We have examined the response of AlGaN/GaN power switching HEMTs to electrical bias stress. Three different gate stack structures were studied. In devices containing a ∼ 5 nm thick AlGaN layer in the ...gate stack, both positive and negative shifts in the threshold voltage were observed following high blocking voltage stress, consistent with a short initial period of electron trapping followed by a longer period of de-trapping. Correlated changes in reverse bias leakage current were also observed, although this also occurred in devices containing only residual AlGaN in the gate stack. The data have been explained by a field-enhanced emission model in which an electron trapping to de-trapping transition occurs. The exact nature of the transition is found to be sensitive to a variety of parameters including trap energy, geometry, and initial and boundary conditions.
The inverter is still considered the weakest link in modern photovoltaic systems. Inverter failure can be classified into three major categories: manufacturing and quality control problems, ...inadequate design, and electrical component failure. It is often difficult to deconvolve the latter two of these, as electrical components can fail due to inadequate design or as a result of intrinsic defects. The aim of the current work is to utilize the extensive background in both inverter performance testing and component reliability found at Sandia National Laboratories to assess the role of component failures in PV performance and reliability. Although there is no consensus on the least reliable component in a modern inverter system, the IGBT is often blamed for failures and hence this was the first component we studied. A commercially available 600V, 60A, silicon IGBT found in common residential inverters was evaluated under normal and extreme operating conditions with DC and pulsed biasing schemes. Although most of the sample devices were robust even under extreme conditions, a few of the samples failed during operation well within the manufacturer-specified limits. Additionally, we have begun in situ monitoring of IGBTs as well as other components within an operating 700 W, single-phase inverter. The in situ testing will guide future device-level work since it allows us to understand the conditions that are experienced by inverter components in a realistic operating environment.
A dedicated design qualification standard for PV inverters does not exist. Development of a well-accepted design qualification standard, specifically for PV inverters will significantly improve the ...reliability and performance of inverters. The existing standards for PV inverters such as ANSI/UL 1741 and IEC 62109-1 primarily focus on safety of PV inverters. The IEC 62093 discusses inverter qualification but it includes all the BOS components. There are other general standards for distributed generators including the IEEE 1547 series of standards which cover major concerns like utility integration but they are not dedicated to PV inverters and are not written from a design qualification point of view. In this paper some of the potential requirements for a design qualification standard for PV inverters are addressed. The missing links in existing PV inverter related standards are identified and with the IEC 62093 as a guideline, the possible inclusions in the framework for a dedicated design qualification standard of PV inverter are discussed. Some of the key missing links are related to electric stress tests. Hence, a method to adapt the existing surge withstand test standards for use in design qualification standard of PV inverter is presented.
This volume contains thirty-five papers from a 2010 conference on landscape archaeology focusing on the definition of landscape as used by processual archaeologists, earth scientists, and most ...historical geographers, in contrast to the definition favored by postprocessual archaeologists, cultural geographers, and anthropologists. This tension provides a rich foundation for discussion, and the papers in this collection cover a variety of topics including: how do landscapes change; how to improve temporal, chronological, and transformational frameworks; how to link lowlands with mountainous areas; applications of scale; new directions in digital prospection and modeling techniques; and the future of landscape archaeology.
Specialized accelerators have recently garnered attention as a method to reduce the power consumption of neural network inference. A promising category of accelerators utilizes nonvolatile memory ...arrays to both store weights and perform in situ analog computation inside the array. While prior work has explored the design space of analog accelerators to optimize performance and energy efficiency, there is seldom a rigorous evaluation of the accuracy of these accelerators. This work shows how architectural design decisions, particularly in mapping neural network parameters to analog memory cells, influence inference accuracy. When evaluated using ResNet50 on ImageNet, the resilience of the system to analog non-idealities-cell programming errors, analog-to-digital converter resolution, and array parasitic resistances-all improve when analog quantities in the hardware are made proportional to the numerical values in the network. Moreover, contrary to the assumptions of prior work, nearly equivalent resilience to cell imprecision can be achieved by fully storing weights as analog quantities, rather than spreading weight bits across multiple devices, often referred to as bit slicing. By exploiting proportionality, analog system designers have the freedom to match the precision of the hardware to the needs of the algorithm, rather than attempting to guarantee the same level of precision in the intermediate results as an equivalent digital accelerator. This ultimately results in an analog accelerator that is more accurate, more robust to analog errors, and more energy-efficient.
PDF
