Control Architecture for Autonomous RF Cavity Filter and Multiplexer Tuning

Abstract

The method of tuning a cavity filter and multiplexer is a stringent process since material and manufacturing tolerances. The post-production process is not only time consuming but also expensive, especially for high-order narrowband complex filters which include coupling and cross-coupling parts. In addition to manual tuning of these filters limits precise tuning and production volumes and it increases manufacturing costs. In this scope, it is inevitable to replace this traditional manual tuning task with some more advanced and automatic methods. In order to overcome these problems, software controlled robotic tuning system is implemented. This paper introduces robotic control architecture for cavity filter and multiplexer tuning based on intelligence computer-aided tuning. The system is specially designed for miniaturized tuning screw filters. It works fully autonomously and its members are COBOT, single and multi-axis robotic arms, and a cartesian platform. Also, it includes a pattern recognition system and force-torque sensors to sense and measure all relevant data during the operation process. Rf tuning and control algorithms architectures are built on data driven model which fed and learned from data derived during the test with optimization approaches. The system works with soft locking principle to prevent damage to tuning screws. In the end, experimental automated tuned filter performance and improvement of tuning iteration times are given.