PAGE CONTENTS
Objectives
Large Scale Optical Cross Connect
The development consists of a MEMS-based Optical Cross Connect (OXC) with about 50 input and 50 output fibres. One important application of this OXC is an enabling component for a reconfigurable switching network within a communication satellite. The 2 year objective was to deliver a breadboard to prove the soundness of the concept and to demonstrate a working MEMS-based OXC.
An important longer-term objective is to use the knowledge developed in this program to create a Swiss industrial source for MEMS-based OXCs.
The following schematic shows the general concept of the complete switching system. Input and output of the system is in the RF domain, while the core switching is in the optical domain. The RF signal are converted to optical signals prior to the actual optical switching. After the optical switching the are re-converted into the RF domain. The optical switching makes the overall system potentially very compact at very low cross talk and low insertion loss levels.
Challenges
The key issue in the project is to develop a 2D electrostatic actuated MEMS micromirror array of 128 mirrors which is extreme stable in position over time to avoid heavy and complex feed back loop. Another task is to develop the necessary high port count collimator. And last but not least a training strategy and software for the switch matrix.
Plan
| Work Package | Expected Duration | Main Actor | Other Actors |
| WP1000: Assessment of large MOEMS switch architecture | 2 Months | Alcatel Space | Sercalo, IMT, EPFL |
| WP1100: trade-off architectures | 2 Months | Alcatel Space | IMT, EPFL |
| WP1200: Conceptual design | 2 Months | Sercalo | IMT, EPFL |
| WP2000: Preliminary Design of breadboard | 2 Months | Sercalo | IMT, EPFL |
| WP3000: Detailed design of breadboard and test set | 14 Months | Sercalo | |
| WP3100: Detailed design of breadboard | 14 Months | Sercalo | IMT, EPFL |
| WP3200: Detailed design of test set | 3 Months | Alcatel Space | EPFL, Sercalo |
| WP4000: manufacturing | 15 Months | Sercalo | IMT, Alcatel Space, EPFL |
| WP4100: MOEMS manufacturing | 15 Months | Sercalo | IMT, EPFL |
| WP4200: test bed manufacturing | 5 Months | Alcatel Space | Sercalo |
| WP5000: Test campaign | 3 Months | Alcatel Space | Sercalo |
| WP6000: Appraisal | 1 Month | EPFL | Alcatel Space, Sercalo, IMT |
Current Status
Completed. We could successfully demonstrate the function of our design and deliver a demo switch system. The testing on the MOEMS switch breadboard and tests-bed was performed on a 10×10 sub-matrix selected among full connected optical inputs and outputs.
| Requirement Object | Target Spec. | Comments |
| RF Requirements | ||
| Frequency band | Ka-band | OK alsofor S band |
| RF Input Power | >0 dBm | OK fr Ka and S band |
| Opto-microwave gain | 0 dB | OK for Ka and S band |
| Opto-microwave noise figure | <28 dB | OK for Ka band |
| C/I ratio (2 tones) | >50dBc | OK for S band, not for Ka band |
| Return Loss | >20 dB | OK for Ka band |
| Static Isolation | >50 dB | >70 dB |
| Requirement Object | Target Spec. | Comments |
| System Level Requirements | ||
| Number of I/O ports | >50 x 50 | Partial 24×24 Full connectivity 10×10. |
| Switching time | < 5 ms | 65 ms |
| Time for reconfiguration | < 50 ms | 100 ms |
| Mass | < 1kg | 0.86 kg |
| Size | < 1000 cm3 | 1400 cm3 |
| Power consumption | < 5 W | 6.01 W |
| Optical Requirements | ||
| Wavelength range | 1.25–1.63µm | 1.25-1.63 µm (measurements performed @ 1.55µm) |
| Insertion loss absolute value | < 7dB | < 6.2 dB |
| Insertion Loss variation | TBD | < 0.5 dB |
| Repeatability of loss | < 1 dB | < 0.06dB |
| Stability of loss | < 1 dB | < 0.9 dB |
| Max Power | 21 dBm | |
| Return Loss (back-reflection) | > 30 dB | 15 – 30 dB |
| Isolation (crosstalk) | > 50 dB | > 70 dB |
| Polarization dependent loss (PDL) | < 0.1 dB | < 0.17 dB |
The results obtained during the project encouraged sercalo to start a new project on single 2D analogue driven tilt mirrors for ground based applications, like tunable filters, micro scanners or low cost spectrometers and to commercialize a 25 port fiber collimator for rotary joint applications. Future work will tackle environmental testing at chip and equipment level, yield of MEMS chips, etc.
