Participants: Baldwin, Cecil, Diaz, Elston, McMahan, Simkin. Absent: Dottori. The telecon lasted 1 hr.
Excavation work goes well, and it is still anticipated that it will be completed by ground breaking -- we can undoubtedly arrange a hiatus in the work and ensure a good sized level playing field at the site even if the work runs over. No cost increase has been incurred and I'm assured that the combination of CTIO, E. Serrano, & P. Gillette (Gemini) are watching out for our interests vis a vis preservation of good bearing strata etc.
J. Baldwin & E. Serrano are continuing preparations for wind velocity & direction measurements to establish boundary layer height and hence SOAR elevation above ground for best seeing.
Our study contractors (L&F, Comsat/RSI, Brashear-Contraves, Raytheon Optical Systems Inc, and M3 are doing a good job of preparing concept designs. We continue to apply pressure to hold the costs down and to explore cost effective alternatives for components and design approaches with them. Good exciting and high quality designs are coming from these contractors, and the next time we meet in person I can give you some good ideas of what SOAR will most likely look like: we are definitely moving toward building the best four meter telescope in the world!
Our meeting with Brazilian industry was quite positive, and we will continue to work with Equatorial to seek ways that Brazil can compete effectively for major parts of the SOAR telescope.
Definition of the SOAR control system has moved forward. G. Cecil arranged and hosted a meeting in Tucson to discuss options and approaches. German Shumacher will be joining the project for the last 3 months of the concept design phase to lead definition of controls & software architecture. CTIO is graciously paying his salary for this participation and we are pleased with his qualifications, cooperative spirit, and the concept that CTIO will be intimately involved with development of all aspects of controls. We are also discussing with CTIO
hiring the full time software person for SOAR with the understanding that they will transition to Chile to become part of the operations staff--conceptually an excellent plan, but perhaps tricky in execution.
V. Krabbendam started work on 16 February, has already made a trip to L&F to get up to speed on their mount design effort, and begun work on the specification for the optical blanks for SOAR. We now are a project team of 6 full time and one half-time (Jeff Barr) people, and on any given day, counting design contractors and excavation contractors there are more than 25 people working on SOAR!
We completed two papers on SOAR for the upcoming Society of Photo Instrumentation Engineers (SPIE) conference in Kona: A status paper titled, The Soar Telescope Project: A four-Meter Focused on Image Quality and a technical paper on design, 4-M Off-Axis Telescope by Moretto & Kuhn which benchmarks much of the work done during the SWG phase.
Simkin asked how large the structural resonances would be if they were moved upwards in frequency. Sebring felt that 12 Hz was significantly higher than any other telescope of its size, it was certainly above the driving range of wind frequencies (generally <4-5 Hz.) Second, jitter is residuals from the drives not from the telescope structure. Comsat RSI is looking a double helical drives with double drives on each gear. The gear teeth are phased to average the jitter induced as they engage over 4 different drives. This will be as smooth as, yet much more rigid than, a capstan drive. As a final line of defense, drive jitter can be removed with the tip/tilt unit.
Baldwin noted that it was best to have the resonance power in the frequency range that can be corrected by the tip/tilt unit. In real tip/tilt systems, the frequency depends on how faint the guidestar is. He felt that 50 Hz was high (would imply sampling at 0.5 kHz which requires a bright star) and that 20-30 Hz was more typically what people got on average stars with a 4m-class telescope. Sebring agreed, but noted that this limitation was really due to sensor noise. The Project would specify the tip/tilt hardware to ensure that a correction was feasible at a high enough frequency to remove wind effects. Elston returned to Baldwin's point, noting that what one needed was not necessarily the stiffest structure, but one where most of the power was within the bandwidth of the tip/tilt system. This was note the case at WIYN, where the M2 supports shake at a high frequency (20 Hz, ARC as well), making tip/tilt correction problematic. It's really a question of how much power is up there. Sebring felt that these effects could be modeled, but that this level of detail would not come in the contract design phase. We'd need a servo-control model and other analytical work. Simkin noted that even after all the structural work done on HST, its solar panels still flapped as predicted by one (but not both) of their analytic models. Sebring noted that the FEA and control models delivered by the contractors would be our property, and could be analyzed by others. However, the contractor teams are very good and SOAR is not a flight system so you should just build the best you can then fix if necessary. Elston felt that the issue of the resonance amplitudes could be determined accurately from models. He asked if a spec. could be placed on the residual spectrum of the drive accelerations. Sebring agreed that we need to spend a lot of attention to specing the drives, and that we should ask that the vendor provide us with a modal analysis that summarizes the frequency response of all the drives. Elston felt that would be great.
Elston noted that the dominant 80-90 Hz residual in the Blanco tip/tilt system is new. He thought it was some interaction with the telescope pointing updates and that it could be ``fixed in software." Initially the camera could be run at 240 Hz, with correction at 1/10th that. However, the computational overhead has increased because there is now more processing on the CCD frames to allow fainter stars to be used. The current limit is 90 Hz with motion corrections therefore at roughly 10 Hz. The Blanco 4m actually had a very smooth power spectrum until this recent effect cropped up; correcting at 10 Hz gave residual image motions <<0."1.
Cecil noted that a copy of the Kona SPIE paper from the Project is now at the SAC website. The ``niche" document is also there, as Part III of the Science Drivers report. This includes comments on rapid instrument selection, and larger diagrams.
There is also a schedule outlining the decision-flow to first light, for discussion. This is in response to a request by Board chair Wolff. There are 3 phases. 1) through most of '98 the SAC converges on science, then emits and receives RFP's. There is a lot of SAC related activity until August, then Project related activity to manage the proposals. Everything is approved by the Board and everyone understands who is building what by the end of '98. 2) There is a 1 year design effort, with purchase of long-lead parts and software development. 3) 2000 is spent actually building hardware, roughly 15 months. Instruments are delivered to CTIO at the end of March 01 and installed on the telescope by Sept. 01.
Cecil asked SAC members to review this schedule and comment on whether they thought the timescales were reasonable based on available institutional resources. The construction/design timelines are adjusted (SHORTENED) from the Gemini GIRS schedule, so those represent activities of a full-time, 6-member professional development team. Your mileage will vary. Elston noted that the 6 fulltime people in the GIRS team did not count machinists. The timelines need to be converted to FTE's. Simkin felt that 3 people working fulltime was about our limit. Cecil asked that the partners circulate the schedule around their institutions for comment, so that people understand what a daunting challenge we face.
The SAC would give an instrument report to the Board by mid-August: these are the instruments that we're going to do. The rest of the year figures out who is going to do a reasonable subset of our goals and hopefully all of our requirements. There is no way to significantly accelerate this, even if political realities has already dictated the outcome of the RFP, until the telescope design is approved by the partners. This will not happen before the end of June. Only then will there be $ available. He felt that the SAC should take the instrument specs & goals to the meeting of the Brazilian Astronomical Society, and present the SOAR concept to their community as a final step in soliciting our requirements. This meeting will last 4-5 days in the first week of August.
Cecil felt that the instrument schedule was very aggressive. He also noted that we were building a mini-Gemini, and that they had found that preserving the superb telescope image quality made for very challenging instruments.
Cecil noted that he was planning a phase that should start soon, to run external contracts to groups unaffiliated with the Project. By April there will be a complete set of instrument requirements, backed by
the science proposals. We can then ask very focused questions to expert builders: of IFU's, fiber systems, etc. We can ask them to design their state-of-the-art solution to our science requirements, to show us what we could realistically accomplish on our budget and timescale. This would ensure that we are completely up-to-date in our thinking. These are more than fishing expeditions, the goal is to get strawman instruments. These studies would be available to all consortium members interested in designing SOAR instruments that would satisfy our science goals. We would not be obligated in any way to work with the external contactors. Cecil felt that 4 months and $5-10K was right for these contracts and that he had notified Board chair Wolff of his desire to proceed. As an example, the AAO is ready to investigate IFU's on small fibers for us, despite their huge committments to the VLT & Subaru. They have much interesting information to convey to us. Durham U. is doing interesting work on image slicers. If SAC members are aware of other innovative approaches, please pass the contact on to Cecil. Kona is a good environment to seek out these people. Simkin noted ad Cecil agreed that the group working at the NOT was also doing good work, but wouldn't be at Kona.
Cecil felt that instrument RFP's was political. It was possible that the Board had already developed ideas for who would build what. He would seek clarification. Simkin felt that there should be an RFP, based on the SAC's science drivers. All agreed that the proposals should be evaluated by external reviewers. Cecil felt that at our meetings in La Serena, the SAC should be able to emit a tentative endorsement of the GIRS clone. There were advantages to doing this by April or May because NOAO would be ready then to order optics, etc. and there would be cost savings if two sets could be purchased. At this point, Cecil felt that the only outstanding uncertainty against adopting a clone of the GIRS was UNC's upcoming instrumentalist hire. It is possible that they would hire someone with expertise in IR instruments, who might propose a better match to the scientific desires of the partners.
Cecil felt that the draft of the Derived Instrument Requirements that he, Baldwin, & Diaz were working on could be circulated to the SAC by early March. It would be refined once the science proposal sifting completes. Simkin has 28 pages from her write-up of the IR imaging science. She felt that the Brazilian synthesis was what was needed from the other partners. You loose too much in a table. Everything is on the SAC website, principaly from the August partner meeting.