Minutes of the EVN Technical & Operations Group (TOG) Meeting Friday 1st July 2005 Onsala Space Observatory, Sweden Introduction ------------ The EVN TOG held a meeting on 1st July 2005 at Onsala Space Observatory, Sweden. Approximately 30 people attended the meeting. The chairman was Alef. Roy Booth welcomed the visitors to Onsala. Lindqvist outlined the local arrangements. 1. Approval of and additions to agenda -------------------------------------- The agenda was approved, with no additions requested from the participants. 2. Review of action items from last meeting ------------------------------------------- The TOG reviewed action items from the previous meeting. 1. Action: Paragi to include cross-hand fringe checks as the default for software correlation - done 2. Action: Paragi and Reynolds to investigate 1Gbps support in the software correlator - new software was obtained from NICT, but it has not yet been tested. 3. Action: Garrington was asked to discuss Mk2/Lovell phase referencing with the EVN PC and form a working process for users - test observations have been analysed and the technique seems to work. Remaining action is to make this readily available to users. 4. Action: Paragi to provide more comprehensive feedback from software correlation but to give immediate 'fringe' feedback to station - efforts are being made to provide this. Lindqvist commented that he found the service very good. 5. Action: All stations should look at data regularly with chchk program, use it to locate significant RFI and report the results to Polatidis, and use it to check pcal throughout the session - Robledo has still to get the software working. Alef commented that continued phase-cal monitoring at all stations is important. 6. Action: All stations should upgrade the decoder and use it to monitor pcal - this has been done at most stations. 7. Action: Urumqi to provide automatic control of pcal as soon as possible - this was implemented before session 3/2004. 8. Action: Noto to check reasons for smaller ftp file sizes - done 9. Action: Himwich to locate total file space error in drudg output and fix - done 10. Action: Alef to suggest to scheduler to provide 6-8 hours in the next K-band NME for calibration tests - done 11. Action: Graham to provide advice on K-band calibration targets - done 12. Action: All stations were encouraged to check amplitude corrections on pipelined experiments available from the JIVE webpage - Alef emphasised that it is important for all stations to check these. 13. Action: Stations were asked to provide detailed Wx and RFI comments in experiment logs - the weather is logged at some stations in standard FS format. Additional comments on the weather, both in logs and feedback, are useful for all experiments. On RFI reporting, it is most important to know at which sky frequencies are the worst interference "black spots". The general consensus seems to be that there is no need to log very short bursts of RFI but anything of longer duration (~10 min or longer) would be useful to log. 14. Action: All stations to continue monitoring Tcal throughout sessions - Reynolds urged all stations to use the available software. 15. Action: Reynolds to look at tpidiff for a range of experiments and recommend a suitable time between diode firings. Update SCHED accordingly - as yet there has been no new release of SCHED; currently SCHED issues a warning message recommending scan lengths be kept to 15 minutes or less for diode firings. 16. Action: Colomer to report on implementation of continuous tpi on VLBA systems - results are to be presented at the next TOG meeting. 17. Action: Noto to resolve problems with automatic flagging of data - done 18. Action: Stations to produce ANTAB and RXG files within 2 weeks of the end of a session - Reynolds reported that timely submission of files was improving. It was decided that all files should in future be placed on vlbeer rather than emailed. Reynolds is to distribute the file naming convention for RXG files. 19. Action: Stations must regularly check for Mk5 software updates and ensure the software matches the appropriate FS release - Alef requested that stations check for new versions of the software ~4 weeks before the start of each session. The software situation seems to have improved. Himwich commented that new Mk5A upgrades will usually work with older versions of the FS. 20. Action: All stations are to provide information to Alef about how much disk space they have purchased to date, what is currently avalable (for February session) and an indication of future plans for buying disk packs - Alef has received information from stations on disk pack purchase to date, but future plans are less clear. Alef requested information from stations regarding plans for future disk pack purchase. 21. Action: Walker to check out the status of Mk5 at the VLBA and report on the use of extended bar codes in TRACK - Alef emailed Walker regarding this action. A new version of TRACK with extended bar codes has been introduced. This works OK for new disk packs. Older disk packs which were labelled as thick tapes need to be corrected by hand in TRACK - there may be a problem with getting the size information for older disks into the system. The VLBA plans to move to disk recording in 2006, but their funding situation is currently not clear. 22. Action: Medicina will send a sample of their supplier's packing foam to Alef - this was done. Alef said that the foam was comparatively soft and maybe not suitable for shipping disk packs. There was some discussion of Mk5 logistics, followed up later. 23. Action: Walker to enquire of NRAO whether Noto could send disk packs to Socorro in the February session since they no longer have a tape recorder - Socorro can't yet accept disks. Himwich informed that Socorro's goal was to have this capability by August. 24. Action: All stations to forward FS error log files to Himwich and include details of what the FS was doing when the crash occurred - Himwich will request log files to tar'ed and sent by email. The cause of FS-Mk5 communication problems needs to be traced as this was evidently a major cause of lost data in the last session. 25. Action: All stations to provide to Himwich detailed descriptions of all bugs encountered in the FS - Himwich commented that it is important to run the scripts for diagnostics. Alef requested that all stations collect debug output and email to Himwich. 26. Action: Stations to provide frequency agility information to Reynolds - this is to be included in future release of SCHED. Action on Reynolds and Walker to incorporate frequency agility information in SCHED. 27. Action: Alef to ask Polatidis to update the EVN status table with new/additional frequency agility information - continued action 28. Action: Stations to investigate whether SCHED slewing model and drive rates in station catalogues are adequate - see next action 29. Action: Jb (Lovell) and Ef should look at wrap limits to see if they can be changed - the problem is occasionally SCHED will calculate that a wrap occurs and skip a scan altogether when in reality the wrap limit was not encountered at all. This stems from the assumed antenna starting positions, which have now been changed in SCHED and the problem should now be much less likely to occur. Garcia Miro reported that Robledo also has a problem with antenna limits in that the antenna is sometimes requested to drive to an elevation that is too high. A comment was made that driving beyond antenna limits should be prevented at the antenna control level, but the limit could also be added in SCHED. 30. Action: All stations to install new firmware for decoder. Update formatter firmware to version 41 - done. Dan Smythe was praised for his work on this. 31. Action: Burgess to establish Mk5/disk spare parts list - ongoing action. 3. Reliability/Performance -------------------------- Reynolds' presentation summary is available (pdf file). Reynolds reported on "near real-time" fringe checks. There was some discussion of strategies to automatically initiate data transfer. If 1Gbps can be verified to work with the software correlator then ftp checks could be done in future for all experiments. Computer resources at JIVE may become a limiting factor in the amount of data which can be checked with the software correlator and the speed of feedback. Alef suggested that the use of ftp fringe checks be ramped up starting from session 1/2006 if automatic transfer is implemented by then. It must be made possible to "switch off" automatic transfer of data if required, however, as automated ftp is not yet possible at some stations. For example, at Shanghai station there is no fibre connection to the telescope yet. There is a connection problem between JIVE and Urumqi station, and currently data is transferred from Urumqi via Bologna to JIVE. Reynolds reported on recent NMEs. These included a test of K-band phase-referencing. The results were summarised in a document by Brunthaler and Paragi. It is planned that in the next K-band session, another NME will be undertaken in order to estimate the zenith delay, and determine whether zenith delay errors are the limiting factor for accurate K-band phase referencing on longer baselines. K-band amplitude calibration is very unstable, which was partly due to bad weather in the last session. Alef suggested that Krichbaum be contacted for advice regarding K-band amplitude calibration as he has substantial experience in calibrating 7mm data. Action Item: Medicina station to look into polarization leakage at 22 GHz. Urumqi had very high polarization leakage at L-band. Subsequently the hybrid was adjusted. Data from this session needs to be checked to determine whether the problem has been fixed. Reynolds and Alef urged stations to look at NME reports sent by JIVE. Action Item: All stations to look at NME reports sent by JIVE. Reynolds reported that there were a number of problems in the February session, many of which were beyond stations' control, e.g. due to snow, or equipment failure. Station feedback shows frequent problems with FS-Mk5 communications. It is important to send all information about such problems to Himwich. Action Item: Alef to send email to Alan Whitney regarding Mk5 communication problems. Noto was down for most of the June session due to a mechanical problem (azimuth engine encoder failure) which may require 5-6 weeks to solve but should be fixed before the next session. Torun still has the problem with BBC#3 which was there in the February session (the problem is in the rack rather than the BBC itself). For testing, they need extender cables for the connectors inside the rack. With Gbps recording, all BBCs are needed. Action Item: Torun station to fix BBC#3. Action Item: Stations with a VLBA terminal were asked to check whether they have an extender cable to loan to Torun. Very few stations need to re-patch cables for higher bitrate observations. The only potential problems are for MkIV terminals to observe a single polarization with more than 128 MHz bandwidth. There was a problem with patching at Jb but this is unlikely to recur as it resulted from a one-off experiment designed to determine the antenna position which required special patching. There was some discussion of the Mk5 communication problems. Szomoru remarked that communication problems are never seen at the correlator during eVLBI observations - the problems seem to only occur when recording. Action Item: Alef to send information regarding monitoring of scan_check during recording. Torun used a Rubidium frequency standard for the June session. The H-maser was not available in time for the last session as it was with a company in Switzerland for repairs - there were some delays in organising this. The H-maser is now back at Torun. Previously the Rubidium maser gave a coherence time at C-band of only ~15-20s which makes the data largely unusable for many experiments (especially for phase-referencing experiments). There was discussion of the role of NMEs for the future. It was decided that ftp fringe tests need to be scheduled at the beginning of each session, preferably at least 2 days in advance of the session in order for JIVE to give feedback to the stations before the start of user experiments. Action Item: Alef to communicate to scheduler that ftp tests should as a rule be scheduled 2 days in advance of each session, and NMEs are to be scheduled at the beginning of each session. For rapid correlation and no waste of disk space, stations should send diskpacks with NMEs as soon as they are full (but not earlier). Action Item: Reynolds to give presentation on assessing station performance from pipeline results - Reynolds demonstrated use of the pipeline plots page in the afternoon session of the meeting. Graham commented that sources known to be unresolved from VLBA Surveys may be most useful for asessing calibration in NMEs. Alef urged all stations to send information on RFI to Polatidis. Graham reported on software developed for monitoring during observations. A new Tcl/Tk script (called 'fs_check.tcl') has been developed which calls monitoring programs and a station-specific checker and displays results in a graphical window, with colour coding to indicate whether or not each item is OK. The display includes disk pack usage and results of the last scan_check. Borkowski asked whether this software would work at stations with a VLBA terminal - Graham replied that only small modifications would be required. Action Item: Himwich, Graham and Gunn are to include Graham's monitoring software with future distributions of the Field System. Graham discussed phase-cal extraction and plotting during observations. The new MkIV decoder firmware allows up to 16 tones to be decoded. There is a script to loop through all VCs. Two methods can be used for phase-cal extraction: (1) using Mk5 track_set and (2) the MkIV formatter is used to select channels. Method (2) seems to be more reliable as some wrong values for phase were decoded using method (1). Alef pointed out that stations should be aware that it is easy to check phase-cal during the observations. Quick pointed out that this is important to do, as it will show if there are any unlocked BBCs. Reynolds stated that it is very important that all stations download the latest versions of all schedules at the end of the JIVE schedule-checking week, to make sure they use the final versions. If stations find any problems with schedules, they should contact Gunn (or JIVE staff), but should not communicate schedule problems directly to the PI as this can cause confusion. Action Item: Gunn to send email before each session when the final versions of all schedules are ready for download. Graham is to handle the special cases where the geodetic SKED program has been used to create schedules as vex files produced by SKED will not DRUDG. 4. Amplitude Calibration ------------------------ Reynolds gave a report on amplitude calibration. Reynolds requested that stations give a high priority to timely submission of calibration files. ANTAB and RXG files should be sent within two weeks of the end of a session, otherwise a bottleneck is created for data processing at JIVE. Reynolds also should improve feedback from JIVE to stations. Some experiments lose many cal-diode firings due to antennas still slewing at the start of a scan. Experiments are sometimes scheduled this way to maximise on-source time without waiting for all telescopes to be on source. There was discussion over whether postob cal diode firings (or both preob and postob firings) should be the default. It was decided that this should be driven by SCHED to give the PI an option to switch. Quick commented that optimally, the diode should be fired mid-scan and these data flagged out. It is important to know how often diode firings are required (and what is desirable to PIs). Reynolds reported on calibration accuracy, showing statistics of the mean absolute amplitude error at stations. The goal is for these to be approximately 10% or smaller. Some stations have scaling problems which may be due to wrong gain curves, in particular Noto and Torun. Alef urged stations to look at the pipeline plots in order to assess their calibration. Lindqvist requested that stations receive timely feedback from NME pipelines. Reynolds replied that there is always some delay in pipelining following correlation but JIVE will make efforts to improve this. It is important that NME pipeline feedback is received well in advance of the next session. Himwich reminded stations of the on-off tool to check calibration, and Alef emphasised the importance of looking at these measurements. Graham requested that the sign of errors be included in feedback as this would give clues to the source of the errors, e.g. low amplitudes might be due to pointing problems. Action Item: Reynolds is to provide statistics on pipeline amplitude calibration including the sign of amplitude errors. Alef noted that feedback from spectral line users who use different calibration methods, such as using autocorrelations, would be useful for comparison. Action Item: Reynolds is to request feedback re. amplitude calibration from spectral line users at JIVE in order to assess results of different calibration methods. K-band amplitude calibration is a problem. K-band amplitudes are unstable within an experiment without any opacity correction. Alef suggested JIVE liase with Krichbaum at Bonn regarding opacity corrections. Alef reported on linearity of BBCs. The behaviour seems good for all stations for which this has been determined so far. The tests are still to be done at Metsähovi. GNPLT problems were discussed. Robledo had a problem but this was solved in the new version of the software which they had yet to install. Gunn and Orlati have course notes from their calibration training course at the Haystack TOW available for anybody who didn't attend the workshop. Alef strongly encouraged stations who did not have somebody attending the course to obtain a copy of these notes. Further discussion on the need for training in calibration software was deferred until later. Graham reported on the implementation of continuous cal with VLBA terminals. The Tsys software was modified and seems to work OK. Further implementation will be done when it is needed (as Yebes station comes on-line). The issue of K-band calibration and opacity will be discussed at a meeting to be arranged between JIVE and Bonn (possibly at the time of the joint meeting in September). 5. Automatic flagging --------------------- Reynolds reported that automatic flagging of off-source data apparently works well. It still needs to be fully verified to be OK for Noto, and is still not implemented for Westerbork. Kuper responded that it would be straight-forward to insert a string into the Westerbork log indicating that antennas are on source. There needs to be a decision as to how many telescopes need to be on source for the tied array to be considered "on source". 6. Mark 5 --------- Alef summarised the status of Mk5. A detailed report by Alan Whitney is available. All EVN stations now have Mk5A systems. All 2005 observations are disk-only except for those being sent to the VLBA correlator. It is planned that the EVN will "ramp up" to 1 Gbps recording as the default for continuum observations. The Directors have agreed to purchase the necessary disk space to accommodate this. In total, around 100 Mk5 systems have been deployed. Mk5B is scheduled to be available in late 2005. The Mk5B system will have built-in phase-cal extraction for both recording and playback, as well as station unit capability. There was some discussion of experience of disk failures. It seems to be the case that a few new disks will fail early on, but those which don't are then more stable. TRACK shows that there is a problem with some EVN disk packs "leaking" into the geodetic area. Lindqvist suggested that stations with mixed astronomy/geodetic operations could prevent this by labelling the packs. The problem is to be discussed later in a tape-shipping teleconference. Mujunen asked where stations which don't already have MkIV systems should get samplers - the response was that digital BBCs will be used. Reynolds commented on shipping costs: if stations think they are spending too much on shipping, they should get in touch with Bob Campbell at JIVE to try to find a cheaper alternative. Schonewille added that JIVE has been using the shipping company UTI, but if stations have a preferred shipper then they should contact him and Bob Campbell. At the moment, correlating 1 Gbps experiments in order to free up disk space is not given much priority. The priorities for correlation are set by the policy of the CBD. Projects involving PhD students and recurring projects are given high priority. Setting priorities for correlation may not be straight-forward, but Directors need to also consider the financial costs of disk space. Action Item: Alef to ask CBD to revise correlation priorities in the 1Gbps era. Alef summarised disk-pack repair policy. Broken disks should be shipped to the correlators. Stations should therefore buy spare disks to replace failed disks in disk-packs. Lindqvist pointed out that if a disk fails mid-way through recording, it may be possible to recover some data for correlation which would be desirable for the users, so the disks should be shipped to the correlator. There was some discussion of warranty problems. Buiter mentioned that JIVE had had a problem with a warranty from Conduant being invalid because some testing had been done on the disk. The best strategy is probably for the correlators to get disks repaired locally and for stations to have spares on hand. In summary, it was decided that we will continue to assemble disk-packs ourselves. There was some discussion of the source of foam for protecting disk-packs during shipping. Bought from Conduant, boxes are ~$35 each (not including shipment) which becomes very expensive. The Bonn correlator has found a German supplier of foam which seems to be suitable. JIVE relies on getting foam from stations for reuse. It may be possible for JIVE to buy foam from the German supplier or to investigate local suppliers. Action Item: Alef to send the specifications of foam for shipping disk-packs to EVNtech. There was some discussion of labelling methods for Mk5 disk-packs and problems with printing labels. DRUDG can print experiment start/stop times and VSN on labels from log information. Dymo label printers can be used to automatically print labels. Alef suggested that stations obtain printer specifications for label printing. Himwich suggested that an option to generate postscript files could be incorporated for cases where labels cannot be sent directly to a label printer. There is a problem with labels becoming easily unstuck from the 8-packs. Alef suggested that all stations label disk-packs with (non-removable) luggage tags containing the experiment names. 7. Field System --------------- Himwich reviewed the current status of the Field System. His power-point presentation is available. Current FS software version is 9.7.7. The release of the next version (9.7.8) is planned for late (northern) Summer 2005, in advance of the October session (i.e. by late September). Some new features incorporated are bit-slip detection for adjusting the MkIV formatter, and provision for "continuous" disk schedules. It is still possible for other small items to be included in the release. Release of version 9.7.9, incorporating more features for disk recording, is planned for the early northern Fall. There are two options for automated ftp of disk2file output: (1) a daemon could be run on the Mk5 to detect the new file and initiate data transfer, or (2) a script could be called from the SNAP file to start a process on the Mk5. The SNAP file would need to tell the Mk5 which files to transfer. A solution involving (2) was agreed upon following discussion. Gunn asked at which point the IP address to send data to JIVE would be read. Reynolds responded that this should not be included at the scheduling stage as it would not yet be known then where to transfer the data. Himwich replied that the IP address would have to be a parameter to be edited during the experiment setup. Orlati asked whether it was possible to write to another directory (rather than the current working directory) using disk2file. Himwich responded that it is possible: the full path name can be inserted (alternatively, start the Mk5 in the directory where the files should be written). Himwich discussed operating system issues for the FS, including firewalls which are relevant to real-time system design for e-VLBI. Himwich reported on diagnostic software developments in preparation for CONT05, an upcoming large geodetic session. The diagnostic software will be useful for all stations. Alef thanked Himwich for his presentation and commented that automated diagnostic tests would be very useful for the EVN in fixing problems. Action Item: all stations are requested to try to run the FS diagnostic tests for the October 2005 session and investigate the results. There were no other suggestions from the TOG regarding Field System developments. 8. SCHED Developments --------------------- Reynolds reported on recent developments in SCHED. SCHED can now support Mk5 recording up to 1 Gbps, although a couple of minor bug fixes are required before the next session. For continuum modes it is still necessary for the user to specify a few parameters - it is planned to include default values in SCHED in future. There has as yet been no progress with respect to identifying fringe finders for fringe checks in SCHED. The latest version of the Field System is needed to interpret Mk5 VEX files, so it is very important that all stations have this. Alef urged all to provide updated catalog information before the session in order for users to plan and schedule experiments. Noto station has fixed the problem where their preob time was too long. Station limitations on the maximum number of scans scheduled per unit time are incorporated in SCHED. Any stations wishing to avail of this should contact Reynolds. A new release is expected 3-4 weeks after the present meeting. Alef enquired about the method of scheduling the Lovell/MkII phase referencing technique. Reynolds said that some care would need to be taken to avoid interpretion as sub-netting in the schedule, which creates problems for correlation. Gunn commented that it would be preferable for PIs to take care of calibration scans for the Lovell in scheduling, as subsequent editing of schedules at Jodrell could be a lot more work. Action Item: Reynolds is to look into scheduling of Lovell+MkII for phase-referencing. Garrington is continuing to test the technique. It is aimed that the user will have a simple recipe to follow for scheduling, and that the technique will be offered in 2006. Action Item: Paragi is requested to arrange a test of the Jodrell MkII/Lovell phase-referencing technique for a C-band NME in the next session. 9. Sessions ----------- Alef requested feedback from stations about the feasibility of more frequent EVN sessions: 1) Can stations do more than the present 3 sessions per year? If yes: 2) Would it be feasible to allocate 1 fixed day per month for VLBI? This would allow the EVN to offer support for monitoring projects which need more frequent observing than the present 3 sessions per year can offer. This day could be given back to the observatory if no VLBI observation can be scheduled. The stations would be informed a few weeks in advance whether the day would be needed or not. 3) What boundary conditions do stations have for scheduling Target of Opportunity (ToO) observations? Regarding allocation of 1 fixed day per month for VLBI, responses were: - Gunn pointed out that some stations are already involved in monthly e-VLBI tests. (The proposed sessions would replace or be added to the e-VLBI tests.) Observing every month is a big strain on resources at Jodrell, but additional short sessions outside of the present 3 per year are doable. - Mujunen also commented that VLBI observing every month at Metsähovi would be difficult. - Alef suggested that sessions could be limited to a certain number (<12) per year. The actual scheduling would be dependent on proposal pressure. - Lindqvist commented that it would be preferable for additional sessions to be scheduled on week days rather than weekends. Extra EVN sessions would have to be coordinated with geodetic observations. - Orlati said that more frequent sessions would not be a problem for the Italian stations. - De Vicente agreed that more frequent sessions may be feasible for Yebes. - Yusup (Urumqi) and Wei (Shanghai) said that their stations would technically have no problems with more frequent VLBI observing. - For Robledo, VLBI observing is restricted to only 5% of available time. Martinez remarked that the DSN typically schedule 6 to 12 months in advance so it may be difficult to reserve additional time for VLBI sessions. However Robledo could participate (also in ToO) if the antenna time became available, e.g. due to cancellations of scheduled tracking experiments. - More frequent sessions are technically possible at Westerbork; Foley would need to be consulted. Regarding ToO observations: - Mujunen commented that dynamic scheduling within 1 week would be feasible at Metsähovi, but ~1 day in advance would not be enough time to prepare. - Borkowski said that in principle Torun could participate in additional sessions, but the antenna is often used for local single-dish projects which means it may not always be available for dynamically scheduled ToO observations. - Lindqvist remarked that the 25m antenna is available more of the time than the 20m antenna at Onsala. The 20m antenna is used heavily for mm observations in the winter. It would have to be left to the discretion of the director whether ToO observations would have priority over scheduled geodetic observations. - Graham reported that at Effelsberg, local single-dish observations are generally scheduled only 2-3 weeks in advance, and it would be up to the scheduler to incorporate ToO VLBI observations. In general the scheduling is flexible. In summary: Most stations could do more than the present 3 EVN sessions per year. The general agreement on question (2) seemed to be that allocating 1 fixed day per month, to be decided a few weeks in advance whether or not it would be used, would be feasible at most stations, subject to the approval of observatory Directors and others involved. Observations on week days are preferable. On question (3), most stations could be involved in ToO observations but would need up to 1 week to set it up. Availability is limited also by scheduled single-dish observations and geodesy. Alef said that once agreement has been reached with Directors, the scheduler and others involved, the options could then be offered to users in a usual Call for Proposals. There was some discussion about polarization calibration using data from WSRT to calibrate the position angle. The Westerbork local interferometry data would then need to be provided to users as a matter of course, and the procedure should be made clear in the EVN user guide - the goal is to make this available for the next session. Action Item: Reynolds is to work with Rene Vermeulen on the guide for polarization calibration using WSRT local interferometry data. 10. IF Switchbox ---------------- Graham presented first results for the prototype MkIV IF Switch installed at Effelsberg in June 2005. The switchbox allows any VC to be connected to four MkIV IF signals. There is a problem with open cables causing bandpass ripple, so unused VCs need to be connected up to avoid this problem. No further action is currently required on this item. 11. Spare Parts --------------- This item was skipped as Burgess was absent. Action Item: Gunn is to present an inventory and status report on Mk5/disk spare parts at the next TOG meeting. 12. Digital BBCs ---------------- A detailed report on digital BBCs is available. Alef reported that the DBBC project had been officially approved by the EVN Board. It is anticipated that 2006 could see the first systems with the new prototype digital BBCs. There was some discussion of polyphase filters vs. board stacking to keep components close enough together for the high speed of connection needed between individual converter boards. Himwich expressed concern that there has still been no communication regarding what the DBBC interface will look like. This needs to be known in advance in order to implement in the software, for example communication and extraction of cal data. Action Item: Himwich to contact Tuccari regarding the DBBC interface. Action Item: Alef to contact Brian Corey regarding DBBCs. 13. e-VLBI Status ----------------- Szomoru gave a presentation on e-VLBI and the EXPReS propopsal. Currently most efforts for e-VLBI are spent working on improving the robustness of the system. Network "weather" monitoring is also being investigated. Research networks such as GEANT are complex and changing fast. GEANT-2 will become available in the coming (northern) Autumn. Currently, network traffic is generally very low. Most of the packet loss experienced in data transfer tests has been caused by equipment failure rather than network congestion. At a demonstration during the recent GEANT-2 launch event, variable transfer speeds were seen, with apparently correlated variations on what should be independent data streams. This behaviour is not understood. The first e-VLBI science observations (a spectral line observation) were successfully performed in Sept 2004 and recently the first continuum e-VLBI science demonstration was performed (observing a supernova). The main goals of e-VLBI are to increase reliability as real-time feedback from correlation can be provided to the stations, to increase sensitivity and to simplify logistics by removing the need for management of recording media. There is also a possibility to improve Target-of-Opportunity capability. EXPReS was submitted as an I3 proposal to the EC, and was very successful, being ranked number 1 out of 43 proposals and granted 3.9M euro out of a requested 4.2M total. The aim of EXPReS is to make e-VLBI an operational instrument at 1 Gbps. Many partners are involved. Szomoru presented a summary of current telescope status, planned connectivity and plans for funding allocation to expand the e-VLBI network. It was pointed out in discussion that costs for connection to the research networks are quite variable from country to country, even within the EU. The cost of "last mile" connections to telescopes is also highly variable. FABRIC (Future Arrays of Broadband Radio-telescopes on Internet Computing) is a joint research activity of EXPReS which will look at GRID computing for correlation. There was some discussion of software correlator performance in terms of what is currently available or will be available in the near future. The LOFAR Blue Gene computer will be the most powerful software correlator available in the near future (~27 Tflops). Alef summarised e-VLBI developments at Haystack. The report from Alan Whitney is available. Alef raised the question of whether the TOG should play a more active role in e-VLBI. Participating stations are already involved, and it was decided there is no need for further action by the TOG at this stage. It may be more important for the TOG to be involved at a stage when user e-VLBI experiments are running. 14. Station and Correlator Reports ---------------------------------- Station reports were not presented orally but are available on the web - see http://www.mpifr-bonn.mpg.de/div/vlbicor/tog_chair/togreps05 Alef briefly summarised developments at the Bonn correlator. The detailed report is available on the web. There was some discussion related to Alan Whitney's report on the status of Mark 5, in particular discussion of available equipment to support higher data-rate recording. 15. AOB ------- None 16. Time and Place of Next Meeting ---------------------------------- Next meeting will be held early next year. The location is still to be decided, but it is likely that Noto station could host the meeting at a hotel in Catania. De Vicente said that in principle there would be no problem with hosting the meeting at Yebes except that it is uncertain if the antenna will be operational in time. It was decided that a small training course held together with the TOG meeting would be useful, in which case a working antenna would be required. 17. Summary of Action Items --------------------------- Action Item 01: Paragi and Reynolds to verify that the NICT software correlator works for 1 Gbps recording. Action Item 02: Garrington to make the Mk2/Lovell phase referencing technique available as a working process for users. Action Item 03: All stations should look at data regularly with chchk program, use it to locate significant RFI and report the results to Polatidis, and use it to check pcal throughout the session. Action Item 04: All stations were encouraged to check amplitude corrections on pipelined experiments available from the JIVE webpage. Action Item 05: Stations were asked to provide detailed Wx and RFI comments in experiment logs. Action Item 06: All stations to continue monitoring Tcal throughout sessions. Action Item 07: Reynolds to look at tpidiff for a range of experiments and recommend a suitable time between diode firings. Update SCHED accordingly. Action Item 08: de Vicente was asked to report on implementation of continuous tpi on VLBA systems. Action Item 09: Stations to produce ANTAB and RXG files within 2 weeks of the end of a session. Action Item 10: Stations must regularly check for Mk5 software updates and ensure the software matches the appropriate FS release Action Item 11: All stations are to provide information to Alef on disk space currently purchased and future plans for buying disk packs. Action Item 12: 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 13: All stations to provide to Himwich detailed descriptions of all bugs encountered in the FS. Action Item 14: Stations to provide frequency agility information to Reynolds. Action Item 15: Reynolds and Walker to incorporate frequency agility information in SCHED. Action Item 16: Alef to ask Polatidis to update the EVN status table with new/additional frequency agility information. Action Item 17: Burgess to maintain Mk5/disk spare parts list. Action Item 18: Medicina station to look into polarization leakage at 22 GHz. Action Item 19: All stations to look at NME reports sent by JIVE. Action Item 20: Alef to send email to Alan Whitney regarding Mk5 communications problems. Action Item 21: Torun station to fix problem with BBC#3. Action Item 22: Stations with a VLBA terminal were asked to check whether they have an extender cable to loan to Torun for use in testing their rack (for BBC#3 problem). Action Item 23: Alef to send information regarding monitoring of scan_check during recording. Action Item 24: Alef to communicate to scheduler that ftp tests should as a rule be scheduled 2 days in advance of each session, and NMEs are to be scheduled at the beginning of each session. Action Item 25: Himwich, Graham and Gunn are to include Graham's monitoring software with future distributions of the Field System. Action Item 26: Gunn to send email before each session when the final versions of all schedules are ready for download. Action Item 27: Reynolds is to provide statistics on pipeline amplitude calibration including the sign of amplitude errors. Action Item 28: Reynolds is to request feedback re. amplitude calibration from spectral line users at JIVE in order to assess results of different calibration methods. Action Item 29: Alef to ask CBD to revise correlation priorities in the 1Gbps era. Action Item 30: Alef to send the specifications of foam for shipping disk-packs to EVNtech. Action Item 31: All stations are requested to try to run the FS diagnostic tests for the October 2005 session and investigate the results. Action Item 32: Reynolds is to look into scheduling of Lovell+MkII for phase-referencing. Action Item 33: Paragi is requested to arrange a test of the Jodrell MkII/Lovell phase-referencing technique for a C-band NME in the next session. Action Item 34: Reynolds is to work with Rene Vermeulen on the guide for polarization calibration using WSRT local interferometry data. Action Item 35: Gunn is to present an inventory and status report on Mk5/disk spare parts at the next TOG meeting. Action Item 36: Himwich to contact Tuccari regarding the DBBC interface. Action Item 37: Alef to establish contact with Brian Cory regarding DBBCs.