So today we met with Quentin and Clémentine (and without Loïk) and thought the pendulum idea over. It clearly came out as too uncertain, complex and risky in terms of mechanics, especially since we want a horizontal center structure that’s not rotating. So we’re about to forget that.
Now what ? We came back to the idea of inner wheels forcing the sphere to rotate, and thought of one of Loïk early ideas : holonomic wheels. These are special wheels with little cylinders around them that allow them to slide sideways when a force is applied perpendicular to the wheel. And if we apply a force that’s partly parallel, partly perpendicular, then it’s partly the actually wheel that turns, and partly the little cylinders (the force is projected along the two axis).
On this drawing the wheels and their structure are on the floor – in our case, the central part with the wheels is put on the sphere, and it just needs to be weighed so that it doesn’t go crazy while the sphere remains immobile – and the sphere must be light. The good thing here is that we can solve the mechanical faults (or part of them) by changing the wheel commands, without having to rebuild the structure. So, hopefully we can set that up ! Stay tuned for more – tomorrow !
And as far as STM32 is concerned, I haven’t quite figured out the right OpenOCD config to write correctly in the memory, I’m investigating known boards, targets and chips configs in /usr/share/openocd and seeing what is done there.