Minutes of the GBT Commissioning Meeting for 21 February 2003 Agenda: 1. Az Track Status -- Bob A. 2. Spectral Baseline work -- Rick 3. K-band and other commissioning -- Ron 4. AIPS++/Dish status -- Joe M. 5. Software status -- Nicole 6. Spectrometer status -- Rich 7. Project scheduling -- John 8. Observing schedule -- Carl 9. Any other business 1. Az Track Status ------------------- This week, we: o Continued bolt test data analysis o Repeated our plate survey so that we can develop a template to drill the replacement plate. Next week, we: o Are expecting to have proposals for the machining and welding trials to review. o Are expecting delivery of replacement wear plates. o Perform pre-modification deflection measurements for the baseplates relative to the foundation, and wear plates to each other. Effelsberg has not done their repair yet. Dennis will be returning to home this weekend. BA Discussion: Jim Braatz asked whether there was any changes with the track cracks and the answer was a welcome "No". Some discussion on March's schedule and the need to schedule at most a week in advance due to the state of the track and weather. More will be known about March by the end of next week. 2. Spectral Baseline Report --------------------------- There's not a whole lot to report this week on spectral baselines. The storms cancelled some test time Tuesday, but that is rescheduled for Friday (today). We made some mods to the Ku-band that we hope to evaluate with the spectrometer tomorrow. Reduction of data from two long (5 and 6 hours) IF noise spectrum stability tests is in progress. The measured ripple periods will need to be matched with cable lengths in the IF system to see where the biggest drifts are occurring. Some cables with good phase stability specs have been ordered for evaluation. Synthesis of all findings so far into an oral and then a written report will begin next week. JRF Discussion: Rick clarified for Barry Turner what Rick meant by "Synthesis". Essentially, after the three month extent of the project there is enough data and enough is already known that getting things onto paper is the next step. 3. K-band and other commissioning ---------------------------------- Like the rest of us, the weather curtailed many of the commissioning activities for the week. Frank made a few attempts at proposal checkouts that were pretty much futile because of the weather. He has determined that proposal 3C61 can be scheduled. I tried to checkout 2C12 but track restrictions kept the most interesting sources just out of reach. Karen had a very successful Wed. night on Spectrometer checkouts Next week we will try to catch up on the proposal checkouts that were cancelled this week as well as more Spectrometer tests. The commissioning schedule for March is almost ironed out. There you will see Q-band as the predominant task. RJM Discussion: Some discussion on the recent report of a 1 arc min pointing offset. The offset was not present in last night's tests implying the culprit was snow or ice somewhere on the structure or over the feed. Jim Braatz asked whether the changes in the track plates will produce a change in pointing. Ron pointed out that an all-sky pointing measurement, performed by PTCS, should follow the wear plate replacement. The last time we did an all-sky measurement was September. John Ford stated that the Q-band receiver will not have accurate thermal cals for March and that the time the receiver is on the telescope is only for engineering tests. He stated that the Q-band system should not be a distraction from out higher priority tasks. Barry asked about the lower frequency limits of the receiver but no one in the room knew the answer to his question. 4. AIPS++/Dish status ---------------------- - New stable to be instated today; it includes support for beam switched calibration - The group (mostly Joe) is busy with technical review - Focus tracking and Az/El pointing scans are now available, and we'll be working on the data reduction code within iards for these types of data JB 5. Software status ------------------- Software Development Division #25 - Friday, February 21, 2003 The state of M&C v3.11 is "green" this week. Testing of v3.11 went well last night. Mark finished testing the IF Manager and its FITS file changes and Toney tested changes to GO. Mark and Toney generated new test data and sent the data to Bob Garwood for verification. Regression testing is scheduled for tonight. If successful, v3.11 will be released Monday, February 24 between 8am and 10am. The key updates for this release are the inclusion of frequency and polarization information into the IF FITS files, updated active surface software which includes a fix of the 100ms problem, . The active surface 100ms fix will not be released on Monday. Instead it will undergo further astronomical testing by R. Prestage as part of PTCS observing and will be patched in to v3.11 next week. Now that testing for v3.11 is complete, developers will begin working on commitments agreed to for the March cycle, which is targeted for March 12, 2003. Key efforts are Operations and Usability, focused on completing improvements for spectrometer reliability, and PTCS, with continued development on the EMS and discussions regarding new instrumentation. For details please review the SDD Plan of Record at http://tryllium.gb.nrao.edu/images/POR_Mar03.jpg. Work has already started on the Caltech Continuum Backend target for March; Three operational scenarios were created by M. Mello and submitted to B. Mason for review. This cycle, 57 new unit tests were written, bringing the total number of unit tests run every day on the M&C system up to 141. Although unit tests do not catch every issue that may be seen when the telescope is in use, they do reduce the SDD's troubleshooting burden while using time on the telescope. Also this week: + the script for configuring one's environment for running the GBT M&C system was modified so glish is always run from aips++ stable. This has a negative impact on any existing glish scripts that depend on the latest Aips++ features. + A Windview (program that allows developers to "see" the timing of tasks on VxWorks machines) licensing problem was identified and partially fixed. The partial fix allows us to use it as is, but the license manager host needs to be switched to sadira when we upgrade to tornado2. + M&C disaster recovery procedures were enhanced. The previous arrangements for disaster recovery were that M. Clark would make a tape once/month of /home/gbt2 and send it to Charlottesville. This tape was sufficient to rebuild the M&C system but other required components were not included. C. Clark is now producing a "Shrink-wrapped box" type of backup. i.e. one that included all components required for the M&C system to function. Two copies of the tape are made each month. One is shipped to Ernest Allen at Charlottesville for storage, the other is stored in the fireproof vault in Green Bank. ALS Discussion: There was a bit of discussion about the lack of a bug reporting facility and that Frank's recent assessment of bugs is therefore incomplete. John and Ron made comments on how the MR process doesn't close the loop with the requestor. Probably the requestor of the MR should be involved in deciding before a release whether or not the work done satisfies the MR. Another example: Even though the K/Ku-band changes to the Configurator are finished, they will not be released this cycle and the signer of the MR was not informed of this. Ron commented on how the 10-day delay in the 3.11 release of M&C was due to just a few items in the release not working. Other items in the release have been working but users have not had the opportunity to take advantage of them. 6. Spectrometer status ----------------------- During the one day available for spectrometer hardware testing, two problems were attacked. The first involved the complete failure of the self-test, where results from all the correlator chips in the spectrometer are slightly off. The test failed several times. We examined the serial communications to the sampler control card which sets up the sampler modes and generates the test pattern. The serial interface performed flawlessly. The self-test then passed with zero errors. The second problem involved a step in the auto-correlation function that is 4096 lags long in 1-sampler, 800-MHz, multi-quadrant mode. Several hours of trouble-shooting led to a faulty sample distributor card. The card was replaced and will be examined on the bench. The first iteration of requirements and a block diagram for an improved test signal source was completed and distributed for comment. A pulsar spigot test run was supported. During this run, we could not get a test observation to produce the correct result. We saw this same problem last week but ignored it. This week, we traced it to a firmware problem in the correlator control cards. Plans for next week: 1 - Attack the multi-integration problem reported last week. 2 - Continue work on the improved test signal source. 3 - Test additional modes in multi-bank mode. 4 - Perform additional tests on the intermittent self-test problem if it crops up. 5 - Follow up on problems reported by observers. RL Discussion: There was some discussion about last week's problem with a bad LTA board for quadrant 2. The replacement board has a similar but less noticeable problem and is in quadrant 1. Since quadrant 1 is always used, it was decided that the semi-bad board will be moved to the less-used quadrant 4. 7. Project scheduling ---------------------- The Project Planning group has been wrestling with the 6 month plan. We have identified a bit of a shift in priority, away from introducing additional capability, to building more robustness and ease of use into the system. This is not a wholesale shift, however, as we still have as priorities the Azimuth track, Baselines, and PTCS. The main impact of this proposed shift is on PTCS, as the effort is mostly in scientists and software, the very resources that are most in demand for the PTCS project. This week saw considerable disruption in planning, as the weather did not cooperate, and stuff had to be moved around wholesale. There has been a drop off of use of the resource calendar, with people just opportunistically getting time. I have asked people again to schedule things so that we know who is trying to do what when. We plan to meet Monday morning and finalize as much as possible the March schedule of commissioning, observing, and maintenance, with much of the observing already tentatively scheduled. JF 8. Observing schedule ---------------------- Last Week ======== Observations for GBT02A-049, GBT02B-021, GBT02A-008, GBT02C-023, GBT02A-031, GBT02A-052. GBT02A-052 completed. February (next week) =============== No scheduled Astronomy except some backup for K band commissioning March ===== Astronomy ~ 325 hours (mostly K band); backup ~ 106 hours Maintenance ~ 146 hours Test+commissioning ~ 278 BB155 (VLBA) was cancelled because Effelsburg repairs not completed Numbers do not include additional maintenance or downtime that may be required for GBT wear plate repair. April === Astronomy ~ 30 hours (VLBI) Just started. RCB Discussion: We ended with comments about the policy for dynamic/backup observing since we've had our first experiences with the new policy over the last few weeks. Dave Hogg pointed out that the current policy won't work well for the typical backup observer who cannot get away from their teaching duties to spend extended time in Green Bank. It worked for the recent observing because the backup observing was shared by local people (Ron and Dave) and a graduate student (Chris Springob). Dave relates a comment from Chris on how there's only a small window between weather that is just bad enough to cancel high frequency observing and weather that cancels all observing. Carl mentioned that John Dickey's graduate student will similarly be a backup observer. RJM February 21, 2003