Originally posted by Michael John Nunnerley
View Post
I'll definitely be building a water splitter... it's on the 'to-do' list, so if I can work on it concurrently with waiting for my electronics guy to build my other projects and waiting for the engine parts to get back from fabrication, that'd be great. Of course, it'll all be completely open source... I have no delusions of getting rich off something like this... the likelihood of ending up dead or in prison is much greater, and I'd much rather avoid that.
As regards your "High voltage nano second pulsed electrodes" comment... that sounds like corona discharge. I'm going to be having one built for an ignition system on a small engine, so I've done lots of research on it.
But in the case of the water splitter, one way to perfectly time the pulsing of the magnetic field and the pulsing of DC voltage on the biasing plates is to add another coil on "The Gizmo". I'll call it the Biasing Coil.
You'll note in the top-view graphic (attached in .PNG and .PDF format) that opposite the water tank is a long magnetic flux path that connects the two permanent magnet banks when the coils are energized... but when they deenergize, the magnetic flux collapses as it goes through the shorter 'short circuit' paths. Adding a coil to that connecting flux path, putting the voltage through a voltage multiplier and full-wave rectifier and connecting that to the biasing plates in the water tank should provide the voltage needed, pulsed at twice the frequency that the magnetic flux is being pulsed at (one pulse when the magnetic field builds, one when it collapses). Thus we get the voltage on the biasing plates just when we need it (prior to the water being ripped apart by the resonating magnetic field) to stretch the O-H bonds to their limits, and (just after the magnetic field has collapsed) to help electrostatically separate the H and O so they can't recombine as easily.
If we were running at 15.812 MHz, that'd give us a pulse time of about 31.621 ns on the biasing plates. Unfortunately, I don't think we'll get to that. We might have a hard time even making 15.812 KHz. That's a pulse time of 31621 ns on the biasing plates.
Anyway, I also attempted to "narrow" the flux path just as it comes into proximity with the water tank, to intensify the flux. Not sure if that's how it works, or if the flux will leak out the sides of those tapers.
As for what types of coils to use for the Flux Steering Coils, I'm researching that... toroidal, poloidal, starship, Helmholtz, bifilar series, bifilar reverse, etc. I'm looking for a coil that has a very intense magnetic field in its center (where the flux short circuit path would be), with little to no flux leakage and a high resonance frequency.
There are other considerations... for instance, do I need to oppose the permanent magnet flux at the full strength of the permanent magnets, or only strong enough to make the flux shift to the long path? When the permanent magnet flux is going through the long flux path, if the coils have lower flux strength than the permanent magnets, will some of the permanent magnet flux still leak back through the short circuit flux paths, despite the coils?
{EDIT: According to Paul Noel, the coils need ~1/4th the power of the permanent magnet array to 'steer' the magnetic flux.
Directory:FPPMT:Paul Noel - PESWiki
So for a 4 Tesla permanent magnet array, we'd need the coils to produce 10,000 Gauss each. Is that doable?}
I've sent out some emails to people who have done extensive magnet and electromagnet research. If they reply, it'll fill in some rather large gaps in my knowledge.
And my brain is still chewing on that "metal ring with coils" thing... almost like a Tokamak, but I can't see how you'd 'steer' the magnetic flux through two flux paths and thereby be able to use the static permanent magnet flux in a pulsing manner using something like that.
I also thought of what would essentially be a 'water motor'... a rapidly spinning magnetic field by steering the permanent magnet flux around in a circle via the two flux paths, with the water tank in the center... but I'm not so sure that'd have the desired effect upon the water molecules. Might just make them spin instead of dissociate.
I settled upon laminated silicon steel as the flux path, for its low hysteresis and eddy current losses. After all the particulars are worked out, I'll order the metal, create a template and have a bunch of them punched out at a metal shop, then laminate them with clear varnish or the like.
Today I learned that MRI machines use anywhere from 1.5 to 7 Tesla magnetic field strengths. It's difficult to build a coil that can produce that field strength, but permanent magnets can already be bought that top 1 T, and using Parallel Path can get us up even higher. So we're using weaker electromagnets to get stronger permanent magnets to do work for us.
Something like this should work well, with 4 of them in a Parallel Path arrangement:
K&J Magnetics - Products
Now, what measurement is used to determine its strength? Surface Field, or brMax? They say brMax can't be used because it would require no gap from the face of the magnet, but that's exactly what we'll have, zero gap between the magnet and flux paths.
Those magnets are about $960 each if you order 4 of them. So $3840 total. I've already spent more than that on this poor little scooter in parts fabrication and having the new electronics built. I'll likely spend that much again in more modifications.
That's about 4900 pounds of pulling force for 4 magnets... so I'll have to be extremely careful with them. Not even sure how I'd secure them... it's going to take a bit of engineering to avoid sticking things together and never getting them apart again, not to mention avoiding getting various body parts crushed.
The good thing about this Parallel Path setup is once the magnets are in their holders and surrounded by the flux path metal, there should be very little stray magnetic attraction outside the machine with the coils deenergized. The magnets have a built-in 'keeper' in those magnetic flux 'short circuit' paths... so no chairs flying across the room and sticking to it. Heh.
I'm going to contact Joe Flynn and see what he has to say. Maybe he can give me a nugget of info that'll help to get this thing started.
But before I go ordering stuff willy-nilly, let's work out that the thing will at least work like it should, and optimize the design of it.
Comment