比较经典:关于driven modal 与driven terminal 的理解,以及两种波端口的区别！

本文转自百思论坛,由sonofsky原创(http://www.baisi.net/thread-53295-1-1.html).
再次转载请说明原出处,在此感谢sonofsky把经验分享给大家！
关于driven modal 与driven terminal 的理解
1. driven modal 模式驱动, 所谓模式驱动就是hfss根据用户所定义的模式数目求解端口模式数目及场分布,并为每个模式分配相等的功率,仿真时用端口场分布做为边界条件对内部进行求解,默认端口阻抗为Zpi 无须定义积分线来求解电压, S参量用入射反射功率来表示
2. 对于分析偶合传输线等一个端口上有多个终端,而求解终端之间偶合问题的模型,driven modal 是不适合的.应用driven terminal ,这里以微带偶合传输线为例子说明这个问题

在这个端口上tem波 有两种模式 1.偶模:V1=V2 2.奇模. V1=-V2 (V1为导体1对接地板等效电压, V2为导体2对接地板等效电压) 如果用driven modal求结 则这两种模式分别被赋予相等功率,而求解出的S11则是整个端口上的每一种模式的反射情况,而不能直接求出两线的偶合状况(例如只激励导体1,求导体2上的端口电压)这显然是不合适的.
(关于偶合传输线问题详情见microwave engineering edition 3 7.6节)
Driven terminal默认的求解终端阻抗为Zvi 故对于每个终端需要定义积分线,例如上图中terminal 的积分线为从接地版到导体1的连线(导体1,接地版都为等势体,路径没有关系),terminal2的积分线为接地版到导体2) 计算机求解时对两个终端分别进行激励,通过电压与电流来计算他们之间的偶合关系.
3总结
1. 如果模型中有类似于偶合传输线求偶合问题的模型一定要用driven terminal求解,
2. driven modal适于其他模型, 但一般tem模式(同轴,微带等)传输的单终端模型一般用driven terminal分析 (tem波电压一般由两导体之间电场积分定义,电流为环线磁场的积分,阻抗Zvi=Zpi=Zpv区别于TE TM) 由于其直接对电流电压求解而避免了对整个面上功率的计算从而比较简便.

HFSS---the difference between wave port and lump port
With lump port=> the excitation is applied at a point/cell, as a voltage or current.
With wave-port=> the excitation is so-called eigen-wave, such as the quasi-tem wave supported by a microstrip line. It applies over a cross-sectional area.
voltage is scalar, wave is vector by nature, hence there are substantial difference between the two. So use waveport whenever possible, because "simulation of wave phenomenon" is what HFSS is designed for. And compare with the "correct" measurement whenever possible (i.e. measure "wave", not simply "voltage").
Why lumped port is there? It is easy to applied and people found that good/reasonable results can be obtained. Why? if the frequency is low enough or the excitation is applied at sufficiently small area, then the "wave" can be described by some "voltage" or "current", which must be "measured"/"calculated"/de-embedded/etc in the correct manner.
if the excitation can be applied on some locally uniform region=>waveport,
if geometry/material discontinuities are near/closer to the point of excitation=>lumped port might be the only way.

The subject of lumped vs wave port is actually complicated. people continue to study related matters, (e.g. arguing what is the "characteristic impedance").
Reading about the various de-embedding scheme should help with the understanding..

个人觉得其在总结上面第２点中，应该是同轴线和偶合线用driven　terminal．可能是作者笔误！

感谢分享

不错,顶顶

真的很经典，真的很经典
。

:29bb 谢谢指点！

好！

多谢老毕啊：17de

原来是。。
那我得顶一个

正在迷惑,谢谢指点

一直不是 很明白

一直不是 很明白

经典，受益匪浅！