Trigger Operation Meeting (12/11/07)

Meeting Information

 Date      :  December 11, 2007
 Time      :  2:00pm - 3:30pm PST
 SLAC Room :  Madrone conf room, ROB 
 
 Dial-in   :  1-510-665-5437, pass code 6649#

--- Minutes ---

Operations

• Paged on Monday, because DCZ failed to configure. After consulting with Debbie, asked pilot to power cycle the crates. It solved the problem.
• Paged, emailed and called due to discrepancy in the reference plot and the JAS plots (Ex 1, Ex 2). I updated the DQM's short-term instruction to send email to babartrg but not page me. But we are not 100% of the cause yet - this is something that ought to be followed up.
• Paged on Friday once for a transient bit of ODF damage showing up in DQM trigger damage plots and also for a problem with l3tpcaserver.

EMT

• Calibration failed Tuesday when switched from background to foreground process. It seems that sometimes the online machine does not see the trigger nfs space and thus fails the calibration. This has not been noticed before - turns out the lookup tables were wrong for a few years but are now fixed.
  Also, the new calibration code found a failure due to 1 UPC. This was confirmed by swapping cables in the crate. This particular UPC has a history of being flakey and a new one was installed today.
• EMT tower mask procedure was tested (the first time since 2003!) and failed the first time due to a database bug. This was fixed, also the online release was updated. A practice tower was successfully masked and un-masked.

DCT

• A DCH-DCT calibration was performed successfully last week, after the online release in the work area was updated.

DQM

• The empty plots problem has been fixed - some bugs in the DQ monitoring code have been fixed.
• In R24 there is an improvement in DCH efficiency compared to previous releases. This is robably due to TrkFixUp improvements but should be verified.

Trickle injection monitoring

• See plots that Al showed on 11/13, now fit with a quadratic function (last page). This works for summed runs but is not so helpful for single runs. Do we want a better fit, and perhaps a new function?