I had another closer look in fullcreen to this video of Steorn:
http://www.youtube.com/watch?v=9JikYfmEdF8They have used really Toroidal-Ferrite coils as Motor-Drive coils.
So as there are magnets inside the rotor,
the rotor is accelerated without any current input into the
direction of the ferrite toroid coils.
Then when the magnets arrive exactly over the toroidal ferrite core coils,
these coils are pulsed by a Reed Switch from the battery,
see:
http://www.steorn.com/demo/rig/Part Number 23
and then repelled, so the rotor can rotate on.
One interesting fact here is, that these toroidal ferrite core coils
normally have their magnet fields completely inside the ferrite core
and only leakage flux is coming out of them..
so the rotor will probably have not much torque,
cause this will not repell strongly the rotor magnets.
Also the mechanical Reed switching of the input current will
probably wear down the Reed Switch pretty fast, as the BACKEMF
of the toroidal coils
also goes through the Reed Switch and will
spark a bit inside the Reed Switch.
Also the rotation RPM is thus not very fast, because the mechanical Reed Switch
could probably switch not faster than about 30 to 50 Hz or so...
So the Motor part is not very powerful in torque and not very durable
cause the Reed switch will wear out pretty soon.
On this Roth Elektronik RE901 PCB board there seems to be
soldered also only a SMD diode,
cause only 2 connections are used.
It seems to be only the blocking diode,so that
the current from the battery will not flow through the
pickup coils, but this diode will only conduct during the
induction spike of the pickup coils recharging the Ansman 10.000 mAh
NiMh accumulator battery.
So all in all it is a pretty easy to build pulse motor
and should be easy to replicate and measure the
input and output currents.
The only effect I could see is, that they use the toroidal ferrite core
coils for the driver part, which could behave stangely,
cause the magnet will also induce current into them,
when they go by and this could reduce the input current...
But how much could only be measured via a scope on a shunt
resistor.
Gruss, Stefan.