Shorting coil gives back more power

[joefr]

Hi Gyula and all others

Thanks all of you for positive comments and suggestions

Now lets answer the questions:
Gyula
Just wondering what is the FET type you use for switching?
I am using this type of FET IRFB5620 PbF
http://www.irf.com/product-info/datasheets/data/irfb5620pbf.pdf
and this type of Hall sensor:
http://uk.rs-online.com/web/search/searchBrowseAction.html?method=searchProducts&searchTerm=680-7481&x=0&y=0
The gen coil resistance is 13.4 ohm and inductance is 123.5 mH

In your photo showing the two Hall sensors above, both sensors are positioned at North poles. I ask one of the sensors should not be facing a South pole instead?
I made new Hall sensor layout which triggers on both magnetic poles south and north, look at the bottom video.

I put small trigger magnets on the side of the wheel. All small trigger magnets are south pole oriented so they trigger mosfets only at positive part of the sine wave.

MAybe a series capacitor in the some uF (or higher) range is the clue for this, as shown in Doug Konzen schematic, see the 40uF cap labeled as 'high bypass filter' in series with the AC input leg of the full wave diode bridge:
http://www.overunity.com/index.php?topic=10398.msg276614#msg276614
Yes I am using this circuit with 40uF series cap but with your circuit mosfet TC4421 driver

I record new video and show all the steps which all of you suggested:
http://www.youtube.com/watch?v=qPagVQGfDDU

Joe

[gyulasun]

Hi Joe,

Thanks for the good informative video.

Please check the FETs gate-source pulse width how many millisecond it is?  (shown as starting in video at 0:55) Also check what is a shortest pulse width you can achive? (by moving the Hall sensor radially away from the trigger magnets)

If I suppose you use the 5 ms/DIV setting in the video when showing the gate pulses then what I can make out is the pulse width changes roughly between 2.5ms to 4ms: this is STILL very wide for the shorting, considering your one full sine wave time is about 8-9ms.  It would mean the shorting pulse would lay over almost for the total half sine wave (if I consider only shorting one half of the sine wave) and this would explain why it is like the gate control pulse would be inverted: using the shorting on both polarities of the sine wave the shorting times cover up both halves of the sinewave and this tricks us all.

(you have chosen good types of MOSFETs and Hall sensors.)

Thanks,  Gyula

[joefr]

Hi Gyula

I tested today the coil shorting with circuit which is using only hall sensor and mosfet ( without TC4421 driver ).
I got the exactly same result, the coil is shorted too long.
So I am thinking that this wheel setup is not good for testing coil shorting, because trigger magnets are positioned on inside part of the wheel.
I put smaller trigger magnets as showed in the video, but because they spin slower the result is almost the same that before I used generator magnets to trigger.

So i will build the new wheel, something in this way at the bottom photo:

I ask this two questions in previous post and I ask for suggestion if anyone has more experience with this:
I will be using 8 coils but I need proposal what will be better to use 7 magnet, 8 magnets or 9 magnets?

And what diameter size of bottom magnets would be fine for good power generation with coil shorting (charging caps with spikes and then dump to battery ):
1. http://www.supermagnete.de/eng/S-20-10-N
2. http://www.supermagnete.de/eng/S-30-10-N
3. http://www.supermagnete.de/eng/S-35-05-N
4. http://www.supermagnete.de/eng/S-60-05-N

Joe

[Feynman]

Hey guys,

I have a 4.5cm x 0.5cm ferrite rod with about 100-150 turns of 30AWG enameled magnet wire in one layer.  From your experience in coil shorting, what is the best configuration for an electromagnet?   One layer of wire?  More than one layer?  Can I wind the seperate layers individually, and connect the layers input leads with alligator clips or is this too much stray inductance for a 50khz pulsed electromagnet? (Should the multiple layers be continuous?)

Thanks,
Feynman

[Arthurs]

I made a simple coil short-circuit test, but it seems just the circuit voltage increases, the total energy has not increased.
Specific:
1) If a direct short reed switch coil, the speed decreased. Speed ​​drop means reduced output energy.
2) a reed switch if the series capacitor (0.47-47uf all done between the experiment), and then in parallel with the coil, the speed has not changed. But the voltage and current output has decreased significantly.

I think we proved that this method can really increase the output energy, you must use incandescent experiments.
Using the LED does not prove anything. While the use of short-circuit mode LED will light, but this is just the voltage increases to the LED's forward voltage.
But I do not have a big magnet for incandescent bulbs in the experiment.

Where is my experiment is not correct?