Parametric excitation
I already do it with one impulse per cycle, but I want to do it with 2 impulses per cycle.
Alternating impulses.
I already tried to achieve this with the Radiant Half-bridge circuit, which didn't work. I learned a lot from it.

Now I have a new circuit that needs to be tested. See attached.

L1a and L1b are the 2 windings of the same bifilar coil.
L1a makes the negative impulse.
L1b makes the positive impulse.

S1 and S2 are the high voltage series mosfet switches, driven 180 degrees out of phase at 50% duty cycle.
I wondered if maybe 25% would also work, but then the L2 won't resonate properly (I think). need to try it.

BogIrish already has made a test with this, which looked promising.

Impulses enter from both ends of L2, making this even more interesting for a solenoid bifilar coil, as it is more symmetrical.

If a solonoid is used, then ideally  the secondary L3 coil will slide INSIDE the L2 coil , as this is what Tesla also did with his oscillators.

Parametric excitation /tuning

series resonant L2 is Parametrically excited into resonance.

First normal resonance is created by the mosfet switch that allows it to charge(switch=open) and discharge(switch is closed).

the parametric excitation happens, at the moment the mosfet switch is opened.

The L1 coil then becomes resonant, and produces the first half wave which is the impulse.

This impulse flows into the series resonant low impedance L2/C2.
In this way the impulse provides the parametric excitation.

This makes the resonant energy increase drastically.

We can tune it so that all the impulse energy is fully absorbed into the resonant energy. making it a perfect sine wave.

but we can also keep a part of the impulse energy available riding on top of the resonant wave.

This is beneficial, as now this wave can be used, to again induce parametric excitation of the resonance in the L3 coil.

L3 is tuned by parallel resonance.
The L3 tuning cap should be a lot smaller then the L2 tuning cap.

this way L3 can again be tuned into parametric excited resonance.
the sine wave of L2 plus the impulse on top of the L2 sine, are both used, to excite L3.

this should even work without grounding L3. the impulse of L2 should then not be visible on L3. leaving only a pure sine wave.

another possibility is to connect the L4 in series with L3.
L4 is the socalled extra coil.
L4 could be made resonant to the frequency of the impulses of L2.

this photo:

https://www.amazon.de/photos/share/ciZVKTiZhGdeJ7smfTaam8wwuMlzY2gPALFkQomuDj4

shows my work of 2019 where I tried to use two L2 primary coils, one on each side of L3, to do the same.
But now I intend to use a single L2 with. both alternating impulses on a single wave

Normally to increase resonant energy,
external energy is added at one of the zero energy point.

Like with a swing. when the swing is at its highest, and starts returning downward,
its speed is zero.
then a push is adding energy.

with resonance there are 4 points per cycle that are zero.
2x the voltage minimum and 2x the current minimum.

So the L2 gets increased energy by the mosfet switch at 2 points.

One charge by the power supply when the switch is open
the other discharge when the switch is closed.
is this actually  decreasing the energy?

but at the charging by the power supply, the impulse also adds energy!

But the L3 can be induced by L2 sine wave, by mutual coupling
plus the impulse.

trick still seems to use 2 alternating impulses per sine wave