Minutes of the 20 March 2006 phone conference

1. News on bare modules and components assembly and deliveries:
   1. deliveries from AMS
      No new deliveries from AMS.
      After the ATLAS-AMS meeting on Tuesday:
      repair of the bonder is expected Wednesday or Thursday;
      remaining 216 tiles for more modules at AMS, number of FE chips matches the number of tiles.
   2. deliveries from IZM
      Of the 18 modules received last week:
      12 tested in Bonn: 10 OK, 1 bump problem, 1 damaged sensor;
      6 tested in Dortmund, all Ok and already assembled.
      16 modules expected to be delivered on Wednesday
      Bare module logbooks for IZM, from which a lower yield that the one claimed by Joerg was apparent, have been updated to properly show the actual situation.
   3. distribution and flow of components (type 0 cables, shipping frames, FlexMCC)
      Some available shipping plates from CPPM have been sent to Siegen.
      31 bare module boxes have been returned from Dortmund to Milano.
      Missing flex module travelers sent from Dortmund to CPPM.
      Maurice should urgently send ARCLAD tape to Dortmund.
      - inventory of trashed flex hybrid
      Report from Siegen
      Updated report from Genova
      Bonn report to be sent this week.
      Claudia reports that she assembled three modules using repaired FlexMCC (backside below the pigtail burnt during pigtail gluing and now covered by kapton tape). These modules have been heavily tested and no defect has been seen.
      She therefore will continue to assemble the 6 remaining FlexMCC.
      There are also few other FlexMCC that have been recovered.
      To assembly such devices, Attilio should send some new bare modules to Genova.
2. Assembly of AMS Modules in Siegen
   Iskander'swritten report and pictures
   In Iskander's report are marked the total number of pixels dead at the source test and the fraction of these pixel which is dead because they cannot be tuned (TDAC=127). From previous experience, these can be channels with shorted bumps. The remaining dead pixels should be disconnected ones.
   The bare module used were all "first class" modules, with very few (<10) disconnected channels at bare module probing, so this is evidence of damage in assembly. This damage is serious in some modules, but not a major disaster. Moreover, since it is not affected by thermal cicling, it is unlikely develop further in the future.
   The assembly team has been warned to be extremely careful with AMS modules and it seems the latest modules have a lower damage.
   Joerg confirms a similar situation in Dortmund, even if there are not data published yet (he will try to get a summary table next week). Unfortunately no action can be taken immediately, since all modules available in dortmund have already been assembled.
3. Inventory of hospitalized modules (link from previous meeting).
   Report from Bonn
   Report from Dortmund
   Report from LBL
   Report from Genova
   Report from Siegen
   Last meeting a more careful debugging report, especially for MCC failures, has been requested.
   Maurice has created a wiki page which is to be filled with a debugging check list
   Some news about debugging in Genova (visible in the above report):
   • some modules with clear MCC problems had the MCC replaced and will be shipped to Bonn for bonding
   • modules with bonding problems will also be sent to Bonn for repair.
   • some modules with short have been probed and the short on the FE has been confirmed. They will undergo standard qualification.
   • finally Claudia reports a strange failure of module 511213: this module lost a chip and started to show a very noisy region (picture) during the FLEX test (exactly between the source test, which was perfect, and the monleak scan).
   
   Joerg will provide an update list of Dortmund hospital, updating the named MCC failures which have been traced to be due to bad elco connectors.
4. AOB.
   • Please notice next weekend there will be the change to DST in Europe. Since the change will happen in the USA one week later, conncetion time from LBL will be different
   • Before this meeting there was a phone conference about stave ranking. The noise penalty weight needs to be retuned for stave ranking, because the noise in the standard treshold scan is higher than the on in the ANTIKILL mode, which is taken by MA for the module test.
     The only change in ranking application for normal module is that double ToT peak modules will not receive a penalty anymore, but will be excluded from the B-layer by setting the BLayer flag to NO. In such cases a message will be printed in the debug file.
   • There are problem in the tracing of STAVE test data for modules in Bonn and/or CERN. This is due to the fact that RDB tracks PDB shipments, but, once a module is on stave, the PDB applications properly track the ownership of the modules, but do not perform explicit shipments for them. As an example, the RDB cannot find the STAVE data for modules on stave 4092, even if they properly appear in the Bonn RDB Server.
     Paolo is investigating which is the easier solution.
   • In Dortmund it is needed the latest firware for TPLL EPROM (version 20 or so). They ask Maurice to contact John Joseph to get the latest one (previously he provided an older version).