Originally posted at
http://www.overunity.com/index.php/topic,2637.78.htmlHi,
If I am on the right track with the spinor resonance then phase, pulse timing and coil configurations wont quite be what you expect -- that doesnt mean they will be complicated, but not what classical intuition would lead you to expect. Simple experimentation is called for...
I won?t have time to do experimental work on this project for a few months. I have some other projects to finish, an overseas trip and an office to move first.
If anyone is interested this is exactly how I will start. I?m giving instructions here so that others can try this if they wish. No guarantees -- this is just what I plan to do based on every bit of practical, hearsay and theoretical information I have to hand.
Construction:
1. Starting with a single collector coil and primary. I would use some heavy duty coax for my collector coils as this provides a perfect core to wind the primaries on. RG218 is a good candidate. Cut 1196mm length and strip the outer sheath and remove the braiding ? this leaves the 15.7mm OD PE dielectric with 4.95mm OD copper core. The reason for using the coax is that you want precise and even spacing of the primary coil from the surface of the collector wire. I think this will work better than a stranded core ? it?s the precise/consistent spacing that is critical. Leave the striped coax straight for now.
2. Wind a primary coil. I would do this on my lathe winding tightly with 1mm OD magnet wire over a 15mm OD steel former. After winding ? when the tension is released the coil will spring back enough to easily remove it from the former. It should slide snugly over the 15.7mm coax core you have prepared.
3. The RG218 core is 5mm OD solid copper ? not very flexible and very hard to solder too neatly. For connecting to it I would drill 0.95mm holes radial into it right near the ends. I would then inert 1mm solid wire that I had sanded to a slight taper.
4. Shrink wrap to hold the primary firmly in place. Alternately wrap tightly with a single layer of electrical tape.
5. I would bend the completed collector with primary coil assembly over a circular former to bend into circular form. Tape the ends together and you can remove it from the form. The heavy copper core of the RG218 will help to stabilize the circle.
6. *Variation: I would actually wind a perfect counterwound primary in anticipation of testing single primary vs counterwound. The reason for this is that we are interested in the spin wave front and not the B field generated by the primary. It may work better to eliminate the B field by applying the pulse to a counterwound primary. It is easy to wind a counterwound primary now and start testing driving just one helix. (I attached a photograph of the first counterwound coil I ever wound -- I found a novel way to make these perfectly and easily and made a whole lot of them a while back
**OK -- you can also try a complete coax solution -- using the shield as the primary and core as the collector. In theory I thik this would work but the speed fo the spin wave front would be to fast to make for a practical design with a managable circumference.
Test procedure:
1. Start with just one primary/collector coil.
2. Connect the collector to a low impedance load ? a 100w incandescent light bulb.
3. Drive the primary coil with the shortest pulse you can cleanly generate. I would use a 600V MOSFET and drive with rectified mains voltage ~ 340V. I would use a good MOSFET driver like TC4421 and do everything else possible to ensure a clean fast pulse.
4. Slowly sweep the drive frequency from 1Khz or so up to 100Khz looking for a peak in output. At this stage any peak should have a relatively low Q so it should be easy to find.
5. I would then vary the pulse width ? looking to see if there is an optimum.
6. I would repeat the previous step with single wound and counterwound primaries. I would try it with the primaries open circuit first and then closed to ground through a modest resistive load ? 1K ? 100K or so. The reasoning here is that we are looking for the optimum timing of the spin wave front pushing (SM talks about kicking and squeezing) a charge pulse down the collector == very fast circulating static field == very fast rotating magnetic field.
7. After finding the optimum drive frequency for the first collector coil I would then add a second coil set. Space it one primary minor diameter above the first coil set ? that will be 15.7mm. I would use acetal or nylon spacers to hold the coil sets apart and tape them firmly together.
8. I would series connect the collectors and check that with just driving the bottom primary things are working as expected. With the first primary (bottom coil set) running I would sweep the drive frequency of the second primary looking for further increase in the output. I would expect the Q to be higher now so I would sweep more slowly so as not to miss a resonance peak.
9. I would repeat the previous step with the third (top) coil set added.
** After getting the three coil sets working as well as possible I would look at adding a bias coil around the whole set ? this is the single outer toroidal winding over SM devices. I would use this winding to control both the static and magnetic bias environment for the three coil sets.
** All of this would be done with three independent signal sources. Only after getting everything tuned and working like this would I attempt to put in resonant feedback ? from one collector to the next primary.
** I would also repeat all of the above with three segment primary coils on a single colector and injecting the three frequencies into those three primaries on a single collector. I would try 3 primaries covering 120 deg and also 3 primaries covering 360 deg but tharting 120 apart.
If any of you are looking for control boards check out
www.futurlec.com ? I would not use the
http://www.futurlec.com/ATMEGA_Controller.shtml even though it has 6 independent PWM?s ? it?s too slow and you cant get good precise pulse width control over 10Khz or so. I am looking at using 3 of these
http://www.futurlec.com/ARM7024_Controller.shtml and syncing them up.
Mark.
