The system i'm building Matt has an oscillator which is the audio transformer and i'm passing the signal through a stand alone choke. The choke is similar to an high pass filter and i'm asking the choke to cut off current that is of higher frequency/higher impedance than the oscillator, the diode half rectifies the signal from the transformer into an higher impedance, higher frequency signal which the choke absorbs as induction so it's perfectly understandable exactly what Stan was trying to do.
But here lies an important factor: The current in the audio transformer lags the voltage by 90 degrees and the voltage has already begun to influence the choke before the current magnetises the choke core which is what we want it to do. Then the signal is terminated by the audio transformer and therefore the choke is completely isolated from the input signal and decoupled from any primary. At this stage you've already won because you've only used primary current to magnetise two cores and nothing else, there is still no power factor involved but it doesn't finish there. The chokes magnetic field will still collapse in the opposing direction from which they were charged which is against the bias of the diode so current isn't a factor but we still need to get the voltage from the choke into the series circuit and the WFC. The only direction is in the bias of the diode back through the secondary of the audio transformer and to do that you'd need to match the impedance of the choke with the cell. But here lies a great observation by Tesla and others: Even though the bias diode blocked the current from the choke reaching any primaries and therefore we have absolute decoupling going on, the choke can be made to resonate in an LC circuit and offload its voltage but a totally new phenomenon appears, the bias diode doubles the LC resonance yet again.
Conclusions to this: If the audio transformer is running at 5khz then the high impedance, high frequency signal going into the choke is 10Khz because of the half rectification. Stan calls this a frequency doubler. The choke is inductive and supresses the 10khz current from the transformer then tries to collapse in the opposite direction from which it was charged but the bias diode stops this, but then the energised choke forms an LC circuit with the cell BUT the diode doubles this frequency and that is why no calculations ever work out. The LC circuit when it is built MUST be resonant at TWICE the self resonance of the choke not it's normal 10Khz self resonant frequency. The path through the secondary and the diode must be capable of resonating at 20Khz and this is why transformers cores are important, some people will have built the secondary core to take 10khz when in fact it needs to take 20khz.
Another piece of the jigsaw is now in place. Tesla was a genius.