The objective of this research is to: 1) Survey commercially-available and development-level materials in three categories: Laminates, Glass and Ceramic (primarily LTCC) and benchmark them based on their electrical properties and processability for their suitability for 5G applications, and 2) Achieve transmission line loss of less than 0.05 dB/mm on these substrates, and 3) Perform a model-to-Hardware correlation study. The first task is being targeted by extensive paper study and collaboration with partners of PRC to extract properties of new materials. The second task is focused on achieving transmission line loss of less than 0.05 dB/mm with 2% linespace tolerance on selected substrate core and buildup layer combinations for 5G applications. The third task focuses on patch antennas at 5G frequencies as test structure to study the correlation of model and hardware and how the process variation affects the response of the test structure. Fifth generation communication systems, abbreviated as 5G, allow much greater data rate capacity than 4G or LTE and more number of mobile broadband users per area. The target bands are 28 GHz (US) and 39 GHz (EU) along with 60 GHz. A new class of substrate technologies with embedded actives, advanced antennas, transmission lines with ultra-short interconnection lengths to actives are needed to realize such 5G systems. Glass presents a remarkable opportunity for 5G due to its low loss, superior dimensional stability, ability to form fine-pitch through-package-vias (TPVs), surface smoothness, stable dielectric constant and low loss tangent, and the advantage of easier panel-scale processing. Innovative 3D structures with fan-out or back-side assembly are developed to harness these benefits of glass, as illustrated. Advanced transmission line structures (Microstrip, CPW, CBCPW and SIW) and antenna structures (patch array, Yagi-Uda and horn) are explored with these 3D glass structures to achieve insertion loss of less than 0.05 dB/mm. A library of substrates is created after an extensive paper study which are suitable for 5G applications. Out of those substrates, three combinations of core and buildup materials are recommended and focused on for further research. A test vehicle for transmission line loss is fabricated on glass substrate with a thin double-sided buildup layer. The results are compared with models on two other substrates and a comparison is performed. For the model-to-hardware correlation study, rectangular patch antennas at four different resonant frequencies are fabricated on an ultra-thin glass substrate. The results are compared with the simulation and the discrepancy between the two is explained in terms of changes in electrical properties of glass at higher frequencies and small dimension variations during the process.
Cite this work
Researchers should cite this work as follows:
MATIN Development Team