#linuxcnc Logs
Aug 24 2025
#linuxcnc Calendar
01:37 AM Deejay: moin
05:51 AM Tom_L: morning
08:33 AM Unterhaus_ is now known as Unterhausen
10:51 AM roycroft: hmm, it looks like we might get a little rain in a few minutes
10:51 AM roycroft: that was not in the forecaset as of last night
10:51 AM roycroft: i won't complain though
10:51 AM roycroft: aqi is 111 right now, and rain would likely improve that
01:18 PM xxcoder: https://www.sciencedaily.com/releases/2025/08/250822073807.htm
01:18 PM xxcoder: wow pretty cool. they figured some of bee problems, and made special feeder algae
01:19 PM xxcoder: err I meant yeast but yeah
01:32 PM roycroft: i could not find 18ga wire yesterday - the stores that used to carry it no longer do
01:32 PM roycroft: but i think i have enough scraps to finish the 24v psu wiring
01:33 PM roycroft: i'm going to give it a go, and if it's not enough, i'll order the wire and get back to the project when i can
01:33 PM * roycroft hopes to get the wiring done today
01:33 PM roycroft: and i still have that cunning plan to get a 5v power distribution block squeeze in the cabinet
01:33 PM roycroft: i should work on that first, as it may take a while, and may affect the other wiring
01:34 PM xxcoder: too bad "inside is bigger" is not a thing at all
01:34 PM roycroft: if only had gone to hogwarts instead of engineering school ...
01:34 PM roycroft: i had, rather
01:34 PM xxcoder: no way. too bad lime lords isnt available
01:35 PM roycroft: i'm pretty confident it will all fit
01:35 PM * Tom_L gives roycroft a cheap extension cord of 18ga wire
01:36 PM Tom_L: usually they're black white and green inside
01:36 PM xxcoder: you sure its copper and not "copper alum", or alum/butter alloy?
01:36 PM Tom_L: no warranty expressed or implied
01:46 PM lfluffywof: cheap extension cords are usually 16awg these days... cheap computer power cords are more likely to be 18awg. :P
01:50 PM Tom_L: i have a few dozen spare pc cords...
04:06 PM _unreal_: so I mad an annoying discovery. my ultra high quality monitor. best most expensive and largest one I have. has no wall mount ability
04:06 PM _unreal_: mad=made
04:07 PM _unreal_: and I just got my dual display mount mounted to this desk with some mods and I cant even use the secondary monitor with it:(
04:08 PM _unreal_: slightly annoyed
04:24 PM lcnc-relay: <lucid.nonsense@> Which is it? I got a an extra heavy duty mounty for my ergotron arm
04:24 PM lcnc-relay: <lucid.nonsense@> Cost a kidney though
04:31 PM roycroft: so that was a couple hours + of unplanned work
04:31 PM roycroft: but i have a nice 5v power bus now
04:31 PM roycroft: now i can wire up the 48v bus
04:31 PM roycroft: er, the 24v bus
04:32 PM xxcoder: nice an ev bus ;)
04:33 PM roycroft: i had run some temporary 5v wires for testing, and they're all neatly routed now
04:43 PM lcnc-relay: <lucid.nonsense@> Is there a consensus on the best way to add a spindle encoder to a spindle with a pneumatic drawbar? So, side mounted in some way
04:50 PM lcnc-relay: <captainhindsight_.@> anyway you can get to the component being rotated, some people add a machined encoder disk around the spindle and use a low cost optical sensor
04:51 PM xxcoder: isnt it clear sheet printable too?
04:51 PM lcnc-relay: <captainhindsight_.@> do you have a picture of your spindle?
04:51 PM xxcoder: cheaper if impossible to get one otherwise
04:51 PM xxcoder: essentally glue printed one to clear plastic sheet
04:52 PM lcnc-relay: <captainhindsight_.@> there is no consensus that i am aware of, it all depends on the circumstances and budget (time and materials)
04:52 PM xxcoder: true
04:54 PM lcnc-relay: <lucid.nonsense@> Not yet, it's an iso40 on a turret mill, I'd just like to do rigid taping and open up for an atc, the drawbar seems to work well, but maybe open to a Pugh conversion
04:54 PM lcnc-relay: <captainhindsight_.@> encoder + belt + pulley, encoder with gear on shaft and spindle, on and on
04:55 PM olfluffywf: watch Hyce's trip to Train Mountain makes me want to build a live steam project... but not a model. I absolutely can not relate to model makers. lol.
04:55 PM olfluffywf: can your spindle stop quickly enough for rigid tapping?
04:56 PM lcnc-relay: <lucid.nonsense@> Yeah, I'm planning to do something similar to the lathe, just wondering if the location would drift if I tried to do an amt absolute encoder due to the belt
04:57 PM lcnc-relay: <captainhindsight_.@> why would the encoder location drift? Thermal expansion?
04:58 PM roycroft: and i have 24v now
04:58 PM JT-Shop: depending on the belt the pitch could be off a tiny bit
04:58 PM roycroft: that went a lot faster
04:58 PM lcnc-relay: <lucid.nonsense@> Just stretchiness/backlash and inaccuracy of the belt, thibking out loud
04:58 PM olfluffywf: a timing belt is very accurate. a vee belt... is not repeatable.
04:59 PM lcnc-relay: <captainhindsight_.@> well you have to understand all the components involved vs just guess at what might work
05:00 PM roycroft: i'd say a timing belt can be very accurate, not is very accurate
05:00 PM lcnc-relay: <captainhindsight_.@> so it's s good engineering exercise
05:01 PM lcnc-relay: <lucid.nonsense@> Yeah, as long as I can make it work, vs polishing a turd of an idea, if you guys think it's a workable idea, I'll give it a go
05:01 PM JT-Shop: round tooth belts are more accurate than square tooth belts
05:02 PM xxcoder: wolf it is definitely possible to tap with spindle that cant instantly stop. just cant bottom out, and will need tap holder that can move tool down and up a little bit, so slight z movement/thread desync is fine
05:02 PM lcnc-relay: <captainhindsight_.@> time to read app notes from timing belt manufacturers
05:02 PM lcnc-relay: <lucid.nonsense@> Is incremental + limit switches the safer approach?
05:02 PM xxcoder: error must be very small though
05:03 PM xxcoder: movement range is like 2-3mm up and same down'
05:04 PM lcnc-relay: <captainhindsight_.@> how many IO's do you have left in your controller? do you already have an encoder? the qustions go on and on
05:04 PM JT-Shop: yup a floating tap holder works well if the spindle z is not super accurate
05:04 PM olfluffywf: Also, sas I said, can your spindle stop fast enough? heh. mine sure can't...
05:05 PM xxcoder: yeah. when reach the point, just turn off spindle and let tap stop it. then reverse out
05:05 PM JT-Shop: spindle stop time is not an issue emc tracks the spindle
05:05 PM xxcoder: if you need it to bottom out, just do the bottoming manually
05:06 PM olfluffywf: JT-Shop: it's an issue if you have to turn the spindle off halfway down the hole and hope it coasts just right to finish your threads and not end up in your parallels or something..
05:08 PM lcnc-relay: <lucid.nonsense@> Or index switches I suppose is more appropriate, atm I have 5i25 setup ready to go, but for the next week I am about 30 minutes away fron EUsurplus, so my options are almost anything, i have 2 lathes, 1 mill and 1 largish router waiting so could switch things around
05:11 PM lcnc-relay: <lucid.nonsense@> I was planning to use one of my two omron emcoders with the lathe, but now I'm thinking a bunch of modern absolute magnetic encoders might be better if the mechanical compensation isn't too unreliable
05:14 PM lcnc-relay: <captainhindsight_.@> it's up to you, your budget and skills
05:16 PM lcnc-relay: <lucid.nonsense@> Budget, not too limited, skills, very fluid, and depends on the amount of squirrel encounters
05:17 PM lcnc-relay: <captainhindsight_.@> https://forum.linuxcnc.org/media/kunena/attachments/23695/EncoderMount.jpg
05:18 PM lcnc-relay: <captainhindsight_.@> https://www.stevenson-engineers.co.uk/files/encoder.jpg
05:19 PM lcnc-relay: <captainhindsight_.@> for others a 40,000 ppr optical encoder from heidenhain is the only thing they will settle for 🙂
05:20 PM lcnc-relay: <lucid.nonsense@> captainhindsight_.@: I think i happen to have exactly the same encoder
05:21 PM lcnc-relay: <lucid.nonsense@> The surgery on the mill might be a bit nerve-wracking
05:23 PM Tom_L: there are still only 360 deg in a circle not matter how fine you split it
05:23 PM Tom_L: the index is what is important for rigid tapping anyway
05:25 PM Tom_L: or consider single point threading
05:28 PM lcnc-relay: <lucid.nonsense@> Yeah, I'm more worried about inherent drift of the system of using absolute encoders, but a repeatable index would probaly be the way to go. Thanks for the sanity check everyone
05:31 PM roycroft: tapping is overrated, unless we're talking about kegs
05:32 PM Tom_L: lucid.nonsense, consider thread mills as well. less cutter stress
05:32 PM lcnc-relay: <lucid.nonsense@> Yeah, so thread milling makes things a lot easier I'm guessing, never tried it
05:32 PM Tom_L: quicker
05:32 PM Tom_L: i've done single point but not a full thread mill
05:33 PM Tom_L: you just make a shorter helix and the whole thread is cut at once
05:33 PM Tom_L: the disadvantage to them is they are made for one thread size
05:33 PM Tom_L: single point can do several
05:33 PM Tom_L: but so are taps
05:33 PM Tom_L: so meh
05:34 PM Tom_L: https://www.lakeshorecarbide.com/uncunfthreadmills.aspx
05:34 PM lcnc-relay: <lucid.nonsense@> Yeah, 99 percent of what I'll do will be metric, so not that many needed
05:35 PM Tom_L: they have metric too
05:35 PM Tom_L: https://www.lakeshorecarbide.com/metricthreadmills.aspx
05:37 PM lcnc-relay: <lucid.nonsense@> Thanks, will have to look at what the tarriffs are at the time of day I order 😂
05:37 PM Tom_L: https://onlinecarbide.com/
05:37 PM Tom_L: they have quality tools as well
05:38 PM Tom_L: or find a foreign source
05:39 PM roycroft: and where stuff is coming from
05:39 PM roycroft: european countries are just not shipping anything to the us any more
05:40 PM roycroft: it started last week, and more countries are stopping shipments to here next week
05:40 PM lcnc-relay: <lucid.nonsense@> Well I'm uk based for the time being, if it's china produced, probably best to get it dorectly
05:40 PM lcnc-relay: <lucid.nonsense@> Directly
05:41 PM Tom_L: https://www.carbidedepot.com/formulas-tap-standard.htm
05:41 PM Tom_L: good resource
05:41 PM Tom_L: pretty sure they have metric as well
05:41 PM Tom_L: i used that when writing my threadmilling macros
05:43 PM lcnc-relay: <lucid.nonsense@> Cool, main need for imperial would be camera/tripod type stuff, otherwise metric for most things
05:43 PM roycroft: what a sucky day
05:43 PM lcnc-relay: <lucid.nonsense@> Why.
05:43 PM roycroft: it's still overcast, aqi is high (109), but it's also 37 degrees
05:43 PM Tom_L: https://www.carbidedepot.com/formulas-threadmilling.htm
05:43 PM Tom_L: if you're new to it
05:44 PM roycroft: yesterday the aqi was high but the smoke kept the heat away
05:44 PM roycroft: today not so much
05:44 PM roycroft: so i finished wiring the 24vdc psu
05:44 PM roycroft: i have maybe 100mm of black wire left
05:45 PM lcnc-relay: <lucid.nonsense@> Tom_L: Thanks, yeah I get the gist of it, still,atc'in the thing would need to need an index of some sort
06:03 PM olfluffywf: urgh... trying to do some gardening, and it is way too fucking hot.
06:04 PM roycroft: i'm heading back to the studio to work on the cnc controller some more
06:04 PM roycroft: because it's too hot and smokey to go outside
06:09 PM lcnc-relay: <.le_potato@> Tom_L: Imma here to install flexguyiii
06:12 PM lcnc-relay: <.le_potato@> what theme i need?
06:12 PM lcnc-relay: <.le_potato@> touch-probe3?
06:16 PM Tom_L: it's a good example yes
06:16 PM Tom_L: there are several examples
06:16 PM Tom_L: jt did an axis lookalike
06:19 PM lcnc-relay: <.le_potato@> one thing i dont see
06:19 PM lcnc-relay: <.le_potato@> where is the keyboard jog
06:22 PM lcnc-relay: <.le_potato@> AttributeError: 'flexgui' object has no attribute 'keyboard_jog_cb'
06:22 PM xxcoder: isnt that a part you needed to add?
06:23 PM lcnc-relay: <.le_potato@> i have to add on gui?
06:23 PM xxcoder: I probably is wrong lol
06:23 PM lcnc-relay: <.le_potato@> You will need a QCheckBox named keyboard_jog_cb and it must be checked for keyboard jog to work Keyboard jog only works when the power is on, in manual mode and keyboard_jog_cb is checked
06:23 PM lcnc-relay: <.le_potato@> oh lol
06:23 PM lcnc-relay: <.le_potato@> otherwise crash? not a nice solution!
06:25 PM lcnc-relay: <.le_potato@> ayy, it works!
06:26 PM xxcoder: awesome
06:30 PM lcnc-relay: <.le_potato@> thank you guys, now im obessed with flexgui!
06:31 PM Tom_L: :)
06:31 PM Tom_L: you can customize it any way you like
06:31 PM Tom_L: that is a test version. he is going to push it to the main repo soon
06:33 PM Tom_L: once he does, i will add the checkbox to the example
06:36 PM xxcoder: flex have been looking awesome
06:36 PM xxcoder: if I was still working on machine, I'd try it out
06:36 PM xxcoder: assuming I still had plenty of energy
06:44 PM lcnc-relay: <skunkworks8841@> scrub until tomorrow morning
06:45 PM Tom_L: it'll blow up anyway
06:46 PM lcnc-relay: <skunkworks8841@> not this time! lol
06:46 PM Tom_L: i guess they are getting better at it
06:47 PM Tom_L: i haven't watched it that much
06:48 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/temp/spacex/spacex_reentry.mp4
06:48 PM Tom_L: a reentry my kid caught with his phone at work
06:53 PM lcnc-relay: <skunkworks8841@> cool - was that a dragon capsil?
06:56 PM lcnc-relay: <skunkworks8841@> spelling
06:57 PM lcnc-relay: <skunkworks8841@> I am wiped. Just spent most of the day putting together book shelves..
06:57 PM lcnc-relay: <skunkworks8841@> thank goodness for impacts
07:04 PM Tom_L: i'm not sure what it was, it was a while back
08:02 PM xxcoder: man burning man seems worse and worse each time I hear of it lol
08:04 PM roycroft: ok, dumb question for the day
08:04 PM roycroft: my x axis travels in the wrong direction
08:04 PM roycroft: y and z are fine
08:04 PM roycroft: i know how to change that with servos - just set the scale to negative for that axis
08:04 PM roycroft: but i don't see how to do that for steppers
08:04 PM roycroft: other than a mention to "switch it in the hal file"
08:04 PM xxcoder: yeah you can flip the direction bit too
08:05 PM roycroft: i don't see a direction directive in the hal file
08:05 PM xxcoder: my first test run had linuxcnc logo drawn backwards
08:05 PM roycroft: or at least not one that i recognize as such
08:06 PM xxcoder: hmm since I just used configurator, I dont know what hal line for that would be
08:06 PM xxcoder: well you could also re-order wires
08:06 PM xxcoder: but yeah there should be config line somehow
08:07 PM roycroft: i'd rather do it in the hal file
08:08 PM roycroft: both because i'd like to keep the wiring consistent
08:08 PM xxcoder: indeed
08:08 PM roycroft: and because if i ever want to change it for whatever reason in the future it would be a lot easier to do it in software
08:08 PM xxcoder: I changed it using setting and not by wiring change too
08:09 PM roycroft: and i used a configurator to generate the hal file initially
08:09 PM roycroft: but i was not asked any questions about the direction the axes move
08:10 PM xxcoder: I recall something about reversing axes checkbox
08:10 PM xxcoder: but its been almost ecade
08:12 PM xxcoder: oh it was not MESA configurator. I dont know what mesa one looks like
08:13 PM roycroft: also, y will rapid for a bit then stop, and i have to power cycle the machine to get it working again
08:13 PM roycroft: i wonder if that driver is on its last legs
08:14 PM roycroft: anyway, that's a later problem - i need to figure out this direction problem now
08:14 PM xxcoder: tb6600 is so dang cheap, I wouldnt be surpised if caps blew in it afer few years
08:15 PM roycroft: yeah, and if i decide to order another i'll get 2 or 3
08:15 PM roycroft: or maybe consider using the gecko drivers that i have, but i'd rather not do that
08:16 PM roycroft: both because the controller is already worth way more than the machine it controls, and because i'd have to do a bunch of rewiring, and i'm tired of doing rewiring
08:16 PM xxcoder: 3 pc of tb6600 now is 25 bucks. which is 10 dollars more than I paid
08:16 PM xxcoder: drivers was so few back then, im sure theres more options now
08:17 PM xxcoder: my choices was reprap type, tb6560, tb6600, or driver that cost 3 times my machine cost so far for single one
08:18 PM Tom_L: flipping scale works with steppers too
08:19 PM roycroft: it does/
08:19 PM roycroft: ?
08:19 PM roycroft: thanks
08:20 PM roycroft: oh, i just found what i need in the .ini, i think
08:20 PM roycroft: DIR_INVERT
08:20 PM Tom_L: just flip scale
08:20 PM Tom_L: it's easier
08:20 PM Tom_L: never heard of that constant in the ini....
08:21 PM Tom_L: must be very new
08:21 PM xxcoder: dunno, I do remember such setting
08:21 PM xxcoder: I simply couldnt remember what it looked and in which file
08:21 PM roycroft: i just flipped the scale
08:22 PM roycroft: and am restarting
08:22 PM Tom_L: JOG_INVERT = - For each axis letter, the jog direction is inverted. The default is "X" for lathes and blank otherwise.
08:22 PM Tom_L: that seems to appear in the display section, not joint though
08:22 PM roycroft: right
08:22 PM roycroft: and i'm assuming that it's all flipped
08:23 PM Tom_L: it must be changed in the axis/joint section
08:23 PM roycroft: i could intentionally jog in the opposite direction that a program would move
08:23 PM roycroft: but i should assume that by default jog direction is the same as program direction
08:24 PM xxcoder: easy enough to test
08:24 PM roycroft: setting the scale negative fixed the x problem
08:25 PM xxcoder: awesome
08:25 PM Tom_L: i was sure of that
08:25 PM roycroft: but y still crashes after a bit of movement
08:25 PM Tom_L: in what way
08:26 PM Tom_L: during homing?
08:26 PM roycroft: linuxcnc keeps telling it to move
08:26 PM roycroft: but it does not move
08:26 PM roycroft: unless i power cycle the driver
08:26 PM roycroft: jogging
08:26 PM xxcoder: I wonder driver being bad is why I had so much problems, besides 24 volt being too weak
08:26 PM Tom_L: there is an override limits check box in axis to jog off a limit if it hits
08:27 PM Tom_L: determine if it's software or hardware first
08:27 PM Tom_L: move the problem to X and see if it's the driver
08:27 PM xxcoder: makes sense
08:28 PM roycroft: yeah, i'll swap x and y
08:29 PM Tom_L: i don't know alot about the dual gantry driver stuff
08:29 PM xxcoder: its not dual gantry driver
08:29 PM Tom_L: it looks like you're using a single though
08:29 PM xxcoder: you probably saw the machine in one of my videos
08:29 PM Tom_L: i know they've improved that but i've never tested one
08:30 PM roycroft: i swapped y and z, because the cables don't reach to swap x and y
08:31 PM roycroft: and it's not the driver, because i can reproduce it with the z driver
08:32 PM xxcoder: does it stapp over same spot each time?
08:32 PM roycroft: no, but i think the same travel distance
08:32 PM roycroft: keep in mind that i'm doing no homing at this point
08:32 PM Tom_L: double check that joint settings
08:32 PM roycroft: yes
08:33 PM roycroft: after i restore the wiring
08:33 PM Tom_L: soft limits won't apply until it's homed
08:33 PM Tom_L: because it doesn't know where it is
08:34 PM Tom_L: you don't get a joint following error do you?
08:34 PM Tom_L: that could be accel too high
08:34 PM roycroft: no
08:34 PM roycroft: it just stops moving, but linuxcnc thinks it's still moving
08:34 PM Tom_L: run from terminal to get more feedback
08:35 PM roycroft: maybe it is accel too high - i just increased that
08:35 PM Tom_L: possibly
08:35 PM Tom_L: you'd want to lower it
08:35 PM roycroft: i can do that for testing
08:35 PM roycroft: and eventually work out what is the right accel
08:35 PM roycroft: y is the longest axis, of course
08:35 PM roycroft: so the most likely axis to show that behavior if accel is set wrong
08:36 PM Tom_L: stepgen max accel needs to be 2x accel iirc
08:36 PM Tom_L: i didn't look up the exact names..
08:37 PM Tom_L: STEPGEN_MAXACCEL must be increased to 1.5 to 2 times the MAX_ACCELERATION for the joint. Excessive backlash compensation can cause an joint to jerk as it changes direction. If a COMP_FILE is specified for a joint BACKLASH is not used.
08:37 PM Tom_L: that's referring to backlash comp but could apply
08:38 PM Tom_L: then lower max accel
08:38 PM Tom_L: max_velocity could play a part
08:39 PM xxcoder: Tom_L: https://www.youtube.com/watch?v=1JMSRF6Q9uw
08:39 PM xxcoder: thats what roy has now
08:40 PM xxcoder: 8 years ago already lol
08:40 PM roycroft: i lowered both MAX_VELOCITY and MAX_ACCELERATION for the axis and the joint and now it travels even less before it stops
08:41 PM Tom_L: i think my x and y are pretty much the same but z is slightly different
08:41 PM Tom_L: try increasing it :)
08:41 PM xxcoder: maybe this is dumb question, but checked y coupler?
08:41 PM xxcoder: would be funny if its slipping
08:41 PM Tom_L: traj also has a max velocity
08:41 PM roycroft: i just lowered it even more
08:42 PM roycroft: i haven't, xxcoder - i'll do that after a couple more software tests
08:42 PM xxcoder: couplers was other issue I wanted to replace with better ones.
08:43 PM roycroft: and i don't know if this is a new problem or if i just discovered this problem
08:43 PM roycroft: i don't think i've ever jogged y from end to end before
08:43 PM xxcoder: y is along the longest axis right?
08:44 PM xxcoder: I think it was when I configured it, but you might have made different decisions
08:45 PM roycroft: yes
08:46 PM xxcoder: ok. same axis I had so much trouble with
08:46 PM Tom_L: you can purchase that mill right?
08:46 PM Tom_L: i think i know a guy here that has one similar
08:47 PM roycroft: i think it's still available
08:48 PM roycroft: when i further reduce those variables it does not move at all
08:48 PM roycroft: i'll increase them
08:49 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/configs/NEW_MILL/new_mill.ini
08:49 PM roycroft: now i get a joint 1 folowing error
08:49 PM Tom_L: there's mine for reference
08:50 PM Tom_L: check for binding in the axis
08:50 PM Tom_L: just tossing out ideas
08:51 PM roycroft: ok, it's not peculiar to y
08:52 PM roycroft: x does the same thing
08:52 PM roycroft: z doesn't have enough travel to be affected
08:52 PM roycroft: so i have some configuration issues
08:52 PM Tom_L: make sure the step timing is correct for that driver
08:52 PM roycroft: i'll try the old, slow as molasses config just to see
08:52 PM roycroft: i double checked that, but i can triple check
08:53 PM xxcoder: too bad I dont remember what I did to improve it, though it wasnt 100% at end
08:53 PM xxcoder: kinda wish I saved that 8 gb drve I had on linuxcnc pc
08:54 PM roycroft: it's all good
08:55 PM roycroft: every time i run into an issue like this it's an opportunity for me to learn more about linuxcnc
08:55 PM roycroft: and i always take advantage of that opportunity
08:56 PM Tom_L: there's plenty to learn
08:56 PM xxcoder: yeah
08:56 PM Tom_L: what's the max velocity?
08:57 PM roycroft: even in slow as molasses mode it stops after a bit
08:57 PM roycroft: 5
08:57 PM roycroft: and MAX_ACCELERATION is 4
08:57 PM Tom_L: mine is 4
08:58 PM Tom_L: vel
08:58 PM roycroft: i tried 4 and 3 and no change
08:58 PM xxcoder: https://forum.linuxcnc.org/16-stepconf-wizard/28553-configuring-tb6600
08:58 PM xxcoder: I wonder if thats what I found and used, which left it to only not enough voltage at end
08:58 PM Tom_L: also a setting for that in display
08:59 PM Tom_L: max linear velocity and default linear velocity
08:59 PM roycroft: i used jt's configurator for my machine
08:59 PM Tom_L: mine are 3.5 and 2.5 respectively
08:59 PM roycroft: but i've thought about using stepconf just to get the numbers it generates
09:00 PM roycroft: even though it does configs for parport
09:00 PM lcnc-relay: <.le_potato@> thats doesnt work with my stepper drivers 🙁
09:00 PM lcnc-relay: <.le_potato@> i had to do it by hand
09:00 PM Tom_L: those are just starting points. i haven't seen one yet that didn't need to be hand worked
09:00 PM lcnc-relay: <.le_potato@> then im chill
09:01 PM xxcoder: 5k, 5k, 20k, 20k. whats yours set to currently, on timings?
09:01 PM Tom_L: i suppose you can but not generally
09:02 PM Tom_L: jt's is pretty good, i dunno about the other one
09:08 PM roycroft: i think i'm about done for the day
09:08 PM xxcoder: is there such thing as timing for MESA?
09:08 PM roycroft: i have to do some more research on this so i understand better what i'm doing when i make changes to my ini file and main.hal file
09:10 PM roycroft: the plan was to do the psu wiring, then confirm that everything still works as expected, then get back to the e-stop and classic ladder issue today
09:11 PM roycroft: but everything is not working as expected, and likely never was
09:12 PM roycroft: life is large
09:12 PM xxcoder: reward of problem solved is a new one
09:14 PM roycroft: that is called "learning and growing"
09:14 PM xxcoder: yep
09:16 PM Tom_L: roycroft, just the stepgen max accel
09:17 PM roycroft: the stone of scone has been attacked again
09:17 PM roycroft: this time by a kilt-wearing aussie
09:18 PM roycroft: last month
09:18 PM roycroft: but i just saw the story :)
09:20 PM Tom_L: you're working with the 7i96s right?
09:20 PM roycroft: yes
09:20 PM roycroft: i'm wondering if this is some kind of limit issue though
09:20 PM Tom_L: it hasn't been homed yet though
09:20 PM roycroft: although i don't have limits set, and don't home or "touch off", i wonder if it refused to jog farther than it thinks it can go
09:21 PM xxcoder: I do feel I had partially fixed it using timing adjustment but cant remember more
09:21 PM roycroft: although an e-stop is not triggered, and that does not explain why i can't move in the other direction after it stops
09:21 PM xxcoder: tried homing it?
09:22 PM Tom_L: did you run from terminal and see anything more?
09:22 PM roycroft: yes and no
09:22 PM xxcoder: just turn it off and manually move it to home position
09:23 PM roycroft: that's irreleveant, xxcoder
09:23 PM roycroft: because when i power cycle the machine, linuxcnc has no idea where the spindle is located physically
09:23 PM Tom_L: it's an 'i can't push it up the hill' issue
09:23 PM roycroft: it assumes it's at 0,0,0
09:23 PM roycroft: wherever that may be
09:23 PM xxcoder: yeah. thats why you manually position it, then set it as home, then try jog it around
09:24 PM roycroft: but it doesn't matter where it is
09:24 PM roycroft: it still stops after it moves so far
09:24 PM Tom_L: what are the axis limits set to ?
09:24 PM Tom_L: min max
09:24 PM roycroft: greater than the size of the machine, at this point, because i increased them for testing
09:25 PM Tom_L: both directions?
09:25 PM Tom_L: pos and neg
09:25 PM roycroft: yes
09:25 PM roycroft: i think y is at -60, 60 now
09:25 PM Tom_L: ok
09:25 PM roycroft: and it's 18" long
09:25 PM xxcoder: where do it think its at when it had stopped?
09:25 PM roycroft: i wanted to eliminate that as the problem
09:25 PM Tom_L: and the scale is correct for the screws
09:25 PM roycroft: different places, it seems, xxcoder
09:26 PM roycroft: the location it thinks it's at does not seem to matter
09:26 PM xxcoder: odd. but still within the range I guess
09:26 PM roycroft: how far it's traveled since i started jogging is what seems to matter
09:26 PM Tom_L: the steppers are hefty enough to drive it?
09:26 PM roycroft: this is all empirical at this point
09:26 PM roycroft: they seem to be
09:26 PM Tom_L: ok
09:27 PM Tom_L: what's the driver current set to?
09:27 PM Tom_L: i
09:27 PM roycroft: i don't recall, but i think maybe 4.5a
09:27 PM Tom_L: i'm not familiare with those
09:27 PM Tom_L: ok
09:27 PM Tom_L: plenty
09:27 PM Tom_L: mine aren't even set that high iirc
09:28 PM roycroft: 3.5a i guess
09:28 PM roycroft: that's the max for that driver
09:28 PM Tom_L: mine are 3 or 3.5 somewhere in there
09:28 PM Tom_L: i could set them for 7 but don't see a need
09:28 PM roycroft: no, 4.0a is the max
09:28 PM Tom_L: not on mine
09:28 PM roycroft: i'm almost certain they're set to max
09:29 PM Tom_L: but the steppers are 4a iirc
09:29 PM roycroft: they are not overheating
09:29 PM Tom_L: they heat more when idle
09:29 PM roycroft: and i do have the specs for them
09:29 PM Tom_L: the steppers
09:29 PM roycroft: i think they're rated higher than the drivers can provide
09:29 PM Tom_L: unless you have idle current shutdown
09:30 PM roycroft: jogging is not idling :)
09:30 PM Tom_L: nope
09:30 PM roycroft: they are just warm, not hot at all, when they stop
09:30 PM xxcoder: wonder it would still stop if you disconnected other 2 steppers
09:30 PM xxcoder: using less power
09:30 PM roycroft: i'm only moving one axis at a time
09:30 PM Tom_L: i doubt it
09:30 PM xxcoder: your psu should be plenty beefy
09:30 PM roycroft: the other two axes should be not using much power at all
09:31 PM xxcoder: only enabled one axis?
09:31 PM Tom_L: standing they'er holding more
09:31 PM roycroft: i'm ony moving one axis
09:31 PM xxcoder: because not moving is highest power use
09:31 PM roycroft: yeah, that's true
09:31 PM roycroft: i seriously doubt this is a power issue
09:31 PM Tom_L: me either
09:31 PM roycroft: but i could test that next time i'm out there
09:31 PM xxcoder: likely no. you got big psu
09:31 PM roycroft: this is a software issue
09:31 PM roycroft: i'm convinced
09:31 PM Tom_L: same
09:32 PM roycroft: but it would be trivial to pull the power to the other two drivers for testing
09:32 PM roycroft: i can guarantee the results
09:32 PM Tom_L: you could tune one at a time if you wanted
09:32 PM roycroft: but it would be useful to prove that
09:33 PM roycroft: guarantees are not proof :)
09:33 PM xxcoder: yeah if it still happens, you have proven it isnt power, nor any inference
09:33 PM Tom_L: grounds are all correct?
09:34 PM Tom_L: star grounding
09:34 PM Tom_L: i had that problem on my sherline
09:34 PM roycroft: so when i run it in slow as molasses mode, which is a config that does not have distance per revolution configured properly, it moves a much shorter distance before stopping than when i run the config that has distance/revolution set properly
09:34 PM roycroft: grounds are good
09:35 PM roycroft: which supports my hypothesis that it is a distance traveled thing
09:35 PM Tom_L: put more money in the slot... you get more time
09:35 PM roycroft: not a hit limits or physical fault thing
09:35 PM roycroft: i think it moves as far as it thinks it's allowed to move then stops
09:36 PM Tom_L: if that were true you could reverse
09:36 PM roycroft: but it could also be a time thing, which makes no sense to me
09:36 PM roycroft: i.e. it moves for 17 seconds then stops
09:36 PM roycroft: no matter how far it's moved in those 17 seconds
09:36 PM roycroft: i should test that by timing it
09:36 PM roycroft: but i don't understand how or why linuxcnc would care about that at all
09:37 PM Tom_L: it wouldn't
09:37 PM Tom_L: that's just a result observation
09:37 PM roycroft: yes
09:37 PM roycroft: that's why i said up front that it would not make any sense to me
09:38 PM Tom_L: next test, run from terminal to see if you see more clues there
09:38 PM roycroft: i already told you i've done that
09:39 PM Tom_L: oh i thought you said no
09:39 PM roycroft: i said yes and no
09:39 PM Tom_L: is debugging set on?
09:39 PM roycroft: meaning yes, i ran it from a shell, and no, i got no more information
09:39 PM roycroft: no, and i don't know how to set that, but that's a good next thing to do
09:39 PM roycroft: so i'll find out how to do it
09:40 PM roycroft: oh, linuxcnc - d
09:40 PM roycroft: that's easy enough
09:40 PM roycroft: -d, rather
09:40 PM roycroft: probably -d -v would be good
09:40 PM Tom_L: # Debug level, 0 means no messages. See src/emc/nml_int/emcglb.h for others
09:40 PM Tom_L: DEBUG = 0x7FFFFFFF
09:40 PM Tom_L: in [EMC]
09:40 PM roycroft: i can do it from the command line
09:41 PM roycroft: i'm not sure if i can set the debug level from the command line though
09:42 PM Tom_L: DEBUG = 0 - Debug level 0 means no messages will be printed when LinuxCNC is run from a terminal. Debug flags are usually only useful to developers. See src/emc/nml_intf/debugflags.h for other settings.
09:42 PM Tom_L: they moved it a bit
09:42 PM roycroft: i'll try -d -v next time i start it up
09:43 PM roycroft: and if that's not enough i'll add a debug level in the hal file
09:43 PM Tom_L: ini
09:43 PM roycroft: or the ini file :)
09:43 PM roycroft: i'm still not always sure what belongs in which file
09:43 PM Tom_L: pins connections etc go in hal
09:44 PM Tom_L: settings usually in ini
09:44 PM xxcoder: hardware abstraction level makes sense yeah when i think about it
09:44 PM roycroft: i still need to get back to the encoder issue, once i get the e-stop figured out
09:45 PM roycroft: so i have two big things to deal with
09:45 PM Tom_L: i don't remember that one
09:45 PM roycroft: and the current problem is preventing progress on either
09:45 PM roycroft: it should not be a big deal
09:45 PM Tom_L: it's an adventure for sure
09:46 PM roycroft: i have a hall effect sensor on the spindle
09:46 PM xxcoder: it'll translate to big router someday
09:46 PM roycroft: i've been able to monitor spindle rpm by connecting it to a fluke
09:46 PM xxcoder: your xp
09:47 PM roycroft: i just have to figure out how to connect it to my mesa board and configure linuxcnc to use it
09:47 PM roycroft: yes, xxcoder
09:47 PM Tom_L: ok
09:47 PM roycroft: this machine is my "learning linuxcnc" machine
09:47 PM roycroft: i will do real work with it
09:47 PM roycroft: but the biggest benefit is learning how to interface linuxcnc with a machine
09:47 PM roycroft: that will translate to all my future builds
09:48 PM * Tom_L calls it a night
09:48 PM roycroft: ciaoder
09:48 PM roycroft: thanks for your help
09:48 PM xxcoder: colander
09:48 PM Tom_L: np
09:48 PM xxcoder: have good night tom
09:48 PM * roycroft has cad work to do this evening
09:49 PM roycroft: i'm almost out of time to work on the minirouter before i start work on the roof
09:49 PM roycroft: oh, i ordered wire today, and it's really weird
09:49 PM roycroft: i bought some 18ga wire on ebay
09:49 PM xxcoder: hopefully it isnt 30ga wire made from tin
09:50 PM roycroft: for a great price - the listing was for 500' spools for almost the price of 100' spools
09:50 PM roycroft: so i ordered a couple, and made an offer well below the asking price
09:50 PM Tom_L: generally get what you pay for or less
09:50 PM roycroft: the offer was accepted about an hour later
09:50 PM roycroft: then i got a shipping confirmation
09:50 PM roycroft: then i got a refund of $20
09:50 PM roycroft: then i got another refund for the entire shipping cost
09:51 PM Tom_L: you ain't gettin no wire sir
09:51 PM roycroft: it's ebay, tom_l
09:51 PM roycroft: i have buyer protection
09:51 PM Tom_L: later..
09:51 PM roycroft: if i don't get anything, or i get not what i ordered, i get my money back
09:51 PM xxcoder: thats really weird
09:51 PM roycroft: yeah
09:52 PM roycroft: and i never asked for any refunds
09:52 PM roycroft: these spools were listed at $25, which is already a bargain
09:52 PM roycroft: i offered $15
09:52 PM roycroft: so $30 total
09:52 PM roycroft: for the two
09:52 PM roycroft: i expected my offer to either be rejected or to get a counter offere
09:53 PM roycroft: but it was accepted
09:53 PM roycroft: then i got a $20 refund that i did not ask for on a $30 order
09:53 PM roycroft: i think i'll email the seller and ask what's up with that
09:53 PM xxcoder: good idea
09:54 PM xxcoder: oh talking about packages
09:54 PM roycroft: since i offered $10 less than the listing price, and ordered 2, the seller may have thought that i paid full price and that they needed to refund the difference between my offer and the listing price
09:54 PM roycroft: if that's the case i want to give them the $20 back
09:54 PM xxcoder: seems I havent gotten any random free package since then. awesome
09:54 PM roycroft: it's already a great deal
09:54 PM roycroft: yeah you mentioned that on friday, i think
09:54 PM roycroft: that you think you're done getting random stuff
09:55 PM roycroft: that's good, especially since it was not useful to you random setuff
09:55 PM roycroft: you just have to pay to throw it away
09:55 PM xxcoder: yeah. I was worried pause was just due to weekend, but so far seems not
09:55 PM xxcoder: car stuff isnt too bad, but film and breathe aid for nose is useless
09:55 PM xxcoder: I had given away skin lotion
09:55 PM roycroft: i'd just donate the stuff to st vinnie's
09:56 PM roycroft: but i have to drive 45 minutes to get to st vinnie's
09:56 PM roycroft: adam savage just dropped another video with nerdforge
09:56 PM roycroft: those have been pretty interesting - they've done a bunch of them recently
09:57 PM roycroft: they're all about process, not about actually making stuff
09:57 PM roycroft: both the process of making stuff and the process of making the youtube videos of them making stuff
09:58 PM roycroft: they are talking anvils and welding right now
09:59 PM xxcoder: adam and nerdforge is both channels I cant really watch
09:59 PM xxcoder: not that theyre bad or anything, just not ones I can watch
09:59 PM roycroft: i know you have difficulty with nerdforge
09:59 PM roycroft: i did not know about adam savage's channel though
09:59 PM roycroft: it is the subtitles? or the chaos?
09:59 PM roycroft: chaos
10:00 PM roycroft: oh, i spelled it right in the first place
10:00 PM xxcoder: dunno just how I felt, if I recall they did some handcams too
10:00 PM xxcoder: its good chaos to correct a correct spelling.
10:01 PM roycroft: adam uses fixed cameras more now
10:01 PM xxcoder: interesting
10:01 PM roycroft: and these discussions are with a single camera, with them sitting at a table
10:02 PM roycroft: i guess i like these shows because of what i was saying yesterday
10:02 PM roycroft: that on nerdforge videos i get why martina is the star, but i wish hansi had more input
10:02 PM xxcoder: sounds like they made some good changes. I remember legaleagle being agressively unwatchable due to zoom changes each shot, but he lessened it to what I accept, but later it got boring to me
10:02 PM roycroft: and in these discussions it's more hansi than martina talking with adam
10:02 PM roycroft: i get where you're coming from
10:03 PM xxcoder: watched eagle one for couple years or so
10:03 PM roycroft: part of the issue with the vertigo that i had (and still mildly have) is that some video effects trigger it, as well as quick positional changes
10:05 PM xxcoder: fun isnt it. now picture having it in extreme level for decade or more, though after 2020 was much worse when my vision stopped tracking
10:05 PM xxcoder: vision tracking finally well changed for better but barely
10:05 PM xxcoder: it now tracks "late" or slow
10:05 PM roycroft: yeah
10:06 PM roycroft: i doubt that i have the issue anywhere near to the degree that you do, but i do have it
10:07 PM roycroft: btw, i was just going through a book i have on roofing, since i'll be doing a roofing job for the first time in about 35 years
10:07 PM roycroft: and it started out with "anyone can shingle a roof (unless you have vertigo)
10:07 PM roycroft: i was like "oh, i'm no longer qualified to do this"
10:07 PM xxcoder: oh boy
10:07 PM roycroft: but the safety harness will make me qualified again :)
10:09 PM roycroft: the lower roof will be, i think, the steepest i've ever worked on
10:09 PM roycroft: 23/12
10:09 PM roycroft: ish
10:09 PM roycroft: but it also won't be too high
10:10 PM xxcoder: yeah safety is always important
10:11 PM roycroft: i still have my old roof jacks from the '80s, when i was doing roofing pretty regularly
10:11 PM roycroft: and i haven't used them since then
10:11 PM xxcoder: I would wear safety harness if I needed to work 5 feet off, because I have very bad sense of balance with dizzyness and fun vision stuff
10:14 PM * roycroft wonders if long pants can even be purchased anywhere in australia
10:14 PM xxcoder: maybe in halloween
10:15 PM roycroft: it just seems that all austalian men wear shorts 100% of the time, no matter what the situation
10:16 PM xxcoder: its quite hot land
10:16 PM xxcoder: but yeah pretyty sure it gets cold in south au
10:17 PM roycroft: and in the mountains in queensland and new south wales
10:19 PM roycroft: but most aussies live on the coast
11:08 PM wolfluffyf is now known as fluffywolf