Parallel coupled microstrip line filter based on microwave circuit simulation software

In the past, when designing various filters, it is often necessary to determine the parameters of each level of the filter based on a large number of complex empirical formulas and table lookups. This method is not only complicated and cumbersome, but the performance indicators of the designed filters are often difficult to meet the requirements. This paper combines the advanced microwave circuit simulation software ADS2008 with traditional design methods to design a parallel coupled microstrip line filter, and conducts modeling, simulation, and optimization design.

Parallel coupled microstrip bandpass filter

The edge-coupled parallel coupling line is composed of two parallel and close microstrip lines. A single bandpass filter unit is shown in Figure 1 (a). According to the transmission line theory and bandpass filter theory, the bandpass filter element is completed by the resonator on the series arm and the resonator on the parallel arm. However, it is particularly difficult to realize alternating series and parallel resonant elements on the microstrip. For this reason, an inversion converter can be used to convert the series-parallel circuit into a resonant element that is all connected in series or all connected in parallel on the line. Therefore, a single coupled microstrip filter unit can be equivalent to a combination of an admittance inversion converter and transmission line segments connected on both sides as shown in Figure 1 (b).

Although this single coupled line segment unit has the typical characteristics of a bandpass filter, it is difficult for a single bandpass filter unit to have a good filter response and a steep passband-stopband transition. Therefore, in general, a plurality of these basic coupled units are cascaded to form a practical filter. As shown in Figure 2, a typical structure of a bandpass filter composed of a cascaded coupled microstrip line segment unit is shown. Each coupled line segment is symmetrical and has a length of about a quarter wavelength (for the center frequency). The bandpass filter is composed of N + 1 coupled line bandpass filter units shown in Figure 1, and each coupled line segment can be equivalent to the circuit structure shown in Figure 1 (b). Therefore, the transmission line segment with a characteristic impedance of Z0 and an electrical angle of 2θ is between the admittance inversion converter. Z0o and Z0e are the odd-mode and even-mode characteristic impedances of the coupled line, respectively, and can be determined by the following formula:

BW is the relative bandwidth of the bandpass filter, g is a standard low-pass filter parameter, Z0 is the transmission line characteristic impedance of the filter input and output ports, and the subscripts i, i+1 represent the coupling section unit as shown in FIG. 2 .

Design of Parallel Coupled Bandpass Filter

In order to design a bandpass filter that meets the requirements, the traditional parallel coupled microstrip line design method can be combined with the advanced microwave circuit simulation software ADS2008 to convert all design requirements into actual filter design. Figure 3 is a flow chart of the design of parallel coupled microstrip line filter.

The designed filter parameters are as follows: the filter center frequency ω0 (ω0=(ωH+ωL)/2) is 2.1GHz, and the relative bandwidth is 9%; the attenuation is greater than 35dB when ω is 1.8GHz and 2.4GHz, the in-band ripple is 0.1dB, and the microstrip line characteristic impedance Z0 is 50W.

(1) First, calculate the filter low-pass normalized frequency based on the filter parameter index

, check the Chebyshev filter attenuation characteristics table to get the filter order N, and check the Chebyshev filter component parameter table to know that the parameters of the 6th-order Chebyshev standard low-pass filter with an in-band ripple of 0.1dB are g0=g6=1, g1=g5=1.1468, g2=g4=1.3712, g3=1.9750.

(2) Obtain filter bandwidth based on filter design requirements

According to formulas (1) and (2), the values ​​of parameter Ji,i+1 and odd-mode and even-mode impedance are shown in Table 1.

(3) The parameters of the selected circuit board are as follows: thickness h is 0.5 mm, dielectric constant is 4.2, relative magnetic permeability Mur is 1, conductivity CONd is 5.88E+7, and metal layer thickness is 0.03 mm. The calculation tool LineCalc in ADS is used to calculate the width W, spacing S, and length L of the microstrip line. The physical size parameters of each coupling segment are shown in Table 2.

(4) Schematic diagram simulation optimization. The above structural dimensions are input into ADS, and the dielectric parameters and frequency sweep parameters are set to perform schematic diagram simulation. The simulation results are shown in Figure 4. It can be seen that the center frequency has an obvious shift phenomenon. This is because the influence of the edge field effect is not considered when designing the parallel coupled microstrip bandpass filter. Therefore, it is necessary to optimize the setting of optimization targets and optimize the controller parameters.

In fact, the main reason why the actual value is lower than the design value is the influence of the edge effect of the open end of the coupling unit microstrip line. For the open end microstrip line, its edge effect is usually equivalent to a capacitor, and this equivalent capacitor can be replaced by an additional transmission line of a certain length. To solve this problem, the optimization function of ADS can be used. The optimized simulation results are shown in Figure 5, and the optimization schematic is shown in Figure 6.

(5) The actual circuit of the microstrip filter is calculated by the circuit board and the microstrip line. The performance of the actual circuit may be very different from the results of the principle simulation diagram. The simulation of the layout uses the moment method to directly simulate the electromagnetic field, and its results are more accurate than the simulation in the schematic diagram. Therefore, the layout generated by the schematic diagram can be simulated by the moment method (Momentm). The moment method simulation results are shown in Figure 7.

If the curve obtained by simulation cannot meet the index requirements, then it is necessary to return to the schematic window for optimization simulation. The reason for this situation is that the line width difference between adjacent coupling line nodes is too large or other parameter values ​​are not suitable. These can be changed by changing the initial value of the optimization variable, or by appropriately adjusting the optimization target side parameters according to the difference between the curve and the index, and re-optimizing.

(6) Debugging and fabrication of parallel coupled bandpass filters

The designed parallel coupled bandpass filter is made into a printed circuit board, and then the actual circuit of the microstrip filter is tested with a network analyzer. The block diagram of the debugging system is shown in Figure 8.

The parameters that need to be debugged are mainly the following: reflection parameters S11, S22 of input and output ports; attenuation in passband and stopband S21. S12; group delay [9~10]. By measuring the above parameters, the various parameters of the microstrip filter can be obtained. After the measurement is completed, observe whether the measurement results of the network analyzer meet the index requirements, and compare the results with the actual measurement results. If the test results differ too much from the design requirements, the circuit needs to be adjusted until it is redesigned and the board is made.

Conclusion

Based on the principle of parallel coupled microstrip bandpass filter, this paper combines the traditional filter design method with the method of designing filters using microwave circuit simulation tools, designs a parallel coupled bandpass filter with a relative bandwidth of 9%, gives the simulation results, and analyzes the simulation results. The simulation results show that the parallel coupled bandpass filter designed using this method meets the required indicators, while greatly reducing the design workload and improving accuracy and efficiency.