Artificial neural networks and statistical modeling for electronic stress prediction using thermal profiling

Electronic components are constantly under stress due to factors such as signal density, temperature, humidity, and high current and voltage. Relatively little research has emphasized stress-level prediction under voltage stress. The purpose of this paper was to develop an electronic thermal profile model for stress-level prediction utilizing neural network (NN) and statistical approaches, such as multivariate regression models. Electronic components were removed from boards, subjected to different levels of stress, then replaced. An infrared camera was then used to capture information about component temperature changes over time under normal operating and stress conditions. Statistical analysis of the captured images suggests a strong correlation between thermal profiles and voltage stress levels. Artificial neural network (ANN) and statistical approaches were used to model temperature change profiles for components that had been stressed at different levels, and their predictive ability was compared. Separate data sets were used for model development and model verification. ANN prediction rates were around 70%, compared to 30% for the statistical approach. Experiments were also conducted to evaluate the robustness of the ANN model to the presence of noise in the data. Results suggested that the ANN model was able to accommodate the presence of noise. Various backpropagation (BP) learning algorithms were also evaluated and yielded similar average error rates. A 3-2-1 ANN topology performed better than 3-3-1 or 3-2-2-1 topologies, perhaps because the 3-2-1 topology has a higher data sample to nodes ratio than the other topologies.     

Ag and Cu migration phenomena on wire-bonding

After storing a plastic packaged sample at 250°C for 588 h, the Au plus 1% Pd wire composition was found to be changed. The Ag and Cu atoms can migrate from the wedge bond through the wire surface and arrive at the ball bond. At the same time, Ag and Cu atoms diffuse into the gold wire itself and form a layer type structure. The atom migration phenomena are due to three driving forces: diffusion, alloy formation, and Galvanic effect. The obtained diffusion rate constant is in the order of 10⁻¹² cm square per sec, which corresponds to an activation energy of 0.7∼0.76 eV. The phases sequence formed by diffusion is inconsistent with the equilibrium Ag-Au-Cu phase diagram which indicates that the present diffusion layer has reached thermodynamic equilibrium.     

Adaptive parallel interference cancellation for CDMA systems—A weight selection and filtering scheme

Parallel interference cancellation (PIC) is a well-known multiuser detection algorithm in direct-sequence code-division multiple-access (DS-CDMA) systems. It is typically implemented with a multi-stage architecture. One problem associated with the PIC is that unreliable interference cancellation may occur in the early stages and the system performance may be degraded. Thus, the partial PIC detector was developed to control the cancellation level by use of interference cancellation factors. Partial PIC can be implemented with an adaptive form, in which optimal weights are derived using the least mean square (LMS) algorithm. In this paper, we propose an algorithm improving the conventional adaptive partial PIC. The main idea is to reduce the number of active weights in the LMS algorithm, and to perform weight post-filtering such that the resultant excess mean square error can be reduced. We also analyze the performance of the proposed algorithm and derive the bit error rate of the second stage output. Simulation results verify that the proposed algorithm outperforms the conventional partial PIC, and derived analytical results are accurate.     

Application of an auto-tuning neuron to sliding mode control

This paper presents a control strategy that incorporates an auto-tuning neuron into the sliding mode control (SMC) in order to eliminate the high control activity and chattering due to the SMC. The main difference between the auto-tuning neuron and the general one is that a modified hyperbolic tangent function with adjustable parameters is employed. In this proposed control structure, an auto-tuning neuron is then used as the neural controller without any connection weights.. The control law will be switched from the sliding control to the neural control, when the state trajectory of system enters in some boundary layer. In this way, the chattering phenomenon will not occur. The results of numerical simulations are provided to show the control performance of our proposed method.     

Investigation of a solid-state detector for advanced computed tomography

Utilization of solid-state detectors for computed tomography (CT) has been the focus of many studies. Previous phantom and clinical experiments have shown that one of the important performance parameters for the solid-state detector is the primary speed and afterglow. In this paper, we present a detailed investigation on the signal decay characteristics of the HiLight (GE Medical Systems, Milwaukee, WI) scintillating detector. The detector primary speed and afterglow are modeled by a multiexponential function and fully characterized by a set of time constants and relative strengths. The sensitivity of these parameters to X-ray photon energy, detector aging, and radiation exposure is then established and analyzed. No statistically significant variation is observed in these parameters due to changes in the above external variables. The impact of various decay time constants on CT image quality, such as spatial resolution, noise, and artifacts, is subsequently illustrated with computer simulations and phantom experiments. Finally, an algorithmic correction scheme is derived to compensate for detector afterglow. The correction scheme employs a recursive filter to remove adverse effects of the detector decay on image quality. Experimental results have shown that the correction scheme successfully restores system spatial resolution, produces a more homogeneous noise pattern, and eliminates ring-band image artifacts due to detector afterglow. The effectiveness and robustness of the correction scheme are demonstrated by extensive phantom and clinical experiments.     

Ball Impact Reliability of Zn-Sn High-Temperature Solder Joints Bonded with Different Substrates

In this study, the high-speed deformation behavior of solder joints formed with Pb-free Zn-Sn and commercial Pb-Sn alloys bonded on different substrates was investigated by the ball impact test method. Overall, Zn-Sn joints exhibited greater impact strength but inferior impact toughness than Pb-Sn joints. This can be ascribed to the high hardness of Zn-Sn solders resulting in partial or overall interfacial fracture. In contrast, the joints with soft Pb-Sn solders all showed a ductile fracture feature. It is suggested that, for the joints revealing brittle fracture, the impact toughness (impact energy) increased with the plastic ability of interfacial intermetallic compounds, while for those showing a ductile fracture mode, the impact energy deteriorated with a hardened solder matrix resulting from substrate dissolution.     

Wideband microwave filter constructed by asymmetrical compact microstrip resonator and floating plate coupling structure

The realisation of a microwave filter is presented by using the asymmetrical compact microstrip resonator and floating plate coupling structures. The filter demonstrates the properties of low insertion loss, wide pass bandwidth and two controllable transmission zeros, which results in fast roll-off responses near to the corner frequencies of the passband. It also has the advantages of compact size, ease of fabrication and absence of via holes. An X-band prototype with 34% fractional bandwidth was designed and fabricated to verify the proposed design concept. Good agreement between simulation and measurement was obtained.     

Analysis of passively Q-switched lasers with simultaneous modelocking

Simultaneous Q-switching and modelocking in a diode-pumped Nd:YAG/Cr⁴⁺:YAG laser is experimentally demonstrated. A general recurrence is derived for the analysis of the temporal shape of a single Q-switched envelope with mode locked pulse trains. With the developed model, the modelocked pulse energy and the total Q-switched pulse energy can be calculated. Excellent agreement was found between the present results and detailed theoretical computations     

Effect of fluorinated silicate glass passivation layer on electrical characteristics and dielectric reliabilities for the HfO2/SiON gate stacked nMOSFET

The superior characteristics of the fluorinated hafnium oxide/oxynitride (HfO₂/SiON) gate dielectric are investigated comprehensively. Fluorine is incorporated into the gate dielectric through fluorinated silicate glass (FSG) passivation layer to form fluorinated HfO₂/SiON dielectric. Fluorine incorporation has been proven to eliminate both bulk and interface trap densities due to Hf-F and Si-F bonds formation, which can strongly reduce trap generation as well as trap-assisted tunneling during subsequently constant voltage stress, and results in improved electrical characteristics and dielectric reliabilities. The results clearly indicate that the fluorinated HfO₂/SiON gate dielectric using FSG passivation layer becomes a feasible technology for future ultrathin gate dielectrics applications.