Testing L3 muon unpacking online

1
Testing L3 muon unpacking online

• Paul Balm
• collaboration meeting, Oct 2000

2
Testing of last weeks

• Scintillator data from cosmics
• 3 crates
• What did L3 do
• First actual look at the data!
• Get the crate/module ID from the data, from the
  geometry obtain pos. of hit in D0 coordinates
• Pass/reject based on hits in event
• Test was successful
• some events rejected (no hits -only trigger
  crate)
• some events successfully sent to Collector/Router

3
Tests remaining

• Repeat this test for PDT data
• currently there is a problem of synchronization
• no idea when this is fixed (being worked on)
• then we can also try our tracking
• Sync bits were not tested on in last weeks test
• that test must be put in the code (MF)

4
Requests for muon commissioning

• Coincidence filter
• Pass if there is coincidence between specific
  scintillators
• will greatly reduce noise (now gt90 noise!)
• Implementation
• using the run config only will be too coarse
  (module level)
• will need RCP config - need to think
• how to list all scintillators in RCP
• how to discriminate internally in filter
• finding the time to do it is a problem

5
Proposal - v0.0

• Examples of coincidences requested
• SFE 22/Chan 19 with OR of (SFE 03/Chans 0-4)
• SFE 22/Chan 35 with OR of (SFE 03/Chans 16-20)
• SFE 23/Chan 03 with OR of (SFE 03/Chans 32-36)
• Use one filter for each type (1/line above)
• Config file (RCP)
• This scheme only allows 2 OR-terms between 2 SFEs
  per filter, and either one channel per SFE or a
  range. Filters can also be OR-ed
• Is that flexible enough?

TrigSFE 22 TrigSFEChanList (19) DataSFE
3 DataSFEChanList (0,4)
6
Conclusion

• Learned from past test
• geometry has to be initialized, stream numbers,
  start thinking about non-existent crates
• no severe problems in filter code
• Remaining testing
• L3 filters should not be a big problem - an
  afternoon should be enough
• waiting for
• PDT synchronization problem
• sync bit test in unpacking code
• Coincidence filter
• Not a huge job once we figure out implementation
• sorry - I cannot give you a schedule