An intelligence driven active loadpull system

Abstract

This paper describes how the application of the PHD model can add intelligence to an open loop active loadpull system. This intelligence driven approach by providing for an improved prediction of the operating conditions required to emulate a specified load speeds up the load emulation convergence process by minimizing the number of iterations to predict the injected signal, therefore making more efficient use of a measurement system. The results were validated by carrying out loadpull measurements on the fundamental tone of a 10×75um GaAs HEMT, operating at 3 GHz.