I would think in terms of adding a high input impedance buffer to the
tank, so you can use that output for both the positive feedback that
generates the negative resistance, and also feeds the rectifier, so
you can use that to measure the amplitude, without loading down the
tank. Then you need a comparator, with hysteresis (some positive
feedback), to make the drive versus coast decision, based on the
rectified and filtered tank amplitude signal. A second comparator and
some sort of voltage reference would be used to shut the drive down
when the supply voltage is too low. You might use a CMOS analog
switch to turn the negative resistance drive on and off, based on
these two control signals. This might all be distilled down to a few
transistors, but doing it with nice, clean functional blocks, first,
will probably get you a working unit, sooner.
As I understand your problem (correct my errors) you want to cause an
LC tank to oscillate. But you don't want it to be driven into
oscillation, continuously, but given a blast of energy only when the
oscillation has damped down to some minimum amplitude. I don't understand
if the blast is to be a single pulse, of a negative resistance that feeds
energy in, slowly, till the amplitude gets to a second, larger amplitude.
yes that is exactly it.
I have been trying to find a solution to this a over a week now. You have