摘要
焊料老化是绝缘栅双极型晶体管(IGBT)模块内部传热能力退化和结温估计偏离的主要诱因。利用壳温与焊料老化程度间的对应规律构建了两者的量化关系,提出了焊料老化状态监测方法。采用与功率损耗无关的参数对恶化Cauer热网络(CTN)有效传热面积进行表征,提出了焊料裂纹诱导的结温低估补偿机制;考虑温度相关的异质材料导热系数及比热容参量,抑制了温升引起的材料传热特性退化影响。在此基础上,通过对传统CTN模型的优化,克服了传热路径无法自适应配置问题。仿真结果表明,所提方法可有效减小传热退化对模型计算结果的影响,实现对IGBT模块热行为动态变化的精确模拟,且结温估计结果相较传统CTN模型的更为精确。
Solder aging is the main cause of heat transfer capability deterioration inside insulated gate bipolar transistor(IGBT)modules and estimation errors of the junction temperature.A monitoring method of the solder aging condition was proposed through establishing a quantitative relationship between the case temperature and solder aging degree.A compensation mechanism of the junction temperature underestimation caused by solder cracks was proposed by adopting a parameter independent of the power loss to characterize the effective heat transfer area of the deteriorated Cauer thermal network(CTN).The thermal conductivity and specific heat capacity dependent of the temperature of heterogeneous materials were taken into consideration to inhibit the influence of material heat transfer property degradation caused by temperature rise.Based on this,the difficulty in adaptive configuration of heat spreading path was vercome through optimizing the traditional CTN model.The simulation results show that the proposed method is able to reduce the influence of heat transfer degradatio on the model calculation results,and realize the accurate simulation of the dynamic changes of the thermal behavior of IGBT modules.Besides,estimation result of the junction temperature is more accurate than that of the traditional CTN model.
作者
崔昊杨
滕佳杰
范煜辉
韩韬
Cui Haoyang;Teng Jiajie;Fan Yuhui;Han Tao(College of Electronics and Information Engineering,Shanghai University of Electric Power,Shanghai 200090,China;State Grid Electric Power Research Institute,Nanjing 211106,China)
出处
《半导体技术》
CAS
北大核心
2021年第3期241-248,共8页
Semiconductor Technology
基金
国家自然科学基金资助项目(61107081)
上海市地方能力建设项目(15110500900)。
关键词
IGBT模块
焊料老化
状态监测
热网络模型
结温
IGBT module
solder aging
condition monitoring
thermal network model
junction temperature