Joule Thief

[resonanceman]

   Guys:
   Thanks for the replies.
   I'll be getting a solar panel and a couple of smaller 12v batteries to work with soon as possible. But the idea is to make a big strong JT-TPU circuit that is feeding back to the run batteries, or run caps, and eventually not needing the battery-solar panel support.  Since once the Jtc drops in its run battery voltage, it is not working properly, if not kept at a constant optimum running frequency. There is no circuit that can work Properly like that, if not kept at a constant input level.
  I have been trying different transistors, resistors, diodes, pots, caps, toroids, Leds, etz... trying to find the winning combination.  It seams like a simple quest, but it's not.
What keeps me hard at it , is the idea that a self running joule thief is more than possible, and that it can be upgraded.
  Otto had said that what makes solid state components interesting is the fact that sometimes they can be fooled...
                                                   NickZ

NickZ

You might want to keep an eye on my UCVC thread .....what I am working on there is very similar about what  you talked about in this post.

I agree it is not as simple as it seems like it should be.
I played  with flyback and feedback a few years ago.
I was able to keep the source battery charging as long as I was willing to tweak the base resistor  every few hours.

Now I know a little more .
As the battery voltage rises.........the JT primary voltages rises too.......part of the problem is the JT primary voltage rises faster than the battery voltage.
So far I have only got to the self charging state at low power levels......If the power level goes up a little to high......the self charging stops.


gary

[poynt99]

I don't know if this has been noted here by anyone yet, but I thought the JT people might find this patent interesting.

.99

[hidave]

NickZ

I have never  tried the core on the yoke of  TV tube as a JT
I believe IST tried them  .......he suggested that I try them...but I never got  around to it.

I am not sure if they are actrually ferrite.......but it probably doesn't matter. If they are ferrite they will run with less windings.
There is  plenty of room for more windings  if they are iron.

As far as lighting 400 LEDs ..........I assume you mean to full or near full brightness. 
It might  be difficult to do with  just 1.5 V .......  you can light  them pretty well with one battery ......but I think full brightness would be dificult

The  yoke core might be a good  thing to try for this ........a flyback transformer core might work well too.

I know I have never been able to get that  much power out of a single 5 for $1 toroid.........but maybe it can  be done.

Personally  I would use a candy cane coil........ I use them for  most of my experments.
I don't remember anyone else every saying anything about having made one.... but I find them very useful........and  they can be scaled........if  it doesn't  have enough power.........just add more toroids.
I am about out of time tonight. 
I will try to take some time and see what it will do with LEDs
I think I have 400 LED christmas tree lights.....
I will try to post a picture tomorrow evening


gary


As for  12 V
I use  12 V quite a bit.
it is MUCH easier to light a CFL with 12 V than 1.5

I found a ferrite cone core from tv deflection yoke.

Big 5" O.D, 1.75" I.D. and 2.25" tall. It has 2 windings on each side, several hundred turns of 24 gauge.

I want to use this to light to full brightness 100w gutted spiral CFL (several) from 12v 220ah deep cycle battery. This battery is hooked up to 200w solar panel and I get about 6 sun hours.

What will be a good JT circuit to use?.  I see many small JT that lights 13w CFL to full brightness and drawing 300 mili amp.  What is the draw on a gutted 100w CFL in full brightness?.

Thanks.

[conradelektro]

Hello hidave,

I did a lot of experiments with the Joule Thief and CFLs.

To light a 13 Watt CFL to its full potential brightness you have to push 13 Watt through it. You can do that conventionally (with the circuit in the CFL for 110V or 220V) or with a Joule Thief, but you can not avoid 13 Watt. If you push less power than 13 Watt through a 13 Watt CFL it will be less bright.

It is kind of difficult to run a Joule Thief circuit with more than 6 Volt because the transistor over heats easily. The transistor is driven into saturation (by high Voltage on the base) and the Voltage over emitter-collector rises to more than 100 Volt (which is too much for most transistors).

Read this web site to learn about circuits to drive high voltage transformers with a 10 to 40 Volt power supply http://wiki.4hv.org/index.php/Flyback#High_Power_Drivers .

One needs two transistors and a zero voltage switching technique to avoid overheating of the transistors. The above web side talks a bout fly back transformers, but this also applies to Joule Thief type coils.

Also observe the diodes and zener diodes in the proposed circuits, which limit and rectify the voltage (coming from the sense coils) on the base of the transistors so that the transistors stay in their intended region of operation.

See this web page about regions of operation of a transistor:

http://en.wikipedia.org/wiki/Bipolar_junction_transistor#Regions_of_operation

The Joule thief circuit is good for supply voltages between 1 and 4 Volt, but then it becomes more complicated. With a 1 to 4 Volt power supply one can reach very high voltages on a properly built secondary. I managed up to 2000 Volt with big air coils. But this becomes impractical. Many things can be done with a Joule Thief but very soon the transistor overheats. I had good success with Joule Thieves driving up to 100 LEDs with a 1.5 or 3 Volt power supply.

But once you have a 12 Volt power supply it is better to drive the LEDs directly from this 12 Volts, no Joule Thief necessary.

It is a myth that a Joule Thief circuit somehow generates more power than you push into it. In fact it looses power in the form of heat in the coils, in the transistor and in the resistor. But one accepts this loss because one can light many LEDs with a 1.5 Volt battery.

Many people do not realise that modern white LEDs driven with the intended power (specified by the manufacturer) are so bright that you can not look at them directly. And if you drive them with a Joule Thief they light brightly, but much less than they could shine (when given the specified amount of power instead of the reduced intermittent power from the Joule Thief). So, many people think the Joule Thief saves power.

One thing one could claim: it is not desirable to drive LEDs with maximum power, because they hurt your eyes. So, driving them with reduced power via a Joule Thief looks nicer.

Greetings, Conrad

[WilbyInebriated]

t is a myth that a Joule Thief circuit somehow generates more power than you push into it. In fact it looses power in the form of heat in the coils, in the transistor and in the resistor. But one accepts this loss because one can light many LEDs with a 1.5 Volt battery.

there are many people in this thread who are well aware that a simple blocking oscillator is not OU. the jt is a simple old circuit that was cleverly applied to the purpose of using the energy left in a 'dead' battery. i agree, the transformer seems to mystify those who don't understand volts, amps and joules. it's a great circuit for flashlights and other things that run on low voltage and don't mind 'noise'...

Many people do not realise that modern white LEDs driven with the intended power (specified by the manufacturer) are so bright that you can not look at them directly. And if you drive them with a Joule Thief they light brightly, but much less than they could shine (when given the specified amount of power instead of the reduced intermittent power from the Joule Thief). So, many people think the Joule Thief saves power.

to be specific and precise, leds driven with a pulse, are more efficient than being driven by a dc supply. two words... duty cycle.

edit: corrected. i had misread conrad's post. my apologies conrad.