#robotics Logs

Aug 21 2017

01:14 AM rue_house: but there was a happy ending
01:17 AM rue_house: so, I have this 4017 code to do
01:17 AM rue_house: I'll rewrite the code for the m328 from using its io pins to using a 4017
01:17 AM rue_house: I think, I can get it to run 16 channels
01:17 AM rue_house: with full resolution
01:17 AM rue_house: if the writes to the OCR are buffered, then its totally a win
01:19 AM rue_house: The OCR1x Register is double buffered when using any of the twelve Pulse Width Modulation
01:19 AM rue_house: (PWM) modes
01:19 AM rue_house: SO that means I can write the next value before the current one finishes, which means the irq service time does not affect the waveform
01:22 AM rue_house: now, the 4017 clocks on the rising edge...
01:23 AM rue_house: if I limit to to service 8 servos, then I can let it interleave the deadtime
01:23 AM rue_house: instead of packing all the pulses and leaving them to the end
01:23 AM rue_house: which should help normalize the gains
01:24 AM rue_house: so what do we change
01:24 AM rue_house: the irq handler
01:24 AM rue_house: thats it?
01:24 AM rue_house: well, the calc of the time might be intersting
01:25 AM rue_house: so, you want the time between rising edges to be the high time for the current servo
01:25 AM rue_house: k
01:26 AM rue_house: so ______^-------_________^-------_______
01:26 AM rue_house: the pwm value is the duration of the first pulse
01:26 AM rue_house: but the servo pulse is the time between the rising edges
01:27 AM rue_house: but, in the normal pwm system those rising edges are at fixed times
01:27 AM rue_house: how about we invert the pwm pins output polarity after a match
01:29 AM rue_house: __________^---I________C------^______?????
01:29 AM rue_house: the pwm cycle starts at ^
01:29 AM rue_house: the I is the software invert point
01:29 AM rue_house: and the C is the start of the next pwm cycle
01:30 AM rue_house: we didn't accomplish anything
01:30 AM rue_house: hold on, I'm thinking
01:31 AM rue_house: I'm not optimistic here
01:31 AM rue_house: sssh THINKING
01:32 AM rue_house: ok , on the m328 we have 2 channels on the high res. timer
01:32 AM rue_house: yea but
01:32 AM rue_house: SSHHHH
01:32 AM rue_house: for a pwm mode, we can set the clocks per cycle
01:33 AM rue_house: toggle
01:34 AM rue_house: damn, OCRB is used as TOP
01:35 AM rue_house: so I cant use top
01:36 AM rue_house: what re you doing now
01:36 AM rue_house: I just use it as a duration for the irq to clear the ports
01:36 AM rue_house: maybe we should bit bang it
01:36 AM rue_house: no, use the timer to toggle the pin
01:37 AM rue_house: oh, half bit banged
01:37 AM rue_house: THATS what I was TRYING to tell you
01:37 AM rue_house: can we do that
01:37 AM rue_house: 0 1 Toggle OC1A/OC1B on Compare Match.
01:37 AM rue_house: YES
01:38 AM rue_house: ok so let me get this right
01:38 AM rue_house: ok, I'm listening
01:38 AM rue_house: I know, I'm trying to put it all togethor, its like I'm low on ram, cause of some other process running or something
01:38 AM rue_house: uhu
01:39 AM rue_house: ok, so we leave the code as is for the most part, but change the timer to a mode that it toggles the pin when it hits time
01:39 AM rue_house: we turn the pin on, and start the timer
01:40 AM rue_house: we use a pwm mode so that it'll double buffer the OCR
01:40 AM rue_house: we write the next value to the OCR
01:41 AM rue_house: when it toggles it, we get an interrupt, by then, the state of the pin has been toggled and the value for the next edge has been loaded
01:41 AM rue_house: so we clear the pin back, and load the next value into the OCR
01:41 AM rue_house: almost right
01:41 AM rue_house: almost?
01:42 AM rue_house: on the first cycle you have to come into it with the pin clear, and set it just before you enable the timer
01:42 AM rue_house: so
01:43 AM rue_house: ____A----B_______C-----D________E------F______G------H___
01:43 AM rue_house: first cycle, irq turns pin on, and starts the pwm thats A
01:44 AM rue_house: B irq turns off pin before it exists
01:44 AM rue_house: C is when OCRx matches, and hardware toggles the pin
01:45 AM rue_house: a bit later, the irq (from the OCRx match) turns the pin off again thats D
01:45 AM rue_house: --
01:45 AM rue_house: at E the OCRx match happens again
01:45 AM rue_house: F irq turns it off
01:45 AM rue_house: G toggled by match
01:45 AM rue_house: times are A to C
01:46 AM rue_house: C to D
01:46 AM rue_house: er
01:46 AM rue_house: C to E
01:46 AM rue_house: E to G
01:46 AM rue_house: so we use fast pwm mode?
01:46 AM rue_house: yes
01:46 AM rue_house: but carefull, time for first cycle starts at A
01:47 AM rue_house: so load OCRx, set the pin high, start the timer, dillydally for a sec, set the pin low, load the next OCRx value, and leave
01:47 AM rue_house: unlike a normal irq that would be:
01:48 AM rue_house: load the next OCRx value, dillydally, set io pin low
01:48 AM rue_house: the last irq needs to stop the cycle or something
01:49 AM rue_house: why dont we just pulse the reset line and carry on?
01:49 AM rue_house: something is wrong here
01:49 AM rue_house: what
01:49 AM rue_house: what about the pwm counter rollover value
01:50 AM rue_house: you have to diff the OCRx value against the current pwm counter value
01:50 AM rue_house: ...
01:50 AM rue_house: its 'time from now' not actual pwm
01:50 AM rue_house: were just using it to time out the io pin
01:51 AM rue_house: ok, I think I got it
01:52 AM rue_house: wait, how do we know what value to load in?
01:52 AM rue_house: ?
01:52 AM rue_house: its not happened yet
01:52 AM rue_house: you add it to the time you knew the last thing would happen
01:52 AM rue_house: but, were on the clock
01:52 AM rue_house: not really, you have 8000 cycles
01:53 AM rue_house: oh, the 1ms porch
01:53 AM rue_house: yea
01:53 AM rue_house: ok so, if we do it like that, we can run 2x 4017 off TC1
01:53 AM rue_house: yea, 16 servos, 24000 count resolution
01:54 AM rue_house: which means the stm32 should be able to take on about 48 servos using the same method
01:54 AM rue_house: witht eh full 48000 count resolution that it has
01:55 AM rue_house: k
01:55 AM rue_house: hows this help with the 12' mecha?
01:55 AM rue_house: do it and you'll see! :)
01:57 AM rue_house: ok, I think I got it
01:57 AM rue_house: just go code
01:57 AM rue_house: wait, how do I test this?
01:57 AM rue_house: thats for you to work out, I'm just here to make life hard
01:57 AM rue_house: oh gee, thanks.
01:58 AM rue_house: atleast your helping
02:01 AM rue_house: hey guys! did I miss anyting!?
02:01 AM rue_house: nope
02:01 AM rue_house: no
02:02 AM rue_house: we just been sittin here listening to the crickets
02:02 AM rue_house: oh cool, hey, I got this idea for an avr program, a neural net simulator!
02:02 AM rue_house: oh gee, no I'm not into coding just now
02:02 AM rue_house: sorry, me either
02:02 AM rue_house: ok, fine fine, maybe later
02:03 AM rue_house: yea, laters good
02:04 AM rue_shop3: m328?
02:04 AM rue_shop3: yup
02:05 AM rue_shop3: fork hobbydriver?
02:05 AM rue_shop3: yea
02:09 AM rue_shop3: we only have 1ms
02:09 AM rue_shop3: no, sorry 0.5ms
02:09 AM rue_shop3: lots of time still
02:10 AM rue_shop3: so, we dont need the 2.5ms restart timer?
02:11 AM rue_shop3: no, were steaming
02:11 AM rue_shop3: damnit man, you wrote the whole thing in like 16 lines
02:11 AM rue_shop3: how the hell am I supposed to work this out
02:12 AM rue_shop3: there are 3 functions
02:12 AM rue_shop3: timerinit(), timer0irq() and timer1irq()
02:13 AM rue_shop3: so, first quteston is, does the timer have to be initialized differently?
02:13 AM rue_shop3: yea
02:13 AM rue_shop3: we need it to operate the io port
02:13 AM rue_shop3: good
02:13 AM rue_shop3: and were only using ... one of the timers
02:13 AM rue_shop3: the high res one
02:13 AM rue_shop3: thats timer1
02:13 AM rue_shop3: thanks
02:14 AM rue_shop3: ... so what do I do?
02:14 AM rue_shop3: *sigh*
02:14 AM rue_shop3: its only midnight
02:14 AM rue_shop3: yea...
02:14 AM rue_shop3: you should be able to handle this.
02:15 AM rue_shop3: just give me a kick, I'm ok, really...
02:15 AM rue_shop3: timer1, so I remove timer0 initialization
02:15 AM rue_shop3: :)
02:15 AM rue_shop3: do we need it to start a new cycle tho?
02:15 AM rue_shop3: if you want, you could set it up to 20ms
02:16 AM rue_shop3: 50hz?
02:16 AM rue_shop3: yea
02:16 AM rue_shop3: so we shouldn't 2.5ms all the channels
02:16 AM rue_shop3: no
02:18 AM rue_shop3: ... the code is over here
02:18 AM rue_shop3: oh, OH! thanks...
02:18 AM rue_shop3: whats 10*20 again?
02:18 AM rue_shop3: ugh
02:18 AM rue_shop3: just run it on timer1 for now
02:20 AM rue_shop3: you need a datasheet...
02:20 AM rue_shop3: uh... yea!
02:21 AM rue_shop3: dude, if you scroll with your eyes closed, you will miss what were looking for
02:22 AM rue_shop3: found it, COM1a1=0 com1A0 = 1
02:25 AM rue_shop3: there is a problem
02:26 AM rue_shop3: in fast pwm, oca1 will toggle, but oc1b wont
02:27 AM rue_shop3: I have a solution
02:27 AM rue_shop3: use the set on compare match
02:27 AM rue_shop3: and clear the pin
02:28 AM rue_shop3: other wise the same
02:29 AM rue_shop3: problem with that, its gonna clear it on its own
02:30 AM rue_shop3: we need to make sure it dosn't hit bottom
02:32 AM rue_shop3: the double buffering may not matter
02:32 AM rue_shop3: we have a half ms to update
02:33 AM rue_shop3: ok
02:34 AM rue_shop3: so, mode 0000
02:39 AM rue_shop3: ok, then we delete the code for timer0
03:08 AM rue_shop3: huh, my good software uses a different bus protocol
03:08 AM rue_shop3: awkward
03:38 AM rue_shop3: if I redo the firmware for controlling these hobby servos, I kinda put them on a different level
11:16 AM rue_bed: whens that eclipse?
11:18 AM codepython777: anyone has any recommendations on where one can buy 4p/6p DF13 cables in bulk in the US?
11:20 AM rue_bed: df13?
11:20 AM rue_bed: orlock, eclipse? when?
11:24 AM codepython777: rue_bed: the pixhawks plug those connectors in
11:27 AM rue_bed: hmmm, I need supply fail notification on an avr from a power adapter
11:50 AM rue_bed: arg, every morning I have to eat breakfast again
11:51 AM rue_bed: none of the breakfasts I'v had last
11:51 AM rue_bed: what do I want for breakfast? I want to be a robot.
11:51 AM rue_bed: ho hum
12:20 PM rue_shop3: its hanging about 50% here
12:20 PM rue_shop3: I"ll wait a bit I suppose
12:36 PM mumptai: any relevant open-source robots electronics projects out there?
12:37 PM veverak: you mean arduino or raspberry pi?
12:37 PM * veverak hides
12:39 PM mumptai: yeah, i'm well aware of stigmata of arduino and similar. but, as long as there are people actually using it and its open source hw, it qualifies from my point of view
12:40 PM veverak: wat?
12:40 PM veverak: :)
12:40 PM veverak: mumptai: I don't understand
12:41 PM mumptai: rpi and that popular AVR board aren't robotics-specific
12:44 PM veverak: yeah, but they are usually close enough
12:44 PM mumptai: but nothing really fancy?
12:45 PM veverak: nothing I am aware of
12:45 PM veverak: mumptai: you would need to be more ... specific
12:45 PM veverak: but I am not aware of anything 'universal'
12:46 PM veverak: of course, some companies make open source designs of battery management circuits and motor drivers
12:46 PM mumptai: decent power & battery management
12:46 PM mumptai: multi-axis servo control
12:46 PM mumptai: computer vision
12:46 PM veverak: yep, that are specific circuits for that
12:46 PM veverak: adafruit/sparkfun make these
12:47 PM veverak: hmm, for computer vision, usually webcam/kinetic + raspberry pi
12:48 PM mumptai: i actually did the later, and the rpi3 is rather limiting (but so are cheap usb cams)
12:48 PM mumptai: and something for sensor interfaces, maybe bus-based (can, rs-485, etc.)
12:49 PM mumptai: the old OAP would qualify, but the project is long dead (afaik)
12:50 PM mumptai: i'd like to extend this list: https://github.com/cajt/list_of_robot_electronics
12:54 PM veverak: mumptai: of rpi3 is not enough, I would go for strongest ODROID
12:54 PM veverak: if that is not enough, I would go for intel's NUC
12:56 PM mumptai: i have enough options, rpi3 is just cheap and availible for many
12:57 PM mumptai: the basic problem is that most robotics stuff is just too much for a single developer
12:58 PM mumptai: so cooperation is required and not at all optional
01:03 PM rue_shop3: huh, interesting
01:03 PM rue_shop3: the eclipse gives me a new outlook on blur
01:49 PM rue_shop3: iiiiiiiiiiiiiiinteresting
01:49 PM mumptai: yes?
01:49 PM rue_shop3: the COB led bulbs I got from work that were no good (internal ballasts blew up) are rated for 65w
01:49 PM rue_shop3: BUUUUT
01:50 PM rue_shop3: they have 12*18 0.5w leds on them
01:50 PM rue_shop3: which is 108w
01:50 PM rue_shop3: maybe they were limited by the ballast size
01:51 PM rue_shop3: should be 2.16A at about 36V
01:53 PM rue_shop3: I wonder if the 36V supplies are be dialed over 36V
01:53 PM rue_shop3: bet they can
02:05 PM rue_shop3: well, it didn't explode when I applied power, I even got a green light...
02:06 PM rue_shop3: ok, I can get 14V from the 12V model
02:06 PM rue_shop3: 117%
02:07 PM rue_shop3: so I should, by unfounded logic, get up to 42V from a 36V supply
02:07 PM rue_shop3: even if I only get 38, I'm ok
02:08 PM rue_shop3: so then, the question is, how do I regulate the current
02:10 PM codepython777: is there a way to get multiple i2c/spi connections to USB converter ? I can of course add an Arduino - but perhaps there is a better way?
02:10 PM rue_shop3: no
02:10 PM rue_shop3: the protocols are all mismatch
02:10 PM rue_shop3: and the timing... just isn't
02:11 PM rue_shop3: look for an asynchronous to synchronous project someone else did tho
02:11 PM rue_shop3: I'm sure its already been written a million times
02:12 PM rue_shop3: see, on i2c, you manually read as many bits as you want
02:12 PM rue_shop3: on async, you get get them
02:12 PM rue_shop3: you dont ask for them
02:12 PM rue_shop3: see the start of where it dosn't match up?
02:12 PM * rue_shop3 waits for an ACK or NACK from codepython777
02:13 PM codepython777: rue_shop3: Pointers? I have an intel linux machine on my robot, and would like to talk to I2C/SPI devices.
02:13 PM rue_shop3: use the i2c bus on the video card
02:13 PM rue_shop3: or the ram slot
02:13 PM rue_shop3: or the hardware monitor bus
02:14 PM rue_shop3: or use a parallel port
02:14 PM rue_shop3: via an ata port if you have to
02:14 PM rue_shop3: the parallel port i2c kernel module works great
02:15 PM * rue_shop3 waits for an ACK or NACK from codepython777
02:15 PM codepython777: rue_shop3: https://electronics.stackexchange.com/questions/165150/intel-nuc-gpio-access-from-ubuntu
02:16 PM rue_shop3: does it have a VGA connector?
02:16 PM codepython777: how do you physically connect multiple i2c devices to USB? no vga - a hdmi
02:16 PM rue_shop3: again, no
02:16 PM rue_shop3: apples - oranges
02:17 PM rue_shop3: like, one of them is using manchester and the other is using FSK
02:17 PM rue_shop3: dosn't just go togethor
02:18 PM rue_shop3: what board is it?
02:18 PM rue_shop3: what connectors does it have
02:18 PM rue_shop3: I bet there's an i2c bus
02:18 PM rue_shop3: did you install lmsensors and have it search?
02:18 PM codepython777: rue_shop3: it is an intel nuc
02:19 PM rue_shop3: I'm doing stuff, paste me a url
02:23 PM rue_shop3: so, I need up to 2.16A of 36V
02:24 PM rue_shop3: I have 18V adpaters rated for... 6A
02:24 PM rue_shop3: so, a 1:2 boost converter and I'm laughing
02:25 PM rue_shop3: hmm 50W boost converter, I think I can do that
02:26 PM mumptai: are the 18V adapters isolated?
02:27 PM rue_shop3: yup
02:29 PM mumptai: please, at least add some diodes to protect them from polarity reversal in the fault case
02:45 PM rue_shop3: yea, I really need a robot arm on a relocatable pedistal
02:47 PM veverak: :)
03:24 PM anonnumberanon: rue_shop3, sup
03:27 PM anonnumberanon: rue_shop3, got the gas motor for my RC airplane removed from the airplane, ready to disassemble and inspect
03:27 PM anonnumberanon: with your help hopefully
04:44 PM SpeedEvil: Any reason for gas, other than history?
04:47 PM anonnumberanon: the plane came with it, i changed it to electrical and it was nice but now i want to fix the gas engine
05:02 PM SpeedEvil: Actual gas, or methanol
05:25 PM rue_shop3: thats the one you couldn't get to run?
05:25 PM rue_shop3: I'm about to go shopping
05:34 PM peepsalot: anyone here use a no-contact laser tachometer? i just fried my super cheap one on accident, putting battery in backwards, so lame there was no protection
05:35 PM peepsalot: i'm wondering which white label product is better
06:44 PM rue_shop3: hah, you cant put the battery in backwards...
06:44 PM rue_shop3: none of the products are completely userproof, you need a smarter user
06:44 PM rue_shop3: :D
06:45 PM rue_shop3: I have to turn the internet off for a bit, is everyone ok with that?
06:45 PM rue_shop3: (please save your work, I'll have it back up in a bit)
07:06 PM Deacydal is now known as Deacyde
07:32 PM rue_house: ok, it looks like the internet is back up, and there are better batteries in the UPS
07:33 PM rue_house: and the data closet is dusted
07:33 PM orlock: rue_house: one of our factories in china lost power
07:33 PM rue_house: peepsalot, sorry, but if you cater to people who dont know what their doing, it just promotes people to know less and less about what their doing untill the whole world is stupid
07:33 PM orlock: rue_house: apparntly ome "custom made high pressure" part failed
07:34 PM orlock: apparently some, even
07:34 PM rue_house: orlock, high voltage, yea
07:34 PM rue_house: they call it pressure, bad translation
07:34 PM orlock: rue_house: yeah, thats what i guess it was, translation, etc.
07:34 PM orlock: ETA for a fix - 3 months
07:34 PM rue_house: I see that from aliexpress
07:34 PM orlock: Ahh, hah.
07:34 PM rue_house: their power service blew up
07:35 PM orlock: Yeah, but - 3 months for a factory substation? they are talking like it's a custom one-off
07:35 PM orlock: generator and/or really long extension cable to the building next door fixed it
07:35 PM rue_house: https://www.aliexpress.com/item/DC-DC-Constant-pressure-constant-current-module/1000001309871.html
07:35 PM orlock: but i was stunned by their comment that it was a custom designed system
07:36 PM rue_house: its china
07:36 PM rue_house: its all good till someone is relying on it
07:36 PM rue_house: probably a 480V transformer
07:37 PM rue_house: we have to deal with silly voltage equipment all the time
07:37 PM orlock: 480 3 phase down to 240?
07:37 PM orlock: is that how it works, or something else?
07:37 PM rue_house: if its big stuff it might need the 480 direct
07:37 PM orlock: well, they could fix the problem with a standard generator
07:37 PM rue_house: I dont know what the distribution voltages are in china
07:37 PM orlock: currently they are running a really long extension lead somewhere
07:38 PM rue_house: not if the equip needs some stupid voltage
07:38 PM orlock: anyway, they seem to be used to it
07:38 PM orlock: and are taking it in stride
07:38 PM rue_house: and a 'custom' transformer is all they have to produce it
07:39 PM rue_house: you would think, in china someone could make a custom transformer pretty quick
07:39 PM rue_house: maybe the quick ones are pricey
07:40 PM rue_house: ok, I'm supposed to arrange a new shop printer
07:44 PM Tom_shop: rue_house, how do you disable screen blanking on wheezy?
07:44 PM rue_house: 1 min
07:45 PM rue_house: turnthedamnscreensaversoffalready.sh
07:46 PM Tom_shop: that is
07:46 PM rue_house: #!/bin/bash
07:46 PM rue_house: xset s off dpms force on s noblank s expose -dpms
07:46 PM Tom_shop: doesn't work on wheezy i'm told
07:46 PM orlock: build a robot to jiggle the mouse every so often
07:46 PM rue_house: xset it like the toaster on red dwarf
07:46 PM rue_house: you cant just tell it you dont want toast
07:47 PM rue_house: you cant just tell it you dont want toast or begals
07:47 PM rue_house: you cant just tell it you dont want toast or begals, or crumpets,..
07:47 PM rue_house: use that xset line and it should disable EVERYTHING
07:47 PM rue_house: try it!
07:48 PM rue_house: tell me if it works, it took me days to develop
07:48 PM Tom_shop: but
07:48 PM Tom_shop: but
07:48 PM Tom_shop: but
07:48 PM Tom_shop: i need my keyboard...
07:48 PM rue_house: it works till you turn your computer off
07:48 PM Tom_shop: xset works only for the current session
07:49 PM rue_house: so it should last between 100 and 300 days
07:49 PM Tom_shop: want a startup solution
07:49 PM rue_house: you cant
07:49 PM rue_house: the session is whats setting it to blank
07:49 PM Tom_shop: I WANT A STARTUP SOLUTION!!!!
07:49 PM orlock: https://xkcd.com/196/
07:49 PM rue_house: and no, its not options you cant turn off
07:50 PM Tom_shop: i want this motor problem fixed worse though
07:50 PM rue_house: didn't the relay work?
07:51 PM rue_house: where is my tea?
07:51 PM Tom_shop: it blows the fuse when i use reverse
07:51 PM Tom_shop: even from a stop
07:51 PM Tom_shop: wires look right!
07:52 PM Tom_shop: it's just a damn dc motor. you reverse the wires and the motor turns backwards!!!
07:53 PM Tom_shop: i'm down to 1 fuse
07:54 PM rue_house: and your reversing the motor wries right
07:54 PM rue_house: have you tested the motor itself with just a plain dc supply?
07:54 PM Tom_shop: yes
07:54 PM rue_house: and all is good eh?
07:55 PM Tom_shop: works fine
07:55 PM Tom_shop: from a 12v battery
07:55 PM rue_house: take pictures, let me stare at it
07:56 PM rue_house: Tom_L, there is a way to put a command in the users profile start
07:56 PM rue_house: I dont do it, and its deep under the hood, but its there somewhere
07:57 PM rue_house: ok I cleaned out the data closet and fixed up the ups batteries
07:57 PM rue_house: the other task was a "new" printer for the shop
07:57 PM rue_house: wehre IS my tea?
07:58 PM SpeedEvil: Fix the replicator, ask it for some.
08:06 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/Control4.jpg
08:06 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/Control5.jpg
08:06 PM Tom_L: that's with the new relay in place
08:07 PM Tom_L: the blue relay triggers the 'inhibit' signal on the control so it's dead during switching
08:07 PM Tom_L: the other relay has a software delay of a few ms before switching
08:07 PM Tom_L: the black thing by the 2 fuses isn't connected
08:14 PM rue_house: waaaait
08:14 PM rue_house: you have the 4 wire cable going to the mtoor
08:14 PM rue_house: right?
08:15 PM rue_house: the cabtire?
08:15 PM Tom_L: 2 wires
08:15 PM rue_house: oh, the other two are L1 L2?
08:15 PM Tom_L: the big black cable is the motor
08:16 PM Tom_L: L1 L2 are 120vac in
08:16 PM Tom_L: white wire
08:16 PM rue_house: the big black cable has 4 wwires
08:16 PM Tom_L: 5
08:16 PM rue_house: blk/red are motor and white blue are power in?
08:16 PM rue_house: (I'm ignoring green)
08:17 PM Tom_L: blue goes in to the switch on the sherline and comes back out the white
08:17 PM rue_house: ok
08:17 PM Tom_L: red and black are the motor
08:17 PM Tom_L: that go to the center of the relay
08:18 PM rue_house: what part number is that relay base?
08:19 PM Tom_L: just a sec
08:20 PM Tom_L: 700-HN122 SER A
08:20 PM Tom_L: i can upload the pdf
08:21 PM rue_house: dude, ...
08:21 PM rue_house: did you check the motor for a ground fault?
08:21 PM Tom_L: umm, no
08:21 PM rue_house: it just occured to me that the conducts are packed in those mtoors soemthing horid
08:21 PM rue_house: meter +- to ground
08:21 PM rue_house: see if it took on a short
08:21 PM rue_house: might explain everything
08:21 PM Tom_L: ok
08:22 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/700-td552_-en-p.pdf
08:22 PM Tom_L: allen bradley pdf
08:24 PM rue_house: commons are on the bottom, not the middle
08:24 PM Tom_shop: no shit?
08:24 PM rue_house: <!>
08:24 PM Tom_shop: i though i ohm'd it
08:24 PM rue_house: page 18
08:24 PM rue_house: of pdf
08:25 PM rue_house: ?
08:25 PM Tom_shop: well that makes sense but it doesn't make any sense
08:25 PM Tom_shop: you sure?
08:26 PM rue_house: oh I might have the wrong one
08:26 PM rue_house: hold on
08:26 PM rue_house: bottom page 19
08:26 PM Tom_shop: i think the center is
08:26 PM rue_house: bottom right
08:26 PM Tom_shop: gotta go back inside to check
08:26 PM rue_house: no
08:26 PM rue_house: common is 3 and 6
08:26 PM Tom_shop: bottom?
08:26 PM rue_house: which are ont eh bottom of the starcase
08:26 PM Tom_shop: DAMMIT!
08:27 PM rue_house: :)
08:27 PM rue_house: page 19 of that pdf
08:27 PM rue_house: combined
08:27 PM rue_house: http://epub1.rockwellautomation.com/images/web-proof-small/GL/1068589.gif
08:27 PM Tom_shop: so move the X to the top 2 rows
08:27 PM rue_house: with that
08:27 PM Tom_shop: and put the motor on the bottom
08:27 PM Tom_shop: i'll ohm it out on the control first
08:28 PM rue_house: I'm gonna go build a loading dock, biab
08:28 PM Tom_L: which drawing was it again?
08:49 PM rue_house: bottom right of page 19
08:49 PM Luminax-Work is now known as LuminaxWk
08:49 PM rue_house: oh boy, LOTS of aliexrpess pages today
08:49 PM Tom_L: i'm gonna go ohm it right now
08:50 PM Tom_L: got held up
08:50 PM Tom_L: you better be right!
08:50 PM Tom_L: i got one fuse left!!
08:50 PM Tom_L: i don't see the drawing that goes with the pin numbers
08:51 PM rue_house: its the png I pasted
08:51 PM rue_house: http://epub1.rockwellautomation.com/images/web-proof-small/GL/1068589.gif
08:53 PM Tom_shop: it's so farking hot out here i can't think
08:54 PM veverak: http://hackaday.com/2017/08/21/open-source-modular-rocket-avionics-package/
08:54 PM veverak: cool
08:55 PM Tom_shop: thanks rue
08:58 PM rue_house: :) got it?
08:59 PM Tom_shop: switching the wires now
08:59 PM Tom_shop: got a spare i'll ohm
09:00 PM rue_house: I never trust the way those bases are wired
09:03 PM Tom_shop: looking very close on the housing the numbers are moulded i it
09:03 PM Tom_shop: very small
09:03 PM rue_house: yup
09:04 PM Tom_shop: my eyes aren't that good anymore
09:04 PM rue_house: my divergence is slowly failing
09:04 PM Tom_shop: so the middle row is the NC set
09:04 PM rue_house: high res, but multiple copies
09:04 PM Tom_shop: top is NO
09:04 PM rue_house: I think so yea
09:04 PM Tom_shop: so i want NC for FWD
09:05 PM rue_house: whatever floats your boat
09:05 PM Tom_shop: and that's how i think i have it wired now
09:05 PM Tom_shop: ready to test
09:06 PM rue_house: put a 60W 12V lamp in the fuse holder eh?
09:06 PM Tom_shop: yay
09:07 PM Tom_shop: fwd & rev work like they're supposed to
09:07 PM Tom_shop: no time for a lamp :D
09:07 PM rue_house: hah
09:07 PM rue_house: what do you think caused the origional failure?
09:07 PM Tom_shop: gawd that's been buggin me for a few days
09:07 PM Tom_shop: a stuck cheapass chinese relay
09:08 PM Tom_shop: like the blue one you see
09:08 PM Tom_shop: cause i was using 2 spdt instead of one dpdt
09:08 PM rue_house: ah, only one stuck?
09:09 PM Tom_shop: probably
09:09 PM Tom_shop: won't happen again now
09:09 PM rue_house: !
09:09 PM rue_house: I only have 18 things left in the mail form aliexpress!
09:09 PM rue_house: this isn't good, I keep a backup of atleast 20 things
09:10 PM rue_house: $7 usd, 3/8" NPT tap
09:10 PM rue_house: :)
09:11 PM rue_house: I dont know why I bought the 200R resistors
09:13 PM rue_shop3: *erp, you know, getting one that was hex shaped would have been a good idea
09:13 PM rue_shop3: have to keep an eye open for a tap handle
09:14 PM Tom_shop: make one
09:14 PM Tom_shop: i've got one
09:14 PM rue_house: yea, I made my last one
09:14 PM rue_house: oh, I have the lathe now too
09:14 PM Tom_shop: 2 bars with a V cut in the middle between 2 screws
09:14 PM rue_house: hah, for hex ones
09:15 PM rue_house: this one is round
09:16 PM rue_house: the resistors are in the middle of the optos and the transistors for the mecha
09:16 PM rue_house: so I can only assume they are for it too
09:20 PM Tom_shop: raining cats n dogs here now
09:20 PM Tom_shop: one just ran thru the shop :D
09:26 PM rue_shop3: heh
09:27 PM rue_shop3: that last command did work for the last computer I did it on
09:27 PM rue_shop3: (screensaver)
09:31 PM Tom_L: now maybe i can get on with my orient code
09:31 PM Tom_L: to orient the spindle for a tool change
11:03 PM Tom_L: i think i figured out the lut5 command
11:03 PM Tom_L: weighted lookup table function thingie
11:03 PM Tom_L: for all the logic gates
11:04 PM Tom_L: http://linuxcnc.org/docs/2.7/html/man/man9/lut5.9.html
11:04 PM Tom_L: i needed a 3 input or gate
11:05 PM Tom_L: 0xFE i think
11:10 PM rue_shop3: ok, I built about half the loading dock
11:11 PM rue_shop3: the other half is a castle, cause with all those bricks and the concrete, its just so much more fun, and ... well, ... I get distracted easy
11:14 PM rue_shop3: according to my hour meter I'm 45 hours into my vaccation
11:15 PM Tom_L: i wish notepad ++ worked in linux
11:15 PM rue_house: nedit
11:15 PM Tom_L: i didn't like it
11:16 PM Tom_L: notepadqq
11:16 PM Tom_L: maybe
11:19 PM rue_shop3: what didn't you like about it?
11:19 PM Tom_L: just not used to the way it worked
11:20 PM Tom_L: may try scite
11:20 PM Tom_L: notepadqq not available for debian
11:24 PM rue_shop3: the way it worked?
11:25 PM rue_shop3: its ui elements are horrid cause their as old as timeitself, but it will do awesome things that most editors wont consider
11:25 PM Tom_L: yeah i'm sure it's good