This is the list of physics runs that had a LV trip problem in MCH during LHC15o.

This list is constructed using 3 sources of information :

  1. LB : via the automatic log of the LV trip from DCS into the logbook. This should be the ultimate source but it did not work for some period of time, hence the double check here.
  2. Using the AliAnalysisTriggerScalers::PlotTriggerEvolution to get the ratio L0A/L0B (aka livetime, LT) for CMUL7 trigger. When there is a LV trip, the readout of the corresponding detector parts goes to a timeout of 5ms. Knowing that livetime = 1 / ( 1 + busyTime x inputRate) = 1 / ( 1 + 5E-3 x 1000 Hz) , we thus look for periods where the livetime was around 16%. The PlotTriggerEvolution is actually looking for LT > 5% and < 60%, so it is getting more runs than just the LV tripped ones…
  3. Finally a visual cross-check is made using the MAD display for the runs spotted above. Here we look for drops in the event rate (not down to zero, those are PARs).

In the table below are all the runs of this list (158 runs at time of this writing) found by the step 2, for the record.

RUN Remark
245349 LB - LT - MAD trip very close to the end of the run
245396 LB - LT - MAD trip for last 30 minutes of the run
245411 LB - LT - MAD trip for the full run basically
245439 LB - LT - MAD trip for the last 20 minutes of the run
245452 LB - LT - MAD this one was not detected by step 2 but it’s because Indra was there in ARC switching back the LV “live” but the 2 trips appear clearly in MAD (@49 min and 1h01)
245453 LB - LT - MAD trip towards the end of the run
246036 LB - LT - MAD not a LV trip but full CH10L occupancied hence a poor LT (busy 2ms)
246271 LB - LT - MAD - this is an instance of the lack of automatic log for a LV trip (can be seen clearly in MAD @ 49min)
246671 LB - LT - MAD not a LV trip but a drop below 40% of LT

So in summary there were 6 physics runs (total duration ~7h, CMUL7 lumi ~7.5 ub^-1) with a LV trip :

245349 , 245396 , 245411 , 245439 , 245453 , 246271