1:4 ratio multi-rotor all magnet motor (catchy name!)

[Ten penny]

I was hesitant to post this because I didn't want to seem like the newcomer who thinks he's a genius.  But it is haunting me...I've got to know.

I just finalized this design last night after many variations.  I'm hoping the many sets of experienced eyes on this forum will help me see any flaws before I waste my money and time building the prototype.  I looked at a lot of previously attempted designs and I didn't see this one.

I was looking at multi-rotor all magnet motors, and they all seem to have two problems: sticky spots and timing issues.  I thought to myself, "What if instead of directly gearing the two shafts to rotate at the exact same speed, you use gears or belts to increase the speed of one of the shafts to quickly get it out of the way when there would be a sticky spot and quickly bring it back into phase for repulsion or attraction?"  That's quite a mouthful, and it's not even the whole story.  I ended up deciding that one shaft should have only two poles interacting with the other four-pole shaft, and that this two-pole shaft must rotate four times faster than the four-pole shaft.  The two-pole shaft must have one North and one South Pole exposed, while the four-pole shaft must have all poles alike.  IN EFFECT, this seems to be the exact same principle that a conventional electric motor functions on - something reverses polarity in a timed way to attract and then repel.  Rather than reversing the charge on a coil, this device would mechanically reverse the charge by flipping a magnet or set of magnets around to face the other way.  I'm guessing this one would be a self-starter with no way to stop.

My first design had only one small shaft with two poles to interact with the four-pole shaft.  The one depicted on the attachment here has two such two-pole rotors, and it seems better this way in my mind.    My theory is that the two 2-pole rotors will limit the rpm potential of the larger shaft, but provide the ability to bear more load even with lower rpm.  I'm interested to try both versions.  It might be that the phenomenon I am describing is why the Whipmag motor seems to increase in speed when two of the three small shafts are stopped- perhaps there is a cost in power output that has yet to be measured.

Theories about this design:
     Con- the two-pole rotor's increased speed reduces the time spent interacting with the four-pole rotor, reducing the amount of pushing or pulling effects that it would have if it were rotating slower.
     Pro- the increased speed and polarity reversing of the two-pole shaft eliminate persnickety timing issues and all sticky spots.  This design also allows greater tolerance for magnet placement, air gap, differences in magnet strength, and synchronization- seemingly very forgiving if all of that is true.  It very well might be that most other designs proposed up until now would have worked in theory, but we can only be so precise in our replicating them in real life.

Well, give it to me straight, docs.

[Floor]

@TenPenny

Go ahead and build your design !
Don't assume that you must use very powerful and expensive magnets.
You can use inexpensive ceramic magnets. and use them again in other
experiments.

I have seen designs similar to your drawing, but still, you may discover
some thing new and at least you can learn and also contribute.

             All of us are still learning

                        welcome

                           floor

[Ten penny]

If you can share pictures or links to anything similar, I'd appreciate it.  I enjoy benefitting from the wealth of previous experiments.

Here is my simplified 1:2 ratio motor.  Same principle as the first one.  I originally designed this one to only have two exposed poles on each rotor, but I added some on the left rotor and it looks better that way. 

I'm not sure whether this design will be better than the 1:4 proposed above.  I'm going to try and make a rig that can accommodate a lot of changes to design. 

Any good sources for tiny belt drives?  I'm currently trying to use sdp-si.com, but their website is a mess.