why do we limit our e-lyzers to 13.8 VDC

[Helz_McFugly]

@phill, Yea makes perfect sense. so both examples I gave would be exactly the same?
I just see a setup like this http://www.youtube.com/watch?v=J1FYGCSDekE and think "WTF arent we using this type of config in our cars with a dc converter?"
but we can, we would just have to make it bunch of +nnnnnnn- cell's in parallel? correct?

[biggy boy]

If you are making a very high output HHO Gen,
the one benefit I can see to using higher voltage is:

The amperage draw! Lower voltage means higher amperage.
Compared to higher voltage lower amperage.

Lower voltage higher amperage:
Bigger wires need, higher amperage rated controls IE: relays, fuse holders, connectors, PWMs....

Higher voltage lower amps:
smaller gauge wire, smaller relays, smaller fuse holders, smaller components for the PWM.............

[IM2L844]

There are several theories about exactly how the phase transition between the electronic phase and the ionic phase at the electrode/electrolyte interface takes place, but that's neither here nor there.

Excluding losses in efficiency due to side reactions, joule heating and so forth, at room temperature and 1 atmosphere of pressure you will ALWAYS get a MAXIMUM of 11.4 milliliters per minute per amp per cell.

This is not speculation on my part. This is a fundemental law of nature. Some of the greatest scientific and engineering minds on the planet have thoroughly and exhaustively studied every aspect of this you can think of with the most precise equipment under the most tightly controlled environments and they have universally come to the consensus that Faraday's Law governing electrochemical equivalents is a fundementally ridged, precise and unbreakable law of nature to which no known exceptions have ever been verified.

I'm not saying that some guy in his garage can never discover an exception, but, lacking independent verification by mutiple qualified professionals, the burden of proof should be extraordinarily nonindulgent.

Anyway, yeah, I took off on a little tangent, but 11.4 mL per minute per amp per cell will let you do the math for yourself or you can just play around with my calculator and it will do the math for you.

[biggy boy]

Just so I understand!
The max production of HHO possible at 100% efficiency is 11.4 millilitres for every amp you pump through an entire Generator or is it per each individual cell space?

[IM2L844]

It will change with temperature and pressure , but, yes it is per each individual cell space. An electrolyzer is NOT "a cell". A cell consists of two electronically conducting phases (at the plates) connected by an ionically conducting phase (through the aqueous electrolyte between them). In a single cell, as an electrical current passes, it must change from electronic current at the surface of the positive face of one plate to ionic current through the electrolyte and back to electronic current at the negative face of the other plate.

The only way to increase efficiency with brute force electrolysis, which is what most of us are trying to do, is to get the maximum amp draw with minimum voltage and according to Ohm's Law the only way to do that is by reducing resistance.

[Philldpapill]

VERY well put, yet again, IM2L844!

@biggy_boy #6
Running 150V with your 200V MOSFETs is still pretty excessive. In one of my early PWM designs, I was getting WAY more power dissipation in my FETs than I had figured. It was a total mystery to me for the longest time until I found the problem - stray inductances.

I was only running 30A @ 12V input, but my FETs were getting really hot to the touch. What I found was a very simple, but hard to fix, problem. The actual 10gauge wires I used were only about 5' in length total, but doing a quick inductance measurement showed that I had about 2uH of inductance in those lines. When my FETs would turn off, they would get slammed with a high voltage spike, causing the internal diodes to reverse bias, and conduct. The FET was a 55V, 0.0085 Rds, power FET, but those voltage spikes were causing the excessive dissipation. In other words, for a fraction of a second, the FET internal diodes were turning on, dropping 55V, and conducting 30A. This happened for a very short period of time, but when I was switching at 10kHz-100kHz, that little bit of dissipation happened thousands of times a second, resulting in alot of heat and sometimes failure.

The solution(which is in my current design), is a rather beefy capacitor bank at the inputs of the PWM. This allows the energy, that is stored in the stray inductance, to transfer to the capacitor bank instead of slamming the FETs. This is reactive energy, so it is returned back to the load. The bank needs to have a low ESR(Equivalent Series Resistance which is a property of the capcitors), so that the caps themselves don't dissipate too much power via the ESR. Big caps, however, are pretty pricey. I've found some on Mouser that should be pretty good. Only trouble is, they are surface mount, and will take up a 5" x 5" area of board. My simulations show that for a 35A ripple, I'd need about 22 of these. 25 would be a very safe bet. That ripple is NOT the actual HHO cell current. Using an inductor like I had said earlier, can yield about 75A of HHO current, with 35A ripple.

Anyway, I've almost lost sight of the point... Oh yeah - don't operate your FETs at 150V... You MIGHT get by with 100V input voltage, with minimal heat.

Just a quick rant... Those PWM's on ebay that boast a whopping 55A and use "6 100A mosfets" are decieving. The PWM could probably handle a completely non-inductive load at 55A, but throw in the tiny bit of inductance in the wires, and good luck working with even 30A without massive power dissipation.

[Helz_McFugly]

I love this place

[biggy boy]

So you use Inductance to control current spikes.
Capacitance to control voltage spikes.

Inductance in series with the load ( HHO gen)

Cap bank in ?parellel? with the load?

[Philldpapill]

Er, sort of..... Inductors resist changes in current, and capacitors resist changes in voltage.

In my setup, I'm going to have an inductor in parallel with the load(the HHO cell), and a capacitor bank in parallel with the INPUT supply. I'll also have a diode in parallel with the inductor/load series circuit. This allows the current to recirculate through the diode with the mosfet turns off.

[biggy boy]

Er, sort of..... Inductors resist changes in current, and capacitors resist changes in voltage.

In my setup, I'm going to have an inductor in parallel with the load(the HHO cell), and a capacitor bank in parallel with the INPUT supply. I'll also have a diode in parallel with the inductor/load series circuit. This allows the current to recirculate through the diode with the mosfet turns off.

Ok
In your original diagram the inductor is in series with the load
relative to the diode!