GBT Commissioning & Operations Meeting Friday, 7 November 2003 AGENDA 1. Az Track and Telescope Status -- Bob 2. Observing News -- Ron 3. Frontends, IFs, and Spectral Baselines -- Roger 4. Spectrometer Status -- Rich 5. Software Status -- Amy 6. Scheduling Status -- Carl 7. Project Planning Status -- John 8. Any other Business PRELIMINARY REPORTS 1. Az Track and Telescope Status Az Track No developments this week. Structural Inspection/Defect Repair Work continues on walkway repairs. Azimuth Bearings New style seals were installed in 2 bearings. Lubrication converted to heavy oil in them, and a 3rd bearing with original seals reversed. Trial period from now until after first of the year. -- RAA 2. Observing News Another full week of various projects with the telescope. A couple of VLB runs, K-band observing when the weather allows, L-band galaxy survey, Spigot Card observing, program checkouts, .... We had the rather standard set of problems: Antenna faults, troubles passing control of the antenna to M&C, a few Spectrometer and Spectral Processor failures, ... About 18 such failures lasting about 10 minutes each. This is about 2% of the scheduled observing time. The average 'setup' time is under 30 minutes, a significant improvement from a year ago. Items to note: o Karen has spent a good amount of time with observers who are piggy-backing the Spigot card with their BCPM observations. Unfortunately, Karen was doing program checkouts last night and probably won't be able to give us a status report. o We had our first attempt at remote spectral-line observing, as I reported in an e-mail earlier this week. Mark Claussen was the observer and Toney and I were in the control room. Toney and I concur that we saw no difference in the work involved and problems encountered from what we would have had if Mark were in the control room. Mark has promised to give his side of the remote observing story. -- RJM 3. Frontends, IFs, and Spectral Baselines -- Roger Q-band: Things were going well until Thursday morning, when three of the new cryo amplifiers were found to be not working. We are going though our test setups to try to locate any sources of voltage transients or anything else that could be damaging the amps; we are adding redundant protection diodes in the bias circuits; and we are trying to arrive at a plan of action to get the receiver on the telescope ASAP. We may need to consider scaling our ambitions back to only two working beams rather than four. Ku-band: Cold testing is resuming. No disasters to report as yet. More testing to isolate the source of baseline structure on continuum sources indicated that noise introduced in the high-speed sampler modules is contributing to the poor baselines. Discussions on the matter continue. Field mods to the synthesizers used as the system LO2 continue, with five modified units now in service. These will make more output power available, which we expect will improve the gain stability of the converter modules. The mods should be wrapped up after a couple more maintenance days. A modified (and improved!) feed defroster is nearing completion. Galen supplied a status report on the 1cm Receiver: Assembly of the dewar section continues. Remaining tasks are fabricating stainless steel coax cables and wiring the temperature sensors and heaters to their connector. I expect to be able to pump down the dewar for leak checks and cooling tests early next week. The 300K section is about 1/3 complete. The TEC system needs to be tested for RFI in the anechoic chamber. A box to contain and insulate the plate carrying the RF components is being fabricated by the shop. Some brackets need to be made and the RF assembly needs some basic tests performed. I have yet to finalize the design of the turret plate and carrier frame. The shop will experience delays in fabricating some parts due to Dwayne's illness. Two multiplexer bodies should be finished by Nov. 18, which is the cutoff for having them plated by the CDL before their move. Another batch of amplifier bodies is expected in mid December, these had to have slight design changes due to our learning more about MMIC packaging. Tests were performed using the Q band RX on some materials for dewar window support. One material looks good: Trx with no foam: 26.3K Trx with blue, LD33 foam: 26.2K (?) Trx with black, HD60 foam: 30.1K Trx after foam tests: 23.22K (!) The drop in receiver temperature during the test has me planning to do more testing to get a better idea how the LD33 performs, as Q band receiver availability permits. Galen Watts. -- RDN 11/06/2003 4. Spectrometer Status There were a few quirks in spectrometer behavior this week. These were summarized in observing reports. It is difficult to get a handle on these because they are not reproduceable. It will probably be necessary to call in hardware and software support at the time of failure to get these problems resolved. Observers don't want to give up the time to do this, since, generally, a reboot of a computer, which takes just a few minutes, fixes the problem. A fair amount of time was used up this week in support of spigot observations. These were fairly successful. A fairly complete setup sheet now exists, thanks mainly to Karen's efforts. This will hopefully result in less staff time required for expert mode support. Short scripts were written and tested in support of 50 and 25 MHz spigot observations. These allow dual polarization observations without the physical switching of cables. Spigot endurance tests were run. With the old kernel and operating system, the write speed to disk is about twice as fast as the with the new software. To the best of my knowledge there were no frame slips during the real observations this week and test observations of 28, 28 and 56 minutes were successful. High speed sampler modifications are now 60 % complete (got one more done this week). Errors were found in one of the setup tables in the pulsar requirements document. It took several hours to make corrections... Continuing work on the FPGA design. -- Rich and Holly 5. Software Status Single Dish Development IPT #58 - Friday, November 7, 2003 Today ends week 5 of 6 in the current development cycle, which is the 8th cycle in 2003. The Plan of Record for the current development cycle is available from the wiki at Software.PlanOfRecordC82003 along with the most up-to-date status information. We are continuing to add features and increase the effectiveness of the beta Configuration tool. Please continue to send all feedback on the tool to Melinda Mello. Users are encouraged to visit the wiki and look under Knowledge.HowToConfigureTheGbt to find up-to-date information on how to use the beta Configuration Tool to configure an observation. We supported on sky tests with the new controller this week. Initial results are promising. Work continues on SDFITS. Our focus is solely on the Spectrometer at the moment. Other backends will follow when it is deemed that sufficient progress has been made with the Spectrometer. Further information on this project can be found on the wiki at Data.SdfitsBetaStatus. Incremental work was completed on Python tasks for PTCS, SDFITS generation, and prototype applications for the "new GO" which will be unveiled during the next cycle for initial community review. 7 out of 10 SOAP items planned for this development cycle are now complete; we are currently adding callbacks to the SOAP server. The Air Temperature Sensor Monitoring Manager is currently under development. Work on the new gbtlogview has recommenced. There was no significant operational support this week. -- AS 6. Scheduling Status Last Week ======== Observations for: GU003, GBT02C-065, BG141, GBT02C-045, GBT03A-024, GBT02A-063, Completed: GBT03A-024, GT005, BG141, GU003 Novmber ======== Scheduled hours [backup] Astronomy ~ 309 [112] Maintenance ~ 100 Tests & Comm ~ 275 [123] Shutdown ~ 36 December ======== Scheduled hours [backup] Astronomy ~ 265 [69] Maintenance ~ 111 Tests & Comm ~ 212 [35] Shutdown ~ 36 Un assigned ~ 121 Proposal Checkout Queues ==================== Awaiting Checkout - 9 Active Checkout - 27 Schedule - 43 On Hold - 5 Completed - 77 -- RCB 7. Project Planning Status November 3rd Planning Meeting minutes Rm 241 10:00 A.M. [0] Observer comments Mark Claussen Summary: Spectrometer setup failures and go rate-changing problem were the only ones encountered. Config tool worked well except for spectrometer problem. Jim Braatz Ditto, plus: Antenna motion problem. No further details specified Pointing model should be updated to include temperature, since pointing required << 1 hour in rapidly changing temperatures Both observers report that the system is much more robust than last spring. [1] Resource calendar schedule conflicts and discussion Azimuth motion for deflection tests will be done the week of November 17th. Likely at least 2 days will be needed. There is a network outage scheduled (anticipated?) for sometime in the next few weeks. We will ensure that it is done on maintenance day. [2] November Observing Schedule discussions Carl asked if the Q and Ku band receivers will be done. They are supposed to be put up next Tuesday and Wednesday. The baseline project does not have much need for tests right now. [3] December Observing Schedule discussions The December NSF/AUI review meeting group desires a tour. This will be Thursday the 18th, from 1 to 3. [3B] January observing schedule discussions We will have 2 days a week, dynamically scheduled. Carl requested input for tests: baselines, PTCS, and software. [4] GBT development planning Staffing Spreadsheet discussion and conclusions [5] AOB John asked about support for the "expert mode" of the spigot card. Carl thought that we should make every effort to get it working, since it has been advertised as available in expert mode. John agreed, and volunteered to work with David Kaplan and Karen to troubleshoot and fix the acquisition code. We may have to make this a formal software group request for cycle 9 if this does not work out. Dennis mentioned that the shop has had a setback in that one of the machinists is out. Projects affected are Ka band multiplexer, and the Penn Array dewar. He will report back when the plan for getting these back on track is finalized. Observer summaries: Mark Claussen Oct 27, 2003 09:14 PM Proposal codes 02A_063, 03C_010 Receivers used K(18.0-26.5 GHz) Back ends used Spectrometer Observing and Data Processing Summary Overall success of observations: Very successful Observations will result in a journal paper: Yes Description of success and problems The config_tool will be a very useful tool for novice observers. It seemed to work well, except for the spectrometer, which sometimes required a parameter reset even after the config_tool was invoked. Jim Braatz had taken VLA calibrators from the U, K, and Q bands, above 1 Jy (?) and also with no extended structure (?) and entered them into a catalog readable by the Skyview & Scheduler. Having them in the Skyview was very useful in determining sources for pointing and focus. Main problem was that the spectrometer would not \"take\" parameters until second time, and the spectrometer screen from CLEO didn\'t appear to have any problems. Jim also had a better TP calibration script that worked better than that in iards. Minor problem in that one IF didn\'t have quite enough power to make the spectrometer happy, so we ran with a spectrometer WARNING most of the time. (My guess is that the power level was ok for the spectrometer, but the software limits are too strict.) GO_LITE is an improvement over GO, but there were a few things that should be remembered from changing observing modes...see below. Finally, when iards calibrates spectra it should give a Tsys calculation... I used a script from Jim Braatz that appears to do a better job of calibration than the default in iards. But there also seems to be a problem with the Tsys calculation done in the DCR pointing calculation done in iards. What Jim\'s script calculated for the Tsys for spectra was always quite a bit lower than the iards DCR calculation... surely we can get this right .... ? In general, there seems to be a lot of improvement to the robustness and ease of use of the system since I was last here in April 2003. Hardware Get the spectrometer to take in the parameters given to it or to at least give some indication that it hasn\'t. There are still quite a lot of things in the setup that your novice user won\'t get or won\'t want to get. We need to somehow have the config_tool just work, and not have to worry about adjusting attenuators etc to get the power levels adjusted properly. This is ok for somewhat experienced users but not for your Joe Astronomer off the street. Software GO_LITE problems: switching from continuum, beam switch, pointing or focus measurements to spectrometer, nod, total power --- for example, the RA and Dec rates do not zero; the velocity frame isn\'t remembered. Without very careful monitoring of these things,one could lose a lot of telescope time. It would also be useful to be able to click on the Skyview sources and have the names and coordinates get to GO_LITE (maybe this is possible, and I just couldn\'t figure it out). Jim Braatz: IN GENERAL: The system has been generally robust and the data have been excellent. The many improvements even from last spring are obvious. ANTENNA: Last night I noticed the antenna misbehaved badly on at least 2 scans. See scans 105 and 107 of AGBT02C_065_05. These were pointing scans, and it was easy to see in the IARDS display that something was wrong, so I was able to simply repeat the scans. However, if the antenna misbehaves on other types of scans, e.g. while taking spectra, it would be difficult to know there is a problem at all. I have not examined such scans from my spectra yet. I alerted Joe. CONFIG_TOOL: I like this tool! Yesterday it configured the system fine the first try, but today it would not configure the spectrometer. I had to issue g.setup several times, with no luck. I changed the spectrometer config in cleo to an arbitrary setting, then reissued g.setup() and it seemed to work. However, I recognized after a few scans that bank C was unhealthy even while Bank A was fine. A "reset parameters" cured it. POINTING & FOCUS MODEL: If I have not emphasized it enough in the past, let me say again that we really need to update these, and the new models should include temperature dependence. The rapidly changing air temperatures the past couple of days make it necessary to refocus (and repoint) even more than once per hour. The overhead is painful. And GO/GO_LITE are not very smooth with frequent switches between point/focus and spectroscopy. GO_LITE: The issues are known, but are more problematic than usual because of the need for frequent point/focus. To summarize: there is too much reconfiguration required to switch between point/focus and spectroscopy. This problem could be alleviated if each procedure remembered its parameters. I've also run into some problems in dealing with large GO tables, and reported these. FOCUS PROCEDURE: It has become my standard mode to issue a "SlewTo" prior to FocusSubreflector. This eliminates the need to issue 2 focus scans. Perhaps we could instruct the focus procedure do the SlewTo (followed by a delay?) for us, until the antenna motion problem itself is solved. -- JF 8. Any other Business