Single AA battery to light WHITE LED for long-long time

[gyulasun]

@gyula

My friend's email to me (sample DC Pump Energy, Maybe from JLN Labs).  Concept only.
I want combining with our "LED" design.
Could u help me and share to us

zon

Hi Zon,

I combined your "LED" design with this latest circuit you showed. I hope it is how you wished?

Naudin made some test on such circuits  (of course without the LED circuit addition) in 1997 he called it TEP projects, see here:
http://jnaudin.free.fr/html/tep50a3.htm

He used a 12V battery and 100mA average current and his battery maintained its starting voltage for one day.  It is a pity he did not include some more days' tests... I guess his battery started to discharge...?

Gyula

[zon]

@gyula

Thanks for your circuit combining
How much value of R2 and R3  ?
I have try with 1K, but it's not working.
Sorry for my poor electronic background.

My friend sugest for TEP ver 50a3, http://jnaudin.free.fr/html/tep50a3.htm to take off C7 , it's would be better.

zon

[gyulasun]

@gyula

Thanks for your circuit combining
How much value of R2 and R3  ?
I have try with 1K, but it's not working.
Sorry for my poor electronic background.

My friend sugest for TEP ver 50a3, http://jnaudin.free.fr/html/tep50a3.htm to take off C7 , it's would be better.

zon

Hi zon,

The values of R2 and R3 depends on the transistor DC current gain parameter (hFE) you are using.  The best is to experiment with  220 kOhm trimmer potmeters in places of R2 and R3 (your 1 kOhm value is too low).  If you look at Naudin's schematics he used 470 kOhms but it was for the 12V battery and you use 1.2V battery. 
Here the choice/type of transistors narrows down to types of low collector-emitter saturation voltages and the best would be to look around your shopping possibilities to choose from.  Hint: see this Fairchild transistor tablet:
http://www.fairchildsemi.com/sitesearch/fsc.jsp?command=toggleSort&attr1=Small+Signal+Transistor%2F%2FSaturation+Voltage+Vcesat&attr2=0&q=50&qt=1207735589&vid=%24__visitId__%24&i=part_number&qid=%24__queryId__%24&s1=Product+Volume%2F%2F0&s0=Product+New%2F%2F0&g=Family&a1=Saturation+Voltage+Vcesat%2F%2Fv%3A0&a0=AAAFamily%2F%2Fv%3A0&as=1&qtid=%24__lastTextQueryId__%24&t=0&ia=1&c2=n3%3A2%2F%2FSaturation+Voltage+Vcesat%2F%2F0..0%2F%2F1..0&c1=e%3A12%2F%2Fpart_type%2F%2F%3Aeq%2F%2Fg&c0=e%3A0%2F%2FAAAFamily%2F%2F%3Aeq%2F%2FSmall+Signal+Transistor&domains=fairchild_db%3Apart_number&text=

It lists the transistors by saturation voltages, starting by the lowest saturation types (so the best in your application).  There is the FJC690 transistor with 80mV saturation voltage at 100mA collector current, this type is developed for camera flash circuits ( http://www.fairchildsemi.com/ds/FJ/FJC690.pdf ) and its 45V max collector emitter breakdown voltage hopefully will be enough for the inductive peaks of your coils if you always run your circuit under a load i.e. with LEDs on the 3rd coil etc.
Of course it is possible this type is not readily avaiable at your local electric components shop but probably you can select from the choices there for the minimum saturation voltage (0.2 - 0.3V is just the bearable max value as a trade-off).  The higher the saturation voltage the worse the circuit efficiency becomes due to the dissipation on the transistor. If you use a transistor with 0.3V saturation voltage, the effect is like you were using a 0.9V battery (1.2V-0.3V=0.9V).

Re on the C7:  in an earlier post above I answered my views on caps in parallel with the battery. It is possible that there are circuits and its components in which it is not needed, experiment can only tell.  Nevertheless, a bank of good capacitors of 22nF 47nF 100nF 220nF values are always a good thing to have in your junk box!

rgds, Gyula

[notsure]

Hey guys and gals.
Just got here yesterday and have been thumbing thru the pages. My main objective is researching fuel efficiency devices/tchniques but I cam across this and found it pretty awesome. I put headlights in my daughters power wheels jeep powered by a big 6v flashlight battery. Does not last long at all. I wanted to switch to a cluster of leds in each headlight and was wondering if anyone could clue me in with some simple direction on how I would go about it using this circuit? ........ Would this be a good experiment for her powerwheels batt and motor as well. If I could extend the batt life between charges that would be awesome. Just curious...
Notsure

[gyulasun]

Hey guys and gals.
Just got here yesterday and have been thumbing thru the pages. My main objective is researching fuel efficiency devices/tchniques but I cam across this and found it pretty awesome. I put headlights in my daughters power wheels jeep powered by a big 6v flashlight battery. Does not last long at all. I wanted to switch to a cluster of leds in each headlight and was wondering if anyone could clue me in with some simple direction on how I would go about it using this circuit? ........ Would this be a good experiment for her powerwheels batt and motor as well. If I could extend the batt life between charges that would be awesome. Just curious...
Notsure

Hi Notsure,

The best powersavings come from energy not consumed. The trick with LEDs (or normal lamps) is our eye does not senses when we switch on and off a LED quickly enough. So if you supply your LEDs with current pulses of appropiate strength instead of normal continuous DC current (in fact you chop up the continuous flow), then your average current consumption will be less,  with nearly the same LED brigthness sensation.
And if you could combine the pulsed current supply with inductive circuits in which the collapsing magnetic field at pulse switch-off induces a socalled flyback pulse across the coil and you make sure this flyback pulse is regained, then this is another possibility for increasing efficiency of your circuit i.e. reduce input energy consumption. 
A good example for the combination of both abovementioned tricks is the joule-thief type circuits.

There are other circuits whereby you quickly charge up a capacitor to a given voltage level under a half periode time of a full period and connect this capacitor in series with the main supply voltage under the other half periode to get quasi the double supply voltage to feed a LED, this solution also gives a reduced average current consumption.  Here you can see circuits applying this principle:  http://www.discovercircuits.com/PDF-FILES/1vled3.pdf or http://www.imagineeringezine.com/PDF-FILES/ac14fls.pdf 

Your specific problem you mentioned is partly solved by this circuit: http://ourworld.compuserve.com/homepages/Bill_Bowden/555.htm#555leds.gif  where the NE555 timer gives out 4.7Hz pulses to drive transistor that switch LED arrays for flashing bicycle lights.  This circuit can only achieve powersawings by the pulsed switching of the LEDs on and off I mentioned first above ( the frequency and the duty cycle of the pulses should be experimented with),  and has no coil with collapsing field to regain some energy from the input energy. 

Further useful links with very good explanations: http://cappels.org/dproj/ledpage/leddrv.htm or http://cappels.org/dproj/PulseBoostLED/Pulse_Boost_White_LED.html or http://www.talkingelectronics.com/projects/LEDTorchCircuits/LEDTorchCircuits-P1.html

Armed with these hints you now can start experimenting with simple, low power circuits first then attack your powerhungry headlights with the knowledge gained.   

rgds, Gyula