Back
[00:03:04] <Triffid_Hunter> sounds like a nightmare for heat dissipation
[00:03:15] <anniepoo> Yes, it's going to be
[00:03:32] <anniepoo> and to be honest, I haven't figured out a good solution for it
[00:04:10] <anniepoo> the robots will be sessile, so that keeps the power way down
[00:05:26] <Triffid_Hunter> sessile? they don't move?
[00:05:49] <anniepoo> they can't walk, not
[00:05:51] <anniepoo> they can't walk, no
[00:06:01] <anniepoo> the legs are just decorative
[00:07:31] <anniepoo> there are some high tech high thermal conductivity polymers, but cost is an issue
[00:08:35] <anniepoo> the arm articulation's pretty crude. The arm's essentially an underactuated finger.
[00:08:46] <anniepoo> they're social robots, not MSD robots
[00:09:10] <anniepoo> think 'doll'
[00:09:25] <anniepoo> the head moves, as do the eyes
[00:09:47] <anniepoo> and the mouth
[00:10:56] <anniepoo> And I'd love to have a channel for facial distortion, but I'm not sure how that'll work or if we can do it.
[00:18:43] <anniepoo> Triffid - any thoughts on solving the heat dissapation problem?
[00:19:15] <Triffid_Hunter> anniepoo: if you're wrapping the thing in insulating materials, the only option is to make as little as possible in the first place
[00:20:07] <anniepoo> Yes, and of course we'll do that.
[00:22:35] <Triffid_Hunter> anniepoo: put a temperature sensor in it, make it sluggish or sleepy or something when hot
[00:24:45] <anniepoo> hmm.. yes, the majority of power is the 'idle animation' (the robot has to move occasionally to seem alive)
[00:25:24] <anniepoo> but that's a brilliant idea - I can have the robot do less when it's power is running down or it's hot
[00:26:00] <Triffid_Hunter> as it's a social robot, giving it needs that reflect actual physical conditions should work wonderfully
[00:27:23] <anniepoo> Yes. On the other hand, when the teacher needs the robot to move, that's a mission critical need, it's gotta happen.
[00:29:39] <Triffid_Hunter> then teach them to grumble :P
[00:32:01] <anniepoo> 8cD sometimes they'll misbehave
[00:32:12] <anniepoo> (under teacher control)
[01:04:44] <anniepoo> Triffid - the 'getting sleepy' thing is also a great way to signal the teacher that the robot's batteries are going
[01:05:35] <Triffid_Hunter> anniepoo: I like it most because it's not artificial at all, what better literal interpretation of low batteries is sleepiness? humans even use battery references when discussing their state of tiredness ;)
[01:06:14] <anniepoo> lol - late in the day, been using the robots all day, you've got a classroom of bored children (who may be disengaged or may be hyperactive) and a bunch of nodding off robots
[01:06:15] <anniepoo> LOL
[01:06:19] <anniepoo> Yes, I like it
[01:14:20] <anniepoo> keeping the teacher's cognitive load down is important
[01:30:24] <anniepoo> anyway, gnite all
[01:30:33] <anniepoo> thanks for the useful conversation
[04:39:40] <Jak_o_Shadows> rue, I don't suppose you've got an example of a timer toggling a pin for the stm32 doyou?
[05:36:33] <Triffid_Hunter> Jak_o_Shadows: dunno about stm32 specifically but usually you have to flip some mode bits for the pin so it takes commands from the timer compare block (and have the timer set up obviously)
[05:37:58] <Jak_o_Shadows> Yeah. I have the timer configured, and one channel is firing off an interrupt. I'm trying to get a different channel to toggle a pin, and I THINK I've set all those bits
[05:38:18] <Jak_o_Shadows> But I dunno, and cause it's all hardware, I can't get in there with gdb to see WHERE it's failing
[05:39:24] <Triffid_Hunter> Jak_o_Shadows: with these bigger chips, the GPIO routing block is totally separate to the timer, the bits you need may be described in an earlier chapter in the datasheet
[05:40:40] <Jak_o_Shadows> Those routing bits are actually pretty simple I think. I'm more concerned about the timer side of things.
[05:41:42] <Triffid_Hunter> Jak_o_Shadows: well flip a GPIO from an interrupt to verify it, then play with the routing once that works
[07:32:25] <veverak> http://www.postavrobota.cz/fotky46704/fotos/_vyr_3320160514_115244.jpg
[07:32:26] <veverak> hmm
[07:32:31] <veverak> nice little small stepper in local shop
[07:32:45] <veverak> around 4$
[08:37:52] <Snert_> Kinda intersting...SPI HBridge
[08:37:58] <Snert_> http://www.infineon.com/dgdl/Infineon-H-Bridge_Kit_2Go_UM-UM-v01_01-EN.pdf?fileId=5546d46250cc1fdf015133ee879d3696
[08:40:40] <Snert_> http://www.infineon.com/cms/en/product/productType.html?productType=5546d46250cc1fdf015124e027823420
[08:42:02] <Snert_> it can push 7amps.
[08:44:48] <Snert_> 7amps for 7bucks
[08:45:01] <Snert_> no, 14bucks
[08:45:06] <Snert_> http://www.digikey.com/product-detail/en/infineon-technologies/HBRIDGEKIT2GOTOBO1/HBRIDGEKIT2GOTOBO1-ND/5878318
[08:50:31] <SpeedEvil> Consider also second sources
[08:50:43] <SpeedEvil> if the part goes away , will you be able to repair in 4 years
[08:50:46] <SpeedEvil> - if you care about that
[08:51:17] <Snert_> I'm looking to mainly be able to heist the design lol.
[08:52:06] <Snert_> do you mean that driver chip has potential to become non-existant?
[08:52:13] <SpeedEvil> yes
[08:55:50] <Snert_> maybe infineon has a track record of vaporware and dissapointment too.
[08:56:23] <SpeedEvil> I'm not saying vapourware - just that chips of this type do go away
[08:56:34] <SpeedEvil> If it diddn't sell as well as tey hoped, ...
[08:57:04] <Snert_> why do you say this type? HBridges must be smaller volume.
[08:57:21] <Snert_> you must mean specialty chips.
[08:58:10] <SpeedEvil> yes
[08:58:38] <jhylands> Snert_, what do you plan on using to control this board?
[08:58:49] <SpeedEvil> It is 100% certain I'll be able to buy a gate driver in 20 years.
[08:58:54] <SpeedEvil> And MOSFETs
[08:59:08] <SpeedEvil> (or something that will work comparably)
[08:59:15] <Snert_> I'm leaning towards bigass mosfets on a compact board anyways.
[08:59:24] <veverak> just can't find this small fittings
http://www.hobbyking.com/hobbyking/store/catalog/WaterPump.jpg
[08:59:50] <veverak> GOT THEM
[08:59:52] <veverak> damn it
[08:59:54] <veverak> as always
[09:00:11] <SpeedEvil> Seperate FETs give an enormous robustness and repairablitiy
[09:00:15] <Snert_> jhylands: just an arduino of some flavor. That 'ticular board may require the infineon wares.
[09:00:48] <jhylands> are you planning on driving a lot of motors?
[09:01:01] <Snert_> I'm not opposed to switching to something else at the appropriate time like stm32
[09:01:17] <Snert_> not alot. 3 or 4
[09:01:46] <jhylands> it seems silly to use SPI to control a motor - most modern microcontrollers have hardware PWM
[09:03:48] <Snert_> drawbacks and considerations. One choice bars your entry to other choices.
[09:04:18] <jhylands> I don't understand
[09:04:31] <Snert_> if it can handle 7amps and be super tiny then the programming language changes.
[09:04:38] <Snert_> 2 out of 3.
[09:06:15] <jhylands> are you talking about just the hbridge chip, or the whole board?
[09:06:47] <Snert_> I'd be buying the whole board for my own purposes.
[09:07:20] <jhylands> that's a big board to drive only one motor
[09:08:41] <jhylands> the micro-controller half of the board could easily drive 5-10 different motors
[09:09:23] <jhylands> I guess if you made your own board with a few copies of that hdbridge chip, and one microcontroller
[09:09:58] <Snert_> it's just something I googled across. Maybe AOE has a good hbridge that I can adapt.
[09:10:19] <Snert_> there a good point about using 4 discrete mosfets.
[09:12:03] <Snert_> at first, I simply would not do my own pcb. Only later if I decided I needed abunch of them.
[09:12:26] <Snert_> I'm sure it's not too difficult.
[20:19:01] <theBear> don't make such bold and absolute statements like that here Snert ! you'll have me lying awake all night until i give up and try to proove that whatever "it" is, is FAR from difficult
[20:19:44] <orlock> heh
[20:19:59] <orlock> heya theBear
[20:20:01] <orlock> hows things?
[20:25:04] <theBear> awlays bustin balls with the tricky questions my way eh ? mmmeh well in defiance of this tyranny and the relenteless and target manner it seems to apprach only meeeMAAAAAAN, for no real reason, and nearly fogotteen already,)i will say cos i'm feeling quietly confident that we watchin storms go floaty away, and ain't in any sneaky eye of a perfect hurrimicane, errm, yep, that's what i say, that i lazily and poorly just roo-eeeeeeened
[20:49:25] <anonnumberanon> Start again without the 7 drinks?
[20:49:47] * anonnumberanon self-righteously takes a shot of water
[20:50:14] <anonnumberanon> here theBear this should be a good (new) read:
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160005963.pdf
[20:58:07] <theBear> wouldn't be a start if i did that, it'd just be err, a boring wait before work
[21:39:42] <anniepoo> I'm POOPED
[21:40:42] <Tom_itx> no just poo
[21:44:20] <anniepoo> true
[21:44:31] <anniepoo> that name follows me.
[21:48:56] <anniepoo> I use SolidWorks already. Is it worth it to figure out the dynamic simulation tools and generate animations and get the needed motions from blender to SolidWorks, then dynamically model the animatronic?
[21:49:09] <anniepoo> Or is that going overboard?
[21:49:44] <Tom_itx> sw can animate to an extent
[21:50:21] <anniepoo> good to know. I haven't looked at all at the dynamic simulation or FEM stuff
[21:51:17] <anniepoo> I'm just wondering if it'd be better to make some manual calculations and guess, which might lead to much suffering with refitting larger servos, or to suffer through all the simulation
[21:51:42] <Tom_itx> seems you're suffering either way you go :)
[21:51:44] <anniepoo> sounds like you've done some of this. Is this much learning curve?
[21:51:58] <anniepoo> yes, a path diverges in the road
[21:52:07] <anniepoo> one side leads to the Mines of Moria
[21:52:10] <Tom_itx> i have done nothing of the sort but i know sw to an extent and catia
[21:56:15] <anniepoo> the other to the Misty Mountains
[21:59:48] <anniepoo> is there a third road?
[22:00:20] <anniepoo> I suppose there's Gazebo
[22:00:48] <anniepoo> though I have to wonder about using something like bullet for this
[22:07:45] <anniepoo> 8c/ well, the simulation doesn't look too intimidating, based on an advert type video
[22:09:12] <anniepoo> okey dokey, I'm pooped, so I'm headed home to a nice bath
[22:09:33] <anniepoo> one that does NOT involve mechanics hand cleaner or using paint thinner on my hands
[22:40:50] <theBear> when i mechanic it up these days i use the same degreaser or whatever was involved in getting the black onto me
[22:42:15] <theBear> also surprisingly citrus based hand cleaney things that come from commercial cleaning supplies kinda places in big (for handwash anyway) tubs and don't got any hippy silliness written on there explaining that dolphins are fine with it, but it doesn't actually do much of anything now they fixed that
[22:42:38] <theBear> that's what every pushbike back room i ever seen/worked in uses
[22:42:43] <orlock> Solvol
[22:42:46] <orlock> thats what i use
[22:42:53] <orlock> which is basically soap with grit in it
[22:43:32] <theBear> yeah, that's a fine example (it DOES have a little lemon or some crap in there for reasons i may never appreciate, doesn't it ?)
[22:43:52] <orlock> yeah, i think you are right
[22:44:16] <theBear> heh, i'm impressed how little hesitation that took to say, i appreciate your confidence !
[22:44:35] <orlock> faint aroma of urinal cakes
[23:12:01] <Snert_> (hmmmm. my deodorant is wearing off)
[23:44:27] <theBear> out of context ambiguous statemen, err, series of words
[23:44:38] * theBear feels like his a part of the fitting-in
[23:49:07] * Anniepoo_ files a bit off of theBear so he fits
[23:49:30] <z64555> ew. bear shavings
[23:49:41] * z64555 sics the roomba on em
[23:49:43] <Anniepoo_> bear swarf!
[23:53:29] <theBear> umm, rue strafe!
[23:53:42] <theBear> everyone, PARLAIS ! for the love of god make it quick !
