Broadband Partially Covered Microstrip-to-Waveguide Transition With Enhanced Radiation Suppression for Millimeter-Wave Transmission

A broadband microstrip-to-waveguide transition (MWT) structure is reported for efficient millimeter-wave (mm-wave) signal delivery. The transition design, incorporating a printed circuit board (PCB) with three copper layers and two dielectric layers to realize a partially covered microstrip line (MSL) and a via fence, achieves low loss and broadband signal delivery by suppressing radiation leakage and by using multiple resonances, respectively. A back-to-back (B2B) measurement was performed to validate the MWT design. It exhibited a return loss (RL) greater than 10 dB in the frequency range of 41–63 GHz, with an operating bandwidth of 42%. The proposed MWT exhibits good RL, low radiation loss, and broadband performances at mm-wave frequencies.


Broadband Partially Covered Microstrip-to-Waveguide Transition With Enhanced Radiation Suppression for Millimeter-Wave Transmission
Hafeez-Ur-Rehman , Ha Il Song, Sean Park, and Jae-Hyung Jang , Senior Member, IEEE Abstract-A broadband microstrip-to-waveguide transition (MWT) structure is reported for efficient millimeter-wave (mm-wave) signal delivery.The transition design, incorporating a printed circuit board (PCB) with three copper layers and two dielectric layers to realize a partially covered microstrip line (MSL) and a via fence, achieves low loss and broadband signal delivery by suppressing radiation leakage and by using multiple resonances, respectively.A back-to-back (B2B) measurement was performed to validate the MWT design.It exhibited a return loss (RL) greater than 10 dB in the frequency range of 41-63 GHz, with an operating bandwidth of 42%.The proposed MWT exhibits good RL, low radiation loss, and broadband performances at mm-wave frequencies.

I. INTRODUCTION
E FFECTIVE signal delivery at microwave and millimeter- wave (mm-wave) frequencies requires high-performance passive and active components and efficient transition structures [1].To enable seamless signal coupling between transceiver modules on a printed circuit board (PCB) and the waveguide used for board-to-board communication, a highquality microstrip line (MSL)-to-waveguide transition (MWT) is necessary.However, radiation losses can significantly impact the MWT's frequency response at mm-wave frequencies.The most common method to reduce radiation leakage is using a quarter-wave metallic short waveguide cavity to surround the radiation zone [2], [3], [4].However, the waveguide back-short cavity increases manufacturing and assembly costs [5].To address this issue, a planar transition composed of a short-waveguide pattern on a single dielectric substrate has been proposed as an alternative approach to reduce the required parts and assembly costs while reducing radiation leakage [6].However, this design approach allows only a single transmission mode to dominate at the transition's input, leading to a narrow bandwidth of 4.53 (5.9%) with a single resonance.
A broadband planar microstrip-to-waveguide transition has been proposed as an alternative solution, involving the excitation of multiple transmission modes to form a double-resonant structure [7].However, this approach offers a limited bandwidth of only 13.8% within the mm-wave band.Another technique for reducing radiation leakage and increasing coupling is to employ double slots and double matching stubs on the microstrip circuit [8].This design faces challenging manufacturing processes due to the sensitivity of the coupling mechanism to the alignment of the double slot, the matching stub, and the numerous parameters involved.
This study introduces a partially covered back-short pattern in the MWT to achieve a cost-effective, wideband, and low insertion loss (IL).The simple MWT structures achieve large bandwidth and low loss coupling by minimizing radiation loss without using a shorted waveguide cavity or additional matching structures.The simulation results are compared with the experimental results.

II. DESIGN AND CONFIGURATION OF THE MWT
As depicted in Fig. 1(a), the MWT designed in this study utilizes a PCB consisting of the three 0.035-mm-thick copper layers (L 1 , L 2 , and L 3 ) and Taconic RF-35A2 dielectric layers (D 1 and D 2 ), whose thicknesses are 0.02λ and 0.05λ , respectively.λ is the signal wavelength at the central frequency of 52 GHz.Through-hole vias (THVs) establish connectivity between L 1 and L 3 .The custom-made flexible rectangular The MWT design includes an aperture-coupled patch for matching and conversion from the quasi-TEM (Q-TEM) mode to the TE 10 mode within the waveguide, as shown in Fig. 1(b).The Q-TEM signal is launched onto the MSL as it approaches to the partially covered part, and a transition occurs from the MSL to the stripline.The matching element at the waveguide aperture further transforms the TEM mode of the stripline into the TE 10 mode in E-TUBE.Partially covered metal layer mitigates radiation leakage, while metallic vias connect the cover to the ground plane.As shown in Fig. 2, the simulation results of a single MWT exhibit the return loss (RL) larger than 10 dB at the frequency band from 39 to 64.5 GHz and the IL of 0.22 dB at the central frequency of 52 GHz.
Quantitative analysis was carried out to show the effectiveness of the design in suppressing the radiation leakage, as shown in Fig. 3. Case 1 without the cover suffers from high radiation loss.As high as 25.4% of the input power is lost due  to the backside radiation.Case 2 with the cover successfully mitigated the radiation loss, but 24.9% of the input power was lost due to the parallel-plate mode leakage.A fence of vias is additionally applied in case 3. The parallel-plate mode leakage was successfully suppressed by the effect of the metal cover and via fence and 1.68% of the input power is lost by the radiation at the end of the MWT structure.By using the partially covered MSL surrounded by the via fence, the percentile radiated power was reduced from 25.4% to 1.68%, as shown in Fig. 3(d).It validates the effectiveness of the short-circuited metal cover in providing effective shielding against the electromagnetic leakage.For the impedance matching between the MSL and the E-TUBE, the length of the probe is found to be the most important parameter.The broadband performance is achieved by the parametric analysis of the structural parameters of the MWT detailed description in given Supplementary Notes Section I-A.The length of the matching element and via placement parameter is tuned to ensure dual resonance within the operating frequency band (Supplementary Fig. 1).

III. EXPERIMENTAL RESULTS AND DISCUSSION
The performance of the designed MWT was evaluated using a back-to-back (B2B) configuration with an E-TUBE.Fig. 4(a) Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.To perform the B2B characterization, two mounting structures were utilized.One is placed at the interface between the PCB and waveguide connector to hold the circuit in place, while the other one is placed at the interface between the PCB and coaxial connector and securely fixed using screws, as shown in Fig. 4(b).The measured and the simulated frequency responses of the B2B setup are shown in Fig. 5.The measured RL is larger than 10 dB within 41-63 GHz, with an operating bandwidth of 42%.For a 1-m-long E-TUBE, the average measured IL is 10.5 dB from 41 to 63 GHz.Although the IL data are satisfactorily flat and in good agreement with the simulations, the measured IL is higher than the simulation.This is mainly due to a manufacturing error as discussed in supplementary notes.It implies that precise manufacturing is more crucial at a high frequency.The B2B setup has a pair of MWTs with the E-TUBE.The total IL for the B2B setup is represented as follows: where IL T is the total IL of the B2B transition, IL ST is the IL for the single transition side, IL W is the waveguide loss per unit length, and l is the waveguide length.

TABLE I PERFORMANCE COMPARISON OF THE VERTICAL WAVEGUIDE TRANSITIONS
The IL T is 10.5 dB for the B2B setup with a 1.0-m-long waveguide.Section II-A of the supplementary notes elaborates the methodology to calculate the IL for a single transition side (IL ST = 2.50 dB) and the waveguide loss per unit length (IL W = 5.50 dB).IL ST include additional losses, such as the end-launched connector and MSL loss.To precisely estimate the IL of the MWT, a detailed calculation was carried out, considering the manufacturing errors [Supplementary Notes, Section II-B].By subtracting the estimated loss associated with MSL (0.15 dB) and end-launched connector (0.15 dB) from IL ST , the MWT loss is calculated to be 2.2 dB.The MWT loss can be reduced to 1.1 dB, if the connector manufacturing error is minimized.As shown in the Supplementary Note, Section II-B, the extra loss due to the manufacturing error is estimated to be 1.1 dB.Table I summarizes the comparisons of our work with the previously reported vertical waveguide transitions, demonstrating that the performance of the designed MWT is comparable with those of the reported structures in terms of the fractional bandwidth and the IL.

IV. CONCLUSION
This study presents the MWT employing partially covered metal and the via fence that resolves the inherent problems of the aperture coupling transitions, such as narrow bandwidth, backside radiation, and higher order modes leakage.By adjusting the position of vias and implementing a patterned metal cover, the MWT achieved wideband performance and low radiation loss.

Fig. 1 .
Fig. 1.Schematic of the transition.(a) Three-dimensional view of the transition consisting of three 0.035-mm-thick copper layers (L 1 , L 2 , and L 3 ) and the two dielectric layers (0.127-mm-thick D 1 and 0.276-mm-thick D 2 ).(b) Side view of the transition and the transmission modes.

Fig. 2 .
Fig. 2. Simulated transmission and reflection coefficients of the designed transition structure.

Fig. 3 .
Fig. 3. Electric field distribution along the MWT.(a) Without a metal cover and vias.(b) With metal cover but without via fence.(c) With cover and vias fence.(d) Comparison of the percentile radiated power for Cases 1-3.

Fig. 4 .
Fig. 4. Measurement setup.(a) Schematic illustrating the measurement setup, including the two MWT, the two 90 • bent board-to-waveguide connector, an E-TUBE, and the two end-launched connectors.(b) Optical micrograph showing the right-hand side view of the measurement setup in (a).(c) Fabricated MWT board and the interface between the MSL and the end-launched connector.

Fig. 5 .
Fig. 5. Simulated and measured transmission and reflection performance of the B2B measurement setup including a 1.0-m-long E-TUBE.
Manuscript received 7 August 2023; revised 20 November 2023 and 4 January 2024; accepted 30 January 2024.Date of publication 12 February 2024; date of current version 10 April 2024.This work was supported in part by the Regional Innovation Mega Project Program through the Korea Innovation Foundation funded by Ministry of Science and ICT under Grant 2023-DD-UP-0015, in part by the National Research and Development Program through the National Research Foundation of Korea (NRF) Digital Object Identifier 10.1109/LMWT.2024.3361918