Joule Thief 101

[resonanceman]

Just thought I would update my information on what I called candy cane coils
I called them candy cane coils because on my first ones the windings spiraled around the stack of toroids like the stripes on a candy cane.

Now I would call them modular coils.
If you have a project calling for a large ferrite toroid but you only have smaller toroids you can split the windings of both the primary and secondary over a number of smaller toroids.

The total number of primary turns VS the total number of secondary turns seems to work out almost as if it was one large toroid.

It can also work to put your primary on 1 toroid and leave a long wire to connect  with the collector........this wire can then go through quite a few toroids......any windings on these toroids become secondarys...... I suggest going for fairly high voltage in the primary if you try this.

If you dont need alot of power from your secondarys just passing the wire going to the collector through the center of the toroids may be enough.....wrapping then around the toroids increases the power but usually lowers the voltage.

[gadgetmall]

Here is a video tutorial from makemagazine for a Solar Joule thief bracelet http://www.youtube.com/watch?v=ghB2irHIN8I&feature=related

[Pirate88179]

Gadget:

Wow, the little JT circuit is really getting around everywhere now.  That is a cool application.

Bill

[dennisk]

Hi sierraloewe,
After you have made your first jt and it works, then it is time to delve into those questions.

Follow the makezine or evilmadscientist youtube as closely as you can.

The gauge I use for the primary bifilar is 24awg. I got this from a telephone extension cord.
The length you will need to make this first one will be about 14 inches of each wire.
This depends on the fatness and diameter of your toroid, so this is just a guess. (I use a small one and it takes 11 inches of each wire and is very generous.)

After you have made a plain joule thief make a second one exactly the same way. Put the led into the led spot but do not solder it.
Remove the led when you prove that the circuit works.
Now, wind a secondary.
The choice of gauge for the secondary is up to you.
The thinner wire will make higher voltage, but the problem with really thin wire (like 30awg from radio shack) is the insulation is easily scratched. If it gets scratched you will have little to no voltage from your output.
One more thing about the secondary.
The side of the primary that connects to the collector coil is the part that is being "transformed" so, if you have 10 turns on the collector side (you do now if you are following the makezine video) Then double that will be 20. So, make your first secondary with 20 turns.
This will give you a starting point for reference voltages.

Now, you have 2 very useful circuits.
1-the plain jt is the best way to get the last bit of joules out of the battery. It makes a terrific emergency light.
2- The battery needs to be a little higher for the secondary to work. There are exceptions to this, but most of the ones that use a secondary run down after the battery is at about 1.1v.
So, you use the jt with secondary then when it is no longer bright you switch that battery to the plain jt.
This does not apply to a rechargeable battery.
Do not use a rechargeable in a plain jt, because it is not good for batteries to be so thoroughly drained.
(I could never figure out why but it seems to be so.)

After you have made these 2 you can start experimenting.

I have found that it is very important to make the wires wound snugly onto the toroid.
It is possible to have a good one that is loosely wound but if you want high voltage... get it tight around and try to have as many winds as you can actually in contact with the ferrite.

(I believe there is a difference in opinion about loosely wound together, but this is not about wound together, this is about being tightly wound on the toroid itself.)

That toroid is a powerful source of "something" that makes high volts.

I hope this helps.

jeanna

I noticed you said not to use deep cycle batteries that have been completely discharged.  I don't know if this is off topic but have you tried recharging them with the Bedini school girl charger?  Just curious.

[sm0ky2]

Hi all,

after months of lengthy conversation with a couple of my guitar-builders
I think I can explain the resonant operation of a joule thief circuit, in a manner which the technicians and builders can more easily understand.
(and replicate)

the biggest problem I have encountered when discussion self-resonant JT operation, is other builders cannot seem to get them to operate
in a resonance mode.

Several heated discussions, have led to the understanding of SRF with respect to the LRC portion of the circuit.
However, the transistor function seems to cause problems for many.
I have tried to explain the process of "tuning" the base resistor, while observing scope image of the primary transformer winding.

What I kept getting as a response, was asymmetric scope images of the transistor switching
(I assumed they were measuring across the LED (diode).....)

These scope images were sent to me and/or posted in multiple JT threads, displaying a sharp increase in voltage during the "switching on" time.
And an elongated (delayed) voltage decrease during the "switching off" time.

Part of this I have found to be a function of certain LEDs. They store energy across the diode (capacitance), which dissipates over time after the LED is switched off. meaning there is still current flowing through the diode after the transistor is off, and the production of "light". This can result in a misleading scope image. I state now - the (indicator) LED is not an essential part of the Joule Thief Circuit.
It is simply an indicator to let you know the JT is in operation.

Remove the LED, and the Joule Thief will continue to operate.
Now the scope image, across the coil, does not include the diode bias discharge.

Still I was receiving similar responses - they could not get their JT to run in a resonance mode.

At some point I gave up trying to explain this to people, because they just wanted to argue rather than try to understand what I was explaining.
Similar scopeshots of an asymmetric transistor switching were presented, and this was a sort of 'road block' to the communication process.

Anyways, my recent discussions with JT builders has given me a new perspective (and perhaps new terminology) that others may understand.

there is a voltage bias (different for each transistor) which sets the switching function into what is called in the transistor world

"linear mode" This term is derived from power graphs, in which the function is linear.
With the voltage biased within this range, an A/C waveform translates perfectly across the transistor.
People in the radio world already know this, because such is necessary to use a transistor for analog frequency response.
But I have found, that digital techs often know nothing about this mode of operation.

Each transistor can switch in a purely sinusoidal fashion when the base resistor is biased to within the voltage range of this "linear mode" of operation.
Within this linear mode, changes in the bias voltage results in a range of operating frequencies.
The SRF of the circuit should be designed to within these parameters.
Most standard JT designs are capable of this.

It is simply a matter of adjusting the base resistor to match the self-resonant frequency rest of the LRC circuit.

This is evident in the scope shot across the coil, As the waveform will become perfectly symmetrical.

When this occurs, the JT circuit will output the highest possible voltage and current for the input.
And as such, this is the most efficient way to use the JT circuit.

This is the difference between a Joule Thief powering a light for a few days,

vs JT's that have been running now for years.

all off what we would consider to be a "mostly dead battery".
meaning, one that has been drained below the operation conditions of most equipment.

So, when starting with two similarly drained batteries, why such a great discrepancy in run times?

In non-resonant operation, a JT circuit wastes a lot of energy, bucking against the natural frequency response of the coil.
This generates heat and EMF radiation. Because the field collapse and the charging function have a gap in time. and/or are an asymmetrical function, due to unsynchronized transistor on/off times.

When the transistor is in linear mode of operation, the sine-wave from the coil (through the base resistor), draws current from the battery, corresponding to the waveform. If you take a current scope shot of the signal in this mode of operation, you find the current also follows the same sinusoidal waveform.
slightly advanced ahead of the voltage. (~90-degrees)

I hope this helps, I haven't quite given up trying to deliver this message... it just gets frustrating with the language barrier, when it comes to technical terminology, differences between engineers and techs, etc...
People either have a hard time understanding what im trying to say, or they take a term in another context than the one in which I was trying to deliver it...

anyways, apparently the proper name for this mode of transistor operation is called "linear mode" in the technician world.
It is listed in the technical specs of the transistor - so you can get a feel for the biased voltage range of the base resistor.
The number of turns in the coil should correspond to this range of voltages, and by proxy, affects the SRF of the particular circuit.