A Self-Adaptive Low-Voltage Current Mode ASK Demodulator for RFID Tags Wei Liu, Yongming Li, Chun Zhang, Zhihua Wang Tsinghua National Laboratory for Information Science and Technology Institute of Microelectronics, Tsinghua University, Beijing 100084, China Email: liu-wei06@mails.tsinghua.edu.cn Abstract Proposed a self-adaptive current mode amplitude shift-keying (ASK) demodulator to meet the requirements of the Low-Voltage RFID tags. This demodulator improves the dynamic demodulation performance by converting voltage signal to more detectable current signals, by employing two-stage current-peak-hold technique, and by introducing leak circuit. It could operate under the supply voltage range from 0.6V to 1.8V, the input carrier magnitude range from 250mV to 1.1V, and the modulation depth range from 20% to 100%. Under the voltage supply of 1.8V, its dynamic demodulation range is from 80nA to 3.96μA. The demodulator is designed and implemented with 0.18-μm CMOS technology process. Index Terms RFID; current-mode demodulator; current peak hold; Low-Voltage EEACC: 1250 CLC number: TN292 Document Code: A 0 INTRODUCTION The Radio Frequency Identification (RFID) system, as one type of electric identification system, is capable of tag reading and tag writing over a long distance [1]. The signal strength of passive RFID tags ranges from several hundred millivolts to a few volts as the reading/writing distance and orientation varies [2]. ASK demodulator used in RFID tags should not only meet the low-voltage and low-power requirements, but also have great dynamic range and high detection sensitivity. To ensure an adequate energy supply in far-field [3], the modulation depth of ASK signal is required to be very shallow, such as 20%, although ISO/IEC 18000-6 protocol just requires 27%~100%. As in RFID tags, digital modules are the major source of power dissipation, reducing the voltage supply of the whole RFID tag could drastically reduce its overall power dissipation. However, it is very difficult to further reduce the voltage supply of analog modules in voltage-mode [4, 5] implementations due to the analysis in [6], which points out that lower node resistance in current-mode circuits allows very low voltage supply. In addition, current-mode circuits have many other advantages over voltage-mode circuits, such as higher speed and larger bandwidth, at the expanse of higher power dissipation. Two current-mode demodulator structures are proposed in [7, 8]. [7] utilizes current edge detection technique, which brings in high power consumption and small dynamic detection range due to its unchangeable reference current. [8] adopts current-peak-hold (CPH) technique to convert small modulated voltage change into large current difference, which is easier to be detected by demodulators. However, it has the following disadvantages: a) The fixed reference current to peak current ratio limits the farthest working distance of tags and the minimum modulation depth of RF modulated signal; b) The demodulator can work properly only when the position of tag is fixed. To resolve these problems, a self-adaptive low-voltage current mode ASK demodulator is proposed, which is composed of a V-to-I convertor (voltage to current convertor), a two-stage CPH and a leak circuit. This demodulator has large dynamic detection range, and could work properly when the voltage supply is as low as 0.6V in far-field. Section 2 first gives an overview of the ASK demodulator and then discusses the design details in 2.1, 2.2 and 2.3. Finally, section 3 and 4 contain preliminary
results and conclusion respectively. 1 DEMODULATOR ARCHITECTURE The block diagram of the ASK demodulator is shown in Fig.1. Envelope detector extracts and filters envelope from RF modulation signal received by a dipole antenna. V-to-I convertor limits the current amplitude, and accordingly, reduces the power dissipation of the whole circuit. The subtraction circuit exports I DIFF, the current difference between I PK1 and I SIG, to CPH2 and current comparator. The current comparator is a transimpedance amplifier used to generate the output voltage V ASK from I DIFF *2 and I REF. The circuit uses two-stage CPH to extend the dynamic detection range and leak circuit inside CPH1 to enable the demodulator to function properly while the tag is moving. envelope detector. As Table І shows, when the input RF modulated signal is as large as 4V, the V MAX is 2.14V, which satisfies the requirement mentioned above. TABLE І Simulation Result of Envelope Detector and V-to-I Convertor 表 І 包络检波及电压电流转换器仿真结果 VDD ANT V SIG (V) I SIG (μa) V DD V IN (V) (V) I dx V MAX V MIN I MAX I MIN 0.6 0.25 0.2 0.53 0.40 0.14 0.06 1.8 0.25 0.2 0.53 0.40 0.28 0.11 1.8 1.1 0.2 1.70 1.56 3.96 3.43 1.8 4 0.2 2.14 2.07 4.81 4.56 Note: V IN is the signal amplitude and I dx is the modulation depth. I MAX /I MIN is the logic 1 / 0 of I SIG, which corresponds to V MAX /V MIN, the logic 1 / 0 of V SIG. Fig.1. Block Diagram of the Current-mode Demodulator 图 1 电流模解调器的结构图 1.1 Envelope Detector, V-to-I Convertor The envelope detector is composed of a Dickson voltage multiplier, a current sink, a voltage limiter and a filter. The function of the envelope detector is to extract the envelope of RF modulated signal. The schematic of V-to-I convertor is shown in Fig.2. In this circuit, the negative feedback introduced by resistance R3 could guarantee that the output I SIG is within the detection range of the post-processing circuit. To limit the amplitude of V SIG below 2.5V, which is a requirement of 0.18μm CMOS process, a voltage limiter is included in Fig.2. Schematic of the V-to-I Convertor 图 2 电压电流转换电路图 1.2 CPH1 Circuit and Subtraction Circuit As is shown in Fig.3, CPH1 circuit holds the maximum value of I SIG, which is subtracted from I PK1 by the subtraction circuit to generate I DIFF. As transistors connected to V P are just used to form a cascode structure to increase output resistance and obtain accurate current ratio, it is directly connected to supply power GND. To guarantee that the transistors connected to V N work in saturated region, node V N is generally set to 0.7V, which is an output of bias circuit. When current is small in far-field, it could alternatively set to 0.6V, the lowest supply voltage. Leak circuit composed of transistors M58~M61 is
shown in the rightmost block (shown by dashed line) in Fig.3. It allows the tag to move and could reenter the proper demodulation state after the tag is settled. Designing the circuit, we consider the following two aspects: a) It is active only when I SIG is lower than I PK1. When I SIG < I PK1, VOUT2= 1, which turns the transistor M61 on and discharge the node Vh1 slowly; otherwise, VOUT2 = 0, which turns the leak circuit off and allows I SIG to charge C 2 and C 3 through M49 quickly. When the tag is at settled place, the leakage current has little impact on the demodulation result. But while tags are moving from near-field to far-field, the leak circuit will be open until Vh1 reaches another stable state. While tags are otherwise moving from far-field to near-field, as the leak circuit is always closed, Vh1 could enter to stable state easily as soon as the tag. In ISO/IEC 18000-6 protocol, it is favorable that the time sine-carrier modulated by the logic of 1 is rarely. b) For shallow modulated signals, the leakage current should be small. Given the accuracy of post-comparator, Vh1 node voltage (0.45V~0.3V), capacitance and the pulse width restricted by protocol, the leakage current I leak range could be derived as: Vh1 Vh1 ) * ( C + C ) I * τ (1) ( H L 2 3 = leak I ave ave I < (2) where I leak is the average drain-source current of the transistors M62 and M65 when Vh1 changes from Vh1 H to Vh1 L. In this design, the leakage current is adjusted to 10nA@Vh1=0.45V. PW small/large as is shown in Fig. 5, n should be large/small to guarantee the proper behavior of the demodulator. Consequently, dynamic range of the demodulator is confined by n, which is a constant in actual design. Another factor that might decrease the dynamic range is the mismatch between I PK1 and the peak of I SIG. To alleviate this problem, the I REF is mutable in CPH2 to support both strong and weak signals. The current comparator compares I REF with I DIFF *2 to obtain the demodulation result V ASK. I DIFF *2 is used in the comparison to ensure the rise time and fall time of the output signal be equal. [7] discusses the design details of the current comparator. Fig.3. Schematic of CPH1 and Subtraction Circuit 图 3 CPH1 及减法器电路图 1.3 CPH2 and Current Comparator CPH2 is shown in Fig.4, whose responsibility is to extend the dynamic demodulation range of the tag by detecting and holding I REF, or the peak current difference of I PK1 and I SIG. If I PK1 /n is set as the reference current like [8], it is difficult for the modulator to cover large dynamic range of input modulated signal because the amplitude of I SIG varies widely, as is shown in Table І. When the logic 0 of I SIG is large/small while the dynamic range of I SIG is Fig.4. Schematic of CPH2 and Current Comparator 图 4 CPH2 及电流比较器电路图
Fig.5. Example to illustrate the dynamic range of [8] under both strong and weak input signals 图 5 文献 [8] 在强 弱输入信号下的动态范围分析 2 PRELIMINARY RESULTS The proposed demodulator, whose layout is presented in Fig.6, is implemented with 0.18-μm CMOS technology. The layout without pads occupies 140μm 190μm. Fig.7 presents the waveforms of the demodulation process of a strong ASK modulated input sine-carrier; while Fig.8 presents the waveforms of the demodulation process of a weak modulated signal, where I tot is the total current dissipation. All the input sine-carriers of both the above cases are of 915MHz frequency. Fig.7. Post simulation results when the amplitude of the RF input signal is 1.1V, modulation depth is 20% and the power supply voltage is 1.8V 图 7 1.8V 工作电压下解调电路输入载波幅度为 1.1V, 调制深度 20% 时的后仿真结果 Fig.8. Post simulation results when the amplitude of the RF input signal is 250mV, modulation depth is 20% and power supply voltage is 0.6V 图 8 0.6V 工作电压下解调电路输入载波幅度为 0.25V, 调制深度 20% 时的后仿真结果 3 CONCLUSION Fig.6.The layout of proposed demodulator 图 6 电流模解调器的版图 A self-adaptive low-voltage current mode ASK demodulator for RFID Tags implanted with CMOS 0.18-μm technology is presented. The circuit is based on current mode techniques and could operate when the voltage supply is as low as 0.6V. Under the voltage supply of 1.8V, the demodulation current range is from 80nA to 3.96μA, which corresponds to all inter-modulated signals under the modulation depth
of 20%~100% that could satisfy the requirements of ISO/IEC 18000-6 protocol. ACKNOWLEDGEMENT This research was partly supported by the National Natural Science Foundation of China (No. 60475018,) and National Key Basic Research and Development Program ( No. G2000036508 ) Beijing Municipal Science & Technology Development Program( D0305003040111). REFERENCES [1] K. Finkenzeller. RFID Handbook, Fundamentals and Applications in Contactless Smart Cards and Identification, Second Edition [M]. West Sussex, U.K.: John Wiley & Sons, 2003: 1-4. [2] Auto-ID Laboratory, Univ. of Adelaide. RFID analog front end design tutorial [EB/OL]. http://autoidlab. eleceng. adelaide.edu.au/tutorial. Aug. 2004. [3] John D. Kraus, Ronald J. Marhefka. Antennas: For All Applications [M]. Beijing, P.R.China : Publishing House of Electronics Industry, 2006: 30-33. [4] BAI R R, LI Y M, ZHANG C, et al. Novel Low-Voltage/Low-Power ASK Demodulator for RFID Tags [J]. Microelectronics, 2007, 37(6):790-793. [5] Cinco-Galicia J C, Sandoval-Ibarra F. A Low-Power 2.7μW, 915-MHz Demodulator for RFID Applications [C]//Electrical and Electronics Engineering, 2006 3rd International Conf. Mexico: Veracruz, 2006:1-4. [6] BARTHELEMY H, Current Mode and Voltage Mode: basic considerations [C]// Circuits and Systems, 2003: MWSCAS 03. Proceedings of the 46th IEEE International Midwest Symposium on. Egypt: Cairo, 2003: 161-163 Vol 1. [7] DJEMOUAI A, SAWAN M. New CMOS Current-Mode amplitude shift keying demodulator (ASKD) Dedicated for implantable electronic devices [C]// IEEE International Symposium on Circuits and Systems. Canada: Sheraton Vancouver, BC, 2004:I441-I444. [8] NKAMOTO H, YAMAZAKI D, YAMAMOTO T. A Passive UHF RF Identification CMOS Tag IC Using Ferroelectric RAM in 0.35-μm Technology [J]. IEEE J. Solid State Circuits (JSSC), 2007, 42(1): 101-110. 一种用于 RFID 标签的自适应低压电流模 ASK 解调器 刘伟, 李永明, 张春, 王志华 ( 清华信息科学与技术国家实验室, 清华大学微电子学研究所, 北京 100084) 摘要 : 针对 RFID 标签低压工作的要求, 设计了一种自适应电流模 ASK 解调器 通过把电压信号转换为电流信号 采用两级电流峰值保持技术以及泄漏电路等技术提高了解调器的动态检测性能 解调器的工作电源电压范围为 0.6V~1.8V, 能对输入载波幅度为 250mV~1.1V, 调制深度为 20%~100% 的信号进行正确解调 电源电压为 1.8V 时, 解调器的动态检测范围从 80nA 到 3.96μA 电路采用 0.18μm CMOS 工艺设计 关键词 : RFID; 电流模解调器 ; 电流峰值保持 ; 低压 EEACC: 1250 中图分类号 : TN492 文献标识码 : A 文章编号 :80722-43
作者简介 : 刘伟 (1982- ), 男, 湖北人, 清华大学微电子学研究所硕士研究生, 主要研究方向为 UHF 频段低压低功耗无源射频识别标签的设计实现 Liu Wei (1982- ), Male, was born in Hubei province, is a graduate of the Institute of Microelectronics, Tsinghua University. His current research interests include analog circuit design for UHF passive low-voltage/low-power radio frequency identification (RFID) tags. 李永明 (1945- ), 男, 四川人, 清华大学微电子学研究所教授, 长期从事模拟及数模混合集成电路的研究与设计 Li Yongming (1945- ), Male, was born in Sichuan province, is a professor of the Institute of Microelectronics, Tsinghua University, where he has been engaged in the research and development of analog and mixed signal integrated circuits for a long time.