Joule Thief

[MarkE]

I think that ridiculous claims made without support deserve whatever response they might engender. But perhaps you can tell me how asking for a demonstration of a claim is somehow "uncivil". Dave has shown his great propensity to misinterpret his "experiments" in the light of his own preconceived notions, which do not jibe with conventional understanding, and which are leading him astray. This has been pointed out to him, civilly and uncivilly, several times before and anyone can look back in the thread to see how he responds.
An experiment is done to test an hypothesis. A demonstration is done to illustrate a point. We are seeing demonstrations, not experiments, from Dave, and he is interpreting them according to his notions, not according to what the data reveal.

I felt that jumping down Dave's throat about the one sided capacitor charging gaffe was not necessary.  That's JMO.

[MarkE]

I make no claims as fact,, just my view.
The battery charges the cap but when the oscillator kicks in the bemf from the collapsing magnetic field charges the pos side of the cap and also depletes the neg side.
I am testing and will post results.
I dont really care if Im right or wrong, I want to find the answer.

Dave it's god that you want to find the right answer.

When the transistor is on the input power source:  capacitor alone or in parallel with the battery, builds the magnetic field until either the transistor turns off or the current builds up to the point that the I*R drop across the coil matches the available voltage and the current builds no further.  When the transistor switches off that stored energy causes voltage to build in reverse across the coil.  Where the secondary voltage was positive from top to bottom, as the transistor turns off the voltage on the bottom of the secondary swings positive with respect to the top of the coil.  When the voltage gets positive enough, the LED forward biases.  This closes the current loop and the energy stored in the magnetic field surrounding the transformer discharges through the LED.  Current stops flowing from the capacitor, capacitor/battery.

In order to get the energy to recharge the capacitor or battery, you have two basic choices:  Wind a third winding on the core and connect that in series with a Schottky diode in the same orientation as the LED.  Connect the cathode of the diode to the positive rail of the circuit and the free end of the coil to the negative rail.  Where the sum of the supply voltage plus the Schottky diode forward drop is less than the forward drop of the LED, the modified circuit recharges the capacitor, or capacitor/battery combination each cycle.  The other alternative is to arrange a second transistor switch in series with the top of the secondary, and connect a Schottky with the cathode on that top secondary connection and the Schottky diode anode to the negative rail.  When the transistors turn off the bottom of the secondary jumps up as before until the LED conducts, while the top of the secondary jumps down until the Schottky diode conducts.  Now the capacitor, or capacitor/battery combination complete the circuit and get recharged a bit each cycle.  You will also notice that the magnetizing current discharges faster than before.  Consequently the LED does not glow as brightly. 

[Dave45]

Dave it's god that you want to find the right answer.

There is only One God and thats with a capital G.
The rest were created and yes I hope and pray He gives us the answer.
dave

[NickZ]

  In the Joule Ringer circuit, the Hv output that can light even a 50 to 75 watt incandescent bulb, is also going back to the battery positive or to the negative rail.
But, this does not help to obtain a self runner, cap or no cap.

  The 10000uf cap on the previously mentioned circuit is what is making it look like a self runner, but it will still drain down in time, as it's not being charge at the same rate as is being discharged, once the battery is removed.
  After all these years, and experiment with the JT circuit, not one person has made an actual self runner.  One that is STILL Running on its own power, years later...
Hopefully this will change, soon.

[Dave45]

I concede, does not self run