3v OU Flashlight

[semenihin-77]

...

~A

Очень похоже на то что надо, только надо емкость подобрать по максимальному пику и контролировать потребление.

Very similar to what is necessary, just need to pick up the capacitor at the maximum peak and control consumption.

[MenofFather]

Амплитуда синуса зависит от частоты, а частота от емкости. Подбирается конденсатор на максимальную амплитуду резонанса, это будет самая эффективная частота. Я пробовал от 10микрофарада  до 1нанофарада.....остальное пока не могу сказать.
Я пробую 2 последовательных CREE светодиода.

О какой ёмкости говориться, которая между 3 и 4 ножкой у 34063 микросхемы или о какой?

[Vortex1]

What if you keep the squeezing force constant and superglue two large masses to the core laid horizontally?
(a heavy glass ashtray for exampe, or tea cups ...or anything else non-magnetic and non-conductive)

Maybe build a 4 sided structure from glass plates with the pot core halves glued to two inside surfaces with a small gap. The tiny amount of flexing should be tunable, as also the gap. A cutoff section extruded aluminum channel might work, but for really hi Q, maybe construct from hi carbon steel with magnetically insulating risers.

I'm thinking of gluing the core halves to the inside poles of a U shaped piece of non-metallic hi Q material, open portions of the cup cores facing each other, gap adjustable. This could make for a dandy parametric inductor.
------------------------------------------------------------------------------------------------------------------------
BTW, the above posted schematic (circuit.png)makes a bit more sense than any of the schematics I've seen, the transistor VT1 acting as a synchronous rectifier switch to utilize the hi Q ringing of the tank circuit and add to the 470uF.

As shown, with a low frequency drive for the transistor, and un-synced to the main oscillator, it is hit or miss rectifier, hoping to  close on the peak of a positive excursion of the ringing frequency and open at the peak of the last positive, it will add 1/2 cycle of the ringing energy to the 470uF storage cap. It will zero to a null, operating randomly unless the LF oscillator can lock to a subharmonic of the main oscillator.

[avalon]

Очень похоже на то что надо, только надо емкость подобрать по максимальному пику и контролировать потребление.

Very similar to what is necessary, just need to pick up the capacitor at the maximum peak and control consumption.

Latest: 5 mA @ 2.6V, 68V ~ secondary (21V~ before)
~A

[verpies]

I'm thinking of gluing the core halves to the inside poles of a U shaped piece of non-metallic hi Q material, open portions of the cup cores facing each other, gap adjustable. This could make for a dandy parametric inductor.

It would.
...and if you had a short HF low current winding there, its inductance would reflect the instantaneous width of that gap. A very useful information. 
The MC34063 can extract this info as an easily integrateable pulse widths, when operating in constant current mode with constant load.

What is your idea how to utilize the kinetic energy gained by the core halves?