A low-power, low-noise, and high-bandwidth mag-1 netometer that utilizes the magnetoimpedance (MI) element as a 2 sensor head is presented. The MI element has a high sensitivity, 3 and it can be implemented in the mm-scale through the MEMS 4 process. The analog front-end (AFE) circuit of the magnetometer 5 includes a digital calibration scheme that automatically enhances 6 the loop gain of the system, resulting in high bandwidth and 7 low-noise characteristics. The AFE circuit is designed based on 8 a switched-capacitor (SC) approach, and its dedicated switching 9 scheme can suppress the folded noise of an amplifier. A single-10 coil magnetic negative feedback architecture with correlated 11 double sampling (CDS) enables to achieve a high dynamic range 12 (DR) and stable passband gain in addition to simplifying the 13 structure of the MI element. The AFE chip of the magnetometer 14 is implemented in a 0.18-µm CMOS process, and it achieves an 15 8-pT/ √ Hz noise floor within a 31-kHz bandwidth and the DR of 16 96 dB, where the power consumption is 1.97 mW. 17 Index Terms-Analog front-end (AFE), biomagnetic, digital 18 calibration, Internet of Things (IoT), magnetic feedback, magne-19 toimpedance (MI) element, magnetometer. 20 I. INTRODUCTION 21 A BIOMAGNETIC sensing technique such as magneto-22 myography (MMG) or magnetoencephalography (MEG) 23 is one solution for capturing biological information with a min-24 imum invasive approach. Implantable MMG has the potential 25 to acquire fast neuronal magnetic activity, which corresponds 26 to the action potential of neurons close to skeletal muscle 27 with high spatiotemporal resolution [1], [2], [3], as opposed 28 to an approach with an optically pumped magnetometer that 29 achieves low noise but a relatively large size because of 30 the optical system [4]. Magnetometers for such applications 31 require of low noise less than 100 pT/ √ Hz, high bandwidth 32 over 10 kHz, low power, and small size because they are 33 implanted. Furthermore, a wide input range over 100 μT is 34 desired because there is a need to accept the geomagnetic field 35 and artifact without saturating the signal. A magnetic negative 36

The magneto-impedance (MI) effect was discovered about 30 years ago and a microsize magnetic sensor utilizing this effect has become commercially available. We make some modifications to the commercially available MI sensors to cover the dynamic range of the geomagnetic field. The total cost of three MI sensors for the two horizontal components and one vertical component including the modification is approximately one-third of the standard price of triaxial fluxgate magnetometer sensors. For the period of 30 March to 27 April 2018, we conducted experimental observations of geomagnetic field variations with the MI sensor magnetometer (MIM) at the Mineyama observatory, which is located about 100 km northwest of Kyoto, Japan. Data obtained with the MIM are compared with those from the fluxgate magnetometer (FGM) that has been working at the observatory. Results show that the MIM can record geomagnetic field variations such as geomagnetic storm, solar quiet variations, low-latitude positive bays, storm sudden commencement, and long-period geomagnetic pulsations with a peak-to-peak amplitude of ≤1 nT that is also detected with the FGM. Power spectra of the geomagnetic field variations measured with the MIM and FGM are almost the same. It is found that the MIM has a larger temperature drift than the FGM. The present study reveals that the MIM is comparable to the FGM in measuring the geomagnetic field variations in a period from a few tens of seconds to a few hours and is useful for researches in upper atmospheric physics or space physics. NOSÉ ET AL.