Back
[00:05:09] <rue_house> oooh I wanted to work on a robot like yours
[00:05:17] <rue_house> missed lots of things this evening
[00:05:31] <rue_house> insulated the boiler firebox today tho
[00:05:41] <rue_house> tommoorw I have to clean the boiler tubes :/
[00:06:28] <z64555> steam engine boiler, or house heating boiler?
[00:06:33] <rue_house> house heating
[00:06:42] <z64555> shame
[00:06:45] <rue_house> wood fired, the tubes get all carboned up
[00:07:04] * Anniepoo once had a job running a steam locomotive
[00:07:22] <rue_house> ? sweet
[00:07:46] <rue_house> there arn't many left
[00:08:04] <Anniepoo> 12" gauge Ottaway, ran around a park
[00:08:15] <rue_house> :) that works
[00:08:18] <Anniepoo> coal fired
[00:11:00] <rue_house> did the tubes ever need cleaning?
[00:13:20] <z64555> huh, there's a furnace that's made to replace a central electric heater
[00:13:41] <Anniepoo> yes
[00:13:48] <Anniepoo> every day
[00:14:14] <Anniepoo> before we started we cleaned the flues
[00:14:18] <rue_house> ah
[00:14:33] <rue_house> I get grumpy if I have to clean the tubes every two weeks
[00:15:29] <z64555> I would think the engine would consume more fuel than a house heating system
[00:15:32] <Anniepoo> while running we'd turn the blower on high and put sand in the firebox door
[00:16:27] <z64555> that... sounds harmful
[00:16:31] <rue_house> oh, no little pipe cleaners?
[00:17:07] <rue_house> I get about a gallon bucket of fine carbon powder from them all
[00:17:12] <Anniepoo> big pipe cleaner, used that before we started
[00:17:16] <rue_house> not worked out a use yet
[00:17:41] <z64555> rue_house: anyway to compress them into rods?
[00:17:57] <rue_house> hmm, I have a 20 ton press
[00:18:19] <z64555> could make your own pencils, or elements for carbon arc lamps
[00:18:21] <Anniepoo> every couple years you have to cut them all out and replace
[00:18:31] <z64555> or an electrode for a plasma torch
[00:20:09] <rue_house> I suppose thats I suppose on a 1/4" rod , thats 816k psi
[00:20:42] <rue_house> Anniepoo, thankfully not so far
[00:20:49] <rue_house> its all electrical conduit
[00:20:52] <rue_house> 1.25"
[00:21:15] <rue_house> 54 of them, about 4' long
[00:21:47] <rue_house> vert, I think I should put wires down them with flat danglers at the bottom
[00:21:58] <rue_house> as they flop around the wires brush the edges
[00:22:06] <rue_house> keep the tubes clean(er)
[00:23:06] <rue_house> its -3c out, the system pipes are losing about 1c/min
[00:23:06] <Anniepoo> we used a bottle brush on a long rod
[00:23:27] <rue_house> I have a plastic pluger thats a snug fit
[00:23:34] <rue_house> it just shaves off all the carbon
[00:25:49] <z64555> this is not helping my desire to make a steam engine
[00:26:31] <z64555> well, more like a self-powered bellows on a forge
[00:26:33] <rue_house> 34k left on the mortguage
[00:28:49] <z64555> original idea was to use a water-tube design that would wrap around the forge
[00:29:28] <z64555> but, that would require a fire that took up the entire hearth
[00:30:11] <rue_house> forge?
[00:30:14] <z64555> so a fire-tube design to collect heat from the chimney might be better
[00:30:28] <z64555> yeah, another project I'd like to build
[00:30:56] <rue_house> have to keep the fire stuff hot, or it carbons up
[00:31:12] <rue_house> which means you have to pull the energy out nicely
[00:31:36] <rue_house> I'm gonna tear into the boiler tommorow to see whats going on
[00:31:49] <rue_house> its burning hot, but I'm not getting heat out of it
[00:31:54] <rue_house> I think I set the choke wrong
[00:33:17] <rue_house> might need to seal some seams up better too
[00:41:38] <justan0theruser> anniepoo: do you think I can pierce t-slotted framing with the axle/shaft and it will remain structurally sound, or will I need a full-metal-beam with no air in it (if that makes sense)
[00:42:14] <Anniepoo> piercing is ok
[00:45:02] <rue_house> hahaha
[00:45:08] <rue_house> with 800 inch-lbs force?
[00:45:12] <rue_house> hahah
[00:45:27] * rue_house shuts up
[00:46:36] * rue_house works on how to reflect the signed values
[00:49:12] <rue_house> so I have 2 pwm channels, one on each side of the motor
[00:49:16] <justan0theruser> Anniepoo: I'm wondering where the roll pin actually goes after piercing
http://imgur.com/a/OUZhn
[00:49:19] <rue_house> if (AdcValues[0] > 512) {
[00:49:19] <rue_house> outputValues[0] = 0;
[00:49:19] <rue_house> outputValues[1] = (AdcValues[0] - 512) << 3;
[00:49:19] <rue_house> }else{
[00:49:19] <rue_house> outputValues[1] = 0;
[00:49:20] <rue_house> outputValues[0] = (512 - AdcValues[0] ) << 3;
[00:49:22] <rue_house> }
[00:49:25] <justan0theruser> spam
[00:49:26] <rue_house> that look right?
[00:49:49] <rue_house> justan0theruser, look for a hub...
[00:49:54] <justan0theruser> hush
[00:49:59] <rue_house> ugh
[00:50:36] <rue_house> wow, google remembered my BD
[00:50:56] <Anniepoo> directly into your picture
[00:50:59] <justan0theruser> Anniepoo: if the material were solid I could just go from basically ane of the middles of the sides, but since it's not I'm not sure it's even possible with t-slotted framing
[00:51:31] <justan0theruser> oh yeah...
[00:51:33] <Anniepoo> how about into short end of arm,
[00:51:35] <justan0theruser> that works
[00:51:43] * justan0theruser feels dumb
[00:52:05] <justan0theruser> that actually makes drilling easier too :p
[00:54:14] <rue_house> http://www.tracepartsonline.net/PartsDefs/Production/TEA_GER/80-16112007-401033/pictures/80-16112007-401033L.gif
[00:54:18] <rue_house> MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM
[00:54:35] * rue_house has interweb issues
[00:55:13] * rue_house dances and frantically points at the image to justan0theruser
[00:56:11] <rue_house> http://www.steeltubesindia.net/images/flange/flange-manufacturer-supplier.jpg
[00:56:19] <rue_house> hhahahahaha its bigger than both of ours
[00:56:51] <justan0theruser> rue_house: no ty
[00:57:17] <justan0theruser> think I've got my cart ready
[00:57:35] <rue_house> anyhow a flange like that is the way to do it
[00:57:45] <rue_house> the stuff just bolts to the side
[00:58:33] <rue_house> http://img.allhaving.com/upload/1961/b/16111.jpg
[00:59:04] <rue_house> https://ae01.alicdn.com/kf/HTB1CPkTJpXXXXX0XVXXq6xXFXXXs/-font-b-wheel-b-font-mounting-bracket-font-b-wheel-b-font-font-b-hub.jpg
[00:59:35] <rue_house> http://www.robotshop.com/media/catalog/product/cache/7/image/900x900/9df78eab33525d08d6e5fb8d27136e95/1/-/1-2-press-fit-key-bore-flange-hub-colson-2-1_1.jpg
[00:59:59] <rue_house> http://assets.suredone.com/1921/media-photos/15010402-dodge-taper-lock-flange-hub-qd-quick-disconnect-bushing-sd-1-1-4-kit-6x6-hole-8.jpg
[01:00:31] <rue_house> http://shared1.ad-lister.co.uk/GetImage.aspx?cid=386995a6-d873-4856-b875-78de128ad194&id=8a641252-ba71-4ba0-98e6-fa42958e6cf1&num=1
[01:00:37] <rue_house> justan0theruser, are you there!?
[01:13:04] <rue_house> rue
[01:13:10] <rue_house> yea
[01:13:13] <rue_house> help me work this out
[01:13:18] <rue_house> k
[01:13:24] <rue_house> bipolar current source
[01:13:37] <rue_house> and your positon is returned
[01:13:51] <rue_house> so you set all motors to the same current
[01:14:01] <rue_house> based on..
[01:14:13] <rue_house> based on the position error to the other servo
[01:14:52] <rue_house> so you have two loops?
[01:15:00] <rue_house> no
[01:15:06] <rue_house> just a current loop then?
[01:15:21] <rue_house> I suppose, the controller does the position delta loop
[01:15:46] <rue_house> how do you want to make a demo
[01:16:05] <rue_house> I suppose over serial to computer and back
[01:16:15] <rue_house> do we have a serial proto for that all yet
[01:16:20] <rue_house> iirc something about two streams
[01:16:52] <rue_house> so each controller would have to take in a current and give back a position
[01:17:00] <rue_house> I thought current had to go both ways
[01:17:28] <rue_house> no, you use the current setting to close the positional gap, but because its breoadcast your control might move instead of your plant
[01:17:57] <rue_house> so, we only have to read in the current values
[01:18:04] <rue_house> yea, and we report the positions
[01:19:27] <rue_house> so, reading the currents
[01:19:49] <rue_house> yea, we have to sync over a few pwm cycles
[01:19:58] <rue_house> a few?
[01:20:03] <rue_house> its a peak detector
[01:20:42] <rue_house> so, maybe over a 1.5x period to garuntee we hit a peak?
[01:20:58] <rue_house> you know tho, you have an interrupt on the finish of pwm
[01:21:04] <rue_house> ah
[01:21:26] <rue_house> so, part of the interurpt can be to collect the peak currents and reset the peak detectors
[01:21:41] <rue_house> I suppose we can use one ouput to reset them all at once
[01:21:55] <rue_house> but we need seperate 2n7000's
[01:22:23] <rue_house> I wonder how long the reset pulse needs to be
[01:23:05] <rue_house> oooh problem
[01:23:09] <rue_house> what
[01:23:18] <rue_house> the adc's cant read them all at once
[01:23:23] <rue_house> hmm
[01:23:52] <rue_house> so maybe 2 cycles, reset, and read them all over the next cycle
[01:24:00] <rue_house> dosn't work, the adc is running async
[01:24:02] <rue_house> hmm
[01:24:54] <rue_house> too bad we cant sync the adc mux with the 2n7000 resets
[01:25:53] <rue_house> you said something earlier
[01:25:58] <rue_house> ?
[01:26:12] <rue_house> about the current and the peak value time being thats important
[01:26:34] <rue_house> no, the average reads proportional to the motor load
[01:27:33] <rue_house> I'll guess the current peaks on the mtoor is proportional to the pwm duration :)
[01:27:39] <rue_house> huh
[01:28:26] <rue_house> is the inrush too fast for the peak detector?
[01:28:29] <rue_house> I dont remmeber
[01:28:35] <rue_house> I hope so
[01:28:41] <rue_house> hmm, good test point
[01:28:58] <rue_house> so, how long does it take the adc to read the current
[01:29:18] <rue_house> yea, almost need an 'adc cycle complete' flag
[01:29:43] <rue_house> and the pwm restart irq would clear them for a new cycle
[01:30:02] <rue_house> complex timing
[01:30:04] <rue_house> yea
[01:30:26] <rue_house> the pwm freq is almost 1Khz
[01:30:39] <rue_house> and the sample rate is almost 8khz
[01:30:48] <rue_house> one avr is doing 6 channels
[01:30:56] <rue_house> the other is doing 10 channels
[01:31:17] <rue_house> so, thats slightly under the pwm rate
[01:31:25] <rue_house> no, do the fine maths
[01:31:43] <rue_house> 976.562 Hz
[01:31:49] <rue_house> and 8khz...
[01:32:10] <rue_house> 8.192 channels
[01:32:25] <rue_house> I suppose .192 channels wont help us
[01:32:55] <rue_house> k, so, how do we read the current values
[01:33:14] <rue_house> what if we reset them in groups
[01:33:20] <rue_house> say, 4 groups?
[01:34:02] <rue_house> syncronization of that is a bit**
[01:35:25] <rue_house> we would need to have flags on each channel to say its been read
[01:35:49] <rue_house> the pwm irq would have to check all the flags for a group and reset the peak detectors if all the flags were clear
[01:36:20] <rue_house> and make sure there adc's didn't read them till after the next pwm cycle, wich means more flags
[01:36:46] <rue_house> so the adc would ahve to be skipping channels not ready
[01:37:15] <rue_house> and the pwm interrupt would have to be clearing channels that were part of a group that was done
[01:37:31] <rue_house> and resetting the ready flags if they had been clear for acycle
[01:37:48] <rue_house> your sure linear drives aren't a good idea?
[01:37:51] <rue_house> yea I'm sure
[01:38:00] <rue_house> cause, this is a lot of trouble
[01:38:14] <rue_house> well 7V, 1A, x8
[01:38:27] <rue_house> yea 56w, evermind
[01:39:02] <rue_house> ... there is no way one of those SG9 motors can take 7w
[01:39:09] * rue_house shrugs
[01:39:26] * rue_house looks at CrazyEddy
[01:39:27] <rue_house> ?
[01:39:32] <rue_house> nothing, nevermind...
[01:40:14] <rue_house> its too bad we cant abstract it more
[01:40:24] <rue_house> the lack of adc is a problem
[01:40:52] <rue_house> what about those audio adcs
[01:40:59] <rue_house> the 16 bit stereo ones?
[01:41:01] <rue_house> yea
[01:41:06] <rue_house> can you spell overkill?
[01:41:13] <rue_house> but, really
[01:41:22] <rue_house> how about a maxim multichannel adc
[01:41:30] <rue_house> you mean dac
[01:41:37] <rue_house> did you say adc?
[01:41:38] <rue_house> yea
[01:41:40] <rue_house> oops
[01:41:44] <rue_house> dac, anyhow
[01:42:02] <rue_house> what happened to the feedback digital dac project
[01:42:12] <rue_house> dunno, what processor you write it for
[01:42:17] <rue_house> ... cant reacall
[01:42:21] <rue_house> did it work?
[01:42:25] <rue_house> ... cant recall
[01:42:33] <rue_house> I think there was a stability issue
[01:42:51] <rue_house> wasn't that using pwm?
[01:43:10] <rue_house> yea, you were using an adc to stabalize a pwm converter
[01:43:18] <rue_house> for the pwoer supply
[01:43:24] <rue_house> that would be... t13?
[01:43:30] <rue_house> or t26?
[01:43:46] <rue_house> m328?
[01:45:57] <rue_house> all I can find is the audio dac code
[01:49:21] <Mindstorm8191> ...I'm confused. are there two people named rue_house?
[01:49:58] <rue_house> yes
[01:49:59] <rue_house> no
[01:50:09] <rue_house> yes
[01:50:10] <rue_house> no
[01:50:18] <rue_house> :P
[01:50:30] <Mindstorm8191> whatever, you / y'all have fun with that
[01:50:39] <rue_house> thanks Mindstorm8191
[01:50:48] <rue_house> its past bedtime
[01:50:54] <rue_house> I dont care
[01:51:02] <rue_house> work tommorow and youhave to do the garbage
[01:51:09] <rue_house> maybe i'll snow
[01:51:17] <rue_house> and maybe it wont, go to BED
[01:51:43] <rue_house> it was for the powers upply where is the power supply code?
[01:51:54] <rue_house> ok, code from bed then, thts why there's a laptop
[01:52:04] <rue_house> k, is the basement light off?
[01:52:09] <rue_house> DOH, seriously?
[01:52:19] <rue_house> you wanted to make sure the hose was off the tap too
[01:52:22] <rue_house> damn you
[01:52:36] <rue_house> get a shirt on, its cold out htere
[01:52:45] <rue_house> -3.5 now....
[14:17:42] <Jaimie__> I'm looking at one of the cheap tracked chassis on aliexpress and I was wondering if there's a simple way to hook up one of the 6wire hall effect geared motors to an RC receiver?
[14:20:14] <Tom_itx> no
[14:20:46] <z64555> "hall effect geared motors" what?
[14:21:13] <z64555> Is that a geared motor with hall effect sensors? Because you can't have a hall effect motor
[14:21:54] <Tom_itx> even so, no 'simple' solution to hook it up
[14:21:59] <Jaimie__> I'll just show you like this
https://www.aliexpress.com/store/product/25-motor-85mm-Plastic-wheel-coupling-motor-bracket-etc-of-ONE-Set-of-Car-Parts-For/1849945_32529530064.html?spm=2114.12010608.0.0.eOKEEx
[14:22:10] <Tom_itx> RC puts out a different pulse than a DC motor is looking for
[14:23:31] <Jaimie__> Thanks Tom_itx, got any ideas how it can be done?
[14:24:12] <Jaimie__> Should I go with Arduino or NodeMCU? They're both in languages I haven't used.
[14:24:16] <z64555> Jaimie__: that's a geared motor with hall effect sensors for position/speed detection
[14:26:36] <z64555> hm, there's two sensors, so you can detect the rotor's angular position to a degree
[14:27:41] <Jaimie__> That's how it comes z64555 but really I just want to control each motor separately so I can drive it like an RC car to start with
[14:28:19] <Jaimie__> the positional feedback isn't really something I desire
[14:28:41] <z64555> You'll still need a driver for the DC motor, as Tom_itx mentioned, the RC reciever outputs position signals used by servos
[14:30:09] <Jaimie__> thanks
[14:31:27] <z64555> I'm trying to think/remember if there's a simple circuit to act as a driver
[14:37:25] <z64555> Yeah, there is, but would need some knowledge of op amps and RLC circuits to get it right
[14:37:54] <z64555> Would be easiest for a beginner to use a uC to detect the pulse width and all that
[14:38:33] <z64555> and then use a pair of power transistors to power the motor
[14:39:25] <z64555> eh, or just go with an H bridge. they sell those as IC's
[14:39:40] <Jaimie__> can you show me one?
[14:39:47] <z64555> https://en.wikipedia.org/wiki/H_bridge
[14:40:08] <z64555> https://www.sparkfun.com/products/315
[14:40:20] <z64555> there's one for sale, fairly low powered
[14:40:35] <Jaimie__> I have some TIP120s, can I use one?
[14:42:03] <z64555> Darlington pairs? sure. one? no. You'd need at least 4 to make the H bridge
[14:42:20] <z64555> If you just want to go in one direction, then a single TIP120 should suffice
[14:43:54] <z64555> be advised
[14:44:22] <z64555> If you want to make your own H-bridge, you'll have to bias all four transistors so they have the same gain
[14:44:53] <Jaimie__> sounds like buying an IC is simpler
[14:45:01] <z64555> You can get away with not doing that, but you'll lose power across the transistors more than what you're putting in the motor
[14:45:12] <Jaimie__> The art of electronics book is still on my todo list
[14:45:44] <Jaimie__> I'm still in the "plug is all together and watch for the smoke" school of artistry
[14:46:07] <z64555> application of electronics is knowing the basic electrical rules plus knowing how semiconductors work
[14:46:55] <z64555> you'll spend time on diodes and triode (transistor) basic operation, then progress into signal analysis of transistors
[14:49:11] <Jaimie__> I once knew it back in 6th form college, im sure it will come back to me, thanks for your time z64555
[14:49:44] <z64555> heh, "Thanks for you input, now go away" :P
[14:49:55] <z64555> I'm just kidding. best of luck
[14:51:10] <Jaimie__> heh, I'll make a snow blower out of this yet, somehow
[15:28:56] <rue_house> I asked myself what would be REALLY exciting to do, and I had no answer
[15:32:39] <SpeedEvil> rue_house: rob a bank
[15:38:02] <anniepoo_> @mbrumlow, you on?
[15:38:05] <anniepoo_> 8cD
[19:24:03] <anniepoo_> 8cD just discovered the continuous ink system on the wide format printer was hanging half loose and spilling ink
[19:24:17] <anniepoo_> 8cD more securely attached now, and my hands are pretty colors
[19:42:19] <anniepoo_> @mbrumlow you about?
[20:34:04] <justan0theruser> is t-slotted framing going to have the same center diameter as its T-gap?
[20:34:38] <anniepoo_> measure it
[20:35:41] <justan0theruser> buy once, then measure twice
[20:35:49] <anniepoo_> (weird factoid - you can make precise measurements with printer paper. Figure out how many sheets thick X is, measure height of a full ream, and look at how many sheets are in ream
[20:36:19] <anniepoo_> googling dimensional standards for T slot framing would work too
[20:36:26] <anniepoo_> but measuring is always safer
[20:36:28] <justan0theruser> will do
[20:51:27] <justan0theruser> would it be fine using a .375in diameter spring pin in a .32in hole?
[20:51:34] <justan0theruser> spring roll pin
[20:51:40] <justan0theruser> I guess if it doesn't work I can drill the hole bigger
[20:52:39] <justan0theruser> oh that table you linked tells me
[20:53:26] <justan0theruser> I guess I'll make the hole bigger
[20:55:01] <justan0theruser> oh wait no, I can get a 1/4, nvm
[22:07:23] <justan0theruser> maybe the machine shop can curve t-slotted framing for me
[22:08:20] <anniepoo_> justa, how big is the axle and have you done anything to the t-slot material yet?
[22:08:31] <justan0theruser> yes, ordered
[22:08:43] <justan0theruser> this is separate t-slotted framing from the procedure we were discussing
[22:08:49] <anniepoo_> ok
[22:08:55] <anniepoo_> we have rollers
[22:09:07] <anniepoo_> you're agonizing too much over all this.
[22:09:13] <justan0theruser> ?
[22:09:20] <anniepoo_> I was going to offer to just do it
[22:09:25] <Tom_L> seems the only alive channel i'm on...
[22:09:35] <Tom_L> testing a new pc. can you see this?
[22:09:37] <Tom_L> http://tom-itx.no-ip.biz:82/~webpage/
[22:09:38] <justan0theruser> no
[22:09:43] <justan0theruser> yes I can
[22:09:46] <Tom_L> thanks
[22:09:47] <justan0theruser> anniepoo_: what do you mean?
[22:09:58] <justan0theruser> I'm just planning a separate step
[22:10:17] <anniepoo_> as in, I'm partners in an engineering firm
[22:10:23] <anniepoo_> we have a machine shop
[22:10:27] <justan0theruser> no, I mean how am I agonizing
[22:10:42] <justan0theruser> I thought since I'm new to this I'd discuss my plans first
[22:11:06] <justan0theruser> https://www.mcmaster.com/#t-slotted-framing/=15dybii
[22:11:14] <justan0theruser> this on a curved t-slotted extrusion
[22:11:22] <anniepoo_> you've spent a lot of time on here talking about this arm-axle joint. I recognize maybe it's just that you want to learn
[22:11:42] <justan0theruser> I just don't want to fuck it up. The parts are ordered though and I have a decent idea I think
[22:11:42] <anniepoo_> and that's great, but if it's becoming a barrier to you doing more interesting things,
[22:12:00] <anniepoo_> I do have the capability to do this basically without thinking about it.
[22:12:10] <justan0theruser> mhm
[22:12:27] <justan0theruser> anniepoo_: are you in chicago area?
[22:12:29] <anniepoo_> so, it's a bit like if you're studying, I don't want to blurt out the answer
[22:12:32] <anniepoo_> no, I'm in Oregon
[22:12:41] <justan0theruser> ok, wouldn't want to trouble you then :p
[22:12:48] <anniepoo_> no trouble
[22:12:57] <justan0theruser> mcmaster is right here
[22:13:06] <justan0theruser> not sure how you get stuff out there, same way?
[22:13:08] <anniepoo_> yes, well, they're right everywhere
[22:13:21] <anniepoo_> I'm in a tiny town, but UPS comes here
[22:14:29] <justan0theruser> I'm gonna pick em up
[22:14:36] <justan0theruser> shipping is crazy on these extrusions
[22:14:45] <anniepoo_> ok
[22:15:04] * anniepoo_ has big chunks of steel shipped to her
[22:16:06] <justan0theruser> the electronics are coming tomorrow and the PSU coming soonish so things are about to get interesting on that side
[22:17:00] <anniepoo_> 8cD
[22:18:15] <justan0theruser> not sure what that is tbh
[22:18:27] <justan0theruser> a smushed nose maby
[22:18:29] <anniepoo_> sorry, it's my smiley
[22:18:46] <anniepoo_> I habds a button nodze
[22:18:47] <justan0theruser> is it michael jacksons nose?
[22:19:22] <anniepoo_> I need to be afk about 10 mins
[22:28:25] <justan0theruser> how much backlash do Replaceable-Center Flexible Shaft Couplings experience
[22:28:44] <justan0theruser> can't seem to find anything on these
https://www.mcmaster.com/#standard-shaft-couplings/=15dyjcz
[22:36:30] <anniepoo_> a lot.
[22:36:46] <justan0theruser> oh really
[22:36:54] <justan0theruser> ok, I will have to get the zero backlash ones
[22:37:46] <anniepoo_> the kind with a spiral cut have less
[22:39:04] <justan0theruser> $150 for the coupling :p
[22:40:18] <justan0theruser> what about rigid couplings? Do they have very little backlash?
[22:50:40] <anniepoo_> that's nuts
[22:50:51] <anniepoo_> (price is nuts)
[22:50:59] <justan0theruser> yes, looking on the ali
[22:51:10] <anniepoo_> yes, a rigid coupling should have effectively none
[22:51:33] <justan0theruser> those ar $110
[22:51:53] <justan0theruser> well the servo ones
[22:52:42] <anniepoo_> https://www.amazon.com/XYXDI-Flexible-Coupling-Stepper-Connector/dp/B01FXJ3RC0/ref=sr_1_5?ie=UTF8&qid=1481257377&sr=8-5&keywords=motor+coupling
[22:53:54] <justan0theruser> my shaft is 20mm and keyed
[22:54:27] <anniepoo_> that's a hefty shaft
[22:55:10] <justan0theruser> ;)
[22:58:47] <justan0theruser> hows $70 sound for 20-35mm
[22:59:16] <justan0theruser> I assume these have relatively low backlash?
https://www.mcmaster.com/#standard-shaft-couplings/=15dyxrs
[23:01:24] <anniepoo_> what kind of motor is this with this 20mm shaft?
[23:02:10] <anniepoo_> that's a 3/4 shaft - I don't think we have any motors in the shop with a 3/4 shaft! And we've got machine tools!
[23:02:14] <justan0theruser> its a stepper servo moter
[23:02:24] <justan0theruser> its actually a planetary gearbox
[23:02:48] <justan0theruser> the motor + gearbox are 20lbs and 1ft
[23:02:59] <justan0theruser> :D
[23:03:19] <Jak_o_Shadows> justan0theruser, if I was less lazy, I would work out what size shaft you need.
[23:03:21] <justan0theruser> I guess $75 doesn't sound so bady anymore
[23:03:23] <anniepoo_> ok, hope you have an appropriate guard around this thing
[23:03:31] <justan0theruser> umm, guard?
[23:04:17] <anniepoo_> uh, if you attach a length of T rail to this monster and turn it on, and it decides to move because you left out a comma somewhere in the software,
[23:04:26] <anniepoo_> and you're in the way....
[23:04:49] <justan0theruser> anniepoo_: oh I see
[23:05:06] <justan0theruser> the guard is distance I suppose
[23:05:16] <anniepoo_> seriously, this is why big industrial robots have a fence around them
[23:05:18] <justan0theruser> and my cat like reflexes
[23:05:54] <anniepoo_> I wouldn't depend on reflexes
[23:06:07] <anniepoo_> this thing's a lot more dangerous than a table saw.
[23:06:41] <justan0theruser> the PSU to drive it is 6A 100VDC
[23:06:46] <justan0theruser> :p
[23:06:47] <anniepoo_> (also, you DO have the motor secured well? If it comes loose, you have a very dangerous machine on the loose
[23:06:57] <justan0theruser> nope
[23:07:02] <justan0theruser> I live life on the edge
[23:07:24] <justan0theruser> the motor will spin at 1RPM
[23:07:50] <justan0theruser> good advice though, I will be sure to keep my distance
[23:07:50] <anniepoo_> if the delay function doesn't have a bug and returns immediately
[23:08:08] <justan0theruser> 1RPM isn't a software constraint
[23:08:22] <anniepoo_> yes, you absolutely have to stay out of the robot's work envelope
[23:08:30] <justan0theruser> motor is max 1000RPM, and the gearbox is 16:1
[23:08:37] <justan0theruser> sorry, 1RPS :p
[23:09:19] <anniepoo_> that could still give quite a thunk
[23:09:31] <anniepoo_> it's a pinch hazard
[23:09:38] <anniepoo_> and an impact hazard
[23:09:42] <justan0theruser> the diameter is 10ft, so metal hitting me at 10ft/s
[23:09:49] <anniepoo_> 10ft?
[23:09:49] <justan0theruser> decent whack
[23:09:54] <justan0theruser> diameter
[23:10:20] <justan0theruser> first stage, right now, the diameter is actually ~5ft
[23:10:39] <justan0theruser> its like 3lbs
[23:10:57] <anniepoo_> and 80 ft lb torque - only allowing 16 lbs torque?
[23:10:58] <justan0theruser> 15lb-ft/s of momentum
[23:11:33] <anniepoo_> ok, not quite as scary as it sounded, but
[23:11:47] <justan0theruser> its not 16 lbs of torque I think?
[23:12:07] <anniepoo_> be sure you're sure the 1RPS won't be exceeded if the motor, for example, does an uncommanded slew to 0
[23:12:37] <anniepoo_> 16 lbs of pressure at the arm end
[23:12:43] <justan0theruser> 3.5ft * 0.5lb/ft * 2(bars) / 2 (avg length) = 1.75lb-ft of torque from arm, no?
[23:12:54] <anniepoo_> I'd point out it's 4 times that 1 ft from center
[23:13:36] <anniepoo_> I was using 5 ft radius, 80 ft-lb torque , so 80/5 = 16
[23:14:00] <justan0theruser> well for now the only torque I'm measuring is the torque of an empty load
[23:14:06] <justan0theruser> empty arms
[23:15:13] <anniepoo_> is this t rail supported somehow?
[23:17:29] <anniepoo_> now I'm understanding why you're asking about the connection
[23:19:09] <justan0theruser> anniepoo_: by the axle?
[23:20:15] <justan0theruser> oh wait, are you talking about the rail that is curved that I was discussing earlier?
[23:20:20] <justan0theruser> which connection are you referring to
[23:21:39] <anniepoo_> the one you spent considerable time discussing with me
[23:21:52] <anniepoo_> at this point we really need a drawing of this thing 8cD
[23:22:27] <justan0theruser> wait, so you think the roll pin idea might not work anymore?
[23:22:29] <justan0theruser> D:
[23:23:04] <anniepoo_> lets see a drawing of what you're making
[23:23:15] <anniepoo_> well, a 1/4" roll pin should be OK
[23:23:24] <anniepoo_> I'm just worried about the arm itself
[23:23:42] <justan0theruser> http://imgur.com/a/HScxw
[23:23:44] <justan0theruser> lol
[23:23:52] <justan0theruser> the flange idea is out tho
[23:23:56] <justan0theruser> that was my previous design
[23:24:41] <anniepoo_> this is how it's oriented?
[23:24:47] <anniepoo_> did you figure in the dead load?
[23:25:20] <justan0theruser> you mean the weight of the arms themselves?
[23:25:25] <anniepoo_> yes
[23:25:32] <justan0theruser> 22:44:04 justan0theruser:3.5ft * 0.5lb/ft * 2(bars) / 2 (avg length) = 1.75lb-ft of torque from arm, no?
[23:25:51] <anniepoo_> ok, but you have to add inertial loads
[23:26:03] <anniepoo_> and any load on the end of the arm
[23:26:23] <justan0theruser> sure
[23:26:30] <justan0theruser> they will be minimal I think
[23:26:43] <anniepoo_> (which might not be a designed load, eg payload, but rather something like accidentally shoving an object into it's path)
[23:27:04] <justan0theruser> I don't know how to model that
[23:27:16] <anniepoo_> what happens if it does?
[23:27:21] <anniepoo_> you might want a 'fuse'
[23:27:35] <justan0theruser> please explain such a design
[23:28:03] <anniepoo_> for example, you could mount the arm on a collar with a rollpin holding it to shaft.
[23:28:16] <anniepoo_> the rollpin is deliberately the weakest member
[23:28:19] <anniepoo_> it's the 'fuse'
[23:28:33] <anniepoo_> so it fails when you collide
[23:28:55] <anniepoo_> the inertial loads are simple physics, sounds like you can do that
[23:29:16] <justan0theruser> collar?
[23:29:43] <anniepoo_> well, if your weakest member is the arm itself, a collision means you buy a new arm
[23:30:12] <justan0theruser> I don't think so
[23:30:16] <justan0theruser> I think it means I buy a new motor
[23:30:23] <anniepoo_> yes, equally pleasant
[23:30:25] <anniepoo_> 8cD
[23:30:38] <anniepoo_> rollpins are about $2
[23:30:50] <anniepoo_> that motor's gotta be worth a big chunk of change
[23:30:55] <justan0theruser> so I buy a rollpin that is somehow guaranteed to break at 100 ft-lb
[23:31:00] <anniepoo_> yes
[23:31:37] <justan0theruser> I thought the torque ratings on roll pins were ratings, not guarantees of failure
[23:31:55] <anniepoo_> i'm using wrong term
[23:32:00] <anniepoo_> "shear pin"
[23:32:11] <anniepoo_> a shear pin is the fuse
[23:32:23] <anniepoo_> but it's just a pin
[23:32:33] <anniepoo_> https://en.wikipedia.org/wiki/Shear_pin
[23:32:48] <justan0theruser> those are cool
[23:32:50] <anniepoo_> you can buy shear pins
[23:33:04] <justan0theruser> but why the hell doesn't mcmaster tell me the shear pin torque
[23:33:09] <anniepoo_> ok, I gotta be afk a bit
[23:33:17] <justan0theruser> ok thanks for all your advice
[23:37:02] <anniepoo_> yup
[23:38:12] <anniepoo_> do show us the plans before executing. You have large forces here, it should be given at least a reasonable intuitive analysis before being constructed, for safety's sake
[23:38:37] <justan0theruser> danger is my middle name
[23:38:41] <justan0theruser> also slip clutches are cool
[23:39:17] <anniepoo_> also, I probably should say "I am not a licensed mechanical engineer" IANALME
[23:40:46] <justan0theruser> you're not?
[23:40:58] * justan0theruser throws away all the stuff he bought due to your advice
[23:42:41] <z64555> we cross train
[23:43:23] <z64555> engineering is applied physics and chemistry, and we all have a pretty solid grasp of the basics
[23:44:13] <z64555> plus there's a few analogs across fields, such as RLC networks and mass/spring/damper systems
[23:45:10] <anniepoo_> I've taken most of the core courses a mechanical engineer would take. I have worked around materials for many years. I have not quite a bachelors degree in physics
[23:45:30] <anniepoo_> I've designed many mechanical devices of similar complexity and size to this
[23:46:02] <anniepoo_> and what z said
[23:46:46] <anniepoo_> I wouldn't design a bridge, a pressure vessel, or a rocket body
[23:46:47] <justan0theruser> sorry, I was being sarcastic if that wasn't clear
[23:46:58] <anniepoo_> 8cD
[23:47:12] <justan0theruser> 8>D
[23:47:16] <anniepoo_> well, it's reasonable to question advice gotten from random people on internet
[23:47:51] <justan0theruser> well the advice makes sense at least
[23:48:23] <anniepoo_> you can get hurt dealing with large forces
[23:49:10] <justan0theruser> anniepoo_: the end goal
https://www.youtube.com/watch?v=pQ2dI_B_Ycg
[23:56:14] <justan0theruser> where should I look for torque limiters other than aliexpress
[23:56:15] <anniepoo_> you want it to be that long?
[23:56:23] <anniepoo_> long/big?
[23:57:32] <justan0theruser> oh no, the video is a joke, I'm not making a knife wielding bot
[23:57:38] <anniepoo_> we're going to do a 'robot day' for the local PFLAG group. One of the members tells me he wants to make a tentacle.
[23:57:43] <anniepoo_> so we'll do that
[23:57:51] <anniepoo_> they're actually fairly simple, and fun
[23:58:07] <justan0theruser> what is it, just a motor attached to a bendy tentacle?
[23:58:19] <justan0theruser> then the rotation introduces chaos to the tentacle?
[23:59:16] <anniepoo_> no, it's 'crinkly' expanding tubing
[23:59:26] <anniepoo_> 3 of them around a non-stretchy core
[23:59:46] <anniepoo_> you put air pressure in them, less in one, and it bends
[23:59:52] <anniepoo_> you have several segments
