Minutes from 02/03 July 2003 Penn Array Project Meeting in Green Bank 04jul03 bsm attendance: Mark Devlin, Brian Mason, Simon Dicker, Mark Supanich, Roger Norrod, John Ford, Phil Jewell, Richard Prestage, Melinda Mello, Michelle Caler I. Cryostat & Packaging Simon gave an overview of the cryostat design, which is in good shape. There are three feed throughs to the top (donut) and how we connectorize things is open. We may want to run cabling down the side of the Rx, in which case we should take care to protect it all. A 2-chambered donut was discussed and thought to be a good thing. We don't yet know whether the PTC stepper motor driver belongs in the donut or in the RF rack-- we discussed a separate shielded box for this and this actually seemed the best option; GB can provide some guidance. Overall 120 to 150 dB of shielding are expected to be needed between the driver and just outside the donut, and this will be a challenge. There was some concern about buckling of the cylinder -- either the vacuum can, or the inner tube of 20 mil G10-- Devlin thought this shouldn't be a problem on the basis of indpendent tests, but they will run through the numbers. We discussed heating/cooling needs of the components in the donut-- probably they should have a couple heaters w/regulating circuits. Dew formation on the outer window was discussed; warm dry air to the donut hole, heaters, and an appropriate design for the outer window will help mitigate this. Our blowers have not bee terribly useful in this regard. The outer window should be removable without too much trouble for maintenance and inspection purposes. The material for the window could be goretex, mylar, or polypropylene; CDL likes HR500. The controlling parameters for the dewar design are the outside shell, which is limited by the hole size, and the heights, as they relate to the Pulse Tube Cooler and the optics. These are all pinned down. There was some discussion of getting the optics built before the cryostat was finalized but we moved away from this. Dennis & Roger will use the old drawings, together with "as-is" updated drawings, to order materials for the new (final) dewar. Simon will provide final drawings for the new dewar by 01sep03. This should be "really final" so as not to slow down the construction. II. Electronics & Software Mark Devlin gave an overview of the electronics control packaging scheme we reviewed Mark Supanich's new wiring diagram. The power supply for the Digitial Electronics Box has been condensed to a single auto-sequenced 3U card in the rack itself. The electronics seems reasonably converged. We do want to leave some spare analog readouts which can be filled in later (eg, a temperature monitor and hydrometer for the donut hole). GB can build the RF crate Just-in-time. The connectors from the crate to the donut (and thus cryostat) will probably be (custom?) steel bellows filled with eccosorb. The fibers need to be better specified (how many; the material; the connectors) Melinda & Brian presented example software documentation. Penn's summer intern (Matt) will use these as templates and will produce two of the three analagous documents for the Penn Array (excluding the "algorithms & procedures" doc) by 15 aug-- in any case, these 2 will be at the CDR. Melinda & Brian will give feedback on drafts of these documents. III. Cryogenics The cryogenics will operate down to 32 degrees elevation with reasonable equilibrium cryostat temperatures attained (no significant degradation in performance beyond that implied by increased atmosphere loading). Below this the cryostat will begin to warm up, but at least two hours at an elevation of 20 degrees is sustainable (perhaps significantly more). This is roughly the elevation required for a 4 hour track on the galactic center. At this limit the He3 hold time is the limiting factor. The currently measured hold times are 72 hours at 15 uwatt loading, which is rather extreme-- Penn hopes to achieve hold times of 100 hours or so with some improvements. There was discussion of the operational implications of the cryogenic constraints. If a low frequency observer was observing for 4 hours at 15 degree elevation, for instance, a full recycling of the sorption fridge would be required. This requires 4 hours and works at elevations above 45 degrees; there is some hope of reducing this to something like 2 hours. A full cooldown from room temperature requires 24 hours. We should plan on scheduling cycling times, or using already scheduled maintenance time to cycle the sorption fridge. We should also track the "fuel" remaining with an appropriate (perhaps conservative) calculation. Some form of human interface may always be required (to track the fuel, punch the cycle button, or enforce observer policies like elevation limits). IV. Optics Simon has a nominally final and reviewable optics design documented in the memo which he presented. In this great attention has been paid to image quality considerations, eg, ghosting. Brian & Roger will review this; it will also be sent to an external reviewer. Dominic & Michelle have both checked the loading calculations. Measurements of the cryostat beam pattern down to the 10 dB point are achievable without too much difficulty and would be a useful check in the lab. The filterwheel was not funded. Penn would like to build a shutter inside the cryostat-- just a sliding piece of metal to allow "zero" point measurements in the field-- and this would be good. Other lab measurements will probably involve SKY/Liquid Nitrogen and SKY/SKY with Neutral Density Filter; we may or may not be able to do these on the telescope. Nominally it will take 2-3 months to construct the optics box, from the point that the design is approved. Cardiff is an unpredictable factor timewise (filters and lens coatings). Simon could probably go and help with the coatings, but for the filters it's not so clear that would be helpful. V. Detectors GSFC is working on a mask, which will go into Fabrication by the end of August. Fabrication takes 1 month, and after this (provided a technician) it takes 1 day of wirebonding to attach the detectors to the mount. This would provide Penn with a mechanical prototype for lab tests by mid-October. *After* this point the detectors are put on the array, and G is chosen. The current mechanical design has 0.3 mm supports, 0.05mm gaps (2.9mm ^2 detectors), and thus 3.3mm pitch overall. Penn expects 3mm photon based timestreams from the detectors in the lab by the end of the year (optimistic) or mid-Feb (pessimistic). GSFC claims the 1/f knee (where gain changes = 2 x thermal noise) of TES's is at about 100 seconds. VI. Schedule & Management We should assess the CDR outlook in mid-August, with the intent of a late-Sep or early-Oct CDR. We seem to be on track for this. Richard points out the fiscal year boundary might need to be considered. If the CDR happens in late Sep, and optics is held for this, the optics integration is shifted to March or April 2004 and this doesn't seem to have bad impacts on the schedule. Some specific items we need for the CDR: a) completed cryostat drawings; b) pre-reviewed and final optics design; c) final loading calculation document; d) detector mask complete; e) a presentation or document from Dominic a couple of weeks ahead of CDR date; f) the hardware/software interface document, and documentation of the python scripts; g) an updated and realistic schedule; h) a document describing the cryostat & cryogenics (perhaps simply modifications to the paper being submitted now). The main schedule risks are the detectors and Cardiff's contributions to the optics. For part of the Cardiff stuff we can send helpers out. A significant project risk is the detector loading, which we will cope with via neutral density filters if needed. Taking the receiver down, installing an NDF, and getting back up, is about a 1 week operation we can conduct during commissioning. Penn expects to have several people at GB 3-6 weeks before commissioning for preparations and testing. They will need one half to one "cubbyholes"; we should commit the appropriate software effort during this time, and we will also designated an engineer to work with them during this period. Melinda should probably visit Philadelphia before this time. Brian will produce strawman acceptance tests before the CDR. The idea here is that these tests are the "gateway to commissioning" and basically, they are lists of documented test results we'd like to see. Another area of risk is the telescope. This is beyond the official scope of the project. VII. Receiver Room Visit Good places for the Penn Array Receiver seemed to be in the current X, Ku, or C slots, roughly in that order (N8, N4, N2). X has an obvious point on the Lband support strut where the RF rack could be mounted but the vibrations there are significant. Generally vibrations seemed significant and this could be minimized by doing straight to the roof or heavy centeral pillar. The LO(?) rack is possibly in the way at C band, although the Penn Array will be quite short and perhaps this isn't a problem. The helium lines we'll provide to the pulse tube cooler are thinner than they currently use and the adequacy of this flow needs to be tested. VIII. Analysis Software The Penn group and Brian had offline discussions about analysis software. The most productive approach for now seems to have Michelle and Brian start to work on some of the "Quick Look" software like pointing and beammaps. We aim to have demonstrations of these by the CDR. For now there is no reason for Michelle not to continue working in the environment of her choice, probably IDL, and we can probably even use this for real data analysis & commissioning if we need to. ACTIONS -Mark Supanich will send us info on the stepper motor ASAP so that we can do realistic RFI tests. -GB: RFI tests with stepper motor within 2 weeks of getting stepper motor info from Mark. -GB Electronics will provide some guidance on the separate shielded box for the stepper motor driver. -Simon, Roger, & Dennis: work together to see that the necessary materials are ordered to build the new dewar when we get the drawings (01 sep), bearing in mind the change to copper for the 3K plate. Dennis-- estimate how long it will take to finish the dewar from 01sep. -Brian & Roger: review Simon's optics document. -Simon: send out the optics memo to an external reviewer: at least Bill Duncan, and if you like, another(s) of your choosing. Get someone else to double check the loading calculations. -Brian: draft acceptance tests by CDR. -Brian+Penn: GSFC meeting on 24/25 July to see detector progress, state of electronics & software.