Excerpts from: http://www.energeticforum.com/renewable-energy/2242-water-sparkplug.html
 

From user  Jetijs about this GoToLuc circuit:
 
 

Proposed modification by Jetijs:

This is because at the first picture the cap is charged up and when the switch is closed, the cap is discharged in the ignition coil, but in the same time the high voltage diode is disconnected from the + terminal of the cap. If you do that, you will have only a small spark. The high voltage diode anode should be connected to the cap + terminal at all times or you will not get the big bang. In the second picture you can see how I did this, first I closed the S1 switch, this made the cap charge up. Then I opened the S1 switch and closed the S2 switch, this discharged the cap into the ignition coil. In fact I did not have the S1 switch, I charged up the cap and just turned the variac off before the discharge.

Lighty, I also tried your reversed polarity circuit - it works just the same, I observed no differences. But to get the big bang in this reversed polarity circuit, you need the diode cathode to be attached to the negative terminal of the bridge. If that switch is open and there is no connection between the diode cathode and the negative cap terminal, you will get just a small, weak spark.

Also, **~Imhotep~**, that is a good idea about trying to capture the inductive spikes from the low voltage side. I thing that we could make a bifilar coil with thick wire and about 100 turns and just pulse the low voltage side through one of the bifilar windings. The we just add a rectifier on the other winding and capture the spikes in a battery. Could that work?

Gotoluc answer

Hello Jetijs, you are right about the cap not completing the circuit. Thank you for bringing this to my attention. I will test it with your suggestion and will inform you of the difference. If it works well or better I will update my circuit (where I originally posted it) with your suggestion.

I was blessed to find this circuit but am a novice in EE and I am very open to any suggestion that can improve this simple circuit. So thank you for taking the time to explain this.

@everyone, I am very happy to see this topic going so well here compare to where I originally started it at Overunity. Keep up the good work and share your finding as Jetijs has done here. This is a great example of finding an improvement and then sharing it. Also a big thank you to Aaron for starting this topic here.

This kind of support and enthusiasm can change our World.