In order to solve the problem of insufficient consideration of power handling capacity in the design of bulk acoustic wave (BAW) filters, a method for evaluating the power handling capacity of BAW filters is proposed. First, the average temperature values of each stack of the unit resonator under the self-heating effect can be obtained by the acousticelectromagnetic-thermal multi-physics simulation of the BAW filter at a specific input power and ambient temperature. Then, the average temperature values are introduced as design parameters into the Mason model to construct the Mason(T) model with temperature parameters. Finally, the transmission curve of the BAW filter can be simulated with the Mason(T) model, and the power handling capacity of BAW filter can also be evaluated according to whether the specifications exceed the standard in the whole temperature range. The simulation results of the case filter show that all the specifications of the BAW filter do not exceed the standard when the input power is 23 dBm and the ambient temperature is in the whole temperature range (-25°C - 85°C). However the input power is 30 dBm and the ambient temperature is 30°C, the upper band insertion loss and out-of-band rejection of the BAW filter have exceeded the standard, and the performance is significantly degraded.
With the increased performance requirements for film bulk acoustic resonator (FBAR) devices, accurate test of FBAR device parameters has become critical. The key of the FBAR on-board test is its test fixture structure and de-embedding methods. In this paper, the research status of FBAR board testing technology is reviewed. The parasitic effects, impedance matching and clamping design of the test fixture structure design process are discussed. The principle and error model of de-embedding and the advantages/disadvantages of each calibration method are analyzed. The accuracy of FBAR on-board test can be improved by reducing parasitic effects, optimizing impedance matching, improving calibration methods, and optimizing error model.