Introduction

Although we have been led towards a digital and paperless life, paper has its own advantages. For example, paper is easy to use, lightweight, inexpensive and ubiquitous. Like digital products, paper is also possible to be used to interact with digital information. Moreover, we live in a ‘paper’ culture, our everyday life is linked to paper objects.

In PaperComp 2010, the first International Workshop on paper computing, participates claimed that paper could become ubiquitous interfaces in our everyday interactions with digital information and this is the dawn of paper computing. There have been several researches done in interactive paper computing. For example, the Easigami, a novel tangible user interface which embeds sensors on the edge of paper, so that user can construct different shapes of model by combining papers and the model will be reflected in 3D virtual representation. However, the drawbacks are that there are massive wires and it misses of the intrinsic flexibility and tangibility of paper material. Thus, we present a method of selective wireless power transferring. Although many researchers have conducted various experiments in high-efficiency wireless power transfer, most of these experiments did not suggest the solution for controlling selective wireless powering. Thus, this project presents a method that when there are multiple receiving coils with different resonant frequencies and impedance, the output frequency in the transmitting coil can be changed to activate different receiving coils. 

Principle of selective inductive power transferring

The fundamental principle of this method is based on the theory of electromagnetic power generation.

The primary coil is connected to a high-frequency AC source and the secondary coil is connected to the load. The energy is transferred from the primary coil to the secondary coil wirelessly relying on high frequency changing magnetic flux.

When there are multiple receiving coils with different resonant frequencies and impedances, the output frequency in the primary coil can be changed to activate different receiving coils. In order to achieve different impedances and resonant frequencies, each receiving coil is connected to a small capacitor with different values to attain different results in resonant frequency. 

Improvement of existing work

The existing equipment manages to distinguish multiple receivers with different capacitors using different discrete output frequencies. However, in order to achieve higher resolution of paper folding/cutting, the transmitter needs to be able to generate more frequencies rather than discrete values, therefore we need to look into continuous frequency generation.  Function generator is easy to use and able to produce precise signals. SEFRAM 4422-20MHz DDS function generator is being used. Make use of the RS-232 connector on the function generator to input a desired frequency from a PC. Realterm software is used as a bridge connecting the PC and the function generator. This software is just for experiments. At the final stage, it will be replaced by a new interface which will be created by us.  However, the power of the function generator output is very small and not enough for our system. Thus, we need a power amplifier to cater to the transmitter. Although there are many ICs used for RF power amplifier, they are not suitable for our system because they are difficult to be controlled and spend very long time for the power being amplified to a desired value. 

4.   Power amplifier

There are three common classes of power amplifier: Class A, Class B and Class AB.

Class A:

The transistor in Class A is biased for the complete cycle of the input signal, therefore it can produce minimum distortion and high linearity. Since it is always on even when no input signal is applied, it consumes or wastes lots of power and hence, it is inefficient.

Class B:

Class B is much more efficient than Class A because there are two complementary transistors, each of which deals with half cycle of the input signal. But it suffers crossover distortion.

Class AB:

Class AB is a compromise between Class A and Class B. To eliminate the crossover distortion, diodes are used to supply the transistors bias voltage (about 0.5V) and keep them in a “standby” mode. Thus, its efficiency is less than Class B, but still much higher than Class A. 

5.   Experiment on impedance matching

The maximum power can be obtained only when the input impedance and output impedance are matching, that is when the impedance of the load is equal to the complex conjugate of the impedance of the source. For two impedances to be complex conjugates their resistances must be equal, and their reactances must be equal in magnitude but of opposite signs. In low-frequency or DC systems, the reactances are zero, or small enough to be ignored. In this case, maximum power transfer occurs when the resistance of the load is equal to the resistance of the source In order to verify the feasibility of the power amplifier, I did an experiment using a similar RF power amplifier in Power Electric Lab. Before I did this experiment, I made two transformers for impedance matching because the impedance of RF amplifier and the transmitter are not matching.

1:2 transformer 

1:3 transformer

actual setting up

The receiver cannot receive enough power to light LED without impedance matching transformer. By adding the transformer, the transmitter outputted a much larger power than last time and the receive coil obtained the enough power to light the LED. 

LED lighted up

Next, I am going to create our own power amplifier.

6.   Design power amplifier

1^st design:

I chose LT1357 operational amplifier to amplify voltage and set the gain at 11. Since the output current of the op-amp is very small, I chose Class AB amplifier to amplify the current. 2N3904 and 2N3906 are two complementary transistors. However, the maximum current they can provide is only 200mA. Thus, the output power will not be large. 

actual board

Simluated result: gain = 11 @200kHz

Table: with different input voltage, the gain is about 11 at 200 kHz

 Because of low output current, the power is restricted. 

2^nd design:

It is similar to the first design, just adding some coupling capacitors to stabilize the supply voltage and use two new transistors: BD911 and BD912. Their maximum collect current is about 15A and therefore, the output power should be larger than the first design. However, their transition frequency is low, only 3MHz and it may affect the performance of the amplifier since our system is working at several hundred kHz which is close to 3MHz. 

distorted output waveform

The actual result verified the prediction. At 200 kHz, the output waveform is distorted. If the frequency is reduced to 20 kHz, the amplifier can perform well.

actual board (top)

actual board (bottom)

Both of these two designs are not good and new design is required.

3^rd design:

This power amplifier can produce a very large power. Basically this is also a Class AB power amplifier, but new transistor 2SC2565 can provide 15A collector current and 80MHz transition frequency. All transformers are self-made and have different functions. The first transformer is an impedance converter because the input impedance of the amplifier is 50 ohms and the input impedance of the transistor is much smaller. The second transformer is to separate the input signal into two small but opposite signals. The third and the fourth transformer are used to pass DC signal. The last transformer is to combine the two signals together to form the output signal. The function of the bottom left part is to provide DC bias voltage (about 0.6V) to the transistors. This structure (Q3&Q4) is to maintain the output DC bias voltage at 0.6V. By adjusting the value of rheostat R3 to set the bias collector current at 50mA, make two transistors work as Class AB amplifier. When AC input signal comes in, transistors produce large current. 

Two 2SC2565 transistors as well as the transistor which provides the bias voltage are going to be very hot, so heatsink is needed. All components are connected by copper because of low resistance.

actual setting up for setting bias current

Experiment result @200 kHz with a 50 ohms load

yellow:input signal; green:output signal

The load current is almost half of the total current which is read from the power supply. This is reasonable since another half becomes heat dissipation. When Vin reaches 2V, the circuit will overload, thus we need an intermediate transformer. Thus, the power amplifier works properly. 

7.   Future work

In the next semester, I need to find out the impedance of transmitter and the output impedance of the power amplifier. According to their ratio, a suitable impedance matching transformer should be chosen so that the largest power is obtained. At last, the power amplifier is to integrate with my partner’s part to complete the whole project. She is making a sensor which can sense whether the receiver obtains enough power or not. If not, a feedback will be sent to power amplifier and it will increase the power accordingly. If there is enough time, we will design some small application in the realm of paper computing.

The above is 1:3 transformer. 

The above is 1:2 transformer

The above is the block diagram of the experiment for testing the transformer.

The above is the actual setting up of the experiment.

After adding the transformer, the receiving coil obtain enough power to light the LED.

I did some researches for chips which are able to generate various and continuous waveforms. For example, AD5930 is a programmable frequency sweep and output burst waveform generator. However, after studying its datasheet, I found it is unsuitable for our project. It is capable of providing programmable waveform sequences whereas our project needs just a single sine wave.  I think it is better to just use the function generator because it is easy to deal with and can provide waveforms more precisely.

 SEFRAM 4422-20MHz DDS function generator is being used. Make use of the RS-232 connector on the function generator to input a desired frequency from a PC. Realterm software is used as a bridge connecting the PC and the function generator. This software is just for experiments. At the final stage, it will be replaced by a new interface which will be created by us. 

I found that the power of the function generator output is very small and not enough for our system. Thus, we need a power amplifier to cater to the transmitter. Although there are many ICs used for RF power amplifier, they are not suitable for our system because they are difficult to be controlled and spend very long time for the power being amplified to a desired value. 

Prof Nii suggested a power amplifier circuit which may satisfy the need. This circuit is mainly constituted by transistors and transformers and the input and output impedances are both 50 ohms. I have tested using a similar RF amplifier in Power Electronic Lab to amplify the signal generating by the function generator, but the receiving coil cannot get enough power because the impedances are not matching. Suitable transformers are needed to be implemented before generating a larger power.

I have made two transformers. The first one is a 1:2 transformer. It is made by a iron core and two copper wires twisted together. By connecting these wires properly, it can work as a 1:2 transformer. Another one is a 1:3 transformer which is made by a iron core and three copper wires twisted together. 

I have retested the performance of the RF amplifier and our system by adding the transformer. This time the transmitter outputted a much larger power than last time and the receive coil obtained the enough power to light the LED. 

This project started from the fourth week. 

Week4: Researched for chips that can generate continuous waveforms and analysed their feasibility.  Decided to use the function generator provided by the lab.

Week5: Studied how to use the RS-232 connector on the function generator to connect with a PC from where user can input a desired frequency. Researched for power amplifiers which amplify the output signal of the function generator to a desired level.

Week6: Studied a power amplifier circuit suggested by Prof Nii. Did experiments using a similar RF amplifier in Power Electronic Lab. Made some transformers for impedance matching. Tested these transformers using the RF amplifier.

A workable power amplifier circuit is expected to be implemented by end of this semester. 

My final year project is Selective Inductive Powering System for Paper Cutting. It is a method of selective wireless power transferring for paper computing. The novelty is that by changing the output frequency of the power transmission and the impedance of the receivers to selectively activate different actuators embedded in paper, paper-craft patterns can be created automatically on normal paper. Although many researchers have conducted various experiments in high-efficiency wireless power transfer, most of these experiments did not suggest the solution for controlling selective wireless powering. Thus, this project presents a method that when there are multiple receiving coils with different resonant frequencies and impedance, the output frequency in the transmitting coil can be changed to activate different receiving coils. 

The existing equipment manages to distinguish multiple receivers with different capacitors using different discrete output frequencies. However, in order to achieve higher resolution of paper folding/cutting, the transmitter needs to be able to generate more frequencies rather than discrete values, therefore we need to look into continuous frequency generation. 

The main part of my job is to improve the system with generating more frequencies, such as continuous frequency generation. At the final stage, we are expected to implement a feedback system which is able to generate suitable frequency and power according to different receivers.