Failure mechanism study of anisotropic conductive film (ACF) packages

Anisotropic conductive film (ACF) consists of an adhesive polymer matrix with dispersed conductive particles. In flip-chip technology, ACF has been used in place of solder and underfill for chip attachment to glass or organic substrates. The filler particles establish the electrical contacts between the interconnecting areas. ACF flip-chip bonding provides finer pitch, higher package density, reduced package size and improved lead-free compatibility. Nevertheless, the interconnection is different from traditional solder joints, the integrity and durability of the ACF interconnects have major concerns. Failures in anisotropic conductive film (ACF) parts have been reported after temperature cycling, moisture preconditioning and autoclave. The failures have not been well understood and have been attributed to a wide variety of causes. This paper investigates the failure mechanism of ACF using finite element simulation. From a failure-initiation point of view, the response of ACF packages to environmental (temperature and humidity) exposure is very different from standard underfilled packages. These differences cause the ACF package to fail in different ways from an underfilled package. Simulation results have shown that moisture-induced ACF swelling and delamination is the major cause of ACF failure. With moisture absorption, the loading condition at the interface is tensile-dominant, which corresponds to lower interface toughness (or fracture resistance). This condition is more prone to interface delamination. Therefore, the reliability of ACF packages is highly dependent on the ACF materials. The paper suggests a new approach toward material selection for reliable ACF packages. This approach has very good correlation with experimental results and reliability testing of various ACF materials.     

Power adaptation for BPSK signaling with average and peak power constraints in Rayleigh fading channels

We consider power adaptation strategies for binary phase-shift keying signals in Rayleigh fading channels under the assumption that channel state information is provided at both the transmitter and the receiver. We first derive a closed-form expression for the optimal power adaptation that minimizes average bit-error rate (BER) subject to average and peak transmission power constraints. Then, we analyze the average BER for channel inversion power adaptation with the same constraints. Our results show that the performance difference between the optimal power adaptation and the channel inversion becomes negligibly small as available average transmission power increases and/or peak-to-average power ratio decreases. We also find that an optimal peak-to-average power ratio exists that minimizes the average BER in the channel inversion scheme.     

Narrow-band interference rejection in DS/CDMA systems using adaptive (QRD-LSL)-based nonlinear ACM interpolators

An Mth order adaptive lattice filter automatically generates all M of the outputs that would be provided by M separate transversal filters. This feature may effectively suppress narrow-band interference (NBI) of either unknown or time-varying bandwidth (or number of frequency bands) in direct-sequence code-division multiple access systems for which the order of the interference rejection filter that achieves the optimal performance is unknown or constantly changing. Moreover, a lattice filter may significantly outperform its transversal counterpart in complex jamming environments in which the adaptive lattice filter must suppress multiple jammers, since each stage of a lattice filter adapts to suppress an orthogonal component of the NBI. The paper develops a computationally efficient and numerically stable adaptive QR-decomposition-based least squares lattice (QRD-LSL)-based nonlinear approximate conditional mean interpolator to suppress NBI effectively. Simulation results demonstrate that both the signal-to-noise ratio improvement and the convergence rate achieved by the proposed interpolators outperform those of other existing prediction-based techniques.     

Multilevel voltage wave-shaping display driver for AC plasma display panel application

A new sustainer circuit with multilevel voltage wave-shaping characteristics for an AC plasma display panel (AC-PDP) drive is proposed. The proposed circuit features half the device voltage stresses and significantly reduced power losses compared with those of the conventional ones. This circuit, realizable without much increased cost of the semiconductor devices, gives a significant improvement in the power efficiency, essential for the design of a drive circuit for the AC-PDP. A comparative analysis and experimental results are presented to show the validity of the proposed sustainer circuit.     

A shifted gamma distribution model for long-range dependent Internet traffic

It is important to characterize the distributional property and the correlation structure of traffic arrival processes in modeling internet traffic. The conventional fractional Gaussian noise (fGn) model fails in characterizing the distributional property when the distribution of the input traffic rates is nongaussian. We propose a shifted gamma distribution model which can solve this problem. A linear-time generation algorithm is also given.     

A low-noise L-band EDFA with a 1500-nm Raman-pumped dispersion-compensating fiber section

This report presents a low-noise L-band erbium-doped fiber amplifier (EDFA) with a dispersion-compensating Raman amplifier. With an optimized prestage and 1500-nm Raman-pump laser diodes, the proposed EDFA achieved an internal noise figure of less than 4.5 dB over a 33-nm flat gain bandwidth within 0.5 dB at -2 dBm of large signal input power.     

A 1.8-V 128-Mb mobile DRAM with double boosting pump, hybrid current sense amplifier, and dual-referenced adjustment scheme for temperature sensor

To verify three important circuit schemes suitable for DRAMs in mobile applications, a 1.8-V 128-Mb SDRAM was implemented with a 0.15-/spl mu/m technology. To achieve an ideal 33% efficiency, the double boosting pump uses two capacitor's series connection at pumping phase, while they are precharged in parallel. The hybrid folded current sense amplifier together with a novel replica inverter connection improved power and speed performances. Also, a dual-referenced adjustment scheme for a temperature sensor was proposed to allow a very high accuracy in tuning. Without loss in productivity, the implemented dual-referenced searching technique achieved tuning error of less than /spl plusmn/2.5/spl deg/C.     

Magnetic properties of Co- and Mn-implanted BaTiO3, SrTiO3 and KTaO3

Implantation of Co or Mn into single-crystal BaTiO₃(K), SrTiO₃ or KTaO₃(Ca), followed by annealing at 700 °C, produced ferromagnetic behavior over a broad range of transition metal concentrations. For BaTiO₃, both Co and Mn implantation produced magnetic ordering temperatures near 300 K with coercivities 70 Oe. The M–T plots showed either a near-linear decrease of magnetization with increasing temperature for Co and a non-Brillouin shaped curve for Mn. No secondary phases were detected by high-resolution X-ray diffraction. The same basic trends were observed for both SrTiO₃ and KTaO₃, with the exception that at high Mn concentrations (5 at.%) the SrTiO₃ was no longer ferromagnetic. Our results are consistent with recent reports of room temperature ferromagnetism in other perovskite systems (e.g. LaBaMnO₃) and theoretical predictions for transition metal doping of BaTiO₃ [Nakayama et al., Jap. J. Appl. Phys. 40 (2001) L1355].     

Design of multi-finger HBTs with a thermal–electrical model

Multi-finger heterojunction bipolar transistors (HBTs) with uniform spacing exhibit a higher temperature at the center of devices. The temperature distribution on the emitter fingers of the HBT is studied with a three-dimensional thermal–electrical model. Using this model, multi-finger HBTs are designed with non-uniform spacing to improve temperature distribution. Depending on the number of emitter fingers, different design approaches are demonstrated. For a six-finger or 12-finger HBT, the design is more straightforward. For a complex structure such as a 26-finger HBT, an efficient design procedure is necessary. In all of these cases, the calculated results show significant temperature reduction on non-uniform spacing devices.     

Unsupervised multistage image classification using hierarchical clustering with a Bayesian similarity measure.

A new multistage method using hierarchical clustering for unsupervised image classification is presented. In the first phase, the multistage method performs segmentation using a hierarchical clustering procedure which confines merging to spatially adjacent clusters and generates an image partition such that no union of any neighboring segments has homogeneous intensity values. In the second phase, the segments resulting from the first stage are classified into a small number of distinct states by a sequential merging operation. The region-merging procedure in the first phase makes use of spatial contextual information by characterizing the geophysical connectedness of a digital image structure with a Markov random field, while the second phase employs a context-free similarity measure in the clustering process. The segmentation procedure of region merging is implemented as a hierarchical clustering algorithm whereby a multiwindow approach using a pyramid-like structure is employed to increase computational efficiency while maintaining spatial connectivity in merging. From experiments with both simulated and remotely sensed data, the proposed method was determined to be quite effective for unsupervised analysis. In particular, the region-merging approach based on spatial contextual information was shown to provide more accurate classification of images with smooth spatial patterns.