Back
[00:01:44] <pokmo> i've tried using an u-joint, but the swivel angle is a bit too small
[00:11:55] <z64555> metal can have lower friction with a drop of lube
[00:12:15] <z64555> debris is still problematic
[00:21:02] <anniepoo> take steel sphere end. axially drill a bit into some plastic rod.
[00:21:34] <anniepoo> heat sphere end, push into rod. Push edges back around ball with screwdriver blade
[00:21:53] <anniepoo> after it cools, trim away yuckky bits on edge
[00:22:02] <z64555> that pretty much requires a lathe to do right
[00:22:11] <anniepoo> 8cD
[00:22:28] <z64555> drill press is real iffy
[00:22:51] <anniepoo> yah, I need to keep a scorecard in here of people's skills and tools
[00:32:17] <pokmo> but if i want to achieve inline rotation, are there other parts i could go with?
[00:34:05] <anniepoo> a U joint comes to mind
[00:34:54] <anniepoo> I take it you don't want to transmit torque
[00:35:02] <pokmo> anniepoo, yeah, i've tried that. but the angle at around +/- 45 deg is much smaller
[00:35:09] <pokmo> anniepoo, that's right. no torque requirement
[00:35:33] <anniepoo> forgetting the name of this thing
[00:35:49] <anniepoo> there's a pendulum, often an exhibit in science museums
[00:36:01] <pokmo> hmm
[00:36:02] <anniepoo> it's on a swivel joint
[00:36:18] <anniepoo> the pendulum of course stays in-plane, and earth rotates
[00:36:22] <anniepoo> so it makes a clock
[00:36:50] <pokmo> hmm but it's on a joint?
[00:37:04] <anniepoo> Foucault pendulum
[00:37:23] <pokmo> isn't it hung on a wire?
[00:37:48] <anniepoo> yes, but it won't work to just clamp the wire at the top
[00:37:59] <anniepoo> bending the wire will drain all the energy out of it
[00:38:42] <pokmo> anniepoo, how about this construction
http://visite.artsetmetiers.free.fr/site_anglais/images/foucault/fixation_m_a.jpg
[00:38:52] <pokmo> a stylus on a cone
[00:39:11] <anniepoo> is the force guaranteed to be in tension?
[00:39:29] <pokmo> but what sort of joint are you thinking of?
[00:39:35] <anniepoo> that'd certainly work if you have a large enough jewel
[00:40:30] <anniepoo> what's the down force on your device?
[00:41:13] <pokmo> Not much. Probably around 0.5N
[00:41:21] <pokmo> I mean 5N
[00:42:06] <anniepoo> then just make a clevis joint with good bearings
[00:42:18] <pokmo> Basically I'm trying to make a small moon tracker, controlled by 3 arms
[00:42:33] <anniepoo> the shaft of the clevis goes through a thrust bearing
[00:42:43] <pokmo> The arms themselves are ball links driven by servos
[00:42:47] <anniepoo> ah, you're reducing sticktion?
[00:42:55] <anniepoo> then the jewel's a good choice
[00:43:24] <anniepoo> the secret to the joint you posted is that the stylus and cone are very hard materials
[00:43:33] <pokmo> But because the ball links could shift the platform, I need a way to hold it fixed
[00:43:39] <anniepoo> this type of bearing used to be used in high quality watches
[00:43:56] <pokmo> Hmm what's the jewel
[00:44:03] <anniepoo> often garnet
[00:44:28] <pokmo> Yeah, it looks pretty high precision
[00:44:43] <pokmo> I was just hoping for something simple but works
[00:44:49] <anniepoo> old watches used to say '7 jewel' or something
[00:45:02] <anniepoo> oh, I know
[00:45:11] <pokmo> The platform is only around 6cm by 6cm
[00:45:20] <anniepoo> hmm.... how accurately does it need to hold the center position?
[00:45:35] <anniepoo> would repulsing magnets work?
[00:46:03] <pokmo> Well, I don't have quantitative figures.. But I guess something like those plastic ball links could do
[00:46:14] <pokmo> Shame that they're so small and they aren't linear
[00:46:32] <pokmo> Hmm haven't thought about magnets..
[00:46:59] <anniepoo> I was a physics major
[00:47:01] <pokmo> Magnets along its axis? But I'm trying to prevent the platform from shifting
[00:47:20] <anniepoo> in advanced lab we had a thing for doing rotation experiments
[00:47:33] <pokmo> That's probably out of my league...
[00:47:44] <anniepoo> the bearing was pretty simple
[00:47:54] <pokmo> I was hoping for some existing part that I could attach to the setup
[00:47:55] <anniepoo> big (4") steel ball
[00:48:16] <pokmo> brb someone's at the door
[00:48:30] <anniepoo> put in a spherical die cast hole, with epoxy on it
[00:48:41] <anniepoo> (and some release agent on ball)
[00:48:53] <anniepoo> hole at bottom admitted air
[00:49:01] <anniepoo> ez to make air bearing
[00:49:05] <anniepoo> hella accurate
[00:49:18] <anniepoo> brb, going to check temp of parts
[00:52:01] <z64555> pokmo: what are you making this time?
[00:57:31] <rue_house> can you use playdough with vacuum forming?
[00:57:52] <rue_house> shape whatever you want and suck it into existance?
[00:58:57] <anniepoo> pokmo, I think you're best off with some low starting friction bearings and a clevis
[00:59:43] <rue_house> I think I figured out how to do a good job with threaded rod as lead screws
[01:01:41] <anniepoo> pokmo, if it's sticktion you're worried about
[01:01:46] <anniepoo> and the angles not huge
[01:01:53] <anniepoo> how about a flexible membrane?
[01:02:10] <anniepoo> https://www.engineersedge.com/bearing_types.htm
[01:03:01] <z64555> rue_house: playdough is essentially colored clay, but I haven't heard of clay being used in vacuum forming as the media
[01:03:10] <anniepoo> you can't be both demanding extreme performance and things you can make on your kitchen table
[01:03:48] <anniepoo> playdough is dough, with dye and a lot of salt in it.
[01:03:57] <anniepoo> 8cD I've made my share of it
[01:06:09] <z64555> huh, you can use it to clean up coal soot
[01:06:21] <anniepoo> ooh, you CAN?
[01:06:34] <anniepoo> cause I have a real use for that!
[01:09:11] <rue_house> what about shapes for vacuum forming?
[01:09:33] <anniepoo> well, I've seen people use modeling clay
[01:09:36] <rue_house> I want to see a vacuum formed playdough shell!
[01:09:43] <anniepoo> lol
[01:09:49] <anniepoo> actually, that's not a bad idea
[01:09:50] <z64555> anniepoo: according to wikipedia, its original purpose was exactly that
[01:09:57] <anniepoo> yah, see that
[01:10:09] <rue_house> modeling snail shells or cleaning up coal?
[01:10:26] <anniepoo> I managed to make a 3D model that's acceptable
[01:12:16] <rue_house> what file types does the vacuum former take?
[01:12:33] <rue_house> does it just use an actuated pin bed?
[01:12:42] <anniepoo> ??
[01:12:56] <rue_house> or goes a system position the pins and then lock them for forming?
[01:13:21] <anniepoo> well, the plan is to cut out layers from cardboard
[01:13:30] <anniepoo> cover with plaster wrap
[01:13:38] <anniepoo> smooth it out
[01:13:39] <rue_house> using a laser powered cardborad fire!
[01:13:50] <anniepoo> then fiberglass mold it
[01:13:57] <rue_house> laser fired... cardboard...
[01:14:13] <anniepoo> I suppose I could make a vacuum form machine, but
[01:14:18] <rue_house> (8-} ooo resin... weeee....
[01:14:34] <z64555> she's not going to be mass producing the buggers
[01:14:43] <rue_house> playdough
[01:14:49] <anniepoo> yah, I've never done any fiberglass, so this seems liek a good way to learn
[01:14:54] <rue_house> meh
[01:15:01] <rue_house> I was planning to start with a boat
[01:15:14] <rue_house> use sand and faom to shape a hull
[01:15:21] <rue_house> (pile of sand)
[01:16:33] <rue_house> shape sand, cover in poly, spray with 2" layer of spray foam, shape, fiberglass, flip, shape, inside, fiberglass = hull
[01:16:37] <anniepoo> well, I was thinking about using the interlocking rows and columns method to make the cardboard. I could use something to stuff in the holes
[01:16:51] <rue_house> oo
[01:17:04] <anniepoo> and you know, play-doh's probably the cheapest option.
[01:17:09] <anniepoo> terracotta's too hard
[01:17:37] <rue_house> :) I'm helping!
[01:17:53] * rue_house goes over and tries to vacuum form his hand...
[01:18:05] <rue_house> IT STINGS A LITTLE!
[01:18:08] * anniepoo rushes rue to emergency
[01:18:31] <z64555> lol.
[01:18:34] <rue_house> OOOooooOOOOoooo thats quite a colour of red!
[01:19:07] <anniepoo> brb, gotta go burn myself, er... take part out of oven
[01:19:27] <rue_house> oh, hey, I know what I could vacuum form...
[01:19:51] * rue_house puts a stool infont of the former and starts taking his belt off...
[01:20:05] <z64555> hm, oh.
[01:20:21] <z64555> anniepoo: play doh's missing the detergent from its wall cleaner formula
[01:20:46] <anniepoo> ah
[01:20:52] <anniepoo> well, we have a laser
[01:21:08] <rue_house> you use a laser for cleaning coal dust?
[01:21:16] <anniepoo> people ask us to do stuff like jigsaw puzzles, it's impractical to manually clean the edges
[01:21:49] <anniepoo> weird factoid - 'magic eraser' sponges are melamine foam.
[01:22:00] <anniepoo> they work because at microscale it's hard melamine
[01:22:12] <anniepoo> but it's foam so it deflects
[01:22:33] <anniepoo> 8cD my part's out of the oven
[01:22:43] <anniepoo> with any luck it's OK and I can freaking go home
[01:24:09] <z64555> hm, according to my reading, there wasn't a whole lot of detergent in it
[01:24:10] <z64555> I bet
[01:24:26] <z64555> If you take play dough and mix in some dish soap
[01:24:32] <z64555> you'll get something workable
[01:24:45] <anniepoo> might be
[01:25:15] <anniepoo> now I'm wondering if I could shred magic erasers and put stuff in the vibratory polisher
[01:25:41] <z64555> You'd need a roller mixer probably to get the detergent nice and evenly mixed in
[01:26:00] <rue_house> a good oppotunity to make a tumbler
[01:26:23] <anniepoo> we have a bowl type vibratory polisher
[01:26:25] <rue_house> you can use it to grind bentonite kitty litter into fine powder to make casting sand
[01:26:34] <rue_house> or to polish rocks
[01:26:45] <rue_house> or to shred magic erasers?
[01:26:55] <anniepoo> and one of those tiny 'rock hound' tumblers
[01:27:00] <rue_house> to mix detergent?
[01:27:01] <anniepoo> and a motorized meat grinder
[01:27:07] <rue_house> hmm
[01:27:20] <anniepoo> the meat grinder's the machine for the magic eraser trick
[01:29:37] <rue_house> why are you grinding erasers?
[01:30:38] <z64555> to make polishing compound
[01:30:54] <rue_house> oh?
[01:31:43] <z64555> ah, didn't know that about the kitty litter
[01:31:56] <rue_house> bentonite clay
[01:32:03] <rue_house> not food quality, but cheap
[01:32:11] <z64555> what kind of casts can it do?
[01:32:12] <rue_house> binds the sand for casting
[01:32:20] <rue_house> I do aluminum
[01:32:46] <rue_house> YOU HAVE TO DEGASS ALUMINUM
[01:32:55] <rue_house> ~:-|
[01:34:22] <rue_house> have you ever picked up an example of some code off the internet and as you were working on it you realized what a frankenstien mess it really was?
[01:34:25] * anniepoo belches really loud - sorry, degassing!
[01:34:41] <rue_house> this file is like less than 58 lines, and its horrid
[01:34:45] <anniepoo> yes - quite frequently
[01:34:48] <rue_house> hehe
[01:36:09] <anniepoo> gnite folks
[01:36:16] <anniepoo> lf8cP finished long job
[01:36:39] <rue_house> it gets worse, I just tired to look for where I got this code from, and there are a million versions of it multilated in different ways on the internet
[01:37:03] <z64555> yeah
[01:37:06] <z64555> that's "normal"
[01:37:31] <z64555> suprising thing, a lot of programmers get paid to make code that works, not necassarily code that's easy to maintain
[01:39:48] <rue_house> http://stackoverflow.com/questions/21717115/using-tk-to-create-a-text-editor
[01:40:20] <rue_house> the code example with 0 upvotes seem to be what I started with
[01:40:22] <rue_house> !?!??!!
[01:41:30] <rue_house> I'v been rewriting stupid things out of it for a while now
[01:41:36] <rue_house> but I have a different goal
[01:41:43] <rue_house> which its helping me with
[01:42:09] <rue_house> I'm rewriting my NC spooler
[01:42:16] <rue_house> I'm ooping it
[01:44:29] <z64555> as long as you boop it too
[01:53:15] <rue_house> its a major rewrite
[01:53:23] <rue_house> 99%
[01:53:51] <rue_house> I was trying to also convert it from pack to grid, but I cant get it to stretch properly
[01:55:35] <rue_house> lets rethink this, exit is under the FILE menu
[01:55:39] <rue_house> WHY!!!!???
[01:55:49] <rue_house> the menu should be called someting else
[01:55:59] <rue_house> its nothing to do with a file
[01:56:25] <pokmo> anniepoo hi again. By low starting bearing with a clevis, is clevis the part that holds the bearing in place?
[01:57:09] <z64555> File->Exit "Exit viewing this file"
[01:57:53] <pokmo> As in
https://goo.gl/images/S31ZMd
[01:57:58] <pokmo> The black part?
[01:58:59] <z64555> https://upload.wikimedia.org/wikipedia/commons/thumb/0/08/Clevis_(PSF).png/220px-Clevis_(PSF).png
[01:59:24] <z64555> In case the arrow isn't enough, it's the U-shaped piece
[01:59:24] <rue_house> new, open, close, save, save ass, exit
[01:59:37] <rue_house> ... save ass? wish more programs had that option...
[01:59:37] <pokmo> But how would that allow the platform to swivel?
[02:00:10] <z64555> It can act as a simple union joint
[02:00:39] <z64555> not very flexibl, though
[02:01:00] <z64555> The clevis can pivot around the pin
[02:01:02] <rue_house> swivel?
[02:01:11] <rue_house> you want a rotating platform?
[02:01:40] <pokmo> Yeah I'm trying to make a moon tracker driven by 3 servos
[02:01:53] <pokmo> Each driving a ball link arm
[02:02:10] <rue_house> you just need 6 hinges, I think
[02:02:31] <pokmo> But because the ball links allow the platform to shift, I need to make it tied to the centre
[02:02:42] <rue_house> it wouldn't be stable
[02:02:58] <pokmo> The ball links drive the motion
[02:03:17] <rue_house> not a stewart platform...
[02:04:11] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:04:25] <rue_house> to maintain diagonal stability, 3 of them have to be hinges
[02:04:53] <pokmo> I think I need a ball joint of some sort
[02:04:59] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:04:59] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:00] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:00] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:00] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:00] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:01] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:05:06] <rue_house> just incase you didn't notice the image
[02:05:13] <rue_house> 3 of them are ball joints
[02:05:38] <rue_house> haha! someone made a 3d printable armidillo
[02:06:07] <rue_house> http://www.thingiverse.com/thing:196372
[02:06:11] <pokmo> Ah the red arms?
[02:06:26] <z64555> the silver bits underneath the arms, at the ends
[02:06:51] <z64555> sorry, underneath the platform
[02:07:04] <pokmo> But how are they hinges?
[02:07:08] <z64555> Each of the arms have a single ball joint
[02:07:19] <z64555> which is between the arm and the platform
[02:07:19] <pokmo> Aren't they ball joints?
[02:07:33] <z64555> that's what I and rue have just said
[02:07:34] <rue_house> ooooh ohohohoho
[02:07:52] <rue_house> 3d printrs use ball joints made with rare earth magnets and bearings!
[02:07:57] <rue_house> bearing balls
[02:08:07] * rue_house feel sorry for the bearings
[02:08:11] <rue_house> anyhow, no play
[02:08:15] * rue_house feel sorry for the bearings
[02:08:35] <rue_house> http://www.thingiverse.com/thing:408100
[02:08:40] <rue_house> pokmo, ^
[02:09:01] <rue_house> oh they make thigns now too
[02:09:02] <rue_house> huh
[02:09:17] <rue_house> you can do it with magnets and balls
[02:09:22] <rue_house> balls aren't hard to find
[02:09:47] * rue_house dosn't know if this conversation is working...
[02:10:28] <z64555> not really
[02:10:34] <z64555> Which one is the magnet?
[02:10:50] <z64555> the black tubes, right?
[02:10:51] <rue_house> its a flat magnet in the end
[02:11:02] <z64555> ok
[02:11:03] <rue_house> you made two plastic cups with magnets in the bottom
[02:11:09] <rue_house> and drop a ball into the middle
[02:12:01] <rue_house> http://www.thingiverse.com/thing:565520
[02:12:06] <z64555> What all does the ball contact?
[02:12:10] <rue_house> pokmo, have access to a 3d printer?
[02:12:19] <pokmo> Yeah I do
[02:12:26] <rue_house> the balls go in the cups
[02:12:30] <pokmo> Still wondering why mine isn't stable
[02:12:40] <pokmo> I'm using ball links rather than ball joints
[02:12:43] <rue_house> pokmo, are you checking out these thingiverse links?
[02:12:50] <pokmo> Yeah I am
[02:12:57] <rue_house> you only need 3 balls
[02:13:18] <rue_house> the other 3 should be hinges
[02:13:31] <z64555> yes, that much I figured out, the balls go in the cups. But where are its contact points?
[02:13:41] <rue_house> ?
[02:13:49] <z64555> Is the cup cavity cylindrical, spherical, or?
[02:14:00] <z64555> does the ball contact the flat magnet?
[02:14:06] <rue_house> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[02:14:15] <rue_house> ball dosn't touch the magnet
[02:14:23] <rue_house> that would cause friction and be gay
[02:14:29] <rue_house> there is a space
[02:14:41] <rue_house> the metal balls are pulled into the cups by the magnets
[02:15:25] <z64555> That's what I was figuring
[02:15:56] <rue_house> so the cups dont need to grab the balls
[02:16:06] <rue_house> which makes beter motion
[02:16:20] <pokmo> I'm using links like these
https://goo.gl/images/1tk7Mw
[02:16:38] <z64555> oh, you mean that the cups don't need to have the balls captive
[02:16:40] <rue_house> those are less free
[02:16:54] <rue_house> yea, free the balls with magnets
[02:16:58] <z64555> so there's less contact area between the cup and the ball
[02:17:02] <rue_house> then you can get more than a few degrees
[02:17:11] <rue_house> (O
[02:17:22] <rue_house> instead of @
[02:18:09] <rue_house> but, if you already have the trapped ball things, you can just use a 90 degree bracket with them
[02:18:23] <rue_house> the motion just wont be as good
[02:18:40] <rue_house> I'm gonna have a hard time getting to sleep tonight
[02:19:59] <rue_house> http://paste.debian.net/912395/
[02:20:02] <rue_house> gettin there tho!
[02:21:19] <rue_house> why do they return the file with the show method!?
[02:22:04] * z64555 tries to figure out how his triangle became a hexagon
[02:22:28] * rue_house puts on z64555 s glasses
[02:23:47] <z64555> i have a function that's supposed to clamp dots to be within a user-definable N-gon
[02:24:08] <z64555> I finally got it working, after fiddling around with it for a few hours
[02:25:01] <z64555> but, uh, the test poly was a triangle, and it came out to be a hexagon
[02:30:51] <rue_house> what langauge?
[02:38:53] <z64555> C++
[02:39:07] <z64555> but the function itself just uses C
[02:40:04] <z64555> I'm using a trigonometric/polar method instead of trying to mess with a vector solution
[02:41:41] <z64555> in particular, it uses the function r = 1/(sin(theta) + cos(theta))
[02:42:19] <z64555> This generates a line in polar form, if you step theta
[02:43:15] <z64555> You can rotate the line around its intercept point, r = 1, theta = 0; by adding to theta in both the sin and cos
[02:43:28] <rue_house> well then you prolly got degrees and radians mixed up
[02:44:04] <z64555> don't think so, I was working with rads from the start
[02:44:12] <rue_house> do you want my 2d libary?
[02:44:29] <rue_house> there are no speedling errs, really, I promise
[02:44:37] <z64555> lol
[02:45:07] <rue_house> you would think someone who cant type better than smashing their head on the keybaord couldn't code
[02:45:12] <z64555> The fact that it's a perfectly formed hexagon tells me that I got the calculation of the arc length wrong somehow
[02:45:20] <rue_house> but hey, man, I'm living poof
[02:45:33] <rue_house> Pi is half a circle
[02:46:11] <z64555> arc_length = ((n - 2) * PI) / n;
[02:46:19] <z64555> that's what I have it as now
[02:46:31] <rue_house> # confirm "Pi is half a circle"
[02:46:35] <z64555> n is the number of sides on the polygon
[02:46:37] <rue_house> yes, pi is half a circle
[02:46:38] <rue_house> #
[02:47:09] <rue_house> yup theres your problem
[02:47:21] <rue_house> PI IS HALF A CIRCLE
[02:47:30] * rue_house jobs z64555 with a cattle prod
[02:47:44] <rue_house> HHHHAAAALLLLFFFF AA CIIIIICCCLLLLLLEEEE
[02:48:24] <rue_house> your know what, its after midnight
[02:49:15] <rue_house> HALF HALf HAlf Halph hal ha a a ...
[02:50:02] <rue_house> I'm going to bed
[02:50:09] <z64555> the formula for the arc length checks out though.
[02:50:14] <z64555> sigh
[02:52:27] <rue_house> what happens is you use =2*PI/n
[02:53:24] <rue_house> maybe its not doing that I think it is
[02:53:30] <rue_house> have to see the rest of the code
[02:53:34] <rue_house> beditme
[05:12:17] <pokmo> rue_house, looking at that tripod again, the ball joints are meant to move linearly - right? if so, could these be linear servos of some sort?
[05:12:38] <pokmo> http://alioindustries.com/wp-content/uploads/2014/03/ceramic-servo-motor.jpg
[07:16:33] <veverak> deshipu:
http://www.gadgetify.com/diy-3d-printed-quadruped-robot/
[07:18:33] <deshipu> looks nice
[07:19:04] <deshipu> a bit too big for my taste
[07:20:31] <deshipu> veverak: I'm thinking about a lasercut body for Tote, but I have problems coming up with good design for the servo brackets
[07:21:08] <veverak> hmm
[07:21:17] <veverak> with lasercut only it has to be tricky
[07:21:28] <deshipu> or use glue :)
[07:21:43] <deshipu> dovetail joins and glue
[07:22:04] <veverak> true
[07:22:18] <veverak> but I usually prefer spacing columns
[07:22:20] <veverak> :D
[07:22:53] <deshipu> the problem is the joint on the back of the servo
[07:23:01] <veverak> deshipu: anyway, sort of like the neopixel ring on the previous one
[07:23:51] <deshipu> did I show you
https://cdn.hackaday.io/images/3394331478778433037.jpg ?
[07:24:14] <veverak> nope, but I saw it!
[07:24:17] <veverak> deshipu: and that looks awesome
[07:24:19] <veverak> :)
[07:24:32] <veverak> https://www.aliexpress.com/item/High-Quality-New-Arrival-8-Bits-LED-Ring-16-x-WS2812-5050-RGB-LED-with-Integrated/32456801618.html?spm=2114.01010208.3.56.7tdIvV&ws_ab_test=searchweb0_0,searchweb201602_1_10065_10068_10000032_119_10000025_10000029_430_10000028_10060_10062_10056_10055_10000062_10054_10059_10099_10000022_10000012_10103_10000015_10102_10096_10000018_10000019_10000056_10000059_10052_10053_10107_1
[07:24:34] <veverak> 0050_10106_10051_10000053_10000007_10000050_10084_10118_10083_10119_10000047_10080_10082_10081_10110_10111_10112_10113_10114_10115_10116_10000041_10000044_10078_10079_10000038_429_10073_10000035_10120_10121-10116_10120,searchweb201603_10,afswitch_1_afChannel,single_sort_3_default&btsid=2fcb6c3f-035a-44e7-aa64-d827a5eedada <3
[07:28:07] <deshipu> I have a bunch of those
[07:28:15] <deshipu> I was initially planning to put them into those eyes
[07:28:25] <deshipu> but in the end decided to use plastic lenses instead
[07:29:46] <deshipu> veverak: I would really want to get this kit:
http://zoobotics.de/project/zuri-01-3/
[07:30:00] <deshipu> just the paper, I mean
[07:30:08] <veverak> same here
[07:30:11] <deshipu> I have my own electronics
[07:30:23] <veverak> exactly :D
[07:30:23] <deshipu> I once wrote them offering help with gait programming
[07:30:29] <deshipu> but they weren't interested
[07:30:48] <veverak> nah, finish exams -> work on actual code for my quadruped -> make it presentable -> find somebody that buys me dynamixel and get this kit
[07:30:50] <veverak> more or less
[07:30:52] <veverak> :)
[07:32:09] <deshipu> I should probably ask them if they wouldn't sell me one kit
[07:32:30] <veverak> deshipu: tell me if you succeed
[07:32:45] <veverak> anyway, I am sort of planning to go 3d printing
[07:32:54] <veverak> and modify my tote so it's easy to attach things on it
[07:33:02] <veverak> another electronics | design parts
[07:33:08] <veverak> but I am worried about weight now :/
[07:33:39] <deshipu> yeah, that's another reason to keep it gangsta
[07:33:55] <deshipu> put too much plastic on it and it becomes heavy
[07:34:44] <veverak> yep
[07:34:59] <veverak> deshipu: was thinking of wheeled carriage with battery
[07:35:16] <veverak> with some way to disconnect it and reconnect
[07:35:31] <veverak> because hey
[07:35:32] <deshipu> ideally I would use the plastic they use for blisters
[07:35:45] <veverak> imagine this
http://www.aspencarriage.com/Graphics/Ben%20and%20Champaigne2.jpg but tote looking like spider instead of horse
[07:35:46] <veverak> <3
[07:36:06] <deshipu> that kinda defeats the advatnages of walking
[07:36:44] <veverak> "disconnectable
[07:36:48] <veverak> * "disconnectable"
[07:36:49] <deshipu> I think it would look better then with this configuration:
https://www.youtube.com/watch?v=-zOUz9VLRj0
[07:37:08] <veverak> wow :D
[07:37:21] <veverak> deshipu:
https://www.youtube.com/watch?v=giS41utjlbU I feel like newbie
[07:37:52] <deshipu> yeah, I saw this this morning
[07:38:29] <deshipu> imagine guns on it and it gets scary
[07:38:40] <veverak> yeah
[07:39:25] <veverak> next world war with robots
[07:39:28] <veverak> definetly
[07:39:59] <veverak> anyway
[07:40:04] <veverak> last week of exams
[07:40:04] <deshipu> nah, world wars are so yesterday
[07:40:11] <veverak> going to code after that
[07:40:22] <veverak> ~month to have something I hope :)
[07:40:26] <deshipu> there will only be acts of terrorisms and acts to prevent terrorism
[07:40:43] <veverak> https://www.youtube.com/watch?v=YfSR4UjtUHc
[07:40:45] <veverak> looks nice
[07:43:53] <deshipu> yeah, he did some serious work
[07:44:00] <deshipu> shame he's not updating anymore
[07:44:15] <deshipu> maybe it would be good to reach out
[07:46:19] <deshipu> he did 3d-printed lightweight brackets:
https://www.youtube.com/watch?v=sj7nQDp89Bo
[07:46:59] <veverak> nice
[07:47:05] <veverak> hmmm
[07:47:30] <veverak> when I think of it, it maybe would make sense to attach laser at the bottom of the robot near center of mass
[07:47:33] <deshipu> but I think he just glued them
[07:48:20] <deshipu> veverak: I have led light that makes a horizontal line
[07:48:37] <deshipu> veverak: was thinking about using it to scan the terrain
[07:48:59] <deshipu> veverak: you know, filter the image so that you only see the line, and from that you have a slice of the terrain
[07:49:08] <veverak> sure
[07:49:13] <deshipu> veverak: and as you move forward, you get more of those slices
[07:49:18] <veverak> deshipu: I still hope SLAM will be usable on rpi3
[07:49:20] <deshipu> then you can plan where to place feet
[07:49:21] <veverak> :)
[07:49:37] <veverak> but found out that ROS does not really have proper storage for 3d analysis of surroundings
[07:49:41] <veverak> or I did not found one
[07:49:47] <veverak> some voxel tree would be nice
[07:49:56] <veverak> that could take input from multiple sources and reasonably merge them together
[07:51:39] <deshipu> veverak: can you 3d print nylon?
[07:54:28] <veverak> should be able in hackerspace
[07:54:41] <veverak> but material 404, experience 404
[09:12:13] <pokmo> i'm looking at delta 3d printer's arms like this
https://s-media-cache-ak0.pinimg.com/originals/6f/18/d6/6f18d66e0e7f11aa77e701df88214eba.png
[09:12:30] <pokmo> my understanding is that the head's stable
[09:12:48] <pokmo> but if i have a design like
http://i.imgur.com/eOEOE5Z.png the head wouldn't be stable
[09:13:28] <pokmo> because the clevises are free to rotate
[09:13:42] <pokmo> how are the two different in design though?
[11:31:54] <z64555> blah, I got the equation wrong
[11:31:59] <z64555> for the arc length
[11:32:44] <z64555> The equation I have/found is actually for beta, the interior angle of each sector
[11:33:14] <z64555> well, beta is actually half that, but anyway
[11:59:37] <z64555> ok, this seems to get the correct number of sides, but the intercept is no longer at 1
[12:34:17] <veverak> \o/
[12:34:24] <veverak> json serialization done
[13:03:59] <deshipu> isn't that like just "import simplejson"?
[13:04:17] <veverak> deshipu: complex data structure in C++
[13:04:41] <veverak> and it does not contain everything ... yet
[13:07:20] <veverak> deshipu: python will be on the other side to visualize it
[13:07:52] <veverak> rpi3 will use C++ code to pack the data and send it over network to laptop that will use python visualizer
[13:09:46] <veverak> https://bpaste.net/show/55e47c375721 for now ;)
[13:10:17] <z64555> hm, the intercept is floating, depending on how many sides the ngon has
[13:10:22] <z64555> why...?
[13:10:26] * z64555 mumbles
[13:38:28] <z64555> well that explains why
[13:38:31] <z64555> shoot.
[13:41:15] <z64555> dr/dtheta changes depending on the input offset
[14:48:58] <z64555> why haven't I been using excell for this?
[14:49:00] <z64555> sheesh
[18:11:47] <anniepoo> somebody was on the other night with an issue with a vacuum pump that arrived DOA
[18:12:12] <anniepoo> I now need basically the same pump. Naturally I'd like to avoid that brand
[18:12:26] <anniepoo> anybody happen to remember what brand they were having trouble with?
[18:14:06] <z64555> Do you know which day this was?
[18:14:15] <anniepoo> hmm
[18:14:26] <anniepoo> maybe 8-10 days ago?
[18:14:28] <Tom_L> zlog
[18:14:33] <Tom_L> search for pump
[18:14:40] <Tom_L> pick the day you want
[18:14:40] <anniepoo> thanks Tom
[18:15:06] <z64555> the zlog only covers like, this day
[18:16:46] <Tom_L> not if you hit the back button
[18:17:10] <z64555> I beg to differ! That's a link!
[18:17:11] <z64555> :P
[18:17:11] <Tom_L> http://tom-itx.no-ip.biz:81/~tom-itx/irc/logs/%23robotics/index.html
[18:17:25] <Tom_L> i have years worth
[18:17:31] <Tom_L> how many do you want?
[18:19:01] <Tom_L> well, back to 2011 anyway
[18:22:00] <anniepoo> very strang
[18:22:09] <anniepoo> I know it was this month
[18:22:23] <anniepoo> but I've now looked thru whole month
[18:23:19] <z64555> oh!
[18:23:24] <z64555> I remember the one you're talking about
[18:23:28] <z64555> search for "diaphram"
[18:23:51] <anniepoo> ah!
[18:24:01] <anniepoo> I looked for pump
[18:25:00] <z64555> Check the log for the 1st of Jan
[18:25:22] <anniepoo> thanks
[18:25:22] <Tom_L> http://www.ebay.com/itm/Thomas-Gardner-Denver-014CDC20-12-E-Compressor-Vacuum-Pump-/231575894189?_trksid=p2141725.m3641.l6368
[18:25:28] <Tom_L> is that it?
[18:25:40] <anniepoo> yup
[18:25:41] <anniepoo> thanks
[18:25:46] <Tom_L> <ferdna>
http://www.ebay.com/itm/Thomas-Gardner-Denver-014CDC20-12-E-Compressor-Vacuum-Pump-/231575894189?_trksid=p2141725.m3641.l6368
[18:29:32] <anniepoo> just seemed silly to not avoid the same pain
[20:07:25] <z64555> ugh, that ain't it
[20:33:57] <z64555> maybe I can guess the function from a set of data
[21:01:33] <z64555> no, not really
[21:31:16] * z64555 dohs
[21:31:28] <z64555> So, I just needed to scale it up by its theta=0 intercept
[21:31:38] <z64555> good to go
[21:32:30] <z64555> er, or not?
[21:59:57] <mun> with backlash, how could they be eliminated? do u joints give backlash typically?
[22:04:14] <z64555> If there's a different in the bearing's ID and the axle's OD, then there will be backlash
[22:04:18] <z64555> *difference
[22:04:56] <z64555> There will also be backlash if the bearing has some play in them, like with a roller or ball bearing
[22:05:23] <mun> hmm
[22:05:41] <z64555> The backlash is in degrees, and thus is much larger if the U-join's diameter is larger
[22:06:26] <mun> i see
[22:06:46] <z64555> hmm, no?
[22:06:48] * z64555 thinks
[22:07:15] <z64555> No, it actually would be _less_ delta theta for a larger diameter
[22:08:17] <z64555> dx / r, where dx is the backlash delta and r is the U-joint radius
[22:10:50] <z64555> mun: make sense?
[22:23:06] <mun> hmm
[22:23:26] <mun> i guess it does
[22:23:55] <mun> so the smaller the difference between OD and ID, the lesser the backlash
[22:23:57] <mun> for all joints?
[22:24:12] <mun> all types
[22:24:51] <z64555> when concering bearings/bushings
[22:24:55] <z64555> *concerning
[22:27:05] <mun> yeah
[22:36:50] <pokmo> with this mechanism, is it 3DOF or 4DOF?
http://i.imgur.com/ua89MkW.png
[22:42:44] <z64555> 3
[22:43:16] <z64555> 4, if the link in the middle can also extend
[22:43:23] <z64555> so you get 3 rotations and 1 translation
[22:49:35] <pokmo> z64555, hmm but would something like this produce the same 3DOF, assuming the centre point is fixed?
http://i.imgur.com/pTcdSbi.png
[22:49:40] <pokmo> i can't seem to get it to yaw
[22:49:56] <pokmo> note the servo arms
[22:58:12] <z64555> do the arms extend like the other, or are they fixed size?
[22:58:42] <z64555> I'm guessing they're fixed size
[22:59:11] <z64555> That's actually a 4DOF platform, but if the center was linked like the other one, then it would also be a 3DOF
[22:59:27] <z64555> buuutt... it probably wouldn't work
[22:59:37] <z64555> if it was linked
[23:00:46] <zhanx> ok just spent 45 dollars on garden mesh network
[23:00:52] <zhanx> usd
[23:56:32] <z64555> netting?
.png/220px-Clevis_(PSF).png){kind=link}
