HCAL Cable Plant Status Report

1
HCAL Cable Plant Status Report

• Rochester HCAL Meeting
• October 19, 2001
• J. Whitmore
• FermiLab
• Personnel Involved J. Elias and J. Whitmore

2
Overview

• Requirements
• Options
• Cable
• Fiber-Optic Connector
• Cost
• Issues
• Summary and Plans

3
Requirements

• Dont Screw up the Physics!
• Map layers into towers
• Organize HTR inputs to yield trigger primitives
• Need adjacent phi and eta segments
• Deal with 3 QIE channels/fiber
• Deal with overlap regions
• Keep cost down
• No for patch panel!
• Every extra connection costs

4
Requirements

• Fit into minimal space
• No room at RBX
• Small connectors (but must be robust)
• Tight bend radius (7 cm)
• Must be rad tolerant
• Components must withstand 4E11 n/cm2 (330
  rads)
• No fiber darkening
• Must meet CERN Safety Standards
• ULSZH (univ. low smoke zero-halogen)-no PVC
• No ferromagnetic materials
• Limits companies that can supply

5
(No Transcript)
6
Cable

• Cable requirements
• Able to transmit 1.6 Gbps over 90 m with low
  dispersion
• 50/125 mm rather than 62.5/125 mm
• Very little darkening of fiber up to 3E11 n/cm2
  and 1kRad TID
• Corning (CMS ECAL) and Fujikura (Atlas)
• No thermal or moisture issues (or rodents!)

7
Cable

• Loose Tube Cable
• Rugged
• Good for long pulls
• Need to ribbonize at connector
• Ribbon Cable
• Samsung

8
Fiber-Optic Connectors

• Connectors at RBX
• Multi-fiber connectors
• RMs
• 6 fibers (3 FE boards w/ 2 fibers each)
• Single-fiber connectors
• CCM and Calibration units
• Also need spares
• Connectors at HTR cards
• Prefer multi-fiber for easier pull
• Mapping is easier with single-fiber

9
Multi-fiber Connectors

• MT/MPO Connectors
• MPO/MTP connector
• MT ferrule housing
• Keyed
• Spring load, locking connection
• MPO Adapter
• Guide pins
• MT ferrule
• Can be used without strain relief
• Not keyed
• Guide pins
• Used by CMS tracker folks
• Rad tolerant

10
MT Ferrule

• Advantages
• Small form factor
• Accommodates 4-12 fiber
• Disadvantage
• Clip rather than adapter
• Not as robust as MTP/MPO
• No strain relief
• Limited suppliers Sumitomo, Fujikura, Furukawa,
  US Conec

11
LC Connector

• Advantages
• Small form factor
• Allows for mapping at HTR
• Disadvantages
• Bulky cable pull

12
RBX-HTR Cable Assembly
13
RBX-HTR Cable Assembly
14
Pigtails

• LC Pigtails
• CCM and Calibration module
• MT Pigtails
• RMs

15
Cost

• HCAL System
• 140 RBXs (36 each for HB, HE, HO 22 for HF)
• RBX-HTR Cable Assembly - 2k/cable
• 300k (including spares)
• RBX pigtails
• ??? (maybe as high as 150k)
• Installation
• Testing
• ? 500k for system

16
Issues

• Installation
• Cable detector and pull to electronics room
• 36 LC connectors Difficult to pull
• Cleave and fusion splice in-situ
• Added cost
• Either way, we need to retest fibers after
  installation

17
Issues

• Mapping
• Ideally want non-adjacent channels at FE board
• But problem is best solved top-down not bottom-up
• Light Budget
• Need to measure acceptable attenuation using GOL

18
Summary and Plans

• There are some options (though we do have a lot
  of constraints)
• Need to close this out or we risk losing the
  attention of the vendors
• Need to measure light budget
• Meeting at CERN in November
• Best option may be to attack problem for the
  entire experiment
• Lots of common issues