Sometimes I was wondering if a Joule Thief could be built without a resistor.

Or without a transformer, just a coil (or two). However this circuit needs extra parts and usually, 2 Transistors.

How about reducing the parts count even further? And using a ready-made potcore inductor.
Such as a Panasonic 10mH component.

One extra winding is added, and this becomes the main winding for the LED current.
A regular 2n3906 PNP Transistor and a 1.5v AAA Battery and a 1W white LED are used as component.

Now, the base is connected to one pin of the 10mH coil.
The other is just left free! A piece of tin foil is attached to the corner of the proto board, so it just covers the other pin of the 10mH Inductor.

By rubbing the tin foil, the Joule Thief actually starts!

Explanation: The tin foil works as Antenna, so the electron holes can flow out from the transistor base into free air.

I used just one connection at first, but brightness wasnt great. I had to use a start button, and though this isnt a resistor, component count isnt reduced. Also the wire arrangement was sensible very much.

The tin foil does not need to be stretched out, it is just right like it is, but it is important it has good coupling to the 10mH inductor (here done with a solder blob).

When the tin foil is lifted a little, brightness goes down!

I think before of adding the foil, energy was flowing via capacitive coupling between the windings. I explain that by when compressing the windings, brightness would increase.

It starts up easily (but not by itself).

Only 3 components: LED, Transistor, and Inductor (with added winding actually just a few turns).
The feedback winding has more turns than the main winding.

Interesting circuit, or is it?

 

It does actually create high voltage, more than 60 volts, so a small neon lights up.

 

isnt that a big breakthrough? No more winding of transformers, no more resistor required.
And you even get a switch that way.

I want to try with a smaller Inductor soon.

It is absurd but would it be possible to reduce parts count further? Maybe no transistor?

 

Some updates here. I got the circuit working without tinfoil now.
It ca be started with a pushbutton! It is important the pushbutton is at the open end of the coil.
Rectification reduces brightness a lot, I think 8 turns is not enough.

Made another circuit with a 470uH inductor, and a wire winding on top of it.
At first it needed a fairly large piece of tinfoil, adding a 110k resistor and a LC network
helped to reduce it a lot. It is self-starting actually just with the 110k resistor,
but adding the LC network and a smaller piece of tinfoil increase brightness.

Interestingly, decreasing the resistor does not increase brightness, about 10k or 110k
dont make a difference, but below that, brightness decreases.

Todo: increase brightness for the white LED circuit, try a larger coil with more turns,
and rectification.

Try even a smaller RF inductor and possibly 470k or 1 MOhm.

 

I know it's a simple circuit, but a schematic with component values would be nice.
Your parts count is low, but isn't an inductor like that fairly expensive compared to other components?

 

I know it's a simple circuit, but a schematic with component values would be nice.
Your parts count is low, but isn't an inductor like that fairly expensive compared to other components?

OK I 'll draw one. No they dont cost that much from Farnell, about 20 cents or so.
The larger one I want to try costs more, though.

 

ok here a schematic

 

The low cost of a resistor in series with the base or base winding is good insurance against damaging the transistor, particularly in case the oscillator does not start because of a reversed winding.

 

The low cost of a resistor in series with the base or base winding is good insurance against damaging the transistor, particularly in case the oscillator does not start because of a reversed winding.

Your reply isnt entertaining me. Transistors are cheap too.

did you notice the feedback winding isnt connected to anywhere? And yes there's high voltage across it, more than 60 volts.

 

actually i am going to sell it as kit- I do use the circuit board, since it can be started with a pushbutton,
and because the feedback coil connects to nowhere, the battery can be left in place.

The circuit needs zero current when it is not oscillating since the transistor is off.

To switch it off, simply touch the pins of the feedback coil.

I use it as small lamp to look at my ants in the night.

I bought countless LED torches but they all must be twisted, and after a while, it does not work right anymore.
The larger torches take larger batteries and well arent just flat.

Its going to be cheaper than all my other joule thief kits, and it is easy to build.

 

no transformer? then what is the inductor with the coil wound around it?

 

no transformer? then what is the inductor with the coil wound around it?

yes but you dont have to make a transformer.

 

two coils magneticly coupled are a transformer.

 

OK. Do you agree it is not a regular Joule Thief? There are differences, how it is built, how it works, and how it is used. For instance it must be started, and draws no current when off.

Buttons are cheaper than switches and no component is better than any cheap component.

I have many resistors here thats not the point, just a challenge to work on this circuit.

 

The circuit has its merit. Thanks. Saving or not a resistor or transformer. Call it a takaothief Sort of a "blocking oscillator"
So the neon bulb is in parallel to the superimposed winding ?

 

The circuit has its merit. Thanks. Saving or not a resistor or transformer. Call it a takaothief Sort of a "blocking oscillator"
So the neon bulb is in parallel to the superimposed winding ?

yes in parallel but not required just for testing.

I am mainly interested how can I make the LED brighter? Possibly I should try a better transistor, or more windings.