Joule Thief

[freepow]

Hello  Crowclaw,  I have tried this particular circuit, works great, however the volts coming from 3xAA (used)
                        which is 3.265v go's through the square red LED then drops down to 2.028v after the LED,
                        then the super-cap charges with a maximum of 2.028v.

                        I have powered a JouleThief with 1xLED bright @ 6.70mA draw through a 4k resistor in
                        my JouleThief, and then I tried powering the JouleThief with the super-cap and the current
                        draw went up to a whopping 10mA ! through a larger 8k resistor, why is this ?

                        Does that mean I have to use probably 20k+ to get the same current draw as when using
                        the 1xAA battery ?

[crowclaw]

@freepow, If we apply Ohm's law to the figures you have given, we get: 6.70mA x 4K = 26.8v and 10mA x 8K = 80 volts!! as measured across the resistors. The resistors are in series with your JT circuit so together form the load across the source, and the source voltage will of course not be as calculated. Firstly you must check some values. We know also of course that a JT circuit produces high frequency, high voltage pulse spikes. The current draw is not smooth DC current however, but switches in sympathy with off on nature of the circuit at the operating frequency. Digital meters can and do give false readings under these conditions. What you could do, and this goes for any one experimenting with JT's is to place a decoupling capacitor across the JT's supply point, and after your series resistor. You need to experiment with values but observe polarity for electrolytic's or tant's. This addition will tend to smooth out the switching ripple. An analogue test meter may give more accurate results. If you want to adjust the current draw used a series rheostat which will enable to set the draw. Jeanna uses one, as does Bill I think. Finally a fully charged cell has a lower internal impedance and is capable of supplying a greater current draw. As a cell discharges it's internal impedance (resistance) rises and its current capability reduces as does it's voltage. The JT circuits can operate at very low voltages and current draw, hence "AA" batteries not capable of power a higher demand for some applications are thus ideal for JT circuits. Remember results also change dramatically depending on your source voltage which can be quite critical. Your JT design characteristics inc. coils, transistor gain and component values all play a part, which is reflected in the results you can expect. regards Merv

[crowclaw]

  Thanks very much Crowclaw !


      Can I have a solar cell with output of 70+ mA  and have the red LED working as a charge light and still
      have the green LED turn on when Cap is charged full ??

That's how I presented my circuit, with TR2 acting as a shunt across the solar panel. The current draw needed to fully charge your cap from the solar panel may destroy your red LED as this is in series and will attempt to deliver what ever current is required! A standard LED has a recommended safe forward current of approx 20mA and is not generally a good idea to exceed this value without a series resistor to limit the current. However, a solar panels output depends on the load connected to it, and of course the amount of light reaching it's surface... LED's are cheap enough to experiment with so give it a try. Connect your meter in series on mA's to monitor the charge current.
I didn't answer your question regarding resistor values, QUOTE " will 20K reduce the current to 4mA" the current reduces as you increase resistance in a linear fashion providing the voltage across the resistance remains the same. As I mentioned  previously, use a pott (rheostat) in series with the JT's,  + ve supply to adjust the draw. Hope all this helps.

[ibpointless2]

Ibpointless:

I am not sure what could be wrong, it sounds as if you did the mod correctly.  I had a lot of trouble lighting a cfl with any of my Fuji's for a while and when I finally got it to light, it was not all that bright.  I did much better with the floro tubes and the leds.

Can you post for us a good close-up photo or two of your Fuji board?  Maybe one of us here can see something wrong?  We will try to help you as much as we can.


Stprue:

Yes, I remember seeing that video a while back.  I never tried what he was demonstrating though.

Bill

@pirate
Thank you for your quick response! I figured it out when I was watching one of your videos, you removed a lot of components that I had not removed. I got it to work better but it seems lasersaber on youtube and you pirate I have gotten the most light and I was wondering if that was just a camera trick. I do believe that you did light up 400 leds but I think the camera makes the brightness more bright then what it really is in real life,but i could be wrong and I hope that i’m wrong! If you made any special mods to yours could you please post a pic of all the things you did I would greatly appreciate it. The only reason why I need my setup very bright is that i’m on the edge of achieving overunity, I already have the current setup running a lot longer on my batteries then before. Thank you! 

[kooler]

ibpointless

when you go buy a camera.. look for the one that says quick snap.. or quick flash.. they have a different transformer in them..
so they charge for flash faster..
there the ones that light cfl's and 100's of leds..
but if you are handy with small things you can wind the primany of the regular fuji with 10-11 turns instead of 5-6..

the better camera's cost about a dollar more than the regular ones..

hope this helps

robbie

ps.. i forgot to add if you want some real power don't strip the board.. you can put a small cap in place of the big one and solder the snap buttons together and wire your cfl to the trigger coil and neg of battery..
then you have a very white cfl..
..secret so don't tell anyone..