We held a Mechanical/Electronic/Optical CDR for this project on November 17 & 18, 1995. The report of the committee and our responses are attached as an appendix.

One of the more significant developments as a result of the review was the commitment to write a new design report by mid-April. The report will cover the detailed mechanical design of the instrument particularly in how it relates to the structure. The outline of the report is included in the response to the CDR Committee Report.

We are thinking that we will have a small review meeting to review the report with Dan Fabricant and a few other outside experts plus the project team. We are tentatively scheduling this meeting for the end of April 1996 and hope to start construction of the structure soon thereafter.

At the end of this quarter we came to the milestone for a decision on whether or not to proceed with the plan to deliver both beams with the instrument on commissioning. As there is no firm commitment from a donor, we have elected to deliver only one beam for first light. We currently estimate that it will require three months in Hawaii to install the second beam, in addition to the time required to fabricate the parts once funding has been obtained. The longest lead time item in fabricating the second beam is the camera elements, which will take approximately a year to fabricate.

The schedule has been revised to show the development of only one beam for first light. The second beam would be installed in Hawaii at a later date. As a result of this change, the projected commissioning date has been moved forward to February of 1998. The least float activities on this schedule are the detector/dewar development, with six weeks. Engineering and Mechanical Fabrication are close behind with only a few more weeks of float.

2. Reports on Specific Areas:

2.1 Optics

Updating the optical design for the melt sheet data and for fabrication tolerances remains one of the project's major concerns. Much effort was spent in December bringing ORA up to date on the existing design and our requirements. Negotiations are in progress for them to rebalance the design for the new melt sheet data and do a tolerance analysis to guide our fabrication efforts (a contract was negotiated and they started work in January). We are also in contact with Jim Burge of Steward Mirror Laboratory, who will review all ORA reports with us. Until we resolve who will fill the optical design role during fabrication, we are not proceeding with fabrication of the camera elements. We received the Ohara glass order for the cameras. An error was made at Ohara on the blank size for lens element number four (SK01Y). The blanks supplied were too small in diameter, and they have agreed to replace the three undersized pieces with correct blank dimensions. We expect the replacements by April 1996. We are awaiting re-optimization of the camera design by ORA, based on the melt sheet data of the Ohara glass, before instructing Optovac on the lens parameters for the calcium fluoride boules. Optovac estimates a 6 to 8 week delivery once this information is supplied. The calcium fluoride lenses will be delivered to Lick generated to the specified radii, oversize in thickness and ready for grinding, polishing and figuring at Lick.

The collimator mirror was received from Kodak with the radius of curvature diamond generated and the central hole cored and stepped per specifications. A Zerodur plug was inserted in the center hole to facilitate ease of fabrication, and the surface was fine ground to the optimum initial sphere. Small grinding tools lowered the surface to match the prescribed asphere to within three waves peak-to-valley. Sub-diameter flexible polishing laps and small lap corrections have polished the surface and brought the peak-to-valley to < 0.5 wave. The additional smoothing, figuring and testing required to meet the tightened slope specification of 2.2 x 10-6 radians (1/3 the strawman specification presented at the CDR) will likely take until mid-March to complete.

The collimator cell is complete except for some required modifications of a few parts because Kodak slightly chipped the blank during grinding and generated an extra 3/8 of an inch off the surface to correct the problem. These changes are minor and will be completed when required in late February.

Ohara has supplied us with 2" diameter sample piece of each glass type for Sol Gel coating experimentation. We have two pieces of each glass type, to be polished and shipped to Cleveland Crystals for coating tests.

Good progress was made in designing and ordering materials for the camera optics tooling.

First results were obtained for silver samples sent to Keck for life testing last August. A bare sample and a sample coated with a thin layer of MgF2 were placed in the Keck dome, a bare and coated pair were place near the HIRES slit, and another bare and coated pair were placed in the HIRES anteroom. Both bare samples in HIRES are now badly discolored, the bare sample in the dome looks good, and all coated samples look good. A plan is being developed to monitor the remaining good samples. So far the plan to use MgF2 coated silver on the collimator and tent mirror looks viable.

Potential fabricators for the filters were contacted. We will probably buy new glass from Schott and have the pieces polished and cemented separately. Livermore anticipates no difficulty dipping the cemented filters in Sol Gel.

2.2 Mechanical

The major mechanical effort was to prepare for the CDR on November 17, 1995. This included many illustrations and closed many open design issues. The DEIMOS full-size model was expanded to include four cardboard grating rotator modules in a slide configuration. A prototype of the slit mask handler was completed and added to the model. A concrete dummy of the collimator was fabricated and mounted in the collimator mirror cell. The cell and dummy mirror were mounted in the model to test the system. Jack Osborne has been working closely with Brian Sutin to check the optical layout and make sure that none of the structure gets in the way of light rays coming into the grating from the tent mirror.

Jack Osborne developed the design of the slit mask handler in reasonable detail, as shown in the CDR report. The remaining outstanding component of this system is the device that moves the mask from the handler onto the kinematic mounts on the drive disk. We plan to have this design done for the April review.

The grating slide was laid out in greater detail. There are four positions: one fixed for the flat mirror, two that rotate over all angles, and one that rotates over a restricted range of angles due to a different encoder mechanism. The odd encoder is necessitated by lack of space for a third on-axis full-range encoder. The adopted encoder plan is sufficiently flexible to accommodate virtually all combinations of grating sizes, blaze angles, and tilt angles.

The FEA model of the structure and drive disk was developed further, including realistic (but not yet final) footprints, masses, and moments of all attached submodules. The basic cylindrical structure is quite stiff, and its sag does not contribute significantly to the image motion error budget. The overturning moments on the drive disk likewise are small. Further details on the mechanical error budget are discussed in Section 3.

The filter wheel was laid out, and exact filter sizes adopted after consulting with Schott Glass on prices and sizes.

Use of an NC machine to cut the masks has proven to be very effective and was well accepted at the CDR. Research was conducted on available machines, with a view to purchasing one during the next quarter. Details are presented in the CDR report.

The CDR Committee suggested moving the electronic components from the stationary vault onto the non-rotating part of DEIMOS. Design studies show that this is feasible, and the electronics have now been incorporated into the instrument carriage. CARA continues to design the switching system on the Nasmyth platform.

The tent mirror cell was completed for the CDR. A minor design change has been made to one of the mounts in response to a suggestion from the CDR Committee.

Construction of the collimator mirror cell was completed prior to the CDR.

The camera cell design evolved significantly since the last report. We are planning to use a system of mounting the elements with RTV and to use fluid coupling between closely spaced elements. We researched elastomeric lens mounts and verified that properly thermalized mounts are feasible for all of the camera lens elements. Radial displacements of all camera lens elements under gravitational loads are small and well within specs. We have designed and started fabrication of a cell to test these concepts. It will be used with the CaF2 - flint doublet that we have used for the previous couplant tests. The new tests should tell us the behavior of the elastomeric mounts, the fluid couplants, and the sealing techniques.

A first-cut thermal analysis of the structure and drive disk was presented at the CDR. The time constraint of the structure is about 50 hours with two inches of external insulation. At this level, internal temperature differences are expected to be less than 0.5C, which makes a modest but tolerable contribution to the image motion budget.

2.3 Detectors

There is progress to report in three detector development areas. The Lick Observatory CCD development effort with Orbit Semiconductor yielded the first wafer run with the DEIMOS CCD design. While a single processing error resulted in charge transfer problems in all devices, the run nevertheless had several good results. The run proved that both the design and the mask set (used in fabrication) are essentially free of flaws. In addition, the overall DC yield of imaging devices was good (>50%). Orbit is now processing another run (not for DEIMOS) which should show whether they have corrected the processing error. When this run comes out in April we will proceed with two additional DEIMOS runs.

The Lick Observatory in-house CCD thinning and packaging effort is making steady progress. Most of the individual steps in the process have been carried out on test wafers. We expect to have a thinned, working CCD by early April. Although the first run of DEIMOS CCDs from Orbit is not science-grade, there are many working devices that are suitable for our CCD thinning development, and it is one of these devices which we hope to have ready by April.

Finally, we have received word that Lincoln Labs has finished several wafers in their CCD development effort. They are awaiting the arrival of a wafer probe card, and when this card arrives they will be able to test the CCDs on the wafer. We have no word yet on when the full wafer run will be finished, as this depends on the results from the tests on these initial wafers.

2.4 Software The Software PDR has been rescheduled to March 22 1996. Many facts cumulatively resulted in less time being available to the DEIMOS project in this quarter than was originally planned. The move of faculty and staff out of NSII and into Kerr Hall and Thimann required a tremendous effort on the part of the software staff to relocate and reconnect everyone and everything. To this was added the effort of recruiting a new software person, preparing for the CDR, and the normal support software gives to the Observatory, which left very little time to make progress on the DEIMOS software. The new people and a new schedule dedicating two or three days a week by Bob Kibrick and Steve Allen is expected to push software development in the next quarter.

Despite the general shortness of time, first steps were taken to define the graphical user interface for instrument control. The HIRES/LRIS cartoon scheme will be adopted, augmented by a separate tabular input format to give users a choice for entering new control values. Considerable thought is being given to avoid incorrect and misleading information on the display.

2.5 Electronics

During this quarter the main thrust of the electronics effort has been toward refining the details in the schematic package and generating a parts list. The parts list has been completed and passed along to the mechanical engineering group. The schematic package is near completion with the following areas still in need of attention: 1) calibration lamps; 2) configuration of the CCD controller/crate system; 3) definition and design of the CCD controller; 4) miscellaneous cable drawings.

2.6 Flexure Compensation

Brian Sutin completed his sensitivity calculations of the effects of camera-element motions (tilt, decenter, defocus) on image quality and image position. These demonstrated that the proposed fiber-optic sources at the edge of the detector focal plane will indicate bulk image motion well enough to serve as fiducial sources for the flexure compensation system. This provided the last step needed for the optical proof of concept for the FC system. In addition, we received and tested the Physik Instrumente piezo translator that we plan to use to move both the tent mirror and the camera/detector. Running it remotely via terminal was easily accomplished, and we had it moving a dial gauge probe to repeatable locations within a few hours.

2.7 Miscellaneous

First walk-throughs were conducted for standard observing procedures such as filter changes, grating changes and slit mask changes. These will involve interactions between the instrument and the astronomer/instrument scientist under the guidance of specially prepared software scripts. Safety issues, manual controls, mechanical locks and interlocks, and auxiliary lifting equipment (for gratings) were considered.

3. Report from the PI's:

The main effort by the PI's was assisting in preparations for the CDR. Support was provided for studies of the elastomeric camera lens mounts, mechanical deflections of the structure, the thermal analysis, and the layout of the grating slide and filter wheel. Rough experiments were conducted confirming published curves for the reflectance of Sol Gel versus incident angle, and these were used to verify previous estimates of camera throughput. Strawman recipes for the B,V,R,I filters were adopted. A complete inventory was made of all gratings, rulings, and blazes that might ever be used, and the predicted dispersions, spectral ranges, and resolutions were tabulated. This table was used to confirm the adequacy of the layout and tilt ranges of the proposed grating slide.

Comprehensive error budgets were developed for image quality and image motion in X and Y coordinates, the latter with and without flexure compensation. Approximately 250 potential error sources were identified, including optical fabrication errors, gravitational flexure of the optics, mounts, and structure, and temperature changes and gradients. These error budgets are now helping to focus our design and fabrication efforts. For example, based on these numbers we tightened the specs on collimator surface quality considerably and are trying hard to stiffen the grating tilt and support mechanisms, which are likely to be the largest contributors to image motion. The error budget suggests that we can realistically aim for a passive flexure of 2-3 pixels, which indicates that the basic design is fairly robust. However, this is 10 times worse than our stringent image stability goal of 0.25 pixels, so that flexure compensation will be necessary.

After the CDR, the focus of the PI's shifted to preparing the optical design contract with ORA and providing them with the necessary background information.

4. Budget:

The budget and expenditures through the end of 1995 are shown in Table 1. To the end of the sixth quarter we have spent a total of $1,056,721, $476,306 of which was on labor and $580,415 on materials and supplies.

During this quarter we spent $136,014 on labor and $47,309 on materials and supplies. The principle labor expenditures were for the review in November, continuing engineering design, optical fabrication of the collimator, preparations for fabrication of the camera optics, and optical design. The major expenditures on materials and supplies were optical materials for fabrication of the camera optics and collimator mirror, expenditures for the review, and a portion of a computer to be used for software development and research.

5. Schedule

Figure 1 shows the new project schedule for a single beam instrument, as noted earlier in this report. The schedule has about six weeks of float, with the dewar/detector development program having the least float. However it will take very little slippage of the mechanical design / fabrication to put that program on the critical path.

The following is a list of the milestones for the quarter from the last Quarterly Report, together with the progress made on them:

1. The CDR was held on November 17, 1995, as planned.

2. A prototype slit mask handler was completed for the CDR, and design of the final unit will begin again once the work currently being done on the grating mechanism is completed.

3. Design of the structure, rotator, and instrument handling facilities progressed on schedule.

4. A test cell was designed for the RTV/Fluid coupling experiment, which will take place sometime nest quarter, as planned.

5. The error budget was completed to the point it could be presented at the CDR. Work on it is expected to resume during the next quarter.

6. There was little work on optical couplants during this quarter. The lenses coupled together with grease were decoupled following the completion of that experiment and in preparation for the RTV support/fluid coupling experiment to be done next quarter.

7. The optical glass was received by the end of the quarter, except for two major pieces. Those pieces, along with the ones previously received that were the wrong size are expected in by the end of this quarter. Melt sheet recalculations were postponed until next quarter.

8. Optical figuring of the collimator began as planned.

9. Construction of the collimator cell is complete and installed in the model, as planned.

10. Little progress was made on defining DEIMOS keywords.

11. A start was made at preparing for the Software PDR and work will continue at increasing levels of involvement during the next quarter.

12. CCD efforts continued, with Orbit being given permission to redo the run that failed.

13. The default plan is now for a single beam instrument.

Milestones for the next quarter:

1. Complete the design of mechanical items including the structure, undercarriage, drive, slit mask mechanisms, grating mechanisms, and tent mirror support in preparation for a detailed review, currently planned for late April.

2. Complete preparations for and hold the Software PDR on March 22, 1996.

3. Complete optical figuring of the collimator mirror.

4. Complete the melt sheet rebalance of the camera optical design and prepare for the start of optical fabrication of the camera elements.

5. Complete the prototype of the grating tilt mechanism and test its performance.

6. Complete the testing of the camera test cell to verify the RTV/ fluid coupling mount.

7. Order the CNC slit mask cutter.