#robotics Logs

Aug 01 2014

#robotics Calendar


12:23 mrdata hi
12:24 Solarlux all about robotics here?
12:28 deshipu yeah
12:33 Solarlux I want looking in CNC field
12:33 Tom_itx cnc robotics?
12:34 Tom_itx or cnc machines?
12:34 Solarlux CNC machines
12:34 Tom_itx join #linuxcnc
12:34 Solarlux I like all this stuff but i want implement in life some stuff
12:34 Solarlux i want learn
12:34 Solarlux Tom_itx All about cnc machines?
12:34 Solarlux Thank u very much]
12:34 Tom_itx it's all about linuxcnc
12:35 Tom_itx a open source machine control software
12:35 Solarlux Why not windows cnc?
12:35 Solarlux windows sucks?
12:35 Tom_itx windows is too slow
12:35 Solarlux uhh thanks
12:47 SquirrelCZECH_ Solarlux: and windows won't allow you to modify everything as you like
12:47 SquirrelCZECH_ which is key point, because as far as i know, in linuxcnc
12:47 SquirrelCZECH_ they modify the kernel of the system
12:48 Tom_itx use RTAI kernel
12:48 Solarlux Open source code ?
12:48 SquirrelCZECH_ everywhere! :D
12:48 Solarlux In linux cnc motor control soft?
12:48 Solarlux Nice
12:48 Solarlux Sounds very good
12:49 Tom_itx http://buildbot.linuxcnc.org/
12:49 Solarlux What type of Linux is most stable?
12:49 Solarlux Linux is from Unix?
12:50 SquirrelCZECH_ well
12:50 SquirrelCZECH_ if you want to pay, redhat
12:50 SquirrelCZECH_ otherwise debian?
12:50 SquirrelCZECH_ dunno :/
12:50 sanman depends what you're using it for
12:50 sanman AIX
12:50 sanman that's from IBM
12:50 Tom_itx for cnc you need RTAI
12:54 deshipu "stable"
12:54 deshipu stable means old versions of everything
12:55 deshipu and if you install your own stuff on top, you lose stability
12:55 deshipu there was that free build of riscos for rpi
12:56 SquirrelCZECH_ well
12:56 SquirrelCZECH_ it's not like you have to install fresh libraries because of every app
12:56 SquirrelCZECH_ but yeah, it got it's downsides
13:02 deshipu I can't see why you would need a stable system for a hobby project
14:44 OracleCarbon Well my workbench went from clean to cluttered in about 3 minutes.
14:51 iveevue :(
14:53 OracleCarbon My servos don't have any model number on them. Ideas on how to tell what voltage to power them with?
15:00 GargantuaSauce_ highest voltage rating i've seen on em is 7.2V
15:00 GargantuaSauce_ sometimes they're just rated for 6 though
15:06 OracleCarbon So...throw 'em on the 5v I've got and call it a day?
15:09 GargantuaSauce_ perhaps, but do be aware of the noise they can introduce
15:10 GargantuaSauce_ given that you're not drawing much current on the control electronics side of things it might be pertinent to use the arduino's linear regulator straight off the battery for that, and the smps just for the servos and other bigger loads
15:10 GargantuaSauce_ (or a second smps for the electronics)
15:11 OracleCarbon That's not a bad idea.
15:13 acleCarbon sighs, realizes he has no pins to connect th
15:14 OracleCarbon Any chance Radio Shack will have pins and headers in store?
15:14 OracleCarbon Or will I have to delay fielding this system for a few business days?
15:14 GargantuaSauce_ no idea, i'm in canada and radio shack turned into 'the source by circuit city' and stopped carrying anything useful
15:15 OracleCarbon It still has a few drawers of electronics parts around here.
15:15 GargantuaSauce_ http://www.ebay.ca/itm/231297016386 http://www.ebay.ca/itm/161381690297 these are very useful
15:16 OracleCarbon Yeah, I've got a pile of loose wires with pins
15:16 OracleCarbon I guess that would work...
15:17 GargantuaSauce_ http://stores.ebay.ca/UPGRADE-INDUSTRIES/Connectors-/_i.html?_fsub=5252610011 these guys might be fast since they're in the states, no experience with them though and the prices aren't awesome
15:17 GargantuaSauce_ better than other e-retailers though i guess
15:17 OracleCarbon I'd like to get this thing built by tomorrow morning, ideally.
15:18 GargantuaSauce_ heh
15:18 OracleCarbon The faster I can field it, the better my advantage.
15:18 GargantuaSauce_ well use the jumpers for now i guess, might need a dab of hot glue to secure them
15:18 OracleCarbon Hrm. I don't actually have strippers, I guess I've just "borrowed" them from the lab before.
15:19 OracleCarbon Looks like I need to take a trip to the store anyways, or steal an awful pair from the lab...and some headers
15:19 GargantuaSauce_ generally you want to solder stuff for robotics applications in any case. way more robust
15:20 OracleCarbon Yeah, I'm just tossing this together off an Arduino so pins seemed helpful.
15:20 GargantuaSauce_ also this is going to be in a water fight right?
15:21 GargantuaSauce_ have you thought about waterproofing
15:21 OracleCarbon Yeah, once I get the electronics made, I'll cut an acrylic box to close them in
15:22 OracleCarbon Or at least put a shield over top. This won't be shot at.
15:22 GargantuaSauce_ the servos will be a bit more complicated
15:22 OracleCarbon It just has to get off a sneak attack
15:22 OracleCarbon Splashback is the worst it'll see.
15:22 GargantuaSauce_ yeah probably sensible to assume it'll get drenched anyway
15:23 GargantuaSauce_ i've seen people pack servos entirely full of grease for waterproofing, that might be something to consider
15:24 GargantuaSauce_ and as for the drive motors i dunno
15:24 OracleCarbon I don't know if I'm going to end up using a servo on the gun. It's pretty big, might be better to fix it.
15:25 OracleCarbon Is there any rhyme or reason to whether I want the drive wheels in the front or back of the treads?
15:25 OracleCarbon I can't see why it would matter.
15:26 GargantuaSauce_ i am not experienced with treads but that would be my intuition also
15:26 OracleCarbon Because with some clever plumbing, I can fix the pump down low opposite the gearbox for the drive motors, which should 1) make it less obvious what this is doing, and 2) make it lower-centered
15:27 OracleCarbon And let me house the control electronics above the water systems
15:27 GargantuaSauce_ good plan
15:28 OracleCarbon Plumbing will still be tricky. it's gonna be kind of tall, unfortunately. I underestimated the size of the pump
15:29 OracleCarbon The phone adds a lot of weight up high.
17:03 rue_mohr2 part of why I jsut start with big robots
17:03 rue_mohr2 my dad is one of the nicest people in the world, 3 days ago he walked into the hospital with a gut problem
17:03 rue_mohr2 he collapsed while there and went into emergency operating
17:04 rue_mohr2 when they opened him up he had major cancer, they couldn't do anything and couldn't put him back togethor
17:05 rue_mohr2 so he's on life support, knocked out on drugs
17:05 rue_mohr2 he had been trying to see a doctor and they couldn't find anything wrong
17:08 Non-ICE i feel you dude
17:19 Non-ICE my GP had cancer 20 years ago...
17:20 Non-ICE he died 2012 from second time cancer... same year my dad got cancer for the first time
17:41 MrCurious rue_mohr2: check PM
17:47 MrCurious my dad died < 1 month ago
17:52 Non-ICE :/
17:53 Non-ICE mine survived, so far, but still isn't working like before after the chemo
17:54 MrCurious best advice, keep things good between you and parents, will save you the specter of regret
17:57 billmania MrCurious: General purpose words of wisdom indeed - live each day as if it were your last and make it count.
17:58 MrCurious yeah, its rare you get opportunity to share a final conversation
17:58 MrCurious same thing when my sister pased back in march
17:58 MrCurious been a not so great year for that for me
18:06 MrCurious to change the topic a bit
18:06 MrCurious bought this last night http://www.ebay.com/itm/141358528163?_trksid=p2059210.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
18:06 MrCurious figure with all i do with steppers, having a quick test device would be worth $7
18:07 Non-ICE and?
18:07 MrCurious and now i wait...
18:07 MrCurious figured a $7 stepper driver/tester might be of use to others here...
18:07 SquirrelCZECH_ looks nice
18:08 MrCurious because sometimes you dont want to drag out a driver, and write some uC code to drive the driver and use a terminal to drive th driver program..
18:08 MrCurious actually, thats EVERY TIME
18:18 rue_mohr2 89C2051? they used an 8051 varriant?
18:25 GargantuaSauce_ pretty sure it's still the most pervasive embedded arch
18:37 rue_mohr2 http://www.ebay.com/itm/Ship-from-German-Warehouse-3-axis-Nema23-stepper-motor-425oz-in-CNC-controller-/190722854683
18:37 rue_mohr2 that almost looks like an ok deal
18:37 rue_mohr2 even with the shippng
18:39 MrCurious i concur
18:40 MrCurious 450oz-in tells me that lead screw will be turning like it or not
18:40 MrCurious or the machine will eat itself first hickup
18:41 MrCurious i am jealous. mine are only 160 oz-in
18:42 rue_mohr2 http://www.ebay.com/itm/2003-2008-Ford-Mustang-Stepper-1pc-motor-MCR-Cluster-Speedometer-Gauges-Temp-/251495379406
18:42 rue_mohr2 I wonder what driver chip that vfd has
18:54 MrCurious dont just explore ebay, remember http://www.aliexpress.com/ often cheaper
18:59 Non-ICE or dx.com
19:11 GuShH http://imgur.com/gallery/PPwidQy
19:13 OracleCarbon Radio Shack had an XBee shield. I was surprised by this.
19:14 OracleCarbon But it makes my life a little easier, I guess.
19:17 OracleCarbon God, Arduinos get huge when you start putting shields on shields.
19:17 Non-ICE thats when you have attinos
19:17 Non-ICE https://hackaday.io/project/1387-Attino
19:18 Non-ICE arduinos have all this unnecessary stuff on them
19:19 Non-ICE poke me if you want some more specs on the attino
19:19 OracleCarbon Eh, maybe in the future or if I take the system airborne.
19:21 OracleCarbon I've just got to figure out the logistics of getting the Arduino block, my phone, the water tank, the pump, and the batteries on the chassis without it getting too big.
19:21 OracleCarbon I don't want it topheavy.
19:22 OracleCarbon This is non-trivial.
19:22 Non-ICE waht are you building?
19:22 OracleCarbon Remote-op water gun carrier.
19:23 OracleCarbon I want to talk to it over the internet using my phone for the onboard vision and comms, connect via wifi to the Arduino (but maybe USB would work better), and then be able to spray someone from the safety of a locked room.
19:25 OracleCarbon Replacing the wifi with USB would drop the Arduino stack down by one shield.
19:25 OracleCarbon But I don't know how to do that at all.
19:25 Non-ICE You know, to get rich you should make your device puch stupid people in the face over the internet.
19:25 OracleCarbon As opposed to this, which I mostly don't know how to do but have an Instructable to help with
19:26 OracleCarbon Punching people isn't in the mission parameters.
19:26 Non-ICE punch*
19:27 OracleCarbon Objective is to wet them, not punch them.
19:27 Non-ICE wet/punch, prolly serves the same mission, question is, how to detect stupid?
19:28 OracleCarbon Stupid is not the target criteria
19:28 OracleCarbon In 4 hours, I will be issued a name and picture of someone in the city.
19:28 OracleCarbon I must find them and spray them with water.
19:28 OracleCarbon Someone has the same on me.
19:28 Non-ICE how big is your city?
19:29 OracleCarbon In the play area, a few hundred thousand I think.
19:45 OracleCarbon God, I cannot get this pin to come off the board. The others desoldered nicely, but this one just wont budge.
19:49 OracleCarbon I swear it's got more solder than it started with.
20:41 OracleCarbon Hmm. The Xbee shield gets rid of the need for the regulator.
20:41 GargantuaSauce_ lol xbee
20:42 OracleCarbon What?
20:42 GargantuaSauce_ overpriced, underpowered
20:42 OracleCarbon I'm using an 802.11 radio that sits on the Xbee's pinout, which is more useful to me
20:43 GargantuaSauce_ mkay
20:43 OracleCarbon But it's still an xbee shield.
20:44 OracleCarbon Though if I can do without that, this gets even better
20:46 OracleCarbon So on chiseling open the super soaker's motor housing, there's a board on the motor with 2 inductors, 4 resistors, and a capacitor. The power wires come out of that, and the motor's connected to the board. What is that likely to be?
20:47 GargantuaSauce_ sure there's no ic?
20:47 OracleCarbon I don't see one, no.
20:48 GargantuaSauce_ any active components at all? other boards in the thing?
20:48 GargantuaSauce_ i cant think of any reason to have just those components
20:49 OracleCarbon Well I know what went over the wires going into that board, so I guess whatever voodoo it does is safe to treat as a black box?
20:50 GargantuaSauce_ is the trigger switch just a switch or is there another pcb behind that?
20:50 OracleCarbon As best I can tell it's just a switch. Maybe I should keep breaking things to make sure.
20:51 OracleCarbon It's a quite usable splashproof 4-AA holder, though. I could power something from it!
20:52 OracleCarbon Or I crack it open and learn its secrets...
20:53 OracleCarbon My hammer is not making progress, though, and I don't have my dremel.
20:56 OracleCarbon I tell you what, they did not skimp on the plastic quality.
21:00 OracleCarbon I think I voided the warranty.
21:01 OracleCarbon Yeah it's just a switch
21:01 acleCarbon harv
21:17 OracleCarbon Can I connect the pump motor through the Arduino's ground pins? I'm powering it off the battery at 7.4-8.4v
21:20 GargantuaSauce_ nah that's a bad idea
21:20 OracleCarbon Just wire it back to - instead?
21:21 GargantuaSauce_ i would use a terminal block or something to distribute power to everything
21:21 GargantuaSauce_ in any case the path for current to flow through the battery and motors should be reasonably beefy
21:21 OracleCarbon It's all pretty low-power, so I was just going to tie it all together on some perfboard
21:22 GargantuaSauce_ that's fine too
21:22 OracleCarbon But then I saw my screw terminal shield has a lot of blank holes in the middle.
21:22 OracleCarbon And some non-blank ones that tie to ground and 5V
21:22 OracleCarbon And I'm guessing I don't want to short the ground pin back to battery -, right?
21:23 GargantuaSauce_ that would likely be fine but it'd be best to use the - on the smps you're powering it with, assuming that's still the case
21:24 GargantuaSauce_ in any case i wouldn't want to sink a bunch of current through the arduino board's ground pin. for one those headers are only rated for 1A I think, and for another in the worst case the board design is not awesome and you'll be pulling its ground up
21:25 GargantuaSauce_ which will make the microcontroller shit the bed and maybe sink a bunch of current through its outputs and so on
21:25 GargantuaSauce_ bad idea in general
21:25 OracleCarbon How do I check the draw of my motor?
21:25 GargantuaSauce_ got a multimeter?
21:25 OracleCarbon Yup
21:25 GargantuaSauce_ put it in current mode (with the probe in the right orifice) and put it in series with the motor and battery
21:25 OracleCarbon Power the motor up and read the curren-yeah
21:26 OracleCarbon I feel like I shouldn't run the pump dry.
21:26 GargantuaSauce_ oh, yes good point
21:26 GargantuaSauce_ many will not do well in that situation
21:26 GargantuaSauce_ i dunno about that little dc one but some will definitely destroy themselves run dry
21:27 OracleCarbon Yeah, I doubt it's an issue since it's a kids toy
21:27 OracleCarbon So it's probably designed to be idiotproof
21:27 GargantuaSauce_ my parents' sump pump let out its magic smoke after the float got stuck on some debris
21:27 OracleCarbon But still
21:34 OracleCarbon I love the realization of not having the proper wire to finish the job well.
21:35 OracleCarbon The whole power system, done properly, should be of the same gauge as the wire coming off the batteries, right?
21:41 GargantuaSauce_ as long as it's appropriate for carrying the current it'll be carrying, doesn't matter
21:42 GargantuaSauce_ what do you have?
21:42 OracleCarbon I've got some 24 AWG wire and a lot of 18 AWG in all the same color, which I'm not a huge fan of
21:43 OracleCarbon Since it's easy to accidentally short the battery that way
21:43 GargantuaSauce_ well 24 for the motors may be pushing it
21:43 GargantuaSauce_ but 18 should be just fine
21:43 GargantuaSauce_ do you have red electrical tape?
21:43 OracleCarbon Nope.
21:43 OracleCarbon Just black.
21:43 OracleCarbon And the wire is black
21:43 GargantuaSauce_ well i'm sure you can come up with something for tagging the positive wires
21:43 GargantuaSauce_ even just a zip tie or something
21:44 OracleCarbon Aha! I have white and blue duct tapes.
21:44 OracleCarbon A little of that on the ends and we're good to go
21:44 GargantuaSauce_ yup
21:46 OracleCarbon My last challenge is going to be figuring out how to best join these batteries. I want to run 2 in parallel since they was $10 cheaper than a double-capacity pack
21:47 OracleCarbon Well, last before trying to make it all work
21:47 OracleCarbon Can you even charge a phone from the Arduino's USB port if the Arduino has power through the straight pin?
21:49 GargantuaSauce_ make sure they're absolutely charged to the same voltage before you connect them in parallel
21:49 OracleCarbon Yeah, I charged them on the same balance charger.
21:50 GargantuaSauce_ check the arduino schematic to see if the usb 5v is actually connected straight to its 5v rail
21:50 GargantuaSauce_ sometimes there's a diode
21:50 OracleCarbon Would drawing the 1A or whatever a phone likes to draw burn out its regulator?
21:52 GargantuaSauce_ okay so you're going to be connecting the arduno's Vin to the battery?
21:53 OracleCarbon Yeah
21:54 GargantuaSauce_ is the usb going to be just for power or do you need to use the serial
21:54 OracleCarbon Just power unless I figure out how to do away with the wireless link and send commands through my phone
21:54 OracleCarbon Which I would like to do, but don't know how to do.
21:54 GargantuaSauce_ i suspect android is going to make that a major pain so yeah
21:55 GargantuaSauce_ maybe bluetooth
21:55 OracleCarbon There's an Android IO board, so I feel like it's not an outlandish thing to want to do
21:55 GargantuaSauce_ anyway
21:55 GargantuaSauce_ time to do some math
21:55 OracleCarbon But it's not something I know how to
21:55 GargantuaSauce_ P=IV
21:55 GargantuaSauce_ 1-1.5A would be a good upper bound for charging
21:55 GargantuaSauce_ the voltage is the difference between Vbat and 5V
21:56 GargantuaSauce_ without a heatsink i am going to guess the dissipation of the regulator is max 1W
21:56 GargantuaSauce_ if that
21:57 OracleCarbon It goes into thermal shutdown at 300mA at 12V, apparnetly
21:57 GargantuaSauce_ yeah so definitely not gonna work out
21:57 GargantuaSauce_ cut up a micro usb cable and power the phone from the smps
21:57 OracleCarbon That adds so much more complexity. 4 components and like...8 wires!
21:58 GargantuaSauce_ well you're going to want to power servos with the same smps eventually anyway right
21:58 OracleCarbon Maybe not.
21:58 OracleCarbon I'm down to one very small servo
21:59 OracleCarbon Moving a small mirror
21:59 OracleCarbon And maybe not even that, depending on how the vehicle geometry works out
21:59 GargantuaSauce_ welll in any case you definitely can't charge the phone from the linear regulator
21:59 OracleCarbon Yeah. That's unfortunate.
21:59 OracleCarbon I wonder how long my phone can run a Google Hangout for.
21:59 GargantuaSauce_ even if you could it'd be pretty dumb
22:00 OracleCarbon Yeah, but it's the simplest approach
22:00 OracleCarbon Which means the fewest components for me to fundamentally misunderstand the usage of
22:00 GargantuaSauce_ hehe
22:00 GargantuaSauce_ and thats not even getting into their theory of operation :)
22:01 OracleCarbon To this day I have never been able to get a transistor to consistently operate. It's a problem.
22:01 GargantuaSauce_ also i feel the need to re-emphasize, make sure you measure the voltage of both batteries before putting them in parallel
22:01 GargantuaSauce_ they should be within .1V or bad things may happen
22:02 OracleCarbon Yeah, I'm hesitant to do it without a switching circuit to draw from the high one if they're imbalanced.
22:02 GargantuaSauce_ you could just use a pair of schottky diodes also
22:02 OracleCarbon And if my phone can't survive onboard long enough to run through both anyways, it might be better for me to just run one.
22:02 GargantuaSauce_ .3V drop is tolerable
22:02 GargantuaSauce_ actually what's the max input voltage of the smps again?
22:03 OracleCarbon Uhh...15 or so, I think
22:03 GargantuaSauce_ could put them in series and power all the electronics with that
22:03 GargantuaSauce_ ah guess not
22:03 OracleCarbon Let me check.
22:03 OracleCarbon Oh, 3
22:03 OracleCarbon 36*
22:03 GargantuaSauce_ that may be the best course of action then
22:03 GargantuaSauce_ assuming the motor drivers also can handle it
22:04 OracleCarbon Unless I also power the motors off the smps, I'm going to burn them up
22:04 GargantuaSauce_ just don't run them at a 100% duty cycle
22:04 GargantuaSauce_ max it out at 25% or whatever the ratio of their power rating to the supply is
22:04 GargantuaSauce_ and it'll be just fine
22:04 GargantuaSauce_ with a reasonable switching frequency
22:05 OracleCarbon I know what those words mean, but not how to do it
22:05 GargantuaSauce_ whatever arduino library you're using for pwm
22:06 GargantuaSauce_ just multiply the value you pass to it by that ratio
22:06 OracleCarbon Also my pump motor is just using http://playwithrobots.com/images/one%20transistor%20motor.jpg
22:06 OracleCarbon I don't know the switching speed on the transistor
22:07 GargantuaSauce_ no oscilloscope? :(
22:07 OracleCarbon Nope.
22:07 GargantuaSauce_ whatever library you're using should have some indication of it in the documentation
22:07 GargantuaSauce_ i say should but i acknowledge that the docs for most of them are fucking garbage
22:08 OracleCarbon I have no idea what libraries I'm using. I'm just copying code out of an instructable
22:08 GargantuaSauce_ it's probably high enough in any case
22:08 OracleCarbon I also want a consistent stream out of the pump
22:09 GargantuaSauce_ that's not feasible without current control which is probably not something you can figure out in a day or two or whatever time you have left
22:09 OracleCarbon I wonder if that's what the circuit on the pump motor does.
22:10 GargantuaSauce_ nope
22:10 OracleCarbon Some sort of a filter to smooth the power to it
22:10 OracleCarbon Oh
22:10 OracleCarbon Okay
22:10 GargantuaSauce_ that's just a low side switch
22:10 GargantuaSauce_ you could try measuring Vbat and increasing the duty cycle as it decreases i guess, but it will be pretty rudimentary
22:10 OracleCarbon Vbat won't change all that much in the few seconds the pump runs
22:11 GargantuaSauce_ well it will definitely decrease over the course of the robot's runtime right
22:11 GargantuaSauce_ and that's what you want to compensate for to get a consistent stream
22:11 OracleCarbon True, but I'm honestly not sure it matters.
22:11 OracleCarbon At 6V, the pump shoots a good 25 feet fairly straight, apparently.
22:12 OracleCarbon I'm running it at 8
22:12 OracleCarbon I figure wind and obstructions will mess me up before that matters
22:12 OracleCarbon I just don't want it starting and stopping
22:13 GargantuaSauce_ oh, well i'm sure the pwm library you're using switches at at least a few hundred Hz so that certainly won't be an issue
22:14 GargantuaSauce_ man i have a $500 ebay cart all of a sudden
22:14 GargantuaSauce_ and i feel like i'm still missing something
22:14 OracleCarbon Oh, the transistor's bag has a figure "f_T: 3 MHz"
22:14 rue_mohr2 its your wallet, its too light now
22:14 OracleCarbon Is that the switching frequency?
22:15 GargantuaSauce_ well it's not the transistor that matters, it's the software controlling the arduino's output
22:15 GargantuaSauce_ i'm sure it's high enough
22:15 GargantuaSauce_ just make sure the duty cycle is decreased accordingly if you double the battery voltage
22:16 OracleCarbon I guess I can assume that the pump was okay for 100% duty cycle at 6V, since it was probably made so you could just hold the trigger?
22:16 GargantuaSauce_ yeah, though i am a little suspicious about the electronics you described
22:17 OracleCarbon The problem I run into with the drive motors operating like that is that apparently their driver has either forward, reverse, or brake.
22:17 OracleCarbon So I feel like I'd have a lot higher current draw needed
22:18 GargantuaSauce_ it is dumb that those h bridges have no floating output mode but it should probably be ok
22:19 GargantuaSauce_ and variable speed is definitely a requirement so it's sort of a moot point
22:19 GargantuaSauce_ get a better driver in the future but for now it'll do
22:19 OracleCarbon True
22:20 GargantuaSauce_ oh also have you tested the latency of google hangouts / skype on your phone?
22:20 OracleCarbon So off one of those lipos, I get about 40 minutes of drive time worst case scenario, and I feel like it's unlikely I'll be driving constantly.
22:20 OracleCarbon No, that's a point I hadn't considered
22:21 GargantuaSauce_ yeah the phone's radio being on constantly will probably be the most significant load overall
22:21 OracleCarbon Yeah
22:21 OracleCarbon I know, for example, it can take about 2 and a half to three hours of Netflix
22:21 OracleCarbon From full charge
22:21 OracleCarbon Which is screen plus radio
22:22 OracleCarbon Is it reasonable to assume I'd get 90-120 minutes of video call time?
22:22 GargantuaSauce_ i would assume the video encoding will be about as energy intensive
22:22 OracleCarbon Because if that's the case, I can just let my phone run off its own battery.
22:22 GargantuaSauce_ yeah perhaps
22:22 OracleCarbon At the cost of not being able to swap batteries and immediately return to service
22:24 OracleCarbon Though having the phone outlive the system is a good idea from the standpoint of being able to monitor the system as I go to recover it. Except I can't...because I don't have my phone to watch it on
22:24 GargantuaSauce_ uhhh
22:24 OracleCarbon I actually didn't think this through very well at all.
22:24 GargantuaSauce_ so what is going to be on the other end of the network
22:25 OracleCarbon A laptop. Which is where it all goes to hell because either I have to sit right next to the robot, or it connects over the internet and I have to be offsite at an internet-enabled location
22:26 GargantuaSauce_ mmmm
22:26 OracleCarbon This was all much cooler the first time I had the idea years ago, where its use would be on campus, which is all the same network
22:26 OracleCarbon Rather than now, where its use is...city-wide, which is less good.
22:27 GargantuaSauce_ so you plan on using 3g on the phone and using it as a wifi bridge for the arduino's wifi shield to connect to?
22:27 OracleCarbon Yes
22:28 GargantuaSauce_ that's gonna be..something
22:28 GargantuaSauce_ have you tested that configuration at all?
22:28 GargantuaSauce_ i would probably prioritize that over getting the power system together
22:29 OracleCarbon If any subsystem doesn't work, the whole project has to be retooled or retasked, so I figured I'd cobble it all together and go from there.
22:29 GargantuaSauce_ hehe okay :)
22:29 OracleCarbon I guess the other choice I have is to bring my home router with me, which is much stronger, have it all talk over that, and network it from my car.
22:29 OracleCarbon Then sit down the street.
22:29 GargantuaSauce_ that sounds like a marginally smaller clusterfuck. good that you have a fallback
22:30 OracleCarbon I also don't know how to use this system to good effect at all. It's more of a shock and awe thing to post to the game website.
22:31 OracleCarbon Really I thought none of this through beyond "You know what'd be cool? Robotic warfare."
22:31 GargantuaSauce_ heh
22:31 OracleCarbon And then resurrected an old plan for a very different mission.
22:32 OracleCarbon A quadcopter-mounted Nerf gun firing and surveilance platform for a campus-wide game that I could operate safely from inside campus buildings. All the same network, all nearby, no need to deploy it on a specific person
22:33 OracleCarbon Also not at all okay from a safety and legality standpoint, but I was foolish in those days
22:33 GargantuaSauce_ considering that you'd need to change the battery every 10 minutes, being indoors sounds pretty silly
22:33 OracleCarbon Well, on a balcony away from the action.
22:33 GargantuaSauce_ and also wifi is fucking awful, i would not entrust the integrity of my quad to it
22:34 OracleCarbon Oh, I planned to control the quad via normal RC aircraft means
22:34 OracleCarbon Which I'm starting to think could serve me better here.
22:34 GargantuaSauce_ yeah quite possibly
22:35 OracleCarbon I've got a 2.4GHz 4ch transmitter-receiver pair
22:35 GargantuaSauce_ power?
22:35 GargantuaSauce_ you'll probably need line of sight unless it's a watt or two
22:35 OracleCarbon Uhh...it's from my old RC foamie, it powered off the BEC in the ESC
22:35 OracleCarbon Oh, radio power
22:35 OracleCarbon Yeah, I'll need to hang around the area
22:36 OracleCarbon But I can get low in the back of my car and watch on the monitor.
22:36 OracleCarbon And that cuts out all the wifi crap
22:36 OracleCarbon It's an SUV, I've got space to set up in there
22:36 GargantuaSauce_ assuming you do use the wifi, what are you going to do for control software on the laptop
22:37 OracleCarbon The instuctable on a wifi-controlled Arduino-driven robot has code
22:37 GargantuaSauce_ so copypaste and pray? :)
22:37 OracleCarbon That's every bit of my software plan.
22:37 GargantuaSauce_ i see
22:38 GargantuaSauce_ i am very much of the write-everything-from-scratch school of thought
22:38 GargantuaSauce_ and that's not ENTIRELY grounded in elitism
22:38 GargantuaSauce_ but i guess if it works it works
22:38 OracleCarbon I tend to be, but this project is beyond my depth and on a short timetable
22:41 OracleCarbon I assume someone's made a tankbot that takes PWM inputs and turns them into outputs to the motors
22:41 OracleCarbon Which should give me somewhere to work from if I use the RC system
22:41 Loshki There just isn't time nowadays to write it all from scratch. Not when you add debugging time to all that. Libraries aren't perfect, but as labor saving devices, they beat sliced bread...
22:42 GargantuaSauce_ i dunno every embedded library i've used has been as much trouble as doing the work myself
22:42 GargantuaSauce_ i haven't done IP stuff yet though
22:43 OracleCarbon The internet-enabled idea would, in theory, work. I would just need an accomplice with their own phone to let me connect in the field, or an accomplice to power up the system and set it on the ground while I sit in a coffee shop somewhere, or at home, or in the office
22:43 OracleCarbon And I don't know anyone with unlimited data, so I'd feel bad
22:44 GargantuaSauce_ i would be worried about things like losing connectivity for a moment and the hangout ending and it requiring user action to rejoin it
22:44 OracleCarbon Yeah, that would be less than ideal
22:44 GargantuaSauce_ maybe it has an option to automatically accept requests though
22:44 OracleCarbon I think Skype might.
22:46 Loshki I admit, quality of libraries is an issue. Just look at what happened to Perl. But it's still true that if (for whatever reason) you're forced to write everything from scratch, it hammers your productivity. But then, so do libraries that don't work the way they should. It's probably intractable.
22:46 OracleCarbon I may have been too eager to break in my new workbench with a project that would let me flex my technological superiority/epeen in a competitive arena.
22:47 OracleCarbon Loshki: Just make everyone who programs program well and in the same way.
22:48 GargantuaSauce_ Loshki: and i am not just talking about trivial stuff. https://www.youtube.com/watch?v=2vBi6LNklRg kinematics, packet radio comms, 24 servo control. no way i would have gotten this far if all the code was from third parties
22:49 GargantuaSauce_ well. i guess i could have spent $40 on one of those servo control boards that are just a big microcontroller with closed firmware
22:49 GargantuaSauce_ but fuck that
22:49 Loshki OracleCarbon: well, that's why python syntax is so constrained I suppose. To minimize differences in programming style.
22:50 OracleCarbon GargantuaSauce_: I don't know, I've got plenty of projects that I tabled upon exiting my area of competence, and then never returned to because their fun wore off before I got good enough to finish them.
22:50 GargantuaSauce_ i guess my attitude is just different
22:51 GargantuaSauce_ the fun is IN being outside my area of competence
22:51 GargantuaSauce_ as soon as i fully understand a project is when i table it <_<
22:51 OracleCarbon Oh, absolutely. I tend to bite off more than I can chew, though
22:51 OracleCarbon It's not "This is something I kind of get, but need to look into more."
22:51 OracleCarbon It's "I have no idea what I'm doing here."
22:51 GargantuaSauce_ baby steps i guess
22:52 Loshki GargantuaSauce_: How odd. I expected to hear you say you couldn't have gotten this far *without* libraries. Other people's code always sucks, you know...
22:52 GargantuaSauce_ the only third party code involved in that video is the operating system, sdl, opengl
22:52 GargantuaSauce_ "only" :)
22:53 GargantuaSauce_ mine always sucks too but knowing how it sucks is the key i think
22:53 GargantuaSauce_ and i find usually learning how a library sucks is just as hard as writing my own code that sucks
22:53 GargantuaSauce_ but then again software is my area of expertise
22:54 acleCarbon is not good at soft
22:58 OracleCarbon I can kluge it up well enough to make things work sometimes (except all of my experience with things that talk to hardware)
22:58 OracleCarbon But it's never pretty
22:58 GargantuaSauce_ and i am not trying to brag or anything. the whole idea of just mashing together a few libraries or code examples written by other people and expecting stuff to work just strikes me as a losing battle
22:58 Loshki GargantuaSauce_: I dunno. But if the answer to the software quality problem is we all write our own from scratch, I think I'll just kill myself now...
22:59 OracleCarbon It's the eternal struggle between good and good enough.
22:59 GargantuaSauce_ welll everyone is welcome to come to their own conclusions of course
22:59 GargantuaSauce_ that's just mine
23:00 GargantuaSauce_ and tbh it's not just software
23:00 GargantuaSauce_ i want my robots to have electronics that are all designed and built by me also
23:01 OracleCarbon GargantuaSauce_: I'm like that with some projects as a pride thing.
23:01 OracleCarbon "That? I did that. Not just putting together other things, ALL OF THAT."
23:01 GargantuaSauce_ and eventually mechanical structures, though i need to get a new living space that has a good workspace first
23:01 GargantuaSauce_ so i can fill it with machine tools
23:01 GargantuaSauce_ it's not about pride for me, it's about being able to manipulate EVERY design parameter to optimize for arbitrary results
23:01 ShH pats GargantuaSa
23:02 OracleCarbon I like to keep the budget down, so I tend to make things for modularity
23:02 GargantuaSauce_ GuShH: you should move up here and be my pet machinist
23:02 GuShH for a price!
23:02 GargantuaSauce_ name it
23:03 GuShH two million schmeckles!
23:03 OracleCarbon Room, board, a competitive industry-based salary, and a bonus for putting up with GargantuaSauce?
23:03 ShH wonders when is Rick & Morty coming
23:03 GargantuaSauce_ pretty soon!
23:04 GuShH There are very few things I love... that show is one of them
23:04 GargantuaSauce_ ya it's pretty fantasic
23:04 GuShH pass the butter!
23:04 GargantuaSauce_ but yeah i think it'll be this winter
23:05 GargantuaSauce_ or summer for you bizarre people on the wrong side of the equator
23:05 GuShH haha
23:10 Loshki Well, it bothers me that you're not wrong. I've had to rewrite perl libraries due to bugs. But maybe the real difference is "the whole idea of just mashing together a few libraries or code examples written by other people and making it work" is high art to me. Standing on the shoulders of giants etc...
23:12 OracleCarbon Chromecasts are so nitpicky about their networks.
23:14 GargantuaSauce_ yeah this is mostly an opinion steeped in the fact that i have a reasonable amount of experience writing pretty complex and high-performance software
23:15 GargantuaSauce_ so it's all the hardware stuff that i have a harder time with
23:15 GargantuaSauce_ working on that though. EE is pretty fascinating
23:16 GargantuaSauce_ and i have no eye for mechanical design at this point at all which is frustrating as fuck
23:19 GargantuaSauce_ and i guess i only really am anti-library in the embedded world
23:19 GargantuaSauce_ where behaviour, timing, and resources are of extreme importance
23:20 GargantuaSauce_ for web or some desktop stuff i am quite happy to mash some stuff together
23:20 GargantuaSauce_ the appserver for the system i'm doing profesionally right now uses nodejs after all