EVN Technical & Operations Group (TOG) Meeting
Monday 4th December 2006 Noto, Italy
Introduction
------------
The EVN TOG held a meeting on 4th December 2006 at the Noto radio
telescope, Italy. 27 people attended the meeting. The chairman was
Alef. Cassaro welcomed the participants and outlined local
arrangements.
All reports and presentations from the meeting are available on the
web at http://www.mpifr-bonn.mpg.de/div/vlbicor/tog_chair/togreps06-2
1. Approval & last minute additions to Agenda
---------------------------------------------
The agenda was approved, with an additional 10 minutes allocated for
Ritakari to present e-VLBI developments at Metsähovi.
2. Review of Action Items from last meeting
-------------------------------------------
The TOG reviewed action items from the previous meeting.
1. Action: Garrington to make the Mk2/Lovell phase referencing
technique available as a working process for users - ongoing. This
has been tested and some progress made, but option is not yet
readily available to users.
2. Action: All stations should look at data regularly with chchk
program, use it to locate significant RFI and report the results to
Polatidis and CRAF representatives, and use it to check phase-cal
throughout the session - continuing. Polatidis mentioned that some
reports had been recently received and would soon be included in
the Status table. Mantovani noted that the CRAF chairman had
complained about not receiving enough information on RFI. Alef
urged all stations to please take RFI reporting seriously as this
information is important for CRAF. Some concerns were raised about
the sensitivity of the RFI monitoring receivers which are installed
at some stations.
Action Item: Polatidis, Lindqvist and Burgess to define standards for
RFI monitoring and report at the next meeting.
3. Action: All stations should look at pipeline results available from
the JIVE webpage, in particular amplitude corrections found by
selfcal on strong, compact calibrators - continuing. Stations
reported that pipeline results are useful.
4. Action: All stations to provide detailed Wx and RFI comments in
experiment logs - continuing. In general this seems satisfactory.
5. Action: All stations to continue monitoring Tcal throughout
sessions - continuing. This has been OK except there is a problem
for Cambridge (to be discussed later).
6. Action: de Vicente to report on implementation of continuous tpi on
VLBA systems - de Vicente could not attend the meeting due to
cancelled flights, so the action remains open. Graham reported
that some progress has been made.
7. Action: Stations to produce ANTAB and RXG files as soon as possible
after the end of a session - permanent action. Alef to request
feedback from users as to whether they would prefer calibration
data in two stages (first quick preliminary and then final numbers
later) or whether they would prefer to wait longer and have high
quality calibration data to start with - feedback to be presented
at the EVN users meeting in Torun. Evidently users do not want the
two-stage calibration option. The deadline for submission of ANTAB
and RXG files is 2 weeks after the end of a session. It is
desirable to send files sooner if possible. Directors were asked
to give calibration file production a higher priority and allow
Friends more time for this if needed.
8. Action: All stations to forward FS error log files to Himwich and
include details of what the FS was doing when the crash occurred -
ongoing action. No FS crashes were reported since the last meeting.
9. Action: Reynolds and Walker to incorporate frequency agility
information in SCHED - stations will be prompted to provide this
information. Polatidis then to update the EVN status table with
new/additional frequency agility information - action remains open
(Reynolds absent). Campbell reported that frequency agility
information is expected to be included in the Spring 2007 release
of SCHED.
10. Action: All stations to look at NME reports sent by JIVE -
ongoing. Stations regularly look at NME reports.
11. Action: Whitney to find out whether a spare extender cable is
available at Haystack to loan to Torun for testing BBC problems -
Smythe reported that this was done.
12. Action: Alef to send information regarding monitoring of
scan_check during recording - how to check for discrepancies in
scan values and flag problems in the display window - there was
uncertainty over what specifically this action meant. Smythe to
explain scan_check in his presentation.
13. Action: Himwich, Graham and Gunn to include Graham's monitoring
software with future distributions of the Field System - Himwich
to discuss the available FS monitoring and display options in his
presentation.
14. Action: Gunn to send email before each session when the final
versions of all schedules are ready for download - ongoing. This
has been satisfactory. Campbell usually prompts Gunn to send the
email.
15. Action: Reynolds to request feedback re. amplitude calibration
from spectral line users at JIVE in order to assess results of
different calibration methods - status unknown at meeting, but
subsequently Reynolds reported that feedback from spectral line
users has been received. A lot of spectral line observations also
need good quality a priori Tsys.
16. Action: All to check that Mk5 modules are placed squarely on a
flat surface when received, otherwise connectors are easily
damaged when bent 8-packs get inserted in Mark 5 units - ongoing.
Campbell mentioned that one or two damaged modules have been
received at JIVE. Alef reiterated that it should be standard
procedure to avoid backplane damage, as replacement is very
expensive. Currently there are no spare backplanes available -
later discussion
17. Action: All stations to try to run the FS diagnostic tests and
investigate the results - ongoing. Lindqvist reported that the
tests had been done at Onsala and worked well. He recommended that
others try the tests.
18. Action: Reynolds to look into scheduling of Lovell+MkII for
phase-referencing - done, see following item.
19. Action: Paragi to arrange a test of the Jodrell MkII/Lovell
phase-referencing technique for a C-band NME - done in N06C5.
20. Action: Reynolds to work with Rene Vermeulen on the guide for
polarization calibration using WSRT local interferometry data -
ongoing. Foley reported that Westerbork regularly monitors
calibration sources locally. A user guide is still needed.
21. Action: Gunn to present an inventory and status report on Mk5/disk
spare parts at the next TOG meeting - Burgess maintains all
information on the Jodrell web site (see
http://newton.jb.man.ac.uk/Mark5A/sparemk5.htm ). Burgess
requested that all updates be emailed to him.
22. Action: Alef to establish contact with Brian Corey regarding DBBCs
- closed. A DBBC test will be done between Noto and Westford.
23. Action: JIVE support scientists to give detailed feedback to
Shanghai regarding problems with the K-band data from session
3/2005 - Bignall discussed plots with An Tao who had some ideas
about possible causes of K-band problems at Shanghai. As yet
there have been no subsequent K-band observations scheduled.
24. Action: Paragi to implement labelling of fringe plots with sky
frequency and sideband - still to be done
25. Action: Orlati to ask the vlbeer maintainer (Mauro Nanni) to
create a subdirectory called e.g. 'latest' or 'checked' for
stations to download schedules from. Alef, Campbell, Foley and
Gunn to distribute an email to all stations detailing the new
procedure - done
26. Action: Reynolds to fix the problems with ANTABFS for VLBA racks -
done
27. Action: Graham, Lindqvist to form a small group to discuss via
email opacity corrections at high frequencies, and prepare a
document to present at the next TOG meeting - presentation by
Graham
28. Action: Alef to contact Phil Diamond regarding calibration of Cambridge
data - done, but there are still some problems
29. Action: Graham, Himwich and Lindqvist to solve problem with flux
estimates for primary calibrators (Cas A etc.) in FS - done
30. Action: Alef to include on TOG webpage the collection of URLs where
stations can look for experiment feedback - done. JIVE webmaster
(Lisa Harvey-Smith) to add this also to 'VLBI links' on EVN web
site - still to be done
31. Action: Campbell to write a summary of data products available
from JIVE - done
32. Action: All stations should condition disk packs if time permits,
especially those which are to be used for 1Gbps recording -
continuing. This is being done on a best-effort basis. There was
some discussion of disk conditioning procedures. Smythe
recommended all to look at the report of conditioning and the
status display in order to identify slow disk drives. Any slow
disks should then be assigned to low data rate experiments. Use
the stickers on the modules for reporting slow disks, as these
will reduce the available bit rate. Burgess noted that Jodrell
tries to replace any slow disks with spares. The main priority is
not to lose any user data. When the capacity of a disk pack
changes, the VSN should be retired and a new one started. Alef
keeps track of changes in the disk inventory.
33. Action: All stations which do both astronomy and geodesy should
clearly distinguish between astronomical and geodetic 8-packs and
ensure there is no "leakage" into the wrong pool - ongoing.
Leakage has not been a big problem, although Alef noted that ~10
TB of disk-space is unaccounted for. It is recommended that
geodetic stations use different numbering systems to easily
distinguish between astronomical and geodetic 8-packs. If any
geodetic packs are found at JIVE, these should be sent to
Effelsberg for return to the geodetic pool.
Action Item: Alef to distribute astronomical/geodetic disk database
information to Haystack.
34. Action: All stations to ship disk packs as soon as they are full
- ongoing. According to the 'Bologna Rules', recorded disks should be
shipped at least once a week.
35. Action: Disk packs should be shipped with one European and one US
address on covers so they can be easily returned in case a
shipment is lost - ongoing. Some clarification of this item was
needed - this does not include any disk pack labelling, simply
requires attaching addresses to the metal cover "If lost in
transit, send to
in the US or in Europe".
36. Action: Foley to start a wiki page for FS documentation, which
all users can contribute to - remains to be done.
37. Action: Himwich to send message to EVNtech regarding the location
of the FS "toolbox" package to install - done.
38. Action: Whitney to check the status of the problem with recording
stopping, or Mk5 not responding for a period of time, when the key
is turned (for changing a module while the other is
recording) - Smythe reported that this has been fixed.
39. Action: All to send e-mail to Alef within 4 weeks (i.e. by 21
April) regarding the feasibility of attending a TOG meeting at
Arecibo. Include any conditions, such as financial support needed
- Alef thanked participants for their positive responses, however
the TOG meeting in Arecibo was not approved by the CBD.
Burgess noted that due to spam problems, the EVNtech distribution list
has been rebuilt, and everyone needs to resubscribe if they haven't
already. Ritakari suggested that an archive of EVNtech mail would be
useful. Others agreed.
Action Item: Burgess to look into making the EVNtech mail archive
accessible on the web.
3. Reliability/Performance of the EVN
-------------------------------------
Campbell reported on Network reliability (on behalf of Reynolds; see
presentation online)
Campbell noted the recent change of roles at JIVE. Paragi has a new
e-VLBI postdoc position at JIVE and is no longer responsible for the
NMEs. Stations should contact Campbell and Reynolds regarding NMEs.
The changed schedule download procedure introduced in session
2/2006 seems to have worked fine.
"Near-realtime" fringe checks have been successful in catching
problems.
The UHF experiments in session 2/2006 suffered from severe
RFI. As the observations were outside protected bands, it may not be
useful to report these to CRAF. However Smythe commented that low
frequency RFI problems can sometimes be solved and cited some examples
at Haystack of fixes, implemented by cooperative neighbours, to stop
RFI due to electronic equipment.
Some data in session 2/2006 were lost due to intermittent
communication problems between the FS and Mk5 unit - Smythe to comment
on this later. Outside the session, there were many problems during
the e-VLBI ad-hoc experiment to observe RS-Oph.
There was some discussion of faulty disk replacement. Sometimes
faulty disks sent back to stations have not been clearly marked as
broken, so better communication (labelling and email) is needed when
faulty disks are returned to stations. Smythe noted that for some
disk brands (e.g. Hitachi, Maxtor, possibly Seagate) it is very easy
to obtain replacements online under warranty from anywhere. A return
merchandise authorization (RMA) can be obtained from the
manufacturer's web site. By entering the serial number, information is
provided on the warranty status and whether the disk can be
returned. Schonewille reported that JIVE had recently obtained 5 new
disks this way.
The introduction of disks has significantly improved the EVN
Reliability Indicator (ERI, fraction of "good" visibilities out of
expected visibilities) by roughly 15%. Also the processing factor,
defined as the ratio of correlation time to observing time, has
significantly improved. For non-Gbps experiments the processing factor
is <1.5, however for Gbps the processing factor is larger. The main
limitation is the station units, so this is expected to improve with
the introduction of Mk5B. Initially following the introduction of Gbps
recording there were some problems with video cable interference, but
these have subsequently been avoided.
Correlation of more than two MERLIN antennas has been demonstrated
successfully. Cambridge and Darnhall were recorded on one pack and
data were copied to another pack at JIVE. This worked OK, only some
editing of the vex file was needed. This could be made a standard
procedure for users. Burgess estimated that the option would be
viable for two years, as the bandwidth to MERLIN outstations is
unlikely to be upgraded for at least a year.
After discussion it was decided that this could be done as the default
for joint EVN+MERLIN observations. Users then should not need to
specifically request the extra MERLIN antenna to be correlated at
JIVE, however they should know in advance what data they will get.
In special cases it may not be desirable to correlate an extra MERLIN
antenna (e.g. if spectral resolution is compromised).
The CALC model has been upgraded. It can now make use of atmospheric
pressure information in calculating tropospheric delay. Thus recorded
weather information is now important.
Action Item: (Non-geodetic) Stations to check weather log in FS, as
atmospheric pressure information can now be used. Updates every 5--10
minutes should be adequate for corrections.
Campbell summarised data products available at JIVE: NME reports,
Standard plots made before FITS files are produced which show small
subsets of the data, and pipeline products. Currently everything is
linked by experiment. It would be useful to also have links to data
products by station. Some thought would need to go into how to set
this up. There was some discussion of effective feedback to
stations. It is important to email individual stations if a previously
unknown problem is found. Also it would be useful for JIVE support
scientists to include mapping of subband to BBC in the PI cover
letter, which stations can also view on the archive. There was a
complaint about some pipeline gain calibration data not being
accessible on the archive. In general these should be public, so JIVE
support scientists should check that the password protection is set
correctly (for proprietory data only) after archiving the pipeline
products.
Charlot reported on phase-cal investigations with geodetic data (see
presentation online). He showed plots of amplitude versus phase of
phase-cal tones. Phase-cal amplitude should be constant with phase,
however a problem was found with Urumqi phase-cals. More recently a
new X-band receiver has been installed at Urumqi, so the problem may
not be there any more. JIVE can now extract two phase-cal tones so in
principle could produce similar plots for experiments with phase-cal
on. Stations can also plot all their own phase-cal tones with the
decoder. Himwich is to add an amplitude versus phase plot feature. It
was noted that inspection of phase-cals is a useful diagnostic for
many things, not only receivers.
Action Item: Campbell to look into producing phase-cal plots at JIVE.
Polatidis noted that more information on RFI is needed from
stations. Updated information for two stations has just been received
and will appear in the status table soon.
RFI feedback from JIVE (and WSRT plots from Foley) can also be sent to
Polatidis and CRAF, especially from spectral line experiments.
Alef reported on feedback from the questionnaire. Stations perform
various checks before the session. A good approach is to treat each
session as a project, and plan several weeks in advance for what needs
to be done. It is useful to produce an easily readable plan of the
block schedule at each station. Some stations use a web-based system
for planning and checks. Jodrell was still to send their feedback to
Alef, however Burgess noted that Gunn has produced a web display for
use at Jodrell - anyone interested in using such a display can contact
Burgess.
One important pre-session check is to route all 64 tracks through the
decoder in order to check all connections etc. It is also important to
establish tight communications so that operators know if anything in
the system has been changed, and can quickly telephone in case of
problems. It is recommended that Friends monitor experiments remotely
(tail log files etc.) whenever possible. PC and FS clock checking is
crucial. An operator checklist is very useful.
During the session, the 'greplog ' command in the FS is useful
to inspect errors logged. The snap command 'lastcheck' is also handy
for a quick look at the last scan_check or data_check. Regular
monitoring of scan_check is important. It is also important to monitor
BBC levels and lock status. Foley noted that at Westerbork, they have
set up for some error messages to ring the control room phone to
alert the operator.
There was some discussion of the use of VNC for monitoring operations
remotely. Burgess said he would post some notes on how to use VNC
through ssh.
There was some further discussion of checking IF levels by sampler
statistics, using the systest procedure library. Alef emphasised that
it is important for all stations to measure sampler statistics every
session.
Action Item: JIVE support scientists to include checking of sampler
statistics as part of NME feedback.
4. Amplitude Calibration
------------------------
Campbell reported on amplitude calibration (on behalf of Reynolds; see
presentation online).
Generally the timely availability of amplitude calibration data has
improved, although the two-week deadline is not always being met. The
latest ANTABFS software version is 3.4 (Effelsberg use their own
software).
The calibration of e-VLBI data needs to be addressed. A calibration
scheme is being worked on at JIVE.
There was some discussion of the problem found with N06M2 pipeline
results - as outlined in Graham's presentation (available online),
gain corrections derived from amplitude self-calibration were
inaccurate because of complex source structure in the source selected
as the calibrator. The automatic pipeline did not work well in this
case. When the "target" source was selected for self-calibration in
the pipeline, amplitude gains were more constant and closer to 1
as the source structure was much simpler in this case. The archive
could be updated with pipeline results after swapping the target and
calibrator in this case. Stations cannot be expected to assess the
effects of source structure in inspecting amplitude gain statistics
from the pipeline, so this needs to be addressed at JIVE. Possibly
closure quantities could be used to assign a quality flag to a
source. Charlot commented that problems could be avoided with careful
source selection.
Action Item: JIVE support scientists to look into a scheme to estimate
the quality of pipeline amplitude calibration results.
Graham also raised some concerns about the gain corrections derived by
the AIPS pipeline. His investigations suggested more accurate results
could be obtained with Difmap using 'gscale' after phase self
calibration.
No major problems from calibration runs were reported by stations. A
minor feature still to be implemented in GNPLT is separate receiver
temperatures for LCP and RCP.
Graham reported on 22 GHz calibration, in particular the issue of
whether gain curves include opacity corrections or not. Alef said
that a uniform approach at all frequencies would be preferable. He
suggested that all gain curves be free of correction for opacity.
This approach of effectively taking the telescopes outside the
atmosphere is used at the VLBA. Graham has checked that GNPLT produces
gain curves free of opacity in the correct way. PIs are then
recommended to apply opacity corrections using APCAL in AIPS
(important for frequencies above 10 GHz). Spectral line measurements
can be calibrated using autocorrelation spectra in which case it is
not a problem if PIs do not apply the opacity corrections with APCAL.
Action Item: Friends to provide gain curves free of opacity
corrections (as if taken outside atmosphere). Campbell to inform
users. Also PIs of 22 GHz experiments to be informed that Onsala has a
different method of calibration at 22 GHz (hot load instead of noise
diode), which means that Tsys values are already corrected for opacity.
Graham's presentation included the status of continuous 80Hz
calibration for VLBA terminals. The code is currently in a local
version of FS 9.7.7. Some recoding is needed to incorporate it in a
more modular structure. Graham presented a plot of Tcal versus
frequency for 80Hz switched cal, showing that 80Hz reported a 5% lower
Tcal than on/off method, probably because of blanking and risetime
effects.
Burgess was asked to report on the status of Cambridge calibration.
Some progress has been made on the hardware. Currently, implementation
is mainly limited by manpower, but some tests may be possible for the
next session.
From stations represented at the present meeting, it was felt there
was no need for a calibration "tiger team" at present.
Use of Westerbork and Effelsberg flux measurements was discussed.
Foley asked where is an appropriate place to put the WSRT flux density
measurements. The availability of these data should be emphasised to
PIs. Effelsberg also regularly measures standard calibrator
fluxes. (PIs still need to make a special request if they want single
dish flux measurements for their program sources.)
Action Item: Campbell to determine an appropriate place to put
WSRT and Effelsberg calibrator flux measurements.
Foley reported on Westerbork status. As the tied array is not
controlled through the Field System, some work was needed to implement
off-source flagging. For the flag file it was was decided to define
"on-source" as when a majority of the antennas in the tied array are
on source. This is now implemented, with 10s granulation.
Comment strings are written to the log file when (1) most antennas are
on source, and (2) when all antennas are on source.
5. Mark 5
---------
Smythe reported on the status of Mark 5A/B/B+. Mark 5B features full
VSI-H compatibility. The system is now in regular use at Westford. The
difference between Mark 5B and B+ is the Amazon StreamStor board,
allowing a maximum data rate of 2Gbps.
Action Item: Smythe to send email reminding stations of procedures for
checking recording.
Briefly, important checks include: scan_check for missing or extra
bytes which could indicate cable interference, in which case it is
recommended to look at errors on the decoder and check that cables are
not near metal objects; tail -F ; checking sampler levels.
Smythe noted that Burgess has a clearly written summary of the
diagnosis and repair of the data "throttling" problem found at Jodrell,
which turned out to be due to a faulty chip in the clock module.
Mantovani mentioned that some stations are entitled to a free upgrade
from Mark 5A to B.
Alef gave a brief summary of EVN Mark 5 status.
Smythe gave a presentation on Mark 5 software status. In the Conduant
SDK 7.5 release, the problem of recording stopping when the key switch
is turned has been fixed.
Szomoru commented that possible unresolved issues for the correlator need
to be investigated before upgrading Mark 5 software.
Action Item: Smythe will contact Brisken to organise a teleconference
with NRAO on Mark 5 software status.
Smythe summarised development of application software to control the
Data Input Module ("dimono") and Data Output Module ("domino")
capability. Mark 5B DOM support is correlator specific.
Smythe also summarised Mark 5 OS Software status. Linux Redhat 9 is no
longer supported, Debian "Sarge" Linux is the new standard OS, with
plans to support Debian "etch". Chet Ruszczyk plans to distribute the
OS update as a standard ISO image.
Points were made that it is critical that the Conduant software and
Jungo driver work with the new Linux distribution. Any problems must
be sorted out for Mark 5B.
Graham is to keep EVNtech informed about solutions to current
problems.
Alef presented the Disk Inventory status. Updates were still needed
from a number of stations. 835 TB were accounted for in the EVN pool,
out of 1038 TB purchased in total. The numbers are approximate as some
stations do mm and geodetic observing. Still, approximately 160 TB of disk
space was unaccounted for. This is a considerable amount as 100 TB
costs EUR 35k. Disk availability is sometimes an issue. It should be
possible with current resources to schedule 400 TB in a session - if
not then we should find out what the problem is.
Campbell reported on the status of disk throughput at JIVE. The EVN
aims to recycle disks within 2 sessions (for comparison the VLBA has a
<30 day turnaround). Campbell encouraged all stations to keep
informing JIVE of preferred shipping companies and to communicate
prices of shipping in order that the costs of shipping to different
stations can be compared.
Alef reported that the Scheduler aims to make 1 Gbps recording the
default for continuum experiments in 2007, although it is preferred to
"ramp up" to this rate rather than make a sudden step.
Mark 5 logistics were discussed. It was decided that if possible, a
faulty disk should be sent directly to the manufacturer for
replacement, from wherever it is found. If the warranty does not allow
for this, then the faulty disk must be sent back to its station. If
a faulty disk is sent back from JIVE to a station, then JIVE replaces
the disk with a new one. Schonewille reminded stations to which faulty
disks are returned to please send replacement disks to JIVE.
Lindqvist enquired about the disposal of tapes. In general tapes can
be discarded, except approval is needed to do anything with those
from JPL/DSN which are US government property.
Action Item: Alef to send email to stations regarding disposal of
tapes.
At the last TOG meeting is was agreed to try plastic envelopes for
disk labels similar to NRAO, but these turned out to be difficult to
find. Most people are happy to keep using luggage-tag labels.
6. Field System, status and new features
----------------------------------------
Himwich reported on Field System status. The current release is 9.9.2,
with 9.9.3 expected in Winter 2007. In the current FS, automated
(schedule-initiated) FTP has been implemented based on the approach
used by Burgess. It is controlled from the 'Data Transfer Override'
menu, and uses a locally customized autoftp script which transfers
data using ssh (RSA key-pair exchange).
7. SCHED Developments
---------------------
Campbell reported on SCHED developments (on behalf of Reynolds).
Stations are urged to check that catalogs are up-to-date.
Implementation of frequency agility information in SCHED is possible,
but it is currently not entirely clear how useful this would be for
the EVN in practice.
There was a brief discussion of how to phase in AutoFTP for user
experiments, e.g. if these should be inserted into fringe-finder scans
at JIVE (which would require SCHED key files to be re-processed at
JIVE) or if PIs should be advised to include ftp scans themselves.
Alef suggested it would be good as an extra check to do periodic ftp
fringe tests throughout the session.
8. Sessions
-----------
Some users require more frequent observing. Friends were asked about
the feasibility of doing more frequent, shorter sessions (e.g. 4, 5,
or 6 sessions per year of about 1 week). The conclusion was
that in principle this would be possible for stations, but would be more
work, as most of the work goes into setting up for a session. There was
some concern expressed that more frequent, shorter sessions would
decrease reliability.
9. Digital BBCs
---------------
Tuccari gave a presentation on the status of the EVN DBBC project. The
most commonly used bandwidth configurations are implemented in DBBC
v.1.1. For DBBC v.2.0, a faster AD sampler and FiLa10G board are being
developed in collaboration with Michael Wunderlich at MPI. An RFI
mitigation board is also under development. DBBC technology is also
being used in the development of an L-band digital receiver at
Effelsberg and Noto. A reduced version of the DBBC was used (with
PC-EVN recording) in recent fringe tests with Evpatoria (Ukraine) and
Irbene (Latvia) in August and November 2006, respectively. With
minimal architecture the DBBC costs EUR 6k. For a complete system the
cost is from EUR 30k up to 65k depending on configuration.
Ritakari briefly discussed the possible use of two 10Gbit/s Ethernets
in DBE boards (as referred to in his email to EVNtech of 15 November).
Smythe reported on DBE development at Haystack (see presentation
online). A test of the DBE prototype, using it simultaneously with
Mark 4, produced excellent agreement between fringes from the two
different backends. The polyphase filterbank is cheaper than Tuccari's
DBBC but not as flexible. Development over the next 2.5 years is being
driven principally by the VLBI2010 vision which aims at flexible
observing between 100 MHz and 15 GHz.
10. Spare parts
---------------
Burgess reported on Mark 5 spares and support. A list is maintained on
the web site http://newton.jb.man.ac.uk
Updates to spare parts and documentation of fixes should be emailed to
Burgess.
There was some discussion over whether it would be useful to purchase
spare backplanes. Funds would be required for this.
Action Item: Alef to ask Mantovani if Directors can donate money to
purchase spare Mark 5 backplanes.
Smythe noted that it is important to check the screws connecting the
voltage supply to the backplane as these connections are a potential
source of problems.
11. e-VLBI status
-----------------
Szomoru reported on e-EVN developments (see presentation online). A
new Network Engineer for EXPReS has just started at JIVE; there was
some delay in filling this position. Many new (web-based) monitoring
tools and web-based post-processing are under development. Options for
adapative scheduling are being investigated; there are a number of
issues to consider. Szomoru summarised connectivity status of EVN
stations and improvements being made in various data transfer
protocols. He also noted that lightpaths across Europe are currently
available and affordable. The final date for delivery of EXPReS is in
3 years' time, however most deliverables in the Document of Work are
due in 2 years.
Smythe reported on e-VLBI projects underway at MIT Haystack (see
presentation online). Automated e-transfers are being done, where
data are transferred from Japanese stations in K5 format, then
converted to Mk5 format, written to disk and correlated. E-transfers
from more stations are planned soon. VSI-E testing is ongoing. e-VLBI
tests with Pie Town are scheduled for 2007.
Ritakari reported on e-VLBI developments at Metsähovi (see
presentation online). He discussed use of a realtime version of the
tsunami protocol with a fast server for large-bandwidth real-time data
transfer. He also discussed the possible use of Playstation 3s as
processors for software correlation. Some of the new iBob technology
will be used in the Mileura recombination experiment, planned to be
installed in 3 months' time.
12. Station & correlator reports
--------------------------------
Foley reported on Westerbork VLBI operations (the presentation is
available online). Westerbork operations have dropped in ASTRON's
priority to favour LOFAR developments. The Digital Tied Array (TADU)
is designed to interface directly to a Mark 5B. A design flaw was
found in the digital filtering which has delayed the project. However,
TADU is expected to be working some time in Spring 2007, at which time
tests with other stations will be requested.
Other station reports were not presented orally but are available to
read on the web. It was noted that urgent maintenaince is required at
Medicina (a drive) and Noto (azimuth rail). Directors should do all in
their power to ensure the continued reliable operations of these vital
EVN stations. It was also noted that the new subreflector at
Effelsberg is now in place.
13. AOB
-------
None.
14. Wiki for the TOG?
---------------------
A TOG wiki page was started on the RadioNet (MoinMoin) wiki but this
has not been used much so far and was not felt to be an appropriate
place for technical documentation.
15. Time and Place of Next Meeting
----------------------------------
Alef suggested that two TOG meetings in the next year would be useful
to evaluate the phase-in of Mark 5B and DBBCs.
It was decided that the next TOG meeting would probably be held in
early May, possibly in Bonn or possibly even combined with the TOW in
Haystack (30 April - 3 May). The meeting should preferably be at least
4 weeks before session starts on 31 May. This would also be close to
the time of DBBC and EXPReS CDRs, and the Global mm session on 10-15
May.
It is anticipated that the TOG will have another meeting combined with
the CBD meeting in Yebes (12-14 November), in which case it may be
desirable for the session to finish earlier than 8 November. Alef will
consult with the scheduler on this matter.
Action Item: All to send constraints for the next TOG meeting to Alef.
16. Summary of action items
---------------------------
Action Item 01: Garrington to make the Mk2/Lovell phase referencing
technique available as a working process for users.
Action Item 02: All stations should look at data regularly with chchk
program, use it to locate significant RFI and report the results to
Polatidis and CRAF representatives, and use it to check phase-cal
throughout the session.
Action Item 03: Polatidis, Lindqvist and Burgess to define standards
for RFI monitoring and report at the next meeting.
Action Item 04: All stations should look at pipeline results available
from the JIVE webpage, in particular amplitude corrections found by
selfcal on strong, compact calibrators.
Action Item 05: All stations to provide detailed Wx and RFI comments
in experiment logs.
Action Item 06: All stations to continue monitoring Tcal throughout
sessions.
Action Item 07: de Vicente to report on implementation of continuous
tpi on VLBA systems.
Action Item 08: Stations to produce ANTAB and RXG files as soon as possible
after the end of a session, within 2 weeks.
Action Item 09: All stations to forward FS error log files to Himwich
and include details of what the FS was doing when the crash occurred.
Action Item 10: Reynolds and Walker to incorporate frequency agility
information in SCHED.
Action Item 11: All stations to look at NME reports sent by JIVE.
Action Item 12: Gunn to send email before each session when the final
versions of all schedules are ready for download.
Action Item 13: Reynolds to report on the feedback re. amplitude
calibration methods from spectral line users at JIVE.
Action Item 14: All to check that Mk5 modules are placed squarely on a
flat surface when received, otherwise connectors are easily damaged
when bent 8-packs get inserted in Mark 5 units.
Action Item 15: All stations to try to run the FS diagnostic tests and
investigate the results.
Action Item 16: Reynolds to work with Rene Vermeulen on the guide for
polarization calibration using WSRT local interferometry data.
Action Item 17: Paragi to implement labelling of fringe plots with sky
frequency and sideband.
Action Item 18: JIVE webmaster to add URLs where stations can look for
experiment feedback to 'VLBI links' on EVN web site (see TOG webpage)
Action Item 19: All stations should condition disk packs if time
permits, especially those which are to be used for 1Gbps recording.
Action Item 20: All stations which do both astronomy and geodesy
should clearly distinguish between astronomical and geodetic 8-packs
and ensure there is no "leakage" into the wrong pool.
Action Item 21: Alef to distribute astronomical/geodetic disk database
information to Haystack.
Action Item 22: All stations to ship disk packs as soon as they are
full, at least once per week.
Action Item 23: Disk packs should be shipped with one European and one US
address on covers so they can be easily returned in case a shipment is
lost in transit.
Action Item 24: Foley to start a wiki page for FS documentation, which
all users can contribute to.
Action Item 25: Burgess to look into making the EVNtech mail archive
accessible on the web.
Action Item 26: All (especially non-geodetic) stations to check
FS weather logging, as the information can now be used.
Action Item 27: Campbell to look into producing phase-cal plots at
JIVE.
Action Item 28: JIVE support scientists to include checking of sampler
statistics as part of NME feedback.
Action Item 29: JIVE support scientists to look into a scheme to
estimate the quality of pipeline amplitude calibration results.
Action Item 30: Friends to provide gain curves free of opacity
corrections (as if taken outside atmosphere). Campbell to inform
users. Also PIs of 22 GHz experiments to be informed that Onsala has a
different method of calibration at 22 GHz (hot load instead of noise
diode), which means that Tsys values are already corrected for opacity.
Action Item 31: Campbell to determine an appropriate place to put
WSRT and Effelsberg calibrator flux measurements.
Action Item 32: Smythe to send email reminding stations of procedures
for checking Mark 5 recording.
Action Item 33: Smythe will contact Brisken to organise a
teleconference with NRAO on Mark 5 software status.
Action Item 34: Alef to send email to stations regarding disposal of
tapes.
Action Item 35: Alef to ask Mantovani if Directors can donate money to
purchase spare Mark 5 backplanes.
Action Item 36: All to send constraints for the next TOG meeting to Alef.