AFE II

1
WBS 1.7AFE II Project Cost and Schedule
Alan Bross DZero Run IIb AFE II Directors
ReviewApril 13, 2005
2
AFE II - Cost and Schedule

• Full Project Plan Has been updated since the last
  review

3
AFE II Cost and Schedule

• Cost MS
• Production cost MS are based on quotes for parts
  and labor contingency estimate is grounds up
• Manpower
• Estimates from AFE experience (from AFE I and AFE
  II RD)
• Total Project cost approximately 1.8M
• Schedule critical path is AFE IIt prototype board
  testing
• Planning for adiabatic Installation
• AFE I and AFEII mixed running
• Will need to test compatibility once AFE II
  prototypes are ready
• Installation will be done 16 boards at a time
• Credibility of installation plan is dependent on
  AFE IIt prototype tests
• How quickly they come up
• Platform test and integration with trigger
• Software readiness
• Physical Installation can occur quickly

4
AFE II Milestones

• TriPt production submission 5/05
• AFE IIt Prototype testing 7/05
• AFE IIt Platform test 912/05
• AFE IIt Pre-production 10/05
• AFE IIt production complete 3/06
• Full Board complement Ready
• for installation 4/06

5
AFE II Schedule

• Credibility/Underlying Philosophy
• AFE IIt boards will not have the uniqueness
  problem of AFE I
• All AFE II boards behave identically
• Prototype testing to date supports this
  contention
• Recent data shows AFE II Proto behaves the same
  on bench and on the VLPC Test Cryostat (as close
  as we can get to the platform for now)
• TriPt testing to date supports this contention
• Calibration of boards occurs once during board
  test and the obtained calibration constants are
  valid under all operating conditions of the board
  and are stable with time
• Time calibration
• Cryo and Bias control calibrations
• The above assumptions make the testing and
  installation schedules credible
• Will be Validated with a rather large number (
  20 10 AFE II / AFE IIt) of prototypes

6
Issues - Schedule

• Bringing up the AFE II prototypes has been harder
  than expected (Paul has covered this).
• But excellent progress has been made and this is
  nearing completion
• STILL - No underlying technical problems have
  been discovered
• FPGA code verified
• No uniqueness issues uncovered
• No major assembly issues uncovered
• Resistor packs on VLPC input need to be changed
  to discrete components
• An additional engineer (Stefano Rapisarda) is now
  on board for the AFE IIt design
• His input has been used to update schedule
• More up-front time devoted to schematic review
  and design
• Layout delayed until the above is well along
• This re-evaluation has introduced an
  approximately 1 month delay from previous
  schedule, but we believe that this approach will
  yield a final board that can meet our test
  schedule

7
Production and Testing

• Critical path remains the AFE IIt layout,
  prototype production and test
• We will produce 250-300 boards. First 200 to
  work are set to be installed in D0
• Muon Ionization Cooling Experiment will purchase
  some of the extras (more later).
• Board Testing
• 200 Fully functional Boards
• 8 weeks
• Personnel
• 3 Engineers
• 5 Technicians
• 3 Physicists
• 1 Computer Professional
• 5 fully functional boards come out of the test
  queue per day

8
Manpower

• We now have all manpower resources indicated in
  the project plan except for test/commission/instal
  l.
• As of Jan 1 everyone on board
• A university group needs to be brought on-board
  to work on testing/commissioning/installation
• Although this is later in the project
• There is a long learning curve for work with AFE
  boards
• Now is the best time to become familiarized with
  the boards
• D0 Spokes are addressing this issue and exploring
  a number of possibilities (Jerry will discuss)

9
Manpower

• We have put explicit resources (i.e., names) in
  for almost all tasks
• Tasks have costed manpower resources and a name
  is now associated with EVERY entry for almost all
  tasks in the project
• Fermilab personnel
• University group for Test, Installation, and
  Commisssioning (ND)
• Over-allocation is obvious if one looks at
  resource graph by name
• Effort is now pretty-well leveled
• In order to maintain the schedule
• Must keep each individuals peak effort 1-1.5
• Allocation plots that follow are peak effort in
  the period (Q), so being over 1 is OK so long as
  it is not constant

10
Manpower Profile - EE
11
Manpower Profile - ET
12
Manpower Profile - PHF
13
Manpower Effort Allocation
14
Manpower Effort Allocation
Expanded scale 1 week/div
15
Manpower Effort Allocation
16
Manpower Conclusion

• Resources are identified and in place
• Again except for T/I/C- D0 University involvement
  
• Manpower leveling looks OK
• At this time no one appears to be over allocated
• At least that we can not deal with
• Engineering resources anticipated to remain on
  the project through commissioning

17
M S Profile
18
Non-DOE Participation in AFE II Project

• The Muon Ionization Cooling Experiment has now
  been approved in the UK and US funds have been
  committed to the project for the next three years
  by NSF (also DOE through the Muon Collaboration,
  but those funds directed at other parts of MICE)
• The MICE commitment (in MICE project plan) is
  250k USD for approximately 30 AFE II boards.
• Based on assuming ½ the cost of the TriPt
  submission
• Plus 30 boards _at_ 4k/board
• 75k in funds available from NSF currently
• Balance in MRI proposal to NSF
• If this proposal is rejected, then UK funds will
  be used once an MOU with Fermilab is in place

19
Impact for D0

• Obvious cost savings
• Also helps mitigate schedule risk
• If 300 boards become the baseline production
  quantity
• First 200 good boards going to D0 will then
  arrive earlier based on model that once a board
  fails a test station it is put into a separate
  repair queue and the testing moves on the next
  board.
• The MICE time frame is late Fall of 2006 for
  delivery of all boards, so this produces little
  pressure on the repair queue. MICE also plans to
  have manpower come to Fermilab to help with
  testing and to gain experience/expertise with the
  AFE II boards (late Fall 05 is target) which is
  in addition to the resources we are currently
  assuming we have in place for the project.

20
Cost Schedule - Conclusion

• No technical issues have been discovered to date
  that would impact either cost or schedule
• Or our philosophy/model regarding Test and
  Commissioning
• MS and manpower costs are well understood
• The contingency (MS and manpower) in the project
  plan is conservative
• There is no schedule contingency in the project
• 4 w possible if Vendor Qualification of AFE II
  production is deemed unnecessary
• Non-DOE help may be forthcoming
• Manpower is on board