Big Joule Theif

[morpher44]

Rodin with JT

SO I put together my first Joule Thief variant today. I used one tn3055 Transistor, 1 9V Duracell and one Blue Playschool Ring Rodin Coil. Dont get too excited. Because unless I did something wrong (which is always a possibility) It did not work. I tried it in different configurations. To no avail.

@redrichie

If your Rodin didn't have ferromagnetic material inside, I suspect
the inductance wasn't large enough. I made a Rodin-like coil
with a tuna can, and managed to get the Joule Thief to work with it.
I've seen more impressive videos than mine whereby folks
have made the Rodin work with Joule Thief using a neo-magnet
to BOOST the inductance, so
it is doable.

[jeanna]

That could be right redritchie.

How many turns did you use on the secondary that gave you 700v?
And, what is the amps draw?

If it were proportional to a 1.2v it would be 700/9 *1.2 so this would be 93.3v on a 1.2v.

Can you check that for me?
Do you have a AA around to try?

In general, you are more likely to get high voltage and low amps draw at the same time by making a many turn secondary and starting with a lower voltage battery.
So, a 700v output is really high, but by using the 9 volt you may be throwing away amps, but nobody I know has actually tested this with the same set-up.

What do you think happened with the rodin coil?
just the difference because it is an air coil?

MK1 and ist have verbally described high results of a sloppy rodin style wind but on a ferrite toroid.
I have it on my list...

nice going,

jeanna

[redrichie]

SO I put together my first Joule Thief variant today. I used one tn3055 Transistor, 1 9V Duracell and one Blue Playschool Ring Rodin Coil. Dont get too excited. Because unless I did something wrong (which is always a possibility) It did not work. I tried it in different configurations. To no avail. So I hooked up a smaller toroid I had wound for one of the first JT. Bifilar on a 1.25 OD, 1" tall toroid. This was single layer, tightly wrapped 20 ga wire. Using 9 V i was able output over 700V. Is this accurate? (ballpark at least?)
I also wrapped one 22 ga wire 10 wraps on top of the bifilar, that charged a 1500uF cap up to 12V in about 4 seconds and then held there..

[theremart]

Ahh Joule of a theif stole my time

RE: earth ground

I have managed to get a 2n3055 to fire with earth ground and was charging up super caps with it. got up to .15 V nothing in terms of voltage, but it did charge up the second cap. The input voltage shot from .8 V down to .30 when it was charging when I flipped the scale to micro amps it did indeed show charge on the target cap very very small but it was there.


RE: Transistor makes a difference.

I fully agree with you the transistor makes a HUGE difference, I found there is worlds of differences just in 2n3055 transistors from one to another depending on who makes them.

[jeanna]

at 1.5 V I am thinking the water battery or even the ground battery should power this up.

The question I have is why can't we get an off the shelf Toroid all wound up to work with our setups?

Hi Mart,
I am with you on the EB. I have not been able to even get a transistor to turn on in my back yard, but maybe you will have a better charge.

The ones all wound up may not work because they were designed as straight transformers with house voltage in mind.
If you can find a toroid where the primary was centertapped, and you can stick the center tap onto the positive battery rail, and get hold of the free ends of that same wire to connect to the base and collector you might be able to use one. I know a lot of folks have taken the windings off these things and rewound them.

Is the winding method that precise that it demands we wind our own?

No, I don't think it is.

I think the transistor makes the biggest difference, but many of the parts need to dance together.

You need something whose secondary, when wound, will give you a solid 450 volts.
So there is a minimum.
I am the minimalist in the crowd and I did manage to get a 1 inch toroid wound that lit a 4 watt tube, but it no longer does. I don't know if the connections have gone south or if the cold could make the difference?

I went to order the toroid at Allelectronics suppy = -8

I made it -12 lol.

I am glad you got on the list, but maybe you did not. They do not backorder afaik, so check again.

I noticed they have a 2 inch one. This might do it.

I have just lit a string of 34 christmas leds with a 0.5 inch one that I squeezed 172 volts out of. [tor-23 10/$1]
There is a difference in these volts, and you need 170-180 volts of spikes to equal a 110 house voltage.
While you are waiting... try something. There is still a lot to be learned here... about 90 years worth!

===

Hi all
I am trying something with a 2N3055,a JT Toroid with a secondary and a recharable AA 1.5 Volt Batt.
... voltage started at 1.10 and went to 1.20,sofar it's stayed at that voltage for almost five hour's....also when I hooked it up My led.s dimmed just a little it may not be anything great but could give more run time to the light's. Tec

So, it might be a self charger?
cool.

====
Hi morpher44,

Interestingly if you change from 2 CFLs in series to 1 CFL,
the amp draw doesn't really change .. only the brightness of the light.

D cells have a typical capacity of 20500 mAh ... so this should run
for almost 51 hours from a fresh battery ... probably less since
some threshold will be reached whereby the JT will no longer
oscillate.

D cells for using is a good idea, but not yet for testing!

This business of the way the amp draw doesn't change but the light does is a really important point.
I have been watching something like this for a while.
[plug for think tanks like ours:
I had just noticed it when altrez mentioned it then someone else added after I said I saw it too.]

This is the beginning of where it started to look like a pattern to me.
I cannot make the frequency = amps equation, but if you force the frequency and volts to stay the same, the amps will be drawn.
If you allow the frequency and the volts to go where they want, they behave like freq = amps.
If you add a series inductor to boost the volts, you can then add a draw of some kind, the frequency will diminish.

There is more to learn about this, but I believe it is the road to 90 years ago and beyond.
Great work everybody,


jeanna

[theremart]

Hooked... on the JT

I went to order the toroid at Allelectronics suppy = -8

I made it -12 lol.

at 1.5 V I am thinking the water battery or even the ground battery should power this up.

The question I have is why can't we get an off the shelf Toroid all wound up to work with our setups? Is the winding method that precise that it demands we wind our own?

Looking forward to 1.5 V lighting a florescent.

Mart

[Mark]

What you must have is the end of one wire connect to the beginning of the other.
It is a center tapped primary, and it is often the case that it is off center.
But for your first one, just do it and see.

Hi Jeanna

I'm not sure what you mean by the quote above. I have the start of my 3T and the end on 13T primaries going to positive the end of the 3T going to base w/resistor and start of my 13T going to collector. I assume thats correct because its working.

My circuit ran all night a total of 14.5 hours on the C cell. The light faded quite a bit but battery voltage was still running at just over .5 volts. I unhooked it and the battery is starting to come back some. My draw was around .180 so about .25 watts. I had an extra C cell so now I have a fwbr and cap hooked up and trying to recharge run down C cell.

Jeanna I think I have those same bulbs I bought them at Sams. Came 3 in a pack. The Sams near me carries 4 different types of bulbs. 2 of them I dont like, the flood light (green color) and a 3.5 watt bulb thats too whitish blue. The other 2 (1.5 watt) bulbs have a warmer yellowish color. The ones on my circuit are the small end bulbs. I have them in my chandalier and one of my ceiling fans. The ceiling fan light fixture has 4 bulbs in it and give decent light in my 11x14 room. (lights of america 19 leds) The other bulb I like also has 19 leds and good color just has a standard size end. I use 4 of those in a 1/2 bathroom and they give off decent light.

I did light up a gutted 14 watt cfl but the light wasn't very bright. I'll try a 10 watt when I can find one.

I also lit up one of the 1.5 watt leds on 2 hearing aid batteries! (very dim though)

Thanks for sharing Jeanna and Lidmotor

[morpher44]

in...

So morpher44... you in?
I need to start winding some toroids.
jeanna

Hi Jeanna,
This weekend I did want to focus more on the Joule Thief.
I looked around online -- briefly -- for a big ole toroid .. but
the ones I found cost a little too much for my budget.
Meanwhile, one of your comments reminded me that I needed
to use a variable resistor to the base of the transistor for
tuning purposes. Doh! I kept trying to just "get lucky" w/o
building in a way to tune.
Anyway, thanks for that. I had a variable resistor on there
a few months back ... but somehow forgot to do it again.

I made a video here proposing a simple circuit that some
might want to replicate ... if they are lazy like me and don't
want to wind a 240 plus turn coil. I've been winding all these
coils and I'm getting blisters.

YouTube - Joule Thief Circuit - Driving CFLs with 1.5v

I went to the TIP31 ... because I had a bunch of them and
because I wanted to experiment with higher voltages and currents.
On advantage to higher voltages here ... is that with those
ferro coils, you can get them way up in voltage ... and
your little transistor needs to be able to handle that.
Yes 2n3055 is good too ... although harder to plug in proto board.

I probably should have a big toroid in my arsenal. Searching!!!

Another thought I had was that "if" we could get the Joule Thief
to produce "exactly" 60 Hz (or 50Hz depending upon location),
then folks could experiment with driving traditional appliances
too. I.e. design the electronics to self-tune to exactly those
frequencies ... and have that be the design goal for resonance.
I found that my CFLs were happy, though, at about 278Hz.

One more discovery to pass along.
If I put my ammeter in the mix, its difficult to impossible to tune
and make the CFLs light. The circuitry in the ammeter alter
the situation to an unusable condition. Hence, measuring
current might be best done with an old fashioned analog ammeter.
So I switched to an analog ammeter and see I'm using about 400milliamps
or just under .6 watts with light that I would describe as "good enough
to read by". Interestingly if you change from 2 CFLs in series to 1 CFL,
the amp draw doesn't really change .. only the brightness of the light.

D cells have a typical capacity of 20500 mAh ... so this should run
for almost 51 hours from a fresh battery ... probably less since
some threshold will be reached whereby the JT will no longer
oscillate.

[jeanna]

I've never wrapped one of these before? Does it matter which direction you make each wrap. And how crutial are the amount of turns. What if your off by 1 or 2 turns on the primarys.

Hi Mark,

The true answer is there is only one way to know and that is to try it.

Now,
Each toroid is different and so is each transistor.
These 2 elements can determine how many turns of the primary wires you need to make.
The 22 or 25 ohm value of the resistors comes directly from the mix of the 2 elements.

What you must have is the end of one wire connect to the beginning of the other.
It is a center tapped primary, and it is often the case that it is off center.
But for your first one, just do it and see.

makezine has an excellent youtube video on it

Are you referring to the lights of america 19 leds bulb?

2 in series.
Any led that is intended to be used in a wall socket is heavily protected from surges. They have to be.
[The overvoltage that will do a led in is so little and the wall surges are frequently more than 10 times too much.]
This means that if you didn't modify it, you are doing well.

The led bulbs I'm using are unmoddified and still have the circuit boards and all inside.

Can you take a pic of the light? I am curious what is inside.
Mine have caps and I think supercaps because when the power goes off, the light lingers for 20 seconds or so.


====


@tec

@ jeanna
Ive done the joule T like you said Got 30 to run in series I had to use the tiny JT Goodnight night and bless us one and all. Tec

Yeay!!

[Mark]

6hrs 15 min now at .970 volts, going to let it run all night and see if the lights are still on when I get up in the morning.

[Mark]

I've never wrapped one of these before? Does it matter which direction you make each wrap. And how crutial are the amount of turns. What if your off by 1 or 2 turns on the primarys.

I threw one together and can light up 2 of the new led 1.2 watt bulbs in pararell. I only had a 25ohm resistor need to get a 22. Its been running for 5 hours on a C cell batttery now. Still going with running volts at .974.

The led bulbs I'm using are unmoddified and still have the circuit boards and all inside.

[jeanna]

I'm really not sure why that 1.2volt AA can do all this.

and



jeanna

ps,
It is the reason I was never stressed about proving ou n stuff like that.
It's in our hip pockets.
It's been there all along

[Lidmotor]

Charging a 12v SLA off the light

@ Jeanna & All
I got somewhat side tracked today replicating an idea that Mart Hale came up with for the one magnet no bearing Bedini. I posted a video of it over there on that thread if any one is interested. I did have some time to try something with the light today and it worked. On my 10 turn winding I put a bridge rectifier and charged up a small 12 volt SLA while the CFL was running. For some reason I am getting very high voltage at that point if I turn the rheostat up. To set the voltage right for charging I used a big cap and turned the dial up just enough to get the light on and have the cap sitting at 20 volts (on full brightnesss the voltage was up over 100 ). I let the system run for about an hour and it charged up my small 12v SLA just fine. I'm really not sure why that 1.2volt AA can do all this. There is some weird stuff going on in that big toroid.

Lidmotor

[jeanna]

@jeanna
...

The stacked toroid idea sounds good.
I would recommend larger gauge wire though.

In fact the one I used is a thicker wire. It was great to see lidmotor used the RS wire, but mine was 27awg which I used because the insulation is a bit thicker on the thicker wire, too.

... The toroid KEEPs
the magnetic field locked in the toroid ferro material ...

...
Also, if the transistor is one that can produce a large gain,

I am putting these 2 together because of the way I am using the basic joule thief circuit.
I am using it only as a source of pulses.
period.
So, the gain is irrelevant and the fact that the mag field stays mostly inside is helpful to loop and re-loop and switch back with every re-loop.
In the primary bifilar, which is center-tapped, there is a switching of the direction of the mag field inside the toroid. So the frequency of pulses can be regulated by the resistor at the base of the transistor. The higher the resistance, the higer the frequency..(and the voltage goes down as the frequency goes up.)

So for the JT circuit, I'm thinking that you can pick
your transistor and your coils so that they play well together --

Yes,I agree it is important, but IMHO the match comes from matching the toroid characteristics with the transistor's to get the effect you want.

I like all the LEDs light ... but I'm SHOCKED at how
expensive those super brite white LEDs are.
The price for these is too high
....So a nice low cost lighting system would use CFLs.

ebay--china 100/$8 including shipping. I missed these for a long time.
It helps when you pop them cuz the voltage spikes get too high.

I am floored that you can light a CFL with 1.5V.
...
. So you want Q very high ... inductances very high.

And maybe more things too.
I am very encouraged by lidmotor's replication and the ease he had in getting it to turn on and even that he can use that rheostat that lowers the voltage...I am wondering about the 2N3055. I have a lot of trouble with those things and I just bought 1 TIP3055 which is what lidmotor used, and what xee2 always uses. xee2 had earlier success than I did (but with a smaller fluoro tube.)

One light bulb that went off in my head re: the Hendershot material
is that he simply is wiring several inductors in series ... winding all
over his circuit ... to create a very large inductance.
So this approach of chaining several inductors down a network
of little tuned nodal areas -- ala Eric Dollard's analog computer approach ---
shows some real promise.
One thing that holds everyone up in that regard is they
want to do circuit analysis.

Who wants to do a circuit analysis???

I was pleased to see Eric's video. I became more confident in my intuition that it was the series part and not the transformer part that adds so much to this already pulsing system.

AND I have not yet managed to set anything up exactly like his with a cap crossing between the lines, but I want to try it.
[edit... I may have, and I don't realize it. I better study my drawings some more.]

Also, please let me repeat this.
If these things can be tested and shown at 1.2v, then increasing the output is easy.
If 1 cfl can light with 1.2v and you use 12 volts... I know it will not increase by 10 with all the components being the same, but I am guessing that the thicker version of this very big toroid will be able to supply much higher volts to the pulses... to be shown...

and silently, please think how obnoxious those sparks are. Imagine not being able to turn that off ever... not for me!

So morpher44... you in?

I need to start winding some toroids.

thank you,

jeanna

[morpher44]

Super inductance!!!

I was studying those wonderful physics lectures from prof Lewin at mit last summer, and according to what he said, this very big and powerful but short toroid should be able to make a much more powerful circuit if 2 (or 3 ) were stacked and then wound together as one.
It seems to be a function of the area of the shape defined by the wire. The shape of the 'surface' that the magnetic flux passes through looks like a spiral staircase when you coil the wire, but this helps me to see why my thicker toroid called the filter is the best one I have.

So, yes... stack those rings like griddlecakes and bring on the super joule thief...
I have been purposely working tiny but I am sure the increase is exponential.

Morpher44...you in?


jeanna

@jeanna

The Lewin stuff is great. I wish I had him when I went to college.
The physics teacher I had was more of a math teacher ...and just
wanted to do math all the time ... never getting the kids excited
with experimentation. In highschool, though, I had a Lewin-like
Earth Science professor who was amazing. A good teacher
can really bring out the best in you.

The stacked toroid idea sounds good.
I would recommend larger gauge wire though.
You would be surprised how much resistance is in that tiny
mag wire. You want to keep resistance down to an absolute
minimum if possible.

In my humble little JT/Aircoil experiments I'm amazed at the
distance inductive coupling can occur. The toroid KEEPs
the magnetic field locked in the toroid ferro material ... w/o
a chance to poke out and interact with the surrounding area.
This may be ideal for a transformer to reduce losses ...
because you cage up the magnetic fields and don't let them
"move" things out side.

But ... I'm thinking really we shouldn't think of the
bifilar toroid coil as a transformer in the JT. I'm thinking
really we have the transistor BASE side ... which is your
feedback coil -- like a tickler coil in the armstrong oscillator --
and we have the DRIVE coil ... which can have a super-duper
high voltage on it since the current is impeded.
The so called base tickler coil can be quite humble.
Also, if the transistor is one that can produce a large gain,
within its operating parameters -- picking the base resistor
to place you right in the middle of the linear current gain slope --
the tickler coils feedback to the base can be a very minimal
current. By winding it with a minimalistic number of turns,
just enough for that feedback --- and with a transistor
that has large gain -- the drive coil can be more BEEFY.
The next consideration, though, is whether you want
that drive coil to have its magnetic field locked up and
contained -- ala toroid -- or if you want it to BURST
out all over the room -- ala Dr. Stiffler ... OR do
you want that drive coil to be a primary to yet another
larger -- more turn -- secondary ala Tesla.
Or ... a less turn secondary so that you can have
high current, low voltage output. This would be the way
to go to run little motors say.

So for the JT circuit, I'm thinking that you can pick
your transistor and your coils so that they play well together --
depending upon what your trying to achieve.

I like all the LEDs light ... but I'm SHOCKED at how
expensive those super brite white LEDs are.
The price for these is too high.

CFLs, on the other hand, are getting real cheap.
There are sales ... and sometimes you can buy them for
as low as $1.00 USD.
Further, CFLs can be very very bright ... at the right frequency.

So a nice low cost lighting system would use CFLs.

I am floored that you can light a CFL with 1.5V.
But on the other hand, I am aware that coils and capacitors
can bring voltages up very high at resonance -- depending upon
Q. So you want Q very high ... inductances very high.

One light bulb that went off in my head re: the Hendershot material
is that he simply is wiring several inductors in series ... winding all
over his circuit ... to create a very large inductance.
So this approach of chaining several inductors down a network
of little tuned nodal areas -- ala Eric Dollard's analog computer approach ---
shows some real promise.
One thing that holds everyone up in that regard is they
want to do circuit analysis.

But I think Jeanna intuitive approach of plugging things in various
ways to see what works ... is the same approach I'm using ...
and is the same approach many inventors of the 18th/19th/early
20th century. I think what we LOST ... in the later 20th century
is that ... "lets just try things" attitude for invention. We got
all bogged down in the theory ... and equations ... and
lost the fun of experimentation. As you read the biographies
of various inventors, though, you find that they didn't
know theory ... they just used intuition ... and figured out
theory LATER -- or had to find those who could help them
with theory.