Joule Thief

[jeanna]

@dasimpson,
Thanks for the pix.

The tiny green one looks a lot like one of mine from a cfl. I think 23w size.

extra inductor in series do you mean another toroid say tenty turns dump coil (meaning the coil that comes from the secondory of the firct toriod) and the turns needed to get the lgiht needed on the pickup

sort of.

I made a long study of this.
I theorized that the pulses which were being made by the joule thief were being picked up by the toroid.
But when I had  3 leds coming from the 20 turn secondary, the 10 leds running from 84T secondary would not turn on. but then I used a toroidal inductor in series and suddently both went on.

I used this idea and eventually by adding series inductors of different types, I got I forget 28 leds on the 84T and 8 leds on the 20T and another 5 in an additional string.

It is documented on the ac from secondary thread, but also on 6 videos on yt.

the point is that when you have the pulses already happening, you can get a free ride from them if you add a series inductor or 2.
I also added caps, but they only worked  after I added the series  inductors.

It starts here:
http://www.youtube.com/watch?v=Y4IMgDRGpHE
And there are 6 parts.
1,2,3,4,4a,6 are their numbers. It will do well to have more people try this.
The toroid there is 10/$10 Tor-23 from allelectronics.coom.
It likes 5T to the base, 8T to the collector and here had a frequency of 205khz.

Thanks for being interested!

jeanna

[resonanceman]

you lost me with this part  "with  a good toroid  you should  be able to run  with  3 or 4 wraps bifilar......  right now  your  secondarys  are working kind of like  step down  transformers..... with 3 wraps   your same secondarys would work like step up  transformers. .

and as far as i know ant psu ones good cos they have to work in the 60mhz range

The 60 Hz range is not  good for JTs........ a toroid  designed for that range is designed for power supplies and stuff like that.
Jts often work in the KHz range........but more important  the spikes a  JT makes are VERY fast......  a  good toroid for JTs should  be able to work  in the MHz range.

I added  a couple of pictures  of  JTs I have made  with flyback  transformers

The  first one was made  with just the core......... it has  bifilar 30 gage magnet wire windings 200 ft long.......these  wires used in parallel  will light  a CFL tube...... the thicker wire is 22 gauge  40 ft........it will light my 90 LED array very brightly.

Notice  the blue and white wires  for the primary.
This  primary was wound in the MK2 style ........it has 3 wraps of each wire on each  side.........later I removed 2 of the wraps  on each  side  of the white wire.... the white wire  was connected to the base.


The  second picture  is  of  a JT powering my 90 LED array
This flyback  transformer was  sealed.......I could not get it apart  without damaging  it .........but the windings  inside work  great if you  hook them up right.

Notice my JT board.......  I have a piece of wood  with a heatsink  and a pot mounted on it....... then I have  screws  for  terminals.....I then use  alligator  clips for connections..............for me this works  better than a breadboard.....

gary

[resonanceman]

dasimpson


As  you  look for parts to salvage do you ever  come  across any ferrite core transformers?

This picture is   of the bottom of  a transformer out of  a monitor.
I drew lines between the pins  with continuity......
I was  playing with feeding  back  part of the output  back to the battery...that is what I used the  connection lables feedback for.....as I remember it was about 50 V
The secondary was  about 200 V ......with 1 AA

IN general......if you can find a ferrite core transformer  with 3 or more  low resistance pins  with  continuity .........and  2 pins with higher reesistance.......it will probably make a great JT

gary

Edit

The  B +  on the  clip  with 2 stripes should be read........battery positive ......  that  would  be the same as  the 2 wires twisted together on a normal JT

[Pirate88179]

@ dasimpson:

Nice work on your jt circuit and also on the photos.

As you will read in the thread as you go over it, we have found that, whatever toroid you use, you will need to tune your circuit to it.

I agree with the others that ferrite is the best, easiest type to get good results from although some, like Jeanna, have been able to get the powered iron ones to work.

You are using a 1k resistor which is a good average shot in the dark for the basic jt circuit.  As you alter your primary and base and secondary turns and, depending upon what toroid you are using, you will need to determine the best resistance for what you are trying to achieve.

A variable resistor is the best way to do this.  Try a 5k VR in your circuit and turn it until you get what it is you are looking for be it low amp draw or brightest light output.  Then measure what the VR is providing at that setting and you can replace it with a resistor of that same value.  This will make a huge difference in your circuit's efficiency.

I hope this helps.

Bill

[conradelektro]

@ Gary (resonanceman) and all other experimenters who have experience with a lot of different "cores", toroids or "transformers" in a Joule Thief circuit:

Gary, what is your experience with "screeching" or "singing" of cores, toroids, wires or transformers?

This happens in the 1 to 50 KHz range of oscillation. Also lower harmonics of the Joule Thief frequency happen and are sometimes in the hearing range of people and pets.

I am in the process of doing a series of measurements with different air core coils in a Joule Thief circuit.

At the moment I can light 10 to 30 LEDs with a frequency of 50 to 150 KHz (from a 1 Volt power source). Best so far is an air core with a diameter of about 110 mm, 20/20 bifilar L1/L2 and a 200 turn secondary (standard Joule Thief circuit with a 500 Ohm resistor on the base of the transistor).

The big problem is high Voltage (e.g. 1200 Volt) from a 1 Volt power source. Even with air core coils I can not reach high frequencies. I am kind of stuck with 20 KHz (which are bothering people with good hearing and pets). With an air core the noises made are very faint, but they are still there. With toroids the noises are rather strong. With toroids (and high voltages on a secondary L3) I am stuck in the 4 to 6 KHz range.

All seems to be good once the circuit swings with more than a 100 KHz.

It is easier to reach higher frequencies when starting with a higher Voltage (e.g. 4 Volt or 6 Volt), but one has to be careful not to overheat the transistor or the base-resistor.

I "see progress" with bigger diameter air coils (200 to 300 mm). The bigger coils show very nice regular patterns when swinging, not spikes like toroids or transformers.

Attached see one example with an 110 mm diameter air core. The big bumps are happening with 108 KHz.

When I put a 3 K resistor on the base (with a 1 Volt power source) I can create nice sine waves with several hundred KHz on a secondary L3, but the voltage stays very low (e.g. 60 Volt) in spite of the many turns of L3 (see second scope shot).

I also want to go into pancake type windings for L2.

The "trigger L1" (going to the base of the transistor) should have fewer turns to increase frequency (e.g. 5 turns). So far "high frequency" has lowered very much the voltage on a secondary L3 no matter what I do.

Greetings, Conrad