#robotics Logs

Jul 18 2015

00:48 rue_shop3 well, with the loom I got, I get 10kg @ 50psi, with 9cm retraction for a 28cm length
00:49 rue_shop3 32% retraction
00:59 rue_shop3 maybe threaded rod
00:59 rue_shop3 or a high pitch ball screw
00:59 rue_shop3 hmmm
00:59 rue_shop3 the length difference would be trivial
01:00 rue_shop3 it wouldn't slip off
01:00 rue_shop3 and it would be free to rotate
01:00 rue_shop3 ah, there isn't 300lbs * leverage on the pin either
01:00 rue_shop3 hmmmm
01:08 rue_shop3 oh and I worked out how to do the star on the wrist union
01:09 rue_shop3 I'll put a hub in the middle that will give the bolts angular stabilization
16:30 Devtron robots
16:30 Devtron ardrino
16:30 Devtron where do i sign up
16:30 e4016 cringes a little at the arduino men
16:31 Devtron okok
16:31 Devtron then i need to learn to do it on the atmel chip it self
16:31 Devtron or use something more powerfull
16:31 ace4016 no, no; whatever you can use is fine :P
16:33 Devtron I have so much ardrino crap , including extra nema 17" stepper motors and such
16:33 Devtron just need a project
16:33 Devtron any big robot arm projects i could try out where i just have to print the åarts
16:34 Devtron parts
17:01 Hyratel1 so, a friend wants to make a centaur cosplay, and they're eyeballing the ride-on strandbeest shown here as inspiration - https://www.youtube.com/watch?v=mGOled86Sjo ; anyone have any useful insights to the idea?
17:21 rue_shop3 :/ I cant see youtube
19:14 wolfmanjm ok so i am drawing a blank here... can someone tell me how to calculate the mechanical advantage on this linkage? https://sketch.io/render/sketch55aae62026222.png
19:14 wolfmanjm compared to driving it direct fromthe servo
19:16 ace4016 Hyratel1, look for the Dutch artist's mechanical linkages?
19:17 SpeedEvil wolfmanjm: What is the movement distance
19:17 wolfmanjm say the servo arm can move 90°
19:18 wolfmanjm or even 180°
19:18 wolfmanjm I know the leg movement is decreased with an increase in torque... but damned if I can figure out how to calculate it :) even approximate it
19:19 SpeedEvil Degrees will confuse you
19:19 SpeedEvil you care about how many cm it moves.
19:19 wolfmanjm well I realize it is movin gin an arc, but an approximation would be nice
19:19 Hyratel1 wolfmanjm, pittman arm
19:19 SpeedEvil If the servo moves 1cm, and the end of the leg moves 2cm, then the force is half
19:21 wolfmanjm so if the servo moves 2cm and the leg moves 1cm then there is twice the torque?
19:21 Hyratel1 yes
19:21 wolfmanjm ok I can estimate the movement I think with trig
19:22 Hyratel1 but because of the weird angles, it's more complex to calculate. got a lot of trig going on
19:23 wolfmanjm ok so it looks like the servo arm is half the length of the leg, so maybe a 2:1 mechianical advantage?
19:24 wolfmanjm eh need to draw this out on paper :(
19:24 Hyratel1 you're taking your distance as perpendicular to the direction of motion of the pushrod
19:24 Hyratel1 wait.
19:24 Hyratel1 gah, this hurts my brain
19:24 wolfmanjm yea I need to look at what happens between two positions
19:24 wolfmanjm yea it hurts mine too :)
19:25 wolfmanjm I figured someone smart would say oh that is simply xyz
19:25 wolfmanjm just need an estimate or average mechanical advantage it obviously changes with angles
19:26 Hyratel1 it'd be easier if the levers were at the same angle when in neutral
19:26 wolfmanjm it may be easier to just prototype it and measure the torque :)
19:26 Hyratel1 neutral position
19:26 wolfmanjm yea the closest I found was the aerofoil mechanism on an R/C airoplane
19:26 wolfmanjm and there they have everythign perpendicular
19:27 wolfmanjm Hyratel1: ok for this excercise you could say the servo horn is parallel to the leg...
19:29 Hyratel1 ok if you set it so the pushrod is at 90deg to the horn and the output arm, then you'll have the same force-at-pushrod on both
19:29 wolfmanjm like this https://sketch.io/render/sketch55aaea1779883.png
19:30 Hyratel1 http://flockdraw.com/8mwxvd
19:31 Hyratel1 wolfmanjm, *everything* gets more complex when you put things at funky angles to start with
19:31 wolfmanjm ok butthat defeats the object :) I want some mechanical advantage at expense of range of motion
19:31 Hyratel1 no, you're missing the point i'm making
19:31 Hyratel1 your sketches all have things twisted anglewise
19:32 wolfmanjm well yea that is becuase that is how it would actually be.... http://stubby.digitalcave.ca/stubby/ <-- like that
19:32 Hyratel1 you start simple
19:32 Hyratel1 remove variables
19:33 Hyratel1 distil the problem to its most fundamental
19:33 Hyratel1 learn how to solve that
19:33 Hyratel1 then add complexity one stage at a time
19:34 wolfmanjm ok thanks :) I'll try
19:36 Hyratel1 once you get a sense of how forces change with angle, then start moving things around
19:39 wolfmanjm ok so I need th eleg to move half the distance the servo horn does to acheive 2:1 mechanical advantage...
19:40 Hyratel1 yes
19:41 wolfmanjm I can work backwards from there
19:57 wolfmanjm ok so I whiped up a model with my trusty erector set ;)
19:58 wolfmanjm seems that with the pullrod the same length as the leg, and the servo arm half the length then a 90° mpove on the servo is a 45° move on the leg... I presume with twice the torque :)
19:59 wolfmanjm so much easier to make a working model than messa round with math!
19:59 Jak_o_Shadows But maths is so much fun.
23:47 solifugus i am terrible with electronics... but does the idea of using a computer controlled relay array directly to control solenoid valves.. sound like a correct direction?
23:47 Jak_o_Shadows for what?
23:47 solifugus to control pneumatic air flows
23:47 solifugus to actuators
23:49 solifugus or.. an electromagnetic linear actuators pushing and pulling on the cable in a bicycle brake cable..
23:51 solifugus or maybe.. two strings in a tube attached to each side of the joint with an electric motor turning one direction or the other...
23:54 solifugus I don't know how bi-directional motors work.. so.. not sure how easy that would be to control from a computer
23:54 solifugus maybe two relays, one for each direction?
23:55 solifugus at least i know relays arrays are inexpensive and easy to program.
23:55 solifugus unlike stepper controllers