There are three separate loops demonstrated in the little girls science-fair model.

Input pulse, emf output pulse, and mechanical load (separate generator coil)

When a viscous load is applied, the rpms drop and the input power decreases until a balanced match is reached between input power and the load.

It runs more like an IC engine than an electric motor, in this respect.

With the simple N-pole motor, a heavy flywheel mass is needed to smooth out the input pulses for smooth storage of angular momentum.

The axle shaft can be used for powering mechanical loads, 1 unit input= 1 converted mechanical unit out + 1 unit cemf out the back end. This is another way the Ed Gray "split the positive", in a mechanical sense.

Adjusting the base resistance allows you to "tune" the motor. Low resistance means that the input current increases and the input pulse-width is wide. Raising that resistance allows you to find the tuning point where peak emf and minimum input current coincide. When you find that spot, the motor will be on the verge of run-away self-oscillation in the stator. This is the "magic" tuning point.

There is no specific value for the base resistor. It varies a bit depending on magnet strength and coil design. As a general rule, if it doesn't "kick over" then lower the resistance a bit more.

I use a 1k 1/2watt potentiometer set all the way up to 1k. Then I adjust it down until it's below the self-oscillation point. You can hear it on an AM radio.

Three critical rules to get the "effect":

1. Very weak magnets
2. Very heavy flywheel with low air friction
3. Very VERY freewheeling bearings

If you think of the Bedini system as an analog of a heatpump, then I becomes more clear what's going on; pulling a VOLTAGE VACUUM over the battery(my interpretation).

This is a method of producing a high floating voltage with a small current input. Hold the voltage high on the receiving battery, then wait as long as it takes for the battery to fully charge. The batteries don't usually "like" this much at first, improving over time. The biggest thing is that you SHOULD SEE the VOLTAGE hopping up on the receiving battery when you start the machine. This doesn't mean that the battery is INSTANTLY CHARGED, it just means you're holding the voltage high.

Once built and tuned, the motor/gen has a known, fixed power input every time you fire it up. This means that while you're "teaching" the batteries to accept this type of charging, you can run it from any 12v source you want to.

If the battery you're charging cannot be pulled up above 12.5v in a fairly short time by your charger, the charger isn't big enough for the batteries you're trying to use AND your battery is probably sulfating. In this case, you'll need to add more windings around the wheel.

Let's say you're stator design draws 80ma a peak tuning. Adding a few more identical windings will simply multiply that input requirement and also multiply the emf collected in the capacitor. This will give you increased capability for POPPING the voltage up on larger batteries.

You have to GIVE to GET in any system, even energy-collection systems. The BAIT is very small compared to the FISH that jumps on the hook.

Set a 1amp input Bedini motor next to a 1amp input commercial charger and then you'll see why it's different.

The small motor is a MODEL to learn from. But, if you build it correctly, there is plenty of room for expansion just by adding more stators and driver circuits.

 Don't be too discouraged. There's a lot to digest.

 In my experience, if the device is not big enough to pop the voltage up on the receiving battery, then you either need to step down a battery size OR add more stators around the wheel. Since his basic device is just a slow pulsing of a the output capacitor across the battery, you could start there also by modifying the pulse rate.

If you look at all the pics, you'll see some things to try.

http://www.icehouse.net/john34/bearden.html

Read the top paragraph and look at the old prototype machine design. Ignore the motor on the front end. Look at the way it CHARGED the capacitor then PULSED that cap across the battery. This is the SAME FINAL CHARGING SYSTEM. But, that old system rips rotating contacts to pieces. SNAP! SNAP!

The only component that's not shown in that diagram is the diode between the ENERGIZER(magneto) and the capacitor. The single diode provides pulsed DC "ramp" charging of the capacitor.

Screw a steel sheetrock screw into a block of wood and leave the HEAD of the screw about 3/4" above the surface. Now wind about 200 turns of magnet wire on it. Now mount the wood block so that the screw head get's pulsed by the rotor magnets as they spin by. Put a single diode on one lead of that small coil and charge an electrolytic capacitor with it. THIS is the OLD SYSTEM for charging the caps. It requires the motor on the front end to use it's emf internally, producing torque to turn the magneto. You'll notice that as the capacitor voltage rises the motor speed will also rise due to the lessening load on that winding.

The new (monopole motor) system just uses the emf directly by sending it back out to the capacitors, greatly improving efficiency since there are fewer drag points.

If you put a load on the small coil the motor will slow down, power input will drop, and the capacitor will still charge with emf in a regular fashion. This is demonstrated in the "shoolgirl" motor posted at:

http://www.keelynet.com/bedmot/bedmot.htm

Here's another modification he made to modernize the output.
Solid state pulse output using power transistors as "finals".

http://www.icehouse.net/john34/index100.htm

The specific pic is:

http://www.icehouse.net/john34/mono-pole11.jpg

In this one a 555 timer is used to control the output pulse-width and pulse-rate. You could just as easily control the transistors with some nice clean low-voltage commutation. Then you don't need to fry your contacts all the time and you get more consistant output pulses.