Why Crates?

1
Why Crates?
Dave Doughty 7/23/03 Hall D Electronics Review
2
Plays to our Strengths/Experience

• JLAB has lots of experience with crates
• Successful deployment of large crate system in
  Hall B
• CODA handles this nicely
• JLAB has built numerous complex boards
• Trigger supervisor all three halls
• Trigger for Hall B level 2 uses FEE backplane
  for communication
• F1 TDC
• JLAB has experience with distributed pre-amps
• Hall B 35,000 pre-amps
• Some pain involving deterioration/access/replaceme
  nt
• Hall D Electronics group is fairly small

3
Simplifies Trigger Design

• Boards in crates leads naturally to crate sums
  of energy, tracks
• Use backplane for passing board sums to crate
  summer
• If backplane exists why not use for data
  concentration at the ROC?
• Each calorimeter crate of 256 channels needs 4
  fiber links for trigger
• Distribution of L1 Trigger signal (for
  hit/energy/feature extraction) is simpler

4
Simplifies DAQ/CODA Design

• Architecture of Hall D with crates is essentially
  that of Hall B
• Slightly larger in crate count
• CODA needs minimal changes
• Handle event blocking
• ROC-FEE change
• Handle distributed event building
• Managing crates and ROC doesnt change

5
Simplifies Power/Mechanical

• Only power to run all over is pre-amp power
• Low power
• Heavy-duty power (think FADC) now has tens of
  endpoints rather than thousands
• Mechanical support is automatic
• Except pre-amps which are light and small

6
Simplifies ADC/TDC Development

• F1 TDC developed for current use at JLAB will
  work with very minor modifications for Hall D
• Flash ADC development in VME or CPCI will be
  easier to test/debug