Uncertainty Evaluation of Calibrated Vector Network Analyzers

Abstract

A method is presented to accurately assess the uncertainty of calibrated one-port vector network analyzers (VNAs), using various techniques necessary to minimize systematic errors in the uncertainty analysis. Methods to experimentally quantify connector pin-gap discontinuity errors are presented and validated. Furthermore, we present a simplified equivalent air-line measurement model together with a multivariate optimization algorithm suitable for determining the model parameters. The unique strength of this algorithm is that it only requires two-port scattering parameter data of the airline and avoids extensive dimensional characterization. Experimental verification measurements confirm that a significant enhancement in VNA measurement accuracy is achieved by the combined use of offset discs to avoid measurement reference plane errors together with the air-line model to account for device imperfections. It is proven that using this approach, VNA uncertainty levels can be reached by industrial calibration laboratories that so far only were achievable at national measurement institutes via very extensive and cumbersome mechanical measurements.