#linuxcnc Logs

Sep 11 2021

01:33 AM Deejay: moin
05:07 AM JT-Cave: morning
07:59 AM t4nk_freenode: morning, https://postimg.cc/F7NK8qRH
08:00 AM t4nk_freenode: made a connector yesterday, so I could test my Ender's psu
08:00 AM t4nk_freenode: it turned out to be my psu after all... just couldn't keep up with several steppers pulling at the same time
08:01 AM t4nk_freenode: it also seems that my homing issue on the second y-axis has disappeared with the new psu
08:03 AM t4nk_freenode: so now I can hopefully concentrate on trying to do some actual routing
08:04 AM t4nk_freenode: I was thinking of designing a new front plate for my z-axis, making four circular slots so I can 'Tram' the spindle
08:05 AM t4nk_freenode: but I'm not sure it would hold securely enough.. so maybe I'll make two slots, then tram with that... and then drill the remaining two holes by hand so it can't move afterwards
08:09 AM t4nk_freenode: or... just make slots, no washers.. and tighten the hell out of them there bolts so the nuts will eat into the mdf
08:10 AM t4nk_freenode: ahhhhhh the life of an unrecognized genius.... *sigh*
08:10 AM t4nk_freenode: :b
09:40 AM t4nk_freenode: .. or was that Unrecognizable
09:40 AM t4nk_freenode: :|
11:22 AM -!- #linuxcnc mode set to +v by ChanServ
12:07 PM norias: hi
12:08 PM Tom_L: hello
12:10 PM norias: how are you?
12:10 PM Tom_L: pretty good
12:10 PM t4nk_freenode: as it should be, for the channel's Tram-master
12:10 PM t4nk_freenode: !
12:11 PM Tom_L: still have most of my parts although some are fading
12:12 PM * t4nk_freenode has a pair of glasses since a year or so.
12:12 PM t4nk_freenode: I even make use of em, sometimes
12:52 PM roycroft: i have a couple questions about running linuxcnc on a raspberry pi
12:53 PM roycroft: 1. would an rpi 4 with 8GB of ram and a 64GB microsd card be sufficient to run linuxcnc?
12:53 PM roycroft: 2. if i want to use a mesa ethernet board do i need a second dedicated ethernet port on the rpi?
12:54 PM * roycroft thinks the answer to the first question is yes, but would like confirmation
01:18 PM Tom_L: 1. yes
01:18 PM Tom_L: 2. no
01:18 PM Tom_L: wifi works good for normal internet
01:19 PM roycroft: i suppose i could do that
01:19 PM Tom_L: dedicate the ethernet port for mesa
01:19 PM roycroft: it would save having to run an ethernet cable to the machine
01:19 PM Tom_L: https://www.amazon.com/gp/product/B07XTQL6YZ/
01:19 PM roycroft: i've been reading some threads and websites that talk about issues with using an rp4 with >4GB of ram
01:20 PM Tom_L: pretty sure that's the one i got
01:20 PM roycroft: usb not working
01:20 PM roycroft: but andypugh persistently posts that he cannot find a problem
01:20 PM Tom_L: i haven't had any issues but also haven't run it for an extended period
01:20 PM Tom_L: afik 4g should be plenty
01:21 PM roycroft: well the issue reported is that usb does not work at all
01:21 PM Tom_L: you sure that wasn't on the rpi not the rpi4?
01:21 PM roycroft: and i'm thinking a 7i76e would be the mesa board for this application
01:21 PM Tom_L: iirc they had issues on the rpi
01:21 PM roycroft: i'm reading about rpi4
01:22 PM Tom_L: my kbd and mouse dongle are usb and work fine
01:23 PM roycroft: i decided that i can probably do x11 forwarding over ssh to remote manage the machine
01:23 PM roycroft: and i may get a pi for testing that while i'm working on the rest of the design
01:24 PM roycroft: i really want to avoid having to connect a display and keyaboard directly to the machine - i've found a place already where i can put all the control hardware inside my router table
01:25 PM Tom_L: i also did get an HDMI to vga adapter but only used it on the old monitor on the mill
01:26 PM roycroft: i just realized that i have another issue - my router table has a 240vac power inlet, and it's 3 wire - i don't run the split-phase wire to it
01:27 PM roycroft: so i may need a step-down transformer for anything that expects 120vac
01:27 PM Tom_L: so when/if you get it, i ran the iso then uninstalled the dev version of linuxcnc (forget what it was called on the iso) then run the .deb for 2.8.2 to be up to date
01:27 PM Tom_L: i'm not sure the rpi iso was updated for the new version...
01:27 PM roycroft: ok, i'll keep that in mind
01:27 PM Tom_L: but that will get you buster installed then you can update linuxcnc
01:28 PM Tom_L: i had to uninstall linuxcnc that was on it because it was a 'pre' release
01:28 PM roycroft: some folks are talking about using arch linux instead of debian on their rpis
01:28 PM Tom_L: that was a while back and may have been addressed by now
01:28 PM roycroft: i use debian for most of my servers, though, and would probably stick with that unless there's a compelling reason not to
01:28 PM Tom_L: meh
01:28 PM Tom_L: this path would likely be the simplest
01:29 PM roycroft: i like simple
01:29 PM roycroft: unfortunately i tend to do a fair amount of overthinking and overcomplicating before i'm able to get to the simple, clean solution for something
01:29 PM roycroft: but i do try to get to simple in the end
01:29 PM Tom_L: https://gnipsel.com/linuxcnc/uspace/rpi4-emc.html
01:30 PM Tom_L: also he's got instructions how to rename the user from pi to roycroft if you want
01:30 PM Tom_L: you can probably figure that out on your own though
01:31 PM Tom_L: the lcnc versions have changed since he did those pages
01:31 PM Tom_L: you don't need the steps to build anything. use the iso and .deb
01:32 PM roycroft: i don't think i'll have any trouble installing linuxcnc
01:32 PM roycroft: i've done that many times before, including building custom kernels back when that was necessary
01:33 PM Tom_L: iso: http://linuxcnc.org/downloads/
01:33 PM Tom_L: deb: http://buildbot.linuxcnc.org/dists/buster/2.8-rtpreempt/binary-armhf/
01:34 PM Tom_L: you don't necessarily want the debug sim or dev versions
01:34 PM roycroft: are folks running it on bullseye yet?
01:34 PM Tom_L: linuxcnc-uspace_2.8.2.39.gade80dd4f_armhf.deb looks like the latest
01:34 PM Tom_L: not yet
01:34 PM roycroft: no big deal
01:35 PM roycroft: i'm never one for running the latest bleeding edge software
01:35 PM roycroft: but i also don't want to intentionall install something old if a newer version is stable
01:35 PM roycroft: buster is fine for now and probably a long time
01:35 PM Tom_L: i'm running wheezy on the mill :)
01:35 PM roycroft: especially for a device that won't be directly connected to the internet
01:36 PM Tom_L: buster on the rpi
01:36 PM roycroft: i think i'll figure out how i want to package the control hardware and get a pi ordered soon
01:37 PM Tom_L: i also like that double sided heatsink personally. no fan
01:37 PM roycroft: i'm not sure i want that bundle you linked to - if i end up buliding an enclosure for all the control hardware i'll just mount the pi directly to that and won't need an enclosure for it
01:37 PM Tom_L: i like it because it has a psu with a switch, hdmi cable etc
01:38 PM Tom_L: i didn't use the case that came with it
01:38 PM roycroft: and i won't need the hdmi cable
01:38 PM Tom_L: you might to get set up
01:38 PM roycroft: and would probably use something else for the power supply
01:38 PM roycroft: i have hdmi cables
01:38 PM Tom_L: mini
01:38 PM roycroft: yes
01:38 PM Tom_L: k
01:38 PM roycroft: i have other rpi4s
01:38 PM Tom_L: oh
01:38 PM roycroft: they're being used for other things though, so i can't easily borrow one for testing
01:39 PM roycroft: and i don't have any spares
01:39 PM roycroft: it would be nice to have a spare rpi4 in case one of the production ones dies
01:39 PM Tom_L: skunkworks has some config changes to make it a little better too when the time comes
01:39 PM roycroft: cool
01:39 PM Tom_L: i had it but not sure which one or where it is or i'd share
01:39 PM roycroft: i was thinking of getting a gecko 540 for the router
01:40 PM roycroft: but i may use the individual gecko drives i bought for the mill conversion for this router
01:40 PM roycroft: and replace them
01:40 PM roycroft: i'm not sure
01:41 PM roycroft: the 540 would be easier to install
01:41 PM roycroft: i have this thing about avoiding combo hardware though
01:41 PM Tom_L: i went with single drivers
01:41 PM roycroft: if one component fails i have to replace the whole thing
01:41 PM Tom_L: 302v? iirc
01:41 PM roycroft: however, the 540 comes with 4 drivers, and i need 3
01:41 PM roycroft: so it has a built-in spare
01:42 PM Tom_L: geckos are pretty bullet proof
01:42 PM roycroft: that's what i'm told, and is why i bought some
01:42 PM roycroft: as i said, i have a space inside the router table where i can mount the controller hardware
01:43 PM Tom_L: the 203v probably has the most protection
01:43 PM roycroft: my initial thought was to just built the hardware moutns into it
01:43 PM Tom_L: and a higher working voltage than the 540
01:43 PM roycroft: but maybe i should design an enclosure that i can then fit inside the router table, so i that i can pull it out if i ever get the space to have a stand-alone dedicated cnc router
01:44 PM roycroft: i'll check the 203v out
01:44 PM roycroft: i don't know yet what voltage i need to drive the motors
01:44 PM Tom_L: they've since introduced some digital versions i know little about
01:44 PM Tom_L: iirc they're good for 80v
01:44 PM Tom_L: 70 or 80 i forget
01:46 PM Tom_L: https://www.geckodrive.com/gr214v-bulletproof-high-resolution-stepper-drive.html
01:46 PM Tom_L: the 203v on steroids
01:46 PM roycroft: i have g201x drivers for my mill
01:47 PM Tom_L: same basic drive with a little less protection
01:48 PM roycroft: maybe 203v for the mill and install the 201x in the router?
01:48 PM Tom_L: not a bad call
01:48 PM Tom_L: i'm only using 48v on mine
01:48 PM roycroft: i have a 48v power supply for my mill
01:49 PM roycroft: which is actually more like 54v
01:49 PM Tom_L: mine is analog because it was surplus local
01:49 PM roycroft: mine as well
01:49 PM roycroft: it's a telco power supply, rack mountable
01:49 PM Tom_L: 3 parallel'd transformers (1 probably would have been sufficient)
01:49 PM roycroft: so it's actually -48v
01:49 PM roycroft: but i can fix that :)
01:49 PM Tom_L: 18A
01:50 PM roycroft: iirc mine is 12a
01:50 PM Tom_L: still overkill probably
02:02 PM roycroft: i built my router table about 5 years ago, as a prototype
02:03 PM roycroft: it's been working well, but there are some design changes i wanted to make before building the prodution table, and i've already made many of those changes in my cad drawings
02:04 PM roycroft: i also stashed a few hundred dollars in a bank account back then for building the production router table
02:04 PM roycroft: that money is still laying around
02:04 PM roycroft: i'm starting to think that instead of bolting a cnc router on the existing design, i should take the time to redesign the whole thing to integrate the router better
02:07 PM roycroft: or maybe move my router lift over to the table saw and minimise the user of the router table, turning the current router table into a dedicated cnc router
02:08 PM roycroft: i have to really think about that, because a cnc router really can't replace the router table completely
02:23 PM t4nk_freenode: seems to have really been a problem with the psu.. I just started up again, and even the homing cycle is fixed
02:23 PM t4nk_freenode: so... I did plot a path with freecad... clamped down some mdf
02:23 PM t4nk_freenode: homed the machine
02:24 PM t4nk_freenode: first thing I do? I ram the spindle into the clamp.
02:24 PM t4nk_freenode: lol
02:29 PM t4nk_freenode: ok, I'm set to cut.. here I go
02:30 PM * roycroft is formulating a cunning plan, and hopes he's more clever than baldrick, so it has a chance of succeeding
02:38 PM roycroft: if i run the output of a vfd through a 3-pole rotary switch (with proper power rating), am i likely to have emf issues due to the switch?
02:38 PM roycroft: or some other kind of 3-pole switch
02:48 PM enleth: roguish[m]: what for?
02:48 PM enleth: damn
02:48 PM enleth: roycroft: what for?
02:49 PM enleth: any kind of switch between VFD and motor is a weird and possibly bad idea
02:51 PM unterhaus: a switch between the motor and the vfd is a good way to blow the output devices of the vfd
02:51 PM unterhaus: if you don't ever switch it, it's okay
02:52 PM enleth: I'm not 100% sure if all VFDs are at risk of damage in this situation, but the manuals certainly say so and there has to be a reason
02:53 PM enleth: specifically, they always say not to put a contactor between the VFD and motor as a part of the e-stop circuit
02:53 PM enleth: as this is the most likely reason for anyone to put anything in there
02:55 PM unterhaus: my servo drive manual has an estop circuit that shorts the motors to ground through resistors, uses 6 contactors
02:56 PM unterhaus: I think they just wanted you to blow your drive and buy a new one
02:57 PM unterhaus: but they also say the drive can take a dead short. Unfortunately, I proved them wrong on that claim
02:58 PM enleth: I suspect the risk of damage comes from the fact that no contactor disconnects all three phases in perfect sync, so there's going to be this one phase that diconnects last and has a chance of getting hit by rising volage generated by the motor
02:58 PM enleth: if the delay is significant enough for voltage to build up
03:00 PM unterhaus: could be that older power devices were more fragile
03:00 PM enleth: certainly, and had control systems that were slower and dumber
03:01 PM enleth: I would not be surprised if a modern servo drive/VFD of good enough make could easily take a deliberate braking short on the output, simply by detecting it fast enough
03:02 PM enleth: so you'd use a single 3-phase contactor to short the motor onto a braking resistor array, without even bothering to disconnect it
03:03 PM JT-Shop: no emf after you switch the output off and blow up the vfd
03:10 PM JT-Shop: these are definitely not M&M's in this bag of trail mix... they melt in your hand and not in your mouth lol
03:20 PM * hazzy[m] uploaded an image: (251KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/rLTKGLuqJuSrMXLWTsxsLiAp/ima_8767889.jpeg >
03:20 PM * hazzy[m] uploaded an image: (266KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/CayHIoPRWUdPaIqHoQHyIfAy/ima_9284a81.jpeg >
03:21 PM JT-Shop: amazing hazzy[m]
03:21 PM hazzy[m]: Need a heavy duty truck toolbox? I'll be forming these as a demo at FABTECH and giving them away to whoever wants one
03:21 PM TurBoss: yo
03:21 PM TurBoss: :P
03:21 PM hazzy[m]: Thanks JT!
03:22 PM hazzy[m]: TurBoss: hey!
03:22 PM TurBoss: :D
03:22 PM JT-Shop: can you email the truck box to me?
03:22 PM TurBoss: but zip it
03:22 PM JT-Shop: lol
03:22 PM hazzy[m]: I'll have to compress it first
03:22 PM hazzy[m]: 😂
03:23 PM TurBoss: tha box has realy nice endings
03:24 PM TurBoss: the machine not the operator should be good
03:24 PM hazzy[m]: It's a $500 box. 12ga mild steel. They hardware (hinges and latch/lock) are readily available. I hope at least a few people take them, or else they will end up in a dumpster
03:24 PM hazzy[m]: TurBoss: hey, I'm a pretty decent operator too! xD
03:25 PM TurBoss: 🤣
03:55 PM Tom_L: roycroft, put a switch there and let us know how it works out :)
03:56 PM Tom_L: just have a spare vfd on hand
04:04 PM t4nk_freenode: hey XXCoder, I'm glad I got those flutes.. just ran some more tests, got no idea about any of this of course ;) but I'm pretty pleased. Did some holes, some engraving, and that flute does a nice job of leaving a nice surface I must say
04:07 PM t4nk_freenode: wasn't sure what rpm to use though.. gave it half power so that should be 6k, don't know if it is ok to take the rpms down a lot but I'd like to
04:07 PM t4nk_freenode: not as loud
04:08 PM t4nk_freenode: well... my vacuum cleaner makes the noise
04:08 PM roycroft: tom_l: i would wire it such that the switch cannot engage while the vfd is powered
04:09 PM roycroft: i'm well aware of what happens to a vfd if the load is suddenly removed
04:09 PM Tom_L: still not a great idea
04:09 PM roycroft: i think it can be done safely
04:09 PM roycroft: i'm just concerned about noise
04:37 PM XXCoder: t4nk_freenode: nice
04:41 PM Tom_L: roycroft, on my first sherline powered spindle, i used relays to get fwd rev from a fwd only control by adding a few ms delay in linuxcnc to the switching time
04:41 PM Tom_L: mostly done to save the relay contacts
04:41 PM Tom_L: of course, the delay wouldn't apply to your situation
05:23 PM roycroft: the idea i'm toying with is redoing my router table completely, to turn it into an over/under cnc router
05:23 PM roycroft: with two spindles
05:23 PM roycroft: the gantry would have a bolt-on upper section to hold the overhead spindle, and a bolt-on fence to use with the lower spindle
05:24 PM roycroft: i'd leave the gantry rails installed permanently, and just switch out the upper spindle and the fence as needed
05:24 PM roycroft: if i did that i'd like to use a single vfd for both spindles, since there would never be a case where both would be used at the same time
05:25 PM roycroft: i love my incra jig - this is the third one i've had, and i've used them all heavily - but if i had a computer to set the fence position for me that would be better than using the incra jig
05:26 PM roycroft: and since i'm planning on building the cnc router anyway, it would be a minimal expense to add cnc control of the router fence
05:26 PM roycroft: it would cost less to add that than i can get from selling my current incra jig
05:27 PM roycroft: and actually, the gantry upper section could be permanently attached - i'd just need to put the fence on the front when i need the lower spindle, and remove it when i need the upper spindle
05:29 PM _unreal_: ok so.... I'm futzing around with my marine stuff
05:29 PM _unreal_: trying to see what I can sell and what is g... and NFG.....
05:29 PM _unreal_: one of thehings is a HUGHES hx200
05:30 PM Tom_L: roycroft, different vfd settings for each spindle. you'd have to reset them each time
05:30 PM _unreal_: would not power on. WTF... so I open the case. it can take 48v or 110 standard PC style cord.
05:30 PM _unreal_: in the other socket
05:31 PM _unreal_: so I'm looking at it looking at it. plug it back in turn the power on again. instantly get a loud ""purple smoke"" POP. though I got no smoke
05:31 PM _unreal_: and then nohting happend WTF
05:31 PM _unreal_: so screw it why not.... turned the power switch off and back on and the thing posted
05:31 PM _unreal_: ????
05:31 PM _unreal_: dumb founded
05:32 PM _unreal_: also is any one looking for a nice low profile rack mount style housings for cnc controllers?
05:33 PM _unreal_: I have 2 I'm looking to get rid of. both are for a different use but are the perfect setup for a CNC controller. already have a 200va 24v transformer with relay power control and other features.
05:34 PM _unreal_: and a second psu providing 5v and 12v
05:34 PM JT-Shop: crap I have a 30" diameter white oak tree to cut down but I have to cut my 24" diameter burled up hickory first
05:35 PM Tom_L: JT-Shop, you keep geting more work to do...
05:36 PM Tom_L: both for lumber i presume
05:36 PM JT-Shop: lumber and firewood, that makes 9 dead trees I have to cut down
05:37 PM XXCoder: tom reward for job well done is more jobs ;)
05:37 PM _unreal_: also is anyone looking for any kind of CNC hardware? I'm trying to get rid of stuff
05:39 PM _unreal_: quiet tonight
05:40 PM Tom_L: dead ones can be tricky, especially if climbing is involved
05:41 PM XXCoder: i remember one time my grandfather had someone come in to cut one tree down that was threatening house
05:42 PM _unreal_: what kind of threats was the tree making?
05:42 PM _unreal_: other then the middle branch
05:42 PM XXCoder: guys used rope and cut high up first, and pulled it as it was cut off to make fall direction certain
05:42 PM XXCoder: then did again for middle. they didnt bother with rope for bottom
05:43 PM XXCoder: man been long time ago. over 30 years
05:43 PM XXCoder: we finally saw house again few years ago. looked pretty different, and road had been moved
05:43 PM XXCoder: i wish we could walk around into backyard
05:43 PM XXCoder: curious if forest area was same still
05:44 PM Tom_L: i'm sure the trees would have been 30 yrs older or gone
05:44 PM XXCoder: yeah
05:44 PM XXCoder: most of trees was "trash trees"
05:44 PM XXCoder: so probably most would be new pones
05:46 PM _unreal_: XXCoder, have you looked at that area using google mapes?
05:46 PM _unreal_: maps
06:07 PM JT-Shop: so solidworks says I can fit the 30" diameter white oak in the saw mill and end up with a 21" square... biggest I can joint is 8"
06:10 PM Tom_L: time for some bigger tools?
06:13 PM JT-Shop: gotta think this through, I don't see a need for any boards wider than 8" but I might be wrong
06:13 PM CloudEvil: chainsaw is pretty easy to rig up for a one-two off
06:13 PM JT-Shop: not really
06:14 PM JT-Shop: I have an Alaskan Chain Saw with a 48" bar... and have used it on a bigger red oak
06:15 PM JT-Shop: https://www.youtube.com/watch?v=aOYR5xgJ0bQ
06:16 PM JT-Shop: I had to cut that into quarters to fit in the sawmill
06:21 PM Tom_L: i bet that took a bit to whittle down
06:40 PM XXCoder: isnt that better done with 2 persons?
06:41 PM XXCoder: i guess theres someone else with camera
08:46 PM * roycroft wonders why folks think he's going to blow up his vfd
08:57 PM * norias shrugs.
08:58 PM Tom_L: disconnecting with _any_ power could cause a spike that would damage the mosfet drivers
08:58 PM norias: there ya go
08:58 PM Tom_L: motors can generate such potential on their own if turning
08:58 PM Tom_L: unless you can assure that won't happen, it's not a good idea
08:59 PM roycroft: the application is such that it wold not happen
08:59 PM Tom_L: back emf is generally calculated at 2x the rated voltage i believe
08:59 PM roycroft: the over/under design is not one that can work both ways at the flip of a switch
09:00 PM Tom_L: feel free to move forward with it then
09:00 PM roycroft: i would have to power down, remove/install some parts, and then power back up
09:00 PM Tom_L: not a project i would take on personally
09:00 PM roycroft: i'm still not understanding the concern
09:00 PM roycroft: what i'm hearing is that once i connect a motor to a vfd it is dangerous to ever disconnect it
09:00 PM roycroft: and that doesn't make sense to me
09:01 PM _unreal_: thats bullshit
09:01 PM Tom_L: if it is under power
09:01 PM _unreal_: unless
09:01 PM _unreal_: err ^^^^^
09:01 PM roycroft: if i install this switch it will be inside the router table
09:01 PM roycroft: inside the control cabinet
09:01 PM roycroft: i would have to open up the machine to turn the switch
09:01 PM _unreal_: roycroft, the only thing with VFD's is that if they require tuning by setting settings by jumpers OR serial setup'
09:01 PM roycroft: it's not something that would ever happen accidentally
09:01 PM _unreal_: then you must re-config
09:02 PM roycroft: _unreal_: i would be using identical spindles for both the overhead and underside
09:02 PM _unreal_: My VFD could be damaged if the wrong settings are used
09:02 PM Tom_L: also swapping motors would require different vfd settings like _unreal_ is suggesting
09:02 PM _unreal_: then there is no issue
09:02 PM roycroft: so the vfd config should be identical for both
09:02 PM _unreal_: so long as there are no wiring issues
09:02 PM _unreal_: mistakes
09:02 PM Tom_L: any pid in it would need tuning for the particular motor etc
09:02 PM roycroft: i'm not discounting the concerns
09:03 PM roycroft: i just don't think the apply to this situation
09:03 PM roycroft: they
09:03 PM Tom_L: put it to the test
09:03 PM _unreal_: Tom_L, true though like motors should be or be nearly identical
09:03 PM Tom_L: my understanding was that they were rather different
09:04 PM roycroft: i just said that i would use identical spindles
09:04 PM _unreal_: if there is more to the story I know nothiing about it
09:04 PM Tom_L: i missed that
09:04 PM roycroft: i have a 2.2kw 80mm 2400rpm er20 spindle in my router table
09:04 PM roycroft: i'm looking at using an identical one for the cnc router
09:05 PM _unreal_: 2400 RPM or 24,000
09:05 PM roycroft: 24000, sorry
09:05 PM _unreal_: ok
09:05 PM roycroft: same frequency, same # of poles, same power, same speed
09:05 PM _unreal_: I was going to say.. typically VFD and low rpm high torque are not mixed
09:05 PM roycroft: the vfd config should be identical for each
09:05 PM _unreal_: ya then its just carbon copy
09:05 PM Tom_L: so why do you need a contactor? just make screw terminals etc to swap them while unpowered
09:06 PM roycroft: i'm trying to streamline the conversion procedure
09:06 PM Tom_L: i still think there is potential for a spike when the contactor engages again
09:06 PM roycroft: turning a switch is faster/easier than unscrewing three wires and moving them
09:07 PM roycroft: but if the switch introduces noise, then i can move the wires every time
09:07 PM Tom_L: no ee here so try and see
09:08 PM _unreal_: BLDC spindles make a shit load of e noise... its good to use shielded wires going to the motor
09:08 PM roycroft: i'm assuming that on the linxcnc side this will be pretty easy to deal with
09:08 PM Tom_L: i have a contactor on my vfd but it's on the mains side
09:08 PM roycroft: i would just have two profiles - one for the conventional router table and one for the overhead spindle
09:08 PM _unreal_: I had some issues with my 7i96 till I used shielded wires to my 1.5KW
09:08 PM Tom_L: linuxcnc can handle just about anything
09:08 PM _unreal_: VFD/spindle
09:09 PM Tom_L: you may need to run 2 configs though
09:09 PM Tom_L: which would require restarting linuxcnc
09:09 PM _unreal_: brb checking on the grill
09:09 PM roycroft: yes, and that should be no problem
09:09 PM _unreal_: my daughter snuck the @#$@!#$%!Q@#$@# ribs into the freezer and curled it
09:09 PM _unreal_: what a pain
09:09 PM roycroft: this is not something i'll be doing constantly
09:09 PM roycroft: i anticipate that i'll have it set up in the conventional router table mode almost all the time
09:10 PM Tom_L: even more reason to use screw terminals :)
09:10 PM roycroft: and have to switch to cnc router mode occasionally
09:10 PM roycroft: most of what i do with a router cannot be done with a standard cnc router
09:11 PM roycroft: much of what i can do with a router that can be done with a cnc router can be done faster manually
09:12 PM Tom_L: cnc just takes the redundancy out of of repeatibility
09:12 PM roycroft: i'll think about it some more
09:13 PM roycroft: but if i can be certain that neither spindle is moving at all, and i use a break before make switch to select the one i want, i don't see any chance of damaging the vfd
09:14 PM roycroft: the incra jig, while very repeatable, is prone to error if one is not very careful
09:14 PM roycroft: it indexes to 1/32" very accurately, but it's also very easy to be off by 1/32" due to parallax
09:14 PM roycroft: i don't make that mistake often, but i do make it
09:14 PM roycroft: having linuxcnc move the fence for me should eliminate the possiblity of that error
09:16 PM roycroft: also, cutters are not very accurate in size, and i often have to make adjustments of a few thousandths to get things like finger joints to mesh perfectly
09:16 PM roycroft: that can be done with the incra jig, but it's tedious to do
09:16 PM Tom_L: cutters shouldn't be off that much
09:16 PM roycroft: i'd think that i could just tell linuxcnc that my cutter is 0.002" undersize, and it would move the fence for me appropriately
09:17 PM roycroft: they are
09:17 PM _unreal_: if your breaking with a vfd/spindle. make sure you have a breaking resistor
09:17 PM Tom_L: yeah mine has an internal one but i added an external brake resistor
09:17 PM _unreal_: USE THE EXTERNAL
09:17 PM roycroft: almost all cutters are off by a few thousandths
09:17 PM Tom_L: and my break routine has a fairly slow ramp time
09:18 PM Tom_L: maybe metal cutters are ground more accurately
09:18 PM Tom_L: for what they cost they should be
09:18 PM roycroft: the vfd for my router table does not have the external resistor circuitry
09:18 PM roycroft: i just have it ramp down in 20 seconds
09:18 PM roycroft: and it's fine
09:18 PM Tom_L: mine is quicker than that but still...
09:18 PM _unreal_: as much as I hate to fund the CCP I tend to purchase the china pos bits
09:18 PM Tom_L: i didn't want to push it
09:18 PM roycroft: i could probably stop mine faster
09:18 PM roycroft: but 20 seconds is not a big deal to me for that machine
09:18 PM Tom_L: i did decrease the ramp up time
09:19 PM roycroft: i ramp my belt grinder down in about 10 seconds
09:19 PM _unreal_: 20 seconds to ramp down?
09:19 PM roycroft: yes
09:19 PM _unreal_: wholy hell
09:19 PM roycroft: for a router table that's not a big deal
09:19 PM Tom_L: mine takes maybe 5 or less
09:19 PM _unreal_: I can ramp my 24k RPM 1.5kw spindle down to a stop in about 5 seconds
09:20 PM Tom_L: just doesn't snap to a stop
09:20 PM roycroft: i can probably ramp mine down faster
09:20 PM roycroft: but it's really not a big deal to me
09:20 PM Tom_L: yeah
09:20 PM _unreal_: but my external breaking resistor heats like a mofo
09:20 PM roycroft: i never tried faster
09:20 PM Tom_L: my first time with a vfd so i was a bit skeptical changing parameters
09:20 PM Tom_L: but i'm glad i did now
09:21 PM Tom_L: the numbers for mine are in my setup pdf
09:21 PM Tom_L: factory, shipped and current settings
09:21 PM roycroft: i don't think you were around earlier when i was discussing my idea, _unreal_, but i think i came up with something that will work for me well today
09:21 PM _unreal_: hence I have 2x resistors in series at half value to give the correct ohms. on a very aggressive heat sink setup with thermal grill epoxy.
09:22 PM Tom_L: i'm sure each vfd would be quite different
09:22 PM roycroft: i was planning on building a cnc router that sits on top of my router table, and can be assembled/disassembled quickly and easily
09:22 PM roycroft: because i don't have the space for a stand-alone one
09:22 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/configs/NEW_MILL_SHOP_2.8.1/New_Mill_IO_Pinout.pdf
09:22 PM Tom_L: last page
09:22 PM _unreal_: roycroft, what is your main use for this new system
09:22 PM _unreal_: ?
09:22 PM Tom_L: basically only changed 3 parameters iirc
09:22 PM roycroft: it occured to me this afternoon that if i rebuild the router table, i can do cnc control of both the underside router and the overhead router
09:23 PM roycroft: for the cnc router primarily to make jigs/fixtures
09:23 PM _unreal_: and ever thought about a vertical mill?
09:23 PM roycroft: 100% of the intended use would be 2d milling of mdf and plywood
09:23 PM Tom_L: space constraints
09:23 PM _unreal_: a vertical mill you can mount to a wall
09:23 PM roycroft: i have a big incra jig on my router table
09:23 PM _unreal_: OH SHIT roy look into a vertical mill them
09:23 PM roycroft: i'm thinking if i attach a fence to the front of the cnc router gantry i can use that as my router table fence
09:24 PM _unreal_: if your milling PLY regardless of the material type.... it would be in your best interest
09:24 PM roycroft: and instead of using the incra jig to index the fence i can use linuxcnc
09:24 PM roycroft: it would use the same stepper motors as the cnc router
09:24 PM roycroft: i would continue to manually control the z on my router table, for now at least
09:24 PM roycroft: but i could control the spindle with linuxcnc
09:25 PM roycroft: and that's why i'm thinking about a switch on the vfd output
09:25 PM roycroft: so i can select the spindle i need
09:26 PM roycroft: with my original plan i would have to remove the incra jig and install the whole cnc router setup every time i wanted to do cnc routing
09:26 PM roycroft: with this new idea i could leave the cnc router installed all the time
09:26 PM roycroft: and just install a fence on the front of the gantry whenever i want to use the router table
09:27 PM roycroft: i'd have to remove the spoil board as well, but that would be no big deal
09:27 PM _unreal_: got photos?
09:27 PM roycroft: of an idea that is in my head right now?
09:27 PM _unreal_: you've got a lot going on descriptively
09:28 PM roycroft: yes, and i'm going to start making some drawings in solidworks
09:28 PM roycroft: i'll post some pictures when i make some progress there
09:28 PM _unreal_: well if your looking for any design aid let me know. you've seen my work
09:28 PM roycroft: the idea of a combination over/under cnc router just came to me this afternoon
09:29 PM roycroft: my initial thoughts are that it would solve more problems than it would create
09:29 PM _unreal_: Your biggest issue would be inversion of GCODE and material SAG depending on what is being milled and how much material loss there is.
09:29 PM _unreal_: as well work area size
09:29 PM roycroft: i don't think that will be a problem at all
09:30 PM XXCoder: i want to see picture of that someday
09:30 PM roycroft: linuxcnc would think i have two completely different machines
09:30 PM roycroft: not one machine with two spindles
09:30 PM _unreal_: yes true. but the inversion would still have to be done
09:30 PM roycroft: and the router table would be single axis, at least at first
09:31 PM roycroft: all i need to do with the router table is move the fence back and forth
09:31 PM roycroft: maybe down the road i'll get clever and build a z axis that linuxcnc can control
09:31 PM roycroft: but right now i have a nice incra router lift with a dro on it that i can dial in to 0.001"
09:31 PM roycroft: and i'm fine with that
09:32 PM roycroft: my need and my use are uncommon, at least
09:33 PM roycroft: probably not unique
09:33 PM roycroft: but certainly uncommon
10:00 PM norias: hmm
10:00 PM norias: been trying to run windows in a VM so I can use mastercam on my ubuntu laptop
10:00 PM norias: tried the same disk image with both kvm / qemu and virtualbox
10:00 PM norias: virtualbox says it can use the GPU, no such luck with kvm
10:01 PM norias: but, it definitely runs better on kvm
10:29 PM roycroft: kvm is fine for running linux instances, but it sucks for windows
10:29 PM roycroft: virtualbox is just useless, period
10:29 PM XXCoder: vbox works fine for me when i need to run windows.
10:29 PM XXCoder: very rare but yeah
10:30 PM roycroft: i tried both on my linux laptop, and ended up just running windows natively on the laptop in its own partition
10:38 PM roycroft: abom's mispronounciations i can usually ignore, but it bugs me that he pronounces mitutoyo "mitch-uh-toy-yuh"
10:38 PM roycroft: that's not even close to how it is pronounced
10:39 PM XXCoder: but not worse possible
10:39 PM XXCoder: if I attempt to speak that, doubt you would reconize it as any word
10:39 PM roycroft: you have an excuse
10:39 PM roycroft: i mean, i suppose he kind of does - he's from the south :)
10:39 PM XXCoder: lol ok
10:41 PM TurBoss: I use to boot a external ssd usb 3.0 from kvm
10:42 PM TurBoss: :-)
10:42 PM TurBoss: very handy to install things
10:42 PM roycroft: kvm per se is fine
10:42 PM roycroft: kvm running windows, not so much
10:52 PM _unreal_: nothing like BBQ ribs
11:44 PM roycroft: i grilled some steelhead for dinner tonight, and stubbs have a new low-sugar bbq sauce that i tried
11:44 PM roycroft: i really like it
11:44 PM roycroft: better than the regular stuff