Sunday, January 1, 2012

Joule Thief

The kids and I are doing some more battery experiments (briefly mentioned in an earlier post).  We have been using copper pennies, aluminum foil, felt, wire and diluted vinegar in old film canisters.  We can generate a few volts but so far have not had any luck powering anything off of them.  To try to convert the output to a higher voltage I made a joule thief pictured below.  The name is a play on Jewel Thief and Joule as a unit of energy.  It "steals" extra energy from the battery. 


It has two wires wrapped around an iron (ferrite) ring (although a bar would also work) and two ends are connected to a transistor (one via a resistor).  It sets up an oscillation which boosts the voltage to a higher level for part of the time.  The key to understand it is that as the current flows from the transistor to the coil, a magnetic field increases in the iron, which causes a current to flow in the second coil, which gives positive feedback to the flow of current from the transistor.  So the flow of electricity builds in the circuit.  However, voltage is only produced in the second coil when there is a change in the magnetic field.  So when the current begins to max out the secondary voltage drops, which again due to positive feedback lead to a very rapid decline of current in the system and, because the magnetic field is changing again, a sharp spike of voltage is the opposite direction from the second coil (from the collapse in voltage in the first coil).  This quick pulse of voltage is higher than what is in the original battery and can light the LED.  The circuit is trading continuous lower voltage for brief moments of higher voltage.  (This is similar to the mechanical action of a Shishi odoshi, a Japanese garden fountain that slowly builds in potential then snaps back to the beginning state making a loud "deer scaring" noise.)



In the picture above it is powering the LED to the upper right from a 1.5 V battery.  Normally 1.5 V is not enough to light up an LED (without the voltage oscillation), the LED to the lower left is connected directly to the battery for comparison and is not lit up. 


So this simple joule thief works; however, four of our homemade copper-aluminum batteries are not enough to light the LED even with the oscillating voltage boost.  Back to the drawing board.  (I suspect the zinc in some of the pennies is causing a problem.  Perhaps I will experiment with some other metal combinations.) 

1 comment:

Watson aName said...

Make sure your pennies are copper, the ones dated 1980 or earlier. Also, try putting the cells in series across the LED without the Joule Thief. Four cells in series should have higher voltage than the LED. The Joule Thief wastes a lot of power and may load the cells down too much. You could try to put the cells in parallel with the Joule Thief, but why bother when they will give you enough voltage in series.

Also make sure the electrolyte in the cells is fresh. You might try different electrolytes such as sodium hydroxide (lye) but it's caustic and can burn your skin.