Dependence of Efficiency on Wire Type and Number of Strands of Litz Wire for Wireless Power Transfer of Magnetic Resonant Coupling

Transcription

1 IEEJ Journal of Industry Applications Vol.3 No.1 pp DOI: /ieejjia.3.35 Paper Dependence of Efficiency on Wire Type and Number of Strands of Litz Wire for Wireless Power Transfer of Magnetic Resonant Coupling Tsutomu Mizuno a) Senior Member, Takuto Ueda Student Member Shintaro Yachi Student Member, Ryuhei Ohtomo Student Member Yoshihito Goto Student Member (Manuscript received Jan. 15, 2013, revised Sep. 26, 2013) Wireless power transfer is expected to be applied to portable devices and electric automobiles in the future. To achieve this, it is necessary to improve the transmission efficiency of the coils used in wireless power transfer, that is, to improve the quality factor and coupling coefficient of the coils. To improve the quality factor of the coils, the authors propose the use of a litz wire with a magnetoplated wire (MPW), which is a copper wire plated with a thin iron film. The MPW increases the quality factor of the coils by reducing the AC resistance owing to the proximity effect. In this study, the effect of the number of strands of a litz wire on the quality factor of the coils and efficiency characteristics is considered. Moreover, the quality factor of the coils and efficiency characteristics using three types of coil a solid copper wire (COW), a litz wire with a copper wire (LCW), and a litz wire with an MPW (LMW) are considered. At the transmission frequency f = MHz, it is experimentally demonstrated that many strands, whose number becomes the highest in terms of the quality factor of the coils, exist. The transmission efficiencies of the COW, LCW and LMW coils at an output power of 5 W and a transmission distance of 9 mm are 89%, 84%, and 91%, respectively, and the efficiency of the LMW coil is the highest. Also, in this case, the temperature increases of the COW, LCW, and LMW coils are 12,16, and 10, respectively, and the LMW coil reduces the temperature increase. Keywords: wireless power transfer of magnetic resonant coupling, magnetoplated wire, number of strands, quality factor, transmission efficiency, temperature increase 1. Introduction The technology of wireless power transfer has already been utilized, for example, in IC cards. The international standard output power of 5 W at a short distance (Qi standard) has been considered and its practical application to portable devices and electric automobiles is expected (1) (4). An improvement in transmission efficiency is required in wireless power transfer. To achieve this, it is necessary to improve the quality factor of the coils and coupling coefficient (5) (6). The quality factor of the coils is proportional to the angular frequency and inductance of the coils and inversely proportional to the resistance of the coils. Therefore, it is necessary to reduce the resistance in order to improve the quality factor. The resistance is represented by the sum of DC resistance R dc, AC resistance due to the skin effect, R s, (7) and that due to the proximity effect, R p. A litz wire, which is a twisted plural copper wire with small diameter (LCW), (8) is generally used to reduce R s. The authors proposed the use of a magnetoplated wire (MPW), which is a copper wire plated with a magnetic thin (7) film, to reduce R p. This is because an alternating magnetic flux flows in a magnetic thin film, which has a higher a) Correspondence to: Tsutomu Mizuno. mizunot@ shinshu-u.ac.jp Shinshu Uiversity , Wakasato, Nagano, Nagano , Japan permeability and a higher resistivity than a copper wire. As a result, the R p of the MPW is smaller than that of the copper wire because the eddy current in the MPW also is smaller than that in the copper wire (7). It is known that the MPW increases the transmission efficiency (9). R p depends on the number of strands of the litz wire, and to further improve the transmission efficiency, it is necessary to consider the effect of the number of strands of the litz wire on the quality factor of the coils and efficiency characteristics (10). In this study, the effect of the number of strands of each of the three types of wire, which are a solid copper wire (COW), an LCW and a litz wire with an MPW (LMW), on the quality factor of the coils and efficiency characteristics is considered. In addition, an output power of 5 W for portable devices and a transmission frequency of MHz in the industryscience-medical band (ISM band) are used. Moreover, the following are discussed. 1) Equivalent circuit of wireless power transfer and structure of coils 2) Quality factor of coil characteristics dependent on number of strands 3) Transmission characteristics 2. Equivalent Circuit of Wireless Power Transfer and Structure of Coils 2.1 Equivalent Circuit of Wireless Power Transfer Figure 1 shows a model of the wireless power transfer of c 2014 The Institute of Electrical Engineers of Japan. 35

2 magnetic resonant coupling. The structures of the transmitting and receiving coils are the same. C r1 and C r2 are capacitors used to induce resonance with the transmitting and receiving coils, respectively. The output impedance of the power source, Z 0, and the load Z L are always 50 Ω. Z i is the impedance of the circuit from the power source side and Z o is the impedance of the circuit from the load side. The transmission efficiency η 21 is shown by the following equation with the input power P i and output power P o. η 21 = P o 100 (%) (1) P i In a high-frequency circuit, η 21 decreases because of the reflection that occurs if Z 0 and Z i or Z L and Z o are different (reflection efficiencies η 11 and η 22 ) (11). Therefore, a matching circuit is necessary for high transmission efficiency. Figure 2 shows an equivalent circuit of wireless power transfer. Figure 2(a) shows the inserted matching circuit in the cases of Z i > Z 0 and Z o > Z L and Fig. 2(b) shows the inserted matching circuit in the cases of Z i < Z 0 and Z o < Z L. To insert the matching inductors and capacitors, L m1 and L m2 and C m1 and C m2, Z i and Z o should be the same as Z 0 and Z L, respectively; therefore, both η 11 and η 22 are expected to be zero. In addition, the value of each of the matching inductors and capacitors is given by the following equations (11). L mj = 1 ω ZB (Z A Z B ) (H) (Z i > Z 0, Z o > Z L ) (2) C mj = 1 ZA Z B (F) (Z i > Z 0, Z o > Z L ) ωz A Z B (3) L mj = 1 ω ZA (Z B Z A ) (H) (Z i < Z 0, Z o < Z L ) (4) C mj = 1 ZB Z A (F) (Z i < Z 0, Z o < Z L ) ωz B Z A (5) Where ω is the angular frequency (= 2π f rad/s), Z A is Z i (Ω) forj= 1orZ o (Ω) forj= 2andZ B is Z 0 (= 50 Ω) for j = 1orZ L (= 50 Ω) forj= Structures of Coils and Wires Figure 3 shows the structure of the coils. The diameter of the coils, d, is 36 mm and the bobbin is made of expanded polystyrene. The coils are tightly wound and, as described later, their number of turns, n, is 8. As shown in Table 1, the length of the coils in the axial direction, l a, depends on the number of strands, N, of the litz wire. Figure 4 shows the structures of the COW, LCW and LMW. The COW has a diameter of 450 μm and is plated with an insulating film of 16 μm thickness. The resistivity of the copper wire used, ρ 1,is Ωm and its permeability μ r1 is (12). An LCW strand has a diameter of 100 μm and is plated with an insulating film of 12 μm thickness. An LMW strand is a copper wire with a diameter of 100 μm that is plated with magnetic thin films (Fe and Ni) followed by a 13-μm-thick insulating film. The Ni film is prepared in order to soft-solder easily. The thicknesses of the Fe and Ni films are 0.9 μm and 0.05 μm, respectively, where the resistivity of iron, ρ 2,is Ωm and its permeability μ r2 is 100 (12). The conductor cross-sectional area A of the COW is Fig. 1. Model of wireless power transfer (unit: mm) Fig. 3. Structure of coils (unit: mm) (a) Z i > Z 0, Z o > Z L Fig. 2. (b) Z i < Z 0, Z o < Z L Equivalent circuit of wireless power transfer 36 IEEJ Journal IA, Vol.3, No.1, 2014

3 Table 1. Axial direction lengths of coils changed experimentally and the effect of the dependence of R p on N is considered to determine R p. R i depends on the current that is output power P o. Also there is no R i of the COW and LCW because the wires do not have magnetic thin films. For the above reason, the quality factor of the coils depends on N, and the highest N in terms of the quality factor of the coils is expected for f = MHz. Therefore, the quality factor of the coils characteristic dependence on N is considered in the 3 rd chapter and the highest transmission characteristic of the coils in terms of the quality factor of in the 3 rd chapter is considered in the 4 th chapter. 3. Quality Factor of Coil Characteristic Dependence on N Fig. 4. (a) Solid wire (COW) (b) Litz wires Structure of wires (unit: μm) mm 2 and those of the LCW and LMW for N = 20 are mm 2 ; therefore, the A values of such wires are similar (refer to Table 1). The N considered is less than or equal to 20 for the A values of the LCW and LMW not to exceed that of the COW. The quality factor of the coils is shown by the following equation. Q = ωl (6) R Where L is the inductance of the coils (H) and R is the resistance of the coils (Ω). In addition, the R of the coils is given by the following equation (7). R = R dc + R s + R p + R i (Ω) (7) Where R dc is the DC resistance (Ω), R s is the AC resistance due to the skin effect (Ω), R p is the AC resistance due to the proximity effect (Ω) andr i is the hysteresis loss of the magnetic thin films (Ω). R dc is inversely proportional to A; therefore, R dc decreases as N increases. The AC resistance due to the skin effect per strand is constant and does not depend on N. However, R s decreases as N increases, namely, the strands are connected in parallel. On the other hand, since R p depends on N, N is When f = MHz, d = 36 mm and N = 20, the quality factor of the LMW coil was the highest for n = 8 (13).Therefore, the highest N in terms of the quality factor is considered for n = 8. Table 1 shows the relationship between N and l a for n = 8. l a was measured with a micrometer (Mitutoyo, IP65). Figure 5 shows the impedance vs. number of strands characteristics of coils with n = 8 and f = 13.56MHz. The impedance was measured with a network analyzer (Agilent, 5061B) and the power level was P = 1 mw. Figure 5(a) shows the resistance R, (b) the inductance L and (c) the quality factor. According to Fig. 5(a), the R value of the LCW coil decreased in the range of N = 1to5asN increased. However, the R value of the LCW coil increased in the range of more than N = 5asN increased. As above in the 2 nd chapter, R dc and R s decreased as N increased. Therefore, the increase in R was due to R p. On the other hand, the R of the LMW coil decreased in the range of N = 1 to 15 and became almost constant in the range of more than N = 15 as N increased. The reason for this is that the LMW coil reduced R p.also, the minimum R values of the COW and LCW and LMW coils were 2.2 Ω,3.8Ω (N = 15) and 1.6 Ω (N = 15), respectively; thus, the R of the LMW coil was the lowest. According to Fig. 5(b), the L values of the LCW and LMW coils decreased as N increased (14). The L of the LMW coil was higher than that of the LCW coil because the magnetic thin film of MPW stocks with magnetic energy. According to Fig. 5(c), the quality factor of the LCW coil increased in the range of N = 1 to 5 and decreased in the range of more than N = 5asN increased. This is because, as above, the quality factor of the coils is inversely proportional to R and proportional to L, andther of the LCW coil increased and the L of the LCW coil decreased in the range of more than N = 5asN increased. On the other hand, the quality factor of the LMW coil increased in the range of N = 1 to 15 and decreased at N = 20 as N increased. This is due to the decrease in the L of the LMW coil at N = 20. From the above mentioned results, it was experimentally argued that the highest N in terms of the quality factor of the coils exists for f = 13.56MHz. Also, the maximums quality factor values of the COW, LCW and LMW coils were 166, 120 (N = 5) and 251 (N = 15), respectively; thus, the quality factor of the LMW coil was 51% or 109% greater than those of the COW and 37 IEEJ Journal IA, Vol.3, No.1, 2014

4 (a) Resistance Fig.6. Coupling coefficient vs. distance characteristic ( f = 100 khz) (b) Inductance Fig. 7. Transmission efficiency vs. output power characteristic at l = 9mm(l/d = 0.25, f = MHz) (c) Quality factor Fig. 5. Impedance vs. number of strands characteristics of coils ( f = MHz, n = 8) LCW coils, respectively. Also, the quality factor of the LCW coil with N = 15 was 104 and that of the LMW coil was 141% greater than that of the LCW coil. The reason for this is that the LMW coil reduced R p. Therefore, the efficiency characteristics of the COW, LCW (N = 15) and LMW (N = 15) are compared in the 4 th chapter. A target output power of 5 W, a transmission distance l of 9 mm, which is industrially practical (transmission distance l/coil diameter d = 0.25) and l of 36 mm, which is expected for long-distance transmission (l/d = 1), are considered. 4. Transmission Characteristics 4.1 Coupling Coefficient vs. Distance Characteristic Figure 6 shows coupling coefficient k vs. distance characteristic. k was measured at f = 100 khz. The k of the COW coil was always the highest and that of the LMW coil was always the lowest among the coils considered. However, the difference in k between the COW and LMW coils was 5%; thus, the dependence of the difference in k on the type of wire is negligible. 4.2 Transmission Characteristics at l = 9mm The transmission efficiency at l = 9mm (l/d = 0.25) was measured with an oscillator (Agilent 33522A), an amplifier (AR 75A250A), a directional coupler (Werlatone C ) and a wattmeter (Agilent U2004A). In this case, Z i and Z o were larger than Z 0 and Z L in all the coils, respectively; thus, the matching circuit shown in Fig. 2(a) was inserted. The matching inductors and capacitors were determined to be less than 1% for η 11 and η 22 at P i = 1 mw using the network analyzer. In addition, η 11 and η 22 were always less than 1% regardless of P i. Figure 7 shows the transmission efficiency η 21 vs. output power P o characteristics for the matching circuit at l = 9 mm. The η 21 values of the COW, LCW and LMW coils were 89%, 84% and 91% at P o = 5 W, respectively; thus, the transmission efficiency of the LMW coil was 2% or 7% greater than those of the COW and LCW coils, respectively. This is due to the highest quality factor of the LMW coil. Also, η 21 of the LMW coil was the highest among the coils, and the change in η 21 of the LMW coil depend on P o is the same as that of the COW coil. Thus, there is no increase in the hysteresis loss depends on P o in the range of less than 5 W of the P o. Figure 8 shows the heat generation characteristic of each coil for P o = 5 W and room temperature T = 25 C, which was measured with thermoshot (NEC Avio Infrared Technologies Co., Ltd., TYPE F30). According to Fig. 8, the temperature increases of the COW, LCW and LMW coils, ΔT, were 12 deg, 16 deg and 10 deg, respectively; thus, the ΔT of the LMW coil was lower than those of the COW and LCW coils. The reason for the reduced heating of the LMW coil is that the LMW coils should be the highest quality factor among the coils considered. 4.3 Transmission Characteristics at l = 36 mm Transmission efficiency at l = 36 mm (l/d = 1) was 38 IEEJ Journal IA, Vol.3, No.1, 2014

5 (a) COW (a) COW (b) LCW (b) LCW (c) LMW Fig. 8. Heat generation characteristic for P o = 5W at l = 9mm(l/d = 0.25, f = MHz, room temperature T = 25 C) (c) LMW Fig. 10. Heat generation characteristic for P o = 5W at l = 36 mm (l/d = 1, f = MHz, room temperature T = 25 C) Fig. 9. Transmission efficiency vs. output power characteristic at l = 36 mm (l/d = 1, f = MHz) measured. In this case, Z i and Z o were smaller than Z 0 and Z L in all the coils, respectively; thus, the matching circuit shown in Fig. 2(b) was inserted. Figure 9 shows transmission efficiency η 21 vs. output power P o characteristics for the matching circuit at l = 36 mm. In addition, η 11 and η 22 were always less than 1% regardless of P i and transmission efficiency did not decrease. Since the temperature of the LCW coil in P o = 2W was 69 C, the experiment was conducted at P o less than 2 W. The η 21 values of the COW, LCW and LMW coils were 63%, 52% and 69% at P o = 2 W, respectively; thus, the transmission efficiency of the LMW coil was 6% or 17% greater than those of the COW and LCW coils, respectively. Also, η 21 of the LMW coil was the highest among the coils, and the change in η 21 of the LMW coil depend on P o is the same as that of the COW coil. Thus, there is no increase in the hysteresis loss depends on P o in the range of less than 5 W of the P o. The difference in η 21 between the COW and LMW coils at l = 36 mm (l/d = 1) was larger than that at l = 9mm (l/d = 0.25). The reason for this is that the effectof the quality factor of the coils considered on η 21 is large since k is small. Figure 10 shows the heat generation characteristic of each coil for P o = 2 W and room temperature T = 25 C. According to Fig. 10, the ΔT of the COW, LCW and LMW coils were 25 deg, 44 deg and 18 deg, respectively. Thus, the LMW coil reduced the temperature increase. The reason for this is that the LMW coil reduced R p ;as a result, the quality factor of the coil and transmission efficiency increased, and the loss was minimized. 5. Conclusion In this paper, the following are discussed. ( 1 ) Quality factor of coil characteristic dependence on number of strands It was experimentally argued that 39 IEEJ Journal IA, Vol.3, No.1, 2014

6 the highest number of strands, N, in terms of the quality factor of the coils exists for f = MHz. The quality factor values of the COW, LCW and LMW coils with N = 15 were 166, 104 and 251, respectively; thus, the quality factor of the LMW coil was 51% or 141% greater than those of the COW and LCW coils, respectively. The reason for this is that the LMW coil reduced R p. ( 2 ) Transmission characteristics The k of the COW coil was always the highest and that of the LMW coil was always the lowest. However, the difference in k between the COW and LMW coils was 5%; thus, the dependence of the difference in k on the type of wire is negligible. The η 21 values of the COW, LCW and LMW coils were 89%, 84% and 91% for P o = 5Watl = 9mm(l/d = 0.25); thus, the efficiency of the LMW coil was 2% or 7% greater than those of the COW and LCW coils, respectively. In this case, the temperature increases of the COW, LCW and LMW coils were 12 deg, 16 deg and 10 deg, respectively. In addition, the η 21 values of the COW, LCW and LMW coils were 63%, 52% and 69% for P o = 2W at l = 36 mm (l/d = 1); thus, the efficiency of the LMW coil was 6% or 17% greater than those of the COW and LCW coils, respectively. In this case, the temperature increases of the COW, LCW and LMW coils were 25 deg, 44 deg and 18 deg, respectively. Thus, the LMW coil reduced the temperature increase. Namely the LMW coil proves useful for longdistance transmission. The reason for this is that the LMW coil reduced R p ; as a result, the quality factor of the coil and transmission efficiency increased, and the loss was minimized. References ( 1 ) Y. Matsuda and H. Sakamoto: Non-contact magnetic coupled power and data transferring system for an electric vehicle, J. Magn. Mater., Vol.310, No.2, Part 3, pp (2007) ( 2 ) L. Xiaoyu, F. Zhang, A. S. Hackworth, J. R. Sclabassi, and M. Sun: Wireless power transfer system design for implanted and worn devices, Proc. IEEE Annu. Northeast Bioeng. Conf., Vol.35, pp (2009) ( 3 ) B.G. Jong and B.H. Cho: An energy transmission system for an artficial heartusing leakageinductance compensation of transcutaneoustransformer, IEEE Trans. Power Electron., Vol.13, No.6, pp (1998) ( 4 ) Wireless Power Consortium, ( 5 ) T. Mizuno, A. Kamiya, D. Yamamoto, S. Yachi, and H. Kanazawa: Theoretical expression of efficiency of wireless power transfer using magnetic resonant coupling, JSAEM, Vol.19, No.2, pp (2011) ( 6 ) T. Takura, Y. Ota, K. Kato, F. Sato, H. Matsuki, T. Sato, and T. Nonaka: Relationship between efficiency and figure-of-merit in wireless power transfer through electromagnetic induction, J. Magn. Soc. Jpn., Vol.35, No.2, pp (2011) ( 7 ) H. Shinagawa, T. Suzuki, M. Noda, Y. Shimura, and T. Mizuno: Theoretical analysis of AC resistance in coil using magnetoplated wire, IEEE Trans. Magn., Vol.45, No.9, pp (2009) ( 8 ) J. Acero, R. Alonso, M.J. Burdio, A.L. Barragan, and D. Puyal: Frequencydependent resistance in litz-wire planar windings for domestic induction heating appliances,ieee Trans. Power Electron., Vol.21, No.4, pp (2006) ( 9 ) T. Mizuno, S. Yachi, A. Kamiya, and D. Yamamoto: Improvement in efficiency of wireless power transfer of magnetic resonant coupling using magnetoplated wire, IEEE Trans. Magn., Vol.47, No. 0, pp (2011) ( 10) T. Mizuno, A. Kamiya, Y. Shimura, K. Iida, D. Yamamoto, N. Miyao, and H. Sasadaira: Consideration on influences of number of strands on AC resistance of litz wire, JSAEM, Vol.18, No.3, pp (2010) ( 11) T. Yoshimoto: Antenna introduction to learn from basic, CQ publisher, pp (2007) (12) T. Mizuno, S. Enoki, T. Suzuki, M. Noda, H. Shinagawa, S. Uehara, and H. Kitazawa: Linearity Range of an Eddy Current Displacement Sensor in Relation to the Fe Film Thickness of Magnetoplated Wire, J. Magn. Soc. Jpn., Vol.31, No.2, pp (2007) ( 13) T. Mizuno, T. Ueda, S. Yachi, R. Ohotomo, and Y. Goto: Quality factor of coil dependent on number of strands of litz wire, SEAD24, pp (2012) (14) D. Sinha, A. Bandyopadhyay, P.K. Sadhu, and N. Pal, Computation of Inductance and AC Resistance of a Twisted Lizt-Wire for high Frequency Induction Cooker, IECR 2010 International Conference, pp (2010) Tsutomu Mizuno (Senior Member) was born in He received his M.E. degree in electrical engineering from Shinshu University in He joined AMADA Co., Ltd., in April of the same year. He became on assistant and assistant professor at the Department of Electrical Engineering, Shinshu University, in 1996 and 1999, respectively. He has been the same Professor since He has focused on linear motors, linear actuators, electromagnetic sensors and wireless power transfer. He is a member of IEEE, MSJ and JSAEM. Takuto Ueda (Student Member) received his B.E. degree in electrical in He is currently pursuing his M.E. degree in electrical and electronic engineering at Shinshu University. He has focused on wireless power transfer. Shintaro Yachi (Student Member) received his B.E. degree in electrical in He is currently pursuing his M.E. degree in electrical and electronic engineering at Shinshu University. He has focused on wireless power transfer. Ryuhei Ohtomo (Student Member) received his B.E. degree in electrical in He is currently pursuing his M.E. degree in electrical and electronic engineering at Shinshu University. He has focused on LLC converters. Yoshihito Goto (Student Member) received his B.E. degree in electrical in He is currently pursuing his M.E. degree in electrical and electronic engineering at Shinshu University. He has focused on wireless power transfer for robots inside the body 40 IEEJ Journal IA, Vol.3, No.1, 2014