Techniques for measuring the effective source reflection coefficient of two-resistor power splitters

Three techniques for measuring the effective source reflection coefficient of a two-resistor power splitter are discussed. Computing the coefficient from measured S parameters is sensitive to small measurement errors. Descriptions are given of two direct measurement methods that avoid such sensitivities. The more convenient one uses a passive open circuit on the input port of the splitter during measurements at the output ports. It yields acceptable results in many cases. Better results are obtained in all situations by injecting the proper signal at the input port. The three techniques are compared experimentally for two typical power splitters.