A Detection System with Spider Web Coil-Based Wireless Charging and an Active Battery Management System

The article presents a detection system with spider web coil-based wireless charging. Commonly available metal detectors are sold as handheld systems, which enable only progressive, lengthy, time-consuming search. Importantly, a part of the investigated area can thus be easily missed, and the probability that a metal object will not be found increases substantially. This problem, however, is eliminable via the automatic position tracking mode embedded in the solution obtained through our research. The proposed system facilitates using the spider web coil simultaneously for wireless charging and metal detection by pulse induction. The topology of the detector can emit variable pulse lengths, thus allowing the device to detect more types of metal and to adapt itself to the permeability of the soil. The coil has a branch in a relevant part of the winding to reduce undesirable electromagnetic interference during the charging. On the transmitting side of the topology, impedance matching is included to maintain the maximum spatial gap variability. By changing the position of the receiving side, the output voltage changes; therefore, a high efficiency DC/DC converter is employed. The individual battery cells demonstrate different internal resistances, requiring us to apply a new method to balance the cells voltage. The system can be utilized on self-guided vehicles or drones; advantageously, a GPS resending the coordinates to a mesh radio allows for accurate positioning. With the mesh topology, potential cooperation between the multiple systems is possible. The setup utilizes the same coil for wireless power transfer and detection.


Introduction
Wireless charging has been used for more than a hundred years influencing the development of Tesla transformer. At the present, wireless charging can be used in small electronics to charge electric cars and it is heavily involved in electrical engineering department. There is no use of charging connectors, but the number of components increases. The trend is a resonant or inductive charging. Both methods use different topologies and geometric shapes to transfer the energy. The resonance method carries the possibility of charging up to several tens of centimeters [1,2], while induction results in higher efficiency and lower interference [3][4][5].
Pilot's intervention is not necessary as the unmanned aerial vehicle's (UAV) flight path is controlled by the real-time image detection where the variability and a number of various detected waypoints need to be processed. It is possible to perform a completely autonomous flight from take-off of the aircraft to landing [6,7].
Various speed and altitude may be set during the flight and additional accessories may be controlled by the drone. The use of real-time image detection on an autonomous flight of the UAV can serve for package delivery, mapping of large areas, or for military missions and other security forces [8][9][10][11][12].

System topology
The design of the wireless power transfer and pulse induction detection system consists of transmitting and receiving parts (Fig. 1). The first block of the transmitting part is a power source for the whole unit. The power source can be realized by an adapter or a power bank that contains standard power delivery and can supply up to 100 W of power. This standard includes the CYPD3135 [13] chip with the CCG3 standard from Cypress which also has a 32-bit ARM processor and can be used as a station monitor. The chip enables the supply of power to the DC/AC block, which generates an alternating voltage for the serial resonant circuit with transmitting coil L tx and capacitor C tx . The impedance tuning block adjusts the impedance of the transmission resonant circuit L tx and C tx and can improve the transmission efficiency even if the coils are not exactly offset. It also adjusts the mutual inductance M between the coils which is effective by the coupling factor k according to the equation (1) where L rx -receiving coil. The mutual inductance in this equation is given only by the coupling factor k, which varies with the distance of the coils and the current consumption of the receiving part. The degree of interaction between the receiving coil L rx and the transmitting coil L tx is a function of the distance and the positional arrangement.
On the receiving side, the coil is realized as a spider web coil with a threaded branch for charging. Part of the coil can be used for charging called L rx or the whole winding L rx + L det can be used for metal detection. The rectifier makes for rectifying the energy. This energy is adjusted to the required voltage to the battery cells by the DC/DC converter. The battery cells are controlled by an active battery management system (ABMS). To suppress the electromagnetic field, an L rx coil with an optimized flowing current according to the charging power is reversed by means of a transistor Q.
For metal detection, the charging system disconnects part of the charging winding L rx and the entire winding L det is used. Then the signal response is processed by the analog-to-digital converter (ADC) part discussed in the next block diagram, see Fig. 2. The combination of subsequent processing and logging the position saves the data on the charger monitoring. Both parties can have a mesh radio that transmits data to the receiving party. The mesh radio is implemented by the XBEE SX 868 module [14]. The structure of the output data frames depends on the transmission mode. Received data together with coordinates from the navigation module (GPS) can be forwarded to the transmitting party. Then, it is stored via the universal asynchronous receiver-transmitter or serial peripheral interface on a storage device and it can send the coordinates of the detected targets.

Pulse induction detector topology
The basic topology of the pulse induction metal detector in Fig. 3 is detected by the L det detection coil using formed pulses from the HV 7361 module [15]. The module implements T/R switching with a voltage pulse height of ±100 V at the current of 2.5 A with the possibility of operating frequency up to 35 MHz. The module also has adjustable switch between reception and transmission using the input logic gates that shape the pulse.
To adjust the received pulse, the AD8331 variable gain amplifier is used for pulse detection and guarantees a dynamic range throughout the amplifying the signal up to 48 dB due to the programmable gain and input resistance setting [16]. The ADL5511 pulse detector [17] can send an envelope of received signal or RMS voltage to the transmitter. Then the signal is further sampled by the ADC 10-bit AD9200 [18] converter with a maximum sampling frequency of 20 Msps. For a conventional embedded system, signal processing is sufficient for the conventional metal detection. For experiments with more complex processing and visualization of more complex structures, it is necessary to use an array of programmable gate arrays. The processed data can be stored on a storage device or sent by the mesh radio together with the coordinates using the ISM band only if an important object is detected or can be sent permanently.

Design of spider web coil
The spider web coil was widely used in older radio receivers on the long waves up to very short waves (λ = 2000 m to 1 m). Wireless charging works on the similar frequencies. The advantage of this coil is its small parasitic capacity due to the method of winding. Another advantage is the high inductance achieved by the presence of a large surface area by which it can receive or transmit. The coil design is based on the Archimedean spiral, where the mass point rotates around the z-axis into three-dimensional space at an angular velocity ω and starts from the point at time t [19]. The position of the point relative to the z-axis is then as follows: where (vt + c) -modulus of the position vector of the mass point at time t, from which the velocity components v x and v y for the x and y axes are derived. If we integrate the given equations (2) in parts, their parametric expression are as follows: sin , where the point at time t must change sinusoidally. The coil model is plotted in Fig. 3.

Active battery management
It is beneficial to choose ABMS due to the tolerances in the internal resistance or temperature conditions of individual batteries. If an undervoltage occurs on one cell, the balancer disconnects the battery from the load. The advantages of the topology are simplicity and complexity at the expense of the size of the hardware design.
A pair of MOSFETs of Fig. 4 switching the coil is used. The pair is charged from the lower cell for a time given by the inductance of the coil and the voltage of the Cell low cell according to (4) [ ] where the length of the MOSFET switching time δt is given by the influence of inductance L and the inductive energy given by the voltage difference δU and current difference δI. This time is crucial for not exceeding the maximum cell voltage.

Fig. 4. Balancing between individual cells
During this time, the N-FET must be turned on. In the next cycle, the coil is discharged into the Cell high cell via a parallel D high diode. The same principle can be used for charging a Cell low cell from a P-FET. The gates of the transistors can be controlled by a controller or a PWM regulator according to the voltage of individual cells, but mostly by the capacity of the accumulators, from which the size of the inductor and the length of switching are derived.

Simulations
The theoretical inductance is compared with the actual inductance in Tabl. 1. As expected, the coils of these values have large tolerance (usually 20 %). Theoretical values of the multiple threads differ. The first factor is winding of the threads, and the second factor is neglecting of the supply wires. The coil will have different properties for different frequencies.
With the entered parameters, the simulation was performed for a step change in voltage (Fig. 5a). There is 1, 5, 10 and 30 µs of the pulse lengths in the simulation. Fig. 5 visualizes the responses of the coil oscillation due to its parasitic capacity. R 1 forms the series of the winding resistance and C 1 the parasitic capacitance. According to equation (5), the induced voltage on the coil is given by the pulse size of the source V 1 , the ratio of the resistor divider R 1 and R 2 , and the exponential function of the negative pulse t length divided by the transient state of the coil τ ( ) By adding the resistor R 2 in the diagram in Fig. 6a, which simulates the input impedance of the HV7360 module [14], the parasitic capacitance of capacitor C 1 to R 2 is not applied and the voltage peak is induced due to the very fast current dissipation by the coil L 1 . The exact value can be determined accordingly to the equation (5), where τ is the time constant. a b Fig. 6. Coil response with discharge resistance for pulse lengths 1, 5, 10 and 30 µs: a -adding impedance; b -overvoltage peak on the coil without oscillations

Simulation for recharging the cell
For a Li-Ion cell with a capacity of 2500 mAh with a nominal voltage of 3.7 V, the switching time for charging an inductor of 15 µH is equal by the equation (4). Equation (6) provides a calculation of the switching time of transistors δt for a specified Li-Ion cell 6 2.5 15 10 10.13 . 3.7 The discharge time in equation (7) For given calculation, the scheme from Fig. 4 in P-spice was performed followed by the time analysis for the course of the voltage on the Cell high . In the connection, the cells represent resistors R 5 and R 6 because the ideal batteries hold a constant voltage; thereby, it represents an ideal voltage source. The pulse source has parameters according to equation (4). The bottom graph in Fig. 7 shows the voltage peak from 95 mV coil as a result of the mean value of the voltage on the Cell high increases by 220 µV and periodizes at 160 µV from 150 µs until the balancing stops and the Cell high voltage is equal to Cell low . The article describes a proposed topology that allows wireless charging without the need for the precise centering due to the fine tuning with the use of the impedance matching which is not addressed in this article. The proposed charging via the USB-C connector allows charging with the use of standardized adapters or power banks. A coil winding like spider web coil also allows the detection of various objects according to the settings and processing of the amount of pulse energy. The active balancing system makes it possible to transfer the energy between adjacent cells in the event of different internal resistance and thus use their full capacity. Advantageously, it can be used for an autonomous system for easy charging and detection of metal objects.