------------------------------------------------------------------------
Summary of the Layout working group meeting of 19/01/2000
------------------------------------------------------------------------
(summary compiled by Burkhard)

1) Paul presented some further results of the muon system optimization.
He looked in particular into the question whether region 1 of stations
M2-M5 could be instrumented with strips. The occupancy optained with
logical strip units in R1 of 6 x-strips of 1 cm and 4 y-strips of
2.5cm is 23% in M2 and 9% in M5. The perfromance is very similar to the
one obtained with pad layout, and the triger remains reasonably robust.
Reducing the x-strip size to 0.5cm in M2 and M3 leads even to an
improved performance with normal (+5%) and high (+2%) background
(numbers for 1% MB-retention).
Paul looked also again into the possibility to drop station 4, assuming
a majority logic would be used insteasd in station 5. Unfortunately,
with the old background parametrization (which we are still using and
which contains no information about the direction and penetration power
of background particles, only very a preliminar conclusion can be
drawn. In order to simulate the real situation Paul substracted the
percentage of uncorrelated background (67%) as obtained with the new
parametrization from the old prediction for station 5. This leads to
a very encouraging results: A majority logic in station 5 could replace
station 4.

2) Olivier presented the status on the L0(muon) performances, not
using the theroretical algorthm but simulating the real processors.
Unfurtunately the positive results of Paul could not be reproduced so
far. This might be due to a different version of SICB he is using
(220 instead of 219) and deiffernt fields of interest in y. The
deterioration in performance is (4+/-6)% for the theretical algorithm
and (16+/-7)% for the FIP+DMP.
On the other hand, the results with the new backgronf are encouraging:
the average number of FIP candidates went down by a factor 2.
The performance of the FIP+DMP is independent of the sector size.
The Marseille group had also a look at high MB statistics and gave a
summary of the various MB samples:
- MB type 51: PYTHIA 6, model 3+QQ, msel=1 => only inelastic pp coll.
- MB type 61: PYTHIA 6, model 3+QQ, msel=2 => 60% inelastic, 20% diffr.
and 20% elastic pp.coll.
For us the sample qith MB type 51 is more relevant.

3) At the end of the meeting we had a discussion, introduced by Werner, on
the overall logic. The proposal of December 99 requires a 7-fold AND.
This means, that with a plane efficiency of 99% the overall efficiency
would be only 93%, and with 98% plane efficiency 87%. Therefore, a
reconsideration of the logic has been suggested, leading to a more
robust system.
As a hit in stations M1-M3 is mandatory for the trigger algorithm,
very high efficiency per station is important. Therefore one should not
use an AND of the planes in station 2 and 3, but rather an OR, at least
in station 2. In order to have a redundant system also in case of
faulty channels or chambers, a second plane in M1 might be also
important.
The situation for stations 4 and 5 is a bit different. On one hand,
equal distribution of active layers is advantageous, on the other hand
the use of the full filter is important to reject low momentum muons,
which is only the case if one requires a hit in station 5. In case
station 4 could be dropped, a majortiy in station 5 looks attractive,
but no conclusion has been achieved, as this requires more simulation
studies.