Pierre Auger Project
Progress Report
Summary (photo
album)
Tanks are once again arriving at the Observatory from
That strategy for completing the base line array is to deploy and commission 1600+ tanks in available positions. When the remaining positions become available by agreement with the land owner or become dry enough for access, the staff will continue the deployment up to about 1650 stations. After that tanks will be moved as necessary to fill and smooth the array.
The cause of the instability in some of the PMTs has been isolated as a grounding problem in the base. A fairly straight forward procedure has been developed to make the necessary repair in the field.
The HEAT project to install three high elevation fluorescence telescopes is moving ahead. Permission to rent the land near the Coihueco fluorescence building has been obtained and construction has begun.
Good data is being taken by the groups developing radio detection of showers. For the first time the LAL, LPSC and SUBATECH group has recorded self-triggered shower events that are in coincidence with the main array events.
The American Institute of Physics has included the Auger correlation result in the list of the ten biggest physics stories in 2007.
WBS 1.1.3 Fluorescence Detector Calibration
(Jeff Brack -
University) With approval of FD task leaders, the calibration group has agreed to form two subgroups, focussing on the absolute drum calibration and on the relative calibration analysis. Rossella Caruso will lead the relative calibration group. The overall organization will remain as a single
calibration group, due to the close relation between the two efforts; the intent is to allow the drum group to return to it's primary task. Two more data disks have been loaded with relative calibration data in Malargue. This nearly completes the copy of ~1 Tb of data from the Malargue server to other sites (Lyon and Catania). Work continues on establishing new calibration epochs to include the recent Loma Amarilla calibration. A parallel effort will establish 5 earlier epochs, extending the usable data for analysis back to January 2004. Work also continues on establishing the nightly calibrationdatabase. Calibration labs in Malargue and Colorado are both moving. The Colorado lab move from CU to CSU is nearly complete. Plans are underway for moving the Malargue lab from the galpon to the AB. Several options for defining dark room space in the AB hall are being considered.
WBS 1.1.3.3 Fluorescence Detector Atmospheric Monitoring
(Stefan Westerhoff –
Lidar (reported by Jorge Rodriguez)
The lightning protection system for Loma Amarilla is now in place. Because of the volcanic ground, the construction was particularly difficult at this site.
Raman lidar status
(reported by Vincenzo Rizi)
The system was operating in its standard configuration since November 2005,
although an intervention to obtain better mechanical stability was done in
August 2006, and a change in the DAQ electronics in April 2007.
In July 2007, probably due to the extremely low temperatures attained in
Malargue, the laser broke and is now in its way to the manufacturer for
reparation.
Anyway, since August 2006 the system is in a (relatively) stable configuration.
It is therefore possible to compare the Raman data with CLF measurements until
January 2007 (official CLF reconstruction) or to
It appears that there is both a large dispersion of the data, and a relative
shift of the values. In particular Raman Vertical Aerosol Optical Depth (VAOD)
values tend to be larger than CLF ones. On the other end, if we normalize the
VAOD measurements in each month to a different reference (clear or Rayleigh)
night, a better agreement is obtained.
The aerosol backscatter measurements of the Raman lidar are less affected by
optical or mechanical instabilities of the system setup, and good aerosol
density profiles can be obtained, which, in most cases, clearly show the
behaviour of the aerosol into the planetary boundary layer. The values of the
integrated backscatter profiles are proportional to VAOD in the case of single
aerosol composition, and other strategies to compare with CLF data are possible
(to obtain VAOD values from integrated backscatter ratios would require
knowledge of the Lidar Ratio).
As said before at present the laser head is broken and awaiting repair. In the
meantime there are projects to improve the stability of the system and the
signal/background ratio. As for the stability, some improvement will follow
from the planned change in cover movement and of the lidar roof in LL.
Concerning the signal/noise ratio we need to optimize the optical coupling
between the telescope and the optical fiber (to the receiver box). This
optimization will be tested in laboratory in L' Aquila and duplicated in LL.
Also work on the DAQ electronics is needed, and this work can be done in
to 5-7 km which is relevant for the fluorescence yield.
Finally, it is important to notice that the Raman Lidar can easily take
measurements in periods outside the FD shifts. These measurements are of little
relevance for Auger, but may be of interest for the larger atmospheric
community. We are presently starting collaboration with the
WBS 1.2 Fluorescence Detector Electronics
digital electronics and readout systems (Matthias Kleifges – FZK-IPE)
DAQ software (H.J. Mathes)
The main developer of FD-DAQ
software – H.J. Mathes – is for about 2 month in Malargüe to work on following
problems:
·
The
most severe problem in the past was a crash of the Event builder part, which
stopped data taking, especially, if the inter-camera trigger was enabled. With
the help of various tools (valgrind, efence, gdb) several memory leaks were
found and removed. In addition, the data structure of the EvB was changed as
ROOT containers were replaced by STL containers. Now the software runs much
more stable.
·
The
latest DAQ version 3.3.-2 for the December shift fixed the bugs #715 (critical
disk space), #714, # 547, # 697 and partly bug # 709 (not completely). It also
implements the new TLT developed by A. Schmidt, which is active since the
September shift.
·
A
rudimental inter-camera trigger was implemented in the DAQ, but tests showed that
the extra events would flood the FD-DAQ in case of lightning. We are working on
a more intelligent solution, which would leave the data of the neighbour
telescope for a longer time in the crate and initiate a readout only, if the
triggering event was accepted in higher levels. This scheme will be implemented
in FPGA firmware of the SLT.
·
It
is cumbersome to operate 4 FD eyes one-after-the-other and start each DAQ
separately. Thus, a Run Control GUI to steer all FD sites at the same time is
needed. Within the frame of AUGER ACCESS several options for the protocol to
steer the communication were investigated. It is intended to provided a GUI
with reduced functionality (compared to the current options), but still
sufficient for normal FD operation.
Hardware work / changes in Malargüe
·
The
high amount of data for the monitoring and calibration filled the hard disks in
the FD buildings to fast, especially, if the replication to the server at CDAS
failed for any reason. N. Kunka installed during his stay 142 GByte raid data
disks which provide a 3.5-times higher capacity.
·
Sascha
Menshikov repaired defective (or suspicious) front-end boards of the FD
electronics. By adding an additional blocking capacitor on all SLTs the
sensitivity of the internal PLL to spikes was reduced. This solved the “missing
seconds problem” of one SLT module.
·
A
problem with the LCU leading to multiple ping responses in the LAN was
localised, but needs further investigation in
R&D for Auger North and the HEAT proposal
·
We
have completed the production of a first versions of the HEAT modules:
- The new SLT is completely tested, i.e. the interfaces between the FPGA and
all periphery (memories, backplane buses, PCI, PrPMC module) are working. Basis
read/write operations into the FPGA internal memory from the processor module
are possible with 40 MHz. Minor changes are incorporated into the circuit
diagram. The PCB design for the modified version is currently done.
- With the new SLT and FLT we have started intensive tests of the high-speed
backplane. With our prototypes we could drive the pixel bus as well as the
address bus with 40 MHZ. The abandonment of a +5V supply allows us to add 6
more bus lines, which increases the flexibility of data exchange. These
modifications are incorporated into a revised circuit diagram. The PCB design
will start after the SLT design is finished.
- The FLT prototype was tested together with the new analog board design. We
found no errors in the design, tried successful the gain adjustment and measured
the noise performance and filter characteristic. Also the full data chain on
the FLT from FPGA to ADC and memory (QDR ram and SDRAM) was successfully
working with 40 MHz. The next PCB design for the FLT will start in January
2008.
·
In
total 5 crates are now procured from Wiener of the new type. Of those, 3 are
wired with the prototype backplane, the remaining 2 have to be equipped with
backplanes of the modified type.
WBS 2.1 SD (Ingo Allekote – CNEA)
By the beginning of November,
the tank delivery to Malargue resumed after a long stop due to financial
problems with the trucking company in
In the period October-November 2007, 15 surface detectors were assembled, 32
deployed in the field and 40 filled with pure water. 31 e-kits were installed.
As of November 30, 2007, there were 1559 SDs in the field, of which 1515 were
filled with water.
Batteries still pose a problem, as a huge fraction does not pass the quality
control.
Two insulated tanks, originally designed for thermal test studies, were
assembled. Due to their different inside shape, texture and thickness, some
changes in the assembly procedure were required. New procedures for assembling
insulated tanks are being developed.
There has been some progress
with landowner issues. The
Contact has also been made with the landowners of the fields south of the
Italians (Arias-Arenas). Although a portion of this field is in dispute between
Arias and the Italians, they are willing to grant us access if Auger repairs
some access roads and some bridges. Work is under way.
Negotiations have been carried on with Pietrelli, who owns the land to be
occupied by Amiga. We plan to lease this land with an unrestricted access
permit, to use it as a test ground for various experiments and prototypes. We
expect to reach an agreement with Pietrelli very soon.
WBS 2.2 Surface Detector Electronics
(Tiina Suomijarvi – IPN Orsay)
Status on the
“raining” PMTs
The following table shows the current statistics (October 31) concerning unstable PMTs (so called “raining” PMTs).
1429 running stations
414 “unstable” stations
261 stations: 1 raining PMT
121 stations: 2 raining PMTs
32 stations: 3 raining PMTs
599/4287 raining PMTs (≈14%)
The potting procedure was revised and the quality increased early this year. There are currently about 170 stations on the field with new potting. It seems that the PMTs with new potting have less instability problems (about 5%). However, this result still needs to be confirmed.
There is a clear winter-summer effect. During cold weather, the problem decreases, as can be seen in the fig. 1.
Figure 1. Number of unstable tanks as a function of date. A drastic decrease is seen around June-July-August.
Measurements were performed by the test tank Jamie (AB). This tank had at one period a “raining” PMT. Interesting conclusions were drawn from these measurements. During the “raining” period, a significant fraction of the total charge is apparently charging an extra parasitic capacitance, which adds a characteristic time of hundreds of nanoseconds to the dynode chain. This has as effect to decrease the peak value and enlarge the signal (see fig. 2) yielding to an increase of the ToT trigger rate.
Figure 2. Average pulse shape. PMT’s with “raining” behavior have wider pulses.
The average pulse shape is available in the online calibration and it could be possible to implement an off-line T1 by using de-convolution. This would also allow to have a uniform detector trigger efficiency (independent of water quality).
Careful tests of the “raining”
PMTs have been performed in
Currently extra grounding is added for all the PMTs deployed to the field. For those who have already been potted, it is possible to add a grounding cable by slightly removing the potting, as shown in fig. 3.
Figure 3. Extra grounding is being added to the dynode cable connection on the base.
Several actions have been taken to understand and to reduce the “raining” PMT effect:
Increase the quality of potting
Insulate against high temperature
Studies in laboratories and on site
Tracing back components
Furthermore, it has been checked that this effect does not affect
analysis (above the acceptance saturation).
It seems now that the problem is located and a plan for software and hardware repairing is being implemented.
WBS 3.0 Comms (Jorge Abraham
– UTN)
My information to November 30Th:
- AntennaKits mounted: 1509.
- AntennaKits made: 1660. We finish this part of Task !!!!!
- I make a control periodic of the SD Signal quality.(See Image:
RadioSignal.JPG).
About bad links, We will change some AntennaKits for High-Gain Antenna Kit to
solve the problems. May be during the first six month from
2008.
WBS 4.0 Online Monitoring (Cyril Lachaud – APC)
Concerning the Auger Monitoring Task:
A new web release has been installed on both moni in
Malargue and apcpaox in
We have some problems with the configuration of moni in Malargue. There are some IO problems that need to be solved.
The second moni computer will be used to determine the best parameters before replacing the actual one (that will then be used for the nagios system to monitor all the hardware/software of the Observatory inside the CDAS network).
WBS 5.0 DPA/Offline- (Bruce Dawson, Markus Roth and Tom Paul)
A tag for hybrid simulation production has been applied, and tests for simulation production have got underway. Most of the SdCalibrator work required to reflect CDAS v4r6 Ec has been completed. An SdCalibPlotter module has been added which provides visualization tools for the various algorithms employed by the new SdCalibrator. The ParticleInjector module has been substantially revised. Simulated tank calibration constants are now retrieved from the detector description, rather than from a configuration file in the SdSimulationCalibrationFiller module. If you change tanksimulations, the appropriate calibration constants are read from the detector automatically, using the simulator name as a key. Note that the standard simulation module sequences have beenupdated accordingly (now the calibration filling happens after the tank sim). Evaluation of options for a more modern and maintainable build system are underway.
WBS 9.0 Observatory Operations – (Julio
Rodriquez Martino - INFN)
FD Operations
Operation of all telescopes in all buildings was normal, with only minor and
known problems.
The variance limit, that allows to automatically close the shutters, has been
fine-tuned and has worked properly. Several new pages have been added to the FD
monitoring system and it's now a very useful tool for the shifters.
- The bug in the FD elog was solved, now the shifters are able the log
information directly to the database and also to recover from the crash of the
browser as the elog itself is periodically saved. There's also the possibility
of searching in previous entries to see how problems were solved.
- During the December shift, the satellite providing weather imagery for the
site (GOES-12) was taken out of line due to maintenance activities. Even not having
a direct impact on the operation, these satellite images are useful for the
shifters to 'forecast' the future data taking conditions for the next couple of
hours. It is also possible to identify lightning storms around the site that
could cause interference in the DAQ which are quite common during summer time.
- The variance limit watchdog has been fine-tuned to prevent unexpected closing
of the shutters.
- A summary page has been added to the FD monitoring website. It includes a
summary of the background loop data, the trigger rates, the latest APF&CLF
shots and the status of the weather from the Slow Control, all the displays
have associated timestamps and are color-coded to easily detect which variable
has gone out of range or if some information is out of date. Besides this, some
debugging of other php scripts was also performed.
- Several scripts that perform checks before the start of the run were
troubleshooted.
This year's operation was generally normal. The largest problems were due to an
extremely hard winter with several snow storms which prevented normal data
taking.
WBS 9.0 Observatory
Operations cont. – (Ricardo Sato – Southern Observatory)
During October,
2007 the SD operation was quite fine. On the other hand during November, 2007
and beginning of December had some changes and one strange instability of the
array. Following we have a small descriptions of the what happened.
During the same period of the Collaboration Meeting (2007/Nov/12 ~
2007/Nov/17), the instability (all array stop to send data) was related with a
test to send a new Unifier Board software. The previous software version could
not handle properly a problem which looks to be
related with the GPS signal and make stations lost synchronization. The new
software version is supposed to handle properly this problem, although it is
still temporary version.
During 2007/Nov/24 and 2007/Nov/25, there was a general power cut in Malargue
and almost all equipments of CDAS have been turned off, except those which are
really essential to the operation of the array. Unfortunately this effort looks
to be almost lost. Although CDAS looks working fine, the event data was not
recorded because a apparent problem related with the main disk used to store
all event data (for some reason it have been remounted as "read only"
and it was not possible to write the event information in it). This problem has
been fixed only on 2007/Nov/26. People which use the T2 information or log
files should take care with this particular period, since the problem do not
affect those files.
On 2007/Nov/27 part of one equipment at the communication Malargue tower stop
to work and the link of SD channel between Los Leone’s tower and Malargue tower
have lost. In the same day (afternoon) the Los Morados tower stops to work. It
looks to be a power protection which
cut the power in Los Morados. A day after the problem has been fixed and the
batteries used to the communication have been replaced in Los Morados.
On 2007/Dec/02 we lost connection with Los Morados again. The problem looks to
be generated by the power protection in Los Morados tower which turned off. It
looks to be generated by a problem in the general power line in Los Leones and
Los Morados buildings. The voltage in
both buildings was too high and it looks damage also one UPS in Los Leones
(Fluorescence Detector).
At the same time which the problem in Los Moarados has been repaired
(2007/Dec/03), one radio equipment has been replaced because it looks to have
problem with synchronization.
On 2007/Dec/06 the total number of stations working at the same time (looking
T2 information) was much bellow the normal and all the array looks unstable.
The situation became worse with time and the number of working stations on
2007/Dec/08 the total stations working as very
close to 0. The strange thing was almost all station was answering to CDAS
requests and it could send data event data; however the trigger information
(T2) was not recognized by CDAS. Looking to the communication system,
everything looks fine but because of the replacement of equipment in Los
Morados few days before, it looks to be something related with communication.
During this period, many actions have been made (reset communication system,
restart the all CDAS process, and so on). The system have recovered only
turning off all CDAS machine and turning on again.
Radio
Detection R&D – (Ad Van de Berg – KVI)
Continued data-taking is taking
place near the BLS and the CLF. The uptime near the BLS is hampered by power
failures and occasional required hardware resets, but valuable data are still
recorded (KVI,
Results from simulations by FZK, KVI, and SUBATECH on the shower development
and subsequent radio emission are being studied in more detail, leading to new
insights. These results are being incorporated by RWTH into an end-to-end
simulation for shower-induced radio signals. Work has started for the
development of the radio-offline software by the
Near the BLS a new array of
antennas and new electronics will be deployed in early 2008. For this purpose
NIKHEF has produced new 4-channel 200 MHz sampling scopes. KVI has designed and
constructed a local station providing wireless connectivity to the BLS and
local power from solar panels. The group from RWTH has made a new design for a
log-periodic dipole antenna, which will be more stable against heavy wind
loads. These antennas will also be used for additional testing at various sites
in
Funding has been received for the purchase and installation of GHz receivers
near Coiheco for the initial studies of the molecular Bremsstrahlung R&D
project (OSU). Meanwhile work is progressing at FZK, LAL, LPSC, NIKHEF, and
SUBATECH to optimize various components of the station hardware and embedded
software.
HEAT enhancement -Hans Klages (FZK)
Preparations for the HEAT hardware installations are moving forward steadily and with good speed. The second 40` container with all the parts for the tilting base plates, hydraulic etc. arrived in Malargue on November 7. All material is now stored at the Central Campus. On the same day the third container with the steel frames and insulation material for the HEAT enclosures left Karlsruhe. This one is supposed to arrive in Malargue at the middle of December. All the heavy stuff will be on site by then. Norberto Fazzini and Alan Watson went to meet the Argentinean representative of the HEAT landowner to discuss the land lease contract. Although they in fact did not meet him, land access was granted for the building company even prior to the final signatures. It means that the site preparation might really start next week, even before December 1, thanks to all this effort.This is a big step forward for the HEAT subtask. We hope that our own hardware installations will commence early in March.
AMIGA - Alberto
Etchegoyen
AMIGA has now three interconnected tanks with WiFi (Jamie, Larabelle, and Didi) in acquisition mode and sending the events to Didi, which registers the coincidence events. In the near future the system will be used in the field. In short SBC appears to be trustworthy for AMIGA. The FPGA AMIGA programming is nearly finished and its front-end PCBs are shortly to be sent to manufactures.