CST仿真T型不等功率分配器有問

來源：edatop 更新時間：2024-07-09 閱讀：

我在使用CST仿真2.45GHz 的T型不等功分器時，出現問題，現將問題描述如下：
首先仿真功率比為1.55:1的T型功分器[attachment=60626]
之所以采取這樣的結構設計，是因為功率比很大的情況下，所需的微帶的電阻很大，所以微帶線很細，不容易設計，所以采取了論文中的改變參考臂的設計思想，所以根據功率比知道，只要Z4=k^2*Z3（K^2=1.55），且1/Z4+1/Z3=1/Z2即可，所以選擇合適的Z3，Z4和Z2，則能夠既能夠保證功率比，又能夠保證實際的可實現性，最終選擇Z2=36.5，Z3=60，Z4=93（滿足Z4=k^2*Z3（K^2=1.55），且1/Z4+1/Z3=1/Z2），Z1，Z5，Z6是起著阻抗匹配的作用（長度均為1/4波長）。
最終結果如圖：
[attachment=60628]
實際功率比1.53:1，馬馬虎虎能夠滿足，接下來，我希望將其中的一個輸出端口再進行等功分，即整個模型有三個輸出端口：
[attachment=60629]
等功分的那個端口的設計思想基本上和之前的一樣，將原來的50Ω直接分成兩個100Ω，100歐姆通過70.7的電阻和輸出的50歐姆進行匹配，沒有等功分的那個端口把原來的50歐姆直接延長，模型如上所示。
[attachment=60630]
并且能量曲線沒有下降到-30dB
[attachment=60631]
仿真有警告，并且S參數圖形很差，警告如下：
Maximum simulation time reached, solver stopped.
Please note that the steady state energy criterion has not been satisfied. More...
1 warning occurred.
通過點擊More，出現：

This page contains a list of the most important warning and error messages in the context of the transient solver together with a detailed explanation of the meaning and proposal for handling and resolution.
Maximum number of pulse widths simulated, solver stopped. / Maximum simulation time reached, solver stopped. Please note that the steady state energy criterion has not been satisfied.
The transient solver operates with a time pulse as excitation, and the simulation stops when one of the following conditions are met:
•
Either the remaining energy in the calculation domain as well as the port signals have decreased to the steady state value specified on the time domain solver parameter page in the Accuracy field.
•
The simulation time reaches the maximum solver duration, either defined as Time, Number of pulses or Propagation distance on the Steady state tab on the Special Time Domain Solver Parameter page.

The warning appears when the latter condition was satisfied first before the accuracy level has been reached. This means the energy level remaining in the structure is still bigger than the specified steady state value. Usually this is the case for resonating (energy stays inside the structure for a long time) or large structures (it takes a long time for a pulse to penetrate through the structure). The extra energy would appear as a truncation error, which causes some ripples in the S-parameter curves.
Recommended solutions are the following:
•
For non-resonating structures, the maximum solver duration setting can be adjusted, with the appropriate choice of defining number of pulses or a time or propagation distance setting.
•
For resonating structures a much better approach would be to activate the Autoregressive Filter (see advanced Topics manual) on the AR Filter tab on the Special Time Domain Solver Parameter page, which allows to shorten the simulation time and in addition to avoid the truncation error.
不知道出現這種情況該怎么解決，因為理論上分析的應該沒錯，實際不知道是模型有問題還是軟件設置有問題，求教，