Rack Layout

Murrough Landon - 14 March 2002
http://www.hep.ph.qmul.ac.uk/landon/talks

Overview

• Rack Allocation Meeting
• Cables and latency
• Receiver/PP layout and cabling?

Proposed Niveau 1 and Niveau 2 Layout

• Trigger rack layout presented (and accepted) at the recent rack allocation meeting.
• CTP is now very central (NB CPM/JEM racks swapped to match - compared to Philippes document).
• Our allocation includes the requested gaps and spare racks.

Cables and Latency

Cable Lengths

• Signal cables (holes to outer Receivers): 7m
  (NB plus length from detector to holes)
• Receiver to PPMs: 5m
  (assumes rack layout shown and most direct route across the front of the racks)
• LVDS cables from PPMs to CPMs/JEMs: 11m
• InterCMM cables: 3m
  (assuming under the floor, 1m via hole between racks)
• CMM to CTP: 7m
• Total: 33m
• More details at:
  http://www.hep.ph.qmul.ac.uk/landon/atlas/racks
  http://www.hep.ph.qmul.ac.uk/landon/atlas/racks/racklayout.html

Receiver and PPM Layout

RX/PP Rack Layout

• All layouts have shown two receiver racks next to two PP racks.
• If we abandon any attempt to have PP crates arranged in phi quadrants (simplifies the RX-PPM cabling as Paul would like) then why not intermingle RX and PP racks?
• Easiest if PP crates (mostly) follow layout of RX crates.

``Octopus Cables'' or Patch Panels?

• We are supposed to document our cabling, including that between Receivers and PPMs.
• How should we handle the awkward cables in the barrel/endcap transition regions in EM and hadronic layers?
• If implemented as cables we need 72 cables with three ``ends'' and 8 cables with nine ``ends''! (Plus 360 standard cables).
• 5U space above each crate was requested for cabling (64 1.3cm diameter cables per crate, bending radius 10cm)
• Patch panels would need 3 connectors per slot per rack: 6U?
• Present layout has 8U spare per rack at the bottom.