A Design Rule for Passive Reconfigurable Microwave Devices

Abstract

Every reconfigurable microwave device places a tuning element - a varactor, a diode, transistor, microelectromechanical or phase-change switch, a ferroelectric or liquid-crystal load - inside a fixed passive network. The network can make a small change of the element produce a large change of the response, and design practice often credits it with the device's tuning capability. This paper shows that a fixed, linear, passive network cannot add to that capability. Each tuning element carries a tuning budget, a dimensionless length computed from its own impedance trajectory, and a fixed passive network can transmit that budget or lose part of it, never enlarge it. The unit of the budget is the hyperbolic distance of the Smith chart: the distance from a match is the logarithm of the voltage standing-wave ratio, and the distance between two states sets their mutual mismatch and is the logarithm of their commutation quality factor. From one inequality we derive closed-form, ready-to-use limits for analog and digital phase shifters, true-time-delay lines, switches, attenuators, vector modulators, impedance tuners, tunable filters, reconfigurable antennas, metasurface cells and multiport networks; the two-state and attenuation limits are attained by explicit circuits, and the phase limit by a constant-loss trajectory. We introduce the $Λ$-chart, a Smith-chart companion on which budgets and requirements are read graphically, and apply it to four published microwave case studies.