Semi-Self Charging Joule Thief ( or oscillator )

This is another circuit that I am testing that have the self-charging capabilities. Maybe another one that is almost self-charging.
And a question: Joule Thief is a concept or it defines a device that uses a toroid?

[mrdovie]

One way to test if the battery is charging would be to run the charge down for a while, unhook the battery and measure its Voltage, hook the battery up again so it will charge (with the diode) let it charge, then unhook the battery and measure its Voltage to see if it is more or not.

[SoheiThoth]

You are right... the white LEDs I use needs 3.4v and 20mA to reach the same brightness. But see how amazing these oscillators circuits are... at 400mV I can make them very bright. And the meter shows the voltage on the battery.

[TvRepairHD]

Great work the 1n5819 has much lower break down voltage at .175 or so. Very high frequency operation

[JSuparman]

@sohei: 400mV and still bright? Please note, Voltmeter reads "average" voltage that in real varies between 0 to around 3 Volt (=conducting WHITE LED voltage), but not at 100 duty cycle. This can be seen on oscilloscope,
Remember this: White LED will not lite at all even connected directly to fresh 1.55Volt battery. Please someone correct me if I wrong. Thanks.

[acmefixer]

When the current drops, it gives the battery a chance to recover, and the battery voltage starts to rise. It will do the same thing if you disconnect the Joule Thief. But when the circuit is running and the voltage continues to rise, it will eventually level off and the start to drop down to maybe a half volt, until the cell is no longer capable of powering the JT.

[acmefixer]

About your Q A Joule Thief is a simple one transistor device that uses the inductive kick of a coil to boost a low voltage up to the voltage needed to light the LED. The coil can be a toroid, a bobbin, bar or rod of Ferrite, or it can be a coil of wire with an air core, or any other non-magnetic core such as a piece of wood or plastic. The conventional Joule Thief uses a 1.5V battery, a single transistor, a single resistor, and a coil and LED.

[SoheiThoth]

And the best part of using base->emitter is that I can connect the diode on the collector. About the solar panels... I already lost a shipment of small voltage ones, I am trying to receive another shipment.... And also some supercaps that I am dying to test :))

[SoheiThoth]

Greetings. This one just decrease the amp draw of the circuit and this make the battery increase the voltage (recovery). No, not this time yet. But the closedloop is getting very interesting when talking about self-charging.

[Slider]

Yeah, looks like recovery due to the lower load on the battery when the diode is in place. You could try the LED's to Collector and Base and compare amperage draw (I prefer it to the general normal Collector and Emitter) as the LED's are more directly over the coil collapse voltage. Increasing your tank circuit resistor from 10K would lower amperage used but also lower the LED brightness, depends what compromise you want, light or run time. Throw a small solar panel on and you may get 24/7 :)

[acmefixer]

I looked closer at the schematic. The voltage at the right side of the 1N5819 has to overcome the 0.4V of the diode and the battery voltage, which totals over 1 volt, maybe as high as 1.9V. But the voltage starts to rise and when the base to emitter junction starts to conduct, the voltage can't go any higher than 0.6 to 0.8 V. So the diode doesn't conduct. I would say that the right side of the 1N5819 should be connected to the collector which as 3.3 V. (cont'd)

[alif.t]

Hi Dear Sohei
I rebuilt this circuit but led did not run !!!! Please explain about inductors and voltages of capacitors. Best regards

[SoheiThoth]

I´ve seen the name of those air coil devices that I use as blocking oscillators, and not as a JT. That is why I asked, maybe I was using the wrong name using Joule Thief. Or when I use the inductors for the coils.

[oudotcom]

Well done Sohei!
Just let it run and see where it settles!
Regards, Stefan.

[SoheiThoth]

But one thing is for shure.... That was the brightest led under 1 volt I ever made. Forgeting about this diode... using the JT, I am testing it with less than 400mV and it is still very bright light.

[SoheiThoth]

Thanks Stefan.. Great job with that carbon/alu batt :) Here carbon fiber is too expensive to create batts.

[MongrelShark]

You are semi charging. There is charge going to battery. But unless you can run it for a time, and get more volts after run than before, its not as much charge as load. What is interesting though, is that you can excite the chemistry in some batterys. So they stay in a recovery state. I have had very flat batterys run for weeks, some people claim years. Even if battery goes over start voltage when running, it's not necessarily a charge. Need to finish higher than start. 3 hours rest to read V.

[acmefixer]

The two separate inductors are really not separate because they are coupled magnetically. The windings of a toroid are coupled magnetically, not electrically, so both are effectively the same. But I do not consider the two inductor JT to be a circuit that is ready to go into production. You will not see this odd way of coupling two windings in commercial equipment. If it was to be used commercially, the coils would have to be mounted solidly next to each other and glued in place.

[acmefixer]

How can you desulfate a battery that doesn't have any sulfate??

[SoheiThoth]

The interesting is that with the meter leads across the schottky there is voltage. I posted this crazy thing to get more brains thinking, and maybe someone would say... give it up... that does not work.

[esesoz]

it is really important to show the battery volts before starting, also look at my video with same effect.