Rough thinking on servos

[jerome]

I had a wild idea about how servos could be made more modular and generic. The problems intended to be solved:
- Not all joints have the same speed, couple, and size requirement. E.g a hand joint needs to be small, and requires little couple. A knee joint must have lots of couple. A head joint should be fast.
- Cost: having a single type of motor/servo reduces costs.

So the idea: being able to assemble motors in series or parallel, just like resistors, etc. Two motors in series would have twice the speed, same couple. Two motors in parallel would have twice the couple, same speed.

Every motor would be in a small package, with a generic demultiplication. They would be mounted in series/parallel to achieve the target speed and couple requirements. A single servo bloc would control the assembly.

Here is what I mean by series/parallel mounting of the motors. This does not relate to the electrical cabling at all.

http://picasaweb.google.com/jerome.jh/Motors/photo#5146022451465864962

Please note that in series configuration, some motors body rotate, which requires to transmit electical power through the servo horns.

The pluses/minuses I can see right now:
+ fulfills design goals
+ less power dissipated in gear trains

- "package overhead": actuator is finally bigger than a specially designed motor
- power transmission reqires brushes

What do you think of this? Is it useful or pointless?

Best regards,

Jerome.

[MYKL]

Sir, your link is dead.

Is this like a linear motor?