#robotics Logs

Dec 21 2020

12:47 AM sorki is now known as srk
04:02 AM rue_mohr: http://ruemohr.org/~ircjunk/distractions/FSMservo/slide.htm
04:18 AM Jak_o_Shadows: For the fun of it?
12:13 PM rue_bed: yep
12:14 PM rue_bed: it worked
07:15 PM orlock: Goddamn
07:15 PM orlock: Circuit misbehaving. Remove it from housing. Behaves again.
07:16 PM orlock: Goddamn.. COouncil contracted plumber is out so suck out the water from the drain
07:17 PM orlock: dude.. it's raining...
07:18 PM ATSystems: Metal housing? Capacitance weirdness? External lighting triggering cheap chips with poor packaging? Backscatter stupid from the council worker?
07:20 PM orlock: The PCB's attached to a plastic carrier, but it is driving a pair of steppers that bolt to the same metal frame
07:20 PM orlock: The specific issue is that the steppers don't reliably change direction
07:23 PM ATSystems: And the PCB is totally electrically isolated from the box? If you put it in but don't screw it down does your driver behave itself?
07:23 PM orlock: it's a 3 axis Arduino Nano CNC board - takes 3x stepper drivres and an Nano
07:24 PM orlock: PCB is isolated except for the motors both attach to the same common metal housing
07:25 PM orlock: https://www.ebay.com.au/i/293031308297 - one of these
07:26 PM orlock: and i put it inside a Skywatcher HEQ5 mount
07:26 PM ATSystems: Neat.
07:27 PM ATSystems: I still wonder if it's ambient light voodoo, plenty of stories about folks seeing their boards reset under camera flash. Can you wrap the board in some fabric or something to rule that out?
07:28 PM ATSystems: wait, I have it back to front, sorry
07:28 PM ATSystems: in case bad, out of case good
07:28 PM ATSystems: Might need to wait for someone smarter than me to pop their heads up
07:30 PM orlock: No flashes in this case, but i remember the issue you are talking about
07:33 PM ATSystems: Ja, hopefully not a thing anymore. But again, had it ass backwards anyway.
07:33 PM ATSystems: I don't know how you'd iso test for capacitance problems inside the enclosure short of wrapping the whole shield in alfoil, but thats more of an RF interference thing.
07:34 PM orlock: At least i figured out that there is a problem - the motors not reversing - rather than trying to debug the system as a whole
07:34 PM orlock: because that leads to trying to code AND debug hardware issues outside in the middle of the night
07:36 PM orlock: my printer electronics are giving me the shits too, grr
07:36 PM ATSystems: Yop. Go snowblind reading code then realise you have something tied high when it needs to be low, been there, did not buy the shirt
07:37 PM orlock: yeah - i ended up going back to the vendor-provided dead-simple "hello world" code for the board - it just rotates X Y and Z motors CW and then CCW
07:38 PM orlock: (my code has a serial parser and crap in there too)
07:39 PM ATSystems: Yeah, I got nothing my dude, I'm a metalworker with too many CAD files and not enough robots to show for it, still getting my head around uC development.
07:39 PM orlock: Nice :)
07:39 PM ATSystems: I can strip and rebuild a gearbox with my eye tied behind my back, but code makes my brain hurt.
07:39 PM ATSystems: The beating will continue etc.
07:40 PM orlock: http://www.astro-baby.com/heq5-rebuild/heq5-m1.htm
07:43 PM ATSystems: ooher, your site?
07:47 PM orlock: No, but that's the mechanics of what the controller is driving
07:47 PM orlock: mine is https://somepicturesof.space
08:10 PM ATSystems: Thats some nice photography though.
08:17 PM orlock: Annoying, i just ran the test sketch with the motors housing touching, and that does seem to cause the odd behaviour
09:30 PM rue_bed: orlock, hows yours doing?
09:30 PM rue_bed: I didn't set up the stereo image yet
09:30 PM rue_bed: I think I have the images saved for it