#linuxcnc Logs

Sep 20 2020

#linuxcnc Calendar

02:01 AM Deejay: moin
04:56 AM JT-Cave: morningh
04:56 AM XXCoder: yo
05:47 AM andypugh: syyl: I was watching your S1 video and remembered that I have found a source for the inverted ER collet nuts. ( https://www.ebay.co.uk/itm/181326100045 ) and have made an adaptor for those intermediate sizes I don’t have a dedicated collet for: https://photos.app.goo.gl/ppuaMDJUGfo9Af2d7
06:51 AM Tom_L: morning
06:51 AM XXCoder: hey
06:51 AM XXCoder: andypugh: interesting
06:52 AM andypugh: XXCoder: I used anothe one for a milling collet chuck in a recent video.
06:52 AM XXCoder: cool :)
06:52 AM XXCoder: bah i miss work lol
06:53 AM Tom_L: i see why you made the spanner now
06:53 AM XXCoder: applied for few jobs dunno if any good
07:41 AM miss0r2: Just when I was about to sell my EDM, I get an order for 10x special bolts made from carbide. All I have to do is add the allen hex in one end. :)
07:49 AM andypugh: What are carbide bolts for?
08:11 AM Loetmichel: i cant see any application for bolts made of brittle carbide either, maybe apart from using them as setscrews in hardened shafts.
08:12 AM miss0r2: andypugh: I didn't ask for details, but I was told that is was for an experiment including some hazardous materials and some sort of furnace
08:12 AM Loetmichel: but even then i would probably take a normal steel bold and set a carbide tip in it
08:12 AM miss0r2: yeah well :)
08:12 AM miss0r2: Its my bread and butter ;)
08:13 AM Loetmichel: hehe, you are aware that carbide is flammable?
08:13 AM Loetmichel: (or more like: it burns before it melts)
08:13 AM miss0r2: I did not know that, no :)
08:13 AM miss0r2: but I am sparking a test as we speak. It seems to go fine, however slow that might be
08:15 AM miss0r2: I might as well... I ordered a "sparkplug repair set, extra big" but it didn't turn up friday, so my engine rebuild is stalled anyway
08:19 AM andypugh: miss0r2: Maybe there is a need for strength at high temperature. But solid tungsten would be better in that regard.
08:20 AM miss0r2: maybe
08:21 AM miss0r2: andypugh: I just did the proper translation. this is solid tungsten :)
08:22 AM miss0r2: this translation is bad
08:22 AM miss0r2: Carbide. solid carbide. not tungsten at all.
08:27 AM andypugh: Which carbide?
08:28 AM andypugh: Carbide is a compound, often of tungsten. Calcium carbide would be a very poor choice for bolts, as it reacts with water to make acetylene gas.
08:29 AM andypugh: “Carbide” tools are typically tungsten carbide powder glued together with cobalt metal.
08:34 AM * miss0r2 has no idea
08:46 AM miss0r2: Unless is it something I know about & can see is absolutly wrong the way a client has requested it, I tend not to poke places my fingers don't belong.
08:50 AM Loetmichel: miss0r2: "carbide" toos are actually Tungsten carbide dust "glued" together with cobalt because you cant melt tungsten carbide before it decomposes
08:51 AM Loetmichel: TIG welding rods are pure tungsten though (sometimes with some radioactive substance in it to help with igniton)
08:51 AM miss0r2: makes sense. I wonder what sort of material I will be delievered.
08:51 AM Loetmichel: its easy to check thoug
08:51 AM miss0r2: In denmark we only have one name for all the composites that make up tungsten carbide tools
08:51 AM Loetmichel: cobalt is magnetic, tungsten isnt.
08:52 AM miss0r2: its called "hard metal"
08:52 AM miss0r2: I hope upon delievery I get some more information.
08:52 AM Loetmichel: meaning if the material is attracted by a magnet its most likely tungsten carbide
08:52 AM Loetmichel: if not its more likely tungsten
08:53 AM Loetmichel: in german "hartmetall" or "Wi(e)dia(mant)" are words for tungsten carbide sintered with cobalt.
08:53 AM Loetmichel: no idea how its in denmark though
08:54 AM miss0r2: probably the same. its a horrible name for something :)
08:54 AM miss0r2: I have to run now, see you atound
08:54 AM miss0r2: around*
08:54 AM Loetmichel: also tungsten carbide is even harder than pure tungsten ;)
08:54 AM Loetmichel: hence the "like diamond" name for the stuff ;)
08:57 AM Tom_L: tungsten is brittle enough by itself
08:57 AM Tom_L: we used to machine ballast wing tip weights from it
08:58 AM Loetmichel: yeah, it has the same density as gold
08:58 AM Tom_L: gold is soft though
08:58 AM Loetmichel: and slightly more expensive
08:58 AM Tom_L: meh :)
09:52 AM centreline: hi all
09:58 AM _unreal_: coming down the line cutting parts for my duaghters mini cnc machine
09:58 AM _unreal_: cutting the parts for the pen/touch probe adapter
10:33 AM jymmmm: morning
10:36 AM centreline: can I use hal oscilloscope to see stepper signals sent to the stepper with a 5i25 and 7i76 daughterboard setup? If so does anyone know the signal names?
10:36 AM _unreal_: https://cdn.discordapp.com/attachments/717411349541683212/757263367814512720/render2.PNG
10:36 AM _unreal_: rendering I did of the part I'm making on my cnc right now
10:37 AM pfred1: I like real oscilloscopes
10:37 AM pfred1: I suggest anyone setting a CNC machine up have one
10:38 AM Tom_L: centreline, google hal scope
10:38 AM Tom_L: yes
10:38 AM pfred1: they're not that expensive today
10:38 AM Tom_L: or even a logic analyzer for that
10:38 AM centreline: I have hal scope open, watching x-vel-cmd but nothing appears on the trace when I move the X axis in the gui
10:39 AM pfred1: actually being able to see signal waveshape can be helpful
10:40 AM _unreal_: what do you guys think?
10:40 AM _unreal_: of the rendering?
10:40 AM centreline: I have a scope, just trying to debug it remote from a X session from office
10:40 AM _unreal_: been screwing with fusion360 a bit
10:40 AM pfred1: _unreal_ looks shiny
10:40 AM centreline: I *think* the hardware is fine, because this was a stepper machine running parport breakout cards and that side hasn't been touched :(
10:41 AM pfred1: I made my own breakout board
10:43 AM pfred1: wacky stuff can happen in control electronics with stepper motors
10:43 AM pfred1: stuff you can only fix with optical isolation but optos bring their own problems with them
10:44 AM centreline: the whole opto speed was a minefield I hoped to avoid by just sending money to someone for mesa kit
10:44 AM pfred1: optos are slow
10:44 AM pfred1: drives need setup times too
10:45 AM pfred1: so what you have to do is hunt for the fastest setup time your hardware can handle
10:45 AM pfred1: start big and work down
10:46 AM pfred1: CNC is a crawl before you run deal
10:46 AM centreline: net x-output is connected to hm2_5i25.0.stepgen.00.velocity-cmd, I'm not even sure I have that connected to the right board, the other names are hm2_5i25.0.7i76.0.0.x indicating the pin is on the daughter card...
10:47 AM _unreal_: pfred1, not to mention UPGRADES lol
10:47 AM pfred1: _unreal_ that's the need for speed trap
10:47 AM _unreal_: My machine in its original setup all hand made builts its own upgrade z parts
10:47 AM pfred1: you get into a loop where you just want to go faster and faster
10:47 AM _unreal_: now my machine is SOOOOO much better
10:47 AM centreline: Ive already crawled on this machine, it was together and working but I replaced the parport bob's for more IO and because the pc was having issues
10:47 AM _unreal_: way more rigid and I can run it faster
10:48 AM _unreal_: human error slows me down the most now :)
10:48 AM pfred1: at 2000 RPMs my lead screws give me pause
10:48 AM _unreal_: 2k rpm lead screw
10:48 AM _unreal_: WOW
10:48 AM _unreal_: ya
10:48 AM pfred1: I ook at that and say that's crazy
10:48 AM _unreal_: I dont know what mine does but prob 500 RPM
10:48 AM pfred1: yeah it gets real over 1K
10:49 AM pfred1: 1K is the cut off
10:49 AM pfred1: anything looks OK up to 1K
10:49 AM _unreal_: I also converted all of my nuts from metal nuts to starboard HDPE nuts
10:49 AM pfred1: over that and your leads had better be straight and well supported
10:49 AM _unreal_: the threaded rod wear is near zero
10:49 AM _unreal_: and resistance has dropped a lot as well
10:49 AM pfred1: yeah i run HDPE lead nuts
10:50 AM pfred1: at 2K I burnt them up
10:50 AM _unreal_: I made mine custom
10:50 AM _unreal_: what do you grease them
10:50 AM pfred1: I use a PTFE dry lube
10:51 AM pfred1: I just looked at the can and that's what it is called, "Dry Lube"
10:51 AM _unreal_: I use white lithium
10:51 AM pfred1: doesn't dust stick to it?
10:52 AM pfred1: HDPE is self lubricating to an extent
10:52 AM pfred1: bbiab
10:57 AM _unreal_: https://cdn.discordapp.com/attachments/717411349541683212/757268573012885597/16006171615693638323083227164281.jpg
11:05 AM * flyback has used superlube on several things
11:05 AM flyback: it's really only for plastic parts and sliders
11:07 AM _unreal_: cant use super lub on a cnc
11:07 AM _unreal_: attracts dust
11:10 AM pfred1: flyback I like superlube for O Rings
11:10 AM pfred1: that's what I have it for
11:11 AM flyback: oh didn't know it attracts dust
11:11 AM flyback: used it to repair a cdrom drive back when they cost $$$
11:11 AM flyback: I sitll have one origional rebranded radio shack tube and one superlube tube
11:12 AM flyback: I doubt it ever goes bad
11:12 AM pfred1: I don't know of gooey greases attract dust but they sure hold onto it
11:13 AM pfred1: superlube is dilectric synthetic grease
11:19 AM flyback: teflon
11:20 AM flyback: oh it is a dielectric?
11:20 AM flyback: YOU GO TO HELL
11:21 AM pfred1: die like a dog
11:22 AM pfred1: I'm trying to remember the stuff a buddy of mine's uncle invented. Oh yeah Slick 50
11:22 AM pfred1: it was teflon
11:23 AM pfred1: I wonder how la la la la larry is doing today?
11:23 AM centreline: kind of like fitting oilers to everything and having covered leadscrews and use oil, but I only have metal cutting cnc in enclosures...
11:24 AM centreline: my vmc wont even power up unless it has air pressure to pressurise the oil pump system, I think it oils at 2bar or something crazy
11:24 AM pfred1: I ran a broach in a machine shop that had a crazy cutting oiler on it
11:24 AM pfred1: the outlet was seven tenths of an inch in diameter
11:25 AM pfred1: it was like a waterfall of oil
11:25 AM pfred1: stuff went all over the place
11:25 AM centreline: building a power hone that uses a suds pump to pump honing oil but that's not pressurised, just to flood the particles away during the cut
11:26 AM pfred1: the craziest part of that machine is it ran on hydraulic logic
11:26 AM centreline: this must have hydrodynamic bearings that rely on x bar of oil pressure in the head or something, I havent opened it up, I just know it beeps sadly and refuses to run until the oiler sends a all ok signal
11:26 AM pfred1: the brains of it looked like wall art in Hell
11:27 AM pfred1: I said to the foreman one day, Why don't you get someone to come in and fix this thing? he said, No one knows how it works.
11:27 AM centreline: my surface grinder is automatic and hydraulic logic, but it has a blown hose inside currently so I'm stuck turning handles on it
11:28 AM pfred1: I looked at the mess on the side of it and I believed him
11:28 AM centreline: its my fault it blew too, one day it stopped working and I stripped everything down and found a broken spring in the pump pressure relief, so I eyeballed it not having any data on pump pressures and turned it on, and joy, it ran but at insane speed
11:29 AM centreline: so I threw the lever to stop and the pressure just popped something inside. I have found a manual with all the specs and pressures since, but its a cow to get in the table to replace
11:29 AM centreline: 1930's tech...
11:30 AM pfred1: I don't know how old that broach was
11:30 AM pfred1: we had a horizontal mill from like 1859 though
11:31 AM pfred1: it was beautiful they had it all hand scraped and frosted
11:32 AM pfred1: the broach was a whole different story
11:32 AM centreline: this isnt beautiful, but its mine and it mostly works, one day I'll fix the hydraulics too because its really boring grinding in a large part :)
11:33 AM pfred1: we had a wooden platform and stairs leading up to it that most resembled gallows
11:34 AM pfred1: before running it you expected somene to say, Any last words?
11:34 AM centreline: if it blew a hose that day, it may well have been
11:34 AM pfred1: oh it didn't have to blow anything to end you that was just SOP
11:35 AM pfred1: the blades were 6 feet long
11:35 AM pfred1: it was a big broash
11:35 AM centreline: I've nearly been scalded by my backhoe popping a hose after its been working hard a few hours
11:35 AM pfred1: broach even
11:35 AM centreline: what did it broach?
11:35 AM pfred1: mostly noses for our clamps I'll find a picture
11:36 AM pfred1: https://i.ebayimg.com/images/g/mF4AAOSwJHJc5CCE/s-l400.jpg
11:36 AM pfred1: it ripped the dovetails
11:36 AM pfred1: one pass
11:36 AM _unreal_: hate stupid mistakes
11:37 AM _unreal_: have one nice little nasty ding in the thing I'm making for my daughter because I set the origin in the wrong location in fusion
11:37 AM _unreal_: had it at work bed height not part height
11:38 AM centreline: ah I saw broaching and I imagined a large internal shape for some part of a ship's final drive :)
11:38 AM pfred1: you can see the mark the cam vise left on the nose
11:38 AM pfred1: see the long dent?
11:39 AM pfred1: it's a faint line about a quarter inch back from the face of the clamp
11:39 AM _unreal_: ARG... 4min left to finish this part
11:39 AM pfred1: the part is locked in the cam vise then the table feeds forward and the broach blade comes down on it
11:39 AM _unreal_: cant wait
11:40 AM centreline: on the left? I think so, it must take some clamping to single pass that
11:40 AM pfred1: yeah we had a cam vise with a long lever
11:40 AM pfred1: the part sat in a pocket with the top of the vise holding it down
11:41 AM pfred1: the table automatically went in and out and you had to unload and load a part
11:41 AM centreline: I bought 3 angle plates once described as fairly large, and when I got there, the largest was larger than my mercedes sprinter load area, I was picturing something like the machine that came from :)
11:41 AM pfred1: so it was like playing beat the clock all day
11:41 AM pfred1: because you didn't want to still be screwing with anything when that blade came down
11:42 AM centreline: thats the sort of things safety systems were invented for (and rightly so, I like having limbs even when I have a stupid day)
11:43 AM pfred1: centreline it had an auto stop bar but if you hit it that'd put the machine into a state
11:43 AM pfred1: so you were encouraged to not use it
11:43 AM pfred1: when that machine acted up it'd shake the whole building
11:44 AM centreline: in my homeshop, different of course, I have a old inertia steel guillotine that has a 6hp motor and it winds up a flywheel and when you push the pedal, drives the blade with the inertia in the flywheel, there's no second chances with that...
11:44 AM pfred1: the broch I ran was hydraulic like gas station lift sylinders
11:44 AM pfred1: cylinders even
11:45 AM pfred1: It was oh maybe 15 feet tall?
11:45 AM centreline: but if I was earning machinist rates, I dont want to risk loosing digits to make someone more profit
11:45 AM pfred1: just the blade plates on it were 6 feet long
11:46 AM roycroft: it's time to go machine my leg vise chop to final size/shape
11:46 AM pfred1: it was one kooky contraption
11:46 AM roycroft: that will make a huge difference in the appearance of the bench
11:46 AM roycroft: hopefully it will turn out as i envision it
11:47 AM pfred1: I need a new workbench
11:47 AM centreline: I have a leg vice with no leg, one day I'll make one for it, but one day a lot of things will be made (in my mind)
11:48 AM centreline: even achieving it is a win roycroft :)
11:48 AM pfred1: when I put a side vise on my workbench it didn't go how I imagined it would
11:49 AM pfred1: better luck next time
11:49 AM centreline: im going to the shop to see if the stepper driver itself actually has signals being sent it, perhaps the 5v ena isnt enough to enable the driver or something stupid
11:49 AM pfred1: it is current that you need
11:49 AM pfred1: drives usually have optos in them and they need current to fire
11:50 AM pfred1: least 12 ma
11:50 AM centreline: the stepper power connections are unchanged from when it was a working machine, theres 60v at as many amps as it can draw available for that side, this is a mesa 7i76->driver issue I think
11:50 AM pfred1: calling pcw
11:50 AM centreline: oh I see, you mean on the signal commanding side of the driver
11:51 AM pfred1: 12 ma is not an insignificant amount of current for logic
11:51 AM centreline: rther than a 7i76 driver issue, I mean I've probably messed up the hal somewhere
11:51 AM pfred1: it is likely not enough setup time
11:52 AM pfred1: every drive needs a different amount
11:52 AM pfred1: usually 5000 ns or more
11:52 AM centreline: hmm I will check that, I think I used the value from my working config previous
11:52 AM pfred1: so you start out at 20K and go down until it stops working
11:53 AM pfred1: then add a bit and that's it
12:00 PM centreline: dirsetup was 24k previously but that was probably way too slow but if it works but is slow I can tune later.
12:01 PM pfred1: the downside to long setups is is lengthens your period
12:01 PM centreline: off out to turn on the dc rail again and try it then find the scope leads etc if it doesnt work
12:01 PM pfred1: you get a slower step pulse train
12:02 PM pfred1: so you want your setup as short as you can make it but it still has to work
12:03 PM pfred1: and setup is one of those htings you can't tell someone what it is because it's different
12:04 PM pfred1: drives have specs but they're not always right
12:06 PM centreline: yup, but I know the 24k worked last time with these drivers because it was a running machine
12:07 PM centreline: and its still not working, setting up scope...
12:12 PM pfred1: 24K setup?
12:12 PM pfred1: that's long
12:36 PM Nyxlayne: hello
12:38 PM pfred1: hi
12:44 PM Nyxlayne: i am trying to bring my system back to life using old config files. i am getting an error pertaining to petting the watchdog that is new. can anybody assist me with this?
12:46 PM Tom_L: what linuxcnc version?
12:51 PM Nyxlayne: 2.7.15
12:52 PM Tom_L: mesa hardware?
12:53 PM Tom_L: setp hm2_[HOSTMOT2](BOARD).0.watchdog.timeout_ns 10000000
12:53 PM Tom_L: from my config
12:54 PM Nyxlayne: 5i20, yes. it is a servo
12:54 PM Nyxlayne: driven lathe
12:54 PM Tom_L: # # # addf hm2_[HOSTMOT2](BOARD).0.pet_watchdog servo-thread
12:54 PM Tom_L: also removed that line
12:55 PM Tom_L: it changed a bit in a prior version and i can't recall which one
12:55 PM Tom_L: but mine is a 2.7.xx config
12:56 PM Nyxlayne: wow that looks right that was fast. ill try it straight away.
12:56 PM Tom_L: http://linuxcnc.org/docs/html/drivers/hostmot2.html#_watchdog
01:07 PM JT-Shop: same watchdog line on the Hardinge with a 5i20
01:08 PM JT-Shop: Nyxlayne, are you sure you're running 2.7.x? in 2.8 or a firmware change the watchdog changed
01:10 PM Tom_L: JT-Shop, i had a question in a sec.. phone call
01:11 PM JT-Shop: hurry, I'm off to Elsinore via Leeper for an ice cream waffle cone
01:11 PM Tom_L: it can wait
01:13 PM Tom_L: it's about an estop setup i got from you long time ago
01:15 PM JT-Shop: I meant Eminence, the external estop setup?
01:15 PM Tom_L: i think so
01:15 PM Tom_L: i'll post the lines here in a sec
01:16 PM JT-Shop: ok
01:16 PM Tom_L: i have them in my config but commented out
01:18 PM Tom_L: mine is very basic but i may want to change that
01:18 PM Tom_itx: https://paste.ubuntu.com/p/JM7PB3yRbp/
01:19 PM Tom_L: i'll have 3 sources of estop. axis, the pendant button and a button on the control
01:19 PM Tom_L: all should do the same thing
01:20 PM Tom_L: i think i'm gonna use the stepper enables as well as the vfd enable but probably need a DP relay for those since the logic is opposite on them
01:20 PM Tom_L: the relay will take one low and the other high to disable
01:20 PM JT-Shop: https://forum.linuxcnc.org/47-hal-examples/25861-external-e-stop?start=0
01:20 PM Tom_L: so the estop loop will just control the relay
01:21 PM JT-Shop: that's the external estop code
01:21 PM Tom_L: but i had that code i got from you and didn't quite get what it did
01:21 PM Tom_L: looks the same as what i posted
01:22 PM JT-Shop: yea I see
01:22 PM Tom_L: so the parport pin is the estop signal
01:22 PM Tom_L: but how do i tie the axis soft estop in with them?
01:22 PM JT-Shop: yea the external estop
01:24 PM Tom_L: if the estop button is twisted and reset can this trigger the axis soft estop to also reset?
01:24 PM Tom_L: i haven't tested any of it yet, just trying to get an idea how it works
01:25 PM JT-Shop: hmm I think so
01:25 PM Tom_L: the button on the pendant is just a momentary switch
01:26 PM Tom_L: so right now i'm just enabling it in axis when tripped
01:26 PM Tom_L: haven't really been that worried about it but i think once i get the vfd i should do something
01:27 PM JT-Shop: gotta run now
01:27 PM Tom_L: it will be a while, i'm just gathering info
01:27 PM Tom_L: bye
01:27 PM centreline: I think I just did related today for two external e-stop but no relays yet
01:28 PM Tom_L: one of mine is momentary and the other is a regular estop switch
01:28 PM centreline: I used the latching estop example in the forum, then added the multiple's via a or2 component, and used the -not state for them
01:29 PM centreline: so any switch issues -> estop, then the latch feeds into the machine estop loop
01:30 PM centreline: do you want me to pastebin it? its part of todays stuff that actually works... unlike the steppers still (but I think I found a broken wire)
01:31 PM centreline: I have a physical e-stop on the monitor/keyboard cabinet, another physical to a different pin on the headstock and soft key on touchscreen
01:33 PM Tom_L: i may ask later but need to run right now
01:37 PM centreline: understood
01:41 PM unterhaus_: I think Gene forgot to tell freecad what he wanted it to export gcode for. Who is with me?
01:58 PM pfred1: I think freecad does not really support US Customary
02:00 PM _unreal_: ?
02:00 PM _unreal_: heh I can run circles around freecad
02:00 PM pfred1: _unreal_ can you tough guy?
02:01 PM _unreal_: wheres my strong arm emoji when I need it
02:09 PM CaptHindsight: i think he meant that is what he usually does, run around in circles
02:09 PM CaptHindsight: gets dizzy and then falls
02:31 PM centreline: hmm so I measure a dc voltage of 2.5v on step and dir when axis is commanded to move, but no pulses or waveforms on either wires
02:34 PM CaptHindsight: centreline: what are you hoping to discover?
02:34 PM centreline: why the axis doesnt move at all
02:35 PM centreline: it increments on screen, but on the axis nothing, I wanted to know if there was a hardware issue but I think its my config somewhere
02:35 PM CaptHindsight: You can measure and compare the Step voltage when you are stepping vs not stepping
02:36 PM centreline: the voltage is 2.5v higher when its "moving" vs when its in estop
02:36 PM CaptHindsight: Dir will be high or low at a steady state unless changing directions so you can measure it in both directions and compare
02:37 PM CaptHindsight: have you swapped the Dir and Step pins to the stepper driver by change? It's a common mistake
02:37 PM CaptHindsight: what are you driving the Dir and Step pins with? Break out board and LPT port or?
02:38 PM CaptHindsight: change/chance
02:39 PM centreline: 5i75 with a 7i76, I have migrated it from a breakout board to the mesa boards and redone signal wiring, so just triple checking now
02:39 PM CaptHindsight: if we know what you are driving it with and which stepper driver you are usng we can figure out if you have it wired correctly
02:39 PM CaptHindsight: thanks
02:39 PM CaptHindsight: and which stepper driver?
02:40 PM centreline: I need to check the pinouts on the 7i76 again, I'm following norberts diagram and already noticed a port mistake with encoder names
02:40 PM centreline: they are dm860a's
02:46 PM CaptHindsight: https://www.longs-motor.com/stepper-motor-driver-dm860a.html
02:52 PM centreline: they are the drivers, I have a print out of same and the signal wires match where they should be on the 7i76
02:52 PM centreline: I measured dir voltage and its 5v in one direction, and -5v in the other
02:53 PM centreline: I thought I should see step pulses on the pulse wires on my (physical) oscilloscope but only a continual dc voltage on pulse wires
02:54 PM CaptHindsight: centreline: did you have this working before on your breadboard?
02:55 PM centreline: I had this machine working previously with a parallel port bob, same psu, drivers, steppers. Ive rewired the signal side of the drivers for the mesa card change
02:56 PM centreline: previously I had the ena hard wired all the time to 5v, the only change is this time I'm letting the mesa send 5v enable to the drives (because its safer)
02:56 PM Tom_L: centreline, it's hard to measure the voltage on a pwm signal
02:56 PM centreline: I checked ena wires and they have 5v going to them when the machine is out of e-stop
02:57 PM CaptHindsight: centreline: but not working yet with the Mesa 5i75 and 7i76? or worked for a bit and then stopped working?
02:57 PM Tom_L: but if you saw it on a scope it may very well be 5v
02:57 PM Tom_L: a 50% duty pulse may very well show up as 2.5v
02:58 PM centreline: no ever had it working with the mesa, I got all the limit switch/mpg and estop done first and stepper the last job
02:58 PM centreline: Tom_L, its just flat dc, there's no variance
02:58 PM Tom_L: what's your config line look like?
02:58 PM CaptHindsight: measuring 2.5v and 5V across what exactly?
02:58 PM CaptHindsight: more questions
02:58 PM Tom_L: the sun and moon
02:58 PM centreline: let me post it to pastebin
02:58 PM Tom_L: duh
02:59 PM centreline: 5v between dir -ve and dir +ve wires as they enter the driver
02:59 PM Tom_L: he said 5v on the ena line
02:59 PM centreline: my multimeter and scope don't agree
02:59 PM Tom_L: which one cost more?
02:59 PM centreline: well the d88 but its older than me and smells funny :)
03:00 PM centreline: but they both see a static dc
03:01 PM CaptHindsight: centreline: are you tying the PLS+ , DIR+ and ENA+ all to 5V?
03:01 PM centreline: no I'm not, I did in the previous wiring though
03:02 PM CaptHindsight: centreline: or are you tying all the PLS-, DIR- and ENA- to Ground?
03:02 PM centreline: the diagram I am following just shows individial wires from the screw terminals on the 7i76 all the way to the steppers
03:02 PM Tom_L: servos?
03:03 PM centreline: no steppers, I built this years ago and they came free with the electronics devioid machine from john stevenson
03:03 PM CaptHindsight: I'll sit back and watch Tom_L help
03:03 PM Tom_L: naw, i got work to do
03:03 PM Tom_L: go ahead
03:03 PM CaptHindsight: having breakfast
03:03 PM Tom_L: at 3pm?
03:03 PM centreline: should the pls+ dir+ and ena+ be commoned together to 5v?
03:04 PM Tom_L: :)
03:05 PM skunkworks: heh - it looks like threads does show the order correctly
03:05 PM Tom_itx: first off, in your ini you should have a config line something similar to this:
03:05 PM Tom_itx: CONFIG="num_encoders=2 num_stepgens=3 num_pwmgens=1 sserial_port_0=20xxxxxx"
03:05 PM centreline: https://pastebin.com/sQA9jtdr
03:05 PM Tom_L: ignore the pwmgens on yours
03:06 PM Tom_L: skunkworks, got stuff running before it's supposed to?
03:06 PM centreline: hmm I have 5 stepgens but same (its a 5 axis machine when I put the table/rotab on but just one axis will do for now)
03:07 PM Tom_L: ok but is your card bit file set up for 5 stepgens?
03:07 PM unterhaus_: the built version of freecad is .01 versions out of date
03:07 PM unterhaus_: ^for windows
03:08 PM Tom_L: the unbuilt one doesn't build right here either
03:08 PM Tom_L: centreline, and how do you know what pins they are on?
03:08 PM centreline: its default 7i76, I can drop that to 3
03:08 PM skunkworks: Tom_L: sorry - that was out of context. I was wondering if I could see actual thread order in hal.
03:09 PM skunkworks: and you can
03:09 PM Tom_L: skunkworks, i saw your post in devel
03:09 PM skunkworks: right
03:09 PM Tom_L: i suppose load order _could_ matter in certain circumstances
03:10 PM centreline: Im making the assumption that stepgen0 comes out of hm2_5i25.0.stepgen.00, so axis 0 is wired to that port on the 7i76
03:10 PM Tom_L: what do you see in hal showpin?
03:11 PM Tom_L: (i'm starting from the software side rather than hardware first)
03:14 PM centreline: in the gui hal configuration I follow pins>hm2_5i25>0>stepgen>00>enable and its connected to x-enable
03:15 PM Tom_L: the wizzard?
03:15 PM Tom_L: i don't use those
03:15 PM centreline: Im slightly confused by that, because I wondered if it should be on the 5i25>0>7i76> section, as the stepper connections are on the 7i76, not the 5i25 port
03:15 PM centreline: whats the cmd line to do a hal showpin?
03:15 PM centreline: I dont appear to have a hal and halcmd doesnt know showpin, sorry to be dull
03:16 PM Tom_L: hm2_5i25>0>stepgen>00.7i76.0.xxxx or something like that
03:16 PM centreline: no its not listing the 7i76 in the path, thats where I think it could be wrong
03:17 PM centreline: but there is no stepgens listed in the 7i76 section
03:17 PM Tom_L: http://linuxcnc.org/docs/2.7/html/hal/halshow.html
03:17 PM Tom_L: you can set the watch window to see certain pins toggle
03:19 PM centreline: all my working gpio is on hm2_5i25.0.7i76.0.0.x but there are no stepgens listed in that section at all
03:19 PM pcw_home: Are you sure you have your drive enables wired correctly? (most step drive "enables" are actually "disables")
03:19 PM Tom_L: umm i don't think they show up as stepgens
03:19 PM Tom_L: but if they're configured right they are there
03:19 PM centreline: so 5v disables the drives?
03:19 PM pcw_home: the stepgens will not have "7i76" in their names
03:20 PM pcw_home: Yes 5V disables the drives
03:20 PM Tom_L: woops
03:21 PM centreline: ok, assumption wrong, I assumed they were disabled until 5v was applied
03:21 PM centreline: previously they always had 5v though
03:21 PM centreline: let me check, it still could be a assumption
03:22 PM centreline: I mean mine, wrong still, not yours
03:23 PM centreline: "Enable signal: ENBL+ is the positive end of direction input pin. This signal is used for enabling/disabling the driver. High level for enabling the driver and low level for disabling the driver."
03:23 PM centreline: but then the next row down... "ENBL- is the negative end of direction input pin. Usually left unconnected (enabled)"
03:24 PM centreline: Ill take the wire off ENBL- and try it
03:26 PM centreline: should ENBL+ DIR+ and PULS+ be linked together as I had it the first incarnation?
03:29 PM centreline: so ENBL- disconnected does nothing
03:30 PM pcw_home: Well in most step drives ENA must disconnected to work
03:30 PM pcw_home: does the step motor hold position? (you can check enable with this)
03:32 PM centreline: yes, when I power up the dc rail the stepper locks up
03:33 PM centreline: hmmm even in e-stop
03:33 PM centreline: I'll pull the ena plug
03:34 PM Deejay: gn8
03:34 PM centreline: its still locked with the plug pulled... now even more confused
03:34 PM pfred1: isn't a locked up stepper motor enabled and just not moving?
03:35 PM pfred1: also known as holding
03:36 PM Tom_L: if it's truely disabled the coil lines should be at 0 potential
03:36 PM pfred1: yeah that'd be free
03:37 PM pfred1: stepper motors are like if an idiot designed a motor
03:37 PM centreline: there's always 5v at the 7i76 enable terminals even when its in e-stop
03:37 PM centreline: and as a result the motor is always locked when the dc psu is turned on
03:38 PM pcw_home: Enable terminals?
03:39 PM pfred1: if you want a stepper motor to not move having it enabled is best
03:39 PM centreline: there are two terminals on each stepper connection marked 0v and 5v, I read somewhere (I thought) that these were to enable the drives
03:40 PM pfred1: by default not moving is what stepper motors do
03:40 PM pfred1: you power them up and they just sit there locked up
03:40 PM pcw_home: No, those are just 5V and ground (for single ended drives)
03:40 PM centreline: yes, so somehow I dont think its being told to move
03:41 PM centreline: pcw_home, ok, so the enable circuit is wired wrong, but its always got 5v and is enabled even when in e-stop it seems
03:41 PM centreline: thats another issue I can address when the axis moves
03:41 PM pfred1: you don't tell steppers to move you send them step sequences
03:41 PM pcw_home: That's true since its just 5V power
03:41 PM Tom_L: the enable is just a gpio afik
03:41 PM Tom_L: unless you wired it to 5v
03:42 PM pfred1: drives have step and direction inputs
03:42 PM pcw_home: Did you use pncconf to create your hal/ini files?
03:42 PM centreline: yes initially
03:43 PM pcw_home: If wired correctly, the most common issue I have seen is that the step pulse length is set too short to be recognized
03:44 PM pcw_home: I would set the steplen to 5000 initially
03:44 PM pfred1: the setup time?
03:44 PM pcw_home: no, just the pulse length
03:45 PM centreline: hm2_5i25.0.stepgen.00.steplen?
03:45 PM centreline: if so I had that set to 1 :(
03:45 PM pcw_home: initially I would use steplength,stepspace,dirsetup,dirhold = 5000,5000,10000,10000
03:45 PM Tom_L: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing
03:45 PM pcw_home: 1 would be 1 ns...
03:45 PM Tom_L: has a list of various drivers
03:45 PM pfred1: yeah too fast
03:46 PM centreline: for some stupid reason, this machine ran with these valves as parport
03:46 PM pfred1: would be cool if it could work that fast though
03:46 PM Tom_L: centreline, parport is alot slower
03:46 PM centreline: values even
03:46 PM Tom_L: and it relied on the base period
03:46 PM pcw_home: 30 ns is the minimum on a 5I25
03:47 PM pfred1: parallel ports are like what 1 MHz?
03:47 PM CaptHindsight: can be
03:47 PM pcw_home: The steplength is not used on parallel ports (in the normal "reset" mode)
03:48 PM CaptHindsight: pfred1> stepper motors are like if an idiot designed a motor, this should be on a t-shirt or coffee mugs
03:48 PM pcw_home: and in the non-reset mode, its just one base thread period
03:48 PM pcw_home: (minimum)
03:49 PM pfred1: CaptHindsight well everything about them defies logic
03:49 PM centreline: ok I set those values to what you said but it still does not move when jogged
03:49 PM pfred1: CaptHindsight a stepper motor is never more powerful than when it isn't moving
03:50 PM pcw_home: well those timing changes are necessary, but not sufficient...
03:50 PM pfred1: yeah go bigger
03:50 PM pfred1: 5000 ns would be good
03:51 PM pcw_home: you might try with ENA connected or not
03:51 PM pcw_home: also verify that direction changes when you jog in different directions
03:51 PM centreline: the manual handwheel is locked
03:52 PM centreline: I verified that dir is +5v one direction, -5v the other
03:52 PM centreline: manual handwheel = the one to move x axis manually
03:53 PM pcw_home: OK so direction changing pretty much guarantees that you are getting step pulses
03:53 PM _unreal_: heh I still need to get my motor controllers functing PERIOD
03:53 PM _unreal_: sigh
03:53 PM centreline: but I dont see pulses on the pulse wires using a oscilloscope
03:53 PM _unreal_: lack of time
03:54 PM _unreal_: after I get my daughter going to try what ve7it mentioned adding 1k resistors as pullups on the encoders
03:54 PM pfred1: _unreal_ I made my first motor drives
03:54 PM pfred1: _unreal_ and it took me a real long time
03:55 PM centreline: maybe the oscilloscope isnt fast enough to see the pulses and just appears as a solid line, its a pretty low frequency one
03:55 PM centreline: but I can see just the dc voltage solid with some tiny amount of noise on it
03:56 PM pcw_home: You can temporarily change the steptype to 2 and you will have square waves on STEP and DIR
04:01 PM centreline: Ah, I can see that clearly now, I'll put it back to type 0 and see if its still visible
04:02 PM centreline: yes, the pulse is very fast, 9ns if my reading is correct in type0, wtih the 0v time much longer
04:05 PM pcw_home: Sounds like you have not succeeded in changing the step timings
04:05 PM centreline: 1us, operator error...
04:06 PM pcw_home: (pncconf created files set these values in the ini file)
04:06 PM veegee: Cheapest way to get empty oil jugs?
04:06 PM pcw_home: and pncconfs default is 1 us (1000 ns)
04:06 PM veegee: Just want to dispose of my waste oil
04:06 PM veegee: So any container is fine. I have a lot that's been accumulating over the years, but it's sitting in reusable oil drain pans
04:07 PM veegee: So obviously I don't want to throw those out
04:07 PM _unreal_: I only made them because the cost of dc servo drivers is just stupid expensive
04:07 PM veegee: Just need any cheap container I can dump the oil into and dispose of it at my local mechanic
04:07 PM CaptHindsight: veegee: I get rid of 200L drums all the time
04:08 PM pcw_home: Somehow I merged your name with getting rid of oil and thought you were trying to get rid of vegetable oil
04:08 PM veegee: CaptHindsight where do you even get those barrels?
04:08 PM CaptHindsight: look for used empty drums at car washes and similar
04:08 PM veegee: and aren't they expensive!?
04:08 PM CaptHindsight: 55gal steel empty is $80
04:08 PM CaptHindsight: for new
04:09 PM CaptHindsight: i throw them out all the time
04:09 PM veegee: You throw away $80 barrels all the time!?
04:10 PM CaptHindsight: last time I had 30-40L of used oil I made a few trips to walemart, dropped off 3-4 of the 5L containers of used as a time to 4 different locations
04:10 PM Rab: veegee, I use a 5-gallon bucket and dump it into the tank at the auto parts store. Used 55gal drums are cheap, they're excess packaging. Less if you don't care about food safe. Check Kajiji.
04:10 PM CaptHindsight: veegee: yes, nobody wants used unless they have been rinsed
04:10 PM CaptHindsight: it costs me $ to clean them to really clean
04:11 PM veegee: Did not know that, thanks for the info
04:11 PM CaptHindsight: so i give them a wipe and take them to a recycler
04:11 PM Rab: https://austin.craigslist.org/for/d/buda-55-gallon-burn-barrels/7188176377.html
04:12 PM veegee: Rab ah yeah good idea. I assumed you have to give them the whole bucket. I guess I can just take my drain pans directly to the mechanic and empty it myself
04:12 PM CaptHindsight: Rab: I ran an add and nobody wanted them
04:13 PM Rab: CaptHindsight, yeah, the big secondary market I know of is DIY drum smokers, and people are looking for food safe barrels for that (if they're smart).
04:13 PM CaptHindsight: veegee: if you fill drum then you have to move 250Kg to a waste area
04:14 PM CaptHindsight: so you push the problem to a full drum and then need to get rid of the full drum
04:14 PM veegee: Ah yeah good point
04:14 PM CaptHindsight: so i just use walmart
04:14 PM veegee: I guess then I can just take my oil drain pans and empty it myself and not worry about getting a drum
04:14 PM CaptHindsight: organic solvents go into the waste oil, they don't seem to care of the oil is a bit thin
04:15 PM veegee: Yeah any hydrocarbon can go into the same barrel
04:15 PM pfred1: make a waste oil burner
04:15 PM veegee: That's what the mechanic shops do
04:15 PM veegee: They use the waste oil to heat the shop in the winter
04:15 PM Rab: I may have concealed a multitude of sins in a bucket of waste oil.
04:15 PM veegee: But I can't imagine how toxic the exhaust is to the environment
04:15 PM CaptHindsight: hah
04:16 PM CaptHindsight: https://www.ntotank.com/330gallon-nto-white-reconditioned-ibc-tote-tank-x8680616
04:16 PM pfred1: veegee someone is going to burn is womewhere down the line
04:16 PM pfred1: somewhere even
04:17 PM CaptHindsight: i read that as "womanwhere" :)
04:17 PM pfred1: I need one of them lighted keyboards
04:17 PM pfred1: I don't want one that breathes though
04:18 PM Rab: Nota bene, those HDPE tanks turn into algae farms if you use them for rainwater. The black plastic ones are better.
04:19 PM pfred1: dump some bleach in
04:19 PM CaptHindsight: mine are all opaque since they contain photosensitive resins
04:19 PM centreline: I put used diesel from the parts cleaner in the paraffin workshop heater once, and the fumes were indescribable
04:19 PM CaptHindsight: 200Kg of solid plastic vs liquid is an expensive mistake
04:20 PM veegee: centreline yeah imagine how much worse waste oil is. And it's full of carcinogens and dirty combustion byproducts
04:20 PM pfred1: centreline waste heaters run so hot they burn everything up
04:20 PM CaptHindsight: UPS does help me at times https://imgur.com/nFAsIhU
04:20 PM centreline: pcw_home, I got a follow error, so I fixed that and it is about 0.5us on the scope, so I connected it all back up and still nothing, perhaps the dc power side has a issue
04:21 PM centreline: if it has pulse and direction, and the handwheel is locked when the dc psu is powered up, it has to be in dc power side for the driver
04:22 PM CaptHindsight: what plastic pails look like when you place a large boulder on top of them ^^
04:22 PM centreline: the garages around here run waste oil heaters and one doesnt have a chimney and it makes my eyes water to go in there
04:22 PM pcw_home: 0.5 usec is way too short.
04:22 PM pfred1: the best PSU is just a toroidal transformer with a bridge and some filter caps
04:22 PM pcw_home: 5 usec is a good minimum for most step drives
04:23 PM pcw_home: some even need 10 usec
04:23 PM pfred1: centreline shoot for moving slowly then get it going faster
04:23 PM centreline: Ill have to revisit pncconf and do it properly
04:23 PM pfred1: even creeping along is progress
04:23 PM centreline: I set setp hm2_5i25.0.stepgen.00.steplen 5000 in hal file to get that
04:23 PM centreline: but that gives 0.5uS on the scope
04:24 PM pcw_home: what does
04:25 PM pfred1: 5000 ns is 5 us
04:27 PM pcw_home: halcmd show parm hm2_5i25.0.stepgen.00.steplen show
04:27 PM pcw_home: show?
04:27 PM pcw_home: show param I mean
04:27 PM pfred1: show us?
04:28 PM pfred1: you can use the /exec command if you're on your CNC machine
04:30 PM pcw_home: "halcmd show param hm2_5i25.0.stepgen.00.steplen" will verify that your setting is correct (and its not been overwritten somewhere else in the hal file)
04:36 PM centreline: 28 u32 RW 0x00002710 hm2_5i25.0.stepgen.00.steplen
04:38 PM centreline: 0x2700 so 10000 (I tried 10k t see if the step pulse is longer)
04:39 PM pfred1: 9984 decimal
04:39 PM centreline: yes I mistyped its 2710 in output so 10k as configured
04:40 PM pfred1: 0x2710 is 10K
04:40 PM JT-Shop: well that was a nice ride
04:40 PM centreline: this could be related to the oscilloscope, I have a pocket dso in the house I can try tomorrow
04:40 PM pfred1: so convenient they output hex there
04:40 PM Tom_L: JT-Shop, good ice cream at the end of the trail?
04:41 PM JT-Shop: we actually went another way today but yea the ice cream at the Dairy Shack is great
04:41 PM pfred1: can't make things too user friendly
04:42 PM Tom_L: JT-Shop, on that estop latch code what causes a reset on it?
04:43 PM pfred1: centreline your motor drives have step and direction inputs right?
04:44 PM centreline: yes they do
04:44 PM pfred1: centreline you don't have them wired wrong?
04:44 PM Tom_L: JT-Shop, iocontrol.0.user-request-enable but what triggers that?
04:45 PM centreline: I have checked and no, but tomorrow I think I'll try them wrong in case something stupid like the silkscreen on the cases is wrong
04:45 PM JT-Shop: I think when you release the estop latch fault bit
04:45 PM JT-Shop: in or out?
04:45 PM pfred1: centreline reversing them can make motors lock up
04:45 PM JT-Shop: oh you said user not emc
04:46 PM centreline: pfred1, am I looking for dir- and dir+ reversed , or dir swapped with pulse?
04:46 PM JT-Shop: I'd have to think on it a bit
04:46 PM Tom_L: i'm trying to clear my head so i can sit down and work on it a bit
04:46 PM Tom_L: there's no rush here, i don't have all the parts anyway
04:47 PM pfred1: centreline usually you hook dir - to ground and just feed your signal to dir +
04:47 PM Tom_L: i just thought i'd mount the estop button and wire up a couple things and see
04:47 PM JT-Shop: I need to refresh my memory I did that 8 years ago lol
04:47 PM Tom_L: should be crystal clear by now :)
04:48 PM centreline: pfred1, on the 7i76 there is a dir- dedicated terminal so I followed that
04:48 PM pfred1: centreline they break it out that way in case yo uwant to do something kooky like negative logic
04:48 PM pfred1: centreline then that should just be ground
04:49 PM Tom_L: JT-Shop, on the pendant i may need a push on/push off instead of momentary button
04:49 PM pfred1: centreline logic is 0 or 1 there's no -1
04:49 PM roycroft: i have finish on my chop now, and it is beautiful
04:49 PM roycroft: it turned out better than i hoped
04:49 PM Tom_L: or a toggle comp
04:50 PM pfred1: centreline in your motor drive there should be an optocoupler which is an LED
04:50 PM pfred1: centreline you're essentially making the light flash on and off
04:52 PM pfred1: there's no such thing as minus light either
04:53 PM centreline: it has two leds, power and fault, all the power leds light all the time, and fault stays off
04:53 PM pfred1: unless they start packing little black holes in gadgets
04:53 PM centreline: if we are talkng status lights
04:53 PM pfred1: no the logic signals go through leds too
04:53 PM pfred1: they're encapsulated so you can't see it
04:54 PM centreline: yes I understand how optocouplers work
04:54 PM pfred1: that's good
04:54 PM pfred1: took me a while to figure it out myself
04:55 PM pfred1: I spent some time tailoring drive strength to my optos
04:55 PM centreline: its nearly midnight and Ive been on this all day, I'm now so tired I potentially will make a stupid mistake so I'll take stock tomorrow and bring out the mini dso because its relatively trustworthy
04:55 PM pfred1: I put lower value resistors on the step as opposed to the direction inputs
04:56 PM centreline: but I do appreciate all the input
04:56 PM pfred1: 220 vs 330 Ohm
04:56 PM * JT-Shop is really tired of that annoying do you really want to quit from axis that is so stupid
04:56 PM centreline: I only normally use this d66a for looking at waveforms on alternators or lissajous figures when tuning stuff
04:56 PM pfred1: do you really wanna?
04:57 PM pfred1: I saw some weird things setting my timings up
04:58 PM pfred1: my drives like a 50:50 square wave
04:58 PM pfred1: half up and half down
04:59 PM pfred1: they run with more of a pulse wave but not as good
05:01 PM centreline: right, thanks for all the assist all, time to rest and work through it with a clear head tomorrow
05:01 PM pfred1: yeah good luck
06:08 PM _unreal_: still such a difference what the upgrades on my machine made
06:28 PM veegee: Ha found one: https://www.kijiji.ca/v-other-business-industrial/city-of-toronto/metal-steel-drum-barrel-205l-55gal-clean-empty-intact-lid-caps/1524141670
06:28 PM veegee: Some people selling barrels for $40 each
06:29 PM XXCoder: 40? that page says $5
06:34 PM roycroft: 55 gallon steel food grade drums with detatchable lids go for $20 here
06:35 PM * roycroft needs to pick one up for his dust collection system, and has been pricing them lately
06:37 PM veegee: XXCoder I know, others are selling for $40
06:37 PM XXCoder: oh ok
06:37 PM veegee: I also see listings for free: https://www.kijiji.ca/v-storage-containers/city-of-toronto/55-gallon-steel-drums/1521683359
06:38 PM veegee: roycroft I also use them for my cyclonic separator
06:38 PM veegee: they do a good job of not collapsing under vacuum pressure. Also easy to ground
06:38 PM veegee: You haven't lived until you've hooked up every vacuum cleaner/dust collector you have with a cyclonic separator.
06:52 PM XXCoder: nice
06:52 PM XXCoder: if I ever build a shop, im definitely building one collector
06:52 PM XXCoder: keeps everything clean
07:08 PM unterhaus_: sorta convenient that kollmorgan bought all the servo motor companies
07:08 PM unterhaus_: it's nice they keep the old manuals and software around as well
07:25 PM jdh: they sent me a scan of a drive manual for some ancient drive I had a few years ago
07:59 PM CaptHindsight: veegee: that's near the Tim Hortons :)
07:59 PM skunkworks: ugh.. Got back to figuring out why our electrical bill is so high..
07:59 PM CaptHindsight: something got left on?
08:00 PM CaptHindsight: leaky electric meter?
08:00 PM skunkworks: measured a bunch of the branches and everything looked ok until I measured the outside air unit (compressor)
08:00 PM skunkworks: and it was drawing 2.5a at 240v..
08:01 PM CaptHindsight: when off or on?
08:01 PM XXCoder: dang thats a lot of power
08:01 PM skunkworks: it has been running off
08:01 PM XXCoder: if it was off anyway
08:01 PM CaptHindsight: oh wow
08:02 PM skunkworks: it had a heater option that was enabled. runs 24/7
08:02 PM CaptHindsight: yeah next question
08:02 PM skunkworks: well - never disabled.. I would guess
08:02 PM skunkworks: this is 1984 tech..
08:02 PM CaptHindsight: why i never run my mini-splits on auto
08:02 PM CaptHindsight: only on cool
08:03 PM CaptHindsight: electric heat is $$$
08:03 PM skunkworks: the heater was just for the compressor
08:03 PM CaptHindsight: deicer
08:03 PM skunkworks: for cold temps.. Who uses their air when it is cold..
08:04 PM CaptHindsight: so ~1Kw every 2 hours
08:05 PM CaptHindsight: x12 x 30 days
08:06 PM skunkworks: yep
08:06 PM skunkworks: for 10+ years ;)
08:06 PM CaptHindsight: 432 KWh for the month.... oh no
08:06 PM skunkworks: The last time I checked - I don't think I even thought of checking the compressor..
08:07 PM skunkworks: when it is off - it is off - right? ;)
08:07 PM CaptHindsight: that sucks
08:07 PM JavaBean: usually
08:07 PM Rab: I use my mini-split as a heat pump in the winter, it's nice...but not much cheaper than running A/C in the summer in Texas.
08:08 PM CaptHindsight: we pay commercial rates so winter summer is not very different since they get you for just having service
08:09 PM CaptHindsight: >$100/month just for service
08:09 PM Rab: What FreeCAD is the best FreeCAD, these days? Stable, or development?
08:10 PM skunkworks: about $50 if I did the math right
08:18 PM unterhaus_: Rab, someone on the ML just said all the tutorials are out of date if you run devel, so maybe stable?
08:21 PM unterhaus_: my big monitor is on the windows computer, so I'm stuck with something fairly old
08:21 PM Rab: unterhaus_, thanks! It looks like there are no appimages built for 18.4, so I'm trying 18.3 now (on Linux).
08:22 PM _unreal_: rab you get your starboard project done yet?
08:23 PM Rab: _unreal_, haha, it's a 2021 project. I did get a piece of white material off of eBay last week to experiment with corner rounding and polishing.
08:24 PM _unreal_: what do you thinki?
08:24 PM _unreal_: it mills wonderfully
08:24 PM Rab: It certainly looks nice. I was thinking I'd sand the surface texture off, but I think I might leave it as is.
08:24 PM _unreal_: espeically if you use a ball end. easy to make edges. unless you use fixed form round overs
08:25 PM _unreal_: you can also heat form it
08:25 PM _unreal_: lets say you need a bend
08:25 PM unterhaus_: going to go braze something
08:26 PM _unreal_: pure starboard https://www.deepbluemarine.com/images/products/599-1-md.jpg
08:27 PM _unreal_: rab https://i.pinimg.com/originals/6d/ce/6c/6dce6c80a7037886256289e77ba6add9.jpg
08:43 PM JavaBean: _unreal_: what is starboard?
08:43 PM Rab: JavaBean, https://www.kingplastic.com/products/king-starboard/
08:43 PM Rab: It's a proprietary HDPE material.
08:45 PM JavaBean: interesting
08:48 PM _unreal_: its used extensivly in boats
08:49 PM _unreal_: the cncmachiune I'm building for my daughter is mosty made of starbard
08:56 PM renesis: so freecad doesnt do assemblies?
08:57 PM unterhaus_: freecad says it does assemblies
08:58 PM unterhaus_: there is an assembly workbench
08:58 PM Rab: renesis, this seems to be the bleeding edge: https://github.com/realthunder/FreeCAD_assembly3
08:58 PM renesis: okay so i guess the answer is kinda
09:00 PM roycroft: either a2plus or a3 are what seem to be the promising assembly workbenches for 0.19
09:00 PM roycroft: i wish to hell that the freecad folks would just declare a winner in the assembly workbench race
09:00 PM roycroft: even if the don't do anything for a while
09:01 PM roycroft: at least we'd know where they're going to go
09:01 PM roycroft: there are a lot of things that are annoying, weird, and broken in freecad
09:01 PM roycroft: almost all of which can be tolerated
09:02 PM roycroft: but the assembly-workbench-of-the-week crap, with zero interoperability between the contestants, makes it infuriating and almost unusable
09:02 PM renesis: hmm sounds like best bet is to get friend to ask ME professor wife for another academic solidworks license
09:03 PM roycroft: all they have to do is say "we pick assembly workbench x for eventual integration into the official freecad distro
09:03 PM roycroft: "
09:03 PM roycroft: they don't have to say when it's going to happen
09:03 PM roycroft: they don't have to do anything else
09:03 PM roycroft: just decide which one is going to be the one, and announce that
09:04 PM renesis: sounds like someone needs to setup a deathmatch
09:04 PM roycroft: it's not quite as simple as i describe
09:04 PM renesis: two developers go into the cage, only one comes out
09:05 PM roycroft: ASSEMBLYDOME!
09:05 PM roycroft: two workenches enter, one workbench leaves!
09:05 PM Rab: KiCad had an almost identical problem (maybe still does?).
09:05 PM renesis: or at least gets sent off to apache to die a slow death
09:06 PM roycroft: the problem is that the different assembly workbenches do thing so differently that it would be impossible to create an assembly with one and open it with another
09:06 PM renesis: i have no friend with EE professor wife to get me altium license
09:07 PM roycroft: and extremely difficult, if not impossible to translate one to another
09:07 PM roycroft: so if you create an assembly now
09:07 PM roycroft: in the next release you may have all your individual parts, but have to recreate the assembly from scratch with the new assembly workbench
09:07 PM roycroft: and that's just
09:07 PM renesis: sadfaces
09:08 PM roycroft: well, i'm not going to say because i'm trying not to use pejorative language
09:08 PM renesis: angryfaces
09:09 PM Rab: roycroft, so which workbench are you using now (if any)?
09:10 PM roycroft: i have a2plus and assembly3 installed
09:11 PM roycroft: i mostly still use solidworks though
09:11 PM Rab: Boo, 0.18-16131 breaks my spacenavigator
09:11 PM roycroft: and unless/until the assembly mess gets sorted out in freecad, i'll probably continue to rely on solidworks and play with freecad as a toy
09:12 PM roycroft: 0.18 crashes on mac os quite reliably
09:12 PM unterhaus_: I think all I need right now avoids assemblies, but maybe not
09:12 PM roycroft: i'm on 0.19-22284
09:12 PM roycroft: most of what i do entails assemblies
09:13 PM roycroft: and the parts that don't, i'd like to be able to use in assemblies if/when needed
09:13 PM roycroft: if i'm honest, if i don't have to fit parts together i can draw them faster in autocad
09:13 PM roycroft: 2d
09:14 PM roycroft: i make 3d drawings when i need to fit multiple parts together and want to make sure they'll fit
09:15 PM roycroft: my use may not be typical, in fact i'd almost bet it isn't
09:15 PM unterhaus_: I never learned autocad
09:15 PM roycroft: but the way i use 3d cad software, assemblies are essential and fundamental
09:15 PM Tom_L: you're better off not knowing
09:15 PM unterhaus_: assemblies need to work, too bad they can't get it right
09:15 PM roycroft: not nice add-ons
09:16 PM roycroft: i've been using autocad for so long i can't remember
09:16 PM unterhaus_: I have been getting autocad drawings from someone and they suck
09:16 PM roycroft: i was first exposed to it in the '80s
09:16 PM roycroft: when it was totally command line driven and ran on ms-dos
09:16 PM jdh: not much different now
09:16 PM unterhaus_: I had an msdos cad, forget what it was
09:17 PM Tom_L: i'l generally take a dxf and throw it into solidworks, add some depth to it and export it to my cad cam via iges
09:17 PM roycroft: it is really geared towards click and drool now
09:17 PM roycroft: and is all ribbony
09:17 PM roycroft: but the cli is still there
09:17 PM jdh: you don't still just type the commands?
09:17 PM Tom_L: turning a 2d to 3d for export purposes
09:17 PM roycroft: yes, i do
09:18 PM roycroft: but most people i know who use it don't know how to
09:18 PM jdh: way faster for me
09:18 PM roycroft: all the young whippersnappers
09:18 PM roycroft: yes
09:18 PM jdh: <- ancient
09:18 PM roycroft: it's 2-3x faster to type the commands than to find them in the ribbon and click on them
09:18 PM jdh: not JT ancient of course, but...
09:18 PM Tom_L: hah
09:19 PM Tom_L: good he's asleep or he might smack you with his cane
09:19 PM roycroft: having just said that i started using autocad in the '80s, i probably qualify as ancient as well
09:19 PM Tom_L: mine came on floppies
09:19 PM * roycroft hands jdh the jar of doan's pills
09:19 PM roycroft: have a few and pass them along
09:19 PM * JavaBean grabs his bowl of popcorn
09:20 PM Tom_L: my first cad cam ran on dos
09:20 PM unterhaus_: I got a pneumonia vaccine 4 days ago, still can tell
09:20 PM unterhaus_: there was cam on dos?
09:20 PM Tom_L: yes
09:21 PM Tom_L: there was ncpolaris addon for acad too
09:21 PM Tom_L: i never got it but i knew a guy that could do it practically with his eyes closed
09:21 PM roycroft: so my workbench no longer has any tenons protruding through the top of it
09:21 PM roycroft: they've all been cut down and planed flush
09:22 PM roycroft: the bench top is not flat yet, though
09:22 PM roycroft: it's not too far out
09:22 PM roycroft: but it's going to take a few hours of hard work (hand planing, mostly with a no. 7 jointer plane) to make it totally flat
09:24 PM roycroft: once i get the chop reinstalled and make some bench dogs for the wagon vise, it will be fully functional
09:24 PM roycroft: i can flatten it over time
09:25 PM Rab: Woot, spacenavigator is back in 0.19-22474.
09:26 PM roycroft: it is?
09:26 PM roycroft: cool
09:26 PM roycroft: i should try mine with it
09:27 PM roycroft: but i'm on 0.19-22284, and i don't know if there's a newer mac os build
09:29 PM roycroft: yes, there is
09:29 PM roycroft: 0.19-22474 has a mac os build
09:30 PM Rab: Graphics performance is also much improved for me, from older 0.18 builds.
09:31 PM roycroft: 0.18 would crash within seconds of launching on mac os
09:31 PM Rab: Wait, no it's not. I just had a simpler model loaded.
09:31 PM roycroft: every version of 0.18 i tried
09:31 PM roycroft: which is why i'm on 0.19
09:33 PM unterhaus_: are there alternate builds somewhere?
09:33 PM roycroft: https://github.com/FreeCAD/FreeCAD/releases/tag/0.19_pre
09:33 PM unterhaus_: so far I haven't crashed .18
09:33 PM roycroft: that's where the latest development versions can be found
09:36 PM Rab: Huh...Assembly3 seems to be included.
09:36 PM unterhaus_: thanks. I wonder if windows versions can coexist
09:37 PM unterhaus_: you can tell when an old manual uses an autocad drawing, barely readable
09:41 PM Tom_L: is it 32 or 64 bit?
09:41 PM Tom_L: or both
09:42 PM roycroft: i don't know if different versions can coexist
09:42 PM roycroft: i'd run 0.19 in a vm
09:42 PM roycroft: and keep 0.18 on the main machine
09:44 PM Tom_L: looks like either 64 or both i can't tell
09:57 PM roycroft: the mac os version is 64-bit only now, because mac os no longer supports 32-bit binaries
09:58 PM Rab: The Linux build is x86-64.
09:58 PM roycroft: i'm not sure many folks run 32-bit linux any more
09:58 PM Tom_L: yeah i saw that but does that mean both or 64 only?
09:59 PM Rab: FreeCAD_0.19-22474-Linux-Conda_glibc2.12-x86_64.AppImage: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.18, stripped
10:00 PM Rab: I don't want to make assumptions, but that seems like something that won't execute on a 32-bit kernel/distro.
10:01 PM Tom_L: you're probably right
10:02 PM Rab: But maybe the source can be built for 32-bit.
10:02 PM MrHindsight: trying the new Buster 2.8 ISO and Debian still sucks
10:03 PM MrHindsight: if your network doesn't use DHCP the network manager is useless
10:03 PM roycroft: i've found debian to be the most sane linux distro for a person coming from the bsd world
10:04 PM MrHindsight: the Rpi sdcard image has the same issue
10:04 PM * roycroft just edits /etc/network/interfaces and all is well
10:05 PM MrHindsight: no DHCP then you have to configure it all manually by editing /etc/network/interfaces.d/
10:05 PM roycroft: or let the installer do it
10:05 PM MrHindsight: heh, they still haven't fixed this so it may be done graphically
10:05 PM roycroft: if you don't want to do it by hand
10:05 PM MrHindsight: the Rpi image doesn't do it
10:06 PM roycroft: oh, i don't know about that
10:06 PM roycroft: but the standard installer lets you configure the network manually if dhcp does not work
10:06 PM MrHindsight: opps i skipped setting up networking during the debian install
10:07 PM MrHindsight: well now you have to do it all with an editor, the network manager won't help you
10:07 PM roycroft: anyway, /etc/network/interfaces does not require rocket science to edit
10:07 PM MrHindsight: not very friendly :)
10:08 PM roycroft: address
10:08 PM roycroft: gateway
10:08 PM MrHindsight: huh being a noob I can't use my PC to search the internet for instructions on how to do this
10:08 PM roycroft: that's not unfriendly at all :)
10:08 PM MrHindsight: but a noob won't know this
10:08 PM roycroft: what's unfriendly is being forced to use a gui that takes 10x as long
10:09 PM roycroft: :)
10:09 PM Tom_L: that noobish you should run mach
10:09 PM Tom_L: :)
10:09 PM MrHindsight: nah
10:09 PM MrHindsight: seriously, is this the mindset to not fixing this?
10:09 PM Tom_L: doubt it's that over manpower
10:10 PM roycroft: i don't see what the problem is
10:10 PM roycroft: you said you forgot to set up networking with the installer
10:10 PM roycroft: if you don't forget that, the installer asks you a bunch of questions
10:10 PM roycroft: and it creates the interfaces file for you based on the answer
10:10 PM MrHindsight: it was a scenario :)
10:11 PM MrHindsight: the words to a script :)
10:11 PM roycroft: it seems to me that the problem is forgetfullness
10:11 PM MrHindsight: a noob can't forget, they don't know yet
10:11 PM Tom_L: each time they reinstall they get a little wiser
10:12 PM MrHindsight: oh fuck
10:12 PM roycroft: if there is not entry-level documentation for the installer then i'd agree that's a need that should be addressed
10:12 PM MrHindsight: nevermind
10:12 PM roycroft: but honestly, the installer works pretty well
10:12 PM Tom_L: i haven't tried the new one yet
10:13 PM Tom_L: so i'm just trolling :)
10:13 PM roycroft: the debian installer has not changed substantially in a very long time
10:13 PM roycroft: probably because it works pretty well already
10:14 PM roycroft: if it ain't broke, don't fix it
10:17 PM roycroft: i should clarify that i've never used the debian gui installer - i use the text installer
10:17 PM roycroft: so maybe the gui installer works a lot differently
10:17 PM roycroft: i should imagine it's the same installer but with a gui front end
10:17 PM roycroft: it would be silly to maintain two different installers
10:26 PM MrHindsight: broken for years
10:26 PM MrHindsight is now known as CaptHindsight
10:27 PM CaptHindsight: https://linuxconfig.org/how-to-set-a-static-ip-address-on-debian-10-buster
10:27 PM CaptHindsight: heh assumes you already have DHCP working
10:28 PM CaptHindsight: oh well
10:29 PM roycroft: the only problem with the debian installer that i've found is that it does not give one the option of dhcp vs. static configuration
10:29 PM roycroft: it tries dhcp
10:29 PM roycroft: if it gets an ip address it just moves on
10:29 PM roycroft: if it doesn't it lets you configure it manually
10:30 PM roycroft: that's not usually a problem for me - i mostly build machines on networks that don't have dhcp servers
10:30 PM roycroft: i just know that when i run the installer, when it comes time to set up the network, i can step away and go get a refreshing beverage while waiting for it to time out
10:31 PM roycroft: and there's likely boot time configuration option to tell it to skip trying dhcp
10:31 PM roycroft: i've never bothered to look that up, because it really isn't a big deal
10:33 PM roycroft: to me, that is
10:41 PM skunkworks: https://www.youtube.com/watch?v=WPTcc-0FIaM