Motionless Generator (Bearden/Flynn/Witts/Naudin/magnacoaster)

[d3x0r]

Questions:
1) Do I have to drive the driving coils with a high current signal?  This neodymium magnet is pretty tough, not sure how much drive I would need to entirely block the (say north-direction) flux without just consolidating(constricting) it's flow through the center of the field the coil generates...
Anyhow.  I've attempted to build something like...http://jnaudin.free.fr/meg/megv21.htm

I didn't go with a 'space age material' for the core, I instead have wound a core using iron fencing wire.  The core is approx 1 square inch in cross section, and I planed on using a 1"x1" cylinder (no hole) neodymium magnet for the fixed magnet part.  The Witts document says they had their 'best' results with 300 windings 26gauge wire for the flux-steering coils and 20 windings 12 gauge on the output.  I wound 33m of 24 gauge wire as the flux control coils and 3.5yards of 12 gague wire for the output (should be something like 300 turns and 24 turns respectively). 

The coils are not as perfect as I would have liked them, but the aren't too horrible I think.  I looked all over for the best electromagnetic windings to generate the strongest gate I could.  It seems that for an electromanget, just a single winding over the length of the magnet is the best(?).  So, I wound the control coils longer (spanning from the pickup coil to the gap for the magnet in the center) than tall.  I was originally considering winding the iron core, then securing it and cutting it in half so I could just slip better wound coils on.  I had a bobbin that was 1.5cm wide for 20 windings across of the 24 gague wire, and then 15 layers would have been 300 windings, so I measured the length of wire for the average circumferance times windings and got something like 33 meters.  I wound 24 windings of 12 gague wire around a 1" diameter pole, and then cut the leads with penty of length, then straightened the coil and took s arough measurement to get 3.5 yards.  I then matched the other 12 gague color wire with the first and wound them on the core.

The permability of this iron wire is fairly low compared to the 'maglass(?)' or 'permalloy', but I figure since it's more permeable than neodymium (600x) that it probably has a good probabilyt of containing at least as much flux as the 1" magnet can produce.

I've been playing with this apparatus with my signal generator and scope to determine its resonances, and test the behavior with and without a fixed magnet.  Some things I've learned (without additional magnet) 1) the output coil that is further away from a driver coil takes longer to receive the signal (by several 10s of nanoseconds).  2) my signal generator goes from minimum volatage to max in 70nanoseconds, and it takes many tens of microseconds for the full voltage to reach the other side of the coil. 

I should try and capture some more screen shots and fill in some details, but here's an attempt to descibe the waves.

I have attempted to drive the coils each with a half wave in the appropriate direction from ground to oscillate them as blocking the appropriate pole of the magnet the driver coils are near.  so from the hot of the signal generator I go to two diodes in opposite directions, then through each coil then all back to the same ground.  When I don't have a coil in the circuit on a side, I put a small resistance instead.  If I drive one driver coil from one of the diodes, as a sine wave, as the voltage goes up to the peek, the voltage goes up the same, once it approaches the peek, the volage on the side of the diode that is the coil begins to drop, then as the signal is going down, the coil drops at 2x the acceleration, and ends up at a negative voltage below the neutral.  I assume this is from the field collapsing back on the same coil.  for my loads; I have across the green output coil a 1m resistor, and I have across the red coil two LEDs wired backwards to each other so they should light in different directions. 

If I then connect the other coil in the correct direction to be appropriate to block the north pole, I get the same sort of signal, but neggative of the first... then of course I get back feedback additionall complicating the signal back on the first driver coil.

If I add a magnet small or large, I get no change, other than as I am moving the magnet into or out of position, then I get feedback on both the output coils and flux gate driver coils.

I tried to add some more diodes to get the back signal that went negative to drain to ground, but it didnt.  But then if I drive this with a square wave, it's just that much harder.

And the next oscillation point I think is like 25mhz and my signal generator only goes to 20 as a sine wave and 3 as a square/pulse generator

[d3x0r]

And with the magnet north to the top

All the coils are wound in the same direction... wound so the outermost coil is from the center, around the bottom and up the side... so like looking along clockwise around the center core, the coils are wound clockwise.

[d3x0r]

Test 1) using signal generator with square wave directly to coil - very low power, lots of delay getting signal through coils...
2) connected signal generator to input of 1500W audio amp, better drive power, more response, enough to test that the coils are mostly-north-south aligned correctly ( a magnet aligned in the same direction is attracted, in opposite direction is repelled (pushed away by more than its own weight even), and in sideways, wants to turn to align north-south with coil.  So, the coils are aligned correctly to oppose the fixed magnet in the center of the coil.

Using the amp I thought I saw a larger induction in the two output coils with a permanent magnet in place, since in theory a larger portion of flux than the coil itself produces should move from one side to the other because of the permanent magnet, right?  But after a few repetions and adjusted levels, it was some other thing I was seeing.

With both coils connected, the signal out of each coil looks like it's getting the full signal wave instead of the halfwave it's directly driven with, because it ends up getting an induced voltage from the compliementary drive....

[Jdo300]

Hi d3x0r,

I find your current research very interesting and have thought about doing some experiments with these flux switching devices myself. Since you are already commencing to experiment, I do have some ideas for you which may help you in determining the performance of your device.

1. I would suggest either two H-bridges to drive your input coils or a low-side switching circuit similar to what Jean Louis Naudin used. The device should operate with the lowest losses when being driven with a square wave input as opposed to a sine wave. According to the MEG information I am aware of, you should see a sine wave if you drive it at the resonant frequency but this is easier to spot with square wave inputs.

2. To get a better idea of your power input changes, I would also recommend using a current shunt (or a 1 ohm, non-inductive resistor) in series with the amplifier so that you can monitor the current. Most likely, the audio amp will mantain the same voltage as the impedance of the core changes so you would need to see the current waveform to know for sure how the power is changing. However, if your amp is running on 12V DC, that would make it even easier to monitor the power input from the DC side of the circuit.

3. Once you get a basic characterization of your device, the next step would be to optimize the switching circuit to keep your output power from back coupling to the input. I have some ideas about how this can be done too if you're interested.

- Jason O

[d3x0r]

I had to put this on hold; after moving on to joule thieves and trying the same stranded iron core as a smaller toroid, and it required a lot more amps to stimulate flux; so I have to upgrade my core was toying with this page...


http://www.surplussales.com/inductors/FerPotC/FerPotC-5.html  (the U/I makes a 4inch square) or maybe 4 I's to make a 4inch high (2 inch in center) 6 inch long core... but the thing that concerned me was, these all have sharp corners wouldn't round things make better flux containers?