#linuxcnc Logs
Dec 13 2022
#linuxcnc Calendar
12:02 AM -!- #linuxcnc mode set to +v by ChanServ
12:15 AM -!- #linuxcnc mode set to +v by ChanServ
12:27 AM -!- #linuxcnc mode set to +v by ChanServ
02:24 AM -!- #linuxcnc mode set to +v by ChanServ
02:41 AM -!- #linuxcnc mode set to +v by ChanServ
04:39 AM travis_farmer: G'Morning
04:42 AM Tom_L: morning
04:43 AM Deejay: moin
04:57 AM JT-Cave: morning
05:03 AM Tom_L: pouring rain & 51F
05:36 AM JT-Cave: supposed to rain here this afternoon
05:46 AM travis_farmer: 19F, with a high of 31F, and clear skies
08:33 AM JT-Shop: someone keeps asking me what the ECCN number is for a 7i77...
08:34 AM Tom_L: give them a number
08:34 AM Tom_L: it's alpha numeric
08:35 AM Tom_L: are they foreigners?
08:35 AM Tom_L: An Export Control Classification Number (ECCN) is an alphanumeric designation (e.g., 1A984 or 4A001) used in the Commerce Control List (CCL) to identify items for export control purposes.
08:37 AM Tom_L: it's possible peter might actually have one
08:37 AM Tom_L: doubtful but possible
08:37 AM JT-Shop: I'm not an exporter :)
08:38 AM Tom_L: 4A7i77
08:39 AM Tom_L: or 4AN7i77 for proper english
08:40 AM Tom_L: for an 7i77
08:41 AM JT-Shop: maybe he will quit bugging me now
08:42 AM pcw---home: Yes, I can look up what we use
08:44 AM solarwind: I have a few 2018 iMacs, the ones with the nice beautiful 27" display
08:44 AM solarwind: I wonder if I can repurpose them to CNC machine controllers haha
08:44 AM cpresser: JT-Shop: you can also google for comparable products. Eg. graphics cards or similar.
08:44 AM pcw---home: I think the only exceptions to the generic number we use is for cabled PCIE systems and industrial temperature rated cards
08:44 AM solarwind: I think a thunderbolt to PCIe bridge should work just fine with the mesa cards, right pcw---home ?
08:45 AM pcw---home: maybe...
08:45 AM cpresser: JT-Shop: I did not know that ECCN exists, as a european, I only know of HTS numbers.
08:45 AM solarwind: pcw---home well as long as the cards are PCIe compliant
08:45 AM solarwind: they'll be running Linux
08:45 AM pcw---home: and the BIOS...
08:45 AM solarwind: what do we need from the BIOS?
08:45 AM cpresser: JT-Shop: those numbers can usually be found on digikey as a reference. See here: https://www.digikey.com/en/products/detail/texas-instruments/LM27762DSSR/6235195
08:46 AM pcw---home: card discovery
08:46 AM solarwind: well I don't think there will be an issue
08:46 AM solarwind: It's just another PCIe card
08:46 AM pcw---home: resource assignment
08:46 AM solarwind: hmmmmm ok I'll have to get an enclosure and test
08:48 AM solarwind: I know for a fact people use external audio interfaces, video cards, network cards this way
08:48 AM solarwind: hopefully this won't be an issue
09:13 AM Deejay: re
09:36 AM travis_f[SHOP]: ok, i added a 560ohm resistor between RS485 A+ and 5v, and B- and gnd, along with the 120ohm termination resistors at both ends of the RS485 wire, as recommend by user rmu. but still the signal gets noisy as soon as the VFD starts the spindle, and stops transmitting from the USB to RS485 device. so the interference from the VFD running is still canceling the data signal. not sure how else i can shield it...
09:37 AM zephyr9900: Hmm so what brand VFD do you have?
09:37 AM travis_f[SHOP]: HY (can't remember how to spell it)
09:37 AM zephyr9900: I can show you how mine is set up
09:38 AM travis_f[SHOP]: mine used to work, i just don't understand why now it does not
09:39 AM rmu: travis_f[SHOP]: how do you measure the noise?
09:40 AM travis_f[SHOP]: with my oscilloscope, gnd lead to B- and probe to A+
09:41 AM rmu: termination doesn't hurt, but you won't need it with 9600bps and "short" cable lengths
09:41 AM rmu: travis_f[SHOP]: put GND lead of oscilloscope to GND one probe to A+ and one to B- and use difference
09:41 AM travis_f[SHOP]: 1m cable length, but without the resistor i get no control at all
09:42 AM travis_f[SHOP]: hang on, will take a bit to setup
09:43 AM rmu: grounding the B- with the oscilloscope can introduce a weird loop via protective earth and make your differential signal effectively single ended
09:46 AM rmu: termination should not make a difference if your bit time is way longer than the length of the cable, 1ns is 30cm / 1ft in free air or around 20cm in cable, depending on cable and phase of moon.
09:51 AM travis_f[SHOP]: ok, it works fine with the scope connected, but not with the scope disconnected...
09:55 AM travis_f[SHOP]: seriously, the VFD responds fine, as it should with the scope connected as you stated, but if i unplug the probes, it fails...
09:59 AM travis_f[SHOP]: and now it doesn't... this is very strange...
10:00 AM zephyr9900: Here is how my VFD is mounted an connected https://photos.app.goo.gl/kARXfaK4mH6uSK8S8
10:01 AM sensille: my 6040 has a "nowfoerever" VFD, probably e-100. what is the best way to connected to linuxcnc? i just the modbus enough?
10:01 AM zephyr9900: VFD Terminals https://photos.app.goo.gl/wSLzAipPLU7UUQL5A
10:02 AM rmu: travis_f[SHOP]: check your connections, esp. grounding. can you post a picture?
10:02 AM zephyr9900: connection at computer https://photos.app.goo.gl/iRKKcwhNLAwRHtEn9
10:03 AM travis_f[SHOP]: just a sec...
10:05 AM zephyr9900: sensille I only have modbus connection but you will need to test with the "nowfoerever" VFD to see of you can get it to work with one of the hal VFD components.
10:08 AM roycroft: zephyr9900: what is the problem? are you getting interference when the vfd is running?
10:08 AM zephyr9900: not me but travis_f[SHOP] is
10:09 AM zephyr9900: I was just showing him how I connected mine it works fine
10:09 AM roycroft: oh, ok
10:10 AM roycroft: well, for travis_f
10:10 AM travis_f[SHOP]: here are my RS485 connections https://www.tjfhome.net/community/index.php?threads/temp-images.43/#post-134
10:10 AM roycroft: rs-485 is a differential signal
10:10 AM sensille: at a first glance it looks like i can control it completely via modbus. nice, it even has a relay, probably to control the pump
10:10 AM roycroft: it will benefit greatly if the a and b signals are a twisted pair
10:11 AM roycroft: it will also benefit if the communications cable is shielded, with the shield connected only at the machine end, and floating at the computer end
10:11 AM rmu: travis_f[SHOP]: black wire is ground? it is not connected on the VFD side
10:11 AM travis_f[SHOP]: hmmm, using the shield as my ground reference...
10:12 AM travis_f[SHOP]: is no black wire
10:12 AM rmu: sorry i confused the cables
10:12 AM travis_f[SHOP]: bad picture...
10:13 AM travis_f[SHOP]: orange is A+, and orange/white is B-, gnd is the cable shield wire
10:14 AM travis_f[SHOP]: the ground on the VFD is connected to DCM (digital common)
10:14 AM rmu: you sure DCM is the GND for rs485 and not ACM?
10:14 AM roycroft: i would use shielded cat 6 ethernet cable that
10:14 AM travis_f[SHOP]: tried both
10:14 AM rmu: strange.
10:15 AM travis_f[SHOP]: it is shielded CAT5e cable
10:15 AM roycroft: pick one pair for a and b, pick one random wire for ground, and make sure the shield floats at the cpu end and is connected at the vfd end
10:15 AM roycroft: cat5e should be fine
10:15 AM roycroft: rs-485 is an extremely robust protocol
10:16 AM rmu: the cable is ok? maybe you have a broken conductor?
10:16 AM travis_f[SHOP]: but i will change the ground so it is not the shield
10:16 AM roycroft: you should be able to run cables for miles in fairly noisy conditions and it should still work
10:17 AM travis_f[SHOP]: let me add a image for the scope...
10:18 AM travis_f[SHOP]: same image link as before, refresh and scroll down to see the scope image. my scope doesn't seem to do diffrential
10:19 AM rmu: the green trace should be more or less the inverse of the yellow trace
10:19 AM travis_f[SHOP]: i know, but it is not
10:20 AM travis_f[SHOP]: green in B-
10:20 AM travis_f[SHOP]: is, not in
10:21 AM rmu: you are sure you don't have some short from B- to GND or common? you could check with ohmmeter
10:21 AM travis_f[SHOP]: one sec
10:22 AM norias: oi
10:23 AM rmu: travis_f[SHOP]: M/R button, math setting, subtract should do it ;)
10:24 AM travis_f[SHOP]: 10.4 M ohms gnd to A+, 580 ohms B- to gnd
10:24 AM rmu: http://www.hantek.com/Product/DSO8060/DSO8060_Manual_EN.pdf pg 30
10:25 AM rmu: i'm out of ideas, i would probably play around and look if communicaton from PC to VFD looks different than the other direction, check scope leads, check cable for continuity, and then start to rip my hair out ;)
10:26 AM zephyr9900: If you have 10.4 Mohms to A+ your termination resistor is not connected.
10:27 AM travis_f[SHOP]: hmmm
10:28 AM rmu: correct, should read failsafe b- to gnd plus termination || termination, 60 ohms IIRC
10:28 AM travis_f[SHOP]: resistor is connected... maybe one went bad...
10:28 AM rmu: so it should read 640ohms
10:28 AM zephyr9900: you can also check the ground terminal which should be DC Common and see if it is connected to the chassis or earth ground in the VFD. If the DC common of the VFD is floating (like it is in mine) I think it will be a source of common mode noise when the VFD is running.
10:29 AM travis_f[SHOP]: hmmm...
10:30 AM zephyr9900: That might be why it works when the scope is connected.
10:31 AM rmu: common mode noise on A and B shouldn't matter? at least if it is within limits allowed by transceivers, that should be -7 to +12 V IIRC but i may misremember
10:31 AM rmu: seems to be a portable scope
10:31 AM travis_f[SHOP]: lot less noise on acm
10:31 AM rmu: travis_f[SHOP]: is the scope plugged into an outlet or is it running from battery?
10:31 AM roycroft: how close is the control computer to the vfd?
10:31 AM travis_f[SHOP]: battery
10:32 AM roycroft: if close enough, check the ground potential difference between the two with a voltmeter
10:32 AM travis_f[SHOP]: 3feet
10:32 AM roycroft: if you have a significant ground potential difference (more than a few volts) you could be getting ground interference
10:33 AM sensille: what is the best way to connect the vfd/modbus to the pc? through the fpga card or with a usb dongle?
10:35 AM zephyr9900: sensille for me I went right to the computer because id don't have any UARTs in my FGPA I/O so i think a USB dongle should work also
10:35 AM travis_f[SHOP]: .6mv
10:36 AM zephyr9900: rmu common mode noise can be converted to differential if the lines are not balanced and twisted.
10:37 AM zephyr9900: but I think you already knew that :-)
10:38 AM rmu: it's twisted pair cat5, should be good enough. my guess is broken conductor in cable or something like that
10:38 AM zephyr9900: we don't really know what is in the RS485 box and the VFD
10:38 AM roycroft: and i think he was using the shield as sc
10:38 AM roycroft: i don't know if he's fixed that yet or not
10:39 AM travis_f[SHOP]: with the scope not connected, i get gnd to A+ = open, gnd to B- = open... i think i am seeing something here...
10:39 AM rmu: the 10.4mohms when it should be 640 clearly shows that something isn't connected quite right
10:40 AM rmu: heisenbugs are the best ;)
10:40 AM travis_f[SHOP]: well, i gotta get lunch. will close out here, and return later
11:11 AM roycroft: i just had a momentary panic attack
11:11 AM roycroft: i got my daily digest of incoming mail from the usps
11:12 AM roycroft: the first item was a big, thick letter from the internal revenue service
11:12 AM roycroft: i'm thinking "oh, i screwed up my taxes. i'm in trouble"
11:12 AM roycroft: but later on down the digest was a letter from the us treasury, and i could read "pay to the order of" through the address window
11:13 AM roycroft: so i think i screwed up the good way
11:13 AM travis_farmer: maybe they sent you my tax return... i still haven't gotten it...
11:14 AM roycroft: i didn't file until october
11:14 AM roycroft: i paid my extimated taxes in april, and filed for an automatic extension
11:14 AM roycroft: then finished things up over the summer
11:14 AM roycroft: so this is a fairly quick turn-around, especially since it was not a super-simple return
11:15 AM travis_farmer: been waiting over 6 months on mine...
11:15 AM roycroft: i thought there was some rule about amortizing capital gains over multiple years, so i wonder if that's what this is about
11:16 AM roycroft: actually, i was also right on the threshold for getting a stimulus check - like a couple hundred dollars over the limit
11:16 AM roycroft: so a really small error could have made me eligible for that
11:16 AM roycroft: i guess i'll find out when i go to the post office today
11:16 AM roycroft: i thought i was going to dread that trip. but maybe not
11:17 AM roycroft: actually, i will dread it anyway, because it's still holiday mailing madness time, and i have to send a certified letter today, so i need to queue up with all the crazy people
11:18 AM roycroft: but if i have a nice check from the treasury department in my hand i won't mind the queue so much, so i'll pick up my mail before sending the letter
11:21 AM roycroft: i've been debating whether to build the rack and pinion drives for my cnc router or just buy them pre-built from avidcnc
11:21 AM roycroft: if the check is for a decent amount i think i'll just buy them and be done with it
11:21 AM roycroft: that would save a fair amount of fiddly machining
11:22 AM roycroft: https://www.avidcnc.com/pro-rack-and-pinion-drive-nema-23-p-227.html
11:22 AM roycroft: that's what i'm looking at
11:22 AM roycroft: and by the time i buy the gears and belt and bearings, i would not save that much money making them myself anyway
11:23 AM roycroft: also, my stepper motors have 10mm shafts, and finding gears that fit those shafts is not easy
11:24 AM roycroft: i also need one more motor than i have, so i might just buy a pair from avidcnc with the 3/8" shaft that fits the gears they provide
11:37 AM sensille: "tram" in cnc context in the mesa source was a bit misleading for me at first :)
11:53 AM zephyr9900: I never looked at that source. What does tram mean in that context?
11:53 AM sensille: i can only guess. transaction RAM?
11:54 AM zephyr9900: Oh OK that might fit.
11:56 AM * travis_farmer is contemplating his plan of attack when he heads out to the shop. will likely completely re-make the RS485 cable, and see if it improves...
11:59 AM * zephyr9900 goes from shop to house for lunch
11:59 AM CaptHindsight[m]: travis_farmer: how much of it works so far?
11:59 AM travis_farmer: how much of what Capt?
12:00 PM CaptHindsight[m]: rs485 project
12:00 PM travis_farmer: it is spotty at best. thinking the current wire has a break in it
12:02 PM unterhaus: solidworks wants $20 for shipping, but nothing for the d/l version. Unfortunately you can't choose the d/l version
12:04 PM * travis_farmer heads out to shop...
12:12 PM sensille: question regarding hm2_eth.c: i want to make sure outgoing packets have a minimum time between them. can i do a blocking (hires) sleep? what would be an appropiate function? or do i have to schedule a timer?
12:12 PM sensille: my card can't handle the full gbit rate, and on startup the driver sends lots of packets back to back
12:21 PM JT-Cave: so debian mate has this new very annoying feature that when you set focus to a second application it dims the other open application
12:21 PM JT-Cave: it's so annoying to try and read one file and work in another file with these stupid feature
12:26 PM CaptHindsight[m]: it's in XFCE as well
12:31 PM CaptHindsight[m]: JT-Cave: you can set in ON or OFF, forget where though :)
12:34 PM CaptHindsight[m]: JT-Cave: look in window manager tweaks
12:34 PM CaptHindsight[m]: Opacity of inactive windows
12:34 PM CaptHindsight[m]: slide to full opaque
12:37 PM travis_f[SHOP]: ok, update on the RS485 issue: new cable, orange pair for RS485, green wire for ground, shield connected at one end only (VFD). 560 ohm from gnd to B-, and from A+ to 5v, and 120 ohm term at both ends of a 3m hunk of cable. still, the VFD interference overrides the signal and causes it to shut down communication.
12:38 PM zephyr9900: sensille I don't know what you are doing but why not just tell you card's phy or PC's phy to not use 1G only 100M?
12:38 PM JT-Shop2: ok I need to install the mate tweak tool
12:40 PM Tom_L: JT-Shop2, figure it out soon so i know how to fix mine :)
12:41 PM Tom_L: i actually haven't noticed it yet
12:48 PM travis_f[SHOP]: is it possible something broke in my VFD, causing the interference and comm shutdown? the waveform while the VFD is running the spindle is very spikey.
12:49 PM JT-Shop: travis_f[SHOP], is the control cable to the vfd shielded?
12:50 PM travis_f[SHOP]: yes
12:50 PM travis_f[SHOP]: shielded twisted pair CAT5e
12:50 PM CaptHindsight[m]: JT-Shop2: i tried mine at 50-70% opacity for a while and it got annoying for the exact reason that you found
12:50 PM JT-Shop: shield grounded on one end?
12:50 PM travis_f[SHOP]: yes
12:51 PM JT-Shop: that's all I had, if you ground both ends you create a current loop...
12:51 PM * JT-Shop goes back to work
12:51 PM travis_f[SHOP]: yeah, thanks anyway
12:51 PM travis_f[SHOP]: :-)
12:51 PM CaptHindsight[m]: multitaskers will understand
12:52 PM CaptHindsight[m]: I'm surprised yet again that the default is in annoying mode
01:01 PM travis_f[SHOP]: i am uploading a video to my youtube showing the scope screen, while the VFD starts running the spindle, and then shuts down. of course it worked correctly this time, but you can see the noise i speak of. let me just grab the video link from my youtube...
01:05 PM travis_f[SHOP]: youtube apparently has to check the content first... how annoying...
01:06 PM unterhaus: solidworks took my $20 and now it's a mystery if they are going to do anything else
01:07 PM Tom_L: that's all they really wanted
01:12 PM travis_f[SHOP]: https://youtube.com/shorts/APEPmVYmpjo
01:15 PM zephyr9900: travis_f so does your VFD have a ground pin on the terminal block for the I/O? if so check to see if that ground is floating with respect to earth and or your DC common. If it is you may need to connect it in one place to your DC common for your system and then in one place your DC common should connect to earth.
01:16 PM travis_f[SHOP]: checking...
01:17 PM unterhaus: all of a sudden, conduit clamps are pvc instead of cheap galvanized steel
01:17 PM travis_f[SHOP]: hmmm... floating
01:19 PM unterhaus: never mind, I should have searched for "emt straps" /emilylitella
01:20 PM CaptHindsight[m]: a rural vs urban thing
01:21 PM unterhaus: there is pvc conduit, although I have never seen any in the wild
01:22 PM travis_f[SHOP]: i used PVC conduit for my underground wire run...
01:22 PM unterhaus: it's very hard to see from here though
01:22 PM travis_f[SHOP]: just squint real hard ;-)
01:25 PM travis_f[SHOP]: well, it is getting too dark to work out here (no lighting yet...) so i am closing out and will debug further another day.
01:26 PM travis_f[SHOP]: thanks all whom provided help :-)
01:26 PM CaptHindsight[m]: i recently looked at the construction quality of homes in a gated community
01:26 PM CaptHindsight[m]: they just look well made vs actually are well made
01:27 PM XXCoder: capt im sure plenty of build code violations too
01:27 PM CaptHindsight[m]: brick or stone front facades but the sides and rear are all wood frame with cheap siding
01:28 PM CaptHindsight[m]: i went back and looked at some from 20 years ago, they are already falling apart
01:29 PM XXCoder: yet we are supposed to treat housing as "investment"
01:29 PM XXCoder: maybe when houses was well built
01:29 PM XXCoder: not now
01:29 PM CaptHindsight[m]: trim loose, siding peeling etc etc. Streets need to be repaved
01:30 PM CaptHindsight[m]: why most manufactured suburban communities are in trouble after 20 years
01:31 PM CaptHindsight[m]: they need new infrastructure so the taxes are high
01:31 PM CaptHindsight[m]: the developers new how to make their money and be long gone
01:41 PM unterhaus: most of the development around here is fairly high density instead of detached homes
01:42 PM unterhaus: still so much of it that we don't get our streets plowed very quickly
01:42 PM unterhaus: solidworks finally sent me download instructions
01:44 PM pere: hi
02:14 PM roycroft: yet they still give huge tax breaks to foreign companies who promise to create good, long-lasting jobs but always reneg on those promises
02:15 PM roycroft: and not only do they give the tax breaks, they toss in all the infrastructure the foreign companies need to support their facility that they will abandon in five years
02:15 PM roycroft: causing the municipality to raise property taxes on the laid off workers to pay for it all
02:17 PM unterhaus: development is generally a bad bargain for municipalities, but they love it
02:40 PM rmu: travis_farmer: when attaching the scope probes, make sure to connect both GNDs, connect them near where you connect the signal, and in general make the area in gnd lead and probe tip as small as possible. in the video it seems you have "proper" A and B signals? hard to see. youtube doesn't let me fullscreen or zoom into the vertical video shorts. it looks mostly like common mode noise, that
02:40 PM rmu: should not affect your rs485. if the blue trace is the difference, try making it smaller, 2v/div or 5v/div. also, your rs485 lines seem quite noisy even without vfd running.
02:49 PM unterhaus: solidworks 3d experience panel is really annoying
03:09 PM JT-Shop: CaptHindsight[m], when you remember where to turn off that feature let me know
03:11 PM CaptHindsight[m]: JT-Shop: in window manager tweaks
03:11 PM CaptHindsight[m]: Opacity of inactive windows
03:11 PM CaptHindsight[m]: slide to full opaque
03:11 PM unterhaus: does laser cutting work for parts 1" thick?
03:14 PM roycroft: that depends on the parts and the laser
03:15 PM CaptHindsight[m]: unterhaus: https://www.youtube.com/watch?v=r5vsUta5lmo
03:16 PM unterhaus: I suppose I should call up diacro and ask how much the parts I want cost, although I'm already afraid
03:16 PM CaptHindsight[m]: hah I forgot that they don't heat much office or factory space
03:17 PM CaptHindsight[m]: inside office space in northern China may be the same temp as the outdoor space
03:18 PM JT-Shop2: CaptHindsight[m], what desktop are you using? can't find that in mate
03:18 PM CaptHindsight[m]: XFCE
03:18 PM CaptHindsight[m]: though I have installed more
03:19 PM CaptHindsight[m]: the app is called: Window Manager Tweaks
03:20 PM CaptHindsight[m]: the setting is in the Compositer tab
03:21 PM JT-Shop2: looks like mate does not have that as far as I can tell
03:22 PM CaptHindsight[m]: JT-Shop2: give me a minute, I'll change desktops and check
03:22 PM unterhaus: I might go back to my plan of getting 2 pieces of 1/2" thick material laser cut and bolt them together
03:22 PM JT-Shop2: ok
03:33 PM CaptHindsight[m]: JT-Shop2: Window Manager Tweaks only seems to function when running XFCE
03:33 PM CaptHindsight[m]: I even tried LXQT and LXDE
03:34 PM JT-Shop2: I wonder if gsettings can be directly changed
03:36 PM CaptHindsight[m]: JT-Shop2: is there still a "Mate Tweak"?
03:37 PM CaptHindsight[m]: https://ubuntu-mate.community/t/how-to-apply-real-transparency-on-the-panel/10863/5
03:38 PM JT-Shop2: hmm nothing there that I could see
03:40 PM Tom_L: JT-Shop2, what version linux do you have this on?
03:41 PM JT-Shop2: Debian 12 for most of the pc's this one is Debian 10
03:42 PM Tom_L: using something like geany side by side?
03:42 PM JT-Shop2: pluma side by side
03:43 PM Tom_L: i'm using geany on bookworm (12) and haven't had that problem
03:43 PM Tom_L: lemme open something with pluma
03:43 PM JT-Shop2: I just looked at geany on this pc and no issue with dimming windows not in focus
03:44 PM Tom_L: it's a pluma thing
03:44 PM JT-Shop2: maybe pluma
03:45 PM JT-Shop2: makes me not want to use pluma anymore
03:45 PM Tom_L: i use geany all the time
03:46 PM JT-Shop2: I did before the great PC crash but took a while to get all the annoying features turned off lol
03:46 PM Tom_L: right away i don't see a setting in it to change that
03:49 PM Tom_L: i don't remember if i put mate tweaks on this yet or not
03:49 PM JT-Shop2: mate tweaks only has a few things
04:17 PM CaptHindsight[m]: the Gentoo LCNC install image is ~1.2GB
04:17 PM CaptHindsight[m]: desktops are XFCE and LXQT
04:18 PM CaptHindsight[m]: and GTK + QT and all the compilers
04:23 PM JT-Shop2: raining pretty hard here now
04:23 PM Tom_L: rain quit this am and was sunny all day
04:26 PM JT-Shop2: next slow down in the rain and I'm making a dash for the machine shop
04:26 PM CaptHindsight[m]: JT-Shop2: I have Mate on Bookworm and I can't get it to have non-opaque windows
04:26 PM JT-Shop2: looks like it's a pluma thing
04:27 PM CaptHindsight[m]: I installed mate tweak and you can change the compositor or turn them all off
04:27 PM CaptHindsight[m]: I didn't see a difference
04:27 PM Tom_L: CaptHindsight[m], did you try pluma?
04:27 PM Tom_L: it does it here just with pluma
04:27 PM CaptHindsight[m]: ah so just Pluma acts that way?
04:27 PM Tom_L: seems so
04:28 PM JT-Shop2: seems that when pluma doesn't have focus it dims
04:28 PM Tom_L: it's still readable but i like geany better anyway
04:30 PM CaptHindsight[m]: JT-Shop2: yes i just installed Pluma and it acts as you say "when pluma doesn't have focus it dims"
04:30 PM Tom_L: i didn't see a setting for that in pluma either
04:31 PM JT-Shop2: very annoying when you're trying to use pluma twice
04:31 PM CaptHindsight[m]: yes
04:31 PM CaptHindsight[m]: Pluma also acts this way in XFCE
04:32 PM CaptHindsight[m]: even with windows set to be opaque when not in focus
04:32 PM CaptHindsight[m]: so pluma does its own thing
04:34 PM acer is now known as _unreal_
04:38 PM JT-Shop: sudo apt remove pluma*
04:38 PM Tom_L: :D
04:39 PM CaptHindsight[m]: good solution
04:39 PM Tom_L: if you're using a color scheme it isn't as bad
04:43 PM * JT-Shop heads inside while the rain is slack
04:49 PM JT-Cave: https://github.com/mate-desktop/pluma/issues/669
04:49 PM JT-Cave: for a good laugh
05:03 PM Tom_L: heh
05:04 PM JT-Cave: the biggest thing about geany is the lack of a undo button on the toolbar
05:04 PM Tom_L: hmm
05:05 PM Tom_L: send them a similar message
05:06 PM JT-Cave: https://github.com/geany/geany/issues
05:06 PM JT-Cave: wow 222 pr's
05:07 PM Tom_L: take a number..
05:08 PM Tom_L: you could use vi or nano :)
05:09 PM JT-Cave: preferences toolbar customize and you can add undo in place of revert
05:09 PM * JT-Cave likes nano
05:09 PM Tom_L: i'm used to ctrl-z etc
05:09 PM JT-Cave: ok I'm set redo and undo are on the menu
05:10 PM JT-Cave: when you have a glass of vino in one had the mouse is easier lol
05:10 PM Tom_L: hehe
05:10 PM JT-Cave: we're golden now with geany
05:12 PM JT-Cave: foxy brought home an automatic sanding machine the other day spent a few hours today figuring out why it would not run... had to take apart a safety door switch and twiddle with it so the insert would go in
05:12 PM JT-Cave: it had a plc I've never seen before
05:13 PM Tom_L: old?
05:13 PM Tom_L: must have been
05:13 PM JT-Cave: didn't look more than maybe 20 years old
05:14 PM JT-Cave: seems like all wood working machines are made in italy
05:14 PM Tom_L: huh
05:14 PM JT-Cave: factory type of machines for spindles and chair legs etc.
05:15 PM Tom_L: wonder if they have alot of fancy stair cases there
05:15 PM Tom_L: that would use alot of turned wood
05:15 PM Tom_L: or iron
05:16 PM JT-Cave: this machine might have made stair spindles or something else
05:17 PM JT-Cave: the only bad thing is Jonas brought it home through a rain storm so the linear rails have a bit of rust on them
05:17 PM JT-Cave: so they won't last long unless the clean them off
05:19 PM JT-Cave: the wire drawing machines we had at mid were all italian
05:28 PM Tom_L: i did go to a place here that made wire products. that was kindof interesting
05:28 PM Tom_L: well they did injection molding etc in the other plant as well
05:29 PM JT-Cave: most of my work lately is at an injection molding plant or foxy
05:29 PM Tom_L: winding & injecting the jacket on the wire
05:29 PM JT-Cave: which makes all kind of stuff
05:30 PM Tom_L: the injection mold side did alot of cosmetic stuff & medical
05:30 PM Tom_L: i was friends with a sales rep that took me thru the whole place once
05:30 PM * JT-Cave wanders upstairs to relax
06:04 PM -!- #linuxcnc mode set to +v by ChanServ
06:43 PM * zephyr9900 wants some help on correct way to use git. I cloned a repository from github. Next before I make any changes should I make a branch for my experimental stuff? Then what is next? do I push the branch back to github before making any changes or does it not matter?
07:10 PM roycroft: you don't need to push it right away
07:11 PM roycroft: if you're just going to experiment you don't even need to create a branch
07:11 PM roycroft: basically, if you think you're going to play with something and then throw it away, don't bother with a branch
07:11 PM roycroft: if you think you're going to keep some or all of your changes, then branch
07:12 PM roycroft: but you can't push your branch back until you do a commit of your branch
07:13 PM roycroft: and you don't need to commit until you've done some work that you want to keep
07:15 PM zephyr9900: right but maybe some day I will add something useful. If I do then do I merge my branch changes that I want back to my local master branch? or do I just always work in my branch and commit the changes to my branch and push back to the remote on github then request a pull for anything that is useful for the master that I cloned. I'm just unsure.
07:19 PM roycroft: you will generally want to use your branches for development, and merge any changes you want to keep back to the master branch
07:19 PM roycroft: i'm not sure exactly what you're trying to do
07:20 PM roycroft: if you're trying to make some local changes and keep them synched with a third-party repository, that can be a bit tricky for a beginner to git
07:20 PM roycroft: and unless you have commit access to the third-party repository, you're never going to do a push to the origin
07:22 PM zephyr9900: Right isn't that what the pull request is for? So t he original owner can take stuff or not.
07:23 PM roycroft: git pull merges changes from the remote repository into your local branch
07:25 PM zephyr9900: not git pull but: Pull requests let you tell others about changes you've pushed to a branch in a repository on GitHub. Once a pull request is opened, you can discuss and ...
07:25 PM zephyr9900: OK I think I know what to do Thanks
07:25 PM roycroft: cool
07:25 PM roycroft: i don't do collaborative projects using git
07:25 PM roycroft: if they are using that terminology it is confusing
07:26 PM roycroft: since there is a "git pull" command, and you say there is a git pull request" protocol that is used to advertise changes
07:27 PM roycroft: i suppose i can see how that terminology came about
07:27 PM zephyr9900: Basically I am asking someone else to pull something from my branch to there repository
07:27 PM zephyr9900: *their
07:27 PM roycroft: you announce something outwith git itself that advertises some changes you've made, and how folks can "pull" your changes from your copy of the repository if they want
07:28 PM roycroft: basically "i hacked some of the code. please come grab it and play with it - here's my address"
07:28 PM zephyr9900: So I think I just work in my branch and push it back to the remote on github and if I really want to mess around with stuff I should make another branch from my branch that I can throw away
07:29 PM roycroft: IF you have commit access you can push your changes to the origin on github
07:29 PM zephyr9900: I don't and wont have push access
07:29 PM roycroft: otherwise your "git push origin master" will fail
07:30 PM roycroft: or "git push origin mybranch" even
07:30 PM roycroft: you just work in your branch
07:30 PM roycroft: you merge your branch into your local master
07:30 PM zephyr9900: Right I think I got it
07:30 PM roycroft: if you want
07:30 PM roycroft: but you don't try to push to origin
07:31 PM zephyr9900: thanks
07:31 PM roycroft: if the master changes on the origin
07:31 PM roycroft: you still have your branch
07:31 PM roycroft: so you git pull from the origin to get the changes
07:31 PM roycroft: and then try to merge your branch into the local master again
07:32 PM -!- #linuxcnc mode set to +v by ChanServ
07:35 PM * roycroft has not had a chance to check the news of the day all day so far
07:35 PM roycroft: the world could have ended today and i'd not have known
07:35 PM zephyr9900: you would know because the IRC would stop
07:35 PM roycroft: but no, it looks like i did not miss anything important
07:36 PM roycroft: but would i know that for sure?
07:36 PM roycroft: it could just appear to be very very lagged
08:38 PM Plaid1: So I'm troubleshooting an old machine as I convert it to LCNC. It's got closed loop servos on it. I'm getting quadature count errors on one servo, on about 50% of my jogs, but not constantly. I'll switch around my X and Y axis and re-test, but assuming that continues, Any thoughts on what troubleshooting I should try, or is the encoder pretty much dead?
08:38 PM Plaid1: When I jog that axis, it is always flakey, and the encoder almost always loses counts in the -X direction, I don't think I've seen it do it in the +X direction
08:39 PM Plaid1: And my homing migrates towards once it's homed.
08:40 PM Plaid1: Is this a classic description of encoder failure? miswiring? else?
08:52 PM Plaid1: I understand what the error is (it's getting an invalid change back from the encoder in one cycle, both a&b changing at the same time) I'm not going too fast for the encoder, I'm seeing this problem happen even on short-low speed jogs.
08:52 PM Plaid1: It's tending to only happen in one direction, and it's kind of bursty...
09:11 PM Tom_L: check for broken wires? bad plug etc
09:12 PM Tom_L: put a scope on it if you have one
09:13 PM Plaid1: no scope
09:14 PM Tom_L: being closed loop, lcnc doesn't see the encoder counts does it?
09:15 PM Plaid1: My encodeer is connected via a 7i77, and that sends the signals back to LCNC. But the board does the counting
09:16 PM Tom_L: check the signal in halscope
09:17 PM Plaid1: That'll only report on/off
09:17 PM Tom_L: if it's missing steps it should show
09:18 PM Plaid1: I know it's missing steps. It's more a question of what could be causing it? Is it likely that the encoder failed, or bad wiring/shielding somewhere?
09:19 PM Tom_L: i'd start by checking all the connections etc for bad plugs, cold solder joints if any etc
09:19 PM Tom_L: broken wires...
09:19 PM Tom_L: if it's shielding that would show as noise on halscope
09:20 PM Tom_L: i would think
09:20 PM Plaid1: You got me thinking. Yeah. It'd more likely be bad wiring/shielding
09:20 PM Plaid1: Like... it didn't fail compleatly, so ti's going to be something like analog noise
10:11 PM Plaid1: Humm. Seems like something noise related. Really wish I had a scope :P
10:13 PM Plaid1: I did a quick test of unplugging the servo amps, and moving the screw by hand, and didn't get any errors. But I think I need to do it more to make it a better test
10:31 PM solarwind: CaptHindsight[m] I wonder with 8N•m NEMA 34 closed loop steppers can do for a 3D printer
10:31 PM solarwind: I think I can get some insane acceleration if I find the right torque/rotor inertia optimized motors
10:31 PM solarwind: Everyone is using NEMA 17
10:37 PM Plaid1: Heh. I've seen videos of people putting nema 34 on a CNC bed. Quite funny.
10:38 PM XXCoder: depends on size of bed ans weight lol
10:38 PM Plaid1: I've seen it on like a 200mm x 200mm Just as funny as you'd expect
10:39 PM Plaid1: The machine jumps around.
11:40 PM CaptHindsight[m]: solarwind: FF/FDM is pretty limited, think about printers for RF electronic parts, antennas for 5G and 6G etc https://3dprint.com/296369/lockheed-martin-invests-in-fortify-for-electronics-3d-printing/amp/
11:41 PM CaptHindsight[m]: https://3dprint.com/296238/nextflex-announces-8-45m-in-funding-for-semiconductors-and-electronics-3d-printing/
11:43 PM CaptHindsight[m]: https://www.youtube.com/watch?v=4kwhjBtzgX8
11:45 PM XXCoder: interesting
11:51 PM Plaid1: So, learned lesson... I tried several things, I reseated a connector and the problem went from "Bad, my machine is almost non-functional" to "The software detectects the error, but I can barely see it." I'll try replacing the connector later, and see if that eliminates the problem.