Circuit Board Manufacturing Process Research Articles

PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes

To increase the reliability of the printed circuit board (PCB) manufacturing process, automated optical inspection is often employed for soldering defect detection. However, traditional approaches built on handcrafted features, predefined rules, or thresholds are often susceptible to the variation of the acquired images’ quality and give unstable performances. To solve this problem, a deep learning‐based soldering defect detection method is developed in this article. Like many real‐life deep learning applications, the number of available training samples is often limited. This creates a challenging low‐data scenario, as deep learning typically requires massive data to perform well. To address this issue, a multitask learning model is proposed, namely, PCBMTL, that can simultaneously learn the classification and segmentation tasks under low‐data regimes. By acquiring the segmentation knowledge, classification performance is substantially improved with few samples. To facilitate the study, a soldering defect image dataset, namely, PCBSPDefect, is built. It focuses on the dual in‐line packages (DIP) at the PCB back side, DIP at the PCB front side, and flat flexible cables. Experimental results show that the proposed PCBMTL outperforms the best existing approaches by over 5–17% of average accuracy for different datasets.

Surface Roughness Effect From Different Surfaces of Microstrip Lines and Reference Plane

Microstrip line structures consist of different conductors, such as trace and reference planes, which have different surface roughness levels due to the printed circuit board manufacturing process. The bottom surface of the trace is often rougher than the top surface of the trace, and the roughness levels of different reference planes vary for different foil types and manufacturing processes. To accurately model the additional conductor loss caused by such differences in microstrip lines, a new modeling method is proposed with different roughness levels on different surfaces and a reference plane, in contrast to the traditional roughness modeling approach, which considers a uniform roughness distribution for all surfaces. The effect of the different surfaces on the resistance is determined by applying additional microstrip models to analyze the effect of surface roughness from different surfaces and to improve the accuracy of insertion loss prediction based on the modeled total resistance.

A Hierarchical XGBoost Early Detection Method for Quality and Productivity Improvement of Electronics Manufacturing Systems

This paper presents XGBoost classifier-based methods to solve three tasks proposed by the European Prognostics and Health Management Society (PHME) 2022 conference. These tasks are based on real data from a Surface Mount Technologies line. Each of these tasks aims to improve the efficiency of the Printed Circuit Board (PCB) manufacturing process, facilitate the operator’s work and minimize the cases of manual intervention. Due to the structured nature of the problems proposed for each task, an XGBoost method based on encoding and feature engineering is proposed. The proposed methods utilise the fusion of test values and system characteristics extracted from two different testing equipment of the Surface Mount Technologies lines. This work also explores the problems of generalising prediction at the system level using information from the subsystem data. For this particular industrial case: the challenges with the changes in the number of subsystems. For Industry 4.0, the need for interpretability is very important. This is why the results of the models are analysed using Shapley values. With the proposed method, our team took the first place, capable of successfully detecting at an early stage the defective components for tasks 2 and 3.

A Hierarchical XGBoost Early Detection Method for Quality and Productivity Improvement of Electronics Manufacturing Systems

This paper presents XGBoost classifier-based methods to solve three tasks proposed by the European Prognostics and Health Management Society (PHME) 2022 conference. These tasks are based on real data from a Surface Mount Technologies line. Each of these tasks aims to improve the efficiency of the Printed Circuit Board (PCB) manufacturing process, facilitate the operator’s work and minimize the cases of manual intervention. Due to the structured nature of the problems proposed for each task, an XGBoost method based on encoding and feature engineering is proposed. The proposed methods utilise the fusion of test values and system characteristics extracted from two different testing equipment of the Surface Mount Technologies lines. This work also explores the problems of generalising prediction at the system level using information from the subsystem data. For this particular industrial case: the challenges with the changes in the number of subsystems. For Industry 4.0, the need for interpretability is very important. This is why the results of the models are analysed using Shapley values. With the proposed method, our team took the first place, capable of successfully detecting at an early stage the defective components for tasks 2 and 3.

Smooth Bend Structures Using Hybrid Compensation for Common-Mode Noise Reduction

The proposed smooth bend structure was implemented with a 150-degree bend to reduce the asymmetrical part of the differential bend and thus prevent mode conversion loss and common-mode noise. The smooth bend structure maintained the differential signal integrity. In addition, we proposed several hybrid compensation methods to enhance common-mode noise suppression and concluded that a smooth bend using the “L-C-L” compensation method offered the best performance. The frequency-domain analysis, from direct current (DC) to 6 GHz, was suppressed below 17.3 dB for the differential to common-mode conversion loss (Scd21) and maintained a differential insertion loss (Sdd21) above 2.18 dB. Furthermore, the 150-degree bend, using the “L-C-L” compensation method, reduced common-mode noise by 48.9% compared with a 150-degree bend structure in the time-domain analysis. The significant advantage of the proposed structure is that it can be easily fabricated using the printed circuit board (PCB) manufacturing process, reducing costs. The measurement results of the proposed design showed good consistency with the simulations. The proposed hybrid compensation structure successfully reduced the undesirable effects caused by asymmetry; as a result, mode conversion suppression was improved, and common-mode noise was reduced.

A Novel Data Mining Approach for Defect Detection in the Printed Circuit Board Manufacturing Process

Abstract This research aims to propose an effective model for the detection of defective Printed Circuit Boards (PCBs) in the output stage of the Surface-Mount Technology (SMT) line. The emphasis is placed on increasing the classification accuracy, reducing the algorithm training time, and a further improvement of the final product quality. This approach combines a feature extraction technique, the Principal Component Analysis (PCA), and a classification algorithm, the Support Vector Machine (SVM), with previously applied Automated Optical Inspection (AOI). Different types of SVM algorithms (linear, kernels and weighted) were tuned to get the best accuracy of the resulting algorithm for separating good-quality and defective products. A novel automated defect detection approach for the PCB manufacturing process is proposed. The data from the real PCB manufacturing process were used for this experimental study. The resulting PCALWSVM model achieved 100 % accuracy in the PCB defect detection task. This article proposes a potentially unique model for accurate defect detection in the PCB industry. A combination of PCA and LWSVM methods with AOI technology is an original and effective solution. The proposed model can be used in various manufacturing companies as a postprocessing step for an SMT line with AOI, either for accurate defect detection or for preventing false calls.

The Use of a Convolutional Neural Network in Detecting Soldering Faults from a Printed Circuit Board Assembly

Automatic Optical Inspection (AOI) is any method of detecting defects during a Printed Circuit Board (PCB) manufacturing process. Early AOI methods were based on classic image processing algorithms using a reference PCB. The traditional methods require very complex and inflexible preprocessing stages. With recent advances in the field of deep learning, especially Convolutional Neural Networks (CNN), automating various computer vision tasks has been established. Limited research has been carried out in the past on using CNN for AOI. The present systems are inflexible and require a lot of preprocessing steps or a complex illumination system to improve the accuracy. This paper studies the effectiveness of using CNN to detect soldering bridge faults in a PCB assembly. The paper presents a method for designing an optimized CNN architecture to detect soldering faults in a PCBA. The proposed CNN architecture is compared with the state-of-the-art object detection architecture, namely YOLO, with respect to detection accuracy, processing time, and memory requirement. The results of our experiments show that the proposed CNN architecture has a 3.0% better average precision, has 50% less number of parameters and infers in half the time as YOLO. The experimental results prove the effectiveness of using CNN in AOI by using images of a PCB assembly without any reference image, any complex preprocessing stage, or a complex illumination system. Doi: 10.28991/HIJ-2022-03-01-01 Full Text: PDF

High-Resolution Permanent Magnet Drive Using Separated Observers for Acceleration Estimation and Control.

This paper proposes a high-resolution permanent magnet (PM) motor drive based on acceleration estimation and control. The PM motor is widely implemented in the printed circuit board (PCB) manufacturing process. To achieve the demanded 1 μm drilling resolution, a sine/cosine incremental encoder is usually installed for motion control. In this paper, several improvements are developed to increase the motion control steady-state accuracy balancing transient response. First, the interpolation of every two encoder counts is proposed to increase the position sensing resolution. In this case, the transient response is improved through the high-resolution position feedback. Second, a closed-loop observer with two independent bandwidths is proposed for acceleration estimation. By using the interpolated position for acceleration estimation, the vibration-reflected high-frequency torque harmonics can be compensated through the acceleration closed-loop control. It reduces the steady-state error under the same sensing hardware. According to experimental results, both transient response and steady-state error can be improved on a PM motor using the proposed position interpolation and acceleration control.

Effect of printed circuit board dust on the workability and mechanical properties of self-compacting concrete: A preliminary study

Electronic waste has had an adverse impact on the environment in this era of fast electronic technology advancement and rapid consumption growth. Therefore, awareness and focus on the sustainable disposal of electronic products should be raised. The objective of this research is to study the feasibility of using the dust from the printed circuit board manufacturing process (PCB dust) as a binder to replace ordinary Portland cement (OPC) for manufacturing of self-compacting concrete (SCC). All of the seven SCC mixtures were obtained by replacing OPC with PCB dust by the volume ratio of 0%, 5%, 10%, 15%, 20%, 25%, and 30% respectively, then the workability and the properties of hardened concrete were investigated. From the experimental results, it can be seen that the workability of SCC tends to decrease with the PCB dust replacing increases, and clear that SCC tends to block during the flow and segregate. The engineering properties of hardened concrete indicates that the efficiency decreases as the amount of PCB dust increases. The highest 28-day compressive strength obtained is about 54 MPa at 5% replacement level and tends to decrease between 11.49% and 57.44% at replacement level between 5% and 30% by volume, when compared to the control mix. However, when SCC is mixed with 20% of PCB dust, it can provide a compressive strength of more than 50 MPa after 180 days. The results of the study point out the possibility of recycling PCB dust as a new filler that can be utilized to reduce the consumption of natural materials.

A Low-Cost and Efficient Microstrip-Fed Air-Substrate-Integrated Waveguide Slot Array

A microstrip-fed air-substrate-integrated waveguide (ASIW) slot array with high efficiency and low cost is presented. The design cuts out the substrate material within SIW, replaces the vias with metallic sidewalls, and uses a simple microstrip line-waveguide transition to feed the slot array. Radiating slots are cut on a 5-mil brass-plate, which covers the top of the substrate cutout to resemble a hollow waveguide structure. This implementation provides a simple and efficient antenna array solution for millimeter-wave (mm-wave) applications. Meanwhile, the fabrication is compatible with the standard printed circuit board (PCB) manufacturing process. To demonstrate the concept, a 4-element ASIW slot array working at the n257 band for 5G communications was designed using low-cost Rogers 4350B and FR4 substrate materials. Our simulation result shows 18% more efficiency than a conventional SIW slot array using the same substrate. The fabricated prototype shows |S11| < −15 dB over 27–29 GHz and a peak realized gain of 10.1 dBi at 28.6 GHz. The design procedure, prototyping process, and design analysis are discussed in the paper.

Dielectric-Loaded SIW H-Plane Horn Antenna With Gradient Air Slots

In this letter, a new structure of a dielectric-loaded substrate integrated waveguide H-plane horn antenna with gradient air slots, fed by a coaxial pin, is proposed. This novel structure aims at improvements in the radiation characteristics, including the gain, front-to-back ratio (FTBR), both E- and H-planes half-power beamwidth (HPBW), and sidelobe level (SLL). The measurements of the fabricated antenna show 15.4 dBi gain, 37 dB FTBR, and narrow E- and H-plane HPBW of 35° and 27° respectively, at 24.15 GHz. The entire antenna structure design is briefly expressed to combine the advantage of one single Rogers TMM4 thermoset laminate layer for reduced manufacturing difficulties and ensured improved performance. The optimized proposed antenna is fabricated by standard printed circuit board (PCB) manufacturing process, with the overall size 61.2 mm × 18.6 mm × 3.175 mm.

Partially Filled Half-Mode Substrate Integrated Waveguide Leaky-Wave Antenna for 24 GHz Automotive Radar

A novel highly efficient leaky-wave antenna (LWA), based on a periodic set of holes created in a half-mode substrate integrated waveguide (SIW), is conceived for automotive radar. The computer-aided design of this partially filled SIW (PFSIW) LWA is carried out using a full-wave electromagnetic simulator. As a proof-of-concept, an LWA prototype having a plate size of 15 × 126 mm $^2$ is fabricated through a standard low-cost printed circuit board (PCB) manufacturing process. Our manufactured prototype yields more than 25% impedance bandwidth, targeting the 24 GHz automotive short-range-radar band. Moreover, the measured peak gain and total antenna efficiency reach up to 15.5 dBi and 85% at 24 GHz. Furthermore, the main radiating direction steers continuously from 38 $^\circ$ to 58 $^\circ$ with an average half-power beamwidth of 12 $^\circ$ when the operating frequency changes from 22 to 26 GHz. In comparison, a 160 mm long dielectric-filled SIW (DFSIW) LWA yields a measured scanning range of 49 $^\circ$ and a total antenna efficiency of 43.8%, whereas an 80.3 mm long air-filled SIW (AFSIW) LWA exhibits a measured scanning range of 9 $^\circ$ and a total antenna efficiency of 80%. Therefore, PFSIW technology provides an optimal tradeoff between DFSIW and AFSIW for the envisaged short-range automotive radar application.

Large-Area Resistive Strain Sensing Sheet for Structural Health Monitoring.

Damage significantly influences response of a strain sensor only if it occurs in the proximity of the sensor. Thus, two-dimensional (2D) sensing sheets covering large areas offer reliable early-stage damage detection for structural health monitoring (SHM) applications. This paper presents a scalable sensing sheet design consisting of a dense array of thin-film resistive strain sensors. The sensing sheet is fabricated using flexible printed circuit board (Flex-PCB) manufacturing process which enables low-cost and high-volume sensors that can cover large areas. The lab tests on an aluminum beam showed the sheet has a gauge factor of 2.1 and has a low drift of 1.5 . The field test on a pedestrian bridge showed the sheet is sensitive enough to track strain induced by the bridge’s temperature variations. The strain measured by the sheet had a root-mean-square (RMS) error of 7 compared to a reference strain on the surface, extrapolated from fiber-optic sensors embedded within the bridge structure. The field tests on an existing crack showed that the sensing sheet can track the early-stage damage growth, where it sensed 600 peak strain, whereas the nearby sensors on a damage-free surface did not observe significant strain change.

Development of a Resonant Microwave Sensor for Sediment Density Characterization.

In this study, a sensor based on the development of a planar antenna immersed in sediments dedicated to water content monitoring in this type of material is proposed and experimentally validated. It is produced by a conventional Printed Circuit Board (PCB) manufacturing process on a double-sided metalized FR4 substrate. The sensitivity of the sensor is ensured by the variation of the real part of the complex dielectric permittivity of sediments with water content at around 1 GHz. As shown, in this frequency range, electrode polarization and Maxwell–Wagner polarization effects become negligible, leading to only a bulk water polarization sensitivity. The sensor operates in the reflection mode by monitoring the variation of the resonant frequency as a function of the sediment density through the S11 reflection measurements. An experimental sensitivity of was achieved. Despite the simplification of data interpretation at the considered frequency, the influence of ionic species such as NaCl in sediments on the real part of the relative complex dielectric permittivity is highlighted. This demonstrates the importance of considering a second parameter such as the S11 level at low frequency or the electrical conductivity to extract the density from the frequency measurements.

Affordable Thin Lens Using Single Polarized Disparate Filter Arrays for Beyond 5G toward 6G.

This paper proposes a novel design approach for a thin lens with the aim of overcoming fineness limits in the commercial millimeter wave printed circuit board (PCB) manufacturing process. The PCB manufacturing process typically does not allow the fabrication of metallic patterns with a gap and width of less than 100 μm. This hampers expanding thin lens technology to 5G commercial applications, especially when such technology is considered for 60 GHz or higher frequency, which requires a finer gap and width of metallic traces. This paper proposes that problematic process conditions can be mitigated when a lens is designed by establishing single-polarized lumped element models where larger capacitance and inductance values can be obtained for the same patch and grid unit cells. While the proposed design technique is more advantageous at higher target frequencies, a 60 GHz application and a wireless backhaul system is selected because of a limited range of frequencies that can be measured by an available vector network analyzer. The required gap or width of metallic traces can be widened significantly by using the proposed single-polarized unit cells to acquire the same in-plane capacitance or inductance. This enables the lens operating at higher-frequency under the process limits in fabricable fine traces. Finally, the effectiveness of the simulated design procedure is demonstrated by fabricating a 60 GHz thin lens that can achieve a gain enhancement of 16 dB for a 4 × 4 patch antenna array with a gain of 16.5 dBi.

A Cross Slot Coupling to Enhance Bandwidth of Dual-Layer SIW Structure

In this paper, design characteristics of cross slot coupling have been explored and realized in a proposed dual-layer SIW prototype for bandwidth enhancement at 10.0 GHz. The assembled prototype consists of two SMA-microstrip input/output interface with low-loss microstrip-taper via transition and two manually stacked SIW structures electrically connected via a small cross slot coupling design. The proposed dual-layer SIW structure is designed using CST software and fabricated using conventional Printed Circuit Board (PCB) manufacturing process on Rogers 4003 C with = 3.38 and = 0.813 mm. The close agreement between simulated and measured results is observed within a frequency range studied of 9.2 GHz to 11.2 GHz with 19.0 % bandwidth performance. The used of cross slot coupling design in the assembled dual-layer SIW structure indicated 9.0 % bandwidth enhancement compared to the conventional multilayer design with rectangular slot coupling. The assembled dual-layer SIW structure with cross slot coupling design shows potential in several RF applications such as radar and satellite communication.

Laser speckle reduction by a hybrid MEMS diffuser device with continuous height profile

The hybrid MEMS diffuser device based on Si wafer and fire resistant 4 printed circuit board (FR-4 PCB) was designed in order to obtain a low-cost device with large aperture for laser speckle reduction. Reflective diffuser with continuous height profile was fabricated on Si wafer by micromachining technology. FR-4 PCB was used to fabricate the vibrating structure by printed circuit board (PCB) manufacturing process. The hybrid MEMS diffuser device including diffuser and vibrating structure was driven to vibrate by electromagnetic force with a frequency of 66 Hz. The laser speckle reduction of 71.8% was obtained for the MEMS diffuser device.

Realization of a Low-Cost Displacement Sensor on PCB With Two-Metal-Layer Coplanar Waveguide Loaded by an EBG Structure

There are some difficulties in the realization of a microwave displacement sensor with conventional CPW loaded by electromagnetic bandgap structure on RF laminates with a regular printed circuit board manufacturing process. These difficulties have made this process impossible in practice. In this paper, for the first time, this kind of sensor has been implemented using a two-metal-layer (TML) CPW as a transmission line to overcome these difficulties. The proposed sensor consists of a TML CPW line loaded by two periodic-air-hole structures at both sides of the transmission line. Sliding the periodic structure at one side of the transmission line makes a change in the maximum insertion loss at the fundamental central frequency of the bandgap proportional to the length of the displacement. Analysis, finite-element method simulation, and fabrication have been carried out, and an average sensitivity of about 9 dB/mm and a dynamic range of 3 mm have been achieved for the 21 unit cells sensor with a period of 6 mm at 14.24 GHz. Furthermore, some practical issues, such as influence of the air gap between the metal traces and the movable structured dielectric on the sensor behavior and robustness of the sensor against the temperature variations, are practically investigated.

Evaluation of process influences on surface chemistry of epoxy acrylate based solder mask via XPS, ToF-SIMS and contact angle measurement

Epoxy acrylate based solder mask formulations were conditioned by different printed circuit board (PCB) manufacturing and PCB assembly process stages. Depending on these different influences the chemistry of the solder mask surface was investigated regarding adhesion to possible adhesion partners. The combination of X-ray photoelectron spectrometry (XPS), time-of-flight secondary ion mass spectrometry (ToF-SIMS) and the contact angle method, for surface energy determination, provided a detailed understanding of the surface near region up to the topmost monolayer, which forms the contact zone in which adhesion takes place. The combination of ToF-SIMS and XPS provided molecular information of surface components comprising quantitative information. The influences of all process steps, like UV, chemical and thermal treatment, on the chemical surface composition and appearance were identified. Based on the results a chemical surface model could be created regarding the different adhesion mechanisms. It has been shown that an enrichment of siloxanes at the surface is generated by different mechanisms that were distinguished based on ToF-SIMS. Even though an oxidation process in the surface near region (10 nm) was indicated by XPS, no increase of the surface polar groups and thus no polarity increase could be observed within the first monolayer. A surface model derived from the analysis results shows generation and occupation of free sites at the surface through all stages of the process. An occupation of free sites by siloxanes from additives in the solder mask formulation results in a siloxane dominated topmost monolayer.

Substrate-integrated waveguide filters based on mushroom-shaped resonators

This paper presents a new class of quasi-elliptic pass-band filters in substrate-integrated waveguide technology, which exhibits compact size and modular geometry. These filters are based on mushroom-shaped metallic resonators, and they can be easily implemented using a standard dual-layer printed circuit board manufacturing process. The presented filters exploit non-resonating modes to obtain coupling between non-adjacent nodes in the case of in-line geometry. The resulting structure is very compact and capable of transmission zeros. In this work, the singlet configuration is preliminarily investigated, and a parametric study is performed. The design of three-pole, four-pole, and higher-order filters is illustrated with examples and thoroughly discussed. A four-pole filter operating at the frequency of 4 GHz has been manufactured and experimentally verified, to validate the proposed technique.