Optimization of Microstrip Lowpass Filters with Three-Dimensional Stubs

Abstract

Introduction. Lowpass filters (LPFs) are used to suppress unwanted harmonics and spurious signals. Microstrip LPFs are widely used in various electronic systems. New, more stringent system requirements demand increased LPF selectivity. In the previous work, we considered the calculation of the fifth-order microstrip LPF with three-dimensional (3D) stubs. According to the results of 3D modeling, the LPF frequency response (FR) has a steepness close (but slightly worse) to the steepness of the FR LPF based on lumped elements. In the presented paper the optimization of the LPF with 3D stubs is performed, the experiment results for 3D stub and LPF are given. Optimized LPF has a steeper FR than LPF based on lumped elements.

1 Features of the fifth-order LPF with 3D stubs. The quasi-lumped inductance is made by a through hole in the dielectric with an overhead conductor above it, and the quasi-lumped capacitance is made by a blind metallized hole on the signal conductor side. In addition to the use of 3D reactive elements, the LPF has the following differences from traditional solutions: 1) the stub is connected to the line by a small contact pad; 2) the stubs are placed on different sides relative to the direction of wave propagation.

2 Optimization of the LPF. Value of the stub’s rejection frequency is affected by the stub-line T-junction parasitic inductance connected in series with the stub. Parasitic inductance value depends on the depth of the stub hole and the contact pad sizes. If the LPF stubs contact pads are different in sizes, their rejection frequencies will be different. This will widen the LPF suppression band. Thus, by choosing stub holes depth and contact pads sizes, you can optimize steepness and suppression band width the LPF FR. As a result of optimization, the LPF FR steepness increased from 20.0 to 22.9 dB/GHz and the suppression band widen from 1.9 to 3.8 GHz at the level of –60 dB. The optimized FR has a steepness higher than the FR LPF based on lumped elements equal to 21.5 dB/GHz.

3 Experimental results. Photos and experimental FRs of the 3D stub and LPF with 3D stubs are given. 3D stub experimental and calculated values of the rejection frequency, the rejection level and the relative error of the calculated values are 6.41 and 5.72 GHz, −51.5 and −49.4 dB, 11% and 4%, respectively. The experimental LPF FR is in a good agreement with calculated one.

4 Results discussion. The presence of the third size in the microstrip elements provides not only a significant increase in their efficiency, but also additional design possibilities. Since the value of the parasitic inductance, which determines the rejection frequency, depends on the 3D stub hole depth, this parameter, as well as the contact pads sizes are optimization parameters to the FR steepness and suppression bandwidth.

Conclusion. The use of 3D stubs as quasi-lumped capacitances allows to optimize the LPF FR by choosing the stubs hole depth and the stub contact pad sizes. As a result of LPF optimization, the FR steepness increased and the suppression band widen. The steepness of the optimized FR is higher than the FR LPF based on lumped elements.