Design of Efficient Rectifier for Low-Power Wireless ... - IEEE Xplore

Design of Efficient Rectifier for Low-Power Wireless Energy Harvesting at 2.45 GHz Tag Jong Lee1 , Pavan Patil1 , Chiao Yi Hu1,2 , Mohammad Rajabi1 , Saeed Farsi1 , and Dominique M. M.-P. Schreurs1 1

2

KU Leuven, Leuven, 3001, Belgium National Chiao Tung University, Hsinchu, 300, Taiwan topology at the frequency of 2.45 GHz in the industrialscientific-medical (ISM) band.

Abstract— This paper reports an efficient rectifier operating at 2.45 GHz for wireless energy harvesting applications with low input power levels. Single diode shunt-mounted topology is adopted for operation with low input power level. The efficiency is measured as 27.7% at -20 dBm of input power, 39.2% at -15 dBm, and 51.2% at -10 dBm. The maximum efficiency of 61.7% is measured at -0.4 dBm input power.

II. R ECTIFIER D ESIGN Main components of a rectifier are the diode(s), a lowpass filter and a resistive load. The heart of a rectifier is the diode. The diode chosen for the present application is a Schottky diode. It is chosen mainly because of its low turnon voltage and small junction capacitance which enables it to work at the required high frequencies. The simulations were carried out in the Keysight Technologies’ Advanced Design System (ADS) software. Both the schematic and EM Momentum simulations of the rectifier circuit were performed, and the two results were found to be in good agreement. The substrate for the rectifier was chosen as Rogers Corporations RO3003 (tanD=0.0013, Er=3) because of its low dielectric loss. Different Schottky diodes were tried in the aforementioned rectifier configuration in simulations to check the most efficient one for the present application. SMS7630 and HSMS286x were compared and both designs result in similar efficiencies within 2% difference. Finally SMS7630-079 was chosen because of a small package size with a small parasitic capacitance and inductance. The rectifier efficiency is regarded as a major factor in the design considerations. The target of the design was set as overall efficiency, defined as in [4]: dc Output P ower ⌘= (1) Incident RF P ower It is known that the input power level is a determining factor in the choice of rectifier structure for efficient RF to DC conversion. Based on simulation results of several rectifier topologies, the single diode shunt-mounted configuration was chosen for its high conversion efficiency at low input power level. The used topology is shown in Fig. 1. It starts with a DC block capacitor, then the input signal proceeds to the rectifying network followed by a low pass filter and an output load resistor. A stub was added near the diode to increase the efficiency via impedance matching. The output DC power in the efficiency equation (1) was calculated by (2).

Index Terms— Energy harvesting, efficiency, nonlinear circuits, rectifiers, Schottky diodes

I. I NTRODUCTION Wireless power transmission (WPT) is increasingly getting more attention for low-power applications. For example, wireless sensor nodes [1], micro aerial vehicles (MAVs) [2], or medical implant electronic devices [3], are devices where the principle of WPT is applied for charging, leading to a reduction of maintenance cost for wireless sensor nodes, less weight for MAVs, and even less danger of losing lives by avoiding surgeries just to replace batteries of the implanted devices. However, the strength of ambient RF signals is weak and the input power of rectifiers is therefore very low for these applications. For this reason, the efficiency of rectifiers subject to low input power levels is important for wireless energy harvesting of these applications. There are various possible topologies for WPT rectifier implementation. Greinacher topology [4] cascades diodes, doubles voltages, and has high efficiency at high input power region. However its efficiency at low-input power is lower than that of a single shunt-mounted topology [5] because it also cascades forward voltage drops of diodes. A single diode shunt-mounted topology consists of only one diode, single forward voltage drop of the diode, small area and high efficiency with low input power, and is simple. State-of-the-art rectifiers published so far, however, had efficiencies less than 30% with Schottky diodes and 33% with CMOS technology at low input power level of -15 dBm [6]. This paper presents a simple single-diode shuntmounted rectifier that has 39.2% efficiency at -15 dBm with a conventional single Schottky diode shunt-mounted

978-1-4799-5507-7/15/$31.00 © 2015 IEEE

47

RWS 2015

(a)

Fig. 1. Schematic of the initial rectifier optimized with an unmodified package model (dimensions in mm).

PDC = Output DC P ower =

2 VDC RL

(2)

The optimized values of capacitors C1, C2, and load resistor RL were found to be 1 pF, 100 pF and 3.24 k⌦ respectively. The inductors L1 and L2 were both optimized to 15 nH. Real models for lumped components such as L, C, and diode were used in the simulations for better optimization and to get the results closer to the real measurements. In addition to the diode model, its package model is also used for more accurate modeling. We observed that the efficiency especially largely depended on the parasitic capacitance of the package model in simulation. Initial board was designed based on a package model of the diode given by the manufacturer. After fabricating an initial circuit board, we measured the efficiency versus input power at various levels and compared it with those from the simulations. Subsequently, we adjusted the parasitic capacitance value of the model. Fig. 2 shows the package model obtained in this procedure.

(b)

Fig. 3. Final design of the rectifier with the modified diode package model (a) schematic representation and (b) photograph of the fabricated rectifier. The dimensions mentioned in the schematic are in mm.

S11 [dB]

dB step. A power meter was used to accurately measure the input power at the rectifier considering the cable losses. The output DC voltage across the load resistor was measured by a multimeter. 0 -2 -4 -6 -8 -10 -12 -14 -16

Simulation Measurement

2

Fig. 2.

Package model of SMS7630-079 diode. Fig. 4.

After the adjustment, a newly optimized rectifier with the modified package model is fabricated as shown in Fig. 3, and is measured and reported in the next section.

2.1

2.2

2.3

2.4 2.5 2.6 2.7 Frequency [GHz]

2.8

2.9

3

Simulated and measured S11 of the rectifier.

Fig. 5 shows the efficiency with regard to the input power of the rectifier. This rectifier is measured to have efficiencies of 27.7% at -20 dBm, 39.2% at -15 dBm, 51.2% at -10 dBm of input power, and has a peak value of 61.7% at -0.4 dBm as shown in Fig. 5. Its efficiency starts dropping above 0 dBm due to the breakdown of the diode. Both graphs match very well when the input power is less than -4 dBm. Measured efficiencies are higher than simulated ones when the input power is larger than -4 dBm. However mismatch of simulation and measurement values at high input power region was not of relevance for this circuit, because the present design is focused on low-power operation. If high input power applications are

III. E XPERIMENTAL RESULTS As to characterize the RF behaviour of the rectifier, the measurement of S11 is carried out with a network analyzer from 2 to 3 GHz, and the measured and simulated results are as plotted in Fig. 4. Both results have low peak values of approximately -10 dB around 2.45 GHz which indicates low reflections at the operating frequency. The measurement of efficiency versus input power was done with a signal generator as the power source at 2.45 GHz with power levels from -20 dBm to +4 dBm with 1

48

efficiencies have higher values with higher input power. However, this work has better efficiency at -15 dBm even when compared with some of the rectifier configurations with higher input powers of -10 dBm.

to be considered, the diode model needs to be corrected further. 70

η [%]

60 50

IV. C ONCLUSION

40

A rectifier design for low input power wireless energy harvesting was reported. A single diode shunt-mounted topology is used and optimized for high efficiency. A modified package model of the diode is adopted. The measurements closely match with the simulation results. The efficiency at -15 dBm was measured as 39.2% with a peak value of 61.7% at -0.4 dBm. Thus this work reports that higher efficiencies compared to prior works can be achieved with a conventional single shunt-mounted topology by a careful modeling of rectifier and its optimization.

30 20

Simulation Measurement

10 0 -20

-15

-10 -5 Pin [dBm]

0

5

Fig. 5. Measured and simulated efficiencies versus input power of the rectifier.

ACKNOWLEDGMENT

45

This work was supported in part by FWO Flanders and the Hercules Foundation. The authors would like to thank Dr. A. Georgiadis for his valuable advice.

η [%]

40 35 30

20 2.4

R EFERENCES

Simulation Measurement

25 2.42

2.44 2.46 Frequency [GHz]

2.48

[1] D. S. Tudose and A. Voinescu, “Rectifier antenna design for wireless sensor networks,” in International Conference on Control Systems and Computer Science, May 29–31, 2013, pp. 184–188. [2] L. M. M. Tan, “Efficient rectenna design for wireless power transmission for MAV applications,” M. Sci. thesis, Naval Postgraduate School, Monterey, California, Dec. 2005. [3] C.-S. A. Gong, M.-T. Shiue, Y.-P. Lee, and K.-W. Yao, “Selfpowered integrated rectifier for wireless medical energy harvesting applications,” in Intl. Symp. VLSI Des., Automat. Test, Apr. 25–28, 2011, pp. 1–4. [4] J.-P. Curty, N. Joehl, F. Krummenacher, C. Dehollain, and M. J. Declercq, “A model for µ-power rectifier analysis and design,” IEEE Trans. Circuits Syst. I, vol. 52, pp. 2771–2779, Dec. 2005. [5] H. Sun, Z. Zhong, and Y.-X. Guo, “An adaptive reconfigurable rectifier for wireless power transmission,” IEEE Microwave Wireless Compon. Lett., vol. 23, pp. 492–494, Sept. 2013. [6] S. Hemour, Y. Zhao, C. H. P. Lorenz, D. Houssameddine, Y. Gui, C.-M. Hu, and K. Wu, “Towards low-power high-efficiency RF and microwave energy harvesting,” IEEE Trans. Microwave Theory Techn., vol. 62, pp. 965–976, Apr. 2014. [7] S. Riviere, F. Alicalapa, A. Douyere, and J.-D. L. S. Luk, “A compact rectenna device at low power level,” Progress In Electromagnetics Research C, vol. 16, pp. 137–146, 2010. [8] K. Lui, A. Vilches, and C. Toumazou, “Ultra-efficient microwave harvesting system for battery-less micropower microcontroller platform,” IET Microw. Antennas Propag., vol. 5, pp. 811–817, 2011. [9] G. A. Vera, A. Georgiadis, A. Collado, and S. Via, “Design of a 2.45 GHz rectenna for electromagnetic (EM) energy scavenging,” in Proc. Radio Wireless Symp., New Orleans, LA, 2010, pp. 61–64. [10] U. Olgun, C.-C. Chen, and J. L. Volakis, “Wireless power harvesting with planar rectennas for 2.45 GHz RFIDs,” in URSI International Symposium on Electromagnetic Theory, Berlin, Germany, Aug. 16– 19, 2011, pp. 329–331. [11] C. M. Ghiglino, “Ultra-wideband (UWB) rectenna design for electromagnetic energy harvesting,” M. Eng. thesis, Escola Tcnica Superior dEnginyeria de Telecomunicaci de Barcelona, Barcelona, Spain, Oct. 2010. [12] T. Le, K. Mayaram, and T. Fiez, “Efficient far-field radio frequency energy harvesting for passively powered sensor networks,” IEEE J. Solid-State Circuits, vol. 43, pp. 1287–1302, May 2008.

2.5

Fig. 6. Measured and simulated efficiencies versus input frequency at Pin = -15.4 dBm.

The power level at the signal source was set as -15 dBm (power meter is read as -15.4dBm) to measure efficiency versus frequency, as shown in Fig. 6. Except the small shift of frequency 0.02 GHz between the measurement and simulation, both results match closely with almost same peak values of efficiencies. TABLE I C OMPARISON OF MEASURED EFFICIENCIES OF RECTIFIERS . Ref

f[GHz]

Pin [dBm]

⌘[%]

Diode

[7]

2.45

-10

34

HSMS2860

[8]

2.45

-10

26.6

[9]

2.45

-10

31

[10]

2.45

-15

29

HSMS2852

[11]

2.45

-15

27

Schottky diode

[12]

0.906

-15

33

CMOS

This work

2.45

-15

39.2

SMS7630-079

HSMS2850 & SMS7630 SMS7630

Table I shows the comparison of efficiencies of various rectifiers published in literature. As observed from the table, this work has higher efficiency compared to others at the input power level of -15 dBm. There is a tendency that

49