#linuxcnc Logs
Aug 26 2025
#linuxcnc Calendar
05:07 AM Deejay: moin
07:52 AM -!- #linuxcnc mode set to +v by ChanServ
08:20 AM lcnc-relay: <skunkworks8841@> morning!
08:25 AM Tom_L: it is
10:26 AM roycroft: hi
11:26 AM lcnc-relay: <roguish> you guys see Linux turns 34 today ????
11:26 AM lcnc-relay: <roguish> actually, yesterday
12:04 PM lcnc-relay: <Mister Wallace> https://jauriarts.org/_heisenbridge/media/matrix.org/dWHqyCOnrbLflGYNNidKQxmn/Jf4Jq1Og1VE/IMG_2969.jpeg
12:04 PM lcnc-relay: <Mister Wallace> Howdy LinuxCNC folk, I and some friends are planning out a LinuxCNC conversion for a Shizuoka ST-N that currently has an old Bandit controller and slo-sun NEMA42 steppers. Anyone able to help us along the first steps of picking out parts?
12:04 PM lcnc-relay: We’re hoping to be able to re-use the existing steppers and maybe even the existing drivers, but flexible on this. We have a little PC that should be powerful enough to run the software side, so we’d need an interface board. Getting somewhat lost in the sea of options at the moment.
12:04 PM lcnc-relay: <Mister Wallace> * slo-syn
12:05 PM roycroft: look at the mesa 7i96s
12:07 PM lcnc-relay: <Mister Wallace> https://store.mesanet.com/index.php?route=product/product&product_id=374&search=7i96 this?
12:07 PM roycroft: https://mesaus.com/product/7i96s/
12:07 PM lcnc-relay: <Mister Wallace> Oh sick that’s one of the Ethernet ones
12:08 PM roycroft: and the proprietor of mesaus.com just joined the channel
12:08 PM roycroft: and just departed
12:08 PM lcnc-relay: <Mister Wallace> Whelp
12:09 PM roycroft: anyway, that board is made for stepper motor setups and has spindle control as well
12:09 PM roycroft: it might serve you well
12:09 PM roycroft: and it's cheap
12:10 PM lcnc-relay: <Mister Wallace> Think it’ll interface with the old drivers?
12:10 PM lcnc-relay: <roguish> Mister Wallace: good morning
12:10 PM roycroft: i don't know that
12:10 PM lcnc-relay: <Mister Wallace> Good morning!
12:10 PM roycroft: discuss the old drivers here
12:10 PM roycroft: jt runs mesaus.com and is here most of the tim
12:10 PM roycroft: pcw *is* mesa, and is often here
12:10 PM lcnc-relay: <Mister Wallace> I’ll snag some more info off the drivers after work today. Only have details on the steppers
12:11 PM roycroft: you'll find a lot of resources here
12:11 PM roycroft: if you're patient
12:11 PM lcnc-relay: <Mister Wallace> https://jauriarts.org/_heisenbridge/media/matrix.org/UjMaABFsJAcLeZXXpakdIBZR/1Y05lSWkRv8/IMG_0323.jpeg
12:11 PM lcnc-relay: <Mister Wallace> https://jauriarts.org/_heisenbridge/media/matrix.org/CHZhiVupVFTXpBaGEOhgnAik/EWrr6NJq-tQ/IMG_0324.jpeg
12:11 PM lcnc-relay: <Mister Wallace> Steppers in question
12:12 PM lcnc-relay: <Mister Wallace> I am somewhat tempted to get new drivers to keep that Bandit complete
12:19 PM xxcoder: whew day certainly started off hotter than yeserday
12:20 PM roycroft: it's pretty warm here, but it won't get as hot as yesterday
12:20 PM roycroft: we have a new red flag warning, though
12:21 PM roycroft: firefighters have established a line at the southern end of the flat fire, btw
12:21 PM roycroft: that fire almost made it to the city limits of sisters
12:21 PM roycroft: 4000 people are still in the evacuation zone, but with a fire line at the south, if that line is held, no more houses should burn
12:22 PM lcnc-relay: <Mister Wallace> roycroft: I didn’t realize the red flag warnings ever stopped… or maybe they just stack?
12:22 PM JT-Cave: Gecko makes the best stepper drivers
12:22 PM roycroft: ours ended yesterday but was reissued today
12:23 PM roycroft: and we may have 65km/h outflow winds from the thunderstorms that reach 10s of km away from the storm
12:23 PM Tom_L: yes, geckos are nearly bullet proof
12:24 PM Tom_L: worth the extra cash imo
12:25 PM roycroft: mister wallace's motors are medium current, but the g201x should work with them
12:25 PM lcnc-relay: <Mister Wallace> Someone previous recommend Leadshine drivers, any comparison to be made?
12:25 PM roycroft: which is one the lower-cost geckos ($117)
12:25 PM roycroft: leadshine are cheap
12:25 PM lcnc-relay: <Mister Wallace> Shoutout to roycroft for getting me this far after randomly stopping at my garage
12:25 PM roycroft: and if you know what you're doing they can be fine
12:26 PM roycroft: if you don't, they readily give up the magic smoke
12:26 PM Tom_L: 203 is basically the same with better short protection
12:26 PM roycroft: geckos like to keep their magic smoke
12:26 PM Tom_L: i'm using 203v on mine
12:26 PM Tom_L: iirc they have a digital version now
12:27 PM lcnc-relay: <Mister Wallace> These can drive the old SLO-SYN steppers? For no reason I can justify I’m worried they won’t
12:27 PM lcnc-relay: <Mister Wallace> Tom_L: Yeah looks like there’s a 203V digital
12:27 PM Tom_L: find a data sheet on them if possible and compare the specs
12:27 PM roycroft: he posted the data sheet already
12:27 PM Tom_L: i just got back
12:27 PM Tom_L: covid test
12:27 PM Tom_L: neg
12:28 PM lcnc-relay: <Mister Wallace> Data sheet
12:28 PM roycroft: they're 6a, and the geckos can do 7a, and voltage should be no problem
12:28 PM roycroft: congrats
12:28 PM roycroft: not having covid is good :)
12:28 PM xxcoder: indeed
12:28 PM Tom_L: geckos are good for 70v
12:28 PM Tom_L: my kid does
12:28 PM roycroft: hence the test ...
12:28 PM Tom_L: maybe 80... i can't remember off the top of my head
12:28 PM lcnc-relay: <Mister Wallace> https://jauriarts.org/_heisenbridge/media/matrix.org/GJrXxyYtXRgMjfvQpckIWVtr/bBIpfW79Wvc/IMG_3123.jpeg
12:28 PM lcnc-relay: Sticker from the motor too if this makes any difference
12:29 PM Tom_L: 9.2A
12:29 PM roycroft: gecko also have really good tech support
12:29 PM Tom_L: you can run them lower probably
12:30 PM roycroft: hmm, i read 6a on one of the spec sheets that were posted earlier
12:30 PM roycroft: but the sticker on the motor overrules posted spec sheets :)
12:30 PM Tom_L: the motor sticker says diff
12:30 PM JT-Cave: there's a calculator on the wiki or at least there was one to calculate stepper stuff
12:31 PM roycroft: really low voltage on those
12:31 PM Tom_L: that's what chopper drives are for
12:31 PM lcnc-relay: <Mister Wallace> The 6 wire connections to them also keep throwing me off
12:32 PM Tom_L: 2 likely can be wired together so you have 4
12:33 PM Tom_L: two are likely a center tap for the coils
12:34 PM roycroft: btw, last night i did not get to the scale calc, but i did jog the y axis back and forth several times
12:34 PM Tom_L: pretty zippy now ehh?
12:34 PM roycroft: it finally stopped like it did before when i was doing 32 microsteps, but it took five full travels back and forth before it stopped
12:35 PM roycroft: i'm thinking this is indicative of the driver missing steps and stopping after reaching a threshold of missed steps?
12:35 PM Tom_L: https://builds.openbuilds.com/projectresources/stepper-motor-hardware-interfacing.14/
12:36 PM Tom_L: there used to be a better site for that....
12:36 PM Tom_L: https://homepage.divms.uiowa.edu/~jones/step/
12:37 PM roycroft: hmm, the tb6600 has an alarm indicator
12:37 PM roycroft: i should open the case and run it again until it stops and see if the alarm indicator is lit
12:38 PM lcnc-relay: <skunkworks8841@> wow - that brings back memories
12:38 PM lcnc-relay: <skunkworks8841@> I made my own stepper drives initally... using jones theries..
12:38 PM lcnc-relay: <skunkworks8841@> theories..
12:38 PM lcnc-relay: <skunkworks8841@> yah - with 6 wire - you can run them in ether series or parralell
12:38 PM lcnc-relay: <skunkworks8841@> iirc
12:39 PM Tom_L: that's what i was thinking
12:39 PM lcnc-relay: <Mister Wallace> Ah is it so you can wire it bipolar?
12:39 PM Tom_L: jones is a good pate
12:39 PM roycroft: i don't know much about stepper motors but i recall reading that
12:39 PM Tom_L: probably so yes
12:39 PM Tom_L: page*
12:39 PM lcnc-relay: <skunkworks8841@> actually - I think they have to be 8 wire.. (unless you can separate the center taps)
12:39 PM lcnc-relay: <Mister Wallace> Not all types of bipolar
12:39 PM lcnc-relay: <Mister Wallace> Ohhh man this is a whole thing isn’t it
12:39 PM Tom_L: you need to ohm the leads to find the right pairs
12:40 PM lcnc-relay: <skunkworks8841@> (to do series / parallel)
12:40 PM Tom_L: Mister Wallace, you're jumping right into the fire :)
12:40 PM Tom_L: welcome
12:40 PM roycroft: if you are thinking about new drivers you might consider replacing the stepper motors as well
12:40 PM roycroft: they are less expensive than good drives
12:41 PM lcnc-relay: <Mister Wallace> Good point
12:41 PM roycroft: i know this can be the start of project creep
12:41 PM roycroft: but besides having an easier time wiring it up, you'll have brand new motors
12:41 PM lcnc-relay: <Mister Wallace> Well, replacing the motors won’t be a huge undertaking if I get NEMA42
12:41 PM lcnc-relay: <skunkworks8841@> https://photos.app.goo.gl/UxMLmVoBv8yEaMhh9
12:41 PM Tom_L: what form factor are those steppers?
12:41 PM Tom_L: nema 34?
12:41 PM Tom_L: 23?
12:41 PM lcnc-relay: <Mister Wallace> NEMA42
12:41 PM Tom_L: bigger
12:41 PM lcnc-relay: <Mister Wallace> They’re chunky
12:42 PM lcnc-relay: <skunkworks8841@> they are huge lol
12:42 PM lcnc-relay: <Mister Wallace> The mill is similarly somewhat huge
12:42 PM lcnc-relay: <skunkworks8841@> you are not going to find 9 amp drivers.. So you would have to run less current
12:42 PM lcnc-relay: <Mister Wallace> Or buy motors.
12:43 PM lcnc-relay: <Mister Wallace> Any recs for motor brands?
12:43 PM Tom_L: https://www.automationtechnologiesinc.com/
12:44 PM lcnc-relay: <skunkworks8841@> at 6 amps - you will have around 550 in-oz
12:44 PM roycroft: to get back to your original question about the 7i96s, it should work with just about any driver you use, and it doesn't care at all about the motors
12:45 PM lcnc-relay: <Mister Wallace> roycroft: Would I use this board either way? Could always start there and try to re-use the drivers, replace the drivers/motors later if it’s unhappy
12:45 PM roycroft: sure, i don't see why not
12:45 PM Tom_L: the 7i96s also has a header to add a daughter card for more IO as well as sserial for even more 'slower' io like limits
12:45 PM roycroft: i'm not the expert here, though
12:45 PM lcnc-relay: <skunkworks8841@> what do the drives look like?
12:45 PM lcnc-relay: <Mister Wallace> skunkworks8841@: I’ll need to take some pictures after work. They’re the original drivers inside the Bandit, but they’re discrete boards.
12:46 PM Tom_L: skunkworks has tread in deeper water than i have on retrofitting
12:46 PM roycroft: you have a pair of dir pins, a pair of step pins and a pair of enable pins on the 7i96s for each driver
12:46 PM Tom_L: if they have step/direction you may be able to use what you have
12:46 PM roycroft: almost any driver will have the same pins
12:47 PM lcnc-relay: <Mister Wallace> roycroft: They definitely had a 6 wire connection. I have the Bandit wiring diagrams at home, I’ll check that after work too.
12:47 PM Tom_L: you would need to figure out the step timing for the driver
12:47 PM roycroft: having just said that, you almost certainly have oddball drivers that don't work that way :)
12:48 PM roycroft: the drivers will have two additional pins for powering the motors
12:48 PM roycroft: and outputs to the motors, of course
12:48 PM roycroft: so you should have at least 12 pins on the drivers
12:48 PM lcnc-relay: <Mister Wallace> There was also a ribbon cable
12:48 PM Tom_L: take some connection pics
12:48 PM Tom_L: that may help
12:49 PM roycroft: two from the psu, six to the controller board, 4 to the motors
12:49 PM Tom_L: they're 6 wire motors
12:49 PM Tom_L: if they're using them all
12:49 PM roycroft: yeah, so his drivers may have six wires to the motors
12:49 PM JT-Cave: https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z000000PAkPSAW
12:50 PM lcnc-relay: <Mister Wallace> Tom_L: It is
12:50 PM roycroft: but i said "at least"
12:50 PM roycroft: :)
12:50 PM Tom_L: 6 wires are usually centertapped
12:51 PM lcnc-relay: <Mister Wallace> Any thoughts on swapping for servos?
12:51 PM Tom_L: servos may need further reduction
12:51 PM Tom_L: belt or gear
12:52 PM lcnc-relay: <skunkworks8841@> the drives may be phase controlled depending on the age..
12:52 PM Tom_L: jt that's a good page
12:53 PM lcnc-relay: <Mister Wallace> Age: extremely old
12:53 PM lcnc-relay: It’s an early generation Bandit
12:54 PM pcw-home: If they are 6 wire motors they likely have 4 transistor drives with large ballast resistors
12:55 PM pcw-home: Also likely 1/2 step
12:56 PM lcnc-relay: <skunkworks8841@> pcw-home: do any of the mesa boards allow for 4 phase control? I see that table look up is available..
12:57 PM lcnc-relay: <captainhindsight_.@> Linux turned 34 today
01:01 PM pcw-home: Yes, you can always use table mode
01:02 PM lcnc-relay: <Mister Wallace> From Tom’s suggestion, thoughts on just doing these and calling it a day? https://www.automationtechnologiesinc.com/products-page/nema42-close-loop-stepper-motor-system-hybrid-servo-kit/nema42-220ac-shaft-19mm
01:04 PM lcnc-relay: <Mister Wallace> It’s sorta sounding like just replacing the drives would make my life a lot easier, and subsequently replacing the motors, so trying to figure out a cost to actually do that.
01:05 PM lcnc-relay: <captainhindsight_.@> The nema34 closed loop stepper kits were ~ $120 last year
01:05 PM lcnc-relay: <Mister Wallace> I need NEMA42 to avoid significant modifications to the mill
01:06 PM lcnc-relay: <Mister Wallace> $1500 is definitely more than I want to spend right now though
01:08 PM lcnc-relay: <Mister Wallace> The other question is power source, the bandit is running this on 120vAC but I also have 3 phase 240 coming in for the spindle
01:09 PM lcnc-relay: <captainhindsight_.@> Delta ?
01:11 PM lcnc-relay: <captainhindsight_.@> https://www.omc-stepperonline.com/ys-series-1-axis-9-00nm-1274-5oz-in-nema-34-closed-loop-stepper-kit-w-power-supply-1-clys90
01:12 PM lcnc-relay: <captainhindsight_.@> <$100 ea now
01:13 PM lcnc-relay: <Mister Wallace> Not bad, think those have enough torque to keep up with the current steppers?
01:14 PM lcnc-relay: <captainhindsight_.@> we could not find BLDC drives for even 2x the price of similar power
01:14 PM lcnc-relay: <Mister Wallace> Too bad Stepperonline no longer sells closed loop NEMA42 steppers. But they do have open loop ones for a reasonable price.
01:15 PM lcnc-relay: <captainhindsight_.@> the 12nm kits are even lower cost https://www.omc-stepperonline.com/ys-series-1-axis-closed-loop-stepper-cnc-kit-v2-0-12-0nm-1699-34oz-in-nema-34-motor-w-2-0m-cables-driver-1-cl86y-s120-v20
01:18 PM lcnc-relay: <captainhindsight_.@> Nema 42 to 34 conversion usually just means making a new motor mount
01:18 PM lcnc-relay: <captainhindsight_.@> how many Nm at what RPM do you need/want?
01:18 PM lcnc-relay: <Mister Wallace> It would mean that here for sure.
01:20 PM lcnc-relay: <Mister Wallace> captainhindsight_.@: I have struggled to answer this question. Was going to try and figure out what’s reasonably priced and see what it could achieve. The mill table is extremely heavy and running on dovetail ways, I don’t think it’ll ever go insanely fast.
01:21 PM lcnc-relay: <Mister Wallace> Could maybe start looking at 400 Ncm at 2000rpm, smack in the middle of the curve for the existing steppers.
01:22 PM Tom_L: acme or ballscrew?
01:22 PM lcnc-relay: <Mister Wallace> Ballscrew
01:22 PM roycroft: in my view, integrating the existing system with the new controller would be a first step, so getting the specs on your drivers would be good - they can likely be controlled by linuxcnc with the 7i96s
01:22 PM roycroft: when you get that working, then focus on the motion control hardware
01:23 PM lcnc-relay: <Mister Wallace> Very reasonable
01:23 PM * roycroft always likes to take baby steps vs giant leaps
01:23 PM lcnc-relay: <captainhindsight_.@> https://www.omc-stepperonline.com/hybrid-stepper-motor?mfp=184-frame-size-mm[Nema%2042%20(110%20x%20110)]
01:23 PM roycroft: and yeah, i know that means i'm no fun any more
01:24 PM lcnc-relay: <Mister Wallace> NEMA42 steppers would at least make the big step less of a hassle
01:24 PM lcnc-relay: <captainhindsight_.@> 30Nm is as big as they seem to offer right now
01:25 PM lcnc-relay: <Mister Wallace> Dang those don't go very fast either
01:26 PM lcnc-relay: <Mister Wallace> What about these? https://www.omc-stepperonline.com/servo-motor?mfp=294-rated-power-w[1000],184-frame-size-mm[80%20x%2080],291-encoder-type[Magnetic%20Absolute],389-with-brake[Yes]
01:28 PM lcnc-relay: <captainhindsight_.@> is that the right link? those are < 10% of what you were asking about for torque
01:29 PM lcnc-relay: <Mister Wallace> 3.81NM?
01:29 PM lcnc-relay: <captainhindsight_.@> Nema42, 40NM @ 2k rpm , is that what you want?
01:30 PM lcnc-relay: <Mister Wallace> I was trying to match the current steppers, which do 4NM @ 2k rpm
01:31 PM lcnc-relay: <Mister Wallace> (Listed on the spec as 400 Ncm)
01:34 PM lcnc-relay: <captainhindsight_.@> sounds suspect since steppers usually don't have much torque by 2K rpm
01:34 PM lcnc-relay: <Mister Wallace> Finding solid information on these old weirdo steppers has been somewhat challenging.
01:39 PM lcnc-relay: <captainhindsight_.@> sorry my new brain read nema 34 for nema 43 earlier
01:39 PM lcnc-relay: <Mister Wallace> No worries
01:39 PM lcnc-relay: <captainhindsight_.@> nema 42, sorry new brain
01:39 PM lcnc-relay: <Mister Wallace> https://www.omc-stepperonline.com/dual-shaft-nema-42-cnc-stepper-motor-30nm-4248oz-in-8a-110x201mm-4-wires-42hs79-8004d These would sure be easy to install. Open loop but equally easy encoder install.
01:40 PM lcnc-relay: <captainhindsight_.@> yeah adding your own encoder is usually easy
01:41 PM lcnc-relay: <captainhindsight_.@> especially with dual shaft
01:44 PM lcnc-relay: <Mister Wallace> Reducer gearboxes actually seem to be pretty reasonably priced
01:49 PM lcnc-relay: <Mister Wallace> Taking a step back, couldn't these drive the existing steppers at 9.2A? https://www.geckodrive.com/product/g320x-digital-servo-drive/
01:49 PM lcnc-relay: <Mister Wallace> nvm servo drive
01:49 PM roycroft: i just found what may be an issue that is causing the drivers to shut down
01:49 PM roycroft: btw, no alarm - when the axis stops moving both the power and alarm lights are off
01:49 PM roycroft: but i just saw the driver is rated to 42v, and i'm pushing 48v to it
01:49 PM xxcoder: yeah? what would be the possible cause?
01:50 PM roycroft: i don't know if i can dial the psu down to 42v or not
01:50 PM xxcoder: oh interesting
01:50 PM roycroft: i have the scale set properly now
01:50 PM xxcoder: maybe dial it down as far as you can, see if movement before it stops is longer
01:50 PM roycroft: yes, i'll try that
01:51 PM roycroft: that unfortunately means i have to remove the psu from the chassis, but i'm getting used to having to do that
01:51 PM lcnc-relay: <captainhindsight_.@> power supplies often have 10% voltage adjust, just add a few diodes for the win
01:51 PM roycroft: yes, the mean well psus usually can be adjusted about 10%
01:51 PM roycroft: which would bring me down to 43ish volts
01:51 PM roycroft: which may be close enough
01:51 PM xxcoder: whats one volt among friends
01:52 PM roycroft: my friends are chinese
01:52 PM roycroft: and they may lie about even going to 42v
01:52 PM lcnc-relay: <captainhindsight_.@> Chinese volts and all
01:52 PM roycroft: i'll dial it down as much as i can
01:52 PM xxcoder: in least it dont seem to have released any magic smoke
01:53 PM roycroft: and if that still doesn't do the job i may have to use the geckos
01:53 PM xxcoder: drivers that is
01:53 PM roycroft: which i don't want to do - that adds another $350 to the cost of the project
01:53 PM xxcoder: im sure theres new cheap options besides tb6600 nowdays
01:53 PM lcnc-relay: <captainhindsight_.@> it's universal design , you can open the supply and change resistors to get a wider voltage range
01:53 PM roycroft: not to mention all the rewiring required
01:54 PM roycroft: let me see what it dials down to
01:54 PM roycroft: iirc, the lower the voltage the slower the motors, right?
01:54 PM roycroft: voltage = speed, current = torque with steppers?
01:55 PM lcnc-relay: <skunkworks8841@> yes
01:55 PM roycroft: i could always swap that 24v psu that came with it back in to drive the steppers, but i don't want to do that
01:55 PM roycroft: even though it's a small machine, i want it to be small and mighty, not small and meek
01:56 PM * roycroft heads out to dial down the voltage
01:56 PM xxcoder: mighty mouse and not his hmm brother? forgot
01:56 PM xxcoder: comic, not that apparently musician?
02:04 PM roycroft: so that was the one thing i could reach on the psu without removing it
02:04 PM roycroft: i dialed it down to 40v
02:05 PM roycroft: and that may have fixed the problem - i was able to use the g1 command to move it 9" and it completed for the first time
02:05 PM roycroft: i need to move it back and forth a few times, but that may have solved my problem
02:05 PM roycroft: and i think i can live with 40v
02:05 PM roycroft: or maybe tweak it up to 42v
02:06 PM xxcoder: mighty mouse!
02:06 PM roycroft: it's not worth investing $350 to get 5 more volts
02:06 PM roycroft: here i come to save the day!
02:08 PM roycroft: https://www.youtube.com/watch?v=4d9N_6YY_mM&list=RD4d9N_6YY_mM&start_radio=1
02:08 PM roycroft: for the youngsters out there
02:09 PM xxcoder: interesting video style.
02:10 PM xxcoder: wow. youtube have completely removed ability to turn auto play though list videos
02:16 PM roycroft: nope, it still stops after a while
02:16 PM roycroft: but i can do about 4 "laps" before it stops
02:16 PM roycroft: power and alarm lamps off on the driver
02:17 PM roycroft: maybe i should current limit it?
02:17 PM xxcoder: ugh. I still cant remember what I did to resolve it
02:17 PM roycroft: it's set to max now, which is 3.5a continuous, 4a peak
02:17 PM roycroft: maybe i should turn it down to 2.5a for testing
02:17 PM xxcoder: good idea to test that parameter yeah
02:18 PM roycroft: i still don't know what i'm doing, but i'm thinking more and more this is an issue with the driver, not the motors or the mesa board
02:20 PM lcnc-relay: <skunkworks8841@> so the drives lose power?
02:20 PM lcnc-relay: <skunkworks8841@> Does the power supply shut down?
02:21 PM lcnc-relay: <skunkworks8841@> Could it be back emf causing the power supply voltage to spike and shut itself down?
02:31 PM Tom_L: now we hear the rest of roy's story
02:32 PM lcnc-relay: <captainhindsight_.@> back EMF with some switchers are a problem
02:33 PM lcnc-relay: <skunkworks8841@> Not as common with steppers - but still possible.
02:37 PM roycroft: i got bored watching the y axis do laps, so i shut it down
02:37 PM roycroft: i'll kick it up to 3a after work and see if it still works ok
02:38 PM roycroft: so the solution seems to be to drop voltage down to the drivers' spec, and to reduce current
02:38 PM roycroft: the psu does not shut down, but the driver would shut down
02:38 PM lcnc-relay: <skunkworks8841@> huh.
02:38 PM roycroft: i.e both the power light and alarm light would be off, and i could easily move the axis by hand
02:39 PM lcnc-relay: <skunkworks8841@> and the power supply was outputting voltage?
02:39 PM roycroft: my psu was putting out 49v, and the drivers are rated at 42v
02:39 PM roycroft: yes
02:39 PM lcnc-relay: <skunkworks8841@> wow. that is odd.
02:39 PM roycroft: i reduced the psu output to 40v, and the driver would work longer, but still shut down
02:39 PM roycroft: i had current set to max - 3.5a continuous
02:39 PM roycroft: i reduced that to 2.5a, and now the driver will work all day long
02:40 PM roycroft: i'm not saying i undestand at all why this worked out the way it did
02:40 PM roycroft: it was pure experimentation and empirical data gathering
02:40 PM xxcoder: very strange. but if it works, it works
02:40 PM lcnc-relay: <skunkworks8841@> what are the drives?
02:40 PM roycroft: so i'm going to bump it up to 3.0a after work and see if it still works
02:40 PM Tom_L: chinaco amps
02:41 PM roycroft: tb6600
02:41 PM roycroft: and they probably don't know what else to do when they're stressed out
02:41 PM roycroft: at least they have the wherewithall to shut down instead of belch out their magic smoke
02:42 PM xxcoder: maybe yours was very different problem to what I had. I used 24v which probably wasnt enough power, and yours is plenty of volt but too much amps
02:42 PM roycroft: and while i do intend to cut/engrave/machine non-ferrous metals at times on this machine, i'm not so worried about having to do the current limiting
02:42 PM roycroft: voltage is more of a concern, as i do not want to have to wait until christmas to pull a part off of it that i start before thanksgiving
02:42 PM xxcoder: :)
02:43 PM roycroft: with 24v it should work fine, but very slowly
02:43 PM roycroft: anyway, i think i have a solution
02:43 PM xxcoder: it definitely was slow. maybe that was my solution, slow it down
02:44 PM roycroft: which is to feed the driver within its specifications
02:44 PM xxcoder: or somewhat below
02:44 PM roycroft: yes
02:44 PM roycroft: so in my office, with the recent rearrangement of the furniture, there's this space in between my desk and a bookshelf that is blocked by the printer
02:44 PM roycroft: if you're on the floor there is no escape route
02:45 PM roycroft: but you can jump into the space from the desk
02:45 PM roycroft: of course, both my cats like to jump into the space
02:45 PM roycroft: and then start whining when they can't get out
02:45 PM roycroft: i'm going to rearrange things again soon, so that problem will go away
02:46 PM roycroft: but for now, whenever my office door is open i have to expect to rescue a cat at any time
02:46 PM roycroft: and i do so often
02:46 PM xxcoder: I probably would put up a book there or something :) or a binder
02:46 PM roycroft: it's a pretty big space
02:46 PM roycroft: and they *should* be able to get out
02:47 PM roycroft: i don't know why they think they can't
02:47 PM roycroft: both of them are perfectly capable of jumping up from the floor onto the desk
02:47 PM xxcoder: maybe try teasers see if they jump off
02:47 PM roycroft: but when they get in that little space, they think they cannot jump back up on the desk
02:47 PM roycroft: one of them has tried multiple times to jump up
02:47 PM roycroft: but she's just not trying very hard
02:48 PM roycroft: and there's plenty of room, so it's not like it has to be a vertical jump
02:48 PM roycroft: it can be angular, like they do from the main floor area
02:48 PM roycroft: i kind of think they just like getting rescued
02:48 PM xxcoder: try teasers lol
02:48 PM roycroft: anyway, problem mostly solved
02:49 PM roycroft: and i put a rule on the minirouter bed
02:49 PM roycroft: i still do not have the scale correct
02:49 PM roycroft: i told y to move 15", and it moves about 13"
02:50 PM xxcoder: what I did was use webcam attached to tool on spindle, and point it down to ruler. use cursor on pc as "mark"
02:50 PM roycroft: that's 13% off
02:50 PM xxcoder: have it repeat a loop
02:50 PM xxcoder: moving back and forth
02:50 PM roycroft: i'll tweak it by 13%, and that should get me really close
02:50 PM roycroft: then, when i have it working, i'll cut some squares
02:50 PM roycroft: and measure them, so i can do the final tweaking
02:53 PM xxcoder: repeatively back and forth is always great way to find any problems of lost steps, or coupler slipping
02:55 PM xxcoder: (it was also what I used webcam for, to see if theres any drafting)
02:55 PM roycroft: yes
02:55 PM roycroft: although that can be done more accurately by cutting
02:56 PM xxcoder: yeah
02:56 PM xxcoder: "video cutting" is great for very early tests though
02:56 PM roycroft: cut a square, then air cut the square a few more times, then go down farther and cut a new square, repeat air cut, etc.
02:57 PM roycroft: if you're losing steps that won't work well, unless you measure the square each time before you cut again
02:57 PM xxcoder: I just remember one test I planned to make but never did. camera to center spot, then go circle around it, then back to point
02:57 PM roycroft: actually, you should be able to tell if you cut a square, air cut the square a few times, and then recut the square
02:57 PM roycroft: if there are any chips the second time you're losing steps
02:58 PM roycroft: and you're less likely to lose steps when you air cut
02:58 PM xxcoder: makes sense. should have essentally zero draft with wood
02:58 PM roycroft: so maybe just cutting the same square repeatedly is the ticket
02:59 PM xxcoder: I suppose, though camera method is zero risk for very early
02:59 PM roycroft: yeah, i'm talking about doing the final dialing in
02:59 PM xxcoder: indeed :D
02:59 PM roycroft: which you can only do by making parts and measuring them
02:59 PM xxcoder: just so glad you figured the drivers issue
02:59 PM roycroft: yeah, it took me a while to get there
03:00 PM roycroft: but i learned some stuff along the way
03:00 PM roycroft: so it's all good
03:00 PM xxcoder: I wonder if I would be able to run machine faster had I figured the amps issue
03:00 PM xxcoder: probably not a whole lot faster but yeah
03:01 PM roycroft: well that little machine is doing for me exactly what i wanted and expected, for now
03:01 PM roycroft: it's getting me hands-on experience with a real machine that is controlled by linuxcnc
03:02 PM roycroft: playing with it in sim mode for years taught me the basics, but i knew that when it was time to build a real machine there would be a whole lot of new stuff that sim mode could not prepare me for
03:02 PM roycroft: like most of what i'm dealing with now
03:02 PM roycroft: and having something that small and simple makes it easier to learn
03:03 PM xxcoder: I bet. its very bog standard 3 steppers 3 axis machine
03:03 PM xxcoder: drivers being the most "quirky" part of it
03:11 PM roycroft: they are very common drivers
03:11 PM roycroft: it's actually good that i have them, because i'll have experience with them, and will likely run into them again
03:12 PM xxcoder: curious if you log stuff
03:12 PM xxcoder: I kinda wish I did lol
03:12 PM roycroft: i keep notes on things
03:12 PM xxcoder: nice
03:12 PM roycroft: not always when i'm in the middle of figuring them out
03:12 PM roycroft: and even then, not usually in a form that i can easily access later
03:12 PM roycroft: i do like to document stuff when it's done, though
03:13 PM roycroft: if i'm honest, though, i have so much stuff going on right now that i'm not doing as good a job of documenting as i usually do
03:14 PM roycroft: i'm kind of curious to see if going back to 32 microsteps would work reliably, now that i've done the voltage and current adjustments
03:15 PM roycroft: if i understand correctly, though, the more microsteps the less torque
03:15 PM roycroft: so if i'm already current limiting more than initially, i should probalby not increase the microsteps to further reduce torque
03:15 PM JT-Shop: microsteps are usually not and advantage above 10
03:15 PM roycroft: but it would be intesting to see if it still works with 32 microsteps
03:16 PM JT-Shop: hence the reason gecko uses 10 and morphs to 1
03:16 PM roycroft: so if i'm current limiting below what i was, reducing microsteps from 16 to 8 would mitigate the loss of torque from the current limiting
03:16 PM xxcoder: yeah good idea to test if it was other factor or not, but yeah I'd retain current microstep
03:17 PM roycroft: the tb6600 does microstepping by powers of 2
03:17 PM JT-Shop: yup I would not use more than 8
03:17 PM roycroft: so i can't do 10
03:17 PM Guest3050: Increasing microsteps does not reduce stall torque
03:17 PM JT-Shop: holding torque is just that staying still
03:17 PM roycroft: right
03:18 PM roycroft: holding torque is what makes one take a nap when one does not want to take a nap
03:18 PM Guest3050: in fact, increasing microsteps increases torque through the resonance zone
03:18 PM JT-Shop: just do a search for microstep myths
03:19 PM roycroft: yes, i'm getting mixed messages now
03:19 PM Guest3050: (by not exciting resonance)
03:19 PM roycroft: or maybe not, but maybe i'm not grasping the nuances of the messages well enough
03:19 PM xxcoder: honestly I dont understand guest
03:20 PM Guest3050: there is a big message around that higher microstep ratios decrease torque, this is false
03:22 PM Guest3050: the torque per microstep is of course lower, but the stall torque is unchanged or higher
03:23 PM * JT-Shop goes to spoil the chickens
03:23 PM Guest3050: (this is not considering ratios as low as 1/2 as these increase the average current)
03:24 PM Guest3050: (because there are times when both windings are at 100% current with 1/2 stepping)
03:25 PM xxcoder: well ultimately the biggest concern is when its moving
03:25 PM Guest3050: also at higher speeds there is no microstepping as the drive never reaches current control mode
03:26 PM xxcoder: and it is torque when its moving we was talking about.
03:26 PM roycroft: you're using stall torque as that required to ovecome load torque + friction torque + coging torque?
03:26 PM Guest3050: Yes, stall torque = better with higher microstepping
03:28 PM Guest3050: stall torque being the amount of torque load in the shaft needed to stall the motor (and thus lose steps)
03:28 PM Guest3050: at any speed
03:28 PM roycroft: so yes, what i said, which are the three counter-torques that must be overcome
03:29 PM roycroft: so i'm not saying i dispute your contention that stall torque is better with microstepping, but i'll say that i don't understand it and that it seems counter to what little i do understand
03:30 PM roycroft: i suspect that when the time comes that i understand well the nuances of stepper motor technology i'll have just moved on to using servos :)
03:34 PM xxcoder: maybe. you'd still have that minirouter lol
03:37 PM * JT-Shop finishes up the last hex
03:38 PM roycroft: and maybe i'll servocharge the minirouter
03:39 PM roycroft: slap some encoders on it, put a spoiler fin on the gantry, and mount a nitrous tank on top
03:39 PM JT-Shop: don't forget to tub the wheel wells
03:41 PM xxcoder: as long as you dont add coal roller to it, im good
03:41 PM JT-Shop: lol
03:42 PM roycroft: the more i read about microstepping the more i'm convinced that i should really dial that back, especially on such a small machine
03:42 PM roycroft: i did not realize that torque dropped off so dramatically
03:42 PM xxcoder: yeah. I could easily stop machine by hand at 32
03:43 PM xxcoder: thats why I told you to lower microstep a while ago :)
03:43 PM roycroft: i thought it probably dropped off linearly, but it seems more like exponential
03:43 PM roycroft: and there goes a cat down into the hole
03:43 PM xxcoder: (stoppable by hand at 32 is with 24v)
03:43 PM roycroft: i should start hearing meowing in about 15 seconds
03:44 PM * roycroft gets up to perform a rescue operation, on queue
03:45 PM xxcoder: try use your cat teasers to coax then out
03:45 PM roycroft: when i have time
03:45 PM roycroft: it's faster just to scoop them up
03:45 PM xxcoder: lol ok
03:50 PM Guest3050: Torque does not decrease with higher microstep ratios, that is pure myth
03:51 PM roycroft: that is what you keep saying
03:51 PM Guest3050: Its true
03:51 PM roycroft: but every resource i've read, including documents from stepper motor manufacturers, says the opposite
03:51 PM JT-Shop: link
03:51 PM roycroft: please present your evidence
03:52 PM roycroft: several articles discuss how microstepping can actually result in poorer performance, as when there is not enough torque generated at a microstep the motor does not move, and cannot move until the cumulative torque of multiple microsteps can overcome the stall torque, resuilting in jerky motor movement
03:53 PM roycroft: with increased vibration and the negatives that go along with that
03:53 PM roycroft: if you have evidence to the contrary i'd love to see it
03:56 PM Guest3050: You are mixing up the torque per microstep (not very important) with stall torque (very important)
03:57 PM roycroft: i'm not
03:58 PM Guest3050: Here's a fairly good reference:
03:59 PM Guest3050: https://www.faulhaber.com/en/know-how/tutorials/stepper-motor-tutorial-eight-facts-and-myths-surrounding-microstepping-operation/
03:59 PM Scopeuk: I believe there is some distinction based on what is limiting your stepper impulse, if everything is working "properly" and the driver is actively limiting the current it is supplying with effect the microstepping has no impact, when you are taking the motor coils up to saturation such that they limit it instead of the motor driver things get more
03:59 PM Scopeuk: interesting. if the coils are at saturation microstepping is only a detriment
03:59 PM roycroft: yes, i just read that
03:59 PM JT-Shop: https://www.geckodrive.com/support/accuracy-and-resolution/
03:59 PM roycroft: i suspect this is a context/perspective thing
04:00 PM xxcoder: https://cdn.faulhaber.com/media/DAM/Images/Logos%20-%20Drawing/Tutorial/faulhaber-tutorial-stepper-motor-microstepping-en-grafik-002.jpg
04:02 PM Guest3050: Yes that's torque per microstep which people mistake for stall torque, leading to the more microsteps=less torque myth
04:03 PM roycroft: perhaps it would be useful to use a less loaded term than "myth" if you're trying to convince someone of something
04:04 PM roycroft: "misconception" or "misunderstanding" might be better words
04:04 PM Guest3050: Imagine I pull a compass needle around with a magnet, microsteps are just the angular distant I move the magnet each step
04:04 PM Guest3050: distance
04:06 PM xxcoder: well reardless, it have been my experence that too high microsteps makes machine not go
04:06 PM xxcoder: or have problems like missed steps, stalling because it cant move etc
04:06 PM roycroft: it puzzles me to read a document that tells me that microstepping torque may be so low that multiple microsteps are required to overcome stall torque, and then for you to point to the same document as evidence that it is a "myth" that microstepping reduces torque
04:07 PM roycroft: i'm trying to understand that
04:07 PM roycroft: and then you claim that i am confusing "microstep torque" with "stall torque"
04:07 PM roycroft: which i clearly am not doing
04:08 PM roycroft: perhaps if you could find another way to express your point it would be easier to understand
04:12 PM Guest3050: When I say stall torque I mean the amount of torque needed to make the motor miss a full step (stall)
04:13 PM roycroft: correct
04:13 PM Guest3050: this does not decrease regardless of how high the microstep ratio is
04:13 PM roycroft: i further defined that as the force required to overcome load torque + friction torque + cogging torque
04:13 PM roycroft: i.e. the forces working against the motor
04:14 PM roycroft: stall torque is the amount of torque that must be over come for the motor to move
04:15 PM roycroft: perhaps the difference in perspective here is that you're claiming that it's the amount of torque that must be overcome to move the motor a full step
04:15 PM Guest3050: That is not how motors are rated. They are rated by maximum load torque
04:16 PM roycroft: and by that definition, the reduced microstep torque alone is meaningless
04:16 PM Guest3050: Maximum load torque is very important for open loop step/dir systems as a lost step is normally a disaster
04:17 PM roycroft: but that the cumulative torque by all the microsteps to effect a full step is the same as that required without microstepping
04:17 PM roycroft: is that what you are saying?
04:18 PM Guest3050: Yes, so the motor stall torque (the curves in a step motor data sheet) is not affected by the microstep ratio
04:18 PM roycroft: ok
04:20 PM roycroft: so if you have 8 microsteps, and each microstep produces 0.1Nm of torque, you can drive a load of <= 0.8Nm
04:21 PM roycroft: if you do not microstep, and each step produces 0.8Nm of torque, you can drive a load of <= 0.8Nm
04:22 PM roycroft: so the question is, with a given controller and motor, will setting it to 8 microsteps produce the same cumulative torque as setting it to 0 microsteps?
04:23 PM Guest3050: well its not linear, more sinusoidal (Its when you feel when you turn the shaft of a powered stepper motor)
04:24 PM roycroft: i have the formula to calculate the incremental torque of a microstep
04:24 PM roycroft: when i'm done with work and have a little more time, i'll get some real world specifications and do the math
04:25 PM JT-Shop: there's a reason gecko starts out at 10 microsteps and morphs to full step as the speed increases and the morphing point is adjustable
04:25 PM roycroft: the articles i've been reading, btw, mostly focus on increased resonance/vibration with too much microstepping
04:26 PM xxcoder: on tb6600 it has no smarts, so you gonna select best balance between too much vibration of 1 to too much vibration and weakness at 32
04:26 PM JT-Shop: yup that resonance is tuned out with gecko by adjusting the morphing point
04:27 PM roycroft: yes, i've read gecko articles on that
04:27 PM xxcoder: I had bashed my head on microstepping so much, because I couldnt hear it
04:27 PM xxcoder: all I could do is feel it, and try pressure it see if it fail
04:28 PM roycroft: you can see it too, in your finished parts
04:28 PM roycroft: especially with a smaller, less rigid machine
04:29 PM Tom_L: i wonder if their digital 203 are much better than their older ones
04:29 PM roycroft: anyway, i think guest050 is not contradicting what we understand about stepper motors
04:29 PM roycroft: i think he's just discussing it from a different perspectdive
04:29 PM Tom_L: it's pcw
04:29 PM Tom_L: he's hiding :)
04:29 PM roycroft: ok
04:29 PM roycroft: then i'm especially glad i was being polite :)
04:30 PM xxcoder: lol I always express what I feel like, reardless of who :)
04:30 PM roycroft: but the real reason i was being polite is that i thought all along that this was not really a disagreement
04:32 PM xxcoder: ypu're always nice anyway
04:32 PM roycroft: apparently i'm supposed to be done with work now
04:33 PM roycroft: because there are currently two cats sitting on my desk, each blocking 1/2 of my primary display
04:33 PM roycroft: my irc screen is on one of the ancillary displays, so i can still see that
04:33 PM roycroft: but i cannot see any of my work terminal sessions
04:33 PM roycroft: and i'm not always nice
04:34 PM roycroft: i try to never be mean
04:34 PM roycroft: but sometimes i'm impatient and a bit snarky
04:34 PM roycroft: i'm not proud of that, but oh well
04:34 PM xxcoder: reminds me of that it guy trick to keep cat off his router, as cat would overheat it. he put another router on top with gap, and powered it on but no wifi and unused
04:34 PM xxcoder: so cat lays on that router and actual working router runs fine
04:34 PM roycroft: i think i'll go up the current to 3.0a and do a test now
04:34 PM roycroft: and hope the cats move while i'm gone
04:35 PM xxcoder: hopefully not reconfiguring servers while youre away ;)
04:37 PM * roycroft decides to tweak the scale while he's at it
04:48 PM JT-Shop: https://pasteboard.co/QDThHHOxSppW.jpg
04:48 PM Tom_L: mine was a gecko keychain
04:48 PM xxcoder: nice work on lizard doll. not sure why 4 parts is there
04:49 PM JT-Shop: lol
04:51 PM Tom_L: no idea where it is
04:52 PM roycroft: ok, the scale is very close now, and i just did 6 full "laps" - 15" back and forth - on the y axis at 3.0a
04:52 PM JT-Shop: nice
04:52 PM roycroft: the travel is as close to 15" as i can eyeball a moving gantry
04:52 PM xxcoder: very nice indeed
04:52 PM roycroft: and i think close enough to go with until i actually cut some test parts
04:53 PM roycroft: i still want to up the microsteps to 32 again just to see if it works
04:53 PM xxcoder: steps should end up nice round number
04:53 PM roycroft: but then i think i'll drop to 8
04:53 PM Tom_L: if you ran the numbers it should be right
04:53 PM Tom_L: numbers don't lie
04:53 PM roycroft: this all assumes i read the lead screw pitch correctly
04:53 PM roycroft: which i needed to measure
04:53 PM JT-Shop: they do on my plasma... it's belt drive and not the expensive belts
04:53 PM xxcoder: 200 steps per 360 degree, which is uhhh 8 mm length?
04:54 PM xxcoder: (whole steps)
04:54 PM roycroft: i ran the numbers earlier, and it was moving about 13" when it should have moved 15"
04:54 PM JT-Shop: so there is a bit of stretch
04:54 PM Tom_L: what's the screw pitch
04:54 PM Tom_L: 5 or 5.08
04:54 PM xxcoder: tom I remember 8 mm, but could be wrong. been too long
04:54 PM Tom_L: ^^ those are typical
04:54 PM roycroft: that's a 13.33% difference, so i increased the scale by 13.33%, and now it's very close
04:54 PM xxcoder: whats current steps now?
04:55 PM roycroft: 16 microsteps
04:55 PM roycroft: iirc i measured 0.1833", which is a bit over 4.6mm
04:55 PM roycroft: on the lead screw
04:55 PM xxcoder: hmm lemme see if I can find info
04:55 PM roycroft: 8mm would be the screw diameter
04:55 PM roycroft: anyway, i have it really close now
04:55 PM xxcoder: ahh thats where 8mm is from
04:56 PM roycroft: when i make a part i'll measure it and i'll know how far off it is
04:56 PM roycroft: and can do the final adjustments
04:57 PM roycroft: i think i want to set the driver back to 3.5a next, and hope that i can reproduce my problem with that setting
04:57 PM roycroft: that would confirm that it's a current issue
04:57 PM Tom_L: you're at 40v now?
04:57 PM roycroft: yes
04:57 PM roycroft: but when i had it at 40v and 3.5a it would still fail
04:57 PM roycroft: at 40v and 3.0a it works fine
04:58 PM roycroft: i'm not going to change the voltage
04:58 PM JT-Shop: the drive shut down?
04:58 PM roycroft: i'm fine with 40v
04:58 PM roycroft: yes, it shut itself down
04:58 PM roycroft: no alarm
04:58 PM JT-Shop: interesting
04:58 PM roycroft: it just shut down completely
04:58 PM Tom_L: next alarm will be blue in color
04:58 PM roycroft: and i could move the axis by hand
04:58 PM xxcoder: so it is decisively amps issue?
04:59 PM Tom_L: i'm remaining undecided
05:00 PM Tom_L: JT-Shop, you did the hex on the lathe with skunkworks' comp right :D
05:00 PM xxcoder: roy is it ballscrews or trapezoid screw?
05:00 PM roycroft: habemus exactoris mortis
05:01 PM roycroft: xxcoder: that's what i aim to determine by increasing the amps back to 3.5
05:01 PM xxcoder: looking at similiar 3040 models trapezoidal is 1404
05:01 PM JT-Shop: lol, no I used a hex collet holder
05:01 PM roycroft: ball screws
05:01 PM xxcoder: ballscrews is "1605"
05:01 PM Tom_L: 5mm
05:01 PM Tom_L: 16mm d
05:01 PM xxcoder: 16 dia 5 mm thread?
05:02 PM roycroft: that would make sense, since i measured (roughly) 4.6ish mm
05:02 PM Tom_L: inch would be 5.08
05:02 PM xxcoder: https://ae01.alicdn.com/kf/S77b1b056f9064e399666b0674f2321e6W.jpg?width=902&height=793&hash=1695
05:02 PM roycroft: and that's an ~11% difference
05:02 PM roycroft: and i was off by ~13%
05:02 PM xxcoder: problem is its for current model, and minirouter is nearly decade old
05:03 PM roycroft: but it doesn't matter much in the end, i think
05:03 PM roycroft: i have it close
05:03 PM roycroft: and i'll dial it in when i make test parts
05:03 PM roycroft: which my understanding is the way everyone does it
05:03 PM roycroft: you calculate to get close
05:03 PM roycroft: you dial in by making parts and measuring them
05:04 PM roycroft: well, almost everyone
05:04 PM roycroft: there are always the whiners who post on youtube and vendor sites saying "i calculated the scale, but my parts came out the wrong size - BAD VENDOR!"
05:05 PM roycroft: anyway, the cats are gone, i just finished what i needed to get done today, and it's less than 6 minutes until quitting time
05:05 PM roycroft: i'm not going to start something new for that short a time
05:06 PM roycroft: i'll go flip the dip switch to 3.5a and hope my machine fails
05:12 PM roycroft: well, it it did not fail at the 3.5a setting
05:12 PM roycroft: i wonder if the time it did before was a fluke
05:12 PM roycroft: or maybe i was jogging it faster
05:12 PM roycroft: i think dropping the voltage was probably what did it
05:17 PM xxcoder: interesting
05:18 PM xxcoder: so you have full benefits of all the amps at say 8 microstep or whatever
05:18 PM JT-Shop: voltage is king with steppers the more you have the faster you can accelerate
05:19 PM Tom_L: technical jargon
05:22 PM lcnc-relay: <ccatlett1984@> well, more of. more voltage = fewer amps needed
05:24 PM Tom_L: i think higher voltage means faster current rise time
05:24 PM Tom_L: on a chopper
05:30 PM lcnc-relay: <captainhindsight_.@> https://www.orientalmotor.com/stepper-motors/technology/stepper-motor-basics.html
05:30 PM Guest3050: dI/dT = V/L
05:30 PM roycroft: are you still about, xxcoder?
05:30 PM xxcoder: yep
05:30 PM lcnc-relay: <captainhindsight_.@> https://www.e-jpc.com/stepper-motor-voltage-explained/
05:31 PM roycroft: hopefully this is not too shaky for you - it was handheld, but as steadily as i could:
05:31 PM lcnc-relay: <captainhindsight_.@> https://www.linengineering.com/news/the-influence-of-operating-speed-on-stepper-motor-selection
05:31 PM roycroft: https://roycroft.us/MiniRouter/MiniRouterTest.mov
05:32 PM roycroft: i don't know if i can make it go faster than that
05:32 PM xxcoder: those 2 links slid right over my silky smooth brain
05:32 PM xxcoder: watching
05:32 PM roycroft: but that's probably an improvement over what you were doing
05:32 PM lcnc-relay: <captainhindsight_.@> https://homepage.divms.uiowa.edu/~jones/step/physics.html
05:32 PM xxcoder: it looks like moking bit faster than I had it at. I bet you could make it move faster
05:32 PM lcnc-relay: <captainhindsight_.@> https://www.linearmotiontips.com/what-is-a-chopper-drive-for-a-stepper-motor/
05:33 PM xxcoder: video was steady enough for me
05:33 PM roycroft: good
05:34 PM roycroft: i could not be bothered to get out a tripod and some duct tape for a 30 second video :)
05:34 PM xxcoder: I woulnt be able to watch more than 2 or 3 videos like that but its plenty for one
05:35 PM roycroft: and of course x and z were moving a lot slower than they could, since i sent the command to move all three at once
05:35 PM roycroft: and y had the farthest to move
05:35 PM xxcoder: ahh thought it was rapid
05:36 PM roycroft: it was just a gcode one-liner
05:36 PM roycroft: g1 x10 y15 z2 f200
05:37 PM roycroft: pretty much my limits
05:38 PM roycroft: i can try pushing f higher to see if it speeds up
05:39 PM xxcoder: yeah
05:45 PM roycroft: i'm starting out conservatively with this
05:45 PM roycroft: and honestly, i think i'd probably be ok if that's the fastest it ever moves
05:45 PM roycroft: but until i start making parts i can't be sure about that
05:46 PM roycroft: i thought the same about my 3d printer until i needed to make a big batch of parts, at which time i bought a volcano hot end with a giant nozzle so i could squirt filament a lot faster
05:52 PM roycroft: so when i ran that job again multiple times, x shut down on the second go
05:52 PM roycroft: so maybe i need to limit to 3.0a anyway
05:52 PM xxcoder: maybe
05:53 PM xxcoder: circle is one of highest dynimatically changing on amp usage etc so might be best, and it can be infinitely repeated
05:53 PM roycroft: oh, i was trying to go faster too
05:53 PM roycroft: i'll reset the job to f200 and see if it works
05:54 PM xxcoder: might want to test both f on 3.0a too
06:08 PM Tom_L: i wrote a test that moves all 3 in all quadrants shifting 45deg then circles cw & ccw
06:09 PM Tom_L: was trying to trigger a following error
06:23 PM roycroft: so at 2.8a, x fails after a minute or so
06:23 PM roycroft: at 2.5a, i ran for about 10 minutes with no failure
06:24 PM roycroft: the psu is rated at 12.5a output current, so there should be plenty of current from the psu, if it's doing what it's supposed to do
06:24 PM roycroft: that's ~4.2a per driver, if they're all drawing maximum current
06:25 PM roycroft: since i have three separate power outputs on the psu it's hard to monitor total current at any given moment
06:31 PM lcnc-relay: <.le_potato@> milling steel, isnt easy
06:31 PM lcnc-relay: <.le_potato@> i broke one tool
06:32 PM roycroft: milling steel just requires the proper tooling, proper feeds, and proper speeds
06:32 PM Tom_L: breaking tools is easy too
06:32 PM lcnc-relay: <.le_potato@> yes
06:32 PM lcnc-relay: <.le_potato@> with aluminium, not that easy
06:32 PM Tom_L: right feed/speed you can cut steel
06:32 PM Tom_L: should have 4 flute cutters for it
06:32 PM lcnc-relay: <.le_potato@> my 18k rpm spindle cant handle low rpms
06:32 PM lcnc-relay: <.le_potato@> no torque
06:33 PM lcnc-relay: <.le_potato@> with 0.1mm and 1800mm/min feedrate
06:33 PM lcnc-relay: <.le_potato@> it was goodish
06:33 PM Tom_L: mess around with flex any more?
06:34 PM lcnc-relay: <.le_potato@> tomorrow! my collet is getting deliverd
06:34 PM lcnc-relay: <.le_potato@> i want to mill something then flip it, with touch get it right position
06:34 PM lcnc-relay: <.le_potato@> and mill from other side :p
06:34 PM Tom_L: get the latest files as they have been changing some lately
06:34 PM lcnc-relay: <.le_potato@> deb or source?
06:35 PM Tom_L: deb
06:35 PM lcnc-relay: <.le_potato@> https://github.com/jethornton/flexgui2
06:35 PM lcnc-relay: <.le_potato@> this one?
06:35 PM Tom_L: should be pretty up to date
06:35 PM Tom_L: no
06:35 PM lcnc-relay: <.le_potato@> oh they added it already!
06:35 PM Tom_L: yes
06:35 PM Tom_L: 1.3.0
06:35 PM Tom_L: iirc
06:36 PM Tom_L: work goes on while you sleep :)
06:38 PM lcnc-relay: <.le_potato@> i love to sleep
06:41 PM roycroft: so it works fine as it is set - 8 microsteps, 2.5a current, 40v
06:42 PM roycroft: and i can move from front to back in <20 seconds
06:42 PM xxcoder: thts pretty fast for that machine
06:42 PM roycroft: so i think i'll leave it set at that for now
06:44 PM Tom_L: is this an inch or metric config?
06:44 PM roycroft: inch
06:44 PM roycroft: for me
06:44 PM roycroft: almost certainly metric for .le_potato
06:45 PM xxcoder: which is little bit funny. when I had that mini router, it was always a metric machine
06:45 PM Tom_L: you could probably get better results than 20 sec
06:45 PM roycroft: yes, and i'll work on that eventually
06:46 PM Tom_L: was looking for my test video but can't find it
06:46 PM Tom_L: your feed was 200 ipm?
06:46 PM roycroft: yes
06:46 PM Tom_L: that's not so bad
06:47 PM Tom_L: iirc mine rapids at 210
06:47 PM Tom_L: but i'm not pushing it
06:47 PM roycroft: but it was really going more like 60ipm
06:47 PM lcnc-relay: <.le_potato@> can i drill with flat mill tool?
06:47 PM Tom_L: if it's center cutting
06:47 PM roycroft: it depends, .le_potato
06:47 PM Tom_L: if not then no
06:47 PM roycroft: if it's a four flute end mill, almost certainly not
06:47 PM roycroft: if it's a 2 flute end mill, quite possibly
06:47 PM Tom_L: some can
06:47 PM lcnc-relay: <.le_potato@> if its 6?
06:48 PM lcnc-relay: <.le_potato@> oh
06:48 PM Tom_L: 2 flutes are cut to the center
06:48 PM roycroft: look at the end of the cutter
06:48 PM Tom_L: on my 3 flute only one is
06:48 PM roycroft: if the flutes overlap so that the center is crossed, then yes
06:48 PM roycroft: if there's a gap in the center then no
06:48 PM lcnc-relay: <.le_potato@> oh, thank you!
06:48 PM roycroft: unless you pilut drill first
06:48 PM roycroft: pilot
06:48 PM lcnc-relay: <.le_potato@> thats looks logical
06:49 PM Tom_L: or helix in
06:49 PM xxcoder: yeah, pilot hole larger than center gap, but smaller than final size you want. yeah helix is also good
06:50 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/cutters/carbide_6flute1.jpg
06:50 PM Tom_L: that 6 flute is center cutting
06:51 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/cutters/ZRN_Alum_3flute1.jpg
06:51 PM Tom_L: as are those but only 1 flute is
06:51 PM lcnc-relay: <.le_potato@> http://6-f.cc/nc/index.php/s/YREwCCB7FzWNC92
06:51 PM lcnc-relay: <.le_potato@> i did this today
06:52 PM Tom_L: very high helix em
06:52 PM Tom_L: for steel?
06:52 PM Tom_L: or hard metals
06:52 PM lcnc-relay: <.le_potato@> i broke it
06:52 PM Tom_L: well i see that
06:52 PM lcnc-relay: <.le_potato@> or this question not to me?
06:52 PM Tom_L: to you
06:53 PM lcnc-relay: <.le_potato@> 0.5mm in to steel
06:53 PM lcnc-relay: <.le_potato@> bamm
06:53 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/cincinnati/2inch_insert_cutter.jpg
06:53 PM lcnc-relay: <.le_potato@> ouch
06:54 PM xxcoder: well now thats with face smashed in
06:54 PM lcnc-relay: <.le_potato@> how
06:54 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/cincinnati/Ti_part.jpg
06:54 PM Tom_L: titanium
06:54 PM lcnc-relay: <.le_potato@> so this happens
06:54 PM xxcoder: yeah not surpised. ti is real mean material
06:55 PM Tom_L: somebody wasn't watching
06:55 PM lcnc-relay: <.le_potato@> i have to mill brake disc tomorrow
06:55 PM lcnc-relay: <.le_potato@> thats other hard material
06:57 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/temp/IMG_2421.jpg
06:57 PM lcnc-relay: <.le_potato@> you like to break tools
06:57 PM Tom_L: fixture was programmed wrong
06:58 PM Tom_L: seems you do too :)
06:59 PM lcnc-relay: <.le_potato@> have to start it somewhere
06:59 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/temp/IMG_2423.jpg
06:59 PM Tom_L: that's what it hit
06:59 PM lcnc-relay: <.le_potato@> haha
07:10 PM roycroft: grr
07:10 PM roycroft: i've been working on designs for my new shop door(s)
07:10 PM roycroft: i wanted a double door that is 60" wide, so i'd be able to have a nice opening
07:10 PM roycroft: but i hate the designs for the two 30" wide drawers that would make up the double door
07:11 PM roycroft: i think i need to either make a single door about 42" wide or make the double door with two 36" wide doors
07:11 PM roycroft: neither of which excites me
07:13 PM roycroft: i'd save a lot of money with a single door
07:13 PM roycroft: i want the door to open to the outside, not the inside, but i do not want exposed hinges on the outside
07:14 PM roycroft: so i need to use hidden hinges like soss, which are ridiculously expensive
07:14 PM roycroft: a single door would cut that cost almost in half
07:14 PM roycroft: at least by 1/3
07:14 PM roycroft: (i would need beefier or more hinges for the bigger door)
07:15 PM roycroft: 42" is wide enough to fit all my machines through the door with ease
07:15 PM roycroft: i had issues moving in with the 36" door - some stuff had to be partially dismantled to fit
07:15 PM roycroft: but i want a bigger opening so i can get more fresh air and sunlight in on nice days
07:19 PM roycroft: i do not need to build the door(s) before the roofing job, but i do want to know what the rough opening will be, as i need to reframe the wall at the front of the shop when i do the roof
07:20 PM xxcoder: wonder if can do "fake wall" which can expand normal door size to wide for when you need it
07:53 PM roycroft: i don't need to build the doors now
07:53 PM roycroft: but i do need to determine the rough opening now
07:53 PM roycroft: i could always reframe the rough opening, but it's a lot easier to do when framing the wall in the first place
07:55 PM roycroft: i was just out sorting the porcelain tiles that i got for my front entry way
07:55 PM roycroft: i think i understand why the seller did not want them - the have a fake stone look, which is actually very good, but there are only 8 patterns
07:55 PM roycroft: and even though there are only 8 tiles per box, there were repeats of the same one in each box
07:56 PM roycroft: fortunately for me, exactly 8 tiles will do my entry way, so i can do it with no repeats
07:56 PM roycroft: i just have 4x as many tiles as i need, but i should be able to sell the rest
07:57 PM roycroft: and even if not, i paid the same price for 4 boxes on marketplace as i would have paid for one box in the store, and i have the luxury of sorting through them and choosing the ones i want to use
07:57 PM xxcoder: nice
08:29 PM lcnc-relay: <ljade31@> Just to make sure I'm posting information that is relevant to this discord. Is hal2udp with esp32 and W5500 questions accepted?
08:30 PM lcnc-relay: <ljade31@> +If so, where would be the best place to ask these questions
08:34 PM xxcoder: hal2udp is lcnc related so yeah this is good place
08:34 PM xxcoder: might need to wait a bit :D
08:34 PM xxcoder: probably ask more specific question, people tend to not answer generic one
08:36 PM xxcoder: im not some ai btw, I'm talking from over the bridge from irc server :D
08:40 PM Tom_L: https://forum.linuxcnc.org/18-computer/52035-esp32-hal2udp-setup-for-linux-cnc-noob
08:40 PM Tom_L: not your post is it?
08:41 PM Tom_L: don't know about hal2udp but someone might
08:55 PM roycroft: i just air cut a bunch of circles
08:55 PM roycroft: the machine went a lot slower, but it did not fail
08:58 PM xxcoder: slower due to set f lower?
09:03 PM lcnc-relay: <ljade31@> Is there any plans to integrate esp32 and w5500 into the standard lcnc platform
09:04 PM roycroft: no, i set it the same
09:04 PM xxcoder: roy so circles at f200?
09:04 PM roycroft: yes
09:04 PM xxcoder: ok
09:04 PM roycroft: g2 j-3 f200
09:05 PM roycroft: and maybe it wasn't really a lot slower, but it seemed to move a lot slower
09:05 PM xxcoder: sum of vector probably stayed 200
09:05 PM roycroft: i should time 10 circles and calculate the actual travel rate
09:10 PM lcnc-relay: <ljade31@> @Tom_L yeah. Thanks for the help
09:24 PM roycroft: aqi is down to 68
09:25 PM * roycroft his happy about that
09:25 PM xxcoder: I bet
09:25 PM roycroft: it needs to get even lower
09:25 PM roycroft: but at least it's going down, not up
09:31 PM roycroft: my doors look a lot better if they are 6" wider
09:31 PM roycroft: i think i can make that work
09:46 PM roycroft: i have a lot of work to do on the door design still, but i can live with the dimensions, so i know my rough opening size
09:46 PM xxcoder: thats good
09:46 PM roycroft: yeah
09:47 PM Tom_L: arcs would seem slower i think because the feed is sent to multiple axis
09:47 PM roycroft: yeah, i agreee
09:48 PM xxcoder: yeah. thats why I think vector was 200 whole time :)
09:48 PM Tom_L: i've watched that in flex showing the per axis feedrates
09:50 PM roycroft: i have all three axes working reliably now
09:50 PM roycroft: they can still use some tweaking, but they work
09:50 PM roycroft: it's time to move on from that
09:50 PM roycroft: for now
09:51 PM roycroft: i'll get back to the e-stop/ladder issue next, when i'm able to work on it
09:51 PM roycroft: but i really need to focus on my shop roof for now
09:56 PM Tom_L: ljade31, no plans i'm aware of
09:56 PM lfluffywof: my replacement Y axis leadscrew came in yesterday. will see if I have time this weekend to put it in.