Research

Recently, I started working on a second quadrupedal rebot build. The first one turned out to be heaps of fun to design and program but ultimately failed to walk which was the goal I set out to achieve.

I still learned a lot from the build and have now a good idea of what I’m looking for when building the second quadruped. A major issue witht the first one was leg assembly which would bend in different ways and which was too weak to lift the quadruped when in trod gate where two legs have to carry the entire robot. A contributing factor was also the weight of the two batteries and all the electronics that I didn’t strictly needed.

Leg assembly

With that in mind, the new version should have more rigid legs, with stronger motors and better mounting of the motors and joints. Additionally, I would like to add a custom PCB containing all the electronics I need - just for kicks.

This time around it would also be nice to do a little bit more research into the various aspects of the assembly which is exactly what I will be focusing on in the couple of next updates.

First up is the lower leg joint. The nomenclature I’ll be using throughout the articles is the following: From the top to the bottom we have the body, shoulder, upper and lower leg. This 4 parts are connected by 3 joints - again from the top to the bottom - shoulder joint, upper leg joint and lower leg joint.

In the previous build I used some form of lever and a spoke which was also not braced. This would lead to bending of the joint under load. The construction also meant that the lower leg could only rotate some 120 degrees which is pleanty for normal walking but it would be nice to have some more travel inorder to recover the robot from a fallen over position. Over the month I have seen a couple of alternatives like the following: