Gating MOSFET with an antenna (adjust voltage to use only peak)

Question

If we pulse a coil/tank circuit it's supposed to ring at it's natural frequency. We could use a pickup coil over a shared ferrite core or a straight antenna, connected to the gate of a MOSFET to time the pulsing signal and remain at resonance autonomously. Problem with that idea is is that the voltage necessary to open the gate could be reached far from peak or not at all.

So how (if possible) could we normalize picked up voltage to reach gating voltage only in the vicinity of the sinus peak?

Edit

Per @Andy Aka’s response, looked up for opAmps and peak detectors, I figured subtracting the peak detector Voltage from the wave and rectifying would show some voltage only around 10V from peak. on this diagram the pulse generator has a very small duty cycle (intended to be the 1st pulse only, should be smaller freq as well).
the resonant circuit is the tank LC, the rest ys gating logic: diode + 10Ohm resistor + small capacitor for peak detection. OpAmp for subtracting max V detected (with a 10V drop) from Tank's wave.

Does this make sense? (need more practice with the tool to make the simulation work.)

EDIT

Found this implementation: https://youtu.be/wWIeUsnqkRk?t=732

Answer 1

So how (if possible) could we normalize picked up voltage to reach gating voltage only in the vicinity of the sinus peak?

You would use a variable gain amplifier and a pulse shaping circuit.

We could use a pickup coil over a shared ferrite core or a straight antenna, connected to the gate of a MOSFET to time the pulsing signal and remain at resonance autonomously.

No need to use a pick-up coil; feedback some of the output voltage into the variable gain amplifier (just as you would in any sinewave oscillator where the frequency is defined by the tuned circuit).

Look up some metal detection schematics to see how they do this. I remember at least one that used this method (a sync pulse).

Answer 2

This is one potential advantage of BJTs over MOSFETs: biasing can be easier at times.

We could construct a fairly simple circuit with a similar behavior. For example, using a MOSFET with R_ds(on) < V_gs(th) / I_d(pk), we could wire a diode from gate to drain (also known as a Baker clamp, at least when the diode is schottky, and applied similarly to a BJT), so that when drain voltage pulls down (approaching saturation), current is drawn out of the gate circuit. Then we can apply a small pull-up bias current, and the transistor will turn on automatically at zero signal, developing gain, and as signal grows, pass into class C mode where it's in cutoff much of the cycle except during the peaks, when just a little current is drawn (from the gate circuit) to keep it in that state.