#robotics Logs

Aug 15 2015

#robotics Calendar


00:01 codepython777 Hyratel1: how much did you run Hormes? How many miles?
00:01 Hyratel1 well its range is around half a mile on a single charge
00:01 Hyratel1 and we just goof around a lot with it
00:01 Hyratel1 and there's no odometer
00:02 codepython777 I need 4 hours of battery life- so thats another pain that i'll have to solve.
00:02 Hyratel1 wrong measure
00:02 codepython777 ?
00:02 Hyratel1 battery life is a function of loading
00:02 Hyratel1 and for a bot like yours, would be properly measured in distance, not time
00:03 codepython777 ah i see
00:03 JEntrep so many idiots on the internet
00:03 codepython777 indeed :)
00:04 JEntrep I am genuinely getting tired of humanity
00:04 Hyratel1 codepython777, Hormes is only 18 inches rail to rail front to back
00:04 JEntrep you know how there are activists for anti-nuclear activity?
00:04 JEntrep I think I am going to go join a pro-nuclear activists group
00:04 codepython777 Hyratel1: it looks bigger in that picture - i was thinking bigger
00:04 JEntrep the majority of humanity blowing themselves up just seems ... like not such a bad thing
00:04 Hyratel1 it's still pretty big
00:05 JEntrep I mean really ... think about it.
00:05 Hyratel1 JEntrep, ROBOTS
00:05 codepython777 Hyratel1: so the holes on your rectagular metal frame are not even one inch part?
00:05 ratel1 distract
00:05 JEntrep nuclear robots?
00:05 JEntrep probably a good thing to research
00:06 Hyratel1 oh, d'uh. the side rails are 18" long
00:07 Hyratel1 so it's 20" to the outside of the rails
00:07 Hyratel1 it's been a while since I messed with it
00:07 codepython777 Hyratel1: what was the spec on the motor controller? how many amps? voltage?
00:08 codepython777 2 or 1?
00:08 Hyratel1 codepython777, 40A, one per side
00:08 Hyratel1 http://www.andymark.com/Victor-SP-p/am-2855.htm
00:09 codepython777 how do you supply pwm to it?
00:10 Hyratel1 by a PWM cable
00:10 codepython777 :)
00:11 codepython777 could you use something like i linked? Of course you would need 2 x 60A
00:11 Hyratel1 http://www.andymark.com/product-p/am-0975.htm
00:11 JEntrep what is AndyMark?
00:11 Hyratel1 why
00:12 JEntrep is it a website for small robotics?
00:12 codepython777 Hyratel1: it can control 2 motors?
00:12 JEntrep small-scale
00:12 Hyratel1 codepython777, sssso?
00:12 Hyratel1 it fails, both sides die
00:13 codepython777 its not great if one side dies :)
00:13 Hyratel1 and you're out $125
00:13 codepython777 that is a good point :)
00:13 codepython777 you are only out $60 :)
00:13 Hyratel1 one Victor dies (which I have never heard of happen), you're out $60
00:13 Hyratel1 now, clarify for me
00:13 Hyratel1 what are you doing
00:14 codepython777 I'm trying to design a truck for a project, which can carry around a payload
00:14 Hyratel1 vague
00:14 codepython777 I cant be precise - behind an NDA - hence vague
00:14 codepython777 except for the payload, i can be precise
00:14 Hyratel1 you need to have a "Thesis statement" for your build, which should effectively describe the entire project
00:15 Hyratel1 while still staying safe with your NDA
00:15 Hyratel1 Indoor? (carpet, hardwood, smooth concrete? nonmarking tires?)
00:15 codepython777 outdoors
00:15 codepython777 grass and road
00:16 Hyratel1 outdoor? smooth concrete, rough concrete, gravel, cobblestone, negotiate curbs?
00:16 Hyratel1 Offroad? Grass, dirt, leaf litter, tree debris?
00:16 codepython777 offroad / onroad
00:16 Hyratel1 payload (as described)
00:17 Hyratel1 for a runtime of (time) at (how fast)
00:17 Hyratel1 including, or not-including Loiter time
00:17 codepython777 4 hours, 1 m /sec
00:17 Hyratel1 2mph
00:17 codepython777 k :)
00:17 Hyratel1 low speed, crawling
00:18 Hyratel1 hormes will do 10MPH
00:18 codepython777 nice
00:18 Hyratel1 but it has limited radio range
00:19 Hyratel1 10mph is trivial
00:19 codepython777 my catch is payload + time - being cheap
00:19 codepython777 still need to be reliable
00:19 Hyratel1 you're going to come up short on 'being cheap"
00:19 Hyratel1 HOWEVER
00:20 Hyratel1 there is a lot of information available out there for your size range
00:20 Hyratel1 you could derive the Hormes platform
00:20 codepython777 cheap = not paying $42 for motor mount, when i pay $60 for motors
00:20 codepython777 that is a no no
00:20 Hyratel1 why
00:21 codepython777 I'd rather pay $100 on motors and $5 on motor mounts :)
00:21 Hyratel1 the parts come to $600 for Hormes and it is a very sturdy design
00:21 Hyratel1 oh. thought you were saying "no" to deriving Hormes
00:21 Hyratel1 http://www.andymark.com/CIMplebox-p/am-0734.htm
00:22 Hyratel1 4.67:1 into a 4:1 chain drive
00:22 Hyratel1 251 max RPM on the wheels
00:22 codepython777 Hyratel1: Problems I see with Hormes: I want something easy to assemble. Seems like you put a lot of effort in assembly. I've to assemble > 5 of these
00:23 Hyratel1 the only "bespoke" parts are the corner brackets
00:23 Hyratel1 and the battery bracket
00:23 Hyratel1 the roll cage is assembled from scavenged practice bends of galvanised conduit
00:24 Hyratel1 behind a machine shop
00:24 Hyratel1 the frame is square steel tubing, avail from big box hardware
00:24 codepython777 Hyratel1: anything easier i can do? I think Hormes can carry more than 20lbs, right?
00:25 Hyratel1 the axles are 5/8" solid steel, anchored with U-bolts through the frame holes
00:25 Hyratel1 certainly
00:25 Hyratel1 but the tire tread, while OK for mild offroad, is NOT very luggy
00:25 Hyratel1 it just skids on wet
00:25 codepython777 i see
00:26 codepython777 i guess AT tires are more expensive
00:26 Hyratel1 the sprockets are http://www.andymark.com/ProductDetails.asp?ProductCode=am-0736
00:26 Hyratel1 AT?
00:27 codepython777 all terrain
00:27 choonway suspension required?
00:28 codepython777 choonway: would be nice , but not necessary - ease of assembly takes priority
00:32 Hyratel1 if you have 4wd and a "bias" in your balance, you'll get stable driving on a solid frame
00:33 Hyratel1 codepython777, does your "Client" have a listed budget?
00:33 Hyratel1 or are you going to submit a proposal to have it confirmed or denied
00:33 codepython777 Hyratel1: nope - But want to spend <$1k on it
00:34 Hyratel1 you can do $600 per if you derive Hormes
00:34 Hyratel1 less if your client is providing the Control Source themself
00:35 codepython777 Hyratel1: remember, i need ease of assembly more than price. Control source + MC is already given all i need is the chasis/ motors and wheels
00:35 Hyratel1 do you have access to a machine shop?
00:35 Hyratel1 or just a power hacksaw
00:36 codepython777 Hyratel1: no - but i can get access to a cnc
00:36 Hyratel1 or a hacksaw
00:36 codepython777 remember - ease of assembly
00:36 Hyratel1 we used no CNC on hormes
00:36 Hyratel1 everything is fucking bolted.
00:36 codepython777 i'd rather use cnc. Print the whole thing out - then print it 5 times or more
00:36 Hyratel1 if you don't have basic hand tools
00:36 Hyratel1 dude
00:36 Hyratel1 dude
00:36 Hyratel1 seriously?
00:36 Hyratel1 not easier
00:36 codepython777 hacksaw is basic hand tool?
00:36 Hyratel1 we had to design from zero
00:36 Hyratel1 yes
00:36 Hyratel1 it's a saw
00:37 codepython777 it is
00:37 Hyratel1 that you push and pull by hand
00:37 Hyratel1 with a blade for cutting metal
00:37 codepython777 but its manual work
00:37 Hyratel1 get over it
00:37 Hyratel1 you need 8 corner gussets
00:37 Hyratel1 1/8" steel plate
00:37 choonway i suggest you can use a powered jigsaw
00:38 Hyratel1 cuts won't be straight
00:38 codepython777 choonway: :)
00:38 codepython777 cnc = straight cuts :)
00:38 Hyratel1 codepython777, the kerf is thinner with a hacksaw
00:38 Hyratel1 seriously
00:39 Hyratel1 why are you so averse to manual labor
00:39 codepython777 Hyratel1: because i might have to make 20 of these - right now i 've an order of 5
00:39 Hyratel1 or you can use an angle grinder with a slitting disc
00:39 codepython777 hence nothing manual if possible
00:39 Hyratel1 I'm sorry did I hear you say "i've never been in FRC"?
00:40 codepython777 indeed - you heard that :)
00:40 Hyratel1 with materials available, I'd bet I could turn out several kits in an afternoon
00:40 Hyratel1 you need to figure out what you're doing, dude
00:41 codepython777 Hyratel1: indeed.
00:41 codepython777 and this is clearly not what i do :)
00:41 Hyratel1 Hormes is a versatile, sturdy design
00:41 Hyratel1 if you want sturdy, you build it
00:41 Hyratel1 if you want fast, but flimsy, you print it
00:41 Hyratel1 Hormes has run full-speed into a steel post. it tweaked the frame a tiny bit.
00:42 Hyratel1 a minute of smacking the corner with a hammer trued it back up
00:42 codepython777 Hyratel1: poor steel post :)
00:42 Hyratel1 nah, hormes kinda bounced back
00:44 choonway powered jigsaw cuts can be straight enough if you set it up correctly
11:10 rue_house http://hessmer.org/gears/InvoluteSpurGearBuilder.html
21:00 xulf HEY
21:00 theBear HO
21:01 xulf You have made some robotic stuff?
21:01 theBear hmmm, is that a euro x or a cantonese or more of a english x in your name, as far as saying it with a mouth
21:01 theBear sure, i mean, not thismorning, but yeah
21:01 xulf uh I'd say it like za-ulf
21:02 xulf is that the english way ?
21:02 xulf or ULF
21:03 xulf what kind of stuff have you made?
21:03 theBear i'm tempted to say i'd call it the european way, which is closer to english than any chinese :) names are hard to describe by area cos they don't change much between languages (for a same name)
21:04 xulf I'd actually say ULF
21:04 theBear little bits and pieces mostly, tho a few years back i had a cool job doing repairs on MANY big expensive robo-disco-lights.. you know all the modern ones that throw themselves this way and that, often built like a giant disco-egg suspended in a big U shaped yolk
21:05 xulf Also if you don't know how to say someone's name don't say it at all, just say the first or last part of their name or say the first letter
21:06 xulf why would someone fix something like that instead of buying a new one?
21:07 theBear mmm, language and sounding not-too-stupid in foreign language situations is a hobby
21:07 theBear 'cos they cost $20-30k to buy maybe
21:07 theBear at least the brandname bigger ones
21:07 theBear that's each
21:07 xulf a rotating light costs 30K?
21:07 xulf I'm in the wrong business
21:08 theBear mmm, it blows my mind even after years seeing the industry from inside and several more to come to terms with it :)
21:10 xulf but on google I'm finding them for 30 dollars
21:10 xulf I googled robo disco light
21:10 theBear if m y stupid browser would start i'd find you one
21:10 theBear meh, ask for a mac2000 maybe
21:11 theBear martin brand
21:11 xulf ok so I should type "Light" and sort by price, high-low
21:11 theBear pretty sure that was the new model introduced a couple years before i finished up at that job, maybe 4 or 5 years ago, debuted for $26,000 AUD at the time
21:12 xulf if I were starting a club or something I'm not sure I'd go for the 30000 dollar disco light
21:12 theBear and if ya wanna really drop your jaw, lookup "GrandMA 2" lighting console/control desk for a price
21:12 theBear if i ran the state theatre i wouldn't even go for 'em, but i don't have to approve to be damned good at fixing them :)
21:13 xulf so this industry has a 99.99 percent product profit margin
21:14 theBear also for lighting generally, i still feel the older style "small X-Y moving mirror" with everything else (big and heavy and hot parts) fixed in the non-moving part.. apart from conceptual sensibleness, even $30k worth of big egg style ones can't move a spot of light around half as fast as a tiny mirror
21:15 xulf oh yeah that makes way more sense, the moving mirror
21:16 theBear to an extent, the big 2 or 3 companies do LOTS of r+d/testing in relation to guaranteeing reliability and spares failure rates etc, but that only makes it 99.19% profit <grin> and that grandma2 is just a stock late-model-p4 mobo in a case with some motorised sliders and 3 or 4 small lcd screens, and a crappy (vxworks from memory, version 1 was, and it still sells and still has a standard k6-3 mobo/cpu !) rs485 interface/shifter
21:18 theBear course the companies/bosses/fools out there with the big companies cheque books in their offices say "but the new ones look so nice, ooooh " instead of "who cares ? the lights are up in the ceiling, noone looks at them ! they're lights, they just hurt you eyes if you try anyway, and damn, i could save so much that my pay could triple and we'd still be saving money !"
21:19 xulf seems like there's major problems in that industry
21:19 xulf and who calls a disco light "Grandma" come on..?
21:20 xulf does grandma go out on the dance floor
21:21 theBear it's a controller, but the company is grandma, and i suspect it's a joke... only guys i met from there have been a bit too foreign and humorless to say either way, but it sure makes me and my old work buddies giggle when you see someone walking away with a big GrandMA on the back of their jacket :)
21:22 xulf I got an even better name for a company so long we're joking.. its called "The Scam Lighting Company"
21:22 xulf that's real funny
21:27 theBear heh :)
21:27 xulf why don't I just put a projector on axis and call it a disco light and sell it for 30K ?
21:27 xulf isn't that what this thing is? a rotating projector
21:28 theBear as a tech tho, it was a fun/interesting industry to be in... lots of robotic/moving stuff, little bits of glass and hid lamps and crap to deal with, long data runs, installs and callouts and troubleshooting, regular chances to go out in the truck for a break from your workshop, maybe even do some sound/lighting operating which is always casual (at least the bit between the setup/packup)
21:30 theBear oh, they recently started doing that too, can charge more and they get to replace the many steppers and glass/metal wheels and things inside with a single lcd panel and a couple fans :) i would be impressed IF they could do vari-focus on the same thing tho, if you could do even a low-res focus-map on a projector like that, you could project accurately on all kinds of shapes/things/angles. which would at least for the first couple of "plays" be lots
21:30 theBear of fun (i find many things seem amazing, but then you actually get your hands on them, have fun, and never get interested again after :)
21:32 xulf is it not just a rotating projector?
21:34 theBear the existing ones are, well, conceptually/basically, just made to look like previous lights more than a projector that rotates :) and there is no common/non-experimental (think mit kinda pre-industry experimental) tech for variable or adjustable focus lenses yet (i know you can focus a pair of lenses <grin> i'm talking about a single "thing" that can alter focus (effectively thickness in a lens-situation) arbitrarily across itself
21:36 xulf so it needs to auto focus fast
21:39 theBear well, i suppose that would be better than slow, but it's really not being worked on or even desired in the "entertainment lighting industry" outside of an idea i had 'cos a sales-dude worded something in a strange way one day when he was introducing the first moving-projector kinda models at some talkey day or other
21:41 theBear i forget what industry it is "useful" for, ooh, probly eye-lens implant/replacements or maybe even spectacles
21:41 theBear thus the world-class/university kinda research being done on the concept in general (just the adjustable/focusable lens/something concept that is)
21:42 Jak_o_Shadows I dunno if you need varying focus in glasses. Even biofocals you probably want distinct regions.
21:42 Jak_o_Shadows That said, it is probably being worked on for eye implants
21:45 theBear mmm you are probably right... the short period i had glasses they weren't really that kind of strong focus thing anyway... and last couple days the eye that they helped to long ago fix has been a little dilated and generally not doing well with close/TINY focusing.. i might have to get me some kinda testage this week
21:46 theBear mmm, i'm not 100% sure why they doing it, but a personal/far from sensible interest/project made me wonder about it and i AM sure that some proper scientist guys with infinite resources are looking into it :)
21:47 Jak_o_Shadows Oh yeah, it's definitely going to be useful at some point
21:48 theBear and a couple years or so after that point, all that money and effort and science can be abused by me for next to nothing :)
22:24 JEntrep how would you get a massive group of people to cooperate on a project?
22:26 Jak_o_Shadows trello
22:26 Jak_o_Shadows some version of source control, probably github
22:26 theBear well, lets just say it wouldn't be "peaceful" or likely legal :-)
22:26 theBear but i'll be damned if they wouldn't work AND pretend to enjoy it :)
22:29 JEntrep guys seriously ...
22:29 JEntrep I mean cooperation
22:29 JEntrep not the tools used by cooperating people
22:30 JEntrep theBear, ACTUAL cooperation
22:30 JEntrep lol
22:30 Jak_o_Shadows oh. Regular meetings
22:30 Jak_o_Shadows lots of commjnication
22:30 theBear i got yer meaning, the real answer would be i never really thought about it much 'cos i don't work well with others
22:30 JEntrep commjnication? hmm, never heard of that one.
22:30 JEntrep :p
22:30 theBear or maybe i work best alone, somewhere in the line between those 2
22:30 JEntrep theBear, I think it is more that you haven't found anyone with a good cooperating methodology
22:31 JEntrep such that you feel natural cooperating in the group constructed by the methodology
22:31 JEntrep I think human cooperation is the single-most diffucult problem in modern society.
22:33 JEntrep Jak_o_Shadows, why did you answer with the things you answered with?
22:33 JEntrep regular meetings and communication?
22:34 Jak_o_Shadows How can you work with people if you don't know what's going o n
22:35 theBear it's easier that way, they don't annoy me so much :)
22:36 JEntrep theBear, how can you know what's going on if you don't work with people? ;P
22:37 theBear JEntrep, maybe, but at all levels, i suspect it's a communication time vs do-it-myself time kinda thing mostly, which means there are probably times/situations where it would feel much better and be useful, but i just seen so many where it isn't... tho you make me think about things like un/loading trucks or setting up sound systems with someone else who does it without thinking and is half-decent at it... i cooperated in those situations lots and
22:37 theBear never even noticed it (i've been known to think/say how i "don't work well with others" one way or the other more than once in the past)
22:38 theBear "i need to be drunk ! helps with the hallibulutions ... lets me see things in clear black and white" <-- quote from unknown shamen
22:38 JEntrep you mean short-term cooperation on something that clearly benefits from having multiple people working on it?
22:38 JEntrep like unloading a vehicle?
22:39 JEntrep theBear, like you would be able to cooperate with a "super--annoying" person if you knew that UNLESS you cooperated something absolutely awesome wouldn't be built?
22:40 theBear well, unloading a vehicle where the things landing or moving in the wrong order is a HUGE hassle, that kinda thing..
22:41 theBear JEntrep, i answer your question with this question... if you were a completely dedicated vegetarian, and i came up to you with some chicken saying "eat this chicken, or i'll kill another chicken !" what would YOU do ?
22:41 theBear this is getting a bit deep and philisophical suddenly
22:42 JEntrep why would I care about you killing another chicken?
22:42 JEntrep let's assume I am a staunch animals activist?
22:42 JEntrep then you say: eat this endangered animal meat or else I will kill another one.
22:42 theBear oh, becuase you chose to be vegetarian because you don't believe in or approve of killing animals for food
22:42 theBear or at all
22:42 JEntrep Then I would say: okay :(
22:43 JEntrep I would munch on that chicken and my tears D:
22:43 JEntrep knowing it was for the best
22:43 JEntrep but good thing I like eating chicken ;P
22:44 theBear mm, one of those two... in which case i may well say ok, or would have once upon a time.. things look different after things like not-dying-when-you-are-sure-it's-about-happen or imagining the rest of your life being just as crappy as it is currently (and it IS currently crappy)
22:44 theBear heh, me too :-) chicken is a tasty dead animal !
22:45 JEntrep oh, well I see you are coming from
22:54 Jak_o_Shadows i2c just needs a clock line and a data line right?
23:00 theBear see, i'm not good for those questions generally.. i'm a sociopath deep down, only have a remote concept of empathy etc 'cos i can appreciate intellectually that some things are necesary so the world and infrastructure i value doesn't collapse overnight, and seperately i just don't enjoy excessive social contact, or (maybe less seperately) feel the need for very much at all, a lot of current-societies common values and views really bother/disgust/offe
23:00 theBear nd me too, so that is another reason for me to want to avoid close contact