Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Highly-Efficient Single-Switch-Regulated Resonant Wireless Power Receiver with Hybrid Modulation

Highly-Efficient Single-Switch-Regulated Resonant Wireless Power Receiver with Hybrid Modulation Highly-Efficient Single-Switch-Regulated Resonant Wireless Power Receiver with Hybrid Modulation Kerui Li, Student Member, IEEE, Albert Ting Leung Lee, Member, IEEE, Siew-Chong Tan, Senior Member, IEEE, and Ron Shu Yuen Hui, Fellow, IEEE wireless power receivers [9][15]. Compared with the single- Abstract—In this paper, a highly-efficient single-switch- stage counterpart, the two-stage solution is bulky, costly, regulated resonant wireless power receiver with hybrid modulation is proposed. To achieve both high efficiency and good inefficient and less reliable because it requires more power output voltage regulation, phase shift and pulse width hybrid switches and passive components. In light of this, considerable modulation are simultaneously applied. The soft switching research efforts have been devoted to developing various types operation in this topology is achieved by the cycle-by-cycle phase of single-stage solutions [16]−[21]. In particular, an active full- shift adjustment between the input current and the gate drive bridge rectifier is introduced, which can concurrently achieve signal and also attributed to the reactive components such as the high-frequency rectification and output voltage regulation [21]. series-compensated secondary coil (L , C ) and the parasitic s s Unfortunately, no soft switching operation is allowed, which capacitor of the active switch (C ). The soft switching operation s1 degrades its efficiency and EMI performance. In addition, the also leads to high efficiency and low EMI. By adjusting the duty ratio of the switch, tight regulation of the output voltage can be synchronization between the resonant input current and gate attained. The steady-state and dynamic models of the resonant driving signals becomes more sophisticated at higher switching receiver with hybrid modulation are analytically derived in order frequencies, which may not be suitable for high-frequency to properly design the feedback controller. An experimental setup WPT applications. In [19], even though the half-bridge-based of a two-coil wireless power transfer (WPT) system, including the solution can attain good output regulation, the use of hardware prototype of the proposed receiver, is constructed for complementary high-low-side active switches on the same experimental verification. The experimental results show the bridge leg increases the risk of a direct short between the DC effectiveness of the soft-switching operation in the receiver with a bus voltage and ground, which unavoidably undermines the maximum AC−DC efficiency of 98% while maintaining good reliability and robustness of the receiver. Moreover, the use of regulation of the output voltage, regardless of line and load pulse density modulation produces significant output voltage variations. ripples. A relatively bulky output capacitor is therefore needed Keywords—Semi-active class D rectifier, wireless power transfer to mitigate the output ripples. (WPT), output regulation, resonant rectifier, hybrid modulation. Recently, a number of single-stage single-switch wireless power receivers have emerged [16], [17], [22], which employ I. INTRODUCTION fewer power switches, thus making them more attractive solutions for WPT by offering higher reliability and lower cost. The rapid advancement in wireless power technology has Nonetheless, these prior art have their own limitations. For completely transformed the way we recharge a myriad of example, the duty-cycle-regulated class-E rectifier in [16] battery-powered portable electronic devices because of its suffers from high voltage stresses on the main switch and the added convenience, durability and safety [1][6]. Nowadays, diodes, which inevitably degrades its reliability and potentially an increasing number of consumer electronic devices such as shortens its operating lifetime. In [17], the phase-shift regulated smartphones, smart watches, headsets, and tablets have already class-E rectifier requires complicated design procedures and incorporated the Qi wireless charging function. According to suffers from poor efficiency at light load condition. In [22], the the latest market survey [7], the total shipment of wireless main disadvantages of the multi-cycle-switching-regulated power receivers is more than doubled that of the wireless power active rectifier include the use of a bulky output capacitor, a transmitter by 2022. To cope with the fast-growing market narrow regulation range, and lower efficiency. demand and more stringent requirements in medium-power In this paper, a single-switch-regulated resonant wireless WPT system [8], it is advantageous to develop a wireless power power receiver is proposed. This wireless power receiver receiver that is efficient, compact, low-cost, and reliable for carries the following merits: 1) concurrent high-frequency practical applications. Hence, the motivation of this research AC−DC rectification and DC regulation; 2) a wide range of work is the introduction of a highly-efficient single-switch- output regulation; 3) robust output regulation against line and regulated resonant power receiver with hybrid modulation, load disturbances; 4) soft-switching operation regardless of the which is characterized by a simple circuit structure, low coupling and load conditions; and 5) highly-efficient AC−DC component count, high AC-DC conversion efficiency, and energy conversion. This paper is organized as follows. Section good output voltage regulation. II presents the circuit topology of the proposed single-switch The conventional two-stage topology, which comprises a resonant wireless power receiver and its operating principle. diode bridge rectifier and a regulated DC−DC converter (or Section III discusses the design consideration, particularly, how LDO linear regulator), is more prevalent in commercial to determine the minimum value of the output capacitor in order © 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. to achieve a nearly-constant output voltage. Section IV the equivalent series resistance (ESR) of the passive component compares the proposed wireless power receiver with its can be neglected; 2) The quality factor of the resonant tanks, predecessors. Section V contains the experimental results. namely, (L , C ) and (L , C ), is sufficiently high so that only the p p s s Finally, Section VI concludes this research work. fundamental component of the current is taken into consideration; and 3) The value of the output capacitor C is II. HIGHLY-EFFICIENT SINGLE-SWITCH REGULATED large enough so that a constant DC voltage with reasonably RESONANT WIRELESS POWER RECEIVER SYSTEM small ripple is produced. Due to the inherent property of series-series compensation, A. Circuit Topology the input current of the wireless power receiver i is only a Ls Fig. 1 shows a generic two-coil WPT system, with the function of the fundamental component of the output voltage of emphasis on the proposed wireless power receiver. On the the VSI at the transmitter side and the coupling coefficient M transmitter side, it is assumed that a voltage source inverter [26], [27]. Consequently, i can be treated as a current source Ls (VSI) (e.g. full-bridge/half-bridge inverter), that generates a at the receiver side, which is expressed as square waveform with a fundamental frequency of f and a | | 𝑖 I sin2𝜋 𝑓 𝑡 . (1) series-compensated primary coil (L , C ), where f = 1/T = p p s s 1/2 C L , is used. On the other hand, the wireless power p p By using i as a reference signal, the ideal waveforms of the Ls receiver is made up of the series-compensated secondary coil key signals (i.e., v , v , v , i , and v ) of the single-switch- gs CS1 CD1 CD1 o (L , C , and f = 1/2π C L ) and the single-switch resonant s s s s s regulated resonant wireless power receiver wireless are rectifier circuit, which consists of a power switch (S ), three graphically illustrated in Fig. 2. capacitors (C , C , C ), a diode (D ), a resistive load (R), a S1 D1 o 1 synchronization circuit, and a microcontroller (MCU). State I State II State III State IV State V T =1/f s s Wir eless Power Wir eless Power Rece iver Transmitter r − v + t CS1 i Ls + v − S1 C Cs Cs s S1 Voltage + C R D1 L L CD1 v DT Source p s D o s − o Inverter D1 i i gs Lp Ls t t t v r gs Sync Cs1 + v t Syn . MCU Circu it s1 Fig. 1. Schematic diagram of a generic two-coil WPT system with the proposed wireless power receiver. CD1 Basically, the topology of the proposed receiver is a semi- D1 active class-D rectifier. S and D operate in a complementary 1 1 manner which determines the nominal value of the output voltage. C and C are parasitic capacitors of S and D , S1 D1 1 1 t nT nT +t (n+0.5)T (n+D)T +t (n+D)T +t +t (1+n)T respectively. C is the output capacitor for minimizing the s s f s s f s f r s output voltage ripple. In contrast to conventional class-D synchronous rectifiers for Fig. 2. Key waveforms of the single-switch-regulated resonant DC-DC conversion applications [23]−[25], the proposed wireless power receiver system. wireless power receiver overcomes the following design challenges: 1) Enable synchronization between resonant current In particular, the gate-to-source voltage (v ) of S , i.e., the gate gs 1 and the gate driving waveform of the receiver; 2) Maintain soft drive PWM signal (with a duty ratio of D), has a time delay of switching and good output regulation for a wide range of load t relative to the zero-crossing point of the positive cycle of i . f Ls and coupling coefficient values by employing hybrid By considering the on/off switching state of S and the modulation; 3) Make use of standard steady-state and small- charging/discharging period of C and C , four distinct states S1 D1 signal models to analyze static and dynamic performance of the can be defined for the receiver, namely, State I for nT ≤ t < (nT s s closed-loop system; and 4) Achieve concurrent high-frequency + t ), State II for (nT + t ) ≤ t < (n + D)T + t , State III for (n + f s f s f AC−DC rectification and DC output regulation by employing D)T + t ≤ t < (n + D)T + t + t , and State IV for (n + D)T + t s f s f r s f the model-based controller design method. + t ≤ t < (1 + n)T , where t and t are the fall and rise time of r s f r the voltage across C (i.e., V ), respectively, n is an arbitrary S1 Cs1 B. Operating Principles nonnegative integer number, and D is the on-time duty ratio of In the ensuing discussion, the following assumptions are the switch. Fig. 3(a)(d) shows the equivalent circuit model in made: 1) For simplicity, ideal circuits are assumed and hence, each of the four states. i charges C and also discharges C at the same time. By Ls D1 S1 − v + CS1 using KCL and KVL, we have CD2 C S1 L C s s 𝑑𝑣 𝑑𝑣 (2) 𝐶 𝐶 𝑖 i + 𝑑𝑡 Ls C R D1 v 𝑣 𝑣 𝑣 . v D o CD1 1 CD1 Meanwhile, the output capacitor C supplies current to the load (a) State I R. Mathematically, we can write − v + CS1 𝑑𝑣 𝑣 C (3) 𝐶 . L C S1 s s S 𝑑𝑡 𝑅 S1 i + Ls C R D1 State II [(nT + t ) ≤ t < (n + 0.5)T ]: s f s D o CD1 1 C Fig. 3(b) contains the equivalent circuit model of the receiver circuit in State II. At t = (nT + t ), v returns to zero volts and s f Cs1 afterwards, S is turned on at the rising edge of v . This enables 1 gs (b) State II zero-voltage-switching (ZVS) turn-on operation. After S is − v + switched on, v and v are clamped at zero volts and v , CS1 CD1 o CS1 respectively, i.e., v (nT + t ) = 0 and v (nT + t ) = v . By Cs1 s f CD1 s f o L C S1 s s substituting these two conditions into (2) and solving for t , where tf is the time interval of State I (tf), we have S1 Ls C R D1 v D o CD1 1 𝑡 . (4) 𝜋 𝑓 |I | (c) State III − v + State III [(n + 0.5)T ≤ t < (n + D)T + t ]: s s f CS1 Fig. 3(c) contains the equivalent circuit model of the receiver CD2 C S1 L C s s circuit in State II. At t = (n+0.5)T , i returns to zero and s Ls becomes negative, while S remains on. During State II and III, + i transfers the input energy to both the output capacitor C and i + Ls o Ls C C R D1 o v the load R via S , which can be written as D o 1 CD1 1 CD1 𝑑𝑣 𝑣 𝐶 𝑖 . (5) 𝑑𝑡 𝑅 (d) State IV To allow proper regulation of the output voltage (v ), the range − v + CS1 of the duty ratio (D) is defined as L C S1 s s (6) 𝑓 𝑡 𝐷 1 2 𝑓 𝑡 . i + Ls C C R D1 o v D CD1 1 State IV [(n + D)T + t ≤ t < (n + D)T + t + t ]: s f s f r D1 Fig. 3(d) shows the equivalent circuit model of the receiver circuit in State III. At t = (n + D)T + t , S is turned off with s f 1 (e) State V zero voltage. Afterwards, i , which is in the negative half cycle, Ls starts discharging C and charging C simultaneously. The D1 S1 Fig. 3 Equivalent circuit model in (a) State I; (b) State II; (c) State charging procedure can thus be mathematically represented as III; (d) State IV, (e) State V. 𝑑𝑣 𝑑𝑣 𝐶 𝐶 𝑖 State I [nT ≤ t < (nT + t )]: . s s f 𝑑𝑡 (7) 𝑣 𝑣 𝑣 Fig. 3(a) shows the equivalent circuit model of the receiver circuit in State I. At t = nT , the diode current i drops to zero s CD1 and stops conducting naturally, resulting in a zero-current- Meanwhile, C continues to supply current to the load R. Hence, switching (ZCS) turn-off operation of the diode. When t > nT , we have 𝑑𝑡 𝑣 𝐶 𝑑𝑡 -7 𝑑𝑣 𝑣 10 5.5 (8) 𝐶 . State V [(n + D)T + t + t ≤ t < (1 + n)T ]: s f r s Fig. 3(e) contains the equivalent circuit model of the receiver 4.5 circuit in State IV. At t = (n + D)T + t + t , v returns to zero s f r CD1 volts and D starts conducting. After D is forward biased, v 1 1 CD1 and v are clamped at v and zero volts, respectively, i.e., CD1 o 3.5 v [(n + D)T + t +t ] = v and v [(n + D)T + t +t ] = 0. By Cs1 s f r o CD1 s f r substituting these two conditions into (6), t can be obtained as 1 𝐷 1 2.5 𝑡 𝑡 arccos cos 2𝜋 𝑓 𝑡 00.2 0.4 0.6 0.8 1 𝑓 2𝜋 𝑓 C (F) -8 D1 10 2𝜋 𝑓 𝐶 𝐶 𝑣 (9) (b) t versus C f D1 |I | -7 5.5 From Fig. 2, Q and Q represent the total charge transferred f r during the switching intervals t and t , respectively. Since Q = f r f 5 Q , the average current of i in State III is higher than that in r LS 4.5 State I. Hence, t becomes larger than t . Note that in this f r particular state, when D conducts, i freewheels through D 1 Ls 1 and no energy is transferred to the output. C continues to 3.5 discharge its current to the load R. By KCL, we can write 𝑑𝑣 𝑣 3 (10) 𝐶 . 2.5 0 0.2 0.4 0.6 0.8 1 C (F) -8 At t = (1 + n)T , the circuit enters State I of the next switching s S1 10 period and the aforementioned state transitions will repeat. (c) t versus C f S1 -7 4.3 C. Steady-State Model 4.25 By invoking state-space averaging on the output capacitor C , 4.2 the relationship between the duty ratio (D) and the output 4.15 voltage (v ) based on the steady-state model can be analytically derived as 4.1 𝑑𝑣 1 𝑣 𝐶 𝑖 4.05 𝑇 𝑅 (11) |I | 𝑣 1.9 1.95 2 2.05 2.1 cos 2𝜋 𝑓 𝑡 cos 2𝜋𝐷 2𝜋 𝑓 𝑡 . f (Hz) 5 2𝜋 𝑅 (d) t versus f f s Since 0 in steady-state condition, the L.H.S. of (11) Fig. 4. Plots of output voltage v versus duty cycle D, t versus o f becomes zero and hence, by re-arranging, v can be obtained as o C , t versus C and t versus f . D1 f S1 f s |I |𝑅 𝑣 cos 2𝜋 𝑓 𝑡 cos 2𝜋𝐷 2𝜋 𝑓 𝑡 . (12) Fig. 4 shows plot of v against D based on (12) and plots of t o f 2𝜋 versus C , C , and t . Clearly, v can be regulated by adjusting D1 S1 f o the variable D. It is observed that the maximum voltage is achieved if D = 0.5 − f t for a given value of R. The fact that s f 1/π│I │R Ls max ×cos(2πf t ) the output voltage is highly dependent on R implies that a s f closed-loop system is needed in order to achieve tight regulation of the output voltage. The value of t increases as C f D1 1/π│I │R and C increased, while t drops as f increased. Ls min S1 f s ×cos(2πf t ) s f R↑ Fig. 5 shows the relationship between the output voltage (v ) and the duty ratio (D) with and without resonant capacitors (C , S1 C ). The latter is referred to as the ideal case since an ideal D1 switch (or diode) contains no parasitic capacitances, i.e., C = S1 0.5−f t 1−2f t s f s f 0 (or C = 0). By comparing the two curves in Fig. 5, the output D1 (a) Plots of output voltage v versus duty cycle D at voltage (v ) with resonant capacitors (i.e., the bottom curve) is different values of R. reduced by 1/π│I │R ×(1 − cos(2πf t )), compared with that Ls max s f t (s) t (s) t (s) 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 without resonant capacitors (i.e., the top curve). In addition, the Hence, the minimum value of the output capacitor is given by minimum and maximum value of D is shifted by f t and 2f t , s f s f |I | respectively, as illustrated in Fig. 5. 𝐶 . (16) 𝑥%𝑣 𝜋 𝑓 C =C =0 S1 D1 1/π│ILs│Rmax ×(1−cos(2πf t )) 1/π│I │R s f Ls max C ≠0, C ≠0 Design of C and C . Since C and C are the parasitic S1 D1 D1 S1 D1 S1 1/π│I │R Ls max capacitance of the diode D and switch S , the values can be 1 1 ×cos(2πf t ) s f obtained from their datasheets. B. Derivation of the Small-Signal Model and Feedback Control for Output Voltage Regulation f t s f 2f t s f Theoretical result Sim ulation result D 10 0.5−f t 0.5 s f 1−2f t 1 s f ‐10 Fig. 5. Plots of output voltage (v ) versus duty ratio (D) with and ‐20 without resonant capacitors (CS1, CD1) at the maximum load Rmax. ‐30 10 100 1000 10000 Frequency (Hz) III. DESIGN CONSIDERATIONS AND IMPLEMENTATION (a) Bode magnitude plot ‐180 A. Design of Reactive Components Theoretical result ‐200 Simulation result Design of L and C . The design objective of L is that the quality s s s ‐220 factor Q of the coil is higher than the desired quality factor Q, Ls ‐240 i.e. Q ≥ Q. Given that the equivalent series resistance (ESR) Ls ‐260 of L is R , Ls is designed as s Ls-ESR ‐280 10 100 Frequency (Hz) 1000 10000 2𝜋 𝑓 𝐿 𝑄𝑅 𝑄 𝑄 ⇒ 𝐿 . (13) (b) Bode phase plot 𝑅 2𝜋 𝑓 Correspondingly, due to the use of series-series compensation, Fig. 7. Theoretical and simulated Bode plots of the open-loop the capacitor C is sized as control-to-output voltage using the small-signal model By linearizing (11) and considering the AC perturbation of duty 𝐶 . (14) 2𝜋 𝑓 𝐿 cycle D, the resulting small-signal control-to-output linearized equation can be written as Deign of the Output Capacitor (C ). The design objective of the 𝑑𝑣 𝑣 output capacitor is to maintain a constant output voltage v with 𝐶 |I |𝐷 sin 2𝜋𝐷 2𝜋 𝑓 𝑡 . (17) sufficiently small voltage ripple ∆v , i.e., ∆v ≤ x%  v . Fig. 6 𝑑𝑡 𝑅 o o o provides a graphical illustration of the root cause of the output Fig. 7 shows the Bode plots of the theoretical and simulated voltage ripple. small-signal models at D = 0.5,│I │= 1 A, R = 30 Ω, f t = 0.1, Ls s f and C = 100 µF. As is evident in Fig. 7(a) and (b), the theoretical and simulation results are in close agreement, ∆v v o thereby validating the accuracy of the derived small-signal equation from (15). Also, the phase plot in Fig. 7(b) shows that the phase of the open-loop system decreases from −180 to −270. It is important to note that the use of conventional Q proportional-integral (PI) controller with positive proportional ripple Ls and integral coefficients (k and k ) is unable to provide p i v / R adequate phase boosting for the uncompensated system, which t t 1 2 leads to unstable transient response because of insufficient phase margin and unacceptably large steady-state error due to relatively low DC gain [28]. To address this issue, a modified Fig. 6. Illustration of the root cause of the output voltage ripple. proportional-integral (PI) controller, as shown in Fig. 8, with Basically, the rise of the output voltage is attributed to the negative values of k and k is employed to attain accurate and p i stable regulation of the output voltage. In addition, an anti- accumulation of charge Q , which is given by ripple windup scheme is used to prevent overflow of the integrator and ∆𝑣 𝑥%𝑣 to ensure linear operation of the PI controller. Specifically, the 𝑖 𝑑𝑡 anti-windup loop is added to avoid integrator wind-up, which 𝑄 |I | (15) can occur when the duty ratio (D) is saturated. In other words, 𝐶 𝐶 𝜋 𝑓 𝐶 Phase (degree) Magnitude (dB) it prevents the actual value of D from exceeding beyond its Ls + GPIO32 Sy nc upper and lower limits, as defined in equation (6). Sy nc Duty cycle (50%) PWM1 Controller Anti-windup InSYNC │k│ i Frequency Synchroni zation Duty cycle D Con troller V v ref gs z Gain −│k│ V i ∫ ref t 2πf PWM2 f s Eq. (4) o + Phase Eq. (6) Gain Hyb rid Mod ul ation −│k │ TMS320F28335 PI controller Fig. 10. Block diagram of the frequency synchronization and Fig. 8. Block diagram of the modified PI controller with anti- hybrid modulation. windup. As can be seen in Fig. 10, an external comparator, which is part The proper values of the proportional gain k and integral gain of the synchronization circuit, is employed to detect the zero- k of the PI compensator can be determined as follows. crossing points of i and generate the corresponding square Ls 2𝜋 𝑓 𝐶 𝑘 waveform labelled Sync [see Fig. 11]. 𝑘 ,𝑘 . (18) Frequency Synchronization | | 𝑅𝐶 I sin 2𝜋𝐷 2𝜋 𝑓 𝑡 𝑠 T Ls t where f is the desired crossover frequency. Fig. 9 shows the resulting Bode plots of the open-loop transfer function and the Sync closed-loop transfer function of the system with k = −1.07 and TBCTR1 k = −356. The numerical results of the closed-loop transfer CMP1(50%) 0.5T function show that the crossover frequency is at 1 kHz with a s PWM 1 phase margin of 90 and a low-frequency gain (at 10 Hz) is 40 0.5DT +t Hybrid Modulation dB. The phase margin of 90 ensures a stable closed-loop s f CMP2(D) response and the static gain at DC is high enough (> 40 dB) to TBCTR2 eliminate the steady-state error. PWM2 50 DT (v ) gs Open‐loop results t Closed‐loop results Fig. 11. Ideal waveforms of the key signals in frequency synchronization and hybrid modulation. ‐10 ‐20 The MCU receives this external synchronization signal Sync via ‐30 10 100 Frequency (Hz) 1000 10000 the GPIO32 pin, which is then used for EPWM synchronization. The rising edge of Sync triggers the counter of (a) Bode magnitude plot the PWM1 (TBCTR1) by counting from zero on a cycle-by- Open‐loop results cycle basis, which leads to a triangular carrier waveform for ‐50 Closed‐loop results PWM1. The duty cycle of the PWM1 is set to be at 50%. ‐100 Consequently, PWM1 is synchronized with i with a constant ‐150 Ls ‐200 phase difference of 0.25T between them. In this way, frequency ‐250 synchronization between i and D is realized. LS ‐300 Subsequently, hybrid modulation is implemented on the 10 100 Frequency (Hz) 1000 10000 PWM2 module using the PWM1 as a reference. The duty cycle (b) Bode phase plot D, which is obtained from the controller output [see Fig. 10], is used to compare with the counter of the PWM2 module Fig. 9. (a) Bode magnitude plot and (b) bode phase plot of the open- TBCTR2 in order to adjust the pulse width of the PWM2. On loop and closed-loop transfer function. top of them, the phase-shift modulation is realized by adjusting the phase difference between PWM1 and PWM2 (based on the C. Synchronization and Hybrid modulation internal synchronization signal InSYNC). The desired time Fig. 10 shows the functional block diagram of the frequency delay t is calculated from equation (4). To reduce the synchronization and hybrid modulation scheme. The frequency computation complexity, the real-time output voltage v and synchronization is implemented by using the PWM1 module │I │ in equation (4) are replaced by a fixed output voltage Ls whereas the hybrid modulation is realized by using the PWM2 reference V and a constant value, respectively. Since the ref module of the microcontroller (part number: TMS320F28335). carrier waveform TBCTR2 is triangular in shape, the center of Fig. 11 shows the ideal waveforms of the key signals in the resulting PWM2 is thus aligned with the valley of TBCTR2. frequency synchronization and hybrid modulation. Hence, to produce a phase shift of t , TBCTR2 is simply shifted Magnitude (dB) Phase (degree) by 0.5DT +t . After performing normalization of the phase s f TABLE II PARAMETERS OF COMPONENTS angle φ of PWM2 by 2π, the effective value of φ is given by Part Value 𝜑 2𝜋 𝑓 𝑡 𝐷𝜋 . C , C 4.5 nF D1 S1 (19) Ls 172 µH Fig. 10 shows the implementation of (17) in the MCU. Cs 3300 pF + 330 pF Co 1000 µF IV. COMPARATIVE STUDY TABLE III. PART NUMBER OF COMPONENTS. Table I compares the proposed receiver with the prior art of Part Part Number single-power-switch solutions [16], [17], [22] in terms of CD1, CS1 Parasitics of TK56A12N1 topologies, compensation of receiver coil, resonant frequency, Custom-made L circular copper air coil etc. As clearly shown in Table I, the proposed receiver uses only s diameter = 29 cm one power switch and achieves fully soft-switching operation B32682A7332K000 (3300 pF) Cs on the switch and diode. The maximum AC-DC conversion PHE448SB3330JR06 (330 pF) efficiency reaches 98%, which is higher than its predecessors. C UVZ1H102MHD The regulation frequency is identical to the resonant frequency, Gate Driver ADuM3223 Comparator LM 393P which eliminates the use of bulky reactive components. The Current Transformer AS-100 switch-diode bridge structure can prevent the direct short circuit MOSFET S TK56A12N1 of the output terminals, thus enhancing the system reliability. Diode D Body diode of TK56A12N1 Indeed, these features make it a very competitive solution for Microcontroller TMS320F28335 future WPT applications [3]. Compared with the existing class- E-based solutions [16], [17], a significant advantage of the Oscilloscope Primary proposed receiver is that it has relatively low voltage stresses Secondary coil L p coil L on the power switch. Even though the proposed solution s requires an additional diode when compared with [17], the relatively low voltage stresses on both the switch and diode can Differential voltage probe justify the slight increase in the cost. Compared with the semi- active class-D with pulse density modulation [22], the benefit of the proposed receiver is that it can achieve ZVS turn-on for MCU Current the power switch, which results in higher efficiency and much probe higher output power. Additionally, the proposed solution Full bridge allows the use of a much smaller output capacitance due to the Current Wireless power MCU Inverter transformer relatively high regulation frequency. Receiver Fig. 12. Experimental setup (including the prototype of the proposed receiver). TABLE I. COMPARISON WITH EXISTING WPT RECEIVERS. Proposed [16] [17] [22] V. EXPERIMENTAL RESULTS Semi-active Modified Active Semi-active Topology A hardware prototype of the proposed WPT receiver with a Class D Class E Class E Class D Compensation Series Series Series Series switching frequency of 200 kHz switching frequency is of receiver coil compensation compensation compensation compensation constructed for experimental verification. The nominal output Resonant 200 kHz 6.78 MHz 200 kHz 1 MHz voltage is 24 V and the maximum power is 16 W. Fig. 12 shows frequency a photo of the whole experimental setup including the prototype, Regulation 200 kHz 6.78 MHz 200 kHz 50 kHz frequency in which the DP832 power suppliers, DSOX3204T oscilloscope, Number of N2790A differential voltage probe, and 1147B current probe 1 1 1 1 power switches are used. Table II lists the design parameters and part numbers Number of 1 1 0 1 of the key components used in the prototype. power diode Maximum V ≈3V ≈3V V o o o o A. Steady-State Performance Voltage stress Hybrid Phase-shift Pulse Density Modulation PWM Fig. 13 (a) and (b) shows the key waveforms of the proposed modulation Modulation Modulation receiver in steady-state condition with v = 24 V and i = 0.63 Implementation o o Yes No No Yes ease A. The equivalent load resistance R is 38.09 Ω. The off-state ZVS for ZVS for ZCS for time interval of v and v are measured to be around 2.66 µs CD1 CS1 Soft switching ZVS for switch switch switch and operation switch and 2 µs, respectively. Hence the duty cycle D is around 0.532. ZCS for diode ZCS for diode diode The current of the coil i lead 0.5π against compensated Output voltage, Ls 24 V, 16 W 10 V, 17 W 24 V, 16 W 3.1 V, 96 µW power capacitor voltage v . The voltage stress of the diode is Cs Maximum 98 % (Power 92 % (Power 93 % (Power measured to be 24 V while the measured peak-to-peak input N/A Efficiency stage) stage) stage) current is 4.7 A. As a sanity check, by substituting the above values of D, t , i , and R into (13), the resulting theoretical f Ls output voltage is 24.56 V, which agrees very closely with the Fig. 15 shows the ZVS operation of the switch and ZCS measured value of 24 V. operation of the diode. Fig. 15(a) shows the turn-on transition of the switch. As soon as i enters in its positive cycle, the Ls 2.64 µs 24 V v voltage of the switch v starts dropping to zero. After v CD1 CS1 CS1 reaches zero, the rising edge of v is applied to the switch which gs Ls turns it on completely. Hence, the switch is turned on with ZVS. 4.7 A The measured falling time t is 336 ns. As a sanity check, by substituting the corresponding values of v , i , C and C into o Ls D1, S1 vo 24 V Eq. (4), the theoretical value is 382 ns, which is in good io agreement with the measured value. Fig. 15(b) shows the ZCS 0.63 A turn-off transition of the diode. At the end of the conduction state of the diode, i , which flows through the diode, reaches Ls (a) zero. Hence, the diode is turned off automatically with ZCS. After a lapse of t = 336 ns, v finally reaches 24 V. f CD1 2.00 µs 24 V CD1 CS1 CS1 ZVS 0.5π vgs Ls t = 336 ns Cs Ls 4.7 A (b) i S1 Fig. 13. Steady-state waveforms of v , i , v , v , v and i at a CD1 Ls o Cs CS1 o rated output power of 15 W. (a) ZVS turn-on of the switch. Fig. 14 shows the Fast Fourier Transformation (FFT) and t = 336 ns THD analysis of the measured waveforms of v and i . The v CD1 Ls CD1 fundamental current and THD of the resonant current i are Ls rd th th 2.27 A and 2.33%. The high-order (3 , 5 , 7 , …) harmonic ZCS Ls components are suppressed by the wireless power coils. The fundamental voltage and THD of v are 15.02 V and 46.30%, CD1 D1 respectively. Since the duty cycle is greater than 50%, the even- order harmonic components are produced, thereby degrading the THD performance. Since i contains very limited harmonic Ls (b) ZCS turn-off of the diode. components, the assumption that i is purely sinusoidal in the Ls Fig. 15. Soft switching operations of the switch and diode. theoretical analysis remains valid. 0 Fig. 16 shows the measured output voltage and efficiency -1 values of the rectifier power circuit across different output -2 -5 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 ×10 Time (s) power under output voltage regulation. The nominal output 2.5 Fundamental (200000Hz) = 2.27 voltage is 24 V. The maximum steady-state error of the THD= 2.33% 1.5 regulated output voltage is only 0.1 V, which is around 0.43% of the reference voltage as the power varies from 20% load 0.5 power to full power. The maximum efficiency of the converter 01 234 5678 9 10 Harmonic order is around 98%, which is achieved at full load condition. At light (a) FFT analysis of iLs. load condition, the efficiency only reduces slightly to 94%. 15 100.00% 24.15 97.50% 24.1 -5 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 ×10 Time (s) 95.00% 24.05 Fundamental (200000Hz) = 15.02 92.50% 24 THD= 46.30% 90.00% 23.95 20.00% 40.00% 60.00% 80.00% 100.00% 0 Pe rce ntage of the nominal powe r 0 1 2 3 4 5678 9 10 Harmonic order (b) FFT analysis of vCD1. Fig. 16. Measured efficiency and output voltage of the proposed Fig. 14. FFT analysis of the key AC waveforms. receiver system. Current (A) Current (A) Voltage (V) Voltage (V) Efficiency Votlage (V) Fig. 17 shows the measured output voltage v and amplitude of i of power stage versus coils’ distance with output voltage Ls regulation and an output power of 10 W. As the separation Ls distance between the primary and secondary coils increases 8 ms from 10.5 cm to 21.5 cm, the amplitude of i increases from Ls 1.45 A (at 10.5 cm) to 2.6 A (at 21.5 cm). The maximum voltage vo 0.6 V regulation error is 0.1 V (which is 0.43% of the reference 0.688 A voltage). Hence, the experimental results in either Fig. 16 or Fig. io 17 indicate that the steady-state output voltage regulation error (a) is relatively small over a wide line or load range. 24.15 3 Ls 24.1 2.25 8 ms 24.05 1.5 0.6 V vo 24 0.75 23.95 0 10 11 12 13 14 15 16 17 18 19 20 21 io Distance (cm) (b) Fig. 17. Output voltage vo and amplitude of iLs of power stage versus the coils’ separation distance with output voltage regulation. Fig. 19. Measured waveforms of the rectifier in response to step changes in the output power with output voltage regulation. B. Dynamic Performance Fig. 20 (a) and (b) show the measured waveforms of the Fig. 18 shows the measured waveforms of the rectifier rectifier in response to the input current i changes in light-load Ls operating with synchronization on/off transition. Prior to the condition. When peak-to-peak value of i increases from 2 A Ls enabling of synchronization, the waveforms of the rectifier have to 3.7 A in 10 ms, the output voltage experiences an overshoot significant fluctuations. As soon as synchronization is enabled, of 0.325 V (i.e., 1.35% of the nominal voltage). Likewise, when the output voltage ramps up to the nominal value of 24 V within the peak-to-peak value of i drops from 3.7 A to 2 A in 10 ms, Ls 69 ms. Note that there is no overshoot during the start-up the undershoot is also 0.3 V. The measured dynamic responses process. Besides, any undesirable low-frequency fluctuations in corroborate the robustness and stability of the output voltage the waveforms of the rectifier due to frequency asynchrony are regulation against variations in load and input current. eliminated. Therefore, the experimental results demonstrate the effectiveness of the synchronization scheme. 2 A 3.7 A Ls CD1 i 325 mV Ls vo io 10 ms 24 V io vo 69 ms 159 mA Syn. OFF Syn. ON (a) Fig. 18. Measured waveforms of the rectifier operating with 3.7 A 2 A synchronization on/off transition. Ls Fig. 19 (a) and (b) depict the measured waveforms of the 325 mV rectifier operating with respect to the step-current changes under output voltage regulation. When the output power is vo increased from 0 W to 16 W, the output voltage experiences a 10 ms 0.6 V dip (which is 2.5% of the reference output voltage). The io 159 mA settling time for the step-up load change (i.e., from no load to full load) is 8 ms. Conversely, when the output power is reduced (b) from 16 W to 0 W, the resulting overshoot is 0.6 V. The Fig. 20. Measured waveforms of the rectifier in response to step corresponding settling time for the step-down load change is 8 changes in the input current with output voltage regulation. ms. Voltage (V) Curreent (A) [13] Fact sheet: WPR1516, “Medium power wireless charger receiver VI. CONCLUSIONS 15 W WPC Compliant Solution,” NXP Semiconductors, 2014, [Online], Available: In this paper, a highly-efficient single-switch-regulated https://www.nxp.com/docs/en/fact-sheet/WPR1516FS.pdf resonant wireless power receiver system with hybrid [14] Y. Fang, B. M. H. Pong, and R. Hui, “An enhanced multiple modulation is presented. The hybrid modulation scheme with harmonics analysis method for wireless power transfer systems,” phase-shift and pulse width modulations is employed to IEEE Trans. Power Electron., pp. 1205–1216, 2019. simultaneously achieve very high efficiency and good output [15] The Wireless Power Consortium, “wireless power the qi wireless regulation. A comparative study with the existing single-switch power transfer system power class 0 specification part 4: wireless power rectifiers highlights the conspicuous advantages reference designs,” pp. 1–160, 2016. [16] M. Liu, J. Song, and C. Ma, “Active class E rectifier for dc output of the proposed rectifier, which include low component count, voltage regulation in megahertz wireless power transfer systems,” simple design procedure, high efficiency, and cycle-by-cycle IEEE Trans. Ind. Electron., vol. 67, no. 5, pp. 3618–3628, May output regulation. A hardware prototype of the entire WPT system has been constructed and tested for experimental [17] K. Li, S. C. Tan, and R. Hui, “Single-switch-regulated resonant verification. The experimental results clearly demonstrate that wpt receiver,” IEEE Trans. Power Electron., vol. 34, no. 11, pp. high efficiency and accurate output voltage regulation of the 10386–10391, Nov. 2019. proposed resonant receiver is achievable. The measurement [18] L. Cheng, X. Ge, W. C. Ng, W.-H. Ki, J. Zheng, T. F. Kwok, C.- results also confirm good transient responses of the output Y- Tsui, and M. Liu, “A 6.78 MHz 92.3%-peak efficiency single- voltage under output voltage regulation. stage wireless charger with cc-cv charging and on-chip bootstrapping techniques,” 2019 Symposium on VLSI Circuits, pp. 12, Jun. 2019. REFERENCES [19] H. Li, J. Fang, S. Chen, K. Wang, and Y. Tang, “Pulse density [1] Z. Zhang, A. Georgiadis, and C. Cecati, “Wireless power modulation for maximum efficiency point tracking of wireless transferan overview,” IEEE Trans. on Ind. Electron., vol. 66, no. power transfer systems,” IEEE Trans. Power Electron., vol. 33, 2, pp. 10441058, Feb. 2019. no. 6, pp. 5492–5501, 2018. [2] C. T. Rim and C. Mi, Wireless Power Transfer for Electric [20] L. Cheng, W.-H. Ki, C.-Y. Tsui, “A 6.78-MHz single-stage Vehicles and Mobile Devices. Hoboken, NJ, USA: Wiley-IEEE wireless power receiver using a 3-mode reconfigurable resonant Press, 2017. regulating rectifier,” IEEE Journal of Solid-State Circuits, vol. [3] S. Y. R. Hui, W. Zhong, and C. K. Lee, “A critical review of 52, no. 5, pp. 14121423, Feb. 2017. recent progress in mid-range wireless power transfer,” IEEE [21] T. Diekhans and R. W. De Doncker, “A dual-side controlled Trans. on Power Electron., vol. 29, no. 9, pp. 45004511, Sep. inductive power transfer system optimized for large coupling 2014. factor variations and partial load,” IEEE Trans. Power Electron., [4] G. A. Covic and J. T. Boys, “Inductive power transfer,” Proc. vol. 30, no. 11, pp. 6320–6328, 2015. IEEE, vol. 101, no. 6, pp. 12761289, Jun. 2013. [22] H. S. Gougheri and M. Kiani, “Current-based resonant power delivery with multi-cycle switching for extended-range inductive [5] I. Mayordomo, T. Drager, P. Spies, J. Bernhard, and A. Pflaum, power transmission,” IEEE Trans. Circuits Syst. I Regul. Pap., “An overview of technical challenges and advances of inductive vol. 63, no. 9, pp. 1543–1552, 2016. wireless power transmission,” Proc. IEEE, vol. 101, no. 6, pp. 13021311, Jun. 2013. [23] M. K. Kazimierczuk, M. J. Mescher, and R. M. Prenger, “Class D [6] A. P. Sample, D. A. Meyer, and J. R. Smith, “Analysis, current-driven transformer center-tapped controllable synchronous rectifier,” IEEE Trans. Circuits Syst.−I: experimental results, and range adaptation of magnetically Fundamental Theory and Appl., vol. 43, no. 8, pp. 670−680, Aug. coupled resonators for wireless power transfer,” IEEE Trans. Ind. Electron., vol. 58, no. 2, pp. 544–554, 2011. [24] M. K. Kazimierczuk and M. J. Mescher, “Class D converter with [7] Dinesh Kithany, “Wireless Power Market Tracker,” IHS Markit half-wave regulated synchronous rectifier,” IEEE Appl. Power Surv., 2019. Electron. Conf. Expo. (APEC), vol. 2, pp. 1005–1011, 1994. [8] The Wireless Power Consortium, “Medium power wireless [25] M. Mikotajewski, “Class D synchronous rectifiers,” IEEE Trans. charging: emerging applications, considerations and Circuits Syst., vol. 38, no. 7, pp. 694–697, Jul. 1991. opportunities.” [26] Y. Zhang, T. Kan, Z. Yan, and C. C. Mi, “Frequency and voltage [9] Datasheet, STWLC68, “STWLC68 Qi-compliant inductive tuning of series-series compensated wireless power transfer wireless power receiver for 5 W applications,” system to sustain rated power under various conditions,”, IEEE STMicroelectronics, 2020, [Online] Available: Journal of Emerging and Selected Topics in Power Electron., vol. https://www.st.com/resource/en/datasheet/stwlc68.pdf 7, no. 2, pp. 1311−1317, Jun. 2019. [10] Datasheet, P9222-R, “Wireless power receiver for low power applications”, Renesas Electronics Corp., 2019, [Online] [27] W. Zhang and C. C. Mi, “Compensation topologies of high-power Available: https://www.idt.com/us/en/document/dst/p9222-r- wireless power transfer systems,” IEEE Trans. Veh. Technol., vol. 65, no. 6, pp. 4768–4778, 2016. datasheet [28] R. W. Erickson and D. Maksimovic, “Fundamentals of power [11] Datasheet, BQ51003, “BQ51003 highly integrated wireless electronics,” Springer, 2006. receiver Qi (WPC v1.2) compliant power supply,” Texas Instrument Inc., 2018, [Online] Available: http://www.ti.com/lit/ds/symlink/bq51003.pdf [12] User’s Guide, WPR1500-BUCK, “WPR1500-buck MP receiver v2.1 reference design user’s guide,” NXP Semiconductors, 2016, [Online], Available: https://www.nxp.com/docs/en/user- guide/WPR1500BUCKMPUG.pdf http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Computing Research Repository arXiv (Cornell University)

Highly-Efficient Single-Switch-Regulated Resonant Wireless Power Receiver with Hybrid Modulation

Loading next page...
 
/lp/arxiv-cornell-university/highly-efficient-single-switch-regulated-resonant-wireless-power-N3r7t0hitx
ISSN
2168-6777
eISSN
ARCH-3344
DOI
10.1109/JESTPE.2020.2986372
Publisher site
See Article on Publisher Site

Abstract

Highly-Efficient Single-Switch-Regulated Resonant Wireless Power Receiver with Hybrid Modulation Kerui Li, Student Member, IEEE, Albert Ting Leung Lee, Member, IEEE, Siew-Chong Tan, Senior Member, IEEE, and Ron Shu Yuen Hui, Fellow, IEEE wireless power receivers [9][15]. Compared with the single- Abstract—In this paper, a highly-efficient single-switch- stage counterpart, the two-stage solution is bulky, costly, regulated resonant wireless power receiver with hybrid modulation is proposed. To achieve both high efficiency and good inefficient and less reliable because it requires more power output voltage regulation, phase shift and pulse width hybrid switches and passive components. In light of this, considerable modulation are simultaneously applied. The soft switching research efforts have been devoted to developing various types operation in this topology is achieved by the cycle-by-cycle phase of single-stage solutions [16]−[21]. In particular, an active full- shift adjustment between the input current and the gate drive bridge rectifier is introduced, which can concurrently achieve signal and also attributed to the reactive components such as the high-frequency rectification and output voltage regulation [21]. series-compensated secondary coil (L , C ) and the parasitic s s Unfortunately, no soft switching operation is allowed, which capacitor of the active switch (C ). The soft switching operation s1 degrades its efficiency and EMI performance. In addition, the also leads to high efficiency and low EMI. By adjusting the duty ratio of the switch, tight regulation of the output voltage can be synchronization between the resonant input current and gate attained. The steady-state and dynamic models of the resonant driving signals becomes more sophisticated at higher switching receiver with hybrid modulation are analytically derived in order frequencies, which may not be suitable for high-frequency to properly design the feedback controller. An experimental setup WPT applications. In [19], even though the half-bridge-based of a two-coil wireless power transfer (WPT) system, including the solution can attain good output regulation, the use of hardware prototype of the proposed receiver, is constructed for complementary high-low-side active switches on the same experimental verification. The experimental results show the bridge leg increases the risk of a direct short between the DC effectiveness of the soft-switching operation in the receiver with a bus voltage and ground, which unavoidably undermines the maximum AC−DC efficiency of 98% while maintaining good reliability and robustness of the receiver. Moreover, the use of regulation of the output voltage, regardless of line and load pulse density modulation produces significant output voltage variations. ripples. A relatively bulky output capacitor is therefore needed Keywords—Semi-active class D rectifier, wireless power transfer to mitigate the output ripples. (WPT), output regulation, resonant rectifier, hybrid modulation. Recently, a number of single-stage single-switch wireless power receivers have emerged [16], [17], [22], which employ I. INTRODUCTION fewer power switches, thus making them more attractive solutions for WPT by offering higher reliability and lower cost. The rapid advancement in wireless power technology has Nonetheless, these prior art have their own limitations. For completely transformed the way we recharge a myriad of example, the duty-cycle-regulated class-E rectifier in [16] battery-powered portable electronic devices because of its suffers from high voltage stresses on the main switch and the added convenience, durability and safety [1][6]. Nowadays, diodes, which inevitably degrades its reliability and potentially an increasing number of consumer electronic devices such as shortens its operating lifetime. In [17], the phase-shift regulated smartphones, smart watches, headsets, and tablets have already class-E rectifier requires complicated design procedures and incorporated the Qi wireless charging function. According to suffers from poor efficiency at light load condition. In [22], the the latest market survey [7], the total shipment of wireless main disadvantages of the multi-cycle-switching-regulated power receivers is more than doubled that of the wireless power active rectifier include the use of a bulky output capacitor, a transmitter by 2022. To cope with the fast-growing market narrow regulation range, and lower efficiency. demand and more stringent requirements in medium-power In this paper, a single-switch-regulated resonant wireless WPT system [8], it is advantageous to develop a wireless power power receiver is proposed. This wireless power receiver receiver that is efficient, compact, low-cost, and reliable for carries the following merits: 1) concurrent high-frequency practical applications. Hence, the motivation of this research AC−DC rectification and DC regulation; 2) a wide range of work is the introduction of a highly-efficient single-switch- output regulation; 3) robust output regulation against line and regulated resonant power receiver with hybrid modulation, load disturbances; 4) soft-switching operation regardless of the which is characterized by a simple circuit structure, low coupling and load conditions; and 5) highly-efficient AC−DC component count, high AC-DC conversion efficiency, and energy conversion. This paper is organized as follows. Section good output voltage regulation. II presents the circuit topology of the proposed single-switch The conventional two-stage topology, which comprises a resonant wireless power receiver and its operating principle. diode bridge rectifier and a regulated DC−DC converter (or Section III discusses the design consideration, particularly, how LDO linear regulator), is more prevalent in commercial to determine the minimum value of the output capacitor in order © 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. to achieve a nearly-constant output voltage. Section IV the equivalent series resistance (ESR) of the passive component compares the proposed wireless power receiver with its can be neglected; 2) The quality factor of the resonant tanks, predecessors. Section V contains the experimental results. namely, (L , C ) and (L , C ), is sufficiently high so that only the p p s s Finally, Section VI concludes this research work. fundamental component of the current is taken into consideration; and 3) The value of the output capacitor C is II. HIGHLY-EFFICIENT SINGLE-SWITCH REGULATED large enough so that a constant DC voltage with reasonably RESONANT WIRELESS POWER RECEIVER SYSTEM small ripple is produced. Due to the inherent property of series-series compensation, A. Circuit Topology the input current of the wireless power receiver i is only a Ls Fig. 1 shows a generic two-coil WPT system, with the function of the fundamental component of the output voltage of emphasis on the proposed wireless power receiver. On the the VSI at the transmitter side and the coupling coefficient M transmitter side, it is assumed that a voltage source inverter [26], [27]. Consequently, i can be treated as a current source Ls (VSI) (e.g. full-bridge/half-bridge inverter), that generates a at the receiver side, which is expressed as square waveform with a fundamental frequency of f and a | | 𝑖 I sin2𝜋 𝑓 𝑡 . (1) series-compensated primary coil (L , C ), where f = 1/T = p p s s 1/2 C L , is used. On the other hand, the wireless power p p By using i as a reference signal, the ideal waveforms of the Ls receiver is made up of the series-compensated secondary coil key signals (i.e., v , v , v , i , and v ) of the single-switch- gs CS1 CD1 CD1 o (L , C , and f = 1/2π C L ) and the single-switch resonant s s s s s regulated resonant wireless power receiver wireless are rectifier circuit, which consists of a power switch (S ), three graphically illustrated in Fig. 2. capacitors (C , C , C ), a diode (D ), a resistive load (R), a S1 D1 o 1 synchronization circuit, and a microcontroller (MCU). State I State II State III State IV State V T =1/f s s Wir eless Power Wir eless Power Rece iver Transmitter r − v + t CS1 i Ls + v − S1 C Cs Cs s S1 Voltage + C R D1 L L CD1 v DT Source p s D o s − o Inverter D1 i i gs Lp Ls t t t v r gs Sync Cs1 + v t Syn . MCU Circu it s1 Fig. 1. Schematic diagram of a generic two-coil WPT system with the proposed wireless power receiver. CD1 Basically, the topology of the proposed receiver is a semi- D1 active class-D rectifier. S and D operate in a complementary 1 1 manner which determines the nominal value of the output voltage. C and C are parasitic capacitors of S and D , S1 D1 1 1 t nT nT +t (n+0.5)T (n+D)T +t (n+D)T +t +t (1+n)T respectively. C is the output capacitor for minimizing the s s f s s f s f r s output voltage ripple. In contrast to conventional class-D synchronous rectifiers for Fig. 2. Key waveforms of the single-switch-regulated resonant DC-DC conversion applications [23]−[25], the proposed wireless power receiver system. wireless power receiver overcomes the following design challenges: 1) Enable synchronization between resonant current In particular, the gate-to-source voltage (v ) of S , i.e., the gate gs 1 and the gate driving waveform of the receiver; 2) Maintain soft drive PWM signal (with a duty ratio of D), has a time delay of switching and good output regulation for a wide range of load t relative to the zero-crossing point of the positive cycle of i . f Ls and coupling coefficient values by employing hybrid By considering the on/off switching state of S and the modulation; 3) Make use of standard steady-state and small- charging/discharging period of C and C , four distinct states S1 D1 signal models to analyze static and dynamic performance of the can be defined for the receiver, namely, State I for nT ≤ t < (nT s s closed-loop system; and 4) Achieve concurrent high-frequency + t ), State II for (nT + t ) ≤ t < (n + D)T + t , State III for (n + f s f s f AC−DC rectification and DC output regulation by employing D)T + t ≤ t < (n + D)T + t + t , and State IV for (n + D)T + t s f s f r s f the model-based controller design method. + t ≤ t < (1 + n)T , where t and t are the fall and rise time of r s f r the voltage across C (i.e., V ), respectively, n is an arbitrary S1 Cs1 B. Operating Principles nonnegative integer number, and D is the on-time duty ratio of In the ensuing discussion, the following assumptions are the switch. Fig. 3(a)(d) shows the equivalent circuit model in made: 1) For simplicity, ideal circuits are assumed and hence, each of the four states. i charges C and also discharges C at the same time. By Ls D1 S1 − v + CS1 using KCL and KVL, we have CD2 C S1 L C s s 𝑑𝑣 𝑑𝑣 (2) 𝐶 𝐶 𝑖 i + 𝑑𝑡 Ls C R D1 v 𝑣 𝑣 𝑣 . v D o CD1 1 CD1 Meanwhile, the output capacitor C supplies current to the load (a) State I R. Mathematically, we can write − v + CS1 𝑑𝑣 𝑣 C (3) 𝐶 . L C S1 s s S 𝑑𝑡 𝑅 S1 i + Ls C R D1 State II [(nT + t ) ≤ t < (n + 0.5)T ]: s f s D o CD1 1 C Fig. 3(b) contains the equivalent circuit model of the receiver circuit in State II. At t = (nT + t ), v returns to zero volts and s f Cs1 afterwards, S is turned on at the rising edge of v . This enables 1 gs (b) State II zero-voltage-switching (ZVS) turn-on operation. After S is − v + switched on, v and v are clamped at zero volts and v , CS1 CD1 o CS1 respectively, i.e., v (nT + t ) = 0 and v (nT + t ) = v . By Cs1 s f CD1 s f o L C S1 s s substituting these two conditions into (2) and solving for t , where tf is the time interval of State I (tf), we have S1 Ls C R D1 v D o CD1 1 𝑡 . (4) 𝜋 𝑓 |I | (c) State III − v + State III [(n + 0.5)T ≤ t < (n + D)T + t ]: s s f CS1 Fig. 3(c) contains the equivalent circuit model of the receiver CD2 C S1 L C s s circuit in State II. At t = (n+0.5)T , i returns to zero and s Ls becomes negative, while S remains on. During State II and III, + i transfers the input energy to both the output capacitor C and i + Ls o Ls C C R D1 o v the load R via S , which can be written as D o 1 CD1 1 CD1 𝑑𝑣 𝑣 𝐶 𝑖 . (5) 𝑑𝑡 𝑅 (d) State IV To allow proper regulation of the output voltage (v ), the range − v + CS1 of the duty ratio (D) is defined as L C S1 s s (6) 𝑓 𝑡 𝐷 1 2 𝑓 𝑡 . i + Ls C C R D1 o v D CD1 1 State IV [(n + D)T + t ≤ t < (n + D)T + t + t ]: s f s f r D1 Fig. 3(d) shows the equivalent circuit model of the receiver circuit in State III. At t = (n + D)T + t , S is turned off with s f 1 (e) State V zero voltage. Afterwards, i , which is in the negative half cycle, Ls starts discharging C and charging C simultaneously. The D1 S1 Fig. 3 Equivalent circuit model in (a) State I; (b) State II; (c) State charging procedure can thus be mathematically represented as III; (d) State IV, (e) State V. 𝑑𝑣 𝑑𝑣 𝐶 𝐶 𝑖 State I [nT ≤ t < (nT + t )]: . s s f 𝑑𝑡 (7) 𝑣 𝑣 𝑣 Fig. 3(a) shows the equivalent circuit model of the receiver circuit in State I. At t = nT , the diode current i drops to zero s CD1 and stops conducting naturally, resulting in a zero-current- Meanwhile, C continues to supply current to the load R. Hence, switching (ZCS) turn-off operation of the diode. When t > nT , we have 𝑑𝑡 𝑣 𝐶 𝑑𝑡 -7 𝑑𝑣 𝑣 10 5.5 (8) 𝐶 . State V [(n + D)T + t + t ≤ t < (1 + n)T ]: s f r s Fig. 3(e) contains the equivalent circuit model of the receiver 4.5 circuit in State IV. At t = (n + D)T + t + t , v returns to zero s f r CD1 volts and D starts conducting. After D is forward biased, v 1 1 CD1 and v are clamped at v and zero volts, respectively, i.e., CD1 o 3.5 v [(n + D)T + t +t ] = v and v [(n + D)T + t +t ] = 0. By Cs1 s f r o CD1 s f r substituting these two conditions into (6), t can be obtained as 1 𝐷 1 2.5 𝑡 𝑡 arccos cos 2𝜋 𝑓 𝑡 00.2 0.4 0.6 0.8 1 𝑓 2𝜋 𝑓 C (F) -8 D1 10 2𝜋 𝑓 𝐶 𝐶 𝑣 (9) (b) t versus C f D1 |I | -7 5.5 From Fig. 2, Q and Q represent the total charge transferred f r during the switching intervals t and t , respectively. Since Q = f r f 5 Q , the average current of i in State III is higher than that in r LS 4.5 State I. Hence, t becomes larger than t . Note that in this f r particular state, when D conducts, i freewheels through D 1 Ls 1 and no energy is transferred to the output. C continues to 3.5 discharge its current to the load R. By KCL, we can write 𝑑𝑣 𝑣 3 (10) 𝐶 . 2.5 0 0.2 0.4 0.6 0.8 1 C (F) -8 At t = (1 + n)T , the circuit enters State I of the next switching s S1 10 period and the aforementioned state transitions will repeat. (c) t versus C f S1 -7 4.3 C. Steady-State Model 4.25 By invoking state-space averaging on the output capacitor C , 4.2 the relationship between the duty ratio (D) and the output 4.15 voltage (v ) based on the steady-state model can be analytically derived as 4.1 𝑑𝑣 1 𝑣 𝐶 𝑖 4.05 𝑇 𝑅 (11) |I | 𝑣 1.9 1.95 2 2.05 2.1 cos 2𝜋 𝑓 𝑡 cos 2𝜋𝐷 2𝜋 𝑓 𝑡 . f (Hz) 5 2𝜋 𝑅 (d) t versus f f s Since 0 in steady-state condition, the L.H.S. of (11) Fig. 4. Plots of output voltage v versus duty cycle D, t versus o f becomes zero and hence, by re-arranging, v can be obtained as o C , t versus C and t versus f . D1 f S1 f s |I |𝑅 𝑣 cos 2𝜋 𝑓 𝑡 cos 2𝜋𝐷 2𝜋 𝑓 𝑡 . (12) Fig. 4 shows plot of v against D based on (12) and plots of t o f 2𝜋 versus C , C , and t . Clearly, v can be regulated by adjusting D1 S1 f o the variable D. It is observed that the maximum voltage is achieved if D = 0.5 − f t for a given value of R. The fact that s f 1/π│I │R Ls max ×cos(2πf t ) the output voltage is highly dependent on R implies that a s f closed-loop system is needed in order to achieve tight regulation of the output voltage. The value of t increases as C f D1 1/π│I │R and C increased, while t drops as f increased. Ls min S1 f s ×cos(2πf t ) s f R↑ Fig. 5 shows the relationship between the output voltage (v ) and the duty ratio (D) with and without resonant capacitors (C , S1 C ). The latter is referred to as the ideal case since an ideal D1 switch (or diode) contains no parasitic capacitances, i.e., C = S1 0.5−f t 1−2f t s f s f 0 (or C = 0). By comparing the two curves in Fig. 5, the output D1 (a) Plots of output voltage v versus duty cycle D at voltage (v ) with resonant capacitors (i.e., the bottom curve) is different values of R. reduced by 1/π│I │R ×(1 − cos(2πf t )), compared with that Ls max s f t (s) t (s) t (s) 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 without resonant capacitors (i.e., the top curve). In addition, the Hence, the minimum value of the output capacitor is given by minimum and maximum value of D is shifted by f t and 2f t , s f s f |I | respectively, as illustrated in Fig. 5. 𝐶 . (16) 𝑥%𝑣 𝜋 𝑓 C =C =0 S1 D1 1/π│ILs│Rmax ×(1−cos(2πf t )) 1/π│I │R s f Ls max C ≠0, C ≠0 Design of C and C . Since C and C are the parasitic S1 D1 D1 S1 D1 S1 1/π│I │R Ls max capacitance of the diode D and switch S , the values can be 1 1 ×cos(2πf t ) s f obtained from their datasheets. B. Derivation of the Small-Signal Model and Feedback Control for Output Voltage Regulation f t s f 2f t s f Theoretical result Sim ulation result D 10 0.5−f t 0.5 s f 1−2f t 1 s f ‐10 Fig. 5. Plots of output voltage (v ) versus duty ratio (D) with and ‐20 without resonant capacitors (CS1, CD1) at the maximum load Rmax. ‐30 10 100 1000 10000 Frequency (Hz) III. DESIGN CONSIDERATIONS AND IMPLEMENTATION (a) Bode magnitude plot ‐180 A. Design of Reactive Components Theoretical result ‐200 Simulation result Design of L and C . The design objective of L is that the quality s s s ‐220 factor Q of the coil is higher than the desired quality factor Q, Ls ‐240 i.e. Q ≥ Q. Given that the equivalent series resistance (ESR) Ls ‐260 of L is R , Ls is designed as s Ls-ESR ‐280 10 100 Frequency (Hz) 1000 10000 2𝜋 𝑓 𝐿 𝑄𝑅 𝑄 𝑄 ⇒ 𝐿 . (13) (b) Bode phase plot 𝑅 2𝜋 𝑓 Correspondingly, due to the use of series-series compensation, Fig. 7. Theoretical and simulated Bode plots of the open-loop the capacitor C is sized as control-to-output voltage using the small-signal model By linearizing (11) and considering the AC perturbation of duty 𝐶 . (14) 2𝜋 𝑓 𝐿 cycle D, the resulting small-signal control-to-output linearized equation can be written as Deign of the Output Capacitor (C ). The design objective of the 𝑑𝑣 𝑣 output capacitor is to maintain a constant output voltage v with 𝐶 |I |𝐷 sin 2𝜋𝐷 2𝜋 𝑓 𝑡 . (17) sufficiently small voltage ripple ∆v , i.e., ∆v ≤ x%  v . Fig. 6 𝑑𝑡 𝑅 o o o provides a graphical illustration of the root cause of the output Fig. 7 shows the Bode plots of the theoretical and simulated voltage ripple. small-signal models at D = 0.5,│I │= 1 A, R = 30 Ω, f t = 0.1, Ls s f and C = 100 µF. As is evident in Fig. 7(a) and (b), the theoretical and simulation results are in close agreement, ∆v v o thereby validating the accuracy of the derived small-signal equation from (15). Also, the phase plot in Fig. 7(b) shows that the phase of the open-loop system decreases from −180 to −270. It is important to note that the use of conventional Q proportional-integral (PI) controller with positive proportional ripple Ls and integral coefficients (k and k ) is unable to provide p i v / R adequate phase boosting for the uncompensated system, which t t 1 2 leads to unstable transient response because of insufficient phase margin and unacceptably large steady-state error due to relatively low DC gain [28]. To address this issue, a modified Fig. 6. Illustration of the root cause of the output voltage ripple. proportional-integral (PI) controller, as shown in Fig. 8, with Basically, the rise of the output voltage is attributed to the negative values of k and k is employed to attain accurate and p i stable regulation of the output voltage. In addition, an anti- accumulation of charge Q , which is given by ripple windup scheme is used to prevent overflow of the integrator and ∆𝑣 𝑥%𝑣 to ensure linear operation of the PI controller. Specifically, the 𝑖 𝑑𝑡 anti-windup loop is added to avoid integrator wind-up, which 𝑄 |I | (15) can occur when the duty ratio (D) is saturated. In other words, 𝐶 𝐶 𝜋 𝑓 𝐶 Phase (degree) Magnitude (dB) it prevents the actual value of D from exceeding beyond its Ls + GPIO32 Sy nc upper and lower limits, as defined in equation (6). Sy nc Duty cycle (50%) PWM1 Controller Anti-windup InSYNC │k│ i Frequency Synchroni zation Duty cycle D Con troller V v ref gs z Gain −│k│ V i ∫ ref t 2πf PWM2 f s Eq. (4) o + Phase Eq. (6) Gain Hyb rid Mod ul ation −│k │ TMS320F28335 PI controller Fig. 10. Block diagram of the frequency synchronization and Fig. 8. Block diagram of the modified PI controller with anti- hybrid modulation. windup. As can be seen in Fig. 10, an external comparator, which is part The proper values of the proportional gain k and integral gain of the synchronization circuit, is employed to detect the zero- k of the PI compensator can be determined as follows. crossing points of i and generate the corresponding square Ls 2𝜋 𝑓 𝐶 𝑘 waveform labelled Sync [see Fig. 11]. 𝑘 ,𝑘 . (18) Frequency Synchronization | | 𝑅𝐶 I sin 2𝜋𝐷 2𝜋 𝑓 𝑡 𝑠 T Ls t where f is the desired crossover frequency. Fig. 9 shows the resulting Bode plots of the open-loop transfer function and the Sync closed-loop transfer function of the system with k = −1.07 and TBCTR1 k = −356. The numerical results of the closed-loop transfer CMP1(50%) 0.5T function show that the crossover frequency is at 1 kHz with a s PWM 1 phase margin of 90 and a low-frequency gain (at 10 Hz) is 40 0.5DT +t Hybrid Modulation dB. The phase margin of 90 ensures a stable closed-loop s f CMP2(D) response and the static gain at DC is high enough (> 40 dB) to TBCTR2 eliminate the steady-state error. PWM2 50 DT (v ) gs Open‐loop results t Closed‐loop results Fig. 11. Ideal waveforms of the key signals in frequency synchronization and hybrid modulation. ‐10 ‐20 The MCU receives this external synchronization signal Sync via ‐30 10 100 Frequency (Hz) 1000 10000 the GPIO32 pin, which is then used for EPWM synchronization. The rising edge of Sync triggers the counter of (a) Bode magnitude plot the PWM1 (TBCTR1) by counting from zero on a cycle-by- Open‐loop results cycle basis, which leads to a triangular carrier waveform for ‐50 Closed‐loop results PWM1. The duty cycle of the PWM1 is set to be at 50%. ‐100 Consequently, PWM1 is synchronized with i with a constant ‐150 Ls ‐200 phase difference of 0.25T between them. In this way, frequency ‐250 synchronization between i and D is realized. LS ‐300 Subsequently, hybrid modulation is implemented on the 10 100 Frequency (Hz) 1000 10000 PWM2 module using the PWM1 as a reference. The duty cycle (b) Bode phase plot D, which is obtained from the controller output [see Fig. 10], is used to compare with the counter of the PWM2 module Fig. 9. (a) Bode magnitude plot and (b) bode phase plot of the open- TBCTR2 in order to adjust the pulse width of the PWM2. On loop and closed-loop transfer function. top of them, the phase-shift modulation is realized by adjusting the phase difference between PWM1 and PWM2 (based on the C. Synchronization and Hybrid modulation internal synchronization signal InSYNC). The desired time Fig. 10 shows the functional block diagram of the frequency delay t is calculated from equation (4). To reduce the synchronization and hybrid modulation scheme. The frequency computation complexity, the real-time output voltage v and synchronization is implemented by using the PWM1 module │I │ in equation (4) are replaced by a fixed output voltage Ls whereas the hybrid modulation is realized by using the PWM2 reference V and a constant value, respectively. Since the ref module of the microcontroller (part number: TMS320F28335). carrier waveform TBCTR2 is triangular in shape, the center of Fig. 11 shows the ideal waveforms of the key signals in the resulting PWM2 is thus aligned with the valley of TBCTR2. frequency synchronization and hybrid modulation. Hence, to produce a phase shift of t , TBCTR2 is simply shifted Magnitude (dB) Phase (degree) by 0.5DT +t . After performing normalization of the phase s f TABLE II PARAMETERS OF COMPONENTS angle φ of PWM2 by 2π, the effective value of φ is given by Part Value 𝜑 2𝜋 𝑓 𝑡 𝐷𝜋 . C , C 4.5 nF D1 S1 (19) Ls 172 µH Fig. 10 shows the implementation of (17) in the MCU. Cs 3300 pF + 330 pF Co 1000 µF IV. COMPARATIVE STUDY TABLE III. PART NUMBER OF COMPONENTS. Table I compares the proposed receiver with the prior art of Part Part Number single-power-switch solutions [16], [17], [22] in terms of CD1, CS1 Parasitics of TK56A12N1 topologies, compensation of receiver coil, resonant frequency, Custom-made L circular copper air coil etc. As clearly shown in Table I, the proposed receiver uses only s diameter = 29 cm one power switch and achieves fully soft-switching operation B32682A7332K000 (3300 pF) Cs on the switch and diode. The maximum AC-DC conversion PHE448SB3330JR06 (330 pF) efficiency reaches 98%, which is higher than its predecessors. C UVZ1H102MHD The regulation frequency is identical to the resonant frequency, Gate Driver ADuM3223 Comparator LM 393P which eliminates the use of bulky reactive components. The Current Transformer AS-100 switch-diode bridge structure can prevent the direct short circuit MOSFET S TK56A12N1 of the output terminals, thus enhancing the system reliability. Diode D Body diode of TK56A12N1 Indeed, these features make it a very competitive solution for Microcontroller TMS320F28335 future WPT applications [3]. Compared with the existing class- E-based solutions [16], [17], a significant advantage of the Oscilloscope Primary proposed receiver is that it has relatively low voltage stresses Secondary coil L p coil L on the power switch. Even though the proposed solution s requires an additional diode when compared with [17], the relatively low voltage stresses on both the switch and diode can Differential voltage probe justify the slight increase in the cost. Compared with the semi- active class-D with pulse density modulation [22], the benefit of the proposed receiver is that it can achieve ZVS turn-on for MCU Current the power switch, which results in higher efficiency and much probe higher output power. Additionally, the proposed solution Full bridge allows the use of a much smaller output capacitance due to the Current Wireless power MCU Inverter transformer relatively high regulation frequency. Receiver Fig. 12. Experimental setup (including the prototype of the proposed receiver). TABLE I. COMPARISON WITH EXISTING WPT RECEIVERS. Proposed [16] [17] [22] V. EXPERIMENTAL RESULTS Semi-active Modified Active Semi-active Topology A hardware prototype of the proposed WPT receiver with a Class D Class E Class E Class D Compensation Series Series Series Series switching frequency of 200 kHz switching frequency is of receiver coil compensation compensation compensation compensation constructed for experimental verification. The nominal output Resonant 200 kHz 6.78 MHz 200 kHz 1 MHz voltage is 24 V and the maximum power is 16 W. Fig. 12 shows frequency a photo of the whole experimental setup including the prototype, Regulation 200 kHz 6.78 MHz 200 kHz 50 kHz frequency in which the DP832 power suppliers, DSOX3204T oscilloscope, Number of N2790A differential voltage probe, and 1147B current probe 1 1 1 1 power switches are used. Table II lists the design parameters and part numbers Number of 1 1 0 1 of the key components used in the prototype. power diode Maximum V ≈3V ≈3V V o o o o A. Steady-State Performance Voltage stress Hybrid Phase-shift Pulse Density Modulation PWM Fig. 13 (a) and (b) shows the key waveforms of the proposed modulation Modulation Modulation receiver in steady-state condition with v = 24 V and i = 0.63 Implementation o o Yes No No Yes ease A. The equivalent load resistance R is 38.09 Ω. The off-state ZVS for ZVS for ZCS for time interval of v and v are measured to be around 2.66 µs CD1 CS1 Soft switching ZVS for switch switch switch and operation switch and 2 µs, respectively. Hence the duty cycle D is around 0.532. ZCS for diode ZCS for diode diode The current of the coil i lead 0.5π against compensated Output voltage, Ls 24 V, 16 W 10 V, 17 W 24 V, 16 W 3.1 V, 96 µW power capacitor voltage v . The voltage stress of the diode is Cs Maximum 98 % (Power 92 % (Power 93 % (Power measured to be 24 V while the measured peak-to-peak input N/A Efficiency stage) stage) stage) current is 4.7 A. As a sanity check, by substituting the above values of D, t , i , and R into (13), the resulting theoretical f Ls output voltage is 24.56 V, which agrees very closely with the Fig. 15 shows the ZVS operation of the switch and ZCS measured value of 24 V. operation of the diode. Fig. 15(a) shows the turn-on transition of the switch. As soon as i enters in its positive cycle, the Ls 2.64 µs 24 V v voltage of the switch v starts dropping to zero. After v CD1 CS1 CS1 reaches zero, the rising edge of v is applied to the switch which gs Ls turns it on completely. Hence, the switch is turned on with ZVS. 4.7 A The measured falling time t is 336 ns. As a sanity check, by substituting the corresponding values of v , i , C and C into o Ls D1, S1 vo 24 V Eq. (4), the theoretical value is 382 ns, which is in good io agreement with the measured value. Fig. 15(b) shows the ZCS 0.63 A turn-off transition of the diode. At the end of the conduction state of the diode, i , which flows through the diode, reaches Ls (a) zero. Hence, the diode is turned off automatically with ZCS. After a lapse of t = 336 ns, v finally reaches 24 V. f CD1 2.00 µs 24 V CD1 CS1 CS1 ZVS 0.5π vgs Ls t = 336 ns Cs Ls 4.7 A (b) i S1 Fig. 13. Steady-state waveforms of v , i , v , v , v and i at a CD1 Ls o Cs CS1 o rated output power of 15 W. (a) ZVS turn-on of the switch. Fig. 14 shows the Fast Fourier Transformation (FFT) and t = 336 ns THD analysis of the measured waveforms of v and i . The v CD1 Ls CD1 fundamental current and THD of the resonant current i are Ls rd th th 2.27 A and 2.33%. The high-order (3 , 5 , 7 , …) harmonic ZCS Ls components are suppressed by the wireless power coils. The fundamental voltage and THD of v are 15.02 V and 46.30%, CD1 D1 respectively. Since the duty cycle is greater than 50%, the even- order harmonic components are produced, thereby degrading the THD performance. Since i contains very limited harmonic Ls (b) ZCS turn-off of the diode. components, the assumption that i is purely sinusoidal in the Ls Fig. 15. Soft switching operations of the switch and diode. theoretical analysis remains valid. 0 Fig. 16 shows the measured output voltage and efficiency -1 values of the rectifier power circuit across different output -2 -5 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 ×10 Time (s) power under output voltage regulation. The nominal output 2.5 Fundamental (200000Hz) = 2.27 voltage is 24 V. The maximum steady-state error of the THD= 2.33% 1.5 regulated output voltage is only 0.1 V, which is around 0.43% of the reference voltage as the power varies from 20% load 0.5 power to full power. The maximum efficiency of the converter 01 234 5678 9 10 Harmonic order is around 98%, which is achieved at full load condition. At light (a) FFT analysis of iLs. load condition, the efficiency only reduces slightly to 94%. 15 100.00% 24.15 97.50% 24.1 -5 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 ×10 Time (s) 95.00% 24.05 Fundamental (200000Hz) = 15.02 92.50% 24 THD= 46.30% 90.00% 23.95 20.00% 40.00% 60.00% 80.00% 100.00% 0 Pe rce ntage of the nominal powe r 0 1 2 3 4 5678 9 10 Harmonic order (b) FFT analysis of vCD1. Fig. 16. Measured efficiency and output voltage of the proposed Fig. 14. FFT analysis of the key AC waveforms. receiver system. Current (A) Current (A) Voltage (V) Voltage (V) Efficiency Votlage (V) Fig. 17 shows the measured output voltage v and amplitude of i of power stage versus coils’ distance with output voltage Ls regulation and an output power of 10 W. As the separation Ls distance between the primary and secondary coils increases 8 ms from 10.5 cm to 21.5 cm, the amplitude of i increases from Ls 1.45 A (at 10.5 cm) to 2.6 A (at 21.5 cm). The maximum voltage vo 0.6 V regulation error is 0.1 V (which is 0.43% of the reference 0.688 A voltage). Hence, the experimental results in either Fig. 16 or Fig. io 17 indicate that the steady-state output voltage regulation error (a) is relatively small over a wide line or load range. 24.15 3 Ls 24.1 2.25 8 ms 24.05 1.5 0.6 V vo 24 0.75 23.95 0 10 11 12 13 14 15 16 17 18 19 20 21 io Distance (cm) (b) Fig. 17. Output voltage vo and amplitude of iLs of power stage versus the coils’ separation distance with output voltage regulation. Fig. 19. Measured waveforms of the rectifier in response to step changes in the output power with output voltage regulation. B. Dynamic Performance Fig. 20 (a) and (b) show the measured waveforms of the Fig. 18 shows the measured waveforms of the rectifier rectifier in response to the input current i changes in light-load Ls operating with synchronization on/off transition. Prior to the condition. When peak-to-peak value of i increases from 2 A Ls enabling of synchronization, the waveforms of the rectifier have to 3.7 A in 10 ms, the output voltage experiences an overshoot significant fluctuations. As soon as synchronization is enabled, of 0.325 V (i.e., 1.35% of the nominal voltage). Likewise, when the output voltage ramps up to the nominal value of 24 V within the peak-to-peak value of i drops from 3.7 A to 2 A in 10 ms, Ls 69 ms. Note that there is no overshoot during the start-up the undershoot is also 0.3 V. The measured dynamic responses process. Besides, any undesirable low-frequency fluctuations in corroborate the robustness and stability of the output voltage the waveforms of the rectifier due to frequency asynchrony are regulation against variations in load and input current. eliminated. Therefore, the experimental results demonstrate the effectiveness of the synchronization scheme. 2 A 3.7 A Ls CD1 i 325 mV Ls vo io 10 ms 24 V io vo 69 ms 159 mA Syn. OFF Syn. ON (a) Fig. 18. Measured waveforms of the rectifier operating with 3.7 A 2 A synchronization on/off transition. Ls Fig. 19 (a) and (b) depict the measured waveforms of the 325 mV rectifier operating with respect to the step-current changes under output voltage regulation. When the output power is vo increased from 0 W to 16 W, the output voltage experiences a 10 ms 0.6 V dip (which is 2.5% of the reference output voltage). The io 159 mA settling time for the step-up load change (i.e., from no load to full load) is 8 ms. Conversely, when the output power is reduced (b) from 16 W to 0 W, the resulting overshoot is 0.6 V. The Fig. 20. Measured waveforms of the rectifier in response to step corresponding settling time for the step-down load change is 8 changes in the input current with output voltage regulation. ms. Voltage (V) Curreent (A) [13] Fact sheet: WPR1516, “Medium power wireless charger receiver VI. CONCLUSIONS 15 W WPC Compliant Solution,” NXP Semiconductors, 2014, [Online], Available: In this paper, a highly-efficient single-switch-regulated https://www.nxp.com/docs/en/fact-sheet/WPR1516FS.pdf resonant wireless power receiver system with hybrid [14] Y. Fang, B. M. H. Pong, and R. Hui, “An enhanced multiple modulation is presented. The hybrid modulation scheme with harmonics analysis method for wireless power transfer systems,” phase-shift and pulse width modulations is employed to IEEE Trans. Power Electron., pp. 1205–1216, 2019. simultaneously achieve very high efficiency and good output [15] The Wireless Power Consortium, “wireless power the qi wireless regulation. A comparative study with the existing single-switch power transfer system power class 0 specification part 4: wireless power rectifiers highlights the conspicuous advantages reference designs,” pp. 1–160, 2016. [16] M. Liu, J. Song, and C. Ma, “Active class E rectifier for dc output of the proposed rectifier, which include low component count, voltage regulation in megahertz wireless power transfer systems,” simple design procedure, high efficiency, and cycle-by-cycle IEEE Trans. Ind. Electron., vol. 67, no. 5, pp. 3618–3628, May output regulation. A hardware prototype of the entire WPT system has been constructed and tested for experimental [17] K. Li, S. C. Tan, and R. Hui, “Single-switch-regulated resonant verification. The experimental results clearly demonstrate that wpt receiver,” IEEE Trans. Power Electron., vol. 34, no. 11, pp. high efficiency and accurate output voltage regulation of the 10386–10391, Nov. 2019. proposed resonant receiver is achievable. The measurement [18] L. Cheng, X. Ge, W. C. Ng, W.-H. Ki, J. Zheng, T. F. Kwok, C.- results also confirm good transient responses of the output Y- Tsui, and M. Liu, “A 6.78 MHz 92.3%-peak efficiency single- voltage under output voltage regulation. stage wireless charger with cc-cv charging and on-chip bootstrapping techniques,” 2019 Symposium on VLSI Circuits, pp. 12, Jun. 2019. REFERENCES [19] H. Li, J. Fang, S. Chen, K. Wang, and Y. Tang, “Pulse density [1] Z. Zhang, A. Georgiadis, and C. Cecati, “Wireless power modulation for maximum efficiency point tracking of wireless transferan overview,” IEEE Trans. on Ind. Electron., vol. 66, no. power transfer systems,” IEEE Trans. Power Electron., vol. 33, 2, pp. 10441058, Feb. 2019. no. 6, pp. 5492–5501, 2018. [2] C. T. Rim and C. Mi, Wireless Power Transfer for Electric [20] L. Cheng, W.-H. Ki, C.-Y. Tsui, “A 6.78-MHz single-stage Vehicles and Mobile Devices. Hoboken, NJ, USA: Wiley-IEEE wireless power receiver using a 3-mode reconfigurable resonant Press, 2017. regulating rectifier,” IEEE Journal of Solid-State Circuits, vol. [3] S. Y. R. Hui, W. Zhong, and C. K. Lee, “A critical review of 52, no. 5, pp. 14121423, Feb. 2017. recent progress in mid-range wireless power transfer,” IEEE [21] T. Diekhans and R. W. De Doncker, “A dual-side controlled Trans. on Power Electron., vol. 29, no. 9, pp. 45004511, Sep. inductive power transfer system optimized for large coupling 2014. factor variations and partial load,” IEEE Trans. Power Electron., [4] G. A. Covic and J. T. Boys, “Inductive power transfer,” Proc. vol. 30, no. 11, pp. 6320–6328, 2015. IEEE, vol. 101, no. 6, pp. 12761289, Jun. 2013. [22] H. S. Gougheri and M. Kiani, “Current-based resonant power delivery with multi-cycle switching for extended-range inductive [5] I. Mayordomo, T. Drager, P. Spies, J. Bernhard, and A. Pflaum, power transmission,” IEEE Trans. Circuits Syst. I Regul. Pap., “An overview of technical challenges and advances of inductive vol. 63, no. 9, pp. 1543–1552, 2016. wireless power transmission,” Proc. IEEE, vol. 101, no. 6, pp. 13021311, Jun. 2013. [23] M. K. Kazimierczuk, M. J. Mescher, and R. M. Prenger, “Class D [6] A. P. Sample, D. A. Meyer, and J. R. Smith, “Analysis, current-driven transformer center-tapped controllable synchronous rectifier,” IEEE Trans. Circuits Syst.−I: experimental results, and range adaptation of magnetically Fundamental Theory and Appl., vol. 43, no. 8, pp. 670−680, Aug. coupled resonators for wireless power transfer,” IEEE Trans. Ind. Electron., vol. 58, no. 2, pp. 544–554, 2011. [24] M. K. Kazimierczuk and M. J. Mescher, “Class D converter with [7] Dinesh Kithany, “Wireless Power Market Tracker,” IHS Markit half-wave regulated synchronous rectifier,” IEEE Appl. Power Surv., 2019. Electron. Conf. Expo. (APEC), vol. 2, pp. 1005–1011, 1994. [8] The Wireless Power Consortium, “Medium power wireless [25] M. Mikotajewski, “Class D synchronous rectifiers,” IEEE Trans. charging: emerging applications, considerations and Circuits Syst., vol. 38, no. 7, pp. 694–697, Jul. 1991. opportunities.” [26] Y. Zhang, T. Kan, Z. Yan, and C. C. Mi, “Frequency and voltage [9] Datasheet, STWLC68, “STWLC68 Qi-compliant inductive tuning of series-series compensated wireless power transfer wireless power receiver for 5 W applications,” system to sustain rated power under various conditions,”, IEEE STMicroelectronics, 2020, [Online] Available: Journal of Emerging and Selected Topics in Power Electron., vol. https://www.st.com/resource/en/datasheet/stwlc68.pdf 7, no. 2, pp. 1311−1317, Jun. 2019. [10] Datasheet, P9222-R, “Wireless power receiver for low power applications”, Renesas Electronics Corp., 2019, [Online] [27] W. Zhang and C. C. Mi, “Compensation topologies of high-power Available: https://www.idt.com/us/en/document/dst/p9222-r- wireless power transfer systems,” IEEE Trans. Veh. Technol., vol. 65, no. 6, pp. 4768–4778, 2016. datasheet [28] R. W. Erickson and D. Maksimovic, “Fundamentals of power [11] Datasheet, BQ51003, “BQ51003 highly integrated wireless electronics,” Springer, 2006. receiver Qi (WPC v1.2) compliant power supply,” Texas Instrument Inc., 2018, [Online] Available: http://www.ti.com/lit/ds/symlink/bq51003.pdf [12] User’s Guide, WPR1500-BUCK, “WPR1500-buck MP receiver v2.1 reference design user’s guide,” NXP Semiconductors, 2016, [Online], Available: https://www.nxp.com/docs/en/user- guide/WPR1500BUCKMPUG.pdf

Journal

Computing Research RepositoryarXiv (Cornell University)

Published: Apr 9, 2020

There are no references for this article.