dutchkiwifruit

That would be fine too but I dislike the idea for aesthetic reasons.

Bobrow’s research mentions a set of equations that can be used to derive the controller gains based on a set of dynamics-related parameters of the cube (like inertia, mass), but the results were unsatisfactory, so I tuned the gains by hand. But in theory, yes.

Thank you!

Fusion 360 calculates things like moments of inertia and weight, and this was very helpful since the controller requires these parameters to convert control inputs to torques, for example. I still had to tune the controller gains, of course. (Which I did by hand.)

Most of the dynamics and kinematics involved are thoroughly discussed in the research work that has been published by the authors of the original Cubli! For example: https://ethz.ch/content/dam/ethz/special-interest/mavt/dynam...

Agreed. If only to enable others to reproduce and improve on the work.

I've considered various concepts for the brake design and think the way forward is to use friction brakes. The original Cubli uses bicycle brake pads, but this has its disadvantages. I'm thinking about using a (possibly bidrectional) band brake for my cube, similar to the one used in the "M-Blocks" [1]. Based on this video and the associated research paper, I'm quite optimistic about this kind of brake.

[1] https://www.youtube.com/watch?v=hI5UDKaWJOo.

My envisioned solution to this - very real! - danger is to attach some very thin round sheets of steel to the wheels such that you can’t stick your fingers in them.

That would be cool. I thought about balancing one cube on top of another balancing cube, but sadly I’ve only got one.

The house always wins!

Thank you! The weight of each wheel is approximately 220 g, whereas the total weight of the cube including wheels is 1700 g.

I’m not sure! You could probably at least double the rib length if the wheels are scaled accordingly. I can however answer the opposite “how small” question to some extent by linking to this really cool project: https://youtu.be/hI5UDKaWJOo

No worries: https://youtu.be/woCdjbsjbPg

Yes, the reaction wheels normally rotate very slowly.

I spent quite some time (almost) perfectly aligning my reaction wheels. Vibrations at multiple thousands of RPMs are no joke. Then again, such speeds are typically unnecessary once the cube has settled into its equilibrium position. ReM-RC on YouTube has built a similar cube which uses 3D-printed wheels with steel bolts in the circumference [1].

My wheels are oversized when you just consider the balancing application, but I want to add jump-up functionality later, and the wheels need significant inertia (and also need to run at significant speeds) to accomplish that.

[1] https://youtu.be/AJQZFHJzwt4

This is definitely something I’m thinking about doing! I would ideally like to keep the reaction wheels visible though. Putting LEDs inside the wheels for a persistence of vision effect might be cool.

I didn’t know about this, but it’s great! It makes me wonder whether I should come up with some kind of funny enclosure for my cube now, haha.

Your comment makes me smile. Thank you! :)

That’s an interesting idea. I have no good answer to your question, but something like a dodecahedron comes to mind, as a ‘trade-off’ between a cube and a sphere.

Some things certainly can be modelled, but for others it is easier to simply try. For example, will applying a positive current to the motor make it spin in the clockwise or counterclockwise direction? It really depends on the behaviour and configuration of the motor controller, and in this case it was easier for me to just try.

The trick is to do these tests at a sufficiently low level, because that’s usually where these issues are, in my experience.