1. 16 Sep 2008 francois.vasey@cern.ch0 Versatile Link Status Report F. Vasey on behalf of the project steering board With input from C. Issever J. Troska J. Ye A. Prosser

2. 16 Sep 2008francois.vasey@cern.ch1 Versatile Link Project Description Optical Physical layer linking front- to back-end Optical Physical layer linking front- to back-end Bidirectional, ~5Gbps Bidirectional, ~5Gbps Versatile Versatile Multimode (850nm) and Singlemode (1310nm) versions Multimode (850nm) and Singlemode (1310nm) versions Point to Point and Point to Multipoint architectures Point to Point and Point to Multipoint architectures Phase 1: Proof of Concept based on tentative specifications Phase 1: Proof of Concept based on tentative specifications

3. 16 Sep 2008francois.vasey@cern.ch2 Workpackage Allocation oRx oTx oRx oTx oRx oTx oRx oTx oRx oTx oRx In detector oTx oRxoTx oRx In detector oTx oRx oTx oRx splitter oTx oRx oTx oRx oTx oRx In detector P2P System PON SystemBoBW WP1.1 WP1.2 WP1.3 Front End Components VTRx WP2.1 WP2.2 WP2.3 Back End Components MRx PON TRx HTx Passive Components Fibre mux split PON TRx split Project Management CERNFermilabOxford SMU

4. 16 Sep 2008francois.vasey@cern.ch3 Deliverables and Timescales Phase 1: Proof of Concept based on tentative specifications Deliverables: Deliverables: a tentative specification based on past experience a tentative specification based on past experience a non-exhaustive portfolio of components meeting (even partially) the tentative specification (SM and MM, P2P and P2MP) a non-exhaustive portfolio of components meeting (even partially) the tentative specification (SM and MM, P2P and P2MP) a front end TRx prototype package a front end TRx prototype package test bench(es) for components and systems test bench(es) for components and systems preliminary irradiation test results preliminary irradiation test results preliminary functionality test results for components and systems preliminary functionality test results for components and systems a set of recommendations for phase 2. a set of recommendations for phase 2. Timescales: Timescales: 18 months for phase 1 (Sep 09) 18 months for phase 1 (Sep 09) Phase 2 to be defined after 12 months (Apr 09) Phase 2 to be defined after 12 months (Apr 09) Phase 2: Feasibility Study Phase 2: Feasibility Study Detailed specifications based on user requirements Detailed specifications based on user requirements Components and Variants Shortlist Components and Variants Shortlist VTRx package definition and fabrication VTRx package definition and fabrication Systems and Architectures Demonstration Systems and Architectures Demonstration Phase 3: Production Phase 3: Production Specification freeze Specification freeze Market Survey Market Survey Pre-Production Pre-Production Qualification Qualification Production Production

5. 4 a) VTRx Customisation Bi-directional Module with connector interface Bi-directional Module with connector interface Based upon an acknowledged standard Based upon an acknowledged standard Work with Industrial partner early-on Work with Industrial partner early-on Low Mass & Volume Low Mass & Volume Minimize material, avoid metals Minimize material, avoid metals Non-magnetic, capable of operating in a magnetic field Non-magnetic, capable of operating in a magnetic field Requires replacement of ferrite bead used in laser bias network Requires replacement of ferrite bead used in laser bias network MM 850nm & SM 1310nm versions MM 850nm & SM 1310nm versions Work-plan Work-plan 1. Identified potential industrial partner 2. Test commercial parts of the type we would like to customize (SFP+) ✓ Evaluated standard 850nm and 1310nm SFP+ modules in early 2008 ✓ Evaluated VCSEL-based 1310nm SFP+ modules in mid-2008 3. Customize package to meet HEP need 4. Ask for CERN-specified parts to replace standard ones Laser Driver, TIA Laser (VCSEL or EEL), PIN photodiode WP2.1 status report, CERN VTRx

6. 5 Major effort has been put into evaluating commercial SFP/+/XFP transceiver modules Major effort has been put into evaluating commercial SFP/+/XFP transceiver modules Functionality testing broken down into three parts Functionality testing broken down into three parts Power Consumption Power Consumption Transmitter portion Transmitter portion Receiver portion Receiver portion Testing of currently available modules has led to starting to define detailed performance specifications & methods Testing of currently available modules has led to starting to define detailed performance specifications & methods For details, see presentation by Luis Amaral For details, see presentation by Luis Amaral 1310nm VCSEL-based SFP+ @ 5Gb/s < 600mW/module Magnetic-field tolerant bias-T @10Gb/s Further signal integrity improvements possible b) Functionality Testing

7. 6 Very Similar overall trend Very Similar overall trend ROSA (solid symbols) not much worse than bare PINs ROSA (solid symbols) not much worse than bare PINs Several orders of magnitude difference in response between devices Several orders of magnitude difference in response between devices BER independent of Data- rate BER independent of Data- rate Burst Errors observed Burst Errors observed max. 10-bits long in PINs max. 10-bits long in PINs max. 00’s bits long in ROSAs max. 00’s bits long in ROSAs Error correction mandatory Error correction mandatory Plans: Plans: Total Fluence testing of Laser, PINs and possibly TIAs and Laser Drivers in early 2009 Total Fluence testing of Laser, PINs and possibly TIAs and Laser Drivers in early 2009 Singlemode 60MeV protons at PSI Flux 10 8 p/cm 2 /s (1-2 orders of magnitude above SLHC Tk/Pix) Multiple device types from multiple Vendors, also with integrated TIA (ROSAs) c) Radiation Testing SEU

8. 16 Sep 2008francois.vasey@cern.ch7 a) Radiation tests of passive components Irradiated up to 700 kGy Irradiated up to 700 kGy 4 multi-mode (MM) fibres @ 850nm 4 multi-mode (MM) fibres @ 850nm 1 single-mode (SM) fibre @ 1310nm (SMF28) 1 single-mode (SM) fibre @ 1310nm (SMF28) MM fused taper splitter MM fused taper splitter SM PLCC splitter SM PLCC splitter MM and SM LC-LC connectors MM and SM LC-LC connectors Preliminary results (details see T. Huffman’s talk) 2 MM fibres and 1 SM fibre candidates Survey of market for more MM fibres will continue Need more splitter tests Connector candidates identified Overall: Very good progress made. Overall: Very good progress made. WP2.3 status report, Oxford Fibre mux split

9. 16 Sep 2008francois.vasey@cern.ch8 MM fiber irradiation test Ericsson 1 Ericsson 3 Infinicor Ericsson 2

10. 16 Sep 2008francois.vasey@cern.ch9 b) Tests of mechanical properties of fibres started Until now un-irradiated SMF28 Until now un-irradiated SMF28 Next test irradiated fibres Next test irradiated fibres

11. 16 Sep 2008francois.vasey@cern.ch10 c) Plans for the next six months Continue market surveys Continue market surveys fibres fibres splitters splitters Prepare next gamma radiation test Prepare next gamma radiation test fibres at low and high dose rates fibres at low and high dose rates Splitters  passive tests Splitters  passive tests Setup environmental tests for fibres, splitters and connectors Setup environmental tests for fibres, splitters and connectors Continue mechanical tests Continue mechanical tests

12. 16 Sep 2008francois.vasey@cern.ch11 WP1.1 status report, SMU Deliverables in phase 1: Deliverables in phase 1: Development of architecture demonstrators for P2P network on commercial transceivers. Development of architecture demonstrators for P2P network on commercial transceivers. Development of test bench(es) for components and systems Development of test bench(es) for components and systems Development of test procedures. Development of test procedures. The status: The status: SFP+ is chosen as candidate for Versatile Link standard (by Versatile Link group). SFP+ is chosen as candidate for Versatile Link standard (by Versatile Link group). SFP+ carrier board has been designed and fabricated. With a loop-back (oTx to oRx via fiber) configuration, a 10 Gbps optical serial link has been demonstrated with AFBR-700SDZ (10Gb, 850 nm) from Avago. SFP+ carrier board has been designed and fabricated. With a loop-back (oTx to oRx via fiber) configuration, a 10 Gbps optical serial link has been demonstrated with AFBR-700SDZ (10Gb, 850 nm) from Avago. Equipment is in place for 10 Gbps tests: Equipment is in place for 10 Gbps tests: 12Gbps BERT. 12Gbps BERT. 20GHz real-time scope with 8GHz differential probe. 20GHz real-time scope with 8GHz differential probe. O/E module with 12GHz bandwidth. O/E module with 12GHz bandwidth. Sampling scope with 10GHz optical, 50 GHz electrical input modules. Sampling scope with 10GHz optical, 50 GHz electrical input modules. New collaborator (IPAS) has been added to this work package. New collaborator (IPAS) has been added to this work package. The plan for the next 6 months: The plan for the next 6 months: With equipment in lab, and commercial transceivers, a full P2P network of 10 Gbps will be demonstrated. With equipment in lab, and commercial transceivers, a full P2P network of 10 Gbps will be demonstrated. With Stratix II GX programmed as the BERT (signal generator and bit error rate checker), a portable 6 Gbps test bench will be developed. With Stratix II GX programmed as the BERT (signal generator and bit error rate checker), a portable 6 Gbps test bench will be developed. Tests will be carried out with the 12 Gbps BERT and with Stratix II GX based test bench. Testing procedures will be studied and defined through this process. Tests will be carried out with the 12 Gbps BERT and with Stratix II GX based test bench. Testing procedures will be studied and defined through this process. oTx oRxoTx oRx In detector P2P

13. 16 Sep 2008francois.vasey@cern.ch12 The SFP+ carrier board for 10 Gbps signal Top Layer Bottom Layer Ground Layer Power Layer

14. 16 Sep 2008francois.vasey@cern.ch13 Preliminary test results Eye diagram tests at 3, 5, 8 and 10 Gbps. Eye diagram tests at 3, 5, 8 and 10 Gbps. BER tests at 5 and 10 Gbps. BER is found to be bettet than 3E-14 at 10 Gbps. BER tests at 5 and 10 Gbps. BER is found to be bettet than 3E-14 at 10 Gbps. The board works at 10 Gbps, but there are issues that needs more careful studies. The board works at 10 Gbps, but there are issues that needs more careful studies. 3 Gbps 10 Gbps 5 Gbps 8 Gbps Bathtub curve at 5 GbpsBathtub curve at 10 Gbps

15. 16 Sep 2008francois.vasey@cern.ch14 Fermilab Has Joined the Project Work Plan Submitted and Approved Presentation at Project Meeting (14-9-2008) 3 Month Plan Document and Communicate Test Strategy Components Procedures Equipment 6 Month Plan Procure Samples of Selected Components Test Lab Configuration Begin Testing Following Documented Procedures WP2.2 status report, Fermilab MRx PON TRx HTx PON TRx

16. 16 Sep 2008francois.vasey@cern.ch15 Conclusion Phase 1: Proof of Concept based on tentative specifications Status after 6 months: Status after 6 months: a tentative specification based on past experience a tentative specification based on past experience a non-exhaustive portfolio of components meeting (even partially) the tentative specification (SM and MM, P2P and P2MP) a non-exhaustive portfolio of components meeting (even partially) the tentative specification (SM and MM, P2P and P2MP) a front end TRx prototype package a front end TRx prototype package test bench(es) for components and systems test bench(es) for components and systems preliminary irradiation test results preliminary irradiation test results preliminary functionality test results for components and systems preliminary functionality test results for components and systems a set of recommendations for phase 2. a set of recommendations for phase 2.