This report is in progress. Today we have used very little beam but have
made progress on a number of outstanding issues. Presumably we should have
all tof signals in, and in the right place now. Work is expected on checking
the coinc timing and the issue of L/R delay (we wish to understand this
before the chambers take data).
Data: Run 744 is a long (ca 50k events, 1.5 hrs) EMPTY target run.
=====
Beam: some time was spent in "optimizing injection" and steering. Some
===== improvements in the first (from 40 to 65-70 mA) while the latter
was required since the beam was (locally) optimized overnight by
the compton as well. Compared to thursday, rates on an empty target at 40
mA went up by 30- 50 % in counters (top-bottom coincs) on the left, maybe a
bit less on the right. The cerenkovs may actually provide a nice background
monitor, also showing a higher rate increase for the forward left sector.
Elog: well, it didn't work with 3 of us and didn't spend time there
==== Where is one supposed to get the .bin file ?
Scintillators: we finally find what was wrong with R11 and L12 !! In either
============= case a bad cable or connector to the TDC (short). If you saw
something funny on the adc of those paddles, it may have been
related to the fact that some adc macros require a tdc cut which obviously wouldn't
have worked. This problem was not diagnosed the other day due to an experimenter
error (I have to remember that after ECL-NIM conversion you start counting from 1,
not 0..) Tavi is remaking the connections to the TDC. Presumably, this was
could have been the last "cabling" issue with the scintillators (both sectors)
or at least the last known one. A more detailled check can now be done off-line
with the dump.C macro
Left and Right sector delay: As you may know, when we timed the L/R sector
=========================== the start signal was fed into channel 16. However
the delay unit after the 3fold coinc was set
differently for the left-16 and right-16 channels, respectively with 1ns and 16 ns
delay. I also find that the meantimer output is different between L and R (about
10 ns, w/ flasher) but likewise to the above delay units, the meantimer
is also fed to the trigger and thus the start. Both of these should not make a
difference, but maybe we should check with data. Indeed all cable length is the
same. Also flasher signals come at equal times up to after the CFD's. (Note that
there are internal trims in the meantimer that can explain the different timing of
outputs). Tavi is looking into this following John suggestion, I leave.
Coincs: With flasher on, I started looking at the 3fold (top, bot,
====== meantime) overlap that gives the prompt for each paddle. Found
that for the left sector the meantimed signal (60 ns wide) sits
comfortably in the middle of the longer top/bot pulses, so for that sector
the overlap seems maximum. HOWEVER, channel L1b showed an interested effect: if
the CFD threshold is low (or rather, the line or signal noisy) you can get
multiple firing from the CFD. In that case the overlap can be as small
as 20 ns. It is clear that this depends on ringing in the cables as well,
and must be experienced with beam. This is a work in progress.
Software: We had a number of discussions and ideas. Chris Crawford has
======== come up with a nice idea to allow for automatic journaling of
the trigger settings and the electronics map. Basically in
the first step these will be copied into an archive everytime we change
them (the filename will contain time and date). In the second step, that
filename will be stored in the configuration database. In the last step,
everytime one starts a run with coda, coda queries the configuration
database and puts that filename into the run database. So the correct file
name will be always be available from the analysis, at which time one can
retrieve the correct file from the archive. There will be no changes in
the way you operate trigger.map and electronics.map: at some point the
download button or an emacs macro will automatically save the
configuration name in the configuration database. Ask chris for details,
he just started.
finally, I get the following report from Adam de G.:
Wire Chamber:
=============
While waiting for the methane bottles to arrive, we ran out of ethane gas
sometime overnight. The channel for that mass flow controller has been
turned off and the pneumatic valve before the mfc has been closed.
Helium flow has been increased to ensure bubbling of gas through the
bubblers. There was no tripping of high voltage and the currents have
remained stable while running beam, except for the sense wire voltage for
box 24, a channel already known to have problems.
This will be investigated when the hall is reopened. The voltage for
that channel has been reduced from 1000 V to 500 V and draws roughly the
same current as the other sense wire voltages. It has not tripped since.
Regards
-- tancredi
________________________________________________________________________________
Tancredi Botto, phone: +1-617-253-9204 fax: +1-617-253-9599
research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST