regards,
Steve Dytman
*******************************
BEAM ISSUES:
We need a beam that is mildly defocussed (spot size of about
1 mm diameter) at the CLAS target location. The accelerator
can be tuned to a tiny spot and we can defocus with small
coils as in the June, 1997 run.
The spot location must be stable to +/- 0.5 mm. There is a
5 mm diameter section in the liquid target where the the beam
is supposed to go. Here, the window is 10 micron thick aluminum;
outside this central section, the window is 10 micron aluminum and
100 micron Kapton. The beam spot *cannot* move more than 1 mm as
a result of any action of the other Halls running, e.g. entries,
beam current changes.
The beam line must have significantly upgraded diagnostics as a
means to establishing low background and stable tunes.
The low current BPMs are a key part of this so that the
beam location can be known at the important locations.
We also need a quick way to know the properties of the beam
at our target. Using the viewscreen at the Faraday cup is not
good enough because the spot size is large there after multiple
scattering in the target windows. There are a few ways to do this--
1) Optical Transition Radiation (OTR) detector. Simple to use
when it is set up right, can be done parasitically? while
running. However, further development is required, there must
be a separate foil from the target which has more stable temperature.
I hear it won't happen fast.
2) Install a viewscreen with a TV camera upstream of the
target. Moderately easy to install and interpret. Need a long-lived
TV camera near the target.
3) Harp is already in and working. It requires a stop in running the
experiment to run a wire through the beam. Information is very accurate.
4) Bscope scans can be done by acelerator people. These accurately
give the beam excursions at any point along the beam line,
especially at the target. They use the BPM data.
***We need to decide on priorities for these 4 strategies.
Beam current must be tunable to within +/-5-10% of requested value
(depending on whether current is low or high) in the range
of 1-200 nA and stable to +/-10% over a week of running.
Of course, there will be down times. The location and current
must be able to be quickly reestablished after down times,
entries, etc. Our typical running condition will be at 5 nA?
Can line B be operated with all other halls off? We want to
continue running when the other hall (while lab runs 2 halls) is
down for a few days. I think that is mode to be used for Sept.
run, so that will be tested.
Beam current monitoring. Our goal has been to measure cross sections
with accurasy of a few percent. Faraday Cup is main beam current
device. What is its accurasy? What is accurasy of target thickness
determination? An obvious goal of the Sept. run should be to determine
cross sections reliably.
CLAS DETECTOR:
***Having CLAS working at high efficiency and stably is a major
requirement. However, it will take an enormous effort by
many people to get it there. We must be prepared to dive
into hardware and software efforts to get what we need.***
We require 4 fully-instrumented sectors. A fully-instrumented
sector has all detectors installed (including Cerenkov and
large-angle TOFs). It also has all associated electronics
installed and tested with cosmic rays. All forward TOF panels
should be working with similar characteristics so that the trigger
efficiency is smooth across the panels. The drift chamber hit-based
tracking efficiency should be smooth (no sharp changes of
50% except at the the detector boundaries and no holes in
the acceptance as a function of theta, momentum, and phi).
(Level 2 trigger?). We anticipate running with triggers
of
1) electron= High thresh in EC and TOF (prescaled)
2) charged= Low EC and TOF (heavily prescaled)
3) electron and 1 charged (NO prescale)
4) electron and 2 charged (NO prescale)
5) electron and 1 neutral (neutral= high EC and no TOF)
6)
7)
8)
We require a stable and easy-to-use data acquisition program.
The uptime should be greater than 80% and the runcontrol
should be able to be used by non-experts with success.
We require some means of getting a real event rate of
greater than 400 Hz with a deadtime less than 20%. Regular
ethernet is inadequate. We have no strong feelings about
whether the solution be ATM or fast ethernet.
We require a liquid target of 5? cm thickness with a good
control system. The control system used in the June, 1997 run
seemed to be adequate. We should be able to tune beam at 100 nA
through the target when it is empty. High currents are also
required for Bscope runs, I hear 3-5 microamp pulsed current
with 1.6% duty cycle=> 50-80 nA average is minimum.
The data taking procedures in June were far too complicated
and flakey. We have lots of complicated systems, so shift
leaders (the persons most responsible for the correct
functioning of CLAS each shift) need more training. Each
subsystem group should be responsible for establishing a
set of monitors (software and hardware) to tell the status
of their system. These monitors must have user documents
that the shift leaders can use and explain to their shift
members. These status monitors should be either tied
to alarms or (more likely) checked by the shift scientists
every few hours. A lot of thought and effort needs to go into
this, it can't be done on the fly.
We need EPICS to function better. There must be documents to
understand what the system is telling us and the suggested
responses. The system isn't self-explanatory now and won't
be for many months.
We need call lists for EACH detector system unless there is
a person on-shift. In December, the detector shouldn't be
in a state where an expert is needed on-shift for, say, daq
or drift chambers.
ANALYSIS:
Do we have requirements that MUST be settled before we run
in December? I don't have any that seem approriate to
write down.
SIMULATION:
Any requirements? I don't see anything that we need for
RUNNING beyond what we did for proposals, but much work
remains before we can have acceptances good enough for
cross sections.