A Method for Comparing Vector Network Analyzers

We present a method of comparing two distinct vector network analyzer systems by taking the differences in calibrated S-parameters over a set of test devices. The maximum magnitude of all S-parameter differences in the ensemble of data provides an estimate of the upper bound on the system differences for the set of test devices measured. If the maximum ensemble difference is greater than the repeatability limits, either the residual errors in the two systems are not negligible, or they do not agree. We demonstrate our method here by making comparisons between two commercial frequency-domain network analyzer (FDNA) systems and by comparing an experimental time-domain network analyzer (TDNA) to a commercial FDNA.