An other 4DOF leg design

Lynxmotion Legged Robots
1

Hi everyone,

I’ve been wondering why most 4DOF leg design around here have 3DOF in the same plane (T-Hex like), and if other 4DOF designs could be feasible.
So with a friend, we jumped on the occasion to explore a less common 4DOF leg and build our first quadrapod.

We settled on doing this progressively :
First, a prototype of one leg with small and inexpensive servos, then the first version of the body + 4 legs, but without any load (external power source , external brain …), and then the “real”
version, bigger, with better servos, and everything inside the body. A last step would be to add foot pressure sensors, IMU, camera, you name it… to add some AI.
We may or may not go all the way with this project, depending on the success of previous milestones, time and available disposable income, but at least it will be fun to go as far as possible.
Each of us has already built a few robots, but both of us are quite new to robots with limbs.

Here is the schema of the leg we want to use :
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/LegIdea.jpg

The axis of the 3 first joints (starting from the body) all cross in one point, making the system nearly equivalent to a spherical joint between the body and the femur, and a hinge joint between the femur and tibia.
We looked around, and did not find projects using this particular design, but if you know of some, I would be delighted to ear about it.

To build this, one of the cool things is that we can make a module with two servos with orthogonal axis and use two of these modules to make a leg.
My first shot at making this module was to use rapid prototyping techniques to build a bracket like this (yellow rods are M2 screws, nuts are not represented):
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/BracketTest01.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/BracketTest02.jpg

But for the first prototype, it was a little too expensive. I used some aluminum sheet to make the part that links together the servos, and for the axis I just made some holes in the servos and fixed M2 screws on the side and back of the bottom servo.
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000237_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000239_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000243_1365x1024.jpg

And here is the first proto of the leg (with an improvised tibia just for the looks…):
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000244_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000245_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000246_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000247_1365x1024.jpg
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/P1000250_1365x1024.jpg

I am currently working on the control of this beast, I think I will try to use a model with a spring attaching the knee to a virtual point to get an additional equation that will allow to solve the IK problem. It’s still a little fuzzy in my head, but I’m working on it ^^

If you have any though about this leg design, the problems it could lead to, a good way to approach the control problem, I’m open to suggestions ^^

Thanks !

2

Interesting, I think it might be similar to mine:

http://www.marconettengineering.com/4DOF_Leg_B.jpg
4DOF Leg Design

lynxmotion.net/viewtopic.php?f=20&t=5444&hilit=4dof

Although I think the two middle joints are reversed. I’ll have to study your sketch more, and see if it can be build with SES components.

Alan KM6VV

3

It has some resemblance, the degrees of freedom are nearly the same. The difference, beside the order of the middle joints, is that in my design these middle joints have crossing axis.
If I remember correctly, this has an impact on the kinematics of the leg and on the complexity of the computations that come with it.
http://img.photobucket.com/albums/v442/Xevel/4DOF%20leg/Equivalent_kinematic_models.jpg
H : hip
K : knee
F : foot
In your’s, the second and third servos’ axis don’t have a common point, so you have to take into account H1-H2, aka the distance between the axis. I don’t know how my version will cope with load, etc… , but the IK resolution seemed easier to tackle with one less (fixed) parameter to handle.

I think with SES brackets, this kinematic chain could be built, but not he way I made it with my prototype. It would have to be much more voluminous if made only with SES parts since the “two servo module” can not be built as compact as mine, afaik. However I could instead use SES for the main parts of the leg, and build my own brackets to fill the gaps (aka make the servo module) while keeping the same servo layout, I guess.

4

Hi,

Since this project will (sadly) probably never be lyxmotion-related, I migrated it here : forums.trossenrobotics.com/showthread.php?t=4341

5

Cool design, I also love the profile picture. :slight_smile:

6

Thanks !
Some updates are on the way, I will post that by the end of the week-end I think :slight_smile: