This reaction, essentially is a pulse. Acid eats metal, creates H2 and O2. Bonds broken, H2 and O2 taken from water (using gathered energy) and reaction starts over.
I noticed that, through almost the entire reaction, we have metal sulfates, already reacted, sitting there taking up valuable space. We've already gathered energy from their reactions, so let's use it now, instead of waiting until the entire amount is done, that's a bit wasteful.
As soons as we appy the gathered energy to the metal sulfates, they'l break apart, and react again just as fast. There'll be a microsecond gap between putting energy in, and getting it out, plus H2 and O2. To keep this as efficent as possible, wasting as little time as we can would keep the H2 flow near constant.
I like the 2-box idea, actually. Energy produced in one box immidiately (well, almost) traves to the other box to break bonds there, and vice versa. Eleminates the need for a complicated pulsing system, or microcontrollers.
As for keeping the liberated H2 and O2 from being absorbed in the regerating process, the only sure way of not having it used again is giving it a way to escape as soon as it's created, out of the top. This would need a large surface area to insure that almost all of the created H2 and O2 is collected, and not re-used. But that's impractical, so we need to find a decent surface area to allow enough of the gas to escape, but still have the unit remain portable and practical. We need something reproducable, not a 10x10 metre pool to make this work, and minimise losses.
Could we make 4+ (relatively, again, we need to have enough water to last a little while in each) thin compartments, stacked on top of one another, and connected in pairs for energy transfer? You bet, and we can keep it a decent size, I think.
Depending on what the strenght of the bonds is, and how much energy is needed or released in creating and breaking them, we can find out if, how often, and how large of a electrical pulse we need to break all of the metal sulfate bonds that were left behind (not broken by the gathered energy that was put back in), and keep all of the available sources of energy and H2 going. As tiny as the losses should be, it will stop it eventually, and keeping it at near 100% capacity at all times should keep the gas coming consistently.
If it's needed (which I think it will be, if not very often), we need a point at which to input the pulse "booster" (think a cup of coffee, keeps you going for a while, but wears off, so you have another to re-apply the effect). We should be able to let the gas production drop anywhere from 10-20% (maybe more, depending on H2 demand, or less) before introducting the pulse. Having it coming in very often, say, if the gas production drops 5% or so, would waste energy. True, the longer we wait, the larger of a pulse we'd need to deliver (more bonds to break, more energy needed), so we need a balance of efficency and energy needed to re-charge the reaction.
, but with your help (mega kudos for you for figuring it out) i think i understand the direction alaskastar is pointing us in ... make a battery with a voltage > 2v, and catch the hydrogen that is evolved in the dis/charging processes