Since BrightAWGs in Lionix 13 run does not reach our goal, AWGs are redesigned for Lionix 14 run by BRIGHT Photonics to improve their performance. This project aims to characterize those AWGs from Lionix 14 run, find out optimal design parameter setting and propose a new test plan for the next run. Firstly, AWGs from Lionix 14 run is measured at Technobis. Measurement system is evaluated and specification range for characteristics of interest is recommended from measurement system gage study. Then, important characteristics from all AWGs are analyzed and some important design parameters are specified. Finally, a new test plan is proposed for new run with design of experiment theory, including new design setting and new concept AWGs. Crosstalk of AWGs in Lionix 14 run reaches up to 30±5dB, much better than the ones in Lionix 13 run with <5dB. Sidelobe correlates with size of AWG, mostly because of array arm length. Array arm width near free propagation region is a confidently critical design parameter and array arm width on the bending is also an influencing design parameter. In addition, it is found that wavelength shift depends on the wafer process and unable to determine any critical parameter for channel spacing shift possibly due to the incapable measurement system. In conclusion, new design for Lionix 14 reaches better performance than Lionix 13, but more potentially critical design parameters and new concept AWG are still needed to be tested for the next run. In addition, through the whole procedure, some suggestions to improve the work flow are given for all parties involved in this project.
|Award date||10 Dec 2015|
|Place of Publication||Eindhoven|
|Publication status||Published - 2015|