Joule Thief

[electricme]

@MK1
Bingo, you got it, this is what I have been saying for some months now, but this refers only to those who want grunt output, not to those who are going for economy with their JTs.

That is a direction i like !

Things to think about , First the cap charge time , someone told me they change depending on the overall charge percentage , if that is true then we can take advantage of it , lets say it charges faster the first 10% , then you get a one rated at 1000v the 10% faster means 100v after that it still charges then but slower.
Also recover time a quick sparks and the time it take to get it back up.

I think the toroid selection is really important it need high flux properties , and smaller then bigger , the freq is a major issue since more cycles is more power for less, and believe me i wish you all a good toroid like my green one inch,
and believe me there is room for more turns and pick up coils , i could go 10 to 12 on each side , all of of those charge caps fast , a 470uf 200 v get over 35 v in seconds , a 10uf would get 100v instantly , no kidding , i connected 4 of those cap and scared my self sparking caps


Actually, caps (electrolytic's) recharge up to about 80% fast, then they soak up the rest in a graduated time scale depending on how much voltage/current is pushed at them.

So the trick is in timing arrangements, does one need the "fully" charged cap? probably not, if you could trade time to charge for time to use the accumulated charge.

Having a adjustable circuit to control this helps.

Picture 8 10uf caps rated at 6.5volts, all setup in a line, then you put the pos and neg lead on the 1st cap, charge it for 1 second, move the leads to the 2nd then 3rd etc etc.
Now connect all the negs together, now connect all the Pos together, then bridge out the Pos and Neg leads, POW.

OK in practice you would need a mosfet in series with each cap, then connect all the mosfets triggers together, then you got POW.

In reality, a timing circuit is used to wiz around to recharge and discharge all the caps, then you got POW, POW, POW, POW
A train of high energy DC Pulses

A bit like a rectifier, rectifying 50Hz, ac to 25 pulses of DC if 1 1N4004 diode is used.

BUT, to do this, you need to take a feed directly from the Pos 1.5v battery to control the pulses.
You need to also take a feed directly to the Torid coils.
You need several in dependant torids or coils, there are a number of configurations, to get a result.

Bill and someone else (forgive me I have forgotten)  way back was charging his special battery's/caps, and when he put them all in parallel, and bridged out the leads, BOOOM.

Yep the JT has a fair bit of potential.

@TheNOP,
Thank you also for the DSN to JPG tips, I just love making circuits, its yummy stuff.

jim   

[nievesoliveras]

Jesus,

I am wondering if the arrows, or some of them are going in the wrong direction?
Aren't the 2 AC points of the bridge for ac INput? If so, is the load pulling from the toroid before it gets to the bridge?

Also, Isn't the neg on the bridge supposed to be coming out of the bridge?

This drawing and the instructions have me very confused.

Today, I used a diamond shaped bridge, made as Jim described a while back. It works really well. Since I couldn't figure what Mark said in the instructions and since I want more voltage if possible,

I put all the pickup wires from the MK2 in series, and ran them through this bridge into a cap.

The cap was able to collect 82 volts from this bridge.

The scope says the wires are oscillating with 109vac. (remember that nice wave? that one.)

My only problem is that the string of leds doesn't light. I remember someone told Bill to short out the plug of his led lights or something. I have tried several things and haven't got it to light yet.   

Bill do you remember what you did? Or does someone have some advice on this?

thank you,

jeanna

(I checked. The string does turn on when plugged into the wall.)

The Rectifier bridge has two ac as input as you said and two outputs. one is negative dc and the other is positive dc.

Jesus

[TheNOP]

a normal light bulb is not much different then a led, only the way to make the light is.

http://www.youtube.com/watch?v=7NUbsQt-G9U
at around 33 minutes till the end.
notice how long the light stay on for the bulb.
now, ask yourself, what if this was pulsed instead of being on all the time.
also compare to how long the light stay on for the flash.

in the last experiment, ask yourself how many amps was flowing trough that wire in a fraction of second.
that wire would probably had be able to witstand 450 joules second or more.
450 joules second = .450 joules per microsecond.
but the time frame was: 450 joules in less that a microsecond...
can we do more work with same energy ? 

an other example is IST video about lighting an unmodified cfl with a jt and a cap.
it work because the cfl transistors are not triggered until a certain voltage is reached.
with a proper delay mechanism, one could use a normal light bulb to get to the same result.

being able to discharge 60 watts of power, be it for a picoseconds, 26 times per second, should be enough for our eyes to be tricked.
flickering would be the only thing to hint us about the faq it is not always on.
well, due to the resistive nature of a light bulb, it might actually take more then 26 times per second.
but that give the basic idea of how it should be possible.

[electricme]

@TheNOP

What you are saying is correct, this is the way we can save 1.5v energy in our batteries.

The same principle is used in the motion picture industry, (not to save batteries).
I once was a Assistant Projectionist at the Huon/Stewart theatre in Lae in New Guinea (during my Hi school days and early working life).
The Gaumont Galee 35mm projectors had a frame rate of 24 or 26 frames per second, this alowed the scenes to be seen as a flowing medium without any flicker, but those people who were sensitive to flicking lights did say say could notice a flicker, hence the words "I'm off to the flicks tonight".

As you mentioned, the LED "appears" to stay on all the time, but in fact it is pulsing at a very high rate.
If you wanted to save more 1.5v DC from your batteries, then lower the hi frequencies to a much lower ON to Off time ratio, I believe the times could be trippled or doubled in battery life.

I asked way back if anyone had a circuit that could pulse the JT to 50 Hz, and silly me didn't write down where the answer was posted, dumb uh.

jim

[electricme]

@all
Here is my first efforts using TinyCAD to make a circuit diagram.
I need to give arknowledgement to the original circuit maker as I used his to make this one so I wrote the thanks inside the png file itself.

My thanks to those who sent in suggestions on how to convert and attatch to the forum pages.
TA

jim