I'm going to show you how to make your own low-power wireless charging circuits that will let you pass electricity through the air (or any other non-metallic medium) over short distances. This is suitable for wireless battery and capacitor charging and powering of very small un-buffered circuits (such as a single LED).
Please make sure to check out the last page as there are tons of references and other sources I managed to gleam from the internet and other instructables. Also note that I spent a GREAT deal of time experimenting and researching to get this right. I'm an electrical engineer, and even still it took quite a while to get my head around some of the technical challenges. As such this is for experienced hobbyists only, unfortunately it's not easy to do although I tried to make it as simple as possible. It doesn't take a lot of skill, just a lot of tinkering to get it to work right.
Now there shouldn't be current patents on any of this (Tesla, Colpitts, Cockcroft, and Walton all made this stuff yeaaaaars ago), but I would look into it first if you wish to sell anything using this design.
If you want the circuit then just skip ahead to step 2 and ignore the theory part :).
Step 1: Theory of Operation
The short story: this is a Cockcroft-Walton generator hanging off a resonant transformer. If you don't mind wasting a couple minutes with detailed theory then charge ahead intrepid reader! Otherwise skip to the next step.
The long story, well, it's not much longer. Take a coil, make it resonate at a particular frequency using a capacitor, then place it near a similarly tuned coil and use the oscillating magnetic field of the first to cause the second to resonate. Use a clever AC to DC converter and voila, you have a method of wireless energy transfer.
After some sleuthing on the internet, I went about devising the first part, an oscillator. Various homebrew methods have been used (see: Wireless Power Instructable) but weren't very good or just temporary solutions. I used the suggestion on wikipedia of using a Colpitts oscillator. This is a decent solution because it's dead simple to build and, most importantly, it's a current oscillator and not a voltage oscillator. As current through an inductor is what generates the magnetic field, this is what will drive both coils.
The second part is fairly easy to understand, that being the two coils. Although they don't have to be the same physical size, they do need to resonate at the same frequency. The combination of number of turns and diameter determine the inductance, and some capacitors were added to obtain the correct oscillating frequency. It gets tricky when you get into the details however (and they get very, very detailed, so I won't put the majority down here) as you need to select the diameter of wire to go with the amount of current going through your coil, which will determine the amount of resistance in the coil, which will impact the viability of your oscillator. To make it somewhat easy, go with 24AWG enamled magnet wire.
You now get to pick a some-what arbitrary frequency for your circuit. This I decided to go with 80KHz, it happened to be a nice middle ground between easiness and efficiency. Then you pick a capacitor value that's commonly available, I picked 150nF. This took a while to select because you need to get an inductance that is within the realm of being hand made. Using the equation:
frequency = 1/( 2 * pi * sqrt(inductance * capacitance / 2) ) (from Colpitts oscillator)
we use the capacitor and frequency values to try to get the inductance in and around 20uH to 70uH. Air-core inductors around those values are easy to make. I used a value of 53uH, which to calculate:
freq = 80k
cap = 150nF
freq = 1/(2*pi* sqrt(ind * cap/2) )
freq * 2 * pi * sqrt(ind * cap/2) = 1
ind * cap/2 = (1/ (freq * 2 * pi) )^2
ind = (1/ (freq * 2 * pi) )^2 / (cap/2)
ind = (1/ (80k * 2 * pi) )^2 / (150nF/2)
ind = 52.77uH
From here you need to use this handy inductor calculator to try to figure out what diameter and number of turns are needed. I used values of ~22 turns at 6cm diameter, with an arbitrary length around 4-5x the wire thickness for the secondary, and ~13 turns at ~15cm diameter for the primary. These values will be your STARTING POINT ONLY. You have to experiment to get it right (covered in the next couple steps).
Note that you are using the same inductance and capacitance for both the resonating coils, this is so it's easy to tune. Don't go crazy with different inductances and capacitances or else you won't get it to work.
OK, the last part of this picture is the AC to DC converter. This is what will shape the received AC into something we can use to charge a capacitor or a battery at a usable voltage. I used a CW generator here to great effect; it allowed me to tune the slave coil to produce exactly the right voltage without going over the charging voltage. I determined (through experimentation) that a two stage generator would be enough, and that will generally be fine when trying to generate ~5V. For the capacitors I arbitrarily chose 2.2uF caps, and for the diodes I chose a nice Schottky diode array with a very low 0.38V forward voltage drop. The P/N is BAS40TW-TP, however these are VERY small parts so you will probably have to order individual schottky diodes for this one. Just use ones with a low voltage drop AND a low reverse leakage current.
OK! Enough of this long-winded theory and background info, let's get to the actual good stuff!
Step 2: Needed Parts and Equipment
If you have access to one, then great! That, a soldering iron, some wire cutters, and pliers are all the tools you need.
As for materials, you will need a few different parts, but nothing too fancy.
24AWG enameled magnet wire
2 150nF capacitors
2 10K Ohm resistors
1 100Ohm resistor
1 100nF capacitor
A bunch of 2N2222 transistors (I used 3, you can use more or less depending on availability)
5V regulator and DC jack to plug it into the oscillator
Slave coil/CW generator:
Another prototyping board
2 150nF caps
4 2.2uF caps
4 low forward-voltage drop schottky diodes (search digikey for ones with Vf<400mV@1A and Ir<1mA@20V)
Note that technically any schottky diode will work here, so if you want to just build it without much efficiency then feel free to use whatever you wish / can get your hands on.
Step 3: Coils and Oscillator
The oscillator is pretty simple and was tested both in simulations and in a practical circuit. I derived it from the one here on wikipedia, however values had to be changed and more BJTs (transistors) were added after I discovered that max current transfer improved. Rough schematic and pictures are below. Note to add more BJTs just connect them in 'parallel' to the one in the schematic, pin-to-pin.
The primary coil started as ~18 turns at ~15cm diameter, then I removed turns based on the final shape/diameter. If you look at the picture of the jig you see how I easily made the coils to a certain size. Just cut out some holes in a cardboard box and use pens on an angle to wrap the wire around.
In the other picture of the completed circuit/coil, I forced the coil into a roughly rectangular shape, so the inductance changed. Simply make the coil with a few extra turns and then connect that up to the oscillator (directly, solder it without cables or any other wires). Place the 'scope probes across the coil and check the wave period. Remove coils until the period matches what you want (in my case, it was 12.5us). By remove coils I mean physically remove a turn by cutting the wire and re-soldering the end. Excess wire will lead to more inductance and you won't get the right value.
After you are finished with the primary coil you'll do essentially the same thing for the secondary coil. Just unsolder the primary and repeat. However when making the coil you are welcome to change the diameter and number of turns. I used my hand for the second one and started with ~30 turns, to make it smaller and easier to fit in things.
Once completed you can wrap the coils, although this is risky as you will change the inductance significantly so you'll have to do a considerable amount of trial and error to get it right. The inductance changes because you are forcing the wires closer together.
Now onto the slave pickup and CW generator!
Step 4: Slave Pickup and CW Generator
Note that in the pictures below, you are looking at my heavily modified original PCB. Because of the simplicity here, this can be made on some protoboard or even dead-bug style.
The schematic and pictures really say it all, simply connect the secondary coil to the capacitors and then connect that up to the diode ladder. The 'bottom' line of the coil is selected as the ground and the top of the ladder is the output, in my case it was a bit under 5V.
The next step is simply putting it all together with some sort of storage element.
Step 5: Putting It All Together (And Caveats)
Of course the purpose of this instructable is that there isn't any 'hooking up'! If you power the oscillator and have the secondary coil inside/on/around the primary coil, you will notice that a fairly decent voltage is generated on the secondary coil. Although usually just over half the voltage on the primary, the CW generator takes that AC voltage and conveniently produces a clean 5V. Ideally this can be used to charge a large capacitor bank like I did, or a low power 5V circuit. Now for the caveats.
This first one being that you can't really use this for any high current applications, such as driving motors or a bunch of LEDs. Charging is a different matter of course and will work perfectly ok for that. When you try to pull too much current out of this circuit the voltage will start dropping considerably. For instance when you connect a fully depleted capacitor bank to it the voltage across the secondary is considerably reduced. If you take a look at the primary during this time you will also see that the frequency and amplitude of the wave is considerably different. This frequency shifting is what prevents you from using high current loads. I'm sure there are better oscillators and other measures that can be taken to improve the voltage regulation, but this works as a preliminary model.
The second caveat is that you can't put ANY metal in between the primary and secondary, particularly iron based metals (steel, stainless or otherwise). Even placing the oscillating circuit inside the primary effects the performance, creating drop-out zones on the upper surface of the picture frame that prevent charging when the secondary is placed in certain spots.
The third caveat is that the distance between the primary and secondary coil must be kept to a minimum. This isn't WiTricity, it can't power anything over a distance of even 20cm.
Working around these limitations is quite easy though. My method was to use the circuit to charge a large capacitor bank (3F @ 5V) and then use that bank to power a switching regulator (to keep a constant 5V even when the capacitor voltage drops) and LDO so I have both 5V and 3.3V to work with. It takes about a full night to charge the capacitors, and I can get a considerable amount of run time with proper power saving attention to the rest of my circuit.
Up next? Well maybe a larger, more robust version to trickle charge a car battery, or a nicer looking version to charge a cell phone battery. Maybe some experimentation in flat-wrapped coils or other methods. Feel free to expand upon my methods and improve this tech, and heck, take my idea and integrate it into some of your existing projects. Just as long as you promise to make an instructable or something out of it, I'm interested to see what people will do with this! Also make sure to respect my creative-commons by-sa license ;).
Step 6: References and Links
Instructables user robotkid249: wireless power transmission over short distances
Instructables user puffin_juice: I need your help. IPT
Robert Coup and Monique Ryan (U of Auckland): Inductively coupled universal battery charger
Now for links from wikipedia (all accessed on or around Dec '09):
Wireless Energy Transfer
Resonant Energy Transfer
Secondary Coil: LC circuit resonance
Other links from around the web:
Common capacitor values
Wireless power links (just fyi):
Neat RGB LED project, similar setup to this one
Using large capacitors and small inductors instead
Quick-and-dirty method for high power output using a CFL bulb.
I'll reserve this space below for anyone that uses this in one of their projects, just let me know!