First-Order Modeling of Errors Due to Coaxial Connector Interface
Arkadiusz Lewandowski , Leszek J. Opalski , Wojciech Wiatr , Arkadiusz Gołaszewski , Adam Abramowicz
AbstractWe present an approach for first-order modeling of errors in vector-network-analyzer (VNA) measurements caused by the coaxial-connector interface. These errors result primarily from the presence of a gap in the inner conductor. Our approach relies on an approximate model of the connector interface that comprises an analytical equivalent-circuit for the inner conductor gap section and a behavioral scattering-parameter model for the socket-contact section. We verify this model using full-wave electromagnetic simulations for the 1.85, 2.4, 3.5 mm, and Type-N coaxial-connector interface for different inner conductor-gap widths. We further apply this model to asses first-order errors in VNA calibration introduced by the tolerance on the inner conductor-gap width. We show that the inner conductor-gap effect can be accounted for with a series inductance inserted at the inner conductor mating-plane while the effect of the socket contact can be lumped into the VNA calibration coefficients. We verify this error model experimentally by measuring a Type-N offset load with an adjustable inner conductor gap. We also demonstrate that including our error model in coaxial offset-short definitions leads to improved consistency of a one-port VNA calibration with 3.5-mm coaxial offset-short standards. Our results can be used in the uncertainty analysis and accuracy enhancement of VNA scattering-parameter measurements.
|Journal series||IEEE Transactions on Microwave Theory and Techniques, ISSN 0018-9480, (A 35 pkt)|
|Publication size in sheets||0.6|
|Keywords in English||Calibration, coaxial connector, error modeling, inner conductor gap, vector network analyzer (VNA)|
|ASJC Classification||; ;|
|Score||= 35.0, 30-04-2019, ArticleFromJournal|
|Publication indicators||= 0; = 0; : 2017 = 1.956; : 2017 = 3.176 (2) - 2017=3.603 (5)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.