I Need Your Help. IPT (Inductive Power Transfer)

THIS INSTRUCTABLE IS NOT YET COMPLETE. I have been trying to build a wireless charger but have come across a problem. I have documented what I have done and learnt so far. Hopefully from this you can see my errors and help me out. Please read the last page to see what I think the problem is and suggest any ways to get around this. THANKS

The Project
I am trying to build a circuit that will allow gadgets that are usually charged by USB to be charged wirelessly. As an example I am reverse engineering an A4tech battery-less mouse. However it is too great a challenge for me and I am seeking help from you. I thought it would be better for me to turn this into a group effort than to ditch the project. I will give a detailed description of what I have built and learnt and hopefully you can tell me where I went wrong.

Transferring power wirelessly is reasonably simple. If you think about it, all transformers are wireless. But we want something that's truly wireless. Like the Splashpad (see picture). It is pretty much a transformer with an air core.
The mouse I was talking about (see other picture) in the introduction is exactly the same as the Splashpad it uses induction to transfer power across an air gap. This is the same technology as RFID; in fact it uses this to communicate with the pad. To make our own wireless device we need to know more about induction.

We need two coils. A primary and a secondary. The primary will be connected to an AC power source that creates an alternating magnetic field spreading out into space to infinity; however as we get further away from the primary coil the strength of the field dramatically decreases. Now, if we hold our secondary coil anywhere in that space an alternating current will be induced. Unfortunately this current is so low that it is too low to do anything useful or should I say too low to even measure. Gigantically low, so low it might as well be nothing. I think you get the point. So the question is; how do we Increase the power that is transferred to the secondary coil?
It is called inductive coupling of tuned circuits (Here is some more detailed information, more tesla coil orientated but more or less the same principles. http://home.freeuk.com/dunckx/wireless/inductive/inductive.html).
We need an AC power source for the primary, any frequency would do, but to make it more efficient we need a high frequency. To explain this think of two solenoids. If we hook up the primary coil to a DC supply. Initially the current will be zero, the current will increase at a decreasing rate due to a magnetic field that is created to oppose the change, once the two reach equilibrium the current reaches a steady maximum (can be calculated by I=V/R) and the magnetic field also becomes constant. This changing magnetic field induces a current in the secondary which produces a magnetic field in the opposite direction. This is just a small blip of current that lasts just for a short period of time. We have just transferred a small amount of power. Now if we did this over and over again we would induce more tiny blips and hence more power. This is why a higher frequency would transfer more power.

The battery-less mouse has a frequency range of about 119kHz to 135kHz, which is what we will use; it is probably a legal frequency???

*The third image I grabbed from a lecture slide, the lecture had no name on it. If someone objects to me taking this slide, please let me know and I'll remove it.

So we have a high frequency power source, now we need a primary coil, this coil shouldn't be too large so that it is more like a pad than a brick. It doesn't really matter too much but we want something like the mouse/Splashpad.
To decrease the resistance/Impedance of the primary circuit we need a coil in series with a capacitor. So the final circuit will look something like the circuit on the left in the image below. The capacitor and inductor make something called a resonator. If the capacitor was fully charged and then connected in parallel with an inductor (right-hand side of image) we would get an alternating current flowing through the two components. This is because the capacitor discharges through the inductor, a magnetic field is created. When the capacitor is fully discharged the current stops, the energy of the magnetic field is converted back into electricity; this recharges the capacitor. This cycle then repeats many times. This type of oscillation is called resonance if the reactance's (see next sentence) of the inductor and capacitor are equal. Reactance is the equivalent of resistance in a DC circuit.

Let's take a look at my high school notes book.

The current and voltage are in phase

The current and voltage are out of phase
The resistance is not easily calculated as AC resistance is affected by frequency, We therefore use a different physical quantity called reactance, it still has the same unit as resistance. The ohm.

It is clear that frequency affects the resistance and hence the current. See second image

This is slightly different from the LC circuit mentioned above, but the same principles apply.
The graph pretty much explains how the resistance of a resonant circuit decreases if the reactance's of the capacitor and inductor are equal. The frequency that this occurs at is called the resonant frequency.
(see picture)

Note: the circle diagrams with the arrows are phasor diagrams that represent the magnitude of the quantity at a moment in time. In the resonant circuit above the reactances of the capacitor and inductor are equal and opposite having the effect of cancelling eachother out.

The Whole picture
This theory helps us understand that if the primary and secondary circuits in our wireless circuit are at resonance we increase the effieciency. No power is lost in the inductor and capacitor.

We need to construct a circuit that can produce alternating current at a desired frequency. My first attempt was using a Multivibrator (http://en.wikipedia.org/wiki/Multivibrator). However that didn't work unless it had a separate power supply, whenever I attached the load, the change in current due to a parallel circuit would affect the rate of the multivibrator. It also isn't very accurate because it relies on the accuracy of the values of the capacitors and resistors. Take a look at the page anyway because it is quite interesting to know that the multivibrator can make a flip flop.
Another way would have been to use a comparator as a square wave generator (http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/square.html#c1). The maths looked complicated so I soon discarded that idea.
In the end I decided to use a 555 Timer. We have probably all heard of this universal device. As far as I know it is quite an accurate timing device and is rather simple to use. We are going to use it in its Astable configuration. Below is a small excerpt from a webpage (http://www.kpsec.freeuk.com/555timer.htm) with everything you need to know about it.

We need to choose resistor values that will give us a frequency of about 120kHz. I happen to have a 181K capacitor.

Lets use the formula from the previous webpage to calculate the value of R2 (in the picture below).
(see second image)
This gives us a result of 32407.41. The closest resistor value I have is 33kOhm, which has a value of 32.5kOhm when measured with a multimeter. We will use this more accurate measurement to work out the true frequency of our circuit using the following formula
(see fifth image)
Our true frequency is 119658.12 Hz

The value of R1 has to be about one tenth of R2 to make the mark and space time approximately equal! So, 3.3kOhm

Now we need something that takes this signal and turns it into AC using a DC source. I have come up with various circuits. Below is the original circuit I designed. I first tested this in combination with the multivibrator but it affected the rate of the multivibrator. I thought it didn't work because of this and decided to look at google patents for another idea. Check these out, there's some interesting stuff.

With a bit of research I found many varieties of DC to AC inverters/converters. Many were complicated and so I went back to the drawing board. The circuit below is what I came up with. It uses a voltage divider to create dual voltage power supply. In my case +3V and -3V with a neutral centre tap. How does it work? (see diagram)

We have alternate current through the resistor by the centre tap.

I have tested this and it does work at low frequencies for testing.

All we have to do now is decide on the coil shape and size and the capacitor that will create a resonant circuit. For testing I have used primary/secondary coils that are of similar size so that they can be wound on the same pipe, just for simplicity. There are many online calculators for calculating inductance (See the external links for some of the calculators I looked at). I decided to use a program called MiscEl (http://www.miscel.dk/MiscEl/miscel.html, its freeware!).

First choose a capacitor and calculate its reactance. Then calculate the required inductance you need to create a resonant circuit. I have used the values I used in my circuit.
(see image)
MiscEl can work backwards. Enter the inductance and the diameter of the coil. It will then give you the number of windings needed to produce that inductance. Make this coil and attach it to the capacitor in series. Ok thats the primary circuit complete. As you may have noticed, all my work has been done on a breadboard. That's because it doesn't work yet. When I get this thing going I'll make all the printed circuit boards and modify this instructable with all the correct values and dimensions. Your help is appreciated.

As you would have guessed. This circuit is also at the resonant frequency. Using the frequency of the primary circuit we will calculate the value of the capacitor and the value of the inductor. The secondary coil cannot be too large as it needs to eventually fit onto one of our electronic gadgets. Therefore we want a small number of turns. When I originally thought of doing this I was going to use a flat spiral coil. I will do that as soon as I get this working. The final secondary circuit will be all on one circuit board, for ease of production.

Anyway. I have a variable capacitor that I should use in conjunction with a fixed capacitor to allow us to fine tune our circuit. I have a variable radio capacitor with a capacitance of up to 220pf when the two internal capacitances are connected. We do this by connecting the two outside leads (A &O) of the variable capacitor. Like in the picture below.
We also attach a capacitor in parallel with the variable capacitor to increase the capacitance. I used a green cap with the markings of 104J (100000pf). Using this information we can once again calculate the reactance's and hence the dimensions of the coil using MiscEl.
(see second image)

We can turn on the circuit and see whether we induce a voltage on the secondary coil. Now unfortunately it doesn't work. I have analyzed the mouse and its circuit to find out what the difference is between my circuit and theirs. One significant difference is the voltage across the primary. I measured a voltage of 8V. The voltage on mine was about 0.4 volts. But how is this possible. A low resistance coil with a high voltage. If we apply ohms law we would find that the current would have to be huge. But it's a USB powered device and is restricted by the maximum current the computer can supply. And this current is very low.

This is where I'm stuck. How can I increase the voltage across the primary coil?
Below are some photos of the A4Tech mouse. The secondary circuit of the mouse is exactly like mine, minus the rectifier which I will add later. The primary is similar too, a capacitor in series with the coil. A coil, which has very little turns and a low resistance.

Please, if you are good with electronics and physics. Could you suggest some ways that would make this thing work and point out any errors I have made?

Below is a short video clip of what I have built so far . All the plans for the circuitry with the values you see in the video clip are below.


For some reason I cant get the video to play. The link below is to the video above
http://video.yahoo.com/video/play?vid=1686520

This file contains click on links with descriptions.

This file was created on a mac with opera. If it doesn't open try opening it with a text program like notepad or textedit

I know its long, but if you're interested in the power transfer part its definately worth reading.

Thank you to Cerincok who brought this to my attention!

If you're interested in using a microcontroller to generate the signal.

See PDF's