Giantkiller moves forward into Full Heterodyning.

[ramset]

GK what a find   WHAT A MIND  once again  the path seems familiar to you  Chet

[Sprocket]

Hi.  That Kunel patent is new to me and the basic setup looked too like an AM coil, (about as far as my coil-winding has got...) so I decided to give it a shot (see pic.)  Unfortunately, despite the suggested 10:1 power gain, I get nothing at all on the collector coil - I was at least hoping to see a voltage of some sort!  FWIW, it's fed a 5V squarewave (which is a bit wonky as coil resistance is too low) @ 500mA-1A. Wire is about 0.25mm and I used neos and normal iron.  Anyone try this and have something interesting happen?

[giantkiller]

Hey Sprocket,
Firstly, great try.
Secondly, I tried that combo too and got nothing. I had only 21 windings like in the patent. But the big misake is the loops are in parallel and not a continuous winding. Look again. Did you fire the pulse coil at resonance or the inductive collector?
You just saved me having to do alot of windings. I thought of that thinking alot of windings would work.
So I thought the concept would be where he shows a loop there would be a resonant coil. Things seemed hidden. I did multiple coils and that produced results.
And here is an ugly one: is the spacing between the coils dependent on the diameter of the iron with respect to the resonant frequency in operation with the operating parameters of the iron induced by the coils? Yes, Its a mobius loop. But that is why my build looks the way it does. I refuse to snip wires to match some moving target.

As I read that patent I noticed the progression in the diagrams. When some references didn't show up at the current diagram I went back to previous diagrams and just surmised. Then I decided that Kunel is progressively explaining the process. The straight and narrow try is not physical. The last 2 diags are loops. Try to connect the ends together with iron. That, at least, would match up to the ring diags. Remember, Deyo shows this too.
My first build of this posted was a closed loop and at resonance. Way good effects. I am going back to that. I had 2 fires and am regrouping from that. The controller took the hits because of a lack of adequate heat sink. Now I can draw 4 amps and no breakdown. But this is incorrect. All I have done is create a bigger window for prototyping. I need to get to closer resonance and the draw will lessen.

The descriptions and pictures are incorrect to be fully interpreted as a complete document of a complete device. Sorry. The patent office just doesn't know.
He also shows square waves as input and sine waves out. Can that be trusted? The circuit is over complicated but you can get an idea of what the intent was.

The bottom line is the disconnecting and reconnecting or squeezing the flux lines. That cannot be patented as that is a display of nature.
Cannonballs!

He basically got a patent without any correct divulgance. Mostly hypothetical. But I have no respect for complying with the patent law. Leedskalnin already open-sourced this. I am going to pursue this design and should I achieve what the intent here was then that will go to production. And nobody will stop this. Because the answer will lie here in OU. The green wire build is the foundation I am still continuing with. Sure beats wrapping the TPU type coils.

If I left anything out or something here is unclear then RFC.

--giantkiller.

[Sprocket]

GK, thanks, and I appreciate the in-depth response.  Yes, I noticed that the collector coils were in parallel - I also noticed there were an awful lot of them, so I latched onto the bit about them needing at least as many turns as the pulsed coil (so 1 big one - took the easiest route basically).  Also, resonance didn't enter into it, I just pulsed it in the low Hertz in the hope something would happen.  To kill an hour or so, I have also just tried feeding it with a slowly increasing sinewave in the range 1000-10000Hz, and also closing the iron-loop as you suggested - nothing interesting noted...

Truth be told, I will probably have trouble replicating your final working design (when you figure it out!) never mind working it out from first-principles.  I sometimes think that patents are formulated specifically to direct you away from a working device - how many FE patents have been 'back-engineered' successfully? - then again, I may just be paranoid

[giantkiller]

@Sprocket,
Welcome to OU. The greatest addiction on the planet!
This was passed onto me: 4904926.
And inside those are 4006401 & 4077001.
By now you should understand what these are talking about.

--giantkiller. I've requested to be buried with my coils come that fateful day.