#linuxcnc Logs

Dec 08 2022

12:51 AM randy: morning
01:29 AM Deejay: moin
02:04 AM saquzi: Hi, anyone can help with my hal file? I have 7i97+8i20, but i cant get my servo to move. only light buzz from servo but nothing when jogging.. Have not found the fault in the resent two days for this.. Here in my latest post is the recent hal file, any help would be really much appreciated :)
02:04 AM saquzi: https://forum.linuxcnc.org/39-pncconf/47359-7i97-8i20-error-function-not-found?start=10#258692
02:11 AM XXCoder: hi
02:11 AM XXCoder: hopefully someone will answer. might b few hours though its in middle of quiet period now
02:13 AM jpa-: saquzi: hmm, so are you using the HAL "bldc" module for controlling the motor? i don't see the outputs of that going anywhere, and typically for mesa cards you would use the hostmot2 3ppwmgen instead
02:14 AM jpa-: saquzi: it might help if you draw a wiring diagram of your hardware
02:16 AM jpa-: hmm no, disregard my first line, seems like you can use the bldc component in combination with hostmot2 like that
02:17 AM jpa-: but i don't see your "x-a-value" net leading anywhere, and the x-c1-in etc. appear unconnected also
03:58 AM -!- #linuxcnc mode set to +v by ChanServ
04:47 AM saquzi: Well the wiring is quite simple right now because of having hardware just laying on the table, i can draw wiring diagram if it helps but its really simple right now so i guess its not maybe needed. Im sure that the 8i20 is wired right, no servo brake connected yet though, and i have just hard wired the enable -+ to 8i20
04:48 AM Tom_L: morning
04:49 AM saquzi: encoder is wired to encoder 0 wich i am pretty sure is also ok. Only thing i am not sure with the wiring is the gray scale pattern output. i have wired its 4 signals (comm1,2,4,8) to the pins of 7i97 tb5 in8, in9, in10, in11
04:52 AM jpa-: when you spin the motor with enable disabled, does x-meas-angle change (view in "HAL Configuration" screen)
04:53 AM saquzi: "but i don't see your "x-a-value" net leading anywhere, and the x-c1-in etc. appear unconnected also" Okay i'll see, any example how they should maybe be connected?
05:00 AM JT-Cave: morning
05:01 AM jpa-: saquzi: it seems i misunderstood how that is supposed to work, apparently x-meas-angle is what you use to control the motor driver?
05:03 AM jpa-: your .hal file is a bit of a mess as it, you may want to clean it up to only the necessary connections; if PNCconf is not managing to generate a functional configuration for you, consider reading the man pages and writing the config files yourself instead
05:04 AM jpa-: though that does require you to understand more of the system, but it's helpful to understand anyway..
05:14 AM TheMightyFozz[m]: Hi folks. I use blender for cnc work a lot. I'm trying to make a plugin to open the .blend file location in the systems default file browser. at the moment it has a really annoying error when nothings saved. Also It's only had one test on windows and nothing on osx. Knowing that LinuxCNC uses python I thought I'd try here to see if i can get it fixed.
05:14 AM TheMightyFozz[m]: https://pastebin.com/4m6fHuFi
05:15 AM TheMightyFozz[m]: please note I've never coded anything before.
05:17 AM TheMightyFozz[m]: had help from the blender chat, but don't want to annoy anyone.
05:17 AM saquzi: I have been doing lots of reading the manuals, but just cant understand most of the things before i get something working and then adjusting/tuning.. And yeah, pnconf dont actually seems to do much what is wanted with cfg:qf and 8i20, have been searching and trying to gather and understand a lot because of that... and yeah the Hal is pretty messed
05:17 AM saquzi: up, I tried to clean it and try other things last night but not good still... Dont even noiw what things in hal are not needed, havent found similar setup hal files yet so i could compare and try to understand
05:21 AM TheMightyFozz[m]: the feral engineer's videos on youtube helped me out a lot. I dont play with hal, my understanding is limited to say the least. https://www.youtube.com/watch?v=9zrCtMX-pU0&t=556s&ab_channel=Theferalengineer
05:22 AM TheMightyFozz[m]: the LinuxCNC facebook page is very active too.
05:24 AM TheMightyFozz[m]: for hardware based questions i usually go to "Home Built CNC Machines and projects" on facebook as well as here. https://www.facebook.com/groups/116706481733986
05:48 AM saquzi23: x-meas-angle shows 0.28125 all the time. x-pos-fb changes when i spin the shaft, though the value changes only when i open the window/pin again, can these pins/signal values be read in real time somehow?
06:24 AM jpa-: saquzi23: you can use the "watch" tab in hal configuration
06:25 AM jpa-: saquzi23: does your system have two position sensors for the motor: one gray code position sensor for commutation and fine resolution encoder for position feedback?
06:26 AM jpa-: based on your file it seems it does, but just making sure
06:32 AM saquzi23: http://pico-systems.com/osc2.5/catalog/links/images/fanuc17.pdf my encoder is like the 17-pin version, wich has both. First I thought that I could use just the abz cfg=q, but andypugh suggested using the qf (grayscalepattern too) so i connected it also...
06:33 AM jpa-: yeah, it is easier when you have the gray scale pattern
06:34 AM jpa-: next trouble-shooting steps would be to check if signals hm2_7i97.0.inmux.00.raw-input-08 .. 11 are changing - based on your hal file these should be the gray code inputs
06:34 AM jpa-: check that each of them has varying signal when you turn the motor, to make sure they all are connected
06:35 AM jpa-: i think your hal file may have a small bug in that in one place the signals are called "x-c1-in" and in other place "x-gray-c1-in"
06:35 AM saquzi23: btw, using the watch tab in hal configuration is just always gray on me, nothing else.. so no realtime nothing or anything else showing in there
06:35 AM jpa-: if you click on a signal it should insert it to the list there, and then update it in realtime
06:36 AM jpa-: https://linuxcnc.org/docs/html/hal/halshow.html#_watch_tab
06:36 AM saquzi23: oh yea now it works, ill check these
06:41 AM saquzi23: raw-input-08-11 will not change.. encoder position/rawcounts etc will change
06:42 AM saquzi23: I remembered, should i connect the 7i97 input common pins also to something? ground or 5vcc or..?
06:42 AM saquzi23: not connected to anything right now, because found no info for them..
06:44 AM jpa-: yeah, those have to be connected to something
06:44 AM jpa-: probably GND, but it won't fry them if you get the polarity wrong, it will just not work
06:44 AM saquzi23: ill try
06:45 AM jpa-: (it depends on whether the gray code signal has open-drain outputs or push-pull)
06:50 AM saquzi23: i wired them to gnd and now they seem to work in hal conf, did not even try 5v
06:52 AM jpa-: once you get bldc.0.C1 .. C8 pins changing, you should see the motor orientation in bldc.0.phase-angle change - then you can rotate slowly to make sure that it changes in logical way and does not just jump around randomly
06:53 AM saquzi23: the bldc.0.phase angle still all the time 0.28125 and will not change
06:53 AM jpa-: are the bldc.0.C1 .. C8 pins changing when you check them?
06:54 AM saquzi23: No they are not changing, i mixed them to those input8-11 that work
06:54 AM saquzi23: so they are not connected right in hal?
06:54 AM jpa-: yeah, i think that is the small naming mistake in your HAL file: x-c1-in vs. x-gray-c1-in
07:00 AM saquzi23: yes it waas that, now the c1-c8 changing. bldc.0.phase.angle changes from 0.000-0.999 and so on
07:02 AM jpa-: i think your bldc.0.scale value -4000 seems weird, so you may want to check that - i would assume 16 instead, but i don't have experience with this motor specifically
07:04 AM travis_farmer: ok, i have seen many people using closed-loop systems for motion control, yet my woodworking CNC router table is open-loop stepper based. i have no money to change my system, so my question is, for woodworking, have i screwed myself over spending money on open-loop hardware?
07:05 AM CloudEvil: There is nothing at all wrong with open-loop.
07:05 AM CloudEvil: It just requires higher margins where you never, ever exceed the drive capabilities at any point in a job.
07:06 AM travis_farmer: ok, thank you
07:06 AM CloudEvil: This can mean you need to considerably step back from maximum productivity than if you have something that can do closed loop and adapt to workload.
07:07 AM CloudEvil: Instead of the cut looking a bit rough in one spot, you end up with the whole job after that point failing if you exceed the load.
07:07 AM travis_farmer: Hmmm...
07:07 AM saquzi23: bldc.0.scale value should be the encoder count? these encoders are 2000p, so its 2000 then?
07:08 AM jpa-: CloudEvil: or alternatively it goes like on my open loop machine, i push it too hard, lose steps and ruin a part :D
07:08 AM saquzi23: Or if 16 should be it im thinking something wrong
07:08 AM jpa-: saquzi23: AFAIK the bldc module does not deal with the high accuracy encoder, only the low accurate commutation sensor
07:09 AM jpa-: hm2_xxxx.encoder.00.scale sets the high accuracy encoder scale
07:13 AM jpa-: travis_farmer: there is also a quite low cost upgrade path later if you need more performance, with "closed loop stepper"; they cost somewhere around 100 EUR for each motor + driver, and mount directly in place of stepper motors
07:14 AM travis_farmer: jpa-: i was thinking that, when i have the money to do so
07:15 AM jpa-: every machine will have some weak point that limits performance: usually either structure, spindle or axis motors; if you get the performance you need, or if your weak point is not the axis motors, open loop is fine :)
07:18 AM saquzi23: jpa-: if i set the scale value to 16, the phase angle will only change 0.125, 0.375, 0.625 and 0.875
07:19 AM travis_farmer: at some point, i should do a circle test. load the machine with some wood, and run a circle cut at a good feed speed, and see at what feed speed the circle cut does not meet up. i have a 3-phase 5HP spindle, that runs around 20Krpm, so i can't imagine it would be the weak point... ;-)
07:19 AM JT-Cave: travis_farmer, most stepper systems are open loop
07:21 AM travis_farmer: i am aware JT-Cave, i am just paranoid about messing up a large panel from missed steps.
07:22 AM JT-Cave: understand, you just have to stay well within the capability's of your machine
07:23 AM travis_farmer: right, and the circle test will tell me what the machine can do, and therefore where my safety-margin should be
07:26 AM CloudEvil: There is also step-loss-detection, which is a halfway house which can let you safely abort somewhat
07:27 AM travis_farmer: how would i implement that?
07:27 AM JT-Cave: you should do a stepper limit test to find the limits of your machine
07:28 AM JT-Cave: http://linuxcnc.org/docs/stable/html/config/stepper-diagnostics.html
07:31 AM travis_farmer: i assume you are pointing out the test code under "3.1. Step Timing"? i will try that later today :-)
07:32 AM JT-Cave: yup
07:41 AM jpa-: saquzi23: hmm, not sure why that would happen, maybe the poles number affects it also
07:42 AM jpa-: saquzi23: i think you should expect 16 values in range 0 to 1
07:42 AM jpa-: saquzi23: and one electrical turn of the motor should equal 0 to 1 change and wrap back to 0
07:51 AM saquzi23: intresting..
08:06 AM saquzi23: linux
08:06 AM saquzi23: oops wrong keyboard..
08:07 AM saquzi23: i dont find any information regarding the bldc.n.scale... Only thing i know is that all my pnconf it has made it to -4000
08:09 AM saquzi23: even tough i cant change any of the values in pnconf bldc... How the scale need to be determined?
08:14 AM -!- #linuxcnc mode set to +v by ChanServ
08:14 AM saquzi23: Found couple different hal examples, scales were -4000, -8192 and -16777216 X)
09:33 AM pcw---home: scale should be set to encoder counts or steps per machine unit (1 Inch or 1 mm)
09:47 AM saquzi23: - just inverts it?
09:55 AM pcw---home: yes, - changes the direction
10:04 AM -!- #linuxcnc mode set to +v by ChanServ
10:27 AM saquzi23: what component will determine the angle for servo with 8i20? I got just a small whine from the servo that changes with the poles by spinning by hand, but no resistant on the servo
10:40 AM saquzi23: could it be that the 7i97 is not enabling the 8i20 and it wont work because of that? i have enable pins hard wired for the 8i20 for now..
10:40 AM pcw---home: The servo can be aligned by applying a small DC current to one winding (thia is basically what one of the BLDC options does)
10:42 AM pcw---home: if the 7I97 communicates with the 8I20, enabling the drive is all up to hal (well and the external enable input to the 8I20)
10:55 AM saquzi23: the only amp-enable i have in hall is: net x-enable <= joint.0.amp-enable-out
10:55 AM saquzi23: should it be hal-wise enabled with that?
10:58 AM jpa-: i don't see a separate enable pin in http://linuxcnc.org/docs/html/man/man9/sserial.9.html#8i20
11:52 AM pcw---home: The 8I20 drive has a hardware enable pin and a software enable pin that must be true for the drive to work
11:53 AM pcw---home: software enable is hm2_[HOSTMOT2](BOARD).0.8i20.0.x.amp_enable (x is the sserial channel number)
12:23 PM -!- #linuxcnc mode set to +v by ChanServ
01:38 PM JT-Shop: that part took about a minute... drill and tap 2 holes and I'm done for the day
01:41 PM * Tom_L checks the clock
02:16 PM saquzi: Yeah looks like its not amp enable... And nvmaxcurrent is ok and currentlimits are enough... I think that it is just really simple and stupid error somewhere in hal or then something is broken
02:23 PM pcw---home: the alignment current is set separately via a bldc parameter
02:24 PM pcw---home: There are at least 4 parameters that need to be correct before you will get proper motion
02:25 PM saquzi: bldc parameter in hal?
02:26 PM pcw---home: yes
02:27 PM pcw---home: getting the commutation right is tricky, requiring both the correct direction and scaling plus the initial alignment
02:28 PM saquzi: yes, but to get bigger current for the servo what can be feel by a hand should be easy?
02:29 PM saquzi: what exactly is the "aligment current" name i should be looking?
02:34 PM saquzi: I added the most recent hal here in my latest post https://forum.linuxcnc.org/39-pncconf/47359-7i97-8i20-error-function-not-found?start=10 ,if you could try to look at it what it is actually maybe missing
03:16 PM * JT[m]1 uploaded an image: (220KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/QGhOSNrctkbTYyDPgahYwKou/20221208_151544_1983072121389153347.jpg >
03:17 PM Tom_L: what is it?
03:18 PM JT-Shop: part of the double chucker machine
03:18 PM JT-Shop: I added the limit switch and actuator
03:19 PM JT-Shop: the double chucker is a machine that shapes and sometimes drills a hole the end of round handles
03:19 PM JT-Shop: it's downstream of the rounder
03:32 PM roycroft: well i figured something out today
03:32 PM roycroft: when i was a wee lad, and was pondering career paths, i thought hard about becoming an attorney
03:32 PM Tom_L: you need a bigger shop?
03:32 PM roycroft: i decided against that path
03:33 PM JT-Shop: me?
03:33 PM roycroft: and as i deal with this petition issue for the non-profit, i'm convinced that not becoming an attorney was one of my better choices in life
03:33 PM XXCoder: lol
03:33 PM Tom_L: JT-Shop, don't we all?
03:33 PM roycroft: i figured out that i need a bigger shop a *long* time ago
03:33 PM Tom_L: but was somewnat directed to roy
03:33 PM JT-Shop: I'm good...
03:34 PM roycroft: i bet i could fill up your shop
03:34 PM JT-Shop: need to get rid of the delta planer and some more stuff
03:34 PM XXCoder: "need a bigger shop?" answer is always yes
03:34 PM JT-Shop: I have filled up both shops and the garage
03:34 PM roycroft: "yes" can never be the answer on jeopardy
03:34 PM XXCoder: unless its "uh my shop needs to exist first"
03:35 PM roycroft: because there are two questions that immediately come to mind
03:35 PM XXCoder: like in my case :(
03:35 PM travis_farmer: my machine filled up my shop before the shop knew it was even a shop... ;-)
03:35 PM roycroft: "do you need a bigger shop?"
03:35 PM roycroft: and
03:35 PM roycroft: "do you need more clamps?"
03:35 PM JT-Shop: 2 more limit switches to engineer mounts for on the double chucker then the fun starts
03:35 PM JT-Shop: my wife asked me once why do you have so many clamps?
03:36 PM JT-Shop: a friend's wife asked me why I had 2 miter saws... my answer was because I sold one
03:36 PM Tom_L: i would (maybe) reply: why do you have so many shoes?
03:36 PM roycroft: https://roycroft.us/Clamps.jpeg
03:36 PM travis_farmer: lol, Tom_L
03:36 PM JT-Shop: she got rid of most of her shoes when she retired
03:37 PM roycroft: the difference is, you don't ever wear all your shoes at the same time
03:37 PM roycroft: i didn't use enough clamps in that picture, btw
03:37 PM roycroft: i should have used at least 2x as many
03:37 PM roycroft: or used some clamping cauls
03:38 PM XXCoder: maybe more long reach clamps?
03:38 PM roycroft: and that was before i had some deep throat clamps, so i had to screw the middle part together
03:38 PM roycroft: and then remove the screws and fill the holes when the glue dried
03:38 PM JT-Shop: https://gnipsel.com/images/wood-working/workbench-mk1/workbench-mk1-19.jpg
03:38 PM roycroft: i could have made cauls that go all the way across, but for some reason i didn't
03:39 PM JT-Shop: https://gnipsel.com/images/wood-working/workbench-mk1/workbench-mk1-18.jpg
03:39 PM roycroft: https://roycroft.us/New_Cabinets/34_Veneer_Glue_Up_1.jpeg
03:39 PM roycroft: even with cauls one needs a lot of clamps
03:40 PM travis_farmer: wow, JT has a HUGE shop, compared to even what my dream shop would be...
03:40 PM JT-Shop: https://gnipsel.com/images/wood-working/workbench-mk1/workbench-mk1-17.jpg
03:40 PM Tom_L: at least you have room enough you don't trip over everything
03:40 PM travis_farmer: me?
03:40 PM XXCoder: interesting clamping on 18
03:41 PM roycroft: i do stuff like that a lot
03:41 PM travis_farmer: my shop doubles as storage...
03:41 PM roycroft: i have some swivel jaws for my bessey parallel clamps, and i can clamp at small angles with those
03:42 PM JT-Shop: https://gnipsel.com/shop/gantry/gantry.xhtml
03:42 PM roycroft: https://roycroft.us/BesseyClamps.jpeg
03:42 PM JT-Shop: that's inside the machine shop before I filled it up with cnc machines
03:42 PM roycroft: that's pushing the limits of direct angle clamping, though
03:43 PM JT-Shop: https://gnipsel.com/shop/gantry/gantry-11.xhtml
03:43 PM JT-Shop: you do what you need to do to finish a project
03:44 PM JT-Shop: it was fun moving the gantry from the machine shop to the big shop
03:45 PM JT-Shop: https://gnipsel.com/images/shop/Siding15.jpg
03:46 PM JT-Shop: that's the garage and machine shop
03:46 PM travis_farmer: maybe someday, i will get my IRS tax return back from 2021... so i can finish my shop... :-(
03:47 PM roycroft: no refund for me this year
03:47 PM roycroft: i had a big capital gain that drove my taxes way up
03:47 PM travis_farmer: apparently, none for me either...
03:47 PM roycroft: i calculated my estimated tax and send the irs a check for that amount plus a few hundred dollars
03:47 PM JT-Shop: https://gnipsel.com/images/equipment-shed/e-shed-19.jpg
03:48 PM JT-Shop: that's the 30' x 50' x 12' shop... they built in 4 days flat
03:48 PM XXCoder: this will be my second year of not filing taxes. fun!
03:48 PM travis_farmer: your making me drool, JT
03:48 PM roycroft: and when i finally filed a couple months ago i had overpaid by a small amount, so i told them to just keep it for next year
03:48 PM Plaid: when dealing with hal, does the order of the pin names matter? (I know the first argument is the signal name, and the arrows are ignored)
03:48 PM JT-Shop: https://gnipsel.com/images/equipment-shed/e-shed-22.jpg
03:48 PM JT-Shop: Plaid, no
03:49 PM JT-Shop: the order of adding functions can make a difference
03:49 PM travis_farmer: the arrows in HAL are ignored??? i have been using them like they mean something...
03:49 PM roycroft: did you buy that shop at menard's?
03:50 PM JT-Shop: which shop?
03:50 PM roycroft: the last one that you posted - shed-22.jpg
03:50 PM roycroft: that looks like a menard's kit
03:50 PM JT-Shop: no, the Mennonites built it with lumber from a local lumber yard
03:50 PM roycroft: oh, nice
03:53 PM JT-Shop: https://thepatriotwoodworker.com/forums/topic/38386-cob-separator/
03:54 PM XXCoder: nice
04:03 PM travis_farmer: so if the arrows in HAL files don't matter, why was i bothering to put them in??? LMAO, i think i self-learned something incorrectly somewhere along the way...
04:05 PM roycroft: grrr
04:05 PM Plaid: Humm. I still haven't figured out how to get LCNC to not start in an e-stopped state. I've tried building a hal net to auto-reset it on startup, but after the one shots finish, it's still in the estop state. https://pastebin.com/2fV1Bvee
04:05 PM roycroft: the jig for my chopstick maker was supposed to be delivered today
04:06 PM roycroft: now it's going to be friday
04:06 PM roycroft: i need that asap
04:06 PM roycroft: i want to make some nice chopsticks for holiday gifts, and i need time for the finish to cure
04:06 PM JT-Shop: Plaid, why do you not want lcnc to start with the estop active?
04:07 PM travis_farmer: it is safer to start with e-stop on...
04:07 PM JT-Shop: yup
04:07 PM roycroft: i think you should always start with the e-stop on
04:07 PM roycroft: let's all gang up on plaid :)
04:07 PM Plaid: JT-Shop If it start with estop active, then it'll trip my estop, which pulls a circut breaker
04:07 PM Plaid: Which I have to manually close
04:08 PM roycroft: that seems to be to be better than potentially starting the machine in an unknown state, and risking life and limb
04:08 PM JT-Shop: you can't use a relay?
04:09 PM Plaid: I've gone through this multiple times with other people. I want to start my machine in the OFF state, not in the ESTOP state.
04:09 PM roycroft: yeah, i would use a latching relay and a momentary contact switch to turn it on
04:09 PM Plaid: roycroft: Come over and rewrite my whole box, and sure.
04:10 PM JT-Shop: seems odd to trip a circuit breaker with the estop
04:11 PM * roycroft guesses he should find something else to do after work today since he won't be making chopsticks
04:11 PM travis_farmer: what is on the circuit breaker?
04:11 PM Plaid: Sorry if I'm frustrated, I just feel like I've gone through this like 3 times already.
04:12 PM roycroft: that's part of the design process
04:12 PM roycroft: i've been working on the plans for my cnc router for close to 2 years now
04:12 PM roycroft: and i've had fully functional models before
04:12 PM * JT-Shop ponders the issue
04:12 PM Plaid: Yeah, I'm past that. I'm into implmentation.
04:13 PM roycroft: but then i learn something new or have a rethink, and come up with a better plan
04:13 PM JT-Shop: how are you tripping the circuit breaker?
04:13 PM Plaid: I'm rewiring an existing machine, I'm trying to minimize the changes. So I'm not buying any new hardware. There's an existing e-stop wired into the system. I'm trying to get it ti work with LCNC
04:13 PM * JT-Shop can relate to that
04:13 PM roycroft: having an e-stop trip a circuit breaker seems unusual
04:13 PM Plaid: The circut breaker is wired to the e-stop buttons, so LCNC has nothing to do with it. But I've added a circut so LCNC can activate the ESTOP as well.
04:14 PM Plaid: pressing the e-stop denenergize the whole machine.
04:15 PM roycroft: yeah, that's good
04:15 PM Plaid: I don't want that circut breaker tripping every time I start LCNC
04:15 PM roycroft: that's how i mostly do it
04:15 PM travis_farmer: understandable...
04:15 PM roycroft: but how does a circuit breaker trip when power is removed?
04:15 PM Plaid: So I need to NOT start in the estop state.
04:16 PM roycroft: that's what i don't understand
04:16 PM Plaid: Otherwise it'll see estop => trip breaker
04:16 PM travis_farmer: or have something else trip the breaker
04:16 PM JT-Shop: setp halui.estop.reset True will deactivate the estop
04:17 PM Plaid: So back to my origonal question, how do I get LCNC to not start in the ESTOP state?
04:17 PM travis_farmer: maybe a large contactor to deenergize the machine, rather than a manual breaker?
04:17 PM JT-Shop: hmm how to do that with something similar to a first scan on a PLC
04:18 PM roycroft: which is how i do it, travis_farmer
04:18 PM JT-Shop: actually setp should only do that once
04:18 PM travis_farmer: that way LCNC can restore power
04:18 PM roycroft: the latching relay + momentary contact switch
04:18 PM travis_farmer: also good, roycroft
04:21 PM Plaid: Yeah. You can design all sorts of other stuff. I'm dealign with an existing machine that I'm trying to minimize alterations to, so I can go back to stock if I want
04:22 PM roycroft: i get that
04:23 PM roycroft: this is a safety issue, and i would consider making an exception for increased safety
04:23 PM JT-Shop: Plaid, # estop loopback
04:23 PM JT-Shop: net estop-loop iocontrol.0.user-enable-out iocontrol.0.emc-enable-in
04:23 PM JT-Shop: setp halui.estop.reset True
04:23 PM JT-Shop: works
04:24 PM * JT-Shop noticed that Plaid has alerted the safety police...
04:24 PM travis_farmer: lol
04:25 PM roycroft: this is not my system, and i don't know what other safety measures are built-in
04:25 PM roycroft: note that i'm not demanding anything - just suggesting that the safety aspect of this change be considered
04:27 PM travis_farmer: imagine a car with a manual trans. now go to start that car without knowing what gear it is in, and all the safety stops disabled. the car could be in gear, you don't know. that is starting a machine with e-stop enabled, in my opinion. :-)
04:27 PM JT-Shop: Plaid, of course in the [HAL] section you need HALUI = halui
04:32 PM travis_farmer: i think we angered him away... (not my intent)
04:33 PM JT-Shop: he's still here
04:35 PM XXCoder: roy its a reason if shop goes into longer blackout we would hit all estops
04:35 PM XXCoder: so none of machine would do random stuff if chances hit just right
04:35 PM roycroft: that's why i use a latching contactor with a momentary-action switch
04:35 PM roycroft: if the power goes out the machine won't turn on when it's restored
04:35 PM travis_farmer: sounds reasonable. i was thinking of adding a e-stop trigger for a power fail
04:40 PM roguish[m]: roycroft: how the weather right now? a bit wet?
04:41 PM roguish[m]: ooo already passed you.
04:43 PM Plaid: Right now it's setup so if you hit any of the physical e-stops is disconnects the circut breaker, whole machine is powered down.
04:44 PM travis_farmer: that include the computer and Mesa board?
04:45 PM Plaid: This is the stock design for the machine, before I changed anything on it. Hit estop-> trips breaker. -> machine powers down
04:45 PM Plaid: Mesa board, but not the computer
04:46 PM roycroft: that it is
04:46 PM roguish[m]: a good way to handle e-stop, is to drop the power to drives, but maintain compute and logic power, that includes encoders. that way recovery is much easier and cleaner, and safer
04:46 PM roycroft: drizzly and dreary
04:46 PM roycroft: and what roguish just said
04:46 PM travis_farmer: right, but if the computer and mesa board is powered, then you can start the computer, LCNC, clear the estop, and hit the breaker
04:46 PM roycroft: an estop should physically kill the power to anything that moves
04:46 PM roycroft: but should keep the cpu and other logic devices powered, if possible
04:47 PM Plaid: I don't want it tripping the breaker if I restart LCNC
04:47 PM JT-Shop: on PLC's that's what I normally do unless it's a hazard to kill output power
04:47 PM * roycroft declares this conversation officially circular, and finds a path to turn out of the roundabout
04:48 PM travis_farmer: lol roycroft
04:48 PM JT-Shop: the hens are doing the roost time rumble
04:48 PM * JT-Shop calls it a night
04:48 PM travis_farmer: night JT
04:48 PM Plaid: Yeah, the real solution is for roycroft to come over to my house, with his own supplies and build me a new controller box.
04:48 PM JT-Shop: say goodnight Gracie
04:49 PM Plaid: goodnight Gracie!
04:49 PM roycroft: i'm not the best qualified for that
04:50 PM * JT-Shop has worked on machines that killing the outputs were more dangerous that not
04:51 PM Plaid: This is just a router, so it's not like this is a massive mill or something.
04:51 PM JT-Shop: in those cases the estop stopped motion and applied the brakes
04:52 PM JT-Shop: alright everyone is on the roost time to read the hens a bedtime story... chicken little comes to mind
04:52 PM Plaid: Night!
04:52 PM travis_farmer: later JT
04:53 PM roycroft: yeah, sometimes that is what is necessary, jt-shop
04:53 PM roycroft: i'm still curious about how the machine in question here trips a breakr on an e-stop event
04:54 PM roycroft: that's not an argument to change it
04:54 PM roycroft: it's genuine curiosity
04:54 PM roycroft: breakers generally trip on overcurrent or short circuit events
04:54 PM roycroft: what does the e-stop button do to cause that?
04:54 PM travis_farmer: good question
04:55 PM roycroft: if it's a gfci circuit it could cause a small current leak to trip it
04:57 PM John_f: Plaid do you really mean a circuit breaker as in an overcurrent protection device or do you mean some kind of contactor that disconnects the power
04:58 PM Plaid: It's a circut breaker that's able to be tripped off via a 24V DC signal.
04:58 PM roycroft: yes, it may be the very latching circuit i've been describing, and not an actual circuit breaker
04:59 PM Plaid: It's a double duty device. It's a circut breaker that's also the switch to turn the cabinet on.
05:00 PM roycroft: so it is an overcurrent device that you can trip out of circuit with a 24v signal?
05:00 PM Plaid: It's an overcurrent device. In additon, there's a seperate 24V signal to externally trip it
05:00 PM roycroft: ok
05:00 PM roycroft: thank you
05:01 PM roycroft: i've not run into a circuit breaker like that before
05:01 PM Plaid: So it has 4 big terminals on it, and 2 tiny ones.
05:01 PM roycroft: a contactor with overcurrent protection, in essence
05:01 PM Plaid: It's not a contactor. There's no way to close it. Just open it
05:02 PM roycroft: but one that has to be manually reset
05:05 PM * travis_farmer has supper ready. time to sign-off for the night.
05:46 PM CaptHindsight[m]: circuit breaker as a switch
05:48 PM CaptHindsight[m]: like dis? https://www.amazon.com/Circuit-Breaker-Automotive-Overload-Protection/dp/B0992F4LRL/ref=sr_1_5
08:02 PM pcw---home: Plaid: Perhaps you should only trip the breaker on an actual fault (not just the estop state)
08:05 PM Plaid: pcw---home: setp halui.estop.reset True got what I wanted I can hit my soft e-stop or my hard e-stop and it'll open the breaker
08:16 PM Plaid: pcw---home: If only there was some kind of state reserved for "Oh shit, I want everything off right now"
08:20 PM pcw---home: I think the issue with most state information is that its safest to be the fault state when LinuxCNC is not running
08:21 PM pcw---home: and that doesn't get along well with a manually set, machine reset circuit breaker
08:48 PM Plaid: Okay, I've now got it so I can trip my physical e-stop from lcnc, or with totally independent buttons. pcw---home: I'd agree with you, but I'm not building this from scratch, I'm converting an existing system, trying not to totally rebuild the machine. So far I've only had to add 3 wires to the existing machines, and moving some ribbon cables over
08:48 PM * roycroft reenters the roundabout and looks for the nearest exit again :)
08:48 PM Plaid: Hi! Bye!
08:48 PM * Plaid waves at roycroft as he passes by
08:49 PM XXCoder: hibye roy
08:50 PM Plaid: Now to setup the servos :P
09:34 PM Centurion_Dan1 is now known as Centurion_Dan
09:36 PM Plaid: For the 7i77, I'm doing the servo amp enable, my servo amps are active high. So for enable, I should put Ena+ to the servo amp, and ena- to ground, right?
09:43 PM pcw---home: If your drives need a positive voltage to enable, you would connect 7I77 ENA+ to a positive voltage and ENA- to the drives Enable input
09:47 PM Plaid: Ah, I see where I was reading it wrong, thanks!
09:54 PM -!- #linuxcnc mode set to +v by ChanServ
11:57 PM -!- #linuxcnc mode set to +v by ChanServ