Joule Thief

[Mk1]

@all

I hope this will help !

First thing tuning, on step one the green coil is put on tight leaving room so that two other coil (for the jt)can be added 180 degrees
apart, the green coil needs to be connected to a diode bridge to get voltage readings .

On step 2 , you start adding jt coils one at a time and check voltmeter every time you put more turns , that way you see the difference it makes continue adding turns until the voltage on the voltmeter start going lower, this will mean you have reached the point where you start messing up with the turn ratio , meaning you add more you get lower voltage.

On more step i do is every time i change the jt numbers of turn i also tune the base resistor , usually a pot of 1k is used, you try to get the highest voltage on the voltmeter . i also keep the voltage data and the resistance at the base by checking the resistance used by the pot in the circuit.

At this point you will know that you have the right number of turns on the jt coil, and the green coil number of turns fitting in the toroid,.

At that point based on results decide to use smaller gauge the get more turns and higher voltage .

You will also have a good idea on the toroid working range.

To make a good design , you need to learn about the toroid first at this point you should have a good idea about doing so.


Step 3 , Start putting pickup coil on the toroid, you need to put the coil on both sides going one way and leave wire to come back over the coil the other way on step 4.Making a cross windings.

If you want more pickup coil , you need to put them all at the same time , so make sure to remember the green coil for the room available  .

Lets say 20 turns fit , then on a mk1 you should have about 20 up then 16 down , mk2 having 2 pickup coils on each side numbers would be 10 up 6 down , you get the idea.

When you need more then 3 pickup coil , twist all those wire into a single one then do it all at once.

Now why the cross windings , you will then have the coil pushing and pulling , you can test it with a led that will now light both ways , it also helpful in charging caps.

I hope did not forget to much, i would really make a video but , i really don't have any money to buy a camera , and none of my friends have one, but honestly most of my time and energy is on putting food in my stomach.

Mark

How to test dead spots use a single turn pickup coil , then connect it to a bridge and voltmeter and move it around the empty space left on the toroid , if you find a spot where there is no voltage that is a dead spot.

[Pirate88179]

@ Jadaro007:

I didn't know there were 2 devices called drinking birds.  I assumed....and we all know what happens then, that Wilby meant the evaporation bird, which is what I ordered.  I got a smaller one that dips about every 30 seconds or so.  I think this will be ideal to test my theory, or my idea anyway.  I am thankful that Wilby thought of this.  Of course, I hope he didn't mean the other one you referenced.


Bill

[electricme]

@ TheNOP

take a 10 volts 10uF cap, fill it to 3 volts and connect a 3 volts led to it.
why did the led blowed up ? you only had 3 volts 10 microamps in that cap.
do we really need lots of charges ?

for me, the talk about the "power gain" is about efficientcy gain.
if it is more efficient you have more power out that you would normally had by other ways.
the "test it and see it by yourself" is a must do thing.
it does not automaticaly imply a ou claim.

one can beleive what he want to beleive, but when facing the test results, one can no longer make assomptions of what was actually meant.

@electricme
i can't make out if you talk about discharging the caps at same time or sequencialy.

i think a cascade discharge would be more appropriate.
i was wondering how to do that.
using leds chaser..., nice idea.
might work
thanks

Hello and thankyou for the question.
As you mentioned in your above statement, use a 10uf 10volt cap etc etc, and the result was as you posted.
The inrush current fried the LED, you didn't state if the LED had a series resistor connected to it to protect it, as the LED is a current handling device, but we here have made it work and very successfully wiht high voltages toooooo.

So to answer you question directed to myself, here goes.

What I want to do is have a lot of identical capacitors, electrolytics, then use a Dot/Bar driver IC to discharge all the caps in one instant of time, which will produce the energy that is required to drive any electrical motor, device or even to recharge a battery.

A 10v cap won't recharge a 12v battery, but substitute a 16v cap willl do nicely.

I want to use a JT, powered by the 1.2v Battery, as this is a NiCad, it can give me the  amps to do it, the AA size carbon battery is incapable of supplying these amps. Why use a JT and not a 555 timer?
Well I want to, and the JT output is used to pulse the caps to charge them.

The switching arrangement to control the time it takes for each cap to refill is done by resistors and capacitor tank circuit, the resistors being a mixture of fixed values and pots for fine tuning the timer between charge time and switching to the next cap to be filled, the discharge and the reset function on the IC.
 
The steps to do this is to have an arrangement of MOSFETS, capacitors and a DOT/Bar driver IC configured in such a way where the timing of the IC itself is controlled, which controlls how much time the pin out of the IC will supply energy to each capacitor in turn.

So the output pins of the IC go from 1 to 10, but I only use 8 outputs to be used to go to each independant cap circuit.

Then I take the 9th output from the IC which usuallys turns on a LED, is redirected to switch on ALL the MOSFETS together (in Parallel) will then dump all the energy from the charged up caps into whatever load one wishes to apply to the circuit.

Now there is one last LED driver, the 10 one which usually drives a LED also, but instead of driving a LED, this part of the circuit is taken back to the "reset" pin of the IC, so the whole process occurs over and over and over again and again.

Just as we all saw in the TinyWeeny Video I posted a couple of Posts back, lets take a quick lok at this now.
The LEDs all are lit in turn by the IC, which directs each LED to be litup in turn, so the string of LEDs appears to light up as the light goes round and round and round.
This circuit has been worknig since switch on over a month ago woithout fail.
I wont be using this identical circuit, rather I will design my own as the one shown in the tiny video is only there to show everyone the principle, which does work as you can see. Go back and take a look.

No doubt there will be refinements along the way, how long before a 1.2v NiCad cell will work, who knows, actually at this point in time I only need to get the circuit working, after having reached that gole, then fine tuning occurs, perhaps between every 10th cycle a burst of discharge energy could be sent back into the NiCad battery, a bit like stooring up energy in a flywheel and using that energy to recharge the circuit.

I am not making any claim to OU, I need proof of a working model over a period of months, even a year or two to do that.
But that is a long way away from this kid right now.

I will begin the circuit shortly, it will be a biggie, but got a little googling to get some more info.

I hoped this explained it for you.
jim
 

[electricme]

@Jeanna,
No LEDs on problem

This time the scope said the combined sides of the MK2 had a peak to peak voltage of 108 vAC. But this is when I tried to get the led string to turn on but it wouldn't.
I am using the probes on each wire of the secondary so this is the curve.

This might be a "loading" of the circuit while your scope probe is in the circuit.

If the X times switch is in the X1 position, this will definatly load the circuit, so switch the probe to X10 and try again.

If you still get a bad result, put a 10meghome resistor between the probe tip and the circuit.

If this dosent help then something else is not quite right.

Hope this helps

jim
 

[TheNOP]

I hoped this explained it for you.

yes it did.

but i am a bit sad.
you don't see it like me.
a mk8 would fill those caps in no time.
no needs to complicate the charging process at least this is the way i think.


anyway, each one is own way.
don't blow up what you are trying to make work.   
depending on the size of caps you want to use, big resistors at the output of your caps should help you.

good luck