Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.atnf.csiro.au/people/Tasso.Tzioumis/sched/Version_11_2.html
Дата изменения: Unknown
Дата индексирования: Sun Apr 10 08:59:57 2016
Кодировка:
Поисковые слова: http www.astronomy.com

Version 11.2

• Released Version 11.2 Update 1 on Jan. 23, 2014 RCW

• Typo fixed (spurious comma) that was tolerated by g77 but not gfortran. Jan. 23, 2014

• Release Version 11.2 on about Jan. 22, 2014. Note that the development and beta versions leading to this release were under the version number 11.1. RCW.

• Deactivate the wrap specifications in the Vex file for all but the VLBA. There is concern that some stations may have a partial implementation of this feature and could get in trouble, so we are going to do more testing before actually using it. Jan. 21, 2014. RCW.

• Give warning and abort if a key file line is more than the maximum of 255 characters after a complaint from a user. Jan 21, 2014 RCW

• Convert the MARK5A setup files for use with the RDBE_DDC and other digital backends. These are the standard setups whose names start with "v" <band>. New example tstsets.key has been added to test some of the files. The setups are sufficiently generic that they would still work for Mark5A scheduling too. Many of the old standard setup files that were too narrow band or otherwise thought to be uninteresting were deleted. Still to be done is to provide standard setups for the RDBE_DDC in the wideband modes. Jan 11, 2014. RCW.

• Provide more explicit warnings about S band channels out of 2200-2400 MHz or in the satellite radio range. Jan. 10, 2014. RCW.

• Expand the warning range for one LO being in or near the RF band for another from 400-1200 MHz IF to 300-1200 MHz after seeing how soft the IF converter filters are on the low side. Jan. 10, 2014 RCW.

• For the VLBA crd files, don't use a setup scan (ie make the main scan span the whole time since the last scan) when the main scan, but not the setup scan, would require a wrap. This is to force the on-line system to do the wrap when it should, which would it would not have in the past. Note that this does not affect recordings since the legacy control system (crd files) no longer controls any recorders. For those who really care about the reasons for this, see the header comments in wrap.f. Jan 9, 2013. RCW.

• Abort if the baseband frequency for the RDBE_DDC is closer than half the bandwidth to a crossover frequency or band edge. Jan. 7, 2014 RCW.

• Add wrap sector to the Vex file. The allowed sectors are in the ``pointing_sector'' lines of the $STATION section for each antenna. The name of the sector is given on the station line of the $SCHED section. Jan. 7, 2014 RCW.

• Test release made to developers and made available in-house. Jan. 2, 2014. RCW.

• Change the limits in warnings from the Vex code from 35 to 40 stations and from 300 to 1000 sources. These match number in an email exchange of June 26, 2013 that was not implemented at the time. Jan. 2, 2014. RCW.

• Set the default for DOPINCR to 15.625 kHz for the RDBE_DDC and to 10.0 kHz for non-RDBE systems. If both are in use, set DOPINCR to 250 kHz. If the RDBE_PFB is used with DOPPLER, abort if DOPINCR is not given. DOPINCR(2) is required and depends on the FIRSTLO. I think using DOPPLER with the PFB is a bad idea (32 MHz tuning steps, which equals the bandwidth) so I did not want put a lot of effort into trying to figure out a reasonable default DOPINCR(2). Dec. 28, 2013. RCW.

• Cause 'INTENT='NONE' to clear the list of intents. Note that VLA phasing intents that SCHED adds when VLAMODE=VA etc are used are not affected. Dec. 27, 2013. RCW.

• Make stations_RDBE.dat and freq_RDBE.dat the default files of those types. Eventually those will be renamed to stations.dat and freq.dat. But the work of changing examples and documentation is being saved for later. Dec. 27, 2013. RCW

• Warn if MARK5A is requested on the VLBA since most of those recorders have been removed from the stations. A significant item remaining on the to-do list is to convert the remaining MARK5A examples to the digital systems. Dec. 27, 2013. RCW.

• Prohibit change of DBE for a station for RDBE stations. The mechanisms for personality switches are not yet smooth enough to do in mid experiment. Dec. 26, 2013. RCW.

• The requirement for a setup with the right bandwidth means that multiple pointing setups are needed for the frequencies at which reference pointing on lines might be done. Those are now provided, at least for the RDBE_DDC. Also the automatic pointing scan insertion has been taught to use the 2 MHz files when
  tt CALCODE='L' for the inserted source. Dec. 26, 2013. RCW.

• The usefulness of the ability to set the bandwidth on a per scan basis is basically lost with the new systems. Those systems require that the sample rate be twice the bandwidth. They don't support over sampling. In principle, SCHED could adjust the sample rate when a new bandwidth is requested, but the software changes to do this looks to be extensive because of the number of places where the sample rate is used. Maybe this will be done some day, but not now. Dec. 22, 2013 RCW.

• Changing the name of the RDBE system at Effelsberg from EB_VLBA to EB_RDBE to help avoid confusion concerning their control systems and other issues. Dec. 2013 RCW.

• Make the defaults MINPAUSE=10 and PRESTART=0 for the DBBC and for an RDBE not under the VLBA control system, such as at EB_RDBE. Dec. 18, 2013 RCW.

• Enforce sample rate = twice bandwidth for RDBE_DDC. This may mean trouble for schedules that do reference pointing on masers. Soon I should make SAMPRATE an in-line parameter, or force it based on the input bandwidth. But not for this release. Dec. 15, 2013. RCW

• Needed to adjust several examples because updated catalog information from Europe switched stations to new digital backends. Those have different tuning constraints than the old systems. Nov. 13, 2013 and later RCW.

• Changed TSCAL default in station catalog to 'GAP', indicating that the station needs gaps between scans for Tsys measurements. All VLBA stations are set explicitly to 'CONT'. Nov 12, 2013 RCW.

• No longer warn when the DDC frequency is not a multiple of 250 kHz as we are no longer concerned about that working, unless older systems limited to N*10 kHz are used. The error when the frequency is not a multiple of 15.625 kHz is still in effect.

• Point out to users when they mix RDBE and non-RDBE frequency and station files. That is generally a bad idea and has caused problems for some observations. Nov. 5, 2013 RCW.

• Change the trigger for warning messages about leaving time for Tsys measurements to include the requirement that TSCAL be 'GAP'. Nov. 4, 2013. RCW

• Change the default firmware for the RDBE_DDC to be VDIF. While at it, split the setting of default values of DBE out of CHKFIRM and put in a new routine SETFIRM. Move both those routines to new places in the calling routines for SCHED logical cleanliness Oct. 31, 2013. RCW.

• Making stations_RDBE.dat complete raised a conflict between the RDBE version of the Effelsberg entry from the NRAO maintained portion of the catalog and the Mark5A version maintained at JIVE. For the RDBE, call the station EB_RDBE with code Er, but expect that to merged back to EB_VLBA eventually. Oct 31, 2013. RCW

• Getting rid of the old stations_RDBE.dat for single RDBE stations. Renaming stations_RDBE2.dat to stations_RDBE.dat since all VLBA stations now have 2 RDBEs. Make stations_RDBE.dat complete by appending all the non-VLBA station catalog segments. Oct. 31, 2013 RCW.

• Update the USNO example n2227.key to the new frequency setup. Oct 29, 2013 RCW.

• Make the warnings about self-generated RFI from LO harmonics less intimidating and a bit more informative. Oct 25, 2013. RCW.

• Improve the scan timing for inserted reference pointing. Before, if there was a long gap between scans, the first inserted scan would start at the beginning of the gap and cover the full time. Especially while Mauna Kea is waiting for the source to rise, that could be hours. The issue had to do with when the final optimization routine call was allowed to adjust times and when it was not. Oct. 24, 2013. RCW.

• Fix bug that caused comments to be lost in many circumstances. Oct. 21, 2013 RCW.

• Example eg3mm_rd2 demonstrates the new ability to deal with reference pointing in the crd files while using the RDBE_PFB and without separate Mark5C and Mark5A passes. Example eg3mm_rdbe.key used the separate passes and has been removed. Oct. 2, 2013.

• Change egrdbe.key to piggyback.key. It is the one remaining example of a file that uses the runsch scheme to get Mark5C and Mark5A files or recordings. Oct. 2, 2013 RCW

• Move peak.cmd to catalogs/Master_NRAO and set it up to be the parent of the related files in the catalog directory. Have Update write peak.cmd, peak_RDBE_PFB.cmd, peak_RDBE_DDC.cmd based on the parent peak.cmd but with altered (with sed) setup file names. Oct. 2, 2013. RCW

• The changes to allow frequency settings different between the legacy and RDBE systems included setting the legacy system data format to a reasonable value for the sample rate, independent of user input (which would be for the RDBE). This cured a problem that happened when changing bandwidths in some satellite tests. Note added Oct. 28, 2012 RCW (work done in September).

• Allow SCHED to make VLBA legacy system (crd) files using reference pointing on masers using Doppler calibration even when using the RDBE_PFB which cannot be tuned. As part of this, the new input parameters CRDDOP, CRDNCH, CRDCH1, CRDFREQ, and CRDBW were added to control just the crd files and not the main setups. This an the item below were a significant code upgrade. Sept. 30, 2013. RCW

• Remove the concept of pcal sets and merge them into the frequency sets. There will not be a separate frequency set for each pcal state used. This simplifies some program logic and should make eventual simplification of the Vex code easier. Sept. 30, 2013. RCW.

• Add Verify_one to test one key file in the same way that Verify tests all of them. Later added Newstd_one to add one example's output to the Stdout directory. RCW Sept. 30, 2013.

• Removed manual_vla.key - it was for the pre-Widar system. RCW.

• Fix synthesizer 2 setting for 2.6 GHz LO entry in freq_RDBE.dat. Sep. 4, 2013. RCW.

• Release a new Version 11.1 Beta on June 19, 2013 RCW

• Trap bad IF names coming from freq.dat under some special circumstances when freq.dat entries cover some, but not all, channels. Also trap errors caused when the vlbasetup stuff tries to print a blank string. June 17, 2013. RCW.

• Fix frequency error in ptd7ghz.set. Jun. 11, 2013. RCW.

• Update the text in the in the PREEMPT section to reflect the 12 hour shift of when the USNO observations are being made. June 10, 2013 RCW.

• Don't allow channels with different bandwidths with the DDC. Adjust some of the warnings having to do with frequency setting. Clarify the warning about duplicate channels. Jun. 6, 2013 RCW.

• Fixed bug in generation of the VLBA legacy system crd files created when scheduling antennas with two RDBEs. June 4, 2013. RCW.

• Removed a couple of continuum sources from the recommended list in peak.cmd because of recent low flux densities. May 2013 RCW. The pointing source list needs to be revisited.

• Changing the code (RDBELEVT) and documentation to reflect the shorter set-and-remember time and shorter down-sampling level setting times on the VLBA. May 22, 2013.

• Changed the compiler (FC) in configureLinux64gfortran from a specific one, newer than the default, needed at the AOC to support pgplot to generic gfortran. Made configureLinux64gfortranAOC to retain the AOC specific path to gfortran. May 10, 2013. RCW.

• Released a beta version on April 26, 2013.

• Reduce TLEVSET in the stations catalogs from 15 to 5 seconds for the RDBE reflecting recent changes in the set-and-remember operation. The VLA still has TLEVSET=60. April 25, 2013 RCW.

• Cause the printout that monitors the building of geodetic segments to go only to sched.runlog, not to the terminal. April 10, 2013.

• Support for the DBBC provided by Cormac Reynolds, April 9, 2013.
  • Basic description of the DDC personality of the DBBC. Allows the usual defaulting of BBCs and IFCHANs. freq.dat entries need to use the new IF names (A[1-4] to D[1-4]).

  • Only supports the 10-512 and 512-1024 MHz filters, and assumes constraints imposed by the 1024 Msps firmware version (a 2048 Msps firmware is planned).

  • Added a `DBBCVER' keyword to the station catalogue to discriminate between different DBBC versions (currently `ASTRO', `GEO', `HYBRID').

  • Placeholder support for the PFB personality - should only be used for testing.

• Default Mark5B track assignments correctly for VSI outputs. This now follows the BBC$\rightarrow$track mapping given in the ``Mark 5B User's Manual" (Haystack). Cormac Reynolds, April 9, 2013.

  Slightly modified by RCW to prevent changes to the Mark5B track assignments when using the VLBA.

• Significant revision of the LBA catalogues (new telescopes, new DBBC and Mark5B systems). Updated positions for the ATCA pads. Added IF/BBC defaulting mechanism for LBADAS and LBADR to allow simplified setups. Cormac Reynolds, April 9, 2013.

• Make it an error, not an obscure warning, when someone tries to assign two polarizations to the same IF. Yes we observed a schedule that did that. April 3, 2013. RCW.

• Adjust the correlator resync times for the RDBE/MARK5C systems go reflect that the media always stops, that it starts at the time good data is expected to start, and that correlation does not take any time to synchronize (no recorded data are lost). Besides reflecting actual timeings better, this will prevent inappropriate warnings from being issued. April 1, 2013 RCW.

• Update the SLALIB to version 2.5-4. This version is under the GNU license so there should be no licensing issues, although I did have permission to use the older version. Tests show that the positions of date change by tens of milliarcseconds relative to the old version. This was traced to changes in the nutation routine nutc.f which contains more modern theory than the previous version. For details see the comments in that routine. The new positions are presummably better. March 28, 2013 with the difference reconciled on Apr. 6, 2013. RCW.

• CORREL=FXCORR will now cause an abort. The correlator no longer exists. March 26, 2013. RCW

• Remove a number of parameters related to the old VLA system from the manual descriptions of the setup files. March 25, 2013 RCW.

• Add VLBA entries in freq.dat and freq_RDBE.dat for S/X that use 7.9 GHz for 4cm. These have the LO in the RF so are not considered good. But the geodesy community is using them. March 25, 2013 RCW.

• Implement the ability to set VLBA synthesizer frequencies to 10 kHz rather than the current 500+-100 MHz in preparation for the new synthesizers that are still under development. March 25, 2013, RCW.

• Write the synthesizer 1 and 2 settings, regardless of which is being used, in the comment at the end of the if_def line in the VEX file. This is in support of the new VLBA synthesizers which will be controlled by the Executor based on information in the VEX file. March 21, 2013 RCW

• Write the first LO frequency in the VEX if_def lines to one more significant figure to support the new VLBA synthesizer that is under development. That synthesizer can be set in 10 kHz steps. March 20, 2013 RCW

• Shift frequencies of pt7ghz.set to avoid RFI. Use different frequency at BR from other sites. March 13, 2013 RCW.

• Add a question about are you using the right station catalog when a mismatch between the DBE and the DAR are found. March 8, 2013.

• The test to warn about overlapping scans that might be intended to be correlated together got confused when there were empty scans. Fixed March 8, 2013.

• Soften the warning about more than 2000 scans if only the VLBA is being used. March 1, 2013 RCW.

• Change read time default of GRABTIME so that GRABTIME(2) can be set to zero. As it was, the read default was zero and, if that was found, the default of 10 seconds was set. The read default GRABTIME(1) also changed, but that should have no effect visible to the user. Feb. 20, 2013 RCW.

• Change the defaulting behavior of GRABTIME to allow GRABTIME(2) to be zero. Feb. 2013 RCW.