Influence of non‐ideal line‐reflect‐match calibration standards on vector network analyzer S‐parameter measurements

In this paper, an improved two‐step method is presented for the sensitivity analysis of vector network analyzer (VNA) S‐parameter measurements due to the non‐ideal line‐reflect‐match (LRM) calibration standards. This improved method is based on the indirect uncertainty propagation mechanism, which is especially suitable for the S‐parameter measurements applying the self‐calibration technique. To further simplify the formula derivation, the deviation matrices [δA] and [δB] are newly defined to represent the uncertainties of the T‐matrices of error boxes. With this definition, formulas for the deviations of device under test (DUT) S‐parameters can be concluded as functions of [δA] and [δB] in a concise form. Eventually, by solving only three linear combinations of the elements from [δA] and [δB], the sensitivity coefficients of DUT S‐parameters due to non‐ideal LRM can be conveniently deduced in an analytical form. Finally, experiments are performed to verify the proposed method. The sensitivity analysis of vector network analyzer (VNA) S‐parameter measurements is performed due to the non‐ideal line‐reflect‐match (LRM) calibration standards. By using the improved two‐step method, the sensitivity coefficients describing how the definitions of LRM affect the device under test (DUT) S‐parameters are derived in an analytical form.