This is the long reply that was delayed over 12h because the forum database was down.
Why is feedback needed for keeping frequency? its quite opposite, feedback is usually needed when you tune frequency to something.
If you have signal generator and set one frequency it will stay the same for ever
I am still confused here ...
Indeed. I also do not know which one of the following generator configurations is meant by the authors:
1) A signal generator forcefully driving a primary winding at a fixed independent frequency
2) The primary winding constituting an integral part in the generator's oscillator's circuit closed by the 1 turn feedback winding.
3) Two signal generators, blindly driving two primary windings at fixed frequencies (as in pt.1), are mutually locked in-phase with an aid of two separate 1 turn feedback windings.
In the beginning I thought the primaries were driven according to the scenario depicted in pt.1, since that scheme requires just 2 wires between one signal generator and one primary winding and that's what I had been seeing on the videos.
Also the following T-1000 statement seemed to support this scenario:
Each generator has feedback with 1 turn on coil and keeps own frequency no matter what changes inside of coil. This is where you get synchronization.
...but then I have read that there are some 1 turn feedback windings, a la pt.2.
This would mean that the primary is not driven blindly by the signal generator and participates in determining the frequency of the oscillation. In this case the oscillation frequency is most likely a
subharmonic of the natural resonance of
the complex LRC circuit, f=1/(LC)^0.5, formed by the inductance of the primary winding, its distributed inter-winding capacitance and the mutual inductance of the other primary and secondary windings as well as their inter-winding capacitances with deliberate capacitances added in the form of physical capacitors.
This scheme would require
at least 3 wires between one signal generator and a primary winding with the feedback winding and would have the desirable side effect of phase locking the two signal generators via the mutual inductance of the two primaries.
In the scenario of phase synchronization of two absolute signal generators described in pt.3, it should be mentioned that it can also be accomplished in other manner, such as supplying the arbitrary signal generators with a common digital clock source or dedicated phase-synch I/O built into the signal generators.
The following Terminator's (T-1000) words suggest that there indeed exists some kind of phase lock established between the Low Frequency (LF) primary winding and High Frequency (HF) primary winding:
The 50Hz is approx frequency because obviously we need standard frequency out of coil. You can tune +/- few Hertz and see where this frequency starts carrying frequency from coil of 51 turns.
Anyway, we replicators really must know if there is:
a) A phase-lock between the natural resonance waveform of the complex LRC circuit and the HF signal generator/oscillator.
b) A phase-lock between the natural resonance waveform of the complex LRC circuit and the LF signal generator/oscillator.
c) A phase lock between the HF and LF signal generators/oscillators, independent of pt.1 and pt.b
d) An integer frequency ratio between the HF and LF signal generators/oscillators, such that LO frequency < HF frequency.
e) An integer frequency ratio between the natural resonance frequency of the complex LRC circuit and the frequency of the LF signal generator/oscillator, such that f=1/(LC)^0.5 > LF frequency
f) An integer frequency ratio between the natural resonance frequency of the complex LRC circuit and the frequency of the HF signal generator/oscillator, such that f=1/(LC)^0.5 > HF frequency
It seems that the natural resonance frequency of the complex LRC circuit, f=1/(LC)^0.5, is the master waveform from which all the other waveforms are causaly derived.
If true, then this waveform relationship could be causaly and diagramatically depicted as
ComplexLRC.freq --> HF --> LF
meaning that:
HF is derived from ComplexLRC.freq and phase-locked to it.
LF is derived from HF and phase-locked to it.
where
ComplexLRC.freq > HF > LF
and
HF/LF = a natural number
and
ComplexLRC.freq / HF = a natural number
...only inventors or experimentation can confirm which of the above is true or false.