EP1564835A1

EP1564835A1 - Inline waveguide filter with up to two out-of-band transmission zeros

Info

Publication number: EP1564835A1
Application number: EP04425096A
Authority: EP
Inventors: Danilo Attilio Gaiani; Pietro Marchisio
Original assignee: Siemens Mobile Communications SpA
Current assignee: Nokia Solutions and Networks SpA
Priority date: 2004-02-16
Filing date: 2004-02-16
Publication date: 2005-08-17
Anticipated expiration: 2024-02-16
Also published as: DE602004026535D1; EP1564835B1; ATE464670T1

Abstract

An inline microwave filter implemented by means of a metallic hollow body having a given number N of resonant cavities separated to each other by coupling means, such as capacitive/inductive irises and/or inductive posts, arranged to obtain a given N^th order Chebyshev frequency response. Two input/output waveguide ports are coupled to the respective ending resonant cavities in a manner that at least one waveguide port is also directly coupled to another resonant cavity adjacent to the ending one, to obtain an extra transmission zero outside the passband of the filter. The distance of the extra transmission zero from the corresponding limit of the passband increases with the offset of the waveguide from the midline between the coupled cavities. In a first embodiment of the filter a capacitive iris delimits the adjacent resonant cavities coupled to the waveguide port, to obtain an extra transmission zero in the lower out-of-band frequency range. In a second embodiment of the filter an inductive iris delimits the adjacent resonant cavities coupled to the waveguide port, to obtain an extra transmission zero in the higher out-of-band frequency range. In a third embodiment of the filter a capacitive iris and an inductive one delimit the adjacent resonant cavities respectively coupled to each waveguide port, to obtain two extra transmission zeros in the out-of-band frequency ranges at the two sides of the passband. In a fourth embodiment of the filter two capacitive irises delimit the adjacent resonant cavities respectively coupled to each waveguide port, to obtain two extra transmission zeros in the lower out-of-band frequency range. In a fifth embodiment of the filter two inductive irises delimit the adjacent resonant cavities respectively coupled to each waveguide port, to obtain two extra transmission zeros in the higher out-of-band frequency range. (fig. 10)

Description

FIELD OF THE INVENTION

The present invention refers to the field of the microwave filters and more precisely to an inline waveguide filter with up to two out-of-band transmission zeros.

BACKGROUND ART

The continuous requests for frequency spectrum occupancy is the main cause of more and more stringent requirements on the design of radio links and specifically for microwave filters, in particular as the shape of the out-of-band frequency response is concerned. For example, adjacent channels are very close to each other and very low interference level can be tolerated, so that high out-of-band rejection is needed especially in correspondence of some intolerable disturbs, such as the local oscillator and its harmonics. Having recourse to filters with sharp attenuations might be excessively expensive and ineffective in case the disturbs are close to one or more out-of-band spectral lines (nearly monochromatic interference). In this case the desired sharp attenuations are better obtained by transmission zeros exactly tuned to the interfering frequencies. The way to obtain these transmission zeros depends on the technology selected for the resonant cavities of the filter; only mechanical cavities are considered thereinafter. According to References [1], [2], and [3] of the annexed Bibliography, direct coupled waveguide filters with canonical frequency responses (maximally flat, elliptic, etc.) are designed starting from an equivalent electrical model of the filter with lumped parameters, including cascaded LC sections and their relative impedance inverters.
According to a known method to obtain extra transmission zeros, the values of the coupling elements (impedance inverters) can be made frequency selective, so as to match in the passband the wanted poles, whereas, at certain frequencies away from the passband the inverter values vanish, thus creating transmission zeros. A multitude of transmission zeros can be generated with this technique but their locations are difficult to control. The papers indicated at References [4] and [5] in the Bibliography constitute current background to the specific problem of synthesizing additional out-of-band transmission zeros.
The Reference [4] discloses a systematic procedure to obtain folded waveguide filters with cross couplings specifically designed to introduce extra transmission zeros. The procedure of the previous paper is a combination of traditional circuit model analysis and full-wave method, where the circuit model acts as a reference. In order to simplify the problem, the whole procedure is divided into a number of steps, each involving the tuning of one dimension (coupling size or cavity length) until the simulated S₁₁ and S₂₁ parameters fit the corresponding target value of the circuit model. The defect of this approach is the need of a mechanically complicated folded structure with the input and the output ports on the same side of the filter.
The paper of reference [5] deals with inline waveguide bandpass filters with arbitrarily located transmission zeros. The design is based on iris-coupled TM₁₁₀-mode cavities utilizing propagating but non-resonating TE₁₀ or TE₀₁ modes to create cross coupling between cavities, input, and/or output waveguides. This type of filters allows simpler mechanical structures than the folded ones, with cross-couplings by resonating mode, besides the maximum number of attenuation poles can be equal that of electrical resonances. The defect of this approach arises from the difficulty of simultaneously and independently controlling the different electromagnetic modes, either resonating or propagating, inside the whole structure including the input and output waveguides. In practice, the geometry of the iris/port centre offsets shall be accurately designed otherwise the TM₁₁₀ mode is not excited.

OBJECTS OF THE INVENTION

The main goal of the present invention is to overcome the defects of the prior art and indicate a simpler mechanical structure of a direct coupled waveguide filter able to introduce up to two transmission zeros in the lower and/or higher out-of-band region/s, preventing the use of propagating but non-resonating modes inside the cavities of the filter.

SUMMARY AND ADVANTAGES OF THE INVENTION

The invention achieves said goal by providing a waveguide filter, as disclosed in the claims. The waveguide filter according to the present invention has a metallic hollow body including a given number N of resonant cavities separated to each other by coupling structures, such as capacitive (or inductive) irises and/or inductive posts, arranged to obtain a given N^th order frequency response, with N attenuation poles, and two waveguide input/output ports respectively coupled to the resonant cavities at the ends of the filter in a manner that at least one port is also directly coupled to another resonant cavity adjacent to the last one. This allows to obtain a transmission zero outside the passband of the filter. The distance of this transmission zero from the passband edge being increased with the difference between the areas of the coupling windows of the input/output waveguide with the two adjacent resonant cavities Stating this condition, the transmission zero contributes to increase the slope of the transition between passband and stopband.
According to a first variant of the invention a waveguide port is coupled to two adjacent resonant cavities separated by a capacitive iris, in order to obtain a transmission zero in the lower out-of-band frequency range. The doubly-coupled waveguide is either connected to the input or the output port indifferently.
According to a second variant of the invention, a waveguide port is coupled to two adjacent resonant cavities separated by an inductive iris, in order to obtain a transmission zero in the higher out-of-band frequency range. The waveguide is either connected to the input or the output port as well.
According to a third variant of the invention, a first waveguide port is coupled to two adjacent resonant cavities separated by an inductive iris, and the second waveguide port is coupled to two adjacent resonant cavities separated by a capacitive iris, in order to obtain two transmission zeros in the two out-of-band frequency range at the two side of the passband.
According to a fourth variant of the invention, the two waveguide ports are respectively coupled to two adjacent resonant cavities separated by an inductive iris, in order to obtain two transmission zeros in the higher out-of-band frequency range.
According to a fifth variant of the invention, the two waveguide ports are respectively coupled to two adjacent resonant cavities separated by a capacitive iris, in order to obtain two extra transmission zeros in the lower out-of-band frequency range.
More than two transmission zeros are often unnecessary in the most of the applications, so that the limit "up to two" of the invention is immaterial and the advantage of this simple architecture can be fully appreciated. Thanks to this simple mechanical shift of the input/output waveguides, one or two transmission zeros are easily obtained: a) without having recourse to folded structures, or similarly complicated, to introduce cross-couplings, b) without the need of simultaneously control resonating and propagating modes, c) without frequency selective stubs, impedance inverters , or other complications of the known art.

BRIEF DESCRIPTION OF THE DRAWINGS

The features of the present invention which are considered to be novel are set forth with particularity in the appended claims. The invention and its advantages may be understood with reference to the following detailed description of an embodiment thereof taken in conjunction with the accompanying drawings given for purely non-limiting explanatory purposes and wherein:

fig.1 shows a top view of the mechanical body (lower part) of a traditional 3^rd order filter (Chebyshev) without extra transmission zeros;
fig.1 a shows an equivalent electrical model of the filter of fig.1;
fig.2 shows a longitudinal cross-section along a plane A-A of the filter of the previous figure;
fig.3 shows a typical plot for the insertion loss (S₂₁) and return loss (S₁₁) of the filter of fig.1;
fig.4 shows a top view of the mechanical body (lower part) of a 3^rd order (Chebyshev) filter arranged to obtain one extra transmission zero in the lower out-of-band frequency range, according to the present invention;
fig.4a shows the main significant dimensions of the filter of fig.4;
fig.4b shows an equivalent electrical model of the filter of fig.4;
fig.5 shows a partial perspective view of the filter of fig.4;
fig.6 shows a typical plot for the insertion loss (S₂₁) and return loss (S₁₁) of the filter of fig.4;
fig.7 shows a top view of the mechanical body (lower part) of a 3^rd order (Chebyshev) filter arranged to obtain one extra transmission zero in the higher out-of-band frequency range, according to the present invention;
fig.7a shows the main significant dimensions of the filter of fig.7
fig.7b shows an equivalent electrical model of the filter of fig.7;
fig.8 shows a partial perspective view of the filter of fig.7;
fig.9 shows a typical plot for the insertion loss (S₂₁) and return loss (S₁₁) of the filter of fig.7;
fig.10 shows a top view of the mechanical body (lower part) of a 3^rd order (Chebyshev) filter arranged to obtain two extra transmission zeros in both the lower and the higher out-of-band frequency range, according to the present invention;
fig.10a shows an equivalent electrical model of the filter of fig.10;
fig.11 shows a typical plot for the insertion loss (S₂₁) and return loss (S₁₁) of the filter of fig. 10.

DETAILED DESCRIPTION OF AN EMBODIMENT OF THE INVENTION

The inline structure of a waveguide filter of the known type is depicted in fig.1 as a basis for the explanation of the successive filtering structures according to the invention. The filter of fig.1 is given as an example, with only three resonant cavities; the reduced number of cavities with respect to the actual number typically used in this type of filters is held ongoing also for the description of the various filters embodying the invention. With reference to fig.1, the three resonant cavities are indicated by RIS-1, RIS-2, and RIS-3. Two capacitive irises IR1-2, IR2-3 delimit adjacent resonant cavities (RIS-1, RIS-2 and RIS-2, RIS-3), respectively. Two input/output waveguide ports (WTA and WTB) are electrically coupled and mechanically connected to the two ends of the filter in correspondence of the resonant cavities RIS-1 and RIS-3, respectively. For reasons of impedance matching, the two input/output waveguide might have reduced cross section in correspondence of their connection to the remaining part of the filter. The mechanical body of the filter includes two superimposed halves joined together by screws (not visible); the lower of these two halves is visible in the figure. Fig.1a shows an equivalent electrical model of the filter of fig.1. In this figure, all the ideal transmission lines, TL1 to TL4, have 90° electrical length and well defined characteristic impedance, so they can act as impedance inverters at the centre frequency of the filter. The physical sizes of the resonant cavities, the capacitive irises, and the two waveguides couplings, are designed to synthesize a 3^rd order Chebyshev response in accordance with the known methods based on an electrical model of the filter, as described in the References [1], [2], and [3]. The fine tuning of the central frequency with respect of the assigned mask is performed by means of tuning screws, not visible in fig.1. The longitudinal section of the complete filter along the symmetry axis A-A (fig.1) is visible in fig.2. With reference to fig.2, two more tuning screws penetrate into the metallic wall of the capacitive irises from the upper part of the filter and protrude into the air-gap between the adjacent cavities; they adjust the bandwidth and the return loss of the filter. The plot of the scattering parameters S₂₁ and S₁₁, to say insertion loss and return loss of the 3^rd order Chebyshev response, is shown in fig.3. With reference to fig.3, the three in-band poles are clearly visible from the S₁₁ plot. The centreband frequency is 19 GHz with 1 GHz 3dB-bandwidth.
Fig.4 shows an arrangement of the filter of fig.1 to introduce an extra-transmission zero in the lower out-of-band frequency range of the 3^rd order Chebyshev response. Fig.4a shows the main mechanical dimensions of the filter of fig.4, and fig.4b shows its equivalent electrical model. With reference to fig.4, the only difference from the filter of fig.1 concerns the position of the input/output waveguide WTA, which now is directly coupled to both the resonant cavities RIS-1 and RIS-2. The two waveguides WTA and WTB are a standard WR42 rectangular waveguide, which dimensions are: 10.7 mm width ("a") and 4.32 mm high ("b"). The three resonant cavities (RIS-1, RIS-2, and RIS-3) and the two waveguide (WTA and WTB) have nearly the same rectangular cross-section area. The lengths of the three resonant cavities are obtained at first by the application of well known analytical design criteria about direct coupled waveguide filters described at Ref.[1], [2] and [3]; they're further optimised by numerical analysis and simulation carried out by specific software tools mentioned at Ref.[6]. The distance "x", or "a - x", of an internal lateral wall of the waveguide WTA from the midline between the two resonant cavities RIS-1 and RIS-2, multiplied by "b", indicates the area of the coupling window with the resonant cavity RIS-1 or RIS-2, respectively. As the area of a coupling windows increases, the area of the other coupling windows decreases of the same amount.
The mechanical part of the filter having greater interest for the present invention is shown in fig.5, where the waveguide WTA carries an input signal; the response of the filter wouldn't change with reference to an outgoing signal. The plot of the insertion loss S₂₁ and return loss S₁₁ of the filter are shown in fig.6. From the comparison between the traces of fig.6 and the corresponding of fig.3, it can be easily appreciated that the slope of the transition zone at the left of the passband is greater for the filter of fig.4 than the filter of fig.1, thanks to the extra transmission zero inserted just in this frequency range. Notice that:

for "x = a" the input/output waveguide has only a coupling window to the resonant cavity RIS-1; the filter of fig. 4 degenerates into the three-poles of fig.1, having a hypothetical pole to the infinite;
for "x = 0" the input/output waveguide has only a direct coupling window to the resonant cavity RIS-2; therefore a two-poles filter is obtained.;
more in general, as the difference between the areas of the two coupling windows increases, the extra transmission zero in the lower out-of-band region goes away from the edge of the passband.

Fig.7 shows an arrangement of the filter of fig.1 to introduce an extra-transmission zero in the higher out-of-band frequency range. Fig.7a shows the main dimensions of the filter of fig.7, and fig.7b shows its equivalent electrical model. With reference to fig.7, the only difference from the filter of fig.4 is replacing the capacitive iris IR1-2 with an inductive one, better visible in the perspective representation of figures 7a and 8. In presence of an inductive iris, the area of the coupling windows of the input/output waveguide WTA to the resonant cavities RIS-1 and RIS-2 still depends by the distances "x" and "a-x"; therefore, the distance of the extra out-of-band transmission zero from the upper edge of the passband depends by the difference between the areas of the two coupling windows, as previously said for the filter of fig.4. The plots of the insertion loss S₂₁ and return loss S₁₁ of the filter of fig.7 are shown in fig.9.
Fig.10 shows an arrangement of the filter of fig.1 to introduce an extra transmission zero in both the out-of-band frequency ranges at the two edges of the passband. The filter of fig.10 includes the two arrangements of figures 4 and 7; fig.10a shows its equivalent electrical model. The plots of the insertion loss S₂₁ and return loss S₁₁ of the filter of fig.10 are shown in fig.11.
Two other cases are possible as a consequence of the preceding description, namely: the insertion of two extra transmission zeros both in the lower out-of-band frequency range, and the insertion of two extra transmission zeros both in the higher out-of-band frequency range. The first case is obtained by a filter which differs from that of fig.4 for the replication to the waveguide WTB of the offset coupling made at the waveguide WTA. The second case is obtained by a filter which differs from that of fig.7 for the replication to the waveguide WTB of the offset coupling made at the waveguide WTA; in this case a second inductive iris replaces the capacitive iris IR3-2. Different values of the coupling offsets between the two waveguide WTA and WTB allow a separation of the additional transmission zeros.

BIBLIOGRAPHY

[1] "Microwave Filters, Impedance-Matching Networks, and Coupling Structures"; G.L.Matthaei, L. Yong and E. M. T. Jones; Artech House Books; 1980.
[2] "Waveguide Handbook; N. Marcuvitz; McGraw-Hill Book Company; 1951.
[3] "Foundation for Microwave Engineering"; R. E. Collin; McGraw-Hill 2^nd Edition; @ 1992.
[4] "Design procedure for waveguide filters with cross-couplings", authors: Jan Kocbach, Kjetil Folgero; Proc. 2002 IEEE MTT-S International Microwave Symposium.
[5] "Inline waveguide filters with arbitrarily located attenuation poles", authors: Uwe Rosenberg, Smain Amari, Jens Bornemann; Proc. 2002 Asia Pacific Microwave Conf., pp. 229-232, Kyoto, Japan, Nov. 2002.
[6] Ansoft - HFSS v. 9.0, http://www.ansoft.com/products/hf/hfss/

Claims

Microwave filter implemented by means of a metallic hollow body having a given number N of resonant cavities (RIS-1, RIS-2, RIS-3) separated to each other by coupling means (IR1-2, IR2-3) arranged to obtain a given frequency response of the N^th order, and two waveguide input/output ports (WTA, WTB) coupled to respective resonant cavities (RIS-1, RIS-3) at the two ends of the filter, characterized in that a first waveguide port (WTA) is also directly coupled to another resonant cavity (RIS-2) adjacent to the ending one (RIS-1), in a way that said coupling means (IR1-2) delimit two coupling windows at the interface between the first waveguide (WTA) and its coupled resonant cavities (RIS-1, RIS-2) to introduce an extra out-of-band transmission zero at a distance from the nearest limit of the passband increasing with the difference between the areas of the two coupling windows.
Microwave filter of the claim 1, characterized in that the coupling means (IR1-2) between the adjacent resonant cavities (RIS-1, RIS-2) coupled to the first waveguide port (WTA) includes a capacitive iris to obtain the extra transmission zero in the lower out-of-band frequency range.
Microwave filter of the claim 1, characterized in that the coupling means (IR1-2) between the adjacent resonant cavities (RIS-1, RIS-2) coupled to the first waveguide port (WTA) includes an inductive iris to obtain the extra transmission zero in the higher out-of-band frequency range.
Microwave filter of the claim 1, characterized in that, the second waveguide port (WTB) is directly coupled to another resonant cavity (RIS-2) of the filter, adjacent to the resonant cavity at the respective end (RIS-3), in a way that said coupling means (IR2-3) delimit two coupling windows at the interface between the second waveguide port (WTB) and the adjacent resonant cavities (RIS-2, RIS-3) to introduce a second extra out-of-band transmission zero at a distance from the nearest limit of the passband increasing with the difference between the areas of the two coupling windows.
Microwave filter of the claim 4, characterized in that:

the coupling means (IR1-2) between the adjacent resonant cavities (RIS-1, RIS-2) coupled to the first waveguide port (WTA) include a capacitive iris to obtain a first extra transmission zero in the lower out-of-band frequency range, and

the coupling means (IR2-3) between the adjacent resonant cavities (RIS-2, RIS-3) coupled to the second waveguide (WTB) includes an inductive iris to obtain a second extra transmission zero in the higher out-of-band frequency range.
Microwave filter of the claim 4, characterized in that:

the coupling means (IR1-2) between the adjacent resonant cavities (RIS-1, RIS-2) coupled to the first waveguide port (WTA) includes a first capacitive iris to obtain a first extra transmission zero in the lower out-of-band frequency range, and

the coupling means (IR2-3) between the adjacent resonant cavities (RIS-2, RIS-3) coupled to the second waveguide port (WTB) includes second a capacitive iris to obtain a second extra transmission zero in the lower out-of-band frequency range.
Microwave filter of the claim 4, characterized in that:

the coupling means (IR1-2) between the adjacent resonant cavities (RIS-1, RIS-2) coupled to the first waveguide port (WTA) includes a first inductive iris to obtain a first extra transmission zero in the higher out-of-band frequency range, and

the coupling means (IR2-3) between the adjacent resonant cavities (RIS-2, RIS-3) coupled to the second waveguide port (WTB) includes a second inductive iris to obtain a second extra transmission zero in the higher out-of-band frequency range.