Jun 16, 2012

Making a Joule Thief Charger Circuit

Continued from the previous article

Here is the joule thief used to test my button cells out after I successfully recharged them.

Here's my battery charger joule thief all installed into a circuit box.

Below is the front view, notice the 12 volts green light at the front panel that is modified to house a neon globe inside it.

While looking through my electronics parts draws I found another 7 or 8 more button batteries and some had charge still on them and some I recharged successfully and 2 couldn’t be charged anyhow a few small ones took a little zapping to get them to take a charge so maybe if they are got some sulphate in them if not then why did it take so long to get them to except a charge.?

Yesterday I found another button battery cell and after a few seconds charging it started to get warm so I stopped it before it got to hot otherwise it would explode
Now I have 24 button cell batteries.

 Below is a picture of the bright white with pink flash of a photo flash capacitor neon circuit test
The outputs of my battery charger joule thief run to the 330 volts photo flash capacitor, and across its leads I placed a neon bulb, loosely so every so long you hear a spark jump but cant see it unless the neon leg is getting pulled in by the high voltage attraction then a short time after that the flash occurs, so if the neon is rated at 90 volts then the cap charge may be the same till it charges up more I filmed this then put it onto my pc and then I had to slow it down so much to capture the super fast flash then pause it then take a picture of it to load back onto my pc.

 Here's what it looks like when it aint flashing

ive soldered one of the neon legs to one of the high voltage 330 volts photo flash capacitor legs and the other neon leg I had it just nearly touching the other capacitor terminal but I didn’t get the arc and flash after it like I got before .
next test was running one of  my jumper leads to a folded peace of aluminum foil and back to the hv capacitor and the other lead I ran to the other hv cap leg to a pencil  and when the neon at the 330 volts 2200uf photo flash cap come on I gently touched the lead pencil tip to the foil edge where its folded and I got flashes etc and tiny holes melted through it and when I tried to unfold the foil it was stuck like it was spot welded , and there sure was this horrible smell made by spot welding aluminum foil.

Next experiment was running the leads from the capacitor to 2x 316 grade stainless steel bolts in a jar of tap water to create any electrolysis effects and I got very fine misty vapours come up to the top of the water and no bubbles and even when the bolts slipped together I was still getting it and even when they are apart to, and the further apart the bolts are the lower the frequency of sound is emitted from the battery charger joule thief and the closer the bolt electrodes are the higher the frequency of the sound the circuit emits gets ,

My watch battery I charged before has stopped I guess the charge didn’t last long enough, I charged it on the 24th may 2010 and it lasted till today at 8past 8 , either in the morning or last night , so it lasted a week only  im gona try re charging it again . Also today at the shops I went to the place they do watches and shoes and there was thousands of button cells inside this Perspex box even large ones next to it , but no one was there to answer my questions about them .
Ive recharged my watch battery again today but I was charging it the wrong way around so I got it right after that now my watch is running again so ill see how long its lasts for this time.
Today is June the 1st 2010 so lets see if it lasts a week, also if it lasts longer then it may be the reverse charging that contributed to the extra running time before I charged it right

 I scored a packet of free button cell batteries including large lithium cells and a few small 12 volts pencil batteries for free from 2 places in Armadale that also replace watch batteries, they have thousands of them and I counted what they gave me in total I had 85 batteries mostly button cells and ive successfully recharged them all except some lithium large button cells so also my 25 batteries I had it makes 110 batteries in total including the few large lithium cells that couldn’t be charged or revived at all  some of the batteries still had power in them to .

Half the button cell batteries couldn’t maintain there charge and went flat overnight
So I recharged them again so ill soon check to see if they hold there charge this time  also I dismantled a lithium battery it has a thin white membrane that reacts with water and foams up and also a thicker black membrane  with what looks like very thin silver wire looking cross patterns on its sides  and when I dismantled  that battery it sure stunk  inside it to
Today Tuesday June 8th 2010 I decided to use my 4 rechargeable batteries from my camera to power my radiant joule thief battery charger, to restore my 12 volts 7.2 amps an hour 20 hour sla battery, and so after an hour or so I used a pair of tweezers to short out the battery terminal and I got some sparks where I didn’t before that,  so I left the joule thief battery charger running all day today and now when I plug in the battery charger the red  charging led light doesn’t go to green after a short charge that’s trickle charge, when green its still on red so im looking forward to how it goes when fully charged then in the green when its charged so if I get plenty sparks when I short the battery terminals then I can say the battery has been revived .

My 12 volts sla battery appears to still have a tiny bit of sulphate on its plates as its voltage dropped 0.7 volts over night so when the battery is low and needs topping up ill put it back on the joule thief circuit.


  1. is it possible to boost the voltage and current from other than a battery source using joul thief ckt?if i have a 1.5 v dc using a transformer is it possible to boost? or this crkt is just for battery ? can you pls explain the working principle?

  2. Yes it can be. It's basically done with the help of the coil and the induced frequency. During the zero pulse sections of the frequency, the coil kicks stepped-up voltage to the LED and makes it glow even at lower supply voltages.

    1. thanks for the reply.i have one more doubt. i have gone thru many joul thief ckts. but no: of turns shown in varios ckt are different. what is the relation ship with no: of turns to the o/p voltage & current?can you post an ideal circuit with maximum efieiency and explanation pls

    2. I don't know the exact data, you may try the number of turns by trial and error, the one which produces brightest of lights and consumes minimum current is the right one.

    3. In Joule Thief, the primary coil (the one that stores and carries almost all currents is the one that is connected from the positive rail to the transistor's collector) is one of the most important part besides the transistor. As you might know, if the resistance of the primary wire is 1 ohm, and 1V across it, and 1A of current is flowing, then the wire is dissipating 1W of energy. Now, a 1.5V battery current can reach upto a half amp or more. Therefore, the resistance of this primary coil must be low to let the current flow freely to your LED instead of being wasted in higer resistance. In order to make the resistance low in your primary coil, the winding wire can be then chosen wisely. For instance, according to the data sheet, copper wire of 30AWG has 0.104 ohm/ft. resistnace whereas 24AWG has 0.0257 ohm/ft resistance. And the last part to determine how many turns, you have to think about the amount of energy you want to store in your primary coil (or inductor). The amount of energy stored in this primary coil is equal to the one half of the inductance (L) times the current (I) squared, that is L/2*I^2. Now, the supply voltage remains the same across the Joule Thief coil, you can increase the energy stored by using fewer turns.
      In a nutshell, doubling the inductance (that is, higher number of turns), doubles the energy stored but lowering the inductance (that is lowering the number of turns) by half, the resistance is lowered by half, that means doubling the current, the energy quadruples. Hope, this will let you determine the number of turns depending on the demand of your load (wheather you want to light a single typical 5mm LED that needs only 20mA current or a 1W LED that requires 350mA current).

  3. Can I use iron powdered toroid, instead of ferrite toroid?, b'cause it's hard to get 1" diameter toroid here.

  4. Iron powder toroid's can be used, and so can air core coils. They don't have high permeability, so they need more turns to have the same inductance as ferrite. 5 meters or 16 feet of telephone wire wound around a AA cell will work, but it's much larger and has a higher DC resistance. Some powdered iron cores are big, too, and more difficult to wind, so the air core coil may be a quicker way to get the JT running.


Swagatam Majumdar
Swagatam MajumdarHi Friends, Welcome to my site, a place where you will discover a massive collection of electronic circuit ideas, mostly requested by the dedicated readers and exclusively designed by me for their customized application needs. I have posted more than 1100 circuit designs in this site, if you have a personalized circuit requirement you may feel free to request it through the comment box, if it seems feasible to me then surely you may find it published here with your credentials attached in the post, thanks and please keep reading

Copyright © 2014 Electronic Circuit Projects • All Rights Reserved.
back to top