[常见问题]Why S21 at dc can be smaller than 1, although electrically connected 2-port structure?原因

[sell=1]Why S21 at dc can be smaller than 1, although electrically connected 2-port structure? `Sx.|`x8  
(Last modified: 24 Apr. 2007) os_WYQ4>j  
If S21 at dc is smaller than 1, this might happen due to the following reasons: zn^v!:[  
`sxN!Jj?  
Reason 1: Port1 and Port2 have different line impedances Z1<>Z2, resulting in an amplitude of S11 at dc > 0 ( S11ampli=abs(Z1-Z2)/(Z1+Z2) ). Therefore also S21 at dc has to be physically smaller than 1. An identical case occurs, when Port1 and Port2 are explictily renormed to different line impedances. `WlH*p)z9  
s7M}NA 0  
Solution: When renorming both ports to the same impedance, S21 at dc should be 1. sDAK\#z  
N(BiOLZL6  
Reason 2: The open boundary condition (PML) acts at dc as a small resistor. ;q&0,B  
,-:a?#f>  
Solution: use magnetic boundaries, where metallic materials are touching the boundary. K @C4*?P  
3W'FcE)|E  
Reason 3: Mismatch of mode pattern at dc (only possible for inhomogenous port) :iEAUM  
?OO%5PSen  
Solution: change port evaluation frequency to 0 (or use full deembedding, which typically runs longer) .FJj  
U/5$%0)  
Reason 4: tan-delta fit of lossy dielectrics does not result in zero conductivity at dc (a leakage current flows through the substrate) raF] k0{  
@it/$>R^)  
Solution: try to modify the material model in a way, so that the curves eps'' and tan delta start at 0 for dc frequency. (robust check: simulate lossless dielectrics) TZBVU&,{Z  
.. QES^^PQe:  
7vq DZg  

未注册仅能浏览部分内容，查看全部内容及附件请先 登录 或 注册

了解一下楼主所说的问题

好像以前遇到过算出来的S21总是为0的情况