Minutes of the BEAST meeting
Friday March 26th, 1999 JST
A meeting was held immediately after the conclusion of the BGM to discuss
issues raised about the current beam abort system as a result of lessons
learned during ongoing BEAST operation and a recent KEKB review.
BEAM ABORT system:
Attendees: Moffitt, Peak, Tsuboyama, Varner
Peak presented a number of figures and diagrams of the current system and
abort diode testing in Australia. This information was disseminated to
Tsuboyama and Varner. A few major issues about the diodes were raised,
and the responses are paired as follows:
1) DIODE I_LEAK AFTER IRRADIATION - testing shows a significantly smaller
increase in leakage current with irradiation for the S3590-08 + readout
than for the BEAST diodes. An offset adjust is provided to allow
maximum dynamic range to be used.
2) GAIN FOR UNBIASED OPERATION - concern was raised about knowing the
actual dose measurement during unbiased operation. Data was
provided from tests with a couple of different radiation sources.
Matching with EGS gives different estimates of the nominal "depletion
depth". Proper knowledge of the sensitivity of unbiased operation
shows an energy spectrum dependence and is thus to be considered as
"under study". Peak has pledged to write up a summary of previous
calculations/studies in a tidy form and distribute for discussion.
3) TEMPERATURE STABILITY - A plot of the temperature stability of the
diode + preamp was shown. Below 30C the thermal-induced shifts are
tolerable. Operation at higher temperatures was deemed unlikely, given
the proximity of the abort diodes to the cooling pipes. At any rate
the diodes are installed and there is little to be done at the
present time. Solutions will be sought only if problems exist.
After this, a general discussion was held on modifications necessary to
the current NIM abort module to be compliant with the recommendations of
the KEKB review, as well as issues of dynamic monitoring. It was decided
that the current module be modified to provide:
= independent analog/digital outputs for monitoring
= a 0.5-2s shaping output ("slow output" to compliment the current fast one)
= possibility to issue a reset from computer control (with a physical
"lock-out" switch to override and require manual reset).
= packaged as either 2 separate single-width or 1 double-width NIM module
= external NIM logic shall be employed to provide separate "beam abort" and
"injection throttle" signals as necessary, from the outputs of the module
Please see the attached set of items from Tsuboyama-san about testing and
cabling issues in the current system. It was decided to leave the NIM
module that has already been provided to KEK and tested by Tsuboyama-san
as a prototype to run with until a "final version" is ready. This
reduces some of the time pressure in developing such a final module.
Communication of the upgrade criterion will be made by Peak, upon receipt
of final specifications.
Summary of action items:
- Moffitt will summarize (in conjunction with Browder) the requirements
on the outputs from the beam abort module. This should happen ASAP.
- Peak will communicate these modifications to the engineer in Sydney and
provide a delivery estimate after discussions there.
- Tsuboyama will "cable in" and prepare the prototype module for
operation after roll-in until an upgraded module becomes available.
- Moffitt (with help from Browder/Tsuboyama) will test the beam abort
diodes that have been installed on BEAST but not read out until now.
Currently there is no DAQ space allocated for this, but space shall be
made, as this is a rather critical issue.
- Varner will repeat baseline shift versus dose rate/annealing measurements
with the HARDCORE system within the next couple weeks.
- Peak will provide a write-up of the testing/simulation results for
knowing unbiased abort diode gain. Peak and Varner will compare testing
results and agree on a general procedure for quoting the dose rate
measurement [mRad/s = C * V_out]
Progress on these issues should be presented during the next few SVD video
meetings.
-----------------------------------------------------------------
Here is the status of "beam dump" from T. Tsuboyama:
(1) Cabling between Accelerator and BELLE electronics hut
is done last week. The module inside the E-hut accepts NIM
"level" signal. Reasonable filtering (response > a few msec?)
is done in the receiver(accelerator) to eliminate noise.
(2) Sydney PINs are installled both into BEAST and BELLE SVD.
Cables are ready in the cage(BEAST) and E-hut(BELLE).
(3) The PIN driver circuit was built by Sydney group as a NIM module.
It was tested by Sydney group before shipping to Japan. Tsuboyama
confirmed the function again in early March.
It accepts 8 PIN-diode signal(*1). Each channel has a comparator and
threshold can be adjustable. Multiplicity logic and latch
are integrated to generate dump request.
It outputs 8 analog signal for data logging(*2) (by IOTECH?)
and a TTL level "dump" request(*3).
--> The module is not computer controlled.
You must go to the E0-hut in order to reset the "dump" request.
This is a kind of fail safe(**).
(4) We need to assemble
(*1) a 26 pin connector to receive signal.
The module has 8 potentiometer which is not described
in the TDR. I do not have an idea how to solder the
cables.
(*2) a 9 pin connector to data logger
Trivial but neither BEAST nor SVD has unused 8 channels
in IOTECH.
(*3) a standard lemo cable.
(5) We might add some logic interface circuits
-> to know which channels were caused a beam dump.
-> to reset the "dump" signal through computer (contrary to the
idea of (**))
(6) How to test:
System test and threshold determination could be done
by BEAST without dumping the beam.
(7) These days people pursue me that not only "beam dump" but also
an "injection pending" signal should be prepared.