To be deleted...

[nul-points]

Bonne année, overunity people!

Got to laugh at the complexity and/or 'scientific' foundation of recent flashlight/lantern offerings from points East

For all you frustrated replicators/hopeful customers, here's a little seasonal present to keep you amused while you wait for something/anything worthwhile to come of yet more fruitless effort, or waste of $100

So - this is my offering in the world of flashlights:  use flyback energy to charge a second battery via LED(s)

Ok, nothing novel, there - except....

Here's a little demonstrator i put together this weekend - and as a 'starter for 10', it's not looking too shabby!

This example uses only 1 AAA 750mAh NiMH as an input 'battery', a similar cell as output, and a single Hi-brite white LED as the lamp - the same type of circuit arrangement could be used for additional cells and LEDs

The circuit is based on a flyback switchmode PSU arrangement  (in this case, a Boost Converter), B2 gets charged with the same flyback current pulses illuminating LED1.

After operating the flashlight sufficiently to discharge B1, the switch S1 can be toggled to swap the i/p & o/p cells/batteries and the process repeats.

Obviously, this arrangement will extend the use of the flashlight, compared to the same initial charge supplied to a single cell without the recharge circuitry

If you wish to experiment with this configuration, you'll see that my circuit is pretty much generic - the transistors i've used have been either high-gain, low power, eg BC337, or medium power, eg BFY51.  Transformer T1 needs sufficient turns ratio to swtch Q1 reliably, my T1 is a Maplin Ferrite Toroid, approx 30mm OD x 25mm high, using 0.45mm magnet wire.  No pulse timing, this thing oscillates at the natural frequency of the assembled parts (approx 500kHz, in my case). Inductor L1 (approx 2.5mH here) may not be necessary, it's a legacy from other stuff i've been doing,  C1 is helping to buffer the input supply, i use a 1500uF electrolytic.  Diodes are Schottky, type BAT42,

I don't think any components are critical or unusual - i have introduced an air-gap into the toroid, mostly as a salute to the late Harold Aspden!  ;-)

I'm hoping to attach the schematic and a sample datalog graph of the discharge/charging voltage trends - i was going to take a photo of my scrappy build, to give an idea of the LED intensity, but i just found (ironically) that both my batteries for my camera are discharged so that will have to wait.  The LED is bright enough to block visual sight of the layout when viewed from approx half a metre above.

PS ...although the graph may LOOK as if the flashlight is already OU, with the charging cell voltage slope being steeper than the discharging cell slope, obviously further (and certainly more rigourous) testing would be required to establish the efficiency of the system

However - this system works - its cheap, simple ...and it's likely to be a whole lot more efficient than some of the offerings being touted as OU!

Enjoy  :-)

np

http://docsfreelunch.blogspot.co.uk/

[SkyWatcher123]

Hi nul-points, thanks for sharing.
I'm working on my own version right now.
Mine will use a gutted filament type led bulb, 3.6 watt, 450 lumen type.
Using flyback ferrite core.
I have around bifilar 220 turn primary oscillator, 6 layers, using 24awg. magnet wire.
Using 3.7 volt lithium ion cell.
Hopefully the 220 turn primary will give some good high voltage spikes to light the 120 volt led bulb to decent brightness and charge the second cell good, we shall see.
peace love light
Edit: oh and for now, i will just manually swap lithium ion cells, until i get proper switches.

[nul-points]

...
I'm working on my own version right now.
Mine will use a gutted filament type led bulb, 3.6 watt, 450 lumen type.
Using flyback ferrite core.
I have around bifilar 220 turn primary oscillator, 6 layers, using 24awg. magnet wire.
Using 3.7 volt lithium ion cell.
Hopefully the 220 turn primary will give some good high voltage spikes to light the 120 volt led bulb to decent brightness and charge the second cell good...
i will just manually swap lithium ion cells, until i get proper switches.

Hey Mr T

that sounds like a cool build - and i know that it will be a good quality one!

...hmmm - that's just reminded me that i have an AA single-cell-powered JT-driven lamp (using a 220V LED 50mm halogen spot replacement bulb) kindly given to me by old OU buddy Nerzh Dishual, when we exchanged project builds a few years back...  ;-)

i look forward to hearing the results you get with your project - i've had good indications already on mine, even with this minimalist version

How do Li-ion batteries interface to flyback spikes?  I've only ever tried NiMH, NiCad & SLA, all of which have been fine at low charging rates in earlier unrelated projects

Keep us posted!
np

[SkyWatcher123]

Hi nul-points, I'm having good indications with mine as well.
However, had to use 2 warm white top hat leds in series, not enough voltage for the 120 volt led bulb.
I think a secondary coil with more turns and then use the 120 volt led bulb as diode to charge lithium cell might work.
Not sure how well that would charge the single lithium ion cell, considering the voltage will be higher, whereas the 2 leds in series probably drops the voltage down from the primary flyback.
I also have mine, for now, with the 2 leds off the collector into positive of charge cell and negative of cell to common of circuit.
The voltages seem to be tracking very closely, as far as the loss on input and gain on charging cell.
The lithium cells seem to take the charge fine, of course they are larger capacity and take about 2 hours at the 20 milliamp input to go from 3.93 to 3.94 volts.
peace love light

[TinselKoala]

Nice work, Doc! The graph is particularly good to see. We are so rarely given real data to chew on. Keep up the good work!