Donald Smith Device self looping no cap no batteries bright Lights !

[poynt99]

Base-Emitter junction of Q1-->Base-Collector junction of Q2 (is not reverse-biased like it normally would be with a positive voltage at the collector).

So the FG output is easily driving the coils to some degree.

[TinselKoala]

How does the signal generator manage to turn the first transistor on, with only one uA of current?

ETA: heh, crossed posts with .99, same intent, questioning the power supplied by the signal source....
great minds do think alike

[poynt99]

How does the signal generator manage to turn the first transistor on, with only one uA of current?

How do you compute 1uA?
ETA: correction:

(5V - 1.4V)/100 + 100 + 50 Ohms = 14.4mA


ETA2:

If it is not a standard function generator, and perhaps a DIY oscillator without a 50 Ohm output resistance (I think he posted that this is the case), then the available current to drive the coils is even greater:

(5 - 1.4)/100 + 100 = 18mA (peak), and an average current of 9mA.

[TinselKoala]

How do you compute 1uA?
ETA: correction:

(5V - 1.4V)/100 + 100 + 50 Ohms = 14.4mA


ETA2:

If it is not a standard function generator, and perhaps a DIY oscillator without a 50 Ohm output resistance (I think he posted that this is the case), then the available current to drive the coils is even greater:

(5 - 1.4)/100 + 100 = 18mA (peak), and an average current of 9mA.

Not me... the conjecture was made above that the signal source might be an ultralow power chip that only draws 1 uA or something like that.... and I'm saying that this could not drive the circuit directly with that low current input to the oscillator. The TS3001 on its demo board, I think it was.

[e2matrix]

I do believe the device is started initially with a 12 volt battery but once running the battery is disconnected.  I may be wrong but I have to assume it gets a jump start from somewhere.  Have you seen the video?