A heightened demand for improved printed circuit board (PCB) fault detection arises with the increasing integration and enhanced functionality of PCBs. Traditional visible light image analysis demonstrates accuracy and safety advantages in nonintrusive detection. However, it exhibits limitations in detecting obscured faults or lack of visible defects. The study proposes a fault detection framework based on infrared thermal imaging, aiming to enhance the practicality and engineering efficiency of PCB fault detection. This article introduces an infrared-thermography-based framework for anomaly detection and fault classification of PCBs. The framework encompasses preprocessing infrared thermal images, extraction of multimodal feature vectors (MMFVs), density-based anomaly detection (DAD), and fault classification based on deviation matrix clustering. The framework extracts multimodal features from the residual temperature scalar fields and residual temperature gradient vector fields, using multiscale detection of global and local images for fault classification. Experimental validation on a two-phase drive (TPD) circuit illustrates a significant enhancement in PCB anomaly detection and fault classification accuracy compared with existing feature extraction techniques. This research provides an innovative and practical PCB manufacturing and maintenance tool, effectively elevating anomaly detection accuracy and operational feasibility in engineering practices.

The electric field is an essential electromagnetic environmental parameter in designing and constructing high-voltage direct-current (HVDC) transmission lines. In this paper, a differential structure is proposed to improve the performance of the field mill sensor. Two groups of vanes with the same characteristics generate induced currents with opposite phases, and the difference operation calculates the ground electric field. We calculated the charges and induced currents and optimized the number of vanes of the sensing electrode using the simulation model. Then we designed the I/V circuit and amplifier based on the differential structure and conducted circuit simulation and sensor calibration experiments. Compared with the single input electric field sensor, the noise of the sensor with the differential structure is significantly reduced, and the sensitivity reaches 0.085 V/(kV/m). The results indicate that the field mill sensor with improved differential structure enhances the anti-interference ability and sensitivity.

High voltage direct current (HVDC) transmission lines have been widely used in China, and the electromagnetic effects of the synthetic electric field generated by them on the surrounding environment have attracted more and more attention. Especially when there are residential buildings around the transmission lines, the distortion of the electric field makes the research and development of synthetic electric field sensors with three-dimensional (3D) measurement capability an urgent technical demand. Based on the basic principle of the field mill electric field sensor, this paper presents a three-dimensional field mill structure to obtain the three-dimensional component of the electric field vector by using unidirectional motion. With the signal processing circuit, we realize the reading of three-dimensional electric field data. The feasibility of the proposed structure is verified by theoretical deduction analysis, numerical simulation, and prototype tests. The proposed structure provides a new scheme to solve the problems of three-dimensional synthetic electric field measurement and can enrich the existing methods.

The increasing integration of electronic components and the complexity of printed circuit boards (PCBs) necessitate more efficient, accurate, and safe PCB inspection requirements. Using infrared thermography instruments to measure temperature fields to identify PCB faults has the advantage of non-intrusive detection compared to the traditional electrical measurement method. This paper establishes a thermal model of PCBs to overcome the mutual interference of heating components and enhance the quality of thermal images. A method for solving the power of heating components and decoupling the PCB temperature field using Green’s functions is proposed. Further, a robust and accuracy-optimized reconstruction method is proposed to reconstruct low-quality temperature fields, including denoising temperature fields with high-level noise and recovering incomplete temperature fields. The study validates the proposed temperature field decoupling method by analyzing a buck DC/DC converter, demonstrating its ability to accurately determine the thermal power of components and separate the temperature fields of different components. The proposed method uses partial temperature measurement points to estimate temperature fields, effectively suppressing the high-level Gaussian and salt-and-pepper noise in thermal images and restoring incomplete thermal images. The research results promise to expand the practical applications of online PCB testing technology and improve fault diagnosis.

Using magnetic field data to characterize the state of radio frequency (RF) printed circuit board (PCB) has been widely used in circuit state detection and fault identification. However, it may be difficult to measure the complete state of magnetic field radiation, so the characteristics of PCB may be challenging to establish, which makes the calculation method of reconstructing other radiation information of PCB with existing magnetic field information urgently needed. Based on the basic radiation principle of the electromagnetic field, we propose an algorithm that has advantages in reconstructing the magnetic field radiation state, in which the new single-layer iterative (NSLI) algorithm provides a more reasonable optimization range and field-source conversion matrix for the hybrid particle swarm optimization (HYPSO) algorithm, and converts the inverse problem of the large matrix into a forward optimization problem, further ensuring the accuracy and objectivity of the algorithm. According to the known magnetic field amplitude, this paper uses a two-phase drive circuit at 50 MHz frequency as a device under testing (DUT) to verify the proposed algorithm’s feasibility by reconstructing the complete three-dimensional magnetic field component radiated from the DUT and estimating the unknown magnetic field amplitude on the predicted plane. The specific experimental data shows that when the magnetic field amplitude is known, the relative errors of the reconstructed magnetic field on the x , y , and z components are 1.9%, 0.02%, and 0.06%. In the meantime, under the actual measurement environment with a maximum relative uncertainty of 0.59%, the relative error of the magnetic field amplitude of the unknown predicted plane is 6.15%.

Non-contact printed circuit board (PCB) fault diagnosis has been widely applied in PCB detection and maintenance. Due to objective factors such as visual blind spots, low-loss circuit structure design, and frequency insensitivity, traditional algorithms based on visual and temperature features are limited in practice. Therefore, the algorithm based on electromagnetic features containing rich physical connotations and prominent frequency features has received attention in PCB fault diagnosis. Based on the basic principles of electromagnetic physics and PCB fault relationship, this article proposes a scalar magnetic field source feature and further improves feature performance by adding topological relationships of multi-faults to generate the fusion feature. The backbone of the PCB diagnosis model is established on the Transformer architecture, effectively utilizing self-attention and parallel computing mechanisms to explore the inner correlation between each group feature. The paper provides a new non-contact PCB fault diagnosis solution that enriches existing methods. Besides, through actual experiments setting up multi-fault PCBs, the feasibility of our process is proved based on the proposed features and models. The specific multiple indicators Overall Precision (OP), Per Class Precision (CP), Overall Recall (OR), Per Class Recall (CR), Overall F1 Measure (OF1), Per Class F1 Measure (CF1), Accuracy (ACC), Mean Average Precision (mAP) are 98.55%, 94.89%, 98.55%, 95.11%, 98.55%, 95.32%, 96.01%, and 97.27%.