Big Joule Theif

[Watson]

@Watson---It sounds like you have much experience and knowledge in electronics. Slayer's Big Joule Thief project was a bunch of guys pooling their thoughts and experimental results. Xenomorph contributed the 555 timer part of pulsing the circuit to reduce the energy comsumption. The idea being that if we blinked the whole circuit (and light) only fast enough to make it appear constantly on, then we could save power.
The whole project spun off a simple joule thief circuit with people adding things here and there. What we were after was bang for the buck---BIG light for the least amount of energy.
Your ideas seem like there are sound and it would be interesting if you made one of these things with your suggested modificatons and posted the results.
There is always room for improvement.

Lidmotor

I've built quite a few JTs none with CFLs (see my blog below). Long ago I built inverters with small fluorescent lamps, run off D cells. They wroked okay. Earlier, I built 12V to 50V DC-DC inverters to power audio power amplifiers, up to 200 watts. I used 12 gauge house wire for the primary.

If I was to build one, I would need some specs to design with:

Supply voltage, xxV +-yy%, or from xxV to yyV and current or if it's gonna be powered from batteries.

Need to use a toroid (low RFI) or okay to use a simpler ferrite rod?

Type and number of lamps or LEDs.

Light output from the lamp(s)

One possible solution is to build a DC-DC converter, say 12VDC to 165VDC. This will allow you to use a CFL, since most of those rectify the incoming AC into 165VDC the first thing after it enters the light.

Probably forgot some other necessities, lemme know.

[Lidmotor]

Halo Light circuit improvements

Could someone helpme with understanding this schematic?

What does the pulsing add to the circuit?

Some of my observations

The 1k resistor controls the base bias to the JT. If you want it to draw less current, increase the resistor.

If the right hand side of the 1k is shorted to the emitter of the right 2N3055, the JT should stop oscillating. I think that if the left 2N3055 is used to short it, the JT should be turned on and off. A much cheaper transistor such as a 2N4401 could do the job.

The charging battery, in addition to accepting charge, is limiting the collector voltage to the supply voltage plus the charging battery voltage. No voltage is shown for the battery, but if both the supply and battery were 6V, then the max collector voltage would be a bit more than 12V.

This charging battery could be replaced by a zener diode, then the zener voltage could be increased (up to the max collector voltage but not higher).

The ratio of the two windings to the transistor is 1:1. If the voltage at the collector is 6V, then the voltage on the base will also be 6V, which is the max for the transistor's base. If the supply voltage goes higher, then the base voltage will exceed the maximum, and may damage the transistor.

The way to reduce this base voltage is to reduce the number of turns. Instead of 1:1, the winding going to the base would have half as many turns.

I hope this makes sense. Let me know if it doesn't.

@Watson---It sounds like you have much experience and knowledge in electronics. Slayer's Big Joule Thief project was a bunch of guys pooling their thoughts and experimental results. Xenomorph contributed the 555 timer part of pulsing the circuit to reduce the energy comsumption. The idea being that if we blinked the whole circuit (and light) only fast enough to make it appear constantly on, then we could save power.
The whole project spun off a simple joule thief circuit with people adding things here and there. What we were after was bang for the buck---BIG light for the least amount of energy.
Your ideas seem like there are sound and it would be interesting if you made one of these things with your suggested modificatons and posted the results.
There is always room for improvement.

Lidmotor

[Watson]

Hi lionhart, we and Lid have had the best success with the Halo light circuit, not sure what your doing wrong with slayers my friend so cannot comment.Just for reference here is Lid's Halo light circuit

Imageshack - 555jtwithswitches

Could someone helpme with understanding this schematic?

What does the pulsing add to the circuit?

Some of my observations

The 1k resistor controls the base bias to the JT. If you want it to draw less current, increase the resistor.

If the right hand side of the 1k is shorted to the emitter of the right 2N3055, the JT should stop oscillating. I think that if the left 2N3055 is used to short it, the JT should be turned on and off. A much cheaper transistor such as a 2N4401 could do the job.

The charging battery, in addition to accepting charge, is limiting the collector voltage to the supply voltage plus the charging battery voltage. No voltage is shown for the battery, but if both the supply and battery were 6V, then the max collector voltage would be a bit more than 12V.

This charging battery could be replaced by a zener diode, then the zener voltage could be increased (up to the max collector voltage but not higher).

The ratio of the two windings to the transistor is 1:1. If the voltage at the collector is 6V, then the voltage on the base will also be 6V, which is the max for the transistor's base. If the supply voltage goes higher, then the base voltage will exceed the maximum, and may damage the transistor.

The way to reduce this base voltage is to reduce the number of turns. Instead of 1:1, the winding going to the base would have half as many turns.

I hope this makes sense. Let me know if it doesn't.

[ashtweth]

Hi lionhart, we and Lid have had the best success with the Halo light circuit, not sure what your doing wrong with slayers my friend so cannot comment.Just for reference here is Lid's Halo light circuit

Imageshack - 555jtwithswitches

[lionhart]

big JT

Hi guys,

I have built several of these over the last 2 months, and was looking on you tube and found lid's big JT.

So I decided to get the components and build one, but I can't seem to get the amps down, I'm currently on 500ma at 14.4V using 4 AA bateries, but if I up it to 1.1A I get 120V. I would like to charge 12V batteries using a solar panel, not light lamps and need to get the current down.

I'm using the dual transistor circuit from slayer007.

Any ideas?

[JohnnBlade]

Hi All,

i got it to work, lighting up the CFL but using the disposeable camera circuit

here is the vid when it went on for the 1e time
YouTube - Lighting up CFL fluorescent light with disposeable camera with1 5v AA batery part 1 JB

but damn, i got my shocks even without cap while testing this


Greets JB

[morpher44]

Would someone please tell me how your 2N3055 power transistors fail?

If you go out of spec, pass too much current thru from Collector to Emitter, they will overheat -- like any transistor.
Also, if you back EMF is going from Emitter to Collector, you have
a breakdown current that can ruin it as well.

I'm using an ice-pack from my freezer -- which helps immensely.

[morpher44]

honeycomb on torroid?

With a small compact coil, especially with a ferrite core, the length of wire needed for a certain inductance is less, and the resistance is lower. It takes dozens of feet of wire to make a 100uH coil on an air core, even more for a honeycomb coil, and the resistance might be more than a dozen ohms.

I believe there is a technique to wrap a honeycomb coil around a ferrite toroid. I'm searching for that. It involves criss-crossing the wires in some way.
It would be nice to have a picture.

The advantage to this sort of coil is a higher Q. Higher Q means
at resonance, there are less losses.
IF, and this is speculation, what is being "tuned" here is a frequency
of interest that has a strong signal, such as the
Earth's field NMR or the Earth's EMR/PMR activity (or some other strong natural carrier wave out there), the resonant frequency might be very SHARP and so you want the bandwidth to be narrow / Q to be high.
A regularly wrapped ferrite core will have high capacitance, low Q, wide bandwidth -- which is not ideal.

Also, the gauge of wire you select and the size of your core is up to you. We should experiment with geometries here.

Anyone have an example of a honeycomb coil wrapped on a ferrite toroidal
to make a transformer?

[captainpecan]

captain, would that component that "looks" like a resistor with its colour bands......actually be an RF choke?......i have a couple..and they look exactly like resistors but of course they are "coils" ( i think )

Very possible. Like I said, I learned something on this one. I would have never guessed there was a coil hiding in that thing. I guess just seeing color bands on a resistor, does not mean it's just a resistor! Hey, I'm a little smarter now, woohoo... I'm hacking this thing up anyway, I just wanted to know why it worked with no inductor. Turns out, it doesn't! I feel better now, lol.

[rave154]

captain, would that component that "looks" like a resistor with its colour bands......actually be an RF choke?......i have a couple..and they look exactly like resistors but of course they are "coils" ( i think )

[captainpecan]

hello Cap

this is what i could fing on the ana608

http://dc87.4shared.com/download/119154918/e63b860/Solar_Powered_Garden_Night_Lights_using_ana608.pdf

and picture

Thanks for the quick response. From reading that link you gave me, I see an extremely similar circuit as is shown there. He mentions he mistook a resistor for an inductor. I did exactly the same thing. This thing only has two resistors, and both are color banded... Upon closer look, I did not notice at first the print on the board. One resistor is labeled R1, the other resistor is labeled L1... So that REALLY IS and inductor. It's color banded exactly like a resistor, green, brown, black, silver. Turns out, this must be some kind of wire wound resistor, used as an inductor?

New one to me, lol, I would have never caught that one! Thanks for the enlightenment!

[JohnnBlade]

Hi All,

i have made the big joule thief.
and i get about 220v with this joule thief on a NI mH 9.6v 700mah

when i hookup one of my cfl lights i only hear this high tone but no light at all, does that mean all is hooked up correctly, but i just need more volts ?

hope anyone can help me out here

thnx

Greets JB

[weri812]

hello Cap

this is what i could fing on the ana608

http://dc87.4shared.com/download/119154918/e63b860/Solar_Powered_Garden_Night_Lights_using_ana608.pdf

and picture

[captainpecan]

Confused, please enlighten me!

Hey all, I don't post real often, but I have been working quite a bit with the JT. First off, I love the work you guys have all done, and I watch each video from Lidmotor, Xenomorph, and many others many times over... Great stuff guys!

Anyway, I was working on a JT hack of a simple garden light. I grabbed a couple really cheap ones at the local walmart for 4 bucks each. Problem is, I was quite surprised with the performance, and I cant figure out why they seem to work at all. There isn't even an inductor in the circuit, and it still only uses 1 "AA" nicad to light 2 LED's? Upon testing, I am sort of impressed, because it is running those 2 LED's when I tested it with 1.23v and only at 20ma. It does put off pretty good light for what it is. Obviously with a joule thief I have easily gotten more usable light, but I have also done it with more amperage, and a good inductor.

I plan on playing with the TLC555 inverted JT on this anyway, so I know I will get it tweaked to be brighter, with less amp draw. But I am still baffled as to why this thing works so damn well to begin with? There is a component on the board I have never seen before, ANA608 and although it looks like a transistor, it must be an IC chip. It does have 4 legs which is kind of a giveaway. It obviously must be the magic that makes this thing push the voltage higher, but HOW?

Could anyone please take a quick look at the circuit, and maybe enlighten me a bit? With no inductors, how is the voltage high enough to light those LED's?

PS.. At 4 bucks, this little made in china piece of junk is a pretty darn good buy for anyone looking to play around. Direct sunlight, the solar panel is showing 2.48v at 30ma measured after the diode voltage drop. Parts alone are worth 4 bucks.

Sorry the circuit pic is a bit blurry, cellphone cam didn't focus well. And no, there are no capacitors in the circuit either. Only 2 resistors, 2 LED's, a switch, a diode, a photo resistor, and that mysterious ANA608... ???

[Watson]

Joule Thiefers, I wonder if using a honey comb coil
might be even EASIER for all concerned to wrap
in a Joule Thief.
Is there such a thing as a bifilar honey comb to create
a 1:1 ratio transformer?

Inductors

What improvement is this going to have over a conventional air core coil? The honeycomb coils I've seen are very large, and when the coil is large and made up of, say, 28 gauge wire, the resistance is higher than a compact coil. The higher resistance limits the peak current that the coil will conduct, when what you want is a high peak current.

With a small compact coil, especially with a ferrite core, the length of wire needed for a certain inductance is less, and the resistance is lower. It takes dozens of feet of wire to make a 100uH coil on an air core, even more for a honeycomb coil, and the resistance might be more than a dozen ohms. I can wind a 100 uH coil with a foot of wire on a ferrite core, with a DC resistance of less than a tenth of an ohm.

The special coils such as honeycomb coils are great for high impedance circuits such as receivers, but for low impedance, the ferrite core coil has the lowest resistance.

Would someone please tell me how your 2N3055 power transistors fail? Do they short from collector to emitter? Thank you for any information on this.

Oh, BTW, instead of honeycomb, it would be easier to wind a spider web coil, where the coil form is a flat piece of wood or fiberboard with notches cut around the outside edge.