#linuxcnc | Logs for 2018-11-10

Back
[00:21:21] -!- Crom has joined #linuxcnc
[00:25:18] <Crom> just popping in to say I won't be around until after my wife passes away from her cervical cancer. She's home from hospital on hospice care. I'd love for her to make it to June so she can see her only child, her daughter, my step daughter walk for graduation from UC Santa Cruz. I'm hoping for her to see Thanksgiving and Christmas at least.
[00:26:00] <XXCoder> hey Crom thanks for letting us know. that sucks. :(
[00:28:51] <Crom> Sure does. We went well together. She cooks, I eat. I fix cars and house, she brakes things (jk). We're both tea drinkers, history nerds, steam punk geeks, and sewers.
[00:30:28] <Crom> been making costumes with her for the last 12 years for Shakespeare in the vine http://shakespeareinthevinesorg My sister in law from my second wife started it and we all joined in.
[00:30:42] <Crom> been making costumes with her for the last 12 years for Shakespeare in the vine http://shakespeareinthevines.org My sister in law from my second wife started it and we all joined in.
[00:31:46] <Crom> Cancer SUCKS!!!! She was stage 4 already when we figured out something was wrong.
[00:32:29] <XXCoder> yeah cancer sucks indeed. :(
[00:34:12] <Crom> I've been fighting the Veterans Administration for the last 10 years trying to get disability. She get's diagnosed Jun 2016, I get rated 100% by the VA in Sep 2017. Now we have money to do things, and she can't do.
[00:34:37] <XXCoder> man timing sucks
[00:34:57] <Crom> effen VA!
[00:35:56] <Crom> May 2016 to Sep 15th 2017 we were living on $860 a month.
[00:37:01] <XXCoder> pretty small budget. used to live on bit less for long time
[00:37:13] <XXCoder> its not even enough to cover rent here
[00:37:41] <Crom> now it's $31xx.xx a month.. received $139k in back pay. spent it on the kids, step kids and grandkids. we couldn't go anywhere...
[00:38:22] -!- ikcalB has joined #linuxcnc
[00:39:21] <XXCoder> heh I almost got the inverse when I tried to stop ssdi
[00:39:35] <XXCoder> they refused to stop and kept sending money. I refused to use it.
[00:39:40] <Crom> house was paid off in Dec2015 after her mom died. Her sibs sold moms house and split the money. Her older bro and we paid off out houses, her sister bought a house outright in Nevada
[00:39:44] <XXCoder> after year of this, they demended money back
[00:40:04] <XXCoder> sure sent em all at once. they lost info, so sent it again. debt-free
[00:40:35] <XXCoder> because ss probably sucks as much as va does.
[00:41:20] <Crom> the same... I filed for SSDI... got denied, haven't worked in the last 8 years
[00:41:35] <Crom> filing for her SSDI now
[00:41:57] <Crom> , I'll get 75% of hers after she's gone.
[00:42:02] <XXCoder> I really hope I wont be on ssdi for 20 more years. I got seriously sick of ss
[00:42:11] <XXCoder> lived on it most of my life
[00:42:16] <Crom> should have filed in 2008 when I filed for the VA
[00:43:34] <Crom> SNAP food stamps counts VA disability pay, the bastards
[00:43:44] <XXCoder> jeez
[00:45:09] <Crom> we got 5 months of CalFresh (SNAP) then the VA money hits.
[00:45:47] <Crom> 5 months of $192 month I think
[00:48:13] <Crom> October our car dies... get it partially fixed $2700, then in June it totally dies... Spend $4000ish getting a 2 year newer car same model *6K miles vs the 280K on the old one.
[00:48:22] <Crom> 86K on the newer
[00:49:33] <Crom> pull all the good parts off the 95 and stick them on the 97. The New radiator, new speakers, stereo, tail light, axles, ...
[00:49:54] <XXCoder> nice wish I had that skill. I just bought 2012 one 2 years ago
[00:50:21] <XXCoder> should be paid off in 2022 whee
[00:51:53] <Crom> Then my wife starts going down hill. 2x Oxycoton 30mg day, 8x 10/325 Norco's day. get a $100 blaupunkt DVD android 2din deck installed so she can be distracted by watching Alton Brown Good Eats
[00:54:04] <Crom> she's now 2x 30mg Morphine Sulfate TER, 8x 5mg Morphine Sulfate liquid for break through pain.
[00:54:31] <Crom> I don't have the strength to do alot of the work now myself.
[00:55:07] <Crom> took me 4 days to pull and replace the radiator.
[00:55:25] <XXCoder> yeah big dismovation :(
[00:55:43] <Crom> strength or stamina...
[00:55:59] <Crom> very dismovational...
[00:56:45] <Crom> and staying with my wife all day is sapping even more strength
[00:57:25] <XXCoder> hard work taking care but better than just send off to die
[01:00:36] <Crom> yep. I'm doing better handling it than I did with Carole
[01:00:51] <Crom> Practice does not make perfect
[01:01:04] <Crom> not the practice I want
[01:01:09] <XXCoder> yeah better handling but doesnt get any easier
[01:01:16] <XXCoder> so many of my family died in 90s
[01:01:20] <XXCoder> and early 2000s
[01:02:34] -!- c-log has quit [Ping timeout: 250 seconds]
[01:02:39] <Crom> but it feels soooo good to be able to help the kids get cars fixed, get presents for grandkids, ... the VA 100% money has really helped.
[01:04:29] <XXCoder> :)
[01:05:06] <Crom> only reason we survived the late summer heat last year was the window A/C I put in. this spring got a HERO loan and had all the windows, the A/C replaced. Also had a Alumawood patio cover put up. Keeps so much heat off the back of the house (where our bedroom is)
[01:05:46] <Crom> so property tax bill due is $2100 in Dec and $2100 in Apr
[01:05:59] <MarcelineVQ> Nice, even just a white tarp over a deck does wonders for keeping a house cool
[01:06:31] <XXCoder> didnt think of deck. interesting. we dont have deck here but other house we had
[01:06:34] <MarcelineVQ> over meaning like a tent, not sitting o the deck floor :>
[01:07:19] <Crom> I want to get Solar on the back of the roof. perfect SW single plane roof. I want to move the vent pipes out of the way as well
[01:07:41] <XXCoder> i wanna try wind power, aliexpress has cheap ones for 400w and 800w
[01:07:45] <XXCoder> $200 or so
[01:07:48] <Crom> the 6" airgap and shading of the panels will keep so much heat out of the attic as well
[01:07:54] -!- c-log has joined #linuxcnc
[01:07:54] -!- c-log has quit [Changing host]
[01:07:54] -!- c-log has joined #linuxcnc
[01:08:25] <Crom> I'm in a hole... wind doesn't blow enough here
[01:08:25] <XXCoder> thats powerful enough for AC. perfect for summer, as well as also add 600w solar since theres fairly cheap fan and power controller that can also have 600w solar on it
[01:08:57] <XXCoder> Crom: other option to consider, there is roofing material that turns white in heat, and black in cold
[01:09:03] <XXCoder> im sure you can see why
[01:09:37] <XXCoder> https://phys.org
[01:10:18] <Crom> 800w is pretty much you want for A/C the fridgaire 5000BTU window something star take 600w. with a battery bank, it's charge, and handle start surge
[01:10:35] <Crom> energystar
[01:11:43] <Crom> batteries will power it past minimum solar to cool the soaked up heat
[01:12:49] <XXCoder> https://www.aliexpress.com example
[01:12:59] <XXCoder> still needs convertor and batteries but quite cheap
[01:14:13] -!- c-log has quit [Ping timeout: 245 seconds]
[01:14:21] <XXCoder> wow everything for me lol https://www.aliexpress.com
[01:14:44] <Crom> wooo I have a concrete tile roof... Their paint version might have posibilities
[01:15:43] <XXCoder> dont look like there is commerical buying for it yet, though maybe different name
[01:17:22] <Crom> probably take 5 years to market. and they'll have better luck licensing it out...
[01:18:21] <XXCoder> news is from 2009 looks like failure to market like so many others
[01:18:41] <Crom> I think my roof is like 34x43 so 27'x35' max solar coverage
[01:19:02] -!- c-log has joined #linuxcnc
[01:21:58] <XXCoder> i want as much solar as possible but expense dont really wortj it in washington state
[01:22:01] <XXCoder> now, wind though
[01:23:01] <XXCoder> summers it shines all time though so nice idea to use it to cover AC usages
[01:28:03] <Crom> I'd want panels that are airgapped from the roof, if they had cooling fins on the back it would be be better
[01:28:50] <Crom> convection cooling... and would keeps the roof dryer as well, making it last longer.
[01:30:05] <XXCoder> you could add tiny solar panel powered fans blowing upwards up the solar back
[01:30:20] <XXCoder> start termal induced air movement going and keep going
[01:30:39] <XXCoder> would work better if radiator has tubes for that
[01:31:36] <Crom> Thermosiphon in the airgap moves quite a bit of air,
[01:32:10] <Crom> almost like building a cold roof for reducing ice dams
[01:33:48] <XXCoder> cold roof is new to me
[01:35:45] <Crom> on youtube 'Purelivingforlife' timber framed house covered in SIP panels with a cold roof for keeping the OSB in the SIPs dry and keeping the roof cold so snow doesn't melt and stick and cause ice dams
[01:36:58] <Crom> 1x2 battens creating a 3/4" gap for air flow. eve is vented as well as the ridge
[01:38:09] <Crom> SIP - batten - plywood - membrane - steel roofing.
[01:38:41] <XXCoder> i also wanna try build air exchanger dual exchangers actually
[01:38:59] <XXCoder> lots of political signs needed, since it uses plastic corggated boards
[01:39:15] <Crom> nows the time to snag them
[01:39:15] -!- Wolf__ has joined #linuxcnc
[01:39:28] <XXCoder> no shop unfortunately to store em
[01:39:40] <XXCoder> oh big one https://ecorenovator.org
[01:40:34] <XXCoder> go to first one
[01:40:45] <Crom> ok 2:21 hours till she gets another morphine oral squirt.
[01:41:07] <XXCoder> first post has dual exchangers I talked about
[01:41:19] <XXCoder> metal corrgated board is best for this but signs is free
[01:41:23] <XXCoder> or very cheap
[01:46:13] <XXCoder> anyway one of ideas is that it exchanges inside air with outside when humidity outside is lower
[01:46:19] <XXCoder> great for shops
[01:46:41] <XXCoder> "pre-exchanger" should be well isulated since it would run at lower temperate when outside is cold
[01:47:37] <Crom> was always thinking about the regenerators dodge used on their turbine cars in the 60's
[01:48:25] -!- ve7it has quit [Remote host closed the connection]
[01:48:36] <Crom> ceramic blades rotating into and out of the diferent temp zones
[01:49:40] <XXCoder> transferring heat?
[01:51:51] <Crom> moving heat from after the power take off turbine and transfering it to the intake air stream after the compressor before the flame can to increase effective compression
[01:52:37] <Crom> helps vaporize fuel as well
[01:55:34] -!- ferdna has joined #linuxcnc
[01:56:08] <XXCoder> interesting
[02:16:41] -!- markus__ has parted #linuxcnc
[02:26:04] -!- Markus_P has joined #linuxcnc
[02:39:14] -!- ferdna has quit [Quit: Leaving]
[02:49:16] -!- Deejay has joined #linuxcnc
[02:49:31] <Deejay> moin
[02:50:08] <XXCoder> yo
[03:14:25] -!- syyl has joined #linuxcnc
[03:40:39] -!- JesusAlos has joined #linuxcnc
[04:29:38] -!- ikcalB has quit [Ping timeout: 245 seconds]
[04:31:13] -!- ikcalB has joined #linuxcnc
[04:52:21] -!- ikcalB has quit [Ping timeout: 252 seconds]
[04:53:51] -!- candyonetoo has quit [Quit: pennen oder so ;)]
[04:56:45] -!- merkatorix has quit [Ping timeout: 252 seconds]
[05:01:23] -!- Akex_ has quit [Ping timeout: 250 seconds]
[05:02:41] -!- Akex_ has joined #linuxcnc
[05:15:59] -!- P1ersson has joined #linuxcnc
[05:48:23] -!- syyl has quit [Quit: Leaving]
[05:53:57] <jthornton> morning
[05:54:11] <XXCoder> yo
[05:57:01] <jthornton> yikes the deep freeze is on it's 28°F with a high of 38°F
[05:57:31] <XXCoder> votex has arrived
[05:57:41] <XXCoder> im at west side, the only area missed by votex
[05:57:47] <jthornton> put a new switch on the test rig and got the same flaky results so I wonder why now
[05:58:02] <XXCoder> cooler though, peak 46f low 36f
[05:58:16] <XXCoder> gets warmer across week, and increasing rain. yay.
[05:59:43] <jthornton> bought a new clock for the night stand and I thought how nice blue LED blue is my favorite color and damn that thing is so bright you can see across the room... had to toss my t shirt over it to get to sleep lol
[06:00:46] <XXCoder> my 20 year old clock with old green lcd has two brown tapes covering it
[06:00:57] <XXCoder> I can still read it in nights but it wont keep me awake.
[06:01:06] <jthornton> the clear brown packing tape?
[06:01:06] <XXCoder> tapes is as old as clock,
[06:01:09] <XXCoder> nope
[06:01:17] <XXCoder> the regular packing tape
[06:01:41] <XXCoder> though its so old its faded a little and I can read it in lit room now
[06:05:28] -!- Akex_ has quit [Ping timeout: 264 seconds]
[06:12:28] -!- Akex_ has joined #linuxcnc
[06:37:58] -!- CubicleNate has quit [Quit: Konversation terminated!]
[06:45:27] -!- homeroot has joined #linuxcnc
[06:45:52] -!- CubicleNate has joined #linuxcnc
[06:45:55] -!- veek has quit [Read error: Connection reset by peer]
[06:46:23] -!- veek has joined #linuxcnc
[06:48:58] -!- sadisticroot has quit [Ping timeout: 252 seconds]
[07:09:42] -!- veek has quit [Quit: Leaving]
[07:10:44] -!- veek has joined #linuxcnc
[07:14:57] <Tom_L> 17�F Hi 44
[07:15:19] <jthornton> another hen has started to moult... dumb birds wait till it's cold to moult
[07:16:24] <Tom_L> dude ^^ it _is_ cold!
[07:18:50] <XXCoder> jthornton: wonder if can use those heated rocks usually used for lizards for those chickens?
[07:19:52] <jthornton> chickens are well insulated balls of down, so long as they have adequate ventilation they are fine
[07:20:11] <XXCoder> cool
[07:24:05] -!- veek has quit [Quit: Leaving]
[07:25:35] <jthornton> a big mistake new chicken owners make is to close up the roost area
[07:27:14] -!- ikcalB has joined #linuxcnc
[07:27:51] <jthornton> moisture from breathing and pooping needs a way to escape or it will condense on the combs and waddles and cause frostbite
[07:31:29] <XXCoder> makes sense
[07:32:58] -!- Akex_ has quit [Ping timeout: 245 seconds]
[07:34:24] -!- Akex_ has joined #linuxcnc
[07:48:11] <jthornton> I still can't figure out why I get multiple hits from the switch... sorta annoying
[07:52:38] <XXCoder> silly question but you added only one event right?
[07:52:48] <XXCoder> if you added 2 listener it would call both
[07:53:14] <jthornton> I added one for each input
[07:53:30] <XXCoder> yeah doubted that was case that you added two
[07:53:43] <jthornton> https://paste.ubuntu.com
[07:53:51] <XXCoder> test out see how riciously large debounce factor get before you get single event?
[07:54:49] <XXCoder> ok I dont know those variables but DOWN on both up and down?
[07:54:55] <XXCoder> lines `6 `7
[07:54:58] <XXCoder> er `6 `7
[07:55:04] <XXCoder> dammit 16 17
[07:55:29] <XXCoder> or is those seperate switches?
[07:55:39] <XXCoder> ah it is nevermind
[07:56:38] <XXCoder> code looks fine?
[07:56:46] <jthornton> something screwy going on, I just clicked one switch with a debounce of 1000000 and got a rising edge on both switches
[07:57:02] <XXCoder> is debounce on MS?
[07:57:15] <jthornton> can crosstalk be the issue at 3.3v?
[07:57:30] <jthornton> yea
[07:57:41] <XXCoder> 3 seconds I wonder if thats too large, try say 300?
[07:58:20] <jthornton> I started at 300, I'm thinking an electrical issue... the cable is a 3 wire from the switch
[07:59:01] <XXCoder> is it packed into single covering or can you seperate em as much as possible?
[08:00:08] <jthornton> I'll rewire it with separate wires or get a chunk of CAT 5 and use that...
[08:01:01] <Tom_L> do you debounce your switch?
[08:01:33] <jthornton> yea for 10 seconds lol
[08:02:02] <Tom_L> 30ms should be plenty
[08:02:06] <XXCoder> do lower volt mean more inference?
[08:02:18] <Tom_L> that or measure the switch bounce
[08:02:24] <Tom_L> then set it
[08:02:30] <XXCoder> Tom_L: that was what I remembered. 300ms is rather large but something up with his setup. code looks fine
[08:02:54] <jthornton> bounce has nothing to do with it, I switched one and got a rising edge on both!
[08:03:20] <Tom_L> it's just eager to get the job done
[08:03:24] <XXCoder> lol
[08:05:34] -!- JesusAlos has quit [Quit: JesusAlos]
[08:07:28] <jthornton> eager to drive me nuts lol
[08:09:22] <Tom_L> double check any var names etc
[08:09:44] <jthornton> https://paste.ubuntu.com
[08:10:22] <XXCoder> his code looked fine tom but then its 5 am im tired lol
[08:10:34] <Tom_L> won't help me.. i don't speak python or arm :)
[08:10:59] <Tom_L> it might come back as klingon
[08:11:05] <jthornton> it reads just like english lol
[08:11:12] <XXCoder> its similiar with C if you know that
[08:11:44] <Tom_L> one thing i learned is sometimes a delay inbetween certain things can help
[08:13:27] <XXCoder> hm I wonder if doing a assignment inside parameter can cause issues expecially if theres 2 of em on same variable
[08:13:32] <Tom_L> those look like interrupt driven events to me
[08:13:49] <XXCoder> move bouncetime =3000 elsewhere and just put bouncetime where bouncetime=3000 is
[08:15:19] <Tom_L> could be
[08:16:09] <XXCoder> im kinda puzzled why it uses assignments in function call anyway
[08:16:35] <XXCoder> callback= can be outright removed since it dont get used elsewhere
[08:16:39] <Tom_L> the while is where it sits most of the time
[08:16:56] <Tom_L> JT-Shop, add a delay inside the while
[08:16:58] <XXCoder> note I might be wrong as i am not fimiliar with embedded
[08:17:07] <Tom_L> just to test it
[08:17:33] <XXCoder> in fact I think bouncetime=, both of em, can be removed leaving just 3000
[08:17:49] <Tom_L> or does the sleep(0.1) do that?
[08:18:25] <jthornton> yea sleep is a delay
[08:18:35] <XXCoder> I dont like "magic" numbers as its not clear what number is from. so bouncetime=3000 somewhere above where its used and just bouncetime in function call is fine
[08:19:02] <Tom_L> i don't like libraries because you don't know what they're doing inside
[08:19:18] <Tom_L> my biggest peeve about arduini.
[08:19:27] <Tom_L> o
[08:19:37] <XXCoder> honestly large part of coding is learning libraries
[08:26:26] <Tom_L> not much help here..
[08:32:05] -!- ikcalB has quit [Ping timeout: 268 seconds]
[08:32:34] <XXCoder> jthornton: still rewiring?
[08:32:55] <jthornton> filling bird feeders lol
[08:33:01] <XXCoder> cool :)
[08:33:09] <XXCoder> well off to bed before I die. laters
[08:33:15] <jthornton> I disconnected input 23 and cycled 22 and got multiple inputs lol
[08:33:19] <jthornton> see you later
[08:33:41] <XXCoder> try refactor code as I discrived, move all assignments out function calls
[08:33:50] <XXCoder> *discribed
[08:33:54] <XXCoder> night
[08:34:45] <Tom_L> jthornton, how does that work?
[08:34:54] <Tom_L> noise?
[08:35:11] <Tom_L> add external pull up/down
[08:35:33] * Tom_L recalls suggesting that a few days ago
[08:36:17] -!- ikcalB has joined #linuxcnc
[08:37:32] <jthornton> looks like the internal resistors are 50k and the suggested external is 10k IIRC so yea time to test that as well as adding a cap...
[08:38:41] <Tom_L> one at a time
[08:40:07] <jthornton> lol I got a rising edge detected on channel 22 just touching the switch terminal with my finger
[08:43:09] <Tom_L> woops
[08:43:26] <MarcelineVQ> XXCoder: it's not assignment, it's providing an argument to a named field. bouncetime=3000 is supplying the argument named bouncetime with the value 3000, this is part of python allowing default values for fields, and allowing you to selectively pick which ones to over-ride
[08:43:27] <Tom_L> alot of times the internal resistors are too weak, that's why i never trust them
[08:44:29] <MarcelineVQ> rather than initializing some variable bouncetime with the value of 3000 before passing it to the function
[08:44:45] <jthornton> yea after breakfast and starting a few fires I'll add a resistor to 22 and see what happens
[08:45:03] <MarcelineVQ> oh he left, darnit
[08:45:16] <jthornton> yea sleepy time for him
[08:45:24] <MarcelineVQ> this RPi.GPIO soruce code is hella weird
[08:45:35] <MarcelineVQ> might be normal for python-land tho ehe
[08:46:39] <jthornton> yea I'm just trying to add a timeout to my door function lol
[08:47:44] -!- veek has joined #linuxcnc
[08:48:56] <MarcelineVQ> Oh I mean the code for the library, no your code
[08:50:11] -!- ikcalB has quit [Ping timeout: 276 seconds]
[08:52:02] <MarcelineVQ> little surprised there's so few options for rpi gpio in python, not that that's really indicative of anything
[08:56:11] -!- ikcalB has joined #linuxcnc
[08:57:22] -!- fragalot has joined #linuxcnc
[08:57:42] <fragalot> hi
[08:57:50] <Tom_L> lo
[09:00:28] -!- JT-Shop has quit [Read error: Connection reset by peer]
[09:01:09] -!- JT-Shop has joined #linuxcnc
[09:01:19] -!- JT-Fshop- has quit [Read error: Connection reset by peer]
[09:02:39] -!- jthornton has quit [Read error: Connection reset by peer]
[09:03:04] -!- jthornton has joined #linuxcnc
[09:04:11] -!- JT-Fshop- has joined #linuxcnc
[09:10:36] <jthornton> MarcelineVQ: are there better/more options for gpio in C?
[09:13:18] -!- mozmck_lp has joined #linuxcnc
[09:13:54] <MarcelineVQ> I've really no idea, just that between the popularity of pi and python I expected to find more options in a quick googling. Could be one of those momentum things, where some few options dominate since they're most used.
[09:14:12] <MarcelineVQ> Or search engine shenanigans :>
[09:16:20] -!- Inline has quit [Quit: Leaving]
[09:20:31] <MarcelineVQ> Though it is noteworthy that python gpio are calls to c gpio anyway, python's job there is to make things slightly nicer to write and ideally safer
[09:20:55] <MarcelineVQ> e.g. https://github.com
[09:23:13] <MarcelineVQ> this is the c gpio from the other main python lib, which I think the above was based on https://gist.github.com
[09:24:37] <MarcelineVQ> Not particularly pleasant stuff to write for one's self
[09:26:13] -!- Inline has joined #linuxcnc
[09:34:53] <fragalot> writing GPIO stuff in C is fine
[09:35:24] <fragalot> it's when you start having to process random text files on a system with limited memory (read: a few K of ram) that it gets interesting :P
[10:03:37] <jthornton> know of any examples of doing gpio in C?
[10:06:40] -!- ikcalB has quit [Ping timeout: 244 seconds]
[10:08:50] <rmu> jthornton: how do you connect your switch to the raspberry pi
[10:09:06] -!- merkatorix has joined #linuxcnc
[10:09:21] <jthornton> 3.3v > switch > input
[10:09:21] -!- ziper has joined #linuxcnc
[10:10:17] <rmu> no resistors anywhere?
[10:13:52] <JT-Shop> just the internal pull down
[10:17:25] <rmu> bounce time of a switch is on the order of ms max, bounce times of 300ms and more seem completely ridiculous
[10:18:19] <rmu> seems the GPIO lib is buggy
[10:18:57] <rmu> JT-Shop: https://raspberrypi.stackexchange.com
[10:21:25] <rmu> what one could do is sample the input pin with fixed frequency, e.g. 10kHz, compare with previous state, if changed, increment a per-input counter, iff counter > 5, report edge, else set counter to 0
[10:28:19] <jthornton> rmu: interesting info about debounce
[10:32:49] -!- ikcalB has joined #linuxcnc
[10:38:33] <fragalot> rmu: smarter way to do it is to use the interrupt, disable it immediately after receiving, and only re-enabling the interrupt after a set filter time
[10:38:58] <fragalot> rmu: that way you don't have to "hold off" on the event like most generic counter based debouncers
[10:40:20] -!- mozmck_lp has quit [Quit: Leaving.]
[10:56:42] -!- veek has quit [Quit: Leaving]
[11:14:31] -!- ikcalB has quit [Ping timeout: 246 seconds]
[11:30:17] <rmu> fragalot: i beg to differ, external interrupt only makes sense if you need to wakeup a µC or some similar. IMHO.
[11:31:00] <fragalot> or if you have strict timing requirements
[11:31:06] <fragalot> which is where my suggestion comes in
[11:31:21] <rmu> with mechanical switches that bounce the timing can't be that strict
[11:31:47] <fragalot> true
[11:32:49] <rmu> consider using RC element and schmitt trigger for debouncing, even that will introduce a delay
[11:36:35] <rmu> if you have a double-throw switch and extra wire you could use a flipflop for debouncing
[11:37:44] -!- Akex_ has quit [Ping timeout: 272 seconds]
[11:40:45] -!- Akex_ has joined #linuxcnc
[11:53:17] <JT-Shop> rmu: a delay is not an issue... what I'm trying to actually do is to have a time out function for the door that fires once as the door transitions
[11:53:44] <JT-Shop> and if it times out to block the door motor until it is manually reset
[12:00:00] -!- swarfer has joined #linuxcnc
[12:00:19] <rmu> JT-Shop: ignore all that fancy GPIO edge stuff, just look at your inputs in a loop, sleep for a second in each iteration, and if input is same changed value for two iterations fire your edge-detected code
[12:01:06] <rmu> btw, using a raspberry pi for this is "mit kanonen auf spatzen schiessen"
[12:01:53] <rmu> adafruit trinket can also be programmed in python
[12:02:11] <rmu> costs <usd10
[12:02:40] <rmu> doesn need sd card
[12:02:46] <fragalot> I think he's already got the rpi set up anyhow
[12:03:57] <tiwake> JT-Shop: do you have a favorite manufacturer for servo drivers and VFD units?
[12:04:31] <tiwake> or anyone else really
[12:06:34] <JT-Shop> I have a GS2 from Automation Direct on the BP knee mill conversion...
[12:07:02] <JT-Shop> rmu: that's what I'm testing now
[12:07:54] <JT-Shop> rmu: yea there are lots of cheaper things I could have used including a drawer full of ardunios lol
[12:08:17] -!- ve7it has joined #linuxcnc
[12:08:17] <JT-Shop> however the RPi has a nice terminal block that just plugs into it...
[12:09:30] <rmu> back in the day, I participated in a university course on some magnetohydrodynamics simulation stuff with mathematica on a machine that was about as capable and fast as a raspberry pi 3
[12:10:13] * JT-Shop wonders why Monique is still in the nest box...
[12:10:20] <rmu> using such a machine to look at some external door switch many millions of times per second seems wasteful...
[12:10:44] -!- swarfer has quit [Quit: swarfer]
[12:11:38] <JT-Shop> well that is not all that it does... but yea it has much more power than needed but the terminal block just makes up for all that extra power
[12:18:56] -!- Akex_ has quit [Ping timeout: 244 seconds]
[12:19:05] -!- Akex_ has joined #linuxcnc
[12:22:46] <JT-Shop> but anyway when code should work but does not you start grasping at whatever you can find like the edge detect stuff
[12:31:42] -!- ikcalB has joined #linuxcnc
[13:11:07] -!- c-log has quit [Ping timeout: 244 seconds]
[13:13:11] -!- c-log has joined #linuxcnc
[13:13:11] -!- c-log has quit [Changing host]
[13:13:11] -!- c-log has joined #linuxcnc
[13:24:02] -!- rburkholder has joined #linuxcnc
[13:27:31] -!- Akex_ has quit [Ping timeout: 246 seconds]
[13:28:53] -!- Akex_ has joined #linuxcnc
[13:43:33] -!- Papagno has joined #linuxcnc
[13:50:08] -!- mozmck_lp has joined #linuxcnc
[13:59:27] -!- ikcalB has quit [Ping timeout: 240 seconds]
[14:05:04] -!- ikcalB has joined #linuxcnc
[14:11:42] -!- Ralith_ has quit [Ping timeout: 252 seconds]
[14:18:51] -!- bedah has quit [Ping timeout: 252 seconds]
[14:29:31] -!- mozmck_lp has quit [Quit: Leaving.]
[14:30:58] -!- bedah has joined #linuxcnc
[14:43:28] -!- ikcalB has quit [Ping timeout: 246 seconds]
[14:44:49] -!- Papagno has quit [Quit: Sto andando via]
[15:24:33] -!- Ross has quit [Quit: ZNC - http://znc.in]
[15:29:05] -!- Ross has joined #linuxcnc
[15:34:13] -!- fragalot has quit [Quit: Lost terminal]
[15:40:44] <Tom_L> JT-Shop, make any progress?
[15:41:26] <JT-Shop> yea, I think I got a working solution but it's a bit clunky
[15:41:41] <Tom_L> what was the issue?
[15:49:55] <JT-Shop> I never did figure out why the edge detect gave multiple hits, I tried something different
[15:50:13] <JT-Shop> I guess I need to try a RC circuit on the inputs just for fun
[15:54:18] -!- Akex_ has quit [Ping timeout: 252 seconds]
[15:54:58] -!- Akex_ has joined #linuxcnc
[15:57:36] -!- KimK has quit [Ping timeout: 252 seconds]
[16:10:26] -!- KimK has joined #linuxcnc
[16:36:29] -!- EntropyWizard has joined #linuxcnc
[16:41:19] -!- syyl has joined #linuxcnc
[16:52:03] -!- Akex_ has quit [Ping timeout: 250 seconds]
[16:52:44] -!- Akex_ has joined #linuxcnc
[17:07:51] <Deejay> gn8
[17:08:42] -!- Deejay has quit [Quit: bye]
[17:13:46] <XXCoder> MarcelineVQ: interesting! never seen that style argument.
[17:14:03] <XXCoder> JT-Shop: looks like I was mistaken. sorry. :)
[17:15:55] <XXCoder> I got idea for debugging, graph the input from your switches
[17:16:03] <XXCoder> see what the heck is going on
[17:16:30] -!- veek has joined #linuxcnc
[17:34:21] -!- CubicleNate has quit [Ping timeout: 260 seconds]
[17:45:59] -!- syyl has quit [Quit: Leaving]
[17:57:47] <jthornton> hey XXCoder
[17:58:44] <XXCoder> hey
[17:58:51] <jthornton> this actually works if not very elegant lol https://paste.ubuntu.com
[17:59:11] <jthornton> going to put a RC circuit on an input and test some more
[17:59:24] <XXCoder> man python coding style is quite different than I expected. I mean, override object internal variables huh
[18:00:06] <XXCoder> yeah that would feed bouncing feed to light but light dont really care
[18:00:07] <Tom_L> the original was interrupt driven wasn't it?
[18:00:13] <Tom_L> this doesn't appear to be
[18:00:31] <Tom_L> just inline code
[18:00:35] <XXCoder> Tom_L: yep INT driven but it dont seem to work well
[18:01:52] <jthornton> yea the one I was messing with before was interrupt driven
[18:02:20] <jthornton> https://paste.ubuntu.com
[18:02:32] <Tom_L> was the interrupt being cleared after each hit? sometimes they must be manually reset. i dunno if that is the case here
[18:03:21] <Tom_L> generally reading or servicing an interrupt will clear it
[18:03:24] <jthornton> hmm no, but if I remove the event detection then next time what happens
[18:03:42] <XXCoder> clear the event not listener
[18:03:49] <jthornton> I think the switch bounce is just the issue
[18:04:18] <Tom_L> that is very typically 20-30ms
[18:04:32] <XXCoder> he went all way to 10 seconds. still double events
[18:04:46] <XXCoder> https://www.raspberrypi.org
[18:04:54] <rmu> "while (not sleep..." looks strange...
[18:05:22] <XXCoder> rmu: its legal. sleep returns 0 an not makes it not 1 ie: same as true
[18:05:29] <XXCoder> err not 0
[18:05:33] <rmu> None
[18:06:40] <XXCoder> jthornton: try this, set sleep(.1) to sleep(.5) see if its race condition between event clearing and checking
[18:06:42] <rmu> sleep doesn't return anything, so in some sense, this is an endless loop in disguise
[18:06:51] <XXCoder> it is endless loop
[18:07:36] <_unreal_> GOT to love trying to log into linux well unlock the display and enter your password only to have it reject it multi times
[18:07:39] <_unreal_> I'm like WTF
[18:07:39] <rmu> i would write endless loop as "while True:" to make it explicit
[18:08:00] <_unreal_> after trying 20+ times. I said ding ding ding.... wonder if caps locks are on :)
[18:08:40] <XXCoder> _unreal_: windows displays "your caps is on" type text when it is on in login screen
[18:08:54] <_unreal_> I'm talking about linux
[18:09:11] <XXCoder> I know
[18:09:25] <XXCoder> I was talking about windows as comprasion
[18:09:46] <_unreal_> I dont have any windows machines (currently)
[18:09:51] <XXCoder> me either
[18:09:56] <rmu> it depends on the display managers greeter, some display a warning if CAPS is on
[18:11:26] <_unreal_> could be
[18:12:44] <XXCoder> it does
[18:12:56] <_unreal_> well monday is going to be fun.... I screamed at my boss today because he abandomed me in a danger work environment.
[18:12:59] <XXCoder> mine don't I think but same time havent tried to login with caps on yet :)
[18:13:13] <_unreal_> today not monday
[18:13:15] <_unreal_> err
[18:13:36] <_unreal_> lot on my mind and my daughter is yacking at me
[18:14:07] <XXCoder> login displays ! speak ballon in password vox when caps is on. cool
[18:14:53] <_unreal_> 3 guys at the sight including me. and nether checked on me in the hole for about 40 min.... high chemical level deadly if you dont have protective gear on....
[18:15:09] -!- veek has quit [Quit: Leaving]
[18:15:11] <jthornton> XXCoder: we found something this afternoon that the bouncetime is not really a debounce
[18:16:06] <XXCoder> really? what does it do?
[18:17:07] <XXCoder> _unreal_: ouch. worksite safety can be important. if I work overtime there must be in least 2 hearing guys and I cant be counted as one for workspace safety
[18:17:10] <jthornton> something rmu found https://raspberrypi.stackexchange.com
[18:18:19] <rmu> google found it ;)
[18:18:19] <XXCoder> man
[18:18:32] <XXCoder> its like anti-debounce
[18:19:33] <XXCoder> so I guess you need to code in debounce yourself
[18:20:00] <XXCoder> make sure value holds true for whatever time you want then its definitely real press. bleh
[18:22:15] <jthornton> it's a toggle switch so it's a real transition but bumpy I guess... I need to find a page that explains how to calculate the rc circuit
[18:22:50] <_unreal_> and I ran out of supplys (my ran out of resin, I needed a new roller, I needed new carbon cartridges), I kept screaming for help. kicking teh walls etc..... when my gloves disintegrated and ripped open I blew up.... I climbed out screamed... NOTHING yet I could hear them talking....
[18:22:55] <_unreal_> going to be interesting
[18:23:04] <jthornton> right now the manual switch works fine but I want to add a time out to the travel time
[18:23:28] <Tom_L> http://protological.com
[18:23:35] <_unreal_> jthornton, are you going to do a hardware debounc? or a software debounce?
[18:23:58] <Tom_L> https://www.avrfreaks.net
[18:24:01] <jthornton> I'm going to try a hardware debounce
[18:24:12] <XXCoder> jthornton: wonder if you can use keyboard buttons. they use various stragies to make bouncing very short
[18:25:16] <_unreal_> a software debounce would likely be less work.... I mean its looking for an interrupt trigger...... then have a wait state, then check if the state is still the same about 7-10ms later..
[18:25:18] <jthornton> the RPi has no keyboard, mouse or monitor but I'm thinking of using a touch screen
[18:25:31] <XXCoder> cherry red MX for example
[18:25:46] <XXCoder> its designed to avoid bouncing
[18:25:58] <Tom_L> here's a whole essay on switch debounce: :) https://pubweb.eng.utah.edu
[18:26:25] <_unreal_> jthornton, if you do a hardware debounce.. your going to need a shcimitt trigger and a few other parts...
[18:26:57] <jthornton> I have a AR trigger will that work?
[18:27:12] <_unreal_> ar trigger?
[18:27:14] <Tom_L> AR?
[18:27:20] <_unreal_> lol
[18:27:36] <XXCoder> hopefully not followed by -15
[18:27:51] <jthornton> of course it is
[18:28:03] <XXCoder> jthornton: https://massdrop-s3.imgix.net
[18:28:31] <XXCoder> mechanical keyboard switches, they are designed to avoid bouncing
[18:31:15] <XXCoder> spring type https://www.leafandcore.com
[18:32:28] <Tom_L> just use a switch with a spring so tight it can't bounce
[18:32:36] <Tom_L> dammit!
[18:32:47] <_unreal_> XXCoder, all keyboards bounch.. any machanical contact bounces period...
[18:33:15] <_unreal_> does not matter WHAT it is or what type it is.... could be carbon contacts.. or hard metal etc...... they call produce a bounch...
[18:33:23] <XXCoder> _unreal_: of course. cant beat ohysics. what I mean is make it short as possible
[18:33:47] <XXCoder> some of examples I posted the bouncing period is very short
[18:34:09] <_unreal_> if there is a bounce period... its still bouncing
[18:34:24] <XXCoder> isnt that what I said?
[18:34:28] <_unreal_> :) yep
[18:35:30] <XXCoder> jthornton: ... simpler than I thought https://hackadaycom.files.wordpress.com
[18:35:47] <XXCoder> I dont pretend to completely understand that but interesting
[18:36:28] <XXCoder> https://hackaday.com
[18:37:03] <jthornton> yea that's a rc debounce circuit with a pull up, I don't understand it either lol
[18:37:24] <Tom_L> did you look at that 'essay' site?
[18:37:32] <XXCoder> reading it current;ly
[18:37:38] <Tom_L> it shows several different debounce circuits
[18:37:53] <jthornton> I missed that one... looking now
[18:37:54] <Tom_L> RC is probably most effective with a gate on the end of it
[18:38:16] <pcw_home> cant you just poll at a decent rate and do software debounce?
[18:38:28] <Tom_L> i probably would
[18:38:31] <Tom_L> https://pubweb.eng.utah.edu
[18:39:09] <XXCoder> lowpass interesting
[18:39:52] <XXCoder> r*c I’m using the microcontroller’s internal pullup resistor, which is around 10K ohms, and a 100nF capacitor, which gives 0.001 seconds — a microsecond — on the release.
[18:40:10] <XXCoder> jthornton: yep should read that. very interesying website. Tom_L's link probably is even better
[18:40:27] <jthornton> pcw_home: this is the code that is running now (the auto part) but I want to add a timeout feature https://paste.ubuntu.com
[18:40:50] -!- Ralith_ has joined #linuxcnc
[18:40:58] <jthornton> to drop out the relay if it takes too long ie the door is stuck or a chicken ate the cord etc
[18:42:18] <jthornton> https://paste.ubuntu.com this also shows the manual part
[18:43:46] <XXCoder> does python has Timer object?
[18:44:08] <XXCoder> if so maybe can just create timer on override if its running too long and interrupt
[18:44:35] <jthornton> so I cobbled this up this afternoon and it seems to work https://paste.ubuntu.com
[18:45:05] <jthornton> the problem is starting something only once like a timer and not restarting it every bounce
[18:45:19] <pcw_home> yeah I would just poll the inputs every time around the loop and wait for N high or N low on the GP inputs before passing the GP in data up the chain
[18:45:31] <XXCoder> jt that code is near bulletproof
[18:45:50] <Tom_L> yeah but interrupts is where the cool kids play
[18:45:57] <jthornton> lol
[18:46:00] <XXCoder> lol
[18:46:14] <pcw_home> Interrupts are silly here
[18:46:20] <jthornton> XXCoder: the last one?
[18:46:26] <Tom_L> pretty much since that's about all it's doing
[18:46:26] <XXCoder> yeah short code on lights
[18:47:16] <Tom_L> jthornton, when you finish... program a stop light cenario
[18:47:23] <jthornton> I think it's faulty because I don't reset lt if I turn the switch off too soon
[18:47:37] <pcw_home> that is you conditionals should only look at filtered GPIO input data
[18:47:46] <pcw_home> s/you/your/
[18:47:50] <XXCoder> it has no handling for < 10 and turned to untrue
[18:48:24] <jthornton> what do you mean by filtered GPIO input?
[18:48:43] <XXCoder> such tiny window of unhandled range (0, 10)
[18:49:04] <jthornton> XXCoder: yea need to drop off the last lt > 10 as I assume by then the switch has settled down
[18:49:32] <pcw_home> all gpio input go through a filter (count up when high count down when low) and only the filtered data is ever looked at
[18:49:41] <XXCoder> maybe make it if untrue then... if > 10 then... else ....
[18:49:51] <XXCoder> so you dont spam yourself with da lights is off!"
[18:50:01] <jthornton> pcw_home: I see now
[18:50:11] <pcw_home> since you have a .1 second loop count of 3 should be fine
[18:50:33] <pcw_home> count needs to be bounded at 0 and 3
[18:50:43] <pcw_home> when 3 report high
[18:50:53] <pcw_home> when 0 report low
[18:50:59] <XXCoder> pcw_home: I was wondering if could make it .05 second and keep lt values where it is so its now half second debounce?
[18:51:09] <pcw_home> this also provide noise rejection
[18:51:23] <XXCoder> or .01 second, making it 1/10 second debounce?
[18:51:35] <XXCoder> but that would be very fast loop so ehh
[18:51:50] <pcw_home> basically you should never make a decision on unfiltered data
[18:51:51] <jthornton> pcw_home: thanks I'll give that a whirl in the morning... chow time now some Texas Red is calling my name
[18:53:01] <XXCoder> hey hey hey you must send me some now. remember the chat rules, if menon food you must send some to XXCoder !
[18:53:04] <XXCoder> ;)
[18:53:34] <jthornton> I'll email you a bowl but you have to provide the onions...
[18:53:38] <XXCoder> lol
[18:54:03] <jthornton> http://gnipsel.com
[18:54:26] <XXCoder> looks decent
[18:54:28] <jthornton> thanks for all the help guys
[18:54:41] <XXCoder> dunno if I actually helped lol but no problem
[18:58:00] -!- merkatorix has quit [Ping timeout: 268 seconds]
[19:25:14] -!- CubicleNate has joined #linuxcnc
[19:27:39] -!- badboynoob1998 has joined #linuxcnc
[19:34:44] -!- matthew1 has joined #linuxcnc
[19:39:31] -!- badboynoob1998 has parted #linuxcnc
[19:39:42] -!- badboynoob1998 has joined #linuxcnc
[19:40:38] -!- badboynoob1998 has parted #linuxcnc
[19:41:09] -!- homeroot has parted #linuxcnc
[19:42:01] -!- evilroot has joined #linuxcnc
[19:44:18] -!- ChunkyPuffs has quit [Remote host closed the connection]
[19:45:12] -!- ChunkyPuffs has joined #linuxcnc
[19:45:20] <ChunkyPuffs> hey, back
[19:45:34] <ChunkyPuffs> I have a garbage CNC machine called "CNCEST"
[19:46:04] <ChunkyPuffs> I have no idea how to use it, have some weirdo controller with it, ultimately hoping to use LinuxCNC, just don't want to bother doing all of that tonight, including rewiring everything. Anybody feeling in the mood to help me get something going?
[19:46:57] <ChunkyPuffs> The thing is plugged in, it shows up as "RNR ECO MOTION 2.0
[19:46:57] <ChunkyPuffs> "
[19:49:42] -!- CubicleNate has quit [Ping timeout: 252 seconds]
[19:53:31] <ChunkyPuffs> So it sees the device in linux, say I have some gcode, how do I send it over usb?
[20:11:27] -!- Markus_P has quit [Ping timeout: 240 seconds]
[20:11:43] <pcw_home> if it emulates a serial device it might be easy to just copy the file to it
[20:11:49] -!- Markus_P has joined #linuxcnc
[20:14:32] -!- CubicleNate has joined #linuxcnc
[20:17:03] <ChunkyPuffs> pcw_home, so I can literally just echo something to it via a shell?
[20:17:07] <ChunkyPuffs> like echo /dev/itsname?
[20:26:10] <XXCoder> cnest yeech
[20:27:42] <gregcnc> looks like something made to run with Mach3
[20:28:06] <ChunkyPuffs> It shows up in linux, but does nothing with universal gcode sender
[20:28:10] <ChunkyPuffs> it's not operating as a serial port
[20:28:38] <XXCoder> looks like standard chiniseium cnc
[20:31:32] <XXCoder> gregcnc: search bears it out. mach3
[20:31:47] <XXCoder> and if can mach3, can do linuxcnc
[20:31:58] <gregcnc> via usb?
[20:32:37] <ChunkyPuffs> Well how do I go about doing stuff on it with linuxcnc?
[20:32:55] <gregcnc> linuxcnc can certainly run the machine, but it wil need parport or mesa, etc.
[20:32:58] <ChunkyPuffs> as it stands, I tried to cat circle.gcode > /dev/usb/hiddev0 2>&1
[20:33:00] <XXCoder> no serial connection?
[20:33:03] <ChunkyPuffs> Which probably means nothing, I'm a noob lol
[20:33:15] <ChunkyPuffs> it doesn't look like it, but where would I look if it was serial?
[20:33:24] <ChunkyPuffs> it shows up at /dev/usb hiddev0
[20:33:49] <XXCoder> hid device 0 interesting
[20:34:06] <ChunkyPuffs> in which case, not much I can do?
[20:34:46] <ChunkyPuffs> the actual connector on the machine is control output, and it looks like serial to me
[20:34:56] <ChunkyPuffs> I think it's a printer cable though? Dunno, a bit beyond my age lol
[20:35:42] <XXCoder> printer is approx 1.5 inch wide port
[20:35:59] <ChunkyPuffs> it is that massive hence why I'm saying it
[20:36:09] <XXCoder> it is then
[20:36:25] <ChunkyPuffs> so what's the steps in figuring out what to do here
[20:36:34] <ChunkyPuffs> it's garbage, its on ebay for like £90
[20:36:36] <XXCoder> figure what pins do what
[20:36:53] <XXCoder> is there pins list anywhere?
[20:38:30] <ChunkyPuffs> https://www.ebay.com
[20:38:44] <ChunkyPuffs> pins list?
[20:38:47] <ChunkyPuffs> There's a weird board inside
[20:38:59] <ChunkyPuffs> https://imgur.com
[20:40:04] <XXCoder> looks like BOB and 3 drivers. nice case
[20:40:44] <XXCoder> on BOB (first picture) there should be pin nunbers by axis letters
[20:41:08] <XXCoder> maybe its on other side
[20:41:55] <XXCoder> 3 outputs and 4 inputs not bad
[20:43:24] <evilroot> Yeah, that's a USB BOB
[20:43:46] <XXCoder> evilroot: rear of box clearly shows parallel port
[20:44:34] <evilroot> Nope, that's the connector to the motors
[20:44:48] <evilroot> Might use a DB-25 connector but its not parallel heh
[20:44:54] <XXCoder> apparently. bleh
[20:45:04] <XXCoder> can still find out pins by breakout board though
[20:45:13] <gregcnc> centronics
[20:45:18] <XXCoder> just not sure if its printed on mount surface or other side
[20:45:39] <evilroot> Each of those three driver boards take step/dir signals from the BOB
[20:46:12] <XXCoder> yep
[20:46:21] <evilroot> The BOB itself fully optically isolated, which is rare on cheaper ones
[20:46:32] <ChunkyPuffs> Yeah I was told by a few people on maker night here at the space I'm at that the port is responsible for power and not any data
[20:46:38] <XXCoder> thats why need to look at BOB for pin number prints if any, so he can configure linuxcnc
[20:46:40] <ChunkyPuffs> that printer cable looking thing is purely re-used as power
[20:47:10] <ChunkyPuffs> what's BOB?
[20:47:16] <XXCoder> i hate it when standard port is used in nonstandard way
[20:47:20] <XXCoder> breakout board
[20:47:32] <XXCoder> first picture in your list
[20:47:34] <evilroot> Um, you can't use LinuxCNC with a USB breakout
[20:47:45] <gregcnc> there are apparently many verions of that particualr board
[20:47:46] <evilroot> That USB breakout likely accepts gCode
[20:47:57] <XXCoder> yeah I was thinking that that port was parallel, guess not
[20:48:16] <XXCoder> but when he tried to feed it gcode nothing happened?
[20:48:22] <evilroot> If you want to use LinuxCNC with that system simply replace that BOB with a parallel one
[20:48:44] <XXCoder> yeah BOB with parallel is quite cheap
[20:48:45] <evilroot> XXCoder: likely incorrect communication protocol
[20:48:56] <XXCoder> lemme find one
[20:49:03] <evilroot> baud rate, etc
[20:49:14] <evilroot> Or it may even ned a specific program
[20:49:23] <XXCoder> https://www.aliexpress.com
[20:49:26] <evilroot> Easiest way is like I said just swap out to a parallel one
[20:49:39] <XXCoder> I didnt even try look for best deal, just first example
[20:49:59] <XXCoder> looks quite a good deal tho, comes with serial, and has port for pendant
[20:50:01] <XXCoder> nice
[20:50:03] <evilroot> Lukily everything is really straightforward, just move over step/dir and input/output, easy peasy
[20:51:13] <XXCoder> *parallel printer cable
[20:51:42] <ChunkyPuffs> I'm interested in linuxcnc, but not tonight
[20:51:56] <ChunkyPuffs> is there any way I can send some test gcode to the machine tonight
[20:51:59] <XXCoder> your machine might require some strange software like evil said
[20:52:06] <ChunkyPuffs> so mach3 on windows?
[20:52:20] <ChunkyPuffs> is there a legally obtainable free piece of software like mach3 I can get?
[20:52:24] <XXCoder> nope
[20:52:29] <ChunkyPuffs> Or alternatively, mach3, for free, legally.
[20:52:29] <XXCoder> ll free mach3 is cracked.
[20:52:34] <ChunkyPuffs> llo
[20:52:51] <XXCoder> thats why they are still pushing mach3 and not 4 or newer
[20:53:06] <XXCoder> mach4 is unbreakable apparently or they'd all push it by now
[20:55:02] <ChunkyPuffs> got a hash?
[20:55:49] <XXCoder> yours kinda looks like this https://www.aliexpress.com
[20:56:13] <XXCoder> yep even has that eagle
[20:56:27] <ChunkyPuffs> so this board sucks?
[20:56:33] <ChunkyPuffs> got a torrent hash for mach3?
[20:56:42] <ChunkyPuffs> one that isn't labelled "Artsoft Mach3 + CrAcK"
[20:56:55] <XXCoder> it might be workable but eh i'd just buy board and bin that bob
[20:57:25] <ChunkyPuffs> that thing from aliexpress that you posted earlier will run linuxcnc?
[20:57:30] <XXCoder> youre supposed to get cd with your machine apparently but I guess not
[20:57:36] <ChunkyPuffs> what do I need to do to get this to run cnc? an arduino or something?
[20:57:41] <XXCoder> not sure if its best deal
[20:57:59] <XXCoder> nah you just need a pc you can install linuxcnc os on
[20:58:13] <XXCoder> and it should do decent on latency test
[20:58:46] <ChunkyPuffs> wait so I can run linuxcnc on this BOB then?
[20:58:58] <XXCoder> yours nah its usb type
[20:59:15] <XXCoder> I thought it had parallel port thats why I said it could earlier
[20:59:21] <XXCoder> but nope
[20:59:22] <ChunkyPuffs> oh, so grab that bob from your aliexpress link and then replace what's in my current controller?
[20:59:41] <XXCoder> hey evilroot what ya think of one I linked?
[21:00:16] <XXCoder> it dont look like it has optical couplers
[21:00:27] <XXCoder> might be wrong tho
[21:00:56] <XXCoder> this one does. https://ae01.alicdn.com
[21:02:01] <ChunkyPuffs> will be back later
[21:02:07] <ChunkyPuffs> going into windows to test mach3
[21:02:16] <XXCoder> ChunkyPuffs: https://www.aliexpress.com
[21:02:47] <XXCoder> see picture of bottom? it has pin numbers
[21:03:00] <XXCoder> ydor is p8 for example
[21:03:04] <XXCoder> *ydir
[21:03:25] <ChunkyPuffs> yeah, what about it?
[21:03:28] <ChunkyPuffs> You want to see the back of my board?
[21:03:40] <XXCoder> thats info you use to find pins on that specific breakout board.
[21:03:41] <XXCoder> nah
[21:03:47] <ChunkyPuffs> I have an irc session on another machine, is it freenode blocking me from talking on it?
[21:03:48] <XXCoder> yours simply wont work for linuxcnc
[21:03:56] <ChunkyPuffs> says I can't get voice in channel, I'm registered and everything
[21:04:00] <ChunkyPuffs> ah ok
[21:04:16] <XXCoder> npot too sure but think you can register in only one location at time? dunno
[21:05:37] <ChunkyPuffs> https://forum.linuxcnc.org
[21:05:42] <ChunkyPuffs> how do I get my mach3 kitted out like this?
[21:05:52] <XXCoder> no idea, I use linuxcnc
[21:05:56] <ChunkyPuffs> lmfao
[21:05:59] <ChunkyPuffs> this is from the linuxcnc forums
[21:06:11] <XXCoder> honestly?
[21:06:18] <XXCoder> I want working ui, not eye candy
[21:07:44] -!- CubicleNate has quit [Ping timeout: 272 seconds]
[21:08:35] -!- ChunkyPuffs has quit [Remote host closed the connection]
[21:09:33] -!- matthew1 has quit [Quit: WeeChat 2.3]
[21:10:35] <Tom_L> hazzy-m and a few others are working on a similar gui
[21:10:44] <XXCoder> saw ad, interesting https://www.gearbest.com
[21:10:53] <XXCoder> Tom_L: really? full of glossy stuff?
[21:11:24] <Tom_L> not quite that busy but similar
[21:12:21] <evilroot> XXCoder: honestly couplers aren't a huge deal unless you really fuck up
[21:14:04] <XXCoder> lol ok
[21:14:45] <Tom_L> https://github.com
[21:14:48] <Tom_L> that might be it
[21:14:51] <Tom_L> not sure
[21:15:20] <Tom_L> i think there's a more recent one
[21:15:25] <evilroot> I've been messing with parallel controlled stuff for over 2 decades and never once have I fried one
[21:15:40] <evilroot> Despite seriously iffy practices when I was younger lol
[21:16:34] <cnomad> is it typical for a stepper motor stage (w/ optical limit sensors) to have the limit sensors pulse when they are actuated?
[21:16:54] <cnomad> It seems like this stage sources some current from the motor leads
[21:17:38] <cnomad> in addition to sourcing it from a wire called "+VS", which is constant 7v
[21:17:57] <cnomad> i can just add a cap in parallel and call it a day, but I'm curious
[21:18:00] -!- matthew1 has joined #linuxcnc
[21:18:20] <Tom_L> https://forum.linuxcnc.org
[21:18:22] <Tom_L> XXCoder
[21:18:23] <cnomad> i also passed the official controller through a scope and noticed there was some pulsing, albeit less of it
[21:18:26] <Tom_L> but there's another one
[21:18:52] <XXCoder> looking
[21:19:05] <XXCoder> I like the look
[21:19:20] <Tom_L> you'd like the newer one better
[21:19:37] <Tom_L> but i can't seem to find it
[21:19:43] <hazzy-m> Tom_L: Here is a thread with screen captures of Lcvettse's screen made with QtPyVCP: https://forum.linuxcnc.org
[21:20:02] <Tom_L> i figured you'd wake up sooner or later :D
[21:20:10] <XXCoder> that ones not bad also
[21:20:12] <hazzy-m> the QtPyVCP repo is here: https://github.com
[21:20:25] <Tom_L> is it working yet?
[21:20:51] <XXCoder> slightly concerned with slightly gray text on gray background
[21:21:02] <hazzy-m> Almost, it can run a machine, but still a lot left to do
[21:21:12] <XXCoder> gcode background should be white
[21:21:25] <hazzy-m> XXCoder: I think he improved some of the colors letter on, look near the end of the thread
[21:21:34] <XXCoder> bcause numbers nad comments is little bit hard for me to read
[21:21:50] <hazzy-m> XXCoder: I agree on the gcode background for sure
[21:22:14] <XXCoder> https://forum.linuxcnc.org
[21:22:55] <Tom_L> that doesn't bother me
[21:22:59] <hazzy-m> Right now we are more focused on the back end coding than the looks, so there will be a lot of refinements latter on
[21:23:19] <hazzy-m> I think the gcode font should be bigger too
[21:23:22] <XXCoder> Tom_L: it does for me, I dont want even slightest problem reading
[21:23:47] <evilroot> cnomad: is everything on the same ground?
[21:24:15] <cnomad> yep
[21:24:28] <cnomad> scope, voltage source, and driver
[21:24:33] <evilroot> Oh, just saw it sources current from the motor leads
[21:24:41] <Tom_L> better than lime green on black
[21:24:42] <cnomad> (maybe it does)
[21:24:44] <evilroot> That's why
[21:24:47] <cnomad> ok
[21:24:49] <cnomad> yeah
[21:24:53] <evilroot> Just throw in a cap
[21:25:19] <cnomad> so pulsing is normal in this case because current is pulsed to keep the magnetic field around the windings up -- is that correct?
[21:25:45] <cnomad> (what's odd is that I also need to connect this mystery +VS wire to see something from the limit sensors)
[21:25:47] <evilroot> If that doesn't solve it use a Schmitt trigger
[21:26:13] <cnomad> sounds good. hoping I didn't do anything else weird that may be causing this
[21:26:27] <evilroot> cnomad: motors generate noise on their leads, that's why they usually usually have caps as a buffer
[21:26:37] <cnomad> gotcha
[21:26:54] <evilroot> Of course you have to attach VS to the sensor
[21:27:05] <cnomad> oh, is VS a reference voltage?
[21:27:14] <cnomad> I have not idea what VS means in this context
[21:27:14] <evilroot> No, it powers the light
[21:27:17] <cnomad> oh
[21:27:19] <cnomad> OHP
[21:27:23] <evilroot> hahaha
[21:27:42] <cnomad> So the current from the motor leads, that's for the detector part of the optical sensor?
[21:27:54] <cnomad> and the emmitter is sources from VS?
[21:27:59] <cnomad> *sourced
[21:28:07] <evilroot> Yeah
[21:28:11] <cnomad> THANK YOU
[21:28:32] <cnomad> now I have a better understanding of what the hell I'm doing
[21:28:48] <evilroot> It's a light on one side (usually an IR LED) and a detector on the other side that's basically a light-activated switch
[21:29:15] <evilroot> Something gets between the two and the dectector no longer has light shining on it
[21:30:13] * hazzy-m goes to build the first fire of the fall
[21:30:37] <evilroot> Since usually they give a sort of sloping analog signal there's usually something involved that changes it to a more pure digital 1 or 0
[21:31:18] <cnomad> right, like a transistor or something else with forward bias?
[21:31:28] <evilroot> At the simplest
[21:31:51] <evilroot> Hence why optical limit switches have a PCB
[21:32:39] <cnomad> Gotcha. I haven't gotten to the point of taking this stage apart yet (that may have answered my question, too)
[21:32:43] <evilroot> The really nice ones have what's called a Schmitt trigger
[21:32:52] <cnomad> but this helps and probably resolves the issue
[21:32:56] <cnomad> gotcha
[21:33:27] <evilroot> Which actively removes noise and bounce via hystersis and feedback
[21:33:58] <evilroot> Also nearly all optical limit switches run of 5v
[21:34:43] <evilroot> And their power needs to be very consistent and well regulated, or you'll get weird results
[21:35:28] <cnomad> that seems to match with what I'm seeing on the datasheet. what's interesting is that the official controller also pulses VS +/- 1 volt, but that may be a side effect as to how it was designed
[21:35:51] <cnomad> i really appreciate the information
[21:37:17] <cnomad> I'm using the same 5v power supply as my BBB, and the scope indicates it's fairly stable. Anyway, going to add a cap, will report back. Thanks again!
[21:52:26] <cnomad> out of curiosity, why wouldn't they source current for both the emitter and detector sides of the light sensor from +VS?
[21:53:02] <cnomad> instead of having the detector on the motor winding circuit?
[22:08:57] -!- CubicleNate has joined #linuxcnc
[22:23:20] <evilroot> because they're stupid?
[22:24:47] <evilroot> The entire thing should be on a separate well regulated supply with shared ground, but people cut corners
[22:32:20] <cnomad> hm. not the first time i'd encounter a stupid design but, *shrug*
[22:32:38] <cnomad> right there's only one GND, not isolated
[22:32:49] <evilroot> http://www.rocketlabusa.com
[22:33:06] <evilroot> This will be a neat one to watch
[22:36:52] -!- justanotheruser has quit [Ping timeout: 244 seconds]
[22:38:16] -!- justanotheruser has joined #linuxcnc
[22:50:54] -!- matthew1 has quit [Quit: WeeChat 2.3]
[22:54:50] -!- makermatt has joined #linuxcnc
[22:55:12] -!- makermatt has quit [Client Quit]
[22:55:34] -!- justanotheruser has quit [Ping timeout: 246 seconds]
[22:55:55] -!- matthew_ has joined #linuxcnc
[22:56:06] matthew_ is now known as Guest5300
[22:57:29] -!- Guest5300 has quit [Client Quit]
[23:00:30] -!- makermatt has joined #linuxcnc
[23:02:09] -!- makermatt has parted #linuxcnc
[23:04:50] -!- justanotheruser has joined #linuxcnc
[23:11:15] -!- justanotheruser has quit [Ping timeout: 252 seconds]
[23:17:59] -!- ChunkyPuffs has joined #linuxcnc
[23:18:02] <ChunkyPuffs> test
[23:18:14] <ChunkyPuffs> Thank God for that, stuck in account creation hell, anyway I'm back
[23:18:22] <ChunkyPuffs> the garbage cnc isn't any further along, I'm afraid
[23:18:34] <XXCoder> welcome back
[23:23:19] <ChunkyPuffs> So, I have to manually get a datasheet or something for mach3 in order to tell to tell it what pins are which?
[23:23:21] <ChunkyPuffs> I'm fucked!
[23:23:39] <XXCoder> yeah it should have had a cdrom with it already configured
[23:23:47] <ChunkyPuffs> I can't figure that out lmfao, there's so many pins. I already manually deciphered one guy's video of his mach3 setup on youtube for a similar cnc
[23:23:49] <XXCoder> tend to have files to read docs and such
[23:24:03] <ChunkyPuffs> Well this is what some guy gave me, no data sheet, no cd, no nothin'
[23:24:10] <ChunkyPuffs> didn't come with one I'm told
[23:24:18] <XXCoder> probably couldnt figure it
[23:24:40] <XXCoder> solution really isnt complex, buy a parallel BOB and configure linuxcnc since that BOB actually has pins in it
[23:24:47] <XXCoder> (pin number printed on it)
[23:27:10] <ChunkyPuffs> So you don't reckon there's much that can be done right now?
[23:27:36] -!- justanotheruser has joined #linuxcnc
[23:29:10] <XXCoder> it doesnt list anything so reverse engineering would take quite a while
[23:32:40] -!- justanotheruser has quit [Ping timeout: 244 seconds]
[23:34:39] <Lcvette> yo
[23:34:52] <Lcvette> whats cookin in here?
[23:35:07] <XXCoder> you
[23:35:15] <Lcvette> lol
[23:35:18] <XXCoder> this is pretty cool https://www.aliexpress.com
[23:35:28] <XXCoder> combine it with VR and you got nice POV setup
[23:35:43] <Lcvette> uh oh.... 11.11 mania in here huh?
[23:35:48] <XXCoder> not really
[23:36:26] <XXCoder> been thinking aboyut something fun with servos but dunno
[23:36:35] <Lcvette> yeah?
[23:36:45] <Lcvette> im researching some servo stuff myself at the moment
[23:36:53] <Lcvette> i just bought a little machine for a retrofit
[23:36:53] <XXCoder> to what endpoint?
[23:37:08] <Lcvette> bought a little rhino ST8 cnc slant lathe
[23:37:34] <XXCoder> https://blog.hurco.com ?
[23:37:52] * Lcvette uploaded an image: big_6_20080610092440_16970.jpg (16KB) < https://matrix.org >
[23:38:00] * Lcvette uploaded an image: chuck and ways.jpg (211KB) < https://matrix.org >
[23:38:07] <Lcvette> smaller
[23:38:08] <Lcvette> lol
[23:38:10] <XXCoder> oh definitely not "little"
[23:38:13] <Lcvette> but with a turret
[23:38:38] <Lcvette> its pretty small
[23:39:00] * Lcvette uploaded an image: s-l500.jpg (39KB) < https://matrix.org >
[23:39:19] <XXCoder> interesting
[23:39:33] <XXCoder> I look forward to lathe rook in future lol
[23:39:45] <Lcvette> hahaha
[23:39:49] <XXCoder> converting it to linuxcnc or?
[23:39:55] <Lcvette> yeah
[23:40:04] <Lcvette> came with fanuc 6T controller
[23:40:17] <Lcvette> servos
[23:40:18] <XXCoder> if I had room I'd want to grab old fadal from work and change it to linuxcnc
[23:40:25] <Lcvette> pretty old school stuff though
[23:40:57] <Lcvette> need to see if the servos are worth keeping around and using new drives or swapping out all together
[23:41:17] <Lcvette> machine hasn't even gotten here yet, won't arrive until Tuesday
[23:44:32] <XXCoder> lol https://www.aliexpress.com
[23:46:42] <XXCoder> is it 3 phase, do you have source if it is?
[23:47:42] <Lcvette> not 100% sure
[23:47:51] <Lcvette> i have 220v 3ph here
[23:47:58] <Lcvette> 25hp RPC
[23:48:05] <Lcvette> runs my big machine
[23:48:59] <XXCoder> cool
[23:49:00] -!- justanotheruser has joined #linuxcnc
[23:49:11] <Lcvette> but the existing controller is a bit wrecked
[23:49:26] <Lcvette> the monitor is broken on it beyond repair
[23:49:37] <Lcvette> there are some cables that have been damaged
[23:49:55] <Lcvette> not sure if i would want to even try and restore an old fanuc 6T control anyhow
[23:50:03] <Lcvette> bit clunky
[23:50:22] <Lcvette> late 80's
[23:50:30] <Lcvette> early 90's control
[23:52:12] <XXCoder> hard to source parts
[23:52:50] <Lcvette> yup
[23:54:53] <Lcvette> seems like a fun little machine from the research i was able to gather on it which is quite limited
[23:55:13] <Lcvette> it has a D1-4 spindle nose with 5C taper
[23:55:14] -!- justanotheruser has quit [Ping timeout: 268 seconds]
[23:55:25] <Lcvette> and has linear rails for x and z axis
[23:55:43] <Lcvette> pretty high tech for a small machine from the 80's
[23:56:12] <Lcvette> fairly light at only 790lbs
[23:56:17] <Lcvette> but not too bad
[23:56:46] <Lcvette> has a 2hp DC spindle
[23:57:09] <Lcvette> and had an optional pneumatic collet closer
[23:57:31] <Lcvette> although finding one is impossible im sure as these machines are pretty rare now days