In the course of this year i make by myself some JT and did some experiments making all kind of variations, having the disadvantage/advantage of my poor knowledge on all the electronics issues i did almost all variations that i have in hand (i more like a informatics dude, so let say that i did ordered and random single and multiple permutations) i find that sometimes shorting together the inductions/coils by a wire or by some kinds of capacitors and coils for a limited time (some components short time, some long time) make the light brighter than when i turned the circuit on.
Also i ask myself, if the JT circuit work like i have read all around the web, it's flickering thousend of times a second (that is best for the wellness and ocular commodity for some kind of fly and birds), and everytime is on (and also off) the base and emitter both are wasting too much amperage, so if we only use the circuit by another regulating circuit that make it turn on and off , like 100 Hz (flickerless), it will be brighter and last longer, theoretically very very longer ...

So, what did you think and how to try to make it?

 

In the course of this year i make by myself some JT and did some experiments making all kind of variations, having the disadvantage/advantage of my poor knowledge on all the electronics issues i did almost all variations that i have in hand (i more like a informatics dude, so let say that i did ordered and random single and multiple permutations) i find that sometimes shorting together the inductions/coils by a wire or by some kinds of capacitors and coils for a limited time (some components short time, some long time) make the light brighter than when i turned the circuit on.
Also i ask myself, if the JT circuit work like i have read all around the web, it's flickering thousend of times a second (that is best for the wellness and ocular commodity for some kind of fly and birds), and everytime is on (and also off) the base and emitter both are wasting too much amperage, so if we only use the circuit by another regulating circuit that make it turn on and off , like 100 Hz (flickerless), it will be brighter and last longer, theoretically very very longer ...

So, what did you think and how to try to make it?

What is your point ?? what do you mean by another regulating circuit ? control this circuit from another ?

If you want higher efficiency you might check out the boost converter using power mosfet N channel. Remember that according your load demand, you have to use a stronger source. You might build 98% efficiency joule thief, but this doesn't mean that you can step up from 2V to 12V 3A, you are always limited to the amount of current you can draw from the source (and it's internal resistance).

If you want your devices to last longer, is a good way to oversize a little, keep use of basic electrical and electronic protections (current limiting resistors, capacitors, inductors, flywheel diodes, forward diodes), respect the specs of the manufacture, and eliminate as much as you can the transient effects.

 

I don't really a english speaker, maybe i don't make my point clear, i mean that in my post i wrote about two things, one of them is the absurd fact that some kind of short-circuit or cut-off of this kind of circuit boost the output for a while, this point i can't try to explain clearly because is over my understanding, but the second point, and is the one that i dont know how to achieve but for most of you is a piece of cake, is the other fact that this kind of circuit is flickering at a nonsense velocity of 1k to 40k a second, and you only need 50 to 100 cycles a second so you are wasting from 90 to 99 porcent of the power in something your senses dont need, so if you add to this circuit another mosfet that shut it down completely for 0.05 secs every cycle, almost without using current, this circuit will last 90 times or more... also the light i think will be stronger.

 

Also, i get what you try to explain, but for my is like an adult speaking with a toddler, some sketch will be helpfull and easier for understand.

What is your point ?? what do you mean by another regulating circuit ? control this circuit from another ?

If you want higher efficiency you might check out the boost converter using power mosfet N channel. Remember that according your load demand, you have to use a stronger source. You might build 98% efficiency joule thief, but this doesn't mean that you can step up from 2V to 12V 3A, you are always limited to the amount of current you can draw from the source (and it's internal resistance).

If you want your devices to last longer, is a good way to oversize a little, keep use of basic electrical and electronic protections (current limiting resistors, capacitors, inductors, flywheel diodes, forward diodes), respect the specs of the manufacture, and eliminate as much as you can the transient effects.

 

Yes you have a point, just bear in mind that in analog electronics everything comes down to RMS power or average power. Don't compare this with computing cycles of a CPU wasting resources of the clocked speed.

Let's say i'm switching an LED 50% duty cycle, i will deliver half of the power so the brightness will be also half.
If you talking about the brightness of the human eye can not detect from 90% to 100%, yes, this might be waste of energy. But believe me, the energy problem of the world is not here.

Now please help me seriously by answering this: you computer guy.
https://uk.mathworks.com/matlabcentral/answers/446711-float-32-bit-number-from-2-modbus-registers

 

Its the avdrage that counts for illumination above the flicker frequency. For slower pulses but with sufficient width you can go by peak power -but don't try to read b y a flashing LED.

 

I don't really a english speaker, maybe i don't make my point clear, i mean that in my post i wrote about two things, one of them is the absurd fact that some kind of short-circuit or cut-off of this kind of circuit boost the output for a while, this point i can't try to explain clearly because is over my understanding, but the second point, and is the one that i dont know how to achieve but for most of you is a piece of cake, is the other fact that this kind of circuit is flickering at a nonsense velocity of 1k to 40k a second, and you only need 50 to 100 cycles a second so you are wasting from 90 to 99 porcent of the power in something your senses dont need, so if you add to this circuit another mosfet that shut it down completely for 0.05 secs every cycle, almost without using current, this circuit will last 90 times or more... also the light i think will be stronger.

You have to keep in mind the goal of the circuit is not to be viewed, it is to operate at the lowest possible voltage and extract as much energy as possible from the cell. It is a demonstration, not a practical circuit. The frequency is not chosen for visual reasons, and the goal isn’t necessarily longest run time but most complete use of the cell.

Also, every component you add will use more power in some way Nd so you are unlikely to make the cell last longer by adding things to the circuit. If the goal was for visual utility, something like the (now defunct) LM3909 flasher IC, that charged up a capacitor with a very low voltage through a charge pump and discharged it into an LED would be more useful. You could build that circuit with discrete components.

 

i really glad for all the answers, but what about that short circuit thing, if i don't wrong the case is if you short base and emiter coils with a small coil o a 47nf, for some seconds, you will have for a while a brighter light, what's that about?

 

what's that about?

You are probably increasing the duty cycle, so the LED will appear brighter. The down-side is that the cell will have a shorter life.

 

You are probably increasing the duty cycle, so the LED will appear brighter. The down-side is that the cell will have a shorter life.

I think it will also not operate at as low a voltage, which is part of the goal of the original circuit.

 

i will need to do a lot of internet diging to take adventage of all your comments...


meanwhile, and like i say, because i a true rookie, i want to try and see for myself what i get from cutting off the circuit, limiting his flickering to 50 or 100 times a second, did some of you can make a sketch for that?


thanks a lot...

PS. perhaps, i also trying in the same funny way to make "better homebrew batteries" and i such naive to want to believe in zero point energy and other almost nonsense ideas, etc. (blame the internet and the lack of knowledge, o maybe i'm victim of the only real conspiration to make the masses to believe in everything that is not logicaly factible, whatever..).

 

The frequency of a joule thief does not affect its brightness.
A 50kHz Joule Thief has a certain brightness before the battery voltage drops. If you turn it on and off at 100Hz with a circuit then you will be dimming the LED at the ratio of the on/off times.

 

exactly that i want to look with my eyes!!! but cuting out the power suply by some circuit that haven't consumption of current from the main power source (the weak to almost dead batteries)
o in place (my other wrong idea that i want to try) turning on-off by external power source a bridge between both coils extremes via a small cap o small inductor in such a way that (for whatever reason, duty cycle?) make the led brighter for a while (from my simple point off view is produced from voltage spikes caused by some back EMF effect that i agree to alec_t that with my missuse short the batt life, but with the right use with short pulses and with much help from the placebo effect you can get some Bedinni, self recharge, Zero point back in time reversal current effect, maybe)

Also everybody including my brother say the same, that flickering less = bright less, but i think that if you take the same bright light point and turn it on and off at a speed that your eye don't see the flickering, will be the same if you turn it on 50 times or 50k times, but if the off time make it brighter for the only reason that you are less stressing the batt, it will last longer and light brighter and if you add a back emf tesla something effect, that make the dipole bring you back more voltage spikes (like tesla's spark ideas, and maybe the JT is all about that, o the opposite the spark is a sort of rudimentary strong and short cascade effect, i sure you understand that better than me).

 

You want a 555 timer driving a led......with adjustable frequency and adjustable duty cycle.

This will show what you want to see.

A jewel thief just produces light, which would have been thrown away.

 

Yes you have a point, just bear in mind that in analog electronics everything comes down to RMS power or average power. Don't compare this with computing cycles of a CPU wasting resources of the clocked speed.

Let's say i'm switching an LED 50% duty cycle, i will deliver half of the power so the brightness will be also half.
If you talking about the brightness of the human eye can not detect from 90% to 100%, yes, this might be waste of energy. But believe me, the energy problem of the world is not here.

Now please help me seriously by answering this: you computer guy.
https://uk.mathworks.com/matlabcentral/answers/446711-float-32-bit-number-from-2-modbus-registers

I can help you with graphic design, drawing, 3d art, image fixing, computer repair, and... all kind of cooking recipes... all self learned
also i can help you with spanish, hebrew, and some aramaic. My only real and very long vocation for years, was very deep talmudic diving (this reference is not my invention, diving in the sea of Talmud, is a very old terminology, but sounds something alike surfing the web ).
i have many other topics of interest as a hobbist... amateur...., nothing too much serious.
And too much ideas, somewhat fine, somewhat weird, almost everytime before somebody else think and make them (indeed i never make an affort to DIY so...) and most important almost all of them don't success, good for me that i don't have wasted my time on them .

 

i

You want a 555 timer driving a led......with adjustable frequency and adjustable duty cycle.

This will show what you want to see.

A jewel thief just produces light, which would have been thrown away.

sorry my ignorance, the 555 can be set to switch by 50 milisec ? how i achieve that? I really new in electronics.

 

A used battery cell has only a small certain amount of power remaining. If the Joule thief causes the LED to be bright then the remaining battery power will be used up quickly. If the Joule thief causes the LED to be dim then the remaining battery power will be used up slowly.

A 555 can be used as an oscillator. It is fairly old so its current is fairly high and it has a 4.5V minimum battery voltage. It is simple to make it produce narrow or wide pulses to drive an LED dim or bright.

50kHz is a good switching speed then the transformer in a Joule Thief can be small and the capacitor in a 555 circuit can also be small.

 

My suggestion about the timer was so you could observe the difference of how frequency and duty cycle effect what you see from the LED.

So when you're tinkering with the jewel thief led, you will know what to expect. Think of it as a simulator.

A jewel thief can be customized for different things. Like the brightest light or the longest life.

Such a timer circuit will be very handy for future projects.

 

Did you follow the thread?, I wanna 100hz.. don't mather worse or better i want to try and find out, but i dunno have a clue how to do it, i need a Scheme D finally i remember that word).

Like i say i'm a true rookie, i achieved to make some regular JT, but strugled to make something else, for example i waste a lot of nights trying to make, in a simulator perhaps, a tesla battery switching circuit, with NPNs PNPs relays etc. finally i achieved that goal but i don't trust the simulator.. i really affraid to make it with actual parts, and blow up many components and stuff..
also in those days (my first steps in electronics) i found a bug in cocodrile (old simple circuit simulator) that seems to generate from a coiled circuit some crazy values that start from Zero and grow up, like a EMF pulsar, back and forth to a voltage and amperage (they go up and down not at the same time) to the point that the program can't give a read and make the computer slow down from overthreading... weird!!! for a while, like a day, i believe that i found the infamous Zero point energy, hehehe.
That remember me that i try to make that circuit, with very simple parts and found some interesting real weird fact about back EMF....

A used battery cell has only a small certain amount of power remaining. If the Joule thief causes the LED to be bright then the remaining battery power will be used up quickly. If the Joule thief causes the LED to be dim then the remaining battery power will be used up slowly.

A 555 can be used as an oscillator. It is fairly old so its current is fairly high and it has a 4.5V minimum battery voltage. It is simple to make it produce narrow or wide pulses to drive an LED dim or bright.

50kHz is a good switching speed then the transformer in a Joule Thief can be small and the capacitor in a 555 circuit can also be small.

 

i


sorry my ignorance, the 555 can be set to switch by 50 milisec ? how i achieve that? I really new in electronics.

You can work with Electrodroid app on android, there are many calculation modules inside you will play a lot, also a module for 555 timer to set duty cycle and frequency. Of course you can always take the more engineering approach and study the datasheet of the 555 timer you can find in your local store, and do a little experiment on your breadboard. Max voltage 18, I recommend run it on 15V or lower.