Yves Ousten

Detection and location of defects in electronic devices by means of scanning ultrasonic microscopy and the wavelet transform

In a highly competitive market, reliable techniques for manufacturing quality control of electronic devices are more and more demanded. In particular, scanning ultrasonic microscopy is nowadays showing itself a suitable, non-destructive tool for detecting and locating defects in a die-attach assembly of a wide range of components. Due to multiple reflections and scattering of the ultrasonic beam in multilayer die-attach assemblies, ultrasonic images can often appear very confused to be correctly interpreted, and the nature of defects very difficult to be pointed out. In the paper, the use of a digital signal-processing method, based on the Continuous Wavelet Transform, is suggested for automatically detecting and measuring the time-of-flight between ultrasonic echoes thus improving ultrasonic image understanding of complex structures. Methods performance is verified by means of numerical tests on several ultrasonic signals simulating different operating conditions; its reliability and efficacy come out from experiments on the die-attach assembly of actual electronic devices.

High temperature reliability testing of aluminum and tantalum electrolytic capacitors

Abstract Nowadays, many innovations in the automobile are enabled by electronics. Ambient requirements can be very stringent especially when the temperature reaches 150 °C or even more. Especially electrolytic capacitors are known to be critical devices at high temperatures. Therefore, it is necessary to validate the performance of such components and check their reliability during high temperature operation. In this paper we discuss how to predict the lifetime of both aluminum and tantalum electrolytic capacitors. In that aim we first review state of the art qualification tests that allow a life prediction. We describe a test setup that we have built in order to investigate electrolytic capacitors by LCR and leakage current measurements at temperatures above current manufacturers specifications. Results for different capacitors after variation of tests conditions will be presented.

Effects of silicone coating degradation on GaN MQW LEDs performances using physical and chemical analyses

This work presents a physics of failure (POF) methodology coupling failure signatures with physico-chemical analyses. The aim is to work out electro-optical failure signatures located in packaged InGaN/GaN Multiple Quantum Wells Light Emitting Diodes (MQW LEDs). Electrical and optical characteristics performed after accelerated ageing tests (30 mA/85 °C/1500 h), confirm a 65% drop of optical power and an increase of one decade of leakage current spreading at the silicone oil/chip interfaces. Through measurements of silicone coating fluorescence emission spectra, we demonstrate that the polymer enlarges the LED emission spectrum and shifts central wavelength. This shift is related to silicone oil spectral instability and the central wavelength of packaged LED appears to be temperature insensitive. In this paper, we discriminate the degradation of bulk silicone oil responsible for optical losses from the polymer/chip interface inducing larger leakage current.

Ultrasonic images interpretation improvement for microassembling technologies characterisation

Abstract This paper deals with two techniques for improvement of ultrasonic images interpretation based on image and digital signal processing. The first technique concerns image processing. We propose an analytical model to evaluate the effective Point Spread Function (PSF) of a reflection scanning ultrasonic system. We present the great interest of the knowledge of the PSF for deconvolution of ultrasonic images based on Lucy-Richardson algorithm. With this technique, lateral resolution is strongly improved and real dimensions of an object can be reached that is necessary to improve the understanding and interpretation of an image. The second technique concerns ultrasonic images showing different nature of defects. A better understanding of these images can be realized by means of time of flight detection. This technique concerns Time-Frequency (TF) analysis to determine nature and depth of defects with accuracy. Two applications of microassembling die-attach characterisation are presented using these methods.

Improved performances of polymer-based dielectric by using inorganic/organic core-shell nanoparticles

BaTiO3/hyperbranched polyester/methacrylate core-shell nanoparticles were studied by varying the shell thickness and the methacrylate ratio. We demonstrated that coalescence typically observed in traditional composites employing polymer matrices is significantly reduced. By modifying the shell thickness, the equivalent filler fraction was tuned from 7 wt. % to 41 wt. %. Obtained permittivities were compared with reported models for two-phase mixtures. The nonlinear behavior of the dielectric constant as a function of the equivalent filler fraction has been fitted with the Bruggeman equation. Methacrylate groups reduce by a decade the loss factor by improving nanoparticles adhesion. The permittivity reaching 85 at 1 kHz makes core-shell nanoparticles a promising material for embedded capacitors.

Vibration lifetime modelling of PCB assemblies using steinberg model

Abstract Automotive requirements are nowadays not only limited to high temperatures but more and more demands are driven towards vibration and higher acceleration values due to the direct mounting of the Electronic Control Units on the Engine. Temperature cycle testing and combined testing methods have been under study for a long time but literature concerning Vibration Modelling is very poor in comparison with thermal models like Arrhenius or thermo-mechanical based models like Coffin-Manson. In this article, a revisit of the Steinberg Model is presented with a direct application on tantalum capacitors populated boards. Experimental results with various sinusoidal excitation g-level are presented; whereas on the other side FEM simulations are performed and results are implemented in the Steinberg Model, in order to identify model parameters.

An improved method for automatic detection and location of defects in electronic components using scanning ultrasonic microscopy

The paper deals with a method for the automatic analysis and characterization of defects due to encapsulation or/and surface mount processes of microelectronic devices. This method is based on digital signal processing of ultrasonic signals in the 10-100 MHz frequency range and in particular used for automatic evaluation of time-of-flight between echoes received by the acoustic transducer. The signals are firstly preprocessed by a new algorithm, based on the Wiener filtering, and then by a numeric algorithm, based on the wavelet transform, already applied successfully to this problem. The preprocessing phase increases the sensitivity of successive numeric algorithm. The theory underlying the preprocessor and the chosen procedure to implement it are described in detail. Furthermore, experimental results obtained applying the proposed method on acoustic signals from an electronic structure acquired through an ultrasonic scanning system are given and discussed.

Localization of defects in die-attach assembly by continuous wavelet transform using scanning acoustic microscopy

Abstract The main goal of this paper is to describe a new method based on Continuous Wavelet Transform (CWT) algorithm and successfully implemented in a Scanning Acoustic Microscope, developed at IXL Laboratory, for the measurement of Time Of Flight (TOF) between ultrasonic echoes and contribute to the help of defect diagnosis and failure analysis of a die-attach assembly. The paper is divided into three main parts. After a brief description of the different methods for the TOF measurement, we give the strong interest to use Wavelet Transform due to the nature of ultrasonic signals. The principle of the CWT algorithm and the TOF measurement method associated are explained. The robustness of the CWT algorithm is evaluated to achieve information concerning its performances in presence of different signal parameters and SNR with a statistical treatment. 1n final, we propose an application of this algorithm to the help of acoustic images interpretation. This application concerns the detection and localization of defects into a die-attach assembly, specially the detection of fine crack.

Power MOSFET active power cycling for medical system reliability assessment

Abstract High voltage generator is one of the critical subsystems of a medical system. So, a methodology assessing its reliability from component to system is proposed. In this context, N-channel power MOSFETs that are used in high frequency inverter for supplying high voltage transformer is tested. These MOSFETs are subjected to power cycling test in order to reveal thermo-mechanical failure modes and at the same time evaluate their lifetime using DoE. The test is done by using an external current supply to induce self-heating through the component. Heating and cooling phase are controlled by a logic controller. The ON state resistance (RDSON) is monitored during the test as the failure indicator and different imaging methods are used for failure analysis and observed failures mechanism are discussed.

Failure mechanisms and qualification testing of passive components

Abstract New electronic architectures and mechatronic integration in automotive and oil-field applications lead to increasing requirements concerning operating temperatures and vibration levels. At the same time, reliability and lifetime have to fulfil strong demands. In the European funded project PROCURE (Program for the development of passive devices used in rough environments) a generic spectrum of passive components needed for electronic control units has been developed. The failure mechanisms, the technological challenges, and the test requirements are highlighted below.