With the development of technology and the continuous improvement of living standards, people have become increasingly concerned about food safety issues. Due to the long-term high-speed operation of food production lines, metal fragments generated by machine wear may inadvertently mix into food. Failure to detect such metallic foreign objects in a timely manner poses a significant threat to human safety. Therefore, the development of high-quality metal detectors is of utmost importance, as they serve as a robust safeguard for people's lives. Aiming to address the unclear relationships among the output signal of metal detection systems, the frequency of the input signal, the size of metallic foreign objects, and the types of materials, this paper constructs a 2D model of a balanced metal detection system using ANSYS MAXWELL finite element analysis software. Combined with MATLAB analytical solution calculations, it conducts an in-depth study of the influence of input frequency, metal materials, and dimensions on the output voltage signal. The analytical solutions provide spatial distribution views of the output voltage signal as it changes with the metal's position and coil spacing, clearly demonstrating the variation patterns of the voltage signal. Through simulation analysis of the differential voltage signal characteristics under different parameters, it is found that a higher excitation frequency and a larger metal size result in a higher differential voltage amplitude and better detection sensitivity. Due to the different electromagnetic characteristics of different materials, the detection sensitivity of iron is higher than that of copper and aluminum. The good fit between the analytical solutions and simulation results clearly presents the distribution laws of output signals with respect to metal position, frequency, material, and size, providing theoretical and experimental verification references for metal detection technology.