Sharing ideas on how to make a more efficent motor using Flyback (MODERATED)

[gotoluc]

It's unfortunate that the scope is still playing a central role here.  It was refreshing to see that you see how well this orthogonal magnet almost completely neutralizes cogging.  You correctly note the relation between the size of the stator and rotor magnet cross section, where the latter should be larger than the prior.  You experienced the almost pulse like nature of the magnet and stator pole piece interaction, extremely important.  There are giants in this area, you would be wise to identify them, one in particular.  That particular individual has mastered this concept, and perfected it to a very high degree.  When looking at his device, at first glance you don't even see the orthogonal relationship!  This principle is not limited to spinning the magnet 90°!  It took me what feels like ages to understand that. 


If you can, reduce your rotor diameter.  Find a better means for locking those magnets in place.  Unfortunately, your desire to make this a strong motor will not pan out, especially when you use solenoids like the one you are demonstrating.  Decrease your on time so that you are only pulsing during the peak of the wave, allowing current to flow only long enough for you to assist the pulse that you get for free.  It should be clear that the pulse is at TDC, should also be clear that the voltage node is peaked at TDC, this is what makes this topology interesting, as under normal circumstances, TDC is the zero crossing.   This concept lends itself for building an electric turbine, a properly built machine will be extremely fast, but will produce very little torque.  It can and take the skin off your hand if the rotor has enough mass, and is spinning fast enough, however, owing to the very short duty, consumption drops when the shaft is loaded.  The device performs best when duty is short.  A torque machine based on concepts learned from spinning the magnets 90° uses "conventional" alternating magnets, these magnets are configured so as to induce orthogonal fields in a properly configured stator.


I spent years investigating this.  Don't make the mistake most make and just throw something together, and broadcast it like its the end all.  That will motivate many to look, and "replicate", but they will not think!  They need to think about whats going on.  They need to try and connect this concept with other concepts which have been demonstrated over the years, find the common links.  Its not enough to throw something together for the sake of just doing it.  Your desire to help others is noble, but you of all folk should know that its best to show a man how to fish so that he can feed himself.  You aren't showing anyone how to fish by just throwing them the bone of turning the magnet sideways...

Regards

Thanks erfinder for taking the time to offer the above.

I'm sure I'll do honor to your Jewel

Luc

[MagnaProp]

In case some of you may of missed it...

I did miss it so thanks for pointing it out. Brilliant idea using the magnet at 90 degrees. Little gems of info like that really help us not spend time reinventing the wheel over and over and help move progress forward in the right direction. Video up voted and credit will be given where credit is due should I use that idea of yours. 

[verpies]

Please give us an update once you have come to a conclusions

I don't have hard conclusions yet because these probes are not connected how I'd like them to be and I don't have any scopeshots of your system when the rotor is not moving.
...but I strongly suspect that the conclusion is Case #3

P.S.
I am also confused how the blue trace can be non-zero when doing a scopeshot with a moving rotor and this circuit.

[gotoluc]

I don't have hard conclusions yet because these probes are not connected how I'd like them to be and I don't have any scopeshots of your system when the rotor is not moving.
...but I strongly suspect that the conclusion is Case #3

P.S.
I am also confused how the blue trace can be non-zero when doing a scopeshot with a moving rotor and this circuit.

Thanks for the update and please find the scopeshots of the coil with the 2 pole rotor not moving.
There is 1.57vdc across the 100 Ohm flyback tank load.

Let me know if you need anything else or have any other conclusions.

Thanks

Luc

[verpies]

OK, that shows me classical inverted exponential current waveform for an LR circuit. 
Was the magnet at the coil's core when this scopeshot was taken...or was it away from the core? - I am asking because of saturation concerns.

Also, the voltage should be constant while the transistors are ON.  Since it is not (it is sloping down), I must assume that R_{DS_ON} of these MOSFETS (or wire resistance) is allowing for this voltage drop.

I expected most ringing when the transistor shuts OFF and interupts the current, but that is not the case because D1 takes over, starts conducting and keeps the current flowing trough L1.
The CSR does not show this current because it is not in the L1, D1, C2 circuit. 
So the question remains: What starts ringing after a delay and what determines this delay. D1 stopping its conduction ? Drift Step recovery?

Anyway, when this current waveform (when the rotor is stationary) is superposed on the current waveform obtained in this circuit (while the rotor is moving), the result should yield the same current waveform as when the stator is powered externally (while the rotor is moving ...at the same speed).

If it does not, then you have an anomaly.

Are you open to moving the CSR and the scope probes according to the schematic below, to clear these questions?