Thanks guys, Gyula the rotor drawing is attached. I think i made the projected part of the field more kinda half circle shaped it seems more  out and also wider cogging is reduced but the magnet attraction to the core is more, when I place the generator core it will help to smooth the cogging as well. I thought if I made the field more projected alone like with stacking it would make more noise due to cogging. But it make lots less and goes faster and speeds up quicker. I can speed it up pretty fast now. I think the field is projected more but is wider and more rounded (effectively) the extra magnets don't seem to attract to the core themselves directly. 

Milehigh I think I get what you're saying about the shielding and I think you're right, it might be more trouble than its worth right now. I can't change the way the driving magnets face but I can add some 10 mm cylinder magnets normally magnetized on the rotor maybe. I'll put that idea away for now. Hopefully my wire for new cores will be ready soon.

I did notice that when the frequency duty of my setup gets high the motor coil "current" does not go to zero even though the mosfet is on for 2.4 mS this is an odd discovery which kinda means the motor coil works in continuous mode rather than discontinuous mode as referred to in boost converters. Still the mosfets could run all day switching an amp each and the inductive energy under control. My boost converter is putting some noise in the circuit but not enough to affect the switching. The current through the charging coil looks very good, it rises sharply and drops at the same rate, the motor coil current rises less sharply and drops off quicker but then at almost no current it continues for a bit then drops off again. For now the input voltage is limited because of the 40 volt rated diode in the boost converter.

Cheers

EDIT: I meant to say in the last paragraph that (the motor coil "current" does not go to zero even though the mosfet is only "on" for 2.4 mS), so I changed it, sorry I was tired. 

Oh and also in the last schematic I posted the capacitor (C3) can be connected to the circuit ground with similar or better effect, when it discharges it can only discharge to the voltage level of the supply anyway so it doesn't go below the supply voltage when connected to the circuit ground, only at start up it is less,.

Try connecting a 100K - 500K ohm or higher resistor between the gate of Q1 and the negative battery supply rail. That may help to determine if the mosfet is switching fully on and off. If the gate voltage of Q1 is floating, that may be the culprit, and a resistor will remedy the problem. (if it is a problem ).

Also, the rotor magnet/s is/are counter inducing a current in one direction through the motor MC1, via D2 D5 L1 and D4. In the circuit you've shown on the previous page, you can pull Q1 out of the circuit, and spin the rotor up to speed by another means, and you will get current through MC1 via the path I just outlined.

I notice also in the same circuit that any collapsing emf from MC1, during off time (from the supply) discharges through the same path.

http://www.overunity.com/11350/confirming-the-delayed-lenz-effect/dlattach/attach/123297/image//

Cheers

Don't worry the mosfet is switching off ok I have a 10 k resistor from gate to circuit ground. I can show the gate signal wave form if you like. If they were not turning off they would be getting warm but they are not. They switch dead clean, they are driven by a TC4420 driver chip. 

The continuous current is because of the coil not fully discharging I think, I need to speed it up a bit more.

Wouldn't it take a south pole magnet on the rotor to induce a current the way you describe ?

The inductive collapse is discharged first into C3 then C3 discharges through L1 back to MC2 (charging coil) in my motor, between the diode D1 and MC2.
I can see the inductive spike go up to about 25 volts for 400 us or so in C3 when the capacitor C2 is at 12 volts.

The shot below shows it with no battery, the spike is only 25 volts but it is in the capacitor C3 and then it discharges to the point at L1-MC2 in my motor, also that cap C3 in my motor is no longer in series with the battery, it's negative is connected to the circuit ground now. 

Cheers

P.S. OH and the current in the motor coil does stop if the motor is not running hard. as you can also see by the attachment. I could run the motor all day with 2 amps input and no heat sinks on the two switching mosfets, they just get a bit warm, which is also evident by the scope shots I think.

The resonant frequency of C3-L1 is 10 kHz it needs to be (higher frequency) and I think I'll use both a bit less capacitance and a bit less inductance, then the coil should discharge it's energy
quicker, hopefully anyway, if not I'll do the opposite. 

EDIT:, actually I just checked and there is a weak south pole between the two north poles, the south poles are very weak though they don't attract anything much but a compass.
Still it might be possible the souths could induce a current that way. The north poles would try to charge the supply I think.

EDIT: 2, ( I was a bit confused the inductive spike charges C3 to about 25 volts when there is only 12 volts in capacitor C2, a fourth capacitor can be placed were L1-MC2 meet and as long as it isn't too big the spike remains. I tried 1 uF there). But it makes little difference because of C3 and L1. Sorry for the confusion, I modified the post to read right. It's a new circuit, I'm still making changes. I've got a much better sketch to explain things, generator coil as well. 

Also below is a shot showing the voltages at the capacitors C2 and C3, and the currents through the coils when running at a reasonable speed. In a way (after the first pulse) C2 is the supply for the motor coil, when running the switch turns on and C2 discharges through MC1, then the switch turns off as the supply charges C2 through MC2 and at the same time C3 is discharging the inductive energy just previously released to it from MC1, into MC2, then the capacitor C2 is charged with a higher than supply voltage ready to discharge through MC1 when the switch is turned on again, because C2 is at double the voltage of the supply and MC2 impedes the current flow, C2 fully discharges through MC1 before any significant current can flow through MC1 from the lower voltage supply feeding MC2. Only 12 volts is applied to MC2 whereas over 20 volts is applied to MC1.

That's how it works in normal operation, and the difference in phase between MC1 and MC2 currents can be used as two driving phases of current. 



..

Just a suggestion for making coils:

Go to the electronics store and buy regular insulated wire spools of different gauges.  Get a nice big spool of very fine wire for sure.  We assume that the spools themselves are regular white plastic spools.  Sometimes the spools have a larger inner diameter so you could add a larger core if you wanted to.  Note the energy storing capacity of the core is proportional to the volume of the core material.  It's arguably more important than the number of turns if your focus is on energy storage in the coil.  I am not so sure that applies for a pulse motor because you assume that you are designing a for maximum rotor push with minimum energy lost per push.  You are probably better off with solid wire also to get the least resistance.  You could also say you are getting a lager cross-sectional area of metal with solid wire as compared to multi-strand wire.

Then for each plastic spool you make two radial cuts on one of the plastic ends about 45 degrees apart and remove the plastic material.  Now you have access to the inner layers of the coil.  You carefully tap into the coil at two or three depths and make connection points.  For example, you could use a small terminal block, the type where you insert the wire into a hole and tighten a screw would be good.  Then you just take hot glue and solidify it all up.  A hot glue gun is the poor man's 3D printer.

Of course you could buy spools of speaker wire and make true bifilar or series bifilar coil configurations also.

Say you buy four largish spools of wire in four different gauges.  And you cut the sides off and you add three taps to each coil.  So that means each coil is four connections, a reference "ground" connection and the three taps.  That would give you effectively 12 coils to experiment with, and you could move your core material from coil to coil.

MileHigh

Farmhand:

Keep doing your thing at your pace.  My posts are intended more to be stream-of-consciousness for the thread itself.

Cheers,

MileHigh

A little bit more stream of consciousness:

You have your pulse motor running at it's steady-speed.  I am talking an ordinary Erzatsbedini type pulse motor with the basic circuit.

Here is a simple thought experiment

You run the motor and you scope the coil current and you measure the pulse width and the final current level.  Let's assume you see a typical exponential type rise in the current during the pulse.

Then you remove the rotor and you just energize the coil with the same pulse width.  For example, you connect a signal generator to the transistor base input.

Will there be a difference in the final current level in the coil when it is driving the rotor vs. when the coil is not driving the rotor?

I say there will be a difference, and some might be surprised when I say that the final current level in the coil will be lower when the coil is driving the rotor.  That would seem counter-intuitive.

Here is the explanation:  When the coil is driving the rotor, you know that some of the electrical power you pump into the coil is being exported into the outside world, it is not in fact charging up the coil.

What follows is this deduction:  The power into the coil is just the instantaneous volts times amps.  So you can say that when you drive the rotor, you are taking a "slice" of the voltage (times current) away from the coil charging function and that "slice" is being exported to the outside world.

In other words, if there are 12 volts across the coil, at a given point in time perhaps four of those volts will be "eaten" by the rotor pushing function and the remaining eight volts will be energizing the coil.

So over the duration of the coil energizing pulse, this "voltage stealing" function means that there is less voltage available to energize the coil, therefore at the end of the pulse there is less current going though the coil.

So, how about all that simplified:  This could explain the current draw going down as the motor speeds up.  It all ties into the notion of the motor being seen as a pure abstract "impedance" by the battery and that impedance is actually an electro-mechanical impedance.

MileHigh

Here's a Video Clip of the magnet proximity (coils driving the rotor more) causing the coil current to reduce, the further away the magnet is when he coil is energized and therefore
the coils are driving the rotor less the "more" the current in the coils. The closer the magnet is to the coil core when the coil is energized the less current goes through the coils.

I think this is what you mean Milehigh ? In a way.

Pulse Timing and Coil Currents
http://www.youtube.com/watch?v=0whkutQ7mNQ

..

Synchro1:

That's a great question.  The thing to keep in mind that in many cases the properties of a circuit are frequency dependent.  For example an ordinary capacitor will change how it responds in frequency.  A low and high frequencies a capacitor looks like a capacitor and passes the signal.  But at very high frequencies the capacitor will act like an inductor and block the very high frequency signal.

So you can imagine a scenario like the following:  If you run a typical pulse motor at low to medium frequency, you may see decreasing input power consumption as the speed increases because of the "voltage stealing" effect from power being exported to the rotor.  (That's what they mistakenly call "the witch" on you know which threads on EF.)  But as you run at higher frequencies the air friction starts to become very significant and then the power consumption will start to go up.  I'm not just sticking these two ideas together for convenience, this is a very real property that you see in many places.  Another example is a typical transformer.  At low to medium input frequencies the transformer will work properly.  But at very high input frequencies, the input capacitance of the transformer primary will effectively short out the input signal and the transformer output is near zero.  The minuscule input capacitance will prevent any current from flowing into the primary windings so the transformer can't work, only if the input signal is at a very very high frequency that you normally wouldn't use anyways.

Tech link about transformers and frequency responses:
http://www.vias.org/eltransformers/lee_electronic_transformers_06_09.html

Again, these are basic principles that would have to be investigated on a case-by-cases basis. 

MileHigh