#robotics Logs
Oct 09 2015
#robotics Calendar
01:39 rue_school hmm I need to weave this in pretty deep
01:47 rue_school ok, so the model I'm looking at, is that you pass the string for the equation you want to solve, you pass a list of variables and values, and a list of functions, it gives you back a result value, and modified variable list
01:50 rue_school I dont know if you should have to pass a list of the functions or if they should be built in somehow
01:51 rue_school I suppose functions can be built in, that saves me having to pass them thru about 7 levels of function calls
02:35 deshipu orlock: cylon basestar?
03:48 veverak hmm,chmm
03:48 veverak 3d print galactica you say
03:48 veverak well, if you insist in doing so
04:03 deshipu veverak: 3d print a board for the bsg board game
04:03 deshipu veverak: complete with character figures
04:04 veverak I see
04:04 deshipu just saying
05:21 veverak nah, worst part about coding: find names
05:41 deshipu there are only two hard problems in programming
05:41 deshipu 1. cache invalidation,
05:41 deshipu 2. naming things,
05:41 deshipu 3. off-by-one errors
05:42 deshipu veverak: have you seen http://zoobotics.de/ ?
05:44 LiohAu deshipu: is it a robotics design company?
05:44 LiohAu or are they actually building the robots?
05:46 deshipu LiohAu: I think they are selling kits
05:46 LiohAu the arms are using servos I guess?
05:47 LiohAu the ARM-01 looks like it uses servos.
05:47 deshipu no idea, I mostly meant the paper spider robot
05:47 Snert think so.
05:48 LiohAu I would really love to find how to build an arm like the ARM 02. But the only thing we find on the net (as hobbyist) are servos based arm...
05:48 Snert yea no good quality. Just cheapass hobby servo stuff.
05:49 LiohAu I assume that we won't see a lot of "robotics" startup before harmonic drives becomes affordable.
05:50 Snert what does harmonic drives have to do with it?
05:50 deshipu there are lots of "not cheapass" stuff out there -- only not for hobbyists
05:51 deshipu I mean, if you can pay for the good servos, you are not a hobbyist anymore
05:51 Snert I'm wanting to biuild an arm - quality stuff. No hobby servos.
05:51 deshipu Snert: well, go ahead
05:51 deshipu Snert: looking for a sponsor?
05:51 Snert littleBy Little, I will.
05:51 Snert not right now on the sponsor thing.
05:52 LiohAu well you can't make a "good" arm (I mean something that can lift more than 500gr..) without using good gear motors.
05:52 Snert I have to get past a certain point first.
05:52 deshipu once you have the budget, it's better to just buy a kuka or whatever :P
05:52 LiohAu from what I read, planetary/spur gearbox have to much backslash for an arm.
05:52 LiohAu so you only have harmonic based gearboxes, that cost 1K$ each.
05:53 deshipu I still have my MeArm here to hack
05:53 LiohAu MyArm = worst than cheap servo :D
05:54 Snert oic - harmnic is the name of a company.
05:54 LiohAu yes, but iirc they have a patent on their gearbox, no?
05:55 Snert I have no idea.
05:55 Snert Looks like pretty high quality stuff.
05:58 LiohAu Snert: how would you build your arm?
05:58 LiohAu using servos? steppers? DC motors?
05:59 LiohAu (Or by making your own servos with DC motors ^^)
05:59 Snert The elbow - a screwjack. (Stepper motor with threaded rod).
05:59 Snert can hold 500 pounds.
06:00 Snert when stationary.
06:00 LiohAu but how fast is it?
06:00 LiohAu stationary is not the hard part :P
06:00 Snert the stepper can spin pretty fast, raising or lowering pretty fast.
06:01 Snert not that fast is a desired thing.
06:01 LiohAu could you make an arm catching a ball thrown on it ?
06:01 LiohAu I mean, enough fast for this task ?
06:02 Snert speedwise, I think so.
06:02 Snert but damned if I could code such a thing yet.
06:02 LiohAu so your problem is programming?
06:02 Snert pololou has a video of the screwjack but I don't think it's fast enuff.
06:03 Snert coding is certainly an issue...so is $ when getting started.
06:03 Snert but I'll get there.
06:03 LiohAu well depends on where you live
06:03 LiohAu if you're in C.A., you can raise funds easily with a good idea.
06:04 LiohAu I don't think it's that easy for company like PAL in spain for example.
06:05 LiohAu Snert: http://www.robotpark.com/academy/robotic-mechanisms-screw-systems/ this is what you was describing?
06:07 Snert https://www.pololu.com/product/2689
06:07 Snert for starters.
06:08 LiohAu it looks like the same mecanism used in electric scissor lift that I wanted to use in a project few months ago.. Indeed it's really powerfull, but that also looked really slow
06:09 LiohAu well I would have used this as a shoulder
06:09 LiohAu (if I were you)
06:09 Snert https://www.pololu.com/product/2345 <--- or this for the elbow.
06:10 LiohAu 40 mm/s too slow
06:10 Snert there are others....thems just examples of what might work.
06:11 Snert I'm not out to catch a ball.
06:11 LiohAu http://sdk.rethinkrobotics.com/mediawiki-1.22.2/images/thumb/4/4d/Baxter_Joint_Names.png/300px-Baxter_Joint_Names.png looks like a linear actuators between S1/S0 right?
06:12 Snert very nice
06:12 Snert alot of what it looks like is bound to what it would do.
06:12 Snert and I haven't really decided what the arm would actually do, sitting on my desk.
06:12 LiohAu well in my opinion, it's just another 7-dof (or maybe less?) arm.
06:14 Jak_o_Shadows linear actuators? Nah, springs
06:14 LiohAu ah?
06:15 Jak_o_Shadows Yeah. Springs can help you lift more weight
06:15 LiohAu a spring that helps?
06:15 Jak_o_Shadows Just gotta be a tad careful about oscillations
06:15 LiohAu compression springs then?
06:15 Jak_o_Shadows yep
06:16 Snert springs to relieve the load perhaps, but an actuator for positioning definately.
06:16 LiohAu I read an article about elastic series actuators, they were talking about springs in it, do you think it's that spring?
06:17 Snert I don't think I have the spring thing correctly pictured.
06:19 LiohAu you don't see it on the picture? that's what you mean?
06:20 Snert I don't have the right concept for where the springs are and how they would be used.
06:20 Jak_o_Shadows Do you know how to draw a free body diagram?
06:20 Snert not that picture. there aren't any springs in that picture
06:21 LiohAu Jak_o_Shadows: this kind of drawing : http://ram-lab.engin.umich.edu/Images/seas.jpg ?
06:21 LiohAu (the thing at the left?)
06:21 Snert Jak_o_Shadows: no, I can draw a stick figure and that's it.
06:22 Jak_o_Shadows That is not a free body diagram. Free body diagrams show force
06:22 LiohAu oh ok, free body diagram means something else (sorry i'm not english ^^).
06:22 LiohAu (just googled it)
06:22 LiohAu well I did this in high school few years ago ^^
06:23 Jak_o_Shadows If you want to do look and do calculations on robot arms, you need to be able to draw a free body diagram
06:23 Snert LiohAu: nice diagrams. I still can't picture how the spring does anything, but no matter, that won't be what I do anyways.
06:24 Jak_o_Shadows http://hyperphysics.phy-astr.gsu.edu/hbase/imgmec/acmfb.gif eg, through you should probably also have a sign convention
06:24 Jak_o_Shadows On the joint: The weight of the arm creates a torque (or moment) that would rotate the arm down
06:24 Jak_o_Shadows The spring provides an opposite torque (or moment), so the motor doesn't need to work so hard
06:24 Snert yes I understand all that.
06:25 LiohAu me too :)
06:25 Snert I need to see actual spring connections to a real world scenario.
06:25 LiohAu but when the arm needs to lower, it needs to work harder (in the baxter case)
06:25 Jak_o_Shadows eh, yes, but the weight of the arm probably does it for you.
06:26 Snert or it could be balanced. force-wise...so it was neutral.
06:26 LiohAu well in the end this is a specific design for a specific use case
06:26 Jak_o_Shadows The main trade-off would be stability: The spring would cause some oscillations, which prevent you from moving fast and accurately
06:26 LiohAu I mean it's not really generic, depending on the weight lifted by the arm, you may want to change the spring, right?
06:27 Snert a sliding counterbalance like a crane?
06:27 Snert naaa unwieldy in an arm scenario.
06:28 Jak_o_Shadows If you were doing it absolutely properly, sure. But a) It's baxter, which is more of a teaching/research robot rather than an industrial robot b) It probably doesn't make too much difference
06:29 LiohAu the VSA (variable stiffness actuator) could be the solution ?
06:29 LiohAu If I understood correctly the concept, they use a motor to alter the spring so the stiffness of the arm changes dynamically
06:29 Snert gear motor with worm-drive sideways on the outer circumference of a drive gear for the shoulder.
06:30 Snert maybe 2 motors.
06:31 LiohAu Snert: that's your solution? a screwjack for elbow, and the worm-drive thing for should?
06:31 Snert not for sure, just 2 ideas. But those are looking good so far.
06:32 Jak_o_Shadows My solution, if I wasn't happy with anything I could buy off the shelf, would be DC motors with encoders
06:32 Snert with perhaps a linear extensor in the forearm. To extend the gripper out if desires.
06:32 LiohAu well you'll tell us in few years :P
06:33 LiohAu Jak_o_Shadows: that's what I had in mind. But the torque that I find for DC motors is lower than the torque I find for servos..
06:33 Jak_o_Shadows What torques are you needing?
06:34 LiohAu I wanted to build an arm able to lift few kg (a 2kg bottle of water was my goal ^^)
06:35 Snert that's what I'm looking to do.
06:35 LiohAu Jak_o_Shadows: this kind of motor http://www.superdroidrobots.com/shop/item.aspx/ig52-04-24vdc-010-rpm-gear-motor-with-encoder/937/ on the should would not be enough right?
06:35 SpeedEvil how long do you want this to last?
06:35 LiohAu on the shoulder*
06:36 SpeedEvil LiohAu: fundamentally, you need to work out how to understand torque
06:36 LiohAu (given the bottle is at the end of the arm that has a 70cm length)
06:36 SpeedEvil 100kgcm - can in principle lift 2kg at 50cm
06:36 LiohAu SpeedEvil: that's why I completed my sentence with the length ^^
06:37 SpeedEvil 10RPm is quite fast.
06:37 LiohAu you think?
06:37 SpeedEvil That is one rev in six seconds, or a quarter rev in 1.5
06:37 Jak_o_Shadows 9.8 Nm is a bit
06:37 SpeedEvil It means you would be accellerating for 0.75s, and decellerating for the same
06:38 Jak_o_Shadows Also, DC motors have a speed torque curve, and you need to look at that.
06:39 LiohAu well as you can see, with 2kg at 50cm, 1st/ I have to think about a shorter arm, 2/ I have to think about a carbon fiber arm (i'm saying this for the weight of the arm itself) :P
06:39 Jak_o_Shadows Alu would probably be easier.
06:40 Jak_o_Shadows Also, please learn statics. It'll make things easier
06:40 LiohAu and 3/ you can only have 1 DOF, because if you add another motor for an elbow for example, you have used you're 2kg limit just with the elbow motor.
06:42 LiohAu Jak_o_Shadows: well alu/carbon fiber, I was just saying it's impossible to make a decent arm with hobby parts.
06:42 LiohAu Torque is way too low
06:42 SpeedEvil LiohAu: Wood
06:43 Jak_o_Shadows I would start thinking about not putting the motor in the joint. I think rue's done some nice ones.
06:43 SpeedEvil Semi-decent wood, with a big section, can be very, very rigid
06:43 LiohAu Jak_o_Shadows: that solves the elbow thing, you're right
06:44 Jak_o_Shadows Indeed. Wood is an really nice material.
06:44 Jak_o_Shadows LiohAu, start small.
06:47 LiohAu My problem is really not about the material rigidity, weight or something like this. It's really that I can't find powerfull enough actuators. You can find affordable motors like the one I linked, and after that there's a gap, you have to spend 1K$ for each motor.
06:48 LiohAu If I recall correctly, the best affordable actuator I found is this one : http://www.topmodel.fr/en/product-detail-18825-servo-topmodel-mammoth-38055 --> 380kg*cm
06:52 LiohAu And also I do not agree with your "start small" :D
06:52 LiohAu I prefer target big, and failing with medium :P
06:52 LiohAu brb
07:06 SpeedEvil in principle, you can do a quite cheap and light reduction with two belt drives.
07:06 SpeedEvil Cogged belts can be quite inexpensive
07:08 SpeedEvil http://www.ebay.co.uk/itm/3D-Printer-GT2-Looped-Belt-2mm-Pitch-6mm-Width-Closed-Loop-Belt-/201406251106?var=&hash=item2ee4bf8062
07:08 SpeedEvil A couple of 10:1, a brushless motor, and you're at a few hundred watt geared drive.
07:29 deshipu I prefer to play with the cheapass servo inaccurate stuff and learn coding first, and *maybe* make something better once I know how.
07:33 LiohAu deshipu: I would agree with you if learning on cheap stuff could let you upgrade when ready. But learning on cheap servo, is completely different than learning on better systems, so there's no capitalization on learning (imo)
07:34 deshipu LiohAu: what is so different?
07:34 deshipu LiohAu: ik is the same
07:34 deshipu LiohAu: motion planning is the same
07:34 LiohAu ok, well algorithm part of the software is the same.
07:34 deshipu LiohAu: the only thing that changes is the low-level servo control code
07:35 deshipu which is literally the easiest part
07:35 LiohAu but I don't think that you will bother with real time kernel and this kind of hard stuff for a servo based arm, right?
07:35 deshipu no, arduino is real time enough :P
07:36 deshipu so there is that trap that people often fall into
07:36 LiohAu well when using things like ROS, the IK/motion planning part is already done :d
07:37 deshipu where they avoid hard problems by finding other problems to tackle first
07:37 LiohAu anyay when I said it's different, I was mostly speaking about the hardware part
07:38 deshipu once you learn how to handle it all, you can replace the hardware
07:38 deshipu it would be a waste if you bought high quality hardware and then realized you can't control it
07:39 deshipu also, small and light stuff (like the MeArm) is actually easier, because you get less inertia
07:39 LiohAu sure, but it would be a waste to learn on cheap stuff, and to realize that you're not able to build/buy quality stuff :P
07:39 deshipu I even replaced all the metal bolts with acrylic ones
07:40 deshipu LiohAu: why would that be? the experience transfers
07:40 deshipu in the worst case you just need to learn a few additional things
07:40 deshipu which you would need to learn anyways, mind you
07:40 LiohAu or maybe you learn everything for nothing because you can't achieve your goal.
07:41 deshipu you still at least had fun with the cheap stuff
07:42 LiohAu well that's the remaining part ^^
07:42 LiohAu little bit expensive :P
07:42 deshipu my cheap-ass hobby-servo quadruped is $25, I pay more for a dinner
07:43 LiohAu from my point of view robotics is an interesting work, not a game
07:43 LiohAu maybe for you it's a hobby
07:43 LiohAu for me it's an interesting topic as gardening is one
07:43 deshipu it is a hobby, yes
07:43 deshipu well, gardening is a fun hobby too
07:44 LiohAu but it's not what I would call a hobby
07:44 deshipu why not? it's pointless, costly and takes a lot of your free time
07:44 deshipu sounds like a hobby
07:44 LiohAu missing the fun part. The fun you get from a paintball game for example ^^
07:45 deshipu nah, anything can be fun, fun comes from your approach
07:45 LiohAu Like when I start coding an app, it's interesting, and I get satisfaction when I succeed, but I'm not calling this a "fun" task.
07:45 LiohAu it's a pleasant work
07:46 LiohAu we should rename that channel in #robotics-philosophy :P
07:56 SpeedEvil http://forum.electricunicycle.org/topic/1385-warning-dont-ride-with-v126-firmware/?page=1 - on bleeding edge firmware
08:07 deshipu LiohAu: philosophy is implied in the fact that it's IRC
08:07 LiohAu :d
08:22 veverak deshipu: nope, but they look cool
08:22 deshipu veverak: they would work with Tote's code :)
08:23 veverak :)))
08:23 verak needs to read tote
08:23 verak made his own leg math finally, wants to com
08:23 veverak by "made" I mean that it is in usable form
08:24 veverak deshipu: that Zuri looks awesome
08:31 deshipu veverak: it would be nice to make a python library at some point...
08:31 veverak deshipu: for.. ?
08:31 veverak for tote?
08:31 veverak that's what I am doing just now!
08:31 veverak :D
08:31 deshipu for the legs and gaits etc.
08:32 deshipu I have code that works and is roughly modular, but the api is far from perfect
08:32 veverak links?
08:32 verak wants definetly to write his own code for controlling the robot, because of practice and fun invo
08:37 LiohAu what is "tote"? a code name ?
08:37 deshipu LiohAu: http://tote.rtfd.org <-- the robot I work on
08:38 LiohAu missing a "fucking"
08:39 LiohAu readthedocs.org => rtfd :)
08:41 deshipu that's "fine"
08:42 veverak deshipu: so many repos O_o
08:42 veverak which one is python-tote?
08:42 veverak :D
08:45 deshipu veverak: https://bitbucket.org/thesheep/ukubik
08:45 deshipu veverak: obviously ;)
08:47 veverak hmm
08:48 veverak deshipu: will try to do something
09:03 veverak http://git.veverak.org/robotz.git/blob/master/tote/algo/leg.py my math so far
09:05 veverak but, want to implement some way how return proper error with proper reason it failed
09:05 veverak and apply min/max angle limits
09:08 veverak so robot can eventually decide how to continue based on when math failed
09:10 deshipu math failed, explode
09:10 deshipu divide by zero, explode
09:10 rue_school veverak, did you do it in C?
09:10 rue_school if so I can help you put it togethor with my library
09:11 rue_school its really easy
09:11 rue_school oh duh, read dude
09:11 rue_school :)
09:11 veverak :D
09:11 rue_school k, nevermind
09:12 veverak rue_school: but I had the creazy idea to write everything in C++
09:12 veverak but nah
09:12 rue_school no, c++ would make hte code worthless
09:12 rue_school after a few layers
09:12 verak loves OOP too
09:12 rue_school oh
09:13 rue_school so you like not really knowing what someones library might be doing?
09:13 veverak wat?
09:13 veverak what does have it similar with OOP ?
09:13 veverak I mean
09:13 rue_school abstraction and overriding
09:13 veverak it is possible way of doing it
09:13 veverak but it's not enforced
09:14 rue_school you can hide and obscure code easy
09:14 veverak and?
09:14 veverak I mean
09:14 veverak it's tool
09:14 veverak it's on you if you use it properly or not
09:14 veverak abstraction is great tool if used properly :)
09:15 rue_school which means that its easy to make a code library that barely works, that someone uses in their library, that someone uses in their library, that someone uses in their project that just barely works and they dont give a **** cause they have no idea how it all works in the first place
09:15 rue_school which is basicaslly the state of almost all arduino libraries
09:15 veverak their problem, not mine
09:16 veverak if programmer just uses library that is shitty... it's shitty programmer
09:16 veverak i mean, I only read "programmers can't handle OOP" ...
09:16 veverak from what you said
09:17 veverak simply: what you described is not using OOP properly ;)
09:17 veverak properly means to check the library you are going to use, look into it and decide, is it good? than OK, I will work with it
09:18 veverak practically, I decide what I want, study it, find implementation that is OK for me or make mine object, document it and than I can forget about most of the stuff inside
09:19 veverak and if problems arise, I will just remind myself the stuff and continue
09:19 veverak *and repair and continue
09:21 rue_school I'v never seen oop used in a) a way that makes usefull use of it b) properly
09:22 rue_school as such with java, I'v never seen a java app that dosn't leak memory and crash
09:22 rue_school unless it adds 1+1
09:22 rue_school and the only beef I have there is that it takes up 20M of space to do it
09:25 veverak java sucks
09:25 veverak :)
09:30 rue_school i mean, I only read "programmers can't handle java" ...
09:31 veverak yeah
09:44 deshipu oop is a useful tool
09:44 deshipu as any tool it can be abused
09:45 deshipu I think that people tend to overuse inheritance, instead of using composition
10:11 rue_school I'v never seen inheritence used properly in a program that wasn't accademic
10:19 verak got some ... "overshooting" things with
10:20 veverak I suppose I sometimes try to isolate object too much
10:21 veverak but, it's more like it makes me to write more, I try to avoid harming functionality :)
10:24 rue_school I do my C with structures, keep all my functions less than one screen page and do light verticle compression
10:24 rue_school so I dont write
10:24 rue_school int
10:24 rue_school main(
10:24 rue_school void
10:24 rue_school )
10:24 rue_school {
10:24 rue_school return
10:24 rue_school 0
10:24 rue_school }
10:24 rue_school when I can write
10:24 rue_school int main(void) {
10:24 rue_school return 0;
10:24 rue_school }
10:25 veverak that's nice
10:25 veverak once, one person hated python that without proepr {} brackets, it's hard to fand where block of code starts and ends if it's long enough
10:26 veverak but I thought... if code is that long I suppose that's different kind of problem
10:26 veverak rue_school: +1 for making short functions
10:29 rue_school yea, shortcomming in python, I know what the intent was tho
10:56 rue_school I hate seeing code thats a 104k file with a 400 line function, with everything and its booty packed into one file
10:58 rue_school usually because if I'm looking at it, it needs a correction of some kind
11:00 deshipu I like the k&R style
11:00 rue_school with or without a 40 column limit
11:01 rue_school with or without real tab characters
11:01 deshipu the k&r brace formatting style
11:02 rue_school when I fit a function on a screen, I allow myself to do that at the finest reasonable font, about as wide a window as it reasonable
11:04 verak made one really long func
11:04 veverak 120 rows?
11:04 veverak delaing refactoring for a long time now
11:04 rue_school ouch
11:04 veverak rue_school: exactly
11:05 veverak it tries to handle incoming data
11:05 rue_school there comes a point where you throw out the code you have and start again, with the knowledge of where you have to end up
11:05 veverak and it was attempt to assert that the one that sends the data is completely dumb and anything can be screwed
11:05 veverak rue_school: yeah, did that too many times
11:06 rue_school pactise makes perfect
11:06 rue_school unless its me on the topic of speeling
11:06 veverak :D
11:06 veverak but I was thinking
11:06 veverak there should be allready figured out mechanism to properly check messages
11:07 veverak and I don't check it entirely
11:07 rue_school what kinda messages?
11:07 veverak jsonrpc
11:07 rue_school can you have a seperate thread run a state machine?
11:07 veverak I can haz threads
11:07 veverak :)
11:10 veverak ocuh, I see
11:10 veverak it is aslo a lot about sending back proper error based on what was wrong
11:10 veverak and handling the "rpc" mechanism
11:11 veverak hmm
11:12 veverak it looks like I didn't implemented timeout for RPC responses
11:12 veverak nah
11:16 rue_school just have a sperate thread that uses wait or sleep to get events
11:17 deshipu ugh, threads
11:17 deshipu :(
11:18 rue_school I have a serial system that uses threads for aggrigating streamed incomming data
11:18 veverak rue_school: wahy separate thread for that?
11:18 rue_school yes, because I can just use a blocking read() call
11:19 verak got event loop and found json pa
11:19 veverak event-loop library
11:19 veverak with TCP support :)
11:20 rue_school for serial?
11:20 veverak serial?!
11:20 veverak too
11:20 veverak but for serial library got different handler
11:20 veverak jsonrpc is for network communication
11:21 veverak :)
11:21 rue_school while(1) {
11:21 rue_school FD_ZERO(&rfds);
11:21 rue_school FD_SET(this->fd, &rfds);
11:21 rue_school FD_SET(0, &rfds);
11:21 rue_school tv.tv_sec = 0 ;
11:21 rue_school tv.tv_usec = 250000;
11:21 rue_school retval = select((this->fd)+1, &rfds, NULL, NULL, &tv);
11:21 rue_school
11:21 rue_school if (retval == -1) {
11:21 rue_school perror("select()");
11:21 rue_school } else if (retval) {
11:21 rue_school if ( FD_ISSET(this->fd, &rfds) ) {
11:21 rue_school
11:21 rue_school // read input
11:21 rue_school /*
11:21 veverak point is... ?
11:21 rue_school if ((*data = malloc(bytes+1)) == NULL) {
11:21 rue_school *data = NULL;
11:21 rue_school return ErrMalloc;
11:21 rue_school }
11:21 rue_school socket, serial port, whatever
11:21 veverak (apart that pasting code into IRC is ugly)
11:21 rue_school the thread runs the while(1) loop
11:22 veverak yeah, propably something I would do if I didn't used event-loop library that does it for me
11:23 rue_school oh that was the old code
11:23 rue_school while(1) {
11:23 rue_school read(this->fd, &data, 1);
11:23 rue_school // printf("-%X", data);
11:23 rue_school this->charHandler(&data, this->callBackPtr);
11:23 rue_school }
11:23 rue_school thats what I'm acutally using
11:24 rue_school serial ports are evil under linux
11:25 verak uses libuv bindings for py
11:29 rue_school if your using other peoples libs, I hope they were written better than most
11:36 deshipu because nobody writes code as well as you
11:39 rue_school no, because "there is no such thing as origional code" nobody writes anything, they just copy and monkey with other peoples code
11:40 rue_school which is why everything is so barely working and broken
11:40 rue_school its why the margins for hackers to exploit are so wide
11:42 rue_school nobody knows how to code a buffer
11:42 rue_school so it seems
11:42 rue_school esp if they work at microsoft
11:43 rue_school "we will recieve exactly 22 bytes in all circumstances so thats how long our buffer will be and we dont need to even bother checking to see if we have hit the end because all anyone will ever send is 22"
11:58 rue_school hmm, I'm waiting to leave
11:59 Kamondelious rue_school, You wouldn't happen to have tutorial on how to properly code a buffer would you?
12:00 rue_school nope, I would say its common sense, but apparently its not
12:00 rue_school most people dont even know how to use a lookup table
12:01 Kamondelious Driving is mostly common sense too and look how that's working out.
12:01 rue_school its in the *ugh, I give up* catagory
12:01 rue_school its no wonder we dont have flying cars
12:01 Kamondelious Indeed
12:01 rue_school it would be the end of the human ravce
12:01 e_school mashes keyb
12:03 Kamondelious We will have flying cars when monkeys give up their aspirations of software development and people no longer feel the need to drive for themselves.
12:03 Kamondelious hehe
12:05 rue_school and the human population is regulated by artificial intelligence to keep it in check?
12:09 Kamondelious sounds wonderful in theory
12:09 rue_school I think too many people would be frightened
12:09 Kamondelious most definitely
12:14 rue_school well, I managed to get most of the part of functionality for functions in my command line calculator working
12:14 rue_school easy from here
12:16 Kamondelious How is it different than bc?
12:18 rue_school its based on a library I wrote that you just pass the string to be calculated, and a list of variables, it gets the results and modifies any variables as required, and its smaller
12:18 rue_school so, if you have a number of calculations to pull on the same set of variables, you can
12:19 Kamondelious Oh, that's pretty cool.
12:19 rue_school calc "A+B(A=3)(B=5)"
12:19 rue_school A+B(A=3)(B=5) -->> 8
12:19 rue_school the brackets increase the priority of the operations, so the variables are assigned first
12:19 rue_school as each operation is done, the bits are deleted
12:20 rue_school so by the time its down to A+B, it already knows what A and B are
12:20 rue_school calc "27(ResultBase=16"
12:20 rue_school 27(ResultBase=16 -->> 0x1B
12:20 rue_school I dont know if those calculators will do other bases either
12:21 rue_school calc "0b010101010101(ResultBase=16"
12:21 rue_school 0b010101010101(ResultBase=16 -->> 0x555
12:21 rue_school I have to fix that bit tho, you cant do arbitrary bases right now, just 2, 8, 10, 16
12:22 rue_school there are no libraries for properly inputting or printing arbitrary base numbers
12:22 rue_school well, that and input base definition challanges
12:23 rue_school the libary was written for a program I made that builds state machines
12:23 rue_school you use equations to set the values in the tables
12:25 Kamondelious How goes the dynamic state machine endeavour?
12:25 rue_school I coiuldn't find any library i could just pass a string and i could get back a numeric result, they all required all sorts of preparsing and ****
12:25 rue_school I need to define my ojectives better
12:26 rue_school do you know how the balancer works in a cdrom?
12:27 Kamondelious I do not
12:29 rue_school ah, there are little grains of sand in a torus, if the spindle is out of balance they rattle around and change position, when they hit a position that causes balance, they stop rattling and stay where they are
12:29 Kamondelious I would guess that it gathers wobble input data as the drive spins up and determines which direction and at what frequency the wobble occurs at and corrects for it.
12:29 Kamondelious oh, neat
12:29 rue_school yea
12:29 rue_school I'm trying to apply that to a self-learning state machine
12:29 rue_school but its not a properly defined problem yet
12:33 e_school expl
14:07 veverak how mad I am
14:07 veverak to define interfaces in python?
14:07 veverak :D
14:18 Tom_itx mad beyond belief
14:32 veverak good
17:14 RobotDuty Is it possible to lift 300 lbs with a drill motor?
17:15 RobotDuty (Cheap and simple)
17:19 Anniepoo__ sure
17:19 Anniepoo__ the question is how fast
17:20 Anniepoo__ a cheap, simple way would be some pulleys, rope, and an long eyebolt in the drill chuck. Rope winds itself around eyebolt
17:21 Anniepoo__ whats the application?
17:23 SpeedEvil 150kg needs 1500J/m, or at 300W, one meter rise per five seconds
17:26 Anniepoo__ Robot, have you measured torque on your motor?
17:28 Anniepoo__ another method: Find a strong person. Hold out the drill and say "strong person, I will give you this nice drill if you lift this heavy object for me"
17:28 Anniepoo__ 8cD
17:34 RobotDuty University's project. Thanks for the feedback. I would like to take the to explain every details later and my idea too, but I don't have the time right now. Are you there in 2 hours? Sometime my ideas aren't realistics, so I need person to tell it's not possible or it's possible and point me some documentations to learn it. (I'm a starter in robotics.)
17:35 RobotDuty (oh forgive my english too ;))
17:36 RobotDuty I'll comeback in 2 hours. Hope you will be there. Thanks again for the feedback.
17:52 veverak deshipu: pingy, you alive?
18:06 Anniepoo somebody will be around
18:07 rue_house I need to build an autonomous water gun centry for keeping the neighbours cat away, suggestions on technology to use?
18:07 veverak nah, solved emantime :)
18:07 veverak rue_house: "dog"
18:07 veverak well
18:07 rue_house nope robotic centry
18:07 veverak where the hell I lost arduino micro pro?
18:07 rue_house dogs need to much attention, eat too much food, and **** on the lawn too much
18:08 ace4016 first is true; second is debateable, and third...depends
18:09 rue_house dogs need attention almost constantly, those who dont pay attention to their dogs are, knowingly or not, abusing them and should have the dog taken away
18:09 jasonj8 I thought about making something like that rue_house, but to train my own cat not to claw at my window blinds.
18:10 rue_house heh
18:10 rue_house I wonder if the motion package still works
18:10 jasonj8 Anytime he wants something, he starts clawing at the window blinds because they're aluminum and it makes a shitload of noise.
18:10 veverak rue_house: true to that
18:11 rue_house oo there is
18:14 verak lost hist ard
18:14 veverak or
18:14 veverak questions is
18:14 veverak if I were arduino, where would I hide?
18:14 rue_house I'd let the pixies or gnomes hide me
18:15 SpeedEvil rue_house: google squirrel python birdfeeder opencv water
18:15 SpeedEvil ^camp
18:15 rue_house they do such a good job, nobody knows they really exist
18:15 rue_house hah
18:15 veverak suckers!
18:19 veverak third order from aliexpress this week
18:20 ace4016 a rolling ball robot is probably sufficient for keeping cats out; it'll get stray dogs and foxes to stay though
18:32 RobotDuty Anniepoo: Ready to learn the story ? ^^
18:33 RobotDuty Ok, the dimension are 30 x 30 x 30 = 27000 (cm³)
18:33 RobotDuty (for the robot)
18:34 RobotDuty We have two robots and they can help each others to get score.
18:38 RobotDuty We have a RobotA. The RobotA must enter into the battle when he heard 2 seconds of 5 kHz. He has 5 seconds to enter.
18:39 RobotDuty We have RobotB. He enters in the battle one minute later with the same conditions as RobotA
18:42 RobotDuty Rules : It's not allowed to throw any kind of liquids.
18:43 RobotDuty It's allowed to modify anything on the actual RobotA-B
18:44 RobotDuty The RobotA-B can't issue sounds.
18:44 RobotDuty The RobotA-B will be removed he they don't enter in the battle after 5 seconds
18:44 RobotDuty after*
18:45 RobotDuty if* (sorry)
18:45 RobotDuty The RobotA-B wheels touches the red line
18:46 RobotDuty (they are removed)
18:47 RobotDuty The RobotA-B will be removed if they starts ahead
18:47 RobotDuty The RobotA will be remove if he touches bowling colors
18:49 RobotDuty The RobotA will be remove if he has 2 WD in another color area
18:50 Anniepoo ok
18:50 Anniepoo that looks like rules
18:51 Anniepoo what are you lifting, and how long do you have to lift it?
18:51 RobotDuty Here is the battle : S = Robot A, and N = Robot B http://postimg.org/image/ip2ctw9vd/
18:51 RobotDuty Anniepoo: The "lifting part" is my idea :p
18:51 Anniepoo what are you lifting and how long to lift it? (tired of repeating that question)
18:52 RobotDuty The RobotA if it's possible to lift 300lbs.
18:53 Anniepoo ok, how are you grabbing A?
18:53 RobotDuty To get the center before other do and stay there until the end of the battle
18:53 Anniepoo sumobot
18:53 Anniepoo ok,
18:53 RobotDuty yep ^^
18:53 Anniepoo so however you lift the bot, you're at the end of an arm
18:54 Anniepoo how far up?
18:54 Anniepoo 660 lb robot?
18:54 RobotDuty What do you mean? Overthrow a robot?
18:54 Anniepoo dude, you claim this robot weighs 300 KG
18:55 RobotDuty Well, I don't know if it's realistic because the Robot looks like that (wait... I will post the picture...)
18:55 Anniepoo you can't do engineering with no idea what you intend to accomplish
18:56 Anniepoo you need to have some idea what sort of mechanism you want to design - "I need a mechanism that attaches to RobotA in X manner,
18:56 Anniepoo lifts it Y cm in Z seconds, and prevents me from turning over'
18:57 RobotDuty http://postimg.org/image/bq22zre2h/
18:58 Anniepoo thats the robot you're lifting?
18:59 RobotDuty The idea is to remplace the wheels (2) by 4 wheels and the engine (motor) to get faster and strong to stay alive.
19:00 RobotDuty Well, imagine everyone wants to get the center because they don't want the Robot A touch the color circle, we must have a strong robot.
19:00 Anniepoo whats that got to do with lifting a robot?
19:02 Anniepoo this is ridiculous
19:02 RobotDuty The idea is the replace the plexiglass by wood, and add a heavy lead shell on the robot to move to the center.
19:02 Anniepoo who cares
19:02 Anniepoo sigh....
19:03 RobotDuty Hum, I said 300 lbs, maybe 100 lbs or less, but imagine 8 Robot going to the center ^^
19:03 Anniepoo ok, this is your robot, or the opponents robot?
19:03 Anniepoo 300 lbs - not 300 kb, got it
19:03 RobotDuty My
19:03 Anniepoo ok,
19:03 RobotDuty This is my strategy to be heavy
19:03 Anniepoo So, if you lift your opponent
19:03 Anniepoo well, is there a weight limit?
19:03 RobotDuty Nope ^^
19:03 Anniepoo is there a size limit?
19:04 RobotDuty Yes, 30 x 30 x 30 (27000 cm³)
19:04 Anniepoo so, you want to be heavy - fill every square inch of that you can with lead
19:04 RobotDuty Did you miss the beginning of the conversation?
19:04 RobotDuty Because I post the dimension and other details up there
19:05 Anniepoo no, just am filtering a lot of your extraneous info out
19:05 RobotDuty Yes
19:05 RobotDuty ok, no problem.
19:05 Anniepoo you want to lift your opponent
19:05 RobotDuty I think it's important to mention how to make points (score)
19:05 Anniepoo have you figured out what a block of lead 30cm on side would weigh?
19:06 RobotDuty I think it's 656lbs ^^
19:06 RobotDuty Well, it's heavy for the human :p
19:06 veverak I am twisted and evil!
19:06 RobotDuty 150 lbs it's ok
19:06 Anniepoo ok, so yes, 300 lbs is a reasonable upper limit for a robot
19:06 veverak because I use dark magic in coding!
19:06 veverak nah nah nah nah
19:07 veverak so many kittens died because of me!
19:07 RobotDuty Yes
19:07 Anniepoo do you really need to lift it 1 meter?
19:07 veverak (I heard that you have to force yourself to like it for guilt to disappear)
19:08 RobotDuty Yes, but maybe I should explain you the points system before, because it's maybe not the best strategy
19:08 Anniepoo no
19:08 RobotDuty system points
19:08 RobotDuty Ok
19:08 Anniepoo don't do that
19:08 RobotDuty Ok
19:08 RobotDuty Yes
19:08 Anniepoo I'm happy to help you with basic mechanical engineering
19:08 Anniepoo but not to design your robot for you
19:08 RobotDuty All right :)
19:09 Anniepoo so, the issue is power
19:10 RobotDuty It's exactly 3 meters (diameter) so 1.5 meters the radius
19:10 RobotDuty Yep
19:11 Anniepoo lifting a bot requires (force) * (distance)
19:11 rue_house arg, I'm always running out of bleach!, I bought this bottle just, just, 3 years ago!
19:11 Anniepoo weight of bot is 136 Kg
19:11 RobotDuty I though a drill under the plank of wood with 4 wheel (like in the car), 2 to move and 2 to keep stability would be a good idea.
19:11 rue_house maybe the bottle is leaking
19:12 Anniepoo now you're talking about moving, not lifting
19:12 RobotDuty lifting my own robot, sorry,
19:13 RobotDuty cheap wheel to lift that and powerful motor to move that RObot and other Robot that are going to hit me by "accident".
19:14 RobotDuty We can't attack Robot, but accidental collisions is ok
19:14 RobotDuty others Robots*
19:15 Anniepoo you want to lift your own robot?
19:15 RobotDuty "support"
19:15 RobotDuty maybe wrong word :S
19:16 Anniepoo English is not your first language?
19:16 RobotDuty Nope :S
19:16 Anniepoo ok
19:16 Anniepoo so, you just want the wheels to hold the robot off the floor a few mm so it can roll
19:17 Anniepoo yes?
19:17 RobotDuty I do the best I can with English. I know, I'm really not an expert with the language. Sorry for that incovenient.
19:18 Anniepoo no I am sorry
19:18 Anniepoo I'll speak plainly
19:18 RobotDuty Imagine, I have 300 lbs on my RobotA. I want the RobotA to move from point A to point B. Distance : 1.5 meters
19:18 Anniepoo ok
19:19 Anniepoo how fast?
19:19 RobotDuty As fast as I can, but, to be honest, it's 300 lbs, so the RobotA have to stop too, not to be dangerous for the other people around the battle.
19:20 Anniepoo well, people should be kept back from two 300 lb fighting robots
19:21 RobotDuty Well, I have to choose. Stability, heavy Robot, or faster Robot like RC auto.
19:21 RobotDuty two*
19:21 RobotDuty I don't think I can combine both under the weight of 300 lbs
19:22 Anniepoo well, you're designing a bulldozer
19:22 Anniepoo a machine to push
19:22 RobotDuty Yep ^^
19:23 RobotDuty A tank :p
19:23 RobotDuty without ammo
19:23 Anniepoo so, Newton's 2nd law - it can push against other robot only as hard as it can push on ground
19:25 RobotDuty Hum, but sounds tricky to detect the robot to your left or to your right to push on them to get faster to the center.
19:26 RobotDuty I might forget this information... the Robot are not done by remote control, but with programmation.
19:26 Anniepoo 8cD make it stupid - at start it rolls forward to center and becomes a lump. No sensors
19:28 RobotDuty 8cD ?
19:28 Anniepoo my smiley face
19:28 RobotDuty ha ok :p
19:29 RobotDuty so, yes it's the idea.
19:29 RobotDuty Do you think that a drill could do the job?
19:30 Anniepoo any motor can generate any amount of force if two conditions hold
19:30 Anniepoo a) there's enough mechanical advantage
19:30 Anniepoo b) nothing in the system slips or fails
19:31 RobotDuty Do you have source or documentation that help me learn more on motor and connection with electronic devices ?
19:31 Anniepoo look up 'permanent magnet motor' on wikipedia
19:32 Anniepoo drill motors are permanent magnet motors (permag motor)
19:33 Anniepoo There's a simple device for measuring torque, called a Prony brake.
19:33 Anniepoo put a wheel of some sort (with a flat tread) on the motor shaft.
19:34 Anniepoo hang a belt over the wheel, and clamp it to desk at one side
19:34 RobotDuty de Prony brake from Wiki ?
19:34 Anniepoo yup
19:34 RobotDuty torque by rotation speed
19:34 Anniepoo hang a weight off the other side
19:35 Anniepoo turn on motor, vary the weight, measur rotation speed
19:35 Anniepoo make a graph
19:36 RobotDuty how do I measure the rotation speed?
19:36 Anniepoo now you have DATA, not guesses
19:36 Anniepoo do you have an oscilloscope?
19:36 RobotDuty yes
19:37 Anniepoo paint alternating black and white lines on face of wheel.
19:37 RobotDuty not right now, but I can use one
19:38 Anniepoo rig up a phototransistor so it shows one voltage when its looking at white, another at black
19:38 Anniepoo measure frequency using oscilloscope
19:38 RobotDuty "rig up a phototransistor" I don't understand
19:39 RobotDuty sounds like an encoder
19:39 Anniepoo yes
19:39 RobotDuty ha ok :)
19:39 Anniepoo if you have an encoder, just use that
19:40 Anniepoo if the encoder is fine enough, you don't even need the scope. Hook encoder to small amplifier and ask music teacher what tone it is
19:40 Anniepoo 8cD
19:41 RobotDuty there is 64 "clicks" on our wheels
19:41 Anniepoo great
19:41 Anniepoo every 64 clicks you have a revolution
19:41 RobotDuty yep
19:41 Anniepoo high school math
19:41 RobotDuty Yep
19:42 Anniepoo now, with a torgue vs speed curve for motor
19:42 RobotDuty But how do I plug the "wheel system" with the drill?
19:42 Anniepoo design a mechanism to move robot - some gears and a wheel sounds right
19:42 Anniepoo I usually don't disassemble the drill all the way
19:43 Anniepoo take the case off, but leave the chuck and the gearing as they were
19:43 RobotDuty like that : http://www.instructables.com/id/Hacking-Drills-CheapPowerful-GearboxMotors-for-/ ?
19:43 Anniepoo now if you can attach a shaft to your wheel, you can put it in the chuck like a drill bit
19:44 Anniepoo yes
19:44 Anniepoo I have a robot that uses 3 of those
19:44 Anniepoo I paid $14 each for them
19:44 Anniepoo they left the gear box on but took the chuck off
19:44 RobotDuty Wow, it's cheap :D
19:45 Anniepoo and they came with batteries
19:45 Anniepoo (I didn't use those, but most people making mobile robots do)
19:45 Anniepoo you can take the plastic case of the battery apart
19:48 RobotDuty How do you connect the drill to the wheel and the battery to the "PCB".
19:48 Anniepoo well, you need basic electronics soldering skills for battery to PCB
19:48 Anniepoo leave the chuck (the gripper for the bits) on the drill
19:48 RobotDuty Yes, I can solder stuff ^^
19:49 RobotDuty Oh wait, I will post the PCB if you want. It's 12 V and 1 A
19:49 Anniepoo bet your motor draws more than 1 amp
19:50 Anniepoo what axle does your wheel have?
19:50 RobotDuty http://postimg.org/image/ake1p9b5b/ (PCB)
19:51 RobotDuty hum, if I decide the design of 4 wheel, I guess, it will be are the top and the bottom under the wood's plank
19:51 veverak hmm
19:51 veverak soemtimes I wonder
19:51 veverak did I really needed function that would tell me how long it will take before my servo finishes it's move?
19:51 RobotDuty I have no ideas about the design of the wheel.
19:53 Anniepoo remember a and b , my rules I said before?
19:53 Anniepoo b) is 'as long as nothing slips or fails'
19:53 rue_shop3 I need bleach
19:53 RobotDuty Ok which means, everthing must work together ?
19:53 rue_shop3 and if this storm picks up I need gas or the genny
19:54 Anniepoo if wheel slips, that is the 'adhesion limit'
19:54 RobotDuty nothing should broke
19:54 Anniepoo wheel slip will be your force limit
19:54 Anniepoo wheel sideforce is sliding friction
19:54 Anniepoo (not rolling friction!)
19:55 Anniepoo so you need to know the 'adhesion factor' for your tire
19:55 RobotDuty Ha ok. So if it's too heavy, the wheel will not turn (rotation), but slip ?
19:55 Anniepoo no, if it's too light
19:55 RobotDuty light ?
19:55 Anniepoo take a toy RC car - childs toy
19:55 RobotDuty yep
19:55 Anniepoo run it against your foot
19:55 Anniepoo it'll slip
19:55 RobotDuty yes
19:55 Anniepoo now take a bulldozer....
19:56 RobotDuty I will loose my foot ^^
19:56 Anniepoo yes - tractive effort is limited by weight on driving wheels
19:59 Anniepoo max tractive effort = coefficient of adhesion * weight on driven wheels
19:59 RobotDuty Well, well, I'm not sure to understand the point with the design of the wheel.
19:59 Anniepoo would you make the wheels of teflon?
19:59 RobotDuty nop!
19:59 Anniepoo correct - because teflon has a low coefficient of friction
20:00 RobotDuty too smooth
20:00 RobotDuty right, ok
20:00 Anniepoo in the US they sell pans coated with teflon as 'non stick' pans
20:00 Anniepoo you want stick
20:01 Anniepoo I'm from rural US. most people drive pick up trucks.
20:01 Anniepoo in the winter they put sand in the truck so the wheels have more weight on them, to compensate for lower
20:01 Anniepoo coefficient of adhesion
20:01 RobotDuty Do you think it's enough the weight those wheel : http://postimg.org/image/ef07rgqxd/b5533915/
20:02 rue_shop3 no USE MORE LEAD
20:02 Anniepoo you need to CALCULATE that
20:02 Anniepoo not ASK RANDOM PEOPLE ON INTERNET
20:03 rue_shop3 MORE LEAD
20:03 Anniepoo it's an ENGINEERING competition, not a ASK RANDOM PEOPLE competition
20:03 Anniepoo 8cD
20:03 rue_shop3 MORE LEAD MORE LEAD
20:04 Anniepoo how much tractive effort do you need?
20:04 RobotDuty I know, but I need to know the basic before starting to build a "bulldozer".
20:04 RobotDuty I don't know exactly what is tractive force or effort :S
20:04 Anniepoo you just need to have numbers for things
20:04 Anniepoo not guess
20:05 RobotDuty I agree
20:05 Anniepoo tractive effort is how hard robot can 'shove'
20:05 Anniepoo it has to shove a little to move itself
20:05 Anniepoo the rest of the tractive effort can be used to shove other robots
20:06 Anniepoo you're in a tractive effort contest
20:06 Anniepoo MORE LEAD
20:06 Anniepoo rue - he should make the thing from woods metal encclosed in silicone bag, have a heater, and have it melt at the center
20:06 Anniepoo then the heater turns off
20:06 Anniepoo 8cD
20:07 Anniepoo it's sorta a good thing htey have students do this
20:08 RobotDuty And he has to 'shove' the other robot... ok, well. Hum, wait where is your formula, ok the max tractive effort is the max 'shove' effort to 'push' the other robots away, so I imagine he has per example to push 3 times me per example?
20:09 Anniepoo well, max TE = (coefficient of adhesion) * (weight on driven wheels)
20:09 Anniepoo pneumatic tires with low inflation have good adhesion
20:10 RobotDuty 600 lbs = X * 300lbs ? I need a coefficient of adhesion of 2 ?
20:10 Anniepoo well, if you want 600 lbs of TE from a 300 lb robot, yes
20:11 Anniepoo the formula is the same for the other robot
20:11 RobotDuty ^^
20:11 Anniepoo so whoever has the higher product of adhesion limit and weight shoves the other one
20:11 RobotDuty Ok :)
20:12 Anniepoo now, if you do start to slide, your adhesion limit goes down!
20:12 Anniepoo like moving a heavy box - once you get it started it slides easier
20:12 RobotDuty But if I slide, I will not be able to brake as soon as I want, right?
20:13 Anniepoo well, if youy have a system designed for high adhesion, braking shouldn't be an issue
20:13 Anniepoo lets ignore braking for now
20:13 RobotDuty ok
20:13 Anniepoo you're having enough trouble with shoving
20:14 RobotDuty You're right, thanks for letting me breathing :p
20:14 Anniepoo so, get busy googling 'adhesion limit'
20:14 Anniepoo you'll find lots of articles about drag racing
20:15 Anniepoo another thing to look at - look at walking drag lines
20:16 Anniepoo https://www.youtube.com/watch?v=PZvquNAXRe4
20:21 RobotDuty Maybe I can do a system like a train :D
20:21 RobotDuty But sounds too complex
20:22 Anniepoo trains have coefficient of adhesion of 0.35 on dry rails
20:22 Anniepoo on wet leaves can be really bad
20:22 Anniepoo trains work by keeping the tractive effort really low for the loads involved
20:23 Anniepoo if you stand near a freight train when it starts, you will hear a boom near the engine
20:23 RobotDuty http://5at.co.uk/index.php/definitions/terrms-and-definitions/adhesion-2.html ? ^^
20:23 Anniepoo the couplers have 'slack' - some deliberate sloppiness in length
20:23 Anniepoo the engine starts itself, then starts the first car, then the second, and so on, one at a time
20:24 Anniepoo A clean dry wheel on a clean dry rail may have a coefficient of friction as high as 0.35 or 35% at zero speed. This can drop to 0.25 in wet conditions, and very much lower if the rail is contaminated with lubricating substances such as ice, oil and leaves.
20:24 Anniepoo don't worry about the 'at speed' part
20:24 Anniepoo do you knwo what the playuing area is made of?
20:25 RobotDuty classroom floor probably
20:25 RobotDuty like a gym*
20:26 Anniepoo so, you can measure the coefficient of adhesion against such a floor
20:27 Anniepoo clamp wheels so they don't rotate.
20:27 Anniepoo weigh vehicle assembly
20:27 Anniepoo pull with rope with spring scale in it
20:28 Anniepoo highest reading before it moves is static coefficient of adhesion
20:28 Anniepoo after it starts moving, this will go down
20:28 Anniepoo be sure the rope is horizontal
20:29 RobotDuty Rope, be we can't touch the Robot :S
20:29 RobotDuty but*
20:29 Anniepoo how do you intend to build it then?
20:30 Anniepoo meant YOUR robot, before the match
20:30 Anniepoo measure things
20:33 RobotDuty Ha, I get it... I'm so stupid. spring scale is an instrument. Ok, I'm in business.
20:33 Snert can you spray paint the other robot's sensors?
20:34 RobotDuty I didn't understand the thing, but now I do. So you mean measure the Robot in static coefficient ?
20:34 RobotDuty nope
20:34 RobotDuty well, it's an attack, right?
20:35 RobotDuty Spray paint can be dangerous ^^
20:37 Anniepoo yes, measure the force required to start dragging robot
20:37 Anniepoo make changes, do it again
20:37 Anniepoo change tire pressure, tire composition,
20:37 Anniepoo tread pattern,
20:37 Anniepoo suspension
20:38 RobotDuty Good tips to know :)
20:38 RobotDuty What I don't understand is the "wheel mechanism" connected with the drill.
20:38 RobotDuty gearing, but how?
20:39 Anniepoo use the gears already on the drill
20:40 Anniepoo decide how much torgue you need
20:41 RobotDuty yep, ok... I get it
20:41 Anniepoo from the diameter of wheel and tractive effort you can compute torque
20:41 RobotDuty and what is the distance from the plank wood ? What are the parameter for a good distance?
20:41 Anniepoo make the graph of torque vs speed for motor
20:42 Anniepoo make torgue at motor be enough to have medium speed for motor
20:43 Anniepoo figure out 'gear ratio' between axle and motor
20:43 RobotDuty wheel axle ?
20:43 Anniepoo whats your wheel look like?
20:44 RobotDuty Actually, it's those wheel : http://s8.postimg.org/na121zfpx/wheel_robot.png
20:44 RobotDuty But, it's maybe better to change, because the screw are maybe not enough strong
20:45 Anniepoo well, I don't see the axle there
20:45 Anniepoo but it's mounted to something
20:45 RobotDuty Hum... plexiglas ?
20:45 RobotDuty plexiglass*
20:46 Anniepoo no, the axle
20:46 RobotDuty There is no axle, just motor and screw and wheel :S
20:47 Anniepoo I need pictures of the other side of wheel, closer in to see the hub
20:47 Anniepoo or a part number or maker name for wheel
20:48 RobotDuty http://postimg.org/image/spzzfki6j/
20:48 RobotDuty The wheel are custom made
20:49 RobotDuty I don't know the number or make name, sorry.
20:49 Anniepoo ok
20:49 Anniepoo well, every wheel is different
20:49 Anniepoo they have a 'hub'
20:49 RobotDuty hub means the things that connect the axle?
20:50 Anniepoo yes
20:52 RobotDuty So, I need to know if the wheel are the perfect condition of support the 300 lbs, the tractive force and the connection with the drill. Sounds a complex problem :)
20:52 RobotDuty Do you know the seller for cheap wheel ?
20:53 Anniepoo where are you located?
20:53 RobotDuty Canada
20:54 RobotDuty RobotShop or things like that?
20:54 Anniepoo http://www.harborfreight.com/
20:54 Anniepoo they sell cheap wheels
20:54 Anniepoo children's ride- on toys
20:55 Anniepoo http://www.harborfreight.com/catalogsearch/result?q=tires
20:56 RobotDuty Maximum Working load (lbs.) 300 lb. ok :)
20:56 RobotDuty Sounds the limit of their cheap wheel product
20:57 Anniepoo http://www.harborfreight.com/8-inch-non-marring-rubber-tire-47638.html
20:58 Anniepoo overloading might be wise for your application
20:58 Anniepoo though of course you don't want a flat during contest
21:00 RobotDuty Does it take time to get the parts ?
21:01 RobotDuty (How long)
21:01 Anniepoo consider getting an inner tube and using it instead of the normal tire
21:01 Anniepoo for me it's a 2 hour drive to nearest store
21:02 RobotDuty ok
21:02 RobotDuty 8 in means 20 cm the wheel ?
21:05 Anniepoo no idea
21:06 Anniepoo welcome to engineering
21:06 Anniepoo you get your keister down to the local tire store and ask questions
21:06 RobotDuty ^^
21:07 RobotDuty I think it's not called a metal, but a piece of circle metal :p
21:07 RobotDuty tire*
21:09 RobotDuty Or the wheel are each side of the robot, but I got only two... hum.
21:12 Anniepoo tire is the rubber part. wheel is the metal center that holds tire
21:12 Anniepoo hub is what wheel bolts to
21:12 RobotDuty and gear?
21:12 RobotDuty "gearing"?*
21:14 RobotDuty I don't get the point of axle to motor :S
21:14 Anniepoo you know, sometimes I htink there's something wrong with using fighting robots as an educational experience
21:14 RobotDuty and hub to axle to motor
21:14 Anniepoo it teaches a sort of optimized engineering that's not good - need to teach satisficing
21:14 Anniepoo anyway, I'm outta here
21:15 Anniepoo time for food
21:15 RobotDuty Bon appetit :p
21:16 RobotDuty (Enjoy your meal)
22:06 RobotDuty Thanks for your time. Good night :)
22:21 rue_shop3 well I'll be,
22:21 rue_shop3 I think the problem with the reprap warping prints has been cause of finger oil on the tape bed
23:08 Snert what did you clean it with?
23:12 Tom_itx fingerprint remover
23:19 rue_shop3 acetone