Studying the Electron Population Effects on Magnetron Oscillation via Particle-In-Cell Simulation

This research examines magnetron physics via the Particle-In-Cell simulations of two magnetron models, the 2D Rising Sun magnetron model and the 3D L3Harris CWM-75kW magnetron model, by using new data analytic schemes. Two data analytic schemes are presented, the electron population analysis and the breadth ratio analysis; these schemes reveal insights into the magnetron physics that are not readily accessible otherwise. Both magnetron models were simulated a) without priming, b) with RF priming, and c) with modulated electron injection. The research found that the 2D Rising Sun magnetron model is sensitive to RF priming; a 50 ns RF priming at less than 1% of the magnetron operating power resulted in a dramatically faster oscillation startup (100 ns with RF priming vs. 350 ns. without). Modulated electron injection led to a fast oscillation startup (80 ns); however, analyses show the oscillation frequency was not stable with the current simulation setup. At the current stage, the model for the L3Harris CWM-75kW does not start oscillation without priming. Oscillations were reached with both RF priming and modulated electron injection in about 160 ns. Analyses based on the results from both simulation models suggest that the electron population became more cycloidal leading to oscillation startup. Although this increased cycloidal motion of the electron population before oscillation did not directly lead to net motion of the electron population toward the anode, a clear correlation was found between the increased electron cycloidal motion and expansion of the electron hub towards the anode. Net electron motion towards the anode does not occur until the electron hub reaches the anode and the device begins to oscillate. A strong correlation was also found between the stability of the oscillation and stability of the net electron motion in the radial direction.