A Shared Channel Design for the Power and Signal Transfers of Electric-field Coupled Power Transfer Systems

Abstract

Electric-field coupled power transfer (ECPT) systems have been proposed as an alternative wireless power transfer (WPT) technology in recent years. With the use of capacitive plates as a coupling structure, ECPT systems have many advantages such as design flexibility, reduced volume of the coupling structure and metal penetration ability. In addition, wireless communications are effective solutions to improve the safety and controllability of ECPT systems. This paper proposes a power and signal shared channel for electric-field coupled power transfer systems. The shared channel includes two similar electrical circuits with a band pass filter and a signal detection resistor in each. This is designed based on the traditional current-fed push-pull topology. An analysis of the mutual interference between the power and signal transmission, the channel power and signal attenuations, and the dynamic characteristic of the signal channel are conducted to determine the values for the electrical components of the proposed shared channel. Experimental results show that the designed channel can transfer over 100W of output power and data with a data rate from 300bps to 120 kbps.

I. INTRODUCTION

Wireless power transfer (WPT) technologies provide solutions for power transfer when direct wire connections are impossible or hard to achieve [1], [2]. Electric-field Coupled Power Transfer (ECPT) is a wireless power transfer technology that has been used in rotating devices [3], mobile robots [4], biological implants [5], cell phones [6], and electric vehicles [7], [8]. With the use of capacitive plates as coupling structures, ECPT has many advantages such as design flexibilities, a small volume of the coupling structure and metal penetration capability [9], [10]. Fig. 1 shows a block diagram of a typical ECPT system.

Fig. 1.Block diagram of a simplified ECPT system.

For ECPT systems, a challenge is to achieve large amounts of output power and high power transfer efficiency [11]-[15]. The signal communication system between the transmitter and the receiver can help achieve a high output power and efficiency.

Many wireless signal transfer strategies have been proposed for inductively coupled power transfer (ICPT) systems. A signal channel whose frequency is closed to the power transfer frequency using two auxiliary coils has been prsented [16], [17]. A power and signal transfer method was proposed for mini equipment with an operating frequency greater than 1MHz and a 12mW transfer power [18], [19]. Additionally, a high-speed and accurate controller is used to modulate a signal for avoiding noise due to the current switching in an inverter, which aids in the simultaneous transfer of power and signals [20]. However, due to essential differences between ECPT and ICPT systems, the power and signal transfer methods for ICPT systems cannot be directly applied to ECPT systems. Therefore, this paper proposes a power and signal shared channel for ECPT systems.

Section II, introduces the shared channel structure and its operation principle based on a traditional current-fed push-pull method. The designed circuits are analysed using both dynamic and static analyses in Sections III and IV. Finally, simulations and experiments are conducted to evaluate the designed channel performance in Sections V and VI.

 

II. TOPOLOGY FOR POWER AND SIGNAL PARALLEL TRANSMISSION

A. Traditional ECPT Topology

A traditional topology of an ECPT system includes a current-fed push-pull resonant converter, a resonant tank, two electric plates and a voltage rectifier, as shown in Fig. 2. In this topology, a dc voltage Edc is transformed into a quasi current by two phase-splitting inductors. The MOSFET drain voltage increases when the current go into the resonant tank. Due to manufacturer imperfections, one of the MOSFET will turn on when the other is off, and the tank circuit starts to resonate to generate an ac electric field. Two pairs of metal plates on the coupling structures can be modelled as the equivalent capacitors Cs1 and Cs2 (in Fig. 2). To transfer power to the load efficiently, the series coupling capacitors are tuned by the compensation inductor Ls at the operating frequency. The rectifier bridge D1-D4 inverts the AC voltage into a DC voltage to supply the load R directly.

Fig. 2.Traditional current-fed push-pull topology of ECPT system.

B. Communication Signal Channel Design

The communication signal channel includes two coupling capacitors Cs1 and Cs2, two branch circuits with two selective LC filters, and resistors that are built for conducting communication between the transmitting side and the receiving side. Fig. 3 shows the communication channel within the traditional ECPT system.

Fig. 3.Improved topology of ECPT system with communication branch circuit.

For wireless signal transmission, the combination of quasi-current source and push-pull inverter services is an AC current source which is equivalent to an open circuit for the signal channel. Therefore, the circuit before the parallel resonance tank is removed, as shown in Fig. 3, according to the simplification. In addition, the rectifier D1-D4, the filtering capacitor Cd, and the load R are together considered as a resistor load which is equal to RL. In this signal channel, the capacitor Cb serves as an isolation capacitor which isolates the low-frequency power wave transfer across the communication branch circuits. To effectively keep the power wave out of the signal channel, the capacitor Cb has to be set as a small value.

A compensation inductor Lb is connected in series with the capacitor to select signals at a signal carrier frequency ωs which can be calculated using Equation (1).

Lb has a negligible influence on the isolation characteristic of Cb because the tuning frequency of Cb and Lb, is equal to the signal carrier wave frequency which is much higher than the power transfer frequency. The signal carrier wave is modulated by the amplitude shift keying (ASK) mode which provides a small changing of the amplitude of the carrier wave during 0-1 shifting. The resistors Rb1 and Rb2, which connect with the signal source can enhance the damping coefficient of the signal channel, and decrease the dynamic response time of the saltus step of the amplitude of the carrier wave which is detailed in a later section. During the half-duplex communicate operation, only one source is connected with Rb1 (or Rb2), and another resistor Rb2 (or Rb1) on the opposite side serves as signal detection resistor. Moreover, the compensation inductor Ls is divided into two inductors Ls1 and Ls2 to enhance the impedance of the branch circuit beside the signal branch circuit, which can increase the communication quality.

 

III. STEADY STATE ANALYSIS OF POWER AND SIGNAL TRANSMISSION

For wireless power and signal transfer systems, the quality of the systems are measured based on the levels of the interferences between the power transmissions and the signal transmissions.

A. Power Transmission

Fig.4 shows a simplified circuit of the power channel in which the current-fed push-pull inverter is considered as an AC current source.

Fig. 4.Simplified circuit of power channel.

If Cp and Lp are fully tuned to the system operating frequency, the AC output voltage of the push-pull inverter can be boosted by π times the input dc voltage, which is shown in Equation (2).

In Fig. 4, the voltage achieved by the load uRL should be equal to the voltage on the parallel resonant tank up at the resonant frequency without considering the resistances of the internal components. The transfer function GRL is presented in Equation (3).

Considering the impact of the signal on the power transfer and the impact of the power on the signal detection, the circuit in Fig. 4 is divided into four parts and the impedance of each part can be expressed as:

Where Zp1, Zp2, and Zp3 are the local impedances shown in the Fig.4. Zb donates the impedance of the communication branch circuit and Cs=Cs1Cs2/(Cs1+s2), Lb=Lb1=Lb2, Cb=Cb1=Cb2, and Rb=Rb1=Rb2. According to the Kirchhoff voltage and current laws (KVL and KCL), the intermediate transfer functions can be expressed as:

Where, iLs1 is the current of Ls1. ub1 and ub2 represent the voltage of the branch circuit of the primary and receiving sides, respectively. ups1 and ups2 represent the voltage of the signal detection resistor on the primary and receiving sides, respectively. uRL is the voltage of the load RL. Gadi(i=1,…,6) represent the transfer functions of the intermediate variables iLs1, ub1, ub2, ups1, and ups2 which are adopted to simplify the solving process and the solution. Based on Equation (4) and Equation (5), the three transfer functions can be obtained as:

Gps1 and Gps2 are the transfer functions from the power input voltage up to the signal detection resistor voltages ups1 and ups2, which are adopted to measure the power interference with the signal transmission. GRL is the transfer function from the power input voltage up to the load voltage uRL after the addition of the communication branch circuit. Comparing GRL in Equation (6) to GRL in Equation (3), the impact of the additional communication branch circuit on the power transmission can be measured.

B. Signal Transmission

For power transfer, the inductors and capacitors of the resonance tank in Fig.3 should satisfy Equation (7).

Where ωp is resonance angular frequency of the power resonance tank, where the parameters Lp, Cp, Cs1, Cs2, Ls1, and Ls2 are marked in Fig. 3. Generally, the signal frequency ωs is much higher than the power wave frequency ωp which is defined in Equation (8).

Where α is the ratio of the signal and power wave frequencies. Define the impedance on the left side of the primary signal branch and the right side of the secondary signal branch as Zleft and Zright, respectively. They can be obtained as follows:

In Equation (9), α/(α2-1) is approximately equal to 1/α since α is much bigger than 1 which means that jωsLp is much bigger than 1/jαωpCp. Therefore, the second term can be ignored. Similarly, for Equation (10), jωsLs2 is much bigger than RL. Therefore, RL can also be neglected. As a result, Equation (9) and Equation (10) can be approximately simplified as:

Since Ls1 is equal to Ls2, the analysis of the communications in both directions are the same due to above analyses. Therefore, for the communication signal channel, both the signal transmission from the transmitting side to the receiving side and the transmission from the receiving side to the transmitting side in Fig. 3 can be simplified as Fig. 5.

Fig. 5.Simplified circuit of signal channel.

The attenuation coefficient of the signal channel is the transfer function from the signal source voltage usin to the voltage of the signal detection resistor usout without power transmission. The circuit in Fig. 5 is divided into four parts with certain impedances which are shown in Equation (12).

Where Zs1, Zs2, and Zs3 are shown in the Fig.5. Lb=Lb1=Lb2, Cb=Cb1=Cb2, Rb=Rb1=Rb2, and Cs=Cs1Cs2/Cs1+Cs2). According to the Kirchhoff voltage and current laws (KVL and KCL), the intermediate transfer functions can be expressed as:

Where iLb1, iCsb1, and iLb2 are the currents of Ls1, Cs1 and Lb2, respectively. uRb2 represents the voltage of the signal detection resistor. Gsi(i=1,…,4) represents the transfer functions of the intermediate variables iLb1, iCs1, and iLb2. Based on the Equation (12) and Equation (13), the transfer function from the signal source voltage usin to the voltage of the signal detection resistor usout can be obtained as:

The transfer function Gs can be used to measure the attenuation coefficient of the signal channel.

 

IV. DYNAMIC ANALYSIS OF THE SIGNAL TRANSMISSION

A dynamic analysis is used to study the saltus step of the amplitude of the carrier wave while 0-1 shifting. After the amplitude of the carrier wave shifts between 0 and 1, the voltage on the detection resistor Rb2 will reach the steady state after a start-up or a decay process (in Fig. 6). The duration of this start-up or decay process is related to the value of the parameters on the communication branch circuit.

Fig. 6.The input signal waveform and output signal waveform.

In Fig. 5, the impedances of the inductors Ls1 and Ls2 can be simplified as an open circuit due to being extremely large under a high-frequency signal excitation. The coupling capacitors Cs1 and Cs2 can be neglected since their values are relative large compared to Cb1 and Cb2. Therefore, Fig. 5 can be simplified as Fig. 7.

Fig. 7.Equivalent circuit model of signal channel dynamic characteristic.

Where Le=Lb1+Lb2, and Ce=Cb1Cb2/(Cb1+Cb2). When the amplitude of the signal carrier wave shifts, a freely damped oscillation occurs in the communication signal channel. During the damped oscillation, the differential equation in terms of the voltage across the equivalent capacitor Ce and the detection resistor Rb2 can be obtained as:

The voltage of the resistor Rb2 can be derived as:

Where, ûsin is the peak value of usin when the modulating signal is 1. The detailed derivation process of Equation (16) is provided in the Appendix. The function curve of the expression in Equation (16) is shown in Fig. 8.

Fig. 8.Curves of signal carrier wave in dynamic procedure.

To measure and define the duration of the dynamic process, the variable τ, which is the time when the amplitude of usout drops to 0.3ûsin , is introduced. The coefficient 0.3 is determined according to the reference voltage of the comparator in the demodulation step. Since the duration of the dynamic process affects the precision of the demodulation and results in a delay in the communication, the duration τ should be limited to within a defined range. In this paper, the duration τ is defined as τ = 15%Tm/2 , where Tm is the modulation period. Based on Equation (16) and the definition of the variable τ, an new equation can be obtained as below.

Based on Equation (1) and Equation (17), the relationship between Rb and Cb can be derived as:

Then the value of the isolation capacitor Cb is the only variable on the communication branch circuit. Both the values of the resistor Rb and the inductor Lb can be calculated according to Cb. Based on the dynamic analysis, the system parameters are chosen, and are shown in Table I.

TABLE ICOMPONENTS PARAMETERS OF PUSH-PULL RESONANT ECPT SYSTEM

Fig. 9 shows the values of the transfer functions Gps1, Gps2, GRL, and Gs whose independent variable is Cb. In Fig.9, the According to Fig.9 (1) and Fig.9 (2), with an increase of Cb, the values of Gps1 and Gps2 decrease a little but the decreasing value range of the isolation capacitor Cb is from 1pF to 50pF.

Fig. 9.The relationship between the values of transfer functions Gps1, Gps2, GRL, and Gs and isolation capacitor.

amplitude is small. Therefore, the effect of the isolation capacitor Cb on the values of Gps1 and Gps2 can be ignored. However, in Fig. 9 (3), the function GRL decreases from 100% to 83% while the value of Cb increases. In Fig. 9 (4), the function Gs also decreases from 100% to 84% while the value of Cb increases. In conclusion, the power interference is kept constant while Cb increases. However, the power and signal attenuate significantly while Cb is set to a large value. To make a tradeoff, the isolation capacitor Cb is set to a small value 4.7pF which is the minimum value of a common silvered mica capacitor. Then the inductor Lb is derived as 53.9μH and Rb is 207.6Ω. All of the values of the elements in the communication branch circuit are listed in Table II.

TABLE IICOMPONENTS PARAMETERS OF COMMUNICATION BRANCH CIRCUIT

 

V. SIMULATION VERIFICATION

The topology which is shown in Fig. 3 has been simulated using MATLAB Simulink. The input dc voltage of the power is set to 15V, and the component parameters on the main circuit are presented in Table I and Table II. The total simulation time was 100μs for the signals and 25μs for the power. In addition, the maximum simulation step was 1ns. The carrier frequency of u1 was 10.0MHz. The amplitude of u1 was 5V. The signal modulation frequency of u1 was 60 kHz.

In Fig. 10(a), the drive signal and voltage of a traditional ECPT system without communication branch circuits are shown. The first waveform is the driving signal of the MOSFET. The second waveform is the voltage of the parallel resonant tank with an amplitude of 45.5V. The third waveform is the voltage across the load with an amplitude of 45.5V. The output power is 103.5W.

Fig. 10.Simulation waveforms of ECPT system with communication module.

Fig. 10(b) presents waveforms of an ECPT system with a communication module. All three of the waveforms are extremely close to those in Fig. 8(a). This indicates that the communication branch circuit has little impact on wireless power transfer.

In Fig. 10(c), waveforms of signal transmission without wireless power transferring are shown. The first waveform is the modulated signal of the transmitting side whose modulation frequency is 60 kHz. The second waveform is the input signal u1 with an amplitude of 5V. The third waveform is the voltage of the signal detection resistor Rb2 with an amplitude of 5V. The last waveform is the demodulation signal whose demodulation frequency is 60 kHz which equalizes a data rate of 120 kbps.

Fig. 10(d) shows the power interference on the communication branch circuits when there is wireless power transferring without signal transmission. The first two waveforms are the driving signal of the MOSFETs. The third waveform is the voltage on the parallel resonant tank with an amplitude of 45.5V. The last waveform is the voltage on the signal detection resistor Rb2 which is excited by up. Since the amplitude of the voltage on the signal detection resistor Rb2 is 1.9V, the impact from the power transferring on the signal transfer is acceptable.

In Fig. 10(e), the waveforms of signal transmission with synchronous power transferring are shown. The first waveform is the modulated signal. The second waveform is the voltage of the signal detection resistor Rb2. The third waveform is the demodulation signal. The last waveform is the voltage on the load RL. Based on the amplitude of the load voltage, which is 45V, the output power is calculated to be 101.3W.

 

VI. EXPERIMENTAL VERIFICATION

A. Experimental Setup

A practical ECPT experimental setup, with the parameters shown in Table I and Table II, has been built and tested.

The experimental setup of the proposed ECPT system is presented in Fig. 11. There are four parts: 1) the push-pull inverter; 2) the power resonant tank and signal branches; 3) the rectifier and load; 4) the signal module.

Fig.11The experimental setup of the proposed ECPT system.

B. Experimental Waveforms and Discussion

All of the waveforms in Section IV are tested on the experimental setup and the experimental waveforms are shown in Fig. 12.

Fig.12.Experiment waveforms of ECPT system with communication module.

Comparing Fig.12 (a) with Fig. 12(b), it can be seen that the branch circuits have a small impact on wireless power transfer since there is a negligible change in the amplitudes of the output voltages. The power efficiencies test results of a traditional ECPT system with or without branch circuits are both higher than 85%. In Fig. 12(c) the signal attenuation is small, which indicates that the signal can be transmitted through the system without changes or errors. Fig. 12(d) verifies that power flow has a small impact on the signal transmission since the amplitude of the voltage at the channel terminals is approximately 2.5 times smaller than the amplitude of the signals. Fig. 12(e) shows that the power and signal can be transferred synchronously in the sharing channel due to the slight signal amplitude change and the high amplitude of the load voltage.

Comparing Fig. 10 with Fig. 12, it can be seen that the simulation waveforms are slightly different from the experimental waveforms since the ultra-high frequency of the carrier wave increase the internal resistance of the inductors Lb1 and Lb2. Therefore, the amplitude of the voltage on the detection resistor in Fig. 12(c) is slightly lower than the simulation results in Fig. 10(c).

Moreover, the resonant voltage and the pick-up voltage on the load are both slightly lower than the simulation voltage. This is due to the fact that the stray parameters of the components in the circuit, such as the internal resistance, were not considered in the simulations. However, all of these differences have insignificant impacts on the system and can be ignored. This makes the experimental results match the simulations. In conclusion, both the simulation results and the experimental results verify the correctness and effectiveness of the proposed wireless power and signal transfer method.

In this paper, the application object is an ECPT system whose control signal or detection signal is transmitting and receiving by the microcontroller units (MCU). Generally, MCU of ECPT systems communicate with external devices by serial communication. The standard data rates of serial communications are specific values in the range of 300 bps-115.2 kbps, such as 9600 bps, 19200 bps, and 115200 bps. In this study, over 100W of power and signal with a data rate from 300bps to 120 kbps can be well transferred without data losses. This is sufficient for the data rate of the common serial communication in engineering applications.

C. Further Discussion

All of the work presented in this paper focus on the topology and characteristics of a system and method to transfer power and a signal in a shared channel. If a higher data rate is needed in some applications, the data rate limitation of the proposed system and its influencing factors make it important to measure the capacity of a communication system. Therefore, a study on the data rate limitation is meaningful for further research.

 

VII. CONCLUSION

Based on the traditional ECPT topology for wireless power transfer, a power and signal shared channel was proposed. The operating principle of the designed channel was explained, and the electrical circuit components of the proposed channel were calculated based on an analysis of the mutual interference between the power and signal channel; the power and signal attenuation; and the dynamic characteristic of the signal channel. Both simulations and experimental results verified the correctness and effectiveness of the proposed wireless power and signal transfer method. Finally, the designed channel can simultaneously transfer over 100W of output power and data with a data rate from 300bps to 120 kbps.