L2beta status

1
L2beta status

• Production 22 boards delivered/tested
• 20 Boards delivered w/ no flaws
• 1 board w/ a single bad solder joint on
  connector
• 1 board has a 1-bit problem on VME section (not
  yet investigated)
• Incorrect ECL connector pin mapping will be
  corrected w/ LAL-designed flex circuit. Ready
  this month. 1st boards tested will use a small
  pcb adapter inline w/ cables to paddle cards.
• All boards on hand by next month.
• Shipping issues. So far so good!

2

• Production
• SBC Status
• 2 usable-model SBC at UVa today
• 1 SBC usable for tests
• 9 more on order (expected in next 2-3 weeks)
• Other orders to go in as CPU is needed
• Disk ordered for full set of boards
• misc hardware on hand to assemble boards at UVa
• Plan to send 1st board to FNAL this Friday for
  planned work next week.

3

• Firmware Core functionality available
• Broadcast, DMA, Mapper, PIO, I/O sections coded
  (all working in tests involving two betas over
  many events)
• PIO w/o complex arbitration working
• Most recent version is starting to address
  issues that will affect multiple card/crate
  configurations
• 1 previous set of MBT/beta Firmware successfully
  tested w/ O(106) MBT broadcasts. However, lost
  MBT words have been confusing. Using latest
  MBT/beta firmware, we have probably spotted the
  issue this week. Consulting on fix now
• Overall firmware is rather mature. Many
  functions such as glitch protection for MBus,
  error reporting registers, timeouts, etc. already
  implemented

4

• Firmware
• Concentrate on functionality for single
  beta/crate operation, then more of multiple card
  competing PIO tests
• Very detailed simulations created for beta
  firmware. Can make much progress consulting
  across the ocean.
• Next week the whole beta gang (including
  firmware engineer) will meet at FNAL Drew from
  the MBT site to iron out our remaining broadcast
  issues.
• Very optimistic, we have much flexibility, now
  are seeing problems due coding choices that we
  control.

5

• Software
• l2 hardware interface nearly complete.
  Everything except mapper interface available now.
  Will be added by end of this week.
• gcc compilation problems fixed in packages that
  hardware interface depends on.
• A private online release area is setup on clude0
  to mirror the alpha release area.
• All hardware interface libraries now build
  transparently in both alpha/beta versions
  depending. But this is not yet released.
• Should be able to try a full build of l2online
  code next week.
• Priorities 1) complete mapper interface, 2)
  complete port of suite of alpha test software to
  beta for next weeks work

6

• Summary Prognostications
• 1st stateside production beta is alive and well
  at UVa
• 2 more expected this week
• 1st beta at D? next week
• When do we plug in the 1st beta? Ask us next
  week! My guess is beginning of March...
• Strategy, replace single alpha subsystems to
  gain factor 3 in processing ASAP
• Longer term to bring multiple cards online
  (software/possible firmware). Good news is
  latest Alpha PIO firmware appears bullet proof.
  Can work out SW issues on Alphas while tweaking
  betas...