Joule Thief

[Pirate88179]

@all

Ok my pnp npn , Darlington came from a old trash Sanyo dual Tape with ,those little tape motor are also real nice to play with there a bigger size and generate up to 6 volt when turned by hand( same time used in crank flash light)like i said you never know what you can find, when recycling yourself If you like this know that there are also motors in printer step motor giving me 12 ac at low speed.

http://www.brookstone.com/bs_assets/images/shop/large_300/544072_p.jpg

@ electricme

I was afraid of that, in my early test with many turn going around i got low results , it could be the overall design of the coil but with more test with transformer i now evaluate coil in resistance not on turn, The resistance/gauge in copper is marginal , so new material may give us more.
Or more resistance may just need more power to be effective and also the freq range.

@ MK1:

I found this on Wiki: (quoted below) http://en.wikipedia.org/wiki/Darlington_transistor


"A similar configuration but with transistors of opposite type (NPN and PNP) is the Sziklai pair, which sometimes called the "complementary Darlington."

You just gave me another crazy idea.  I like your idea of measuring the resistance in the coil as opposed to just counting the turns....so...what if we put a VR in the coil windings themselves and therefore allow us to "tune" the resistance in order to tweak the best settings?

Maybe someone already suggested this or is doing this, I am still not up to speed on this topic. (I have a few pages to go)  I am going back out into the field in a few minutes, I was just checking in.

would a VR of the correct range in-line with the coil windings, or on the core itself, help us do you think?

Bill

[jeanna]

Hi everybody,

So, as you heard yesterday I took my joule thief#1 which I recently upgraded to use a white light, and wrapped a secondary/pickup coil and attached another white led. The led on the 2dary lights but the basic light of the system never lights.

Then last night I made another joule thief in the same manner. It is different in 3 ways.
1- the torroid is called 'charcoal'
2- I wound 11 bifilar turns
3- the transistor is 3904
oh and also, I made the resistor in 3 parts adding up to 929 ohms and 2 are variable.

The light goes on fine as did the other one.

Then I wrapped 44 turns of red mag wire and soldered on a white led between the ends of the mag wire.
I confirm the action that the original joule thief light goes off when I add this secondary and the light.

Next, Iwound another 44 turns and soldered another white led.

both secondary lights shine brightly and still the original led stays off.

Then I put both circuits in || and all 3 lights shine brightly.

Then I put the 2 in series and the pair on the board with two lihts shine verrry dimly - and the jt with the single secondary does not shine at all.

This means to me there is plenty of 'fire' to act like amps but not enough for voltage on these ckts.

I will try to add some pics.
1-new ckt dark so it its parts can be seen.
2-new ckt with 2 secondary lights, and
3-new and old jt circuits in parallel showing the 3 leds of the secondaries.

jeanna

EDIT. Darn. I will be back to finish the pics.

[xee2]

@ jeanna

The LED attached to the collector gets its voltage from the collapsing magnetic field of the collector coil. With the secondary coil present it seems that the secondary coil is absorbing most of the magnetic field energy of the collapsing field and so there is not much energy left to generate a voltage spike to drive the LED on the collector. Mk1 must have had weaker coupling between the secondary coil and the collector coil. Try moving them farther apart or use less turns for your secondary.

EDIT: Before you modify it, check the open circuit voltage of the secondary coil.

[jeanna]

@xee,

I like the less turns idea. That way I should be able to end up with more secondary leds. I have run out of room in the center with 2 leds.

I should make another one with more distance.

The resistors didn't seem to make any difference that I could see in the brightness.

Thank you,

jeanna

[Mk1]

@jeanna

I don't see the diodes, remember that by connecting leds directly to pickup coil is working but a no no because when the current goes to led it is coming out at the other end and you send it back in the coil,reversing the ripples in the bucket killing efficiency .

@electricme

My 2 x48 turn pickup coil are only covering half of the toroid and back , each one have there half of the toroid.