Just another sFOC hub motor project

I found another usecase for an mid power range sFOC compatible controller using hoverboard motors.
An ironless multi-station fitness tower.
It started with the idea to use sFOC’s angle-loop with a hub motor for an indoor rowing machine.
But the same motor could work for Biceps/Triceps/Latissimus curling, benchpress and butterfly a.s.o.
It only takes strategically placed guiding rollers for the rope.

A few years ago I was going to a gym with this concept, I really liked it.

Not only it was resisting when you were pulling, but it was also pushing back when you were pushing. So it was much more effective.
You could set up the parameters on screen, and it was saved on your account.

I have a membership in a fitness studio too, but Covid-lockdown was a massive drawback.
Training at home with a DIY machine would be a good way to restart.
It won’t be as elaborated, but the eccentric force feedback mentioned in the video, I already had in mind for the rowing machine.

I also do the resistance exercise thing, and a module which provided commandable resistance to a cable would be very useful for many things. You could make many different machines, or just use it directly. I would tie it to an object and pull on the cable. You would need a separate module for each arm probably. Right now I use elastic bands to do this, it is a lot cheaper I have to say. But you can do a very wide range of exercises with it. Part of the value is that I can adjust resistance just be stepping slightly farther away. Thus I can switch between exercises.

Especially if you could use some wrist and ankle mounted accelerometers to monitor which exercises had been done, and which need to be done, and at which resistance level, that could be a powerful tool for training. Athletes would probably pay a ton for that.

But actually come to think of it, there are competing approaches that could be better. If you mounted emg electrodes that can tell you how much a muscle is contracting, the frequency of emg waves actually increases with relative contraction level. With an algorithm you could make a system that just uses elastics for resistance and got most of the value of such a system. A little motor based resistance module as described would probably be like $500 I would think, at least. Elastics are $30.

1 Like

I think for this use case you need encoders

What about hall sensors and smoothingSensor?
I could probably use an reflective optical sensor and engrave encoder lines on the hub motor. The side covers are black.

Hall sensors and smoothing won’t help at 0 speed unfortunately.
[EDIT] I mean the torque won’t be optimal.