由於目前高科技的快速進步，使得儀器結構設計越加複雜，需藉由多個零組件利用夾具來相互連接與合成。一般訊號在通過夾具時在不同頻率下會造成儀器訊號的衰減與失真。散射參數是用於評估儀器或傳輸線性能的指標，通常由網路分析儀量測。量測結果包含了訊號通過儀器本身與連接夾具的效應。如果要評估儀器的傳輸性能而不包含夾具效應，則需要使用校正技術將夾具效應從測量結果中移除。本論文將傳統兩埠夾具的表徵方法擴展到四埠，給予構件與實現一個新型四埠夾具表徵的方法。因網路分析儀量測的頻域資料S參數可以計算出反射儀TDR量測的時域資料，首先我們將2x-Thru結構的S參數轉換為TDR，其次透過時域門控方法分離出兩側夾具的部分S參數，然後使用梅森增益公式將剩餘S參數求出，最後使用去嵌入方法，將S參數轉換為T參數並透過矩陣運算結果來移除夾具效應，藉此得到待測物的S參數。所提出的方法將與著名大廠Keysight Technologies開發的Automatic Fixture Removal (AFR)演算法比較以驗證其效果。;Due to the rapid advancement in high technology, the design of instrument structures has become increasingly complex, requiring the use of fixtures to interconnect and assemble multiple components. When signals pass through these fixtures, they can experience attenuation and distortion at different frequencies, resulting in measurement inaccuracies. Scattering parameters are indicators used to evaluate the performance of instruments or transmission lines. They are typically measured using a network analyzer. The measurement results include the effects of both the instrument itself and the connecting fixtures. If one intends to assess the transmission performance of the instrument while excluding the influence of the fixtures, it becomes necessary to employ calibration techniques to remove the effects of the fixtures from the measurement results. This paper extends the conventional two-port fixture characterization method to a four-port configuration, presenting a novel approach for four-port fixture characterization. Since the frequency domain data, represented by S-parameters, measured by a network analyzer can be converted to time domain data obtained from reflectometer time-domain reflectometry (TDR) measurements, our method first converts the S-parameters of a 2x-Thru structure to TDR data. Then, using the time domain gating method, we separate the S-parameters corresponding to the two sides of the fixture. The remaining S-parameters are calculated using Mason’s gain formula. Finally, through the de-embedding process, the S-parameters are transformed into T-parameters, and the fixture effects are removed using matrix operations, yielding the S-parameters of the device under test (DUT). The proposed method will be compared with the Automatic Fixture Removal (AFR) algorithm developed by the renowned company Keysight Technologies to validate its effectiveness.
