Understanding Bedini AOP Improvements

[TinselKoala]

Try putting a fast, highcurrent diode like MUR1560, UF4007, or similar at the connection of Coil to NMOS Drain pin. Anode of diode to Coil-Drain connection. Cathode end of diode then will present "high voltage" pulses with respect to mosfet Source (circuit negative rail) on mosfet turn-off which can be used to charge capacitors or batteries, light neons, etc. Mosfet must turn off cleanly; a 1 megohm resistor from Gate to Source will help it turn off.

As you can see from my MHOP videos, when it's set up properly the motor will self-start just by turning on the power.

[Raccoon]

So after boosting the trigger and with your recommendations, I've retry different things.
Now the rotor could get very high RPM & torque, and start itself with this setting.

I have set this 1 Mega Ohm resistor, but seem to do nothing for me.
I was able to get voltage from a capacitor from drain to source=mass. (with a 1N4002 diode)
I try to put the Neon light (with or without the cap in Parallel): No success.

Because I don't have high voltage rating cap, I don't know how far I could do (without blowing it in my face) ?
The best I could use is 1500 uf rating at 6.3V. I checked 16V "without blowing it" with the multimeter, but I've fear going higher.

Also my cheap multimeter wasn't able to read high spikes (without the cap).
With the NPN transistor, I could get 20V spikes (unstable reading), and also neon keep fire.

I think I have understand why this don't work : Also you have already tell me TinselKoala, and you're probably right.
All MOSFETs I try, have maximums VGS rating for 20Volts, so I think it kill or protect from over voltages.
My NPN 2N5551 was probably unprotected (no internal diode), and could get maximum VCE rating for 120 Volts.
So, before buying some, I would probably search for better MOSFETs or Transistors into olds boards.

Actually, I've build a simulation for the one-shoot NE555 time duration, and could successfully offset in time thanks a capacitor charge and a second voltage comparator (I will share when the final simulation was ok for me).

[TinselKoala]

If you have a CFL bulb you can take apart, inside there is a nice capacitor, usually 22 uF and 200 or 250 volts. You can also put electrolytic capacitors in series, this will "add" the voltage ratings (while reducing the total capacitance).  Yes, it is possible to literally blow up small electrolytics by overcharging them, so stay out of the line of fire!

The one-megohm resistor from Gate to Source will help your mosfet turn off, but you may not be able to see any effect yet, or without an oscilloscope.

You shouldn't be coming anywhere close to exceeding the Vgs of the mosfet using the simple op-amp triggering system; in fact that is one of the main advantages of using this system instead of the regular Bedini method: you always know that your driving pulse to the power transistor is within its limits, since it is just coming from the op-amp's output, not some direct spike from a coil's field collapse.

There may also be a couple of good mosfets inside the CFL, or bipolar transistors of good ratings like "13003" which is a good HV NPN BJT.  The mosfets found in PC power supplies are usually low voltage, high current units. For coil drivers you want modest current handling but high Vds voltage, to withstand the spikes. The Gate voltage maximum for all mosfets will be 20-30 volts, but you should use one with 200 volt or higher Vds for your coil driver. Same with BJTs. Some TV/CRT horizontal deflection transistors will be good coil drivers, but I think mosfets are better and easier to use.

[Raccoon]

Yep, I've disassemble one varilux CFL (which is already dead).

I know that this thing contains good components, but I couldn't expect all of that.
My treasure find list is :
• 1.5uf 400v capacitor
• x2 NPN transistor 13001S (400v 0.5A, only 1 Watt)
• x5 Diode 1N4007S (very useful for me)
• one high speed germainum diode (dont know exactly)
• one transformer + one inductance + one thermistor

For those who want doing the same, be really careful, the thing is it could have keep a lot of energy (thanks the capacitor). So my recommendation is to let it unplugged and unused for a month. Also I checked the capacitor, even if I don't use that for a long time, it remind at 8 Volts ! I discharge it thanks a resistor and a led.

Thanks TinselKoala, you've successfully tell me I already own all that I need
Also thanks for the transistors limitations explanations, before I try searching things like the MJL21194 (only 200V but for 16A !)

So I will test with all of that, and i should probably buy one similar IRFP360 Mosfet.

[Raccoon]

I've made the one-shoot pulser with the NE555.
It works but really strangely, i was able to run the motor only with the MOS. It do somethings really inaccurate, the motor start increase and decrease, and this occur as an unlimited loop. The fastest perfectly synchronize speed I could get seem less than without the one shoot. I think this circuit could be useful on simple monopole bedini, but with my 3 phases modified motor, it do less than the simple MH circuit (which is great for me).

For me moment, I will stay with the voltage comparator, trying different transistor (I'm in reflection about buying one good MOS & one good NPN). Also, I should probably build my own pulse motor. There's a youtube video which inspire me a lot, see the first video in this thread : http://www.overunity.com/14129/any-one-checked-into-the-skyship-motor/msg395281/#msg395281

In this design, I think we could see an Hall effect sensor, and it reminds me that you've made TinselKoala. I will love to get more informations about it ! I will probably do the same with my new motor, it has 2 advantages : perfect timing offset, simple circuit, the one-shoot pulser could be forget.