Penn Array Status Report Minutes, 10jul02 Attendance: Simon Dicker, Mark Devlin, Phil Jewell (chair), Richard Prestage, Eric Sessoms, John Ford, Dennis Egan, Joe McMullin, Roger Norrod, Brian Mason, Nicole Radziwill DETECTORS ============ SD has completed and circulated a more thorough noise/loading spreadsheet as a result of discussion at the last detector meeting in Goddard. These calculations are over the range 0 < Zenith Angle < 65. Progress at GSFC is generally good. OPTICS/FILTERS ============ Cardiff will build the filters, with expected delivery spring or summer 2003. Penn may send someone to help with production. A clarification of the final desired bandpass is desired well in advance. SD has done much work on the optics. A critical (unresolved) point in the design is whether a filter can be put immediately in front of the array; if not feasible it will be desirable to cool the entire cavity (with array, lens, and Lyot stop) to 300 mK. MD notes the possibility of putting a filter (possibly one of several) in front of the entire assembly to keep the detector loading happy over a wide dynamic range on-sky. It was agreed that this type of arrangement is best explored with the instrument on the telescope, although some forethought will be helpful. The option of HDPE -vs- perilene-coated quartz lenses was discussed; HDPE is plan A: quartz is plan B. CRYOGENICS ============ The cryo monitor & control scheme has been simplified from earlier proposed versions. 1) the temperature sensors will all be ROX, no diodes. 2) all of the cryo control + monitoring will be done through a Lakeshore 370 unit. A 2-stage He3/He4 closed-cycle fridge has been adopted in place of the He3 fridge discussed in earlier proposals. This should provide 20 uW cooling, 60 Hr hold times, and 1-1.3 hour cycle times. It should produce most of the 20 uW cooling over a large but currently not-known-precisely range of zenith angles. DE expressed some concern over the mating of the square flange plate to other components; as a result it was decided to round the corners of the flange. The goal is to assemble & test the test dewar in August. It was realized that shop scheduling may delay this; PJ & DE will consider approaches to speed this up. ELECTRONICS ============ For SQUID reasons, the detectors are heavily oversampled (~100 kHz). A sampling of ~1 kHz was generally thought to be desirable and feasible, and will provide a safe margin of oversampling. At some point in the system the mux samples will want to be integrated down to this level, perhaps with a variable integration constant. The individual detectors will nominally be sampled at 16 bits, although the exact scheme is TBD and probably won't involve all 16 bits being read out in phase. SOFTWARE ============ In order to provide a starting point for Monitor & Control software development in Green Bank, SD will provide a spreadsheet with a list of readout points and data rates which the Rx will require. BM, NR, JF, & others @ GB will use this to start specifying the M&C architecture. There was general discussion of analysis software. MD & BM indicated that in addition to the implementation of bolometer data analysis packages along the lines of what exists for other (largely source-finding) telescopes, there is the whole unexplored territory of implmenting CMB-style map-making algorithms which calculate & carry the data correlations through the whole pipeline. This is a big job and it was agreed (partly "online", and partly in discussions afterwards between MD, SD, & BM; also BM & JM) to focus first on implementing a more "standard" package along the lines of what's described in the CoDR document, with a spigot for further development in the future as resources permit. It was generally agreed to make the software system as similar to existing software as the science allows. PJ & RP endorsed the idea of "Use Cases" as a starting point for the development of an overall software architecture for the array. BM will write these, and later develop the overall architecture with feedback generated by them. SCIENCE ============ The bandpass (as noted above) needs to be clearly defined. MD & SD point out that the lower edge of the band is limited by the extended wings of the 22 GHz water line. JM related difficulties he had encountered calibrating sub-mm data with very broad bandpasses. SD & BM will pursue a specification. MD points out that most planets will saturate the bolometers. He is investigating using asteroids as primary GBT calibrators (e.g., bootstrapped off of standard 3mm calibrators via arrays etc). Some clarification of the specific requirements of Galactic plane observing with the array was thought to be desirable. ORGANIZATIONAL ============== We decided to have telecons (or videocons?) monthly. SD & BM will try to get the videocon system working between Penn and Green Bank for the upcoming (August) mtg. We agreed that quarterly face-to-face meetings would be useful, with the next one nominally in Philadelphia sometime in October.