← 返回
储能系统技术 储能系统 有限元仿真 可靠性分析 ★ 4.0

功率半导体器件芯片表面横向温度梯度估计方法

An Estimation Method for Lateral Temperature Gradient on Chip Surface of Power Semiconductor Devices

作者 Maoyang Pan · Erping Deng · Xia Wang · Yushan Zhao · Hongyu Sun · Yuxing Yan
期刊 IEEE Journal of Emerging and Selected Topics in Power Electronics
出版日期 2025年4月
技术分类 储能系统技术
技术标签 储能系统 有限元仿真 可靠性分析
相关度评分 ★★★★ 4.0 / 5.0
关键词 横向温度梯度 功率半导体器件 估算方法 实验验证 可靠性
语言:

中文摘要

如今芯片表面严重的横向温度梯度可显著降低功率半导体器件寿命。目前提取芯片表面横向温度梯度的主要方法包括红外IR相机和有限元仿真。然而前者需要破坏封装,后者可能耗时费力且精度存在挑战。本文推导了横向温度梯度的表达式,研究了影响因素及其含义。以常见三类线键合功率半导体器件为例,提出一种估计横向温度梯度的方法。实验验证了该方法的有效性和准确性,分立器件误差在13%以内,模块误差在1%以内。尽管验证需要破坏性封装进行IR测量,但后续应用可通过主要取决于器件类型的预校准有效性因子实现非破坏性估计。实际应用中该方法有助于在不开封情况下有效预防或缓解与横向温度梯度相关的失效风险,理论上有助于推进失效机制分析并帮助制造商改进器件可靠性。

English Abstract

Nowadays, serious lateral temperature gradient across the chip surface can significantly reduce the lifetime of the power semiconductor devices. Currently, the primary methods for extracting lateral temperature gradient on chip surface include infrared (IR) camera and finite element simulation. However, the former requires destroying the package, while the latter can be time-consuming and labor-intensive, with challenges in accuracy. In this article, the expression of lateral temperature gradient is derived. Then, influencing factors and their implications are also investigated. Finally, taking the common three types of wire-bonded power semiconductor devices for example, a method for estimating lateral temperature gradient is proposed. The effectiveness and accuracy of the method are validated by the experiment, showing agreement within 13% for discrete devices and 1% for modules. Despite validation requiring destructive package for IR measurements, subsequent applications enable nondestructive estimation through precalibrated effectiveness factors that mainly depend on the device type. In practical applications, this method helps facilitate the effective prevention or mitigation of failure risks associated with lateral temperature gradient without decapsulation. Theoretically, it contributes to advancing failure mechanism analysis and helps manufacturers improve the device reliability.
S

SunView 深度解读

该功率器件横向温度梯度估计研究对阳光电源功率模块可靠性设计有重要参考价值。非破坏性估计方法通过预校准有效性因子避免开封测试可应用于阳光ST系列储能变流器和SG系列光伏逆变器的SiC/GaN功率模块在线健康监测。1%模块误差和13%分立器件误差的高精度估计为阳光iSolarCloud平台的数字孪生热管理系统提供了关键输入数据。该方法对横向温度梯度影响因素的深入分析可指导阳光电源优化功率模块的线键合设计、降低热应力集中并延长器件寿命,对阳光PowerTitan储能系统的长期可靠性和全生命周期成本控制有重要意义。