II. RECTENNA DESIGN

The proposed rectenna printed on a Taconic RF-35 substrate with a relative permittivity of 3.5 and loss tangent of 0.0018 is presented in Fig. 1. The receiving antenna is a microstrip-fed circular ring slot antenna (CSRA) with a CMRC, which would realize wideband harmonic rejection. According to the equation as follows [Reference Wong, Huang and Chen11], the inner radius r ₁ and outer radius r ₂ can be calculated, respectively:

(1) $$f_0 \approx \displaystyle{c \over {\pi (r_1 + r_2 )}} \times \left( {\displaystyle{{\varepsilon _r + 1} \over {2\varepsilon _r}}} \right)^{{\rm 1/2}}, $$

where c, ε _r , π(r ₁ + r ₂) is the speed light in free space, relative permittivity of the substrate, and mean circumference of the annular-ring-slot antenna, respectively.

Fig. 1.

The proposed rectenna with harmonic suppression (r ₁ = 10, r ₂ = 13.8. Unit: mm).

The core mechanism of harmonic rejection is the microstrip-fed CMRC as shown in Fig. 2, which has been reviewed in [Reference Kurgan, Filipcewicz and Kitlinski12], In order to investigate the property of CMRC, we only adjust c ₁ with other parameters fixed in order to simplify the design.

Fig. 2.

The designed CMRC (s ₁ = 0.1, s ₂ = 0.2, s ₃ = 0.2, w = 1.13. Unit: mm).

As shown in Fig. 3, changing c ₁ could affect the bandpass response of the CMRC. Figure 4 shows the simulated results and the corresponding measured results. The measured bandwidth of less than 3-dB insertion loss and better than 10-dB return loss is approximately 1.5 GHz when c ₁ = 13.6 mm. The insertion loss and reflection coefficient are 0.68 and −18 dB, respectively. The return loss in 2nd- and 3rd-order harmonic are 0.56 and 0.69 dB, respectively.

Fig. 3.

The effects of c ₁ on S ₁₁ (Unit: mm).

Fig. 4.

The simulated and measured S-parameter of CMRC (c ₁ = 13.6. Unit: mm).

Then the proposed receiving antenna is simulated on high-frequency structure simulator (HFSS), and the simulation result is demonstrated in Fig. 5. Here, a fundamental mode of the antenna is excited at exactly 2.45 GHz. Inset in the Fig. 5 shows the radiation pattern with or without the CMRC under the fundamental mode, which denotes that the CMRC does not have much effect on the radiation pattern. From the Fig. 5, a good reflection coefficient approximately −25 dB is achieved, while the harmonic waves, at the same time, are suppressed that the return loss at 2nd- and 3rd-order harmonic were 0.4 and 0.7 dB, respectively. Notably, the return loss is less than 1 dB over 3–8 GHz.

Fig. 5.

The simulated radiation pattern and S ₁₁ of the proposed receiving antenna.

To visualize the harmonic suppression of the proposed antenna, the current distributions in the antenna are plotted in Fig. 6. It is clearly seen that the CMRC has blocked the signal at the 2nd- and 3rd-order harmonic, while only the fundamental frequency goes through successfully.

Fig. 6.

Simulated current distribution diagrams of the proposed antenna and slot antenna without CMRC at 2.45, 4.9, and 7.35 GHz.

We have already reported a series-parallel rectifier with an inductance [Reference Kang, Lin, Jiang, Chen and Jiang13], which can develop the RF–DC efficiency notably as shown in Fig. 7. The series-parallel rectifier loads a proper inductance to compensate the capacitance of the diodes so that the loop of the RF can be matched quite well, and it directly connects with the receiving antenna.

Fig. 7.

The series-parallel rectifier with an inductance loaded.

III. EXPERIMENTAL RESULT

To validate the concept proposed above, a rectenna operating at 2.45 GHz is fabricated on a 0.5 mm Taconic RF-35 substrate (shown in Fig. 8). The size of the rectenna is 78 × 50 mm² and the diodes are HSMS-8202.

Fig. 8.

Photograph of the fabricated antenna.

Using Agilent E8267D as signal source and standard horn antenna (about 5 dBi gain at 2.45 GHz) as transmitting antenna, we calculated the efficiency of the rectenna by measuring the output DC voltage, as shown in Fig. 9. A power amplifier with a maximum gain of 20 dBi is used so that the available transmit power is large enough, and a standard horn antenna A₁ was used as transmitting antenna. Firstly, the received RF power was measured by a spectrum analyzer (Agilent N9030A) at a distance of 15 cm (far-field condition) from the transmitting horn antenna, and the receiving antenna is the designed microstrip-fed CSRA A₂. Then the prototype CRSA was replaced by the rectenna A₃ at the same location. The RF–DC conversion efficiency is defined as:

(2) $$\eta (\% ) = \displaystyle{{V_{DC}^2} \over {R_L P_R}} \times 100\%, $$

where V _DC is the output DC voltage of the rectenna, R _L is the load resistance and the P _R is the RF power received and measured by the designed microstrip-fed CSRA A₂. When the received power is about 10 dBm, the measured results are depicted in Fig. 10. It figures out that the output voltage increases as the load resistance increases while the maximum value of conversion efficiency could achieve up to 70.2% when the optimum load is 1 kΩ. Good agreement is obtained. Moreover, in Table 1, the performance of the proposed rectenna is summarized and compared with the existing reports.

Fig. 9.

The measuring experiment. (a) Experimental scheme (b) Experimental photograph.

Fig. 10.

Measured results. (a) Output voltage versus load resistance (b). Conversion efficiency versus load resistance (P _R  = 10 dBm).

Table 1.

Comparisons of the fabricated rectenna with the reported ones.

Table 1 illustrates the comparisons between the proposed rectenna and the previously reported ones. Compared with the works, the proposed design has the merits of compact size and relatively high efficiency.

IV. LARGE-SCALE RECTENNA ARRAY

When the transmitting RF energy is weak, antenna arrays as the receiving antenna is preferred. The high-gain antenna arrays could induce high voltage applying to the diodes. Figure 11 shows the simulated reflection coefficient and antenna directivity, respectively. Inset in Fig. 11 demonstrates the radiation patterns of the two antennas almost identical.

Fig. 11.

The simulated S ₁₁ and radiation pattern of the two antennas.

In this paper, it is very convenient to stack the proposed rectenna as a large-scale rectenna array, as shown in Fig. 12(a). There is only one rectifier combined with the array, which may limit the performance of the diode because the input voltage of the diode must be lower than the breakdown voltage. Hence, the DC accumulation is a good method to achieve a high-voltage rectenna. Figure 12(b) shows the proposed rectenna array by connecting output DC lines with the ground planes in parallel. Furthermore, it is also possible to improve the performance by increasing the slot ring antenna units in series.

Fig. 12.

The large-scale rectenna array. (a) Series-stacked rectenna array (N = 2, 3, …). (b) Parallel-DC connection rectenna array.