Quantifying Device Degradation in Live Power Converters Using SSTDR Assisted Impedance Matrix Most Sultana Nasrin, student member, IEEE, Faisal H. Khan, senior member, IEEE and Mohammed Khorshed Alam, student member, IEEE sultana.nasrin@utah.edu, faisal.khan@utah.edu, khorshed.alam@utah.edu Abstract  A non-interfering measurement technique designed around spread spectrum time domain reflectometry (SSTDR) has been proposed in this paper to identify the level of aging associated to power semiconductor switches inside a live converter circuit. Power MOSFETs are one of the most age sensitive components in power converter circuits, and this paper demonstrates how SSTDR can be used to determine the characteristic degradation of the switching MOSFETs used in various power converters. SSTDR technique was applied to determine the aging in power MOSFETs while they remained energized in live circuits. In addition, SSTDR was applied to various test nodes of an H-bridge ac-ac converter, and multiple impedance matrices were created based on the measured reflections. An error minimization technique has been developed to locate and determine the origin and amount of aging in this circuit, and this technique provides key information about the level of aging associated to the components of interest. By conducting component level failure analysis, the overall reliability of an H-bridge ac-ac converter has been derived and incorporated in this paper. Keywords: Reflectometry, SSTDR, aging, converter, matrix, reliability, converter, degradation, MTTF, failure rate. I. INTRODUCTION Power converter circuits have a wide range of applications, and ceaseless operation of these converters is imperative in most cases. However, these converters may be exposed to stressful environments, i.e. over voltage, over current, high temperature, or switching impulses during regular operations and these external forces may result in severe degradation or complete failure of critical components inside a converter. Over longer periods, these conditions coupled with various environmental factors such as mechanical vibration, radiation and thermal cycling may lead to catastrophic failure of converter components. A failure survey of various components in a power converter shown in Figure 1 [1] clearly demonstrates that electrolytic capacitors and MOSFETs are the most aging-affected components in power converter circuits. Photographs of some actual damaged MOSFET and electrolytic capacitor are given in Figure 2. Component parameters such as ON-resistance and switching characteristics of MOSFETs/IGBTs, and equivalent series resistance (ESR) of capacitors degrade with time, and the accumulated aging eventually leads to deterioration of several operating characteristics of a power converter such as output voltage ripple, switching loss, conduction loss etc. [2], [3], [4]. Therefore, the aging profile i.e. the state of health of the entire power converter could be obtained by studying the component level aging although the level of aging associated to any component is difficult to calculate because multiple components are interconnected in a converter circuit. Therefore, two hurdles need to be crossed – the component level aging needs to be identified with reasonable precision; and a converter specific model needs to be derived that could be used to predict the overall reliability of the converter based on component level aging. This situation is even more challenging when components need to be characterized while they are energized in a live converter, and conventional measurement techniques could not be used with ease. Moreover, it is of paramount importance to identify which component parameters to look for. In this regard spread spectrum time domain reflectometry (SSTDR) can be used to measure impedance discontinuity in various current paths in a live converter circuit. As an added feature, several faults inside the circuit can be identified without interrupting the circuit’s normal operation although the discussion on this capability of SSTDR (fault detection) is beyond the scope of this paper. Authors’ previous work demonstrated how SSTDR can be applied to identify degradation in MOSFETs, IGBTs, and electrolytic capacitors [5]. Once the component level analysis was done, authors are particularly interested in the reliability estimation of a converter circuit using the SSTDR generated data. This paper demonstrates that major parameters of power MOSFETs exhibit measurable changes over time in a real converter circuit, and SSTDR can be effectively used to measure these changes. These SSTDR generated data can be Figure 1: A failure survey of different components responsible for converter failure [1]. This is the author's version of an article that has been published in this journal. Changes were made to this version by the publisher prior to publication. The final version of record is available at http://dx.doi.org/10.1109/TPEL.2013.2273556 Copyright (c) 2013 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing pubs-permissions@ieee.org.