The major item of progress during this Quarter was finalizing the collimator and camera optical designs and ordering the optical glass for the cameras. Three sets of optical glass were ordered, and delivery is expected approximately December 1995. We ordered three sets to ensure that both cameras would be identical and to ensure the availability of the glass. The extra set of glass was ordered as insurance against breakage both during manufacture and for the lifetime of the instrument. In the Optics report (Section 3), the reasons for this decision are expanded on. At the end of the quarter, we also placed the purchase order for the CaF2 for a single beam plus one spare lens blank.
The optical glass cost was significantly more than what was originally budgeted, and adjustments to the budget and the scope of the project were necessary. Our single-beam budget allowed $40,000 for each set of two sets of optical glass for the cameras. This is consistent approximately with the $43,166 per set in the May 12, 1994 quotation from Ohara for the PDR camera model. To enable the purchase of three sets, the SSC ok'd our taking $40,000 out of contingency (in May 1994). This brought our available total to $120,000.
From the time the budget was set up until the optical glass was ordered, two significant developments altered the price of the glass. The design of the camera was changed to make it efficient down to 3900Å rather than the original 4400Å. Additionally the U.S. dollar lost considerable ground on the Japanese yen, and Ohara of Japan is the only possible supplier of the required glasses. However, this was mitigated somewhat by the fact that Ohara gave us a 100 yen per dollar exchange rate, though the actual rate was 83 yen per dollar at the time of our order. Thus, approximately $35,400 was actually saved by this courtesy! The final cost of the glass turned out to be $64,500 per set. The total cost of the three sets is $193,500, or $73,500 more than currently budgeted.
We have opted to purchase three new CaF2 blanks to accompany the one we presently own. One of these will be a spare. A set of three lens blanks will be generated by Optovac to the rough lens forms required. The total cost for the new blanks and for generating is $72,000. The spare blank can very likely be sold back to Optovac if it is not needed or wanted later as a long-term spare to be held throughout the lifetime of the instrument.
To offset the increased cost of the glass, the project elected to reduce the number of TV cameras planned back to one, delete the purchase price of the glass window at the entrance to the instrument, and cut back on some upgrades of optical equipment. These changes more than offset the additional costs, and $16,000 was returned to the contingency fund. The existence of a window of some sort is still in our baseline plan. Either we will obtain a blank for free (e.g., from Livermore), or we will install a temporary window out of Mylar. Remarks on the advisability of the second TV are found in Section 4.
On March 15, 1995, a meeting with Frank Melsheimer, the PI's, and the Engineers was held in Santa Cruz. The purpose of the meeting was to review the design concepts for the Structure, the Mask Handler, and the Grating Slide. As a result of this meeting we have decided to design the main part of the instrument as a monocoque shell with bulkheads as the main transverse elements. We are leaving the option open to design the nose of the instrument, which houses the TV cameras, as a space frame. A decision on this will be made later in the year.
An option discussed during this meeting was to have the mask handler designed as a link belt, with each of the positions for the slit masks being one link on this belt. This variation of the handler allows a more flexible geometry and reduces the out-of-focus distance (see below). A detailed design study of this option is planned to begin during the current quarter.
Two basically different variations of how to store and insert the gratings were discussed; a slide, which could accommodate up to three gratings and a flat mirror, or a grating module handler that could conceivably handle up to four gratings and a mirror. Design studies of these two options are currently in progress, and a decision as to the best method of handling and placing the gratings is expected early in the next quarter.
The optical layout of the grating/focal-plane area was finalized, and clearances were adopted between the gratings and the bulkhead of the telescope. The adopted focal position is 3 inches beyond nominal focus. The optical effects of this were investigated and found to be tolerable. The suggestion at the PDR to move a long way out of focus and correct by restacking the primary proved optically non-viable. We have agreed with the Project that we will be on the right Nasmyth Platform of Keck II.
A revised cost estimate of the electronics portion of the project was completed, resulting in a further $13,000 being returned to the contingency fund. Because of this review, a major improvement is likely to occur in the way we control the servo motors. The baseline design had the motor controllers in the vault and many multi-conductor cables running into the instrument. New stand -alone Galil servo motor controllers may allow us to route only power cables and an Ethernet cable into the instrument. There may be further savings, not yet figured, in the motor-control budget.
Steve Allen is back and working full time, much of it on DEIMOS. He and Bob Kibrick are currently preparing for the Software PDR planned for the first week of July 1995.
Sandy Faber, Garth Illingworth, and David Koo are currently preparing a donor prospectus for funding the second beam. An estimate of the cost of the second beam was completed and comes to a little under two million dollars, including a 15% contingency. The figures and text are nearing completion, and the intent is to have a finalized document ready by the end of May 1995.
2. Status of Risk Items
Flexure Control System
The new software budget assumes that we will use Physics Instrument (PI) Piezo Actuators with integral position feed back. The actuators would have a stroke of 90 m. PI is well acquainted with astronomy and participated with the University of Hawaii in their development of an active secondary system for the 88" telescope on Mauna Kea.
Optical Fabrication
We are now planning to fabricate the Collimator before the Camera Optics in order to ease the optics schedule in the later part of the project. Fabrication of this mirror is planned to start in mid summer 1995. Slit mask Handler We are now considering a Link Belt design to alleviate congestion around the drive ring and the grating handler. Design studies of this option are planned to start in the next quarter.
Detector Mosaic
As will be discussed under Detectors (Section 3), Richard Stover's group continues to make good progress on development of thinned CCDs. Since the Lincoln Lab CCD development effort will soon be under way, some engineering time may be needed in the next quarter to help define a suitable CCD mounting structure for their CCDs. CCD Data handling, and Total Software Burden This is reported on under Software in Section 3.
Slit mask Cutter
The Request for Quote (RFQ) on the cutter has been re-drafted. We are preparing to send out an additional RFQ to cut a number of demonstration masks to both laser cutting shops and laser vendors. The intent of getting the demonstration masks is to allow us to assess their quality and to be able to test the repeatability of the mounting systems. Among other things, this should give us a better understanding of what the cost to contract the mask cutting might be, and what sort of turn around time is likely.
Total Budget Overrun
We have returned some money to the contingency fund this quarter, but have had to cut some capability from the instrument to do this and also fund the cost increase for the optical materials. We will have to continue to be very realistic in what we can afford to do for this instrument in the way of enhanced capabilities over the original proposal. The trend to downscope the instrument's capabilities may have to continue. The intent is to suspend areas that can be later added back to the instrument after it has demonstrated its science capabilities and can be justified and funded at a later date.
3. Reports on Specific Areas
Optics
The Optical Lab has received the 24" Strasbaugh Generator ordered in December. It has been set up and used in a test run on in-house optics. It works beautifully, yielding excellent surfaces with accurate control of feed rate, radius, speed, and finish, giving us a warm and confident feeling about processing the large and delicate DEIMOS optics on it. We also used the newly purchased diamond wheels in our test run to assess cutting efficiency and quality of finish, and we were happy with the results.
Other items ordered but not yet received are the Mildex digital direct reading spherometer, the Zygo f/.65 Transmission Sphere and two 16" capacity Strasbaugh precision polishing machines. All these items are expected to be delivered within the next quarterly period.
We have also requested quotations from three glass manufacturers (Schott for Zerodur, Corning for ULE and Ohara for E6), for the collimator mirror blank. This is a mirror 46.3" in diameter by 4" thick with a concave sagitta of about 1.5". We are also seeking quotations for the diamond generation of this curve prior to grinding, polishing and figuring the ellipsoid. Currently, it seems likely we will use ULE and have Corning do the preliminary work before the blank leaves their facility, but an outstanding quote from Kodak has yet to be considered. We expect to place the order for the blank within the next week or two, with delivery expected in about 14 weeks after order placement.
A complete redesign of the 15-inch focal length PDR Camera-11 Run No. 8622 (04/28/94) was undertaken during this quarter. The main purposes were: to increase the camera's chromatic range to include the full photometric "B" band (down to 0.39 microns); to include realistic filter material(s); to improve blue transmission; to assure with firm price quotes that all of the required optical materials were available; to adjust some of the mechanical details such as vacuum window thickness, b.f.d., operating temperature, coupling material; reduce maximum aspheric deviations and the like.
The goals of the work were to adopt and to freeze a specific preconstruction optical design and to order the optical materials for that design. These goals were in fact achieved. The Appendix to this Report includes a Lick Observatory Technical Report on this work titled, "Adopted Preconstruction Optical Design for the DEIMOS 15.0-Inch Camera Lens" (Harland Epps, March 30, 1995). The optical glass order to Ohara Corporation was placed near the end of March. The RFQ response for the CaF2 from Optovac was also received near the end of March and the purchase order for that material is being processed by the UCSC purchasing department for release to Optovac in early April 1995.
During the course of this redesign process, alternative glasses were studied to find a set of available glasses with improved UV transparency relative to the PDR camera. An attractive model which called for (Ohara) @BAK5Y was abandoned because that material proved to be unavailable. The alternative model used (Ohara) @SK01Y instead.
Three sets of optical glass were ordered so as to provide enough glass from the same melt of each type to construct two cameras and keep a spare piece of each glass (and two additional blanks for the small, thin and somewhat fragile field flattener). This was considered necessary as a matter of prudence. It is very likely that these materials in the required sizes will not be available in the future. Ohara accepted this order as a special courtesy and only after considerable "prodding" of Ohara's president, Mr. Brion Hoffman. The total cost for the optical glass will be $193,500 including state tax and shipping.
It was decided to purchase three new CaF2 lens blanks to accompany the one we already own. A set of three blanks will be generated to the rough lens forms by Optovac. The fourth blank will be held by Optovac as a spare and may be "sold back" to Optovac if not needed by the project. The total cost for the new CaF2 blanks and generating will be $72,000 including state tax and shipping.
One slight "downside" to the camera's wavelength extension is that the (theoretical) resolution of the camera has been degraded somewhat as compared with the PDR design. However this effect is negligible over much of the spectral range and minimal (at the level of about 0.5 pixel increase to the expected image diameter or less) over most colors and field angles. Details of this are seen in the aforementioned Technical Report. In practical terms the consequence is nearly negligible and any residual "downside" in resolution is far outweighed by the increased chromatic performance in the UV.
On February 27, 1995 H. Epps and B. Sutin presented a comprehensive review of progress on the DEIMOS optical design to the Science Steering Committee. The outline for that presentation is seen in the Appendix of this Report. The only substantive revision of that information is in the area of the projected "time line" for the camera lens production. Ohara now gives us a 7 to 8 month estimated delivery time for the optical glass instead of the 4 month time estimated for the PDR optical glass. The delay is primarily due to current production commitments at the Ohara factory. Thus, we now anticipate that the glass will be received by December 15, 1995 (instead of August 15, 1995) and that the start of lens production will occur by January 15, 1995.
This "slip" in schedule will be accommodated by advancing the start of the collimator production, and it will probably also be possible to work on some of the CaF2 surfaces before all of the melt sheet data for the glass arrives (if necessary to keep our opticians "busy" in the shop). Thus, it is not anticipated that the glass and/or CaF2 delivery date(s) will play any negative factor to the overall optical production schedule for DEIMOS.
Collimator, Slit Mask, and End-to-End System Performance:
Varying the conic constant of the collimator was studied, and the best value found was a conic constant of -0.75. Pulling DEIMOS away from the focal surface by 3.0 inches degraded the rms diameter spot size by about 0.01 arcsec, or less than 5.0 percent.
The collimator was reimaged with the adopted Epps camera design Run No. 3761 (03/02/95), and the images were always better than 0.38 arcsec rms diameter in the photometric "V" band and always better than 0.42 arcsec rms diameter in the photometric "B" band. The larger part of these aberrations are due to the telescope, rather than the collimator or the camera.
A spherical slit mask surface was compared to a cylindrical slit mask surface. The cylindrical surface which gave the best rms spot sizes over the entire field degraded the images by no more than 0.03 arcsec at worst, while being closer to 0.01 arcsec over most of the field.
The worst case end-to-end spectrographic performance was studied by using the 1200 line/mm grating centered at 0.8000 microns. The worst spot was in the red end of the spectrum at the end of the slit, where the camera is driven the hardest. This spot had an rms diameter of 0.37 arcsec, but also had a considerable tail, consisting of about 15.0 percent of the total energy.
The details of all previously mentioned collimator related calculations are given in a Lick Observatory Technical Report titled "DEIMOS Collimator and Slit Mask Surface Design and End-to-End System Performance", (Brian Sutin, March 31, 1995) which is included here in the appendix.
Detectors
We have been working with Orbit Semiconductor as one possible source for the DEIMOS CCDs. The design of the masks used in the fabrication of those CCDs has been completed, and the first wafer run using the masks is under way. We should see the results of this first run by July. Several more wafer runs are planned following the completion of the first run.
To use the Orbit CCDs in the spectrograph they will have to be thinned, and the effort to set up a thinning facility at Lick for this purpose is continuing. We recently received a donation of equipment from Intel valued at over $130,000. In addition to setting up a clean room at Lick, we are working with faculty, staff, and students in the Electrical Engineering department at UC Davis to develop the thinning process. Using their extensive clean-room and wafer processing facility we have been testing various aspects of the process and have even thinned several silicon samples. We expect to have some thinned Orbit CCDs before the end of the year.
The CCD development effort at Lincoln National Lab has gotten final approval from the Air Force, and Gerard Luppino, as head of the development consortium, is negotiating final details of the contract. Development work at Lincoln is expected to take at least a year, although initial results will be available in six to nine months.
Software
Steve Allen contacted several vendors of RAID disk drive systems. The goal was to determine which technologies were available now and what may become available within the next 2 years. He found that current pricing is approximately $1.00/Mbyte. Of more importance was the bandwidth expected for these disk systems. The vendors made it clear that the only connection technology which can be guaranteed to be standardized in 2 years is Fast Wide SCSI. This will permit transfers of about 37Mbytes/sec between CPU and disk, on 14 sec per 8k x 8k image.
He also began the investigation of feasible schemes for storing the DEIMOS images on disk. No existing, tested scheme for storing such large images as FITS seems satisfactory. Steve has established contact with a team at GSFC and NCSA who have proposed an additional element in FITS files which will permit cross referencing related subimages in a manner that makes sense for DEIMOS images.
As mentioned in Section 1, we are planning to switch to stand-alone Galil servo motor controllers, which would eliminate the need for the motor controller VME crate. Pending evaluation of the Leach S-bus interface board, we may also be able to eliminate the CCD controller VME crate. Given the likely elimination of both VME crates, it is unlikely that we will be using the EPICS control system for DEIMOS. Our current plan is to use the MUSIC-based messaging system that was used in HIRES, and to inherit software for the stand-alone Galil motor controllers from the MOS spectrograph at Lick, which is also a MUSIC-based system. Kibrick and Tucker are currently evaluating the schedule and budget implications of this change in the DEIMOS software architecture. Our preliminary judgment is that this change should simplify the overall software burden.
Mechanical
The laser-cutter RFQ was outlined. We have studied a cylindrical focal plane and found it to be adequate. We have re-worked the preliminary slitmask handler for simplicity. The mechanical error budget for the entire instrument was expanded, and it is currently 80% done. Two more camera lens designs were drawn and investigated. A solid optical drawing was generated to define "stay-out" zones for the structural elements.
We investigated moving the instrument back from the nominal focus in the following increments: 20 inches, 8 inches, 4 inches, 3 inches and 2 inches. (Each of these 5 distances required a corresponding slit mask mechanism and in some cases, grating slide geometry. There was the "side-wheeler", "mill", "petal" and "caterpillar", which now has two variations: caterpillar with masks that always face outward and caterpillar with masks which remain parallel to each other.) We have decided to adopt 3 inches as the distance to move.
We have compiled and plotted the temperature history of the inside and outside of HIRES for 9 months. Based on this gentle profile, it was decided not to regulate the temperature of the DEIMOS lens.
We looked at increasing the number of positions in the grating slides from three to four or possibly five, with a flat mirror in one of the positions. We decided to use a motor and encoder to rotate the gratings instead of having a limited number of fixed grating tilt angles. We also looked at non-standard grating shapes (other than the 6 by 8 and 8 by 12 sizes sold by Milton Roy. New angles for the tent mirror were investigated in order to make more room in the area of the grating, but no improvement could be made
Frank Melsheimer (our Structural Consultant) visited Lick on March 15, to help with the structural design. His expertise was a great help and we are still evaluating the input we have received from him and plan to report to him on a regular basis. (He also helped with bearings, drives, stages, and optical support.)
4. Report from the PI's
The major work of the PI's this quarter focussed on preparing a prospectus for possible donors for the second beam of DEIMOS. Faber, Illingworth, and Koo have been working with the UCSC Development Office to prepare an illustrated text with budget and schedule. A final document should be ready by the end of May.
Observing at Keck I with LRIS continues to be an excellent development platform for DEIMOS. To circumvent possible flexure between the LRIS offset guider and slitmask holder, we prepared 6 small pick-off mirrors that mount directly on the LRIS slitmask holders and feed the slit-viewing TV. Guiding is accomplished by holding a guide star fixed with respect to the boundaries of the pick off mirror. A first test at the end of March was very promising and another test will occur at the end of April. This may be a good method for guiding in multi-slit mode with DEIMOS and would reduce need for a second TV, as the guide point would be closer to the center of the focal plane.
Tests to understand red fringing in thinned CCDs are continuing using the Kast spectrograph on the Lick 3-meter. A removable white card has been installed in front of the secondary mirror, and it has been verified that the observed fringe pattern is stable as to how this card is illuminated. The idea is to use direct illumination of this card as a substitute for flat-field illumination through the telescope optics. Tests of this card as a means to flatten the daytime sky are now commencing.
5. Budget
The Budget sheet for the third quarter of this project is shown as Figure 1. The largest single expenditure during this quarter was the purchase of the optical glass. The project had to re-assign approximately $74,000 from other areas to fund the purchase. This was done largely by eliminating the purchase of the glass window and the second TV camera.
The other area of significant expenditures was on materials for Optical Lab equipment including a transmission sphere, two polishing machines and a spherometer. The cost of these totaled approximately $41,500.
We are expending labor hours at about the expected rate. The areas of principal effort were mechanical and electrical design, project management and administration. Approximately 1,000 man hours of effort was expended during the quarter.
We are planning to use an updated motor controller over the one used in HIRES, and this will possibly result in further savings in both materials and manpower.
6. Schedule and Goals for Next Quarter
Figure 2 shows the overall project schedule with a completion date in mid 1998. Figure 3 shows the detail of activities planned to start before the end of this year.
We include a list of the tasks we said we would complete during this quarter from our last report and assess the progress made against each one.
1) Finish the mechanical error budget: We worked more on this and it is about 80% done.
2) Finalize the camera optical design and order the optical materials: Done.
3) Start optical fabrication of the camera: Delayed owing to delay in receipt of optical glass. To compensate for this, optical fabrication of the collimator has been advanced to start in late summer of this year. We are likewise advancing the design and construction of the collimator mirror cell. We will also start working the CaF2 blanks earlier than originally planned. The net effect should have little impact on the completion date of the optics.
4) Hire a structural consultant and have a get-acquainted meeting: Done.
5) Make major progress on the mechanical design of the structure, the grating holder/slide mechanism, and the slitmask handler: Major conceptual progress was made in all three areas, and a detailed design has started on the grating mechanism. Overall we are probably not quite as far along here as we hoped.
6) Decide on a final focal position relative to the nominal Nasmyth focus: Done.
7) Test image displays of large images: A start made but not much done yet.
8) Purchase optical fabrication equipment: Done.
9) Functional specs and prototype test for rotator drive: Postponed until basic structure better defined.
10) Prepare updated electronics budget: Done.
11) Prepare slitmask cutter RFQ: Mostly done. Test masks from cutting houses are also being solicited.
12) Lincoln Lab CCD contract to start: Done.
13) Software ramp-up: Still delayed. Getting started on requirements this quarter.
Milestones for the next quarter include:
1) Functional requirements for several software areas.
2) Order collimator blank.
3) Major work on the mechanical design of the structure, the grating mechanisms, and the slitmask handler.
4) Begin work on the Drive and Support system. Test run the DEIMOS rotator motor control electronics using the HIRES image rotator as a prototype.
5) Start a checklist for the CDR.
6) Receive and evaluate test slitmasks from vendors.
7) Complete donor prospectus for second beam.