Ruf opened the meeting at 09:06. Alef told us some latest news about correlation of the May/June 1997 CAL experiment.
Gurvits presented a status report of EVN operations. The report was distributed in the meeting. General performance of the EVN shows signs of improvement, but problems remain. Major problems to be addressed include pre-session setup and checks, near-real-time diagnostics, switch to thin tape, and site-specific problems (particularly for new network telescopes).
Action: Gurvits contacts Richard Schilizzi (JIVE) to enforce the EVN Board of Directors' meeting decision in March 1997 to provide 24 hours before the beginning of a given session for system checkout, plus an additional 12 hours before each frequency change.
Porcas argued that a permanent physical setup would require less system checkout time, but Foley answered that for example for Westerbork a permanent setup was impossible. Tuccari suggested to minimize points that need to be changed/reconfigured for VLBI.
Perea described how schedules reach DSN/Madrid via JPL and how
many times details are missing. He asked where he should
ask for clarifications. Porcas answered that if
proposal information doesn't reach a given station then the
station may contact the PI directly---contact information
(email) should be present on the cover sheet and schedule
files. Foley confirmed that this is what he has been
doing whenever there is something unclear as to what the PI
wants the station to do in a particular experiment.
Porcas warned that many PIs are ``amateurs'' in the
VLBI technical sense and that they may be relying on JIVE
experiment support, and in this case it may be appropriate
to send the question to the PI and at the same time
``CC:'' it to ``jive@jive.nfra.nl''.
Gurvits asked if the ``.drg'' contains enough information
and Wiik replied that unfortunately not.
Polatidis stated that a standard cover letter would
usually be enough, but Wiik complained that recently
the only standard thing in cover letters has been that they
have been empty.
Ruf reminded that the schedule files should be available two weeks before the session start and this should leave enough time to get clarifications from PIs---but this deadline seems to be slipping.
Action: Porcas to ``publicize'' on the
``evntech@nfra.nl'' mailing list the email addresses of
the PIs of a given session at the moment when the schedule
files should be frozen on ``astbo1''. This enables
station technical friends to easily contact PIs for
clarifications. (Editor's note: Actually why couldn't this
email message be the aforementioned ``block schedule with
cover sheet information''? The appearance of this message
could also signify that the schedule files are frozen...)
Gurvits asked about the usefulness of the station checklist. Ruf commented that since we encounter wrong polarization how the checklist can help. Altunin explained that DSN schedules pre-calibration time 1.5h before each session and he would like to know if there is a recommended procedure for checking dual-pol 1.3cm, maybe with W3OH. Porcas thought that each station should do its own station checks and Alef added that we could compare this to how GPS was implemented in EVN.
Tuccari described how they add extra noise to only one IF at the receiver and check the IF with a spectrum analyzer at the IF distributor in the rack. Ruf was skeptical since this wouldn't have resolved the recent problem at Jodrell when a primary focus receiver was accidentally mounted at secondary focus and polarization was thus reversed. Foley suggested irradiating the whole dish. Conway suggested that a check of the cables is necessary in the checklist but not a guarantee that polarization will be correct. Pazderski told us that Torun uses a helical antenna to irradiate the disk. Conway summarized that 1) cables, 2) polarizer, 3) reflectors can be wrong and this induced Burgess to suggest and favor an end-to-end test for checking polarization. The discussion continued to ponder pros and cons of an irradiating end-to-end test. Alef asked what has changed recently which could explain recent problems with polarization. Burgess told he had changed his working location at Jodrell, it is not anymore next to the receiver guys. Conway commented that usually something gets touched which has not been touched for 20 years.
Ruf proposed a test tape one week before session start, Gunn told that this wasn't possible for instance at Jodrell. Gurvits asked if we will organize polarization check observations for those who can participate. Tuccari commented that this test would ensure the first part only, since any changes such as a frequency change will invalidate the results of the ``week before'' test. Alef stated that the TWG came to the same conclusion years ago. Gurvits recommended trying W3OH at 18cm.
Alef requested recording quality checks to be made at stations. Ruf concluded that we should stick with the checklist, especially regarding to recording quality issues.
Gurvits noted that upgraded Mark IV would be fully utilizable only after we have new features in scheduling software. We have lack of manpower here, since Huib van Langevelde (JIVE) (who has been enhancing SCHED together with Craig Walker, NRAO) should be working on JIVE correlator software. Also NRAO/VLBA is not directly interested in adding Mark IV-specific features into SCHED.
Action: Gurvits: to recommend to the EVN Board of Directors (BD) that manpower to scheduling software should be allocated.
Porcas noted that in the past scheduling software emerged from the user community. Ruf suggested that this issue should be brought up in the next BD meeting in October in Onsala (held together with EVN Users' meeting).
Gurvits asked if it has been easy to access
``astbo1.bo.cnr.it'' for experiment feedback and
depositing logs and GPS information. Venturi stated
that the VAX machine may be eventually replaced and the
communications line improved. Gurvits encouraged all
stations to report access problems to JIVE and they'll try
to find a way to help. He further told that a QASP
(Quality Assurance Program) team is on its way to the
Chinese stations Shanghai and Urumqi in summer 1997,
1+5 people.
Gurvits reminded that tape shipments are to occur weekly during
sessions on Mondays to Socorro and on Tuesdays to Bonn.
He also promised to take care of the action item on him to
have the EVN block schedule in ASCII format on JIVE Web
pages. Gunn liked to know if we can rely that this
will be the master copy which will be absolutely
up-to-date. Gurvits also reminded stations to post a
dummy log onto ``astbo1'' whenever they did not
observe although they had been scheduled.
Gurvits asked if it is clear which subdirectory of
``astbo1'' to use and Venturi answered that the
starting day of a given experiment dictates the
subdirectory name, if it is, say, ``EVN.MAY97'' or
``EVN.JUN97''. Porcas noted that this is said
on the PI instruction sheet. Conway asked about the
leading zeros in experiment codes and Alef suggested
that they should be eliminated, i.e. ``EL20'' instead
of ``EL020''.
Burgess presented the key ares in which FS developments are desirable. These are:
drudg''.)Mujunen described shortly that a Linux kernel and distribution update is necessary to support the latest PC hardware, to avoid certain network security problems with the old distribution, and to enhance interoperability in a larger network of other Linux computers. Foley asked if this upgrade should be used in November session and Porcas reminded that three stations are not yet using FS9. Hase commented that Wettzell will start using FS9 before end of 1997. Himwich stated that Urumqi may be upgraded in conjunction with the QASP visit and that Simeiz may get a loaner disk which is currently in Brazil.
Perea wanted to know if a single VEX/``.drg'' is at the
moment sufficient for FS9 observing and Himwich
stated yes---if this doesn't seem to work, he and Nancy
Vandenberg (NVI/GSFC, ``nrv@gemini.gsfc.nasa.gov'')
would like to know about the problem with ``drudg''.
Mujunen added that although the Linux upgrade will most probably be available in November 1997 it will not be necessary to upgrade to run the November session. At some later date the FS may refuse to compile/run on the old 1.2.12 Linux platform and require the new Debian 1.3.1.r3-based kernel version 2.0.30. Altunin commented that it will be difficult for DSN to keep up with these upgrades as the people responsible for setting up the stations will be leaving in September 1997 and no software changes were anticipated after that.
Porcas asked about FS support for frequency agility and receiver changes. Himwich felt controlling station-specific receiver change hardware was best left to station-specific FS software. Mujunen complained that many of the apparent ``FS enhancements'' would actually require no modifications to the Field System but that they would require massive revisions in the ways in which information flows in the VLBI observing process as a whole. This would have to cover the whole chain of PIs, stations, correlators, and PIs again in the data post-processing step. One step towards this direction is the adoption of VEX which is designed to convey information from the scheduling stage all the way thru correlation and back to the PI.
Himwich noted that there is new support for continuous tape
motion in ``drudg'' but that due to lack of such
schedules these features are mostly untested. Alef
noted that continuous tape motion works with the VLBA, but
since SCHED VEX-to-drudg support is under
development, FS9 users can test it only later.
Porcas commented that one can get continuous tape
motion with PC-SCHED and some hand-editing and
Mujunen used the opportunity to clarify what he meant
with the main ``FS'' problems being actually problems in
getting the whole chain of scheduling
software--FS--correlator to agree on the same concept
such as ``continuous tape motion''. This is a perfect
example because as far as the FS goes, ``continuous tape
motion'' is nothing but making sure the schedule will not
contain certain ``et'' stop tape commands which would
normally be there---no changes to the FS are actually
required.
Porcas reminded us that Huib van Langevelde cannot be much involved in enhancing SCHED, but nevertheless Himwich estimated that VEX support will most probably become available in September 1997. Porcas summarized that having VEX support and thus SCHED support together with continuous tape motion brings Mark IV and EVN VLBA stations much closer to true VLBA compatibility. Gurvits asked how this new capability can best be explained to prospective PIs and:
Action: Porcas promised to make sure that next call for proposals will include enough pointers that the PIs can understand the new features available with SCHED+VEX+Mark IV.
Himwich wanted to know what exactly is meant by flagging and it
was explained that the VLBA uses this term to mark in the
logs the time ranges during which no good data is to be
expected. The most frequently asked ``flagging
information'' is ``offsource'', the periods during which
the antenna was not pointing to the source. Himwich
noted that if this feature is to be included in the base
FS it will make it even more important that the ``silent''
status query mode of ``antcn'' be implemented
correctly at every station.
Himwich explained that a few NASA stations would like to run the
FS standalone all the time and log information such as GPS
offset using the FS. Conway suggested using multiple
logs for different purposes, but it was felt that one log
file is simpler to manage and whenever subset information
is required, ``grep''-style filters can be used.
Himwich warned that the next (September 1997) version of the FS
will use new `'now()'' routines to calculate the
apparent coordinates with 10 milliarcsec accuracy,
compared to the old 10 arcsec accurary. These routines
are also more independent of 1950/2000 epoch.
Burgess expressed the wish to have the correct time directly
transferred from TAC into the FS computer, perhaps with a
manually invoked setup program along the lines of the
existing ``fmset''. Himwich described Tom
Clark's (NASA/GSFC) goal to setup TAC as a Stratum 0 NTP
network time protocol daemon reference with Linux
``xntpd''. Ritakari mentioned that
Metsähovi has developed a simple and inexpensive
PLD-based digital time difference counter with 10--20ns
resolution which can be connected to a LPT port of the FS
computer and which can monitor the time difference between
formatter 1pps and GPS 1pps.
Gurvits asked about the status of pcal extractor units. Burgess answered that they do have one ``Onsala extractor'' at Jodrell but that it only has DOS-based software. Conway and Burgess pleaded for manpower to convert the software into a ``FS-friendly'' Linux version. Burgess noted that there are rumors about a new Mark IV decoder being developed at Haystack and Porcas added that the JIVE correlator will be able to extract the phase calibration tones as it correlates. Conway, Alef, and practically everybody agreed on the usefulness of extracted phase cal information after a long discussion. No immediate action could be resolved, however. Burgess could not commit to providing the software driver for the Onsala device.
Conway stated that all units have been built and that they are waiting shipment to stations. Hardware has been verified to be functional with DOS-based software.
Action: Burgess will talk to Ralph Spencer (Jodrell) about a possible candidate for writing Linux/FS-compatible control software and asks Ralph to add this to the EMU/VIV meeting agenda in September.
Himwich noted that if automated IF patching at Mark III/IV
stations is required, then the JPL switchbox may be the
best solution as it already has software written for the
FS and integrated into a station-specific ``patch=''
command. Conway and Orfei volunteered to accept
such switchboxes for testing. Foley asked if
Westerbork would badly need it for geodesy observations.
Ruf confirmed that not absolutely but that manually patching
between geodesy IF setup and astronomy dual-pol setup is
very inconvenient. Burgess suggested a larger patch
panel with high-quality connectors that are really
intended for continuous use. Himwich and Alef
favored the automated solution and Hase confirmed
that it is useful for switching between geodesy and
astronomy IF patching.
Porcas asked about IF patching with VLBA racks with only eight
baseband converters. ``drudg'' currently supposedly
can generate SNAP for these racks for modes that require
less than 8 BBCs. IF patching is actually easier for
these racks if they have space for only 8 converters: in
this case the rack is a genuine VLBA rack and every BBC
has access to every of the four IFs, be it A, B, C, or D.
The so-called ``geodetic'' VLBA racks have limitations and
the easiest way to get over them is to connect LCP with a
power splitter to both A and B, and RCP to both C and D.
Now every converter can access both LCP and RCP, BBC01--07
use A(LCP) and C(RCP) and BBC08--14 use B(LCP) and D(RCP).
Conway gave a talk about EVN amplitude calibration. The main
objective is to get amplitude calibration data to PIs in a
format which is directly useful to them---at the moment
this is the AIPS ``ANTAB'' file. Fredrik
Rantakyrö runs his ``log2ant'' script on log
files found on ``astbo1'' and produces a single ANTAB
file per each experiment, ``expt.antab''. The script
is on ``astbo1'' in subdirectory ``[EVN.CAP]''
and the stations can also run the scripts themselves to
check how well the Tsys lines of their logs get
translated.
At the moment the first priority is to improve overall gain scale factors. Information has been collected and testing its quality with long runs is being conducted.
Action: Whenever the amplitude calibration of a given station is off by more than 5%, Conway feeds this back to station(s) involved.
In many cases the Tcal stability of the calibration source is not ideal (for example in Onsala one source was discovered to vary at 10% level). Another source of errors is the variability of the cal source across the band.
The actions for each station were summarized by Conway as follows:
There were some suggestions as how to calibrate amplitude with >8GHz receivers. No real solution was found but 1) accurate absolute Trec, 2) hot load switching, 3) tipping scans, or 4) forgetting about the whole thing may be applicable.
Conway and Foley suggested using a stable phase cal injector in conjunction with the Onsala phase cal device, since it can extract the phase cal amplitude and thus this could be used for continuous Tsys monitoring. Mujunen commented that it is possible to get continuous TPI readings during scans from the FS to support interpolating Tsys changes during scans. The discussion continued to stress that every experiment must include a few known good calibrators and:
Action: Conway to make sure that all JIVE support scientists know about good calibrators and will recommend putting them into schedules.
Alef reported on the EVN polarization performance in February 1997 session. Polarization purity of 2--3% is considered ok but in the EVN percentages such as 5--10--15% are common. These degrade seriously the dynamic range of polarization maps. ``If all EVN receivers achieved 2--3% polarization purity Maria Massi (IRA) could stop her work on correcting instrumental polarization effects'', said Alef.
The meeting broke for lunch and continued 14:05.
Ruf asked for opinions regarding ``TOO'', target-of-opportunity proposals. Foley commented that Westerbork has a comparable system in place which requires predefined setups. Ruf anticipated that EVN-wide support for ``TOO'' would be difficult to arrange. Porcas used the latest session as an example: there could have been schedules created for the gaps in the block schedule and if the source were bright enough it would have been observed. Ruf asked if ``TOO''s are to be conducted outside EVN sessions and Porcas said no. Mujunen thought that it would be possible to schedule ``TOO''s in the normal way and that they would simply be omitted if so instructed in advance. Porcas asked if 2 days would be enough for advance notification and this was thought to be enough as long as the experiments always stay within sessions.
Porcas asked if it would be necessary to provide schedules for the ``gaps'' and Foley commented that this would be much better than to get schedules in email one day before the observation. Venturi commented that for Medicina omitting observations would be okay, and Westerbork, Onsala, and Metsähovi all agreed as these stations do not have any assigned operators and the station personnel must be flexible anyway. Perea noted that last-minute changes do not work well at DSN station but Altunin revealed that DSN has committed to react to VSOP changes presented 3 days in advance. Altunin envisaged that also DSN can accommodate changes provided that 3 days is enough for JPL to re-process the schedules.
Action: Porcas to distribute guidelines to proposers as how to propose a ``TOO'' to the EVN.
Altunin presented the facilities of DSN Radio Observatories and the cooperation possibilities with the EVN. DSN will be soon ``privatized'' to a company (Lockheed/Boeing?) and it would make sense to ensure the status of radio astronomy by establish a few separate DSN-EVN projects.
Interestingly enough, all DSN stations are equipped with sensitive high-frequency receivers, for instance DSS13 (Goldstone) offers 43GHz and all 70 meter dishes will get a new K-band (22GHz) dual polarization receiver by the end of 1997. 80--100GHz will become available in Goldstone in 1998. EVN-DSN projects can be run at all three stations, not only Madrid. (All three are Goldstone, Tidbinbilla, and Madrid, and each of these locations has one 70m and possibly several 34m antennas.) All stations will only run thin tapes and accept VEX format schedules which get centrally processed in JPL before shipping to stations.
Mujunen asked how these new possibilities get communicated to the prospective PIs and Porcas promised to take care of this in the form of the status table. Altunin told that he would like to get an ``all-modes'' verification at the VLBA correlator much like the EVN ``FT1'' test.
It was asked how the schedules shall be sent to DSN and Altunin
answered that VEX format schedules should be sent to the
JPL/DSN schedule processor, and since VEX schedules will
be generated with SCHED and SCHED doesn't directly support
Mark III modes, it will be natural to use Mark IV/VLBA
modes supported by Mark IV formatter. Porcas asked
what happens to the schedules in JPL and Altunin
answered that JPL probably thought they could save money by
running ``drudg'' centrally.
Altunin continued to describe the availability of DSN stations. It is envisaged that DSN could participate three times a year, 2x 18h blocks at (at least) L/S/X/Ku bands. In addition to this each period could contain 3--4h of joint CAL sessions, and Goldstone could participate EVN 43GHz sessions. Altunin was not sure where the thin tapes required for these observations would come from nor he wasn't sure about amplitude cal and pcal tone compatibility.
Action: Altunin to send DSN scheduling and station contact
information to ``jive@jive.nfra.nl'' for inclusion in
JIVE/EVN Web pages
http://www.nfra.nl/jive/evn/network/members.html.
Much of this information can be found on DSN Web page
http://dsnra.jpl.nasa.gov.
Porcas suggested putting at least Madrid in EVN status table. Alef commented that these antennas are very attractive to any VLBI user and that 3x 2x 18h is the only bad point... Mujunen requested that the different-sized antennas at each location be put as separate entries in the EVN status table and Porcas replied that they already are listed in this way. Porcas also asked if the 3x 2x 18h applies to using the 70m antenna and Perea acknowledged yes. He also noted that if requested, we could perhaps get more, especially 34m time. Porcas also asked if it is okay to require Madrid and/or Goldstone without (or with only) VLBA participation. Altunin commented that they do not have any direct arrangement with the VLBA but EVN/Global experiments would be okay. Porcas also suggested that DSN should send a representative to the Board of Directors' meeting at Onsala in October.
Hase asked if it would be important for geodesy people to get active to attain Madrid involved in Euro-Geo projects. Altunin suggested that a ``special project'' would be a good approach.
Porcas (the new Program Committee chairman) presented the PC report. After Alcala users' meeting and advertising conducted there on EVN special abilities (5cm methanol, joint EVN-Merlin, Mark IV modes) the number of proposals is again up, in February 1997 42 proposals, a threefold increase. In June 1997 EVN got only 14 proposals and this proves EVN is suffering from ``competition'' from the VLBA.
Gurvits asked if there were any 50cm proposals and Porcas commented that a few were there but that EVN frequency sequences are now somewhat ``VSOP-dictated''. He also commented that he was afraid to endorse a technically ``difficult'' spectral line project as the like had resulted in bad schedules in the past. Ruf noted that experiments like this are something to worry about in the future.
Gurvits presented the status of the JIVE correlator project. Active development concentrates on the Station Unit (SU) which interfaces the playback drives to the correlator proper. The correlator boards are functional and ``first fringes'' test is about to be executed with a Jodrell--Cambridge baseline and a simplified delay model.
The Station Unit and the playback drive (both of which come from P&G) are on the critical path of the project plan. At the moment there are two playback drives at JIVE but these units have not yet passed the Factory Acceptance Tests (FAT). The SU has software and firmware problems due to complexity. Porcas asked how good is the correlator online software and Gurvits had to answer that nobody really knows. Porcas added that online software was found to be on the critical path when the VLBA correlator was built.
Mioduzsewski was mostly happy with EVN stations regarding to how they interact with the VLBA correlator: tapes arrive mostly soon enough, Track is being used, logs and GPS files are okay. There are occasional truncated log files and often recording quality is significantly poorer than with ``native'' VLBA stations.
When testing correlation of Mark III data versus VLBA data the tests showed that cross-spectrum was identical but that autocorrelation showed bumps which are due to Mark III DR headers. A fix was developed which blanks out headers at the correlator but it is not yet determined when (and if) this fix will be taken into use.
The correlator has found fringes between the baseline Effelsberg--HALCA.
Graham requested station by station recording quality information. Alef and Mujunen asked if there was a way stations could look up the files via FTP or Web.
Action: Mioduzsewski to look up this information, i.e. how
to arrange access to ``weight plots'' and send the
results of this to ``evntech@nfra.nl''.
Porcas asked what modes were uncorrelable at the moment and Mioduzsewski replied that Mark III dual-pol was. Himwich asked about typical problems with tape drives and Alef commented that outdated head calibration, wear of equipment, and introduction of 14 passes (vs the conventional 12) were the foremost problem causes. He noted that first VLBA recordings were worse in these regards but that VLBA recording quality has steadily improved and that 14 passes at the VLBA is now better than 12 at the EVN.
Conway asked what would be an appropriate way to close the recording quality feedback loop. Alef reminded that there is a plenty of information on checks and tests in the FS documentation. Himwich suggested tape drive experts visiting EVN stations and Ruf commented that all EVN stations are bound to have tape drive experts for operation at all. Foley and Conway stated that expertise is needed to teach operators, especially for handing thin tape. Alef estimated that it may be cheaper to have expert visitors at stations than it will be to fail with the introduction of thin tape but Ruf hesitated to ``centralize'' the responsibility. Himwich recommended a ``tape drive workshop'' to be arranged where tape drive expertise can be spread. Alef complained that so few experts are actually available in the EVN to arrange workshops like this and Foley agreed describing the situation bad and even worse with the introduction of thin tape. Porcas concluded by suggesting that stations which do not qualify in the VLBA thin tape test should arrange a visit to their station to check the recorder.
Ruf told the meeting that the prototype P&G playback drive at
Bonn sometimes had reliability problems and Graham
commented that in particular the capstan motor needed
regular maintenance. Ruf continued by explaining
that the Bonn correlator prepares to switch over to
using ``M3TAR'' (/``MK4IN''?) as its output
media format, that is, Unix ``tar'' archives.
Gurvits asked when this would happen and Alef
said that part of February data may be in the new format
and it can be recognized by a new naming convention.
The correlator received the largest EVN session ever in 1/97 and when 2/97 tapes started to come in the correlation of the first 1/97 experiments had begun. Thus, the correlator struggles with about a year of backlog.
The meeting had a coffee break and continued at 16:15.
Each of the station representatives conducted a short verbal report.
Hase reported that no significant changes have occurred apart from a K4 recording system being installed.
Vicente reported about a new antenna control computer and a new 22GHz receiver coming up in 1998.
Orfei reported that the track and wheels substitution and phase 2 of the frequency agility project have been successfully completed. This allows switching between primary focus receivers in seconds and between primary and secondary focus in minutes by remote control. A second version of this design is in progress for Noto antenna.
The final step of this project is to permanently keep mounted all secondary focus receivers and point to each of them with the subreflector. This step has already started at Medicina.
The IF band of the 8GHz receiver has been widened at 800MHz to join the request of geodetic projects and an IF3 module was bought for this purpose.
VLBA terminal is now complete with 14 base band converters and the ground unit for the cable length measurement between the acquisition terminal and the receiver.
Formatter has been sent to Interferometrics laboratories in order to be repaired. Indeed no possibility was to use the digital switch board. Such a board presented a problem and was repaired in warranty, but the operation took about two months.
It's started a project that will bring to have in the Noto antenna the automatic subreflector positioning as developed for the Medicina antenna. The system is expected to be operative in the summer 1998. In the same period a new cooled S/X/L/92cm receiver should be operative in primary focus.
Reduced efficiency has been found at 22 GHz. The problem is due to the subreflector mirror that has been placed not in proper way during the use of the primary focus receiver. Such a way not to keep suspended the mirror produced a deformation able to affect the performances at 1.3 cm. The mirror will be replaced in the new subreflector positioning system.
Field System now used is 9.3.9 for VLBA, MKIII, S2 recording terminals.
Receiver status: 22GHz has been repaired and Tsys disuniformity in the band was eliminated. The isolators to optimise the polarisation performances have not yet been inserted. Such components have been placed in the C band receiver, but during the operation the front- end of the right channel was damaged. L band receiver was modified in order to have uniformity in the Tsys in the full band and in both polarisation; Cal signals have been added.
A Canadian S2 recording system has been installed and this system will be used to observe in the VSOP program. The terminal has been tested and fringes have been detected in L band with Tidbinbilla and Hobart stations.
A new GPS receiver system is going to be acquired in order to optimise the performances in terms of rms for the time offset. The TAC receiver indeed cannot be synchronised with an external reference and the internal clock is not very stable. It's then necessary to integrate for long time (1000 s) in order to mediate the variation due to the internal clock. A receiver with external synchronisation possibility will then be chosen.
Thin tape upgrade has been performed by Michael Wunderlich and now Noto recording system is able to run both kinds of tapes.
Porcas and Alef asked if there was a 2cm receiver at Noto and Tuccari replied not yet. The 6cm receiver is also still the old one and a new one will be available by the end of 1998.
Reported by Burgess.
Derek McKay has arrived and commenced work on the Cambridge phase stabilising system. Overall we now have enough people to have a good chance of maintaining a stable operation.
Recent cutbacks have resulted in a substantial reduction in the activities of the maintenance staff being likely as of September this year. This may result in limits on telescope availability in the future.
Our previous difficulties due to the loss of the HP and the addition of the Cambridge station are for the most part resolved. The only major area causing us problems is that of correct polarisation. A project has been initiated with the RF lab to set up feeds at 5 and 22GHz aimed at the MK2 telescope, allowing a check to be performed as easily as possible. We have targeted the MK2 at present but the Lovell telescope will be included as resources permit.
There have been no major problems with either terminal or the software. Most difficulties are to do with recording and bad tapes. The gain curve and calibration (local) systems have been restored but we have not yet implemented FS9 'aquir' logging.
We have ordered a triple-cap headstack form Spin Physics to replace the P&G one which fell to bits. We plan to order a second stack in October. It is suggested in a recent Email for Alan Whitney that vacuum switching will not be required with these heads.
Remains as-is for the present. In the future we intend to move timing and phasecal data to a Linux machine in the VLBI room for automated transfer. The VLBA timing point was wrong by 8uS due to triggering on the 'wrong' pulse edge: we have corrected this and intend no further changes in this area.
Graham reported that Rainer Boesel
(rboesel@mpifr-bonn.mpg.de) is now working
permanently at Effelsberg on the VLBI terminals.
The new azimuth track and new elevation encoders have improved performance in general.
This summer the wire outer panels of the antenna from 80--100m will be replaced by perforated aluminium panels. This will improve gain at wavelengths 4cm to 1cm and reduce ground noise, but increase wind loading.
Work on frequency agility for VLBA operation in the range 13cm--7mm is complete. Efforts are being made to improve performance at 1.3cm and 7mm.
The VLBA terminal is used with thin tape provided by AOC at the moment, Mk4 runs with thick tape.
We now have many VLBI observations outside sessions, these have shown a need for improved operator training.
Conway reported on new 6cm dual-pol system, new 6cm and 18cm receivers, new 5cm system and 25m autocorrelator (for single-dish spectra), new 25m antenna control system completion, new Russian CH1-75 spare maser, and SEST 3mm fringes with the P&G VLBA terminal. (Editor's note: I couldn't find a machine-readable version of this station report.)
Perea apologized that he didn't know a station report would be appreciated in the meeting and he promised to produce one for future meetings.
By Kaj Wiik,
Metsähovi Radio Research Station, Finland,
``vlbi@hut.fi'' (email address for VLBI-group)
Some of the BBCs are not working mainly due to VCO problems.
The thin tape upgrade was installed in May and seems to be working well when used at high vacuum.
3 mm fringes were found to Metsähovi in the April CMVA session. Starting from the June session Metsähovi will participate routinely to the CMVA experiments at 3 mm.
A new HP-58503A GPS receiver has been installed (GPSMH2) and it shows about 13 ns RMS noise in the daily averages compared with the 40 ns of our old Magnavox (GPSMH1). GPSMH2 shows currently about 150 ns lower offsets than GPSMH1 due to different cable lengths.
Kirsi Karlamaa (kirsu@kurp.hut.fi) is joining the VLBI group, gradually shifting from receiver design to FOV. Jan Engelberg (jen@kurp.hut.fi) is our new contact person in the maser and timing issues.
Pazderski presented Torun progress on their antennas (15m and 32m), their receivers (21/18cm, 6cm, 5cm; 22GHz is being designed), their P&G VLBA terminal, phase cal, and interference problems. (Editor's note: I couldn't find a machine-readable version of this station report.)
Foley presented the latest developments of the Westerbork frequency-agile MFFE (Multi-Frequency Front End). It has linear feed except for S/X and circular polarizations are generated in IF. Full specification receivers are being manufactured approximately one per month. Replacing the MFFEs will mean that there will be no tied-array in 1997. The current 8x 10MHz tied-array bandwidth will be available in March 1998 and the new array processor with 160MHz bandwidth is planned 1999. The new ``DZB'' 8x 20MHz correlator will get its first crate in August 1997.
The antenna dishes will be constantly renovated and this means that 2--3 antennas will be out of operation for the foreseeable future. Phase cal injection from apex of one dish is being considered, as well as single disk wideband.
Porcas asked if sessions 3 and 4 of 1997 will be single-dish at Westerbork and Foley confirmed this. S/X will be also done. Session 1/1998 will also most probably be single-dish but 2/1998 may get the tied-array.
Action: Foley to send these Westerbork updates to Porcas for inclusion to the EVN status table.
June 1997 by Liang Shiguang.
A new PC586 replaced old 8-bit computer to control the antenna in May. Pointing has been improved to be better than 20 arcseconds (r.m.s). This improvement of pointing will be continued to meet higher frequency observations.
The problem of vacuum loss, that occasionally happened in past years, has been solved by dropping down the value of vacuum around 25 percent, from 10 to 7.5. This action is of no effect on parity error rate of tape reverse pass, but it reduces the parity error rate of forward pass by 1 to 3 order. It is also expected to benefit the longer life of headstacks. We found out that after recording the headstacks are much cleaner than before.
Spectrums of LO of BBCs have been cleaned up to be 30-35 dB stronger than phase noise by introducing a amplifier buffer located at terminal room, which gains and filters the reference 5 MHz signal from hydrogen maser.
There is possible to get funds to buy some thin tapes this year.
Three cooled receivers, 1.3 cm, 3.6 cm and 6 cm bands, are in operation this year. These are single polarisation receivers except for 1.3 cm receiver. The Tsys for 1.3, 3.6 cm and 6 cm are about 100, 50 and 40 K repeatedly.
No report was received from Urumqi.
Gurvits reported on the VSOP program, now that the satellite itself has been renamed to ``HALCA'' after its successful launch.
There had been a long period without fringes but on the 16th of May the Mitaka correlator found fringes on the baseline Usuda--Halca with the source 1055+018. This recording was done with the K4 system, a descendant of video recording technology. Later fringes to DSN (Tidbinbilla on 11.06.1997 with the S2 correlator), Greenbank, and VLBA (SC, 06.06.1997 with the VLBA correlator) have been also found. 15.06.1997 resulted in the first image of 1519-273 and on 03.07.1997 1156+95 will be released.
What happens next? There has been only one experiment which has fringed with an EVN antenna (Effelsberg) and VT801 has been cancelled. The next task is to prove that EVN L band is fully operable with Halca, and the assumption is that C band will most probably also be useful, but that the K band is unlikely to work at all. Due to an unknown reason (but a loose waveguide is suspected) there is 10dB of extra attenuation between the feed and the LNA, and a decrease this substantial is bound to affect K band observation programs.
Porcas asked if there were any fringes from tapes translated from one recording system to another. Gurvits explained that not yet, but that the VSOP--VLBA copying machine is about to be ready. Conway asked if there were any thermal problems with Halca but Gurvits couldn't say. Foley asked if there will be any 1.3cm VSOP in September and Porcas answered that the session will have a couple of multi-epoch EVN-only spectral line projects. Wiik asked if Halca had detected any astronomy objects at 22GHz and Gurvits told that Earth was detected but Moon was undetectable. Conway queried if the pointing was good enough for 6cm and Gurvits explained that only center of galaxy could be observed on L band this far.
Ruf drew some conclusions of this meeting as follows:
Foley suggested that the polarization check problem should be resolved at 6cm first. Porcas suggested making polarization checks an issue at the Directors' meeting but Ruf noted that it will held too late (in October) to benefit the September session. Ruf asked if it would make sense to create a TWG subgroup devoted to polarization but Porcas thought that each station has the responsibility to check it.
Ruf asked if the TWG should be held more often and the consensus
was that perhaps January 1998 would be appropriate for the
next meeting. January 16th--17th 1998 (Friday--Saturday)
was suggested and Vicente from Yebes kindly offered
to host the meeting. Ruf left it to the
consideration of Yebes whether to actually have the
meeting in Alcala. (Editor's note: Gurvits has later
suggested if the meeting could be held on January
19th--20th (Monday--Tuesday). Please indicate to
``kruf@mpifr-bonn.mpg.de'' if this is not
possible for you.)
This document was written using the linuxdoc-sgml DTD. SGML
stands for ``Structured Generalized Markup Language'' and I chose SGML
for this document because SGML is made specifically for translation to
other formats. SGML allows you to specify the structure of a
document---that is, what kinds of things make up the document. You
specify the structure of a document with a DTD (Document Type
Definition). linuxdoc-sgml is one DTD that specifies the
structure for Linux HOWTOs and other docs. QWERTZ is another DTD; the
SGML standard provides DTD's for books, articles, and other generic
document types.
For further information about software used to process this document,
the linuxdoc-sgml package by Matt Welsh, please refer to the text
file
ftp://kurp-ftp.hut.fi/pub/emu/linuxdoc-guide.txt.