#linuxcnc Logs
Mar 18 2023
#linuxcnc Calendar
12:00 AM iggw2795[m]: What pinout should i use to wire up my dny4 servo driver to the measa 7i96s
12:10 AM iggw2795[m]: ?
12:16 AM mits[m]: What do people use to provide motion when using hal scope?
12:17 AM pcw-home: iggw2795[m] If the drive has Opto coupler inputs I would wire them single ended, if it has line receiver inputs, I would wire it differentially
12:17 AM mits[m]: like if I'm wanting to tune FERROR
12:32 AM * iggw2795[m] uploaded an image: (433KiB) < https://libera.ems.host/_matrix/media/v3/download/jauriarts.org/lxglldCbONBgoePfrZYClcvv/image.png >
12:32 AM iggw2795[m]: They sent me this
12:32 AM iggw2795[m]: The problem is that the pin names on this dont line up to the names on my 7i96s
12:33 AM iggw2795[m]: Like what pin should DIR+ be
12:33 AM iggw2795[m]: Also not sure about DIR-, ENA+, and ENA-
12:44 AM ZincBoy[CAON][m]: The 7i96s does not look to have a spindle interface on it. You will need to use field IO to enable the drive and control direction. There does also not seem to be an analog out so you will need another method of controlling the drive speed. You could use one of the step generators to control the speed by generating pulses but I haven't done this. The encoder is easy as that is present on your card.
12:46 AM ZincBoy[CAON][m]: I use a simple gcode program that just moves back and forth with delays between each motion. You can just do an infinite loop and stop the program when done.
03:09 AM menace is now known as Deknos
05:24 AM -!- #linuxcnc mode set to +v by ChanServ
05:52 AM Tom_L: morning
05:54 AM XXCoder: oh 30 days finally arrived since that first enmasse join
06:00 AM Deejay: moin
06:22 AM Contract_Pilot: Morning...
06:37 AM travis_farmer[m]: G'Morning ☕️ (been up since 12:45am)
07:06 AM TurBoss: morning 🌲
07:57 AM menace is now known as Deknos
08:25 AM JT[m]: Morning
08:32 AM JT[m]: Starlink router shipped Friday but shows delayed delivery until Tuesday
09:59 AM iggw2795[m]: How would I set this up?
10:33 AM roguish[m]1: good morning all. well, bright and sunny here, at least until tonight. then more rain for a few days. last year, Jan thru Oct we got like 4.5" rain. this year we already have over 20"............and it's only mid March. and of course people are whining and crying about too much water and snow everywhere...... fraking ingrates.
10:33 AM roguish[m]1: La Nina is gone. El Nino is forming. coool.
10:36 AM roguish[m]1: JT: Starlink will soon have competition https://www.cnbc.com/2023/03/14/amazon-first-look-project-kuiper-satellite-internet-antennas.html
10:50 AM JT[m]: Interesting
11:03 AM roguish[m]1: sitting here reviewing an MRI report for my left knee. doesn't sound good. nothing huge, just lots of little problems. all pointing towards A G E
11:04 AM roguish[m]1: last June my right knee got tears in the meniscus. had it cleaned up, but it still hurts...........
11:04 AM roguish[m]1: now the left knee is bugging me.
11:04 AM roguish[m]1: A G E
11:11 AM Tom_L: getting O L D is fun
11:14 AM roguish[m]1: better than the alternative.
11:16 AM roycroft: much of getting old is fine
11:16 AM roycroft: the warranty expiration of the physical body part not so much, though
11:17 AM Tom_L: roycroft, what did the gcode look like on that site?
11:17 AM roycroft: what site?
11:17 AM Tom_L: that viewer site i posted for you
11:17 AM roycroft: oh, that one
11:18 AM roycroft: i did not see a way to upload it, but i could copy and paste it
11:18 AM roycroft: i did so for the first couple hundred lines, and it displayed nothing
11:18 AM Tom_L: i uploaded a file to test it
11:18 AM roycroft: hmm, i did not see an upload button, but after dinner last night i wasn't terribly interested any more
11:18 AM roycroft: i might take another look at it today
11:19 AM roycroft: but i'm more inclined at this point to install a trial version of lightburn and see how that goes
11:19 AM roycroft: but it's sunny today, and it's going to get up to 19 degrees
11:19 AM roycroft: i'll be spending most of the day doing chores outdoors
11:29 AM roycroft: lightburn has a linuxcnc profile built-in
11:29 AM roycroft: it does not let me set the origin where it needs to be for my device, though
11:29 AM roycroft: i'll have to figure out a way to work around that
11:34 AM roycroft: my device has the origin in the middle, which is not an option with lightburn
11:34 AM roycroft: but i suppose i could set the origin to the upper right corner, and reset it to -50, -50 at the beginning of the gcode to move it to the middle
11:36 AM roycroft: it's odd that lightburn doesn't have that option, as there are a lot of galvonic mirror laser devices, and setting the origin in the middle is pretty standard for them
11:38 AM Tom_L: is lightburn windows or linux?
11:43 AM roguish[m]1: roycroft: you maybe could use a G92 to offset the coord system
11:44 AM Tom_L: i'm betting on a limited gcode set on that thing
11:52 AM roycroft: windows, linux, and mac os
11:52 AM roycroft: for lightburn
11:53 AM roycroft: the laserpecker is marilin-based, but it's probably stripped down and has some proprietary tweaks
11:54 AM Tom_L: afik marlin is based on grbl
11:58 AM roycroft: it's partially based on grbl
11:58 AM roycroft: but who knows what the laserpecker did with the marlin code
11:59 AM roycroft: the test i did last night i set to run two passes
12:00 PM roycroft: i.e. the same code run twice, to burn deeper
12:00 PM roycroft: and when it was done i could see a very faint outline of the text, so it is trying to engrave the text, but the power setting is wrong
12:00 PM roycroft: and it's engraving the text in the correct place
12:01 PM CaptHindsight[m]: roguish: satellite internet is looking like cell service only with space junk vs towers
12:01 PM CaptHindsight[m]: I wonder how many providers there will be in 10 years
12:01 PM roycroft: 1
12:02 PM CaptHindsight[m]: with 5,000 satellites each
12:02 PM roycroft: *sigh*
12:02 PM roycroft: the former president has announced that he'll be arrested on tuesday, and is calling for people to "protest" and "take our nation back"
12:02 PM CaptHindsight[m]: but with direct connection from phone to satellite
12:02 PM roycroft: so we may have another january 6 on tuesday
12:03 PM CaptHindsight[m]: sounds political
12:03 PM roycroft: no, none of that has anything to do with politics
12:03 PM CaptHindsight[m]: he can still run for office from jail
12:03 PM roycroft: it's about idolatry
12:04 PM CaptHindsight[m]: does the pope shit in the woods?
12:04 PM roycroft: jim traficant ran for congress from jail
12:04 PM roycroft: he lost bigly, but he did run from jail
12:05 PM roycroft: he was released before the election, though
12:05 PM CaptHindsight[m]: tell me what to think
12:06 PM roycroft: anyway, i'll get back to gcode now :)
12:34 PM mits[m]: Does anyone here drive a laser head PWM from a mesa card, do you have any suggestions on safeties?
12:35 PM * Loetmichel walks in the kitchen, puts a mug under the Saeco machin *button press* *RUUUIIIIIIIII* *clackclack* *clackclack* *miiiieeerrrrk* *prrrrrrr* *prrrrrrrrrrrrrrrrrrrrrrrrrrr* *miiiiiiiifubb* ... adding caramel syrup, adding milk... *sip* "aahh, guuuuud!"
12:35 PM mits[m]: I'm trying to find if there's an easy way to add a safety for if the watchdog to kill the PWM output... or if it actually will just do that
12:35 PM mits[m]: I would like to be reassured that I'm not gonna end up burning holes in things if I lock up
12:39 PM Tom_L: https://linuxcnc.org/docs/html/man/man9/watchdog.9.html
12:39 PM mits[m]: isn't this for parport output and the like?
12:41 PM Tom_L: setp hm2_[HOSTMOT2](BOARD).0.watchdog.timeout_ns 10000000
12:41 PM Tom_L: read about it in the hm2 man page i suppose
12:41 PM mits[m]: I'm asking if when the watchdog fires does it set pins to a defined behavior, and how do you set those
12:42 PM Tom_L: http://linuxcnc.org/docs/stable/html/drivers/hostmot2.html#_watchdog
12:42 PM mits[m]: mmmk
12:44 PM pcw-home: You should verify that PWM is off when linuxCNC is not running
12:44 PM mits[m]: you mean just check the pin?
12:44 PM mits[m]: that's neat if that's the simple answer, thank you
12:44 PM pcw-home: bare FPGA pins will be high at power up and after a watchdog fault
12:45 PM pcw-home: so that state should be safe
12:45 PM mits[m]: ok
12:46 PM pcw-home: if you have a card like a 7I96/7I96S/7I76/7I76E and are using a step/dir pin for Active high PWM. you should choose the - pin
12:46 PM pcw-home: and then invert the PWM polarity in the hal file
12:47 PM mits[m]: oh, hmm, that's a good idea then
12:47 PM pcw-home: so its low at power up and after a watchdog bite
12:47 PM mits[m]: pcw-home I don't mean to trouble you, but I sent an email earlier in the week about repairing a 7i76e and 7i89 to you at work and didn't get a reply... did I maybe end up in the spam folder?
12:48 PM Loetmichel: ehm. A PWM pin for spindle RPM or a PWM pin for driver unlock is supposed to be AC-coupled so that it will NOT engage regardless if the pin is at static hi or static low
12:48 PM Loetmichel: its only active if the pin toggles regularily
12:48 PM pcw-home: Possible plus I have been really busy the last week
12:48 PM mits[m]: but thanks for the - pin suggestion, I didn't even think of using it as just a straight up inverted output
12:48 PM mits[m]: even though that's logical and correct
12:49 PM mits[m]: that's fine then.. it's not critical and also I don't expect an answer
12:49 PM pcw-home: whats wrong with the cards?
12:50 PM mits[m]: I plugged them together with a pin 1 -> pin 1 style cable... it popped a couple chips and I can't read the part number on them to replace them
12:50 PM mits[m]: this was an accident, of course
12:51 PM mits[m]: sorry I mean pin 1 - 25 -> pin 1 - 25... not interleaved like it should be
12:51 PM pcw-home: Ahh must have reversed the 5V power
12:51 PM mits[m]: yes
12:51 PM mits[m]: well, something
12:52 PM mits[m]: it blew 3 chips on the 7i89 and 1 on the 7i76e
12:52 PM mits[m]: 2 of 4 I can read the part number to repair
12:52 PM mits[m]: oh... actually I wonder if there's hires images on google....
12:53 PM pcw-home: I can send you BOMs
12:54 PM mits[m]: that would be great, I'll reply to the email and ask for the BOM directly
01:05 PM mits[m]: oh, super thanks... I was thinking you'd get to it next week
01:09 PM pcw-home: well stuck home today, bad cold, otherwise probably would working in the garden as there is no rain today
01:20 PM -!- #linuxcnc mode set to +v by ChanServ
01:22 PM mits[m]: oh jeeze, it was the same part on both boards
01:23 PM mits[m]: that explains why the crater was in the same spot
01:39 PM skunkworks[m]: I have not blown a mesa card yet. which is surprising.. for me..
01:39 PM skunkworks[m]: I guess I never blew out a printer port either..
01:39 PM JT[m]: LOL
01:42 PM skunkworks[m]: Had to grind a new boring bar for the non-circular boring. Lol - took me a while to figure out how to set it up.. There is a parameter that is toolang that is to set the angle between the index and tool tip..
01:42 PM skunkworks[m]: (needed a longer one to make the ball nut assembly.
01:43 PM skunkworks[m]: I think I got it though..
01:45 PM skunkworks[m]: the effective tool diameter is almost twice as big.. the actual tool path to make the D shape is interesting.. 😉
01:48 PM mits[m]: skunkworks (@_discord_1020005159679578132:jauriarts.org) I was curious what motion is happening... because I can't tell from the video
01:49 PM mits[m]: does the spindle move, or is the table moving
01:49 PM mits[m]: not that it matters, but I just couldn't contextualize it properly
01:49 PM travis_farmer[m]: with all my freakish wiring, i have yet to blow a Mesa board, and hope i never do ;-)
01:49 PM * skunkworks[m] uploaded an image: (187KiB) < https://libera.ems.host/_matrix/media/v3/download/jauriarts.org/EWJfAoiiZzEGHuAxitSshBnk/PXL_20230318_170246218_exported_0.jpg >
01:49 PM mits[m]: ah
01:50 PM mits[m]: the pacman toolpath
01:51 PM mits[m]: skunkworks I mean is the camera fixed to the workpiece or to the spindle... because the other one is moving 🙂
01:51 PM skunkworks[m]: the table moves (table is x/y spindle is z)
01:51 PM mits[m]: ahh
01:51 PM mits[m]: ah, so camera was on the table too
01:51 PM skunkworks[m]: I was holding the camera. The actual motion is pretty small..
01:51 PM mits[m]: OH!
01:51 PM mits[m]: okay
01:53 PM mits[m]: skunkworks how did you calculate the toolpath?
01:56 PM skunkworks[m]: math?
01:56 PM mits[m]: okay, not from a software
01:56 PM skunkworks[m]: it is hal component that calculates a polygon or d shape or an actual scanned in shape..
01:57 PM skunkworks[m]: in effect - spindle synced motion
01:59 PM skunkworks[m]: this isn't the latest version.. but pretty close. https://electronicsam.com/images/greenmachine/config/polygon.comp
02:00 PM skunkworks[m]: I don't think it has the scan functionality in it.
02:02 PM mits[m]: it's perfect for reading
02:02 PM mits[m]: thanks
02:12 PM andypugh: pcw-home: I am a bit confused by the PktUART regmap
02:13 PM andypugh: Bits 20..16 Frames to send RO
02:13 PM andypugh: Bit 17 Parity enable WO
02:13 PM andypugh: Bit 18 Odd Parity WO (1=odd, 0=even)
02:13 PM andypugh: Bits 15..8 InterFrame delay in bit times RW
02:14 PM andypugh: That puts the parity bits inside the frames to send. Is this an FPGA dual-port thing where reading and writing have different effects?
02:16 PM mits[m]: I'm reading conflicting information in docs:
02:16 PM mits[m]: in stepgen it says "Usually, position mode is used for machine axes. Velocity mode is reserved for unusual applications where continuous movement at some speed is desired, instead of movement to a specific position."
02:17 PM mits[m]: but then config practices are using a PID loop to control the velocity of things, even if there is no encoder
02:20 PM mits[m]: am I mistaken about the convention of using PID loop, or is the doc just misleading on what the usual use case is?
02:20 PM mits[m]: or am I misunderstanding the interaction of the two conepts
02:20 PM skunkworks[m]: velocity mode systems - using pid are more immune to latency.. (vs using position mode step gens)
02:22 PM mits[m]: that is logical, but I'm trying to comprehend when that wouldn't be the case then
02:23 PM mits[m]: extremely trivial machines?
02:23 PM skunkworks[m]: both work - and you might not have a problem normally.
02:23 PM skunkworks[m]: with both systems.
02:24 PM skunkworks[m]: Pid is just more robust.. and more complicated..
02:30 PM mits[m]: there is the hal component 'threads'
02:30 PM mits[m]: is there any alternative to that? I can't imagine what it would be, but I'm just asking if anyone knows of one
02:32 PM skunkworks[m]: rephrase the question? I don't understand what you are asking.
02:33 PM mits[m]: the hal component 'threads' creates threads... is there alternate implementations of any kind that also provide the functionality?
02:33 PM ZincBoy[CAON][m]: Why?
02:33 PM mits[m]: since it's pluggable
02:34 PM skunkworks[m]: lol - right - why? Threads creates realtime threads. that is what it does.
02:34 PM mits[m]: I'm not sure if anyone has invented some weird thing like non-cpu processing
02:35 PM skunkworks[m]: you mean outside of linuxcnc?
02:35 PM mits[m]: you could execute hal components on a non-OS controlled CPU
02:35 PM mits[m]: I mean, the answer now is obviously 'no'
02:35 PM mits[m]: I was just asking
02:36 PM mits[m]: I don't know what sort of exotics might end up in open source systems
03:07 PM roycroft: i do non-cpu processing all the time
03:07 PM roycroft: i spin the handwheels on my milll around with my hands
03:07 PM roycroft: my hands are controlled by my brain, which i think resembles pudding more than a cpu
03:08 PM XXCoder: hopefully with many wrinkles lol
03:20 PM andypugh: Can anyone decipher Manchester Encoding (?) with the naked eye? Are these the same bytes?
03:20 PM andypugh: https://photos.app.goo.gl/UUp9wKGiUB53BFVH8
03:20 PM andypugh: https://photos.app.goo.gl/zJrT1PnKnU9LryrX7
03:20 PM andypugh: Une seems to idle high, and one low…
03:22 PM roycroft: they look very similar
03:22 PM roycroft: why don't you flip one of the images along the x axis and overlay it on the other?
03:23 PM andypugh: But does it matter that the polarity seems to be different?
03:23 PM roycroft: i don't know
03:23 PM roycroft: polarity is a point of view, isn't it?
03:25 PM roycroft: the wikipedia page on manchester encoding suggests it doesn't matter
03:25 PM roycroft: that there are two conventions used
03:26 PM andypugh: I am looking at the same page, though, and think that this is not Manchester encoding
03:38 PM pcw-home: andypugh: the "Frames to send: is unrelated Read Only status information in the setup register
03:39 PM pcw-home: (normally only written once)
03:42 PM pcw-home: (its how many frames are still left to send, probably never used in synchronous protocols like modbus etc)
03:49 PM * roycroft is starting to think about finding a different laser engraver and sending this one back
04:08 PM roycroft: yes, it's going back - i found another one that will work with lightburn, and i can upgrade the laser later if i want
04:09 PM XXCoder: and hopefully not cloud locked
04:11 PM roycroft: sweet
04:11 PM roycroft: the one i found was $70 more than the laserpecker
04:11 PM roycroft: and i got a $60 coupon when i checked out
04:12 PM roycroft: i got this:
04:12 PM roycroft: https://www.amazon.com/gp/product/B0B2R9Y6Y2
04:13 PM roycroft: it's already 10x more powerful than the laserpecker, and i can upgrade from a 5w to a 20w laser later if i want
04:13 PM roycroft: i really liked the idea of the small, tripod mount device, but i don't have time to mess with it
04:14 PM roycroft: the new machine will be here thursday
04:14 PM XXCoder: hopefully that ones not locked to cloud too
04:14 PM XXCoder: if it works great i would want to evenually get one
04:15 PM XXCoder: I want to do cardboard builds for fun
04:15 PM XXCoder: oh theres laser board for it too
04:15 PM XXCoder: so it dont cut up stuff under stuff
04:16 PM roycroft: the one i just ordered says it can cut wood 3mm thick in a single pass
04:16 PM roycroft: the laserpecker cannot cut at all
04:16 PM XXCoder: and it barely costs more.
04:16 PM roycroft: yeah
04:16 PM XXCoder: only con I think is large size compared to tripod
04:17 PM roycroft: yes, and i did not want that for two reasons:
04:17 PM roycroft: 1. the space it needs to work
04:17 PM XXCoder: you can always print a wall holder for it
04:17 PM roycroft: 2. the inability to use it with large pieces
04:17 PM roycroft: but i can just put extension legs on it or something
04:17 PM roycroft: and i'll make the space
04:17 PM roycroft: i have a little space where i keep my 3d printer
04:17 PM XXCoder: large peices hmm you probably can make adoptor to lower laser head
04:18 PM roycroft: there are boxes of filament sitting on the shelf
04:18 PM roycroft: i can relocate those, and fit the new laser there
04:18 PM roycroft: that space is also right next to a window, so it will be easy to vent
04:18 PM XXCoder: nice idea
04:31 PM roycroft: i'm not sure i'll want or need this, but the one i just ordered can also take a rotator attachment so i can engrave on round things
04:32 PM roycroft: and the last good thing is that i'll never have to say the word "laserpecker" again :)
04:34 PM XXCoder: yay no more pecker
05:03 PM Tom_L: hope they'll take the return without issue
05:20 PM roycroft: it's already returned
05:20 PM roycroft: my money should be refunded by tonight
05:21 PM roycroft: it was one of those "no box needed" amazon returns
05:21 PM roycroft: which means they don't even return it to the warehouse - they just collect everything on a pallet and auction it off
05:22 PM roycroft: i could have kept it and returned an empty box and gotten my money back, but i didn't, as i have no use for the thing
05:22 PM roycroft: this is not even considering the moral implications of doing so
05:23 PM roycroft: https://www.xtool.com/products/20w-diode-laser-module-for-xtool-d1-pro
05:23 PM roycroft: if i ever want to upgrade the laser there's that
05:23 PM roycroft: and that's powerful enough to deeply engrave metal and even cut thin metal
05:24 PM roycroft: the 5w that i ordered is better for now, though - it's not only less expensive, it's quite a bit finer resolution, so it will do a better job engraving wood
05:31 PM roycroft: the device i just returned has a 0.15mm pixel size, and the one i just ordered has a 0.1mm pixel size
05:33 PM JT[m]: Hopefully it works
05:34 PM roycroft: i have more confidence that it will than the last one
05:34 PM roycroft: but if not, it also goes back with no questions asked
05:35 PM XXCoder: also nice
05:35 PM roycroft: i'm not sure if i can engrave metal with the 5w laser
05:36 PM ZincBoy[CAON][m]: No. A diode laser can't engrave metal.
05:36 PM roycroft: but at the very least i can engrave an etching mask with it, so i can still do metal etching
05:37 PM XXCoder: you could do interesting stuff, like masking metal sheet, burn off whatever you want processed, then do acid to make those areas matte
05:37 PM ZincBoy[CAON][m]: And there are some interesting coatings that you can use to mark metal.
05:37 PM XXCoder: can do it for mirror too I suppose but you'd be super careful not to burn diode or put eye out
05:38 PM djdelorie: ZincBoy[CAON][m]: my diode laser had no problem etching steel...
05:38 PM ZincBoy[CAON][m]: mirror == metal
05:38 PM XXCoder: glass is fine
05:38 PM roycroft: ah, i see now
05:38 PM XXCoder: zinc yeah though actual mirror is dangerous
05:38 PM ZincBoy[CAON][m]: No more so than metal.
05:38 PM roycroft: i'd read that a diode laser cannot etch metal, but xtool say their machines can
05:38 PM XXCoder: hmm ok
05:38 PM roycroft: reading the finer print, their diode lasers can etch stainless steel and coated metals
05:39 PM roycroft: but they also sell an ir laser for their machines, and that's what they primarily recommend for metal
05:39 PM djdelorie: but not aluminum/copper/heatsinks etc because the heat dissipates too fast
05:39 PM ZincBoy[CAON][m]: You can use some spray coatings to mark metal but the normal diode lasers can't mark metal.
05:39 PM ZincBoy[CAON][m]: Co2 lasers also can't mark metal.
05:39 PM roycroft: well my original plan was to etch metal
05:39 PM ZincBoy[CAON][m]: You need a fiber laser to mark metal.
05:40 PM roycroft: i.e. coat it with a resist layer, use a laser to burn off the resist, and then dip it in an etching solution
05:40 PM ZincBoy[CAON][m]: (Co2 lasers can if the power is >150W. Not a typical home system).
05:40 PM roycroft: and that's still my plan
05:41 PM ZincBoy[CAON][m]: It will do that just fine. Just be aware that metal surfaces are pretty much mirrors once you burn off the coating.
05:41 PM roycroft: yes
05:42 PM ZincBoy[CAON][m]: I have be playing with selective powder coating using the laser to partially melt the coating and then finishing with a cure bake after washing off the rest of the powder.
05:42 PM roycroft: that sounds interesting
05:43 PM ZincBoy[CAON][m]: Some success but still playing with settings. There is a very fine line between doing nothing and vaporizing the coating 🙂
05:43 PM roycroft: my interest is more in etching the metal and then filling it
05:43 PM roycroft: so that i can keep the surface level
05:43 PM ZincBoy[CAON][m]: I do that with acrylic all the time. Etch with laser, add powder coating powder, melt in place.
05:43 PM roycroft: but for applications where the badge is not subject to wear, your approach might work very well and would be more efficient than etching and filling
05:44 PM ZincBoy[CAON][m]: Doesn't work with metal though without a post cure cycle.
05:44 PM roycroft: sure
05:44 PM roycroft: the laser just turns it from powder to strings that semi-adhere
05:45 PM roycroft: the ez-bake oven cycle hardens it and adheres it firmly to the metal
05:45 PM XXCoder: powder coating then burning on coating to fuse it is interesting
05:45 PM XXCoder: so does its inverse, bake it fuse to whole sheet then burn off you dont want there like logo
05:45 PM ZincBoy[CAON][m]: Yup. I almost have it working but I really need a powder coating gun as trying to spread it by hand is not consistent enough.
05:45 PM roycroft: i think that after the laser cycle you brush or blow off the excess powder
05:46 PM roycroft: harbor freight sell powder coating guns
05:46 PM roycroft: they are probably better than spreading by hand
05:48 PM roycroft: i already found a lightburn profile for the laser i ordered
05:48 PM XXCoder: i guess its better to see program compitability list and order machine in it
05:48 PM XXCoder: rather than get machine and hope it does
05:48 PM roycroft: so i suspect i'll have the new one up and running pretty quickly when it arrives
05:49 PM roycroft: well i was dubious about the one i just returned
05:49 PM roycroft: but i had read that folks were able to make it work
05:49 PM roycroft: and i believe that's possible, but i ran out of patience
05:50 PM roycroft: my goal is not to figure out the internals of a proprietary engraving machine
05:50 PM roycroft: my goal is to engrave on parts
05:53 PM Tom_L: 'able to make it work' doesn't sound too promising
05:55 PM -!- #linuxcnc mode set to +v by ChanServ
06:02 PM ZincBoy[CAON][m]: Too many hobbyist tools are a hobby in themselves. Just got a bambu lab X1c and it is really close to the ease of use of a paper laser printer.
06:05 PM roycroft: it can be a difficult tradeoff
06:05 PM roycroft: i don't have the need nor can i justify the purchase of a commercial laser right now
06:05 PM roycroft: but i also don't have the time to tinker with a hobby tool forever
06:17 PM skunkworks[m]: some of us live for the tinker - as long as it isn't stupid.
06:41 PM roycroft: i like to tinker with some things
06:48 PM CaptHindsight[m]: it's almost like you can use a laser to pin powdercoat in place just like toner with a laser printer
06:50 PM CaptHindsight[m]: or UV type powdercoating that is used on substrates with lower melting temps
06:51 PM CaptHindsight[m]: https://www.youtube.com/watch?v=ls4zZG1C5kU
06:54 PM CaptHindsight[m]: just imagine being able to use process color powder coatings to form images on non-planar surfaces with a laser
09:37 PM mits[m]: How do I do a tool change in a running program if I need to touch off the tool? I can't jog the machine at all during pause or touch off again, right?
10:38 PM djdelorie: mits[m]: in my limited experience, the right answer is "use separate gcode files"
10:42 PM mits[m]: ok
10:43 PM mits[m]: there's also this feature in axis for starting gcode at a specific line, but how does that even work
10:43 PM mits[m]: if you don't have the starting lines in the gcode it won't start the spindle, for example
10:44 PM XXCoder: you manually start spindle
10:50 PM roycroft: that was my first thought
10:50 PM roycroft: split the job up into separate files when you do tool changes
10:51 PM roycroft: or get an indexible tool mounting system, so you don't have to touch off on every tool change
10:58 PM djdelorie: or get an automatic toolsetter, so that you can ;-)
11:02 PM mits[m]: not sure how to do that in a mill spindle at my size
11:02 PM ZincBoy[CAON][m]: The other option is to override m6 with tool touch off code. That way the tool will automatically touch off on each tool change. Requires a fixed tool probe and some python or gcode to work.
11:05 PM * djdelorie wishes laser toolsetters weren't so expensive...
11:07 PM XXCoder: I remember one who made laser measure rough position then touchoff thing
11:07 PM XXCoder: but I long since lost video
11:07 PM XXCoder: sucks as it was great idea
11:13 PM Thorhian7490[m]: High precision metrology equipment is expensive? Who knew… 🚎 Meanwhile I want to buy a Mazak EZ VC 20. Also stupidly expensive.
11:13 PM djdelorie: I need low precision metrology equipment
11:28 PM djdelorie: on a low-cost machine, there's no low-cost equivalent - you have to use different technology
11:29 PM djdelorie: I would think "beam of light" could be made to pretty much any precision and price point
11:44 PM roycroft: what about points of light?
11:44 PM roycroft: they could be pretty precise, especially if you use a lot of them
11:44 PM roycroft: say, a thousand points of light
11:45 PM XXCoder: just remember, on precision dot move to left, price dot move to right
11:47 PM roycroft: with your hands on the lens, you focus the beam in tight