DFEw firmware update for remote monitoring:
Jamieson developed new firmware for double wide boards
that generates diagnostic information on the front panel LEDs,
and allows remote reading of this information. This firmware, which needs to be loaded
instead of the 'normal' firmware, will be kept on the compact
flash cards, in addition to the production firmware.
Faster DFEA tracking firmware:
Jamieson thinks he can chop off 94 ns of the L1MUON signal latency,
by removing the option of finding a second track per block; see his
updated note http://d0server1.fnal.gov/users/jamieson/www/notes/1999-03-02a.pdf
for details. (Note added 2002-7-18: Vivek has already implemented this as an option in the
trigger simulation. Kyle will
help with the studies.)
Sequencer status after last access:
Fred implemented 264 ns firmware on three of the four crate 52 sequencers.
This should have zero impact on the trigger. (Note added 2002-7-19: not true...)
His plan is to study the basic functionality with
crate 52, then to update crate 50 (next week), and then to study SIFT integration timing etc. with
crate 50.
Global plan for DFEA-L1MUON latency problem:
Ron is putting together a global list of options
and decision points for this problem. See his
latest email for the details.
Mixer firmware:
Stefano is working on further diagnostic firmware. It will allow a wider choice of control bit/link to LED assignments. He reminded us to use the integrated register also for sync and clock error determination (e.g. a spurious L1_accept could actually be due to a bad clock). He remeasured the total transit time, including LVDS (de)serialization, to be 10.25 ticks (not 12 as reported earlier).
Linktest:
Makoto is working on AFE-Mixer and Mixer-DFEA linktest software. Makoto and Yuri will integrate all diagnostics into DFE_ware, so that everything is kept in a single GUI.
2002-7-18: The PW03-2 problem is solved. See
Yuri's
email message for details.
There have been some updates to
DFE_ware
and the accompanying web page. For various reasons (difficulty of keeping it up to date;
difficulty of keeping DFE_ware aware of loaded firmware) the autoexec functionality will
be replaced by a DFE_ware 'restore' button.
Firmware:
L2 CTOC,STOV,STOX: Carsten and Nick
are working on a general (L2 only) firmware problem.
DFES: Quichun will be at FNAL next week.
Readout and Offline Verification:
The CTOC-L3 problem is now believed to be a problem
in the way we treat the G-link sender. We are repeating basic G-link studies.
(Kyle has found one out of our eight CTOC fibers that occasionally seems to allow the
G-link receiver to synchronize. Unfortunately not often enough to make looking at the data
meaningful.)
Note added 2002-7-19: See Jamieson's email message
for details of a possible solution.