October 17, 2007 online meeting minutes

present: Sergey Boyarinov, Eugene Pasyuk, Benjamin Raydo, Sergey Pozdnyakov, Stepan Stepanyan, Nerses Gevorgyan, David Jenkins, Chris Cuevas

DVCS Trigger

• Ben presented recent results obtained by processing DVCS raw data through trigger system simulation tools; we discussed several approaches to the trigger based on production and calibration run requirements; based on that discussion the number of changes will be applied to the 'DVCS Trigger System Requirements' document, most important listed below

• for calibration runs (search for pi0, 2 clusters with relatively low threshold, for example Ben measured 88% efficiency with threshold=50 and requiring minimum 3 cristals in the cluster) efficiency can be about 90%

• for production runs we will be triggering on one cluster matched with hodoscope hit; output will go to the separate trigger bit (need to make sure latch is working, although electron will go to one of them, CLAS or DVCS, not to both so most likely we'll have only one of trigger bits fired); we are expecting powerful clusters so 3+ cristals per cluster can be set as a requirement for the trigger logic

• we'll have relatively low thresholds on discriminators, because the same discriminators are used by TDCs, so that parameter cannot be changed a lot

• new hodoscope will be designed in few following months, it will have 60-80 channels with single element size about 3x3 DVCS calorimeter cluster size; match between two detectors will be enforced by DVCS trigger system; overlaping area will be another parameter for that system

G9-FROST preparations

We discussed what should be done before G9A run start. Following issues came up:

• Franz implemented new monitoring features into fast online reconstruction; Sergey will talk to Franz to see what should be done to make sure it works

• we'll conduct counting house cleanup (remove rack, clean table, exsess old equipment)

• fan was fixed in one of LAC HV mainframes by Danny

• another sy527 HV mainframe was sent to repair by George Jacobs

• Lei tried to fix EC sec4 mainframe problem, it still alarms ...

• we asked Volker to deal with Universal Voltronics, the company which does Lecroy mainframe repair; we are not happy with both repair quality and schedule; it maybe necessary to involve procurement division

• Nerses labeled all HV mainframes in the Hall; need to make a list for shift takers, which mf correcponds to which ioc etc

• Chris will check L2 electronics and DC TDC stop signal pulsers, especially on SF south side where missing gates were observed in dc3, dc4, dc9 and dc10

• request was submitted to computer center to create 'g9a' SILO partition

• epics_monitor switched from CDEV to CA by Sergey, former gamma_monitor incorporated into it, everything is driven by config files in $CLON_PARMS/epics/ area, looks good but have to make sure all output are Ok; Eugene is looking into the set of EPICS channels in those config files making sure all channel names are correct

• $CLON_LOG area moved to 300GB /data disk to avoid $CLON_PARMS overflow

• work in progress to implement IOC resets/heartbeats for new LAC ioc and Linux-based ioc's; it will be done for tagger_mac as well