0 Introduction Commonly used data acquisition systems sample large signals, meaning the amplitude of the useful signal is greater than the noise signal. However, in some special situations, the acquired signal is very weak and submerged in a large amount of random noise. In such cases, general acquisition systems and measurement methods cannot detect the signal. This acquisition system's hardware circuit is optimized for weak signals, with a front-end conditioning circuit designed to effectively suppress common-mode signals using an instrumentation amplifier, ensuring the accuracy requirements of the acquired data. Magnetoresistive sensors are sensitive elements that detect the presence or magnetic strength (within an effective range) of magnetic objects. These magnetic objects include permanent magnets, paramagnetic materials, and can also sense the magnetic field around energized coils or wires. This paper selects the Honeywell HMC1002 magnetoresistive sensor, a two-dimensional magnetoresistive microcircuit, a dual magnetic field sensor in a small integrated circuit package (SOIC). The sensing directions of the two sensors are perpendicular to each other. Sensor A senses a magnetic field parallel to the long side of the outer package, and sensor B senses a magnetic field perpendicular to the long side of the outer package and parallel to the surface. The HMCl002 is highly sensitive to magnetic fields within the range of ±2 Gs (1 Gs = 10-4 T), exhibiting a linear output with a sensitivity of 3.2 mV/V/Gs and a resolution of 27 μGs. 1. Magnetoresistance Effect Principle and Sensor Working Principle 1.1 Magnetoresistance Effect Principle The phenomenon of a material's resistance changing in a magnetic field is called the magnetoresistance effect. There are ordinary and anisotropic magnetoresistance effects. Anisotropic magnetoresistance refers to the fact that when the applied magnetic field deviates from the magnetization direction inside a strongly magnetic metal (iron, cobalt, nickel, and their alloys), the metal's resistance decreases, while when parallel, there is basically no change. The resistivity ρ of the permalloy thin film depends on the angle between the magnetization intensity M and the current I direction. That is, the resistance change of permalloy (Fe20Ni80) is relatively large under weak magnetic fields, therefore, it is suitable for use under weak magnetic field conditions. 1.2 Sensor Working Principle 1) The most critical part of the entire sensor is the Wheatstone bridge. When an external magnetic field is applied, the resistance of the bridge changes, as shown in Figure 1, causing a change in the sensor output voltage Uout. Uout = (ΔR/R)Ub, where Ub is the sensor operating voltage. 2) The set and reset current bands are used to correct the sensor sensitivity. Magnetic fields exceeding 10 × 10⁻⁴ T in the external field will disrupt the polarization direction of the internal magnetic domains of the sensor, changing the sensor's output characteristics and reducing sensitivity. Applying pulses to the set and reset current bands unifies the polarization direction of the internal magnetic domains, improving sensitivity. 2. Data Acquisition System Design The entire data acquisition system consists of three main parts: the set/reset section, the signal conditioning section, and the acquisition section. The main function of the front-end conditioning circuit is to eliminate common-mode interference, amplify, filter, and differentially output weak signals, which are then transmitted to the data acquisition section via shielded wires. The data acquisition section completes data acquisition and storage. The set/reset section avoids the output attenuation caused by strong magnetic fields in the environment, ensuring the correctness and stability of the data output. 2.1 Set/Reset Circuit When a magnetoresistive sensor is exposed to an interfering magnetic field, the sensor element will be divided into several randomly oriented magnetic regions, resulting in sensitivity attenuation. Strong magnetic fields in the environment (greater than 5 × 10⁻⁴ T) will cause variations in the magnetic sensor output signal. To eliminate this effect and optimize the output signal, magnetic switching technology (SR+/SR-) is needed to cancel the remaining magnetic field. The HMC1002 uses a bias magnetic field to compensate for the interfering magnetic field. Specifically, by applying a 3–4 A, 20–50 ns pulse current to the thin film through the set/reset alloy strip integrated inside the chip, the magnetic regions can be re-aligned and unified in one direction, thus ensuring high sensitivity and repeatable readings. The set/reset circuit used in this system generates a strong current pulse of 11.2 A (>4 A), which meets the system requirements, thus enabling low-noise and high-sensitivity magnetic field measurement. 2.2 Signal Processing Circuit The front-end signal processing circuit consists of an instrumentation amplifier, a fourth-order Butterworth low-pass filter, etc., and its basic block diagram is shown in Figure 2. In data acquisition systems, if the signal to be measured is a very small signal, the resulting voltage signal will also be small and have variable amplitude. To make the signal amplitude appropriate, it needs to be amplified. General-purpose operational amplifiers cannot directly amplify weak signals; instrumentation amplifiers must be used. Instrumentation amplifiers have characteristics such as high input impedance, low output impedance, strong common-mode interference immunity, low temperature drift, low offset voltage, and high stable gain. This system uses the AD623 integrated instrumentation amplifier from Analog Devices (ADI). The AD623 has excellent CMRR (CMRR increases with gain), minimizing error, suppressing power line noise and its harmonics, high-precision DC/AC performance, and the ability to easily vary the gain from 1 to 1000 ohms using an external resistor. It is widely used as a preamplifier in systems detecting weak signals. This system uses two-stage amplification: the first stage amplifies by 10 times, and the second stage amplifies by more than 100 times, for a total amplification of more than 1000 times. Before sampling, the signal must be subjected to anti-aliasing filtering and high-frequency noise removal. This system primarily targets magnetic signals with frequencies below 1 kHz; therefore, the filter's cutoff frequency does not exceed 1 kHz. The filter exhibits a flat amplitude-frequency response and good attenuation characteristics, making it widely used in many filter circuit designs. Testing has shown it to achieve excellent filtering results. 2.3 Data Acquisition Section The system's data acquisition section utilizes a UA306 A/D converter. This converter boasts a 16-bit resolution, 16 or 32 channels, and a real-time sampling rate of 250 kHz. Furthermore, it offers numerous advantages, including high measurement accuracy, high speed, no need for an external power supply, convenient programming, and connection to a computer via USB. Data can be directly imported from the PC into the analysis software, making it very convenient to use. 3 Experimental Results and Analysis The system circuit was tested through indoor and outdoor experiments. 3.1 In the indoor experiment, when the magnetic sensor was placed horizontally, the tweezers moved parallel to the A-axis of the magnetic sensor at a height of 20 cm. Under the same conditions (except for the movement speed), the influence of the magnetic object's movement speed on the magnetic sensor's output signal was easily observed by comparing the test results: the movement speed affects the width of the magnetic signal; when the speed is higher, the width of the pattern is narrower; conversely, the width of the pattern is wider. Multiple comparisons of the output signal amplitude showed that the amplitude did not change significantly with speed variations; the output amplitude of the magnetic signal depended only on the different magnetic media being measured. 3.2 In the outdoor experiment, signals were collected from different targets (Mitsubishi cars and minibuses) using the linear and circular methods, with a sampling rate of 3 kHz. In the linear method, the distance between the target and the sensor was 4 m, and the speed of the Mitsubishi car was 20 km/h, while that of the minibus was 14 km/h. Figures 3 and 5 show the signal collection results for the Mitsubishi car, and Figure 4 shows the signal collection results for the minibus. 1) Comparing Figures 3 and 4, it can be seen that the effective signal span in Figure 3 is smaller than that in Figure 4. The amplitude of the output magnetic signal in Figure 3 is greater than 0.4V, while it is less than 0.3V in Figure 4. This means the output signal amplitude in Figure 3 is larger than that in Figure 4. This is because the target in Figure 3 moves at a higher speed. Additionally, since the two targets are different, their influence on the magnetic field around the sensor is different, resulting in different measured output signal amplitudes. 2) Comparing the linear and circular methods, it can be seen that the curve variation patterns in Figures 5, 3, and 4 are significantly different. The large variation in curve shape is due to the different target movement patterns, i.e., the different ways they influence the magnetic field, leading to differences in the measured curves. The large peak in the curve in Figure 5 is because the vehicle passes through the positive direction of the sensor's sensitive axis once during signal acquisition while traveling in a circle. Furthermore, due to the limited acquisition time, the curve does not reflect the periodic changes in the signal. If the acquisition time reaches two cycles, two main peaks will appear. Furthermore, comparing the curves of the Mitsubishi car traveling in different directions reveals that the main peaks of the curves in the graphs are biased in opposite directions because the target's movement direction is opposite. This is determined by the characteristics of the magnetoresistive sensor. 4. Conclusion This paper utilizes effective analog signal conditioning methods to design a weak magnetic signal acquisition system based on a magnetoresistive sensor. To ensure the accuracy of the acquired signals, this system is equipped with a set/reset circuit. In addition, the system also features the ability to detect minute signals and simple data acquisition.