#linuxcnc Logs
Aug 25 2025
#linuxcnc Calendar
12:17 AM Deejay: moin
04:23 AM Tom_L: morning
05:22 AM -!- #linuxcnc mode set to +v by ChanServ
07:26 AM rdtsc-w: wow, there used to be qucs and qucs-s (quite universal circuit simulator, same with spice), then there was qucsstudio, a completely separate Windows project. well, qucsstudio finally renamed themselves to something less confusing. https://qucsstudio.de/News
08:53 AM Unterhaus_ is now known as Unterhausen
10:47 AM JT-Cave: Flex GUI now has keyboard jogging
10:51 AM JT-Cave: Flex GUI now has keyboard jogging
10:51 AM roycroft: i hope to try flex gui at some point soon
10:51 AM JT-Cave: https://gnipsel.com/linuxcnc/flexgui/controls.html#jogging
10:51 AM roycroft: but i think i'll stick with axis until i get my machine working
10:52 AM JT-Cave: axis just works tm
10:52 AM roycroft: exactly
10:52 AM roycroft: and i have other things to worry about right now
10:52 AM JT-Cave: when you need a gui to suit you it's time for Flex
10:52 AM roycroft: i still haven't decided whether to use one of my touch screens with the minirouter
10:52 AM roycroft: if so, then that will be the time for flex gui
10:53 AM roycroft: i got wrapped up in another project last night and never got around to doing the research on my limited travel issue
10:53 AM roycroft: i'll try to do that after work today
11:00 AM JT-MShop: there is a couple of touch screen sims, main difference is the screen size
11:01 AM JT-MShop: one is sized to fit my elo touch screen
11:02 AM -!- #linuxcnc mode set to +v by ChanServ
11:06 AM roycroft: yeah, i know there are a couple
11:06 AM roycroft: i used one of them in sim mode a while ago
11:06 AM roycroft: my touch screen displays are smaller than the one i'm using with the minirouter right now
11:06 AM roycroft: and i'll probably want them to be pretty customized
11:06 AM roycroft: so flex gui seems a good fit for that, when the time comes
11:08 AM roycroft: my portable garage is in missouri now
11:08 AM roycroft: and it looks like it may arrive a day early, which would be good
11:10 AM * JT-Cave looks out the window to see if he can see it
11:14 AM roycroft: it's in earth city
11:14 AM roycroft: wherever that is
11:15 AM JT-Cave: that's a bit far to see
11:15 AM roycroft: it's in a big box
11:15 AM JT-Cave: ups?
11:15 AM roycroft: probably two big boxes
11:15 AM roycroft: yes
11:15 AM JT-Cave: that's one of their hubs
11:16 AM roycroft: oh, it's near st louis
11:16 AM roycroft: by the airport
11:18 AM JT-Cave: I think a chipmunk is eating my marshmallows I'm using for coon bait
11:19 AM JT-Cave: they are small enough to not trip the door
11:40 AM roycroft: i ordered a router base that i need for my upcoming roofing project
11:40 AM roycroft: it arrived in eugene on wednesday last
11:40 AM roycroft: the status was changed to "en route to next destination"
11:40 AM roycroft: then it arrived in eugene
11:40 AM roycroft: then the status was changed to "en route to next destination"
11:40 AM roycroft: it's been doing that every day since
11:41 AM roycroft: this is the usps, of course
11:41 AM JT-Shop: I've seen that before with usps
11:41 AM roycroft: i hope it gets her soon
11:41 AM roycroft: i need to build a jig for it
11:41 AM roycroft: and i'd like to do that this week
11:47 AM xxcoder: well certainly cooler start today
11:48 AM JT-Shop: today is amazing temperature after the heat wave
11:49 AM Tom_L: cool but raining
11:49 AM JT-Shop: chance of rain here but nothing on radar
11:55 AM roycroft: the heat advisory runs through tonight here, but we should see significant relief starting tomorrow
11:56 AM xxcoder: yeah still in heat warning here too. 88f predicted here today, little bit hotter tomorrow
11:56 AM xxcoder: it drops to 82f after
11:56 AM roycroft: and we have a red flag warning now as well
11:57 AM xxcoder: interesting. had that last 2 days but not today. just extreme heat now
11:57 AM roycroft: and a flood advisory in northwest oregon, including eugene, but not here
11:58 AM roycroft: i expect the floodgates will be partially open on the dams between here and town
11:58 AM lcnc-relay: <skunkworks8841@> 64f here
11:58 AM roycroft: they've been completely closed since early june
11:58 AM xxcoder: aqi is poorest here for the year. 28 aqi, or 54 depending on site
12:00 PM roycroft: aqi is down to 57 here
12:00 PM roycroft: it's been in triple digits since last week
12:00 PM xxcoder: i'm wrong, 28 is PM units, not aqi. 50s is aqi here
12:01 PM xxcoder: https://atmotube.com/blog/particulate-matter-pm-levels-and-aqi
12:01 PM roycroft: the flat fire is still growing, and is now consuming homes
12:02 PM roycroft: i think that's the first time this year an oregon wild fire has burned a home
12:02 PM roycroft: "only" four so far
12:02 PM roycroft: but if it's reached civilization, there will probably be many more
12:03 PM roycroft: i'm pretty sure that's the fire that brough the smoke here
12:11 PM JT-Shop: guy just bought two items from my marketplace listing
12:13 PM roycroft: i need to list more stuff
12:13 PM roycroft: it doesn't sell very well when it's in storage and not listed for sale
12:13 PM xxcoder: oh boy. this is new autovomit stabilizer. https://www.bbc.com/future/article/20250822-youtube-is-using-ai-to-edit-videos-without-permission
12:13 PM roycroft: at least that's been my experience
12:14 PM roycroft: ai is making the internet useless
12:16 PM xxcoder: yeah. thats why I wanted to move on from browser that have ai in it
12:17 PM xxcoder: seems theres prompting exploit where just scrolling can make browser send email with bank info, if you happen to have that bank account
12:17 PM xxcoder: prompt injection
12:19 PM JT-Cave: actually googles ai can be quite helpful with programming questions
12:20 PM xxcoder: sure. not questioning that its helpful for some people
12:20 PM xxcoder: prompt injection is still a big concern
12:24 PM xxcoder: its same for when it is not wanted. like youtube using ai to "improve" videos. they didnt learn lesson with what I call "autovomitor stabilizer"
12:25 PM xxcoder: for short time youtube enforced stabilizer to all videos, and results was pretty horrible
12:27 PM xxcoder: man it have been horrible time for me. I had almost entirely stopped watching videos that didnt have stable single camra stand
12:27 PM rdtsc-w: shakiness can make ppl physically ill
12:28 PM xxcoder: yeah. and thats before any health issues that makes it so much worse
12:29 PM xxcoder: but. I rather have shakiness over autovomitor
12:29 PM rdtsc-w: I see the word vomit in there :)
12:29 PM xxcoder: thats for a reason.
12:31 PM xxcoder: https://www.youtube.com/watch?t=75&v=g0a4L3ExQlU&feature=youtu.be this is from autovomitor era
12:34 PM xxcoder: hopefully nobody had to make sudden trip to restroom
12:35 PM JT-Cave: to modify content without permission of the owner is wrong in every sense
12:35 PM xxcoder: indeed
12:35 PM JT-Cave: a class action suit is needed to get their attention
12:39 PM roycroft: i was listening to a story about ai the other day on npr, and they were interviewing an attorney for some artists who claimed that 80% of pintrest pictures are now ai-generated ripoffs of real artists' works
12:39 PM roycroft: i'm pretty sure, jt-cave, that youtube's terms of service are "all your videos are belong to us"
12:40 PM JT-Cave: and that's another reason I never look at anything on pintrest
12:40 PM JT-Cave: prob so
12:40 PM roycroft: i used to try to look at pintrest
12:40 PM roycroft: but it was useless
12:41 PM roycroft: there would be 2 second teaser with something potentially interesting, then a pop-up demanding one log in to do anything else
12:41 PM JT-Cave: https://www.youtube.com/t/terms#27dc3bf5d9
12:41 PM xxcoder: yeah always was useless to me. that site is very agressive on login
12:42 PM JT-Cave: anyway it's nap time
12:42 PM Tom_L: agreed
12:42 PM xxcoder: removing elements help me get to content when pinterest is only matchibng result and i have to find more info. otherwise I never visit that crap site
12:52 PM roycroft: i phoned the usps over two hours ago and got put in a 25-minute hold queue
12:52 PM roycroft: i was given the option to take a call back when it was my turn, and executed the option
12:52 PM roycroft: i'm still waiting for the call
12:53 PM roycroft: make america wait again!
01:03 PM xxcoder: I have never had some company call back. ever
01:03 PM xxcoder: so I always wait
01:08 PM roycroft: most companies do call me back
01:08 PM roycroft: some a lot faster than their hold queue indicates
01:08 PM roycroft: but apparently not the usps
01:08 PM xxcoder: bah lucky you lol
01:09 PM roycroft: they probaby did call me, but the call got routed back to eugene, then forwarded on to san francisco, and then to new york, and the call just hasn't made it back to the west coast yet
01:09 PM xxcoder: lol
02:20 PM roycroft: and life just gets weirder
02:21 PM roycroft: i was calling the usps because they claimed my box rental payment did not go through, twice, and when i went online to pay it manually i was informed that the url they redirected me to to manage my po box was invalid
02:21 PM roycroft: so i had to go the post office at lunch time anyway, and while i was there i figured i would renew it in person
02:21 PM roycroft: but they said it was not due until february, meaning the payment had gone through after all
02:21 PM roycroft: and still no call back
02:40 PM andypugh: pcw-home: Do you know if the 8i20 is expected to come back into stock?
02:41 PM andypugh: I may have a dead one, though if so it is dead in a strange way.
02:44 PM roycroft: we live in strange times :)
02:45 PM andypugh: It will hold position, ie the PID works, and so presumably current commands get through, but it gives communications faults and faults out if I try to move it.
02:46 PM andypugh: I spent ages trying to free up the ball screa and slides, but the problem still exists with the motor loose outside the machine.
02:46 PM lcnc-relay: <skunkworks8841@> through smart serial I assume?
02:46 PM lcnc-relay: <roguish> andypugh: I notice the wiki front page was wierd yesterday. I was showing someone who wants to use linuxcnc
02:46 PM andypugh: Yeah
02:47 PM lcnc-relay: <roguish> looks better now. thanks
02:47 PM roycroft: i still need to email the guy who maintains the wiki page to see if he would like some help with it
02:47 PM roycroft: it's in pretty bad shape
02:47 PM lcnc-relay: <roguish> like it used to as far as I see
02:48 PM lcnc-relay: <skunkworks8841@> that is so weird.. If it holds position - but fails when moving - almost sounds like a wierd noise issue.
02:48 PM roycroft: maybe he's been trying to fix it himself
02:48 PM roycroft: i should mention my problem, now that there are more folks here who are more experienced than me
02:48 PM lcnc-relay: <skunkworks8841@> will it error if you try to force the lead screw?
02:48 PM roycroft: i ran into this last night, after adding a new psu to my controller
02:48 PM roycroft: i have a 7i96s, and stepper motors
02:49 PM roycroft: i can jog for a bit but then the axis stops moving, and i cannot move it again unless i power cycle the machine
02:49 PM roycroft: i've observed this behavior on two axes, and the behavior follows the axis, not the driver or motor
02:50 PM roycroft: it seems to me it's a software issue, but i haven't a clue what could be happening
02:50 PM roycroft: i don't think it's a limit issue, as if it were, i should be able to move in the opposite direction once it stops
02:50 PM lcnc-relay: <skunkworks8841@> do the numbers move in axis? like linuxcnc thinks it is moving?
02:50 PM roycroft: and i do not have any limits defined anyway
02:50 PM roycroft: yes, they do
02:50 PM roycroft: linuxcnc thinks it's moving, but it's not
02:51 PM roycroft: this happens on both the x and y axis
02:51 PM roycroft: z does not have enough travel to trigger it, so i don't know if it would happen there
02:51 PM lcnc-relay: <skunkworks8841@> but no communication errors with mesa?
02:51 PM roycroft: non reported
02:51 PM roycroft: none
02:52 PM lcnc-relay: <skunkworks8841@> that is odd..
02:52 PM roycroft: i'm going to run in debug mode after work today to see if i get any more information
02:52 PM lcnc-relay: <skunkworks8841@> that sounds like a electrical problem..
02:52 PM roycroft: if it is, it's an electrical problem with at least 2 axes on the mesa board
02:52 PM roycroft: because i've switched drivers around and i've switched stepper motors around, and the problem follows the axis, not the driver or motor
02:53 PM roycroft: i first observed it on y, and xxcoder suggested i check the coupler
02:53 PM roycroft: i hadn't gotten around to that, but also experienced it on x
02:53 PM roycroft: and it would be really odd if both the x and y couplers loosened up at the same time
02:54 PM roycroft: but i'll verify that as well
02:55 PM roycroft: and sorry, i did not mean to hijack andypugh's issue
02:55 PM roycroft: i thought both might be discussed simultaneously :)
02:55 PM roycroft: but maybe his is too weird for anyone to want to speculate further
02:56 PM lcnc-relay: <skunkworks8841@> the mesa<>linuxcnc communication would be ok.. (if the numbers are changing in axis) because either way the there would be loop closed between mesa and linuxcnc (either position or pid) depending on your setup. It would be interesting to scope the step pin...
02:57 PM roycroft: i can do that
02:57 PM roycroft: i get off work in a bit over 2 hours, and plan on working on it further then
02:57 PM xxcoder: thats great idea. see if steps signal is going or not
02:58 PM roycroft: yeah
02:58 PM roycroft: if the mesa stops sending the step signals, then we need to figure out why
02:58 PM andypugh: roycroft: Does it move in both directions until it doesn’t?
02:58 PM roycroft: if it is, then we need to figure out why the driver is not responding
02:58 PM roycroft: yes
02:58 PM roycroft: it moves in both directions until it doesn't
02:58 PM xxcoder: I guess its good idea to check both in and out from driver
02:58 PM roycroft: and while i've not explicitly measured, it seems to move about the same distance before it stops each time
02:59 PM andypugh: Keep a watch on hm2…….serial-port-0.error-counts
02:59 PM roycroft: this is when fast jogging
02:59 PM roycroft: ok
02:59 PM andypugh: I think that comms errors are masked until they hit the error threshold.
03:01 PM roycroft: i was going to do some more studying last night to try to figure out what might be going on, but after i got my cad work done i fell asleep
03:02 PM roycroft: i'm finding that when you get old, even if something is interesting and you want to work on it, when it's time to sleep it's time to sleep and you have no choice :(
03:05 PM roycroft: monitoring those pins seems like a better next step than running in debug mode
03:05 PM lcnc-relay: <skunkworks8841@> it is weird though if you are not getting any errors from linuxcnc.. atleast in my experience. (makes me think it is wiring side of the whole thing) Unless you are having a wierd noise issue that somehow puts the mesa board in some odd state. But I have not seen that
03:06 PM roycroft: i did rewire the mesa board to drivers connections yesterday, but i was very meticulous, and redid one at a time, using the others as references
03:06 PM roycroft: and since the motors work (for a bit) it seems it's not a wiring fault
03:07 PM roycroft: i had to move the mesa board back about 10mm in the cabinet, and the cables to the drivers were no longer long enough
03:08 PM roycroft: that's a tiny difference, and one would not think it would be the cause of any interference, since the axes seemed to work before
03:08 PM roycroft: although i'm not sure i ever did a long rapid before
03:08 PM roycroft: so this may have been a problem all along
03:08 PM roycroft: the cables are all different lengths, btw
03:09 PM roycroft: so it's not like i accidently made them all the magic length to pick up unwanted harmonics, etc.
03:09 PM roycroft: well, i could have, but that would have been three simultaneous "accidents"
03:10 PM xxcoder: isnt there ferrious rings something that gets rid od those noise
03:10 PM roycroft: and it seems highly unlikely to me that i managed ot do that
03:10 PM roycroft: yes, and if i suspected that noise was involved i'd add them
03:10 PM roycroft: i was pretty careful in my layout
03:11 PM roycroft: in fact someone here (i forget who) who is an ee commented that it was well thought-out
03:11 PM roycroft: like my ac mains lines are routed far away from the signal lines, and where they do cross, they cross at a 90 degree angle
03:12 PM roycroft: there are some ac and dc lines that run parallel and close to each other
03:12 PM xxcoder: interesting mitutoyo product https://makersmarket.au/product/tooling/mitutoyo-wrist-watch/
03:13 PM xxcoder: yeah. I guess you will have to do signal and power analsys to see whats going on
03:18 PM roycroft: https://roycroft.us/MiniRouter/Wiring4.jpeg
03:18 PM roycroft: that's the current layout
03:20 PM roycroft: if anyone sees any major wire routing flaws please let me know
03:20 PM roycroft: because it seems that my favorite thing to do is rewire that controller cabinet
03:20 PM xxcoder: wonder if can move drivers up a little bit, so you can also move 24v psu up a little?
03:21 PM roycroft: that would be about 3 hours of work
03:21 PM roycroft: what would it gain me?
03:22 PM roycroft: maybe only 2 hours
03:22 PM xxcoder: yeah true, only improvement is ease of access to one row of terminals and should be almost never needed
03:22 PM roycroft: right
03:22 PM roycroft: and i have to remove that psu to access them anyway
03:23 PM roycroft: even before i installed the 24v psu i had to remove the 48v psu to access the terminals
03:23 PM roycroft: i'm not excited about the layout, but it was an evolving thing
03:23 PM roycroft: if i were to rebuild the cabinet from scratch i'm sure i'd come up wiht a better layout
03:23 PM xxcoder: well it looks reasonable to my unexpert eyes
03:23 PM roycroft: but what i have works ok
03:24 PM roycroft: and btw, you can see that 3 conductor ribbon cable i was talking about yesterday
03:24 PM roycroft: going from the potentiometer connector on the spindle control board to the mesa card
03:24 PM xxcoder: yeah I see it
03:25 PM xxcoder: from spindle board to mesa
03:26 PM roycroft: the purple/blue wire pair is for the e-stop circuit
03:26 PM xxcoder: is the drivers to mesa wires the orginal wires but cut down?
03:26 PM roycroft: and since i don't have that all worked out yet they are routed temporarily
03:26 PM roycroft: those are the wires that were originally on the drivers when i got the machine from you, yes
03:26 PM xxcoder: lol ok. I felt I reconized those, but wasnt 100%
03:27 PM roycroft: i briefly considered installing shielded cables, but i figured that those flat wires would probably work fine
03:28 PM roycroft: the blue and red wires from the 48v psu to the drivers are also from the stuff i got from you
03:32 PM xxcoder: lol ok. should be perfectly fine for fixed wiring
03:34 PM roycroft: i'll keep an open mind and investigate, of course, but it does not seem at all obvious to me that interference could be the culprit here
03:34 PM roycroft: especially since the axes move about the same distance every time before stopping
03:34 PM roycroft: interference is generally more random
03:37 PM xxcoder: yeah. much stuff needs to be measured
04:57 PM JT-Shop: crap I got find my z again
04:58 PM xxcoder: maybe need little bit more nap :)
04:58 PM JT-Shop: and my x...
05:11 PM roycroft: so i'm done with work
05:11 PM roycroft: for the day
05:12 PM roycroft: i can go work on the controller again in a few minutes, and monitor some pins to see what the mesa board is doing
05:21 PM JT-Shop: one more hex to cut
05:24 PM JT-Shop: maybe tomorrow I can work on the deck repairs... today was totally shot
05:25 PM roycroft: i'm looking at my list of indoor projects and realized that i need to build most of the new doors for the house before i can install the new flooring
05:33 PM rdtsc-w: JT has no problem finding the Z's around lunchtime :)
06:04 PM xxcoder: fun. had to go out to dump recycling stuff and lower awning some more. had t0 go as fast as I can, very hot
06:04 PM xxcoder: yet predicted max keeps going down. its 85f now
06:05 PM xxcoder: tomorrow was as high as 92f, now 88f
06:05 PM lcnc-relay: <ccatlett1984@> Its an amazing 68F here
06:05 PM lcnc-relay: <ccatlett1984@> finally
06:05 PM lcnc-relay: <ccatlett1984@> cool tomorrow as well
06:06 PM xxcoder: nice. tommorrow should be last hot day for week. rest of week is quite warm to kinda hot
06:06 PM xxcoder: like 75f to 80f
06:07 PM lcnc-relay: <captainhindsight_.@> Big shift in the weather here
06:07 PM Tom_L: 59 overnite 77 for the high
06:07 PM JT-Shop: 55f tonight
06:07 PM roycroft: so when i slow down jog speed and max velocity the axis travels less distance before it stops
06:07 PM roycroft: it seems to be time-related
06:07 PM lcnc-relay: <captainhindsight_.@> All 70's for highs for the next week or two
06:08 PM roycroft: monitoring the pins i thought would be interesting shows no errors and the mesa board keeps sending position signals to the drivers when the axis stops moving
06:08 PM lcnc-relay: <captainhindsight_.@> Even some 40's at night
06:08 PM JT-Shop: if you want the jog to stop at once it's a bit different
06:08 PM roycroft: today should be our last super hot day for a while
06:08 PM Tom_L: roycroft, steppers wired right? not faulting due to some crazy short?
06:08 PM roycroft: but we still have a red flag warning, an extreme heat warning
06:08 PM Tom_L: thermal shutdown
06:08 PM xxcoder: roy sadly one more day for me
06:08 PM roycroft: nothing is getting hot
06:09 PM roycroft: not the drivers, the psus, the mesa board, the motors
06:09 PM roycroft: and the motors are wired the same as they have always been wired
06:09 PM Tom_L: did they ever work like they should?
06:09 PM roycroft: i'm not sure i ever got more travel than this, but i think i idid
06:10 PM roycroft: i'm pretty much at a loss to understand what's going on at this point
06:10 PM xxcoder: I know that feeling
06:10 PM xxcoder: it might be timing still
06:10 PM Tom_L: got any geckos?
06:12 PM JT-Shop: iirc there is two types of jog, one that keeps going until it reaches the jog point and one that just stops when you do a key release
06:12 PM roycroft: yes, but i don't want to use them for this application unless i really have to
06:12 PM roycroft: i even tried "stutter" jogging
06:12 PM roycroft: i.e. i set the jog to continuous
06:12 PM Tom_L: might be useful for a test
06:12 PM roycroft: and i'd click on and off and on and off
06:12 PM xxcoder: mesa output still correct?
06:12 PM Tom_L: the signal is still getting out
06:13 PM JT-Shop: I can't remember how to select which jog you want, you may have to do some reading of the docs
06:13 PM JT-Shop: of course Flex only does stop at key release
06:13 PM * JT-Shop calls it a night
06:13 PM roycroft: in axis i can select 0.001", 0.010", 0.100", or continuous
06:14 PM roycroft: i'm using continuous
06:14 PM roycroft: i suppose i could try setting to 0.100" and repeatedly clicking on it
06:14 PM roycroft: the mesa is still sending signals
06:14 PM roycroft: or so it reports
06:14 PM xxcoder: yeah wonder if that would go past that distance
06:14 PM xxcoder: no olloscope?
06:15 PM roycroft: and btw, i do not have to power cycle the whole machine
06:15 PM roycroft: i've found that if press the e-stop, which kills power to the 48v psus, then shut down linuxcnc, reset e-stop, then restart linuxcnc it works again
06:17 PM Tom_L: is there a lowpass filter on the jog?
06:17 PM roycroft: so when i set jog to 0.100" i can move about twice as far as when it's set to continuous, but it still stops after a bit
06:17 PM Tom_L: i use one on my pendant to have it finish the steps requested
06:17 PM roycroft: i don't think so
06:17 PM Tom_L: ilowpass
06:18 PM xxcoder: could you move it via g code commands see if it also stops?
06:18 PM roycroft: unless measct added one
06:18 PM Tom_L: i doubt it
06:18 PM roycroft: i haven't done that yet - i'm still jogginb
06:18 PM roycroft: jogging
06:18 PM roycroft: but yes, i should send it some gcode and see what it does
06:19 PM Tom_L: request a G91 G1 X1
06:19 PM Tom_L: see if it moves an inch
06:19 PM Tom_L: set F to something first
06:20 PM xxcoder: by habit. I would make it x1. because x1 have different definitions depending on machine. it should be fine and one inch on linuxcnc but yeah
06:20 PM Tom_L: inch or mm
06:20 PM Tom_L: G91 is incremental
06:20 PM xxcoder: or 0.001"
06:20 PM xxcoder: I ran few machines where its like that
06:20 PM Tom_L: not lcnc
06:21 PM xxcoder: like I said earlier :)
06:22 PM xxcoder: frankly I like how lcnc interpets it. x1 = 0.001" never made any sense to me.
06:23 PM roycroft: so i touched off
06:23 PM roycroft: with the spindle in the lower left corner
06:23 PM roycroft: then went to mdi and sent "G1 X0 Y9 Z0 F30"
06:23 PM roycroft: it started moving y
06:23 PM roycroft: but stopped after about 4.8"
06:23 PM roycroft: which is where it stops at that speed when jogging
06:23 PM roycroft: so it's not a jogging issue
06:24 PM roycroft: linuxcnc thought it went all the way to 9"
06:24 PM xxcoder: while it didnt find specific reason, its good to know
06:24 PM xxcoder: 4.8" dont seem to match any microstep error at all
06:24 PM xxcoder: 4.5" maybe but not 4.8".
06:25 PM roycroft: but if it were a microstep error the axis would keep moving until linuxcnc told it to stop
06:25 PM roycroft: the distance would be off, but it would keep moving
06:25 PM roycroft: that's not what happened here
06:25 PM roycroft: linuxcnc started moving it 9", and axis showed it moving
06:25 PM xxcoder: yeah thats another indiction. it would move constantly to finish "9"in
06:25 PM xxcoder: but it stopped too early
06:25 PM roycroft: but at about 4.8", the axis on the machine stopped, but linuxcnc kept "moving" it until it reached 9"
06:26 PM xxcoder: could you do "assistance" of gently pushing it in travel direction
06:27 PM xxcoder: not enough to move y on your own but just enough to offset load
06:29 PM Tom_L: did lcnc keep moving to 9 or was it at 9 when the cnc was at 4.x?
06:29 PM roycroft: it kept moving to 9
06:29 PM * roycroft wishes a fripping gopro would hold a charge for more than 15 minutes
06:29 PM Tom_L: no errors
06:29 PM roycroft: i'm going to try to do a short video
06:29 PM roycroft: none reported that i sawa
06:30 PM Tom_L: did you set the debug level in the ini?
06:30 PM Tom_L: then run from terminal
06:30 PM Tom_L: after that, run show and watch all the associated pins in hal show
06:30 PM Tom_L: see if anything changes when it stop
06:30 PM Tom_L: s
06:31 PM Tom_L: stepgens, enables anything that could cause it to stop
06:31 PM Tom_L: even the limits
06:32 PM roycroft: i watched a bunch of pins with halshow and nothing changed when it stopped
06:32 PM Tom_L: it seems to be an issue outside of linuxcnc or it would fault in some way
06:33 PM Tom_L: but it keeps going
06:33 PM roycroft: but i did not monitor all of the pins, because there are way too many
06:33 PM xxcoder: I would somehow check output to steppers
06:33 PM roycroft: i'm looking for an extension cord
06:33 PM Tom_L: not all, just the associated ones
06:33 PM roycroft: then i can do a short video
06:33 PM Tom_L: the enables to the drivers don't change?
06:33 PM Tom_L: no relay is bouncing etc
06:34 PM Tom_L: monitor all the associated power signals
06:34 PM Tom_L: something is making it stop
06:34 PM xxcoder: enables would be interesting cause, it flipping when its moving
06:35 PM roycroft: i don't have enable wired
06:35 PM roycroft: becaue jt told me not to, and it worked when i did what he told me
06:35 PM Tom_L: i had a noise issue on the sherline that would cause odd behavior
06:36 PM Tom_L: you may want to tie them either up or down with a resistor
06:36 PM Tom_L: if you're not using them
06:37 PM Tom_L: i don't leave anything floating
06:37 PM xxcoder: looking at your picture again, yeah wow not wired at all.I just tied all ena- t ground and ena+ to pin to be enabled
06:38 PM xxcoder: or was it v+ to ena+ and ena- to pin? been too long
06:39 PM xxcoder: honestly im sure you can just wire v+ to + and ground to -, and power supply is what turns v+ on and off
06:39 PM xxcoder: definitely check it but yeah
06:40 PM Tom_L: i don't rely on internal pullups either
06:40 PM Tom_L: if it needs one i add it
06:48 PM roycroft: i had wired it initially, but i could not get the motors working
06:48 PM roycroft: jt told me not to wire it with the mesa board and those drivers, so i did not
06:51 PM xxcoder: its possible that its floating so it "works" till it flips below and stops
06:51 PM xxcoder: but I dont know
06:51 PM xxcoder: should do no harm to just wire ena- to ground and see
06:51 PM xxcoder: since it would stop floating
06:52 PM xxcoder: if it refuses to run with ground, then its probably set to disabled
06:53 PM lcnc-relay: <ccatlett1984@> I was very sad to see the 7c81 doesn't have a switching pin for each stepper enable
06:53 PM lcnc-relay: <ccatlett1984@> ended up tieing them together, and using 1 of the isolated outputs to switch
06:54 PM roycroft: i think i officially hate gopros
06:55 PM xxcoder: that was how I orginally wired it also ccat. just all 3 drivers enable together to pin, and other to ground (or other way)
06:55 PM roycroft: maybe i'll do better with duct tape and my iphone
06:56 PM xxcoder: hm just stupid test idea. move it 3 inches, then wait few minutes, then do it again, then third time
06:56 PM Tom_L: doesn't need to be production quality :)
06:56 PM xxcoder: isnt iphone automatically prod quality? ;)
06:57 PM roycroft: i just wanted something i could mount on a tripod so my hands would be free to type gcode
06:57 PM Tom_L: galaxy does pretty good
06:57 PM rdtsc: some (most?) stepper drives default to enabled, and you have to give them some sort of signal to disable
06:57 PM Tom_L: probably
06:57 PM roycroft: the gopro has a tripod mount
06:57 PM roycroft: my iphone does not - i should get one for it
06:57 PM xxcoder: rd that was what I was wondering. it might flip when its being moving and stay disabled
06:58 PM xxcoder: theres probably 3d printable mount for that tripod for iphone
06:58 PM xxcoder: if thats the case, then tom finally found why :)
06:59 PM Tom_L: i've just found it best to _know_ the state of everything
06:59 PM rdtsc: tom is a good finder of all things
07:00 PM rdtsc: If Tom was at JT's shop, would have found the 7c81 like a month ago :)
07:00 PM Tom_L: i've made enough errors
07:01 PM xxcoder: plenty of sessions in school of hard knocks eh
07:07 PM Tom_L: maybe there's a shipping strap under the axis that needs removed :)
07:14 PM roycroft: it travels different distances, depending on how fast i move it
07:15 PM lcnc-relay: <ccatlett1984@> that sounds like skipped steps
07:15 PM roycroft: but the axis stops moving way before linuxcnc thinks it's done
07:15 PM lcnc-relay: <ccatlett1984@> no alarm on the driver?
07:16 PM roycroft: if it were skipped steps the axis would move as long as linuxcnc keeps sending it step commands
07:16 PM Tom_L: when it stops, are the steppers stiff or can you turn them
07:16 PM roycroft: it just would not move the correct distance
07:16 PM roycroft: i can easily move the axis by hand
07:17 PM Tom_L: no current to the motor then
07:17 PM lcnc-relay: <ccatlett1984@> this closed or open loop steppers?
07:17 PM Tom_L: open i bet
07:17 PM roycroft: yes
07:17 PM lcnc-relay: <ccatlett1984@> betting the same
07:17 PM roycroft: i should try moving by hand when it has stopped but linuxcnc is still sending
07:17 PM lcnc-relay: <ccatlett1984@> i think your drivers are alarming, due to the missed steps, and turning off current
07:17 PM roycroft: let me do that while my little movie tries to upload
07:18 PM Tom_L: add a resistor to the enable lines to eliminate that for sure
07:18 PM lcnc-relay: <ccatlett1984@> my drivers have a alm led on them
07:18 PM Tom_L: or tie them 'on'
07:18 PM lcnc-relay: <ccatlett1984@> for testing, i'd tie them on
07:18 PM roycroft: but even when i slow the speed down to about 1/1000 the normal jog speed it does this
07:19 PM roycroft: it just takes a very long time
07:19 PM lcnc-relay: <ccatlett1984@> my drivers, enable when NO current is on the enable pins
07:19 PM roycroft: but i run it so slow it would seem unlikely that it's missing steps
07:19 PM Tom_L: if it expects to see a particular timing it would
07:20 PM Tom_L: what drivers are they?
07:20 PM roycroft: when the axis stops but linuxcnc is still sending i can easily turn the axis by hand
07:20 PM roycroft: tb6600
07:20 PM roycroft: i think that's what they are
07:20 PM roycroft: i forget :)
07:21 PM roycroft: yes, that's what they are
07:21 PM Tom_L: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing
07:21 PM Tom_L: last entry
07:22 PM xxcoder: yeah. I found post that says 5k 5k 20k 20k works
07:23 PM Tom_L: what is it currently
07:23 PM roycroft: that's way off what i have
07:23 PM Tom_L: change one axis and test it
07:23 PM roycroft: diresetup = 20000, dirhold = 20000, steplen = 5000, stepspace = 5000
07:23 PM roycroft: which is what xxcoder just said
07:24 PM roycroft: oh, i guess it is pretty much what the wiki says
07:24 PM roycroft: but the wiki only gives half the story
07:24 PM Tom_L: is there a way in the config to invert the signal?
07:24 PM lcnc-relay: <ccatlett1984@> # these are in nanoseconds
07:24 PM lcnc-relay: DIRSETUP = 10000
07:24 PM lcnc-relay: DIRHOLD = 10000
07:24 PM lcnc-relay: STEPLEN = 5000
07:24 PM lcnc-relay: ... long message truncated: https://jauriarts.org/_heisenbridge/media/jauriarts.org/XJlYApoDCiQbnzpGenhDbdrO/8HX8TEaUSNw (7 lines)
07:24 PM Tom_L: rising/falling edge
07:25 PM roycroft: my movie is too big to email :(
07:25 PM Tom_L: try catlett's numbers
07:25 PM Tom_L: then see if you can invert the pulse train
07:25 PM Tom_L: in the config
07:27 PM Tom_L: http://linuxcnc.org/docs/stable/html/drivers/hostmot2.html#_stepgen
07:27 PM Tom_L: invert_output
07:28 PM Tom_L: i'm not sure if that affects the pulse
07:29 PM Tom_L: mine is set to step_type 0
07:30 PM roycroft: so i changed dirsetup and dirhold to 10000, from 20000, and now it thinks it moves about 8.7" before it stops, but still not the full 9", and the axis really only moves about 4.5", like before
07:30 PM roycroft: so halving those numbers had no affect on actual axis movement, but made linuxcnc off by 50% on how far it thinks it moved
07:30 PM xxcoder: wonder if can go other way. lik 25k
07:30 PM roycroft: i'm sure i can
07:31 PM roycroft: i anticipate the axis will still move about 4.5", and linuxcnc will think it moved about 3.5" before it stops moving
07:31 PM roycroft: so i'll test that
07:32 PM lcnc-relay: <ccatlett1984@> how many steps are the drivers set to?
07:32 PM lcnc-relay: <ccatlett1984@> should have dip switches
07:32 PM roycroft: i forget
07:32 PM roycroft: let me check
07:33 PM Tom_L: that wouldn't make it stop with pulses still going out
07:33 PM Tom_L: unless it's getting them too fast
07:33 PM Tom_L: but you would hear the stepper bitch about that
07:35 PM roycroft: it's really hard to see the dip switches - i need to get a mirror and a flashlight
07:35 PM roycroft: brb
07:37 PM Tom_L: make sure the stepgen is set to type 0 in the hal file
07:37 PM Tom_L: setp hm2_[HOSTMOT2](BOARD).0.stepgen.00.step_type 0
07:37 PM Tom_L: step_type - (u32, RW) Output format, like the step_type modparam to the software stegen(9) component. 0 = Step/Dir, 1 = Up/Down, 2 = Quadrature. In Quadrature mode (step_type=2), the stepgen outputs one complete Gray cycle (00 -> 01 -> 11 -> 10 -> 00) for each step it takes.
07:40 PM roycroft: ok, i had to pull one of the drivers to read it
07:41 PM Tom_L: if you change that, it will affect the scale number
07:41 PM roycroft: 32 microsteps
07:41 PM Tom_L: doesn't need to be that much
07:41 PM roycroft: or 6400 pulses/rev
07:41 PM xxcoder: pretty high. odd
07:42 PM Tom_L: 16 is way more than enough
07:42 PM roycroft: does it hurt to have it higher?
07:42 PM roycroft: i can set to 16
07:42 PM roycroft: then i divide the scale in half
07:42 PM xxcoder: weaker torque, and it drafts a little bit, so it isnt "more precise" past certain point
07:42 PM Tom_L: would it be half?
07:43 PM Tom_L: considering the screw, drive ratio etc
07:45 PM roycroft: i think so
07:45 PM roycroft: the scale is pulses per inch
07:46 PM xxcoder: hope it works better with lower microstep
07:46 PM roycroft: if i'm doing half the microsteps then i'm moving twice as far per pulse
07:46 PM xxcoder: franjkly I wonder why it was at 32, maybe I was testing something dunno
07:46 PM roycroft: i grabbed a dental pick when i ran out to the shop
07:46 PM Tom_L: stepper have a tooth ache?
07:46 PM roycroft: i'm glad i did - i would have had a hard time with the dip switches otherwise :)
07:46 PM roycroft: i probably set that, xxcoder
07:47 PM xxcoder: yah that was annoying lol
07:47 PM roycroft: and i think it's because i was worried about having enough power
07:47 PM roycroft: if i'm moving half as much per pulse i need less power per pulse
07:47 PM xxcoder: microstep of 1 would be most torque, but also loudest and most vibration
07:47 PM xxcoder: less torque but quieter
07:48 PM xxcoder: for higher microstep that is
07:48 PM Tom_L: check the stepgen type while you're at it
07:50 PM roycroft: setp hm2_[MESA](BOARD).0.stepgen.01.step_type 0
07:50 PM Tom_L: good
07:50 PM roycroft: and i think i am thinking backwards
07:51 PM roycroft: because i just ran y all the way back and had to e-stop it
07:51 PM Tom_L: well, estop works.
07:51 PM xxcoder: good thing you did estop first
07:52 PM roycroft: i did not want to turn the machine on even once without a working e-stop button
07:52 PM xxcoder: machine dont get damaged on hitting ends, but not good for coupler, you probably would have to retighten it after impact, or replace it
07:53 PM xxcoder: so at the worse, you destroy the coupler
07:54 PM roycroft: ok, i was just able to rapid all the way back on z - about 18"
07:54 PM roycroft: now i can test with gcode
07:55 PM roycroft: but the scale is still way off
07:55 PM Tom_L: yeah you need to put a pencil to that
07:55 PM roycroft: i think 16 microsteps has promise
07:55 PM xxcoder: if it dont just stop, it sounds like its big improvement
07:55 PM roycroft: but yes, i need to do the calcs on the scale
07:55 PM roycroft: yes
07:55 PM roycroft: i can move the full distance of the axis now
07:56 PM lcnc-relay: <ccatlett1984@> just got back, setting the steps higher. evey time you double the microsteps, you halve the holding torque at a microstep
07:56 PM roycroft: and very quickly
07:56 PM Tom_L: once you get it moving you can adjust max accel etc
07:56 PM roycroft: i need to get the scale set first
07:57 PM xxcoder: so it really dont need ena wired at all. intresting!
07:57 PM roycroft: it doesn't
07:57 PM roycroft: it defaults to enable
07:57 PM Tom_L: i do as a sanity factor
07:57 PM lcnc-relay: <ccatlett1984@> yeah, powering ENA, disabled the motors
07:57 PM roycroft: it's a bit confusing, because setting enable disables it :)
07:57 PM lcnc-relay: <skunkworks8841@> that is pretty common...
07:57 PM lcnc-relay: <skunkworks8841@> suprisingly
07:58 PM roycroft: so you have to disable enable to enable
07:58 PM xxcoder: I certainly wired it when I had it.
07:58 PM Tom_L: sounds like you're on your way then
07:58 PM roycroft: anyway, i think i made progress
07:58 PM roycroft: thans
07:58 PM roycroft: i'm going to go fix dinner
07:58 PM lcnc-relay: <skunkworks8841@> scale is pretty easy - do you know the pitch of the lead screws?
07:58 PM roycroft: and then figure out how to calcuate scale again
07:58 PM xxcoder: yep. I guess you would need fix other 2 drivers later
07:58 PM roycroft: i reset the dip switches on all 3
07:59 PM roycroft: so no worries there
07:59 PM xxcoder: HH OK
07:59 PM xxcoder: *ahh ok
07:59 PM roycroft: i really want to keep them all set the same if i can
07:59 PM roycroft: it just makes life easier
07:59 PM roycroft: i hae that info, skunkworks9941, but not at the top of my head
07:59 PM roycroft: i calculated scale before, and got it pretty close
07:59 PM Tom_L: my z scale is different than x & y because it's 3:1
08:00 PM Tom_L: i probably didn't need to do that
08:00 PM roycroft: and mine are all direct drive, with the same lead screw pitch on all of them
08:00 PM xxcoder: you might want to try faster ones like 5k 5k 10k 10k see if it even works later. the faster the better control
08:00 PM roycroft: so the scale should be the same for all of them
08:00 PM Tom_L: right
08:00 PM xxcoder: but honestly pretty sure 5k 5k 20k 20k is needed for cheapie tb6600.
08:00 PM roycroft: i'm back to 5k, 5k, 20k, 20k for now
08:00 PM roycroft: i change one variable at a time
08:01 PM roycroft: if i want to change another one i set the first one back to its original value first
08:01 PM xxcoder: wiser than me
08:01 PM roycroft: then, when i understand the effects of all the variables, i can start the real tweaking
08:02 PM * Tom_L wanders off
08:02 PM roycroft: so i'll go make dinner now, then recalculate the scale
08:02 PM roycroft: thanks again, all
08:03 PM xxcoder: welcome
08:03 PM roycroft: i'm sure i'll have other questions later on, but this helps a lot
08:03 PM xxcoder: I probably didnt help a whole lot lol
08:28 PM xxcoder: finally getting past the hot peak
08:43 PM roycroft: i think it's over here too
08:43 PM roycroft: the heat advisory and the red flag warning are both gone
08:44 PM roycroft: yeah, only 30 tomorrow
08:44 PM roycroft: and down to the upper 20s by week's end
08:45 PM xxcoder: sadly still whole day more for me
08:46 PM xxcoder: 30c predicted tomorrow
08:50 PM xxcoder: interesting. I bought really cheapie hand-cranked shredder, as I cant really afford a good on
08:51 PM xxcoder: but it easily handled all paperwork I threw to it
08:51 PM xxcoder: the biggest con is I have to tear paper in half to fit in it but yeah
08:54 PM roycroft: does it do cross shredding?
08:54 PM xxcoder: sadly no, but very fine cut
08:55 PM roycroft: so it would be hard to reconstruct documents
08:55 PM roycroft: and that is what matters
08:55 PM xxcoder: its good enough to block casual info theft, but if someones dermined to steal my 10 dollars theyre welcome to
08:56 PM xxcoder: reminds me of that funny story my dads friend told me years ago
08:57 PM xxcoder: someone did an id theft on them, and got bunch of cards and got into debt. they told police they got permission from that friend
08:57 PM xxcoder: police asked them how they got the permission, and apparently it was verbal via phone
08:57 PM xxcoder: which is impossible, dad and his friend was deaf
08:57 PM xxcoder: no spoken language in friends case
08:58 PM xxcoder: he had problems for couple years, but ultimately resolved
09:25 PM xxcoder: https://www.ebay.com/itm/146736931339 price have increased since I bought it. lol ok
09:28 PM roycroft: welcome to our national sales tax
09:29 PM xxcoder: other listings is almost as low, so I doubt it
09:30 PM roycroft: it all depends on when the stuff arrived and passed through customs
10:02 PM xxcoder: man engineer bo is still working on that scroll thingy
10:02 PM xxcoder: I still want one lol
10:21 PM lfluffywof: what thing is this?
10:23 PM xxcoder: a second
10:23 PM xxcoder: https://www.youtube.com/watch?v=tzqJ1rJURgs&t=2s
10:23 PM roycroft: is it from the dead sea?
10:24 PM xxcoder: still under development. biggest issue is macos and linux, both dont seem to allow smooth scrolling
10:25 PM lfluffywof: oh, I was assuming some physical scroll-shaped device. lol
10:25 PM xxcoder: lol
10:25 PM roycroft: why not just use a 3d mouse?
10:25 PM roycroft: that's a lot more flexible
10:25 PM xxcoder: sure but I just want to scroll
10:26 PM xxcoder: majority of my pc usage is scrolling
10:26 PM xxcoder: my mx 3s probably have spun more than my car tires last few years
10:26 PM lfluffywof: I don't even have a scroll input device.
10:26 PM xxcoder: of you have mouse with wheel, you do.
10:26 PM lfluffywof: (two-button touchpad)
10:26 PM xxcoder: ahh ok
10:26 PM lfluffywof: (without stupid corner crap)
10:27 PM roycroft: usually you can scroll on a touchpad by using two fingers
10:32 PM xxcoder: btw engineer bo have free patreon tier if want to follow the scroll thingy
10:32 PM lfluffywof: some of my other laptops have that. I hate it immensely. heh.
10:39 PM lfluffywof: I should be doing useful things, but instead I'm really tempted to try fitting a 3d rendering engine with map and some sprites in 2k words of ram.
11:12 PM lfluffywof: 2k 29 bit words... that's pretty tight...
11:16 PM xxcoder: I bet
11:49 PM lfluffywof: sleep time