The Characterization of the Analog Devices Inc. (ADI) Magnetometer by Jefri Mohdzaini Submitted to the Department of Electrical Engineering and Computer Science in partial fulfillment of the requirements for the degree of Master of Engineering in Electrical Engineering and Computer Science at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY Feb 2000 ENG @ Massachusetts Institute of Technology 2000. All rights reserved. MASSACHUSETTS INSTITUTE OF TECHNOLOGY JUL 2 7 2000 A uthor ....... .................. ....L IBRARIES Department of Electrical Engineering and Computer Science Dec 16, 1999 Certified by John Geen il Company Supervisor Thesis Supervisor Certified by..... Martin Schmidt Professor ,Thesis, Supervisor Accepted by ...... Arthur C. Smith Chairman, Department Committee on Graduate Students The Characterization of the Analog Devices Inc. (ADI) Magnetometer by Jefri Mohdzaini Submitted to the Department of Electrical Engineering and Computer Science on Dec 16, 1999, in partial fulfillment of the requirements for the degree of Master of Engineering in Electrical Engineering and Computer Science Abstract The Analog Devices Inc. (ADI) magnetometer is a micromachined magnetic field sen- sor (MFS) that uses the Lorentz force to detect an external magnetic field. A circuit is designed to characterize the ADI magnetometer. The circuit does the following: drives an input into the magnetometer, and amplifies, rectifies and filters the mag- netomer's output. An average measured sensor output of 0.5mV was obtained in the presence of a bar magnet with a magnetic field of 8.2mT. The subsequent cir- cuitry (amplification, rectification and filtering) separates this output from interfering signals and raises its magnitude to about 150mV. Thesis Supervisor: John Geen Title: Company Supervisor Thesis Supervisor: Martin Schmidt Title: Professor 2 Acknowledgments First and foremost, I would like to thank my supervisors at Analog Devices, John Geen and Steve Lewis, who were always around when I needed them. I can't ever thank them enough. I learnt so much from them in so little time. I would also like to thank my MIT thesis supervisor, Professor Martin Schmidt, for his advice and quick turn-around time in helping me hand in this thesis in record time. My utmost appreciation to John Chang who was always willing to help me with non-user friendly oscilloscopes, microscopes and spectrum analyzers; Esther Fong who walked me through the pains and tribulations of drawing figures in Adobe Illustrator; Nilmoni Deb who helped me with the formatting in ITEX. Chani Langford who helped scan in all my figures; and last but not least, all my workmates in the 4th floor ADI Cambridge lab who were very supportive and kept life in perspective. From them I learnt a backup career option: "And would you like fries with that?". 3 Contents 1 A General Overview of Magnetometers 9 1.1 The Underlying Mechanisms of Magnetometers . . . . . . . . . . . . 9 1.1.1 The Hall Effect . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.1.2 The Magnetoresistive Sensor . . . . . . . . . . . . . . . . . . . 10 1.1.3 The Fluxgate Sensor . . . . . . . . . . . . . . . . . . . . . . . 11 1.1.4 The Lorentz Force Mechanism . . . . . . . . . . . . . . . . . . 11 1.2 Patented Micromachined Magnetometers . . . . . . . . . . . . . . . . 12 2 The ADI Magnetometer 16 2.1 Structure of the ADI Magnetometer . . . . . . . . . . . . . . . . . . . 16 2.1.1 Structure of the Sensor . . . . . . . . . . . . . . . . . . . . . . 16 2.1.2 Structure of the On-Chip Circuitry . . . . . . . . . . . . . . . 18 2.2 The Mechanism of the ADI Magnetometer . . . . . . . . . . . . . . . 18 2.2.1 The Underlying Mechanism of the Sensor . . . . . . . . . . . . 18 2.2.2 The Underlying Mechanism of the On-Chip Circuit . . . . . . 21 3 Characterizing the ADI Magnetometer 26 3.1 Designing the Driving Circuit . . . . . . . . . . . . . . . . . . . . . . 26 3.1.1 Calibration Technique . . . . . . . . . . . . . . . . . . . . . . 26 3.1.2 Calculating the Input Voltage . . . . . . . . . . . . . . . . . . 27 3.1.3 Calculating the Relative Displacement of the Sensor . . . . . . 31 3.1.4 Calculating the Expected Output . . . . . . . . . . . . . . . . 32 3.1.5 Stage 1: Implementation of the Driving Circuit . . . . . . . . 35 4 3.1.6 Method for Powering Up the Driving Circuit . . . . . . . . . . 36 3.2 Stage 2: Amplifying the Output of the Magnetometer . . . . . . . . . 36 3.3 Stage 3: Rectifying the Amplified Output . . . . . . . . . . . . . . . . 36 3.4 Stage 4: Filtering the Rectified Output . . . . . . . . . . . . . . . . . 37 4 Results and Discussion 38 4.1 Results of the Characterization . . . . . . . . . . . . . . . . . . . . . 38 4.2 Discussion in Discrepancy Between Expected and Experimental Results 41 4.3 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.4 Sum m ary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.5 Future W ork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 5 List of Figures 1-1 UK Patent GB2,136,581 by F. Rudolf filed on March 7, 1984 . . . . . 13 1-2 European Patent EP389,390 by E. Donzier et al. filed on March 19, 1990 13 1-3 European Patent EP392,945 by E. Donzier et al. filed on October 17, 1990 14 1-4 US Patent US5,036,286 by Holm-Kennedy et al. filed on July 30, 1991 15 2-1 The ADI Magnetometer. The sensor is located in the center of the magnetometer while the inputs are on the right side of the magne- tometer. The on-chip circuitry located on the bottom and the left side of the magnetometer are identical to each other. The plane of the magnetometer is defined as the x-y plane. . . . . . . . . . . . . . . . 17 2-2 Top view of the sensor. Since the only supports for the tethers are the anchors in the center, the structure of the sensor is similar to that of an um brella. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2-3 Side view of the sensor . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2-4 Route of current flow through the tethers and paddles. The dashed line represents the current going through the polyground located un- derneath the tether. The solid line represents the current going through the tether........ .................................. 22 2-5 Side view of the movement of paddles B and D to B' and D' respec- tively in the presence of the Lorentz force. . . . . . . . . . . . . . . . 23 2-6 Schematic of the on-chip circuit . . . . . . . . . . . . . . . . . . . . . 24 6 3-1 The off-chip circuit. It has four stages (from left to right): the driving circuit, the amplifier, the rectifier, and the filter. The magnetometer is located within the dotted lines. . . . . . . . . . . . . . . . . . . . 28 3-2 Plot of Voltage Difference vs. 6 (reduction in gap h) . . . . . . . . . . 30 3-3 A circuit model representing the relative displacement, A, and the out- put of the on-chip circuit . . . . . . . . . . . . . . . . . . . . . . . . . 33 4-1 A sample output of the magnetometer, outl . . . . . . . . . . . . . . 39 4-2 A sample output of the amplifier, outla . . . . . . . . . . . . . . . . . 39 4-3 A sample output of the rectifier, outir . . . . . . . . . . . . . . . . . 40 4-4 A sample output of the filter, out1f . . . . . . . . . . . . . . . . . . . 40 4-5 The magnetic field being bypassed away from the magnetometer chip due to the kovar leads. ...... ......................... 42 7 List of Tables 4.1 Voltage outputs of the on-chip magnetometer (outi and out2), amplifier (outla and out2a), rectifier (outir and out2r) and filter (out1f and out2f) in mV (peak-to-peak). Outi measures the B-field in the x-direction while out2 measures the B field in the y-direction. . . . . . . . . . . . 41 8 Chapter 1 A General Overview of Magnetometers The magnetometer is exactly what its name suggests it to be: a magnet meter. A magnetometer (also referred to as a magnetic sensor) measures the magnitude and direction of an external magnetic field. A magnetometer is useful in many areas, including navigation, automotive products and electronic devices. The Analog Devices Inc. (ADI) magnetometer is a micromachined magnetic field sensor (MFS) which uses the Lorentz force to detect an external magnetic field. How- ever, before delving specifically into the ADI magnetometer, this chapter focuses on gaining a general understanding of the many different underlying mechanisms of mag- netometers, and also on providing some background research on patented microma- chined magnetometers. 1.1 The Underlying Mechanisms of Magnetometers This section focuses on the classification, background, and the advantages and disad- vantages of various kinds of magnetometers. There are two main ways to classify a magnetometer: 1. Classification according to the type of output signal such as voltage, current, or frequency output. 9 2. Classification according to the underlying mechanism such as Hall effect, mag- netoresistive sensing, Lorentz force etc. The more common classification in literature is the classification according to underlying mechanism. This chapter will elaborate on the more popular underlying mechanisms for the magnetometer: the Hall effect, magnetoresistive sensing, fluxgate, and the Lorentz force. 1.1.1 The Hall Effect The discovery of the Hall effect by E.H. Hall in 1879 led to the birth of solid-state magnetic sensors. Hall was able to measure a cross-current in a thin gold layer on glass under the influence of a magnetic field. This was proof that the magnetic field exerts a force on the electric current in a conductor and not on a conductor itself, as was claimed by Maxwell[10]. The cross-current accumulates an excess charge, resulting in a transverse electrostatic field between the opposite edges of the gold layer. However, there was not much progress on solid-state magnetic field sensors until the silicon-based integrated circuit (IC) technology came of age 30 years ago. Since the 1970s, the Hall effect has been the underlying mechanism of magnetic field sensors due to its cost-effective way of measuring magnetic fields in the range of 1mT. The Hall sensors' range of 1mT is sufficient in most applications. For exam- ple, they can be used as non-contact switches, brushless electromotors (instead of commutators), and also to determine the position of the crankshaft in engines[7]. Unfortunately, for special applications such as the measurement of the deviation of the Earth's magnetic field or the measurement of biomagnetism, the Hall sensor's res- olution is insufficient. Besides poor resolution, the other problems with Hall sensors are the offset and the often limited dynamic range[6]. 1.1.2 The Magnetoresistive Sensor Magnetoresistive (MR) sensors are made up of thin strips of permalloy (NiFe magnetic film about 25nm thick). If a bias current is applied to the permalloy, the magnetic 10