本論文為了解決微波探針量測成本過高及其量測接腳脆弱，所以開發量測成本低、可重複使用以及有堅固量測接腳之探針。探針是利用市面上常見之SMA接頭母-母(R)型改良而成，分別開發兩種版本，二者主要差異為兩邊接腳與中心針的間距可加工調整與否，自製探針1可以藉由加工技術調整間距大小。接著使用微波探針及兩種自製探針進行量測比較，並於量測前利用SOLT校正方法移除探針之電氣效應，最後以CST Microwave Studio電磁分析軟體驗證實際量測結果。兩種自製探針量測結果差異不大，二者介入損失於頻率10 GHz相差0.11 dB，自製探針2略優於自製探針1。量測頻率0~10 GHz中，反射損失皆位於-20 dB以下。與微波探針相比，兩種自製探針與微波探針在介入損失及反射損失量測結果之損失值差異不大。我們進一步透過電磁分析軟體模擬不同的基於SMA的探針金屬外殼、中心針結構與鐵氟龍構造之影響，發現銅金屬不能設計於電磁波傳遞路徑上，以免阻擋電磁波的傳遞。中心針切半對電磁波傳遞影響不大，切半的中心針結構反而有利於實際量測的進行。不當地增加或減少鐵氟龍也會大幅影響電磁波的傳遞，造成能量的損失。

This paper not only focus on the high measurement cost and weak measurement pins of microwave probe, but also focus on the signal integration problem of SMA at solder place. Therefore, we develop the two different types of probe that contain lower measurement cost, reusable and strong measurement pins. Both of them made from common market’s SMA female-female(R) type connector. The main difference of them is the gaps that between middle signal pin and two ground pins. The improve probe1 can adjust the gaps by manufacturing technique. Furthermore, we took SOLT calibration method to eliminate the probe’s electrical effect before measurement, then we compared the measurement results of microwave probe and two improve probes. Finally, we used the CST Microwave Studio electromagnetic analysis software to test and verify the measurement results. As the S parameters results show that there was a difference of insertion loss about 0.11 dB or so at 10 GHz frequency, and under -20 dB of return loss during measuring frequency 0~10 GHz. So the measurement ability of improve probe2 is a little bit better than improve probe1. In consequence, there were similar S parameters results between microwave probe and two improve probes.
This paper also discussed that structure of SMA about different copper shell, middle signal pin and Teflon effect. In order to avoid that obstruct the energy pass, we can’t design copper shell on a path of electromagnetic waves transmission. The half middle signal pin is an easy way for measurement process, and it doesn’t quite impact the measurement results. Modify the Teflon material structure will cause the significant energy loss.

[1] T. Jamneala, et al., "Employing a ground model to accurately characterize electronic devices measured with GSG probes," Microwave Theory and Techniques, IEEE Transactions on, vol. 52, pp. 640-645, 2004.
[2] J. Palecek, et al., "Examination of SMA connector parameters," in Intelligent Engineering Systems (INES), 2012 IEEE 16th International Conference on, 2012, pp. 259-263.
[3] Agilent, "Agilent Network Analysis Applying the 8510 TRL Calibration for Non-Coaxial Measurements," Agilent Technologies Product Note vol. Product Note 8510-8A.
[4] 陳昌昇, "射頻元件電氣特性量測及網路分析儀校正技術簡介."
[5] M. Magerl, et al., "Broadband characterization of SMA connectors by measurements," in Information and Communication Technology, Electronics and Microelectronics (MIPRO), 2014 37th International Convention on, 2014, pp. 104-109.
[6] R. L. Eisenhart, "A Better Microstrip Connector," in Microwave Symposium Digest, 1978 IEEE-MTT-S International, 1978, pp. 318-320.
[7] H. Johnson, "Tapered transitions," EDN, p. 34, October 11 2001.
[8] C. Jui-Ching, et al., "Improving the High-Frequency Performance of Coaxial-to-Microstrip Transitions," Microwave Theory and Techniques, IEEE Transactions on, vol. 59, pp. 1468-1477, 2011.
[9] 洪瑞禧, "量測連接器電氣特性之探針模組之開發," 國立台灣科技大學, 2013.