Overunity motor, part3, all 4 recharging bats reading at 1.400 volts now.

[sm0ky2]

Very simple, is the set-up shown in this thread.

1) Motor + Battery = run time x.

2) Motor + Battery + recycling circuit and batteries = run time x, + run time y, + run time z, ...... etc.

and what do you think that means?

I think that means that some of the power that is not used by the motor, but is lost through the circuit, may be being "recycled"
We should investigate this, instead of blindly dismissing it.

Is that the only possible answer?
absolutely not, it could be the result of power distribution, leaking from the run-batteries, into the re-charging batteries.
If this is the case, it should effect the run-time

There are other possibilities as well. But when the entire concept is thrown out the window because of some preconceived notion that everything we waste "must be wasted", and recycling the power drawn through our admittedly inefficient circuitry is "impossible" we shut the door to such investigation.

[tinman]

Well... hmm.

So I built the LidMotor version from MarkE's redrawn diagram above. I wound two toroidal inductors to measure 1.0 mH each on my ProsKit meter; this required 32 turns of #33 on each toroid. I used a BC337-25 transistor as I do not have any MPSA06 on hand. A blue LED, a 1n4004 diode and a 220 ohm resistor completed the circuit. I used two depleted batteries for power instead of supercaps. The circuit needs to be "tickled" to get it started, and I found the easiest way is to tickle the cathode of the LED with a little piece of solder. The collector of the transistor also is a good place to "tickle" to start oscillation. I could not get it to stay on with a Red LED, just single flashes when tickled but no sustained oscillation. It works with Blue LED just fine. Have not tried other colors.

My impression is that the circuit does _NOT_ appear to work by coupling between the inductors! At least, moving or reorienting the L1 inductor appears to make no difference in behaviour of the circuit in terms of startup or LED brightness. I have not yet scoped the circuit.

(I'm still waiting for the "friend-funded" Rigol scope to arrive. Supposedly things have been delayed by the Longshoreman's strike on the West Coast container ports and it is not expected to get to me until the first week of April sometime.)

ETA: It still works with the L1 inductor 2 feet away connected by a twisted pair to the solder pads. Still needs to be tickled to start but once it starts, LED brightness, etc. is unchanged from the previous test.

TK
Remove the 220 ohm base resistor,and you have my original circuit.

[tinman]

I thought I saw a label on the tube that said "Aluminum".  I agree that with the circuit as represented, the LED will not light unless the transistor oscillates.  The but for the D1 diode, the LED would be reverse biased when L2 is not flying back. 

I constructed the circuit on a solderless breadboard using a 2N2222A transistor, 1N4005 diode, and OVLBR4C7 red LED.  I used two NiMH cells.  I used several choke configurations with the following results:

1) 1812 1mH 42 Ohm unshielded chokes 6" apart.  No oscillations.
2) 1mH 2.9 Ohm shielded choke L2 for the flyback, and 1mH 1812 42 Ohm choke L1 for the base-emitter.  No oscillations.
3) 1mH 2.9 Ohm shielded chokes both positions.  No oscillations.
4) 470uH x 2 coupled choke. 120 Ohm series base resistor.  Oscillates with coils oriented as in the graphic below, LED glows brightly, but the frequency wanders.
Peak collector to emitter voltage is just over 6V.  2.5V for the NiMH batteries + ~2V for the LED Vfw and ~0.8V for the 1N4005 and the rest is resistive drops in the choke and LED.

Mark
Remove the 120ohm base resistor-no need for it,as the inductor is already a resistor,and the circuit is already low powered.

[sm0ky2]

Go ahead, put up a timing diagram of a "Joule Thief in resonance"

MileHigh

you know, im really trying not to get too deep into this ridiculous argument here, because it's off-topic for this thread.
This should be, and has been many times, discussed in the JT threads.
but since our benevolent author also uses a JT circuit with recharging batteries in a similar manner as the video posted here,
i'll show you this.
When you send a pulsed DC signal through a transformer, there is a reluctance through the core, due to timing differences during charging of the core.
When this signal is at the resonant frequency (adjusted by the resistance through the transistor), the function becomes a purely resistive factor, and a clean waveform is produced, at maximum amplitude.
My lab was lost, and I don't have the tools to do this myself, so I dug up someone elses.

TK makes a great demonstration of this effect in his video, using a fairly accurate signal generator, and his O-scope.
https://www.youtube.com/watch?v=y9ZN5QJZClY

as you can see here, this greatly alters the effect of the induction through the secondary coil, which will increase the efficiency of your joule thief circuit. 
This is how a JT circuit was intended to be used. This is the effect described by Steven Mark.
The "toy" that has become so famous, makes no reference to this critical factor, and therefore, the quickie-circuits produced in the How-To instructionals are not resonant, and inherently inneficient.

[tinman]

This circuits opperation was found quite by accident. It started out as a simple pulse motor circuit where the trigger coil was sepperate from the run coil. I designed it like this so as i could adjust the trigger timeing. Anyway,i gave it a run with the rotor,and it worked quite fine. But when i stopped the rotor,the circuit continued to oscillate. So i removed the rotor altogether,and found i could move the trigger coil to any position,and the circuit would continue to oscillate-->and thus,the birth of the cool joule circuit.