#linuxcnc Logs

May 17 2019

#linuxcnc Calendar

01:49 AM Loetmichel: mornin'
01:56 AM * Loetmichel ist just "shielding" cat5 cables in our devices. Systems failed the CE-measurement. :-( Now there is probably more copper around that cat5 than inside... and yes, it gets bolted to the aluminium enclosure in 2 places, so no need to connect it to the plugs) ... -> http://www.cyrom.org/palbum/main.php?g2_itemId=17382&g2_imageViewsIndex=1
01:58 AM Deejay: moin
02:44 AM pink_vampire: Loetmichel: lust take cat7 or cat6A
02:55 AM Loetmichel: pink_vampire: its a rework of an existing system
02:55 AM pink_vampire: i see
02:55 AM Loetmichel: and the cat5utp was uses because of the space contraints inside the enclosure.
02:55 AM Loetmichel: the shielded plugs are simply to long
02:56 AM Loetmichel: http://www.cyrom.org/palbum/main.php?g2_itemId=17207&g2_imageViewsIndex=1 <- second picture from top, center front to left side
02:57 AM pink_vampire: i see the cables
02:58 AM pink_vampire: now days cat6A is like cat5, very slim and can flex
04:08 AM XXCoder: boo
04:11 AM pink_vampire: baaa
04:11 AM Loetmichel: pink_vampire: emphasis on UTP... S-FTP hasl longer plugs, wouldnt be able to fit
04:11 AM Loetmichel: and cat6/7 isnt aviable in UTP to my knowledge
04:15 AM pink_vampire: the shielding is what make the cable about to work as cat 6 / 7
04:15 AM Loetmichel: indeed
04:16 AM pink_vampire: UTP is just a cable without shield, by wrapping it in copper tape, you made it STP
04:59 AM Tom_L: morning
05:31 AM jthornton: morning
05:32 AM XXCoder: morning
05:33 AM Loetmichel: pink_vampire: not copper tape, copper braid tube. but correct.
05:34 AM Loetmichel: wasnt planned that way but that jobn is botched anyways, those PCs are back from the customer for the 10th time or so because "customer requested upgrades" and Failures.
05:35 AM Loetmichel: we had cpus die on them like fruitflies, then problems with the external CE-lab, then customer request to change the HDD cages to something that cant be ejected "accidentally" and so on...
05:36 AM Loetmichel: and now the failed CE test.
05:36 AM Loetmichel: its like murphy had a field day on those systems
05:44 AM Loetmichel: *meh* ... just went to the pharmacy and home to get some drugs for my wife. came back to the company: "nope, we already ordered pizza for lunch, you werent there so you get nothing... $me is eating some "delivery gummy bears" and some military ration cookies to not faint from low blood sugar... nice coworkers :-(
05:46 AM pink_vampire: omg..
05:48 AM pink_vampire: i need to get key chuck for 2 chucks, and arbor for one of them
05:49 AM jthornton: my coworkers poop all over the place and lay eggs
05:50 AM pink_vampire: jacobs chuck no 34
05:51 AM pink_vampire: jacobs chuck no 2a
05:51 AM pink_vampire: the No34 need an also an arbor
05:52 AM pink_vampire: how do i find what sizes of keys they need?
05:52 AM XXCoder: do they have model #s?
05:53 AM pink_vampire: yes
05:53 AM pink_vampire: No34
05:53 AM pink_vampire: No2A
05:58 AM XXCoder: any luck if you search for say "n034 key"
05:58 AM pink_vampire: nope
05:59 AM pink_vampire: what size it is?
05:59 AM XXCoder: https://www.mig-welding.co.uk/forum/threads/key-size-for-34-jacobs-drill-chuck.73126/
06:00 AM XXCoder: k3 or k3c
06:00 AM pink_vampire: mm
06:00 AM pink_vampire: lets see
09:41 AM pink_vampire: ih
09:41 AM pink_vampire: hi
10:06 AM jym: pink_vampire: lo
10:07 AM pink_vampire: i want to get lathe tools
10:14 AM mozmck is now known as mozmck_m
10:14 AM pink_vampire: but i don't want to get too many inserts
10:14 AM pink_vampire: what is the most versatile insert?
10:23 AM mozmck_m is now known as mozmck
12:45 PM serp: Just using plain axis now. I would like to add a button that moves the head around the xy bounding box of the loaded gcode. Is this a solved problem or any hints about what to start reading up on here.
12:45 PM JT-Shop: machine extents or program extents?
12:54 PM mozmck: JT-Shop, Tom_L: I just finished merging and pushing time.comp to both 2.7 and master. I merged my changes with the ones from itaib and I think it's better yet.
12:56 PM mozmck: For an existing config, simply connect time.N.pause to halui.program.is-paused, for new configs you don't need the NOT component - just connect time.N.start to halui.program.is-running
12:56 PM JT-Shop: cool
01:06 PM serp: JT-Shop: program extents
01:07 PM serp: using the term extents i found this post which describes exactly what i want it for https://forum.linuxcnc.org/10-advanced-configuration/29382-move-on-extents?start=10
01:09 PM Tom_L: mozmck, thanks
01:29 PM andypugh: serp: Does that work for you?
01:32 PM R2E4: hello all
01:32 PM R2E4: JT you around?
01:37 PM serp: andypugh: it was enough of a push that we're going down the rabbit hole now
01:38 PM serp: we can tie the extents to pins and then use gladevcp i think to add a button to traverse those values
01:39 PM jthornton: for a bit
01:39 PM jthornton: I'm in and out
01:41 PM andypugh: serp: I was looking to see if the python interface had program extents in it, but it seems not.
01:42 PM andypugh: The graphical preview knows the answers, but doesn’t export them AFAIK.
01:43 PM serp: yeah... do you feel like the way we're talking will work?
01:43 PM andypugh: Yes.
01:43 PM serp: ok thank you
01:54 PM andypugh: serp: Are you any good at Python?
01:54 PM serp: I have not done anything with linuxcnc specifically but i use python daily at work
02:06 PM andypugh: OK, then perhaps you can figure out how to use the gcode.calc_extents() function?
02:07 PM andypugh: It is pretty easy to call a Python routine from a button press, and then that can send MDI commands to teace your bounding box.
02:07 PM andypugh: This is as far as I got:
02:07 PM andypugh: python
02:07 PM andypugh: import gcode
02:08 PM andypugh: >>> gcode.calc_extents()
02:08 PM andypugh: ([9e+99, 9e+99, 9e+99], [-9e+99, -9e+99, -9e+99], [9e+99, 9e+99, 9e+99], [-9e+99, -9e+99, -9e+99])
02:08 PM andypugh: There is a gcode.parse that might be needed. But I can’t figure out what arguments it takes
02:12 PM andypugh: serp: As a Python programmer you might find that this page opens a shole new world of possibilities: http://linuxcnc.org/docs/devel/html/config/python-interface.html#_reading_linuxcnc_status
02:18 PM serp: andypugh: is that gcode module part of linuxcnc?
02:18 PM andypugh: Yes
02:19 PM serp: I don't see it documented there
02:22 PM andypugh: No, I only just founf it looking through gremlin / axis / glcanon
02:23 PM andypugh: So I suspect it isn’t documented, and you need to figure it out with the Python tools (dir, inspector….)
02:23 PM andypugh: I haven’t even found it in the source code.
02:25 PM enleth: do you know of a simple design for an anti-restart relay circuit that prevents the load from being powered when the "start" button is pressed as power is applied?
02:25 PM serp: andypugh: thanks.. i will see what i can find out
02:25 PM enleth: so that it has to be released and pressed again
02:25 PM enleth: normal anti-restart is simple, but this gets a bit more complex
02:26 PM enleth: oh, and there might be more than one start button in parallel
02:27 PM enleth: the best I can come up with is a latch relay activated by a circuit running in series through the NC contacts of all start buttons
02:28 PM andypugh: Arduino :-)
02:29 PM enleth: andypugh: if I wanted the machinery this is for to kill someone, I can think of many simpler ways
02:29 PM mozmck: andypugh: is it the python interface you haven't found in code?
02:30 PM andypugh: mozmck: No, the g-code parser
02:30 PM mozmck: Oh, ok
02:30 PM andypugh: “import gcode” works…
02:30 PM mozmck: I see.
02:36 PM mozmck: andypugh: src/emc/rs274ngc/gcodemodule.cc
02:39 PM andypugh: Ah, yes. So looking for “def parse” was never going to work, it’s in that baffling PyObject syntax:
02:39 PM andypugh: https://github.com/LinuxCNC/linuxcnc/blob/master/src/emc/rs274ngc/gcodemodule.cc#L1032
02:40 PM FinboySlick: andypugh: Rediscovered your youtube channel a couple days ago. Fun content, you're good at it.
02:41 PM andypugh: I tried to make a new video last night, but got back inside to find that the camera was in timelapse mode :-)
02:41 PM andypugh: And couldn’t be bothered making the widget again just for the camera
02:42 PM andypugh: But can you guess what it is for?
02:42 PM andypugh: https://photos.app.goo.gl/5MBvKxk5ioThiUHu5
02:42 PM FinboySlick: Well, if it's a couple per second timelapse, you could use a Stephen Hawking voice for effect :)
02:42 PM FinboySlick: (couple frame per second)
02:42 PM andypugh: I deleted them all, so it’s too late
02:43 PM R2E4: jthornton: You in or ouot right now....
02:56 PM JT-Shop: trying to get my instrument panel working again on the bike lol but here at the moment
03:54 PM Deejay: gn8
03:59 PM Tom_L: mmm 4.4 quake just now
04:25 PM skunkworks: 4.4 quake? I didn't know you got them around there
04:29 PM Tom_L: just now had another one
04:29 PM Tom_L: due to oil fracking in Ok
04:29 PM mozmck_m: Where is that?
04:29 PM Tom_L: ks ok
04:29 PM Tom_L: most start in Ok
04:29 PM mozmck_m: Huh
04:30 PM Tom_L: https://earthquake.usgs.gov/earthquakes/map/
04:30 PM Tom_L: last one hasn't been recorded yet but i'm sure i felt it
04:31 PM Tom_L: 4 small ones yesterday apparently
04:31 PM gambakufu: mozmck_m, dunno if your getting notified about the time.comp. your merge commit isn't working as it should - pause/resume is resetting the counter. your should use your 2.7 code.
04:31 PM gambakufu: you're
04:36 PM mozmck_m: gambakufu: odd, I'm out right now but will look at it shortly
05:10 PM gambakufu: ah, I see the issue - the 2.7 'running' variable is missing from the master branch version.
05:19 PM mozmck: gambakufu: are you itaib as well? oops did I miss something?
05:21 PM gambakufu: I am.
05:22 PM mozmck: boy, I did mess up that merge in master.
05:31 PM mozmck: gambakufu: I pushed a fix - not sure how I messed the previous one up
05:35 PM gambakufu: cool
05:42 PM jym: Concrete slab for $1/sf https://www.youtube.com/watch?v=40uCoIgxPC0
05:53 PM SpeedEvil is now known as Guest37555
07:31 PM BitEvil is now known as SpeedEvil
07:31 PM jthornton: R2E4_: check your email
08:24 PM gregcnc: CRAP! Did losing work offsets by estopping get fixed?
08:25 PM Tom_shop: check your var file
08:25 PM jthornton: e stop should not change G5x offsets
08:25 PM gregcnc: it does and we've talked about it here
08:25 PM gregcnc: just killed a part
08:25 PM jthornton: I missed that, in master or 2.7?
08:26 PM gregcnc: that's already late, that I don't have material for.
08:26 PM mozmck: first I've heard of it - I'm sure I've estopped and not lost G5x offsets plenty...
08:26 PM gregcnc: i'm done for the night
08:26 PM mozmck: gregcnc: is there a bug report for that?
08:27 PM jthornton: yea me too, I've it estop a bunch of times but I've also had some strange things happen from time to time
08:27 PM Tom_shop: those should be written back to the physical .var file once set i'd think
08:28 PM gregcnc: it's broken tools before, but this time I'm hosed
08:28 PM mozmck: Yeah, I've had a time or two now when it seems that linuxcnc missed some settings in the INI file. Re-starting linuxcnc fixed it.
08:28 PM jthornton: I think sometimes there is a delay in writing to the var file
08:30 PM jthornton: gregcnc: a bug report is the best way to document this
08:31 PM Tom_shop: if you set the offsets at the beginning there shouldn't be
08:34 PM jthornton: can't even try and duplicate it without knowing the details...
08:34 PM gregcnc: https://github.com/LinuxCNC/linuxcnc/issues/126
08:35 PM gregcnc: I guess that's not it
08:38 PM jthornton: are you using something other than G54 coordinate system?
08:39 PM gregcnc: no. started machine, set offsets for g54. started program, estopped program, set Z for g54, scrapped part
08:41 PM jthornton: you should file a bug report with all the details like GUI what e stop is used and what version of linuxcnc your using
08:41 PM jthornton: in 2.8 axis sim I'm unable to duplicate
08:42 PM jthornton: chickens are all tucked away so time to retire for the night
08:42 PM jthornton: say goodnight gracie
09:19 PM Tom_L: andypugh do you know much about pouring aluminum molds??
09:21 PM andypugh: Not _specifically_
09:21 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/Can_Holder/A3/Mold2.jpg
09:21 PM Tom_L: test part for a steel mold i'd like to try
09:22 PM andypugh: I have made patterns for a lot of cast iron parts. And I have made moulds, and cast parts, in pewter. But I have done neither for aluminium
09:22 PM Tom_L: wonder if i need vent holes in the bottom
09:22 PM Tom_L: gettin ready to run it in steel
09:24 PM andypugh: I would have thought that a small groove from the bottom to the fill surace would be a good idea. If only to indicate when the mould was full.
09:24 PM Tom_L: hmm
09:25 PM andypugh: Do you understand what I mean? \
09:25 PM Tom_L: this is a test part, if it works there will be multiple parts in the final mold which i will probably add that in
09:25 PM Tom_L: yes
09:25 PM Tom_L: like a J from the bottom to the top
09:25 PM andypugh: You almost never try to make the air come out against the metal floe direction.
09:25 PM andypugh: yes, exactly
09:25 PM Tom_L: makes sense
09:26 PM Tom_L: or even a trace out the side at an upwards angle
09:27 PM Tom_L: i suppose it should go to the top though or it will empty the mold down to the exit level
09:27 PM andypugh: The only reason not to is if you really care about de-sprue time after demoulding.
09:27 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/Can_Holder/A3/Mold1.jpg
09:28 PM Tom_L: it's easy for me to machine them but this guy doesn't have a good mill so he wanted to mold the part instead
09:29 PM roycroft: i designed my belt grinder stand to have very little clearance with the swiveling part of the grinder
09:29 PM roycroft: i temporarily mounted the grinder on the stand tonight to see if it would work
09:29 PM andypugh: That part is really on the cusp between makes-sense-to-cast and makes-sense-to-machine
09:29 PM roycroft: i could not swing it to the horizontal position
09:29 PM Tom_L: yeah
09:29 PM roycroft: then i saw that i had forgotten to grind down one of the weld beads
09:30 PM roycroft: now it works fine :)
09:30 PM Tom_L: the edge is supposed to have a radius but all i had was a chamfer tool and i didn't feel like ball nosing the whole edge for that
10:16 PM pink_vampire: hi
10:16 PM pink_vampire: it was a looong day today
10:20 PM pink_vampire: Tom_L: XXCoder roycroft??
10:20 PM pink_vampire: https://i.imgur.com/snxoG7M.png those need some glue
10:58 PM jym: pink_vampire: deep fry and spicy peanut sauce