[Al Skierkiewicz]

Cody,
Don't stop thinking or learning and don't listen to any of us, find out for yourself.
Thanks to the other contributors of this thread for the article links. The more important of course, is the one by Soljačić. Now that I see the experimental apparatus it is easy to see the mechanics of what is taking place. I disagree with his method, however. He measures his input power at 400 watts (wall) and determines output power by the brillance of the light bulb. All things being equal I call that 15% max. I would much rather see electrical power measurements in terms of current and voltage at both input and output.
The way I see this experiment, a small loop is used to excite a self resonant and unloaded coil at 9.9 MHz. (As pointed out in the article, current at the ends of the coils is zero, but any ham radio antenna enthusiast could have pointed that out, but it has nothing to do with the experiment.) That coil ought to be able to produce a rather intense magneic field. Another self resonant coil in line with the magnetic field ought to also reinforce the first magnetic field and this reinforced field is then able to induce current flow in an adjacent conductor. However, as any of us can realize, any body in or near any of the coils will affect the self resonance of the coils, thereby affecting the coupling efficiency and the energy transfer. Likewise, ferrous material in the magnetic pathway will also alter significantly the energy transfer. What makes this experiment unique is that it is undertaken at a frequency which is conducive to self resonance in easily formed coils at room distances of 1 to two meters. I believe the distance used in the experiment allows for two variables to be satisfied. One is that the two coils have minimal interaction at self resonance while simultaneously achieving a reasonable power transfer due to proximity of the two coils. I believe Maxwell's equations should give a reasonable prediciton of the power transfer in this experiment. As to the device being non-radiating, I don't see how it would be possible to not have significant radiation and therefore produce both interference and possible exposure problems. If we make the rash assumption that the oscillator is at least 50% efficient, then it is easy to assume the remaing lost power is radiated. If all these assumptions are correct, I make that radiated power at about 50 watts.