Big Joule Theif

[morpher44]

Honey Comb Coil in a Joule Thief?

Joule Thiefers, I wonder if using a honey comb coil
might be even EASIER for all concerned to wrap
in a Joule Thief.
Is there such a thing as a bifilar honey comb to create
a 1:1 ratio transformer?

Inductors

[Watson]

I wonder if diode will reduce the output since our kind of circuit do punish the transistor by forcing it to stop the current abruptly. I think adding diode will add leak point that reduce the power that we can use. I think we should AVOID using diode and protect the transistor another way.

Yeah, stop using the 2N3055, which can handle only 60 volts on the collector. Get some transistors made for this kind of work, the horizontal output transistor. Or if you have an old PC power supply laying around, use the big switching transistors (and the heatsink they're on) instead. Many of the PSes I've cannibalized use the 2SD2646, IIRC.

Then it might just be better to buy an already made power supply such as this one. For $6.50 plus S&h you get a PS that accepts 12V, 1A input and puts out 2000 volts at 10 mA, 30 kHz.

GAS TUBE SIGN POWER SUPPLY | AllElectronics.com

Or see my latest blog dated July 16 for info on building a 12V Fluorescent tube Inverter.

[Lidmotor]

Slayer's winning design----A five day test at sea

Guys we built this mainly as a homage to all, its Lid's slayers/rick's /ren's and John's AIO

In this AIO (switchable) We have a pulsed inverted joule thief , plus solid state, Fan and Lid's light. Schemo and video coming soon. We could operate all (Fan, light solid state charger) on under 1 amp. 500mA with good light and charge two 7aH batteries from one.

Imageshack - aio1
Imageshack - aio2
(full light with fan and charge 900mA

We could get it down to 100mA with no light on and a good charge (still 2 7aH batteries). Fan has good RPM

More details soon guys sorry about the short post.

Ash

Hi Ash,
I just got back from a five day sailing vacation to Catalina Island here in California. I just saw your post here and I was glad to see that round Fl in the your photo. On my trip I took a few of my light projects along to try them out and it was my "Halo Light" that really worked. WHY?? It lit up bright and gave out lots of light. I ran it every night to light up the boat and charged it with a little solar panel only a couple times. It just worked. On a 32' sailboat rolling around at an anchorage 30 miles from home for a week you need something that gets the job done and this thing did.
I'm anxious to see what you did in your design. Round is beautiful.

Lidmotor

[ashtweth]

Guys we built this mainly as a homage to all, its Lid's slayers/rick's /ren's and John's AIO

In this AIO (switchable) We have a pulsed inverted joule thief , plus solid state, Fan and Lid's light. Schemo and video coming soon. We could operate all (Fan, light solid state charger) on under 1 amp. 500mA with good light and charge two 7aH batteries from one.

Imageshack - aio1
Imageshack - aio2
(full light with fan and charge 900mA

We could get it down to 100mA with no light on and a good charge (still 2 7aH batteries). Fan has good RPM

More details soon guys sorry about the short post.

Ash

[sucahyo]

In the schematic, notice the 1N4005 protection diodes around the IRF640. But I think I would use some 3 amp diodes because the current through the ignition coil might be higher than the 1N4005s can handle. But come to think of it, you could put two or three 1N4005s in parallel and get the same result. They're cheaper and common as dog doo-doo.

I wonder if diode will reduce the output since our kind of circuit do punish the transistor by forcing it to stop the current abruptly. I think adding diode will add leak point that reduce the power that we can use. I think we should AVOID using diode and protect the transistor another way.

[morpher44]

Super Joule Thief Proposal - ala Hubbard

I had an interesting idea that someone out there might want to try.
Take a close look at the so called Hubbard coil.

What has bothered me for about 2 weeks now is this mysterious
2.8Ghz declared to be the "Natural Resonance Frequency of Magnetism"
that Hubbard derived all his frequencies and golden ratio geometry from.
So ... crack that ... I thought.

Today I had read up on EMR and PMR ... and WOW!
EMR has the electron spin frequency at 28Ghz if the electron
is in a field that is ONE Tesla.
Ok, so to get 2.8Ghz for the spin, you need 0.1 Tesla.
Did Hubbard have 0.1 Tesla in his coil? If yes, the electrons would
SPIN perfectly AROUND its periphery -- a most optimum situation.

So could a SUPER Joule Thief be constructed with a Hubbard
coil?
Hubbard claimed COP = 3 or so.

[Watson]

So the "big joule thief" is using 2n3055.
Question: Anyone try using an IGNITION coil or
flyback coil?
I guess one problem with ignition is that NEGATIVE is common
for primary and secondary and joule-thief is not happy with
that arrangement.
A flyback, on the other hand, might work -- especially if it doesn't
have a diode on the high-voltage lead -- like some of them do.

-- morpher44

Neither terminal is grounded if you're using a standard 'Kettering' type ignition coil - the one used with points and condenser.

Here's another ignition coil circuit. It gives you an idea of the internal coil wiring. If in doubt, check the windings and windings to case ground with your ohmmeter.

Read the text carefully - it's very informative.

Ignition Coil High Voltage Circuit

And here's how to drive a flyback coil from an old TV or monitor.

Sam Goldwasser's Flyback Transformer Circuit

There is nothing that says that you must use two transistors. You can "halve" the circuit and turn it into a conventional Joule thief circuit. But two transistors gives you more POWER!

In order to view these ASCII schematics you must use Courier or other monospaced font.

[Watson]

So the "big joule thief" is using 2n3055.
Question: Anyone try using an IGNITION coil or
flyback coil?
I guess one problem with ignition is that NEGATIVE is common
for primary and secondary and joule-thief is not happy with
that arrangement.
A flyback, on the other hand, might work -- especially if it doesn't
have a diode on the high-voltage lead -- like some of them do.

-- morpher44

Here's a schematic of an ignition coil driver circuit.

One thing that's great about using a power MOSFET is you can get them in a lot higher voltage. The ones that can handle 200 or more volts and tens of amps are not much more expensive than a 2N3055. But as with a 2N3055, you can still zap them, and then they're good for a paperweight or doorstop.

In the schematic, notice the 1N4005 protection diodes around the IRF640. But I think I would use some 3 amp diodes because the current through the ignition coil might be higher than the 1N4005s can handle. But come to think of it, you could put two or three 1N4005s in parallel and get the same result. They're cheaper and common as dog doo-doo.

[teslaproject]

@ Slayer

Thanks for the reply Slayer, I appreciate you taking the time.

Kind regards.

TP

[slayer007]

Slayer!! where art thou?

Really fantastic input on the thread (also Lidmotor)...therefore thanks. Couple queries whether some clever people might be able to help me out.

Firstly, Slayer you mentioned that ferrite toroids have given you the best result. Whats the heaviest current and voltage you have run through the big coil (Im actually having trouble keeping up which version you are up to ) and is your ferrite generating considerable heat? I was thinking of giving MPP cores a go. What do you think?

Have you (or anyone else here) attempted a bifilar/quad/single/other wound 'standard' coil but make either the primary/pickup (or both) 'directional', rather than laying the magnet wire back and forth in a standard overlay?

I was wondering whether you came up with a good idea other than mechanical (I'm trying to stick to 'solid state') for high speed switching. I'm attempting a number of things but seem to keep burning out my transistors or 555 circuits on the longer term 'heavy duty' voltage tests. Too much power perhaps
Either way, I would love something more robust than these heat prone bits of silicon.

IF I wanted to protect the circuit from a either spikes/surges and keep the voltage steady at one level when the 'input' might typically be ranging beyond the realms of a standard resistor, whats a more dynamic buffer I could place on the front end? I was toying with the idea of MOV's but they will be no good in the long run because they have a limited lifespan.

Thanks again.

TP

Sorry teslaproject I didn't see your post till now.

The most voltage I sent through the ferrite toroid JT was 12 v the most current was close to around 1 amp.

It didn't seem to run much better with the higher voltage and current.
At least with the coil I was using.

The ferrite toride does not get hot at all even at 12v.

For the best results or preformance I was using two or three 1.5v AA batterys.

[morpher44]

ignition coil or flyback?

So the "big joule thief" is using 2n3055.
Question: Anyone try using an IGNITION coil or
flyback coil?
I guess one problem with ignition is that NEGATIVE is common
for primary and secondary and joule-thief is not happy with
that arrangement.
A flyback, on the other hand, might work -- especially if it doesn't
have a diode on the high-voltage lead -- like some of them do.

-- morpher44

[teslaproject]

@Watson

Thanks Watson for the informative reply.

I had to chuckle at your 'science fiction' comment regarding my resistance desires. Perhaps I hadnt worded it correctly and therefore deserved a good clip behind the ear.

Anyway, thanks again, some excellent suggestions regarding 'precautions' I obviously havent been implementing. I might take you up on that transistor offer also.

Kindest

TP

[Watson]

"beyond the realms of a standard resistor"

Slayer!! where art thou?

[snip]

I was wondering whether you came up with a good idea other than mechanical (I'm trying to stick to 'solid state') for high speed switching. I'm attempting a number of things but seem to keep burning out my transistors or 555 circuits on the longer term 'heavy duty' voltage tests. Too much power perhaps
Either way, I would love something more robust than these heat prone bits of silicon.

IF I wanted to protect the circuit from a either spikes/surges and keep the voltage steady at one level when the 'input' might typically be ranging beyond the realms of a standard resistor, whats a more dynamic buffer I could place on the front end? I was toying with the idea of MOV's but they will be no good in the long run because they have a limited lifespan.

Thanks again.

TP

I use some high speed rectifier diodes I got from a PC power supply to clamp the outputs to the power supply + and -. For instance, a 555 should have a HS diode from pin 3 to ground, cathode to pin 3, and another from pin 3 to +, cathode to +. If the pin 3 voltage goes 0.6V higher or lower than the + or -, the diode conducts and clamps it.

The base of the power transistor should have a HS diode from base to emitter, cathode to base. This prevents the base from going negative in relation to the emitter.

If the voltages at the collector of the power transistor are going to be high, then it should not be a 2N3055. Use a horizontal output transistor that can handle several hundred volts. Or else clamp the 2N3055 with zener diodes so that it can't et higher than 60 volts.

This type of thing where a power transistor has to handle high voltage and current is known as SOA, safe operating area. The datasheet shows where the limits are. Exceeding these limits usually lets the smoke out.

For many years, you have turned on your TV set and watched a picture that was made by 30 thousand volts on a CRT tube. This was generated by a flyback switched by a power transistor. There is no secret to doing it right. Just don't do it wrong.

Also it is mandatory that the 555 have a 0.1uF bypass cap across the power pins, close as possible to the chip. There is a high current during switching and this cap is needed to filter it out. If this isn't taken care of, the 555 output can do some rather strange things.

The term "beyond the realms of a standard resistor" that I saw above is not science, it's science fiction. Your resistor is a resistance, and has a maximum power limit, which has to be observed. There's no magic to that.

It's not that the silicon isn't robust, it's that the circuit is subjecting the silicon to voltages and currents outside its maximum limits. If you need twice as much current through a transistor, perhaps you should be using two transistors in parallel, to split the current between them.

If you can't tame your circuit, I've got a box of 100 2N3055s here that I could sell to you for maybe $50, that should take care of your power transistor needs for awhile.

[lanenal]

After a week or so Paypal made a decision in my favor, so I logged on to PP and found $30 credited to my account.

I found this Chinese Co. that sells cheap LEDs for throwies, but you have ro buy a lot of them. Click on fhe x-type box. Most other types, too.

Power LEDs

So it ends well -- at least not bad .

[teslaproject]

MPP and a million other little things

@ Guruji

Running at 3v with two 1.5v batterys.

The rectified voltage across the sedondary going into a small capacitor is around 380 volts.

It should take close to 300v to run a CFL.

This was tried with the basic JT circuit with only one transistor.

One thing you could also try is to take out the 1k resistor and put a 5k pot in it's place.
That way you can go below 1k if you need to for tunning.

Also for best results the toroid should be a ferrite toroid.

P.S. I used a rectifier across the secondary because the AC comming off the secondary isn't true AC it might not be right at 60mhz.
So my meter wouldnt read it correctly.

Slayer!! where art thou?

Really fantastic input on the thread (also Lidmotor)...therefore thanks. Couple queries whether some clever people might be able to help me out.

Firstly, Slayer you mentioned that ferrite toroids have given you the best result. Whats the heaviest current and voltage you have run through the big coil (Im actually having trouble keeping up which version you are up to ) and is your ferrite generating considerable heat? I was thinking of giving MPP cores a go. What do you think?

Have you (or anyone else here) attempted a bifilar/quad/single/other wound 'standard' coil but make either the primary/pickup (or both) 'directional', rather than laying the magnet wire back and forth in a standard overlay?

I was wondering whether you came up with a good idea other than mechanical (I'm trying to stick to 'solid state') for high speed switching. I'm attempting a number of things but seem to keep burning out my transistors or 555 circuits on the longer term 'heavy duty' voltage tests. Too much power perhaps
Either way, I would love something more robust than these heat prone bits of silicon.

IF I wanted to protect the circuit from a either spikes/surges and keep the voltage steady at one level when the 'input' might typically be ranging beyond the realms of a standard resistor, whats a more dynamic buffer I could place on the front end? I was toying with the idea of MOV's but they will be no good in the long run because they have a limited lifespan.

Thanks again.

TP