i thought this thing is for kids to play but with some peculiarities involved ... voila
The concept test F;q

and the partial app

// the graph drops down at 10 V DC ---------------_^
the thing can be utilized to limit output from Solar Cells
(the TEST's aim is the empty batteries draining device's
CMOS logic regulator ...)
the peculiarity here was that if i added a V.ref Diode to T2 E
it also wanted 1M resistor from E1+ to be able to compute
(this is just amazing ... (don't have a scope so i have to rely on such))
• so the point is the Buck converter can start up at about 380mV
but to opt the futhers it requires a control logic ***

(stuped test (for osc startup point))
(!) 380 mV -- seems to be the oscillator independent low-bound limit
i don't get it **the BJT-s can operate at nano amps (10-9A)
why the oscillations do not start below (!)
(( 400(mV) / 50(Ω) = 8(mA) that drops fast below 1mA (1000µA) but** ))