Permanent magnet assisted motor coil designs

[captainpecan]

Very good response, thank you. A lot of info in there. I will read it over a few times when I get done working for the day and get home.
I have no idea why I said puffer coil, I just mistyped. I meant puffer capacitor. What I was asking is how it worked for sure. Let's say the peak voltage of the wave without the spike is 12v, once the cap fills to 12v, and I have adjusted the resistance for the cap to hold at that voltage, then wouldn't that 40v spike roll right past the Capacitor to the resistor or does it just seem that way? I just want to understand the physics behind it.
As for the other questions, it's entirely possible I have some setting wrong on the scope as well. But I have been leaning on the auto functions before the minor adjustments I do. As far as the gate resistance, that is a very good idea and easy for me to test.
As far as a mosfet driver, are you referring to a simple motor driver circuit instead of just a reed switch on the gate? I was thinking this would use the least amount of energy but of course maybe I'm off base there. I tried hall sensors but I simply have bad parts so until I get more I can't really even do that.
If I missed anything, I'll catch it later when I get time to read more closely and follow point by point. Thank you.


One more thing, is there a better way to get only that spoke collected than what I'm doing? I just found something that seemed to work, I'm open to more options for sure. Thank you so much.


Edit... also, I wired each coil separately in a terminal block so I can test parallel as well. Tests so far have shown series being more efficient, but the trade off is higher voltage of course. But i will test everything!

[gyulasun]

Let's say the peak voltage of the wave without the spike is 12v, once the cap fills to 12v, and I have adjusted the resistance for the cap to hold at that voltage, then wouldn't that 40v spike roll right past the Capacitor to the resistor or does it just seem that way? I just want to understand the physics behind it.

As I mentioned, it is a question of the (frequency dependent) impedances of the puffer cap and that of the series resistor + charge battery, these are connected in parallel so the 40V spike appears across both.  Suppose the frequency of the spikes is 40Hz and the puffer cap is say a 100 uF electrolytic, so it has roughly 40 Ohm capacitive impedance. Suppose you find the best resistor value to charge the battery is higher than 40 Ohm, then the 40V spike will cause higher current towards the cap and less current towards the resistor+battery.  Of course the charge current in the cap will be exponentially decreasing as the voltage increases across the cap, so the battery charge current will be increased accordingly by the current the remaining spike amplitude is able to provide via the series resistor which limits it too of course. This is the dynamics of the process. 
IF the cap is say 470 uF, its reactance would be around 8.5 Ohm at 40 Hz, hence more energy will be captured by the cap if the series resistor value is likely way higher than this (the charge battery has very low internal resistance of course).   
 
Yes the reed switch with the 10 kOhm gate-source resistor uses the least amount of energy for the operation. A 1 kOhm would use 10 times as much and a dedicated gate driver IC would use roughly in-between. A motor driver circuit sounds good if its output has max a few Ohm resistance (say < 2 - 3 Ohm). The input of this driver circuit could still receive the reed switch because the switching speed is always governed by the output speed of the driver circuit.  (rise and fall time of the driver)
The way you collect the flyback spikes is basically good I think.   
Regarding the parallel or series connection of the coils: the series connection reduces input power by the higher series coils impedance but as you find you have to increase supply voltage which works against input reduction. However the still lower currents involved versus the parallel connections give less overall losses, this explains the higher efficiency you find.

[captainpecan]

Very good clear response. That makes perfect sense. Thank you! I'll go over everything you said carefully and try some things.

[floodrod]

@ Seychelles
This is a coaxial transformer I put together a few years back. Since I tested with 60 hertz I was only able to apply around 3 vac before core started to saturate. I used 1/8" tubing with 23 gauge center conductor. Don't remember much results other than coupling between primary and secondary unloaded was near perfect.

Edit: A simple experiment would be wrap some TV coax cable on a core which is what I did at first.

Interesting!  I have access to truckloads of free coaxial daily. Literally..

[captainpecan]

Well, I made a change to the motor that did not go well at all. I had a huge gap between the rotors and coils because the rotors were not as straight as needed. I decided to back them with steel and straighten them up so I can decrease the gap. Big mistake. It barely runs at all now. Guess I will be changing back. I expected an increase in rpms and power. But clearly it changed the magnetic fields so much, nothing works right anymore. Really wierd. Maybe its because they are all like poles? Ill keep playing with it a bit to make sure I didnt just do something stupid i havent caught yet before i switch back. Lesson learned I guess.