Monday, April 7, Notes

Present:  E. Barsotti, A. Byon, R. DeMaat, G. Drake, J. Elias, S. Hahn,
          T. LeCompte, C. Nelson, S. Orr, G. Ott, K. Schuh, A. Tollestrup

     Tom LeCompte and George Ott were present to discuss the grounding issues
for the CMX and IMU systems.  In both cases, the concerns follow from the
scintillator coax connections, the chamber high voltage connections, and the
digital links to the counting rooms.  For the CMX, there is also the nasty
business of the preamp oscillations experienced last run.
     The documentation for CMX is still very incomplete.  The working group only
has a sketch made by J. Lewis at the last minute for the November workshop which
he said wasn't the whole story.  Preamp oscillations were discussed in some
detail.  During the followup discussion for CMU and CMP, V. Simiatis had said
there was evidence that the preamp to ASD link could radiate into the test pulse
link leading to oscillations.  This was not the only possible path, but one
which had the right timing and was capable of such feedback.  The test pulse
network has been redesigned for CMP to avoid this vulnerability.  For CMX, a
similar rework of the test pulse network seems prudent.  Also, the system sketch
shows that the preamp-ASD link has a braid which is completely floating; it
should be grounded at one end at least.  Suggestions are: improved documentation
on the system, checking with P. Schlabach about oscillations, and reducing the
vulnerability of the test pulse network.  The CMX ANSLEY cable issue will be
taken care of indirectly as the CMU and CMP groups are participating in the CDF
digital links working group and have similar links to tame.
     The documentation for IMU doesn't exist (yet); we were discussing G. Ott's
descriptions. In IMU, the links are from the downstairs TDC's to the trigger; U.
of Michigan (E. James) is responsible and proposes serial links based on the
"Hot Link" chips.  These need to be brought into the digital links working
group.  The preamps are like CMP (and CMX) and are potential oscillators.  There
is time, however, to change to the new test pulse network design.  Configuring
the system for a single point ground, no loops, is complicated by the
requirement to move each half of the toroids independently without uncabling
anything.  The scintillator systems are not a problem as they can be floated,
the discriminators are downstairs, and the high voltage cable braids can be
resistively interrupted.  The chambers are to be hard grounded to the steel
(which we agree is the best choice), and will use ASD's with transformer inputs.
Again considering oscillations, the question of termination of the preamp-ASD
links is relevant because of the transformers.  A test is needed to see if there
is a problem, to see if reflections occur.
     There is an open question of responsibility for the Pisa boxes which
distribute phototube high voltages.  All agree that modifications are needed to
break up ground loops; the question is who?