Abstract

The s-parameter characterization of differential four-port networks using a two-port vector network analyzer (VNA) involves the application of an electrical stimulus to one of the four ports of the network and measuring the reflection of the signal at that port or the transmission through to any of the other three ports, under the condition that the two idle ports are terminated with a chosen load. The single-ended s-parameters obtained in this manner can be converted to the desired differential s-parameters using well established numerical combinations. However, a consequence of this technique of four-port network characterization is the return losses of each port (S11 , S22, S33 and S44) are measured thrice. Ideally, all three return loss measurements for each port must be identical. Non-ideally, however, the three return losses could be inconsistent. This work provides two methods to reduce these inconsistencies. The first involves the averaging of the return losses at each port, as well as further averaging depending on the symmetry of the device under test (DUT). The second method describes theoretically the removal of the reflections, due to the necessary loads used during the measurements, from the return loss parameters. The DUT used in this work is a pair of symmetrical coupled microstrip lines of characteristic impedance 55Omega each at 1 GHz

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call