диафрагмированные волноводные фильтры / 167bde91-8f36-47a0-a8ac-42f9aca87e2c

2020 IEEE Asia-Pacific Microwave Conference (APMC 2020) | 978-1-7281-6962-0/20/$31.00 ©2020 IEEE | DOI: 10.1109/APMC47863.2020.9331599

Fig. 2. Set-up of the third order extracted pole waveguide filter with one TZ. The dimensions are as follows (all in mm): a = 22.86, b = 10.16,

lep = 26.2, lNRN = 33.6, l2 = 19.7, l3 = 18.8, dS,NRN = 15.8, lNRN,ep = 17.3, oc = 3.9, dNRN,2 = 11.1, d2,3 = 8.6, d3,L = 12.2. The width of the coupling aperture between the NRN and the extracted pole

resonator is 15 mm and the height is 2.5 mm. The red dashed line indicates the cutting plane for the manufacturing.

Fig. 3. Measurement results of the filter from Fig. 2.

In the proposed design the non-resonating node is realized by an oversized cavity, whose resonance frequencies are far away from the passband of the filter. As e.g. shown in [8], it is also possible to directly use the source / load port as a NRN in order to omit a further blend pair (here described by the coupling factor MS,NRN and the dimension dS,NRN ). However, in this case the extracted pole resonator might have some larger distance to the rest of the filter. The NRN in the current design behaves like a constant reactance as desired by the coupling matrix coefficients. The extracted pole cavity is stacked by an iris on top of the NRN. The tuning takes place by the width, length and offset of the coupling aperture.

Each extracted pole section is per definition able to realize one TZ. The position strongly depends on the resonance frequency of the extracted pole cavity. Note, that the stacked extracted pole design proposed here enables a further TZ relatively far away from the passband. Therefore, the filter from Fig. 2 has an extra TZ at 10.45 GHz. The reason for this TZ is a resonance within the NRN - extracted pole section as also discussed in [8]. However, the position of the extra TZ depends on the dimensions of the non-resonating node and the extracted pole resonator and can therefore not be chosen arbitrarily.

The filter from Fig. 2 is printed in two halves along the cutting plane as indicated in Fig. 2. As only plastic materials

can be processed by the common FDM technique, the halves are subsequently sprayed with a copper spray and galvanized with a pure copper layer as comprehensively described in [9].

The measurement results of the filter from Fig. 2 are shown in Fig. 3. The filter is matched to 22 dB within the

One obvious reason are increased cavities due manufacturing tolerances and a rough cutting plane. The insertion loss is lower than 0.25 dB and the Q-factor can be calculated to be Qu ≈ 2800, which is a valuable result considering the high surface roughness of FDM printed workpieces [10]. Furthermore, it is worth mentioning that the high return loss level was achieved without tuning screws.

III. OPTIMIZED DESIGN WITH FREQUENCY-DEPENDENT

COUPLINGS

The realization of waveguide filters with additive manufacturing techniques enables additional degrees of freedom in the design process, e.g. the insertion of frequency dependent couplings to increase the near passband selectivity. For example, in [11] the realization of a frequency-dependent coupling aperture based on a capacitive and inductive window is proposed (Fig. 4). The realization of this iris is extremely difficult in e.g. traditional milling techniques due to the small thickness of the aperture and the resulting high aspect ratio. Therefore, the filter must theoretically be further segmented at positions where this aperture should be introduced. Otherwise, a segmentation further increases the losses. However, the realization of this coupling aperture in 3D printing techniques is quite simple. Due to the symmetry, the filter can still be cut in the E-plane. Fig. 4 shows the frequency-dependent coupling aperture in accordance with [11]. The manufacturing cut is

Fig. 4. Frequency-dependent coupling aperture (exemplary) between cavity 3 and the load port. (a) Side view, the red line indicates the manufacturing (E-plane) cut. (b) schematic front view [11].

along the red dashed line on the top of the structure. Each frequency-dependent coupling aperture is able to realize one TZ in the vicinity of the passband and replaces a classical inductive iris in the filter set-up.

Fig. 5 (a) shows the measurement results in comparison to the simulation, if a frequency-dependent coupling as shown in Fig. 4 replaces the coupling aperture between the source port and the non-resonating node. The dimensions of the coupling aperture are shown in the caption of Fig. 5. The filter is slightly de-tuned and shifted to lower frequencies (around 80 MHz). However, an additional TZ below the passband can

be observed. This TZ is also shifted to lower frequencies due to manufacturing inaccuracies of the frequency dependent coupling aperture. The insertion loss is again in the order of 0.25 dB, however, a slightly decreased Q-factor compared to the first measurement is achieved (Qu ≈ 2200).

A second filter was subsequently manufactured, in which additionally the coupling aperture between cavity three and the load port was replaced by a frequency-dependent coupling for the realization of a second TZ below the passband. The measurement results are shown in Fig. 5 (b) while the dimensions are indicated in the caption as well.

No noticeable frequency shift can be observed. Otherwise, the filter is only matched to RL ≈ 10 dB and the insertion loss is increased as shown in the inset of Fig. 5 (b). The Q-factor can be estimated to be Qu ≈ 1800. Compared to the first measurements, the alignment of the two halves is assumed problematic especially concerning the two frequencydependent couplings.

Note, that also the frequency-dependent coupling aperture is cut in the E-plane in order to reduce the losses. Fig. 6 shows a photograph of the third order extraced pole filter, where the first and last coupling is replaced by a frequency-dependent coupling.

Fig. 6. Photo of the third order extracted pole waveguide filter with a frequency-dependent coupling in the source and load port after galvanization (one half is shown).

IV. CONCLUSION

In this paper, third order additive manufactured extracted pole waveguide filters in the X-band are proposed. Due to the stacked configuration of the NRN - extracted pole section, the filters can be cut in the E-plane, which allows to utilize the low cost FDM manufacturing technique. The basic set-up shows two TZs, one of which is generated by the extracted pole cavity while the second one arises due to a resonance in the NRN - extracted pole section. Additional TZs can be introduced, if classical inductive irises are replaced by frequency-dependent coupling apertures. Each of these coupling apertures introduces one further TZ. The filters can still be cut in the E-plane due to the symmetry of the frequency-dependent coupling aperture. The measurement results are in good agreement with the simulation apart from a frequency shift. The Q-factor is quite high considering the surface roughness of the FDM manufacturing process.

REFERENCES

[1]J. Rhodes and R. Cameron, “General extracted pole synthesis technique with applications to low-loss T E011 mode filters,” IEEE Trans. Microw. Theory Techn., vol. 28, no. 9, pp. 1018–1028, Sep. 1980.

[2]R. J. Cameron, C. M. Kudsia, and R. R. Mansour, Microwave Filters for Communication Systems. Wiley, 2007.

[3]G. Macchiarella, M. Oldoni, and S. Tamiazzo, “Narrowband microwave filters with mixed topology,” IEEE Trans. Microw. Theory Techn., vol. 60, no. 12, pp. 3980–3987, Dec. 2012.

[4]S. L. Romano and M. S. Palma, “Implementation of extracted pole filters in rectangular waveguide,” in 44th Eur. Microw. Conf., Italy. IEEE, Oct. 2014.

[5]S. Cogollos, R. Cameron, R. Mansour, M. Yu, and V. Boria, “Synthesis and design procedure for high performance waveguide filters based on nonresonating nodes,” in IEEE MTT-S Int. Microw. Symp., Jun. 2007.

[6]J.-S. Hong and M. J. Lancaster, Microstrip Filters for RF/Microwave applications. John Wiley & Sons Inc., 2001.

[7]A. G. Lamperez´ et al., “Generation of accurate rational models of lossy systems using the cauchy method,” IEEE Microw. Wireless Compon. Lett., vol. 14, no. 10, pp. 490–492, Oct. 2004.

[8]Y. Feng et al., “WR-2.8 band pseudoelliptic waveguide filter based on singlet and extracted pole resonator,” IEEE Access, vol. 7, pp. 54 705– 54 711, Apr. 2019.

[9]D. Miek, S. Simmich, F. Kamrath, and M. Hoft,¨ “Additive manufacturing of E-plane cut dual-mode X-band waveguide filters with mixed topologies,” IEEE Trans. Microw. Theory Techn., vol. 68, no. 6, pp. 2097–2107, Jun. 2020.

[10]A. Gomez-Torrent et al., “A study of the additive manufacturing technology for RF/microwave components,” in 11th Europ. Conf. Ant. Propag. (EUCAP). IEEE, Mar. 2017.

[11]U. Rosenberg, S. Amari, and F. Seyfert, “Pseudo-elliptic direct-coupled resonator filters based on transmission-zero-generating irises,” 40th Eur. Microw. Conf., Sep. 2010.