Joule Motor

[TinselKoala]

Hall effect sensor commutation, then, will be the most precise and adjustable method. You just need to find the right place for the Hall sensor, and perhaps use a Schmitt trigger stage or CMOS inverter to square up the drive pulse. The Hall sensor can be used to fire a monostable 555 timer to give pulse width control. So you vary the sensor position for timing and the pulse width for dwell, or just use the edge of the Hall pulse to trigger the 3055.

Most of what's shown here may not be relevant to your project; I just wanted to show the use of the Hall sensor.
http://www.youtube.com/watch?v=90rMGmskqXQ

[Lynxsteam]

Thank you TinselKoala,

I will try a more sophisticated triggering on the next build.  This weekend I will work on a machined version with much closer coil to magnet arrangement and strong cylindrical Neo magnets on a stainless shaft in proper bearings.

Here is what I think the cycle looks like through 225 degrees rotation.  Hopefully you will see any mistakes in current flow direction.  It is very confusing in 3D.  I don't remember where I tapped the AC.  This diagram reveals that it makes a difference.  I know that the motor runs faster with the generated voltage feedback.

[wings]

solution without connector ?

coreless, brushless, sensorless, controllerless DC pulse motor based on transistor oscillation

http://www.youtube.com/watch?v=2vOviGW3gFo&feature=plcp

[synchro1]

Very nice use for the Zener Diode.

[Lynxsteam]

Very impressive video.  That type of motor works well and is related to the Bedini SSG and Window type motors.  Any time you trigger the transistor with the stator coil you get a quick pulse of voltage to the drive coil.  The faster it runs the more power you use.  These types of motors aren't very torquey, so while they go around you can't drive much with them. 

I am trying to do something a little different.  I want to power the drive coil for a certain amount of the rotation, use the flyback, and then see if some of the induction that would normally occur can be used to amplify the voltage.  In order to do this the induced voltage and the spike have to be in phase.  Also I am hoping I can use the induced voltage run back through to help speed the rotation.  It is probably another goose chase.  But so far the results have been interesting enough and surprising enough to keep me going.

DC positive stays positive until it passes back to the battery and across the plates.  It can be on its way to the negative side and still do work.  There are two sources for current: 1.  The battery and 2. the induced current from magnets passing the coils.  It takes work to move magnets past coils and the power out is always less than the power in.  But at least I am able to get higher voltage out than in, and I can back charge the drive battery.  Now if I can demonstrate this motor is efficient that would be an accomplishment.