Check out this picture, it is hard to describe. The magnets on the top and bottom support the shaft and flywheel, the two magnets on the sides above and below the vertical support magnets, support the shaft horizontally. I'm still trying to design the control hardware, but for the most part the drawing is pretty complete. The control circuit will read the speed from the marked RPM sensor and turn the power on long enough to keep the RPMs up. Anyone who wants all of the drawings of this idea, PM me, and I will send them to you. They are the different sections of the generator with all measurements.

I've been working on this idea for a long time. I haven't built a prototype unit yet, but I do want to. I started drawing this idea waaay back in the late 90s when I was in Community College. It has gone through a lot of changes since the original drawing. The drawing here is pretty much the final version of the drawing that I drew using the CAD program called DeltaCad.

My generator is totally electric. I designed a simple relay switching circuit that would change the operation from external power to internal power. What would be good to measure the RPM of the flywheel so I can create the control circuit that would switch the power on and off to keep the flywheel at a steady speed? That is my holdup. I still have to design my control circuit so I can control the speed w/o being dependent on a steady flow of power to keep the flywheel at speed. I was also thinking of sealing the case and reducing the air pressure to a near vacuum. That way the air resistance would be down to a low level and that would allow the flywheel to remain at speed longer. I have changed my profile so if you want to IM me using Yahoo, 'cause I am on there most of the time.

Mh-mhh ... the main novelty there is you are trying to make it self run.
When to speak about motor-generator, then losses by mechanical friction are the smallest problem compared with motor losses and generator losses.
At that to support magnetically this visually huge iron flywheel ... all this together with magnetic bearings, when you are not respective educated,  will spend energy and not so small amount. Your huge iron flywheel and miniature lower magnetic support looks funny feeling.
All contemporary flywheel energy storages use composite flywheel instead of iron you use, this is deadly danger when speaking about high RPM used in flywheel energy storages - http://www.aspes.ch/faq.html - mainly only when high RPM then come up needs for vacuum and magnetic bearings. 
Compare your small diameter of motor and generator and large diameter of flywheel  ... have you built any motor, any generator, have you any idea about RPM/V, about iron losses, copper losses, Eddy losses ...  http://www.free-energy-info.co.uk/P25.pdf , originally this pdf file was not meant for free energy sites.
All this rotating unit you did draw need to be two plane precision balanced  -  first at all you need to know what it means. Yeah, it looks the same like tire balancing but it is much(!) more complicated.
And about sensors ... when you have motor and you have generator then why you need more sensors at that on so large diameter?
Anyway I hope success for you,
cheers,
khabe

I was thinking about using a composite flywheel. The one sensor is just to detect how fast the flywheel is moving and pulse the power to keep it up at its operational speed. I do understand about the needs of balancing the system to keep mechanical friction to a minimum and to keep the flywheel from tearing itself to pieces explosively.

Even you need separate sensor "to detect how fast the flywheel is moving" you can find the place for this sensort somewhere on the shaft,
When PM brushless motor principle, then RPM could be measured direct from motor wire (or gen.) considering number of magnet poles, of course.
You did not explain how this small magnetic support on the bottom holds "in air" comparably huge flywheel.
What are appr. sizes and what will be RPM range?
cheers,
khabe

You have no idea what I have seen or not seen, my young friend,
Visually your support magnet as 25 times less diameter than flywheel.
For example when magnet diameter is 10mm (1 cm) then flywheel dimensions are D-240mm x h-40mm (24x4cm)
appr. 1800 ccm what means that flywheel weights appr. 14 kg, shaft looks like ... weights one more kg, total 15kg
Your 10mm diameter x 10mm cylindrical magnets will keep this weight up easily 
I even will not try with similar tiny magnets - surely too small for - but with pair of d-13.5mm x h-100mm it works    but ... you need to think about upper magnet pair ...   contra-force from upper support pair will press your flywheel down    For real unit these support magnets must to be much more bigger.

cheers,
khabe

PS:
Perhaps question is about disproportion (vs. in proportion) of your draw - I have no idea,
And you do not need to explain me about "... rare earth ... Neodymium ..."   - I´m familiar with from the ground up.