Overunity electrolysis - 31 times more effective gas production than with DC

[photonius]

Indeed, it is not so simple to measure the power of the pulse formations at the exit of the circuit. First it`s important to know:
"Have we OVERUNITY: yes or no?"
We can check it with a simple test.
We measure the current from the battery into the circuit and the voltage cross the battery. Now, we get the input - power of the circuit. The power of electrolysis by Faraday divided by the input - power of the circuit is the efficiency between both methodes. All measurements are taken by the same gas volumina.
Okay, this calculation isn`t a perfect one; but better than nothing!

Hans (Photonius)

[MarkE]

Sorry mark this cant simulate current surge capability of SITh thyristor.
Right now only way i see is with array of IGBT in paraller

Anyway we all know that indian reaserch is not best. Thats why i post you Japanese one here.
Very similar one is "Water Electrolysis with Inductive Voltage Pulses ". I think this topic is worth investigating. I dont expect any overunity but it will be nice to have ability to produce more hydrogen with smaller electrolysis without producing lot of heat.

1) There is no high surge current requirement though the primary.  This is a flyback:  Current build up in the primary is slow, and then a rapid voltage voltage spike builds as current is redirected first into the parasitic capacitance and then into the load.
2) Being a cascode flyback, it is the MOSFET rise-time that limits the turn-off and hence pulse generation rise time in the secondary.  That rise time can be made very fast with the correct MOSFET selection.
3) The MOSFET appears in series with the thyristor in all schematics. 
4) Thyristors by themselves have poor turn-off times.  This is why a cascode configuration helps greatly:  The current is turned "off" in the primary by the fast low-voltage MOSFET, while the thyristor or IGBT bears the voltage withstand..

I read the Japanese paper.  The Indians lifted their circuit diagram directly from the Japanese.  The Japanese secondary current waveform should be studied carefully.  Note that current appears to flow at a decaying rate throughout a prolonged discharge.  I would also pay close attention to their plots.  I question that their data supports their claims of significantly improved efficiency.  They were after a legitimate problem of preventing formation of a double barrier.   I question that their data shows success due to both the efficiency plots and the current flow captures.  They did not make any outrageous 31X OU claims, or any OU claims at all as did the Indians.

[Marshallin]

OK i will give it try (i mean your schematic).

Can you please advice what value of electronic components need to be used to get it work properly?

I am using right now 12v 89A power supply. I want to go up to 100khz.

[mscoffman]

I think what the Japanese 2005 paper is trying to do is to show that a relativley low component count circuit
can support pulsed generation of hydrogen at the 60% of unity efficiency level. Unfortunately they didn't specify
a number of critical parameters in their circuit so that the 2007 researchers could reproduce it correctly.

There are a couple major problems that I see;
1) If your think energy efficiency level is important one should make sure their circuit uses only one easily measurable
source of power. This circuit has two! One is the power supply the other is the pulse generator. What they
should do is use a high speed optoisolator so that the fet is triggered resistively from the system power supply.
Historically Fets have had high capacitance gate circuits. How come only JLN seems to knows how to isolate trigger
inputs correctly in a potentially OU circuit where power readings are critical? I think this is the primary error.

2) If you want something that is repeatable and understandable use a generator + driver non minimized circuit
form so you can clearly see what the driving waveform looks like and then the amplfier/driver isolated final waveform
is separate and comprehensible. Is the double trigger of the STI the function of LC resonance or something feeding
back from the complex physical exectrolysis cell? The author should tell *me*. The author is the researcher.

3) How come there are no specifications on the output pulse transformer. inductance? wire guage? core form? core material?
turns count? If the transformer has a step-up transformer form of the schematic, how the heck does the 2007 paper
get .5Vpp when the above photo shows 118Vpp. result without a tremendous output impedance mismatch?
This is a heck of a lot different. Give me some evidence. It should show the approximate initial cell resistance reading
as well. What ever happened to the standard practice of putting the internal resistance as a resistor in series with the
signal generator?

4) there should be an I current trace/ of voltage across sample shunt resistor/  in the 2007 paper.

5) At higher power you d*mned well better use a transmission line feed to cell at these frequencies.

6) How come I do not see the final diode's function in the 2007 paper's output waveform?

So I think reading these two papers is a bit like playing wip the information tail off the snake. The 2007
paper is loosing in this race *big time*.


:S:MarkSCoffman

[itsu]

If the transformer has a step-up transformer form of the schematic, how the heck does the 2007 paper
get .5Vpp when the above photo shows 118Vpp. result without a tremendous output impedance mismatch?

Mark, concerning this statement of yours, to avoid any confusion, that "photo showing 118Vpp" is a screenshot of my nano-pulser output, and thus has nothing to do with any of the experiments in the mentioned PDF's.

I just was commenting on the pulse mentioned in the first PDF showing some 200mV ac like pulse.
My understanding was that it needs to be a strong DC pulse and MarkE kind of confirmed this.

To be complete on this screenshot, i was using 12V for the MOSFET driver logic and 24V on the MOSFET (IRFP260N) drain into a KD226D DSRD into a 56 Ohm resistor.

Regards itsu