GBT Operations & Commissioning Meeting 26 September 2003 AGENDA 1. Az Track Status -- Bob A. 2. PTCS Report -- Richard 3. Observing news -- Karen 4. Spectrometer status -- Rich 5. Spectral Baseline, Front-end, and IF work -- Roger 6. Software status -- Nicole 7. Schedule -- Carl 8. Project scheduling -- John 9. Any other business PRELIMINARY REPORTS 1. Az Track Status Azimuth Track * We discussed field measurements to verify the finite element analyis with Dr. Antebi on Friday. We will begin obtaining those when we can get through with the structural inspection repairs. * Dennis continues materials research for new track design or current wear plate replacement. 140' (43m) inspection * We performed a cursory inspection of the declination structure and found no new conditions. The coating on the dish panels is very worn. We will develop a test to determine how much heat a receiver could receive when pointed at or near the sun. * Structural repairs * Walkway repairs are in progress. * We have received a preliminary recommendation from a fatigue analysis and will have discussed it by meeting time. * M&M returned 9/25 to clear up some inspections. We expect the final information database soon. -- RA 2. PTCS Report A major activity identified at the CoDR, and which has been the focus of a concerted effort over the last six months, has been to demonstrate the potential of the laser rangefinders for delivering effective 3mm operation of the GBT. The two main threads of activity have been led by Kim Constantikes, working with Ray Creager and Paul Marganian, to develop the Engineering Measurement System (for real-time analysis of rangefinder data) and by Dave Parker, working with John Shelton and Bill Radcliff, to improve the reliability and performance of the twelve ground rangefinders. At the same time, we have been pursuing alternative strategies for improving the antenna performance (in particular the implementation of the antenna temperature sensor system), as well as characterizing the current, delivered performance. These activities have resulted in a number of outcomes: - Despite extensive efforts, we have failed to achieve full operation of 12 rangefinders, for any trilateration experiment, due to a range of equipment failure and reliability problems. - In the trilateration experiments we have managed to perform, we have achieved poor agreement between rangefinder results and simultaneous traditional surveying measurements. - In order to understand the above, Kim Constantikes has developed an error analysis for the trilateration experiments, which accounts for the geometry of the experiment, and the contributions of both monument position and range errors. This analysis indicates that, combining the geometries currently available with even optimistic estimates of potential range accuracies, the current rangefinder system will not provide the performance required for 3mm or even 7mm operation. We do not believe we can overcome these limitations in time to improve GBT performance during the Winter 2003/4 observing season. - The current performance of the antenna is better than was hoped for during the construction phase, in terms of both pointing performance and surface accuracy, and the alternative techniques for improving these (astronomical observations and temperature sensors for pointing and collimation, traditional holography and out-of-focus holography to reduce surface errors) show great promise. Accordingly, we have decided that continued development of the rangefinder system should be postponed at this time, so that we can devote our efforts over this Winter (2003/04) to improving the antenna performance via a combination of both traditional, and more innovative astronomical techniques. We will reassess the potential of continued development of the rangefinder system, most probably to address the surface accuracy and collimation issues, in April 2004. By that time, we will have made the required short-term improvements to the antenna for Q-band operation for the Winter 2003 observing season, and will have a good understanding of the strengths and limitations of the alternative techniques. --- RMP 03/09/25 3. Observing news Observing report, 19-25 September, 2003 Four different observing projects and 2 project checkouts took place during the last week, with the rest of the telescope time going to maintenance, hardware tests and PTCS. Two of the observing projects were pulsar projects (3B14 - Stairs and 3B-15 - Ransom) and both used the BCPM backend. The 3B15 observations had one BCPM failure (no lost time) and otherwise all projects went well. The Robishaw and Heiles observing (2C46,1A7) had approximately 62 hours in total. Of these, 5.5 hours were lost to high winds during Hurricane Isabel, ~1.5 hours were lost to a spectral processor failure, 0.3 hours was lost to a software hang with the spectral processor, and 0.5 hours was lost to titania failing. Two project checkouts occurred. O'Neil spent 4 hours on 2C8 (Darling, et.al) The project uses the 800MHz feed and the spectrometer in four bank mode. A number of problems were encountered, all at the software end, resulting in a loss of almost two hours. The problems included: - Antenna Control software failure - Turning on individual banks in Spectrometer manager caused failure & hang-up of software [NOTE: cleo now does not allows the option of turning on/off individual banks] - Took 3 1/2 minutes to balance all 16 IFs (far too slow for a 5 minute observation) - Spectrometer started 1st dump 1-6 seconds late - Duty cycles for the spectrometer read fine when the balance button is pushed, but the readings appear to be bogus at the beginning of most scans Mark Clark is working to sort out the spectrometer software problems. In addition, the observations were plagued by the usual RFI. The PI of the project is going through the data now to determine project feasibility. Ghigo spent 4 hours checking out 2C-38 using the 800 MHz receiver. The only problem was at the beginning: the Spectral Processor was in a strange state and complained of having no 5 MHz reference. Normal re-booting was not enough. The operator on duty had to cycle the power on the VME computer in the SP, after which everything ran fine. Further analysis is necessary to determine if the project should be scheduled. Finally, O'Neil & Maddelena spent approximately 12 hours running K-band receiver checkouts. The only two problems encountered were: (1) naid needed a number of disks remounted after the prospero reboot. As neither Chris Clark nor Wolfgang could be found, the observing had to proceed on titania; (2) The 3 ½ minutes which it now takes to balance the spectrometer when in the four bank mode (see above) significantly slowed down the project, as the spectrometer need to be re-balanced whenever the source or center frequency was changed. At first glance the K-band data look good, although a stability problem seem to appear in one channel (R1). Roger and Mike are already aware of the problem and have put a fix (hopefully) in place -Karen O'Neil 4. Spectrometer status The spectrometer worked flawlessly for Karen for about 12 hours! Most of the effort this week went into spigot testing and High-Speed-Sampler Phase-Locked-Loop testing. With respect to spigot testing, there is some good news and some not so good news. The good news is that three twice-Nyquist modes were successfully tested. The bad news is that tests of modes 15 and 16 (2048 frequency channels, 40.96 usec resolution were hampered by some sort of problem with the command protocol. Some trouble-shooting was done over the phone, but to no avail. Ray will come up to Green Bank some time after the ALMA Correlator CDR. (The CDR is on Oct. 2 and 3.) The bottom line is that we have some useful modes available to observers, but they are not the observers first choice. Testing of the High Speed Samplers is still showing problems. The reason for the modifications in the first place was to clean up the 1600 MHz samling clock. This worked well on the prototype we did about 1.5 years ago. However, the two we have modified recently have acted differently and it's not clear why. The spectrum of the 1600 MHz appears clean but the samplers work poorly in the system, introducing illegal samples into the data stream and ugly features into the resulting spectrum. Further modifications are on hold until we can get a handle on this. The pulsar-programming-requirements document was modified this week per comments on the draft and issued as NRAO Drawing A35208D045, Rev. -. A few modes still need some set up parameter information from Escoffier. However, the document is a sufficiently good starting point for the SDD. Plans for next week: - Continued LTA development - Continued trouble-shooting of the High Speed Samplers - Attend the ALMA Correlator CDR. -- Rich 5. Spectral Baseline, Front-end, and IF work The K-band receiver was installed on Tuesday the 23rd. Commissioning has started. The Penn Array cryostat was brought to Green Bank and successfully cooled in the Receiver Room, reaching a temperature of 325mK. This was the first local experience with a pulse-tube cooler (used to reach 3.5K) running on our compressors, but the system seemed to work fine. The cryostat was left in place for two nights of regularly scheduled observations, which will provide an opportunity to characterize the cryostat performance as the antenna tips. UPenn will produce a written report. One of the Q-band LNAs has failed, and was returned to the CDL for repair. Other troubleshooting continues. The MMIC amplifier prototype for LO amplifiers in the mmConverter system and possible use in the Converter Racks was completed and partially tested this week. It appears we need to do some R&D on the coax-microstrip transition which is not performing very well. -- RDN 9/25/03 6. Software status Single Dish Development IPT #52 - Friday, September 26, 2003 This week ends Week 6 of the 6 week development cycle which is the 7th cycle in 2003, so efforts are now focused on ensuring the successful release of M&Cv3.17 and preparing for regression tests this weekend. Pending successful tests, the new version will be made available the evening of September 29. The Plan of Record for the current development cycle is available from the Project Office web site at http://tryllium.gb.nrao.edu/docs/POR/POR_Sept03.pdf. PTCS related work shifted to scoping out quadrant detector upgrades and gaining access to the M&C system via Matlab. The ability to get and set parameters was demonstrated this week, and work is underway to access data from samplers. This is being done using a new SOAP adapter to the M&C system, also in progress, which concurrently serves as the means for configurations to be sent down to the system. Providing feedback to the antenna could also be done with this adapter, and is presently being explored. Development work on the antenna MRs for this cycle continued, and with the exception of subreflector actuator calibration, will be ready for examination during PTCS tests throughout October. Configuration work continues on schedule with enhancements and added testing. Usage of the command line utility was partially documented this week, and Jim, Paul and Amy learned how to use the configuration programs this week. Jim is testing it out for his observing this morning. As more and more people test out the utility, the documentation will continue to be augmented. Frank reports that a wide range of Spectrometer modes were checked, including 2, 4, and 8 sampler modes with the high speed samplers, and 2, 4, 8, and 16-sampler modes with the low speed samplers. This is only the first of three development cycles dedicated to the configuration task; during the next cycle our focus will shift to making changes based on feedback from observer support. The transition to this next stage will be possible in mid October after all 42 cases have been checked by Frank and Melinda. The new gbtlogview is now demoable, but it has been discovered that the additional optimizations and enhancements to FQL now required to support PTCS work are also required to make this a robust application. Additional work has been completed on preparing SDFITS compliant output format for spectral data and an SDFITS related output for continuum data. A guide to Using IDL for reducing GBT data has been started. Operational support this week included two updates to GO LITE which were patched in on Monday 9/22: adding a time-out mechanism to the GO/Python script interaction so that GO doesn't hang while waiting for Python scripts to run, and fixing the bug where GO_LITE was setting the wrong 'obs_mode'. -NMR 7. Schedule Last Week ======== Observations for: GBT03B-014, GBT02C-009, GBT01A-007, GBT02C-010 Completed: GBT01A-007, GBT02C-009, GBT02C-010 October ====== Scheduled hours [backup]: Astronomy ~ 245 [144] Maintenance ~ 123 Tests & Comm ~ 355 [38] Un Assigned ~ 23 November ======= Scheduled hours [backup]: Astronomy ~ 207 [40] Maintenance ~ 102 Tests & Comm ~ 192 [10] Holiday shutodown ~ 36 Un Assigned ~ 182 -- RCB 8. Project scheduling September 22nd Planning Meeting Minutes Rm 241 10:00 A.M. [0] Observer comments None available on the Web Minter's email comments: Two major problems for the setup for AGBT01A_007_09 were encountered on Monday night. 1) Measurements (software) would not run at beginning of setup due to a file permission problem. Joe Brandt fixed this and can provide more details. 2) No power was being sent over the optical fibers to the converter rack due to cabling not being reconnected. Galen W. and Steve W. came in and fixed this problem. They can provide more details. In all setup was delayed for about 2.5 hours. The setup only took about 5-10 minutes once all the problems were cleared out of the system. We decided that we need to have a more structured approach to how we control maintnance activities. A work plan for each item of maintenance was discussed. Bob Anderson is implementing this with the mechanics. We will work on incorporating this into our daily activities over the next few weeks. Software people now use a checklist before releasing the system for operations in the evening. Electronics people will be testing things more thoroughly before turning systems back for production. [1] This week's schedule Swapping PTCS time around This is a go. Richard's crystal ball is working. Tim would like to do some moving telescope tests. We suggested that he fill in where bad weather keeps the painters off the GBT. [2] Next week's schedule May or may not do the weld repair on the axle. M&M and P. Keating have yet to weigh in on the repairs. [3] October Observing Schedule discussions About 96 hours of commissioning time are on the schedule. Ron will be working to schedule that. [4] November Observing Schedule discussions Just getting started [5] GBT development planning HF workshop discussions We discussed the outcome of the HF workshop. The concensus is that it was a very constructive workshop. Phil mentioned 3 things that we should note: 1) There is a keen interest in Ka band. We will attempt to enlist help from CDL engineers to work on characterizing and testing the receiver. 2) Weather statistics. We need to develop a better handle on the weather statistics for HF observing. It was noted that we have no staff effort to allocate to this. It would be an ideal project for someone outside GB to pursue. We also noted that there was no engineering or technician help available in the near term to work on the tipper or site interferometer. 3) There is a need for the wideband spectrometer at Ka and 3mm. Priority discussions HF vs ease of use PTCS finish date 3mm heterodyne receiver 3mm bolometer receiver Advanced spectrometer modes Software reliability improvements Software data processing and ease of use improvements In connection with the HF workshop, we discussed the relative priority of all the components of the HF program. Since this is the end of the fiscal year, end of the quarter, and end of the development cycle, it is a good time to reassess our path. Planning Cycle 8 planning meeting This meeting will take place Monday and Tuesday of next week. Monday we will have a short presentation on each project, and Tuesday we will discuss the resource allocations to the projects. See calendar for times and places. [6] AOB The 140 ft restart was discussed. Phil will recommend to Fred Lo that we propose the project as having a net gain for GBT development, in spite of the short-term staff overload it will cause. No work will be done until NRAO management approves the proposal. -- JF Addendum to minutes: Project management meetings next week have been set: Monday, Sept 29th, 1:30 to 3:30 JL Auditorium Project status summary presentations Tuesday, Sept 30th, 1:30 to 3:30, JL rm 137 Project resource allocation meeting 9. Any other business