Joule Thief

[jonnydavro]

@ electricme.Hi jim .Take another look at the circuit, i have reposted it here and highlighted the pulse motor and bemf recovery part.The coil i am using is 250 meters off 28swg enamelled copper wire,single wound.This equates to a 250gram spool off wire from the electronics shop,just wind it all onto another spool unless your lucky and they have left the start and the end showing and then you have a ready made coil.I am not feeding the bemf back to the battery yet i am just collecting it in a 4700uf cap.The bemf recovery is not shown in the video but one thing i will say is when i let the motor run on  i connected the bemf recovery cap and diode and when i pulled the plug the main 60000uf cap was at4.54v and the bemf cap (4700uf)was at  4.65v.I was thinking maybe it might be better to try and recharge a separate aa battery.What do you think?
Now safety.Thanks for the heads up on the magnets and i may implement some off your sugestions as eye's and teeth are not as replaceable as a blown transistor but the magnets we are stuck with.Not only do they trigger the reed switch but they are what creates the rotors rotation when oppossing the coils magnetic field.
The use off 2 dc motors as generators in a top and tail fashion is a very good idea and you may see that in a future version  credits to you.Anyway if anyone still has questions about the circuit i will be only to glad to help.
                               Regards jonnydavro

[mdmiller]

@jonnydavro  - very impressive, what is the core in your bemf pulse coil?  I noticed ist was using aluminum which although non-magnetic can function as a flux shield when electricity is involved but I don't understand all the details.

This JT has run for 48 hours on a dead battery which has stabilized the past 12 hours at 0.4V and running strong -- I had pulled the 1K resistor from its schematic position and just had a jumper - the bulb was brighter with the jumper.  Two hours ago, I put a secondary 6-wrap coil on the toroid and ran it through a diode bridge, and ran the DC from the bridge right back to the source battery.  Nothing really happened, but when I pulled the jumper and replaced the 1K resistor although the led dimmed, the battery has slowly been self-charging and is currently up to 0.7V.  Crazy little beast this JT is.

[jonnydavro]

@mdmiller.The core in the coil is steel welding rods.This is probably the best matirial to use considering the main function off the coil is to spin the rotor via its magnetic field .The bemf is just a useful bi product which we can put to good use. regards jonnydavro.

[Mk1]

@All

I see every one getting real busy, congratulations all, this is amazing ,great results everywhere , I have really proud to have you all as human brothers. This is only the beginning, many will come stand with us! What makes this great is that nobody needs to put his name on it , and we all benefit.


@ electricme , jonnydavro , gadgetmall , mdmiller

Great work , you have done in the last few days many fans, keep it up.

@all

There is a reason we are here don't think chance is so kind.

All can help in there way even small details.

Ok some test result , my regular jt with 6 turn of aluminum wire light a 7 w luxon led with no problem. There is also a 42 turn copper coil(30volt) on the jt it lights it brighter , I tried getting electricity out of the aluminum coil in a cap , with no success i still can't see anything on voltmeter ac or dc gives me nothing, did manage to get a cap charged when having a led in series with the cap, so i guess a spark gap is a good way to go, a small light bulb should do . Current may be limited by the bulb amps capacity.I have not been to calculate any amps but the number of led in parallel don't seem to create any issue i only got 20 of them so its early to assume that since the voltmeter doesn't see anything
that the amps are limited since there are none to begin with., I also tried a copper wire connected at both ends of the aluminum coil and the led would not light.


Edit: on my joule meg , the luxon screams for mercy 24 turns copper(only one of the 2 coil is used on the Jmeg  ) , i posted how i did it a couple of pages back 10,20,30 pages ? It goes so fast.

[jeanna]

Hi everyone,

The scientist in me took over for the past couple of days. I have done my best to make some order out of what I have made so far but...

I made a joule thief with 11 bifilar turns and set it up in a breadboard using a bright white LED, a 3904 transistor and a 850R plain plus 220R-trimpot resistors.=, just in case...

That worked fine.

Then using 1 meter length of 30 gauge mag wire (red from RS) and 29 leds (all I had) I began to measure the light and voltage used by the array.

First, I must tell you, I could get no reading across any individual led, nor could I get any voltage across the neg of one to the pos of the next in series.

I think this may be important so I want to stress it.

There are 2 voltage readings from the array. One which reads from one neg pin to the next (or any) neg pin; and also, from one pos pin to the next (or any) pos pin. These vary and are individual according to where they are in the array.

There are 10 leds in series. Each led has 2 more (=3) leds in parallel to it. -except the last one cuz I ran out.

The array looks something like this looking into the lights. The flat side of the D is neg.

-D+, -D+, -D+, -D+
-D+, -D+, -D+, -D+
-D+, -D+, -D+, -D+
...and so here are 4 in series going across (but I have 10),
and looking down here are 3 in parallel.

I made 1 loop of the secondary/ pick up:

I put 1 LED  in the board at a time starting from the upper left corner. Then I added the 2 in parallel.

It is not possible to see the voltage across the first column because nothing shows across the led from - to +. But it was very bright and the original in the toroid did not change at all. So, I had 4 very bright leds.

So, I added one in series and 2 more in ||. All 6 lights went dim.. I made one more turn and the array brightened a bit, and 1 more and it was bright. The voltage said
- to - 0.34v and + to + 0.04v. and between - and + there was zero.

So, I kept adding more leds and the lights would dim, and more turns to brighten it.

In the end I had 10 columns of 3 ||  leds each except the last for a total of 29 bright leds.. When you add the original light which is part of the toroid jt it makes 30 bright lights.

It took 20 turns to make these lights be bright.

The voltage from the first - to the last- was steadily rising to around 1.9v after around 10 turns. but the + to + quickly went to 11.7v etc.

Finally all the lights were in but not bright anymore. This was at 15 turns. So, I wound one at a time and each time it got brighter to about 20 turns. but after 15 turns the voltage started to go down.

Now, I don't know if this is voltage I am catching or voltage I am losing, so I don't know if this is good or bad. hmmm


At 20 turns and 29 lights The voltage showed:
- to - 8.35V and + to +  4.45V.

I got the screwy idea to change some whites to yellow and red.
so, with the first column red and the last column yellow,
The voltage is - to - 1.97V and + to+ 13.1V.

end of day. I mentioned a little to you and tucked in.
=====
next day I wanted to see how much drain off the battery there would be. I recharged 2 AAA batteries together, One would be to test this whole array and see how fast the voltage drained from the battery. The other would be for a plain toroid on the same breadboard but without the 29 lights.

So,

in 10 hours the array drained the battery from 1.43v to 1.25v
but starting with hour 2 the voltage was drawn down from 1.35V to 1.25V (you will see why in a moment)

Last night on the same breadboard but with a plain toroid same specs and the battery had lost a bit to time and started at 1.42V and after 10 hours was 1.27V.

Because that first hour of the battery might be a weak start point, counting this from the second hour to 10th the voltage went
from 1.35V to  1.27V.

The array drained the battery 0.1V And the regular jt drained it 0.08V

So, the difference in drain on a battery between a jt with 20 turns and 30 leds, and a jt with no  secondary and only 1 led is 0.02V.

that is 20 millivolts in 9 hours cost of running the extra 29 leds.

wow.

Please repeat this experiment, folks.

MK1 is this the kind of thing you saw?

Anybody else doing this kind of study?

jeanna

EDIT:
I am adding this anachronistically. the ac voltage across the toroid wound with 20 turns but not connected to the array claims that there is 1.3VAC. coming from its leads.