802.3ca Task Force,
At yesterday’s task force meeting in Whistler, Chris Cole (Finisar) presented slides from SiFotonics on 25G APD performance. The group had several questions which were forwarded to SiFotonics. Below are the answers. If you have any questions,
you can contact Dong Pan (dp7u@xxxxxxxxxxxxxx) directly.
· Is there a manufacturing distribution spread for the sensitivity data?
Answer: The advantages of Silicon is high uniformity. We have very large database from 10G APD to show very uniform sensitivity. For 25G APD, We have built <100pcs ROSA, the sensitivity range is from -23.5 to -21.5dBm, I think this range is related to package
capability and light source conditions.
· How is the APD packaged in the experiments?
Answer: We tried both BOX type and TO-can type experiments, BOX present better performance at low BER (1e-12) but TO-can can perform similar good performance at high BER (1e-3)
· What is the fundamental noise contributor to the BER curve?
Answer: TIA noise and shot noise from photo current
· Was the Tx source a EML or DML? (Group believes it is EML based on ER.)
Answer: Tx source is EML in this presentation. Chris Cole of Finisar presented at Macau results of SiFotonics 25G APD performance using DML.
Thanks,
Shawn M. Esser
Director of Product Management
CATV and Endurance
Office: +1 267-803-3754
www.finisar.com
