自动测试设备接口板环回线路性能快速预测方法
DOI:
CSTR:
作者:
作者单位:

1.南京邮电大学集成电路科学与工程学院(产教融合学院)南京210003;2.航天科工防御技术研究试验北京100854; 3.元器件应用验证技术创新中心北京100854;4.南京邮电大学南通研究院南通226001

作者简介:

通讯作者:

中图分类号:

TN41

基金项目:

国家自然科学基金企业创新发展联合基金重点支持项目(U23B2042)、国家自然科学基金面上项目(62371256)、中国航天科工防御技术研究试验中心项目(F00225MD2024001884)、集成电路与微系统全国重点实验室开放课题(NICL2026KF1001)、南京邮电大学南通研究院有限公司自研项目(ZY-001)资助


Rapid prediction method for loopback circuit performance of automatic test equipment interface boards
Author:
Affiliation:

1.School of Integrated Circuit Science and Engineering(Industry-Education Integration School), Nanjing University of Posts and Telecommunications, Nanjing 210003, China;2.Aerospace Science & Industry Defence Technology Research and Test Center, Beijing 100854,China;3.Innovation Center for Component Application Derification Technology, Beijing 100854,China; 4.Nantong Institute of Nanjing University of Posts and Telecomm-Unications,Nantong 226001, China

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    面向高速数字自动测试设备(automatic test equipment, ATE)接口板环回链路的信号完整性分析,目标是在保证物理一致性的前提下实现宽频散射参数的快速预测与端到端性能评估。针对接口板互连结构复杂、参数组合数量庞大的特点,将高速差分环回链路按物理结构划分为过渡、带状线一、交流耦合电容、带状线二和过渡5个模块,分别构建模块级深度学习代理模型,并通过散射参数级联重构端到端通道响应。模型训练过程中引入相位连续表示与角度一致性损失函数,以提升宽频相位预测的稳定性与一致性,使得在模块级预测中实现复数形式全局均方根误差为0.024 5。在此基础上完成模块级4端口散射参数的级联重构,端到端通道的复数形式预测的均方根误差为0.012 2。结合时域仿真结果对比可知,预测级联通道在时域反射计响应中主要反射位置及反射幅度变化趋势与参考级联结果保持一致;在非归零码条件下的眼图分析中,预测结果的眼宽与参考结果基本一致,眼高相对偏差控制在0.5%~5%范围内。结果表明,该模块化预测与散射参数级联方法能够较好保持链路关键不连续点的表征特性,可作为接口板设计早期的快速评估手段,用于缩短仿真迭代周期并辅助结构方案筛选。

    Abstract:

    Signal integrity analysis for loopback links on high-speed digital automated test equipment (ATE) interface boards aims to achieve rapid prediction of wideband scattering parameters and end-to-end performance evaluation while ensuring physical compliance. Addressing the complex interconnect structure and vast parameter combinations of interface boards, the high-speed differential loopback link is physically partitioned into five modules: transition, stripline 1, Alternating Current coupling capacitor, stripline 2, and transition. Module-level deep learning surrogate models are constructed for each, with end-to-end channel response reconstructed through cascaded scattering parameter reconstruction. During model training, phase continuity representation and angular consistency loss functions were incorporated to enhance the stability and consistency of wideband phase prediction, achieving a global complex-form root mean square error (RMSE) of 0.024 5 at the module level. Building upon this, cascaded reconstruction of module-level four-port scattering parameters was completed, yielding an end-to-end complex-form prediction RMSE of 0.012 2. Comparison with time-domain simulation results shows that the predicted cascaded channel maintains consistent trends in primary reflection locations and reflection amplitude variations with the reference cascaded results in the time-domain reflectometer response. In eye diagram analysis under non-return-to-zero conditions, the eye width of the predicted results aligns closely with the reference results, while the relative deviation in eye height is controlled within 0.5%~5%. The results demonstrate that modular prediction and scattering parameter cascading method effectively preserves the characterization properties of critical discontinuities in the link. It can serve as a rapid evaluation tool in the early stages of interface board design, helping shorten simulation iteration cycles and assisting in structural scheme selection.

    参考文献
    相似文献
    引证文献
引用本文

魏炜,江昊儒,傅成城,严大鹏,蔡志匡,刘轶.自动测试设备接口板环回线路性能快速预测方法[J].电子测量与仪器学报,2026,40(9):13-25

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2026-11-13
  • 出版日期:
文章二维码
×
《电子测量与仪器学报》
关于防范虚假编辑部邮件的郑重公告