Minutes of the Tracker Planning Meeting

November 21 and 22, 2000

 

Robert presented a summary of the schedule and plans for the engineering model (see http://scipp.ucsc.edu/groups/glast/meeting/non_periodical/20001121/robert.pdf).  Ronaldo expressed concerns about the late availability of ASICs to use in making prototype trays.  Robert explained that we plan to make the ASIC production run as soon as we are confident that we have good prototypes.  It is possible that the full set of chips will be available before the end of FY2001, but it could be as late as June 2002.

 

Ronaldo presented a study of work and manpower required for tracker fabrication (see http://scipp.ucsc.edu/groups/glast/meeting/non_periodical/20001121/ronaldo.pdf).  From this he derived a budget request to present to INFN.  There was some discussion of some of his time estimates: in particular, some thought that his estimates of time for ladder testing and electronics attachment were pessimistic.

 

Eduardo presented his own ladder assembly time estimates, which were in reasonable agreement with Ronaldo  (see http://scipp.ucsc.edu/groups/glast/meeting/non_periodical/20001121/eduardo2.pdf).

 

Thereafter followed a discussion of division of work between SLAC and INFN:

In response to Tune’s proposal to have INFN take responsibility for tower assembly but do the work in SLAC facilities, Ronaldo said that it is impractical to have INFN working on tower assembly at SLAC when they still are assembling trays in INFN.  It is also expensive and difficult for them to relocate their personnel to SLAC.  Furthermore, they are prepared to do environmental testing of towers plus functional testing in Italy.  For example, they could have access to the vibration table in Terni for free.  Also, they have available thermal vacuum chambers.

 

Tune said that he is mainly concerned about doing assembly at SLAC asap to allow testing early in conjunction with the Grid, DAQ, Cal, etc.  Hartmut also expressed concerned about the first towers spending a lot of time in Italy during testing when we need them here asap.

 

There was some discussion of the testing planned at SLAC.  SLAC will have a thermal-vacuum chamber, perhaps big enough for the whole instrument but certainly big enough for a tower.  Cosmic-ray testing will be done at SLAC with the instrument in thermal vacuum.   Bill said that thermal vacuum testing will first be done at the tower level (not individual trays).  We will do it on all towers.  All towers will also get an acceptance vibration test.  NASA requires 8 thermal cycles, in total.  This could be 4 at the tower level, plus 4 at the instrument level. 

 

It was noted that tray production cannot stop for testing of the qualification tower but must go in parallel.  The tray design will be qualified much earlier (even before PDR), and the engineering model testing will be to (or beyond) qualification levels.  That must be coupled with good control of the manufacturing of tray panels to ensure that trays simply cannot fail during the qualification tower testing.

 

 

Bill suggested that maybe we could assemble the qualification tower here and rest in Italy.  We then would do qualification tests at SLAC and acceptance tests on the remaining towers in Italy.

 

Guido agreed that it is a good idea for SLAC to do qualification tests on the 1st tower(s).

The consensus to integrate the 1st n towers at SLAC while INFN concentrates on tray production.  The remainder will be integrated and tested in Italy.  The number n should be 2 or 4.

 

There was some concern that we are not supposed to start production of flight parts until after CDR.  Robert insisted that in fact a CDR on each item must be done before flight production.  However, those CDRs can be held whenever we are ready.  They don’t need to wait until the time of the nominal LAT CDR in July 2002.  It was noted that production of flight detectors is about to start, and the relevant documents still are not under configuration control and an appropriate CDR has not been scheduled.

 

It was noted that Tom’s tracker schedule doesn’t agree with Martin’s schedule on when towers should be ready for integration.  Robert and Tom said that they so far had been detailing the pre-CDR schedule.  They will start looking closely at the production schedule soon.

Sandro warned that we need to take care in the schedule about procurement of carbon-fiber materials, which tend to have long lead times.

 

Ronaldo said that in mid December there will be a joint committee meeting between INFN and ASI, where they will agree on how to share GLAST funding.  Then it will go back to the INFN funding committee in January.   He expects that in January we will have a much more clear view of the funding situation.  The year 2001 is the most critical in the funding profile.  He will work to adjust the MOA to the new plan. 

 

Guido said that ASI accepts INFN management of GLAST fabrication. 

 

Hartmut talked about detector testing (see http://scipp.ucsc.edu/groups/glast/detector/GLAST2000/SLAC_INFN_11_21.pdf).  Purchase of 8000 detectors is committed to HPK.  That leaves up to 30% for other vendors.  Hartmut is working on getting the specifications into the GLAST configuration control system.  We will receive 35 detectors in December, which need to be tested by Jan 31 in order to give HPK a go-ahead to make 400.  Ronaldo wants some of the HPK prototype detectors and is ready to contribute to testing, including capacitance measurements.  He can give a 1-week turn-around to get results.   Hartmut said that HPK is testing capacitors up to 100V, with IV curves up to 200V.  The leakage current is specified at 150V.  He will send 5 detectors to Pisa.  He stressed the importance of recording temperature and humidity during the testing.  The standard is 25 degrees and less than 50% relative humidity.  Eduardo does not want to be responsible for testing prototypes at SLAC.  The testing will primarily be done at Hiroshima and Pisa, with work also be T. Handa at SLAC.  During production testing, Hartmut would like not to measure individual strips on detectors (except by the vendor).  Strip testing for shorts, opens, and missing wire bonds will be done on ladders. 

 

Work needs to be done on shipping and storage soon.  It was suggested that we store detectors at HPK until we are ready and ship about 100 at a time (i.e. once a week).  We need to think about tracking and labeling of ladders.  Pisa will look into how and when to read the scratch code.  Possibly this task could be done at the same time as the check of the alignment of the wafer edge. 

 

Hartmut said that most QC testing of the detector production can be done by measurements of structures on the wafer cutoffs.  This will be done on every lot.  It can be done even before HPK has finished testing the real detectors, giving us a quick control of the process.  This testing doesn’t need to be done by the institution doing tray assembly.  For example, Hiroshima could do this for the HPK detectors at least.  INFN should make use of the wire-bond testing structures on the cutoffs. 

 

Next there was a discussion of the database (see http://scipp.ucsc.edu/groups/glast/meeting/non_periodical/20001121/eduardo1.pdf).  Pisa representatives are planning to attend the software meeting at SLAC January 16 to 19. They should get together with Masa at that time to discuss tracker database planning.  Masa has written a lengthy document with ideas about the database, based upon experience with the BTEM

(see http://scipp.ucsc.edu/groups/glast/system/tracker_database2.pdf).

Tim Thurston was asked for his views on what needs to be done.  He said that his immediate concern is that we are producing soon some flight hardware (SSD) and need to get a database set up in time.  NASA requirements are secondary to what we need.  Tim will have to think about what arbitrary requirements NASA might still foist upon us.  Some obvious examples of what we need are information for physics (calibration) plus a complete pedigree of each detector, etc.

 

Eduardo talked about final data products needed: leakage current per tray; strip coordinates for alignment; dead strip maps; noisy strip maps.  He would like there to be one database with web access for everyone.  He also proposed a grading system for ladders.

 

Karen is redefining Masa’s tables for input into the SLAC Oracle database.  She also needs input from Italy.  Oracle at SLAC can interface to any sequel-based database.

Tim said that it is most urgent is to understand the electronic form in which HPK will deliver test data.  Database work at SLAC should continue as it is going, but we need to start communication with Pisa.  Eduardo will be the SLAC contact person.  Hartmut suggested that we also include Hiroshima in the discussion.

 

There was a short discussion of the Italian tandem Van-de-Graaf machine and possible use of it for SEU and SEL testing of the tracker ASICs.   UCSC and SLAC can supply the test hardware.  Ronaldo said that he would like to support this but needs to look into it.