LHC BI fibre infrastructure E Calvo BEBIQP 12102015

LHC BI fibre infrastructure E. Calvo (BE/BI-QP) 12/10/2015

Aim and Outline Aim: The objective is to evaluate if the exchange of information and the way BE/BI and EN/EL collaborate can be optimized. Outline: Calvo (BE/BI) E. E. Calvo (BE/BI) • • BI fibre infrastructure overview Context Proposal Needs, Wants, Likes and concerns… 2 / 11

BI fibre infrastructure overview : LHC: SR 5 SX 4 SR 6 SR 3 SR 7 CCR SR 2 SR 8 Calvo (BE/BI) E. E. Calvo (BE/BI) SR 1 Initially, proposed for the BPM system for transmitting the position signals from the front-end cards (in the tunnel) to the acquisition cards (in surface) ~2176 operational links @LHC ~1088 spares Very early, extended to the BLM system, with a very similar distribution. ~1600 operational links And to the BST system: Broadcast distribution of timing signals from the CCR towards many surface and underground LHC and SPS locations. 1: 16 splitters in surface buildings >200 operational links more 3 / 11

BI fibre infrastructure overview : LTI & SPS TI 2 & TI 8: 867 866 BPM as in the LHC: ~200 operational links 865 SPS : CCR BA 3 BA 2 BA 4 BA 1 BA 5 BA 7 BA 6 Very old infrastructure for distributing timing signals (BST): ~24 operational links The new fibre distribution for the BPM system (MOPOS) ~432 operational links And future BLM? Calvo (BE/BI) E. E. Calvo (BE/BI) Prevessin: Distribution of timing and signals between the CCR and BI laboratories in the CERN Prevessin area. 4 / 11

Context and a bit of legacy • During LHC installation, the termination patch cords (tunnel and surface) were handled by BI (labelling, installation, etc. ) → Mapping (documentation of which signal is on which panel/connector, available spares, etc. ) was generated and maintained by BI. • Mini-tubes came with a standard number of fibres → BI had many spares. • Every small request from new BI users was tried to be fulfilled within the BI available infrastructure → Faster, cheaper, more efficient. Calvo (BE/BI) E. E. Calvo (BE/BI) • The number of BI users increased with time: Ex: - DIDT and MIM (in UA 47), Prepulses (in Pt. 2 & 8), TTC signals for BGV (UA 43), Diamond detectors in (Pt. 2, 4, 5, 8), 400 MHz for Schottky (Pt. 4), White rabbit for MIM, BRAN signals (Pt. 1), Ethernet links for DOROs (Pt 1 & 7), etc. 5 / 11

Current situation Calvo (BE/BI) E. E. Calvo (BE/BI) • Frequently, new requests can not be fully route with existing BI infrastructure → Part of the routing is requested to EN/EL → Documentation is split between both teams and routing requires several iterations. • The documentation, initially organised for BPM and BLM systems, is not appropriate for other users (multiple racks). • Information is distributed through several files: surface rack, tunnel racks, patch cords labels, etc. • Should EL-CF and BI continue having a BI link person ? • Is this splitting of information the most efficient way of managing the BI fibres? 6 / 11

Proposal • Integrate BI fibre infrastructure in the EN/EL fibre database (or documentation system). Calvo (BE/BI) E. E. Calvo (BE/BI) So that… 1) BI users contact directly EN/EL-CF 2) EN/EL-CF can completely route the new requests (using BI infrastructure when possible) 3) EN/EL-CF keep the information updated, complete and in a standard way. Not important differences in the way of working for EN/EL … but an important effort in - understanding the current BI documentation - and (eventually) adapting it to EN/EL standards 7 / 11

Needs, Wants & Likes (I) BI needs: ü Be able to connect/unconnect cards at the link ends for maintenance/upgrade/test purposes. Usually <10 links per TS ü Flip patch cords for disentangle tx/rx issues. BPM can verify correct routing. (BST also can). This is not the case for other systems. ü Access to the database/documentation system. Know about links routing and links availability. BI does not need: Calvo (BE/BI) E. E. Calvo (BE/BI) ü EN/EL “Piquet service” for patching or repairing links. ü EN/EL particular patch cord store. (BI wishes to keep its small local stock of patch cords as till now). BI wishes: ü Access to additionall information about the link : length, full attenuation, number of connectors, OTR data ü Update information in the database: Ex. routing information, power @rx. 8 / 11

Needs, Wants & Likes (II) CF team needs and/or concerns ? Ø Clear and complete documentation: BI has different levels of documentation quality: - Recent links follow EN/EL standards (visio templates, with panel names, connector types, patch cords lengths, etc) + photos (notes) - BPM, BLM and BST links @ LHC are mainly rack layouts and patch cord excel tables. https: //edms. cern. ch/project/CERN-0000115916 Calvo (BE/BI) E. E. Calvo (BE/BI) - BLM has also all the links in the layout database. Old BST links @ SPS have an excel table with building and rack numbers (there is not information about repeaters, optical attenuators or available spares). If the proposal is retained/considered, first steps will be study the documentation and probably make a series of visits to the facilities. 9 / 11

Needs, Wants & Likes (III) CF team needs and/or corcerns ? (cont. ) Ø Teams should negotiate and stablish a rigorous procedure to keep the documentation up to date. Methods are diverse: direct access to the data base, email, some kind of “DIF” form (“DMF”: Demande de Modification de Fibres? ), etc Calvo (BE/BI) E. E. Calvo (BE/BI) Something else? 10 / 11

/ 11

Spare slides / 11

Calvo (BE/BI) E. E. Calvo (BE/BI) Example of SX 4 13 / 11

Calvo (BE/BI) E. E. Calvo (BE/BI) Ex. of SR 1 star point rack 14 / 11

Calvo (BE/BI) E. E. Calvo (BE/BI) Ex. Diagrams for new BI users 15 / 11

Kind of connectors used Calvo (BE/BI) E. E. Calvo (BE/BI) BPM, BLM and BST panels @ LHC FC/PC: Some old patch panels @ SPS E 2000/APC 90% ST: BST rx @ LHC LC: Rx @SPS new MOPOS and growing community of new systems 16 / 11

Calvo (BE/BI) E. E. Calvo (BE/BI) BLM fibres in the Layout DB 17 / 11

Calvo (BE/BI) E. E. Calvo (BE/BI) CO rack configuration tools 18 / 11