In this paper, a SIW (Substrate Integrated Waveguide) operating in the Ka–band is fabricated using a photosensitive glass substrate and measured. There is a disadvantage in the existing TGSV (Through Glass Silicon Via) technology that the process is...
In this paper, a SIW (Substrate Integrated Waveguide) operating in the Ka–band is fabricated using a photosensitive glass substrate and measured. There is a disadvantage in the existing TGSV (Through Glass Silicon Via) technology that the process is relatively complicated, and thus the manufacturing process cost could be increased. To compensate for this, the advantage of selective etching property that the photosensitive glass has was utilized for SIW production. In order to compare the transparent or translucent properties of the photosensitive glass substrate of the SIW, the dimensions of the SIW were designed based on those two properties. The L1 values, which are transition lengths optimized for transparent/translucent glass substrate characteristics, are 0.8 mm and 1 mm, respectively, and SIW was prepared to confirm both values. A via hole of the designed size was etched on the photosensitive glass substrate, and aluminum was deposited on the top/bottom metal layer and the sidewall of the via hole. There was a problem that aluminum was not well conformally deposited on the sidewall of the via hole, but it was solved by depositing copper with higher conductivity than aluminum using the sputtering technique. For the fabricated SIW with L1=0.8 mm, the measured 3-dB cut-off frequency is 20.5 GHz, insertion loss is 2.59 dB at 23.9 GHz, and 4.85 dB at 40 GHz, respectively. For the SIW with L1=1 mm, the measured 3-dB cut-off frequency is 20 GHz, insertion loss is 2.17 dB at 24.7 GHz and 5.28 dB at 40 GHz, respectively. Although the measurement results showed some similarities to the simulation results, it was confirmed that the overall insertion loss and return loss were worse than the simulation results. Since the simulation was performed using the substrate characteristics known at a low frequency of 1 MHz, it is believed that the losses became worse in the high frequency band due to the difference in dielectric constant and loss tangent. Therefore, it is expected that the loss can be improved by accurately measuring the dielectric constant and loss tangent in the high frequency band, optimizing the design of the SIW, and then depositing a thicker conductive metal.