Slide 0 - iNEMI

Development of cleanliness specification for lens-based TOSA/ROSA iNEMI F2F OFC/NFOEC 2012 Mar 5, 2012 iNEMI Confidential for member organization use only

Project Contributors •Glenn Victor •Doug Wilson & Tom Mictheltree •Tatiana Berdinskikh

•Mark Marino •Toshiaki Satake

•Ed Stuart •David Lach

iNEMI Confidential for member organization use only

1

Presentation Outline

•Background •Project Challenges •Different technologies of TOSA/ROSA •Synthetic Defects Experiments •Initial data on impact of contamination on optical performance of lens-based SFPs •Next steps


2

Introduction • No industry specification for lens-based TOSA/ROSA • Sumitomo collected data on OC-48 SFPs in 2005-2007 • iNEMI team combined modeling and experimental data (20082009) to study contamination effects on lens-based devices

– 10G SFP+ Sumitomo transceiver was used a test vehicle for the project – Paper has been published at OFC/NFOEC09 conference ( San Diego, CA, Mar, 09) • Further research is required for development of cleanliness specification of lens-based receptacles


3

Pluggable Optics – Common MSA Package Types

SNAP12 MSA Parallel Optic Device

4


4

Challenges •Multiple designs of lens-based transceivers in use –The collaboration efforts between different module manufactures, CMs and OEMs are required •Difficult contamination techniques – Manual process with low repeatability and reproducibility •Cleaning process of lens-based receptacles is challenging due to limited access to lens surface


5

ROSA/TOSA Technologies • Many different internal structures within OSA devices have been discovered

• The differences, amongst multiple suppliers, create a huge challenge for defining inspection and cleaning requirements

• Several of these designs are shown in next slides


6

Integral Lens Sample •This sample separates the LC cavity and endface from the integral lens surface.


7

Multiple lens Sample •This sample utilizes a “Contact Type” lens that makes physical contact with the LC end-face. •The lens floats, and is

not hermetically sealed within the OSA structure. •Cleaning of the contact flat lens is practical. iNEMI Confidential for member organization use only

8

Spherical Lens Sample

•This sample utilizes a full spherical lens. •Adjacent volume is open, creating an “air” interface for the installed LC

connector. •Distance from the LC aperture, to the leading surface of the “Ball Lens” is ~ 2.28mm


9

Semi-spherical Lens Sample •This sample utilizes a

semi-spherical lens, along with an “air” interface for an LC connector. •A lens is also present on

the plate glass surface of the TO-Canister. •The distance from the LC aperture to the leading lens surface is ~


10

Semi-spherical Lens Close to Aperture Sample •This sample represents a structure where the semi-spherical lens surface is near the LC aperture.

•Open volume surrounds the lens •The distance from the LC aperture to the leading lens surface is ~

1.31mm.


11

Sophisticated Structures Sample •Sophisticated structures, and optical components, other than lenses, may reside. •This structure includes

what is believed to be an isolator, within the optical path. •As the isolator emerging surface is angled,

cleaning and inspection


12

Optical Subassembly Construction •TO can assembly into OSA housing •Photodiode / LED / laser position inside TO can

•Fiber into OSA housing •Theoretical path of light


13

Transceiver De-construction for Testing

SFP transceiver with lens-based transmitter and receiver

14


14

Transceiver de-construction for testing • Cutaways made in metal housing to expose lens without affecting the optical transmission

• Facilitates easy contamination and re-cleaning. • Maybe re-assembled back into the module for performance testing.

15


15

Receiver signal loss with contamination, dB Images at (200x)

Clean Lens

16

Conductive Foam

Toner

-1.54

-2.57

-1.80

-6.40

-1.04

-4.30


16

DOE for Controlled Blockage Using Synthetic Defects •Measure output power with no interference •Use microscope to measure active area of lens •Capture image of synthetic defect at each pre-defined

measurement location (X and Y) with XMTR ON and OFF •Reconnect to power meter, move to each XY location, and measure output power •Plot output power in 2D intensity graph for analysis •Repeat for each defect type


17

Synthetic Defect Patterns • Patterns block the XMTR output to varying degrees based on the defect pattern and its location within the part • Several defect types have been designed – opaque circles of diameters (50-600 um) – opaque circles with holes, “donuts” (50-450 um holes) – Lines >500 um long with widths of 50-600 um – Clusters of randomly distributed opaque circles with varying density and total occluded area


18

Photos of Mylar Film Synthetic Defects

• Opaque Circles 50 m

200 m

450 m

50 m

200 m

450 m

• Donuts


19

Mylar Strip with “Donuts” in TOSA Device • Slotted TOSA device provided means to insert synthetic defect

strips • Initial sample of Mylar film with donut shapes • Final version will have supports to enable repeatable motion of the strip within the slot • Microscope can image the defects with or without the transmitter

enabled • Output power can be measured by iNEMI Confidential for member organization use only

20

Initial Images of Synthetic Defects • Transmitter was ON, which is source of circle of light. • Top row is 265 um defect, lower row defect is ~540 um in dia • Illuminated area is ~600 um in diameter, as viewed through the aperture in device. • Size of defects and transmitter circle are both increased by same proportion, most likely correct for separation distance from

device. • Roughness at edge of circle is defocusing of the sharp edge of the aperture


21

TOSA Focus Journey • Movie not included in this version


22

TOSA Synthetic Defects • Movie not included in this version


23

Preliminary Power Readings • Quick readings were made of output power as a synthetic defect strip was randomly move around

• Observed about 20-25 dB decrease in output power as large circle was moved into the main path

• While not a controlled experiment, it was encouraging to see the defect quench the output as expected • Next step is carefully controlled motion with quantitative measure of occlusion to match with output power measurements


24

Next Steps • Collect additional images and associated optical performance data on contaminated samples

• Conduct controlled synthetic defect experiment to quantify relationship between occluded area and optical performance

• Optical performance metrics include output power loss and receiver loss


25

www.inemi.org Email contacts: Bill Bader [email protected] Bob Pfahl [email protected]
