Constructing and benchmarking a pulsed-RF, pulsed-bias S-parameter system

Abstract

A pulsed-bias, pulsed-RF S-parameter system has been assembled and validated through a systematic benchmarking process. The system was constructed using an Anritsu 37397C Lightning Vector Network Analyzer, along with an external switch, a digital delay generator, and a custom bias tee. A series of benchmarking tests, using passive and active DUTs, was performed to exercise and validate the system. A range of test setup conditions, including varied pulse lengths and duty cycles, were examined. A passive filter was used to explore the dynamic range and effective noise floor of the pulsed S-parameter measurement as compared to CW measurements. Finally, a Si LDMOSFET was measured and the pulsed- and continuous-bias measurement results were found to show dispersion in |S21| due to device self-heating in the continuous-bias case.