Minutes of the GBT Commissioning Meeting 20 December 2002 AGENDA 1. Spectral Baseline work -- Dana 2. PTCS project work -- Richard 3. Az Track work -- Bob A. 4. K-band and other commissioning -- Ron 5. Project and observing program scheduling -- John, Carl 6. Software status -- Amy 7. Spectrometer status -- Rich 8. AIPS++/Dish status -- Joe M. 9. Any other business 1. Spectral Baseline Work (Dana) L-band: 12/16/02: A 5 Jy continuum source on-off spectra show a 2.33-MHz baseline ripple with a p-p amplitude of about 0.5% of the continuum source intensity. This periodicity corresponds to the round-trip delay between the subreflector and main reflector. About 5% variation is seen in Ta/Tsys for continuum sources. The frequency dependence of this spectral structure is fixed in sky frequency, indicating that it is a frequency dependence in system noise or gain ahead of the mixer. Similar spectral structure is seen in the cal spectrum, indicating that much of the effect arises after the cal coupler. Roger and Rick will visit CV to talk to Marian and Rich B on Friday morning 12/20. 12/17/02: Moving the subreflector changes the phase of the 2.33-MHz ripple, confirming the suspicion that it comes from the optics. In addition to the 2.33-MHz ripple we are also investigating a stronger and more irregular baseline distortion of the (on - off) / off continuum source spectrum. To this end we measured a composite continuum source spectrum for the two receiver channels from 1300 to 1500 MHz. This baseline distortion is directly proportional to continuum source strength, is fixed in sky frequency, and appears not to change with subreflector motion. We are talking to Marian and Rich B about the possibility that this could be an LNA/OMT/feed impedance match effect. S-band: 12/18/07: As a follow-on to the L-band measurement of the 2.33-MHz ripple in continuum source spectra a similar measurement was made with 200 MHz bandwidth at S-band. The ripple phase changed with subreflector position and was detected at a p-p level of about 0.1% of the continuum source intensity. One expects this ripple to be a smaller fraction of a continuum source at higher frequencies, which is what we see. A composite continuum source spectrum was measured for the two receiver channels from about 1700 to 2500 MHz to provide more information on the Ta/Tsys frequency dependence. IF (some of this was reported last Friday): 12/11/02: Survey of 2.4 MHz ripple in optical fiber modules 1, 2, 3, 5, 6, and 8 at an IF frequency of 1500, 3000, and 6000 MHz using the Spectrometer with an 800 MHz BW. The rs ripple amplitude is roughly 1% in gain, and there is about a 5 dB spread in the amplitude of the 2.4 MHz ripple amongst the five active external modulators. The ripple amplitude is about 8 dB higher at 1.5 GHz than at 6 GHz. The gain ripple is not purely sinusoidal with the second and third harmonic being 6 and 10 dB weaker than the fundamental, respectively. Longer (30 min) integrations at 3000 MHz were made to characterize an intermittent ripple (~40-70 MHz) thought to reside in the IF system. This ripple was found to have an amplitude cycle of about 11 minutes in three of the six IF channels measured. We are in the process of narrowing down the source of this instability. 12/12/02: Spectrometer measurement of the new optical modulator in IF channel 6 using an 800 MHz bandwidth centered on 3 GHz shows that the 2.4 MHz gain ripple is still present, but it is about 20 dB weaker in the new modulator than in the one it replaced. In a 30-minute integration the fundamental spike in the new modulator's ACF is about 50 times rms. The new modulator has about the same gain stability as the old one, as measured with the DCR. Some more thought needs to be given to whether this improvement is sufficient to justify the cost of modifying the other seven modulators. 12/19/02: The new noise source was used to investigate the ~40-70 MHz intermittent ripple. The noise source was connected upstream of the converter rack and analog filter racks. Spectra were taken with the spectrometer. This ripple may be correlated with the observed temperature cycling in the equipment room but the tests were inconclusive. 2. PTCS Work (Richard) PTCS report: An internal review of the Active Surface Improvements project charter was completed this week, and a number of comments made. Once these are incorporated a review for a wider audience will be held early in the new year. This work will incorporate the correction from FEM predictions to actuator demands as noted by Fred Schwab. Work on the high-level system design continues. A dry-run check of observing procedures and data analysis was successfully performed on the 18th, and additional commissioning time is scheduled over Christmas/ New Year. 3. Az Track Work (Bob A.) Bob reported on the state of a number of projects connected with the Az track work. A sample gallon of zinc paint has been received from the vendor to try out as an anti-fretting material. This paint will not cure very fast at cold temperatures, so we will need to find a day of mild weather to try it. Bob and Dennis have been developing a work plan for the az track project over the next 1+ years. This includes a project plan, budget estimate, and work flow plan for a number of tasks including track modeling (static and dynamic), flex plate analysis, track modifications and evaluation, and design work for a new track. Members of the review panel are continuing to contribute ideas and analysis to the project. 4. K-band and other commissioning (Ron) Scheduled Commissioning activities this week consisted of: o Polarization tests (Anish & Toney) -- by the sound of the e-mail from Toney, his tests did not go over very well. o Spectrometer mode checkouts -- didn't happen since Frank had already put in over 40 hours last week and is on vacation part of this week. o "Training" sessions for Toney (Anish - Polarization), Glen (Frank - VLB & Radar), Richard (Ron - K-band). o The 2B10 Ku-band observing session turned into a commissioning session due to a number of problems (gain instability, inability to balance gains between beams, spectrometer not starting, insufficient gain to drive the Spectrometer, ....). The full list of problems was covered in an e-mail early this week. o We were able to verify that my calculated values for the Ku-band feed offsets are within the measurement errors. o Planning out commissioning activities in January in detail and February in less detail. The training session with Richard on Wed. showed further evidence that beam offsets are probably elevation dependent with a change in offsets of about 9 arcmin from horizon to the rigging angle. This is close to the value I estimated from the plate scale measurements described in commissioning memo 11. There may also be some indication from Wed. data that the orientation of the feeds rotate with elevation. The feed offset measurements that the PTCS group want to make should be able to tell whether these conclusions are right. Al Wootten is in Green Bank Friday and Saturday for 'training' and checking out one of his 22 GHz proposals. Next week, the plans are for Frank to continue the K-band commissioning activities for at most two sessions. There are no activities planned until the weekend of Jan 3. [Glen noted that after a couple of nights of training, Greg Black is now able to do his own system setups for his planetary radar observing program.] 5. Project and observing program scheduling (John, Carl) John reported on the progress of the new telescope activities and project scheduling scheme. This appears to be working well: the projects that are supposed to be getting priority are getting it. The various team leaders have been providing input on short and medium range planning targets in their areas. From Carl: Last Week: Completed observing for the VLBA project BP102. Carried out observations for four other projects (including the on-going radar observations of Titan; pulsars and spectral line). December: Pulsar and VLBA observations scheduled for three proposals. Remaining time dedicated to tests and commissioning. January: Scheduled- Astronomy: 35 hours (~120 of high freq and backup projects yet to be scheduled) Maintenance: 112.5 hours (*new* maintenance starts at 9:30 or 10:00) Tests: 299 hours Yet to be scheduled: ~ 273 hours [Phil noted that Carl has been enhancing scheduling statistics and other information about use of the telescope. We want to consolidate this and other information on commissioning plans and status, proposal status, and project plans into a web page that is easy to find and available to the public. Phil will suggest the things that should go on this page.] 6. Software status (Amy) Software Development Division #17 - Friday, December 20, 2002 We are in Week 1 of a 5 Week Development Cycle, preparing for the release of M&C v3.10 on January 15, 2003 M&C Release v3.9 was released on Thursday, December 19, 2002 Start thinking about future M&C releases such as Wednesday, February 12, 2003 (v3.11) - Get your requests in prior to the release planning deadline on January 13, 2003 - Use the Software Project Office online to submit your request (To be unveiled in January) Data Export Improvements (Filler): In the next two weeks, Eric will be enhancing the prototype Filler with the intention of transforming the prototype into the new production Filler. Baseline Investigation Support: There are two active items. The first is enhancing the DCR so that observers/engineers have the ability to record continuum data when observing with the Spectrometer and VLBI backend. The second is the addition of continuous real-time total power displays (strip charts) in the control room. Antenna Control Unit Improvements: Tim Weadon created QNX test environment to facilitate review of the interface between users and the CCU/SCU. Joe and Melinda are currently focusing on understanding the mechanics of this interface. Requirements have changed since the MR was signed. There is a new MR with the revisions that has not been approved by the sponsor. CME0210-001 Multibank Capability Active: Multibank improvements were released this Thursday. A number of issues raised in the code review and the testing performed by Rich Lacasse have been addressed and will appear as a patch to 3.9, which will be released after Mark and Eric have completed testing. Regression Test Suite: Waiting until Jan 10-11 to initiate regression testing on MCv3.11 post-AAS. Software Project Office Online: Project Office web site unveiling was moved to sometime in January in order to incorporate suggested changes, which will make it more responsive to user needs. Configurator Development: Melinda has two MRs written and ready for review by Ron describing the requirements for adding the K and Ku band receivers to the Configurator. Q Band Receiver: Deferred by Ron; Release will occur approx. March 2003. Engineering Measurement System: Project Charter underway. The charter will be completed and presented by Ray Creager in January. Active Surface Improvements: Project Charter has been reviewed and comments are currently being incorporated into a final product. Amy has written an MR for fixing the 100ms control loop, which has been approved by John Ford. Work on the MR will commence shortly. We are also hoping to do a complete code review in early January, time permitting. 7. Spectrometer status (Rich) Two days were spent on the spigot card. On Ray's previous visit, "ghosts" of test pulses were seen in time bins that should have been empty. Upgraded Memory Card Xilinx personalities were designed to eliminate these ghosts. These were first tested on the bench on Monday. Some problems were found. Ray worked on these in CV Tuesday and Wednesday and fixed the problems. Thursday, the upgraded personalities were tested in the system. The upgrades appeared successful for all modes except possibly a 512-lag mode. Quite a bit of time was spent typing in very low level commands to test the cards. This was found to be quite error-prone. Simple software scripts would be a real time-saver. Ray will be quite busy with the ALMA correlator from now on. He feels that the [spigot card] project is at a good point to hand over to Green Bank. Many, but not all modes have been tested in an engineering sense. What remains to be done is to finish the engineering testing of the untested modes and then test the modes of interest with either a fake or real pulsar. The upgrade of the Low Speed Samplers was begun this week. A modification procedure was written by a technician (Nathan) and four samplers were modified and tested. Two of these were tested in the system as a final verification of the procedure. The result of this work is significantly flatter passband for the 50 MHz bandwidth. Some parts for the modification of the High Speed Sampler Phase-Locked Loop were ordered this week. A filter printed circuit board was modified and 10 copies are now on order. All parts are now on order except for a higher output 15-volt supply. Glish procedures for data evaluation are being upgraded. This should result in time savings when evaluating modes. 8. AIPS++/Dish status (Joe M.) New stable functionality - available Tuesday in GB. --------------------------------------------------- ips: support abcissa ranges braatz, jim 80 15/12/02 d.gauss function supports ranges (specified, interactive). The GUI version will be provided as an enhancement after an outstanding defect with the GUI has been resolved. For documentation, see: http://aips2.nrao.edu/daily/docs/user/Dish/node33.html#dish:dish.gauss.function enhance aver task for improved weight garwood, bob 40 15/12/02 A 'weight' averaging option was added which uses the WEIGHT column which is by default set to 1/sigma^2; sigma is calculated based on the 'k' values from the GBT on-line users guide. The full averaging solution (e.g. per channel averages for spectra) in February release. dish assay functions garwood, bob 80 15/12/02 Initial assay of 'core' dish functions is now done as part of routine build assessment. The thoroughness of the tests will continue to expand. Example execution is: aips++ -l assay.g dishalltest(); multi-if implementation design mcmullin, joe 40 15/12/02 Document was submitted to aips2-dish. It can be viewed at: http://aips2.nrao.edu/mail/aips2-dish. multi-beam implementation design mcmullin, joe 40 15/12/02 Document was submitted to aips2-dish. It can be viewed at: http://aips2.nrao.edu/mail/aips2-dish. Once these are finalized they will be checked into the system as notes. v2.0 newsletters mcmullin, joe 20 15/12/02 Checked into the system. It features articles on: New OH Megamasers detected with the Green Bank Telescope Cataclysmic Variable Star Observations Coordinate Conversions in AIPS++ - The measures Tool Combining Single Dish and Synthesis Data in AIPS++ Adding Inheritance to glish Scripts sd benchmarks mcmullin, joe 80 15/12/02 An initial suite of benchmark scripts will be checked in. This will be expanded upon. The initial values are: Non targeted work: d.gbtsum function (modified version of Frank Ghigo's program for summarizing data sets in /home/gbtdata prior to filling). Filler: Bob added an algorithm in the filler for processing lag data with DC offsets (based on comments from Fred and Glen). Initial look indicates no significant difference in structure or noise level across the band using this option. The optional argument (usebias) is available only at the unix prompt currently: e.g., gbtmsfiller project=/home/gbtdata/my_project usebias=True User Support ------------ Braatz worked in support of the Turner et al. observing session. Braatz also visited CV for additional training/support work. 9. Any other business Thanks to all those who contributed individual reports that appear in the minutes above. Owing to the Holidays, there will be no Commissioning meeting next Friday, 27 December. The next meeting will be 3 January 2003. Happy Holidays to All! PRJ 23 December 2003