years, because that is how he approaches everything
04:31 PM Tom_L: sign of ignorance
04:31 PM CloudEvil: I'm betting this started age 5 or so.
04:31 PM roycroft: when we first set up a trouble ticket system, many many years ago, i got it all set up and working fine, with a form to document the issue
04:32 PM roycroft: the very first thing he did when i put it online was to update the form
04:32 PM roycroft: right after the customer name he added a checkbox field: "Blame"
04:32 PM roycroft: with the values "customer" and "employee"
04:33 PM CloudEvil: wow
04:33 PM roycroft: and he made it a mandatory field
04:33 PM roycroft: he really looks at everything that way
04:35 PM XXCoder: geez
04:36 PM roycroft: this relates to the bullshit that happened at the beginning of the pandemic
04:36 PM roycroft: i found that folks on the right could not understand me at all
04:37 PM roycroft: they were all going around blaming china, and when i would respond "who cares right now how the pandemic started - what we need to do know is figure out how to stop it. we can worry about how it started later"
04:37 PM roycroft: they would totally freak out
04:38 PM roycroft: all they cared about was who was to blame for it
04:38 PM roycroft: and that i did not care about that AT THE MOMENT was incomprehensible to them
04:38 PM roycroft: i DO care about it
04:38 PM CloudEvil: The 'china caused a two week delay by not reporting ...' was amusing when the local government was delaying months.
04:38 PM roycroft: again, who cares?
04:39 PM roycroft: when a pandemic starts, the only thing to care about is how to stop it
04:39 PM roycroft: once it's under control, then it's time to figure how how it started, and perhaps how to prevent it from happening int he same way again
04:39 PM CloudEvil: Quite.
04:40 PM roycroft: that's not how the right see it at all
04:40 PM roycroft: all they see is that someone must be blamed
04:40 PM roycroft: maybe there's merit to the way they think
04:40 PM roycroft: but i'm incapable of comprehending it
04:41 PM _unreal_: There is ALWYAS merit.... the question is who owns the rational side
04:44 PM roycroft: please try to explain to me in a way that makes sense what benefit there is in focusing 100% of one's energy on finding someone to blame for a problem and 0% of one's energy fixing the problem
04:45 PM CloudEvil: The sole case where it's productive pretty much is if they are legally liable for the cleanup.
04:45 PM roycroft: and let me repeat yet again, lest i be thought to not care about this: after the crisis itself has been addressed, it is time to look into the cause of the crisis and act/learn from it
04:46 PM roycroft: cloudevil: even that has no benefit if whatever needs to be cleaned up is a current threat to life and property
04:46 PM roycroft: it just needs to be cleaned up NOW
04:46 PM roycroft: figuring out who gets the bill can come later
04:46 PM CloudEvil: If it's urgent, yes.
04:46 PM roycroft: if it's not an immediate threat, then sure, it's fine to sort things out first
04:47 PM roycroft: well this pandemic was pretty urgent, i should think
04:47 PM _unreal_: I have not read though what you guys are rambling about yet... what is the argument?
04:47 PM roycroft: i'm talking about my boss' mindset where whenever there is a problem all he cares about is blaming someone for it, and not fixing it
04:48 PM _unreal_: :| lol THATS NORMAL
04:48 PM roycroft: i was relating that to the right wing reaction to the pandemic - all they cared about was how to blame china for it
04:48 PM roycroft: and did not care about stopping it
04:48 PM roycroft: it's not normal in my world
04:48 PM _unreal_: I've worked for very few businesses that didnt have a POS in power. that followed that mind set
04:48 PM roycroft: it's incomprehensible
04:49 PM roycroft: when encountered with a problem, my primary and immediate focus is on fixing the problem
04:49 PM roycroft: debriefing comes later
04:49 PM roycroft: imagine if the military worked that way
04:49 PM _unreal_: roycroft, that would be considered a six sigma approach
04:49 PM roycroft: "someone is shooting at us - should we shoot back?"
04:50 PM roycroft: "who is shooting at us? is it the chinese or the north koreans?"
04:50 PM roycroft: "..." (soldiers are dead)
04:53 PM roycroft: this all started when my boss reported that a website he was working on broke
04:53 PM roycroft: i asked him what he was doing when it broke, as that would be useful information in my troubleshooting
04:53 PM roycroft: and his response was "i didn't do anything wrong"
04:55 PM NoGodDamnIdea: guys can someone help me convert to metric #4 x .25 and #6 x .375
04:56 PM NoGodDamnIdea: these are screw dimensions
04:56 PM _unreal_: define convert
04:56 PM _unreal_: how close do you need
04:56 PM NoGodDamnIdea: something I can buy in europe
04:56 PM NoGodDamnIdea: #4 x .25 Flat Head Screw and #6 x .375 Thread Form Screw
04:57 PM _unreal_: ahhh... not in UK. speaking of how is your machine doing?
04:57 PM NoGodDamnIdea: the cnc? I had to take a break from extra stuff to finish my masters thesis
04:57 PM NoGodDamnIdea: nope, not uk
04:58 PM NoGodDamnIdea: I am buying from https://www.tme.eu/en/ if it helps
05:01 PM roycroft: a #4-40 thread is larger than an m2.5x45 thread, and smaller than an m3x0.5 thread, closer to the latter
05:02 PM roycroft: a #6-32 thread is larger than an m3x0.6 thread, and smaller than an m4x0.7 thread, closer to the former
05:02 PM roycroft: 0.25" is roughly 6mm
05:02 PM roycroft: 0.375" is roughly 9mm
05:06 PM NoGodDamnIdea: thanks roycroft
05:19 PM NoGodDamnIdea: ill just take the closest locally and make it work
05:19 PM NoGodDamnIdea: and _unreal_ i have not abandoned it, will start soon on it again
10:35 PM safarir: Hi everyone, I am looking into implementing a way to access the interpreter parameters from the UI process. I was wondering is anyone have any thought on the subject. One way of doing it would be to add a different message type to the emcStatus buffer (RCS_STAT_CHANNEL) but I am afraid that some python code seem to only expect one message type on
10:35 PM safarir: that channel. Any idea ?
10:45 PM roycroft: i've no insights/ideas that may help you directly, safarir, but i'll mention that there are more people active here during daytime hours in north america
10:46 PM roycroft: so be patient and perhaps wait a few hours if nobody speaks up now
10:46 PM safarir: Good to know
10:46 PM roycroft: and i should say daytime in the americas
10:46 PM roycroft: as i would not want to sleight central america or south america