When AlN thin films are deposited directly on the high-resistance silicon (HR-Si) substrate, a conductive layer will be formed on the HR-Si surface. This phenomenon is called the parasitic surface conduction (PSC) effect. The presence of the PSC effect will increase the power consumption of electronic components. Therefore, it is necessary to reduce the PSC effect. In prior technology, the polysilicon layer is usually used as the trap-rich layer to reduce the PSC effect. Experiments show that compared to AlN films deposited directly on HR-Si, the AlN substrates with polysilicon introduced on HR-Si have less radio frequency (RF) loss. To verify the effect of polysilicon on RF loss, polysilicon films of three different thicknesses and several different roughnesses were introduced. The results show that the thickness of the polysilicon will affect the RF loss, while the roughness has almost no effect on it. The polysilicon trap-rich layer can reduce the RF loss, which gradually becomes smaller as the polysilicon thickness increases.

Spectrum resources are becoming increasingly crowded, and the isolation interval between different systems is getting smaller and smaller. This puts forward higher requirements for the duplexer. The duplexer is an important part of the radio frequency front end, and the isolation requirement is becoming higher. This paper presents a phase canceling circuit to improve the performance of the duplexer to meet the requirement of the communication system for isolation. A phase canceling circuit is an effective method to enhance the isolation through use of a surface acoustic wave (SAW) on-chip circuit. It contains a duplexer and a branch. The branch is designed for diminishing the leakage signal of the duplexer. Compared with the leakage signal, the branch consists of two attenuators and a phase shifter to generate a signal which has equal extent and reverse phase. As a result, this method is capable of increasing the isolation of band 5 by 12 dB in the downlink frequency. Meanwhile, it neither affects other factors, such as insertion loss or return loss, nor increases the size of the chip. The phase canceling circuit is expected to promote the quality of duplexer to satisfy the strict requirements in 4G and 5G systems.

Due to the influence of the optical proximity effect (OPE), it is easy for a pattern of photoresistance to be inconsistent with a design pattern, thus damaging the performance of a SAW resonator. To solve this problem, this paper proposes an optimization method for SAW filters based on optical proximity correction (OPC). This method can avoid the tip discharge problem of SAW filters by suppressing the problem of rounding and shrinking of dummy electrode and electrode tail caused by OPE. This method increases the quality factor (Q) of the SAW resonator and thus decreases the insertion loss of the SAW filter. The filter increases the bandwidth by 1.8 MHz at −1.5 dB after applying the OPC method. Additionally, it improves the stability of the filter under high power conditions.

This paper proposes an improved method to calculate the mutual capacitance between interdigital transducer (IDT) electrodes to enhance the accuracy of the traditional coupling-of-modes (COM) model, which is commonly used to simulate surface acoustic wave (SAW) filters and duplexers. In this method, the boundary element method (BEM) is adopted to obtain the capacitance per unit length in a layered medium, while the partial capacitance (PC) method is used to derive the effective relative permittivity of the multi-layered IDT. Numerical results from commercially available software are provided for comparison with the results calculated using the proposed method. The consistent results verify the validity and accuracy of this method, which also demonstrates significantly faster calculation speed compared to commercially available software. Precise electrical response prediction of a dual-mode SAW (DMS) filter can be achieved by applying this method to the COM model, and this ultra-fast calculation method can also be included in filter design optimization.