ELEMENT REMOVAL DESIGN IN MICROWAVE FILTERS

Abstract

A method of designing a microwave filter using a computerized filter optimizer, comprises generating a filter circuit design in process (DIP) comprising a plurality of circuit elements having a plurality of resonant elements and one or more non-resonant elements, optimizing the DIP by inputting the DIP into the computerized filter optimizer, determining that one of the plurality of circuit elements in the DIP is insignificant, removing the one insignificant circuit element from the DIP, deriving a final filter circuit design from the DIP, and manufacturing the microwave filter based on the final filter circuit design.

Claims

1. (canceled)

2. A method, comprising: receiving, by a processor of a computing system, a first filter circuit model comprising a plurality of circuit elements, wherein the plurality of circuit elements comprises a plurality of resonant elements and one or more non-resonant elements; selectively removing one or more circuit elements of the plurality of circuit elements responsive to a comparison of a characteristic of each of the one or more circuit elements to a threshold value; and generating, by the processor, a second filter circuit model comprising a subset of the plurality of circuit elements that were not removed.

3. The method of claim 2, wherein selectively removing one or more circuit elements further comprises iteratively removing one or more circuit elements responsive to the comparison and regenerating the first filter circuit model until no circuit element is removed.

4. The method of claim 3, wherein, in a first iteration, a first circuit element is a non-resonant element.

5. The method of claim 4, wherein regenerating the first circuit model in the first iteration further comprises one of: replacing the first circuit element with an equivalent capacitive element, responsive to the first circuit element comprising an inductive element; or replacing the first circuit element with an equivalent inductive element, responsive to the first circuit element comprising a capacitive element.

6. The method of claim 5, further comprising, in a second iteration: determining that the first circuit element has previously been transformed between capacitive and inductive elements; and regenerating the first circuit model without the first circuit element, responsive to the determination that the first circuit element has previously been transformed between capacitive and inductive elements.

7. The method of claim 2, wherein the characteristic of a first circuit element of the plurality of circuit elements is an impedance of the circuit element.

8. The method of claim 7, wherein the first circuit element is a series inductor or shunt capacitor; and further comprising removing the first circuit element responsive to an absolute impedance of the first circuit element being less than the threshold value.

9. The method of claim 2, wherein the characteristic of a first circuit element of the plurality of circuit elements is a resonant frequency, and the threshold comprises a frequency of a passband or stopband of a performance specification of the first filter circuit model.

10. The method of claim 2, wherein a first circuit element is a resonant element, and wherein removing the first circuit element comprises replacing the first circuit element with a capacitor in the first filter circuit model.

11. A system, comprising: a processor executing an element removal design engine and a memory storing a first filter circuit model comprising a plurality of circuit elements, wherein the plurality of circuit elements comprises a plurality of resonant elements and one or more non-resonant elements; and wherein the element removal design engine is configured to: selectively remove one or more circuit elements of the plurality of circuit elements responsive to a comparison of a characteristic of each of the one or more circuit elements to a threshold value; and generate a second filter circuit model comprising a subset of the plurality of circuit elements that were not removed.

12. The system of claim 11, wherein the element removal design engine is further configured to iteratively remove one or more circuit elements responsive to the comparison and regenerate the first filter circuit model until no circuit element is removed.

13. The system of claim 12, wherein, in a first iteration, a first circuit element is a non-resonant element.

14. The system of claim 13, wherein the element removal design engine is further configured to regenerate the first circuit model in the first iteration by either: replacing the first circuit element with an equivalent capacitive element, responsive to the first circuit element comprising an inductive element, or replacing the first circuit element with an equivalent inductive element, responsive to the first circuit element comprising a capacitive element.

15. The system of claim 14, wherein the element removal design engine is further configured to, in a second iteration: determine that the first circuit element has previously been transformed between capacitive and inductive elements; and regenerate the first circuit model without the first circuit element, responsive to the determination that the first circuit element has previously been transformed between capacitive and inductive elements.

16. The system of claim 11, wherein the characteristic of a first circuit element of the plurality of circuit elements is an impedance of the circuit element.

17. The system of claim 16, wherein the first circuit element is a series inductor or shunt capacitor; and wherein the element removal design engine is further configured to remove the first circuit element responsive to an absolute impedance of the first circuit element being less than the threshold value.

18. The system of claim 11, wherein the characteristic of a first circuit element of the plurality of circuit elements is a resonant frequency, and the threshold comprises a frequency of a passband or stopband of a performance specification of the first filter circuit model.

19. The system of claim 11, wherein a first circuit element is a resonant element, and wherein the element removal design engine is further configured to replace the first circuit element with a capacitor in the first filter circuit model.

20. A non-transitory computer readable medium comprising instructions that, when executed by a processor of a computing device, cause the computing device to: receive a first filter circuit model comprising a plurality of circuit elements, wherein the plurality of circuit elements comprises a plurality of resonant elements and one or more non-resonant elements; selectively remove one or more circuit elements of the plurality of circuit elements responsive to a comparison of a characteristic of each of the one or more circuit elements to a threshold value; and generate a second filter circuit model comprising a subset of the plurality of circuit elements that were not removed.

21. The computer readable medium of claim 20, further comprising instructions to cause the computing device to iteratively remove one or more circuit elements responsive to the comparison and regenerate the first filter circuit model until no circuit element is removed.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings illustrate the design and utility of preferred embodiments of the present invention, in which similar elements are referred to by common reference numerals. In order to better appreciate how the above-recited and other advantages and objects of the present inventions are obtained, a more particular description of the present inventions briefly described above will be rendered by reference to specific embodiments thereof, which are illustrated in the accompanying drawings. Understanding that these drawings depict only typical embodiments of the invention and are not therefore to be considered limiting of its scope, the invention will be described and explained with additional specificity and detail through the use of the accompanying drawings in which:

[0023] FIG. 1 is a block diagram of a prior art wireless telecommunications system;

[0024] FIG. 2 is a schematic diagram of a prior art acoustic ladder filter that can be used in the prior art wireless telecommunications system;

[0025] FIG. 3 is a frequency response plot of the prior art acoustic ladder filter of FIG. 2;

[0026] FIG. 4a is a schematic diagram of an initial circuit design of the acoustic ladder filter of FIG. 2 that can be optimized using a conventional filter optimization technique;

[0027] FIG. 4b is a frequency response plot of the initial circuit design of FIG. 4a;

[0028] FIG. 5a is a schematic diagram of an optimized final circuit design of the acoustic ladder filter of FIG. 2 resulting from the optimization of the initial filter circuit design of FIG. 4a using the conventional filter optimization technique;

[0029] FIG. 5b is a frequency response plot of the final circuit design of FIG. 5a;

[0030] FIG. 6 is a flow diagram illustrating an Element Removal Design (ERD) technique used to optimize an acoustic ladder filter in accordance with one method of the present inventions;

[0031] FIG. 7a is a schematic diagram of an initial circuit design of an acoustic ladder filter that can be optimized using the ERD technique of FIG. 6;

[0032] FIG. 7b is a frequency response plot of the initial circuit design of FIG. 7a;

[0033] FIG. 8a is a schematic diagram of an optimized final circuit design resulting from the optimization of the initial filter circuit design of FIG. 7a using the ERD technique of FIG. 6;

[0034] FIG. 8b is a frequency response plot of the final circuit design of FIG. 8a; and

[0035] FIG. 9 is a block diagram of a computerized filter design system that can implement the computerized steps of the ERD technique of FIG. 6.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] The microwave filter design optimization technique optimizes an acoustic wave (AW) microwave filter (such as surface acoustic wave (SAW), bulk acoustic wave (BAW), and film bulk acoustic wave (FBAR) filters), by changing the circuit element values, changing the circuit element type, and/or discretely removing extra or unnecessary circuit elements. These elements may include inductors, capacitors, and acoustic resonators (modeled, e.g., using the modified Butterworth-Van Dyke (MBVD) model).

[0037] This optimization technique utilizes several traditional computer optimization methods to enable the improved optimization of more complex circuits in the initial design than may be possible in the prior art. These initial filter circuit designs may be produced using any design method, for instance image design or network synthesis. This optimization technique results in a final filter circuit design with a reduced number elements compared to the initial filter circuit design, while simultaneously improving the frequency response of the filter. For the purposes of this specification, a frequency response improvement may refer to an improvement in desired microwave performance of a filter (e.g., a lower insertion loss, steeper rejection slope, higher out-of-band rejection, lower node voltage, more linear group delay, etc.). Therefore, a smaller footprint, lower cost, lower insertion loss, and greater selectively filter may be achieved, while allowing the use of traditional manufacturing processes and infrastructure.

[0038] While the microwave filter design optimization technique is described herein with reference to AW filters, it should be appreciated that this technique can be utilized with other types of microwave filters where the resonant elements can be model at some level of accuracy with a MBVD model. The optimization technique may also apply to other technologies with complex resonator elements (e.g., multi-mode dielectric resonators and other similar technologies).

[0039] Referring now to FIG. 6, one Element Removal Design (ERD) technique 100 that can be used to design an acoustic wave filter will be described. During implementation of the ERD technique 100, a design in process (DIP) is modified from an initial design to an improved final design. The method, in steps, uses computer optimization and decisions to incrementally improve the DIP to reduce circuit element count and improve performance metrics.

[0040] To this end, the ERD technique 100 first generates an initial filter circuit design based on performance specifications, such as center passband frequency, passband width, return loss, and out-of-band rejection may be generated using and image filter technique or a network synthesis technique (step 102). For the purposes of this specification, a circuit element may refer to an inductor, a capacitor, a resonator, switch or a resistor within a network of circuit interconnections.

[0041] For example, as shown in FIG. 7a, the initial filter circuit design may be a band-pass ladder 200, which comprises a number of alternating shunt resonant elements 202a and 202b and series resonant elements 202c. In the illustrated embodiment, each shunt leg of the ladder filter 200 includes two parallel shunt resonant elements 202a and 202b that may resonate at different frequencies. Each of the resonant elements 202 can be modeled with the MBVD model 54 illustrated in FIG. 2. The ladder filter 200 also comprises a plurality of non-resonant elements 204 in the form of capacitors respectively associated with the series resonant elements 202c. For the purposes of this specification, a non-resonant element may refer to a passive component in a circuit. Exemplary non-resonant elements may include inductors, capacitors, switches or resistors. Non-resonant elements may have resonances far from the frequency of interest. For instance, an inductor may have a resonance greater than 50% of the passband frequency. Thus, the filter 200 includes nine resonant elements and three non-resonant elements.

[0042] The filter 200 may initially be designed with the resonant frequencies .sub.R and static capacitances C.sub.0 for each resonator 202 illustrated in FIG. 7a, which when simulated, results in the frequency response illustrated in FIG. 7b. This frequency response is shown characterized by the following markers: R1 of Mag S21=24.627 dB at frequency =1.770 GHz; R2 of Mag S21=64.652 dB at frequency=1.830 GHz; R3 of Mag S21=3.857 dB at frequency=1.850 GHz; R4 of Mag S21=0.987 dB at frequency=1.881 GHz; R5 of Mag S21=3.039 dB at frequency=1.910 GHz; R6 of Mag S21=87.468 dB at frequency=1.930 GHz; and R7 of Mag S21=28.429 dB at frequency=1.990 GHz. It is assumed that the circuit elements of the initial filter circuit design 200 have the following parameters: gamma =12, QC.sub.0=200, Qcap=200, QL.sub.m=1000, and Rs=0.5 ohms.

[0043] Next, the ERD technique 100 inputs the DIP (in this case the initial filter design) into a computerized filter optimizer (e.g., Agilient ADS) (step 104). For the purposes of this specification, computer optimization of a DIP may refer to improving the frequency response by changing values of the circuit elements in a computer-based circuit simulator. The circuit simulator may use goals in order to compare the simulated response against a desired result. It should be appreciated that the DIP to be computer optimized is the initial circuit design at this point, the ERD technique 100 may be implemented to improve a DIP during any point of a filter design process. At any event, the resulting DIP has the same number of circuit elements as the number of circuit elements in the input DIP.

[0044] The ERD technique 100 then determines whether any of the non-resonant elements 204 in the DIP has become insignificant (vanishes) (step 106). For the purposes of this specification, a circuit element tends to vanish if a series element reactance value and/or shunt element susceptance value becomes very small during computer optimization at step 104 and/or the circuit element may be removed from a filter design without a significant impact on filter performance. A non-resonant element 204 may tend to vanish according to its kind and placement within the filter circuit design.

[0045] For example, if the non-resonant element 204 is a series inductor or a shunt capacitor, it will tend to vanish as its absolute impedance value becomes small; for instance, less than 0.1 nH (inductor) or 0.1 pF (capacitor). In this case, the non-resonant element 204 will be determined to be insignificant if its absolute impedance value is less than the threshold value. In contrast, if the non-resonant element 204 is a shunt inductor or a series capacitor, it will tend to vanish as its absolute impedance value becomes large; for instance, greater than 100 nH (inductor) or 50 pF (capacitor). In this case, the non-resonant element 204 will be determined to be insignificant if its absolute impedance value is greater than a threshold value.

[0046] As another example, a non-resonant element 204 will tend to vanish as its relative value (impedance or susceptance) becomes small (e.g., less than 10%) compared to other circuit elements of the same type (series or shunt) connected to them. Thus, if the non-resonant element 204 is a series circuit element, it will be determined to be insignificant if a percentage of its absolute value relative to an absolute value of another series non-resonant element 204 in the DIP is less than a threshold value. Similarly, if the non-resonant element 204 is a shunt circuit element, it will be determined to be insignificant if a percentage of its absolute value relative to an absolute value of another shunt non-resonant element 204 in the DIP is less than a threshold value.

[0047] As still another example, if the non-resonant element 204 is a series circuit element, it will tend to vanish as its impedance is less than a percentage (e.g., 10%) of the impedance seen in either direction from the non-resonant element 204. In this case, the non-resonant element 204 will be determined to be insignificant if a percentage of the absolute impedance value of the non-resonant element 204 relative to an impedance seen in either direction from the non-resonant element 204 is less than the threshold value. In contrast, if the non-resonant element 204 is a shunt circuit element, it will tend to vanish as its susceptance is less than a percentage (e.g., 10%) of the susceptance seen in either direction from the non-resonant element 204. In this case, the non-resonant element 204 will be determined to be insignificant if a percentage of the absolute susceptance value of the non-resonant element 204 relative to a susceptance seen in either direction from the non-resonant element 204 is less than the threshold value.

[0048] As yet another example, a non-resonant element 204 may tend to vanish when removing it results in less than a percentage degradation change (e.g., 10%) in a performance parameter of the filter circuit (e.g., insertion loss, rejection slope, out-of-band rejection, node voltage, group delay flatness, etc.). In this case, the non-resonant element 204 will be determined to be insignificant by removing the non-resonant element 204 from the DIP if the value of the performance parameter without the non-resonant element 204 degrades the value of the performance parameter with the non-resonant element 204 is less than a threshold value.

[0049] If it is determined that one of the non-resonant elements 204 in the DIP has become insignificant at step 106, the ERD technique 100 determines whether the sign of the insignificant non-resonant element 204 has been previously changed (i.e., whether the non-resonant element 204 has been transformed from an inductance to a capacitance, or from a capacitance to an inductance) (step 108). Of course, in the case where DIP is generated for the first time, there will be no such previous transformation. Notably, the ERD technique 100 makes this inquiry to ensure that removal of the insignificant non-resonant element 204 is preferable over transformation of the non-resonant element 204.

[0050] If the insignificant non-resonant element 204 has been previously transformed (indicating that the non-resonant element 204 vanished as both a capacitor and an inductor) at step 108, the ERD technique 100 generates a reduced filter circuit design by removing the insignificant non-resonant element 204 from the DIP (step 110). If the non-resonant element 204 has not been previously transformed (indicating that the non-resonant element 204 vanished as one of a capacitance and an inductance, but not yet as the other of the capacitance and inductance) at step 108, the ERD technique 100 modifies the DIP by changing the sign of the non-resonant element 204 (i.e., changing it from a capacitance to an inductance) (to the extent that the non-resonant element 204 is initially a capacitance) or from an inductance to a capacitance (to the extent that the non-resonant element 204 is initially a capacitance)) (step 112).

[0051] The ERD technique 100 then returns to step 104 to again optimize the DIP by inputting either the reduced filter circuit design generated in step 110 or the transformed filter circuit design generated in step 112 into the computerized filter optimizer, and then determines whether any of the remaining non-resonant elements 204 in the DIP has become insignificant at step 108. If it is determined that one of the non-resonant elements 204 in the DIP has become insignificant, the ERD technique again determines if the insignificant non-resonant element 204 has been previously transformed at step 108. In the case where the insignificant non-resonant element 204 has been previously transformed, then it is deemed that the insignificant non-resonant element 204 should have been removed from previous DIP, and thus, it is removed at step 110. In the case where the insignificant non-resonant element 204 has not been previously transformed, then the ERD technique 100 changes it at step 112.

[0052] If it is determined that none of the non-resonant elements 204 in the DIP (whether it is the first or a subsequently generated DIP) has become insignificant at step 106, the ERD technique then determines whether any of the resonant elements 202 in the DIP has become insignificant (vanishes) (step 114). A resonant element 202 may tend to vanish when the associated transmission zero associated with the resonant element 202 is relatively far from all passbands and stopbands; for instance, when the resonant frequency .sub.R and anti-resonant frequency .sub.A, as given in equations [1] and [2], move to more than 10% from the edge frequency of the nearest passband or stopband.

[0053] If it is determined that one of the resonant elements 202 in the DIP has become insignificant at step 114, the ERD technique modifies the DIP by replacing the insignificant resonant element 202 with a static capacitance C.sub.0, which preferably has a value equal to the value of the static capacitance of the insignificant resonant element 202 (step 116). Notably, a resonant element 202 that becomes insignificant will still affect the circuit because of its static capacitance, and so is better replaced by a capacitor than removed.

[0054] The ERD technique 100 then returns to step 104 to again optimize the DIP and then determines again whether any of the non-resonant elements 204 in the DIP (including any static capacitance C.sub.0 transformed from an insignificant resonant element 202) has become insignificant at step 106.

[0055] If it is determined that one of the non-resonant elements 204 in the DIP has become insignificant, the ERD technique again determines if the insignificant non-resonant element 204 has been previously transformed at step 108 and proceeds as discussed above. In the case where the insignificant non-resonant element 204 is a static capacitance C.sub.0 that has replaced an insignificant resonant element 202, then it is confirmed that the insignificant circuit element (which previously was a resonant element 202, but is now a non-resonant element 204) should be entirely removed, and thus, is so removed at step 110.

[0056] If it is determined that at step 106 that none of the non-resonant elements 204 in the DIP has become insignificant, the ERD technique 100 again determines whether any of the resonant elements 202 in the DIP has become insignificant at step 114. If it is determined that one of the resonant elements 202 in the DIP has become insignificant, the ERD technique replaces the insignificant resonant element 202 with a static capacitance C.sub.0 at step 116, and proceeds as discussed above.

[0057] If it determined that none of the resonant elements 202 in the DIP has become insignificant, the ERD technique 100 deems the DIP to be the improved final filter circuit design (step 118), which in the exemplary embodiment, has the resonant frequencies .sub.R and static capacitances C.sub.0 for the remaining resonant elements 202 and non-resonant elements 204 illustrated in FIG. 8a, which when simulated, results in the frequency response illustrated in FIG. 8b. This frequency response is shown characterized by the following markers: P1 of Mag S21=30.080 dB at frequency=1.770 GHz; P2 of Mag S21=34.193 dB at frequency=1.830 GHz; P3 of Mag S21=1.394 dB at frequency=1.850 GHz; P4 of Mag S21=0.761 dB at frequency=1.872 GHz; P5 of Mag S21=1.406 dB at frequency=1.910 GHz; P6 of Mag S21=45.227 dB at frequency=1.930 GHz; and P7 of Mag S21=45.227 dB at frequency=1.990 GHz.

[0058] As can be seen from a comparison between FIG. 7a and FIG. 8a, the improved final filter circuit design includes fewer circuit elements, and in particular, two less resonant elements 202 and one less non-resonant element 204. As can be seen from a comparison between FIG. 8a and FIG. 8b, the final filter circuit design yields a flatter frequency response and loss at the pass-band. It should also be noted that a conventional optimization technique (i.e., without removing circuit elements) was performed on the initial filter circuit 200 illustrated in FIG. 7a, resulting in a final filter circuit having a frequency response performance that was not as good as the frequency response performance yield by the ERT technique illustrated in FIG. 8b. Thus, contrary to conventional wisdom, the removal of circuit elements using the ERT technique not only decreased the cost and size of the microwave filter, it improves the frequency response performance of the microwave filter over prior art microwave filters that have more circuit elements.

[0059] Notably, the ERD technique 100 may analyze the frequency response of DIP resulting from step 104 in order to determine whether a previous step should be undone or redone.

[0060] For example, if the frequency response performance of the DIP is not better than the frequency response performance of the initial filter circuit design, this means that the initial filter circuit design is not acceptable, and thus, a different initial filter circuit design may be considered, which may then be inputted into the computerized filter optimizer at step 104.

[0061] As another example, if the frequency response performance of the DIP as it existed after a non-resonant element 204 has been transformed (from a capacitance to an inductance, or from an inductance to a capacitance) is worse by a threshold amount relative to the frequency response performance of the DIP as it existed before the non-resonant element 204 was transformed, the ERD technique 100 may simply return to the previous DIP and remove the non-resonant element 204 from that design.

[0062] As still another example, if the frequency response performance of the DIP as it existed after a resonant element 202 has been transformed into a static capacitance C.sub.0 is worse by a threshold amount relative to the frequency response performance of the DIP as it existed before the resonant element 202 was transformed, the ERD technique may simply return to the previous DIP and restore the resonant element 204 back into that design.

[0063] Once the improved final filter circuit design is achieved, the ERD technique 100 manufactures an actual microwave filter based on the final filter circuit design (step 120). Preferably, the circuit element values of the actual microwave filter will match the corresponding circuit element values in the improved final filter circuit design.

[0064] Referring first to FIG. 9, a computerized filter design system 300 may be used to design a microwave filter using the ERD technique 100. The computerized filter design system 300 generally comprises a user interface 302 configured for receiving information and data from a user (e.g., parameter values and filter specifications) and outputting an optimized filter circuit design to the user; a memory 304 configured for storing filter design software 308 (which may take the form of software instructions, which may include, but are not limited to, routines, programs, objects, components, data structures, procedures, modules, functions, and the like that perform particular functions or implement particular abstract data types), as well as the information and data input from the user via the user interface 302; and a processor 306 configured for executing the filter design software. The filter design software program 308 is divided into sub-programs, in particular, a conventional network design synthesizer 310 (which can be used to generate the initial filter circuit design at step 102), a conventional filter optimizer 312 (which can be used to generate the DIP at step 104), and an element removal design engine 314 that controls the network design synthesizer 88 and filter optimizer 90 in accordance with filter circuit design aspects of the ERD technique 100 in order to generate the optimized final circuit design.

[0065] Although particular embodiments of the present invention have been shown and described, it should be understood that the above discussion is not intended to limit the present invention to these embodiments. It will be obvious to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the present invention. For example, the present invention has applications well beyond filters with a single input and output, and particular embodiments of the present invention may be used to form duplexers, multiplexers, channelizers, reactive switches, etc., where low-loss selective circuits may be used. Thus, the present invention is intended to cover alternatives, modifications, and equivalents that may fall within the spirit and scope of the present invention as defined by the claims.