The new generator no effect counter B. EMF part 2 ( Selfrunning )

[MarkE]

MOSFET or transistor doesn't really matter, both will act as a switch to energize the drive coil.  I agree with you about the energy required to drive the base resistor to switch on the transistor possibly being an issue.  That is a design issue for any pulse motor based on a trigger coil and a base resistor - can you find a trigger coil configuration and a value of the base resistor that gives you the most efficient operation for the triggering system?   However, that's an issue that is above the level of a typical pulse motor builder so it never gets discussed.  See, that's a "rock and a hard place" moment right there.  You get criticized for not being a "builder."  If you are not a "builder" then you have "no right" to talk about something.  Then, if you talk about something that is above the level of a typical pulse motor builder, then it is "just words" and more "laws" stuff.  It's a no-win situation.

You can hear some effects of the Lenz drag from a generator coil and observe some changes in the pulse motor.  The most obvious one is that the pulse motor RPM will decrease.  However, there is a "problem" with typical pulse motors in that they have notoriously little average torque, and they also have very limited capacity to respond to an increased load situation, like adding a generator coil, and output more power to try to maintain the same rotor speed.

You say the rotor slows down and that means the drive coil is energized for a longer time.  That's true, but the ON percentage time per revolution stays about the same.  So the rotor slows down under load, but the power duty cycle that energizes the drive coil stays about the same.  Therefore the input power to the motor stays about the same.

That suggests another potential winner for the Pulse Motor Build Off 2015:  Design a pulse motor that senses when a load is put on the rotor and then compensates by increasing the input power to bring the rotor back to the approximately the same speed.  In other words, design a pulse motor that fixes the "flaw" in the basic pulse motor architecture.  You could do it all-analog, or do it with a microcontroller.  Personally if I was going to take up the challenge, I would go the analog route.

MileHigh

A PLL or just a high DC gain speed regulator would do that.  The control loop can be implemented in software or hardware.

[MileHigh]

The duty cycle on time will be the not the same...going faster means less time on(and less input) and going slower means more time on(and more input) needed to reach the same distance of each duty cycle... (http://overunity.com/Smileys/default/wink.gif)
The distance each duty cycle travel at each trigger magnet in each revolution will stay the same which is independent of the speed/rpm...

Well let me just show you my reasoning.  We can just look at this in a generic sense "on paper."  It does not necessarily apply to any specific pulse motor.  We know that in most cases the current consumption of a pulse motor does go up when it slows down because of a generator coil load and I am about to show you some reasons why it shouldn't.  I will deal with that later.

Please see the two attached graphics.  We are working with a trigger coil and base resistor and transistor configuration for this example.  A Hall-effect version will be slightly different which I will discuss later.

The first graphic shows the pulse motor running at high speed with no generator coil load.  The voltage above the green threshold represents when the transistor switches ON.

The second graphic shows the pulse motor running at low speed with a generator coil driving a load resistor.  Because the motor has slowed down, the voltage from the trigger coil decreases.  Therefore less of the waveform is above the trigger threshold an the ON duty cycle decreases.

So notice, at first glance, the duty cycle for switching the transistor ON is actually higher when the motor is running at a higher RPM.   That would suggest that the current consumption should actually be higher and the power consumption be higher when there is no generator coil.  We know that normally that is not the case and the current consumption goes up when driving a generator coil.  So there must be other factors at play.

Some other factors affecting the current consumption:

When the transistor switches ON, the current does not instantly flow through the coil, it takes time to build up.  That can mean if you are switching too fast, the current never has a chance to reach higher levels.  So this factor will tend to reduce the current consumption as the RPM increases.  The way check this is to put a current sensing resistor on the drive coil.  You can't take anything for granted when it comes to investigating the timing of your pulse motor, you have to measure it yourself.

When the transistor switches ON, some of the energy supplied to the coil is the "payload" energy that actually pushes on the rotor.  That means that this "payload" "eats" some of the voltage drop associated with the current flow.  That will also act like an inductance and slow down the increase in current.  So that is a factor that reduces the current consumption also.

When the rotor shows down, the transistor is ON for longer periods.  That means that there is more time for the current to increase in the coil.  If the coil is on "too long" then that can become an issue.  You know your maximum current in the coil, it's just the battery voltage divided by the coil resistance.  If you see on your scope that when the rotor RPM is low that a big chunk of the current through the drive coil is at its maximum, then for a big chunk of the time the drive coil is mostly acting like a resistor dissipating battery power as waste heat.  This reduces the efficiency of the pulse motor and will tend to increase the current consumption as the rotor RPM decreases.

Here is a possible "danger":  The resistance of your drive coil is very low, and so the current can get very high.  If you are not monitoring the current though the drive coil with your scope (assume a very low resistance coil), then you don't know if the current through the drive coil is getting excessively high at low RPMs.  It's possible that that is the reason the current consumption is going really high and it has almost nothing to do with the Lenz drag from the generator coil output.  If this is taking place it's seriously decreasing the efficiency of your pulse motor.

So there is the reasoning.  The truth is that very few pulse motor builders do these kinds of investigations because they don't have the "hard core" electronics experience.  But if you look over what I just wrote, it's not really that hard to do.  I don't build, but if I did build, that would be the fun for me.  It would be to check the current through the drive coil and look at all of the timing issues and stuff like that to get to "know" your pulse motor.

For a Hall-effect sensor, it is only is affected by the magnitude of the magnetic field of the rotor magnets.  The magnitude of the magnetic field does not change if the rotor spins faster or slower.  That's in contrast to the pick-up coil voltage that is affected by the RPM, and that does have an affect on the ON duty cycle.  So in theory the ON duty cycle when using a Hall sensor will not change as the RPM changes.

MileHigh

[MarkE]

There are assumptions about the control going on there.

Coil inductance slows the rate of change in current for a given applied voltage.
Kinetic BEMF subtracts from the available applied voltage for a given power supply voltage.

A sophisticated control would use a relatively high supply voltage in order to overcome both effects under the intended operating conditions.  It would also chop the voltage when the current reaches a desired level.  Even more sophistication would alter that current depending on the rotor position.  That is in essence what chopping microstepping motor controls do.  Those little boards that you can buy for just $3.00 per axis do a really nice job of synthesizing sine and cosine current patterns in the two windings as the motor rotates.   You can typically hook up a 35V supply to one of those boards and run a motor that can only stand a few volts DC across either winding.

[Jimboot]

Nice going Jimboot,
Your setup is as different as mine was...but that's a good thing not to be scared to go in different directions...
Mine setup can produce alot more energy but my issue was...massive cogging and lenz effect...

I have a few idea's so hopefully i can reduce the cogging and lenz effect without reducing the output...i've already collected the stuff i need to build them...just need time now...lol

Keep at it mate...

Thanks mate. I found that a single layer of mags gives much less cogging but the extra Rpms you get makes up for any loss. Also my ratio of mags to steels is 9:2 . I've also run it on a pulse motor with very little torque. I'm rebuilding now with a totally new rig using a Siemens perm mag ac motor rotor as my internal stator. Just trying to work out my cylinder rotor for the steels. I don't want to use plastic because of the potential heat from the eddy currents as grum found out.

[MileHigh]

There are assumptions about the control going on there.

Coil inductance slows the rate of change in current for a given applied voltage.
Kinetic BEMF subtracts from the available applied voltage for a given power supply voltage.

A sophisticated control would use a relatively high supply voltage in order to overcome both effects under the intended operating conditions.  It would also chop the voltage when the current reaches a desired level.  Even more sophistication would alter that current depending on the rotor position.  That is in essence what chopping microstepping motor controls do.  Those little boards that you can buy for just $3.00 per axis do a really nice job of synthesizing sine and cosine current patterns in the two windings as the motor rotates.   You can typically hook up a 35V supply to one of those boards and run a motor that can only stand a few volts DC across either winding.

Thanks for that info on the controller boards.  Indeed, that would be a great design and I had often thought about doing something like that for a pulse motor design in a homebrew fashion.  I guess in engineering parlance you are trying to make a "matched filter" for the coil excitation.

You could do it with a microcontroller running a real-time OS.  Or even develop your own real-time OS from scratch.  The microcontroller inputs clock ticks for the RPM and a once-per-revolution reference tick.  You run a process to measure and filter the RPM clock tick signal and create a corresponding high frequency internal RPM clock.  Let's say you slice the 360 degrees of rotation into 4096 "approximately synchronous" clock ticks.  You literally just have to crunch some numbers and then poke some bits into an internal timer register.  Then, you just allocate say 128 or 256 bytes of memory for a look-up table in memory that stores the coil voltage excitation waveform.  So every time you pass TDC for a magnet fly-by, you synchronously clock out the 128 or 256 bytes of data to make the voltage waveform to drive the coil.

Once you have all of that set up, then you could experiment with loading different coil excitation waveforms into the look-up table.  That's where the real fun would come in.  You could use "Human Fuzzy Logic" (TM) to experiment with the waveform.  Like the proverbial squirrel learning to crack open a nut, eventually you could converge on the perfect voltage waveform for driving the coil with maximum efficiency.  That would be "The Mother of all Sweet Spots."  That would have been a blast for me way back when!

Another strong contender!  lol

MileHigh