Minutes of the ST-Testbeam Run-Meeting - July 3, 2004
=====================================================

General
-------

 - Shifts between Monday 8:00 and Tuesday 8:00 have been
   cancelled, since due to CPS intervention there will be
   no beam that day.

 - After the Sunday meeting the Run-Coordination will
   be taken over by Johan.

 - next run meeting on Sunday, July 4, 15:00


Actions since yesterday
-----------------------

 - the power net at the testbeam setup was
   split into two segments

 - the beam telescope has been commissioned
   and provides tracking for the test modules

 - the time window for the trigger was adjusted
   to a width of 7.5 ns, corresponding to an
   rms width of 2.2 ns. At that level the latency
   scans around the peak shall be done with a
   spacing of 5 ns.

 - all 3 chips of the CMS reference module are now
   operational after recabling the optical link to
   another part of the Beetle1.2 mini crate.

 - first hits in all Beetle modules were seen after
   fixing a bug in the map-files and going to 300V
   for the irradiated sensor. Average S/N values are
     <S/N> = 19.4  for CMS-reference
     <S/N> = 12.7  for CMS-flex
     <S/N> = 16.2  for LHCb1-reference
     <S/N> = 13.0  got LHCb1-Irrad
   The latency is not yet optimized.

Testbeam Program as agreed at the meeting
-----------------------------------------

 1) complete the comissioning of the system

    > comission beam setup (spot size and position)

    > determine nominal latency for the Beetles

    > complete updated documentation


 2) mandatory program

    I) HV scan
       for Vfs=400 and Vfs=100
       peak scan +-20ns in steps of 5 ns around the nominal timing

       - 50k events per point
       - proposed scanning strategy:

         a) start with HV at a central value
              CMS-type sensors : 250 V
              LHCb-type sensors: 120 V
            and go in steps  of
              50 V  for CMS type sensors
              30 V for LHCb type sensors

         b) first go down in steps until the signal is gone

         c) then go up, if possible as high as 500V, at setting of
            the intermediate values

         d) if time permits go back to the nominal voltage, again
            filling the intermediate points

            example of a possible scan sequence
            for the case of CMS-type modules:

            250, 200, 150, 100, 50,
             75, 125, 175, 225, 275, 325, 375, 425,
            400, 350, 300, 250

    II) pulse shape scans
        for Vfs=400 and Vfs=100
        for HV setting of 1.5x and 2x the depletion voltage
        100 k events per point

        - proposed scanning strategy:

          a) start at nominal latency
          b) go back 30 ns in steps of 10 ns
          c) go forward 80 ns in steps of 10 ns at intermediate points
          d) continue an additional 100 ns in steps of ns

          starting with 0 at the nominal latency this wouldn yield the
          following sequence:

             0 -10 -20 -30
           -25 -15 -5 +5 +15 +25 +35 +45 +55
           +75 +95 +115 +135 +155


   Resource estimates:
   -------------------
   The total number of events for part I is 15.3 Million, for
   part II it is 7.2 Million. Assuming 1000 events per spill,
   this requires 22,500 good spills, or 378,000 sec up-time
   corresponding to about 13 shifts. Currently there are still
   22 complete shifts scheduled, i.e. if everythings goes
   smoothly, the foreseen program looks feasible, leaving
   even the possibility for some extra studies with the higher
   irradiated sensor or the GLAST ladder at the end. With
   a data volume of 15 kByte per event the required disk
   storage would be about 340 GB. With 500 GB per PC (plus
   two spare disks) this should be OK.



minutes by Michael Schmelling