Impedance Matching for RFID Tag Antennas Chye-Hwa Loo 1, Khaled Elmahgoub 1, Fan Yang 1, Atef Elsherbeni 1, Darko Kajfez 1, Ahmed Kishk 1, Tamer Elsherbeni 1, Leena Ukkonen, Lauri Sydänheimo, Markku Kivikoski, Sari Merilampi, and Pekka Ruuskanen 1 Department of Electrical Engineering University of Mississippi, University, MS 38677-1848, USA cloo@olemiss.edu, kelmahgo@olemiss.edu, fyang@olemiss.edu, atef@olemiss.edu, eedarko@olemiss.edu, ahmed@olemiss.edu, and taelsher@olemiss.edu. Tampere University of Technology Kalliokatu, 61 Rauma, Finland leena.ukkonen@tut.fi, lauri.sydanheimo@tut.fi, markku.kivikoski@tut.fi, sari.merilampi@tut.fi, and pekka.ruuskanen@tut.fi Abstract: Passive UHF RFID consists of a microchip attached to an antenna. Proper impedance match between the antenna and the chip is crucial in RFID design. It directly influences RFID performance characteristics such as the read range. It is known that an RFID microchip is a nonlinear load whose complex impedance varies with the frequency and the input power. We investigate the effect of using constant chip impedance while designing UHF RFID s for specific frequencies. Commercial Avery Dennison AD- was simulated by Ansoft HFSS software. Afterwards, the read range of the was obtained numerically and compared with the measured values using the Voyantic Tagformance RFID measurements system. Keywords: UHF RFID, antennas, RF identification, integrated circuits, complex impedance matching 1. Introduction Passive radio frequency identification (RFID) is an automatic wireless data collection technology where RFID reader transmits a modulated RF signal to the RFID which consists of an antenna and an integrated circuit chip. The chip receives power from the antenna and responds by varying its input impedance and thus modulating the backscattered signal with data. Important RFID characteristics are maximum read range and orientation sensitivity. In order to achieve optimum operating condition, the antenna impedance is essential to be matched correctly to the chip impedance that is changing with respect to both received power level and frequency of operation. Calculating an accurate reflection coefficient for antenna design is a challenging process especially with the complex varying chip impedance. Most commercial software packages are still unable to calculate correct S-parameters needed for complex port impedance matching. In this paper, a proper calculation of the reflection coefficient taking into consideration the complex chip impedance is presented. Also the effect of using constant chip impedance at certain frequency in designing RFID s is investigated. A commercial (Avery Dennison AD-) [1] was simulated and the numerical results are compared to the measurement result obtained from the Voyantic Tagformance RFID performance measurement instrument []. 749
. RFID Tag antenna Impedance Matching The goal of a antenna designer is to design an antenna that could increase the maximum read range of the RFID. This is mainly governed by Friis equation [3]: ( θ, φ) G ( θ ', φ' ) λ PG t t r pt r = (1) 4π P where λ is the wavelength, P t is the transmitted power by the reader, G t (θ,ϕ) is the gain of the transmitting antenna facing the, G r (θ,ϕ ) is the gain of the receiving antenna in the direction of the reader antenna, p is the polarization efficiency, T is the power transmission coefficient of the and P th is the minimum threshold power necessary to turn on the chip. Inspections of (1) shows that there are not many antenna parameters the designer could change to improve the performance of the antenna. RFID antenna is typically a small dipole antenna of about dbi gain and can not be easily improved without losing the omni-directional property of the antenna. As a result, the design of a good antenna depends heavily on the transmission coefficient T, which for optimal design, a good matching between the antenna impedance and the chip impedance should be implemented. The definition for the reflection coefficient between a complex generator and a complex load impedance is given in [4] as * Zc Z Γ a =. () Z + Z c The power reflection coefficient is then equal to Γ and thus the power transmission coefficient is a th T = 1 Γ. (3) The graphical approach of calculating the reflection coefficient for RFID complex matching using Smith chart is recently determined by Nikitin in [5] with reference to Kurokawa s original paper [6]. For maximum power transfer matching, the rule of thumb is to employ the complex conjugate matching, thus the antenna input impedance should be equal to the complex conjugate of the chip impedance. However, the impedance of the microchip is not a constant value and it is a function of both frequency of operation and the received power by the chip. As an example, the input impedance plot of Impinj Monza chip (which is used in the commercial Avery Dennison AD-) versus frequency of operation at different sensitivity is measured and shown in Fig. 1. The matching should be done using the chip impedance curve of minimum received power level. Conventionally, chip vendor supplied constant value of chip impedances for the three center frequencies that correspond to the primary regional frequencies of operation: Europe (866.5MHz), North America (915 MHz), and Asia (953 MHz). Nevertheless, the calculated power reflection coefficient curve from constant chip impedance value is not accurate for proper antenna design and tuning. Figure shows the AD- as simulated using Ansoft HFSS commercial software package. The simulated antenna impedance is plotted with the conjugated chip impedance in Fig. 3. Figure 4 is the calculated power reflection coefficients using the constant chip impedance values at various frequencies. This AD- was designed for North America operation in the frequency range of 9 MHz-98 MHz. It is a silver ink of (7 microns ink thickness) printed on Polyethylene Terephthalate (PET) substrate (3 mils). The conductivity of silver ink is assumed to have 1% conductivity of silver. At least 14 passes were needed for the imaginary part of the impedance matrix to converge in this simulation. Performance measurement of the was conducted in the RFID anechoic chamber at The University of Mississippi using the Voyantic Tagformance Lite RFID measurement system. Figure 5 compares the measured read range with the theoretical read range using the constant chip impedances for the three regional frequencies. The impedance at 866.5 MHz shows good agreement at the lower frequency range, 75
while 915MHz and 953 MHz impedances show good agreement at their corresponding frequency ranges. However, none of them have good agreement with the measurements for the entire range of frequency. Chip Impedance [Ohms] 4 - -4-6 -8-1 Re +6dBm Im +6dBm Re dbm Im dbm Re -6dBm Im -6dBm Re -11dBm(Extrapolated) Im -11dBm(Extrapolated) 85 9 95 1 Fig. 1. Input impedance of Impinj Monza chip for different received power levels. Fig.. Image of the Simulated Avery Dennison AD- RFID using HFSS. 1 Impedance [Ohms] 8 6 4 Re Chip Im Chip (conjugate) Re Antenna Im Antenna 85 9 95 1 Fig. 3. Antenna impedances based on HFSS simulation along with the conjugate of the chip impedance at -11 dbm. 751
Power Reflection Coefficient [db] -5-1 -15 - -5 866.5 MHz 915 MHz 953 MHz -3 84 86 88 9 9 94 96 98 1 Fig. 4. The AD- power reflection coefficient Γ versus frequency. 7 6.5 6 Read Range [meter] 5.5 5 4.5 4 3.5 3.5 Measurement Theory 866.5 Mhz G =dbi Theory 915 MHz G =dbi Theory 953 MHz G =dbi 84 86 88 9 9 94 96 98 Fig. 5. Experimental and theoretical read ranges versus frequency when is matched to the chip impedance at 866.5MHz, 915MHz, and 953MHz. 75
3. Conclusion With specialized RFID analysis equipment Voyantic Tagformance the performance of the UHF RFID can be measured as a function of its maximum read range. Using single valued chip impedance for the entire frequency band is not an accurate way to calculate the power reflection coefficient due to the variation of the chip impedance with both power level and frequency. To design a good global RFID, the design must be based on the fact that the chip impedance is variable with both received power level and frequency of operation. Therefore, for having good antennas for RFID s variable chip impedance should be a parameter to consider in the design process. Acknowledgements The authors would like to thank Voyantic Ltd. for providing access to Voyantic RFID Tagformance measurement instrument which enabled us to perform the measurements presented in this paper. References [1] Avery Dennison http://www.rfid.averydennison.com/us/index.php [] Voyantic Tagformance TM Lite :http://www.voyantic.com [3] H.T.Friis, Proc. IRE, Vol. 34, p.54. 1946. [4] D. C. Youla, On Scattering Matrices Normalized to Complex Port Numbers, Proc. IRE, Vol. 49, p. 11, July 1961. [5] P. V. Nikitin, K. V. S. Rao, S. F. Lam, V. Pillai, R. Martinez, and H. Heinrich, Power Reflection Coefficient Analysis for Complex Impedances in RFID Tag Design IEEE Trans. Microwave Theory and Techniques, Vol. 53, No. 9, pp. 71-75, Sept. 5. [6] K. Kurokawa, Power waves and the scattering matrix, IEEE Trans. Microwave Theory and Techniques, MTT-13, No. 3, pp. 194-, Mar. 1965. 753