Minutes
of the FMD workshop, 29- May-2001,
CERN.
By Jens Jørgen Gaardhøje, NBI.
Agenda
29-05-2001:
1. Approval of agenda and minutes of last meeting
2.FMD overview. JJG (13.05-13.30)
3.Si-FMD Background simulations. A. Maievskaia (13.30-13.50)
4.Si-FMD mult. response. A.Karpechev (13.50-14.10)
5.Si-FMD FE and RO electronics. B.S.Nielsen
(14.10-14.35)
6.Si Prototype. A. Angelis (14.35-15.00)
7.PMD design. Y. Viyogi (15.00-15.20)
8.FMD reaction plane/flow simulations S. Raniwala (15.20-15.40)
9.Integration. L. Leistam (15.40-16.30)
BREAK
10.T0 design options/summary from
morning
V.Grigoriev/ H. Trzaska/G. Feofilov/ Chris Fabjan (16.50-17.30)
11.V0 and role in trigger J.Y.Grossiord/L. Montano (17.30-18.00)
12.Summary- Conclusions,
Recommendations for ALICE Forum, TB and MB. JJG+All (18.00-18.45)
13.AOB
Minutes:
- The minutes from the previous
meeting in February (29-02-2001) were approved without comments.
- JJG presented the current
design status of the Si-FMD. The salient issues were the high background
from secondaries in the most forward detectors, and the role of the Si-FMD
in the trigger. It was proposed to abandon the Si4 detector located at
630cm from the IP to the left side, which sees a background of about
300-500 % of the primary signal due to interactions in the beam pipe. This
downscoping will reduce the pseudorapidity coverage from approx. 1.5-6 to
approx. 1.5-5.3. It was pointed out also by Chris Fabjan that it would be
desirable to move the Si3 detector to larger distances (approx. 345 cm
from the IP) so that this detector, the left T0 and V0, and also the PMD
could be mechanical mounted together on a common support structure. The
possibility to optimize the pseudorapidity coverage by decreasing the
inner radius of the detectors to 3.5 cm and adjusting detector wafer size
and distance to IP was discussed and will be pursued. It was also proposed
to not to pursue further considerations regarding the inclusion of the
Si-FMD in the LVL 0 and 1 trigger, due to the difficulties related to
producing a useful hardware trigger signal within the approx. 80 ns defined by the trigger
group. It was pointed out that the V0 detector can easily provide this
functionality within the required time limits in a much simpler fashion.
The segmentation of the individual wafers was discussed. The segmentation
was proposed to be (at least) 128 strips/wafer leading to an average
occupancy per strip of the order of 2-4. The issue of the integration of
the Si1, Si2, T0-R and V0-R detectors with the ITS frame was discussed.
Detailed work is needed here (See also item 9).
- Alla Maievskaia presented new
simulations of the occupancy of the various detector elements, which
evidenced the high background in the elements at small angles. A
discussion ensued regarding the design of the beam pipe, flanges, pumps
etc. It was recommended that the integration group should continue
detailed studies and work to reduce the amount of material in the path of
the Si detectors as much as possible. New simulations for the new inner
radius and new detector placements should be undertaken for the next
meeting.
- A. Karpechev showed studies of
the multiplicity response of the proposed system, in particular under
various assumptions regarding the of-line grouping of various detector
segments and an analysis of the multiplicity starting from the deposited
energy The overall conclusion appears to be that the multiplicity
resolution is better than 10% and
thus within the design goals for the system.
- Børge Svane Nielsen reported on
the studies of existing suitable radiation-hard front-end(FE) and RO
electronic system for the Si-FMD. Two FE chips are retained as interesting
for this project. The Viking-RICH chips from IDEAS, Norway and the Hal-25
chip used for the ALICE Si-drift detectors. Both chips may require
modifications for the present implementation in order to match the
required strip capacitances, gain and dynamic range, number of channels
etc. BSN and JJG have visited the IDEAS company in Oslo and started
concrete discussions of performance and price. It appears that a modified
Viking chip can do the job. IDEAS can also deliver hybrids, handle the
gluing to the sensors , bond and provide test of the assembled wafers with
a factory guarantee. A test readout system is available for prototyping.
It appears that the HAL 25 chip may be more difficult to adapt (dynamic
range). It also appears that the HAL25 preamplifier shaping time is so long
(ms) that the chip may not be able to handle the necessary rate for this
high occupancy application. These
issues should be explored until
the next meeting, so that a choice of technology can be made and so that
the project may proceed to initial tender and prototyping. Paolo suggested
to consider two layer bonding for
the inner rings of Si1,2 where it is difficult to connect sensors and
hybrid cards.
- Aris Angelis reported on the
Athens prototype effort. A segmented (16 sector) Si detector has been
purchased from Canberra and has been delivered. A test FE circuit based on
conventional (not integrated) circuitry has been developed. It was
discussed whether this system could be ported to radiation hard highly
integrated form. In view of the complexity of such a project and the tight
timescales (the full system must be installed in ALICE in early 2005) this
does not seem to be a desirable avenue. It was discussed how the Greek
groups could utilize their electronics expertise in the final project. A
possibility is a direct contribution to the RO system (ADC’s- VME – DAQ intefacing). The Athens groups
will discuss this internally for the next meeting.
- Yogendra Viyogi reported news
from the PMD group. The integration of the Left FMD detectors (Si3 etc..) with
the PMD was discussed. The PMD group has looked closely at the beam line
situation and explored various options (Al, Al-Be, longer Be ) for the
beam pipe. A placement at d=350cm for all forward left detectors seem
feasible. The various groups should keep in close contact on these issues.
- Sudhir Raniwala presented
simulations of the sensitivity of the present Si configuration for flow
studies (directed and elliptic). It was pointed out that the sensitivity
to f.ex. V2 depends markedly on the azimuthal segmentation and that it
would be desirable with an increased segmentation in phi for the Si
detectors. The segmentation is currently 16 for the inner rings and 24 for
the outer rings. The Si-FMD group will revisit the segmentation in
connection with the optimization of the geometry and coverage.
- Lars Leistam showed current
plans for the beam pipe and ideas related to the integration of the FMD
detectors with the ITS. It is clear that the mounting of the FMD detectors
in the conical volumes on either side of the ITS poses a difficult
challenge. Detailed work on this issue should commence ASAP. A detailed
plan for the sequence of steps associated with the installation must be
worked out. LL and the integration group will study the beam line further
and attempt to reduce the amount of material in the path of the FMD
detectors as much as possible. A major concern is also the background from
ITS cables. The contribution from this source (f.ex in radiation lengths)
is unknown and must be evaluated. The cooling requirements of the FMD
detectors were discussed. Preliminary figures for the power requirements
of the Si wafers were presented based on Viking chips. It appears that the
total energy release in the FE electronics is low (5-10W). Thus cooling
needs are small. In any case, the Si detectors need to be operated in a
dry environment wherefore a small N2 flow will be
required. The amount of cables
from the FMD detectors were discussed. The VIKING chip system requires
about 20 (HV, multiplexed signal,
and control) cables. However, chips can be daisy-chained and output
signals multiplexed so the total number of cables is limited. The T0
detector (12 PM’s on either side)
will generate its HV in generators on the detector so the number of cables is also small.
- A main issue was the choice of
technology for the T0 detectors. A pre-meeting had taken place in the
morning and Chris Fabjan reported on the discussions on behalf of the
proponents of various systems (MCP- G. Feofilov, radiator-PM by V.
Grigoriev). It is recognized that the MCP system would need significant
additional R&D work in order to be brought to a stage were it can be
considered as a realistic option. It was felt that this is not realistic in view of the time scale of the entire project.
On the other hand, the radiator-PM solution can provide the necessary fast
timing required for the main ALICE T0 trigger and adequate efficiency with
existing and proven technology.
The discussion continued under item 12 after dinner. It was then
decided to recommend to the TB, the Forum and the MB that the T0
radiator/PM option be retained as the baseline option for ALICE. It also
turns out that the funding of this detector is now assured via the
Jyvaeskyla group (Trzaska et al.). The Finnish and Russian groups will
collaborate on this project and bring it rapidly to the prototyping stage.
A two-arm T0 is needed for good vertex for runs where TPC an pixels are
absent.
- Jean Yves Grossiord presented the status of the V0 detector.
The detector will be built in a collaboration between the groups from Lyon
and Mexico. V0 is now adapted in size to match the occupancy of the Si
detectors providing desirable redundancy. It was pointed out that the V0
detectors, now that the Si detectors need not deliver a LVL 0 or 1 trigger,
must provide the main on line LVL 0 multiplicity trigger. The necessary
trigger scheme will be developed for discussion at the next meeting.
- The main decisions from
the meeting are as follows:
-Abandon Si4.
-Position Si3 together with T0-L,
V0-L and PMD in a common group at about 350 cm
from the IP.
-Optimize the geometry of Si1,2,3
for best pseudorapidity coverage.
-The electronics for the Si-FE must
be frozen soon so that realistic prototyping can
commence.
-The beam pipe layout must be
optimized and material strongly reduced to suppress the
production of secondaries.
-The integration with the ITS must
proceed to a detailed level ASAP and the material
budget of its cables evaluated.
-Abandon the Si-FMD as a participant
in the low level trigger.
-The two arm T0-radiator/PM option
is selected as the baseline detector for providing a fast
T0 trigger signal in LVL 0. Trigger scheme must be explored and
developed in detail.
-The V0 detector will provide the
main LVL 0 trigger multiplicity selection. Trigger scheme
must be explored in detail.
- Nothing.
PS: These
recommendation were subsequently presented by JJG at the Technical Board and at
the ALICE forum, and by FB at the Collaboration Board on behalf of the
management board. The decisions were endorsed at all meetings.