Minutes of the Apache Point 3.5m telescope User's Community Meeting September 19-20, 1997 Visitor's Center, Sunspot, New Mexico Rene Walterbos, Chair Attended by ~40 representatives of all ARC institutions Minutes by Michael Strauss The agenda was posted to apo35-general #167. Table of Contents: -Status of 3-year plan -Status of the Current Shutdown -Status of Secondary -Status of DIS -Scheduling Issues -Operating Systems -Science Programs -Educational Outreach -New Secondary Discussion -Feedback from Observers -Instrumentation ---Echelle ---Guest Instrument ---New chips for DIS ---A new instrument from JHU ---Wide-Field Imager ---Possible collaborations between ARC and Durham ---Possibility of a NIR Spectrograph -Sources of Funds -Improving observing efficiency ---Dust ---Snow shutdowns ---Focusing ---Efficiency Action Items -Guest Instrumentation -Institutional collaborations -Overall Meeting Recommendations and Action Items *********Status of 3-year plan********************* Chris Stubbs Our telescope is doing science, but performance can be improved Improvements are more than just those listed in the 3-yr. plan Coordinated effort between site staff and universities What's Been Done: Top end bracing (gets rid of oscillations) New primary mirror support Realuminization of mirror surfaces Collimation Plan has been developed Slit viewer for DIS has been installed A new imaging instrument, SPICAM New guider has been installed Enclosure wheels replaced Drive system work New secondary procurement More disk space What's Next? Implement collimation plan DIS upgrades Install Baffling Observing efficiency issues Thermal management Entirely new top end Tip-tilt secondary Software upgrade Chris's Attitude: Engineering tasks take priority. Coordination of the tasks is not outstanding; to improve this would require a Project Manager working at least 1/2 time. Note all the people involved with 3.5m upgrades are also involved with Sloan. Given the difficulties of scheduling engineering time, we are starting a policy as of 1 October of letting engineering time "fall like rain" on the schedule, pre-empting scheduled observers. Note that observers can indicate in their proposal if their program is time-critical, and thus *must* not be pre-empted. The 3-yr. plan is on budget, except that the new secondary has turned out to be more expensive than originally hoped (see below for further discussion). There is evidence for a slow degradation of throughput (much more on this below), prompting calls for more frequent realuminization. It takes close to 1 month shutdown, and roughly $25,000, to realuminize the primary (much of the time taken in the transportation to and from KPNO). A new aluminizing tank would be 10-40 times this cost, but would save much in transportation time. *************Status of the Current Shutdown******************** Mark Klaene There is an Improvements group, which covers various issues not necessarily covered in the 3-yr. plan. Some of the recent work this group has done and/or thought about: Counterweights for mirror covers (so telescope stays balanced when mirror covers are closed). Replace mirror vent tubes. Moth abatement. Secondary drive mechanical unit. Automatic LN2 fill. ways to remove snow. A new Hartmann mask Ro telescope Instrument Calibration system. Automatic seeing measurement from guider. Automatic focus correction. Primary mirror cell temperature measurement. Automated night log. Much of the improvements in the current shutdown is to reliability, robustness of systems; only some of the changes will be visible to the observers. (Editor's note: the list of what has been accomplished during the shutdown is for the time of the meeting; more of course was done since that time!). Rotator Remove and clean main bearing / done. Replace drive shaft. Adjust alignment (remove burr) / done. Tertiary Install dust shields / done. Clean bearings. / done. Machine bottom of assembly flush so the assembly can be centered /done. Center and pin wrt primary center fixture. Wash and realuminize / done. The washing introduced some fine scratches, which is why it was then realuminized. The reflectivity then improved from 84% to 90%. After realuminization, the scattered light was way down. It would take 7-10 days typically to realuminize secondary and tertiary. Secondary: Not yet done (because parts didn't show up): Disassemble and repair harmonic drives " " gimbals " replace lead screws Put in limit switches. Clean up lamp installation/wiring; remote lamp operation / done. Install new cross bracing / done. Primary mirrors: Repair cateye mask / done. Remove temp sensors and reinstall / done. Reroute temp sensor cabling / done. Remove remaining vent tubes and clean for installation of new tubes / done. Clean and inspect Optical Servo System (OSS)/ done. Drill inspect holes in 0SS / done. Repair transverse links / done. Repair P-gasket seal / done. Measure load cell hard points / done. Calibration measurements for LVDT's / done. Remove damaged pressure transducer / done. Remove and repair WSMR instrument/mirror mount / done. Service altitude bearing / done. Repair leaks in PMSS plumbing / done. Removal of PMSS electronics, cabling, etc. / done. Repair mirror covers - not much work needed Repair eyelids - not much work needed Various repainting, refoam insulation. Limit switches on mirror covers / not done. Plumbing modification on eyelids. Install new vent tubes. On-Sky testing Telescope collimation. Pointing model New instrument block. Complete guider characterization. Throughput measurement. ~ 95% of planned activities will be done by the end of shutdown. Expect improvements in collimation and image quality. Shutdown extended a few days to make up for sick days. Telescope drives are in continual maintenance mode. There are probably problems w/controllers; intermittency in the problems makes it difficult to diagnose. Telescope needs more monitoring to help trouble shooting (part of 3-yr. plan). System is complicated; difficult to diagnose. Sloan has same system! People agree that this is high-priority. ******************Status of Secondary:************************** Uomoto: Current mirror not fully polished (hazy). Honeycomb pattern print-through. Best performance limit ~ 0.7" from optics. Not 100% clear it is all the secondary's fault. New Mirror: Polishing vendor is Steward. Delivery Aug. '98 - June '99. Still need to sign contract. Build polishing and testing support. Contract this out. Ship blank by Dec. '97. We want to get median seeing to 0.8". (See more on secondary below). **************Status of DIS:******************************* Jon Brinkmann fixed 60 Hz noise pick-up problem. Airgap slits 1.6", 1.9" wide. Good performance. Now medium dispersion gratings have been ordered. Planned: 3" slit, slit wheel, remote control of calibration system. Fix the problem with the UV response. This will take 3 months to fix, and entails some real risk to the optics. We're still very much read-noise limited on DIS, especially at high-res. Wish List: New CCD's! Cheap CCD's are out there; packaging is non-trivial, still $100K to do it all w/free CCD's. A strong sense that new CCD's would be a good thing. Stubbs: wouldn't it be easier to more or less rebuild DIS? I'm getting chips for my wide-field imager from Lincoln Labs w/1 e- read noise, 90% QE from 3500-10000 A. (more on this below under instrumentation). Szkody: closing mirror covers for calibration is a big time sink. DIS doesn't have internal calibration - that's what we really need. Use fibers to feed calibration lamps to Naysmith. Trying to automate mirror covering opening; not yet done. (Editor's note: the lamps themselves are now automated on the Web, and arcs can be taken in dark skies w/out closing the mirror covers. See apo35-dis #38 for details). (See more on DIS chip issues below) *********************Scheduling Issues********************* Ed Turner: Level of unhappiness with current scheduling is basically tolerable. Low-ranked proposals w/complicated requirements will always be difficult to schedule. More flexibility, info, iterations with institutional TAC's would help. People are doing KPNO style science: Lots of people getting a few nights/yr. We as a telescope community need to find our unique niche. But people are pushing for more "traditional scheduling"; not many people are taking advantage of our scheduling freedom. Chopping things up with monitoring programs cuts back on regular programs, calibration time. UW: how about institutional block scheduling? Response: Would make monitoring, synoptic programs impossible, and is quite disadvantageous to those institutions with a smaller allocation of time; it leaves much less flexibility for permutations. Calibration/twilight questions. Don't try more risky scheduling until more of the telescope stuff is working. We as a community clearly need more preparation for runs beforehand for planning observing, especially for remote observing. We have not yet taught ourselves how to maintain optimal observing efficiency while taking advantage of remote and/or shared night observing modes. Take more advantage of expertise of observing specialists. Do we want to go the route of having observatory-wide Key Projects? There is a sense that the ESO Key Project approach has not made a big impact. How about Targets of Opportunity? We need to talk more about this. (From further discussions later in the meeting): 1. Block scheduling, each institution assigned a day of the week. Each institution can schedule their own time. Going further in direction of splitting up the pie; people don't like this idea? 2. Should we go to assigning whole nights in the summer, given that they are shorter? Remember you can always ask for full nights if that is what you need! Everyone agrees that current system is maximally flexible. We need a "Figure of Merit" meter: fraction of useful time on the sky. Real figure of merit: Number of useful scientific papers out. ***************Operating Systems******************* Bob Loewenstein Software/architecture/interface. Currently: Heterogeneous system. external protocols are only restriction on instrument software. Remember that this was all thought out 11 years ago! Problems: No simultaneous instrument control Latency Platform dependence Maintenance - need experts on every system. Incorporation of new instruments Crashes ... unexplainable; sometimes magically fixed by rebooting; what is going on? We would like platform independence - Remark does not do a good job of keeping instrument parameters up-to-date. Conventions with things like rotator angles are confusing. Where does the data go? Not clear for all instruments. Each instrument has its own language at the command line. The instruments don't all follow protocol. Is this OK? Can each use different software? Do we want buffering of commands; translating to standard protocols? We need the software people on-site continually working on this. SPICAM people, e.g., have no desire to interface to unstable mc stuff. REMARK and APPLE are becoming obsolete. Real problem is that there is no person on-site working on software. We need to decide on commonality of instrument interface, and how to handle upgrades, platform changes, etc. Ed Turner will resurrect the old software advisory committee for all of this. *************Science Programs*********************** (Editor's note: There were a series of talks on various science programs people are carrying out with the 3.5m; I simply list them briefly here.) Ed Turner: QSO gravitational lens monitoring Rene Walterbos: Luminous Blue Variables in M31 Gordon Richards: Followup of QSO's from FIRST radio survey. Paul Mason: Observations of CV's; distinguishing magnetic CV's. Paula Szkody: More on CV's; synoptic observing. Nancy Chanover: Planetary Science using APO from her group and others. Don York: Finding suitable Quasars for D/H studies with FUSE. *************Educational Outreach************************* Kurt Anderson: Sunspot Astronomy and Visitor Center Part of Southwest Consortium of Observatories for Public Education (SCOPE) People like to interact with exhibits! Planned exhibits and activities: Heliostat feed: Solar images and spectra APO exhibits Interactive imaging Hands-on optics Meteorite Display Public Telescope a la Kitt Peak? Teacher Resource Center? Meetings and workshops for Scientists and Teachers ***************New Secondary Discussion****************** Ed Turner: Is everyone happy with idea of spending money on this? Evidence of Deficiencies of 3.5 m Secondary: 1. Opticians did not claim to have achieved figure goals. 2. We know there were Kodak ion-polishing "errors". 3. Lick profilometry. 4. Hartmann tests in telescope indicate 0.7" static errors + lab indication of excellent primary. 5. Visible "haze" on mirror surface. However, there have been no holographic tests of current mirror surface, so we do not know its actual figure apriori. Remember that the case for the secondary being the problem with the seeing is not completely ironclad. With only a new secondary, we imagine that the median seeing would improve by a few tenths arc seconds. Costs for secondary add up to $430K Original Cost Estimate - $280K in 1996 Budget $393K in 1997 Budget Total 3-yr. plan secondary budget (not including secondary) is $600K. Final numbers put us $125K in the hole for 3-yr. plan. The two next most expensive items in 3-yr. plan are software and DIS upgrades. The time is right for the secondary in terms of availability of Steward Labs; if we don't do it now, we will not be able to do so for sometime, even for quite a bit more money. Will there be a cash flow problem in one year or so? Could we sell our own secondary? Telescope top-end will need to be rebuilt eventually, and that will be quite a bit more expensive. But we can put new secondary on our current top end for now. ********************Feedback from Observers******************* Michael Strauss: We need better feedback from the observers as how much science they are getting out, how efficient the observing is, and what the problems are. Strauss will post a series of suggestions to apo35-general, including writing a summary of each night or each run at the end of the run, to be included in the night logs, and use of the problem reports form off of the APO home page. END OF THE FIRST DAY OF THE MEETING DAY 2 ***********************Instrumentation*************************** *******Echelle Doug Duncan: Echelle - A junior Keck HIRES. Very effective, and very large, instrument. Innovations: Most of volume is vacuum for stability. Camera is Schmidt. Usual thing is to put a folding flat in to send light to CCD in dewar. Here the CCD is put at the focus; the whole assembly is in vacuum. The optical alignment is happening now. A low-grade CCD has been put in for testing. 1.8" entrance slit gives R ~ 38,000 = 8 km/s resolution. Prism cross dispersion ==> more throughput than mirror. Even distribution of orders Can observe 3400-10000A in one observation Design goal throughput 9% or more. Stable to < 1 kms in a night. Calib lamps + flats built in. Slit viewer. The camera optics assumed 15 micron pixels, the existing chip has 24 micron pixels, and the spectra are thus undersampled. With 15 micron pixels, the resolution would double. The readnoise is 5 to 10 e-; if it were 1 e-, this instrument would be competitive with the Keck HIRES instrument. Shu-i Wang is coming to APO at the end of October to work on the instrument. Thermal vacuum test is next. User software will be similar to GRIM. *******Guest Instrument Nancy Chanover: Acousto-Optic Tunable Filter which works in the Near IR. Development for planetary exploration satellites by folks at Goddard. Technology demonstration. AOTF: RF signal applied to birefringent elasto-optic crystal (TeO2) ==> modulation of index of refraction ==> phase grating. One specific wavelength of input light deflects at fixed angle. Wavelength depends on RF frequency. Polarization splitting as well. Band pass is a sinc-squared function, resolution 200-500, narrower than a CVF. A throughput of 70-80%. No moving parts, very compact and rugged. Wavelength coverage to 5 mm. FOV is 50". Mutually beneficial collaboration between NASA and APO. More opportunity for collaboration. Another instrument: CELESTE 5-25 micron grating spectrometer, with resolution > 4000. This is available to community. **********New chips for DIS Rene Walterbos: With the current chips: The readnoise is 15e- on blue side; 28e- on red. We need to work harder on current chip electronics, could certainly drop red read noise by factor of 2. Limited wavelength coverage Limited slit length In high-res mode, it takes a ~ 6 hour exposure to be sky-limited in dark time. Clearly inefficient! Where can we get new chips? There is an NSF program to supply 4 e- readnoise chips from Loral to the community through Lick; we would have to pay for all the packaging ($200K + 1 man-year of work?). If thinned, the QE is 80%, but thinning is risky (1 out of 3 are successful). Other possibilities, left-over chips from the ST Advanced Camera or Stubbs wide-field camera. Stubbs: we should really get state-of-the art. Expertise for building dewars exists at Stubb's lab (but they won't do it for free!). More wavelength coverage will require new optics. Stubbs: Lincoln Labs is building 4Kx1K chips w/1 e- (!) readnoise, and with good response to 1 micron, QE's of 80% at all wavelengths. Stubbs is using these chips for his wide-field camera. As mentioned above, these would be great for the echelle, making it competitive with Keck. Consensus: Lick "free" chips is not the way to go, w/superior chips from Advanced Camera or Stubbs. *********A new instrument from JHU Alan Uomoto: We need to keep the costs below $500K. Wide-field spectrograph: too expensive. Instead, Integral Field Spectrograph. 20 x 20 or 40 x 40 array of lenslets at focal plane, each 1/2" or 1" across. Use 2K x 4K x 15 micron CCD. Use resolution between 2000 and 10,000. Wavelength coverage is small, a few hundred Angstroms. Relative to a fiber-fed dense pack, this has greater throughput and no dead spots. Lenslet arrays are available, but not in the size we need. Science; small galaxy internal kinematics, AGN kinematics. How do you do sky subtraction? "This is not a replacement for DIS." ***********Wide-Field Imager Chris Stubbs The telescope delivers a 30' field; let's use it! Array of 24 2Kx4K thin chips (with small gaps), with 0.08" pixels, with no re-imaging optics, so no distortions. 2x10^8 pixels! Focus monitor Tip-tilt pick-off Optical design is easy Detectors on order Electronics from SPICAM 12-18 months away No atmospheric dispersion correction Baffling is a challenge If observing remotely, write tapes at APO, and send them by mail; the internet can't handle this much data! Other schemes Chris is into: Ways to defeat fringing on CCD's Ways to increase dynamic range > 16 bias Tonry's method of doing tip-tilt on-chip by moving charge around Deep depletion CCD's - 90%QE @ 1 micron. 2.5m telescope at CTIO, possible MACHO 2nd generation telescope. *********Possible collaborations between ARC and Durham Bernie Rauscher Various levels of collaboration : 1. Consortia to buy expensive components (clips, filters). 2. Obtain specific + unique component for an instrument in exchange for a few nights/yr. 3. Complete PI instruments or larger blocks of time. 4. Couple facility-class instruments. Benefits to ARC: 1. Timely access to cutting edge instruments. 2. Access to instruments ARC might otherwise not be able to afford. 3. Greatly reduced costs for critical technologies. 4. New people will bring in new ideas. The Durham Group is working on adaptive optics, on various strategies for integral field spectroscopy, and various IR instruments. Interested in collaboration for an IFS in the near IR, and wide-field J and H spectroscopic surveys. ************Possibility of a NIR Spectrograph Tom Harrison - GRIM has limited spectroscopic capabilities. To get R = 800, you need a 0.35" slit. The read noise is 110 e-, and no multiple reads (Industry standard is 30 e-; Hawaii folks are getting 10 e-!) Lambda calibration is hard; lamps are faint. Some order overlap problems. So let's think about new instrument. Cheap fixes: new grism? multiple reads? Should we build a simple stand-alone IR spectrograph? For example, R = 800, 1500. Cost, $500K if built from scratch, $350K if you copy Rieke's design. Emissivity of telescope is important on 3.5m. Not an infrared optimized telescope! *************Sources of Funds Tim Heckman Resource estimate: steady state is 4 "facility" instruments The productive lifetime of a given instrument is 10 yrs., at $1 million/instrument, this is $400K/yr. Compare with the 3-year plan, $250K/year, and the operating budget of the observatory, $600K/yr. Instruments are a non-trivial cost! This is really a post 3-yr. plan issue, as we don't have money now. Questions - Should instruments be fully funded in ARC? Should we offer matching funds? Should we have a best instrument competition? Would user's committee suggest specific needs? Who decides what to build? Oversight and issue of cost overruns / schedule slip; who has authority? Acceptance criteria of new instruments? How do we keep instrument builder involved following delivery? Other Models: 1. ARC PI instrument What rewards/inducements for PI to share instrument? Financial - observing time - collaboration with users 2. Visitor Instrument (provided by outside party) Issues: Long-term vs. short-term Impact on operations costs: long term is better Support / documentation Acceptance criteria Scientific vs. defense-related (cf, WSMR) 3. McCray committee report approach; Apply for funding from NSF, some of the telescope time goes to the public. This program got 0 proposals last year! There are real operation costs involved in getting outside people observing at APO. We don't need or want 10 instruments! We are not necessarily trying to cover all bases like KPNO. But 4 facilities instruments, cutting edge, sounds like a basic need. We also need to find ways to be unique. Too many partners gets awkward. The consensus is that we need to continue 3-yr. plan level of development funding. ***************Improving observing efficiency***************** ***********Dust Many people are concerned that there have been too many closures on good nights due to dust. In Tom Harrison's words: "Why are the only observatory in the world that monitors dust?" Bruce Gillespie argues strongly that dust monitoring is a good thing. We know we live in a dusty environment, and that dust cuts throughput; cleaning and realuminizing mirrors is *very* expensive. Mark Klaene: Where the dust policy comes from: Dust clearly cuts down throughput. As soon as it gets moist, we get permanent degradation of surface. Dust counts (> 1 micron particles per 0.1 cubic foot) oscillates around ~ few hundred, w/peaks to a few thousand. But integral seems to be dominated by low dust; as peaks are very narrow (albeit often very high). John Varsik of Sunspot has studied this problem in detail, and can quantify what fraction of dust is avoided by closing what fraction of the time: Based on data from May-December'96: Dust Cut off Fraction of Fraction of time (1000 particles Dust Avoided Closed per 0.1 cubic foot) 1 0.5 0.13 1.5 0.33 0.06 2 0.23 0.03 2.5 0.18 0.02 3 0.14 0.009 3.5 0.12 4 0.11 Dust Policy: At 1000 counts, the Observing Specialist starts paying attention. At 1600 counts, use flashlight beam, humidity to make judgment call. Need more work w/witness samples. Dust is worst in April-June, when weather is otherwise the best. Throughput is now measured every month w/DIS spectrophotometry. There seems to have been a drop of throughput of 15% between March to June. Ouch! We discussed possible errors in this measurement. Suggestion: Close when dust flux is, say 10x the previous five day running mean. Karen shows that dust alone has shut us down ~ 5% of all time over last 8 months. Weather has really killed us. Fraction of time shut by dust only: 0% Feb. - April, 20% May - July Weather in general: ~ 40% lost; a bad spring! ~ 10% of time lost to equipment. This was an unusually dusty year. Dust monitor is not very sensitive to pollen. Gypsum (calcium sulfate) is water-soluble, and therefore especially bad news. Keep in mind that realuminization can't be done often, for love or money, given limited availability of the KPNO tank. Gloria: Estimated time per year lost to realuminization. Once every 3 years, 3 weeks per shutdown (in winter), plus one wash per summer. Minus time lost to weather anyway ==> 53 hours lost per year on average. If we had an on-site aluminization tank, it would be 26 hours per year. Vague possibility of aluminization for Starfire Optical Range. (Editor's note: see the recent discussion of this by Eric Deutsch: http://www.astro.washington.edu/deutsch/apoinfo/cloudscan/duststudy.html) *************Snow shutdowns Mark Klaene Why is snow bad? 1. Snow occasionally gets through the slit; snow massed on the roof can fall in when we open the slit; could damage the secondary. 2. Ice forms on top of roof; pulling open shutter could damage shutter mechanism, or could make it difficult to close. 3. Snow crystals blowing off the roof onto the primary. Policy is: OS evaluates roof condition. Don't get on roof when it's icy! Look for ice formation at the pitch of the roof. During snowstorm, we run snow program; running the telescope in azimuth to prevent drifts. Or point it into the sun. Also roof heaters, which don't do much below 29 degrees F w/out sun. Relatively few nights have been lost in the past just due to snow on roof. But upcoming El Nino may change this..... Suggestions: Put in more heaters? A real power hog. Perhaps put video camera on roof to look at ice. Use blowers to blow snow off while it is falling? Put "Flat field screen" over telescope while you open up? *******Focusing Focusing offset done by Tcc based on the truss temperature. The coefficient is empirical, but doesn't work so well. Use vane temperatures rather than truss? Primary temperature is also a factor. Experience shows sudden focus changes happens often. Aim: closed loop focusing w/guider. Requires collimated optics! ********Efficiency Action Items Chris Stubbs: ==> twilight time! Let's use it; not let it go to waste. ==> Instrument changes ==> tertiary rotation. 30 minutes a shot. ==> Fraction of time on which photons are hitting detectors. It is low in remote observing. Let's train ourselves to become efficient remote observers. ==> Better detectors. ==> Programs appropriate to conditions. ==> Improved communications over the internet. DIS - A real-time slit viewer really would speed things up. Hang-ups due to: rotator problems; computer crashes and telescope run-aways. ********Guest Instrumentation We pay a lot for site staff and operations for guest instruments. ARC should get more benefit, by e.g., making guest instruments available to the whole community. How would collaboration work? After all, guest instruments usually are not user-friendly. We have to get away from model that telescope is a pie, separated into six pieces, completely independent. 10% of Operations cost goes to visitor instrumentation. Distinction between "visitor instruments" from ARC institutions, and those from outside. Only the latter should have a requirement that they share with other ARC members. Much discussion on nuances of all this. *************Institutional collaborations: Formal Observatory-wide Key Projects? Not much enthusiasm. Informal collaborations between institutions? Already happening; we should work hard to make more of this, and do some unique science! Target of opportunity: people are enthusiastic about gamma-ray bursts. Use director's discretion to pay people back who get bumped for this. *************Overall meeting Recommendations and Action Items*************** Community is enthusiastic about 3-yr. plan! New secondary: go ahead. Action item: develop commissioning plan for echelle. Scheduling - current system is flexible. Can ask for full nights! Shutdowns: Let them fall like rain. Need Director's Discretionary time; if the sum of engineering + DD time > 20%, then no DD time. Otherwise, DD time is 20% - engineering. (or less) Are shutdowns included in this calculation? It is not clear. Software upgrades are important, but not highest priority. List of publications should be brought up-to-date. List of users should be updated on apo35-general. (Editor's note: this is in progress.) Strauss will post to apo35-general a statement about how to report problems with observing, and how to get a better measure of efficiency in the logs. (Editor's note: Strauss is waiting on response from observatory personnel on his draft statement, before posting it.) Guest instrumentation: give the community a chance to use them. We need new chips for DIS! In the meantime, we should work harder on the existing chips to beat down the read noise. We will probably not use the Lick chips. It is difficult to get member institutions to agree to outside instruments in exchange for telescope time. No consensus on IR spectrograph. Multi-read on GRIM should be implemented! Instrument scientists need to be more involved: DIS: Alan Uomoto has been "acting instrument scientist", but is not formally in that role. GRIM: Mark Hereld? Echelle: ?? SPICAM: Chris Stubbs. Continue 3-yr. plan level of development funding for hardware in the long term. Continue quantitative study of dust, snow shutdown policy. Let's get more rigorous about DIS throughput measurement, figure out if throughput loss is indeed due to dust. Implement Close loop focus. Continue to consider aluminization tank. Let's use twilight time as best we can. Let's educate ourselves on how to observe efficiently. Engage observatory folk in preparations for runs one day in advance. Remember that the OS is ready to go at sunset. Let us encourage informal collaborations. Stubbs will write up a plan to do gamma-ray bursts. APO APO APO APO APO Apache Point Observatory 3.5m APO APO APO APO APO This is message 193 in the apo35-general archive. You can find APO the archive on http://www.astro.princeton.edu/APO/apo35-general/INDEX.html APO To join/leave the list, send mail to apo35-request@astro.princeton.edu APO To post a message, mail it to apo35-general@astro.princeton.edu APO APO APO APO APO APO APO APO APO APO APO APO APO APO APO APO APO