Rotopods and other kinematics ...

Hi all,

... i found a video of a very interesting planar-tripod/rotopod: [155.69.254.10]

From here: [155.69.254.10]


*** start EDIT

... and here an even older and very fast demonstration: [www.ifr.ing.tu-bs.de]

From here: [www.ifr.ing.tu-bs.de]

*** end


This seems to be very easy to assemble and could solve the problem of reprapping parts for hierarchical bigger systems ...

Viktor

Edited 1 time(s). Last edit at 01/14/2008 04:31PM by Viktor Dirks.

The one problem that I foresee is the complications and inaccuracies from a non-Cartesian robot.

Never mind that, after thinking about it, all you have to do in order to keep track of where the stage is in space is triangulate it from the angle off of the motors. This is quite intriguing actually. Great find!! I might get my dad to attempt to write some simple software for a machine like that. it seems to use 1 extra motor over the Cartesian setup (3 in the XY plane and on for Z), but costs saved from excess nuts and bolts could make this feasible, at least on the small scale.

---

Jay

... i'll post some more images and infos, when i have finished my 'tripod-toy':


look here [builders.reprap.org] under "Assembling the Tripod..." for more images and a short video of manually playing with the kinematics ...

Actually i'm at optimizing my 3D-CNC-repstrap (in the BuilderBlog-link above under "Got the perfect RepStrap" too), so i can mill the needed parts for the tripod and some equipment for applying the diode-laser- and dispenser-heads ...

Viktor

After thinking about it, i think that there is a three motor setup for this type of thing. Imagine a two jointed arm, with a motor in each joint and the stage moves under a stationary extruder which then moves up and down for the layers. The mechanics of getting a straight line would be complicated, but it could defiantly work.

---

Jay

Hi Jay,

... here: [www.harmonicdrive.de] you have a parallel-scara-robot in a similar setup with two motors for XY and a third motor for Z ...

Orthogonal moving with angular driving isn't so complicated - you have clean conversions, what's called "Inverse Kinematics", where for every point in the working area you have a corresponding position of the motors.

So you have only to brake down your line into elements small enough, so you can draw a straight line with many small arcs too.

With a linear interpolation between the two (or three axes for a 3D-line) you can move as exact as you wish - in this special setup you have an overall accuracy of some microns!

Viktor