Let's Make Robots!

Joule Thief

There is not much to document in this project. I built a Joule Thief....who didn't ;-)

Mine is powered by a LR44 button cell (1.5V) and it light up a white LED which usually needs a minimum of 3.3 voltes to light up.

I just built it to test how long the LED can light with this one battery (it might be interesting for throwies). 

Will update the blog when the battery is completely depleted.

Power on time: January 14, 2012, 3:30PM

DIY coil, double side tape to attach the battery.

Transistor soldered directly on the battery to make the whole thing a bit stable.

I am not posting the schematics here since everything is already available in the famous internet but if you want to know why this is working then I recommend this http://en.wikipedia.org/wiki/Joule_thief

Test failed. After 3 hours the battery was depleted. I don't know why it happens but have to check the windings on the ferrite core...maybe have to put more on...

The circuit is working good. It lit up a white LED with a tiny solar panel by 0.45 volts but failed in the long run with the button cell.

UPDATE Jan. 30:

Strange things happend. After the button cell died i attached the JT to a almost full AA battery and it was working for a whole week with full brightness. Then the LED faded away to a weak glow until it went off. This was just before I left for my holidays on Jan. 21. Now I am back home and the JT was still lying dead in my book shelf. Last night I saw it lit up again...wondering what happend. Then after about one minute the LED faded away, went off and after another 5 minutes it lit again. The brightness to the beginnuing was about 50%. 

It seems that battery is recovering again and again but how long it will last? More to discover ...


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Ok i have a question .... these seem to be used to get more out of a battrie but also to power something that requires a higher voltage eg using a 1.5v to power a 3v+ led ..... 

i have a solar bulb running from a 1.2v rechargable battrie .... so i charge battrie with the solar panel and when theres not enough light the battrie kicks in and lights the bulb .... so i connect this to the circuit .... eg i have the 1.2 v rechargable battrie as the power sourse and i connect the led to the transistors legs and connect the core wrapped in wire to the positive side of battrie .... the transistor (base) and the transistor (collector) .....  now is my solar bulb getting extra volts and is this bad for it ..... i can disconnect the core wrapped in wire from the resistor and transistor base and bulb still works (as there is a complet circuit) but is the jewl thief still working now as all the connections arent made ... 

Sorry to reply you so late. Can you send a circuit diagram of your circuit? I might be lost in your explanation ;-)

Nice to see joule thief here.... i  built a quite a few last summer..

Edit :-I dont think extra windings will make much difference - just to the time_ing. Just made one with 2x number of turns and it seems to be twice as bright ........ They are stable but temperamental at the same time.

They are active circuits for want of a better word .... ie it oscillates its-self , if too much load is taken out it stops this process as far as i can gather - if you hook up an oscilloscope you can watch the very messy waveform - its a wonder how they work at all..... and the LED rectifies the voltage at same time grrraoooooaaaa

..... i experimented last summer by connecting one up to two probes connected to metal plates stuck into the ground 5 meters apart, however i could only suck circa 0.3V out of the ground (yes i am serious and it can be done - though trying to explain what your doing to the neighbours is no so easy ).... need to experiment with different metals for the ground probes.

BTW :- a good source of ferrite rings can be found in dead (or alive)  motherboards. in a multitude of colours.

Here are 4 such examples confirmed as good candidates.....

Wow Gareth...a ground battery or soil battery...i thought about a Thermocouple some weeks ago...

Ok, if you get 0.3 V then it's possible that you even can drive a LED with that, assuming you got an efficiant Joule Thief connected.

Lumi it looks like what most pipeline engineers dread, would work to our advantage .... in one connected word Galvanic-Corrosion.

Take 2 dissimilar metals and the salts of the earth will do the rest.....

NB :- the further apart you choose your metals in the list ..... the higher is (if i understand correctly) the potential...... so graphite & zinc would be good starting bet.

This would be a neat 24/7 solution .......well...... until your electrode erode.....

Metal                       Potential using a  Cu/CuSO4 electrode and one of the below

Magnesium (pure)           -1.75
Magnesium (alloy)          -1.60
Zinc                              -1.10
Aluminum (alloy)            -1.05
Aluminum (pure)            -0.8 22
Steel (clean)              -0.50 to -0.80
Steel (rusted)             -0.20 to -0.50
Cast Iron                        -0.50
Lead                              -0.50
Steel (concrete)              -0.20
Copper                           -0.20
Brass                            -0.20
Bronze                          -0.20
Steel (mill scale)            -0.20
Cast iron (high silicon)    -0.20
Carbon                         +0.30
Graphite                       +0.30
Coke?                          +0.30

So this looks like it should work........

Cool, it's similar to the thermocouple...i am up to build a earth battery ;-)

The thing is I am living in the 6th floor....but might setup one in the office area (ground floor).

Graphite is a good choice since it corrodes not that like a metal. I will try to test run this after the Chinese New Year with differnt electrodes. Also what is the best shape for the electrodes? Rod, plate, mesh wire???

Me too.... and after watching this YouTuber Pirate88179's Earth batteries connected to joule thiefs........ i am well impressed....

So lumi i accept the challenge too........so lets see how much "juice is loose about the hoose"....

Plan A is to sharpen a dozen or so graphite pencils (minus wood hull) for one electrode and zinc coated nails for other.....

Just back from holiday...Chinese New Year...it means one week offline ;-)))

Ok, I will take this challenge by dismanteling the pensils first... without breaking the lead :-S

This is something that was done over 150 years ago according to this link. Maybe the information here will help.

Thanks, very helpful :)