DEIMOS Quarterly Report

Number 13

July 1, 1997 - September 30, 1997

 

 

1. General Items:

 

The CaF2 Element 3 of the camera remains the most serious potential problem for the project. The boule for this element is currently in the cool down cycle in the furnace at Optovac. It is expected out of the furnace on November 7th, and will then be generated at United Lens. If all is successful, we expect the generated element in Santa Cruz in mid-December, and will complete the optical fabrication of the camera in mid-February.

 

However, if Optovac is not successful in delivering this element, it will take about four months to produce a replacement.

 

We know of a possible replacement optical glass for the CaF2, which is an Ohara product (FK-03). There will be several challenges if we need to use this substitute. FK-03 is approximately 1.5 inches thick and 14.5 inches in diameter. We would need to optically couple two disks of the FK-03 material to get the needed thickness required of Element 3, and we would have some vignetting in the camera. ORA is studying the optical consequence of this substitution, and initial results suggest the substitution is favorable. Delivery time of the FK-03 is estimated to be about six months from date of order.

 

Optovac has not started growing the replacement crystal for Element 5 that we ordered, as the furnace is currently being used to grow Element 3.

 

David Hilyard has plunge ground the aspheric surface of Element 1 and completed the touch up grinding, and polishing is set to begin. Generation of Element 4 is about to begin.

 

Only Elements 1, 3, 4 & 6 remain to be completed before a full set of lens exists to assemble the camera. One side each of Elements 1 & 6 have been completed. Element 5 will be re-fabricated when we get a replacement and exchanged into the camera when possible. Although the current Element 5 remains structurally stable with its fracture, the concern is that when cooled, the fracture may grow.

 

Sandy Faber has identified several optically acceptable couplants for the camera, and we will be testing for any possible reactivity with the RTV used to mount the camera lens. This testing is planned for November - February.

 

We have delivery of three 6x8-inch gratings; 600, 800, and 1200 lines per mm. Design and fabrication continue on the grating system, with the expectation we can start tests on the spectrograph by the end of March. The grating system is not expected to be complete until mid-1998, which is a slip in this part of the schedule.

 

In September we successfully placed a slit mask in the focal plane using an air cylinder. This was a major accomplishment. The next major step will be to power the cassette, which we expect to do by the end of October.

 

The framework for the rotating part of the electronics enclosure was installed on DEIMOS. We rotated DEIMOS under computer control and determined that the friction encoder would not give us the desired pointing accuracy. We have designed a replacement encoder mount and plan to install it during the current quarter.

 

The design of the camera barrel was transferred to a consultant, Alan Schier. Alan is proceeding well with the design and a formal review is planned for November 14th.

 

The Leach-2 controller system for the flexure compensation (FC) system is now complete and being tested. The test dewar is complete and will go to the CCD lab with the FC system. [Later it developed a leak and is being fixed.]

 

The detailed design of the dewar system is complete and fabrication has begun. Work on the mosaic continues and we expect to sign off on the material for the back plane by the end of November.

 

Significant progress has been made in software in the last quarter. A homegrown user interface development tool was released, called Dashboard, and 80% of the low-level code to operate the motor controllers has been completed and was tested. Automatic code generation programs were completed which will generate much of the low level motor control code directly from the keyword database. Work was started on extending and improving the existing CCD image readout and display software for use with DEIMOS.

 

We are currently on schedule for commissioning of the instrument starting in January 1999. This is the earliest possible commission date and depends on little or no interference in the shops by ESI. The mosaic/detector program is on the critical path of the project at this time.

 

This quarterly report has a revision to the budget. The detail will be shown in the budget section of the report, but in general terms we are reducing the contingency to $78,000 and reducing the estimated manpower for Software by 4,300 hours to 13,900 hrs. The revised software manpower estimate is intended to take us only to commissioning, and at about that time we will make a request for additional funding for post commissioning software development. The change was made by the project to redirect funding to pre-commissioning activities, and is not intended to imply that the original software estimate was inaccurate.

 

 

2. Reports on Specific Areas:

2.1 Optics

Plunge grinding of the Element 1 aspheric surface was completed using the new set-up on the Optical Labs Bridgeport to better control the centration of the asphere on the blank diameter and to control the front to back surface wedge. The results were a wedge of .00078" between the back spherical surface and the front aspherical surface (T.I.R. at the edge), and an OFFCTR (decenter) from the CURVMON output of .0016". The OFFCTR goal was originally .001", but ORA has since reported as much as .003" would not be detrimental to the image. Touch-up grinding of the asphere has begun, and polishing will be started soon.

 

Terry Mast and David Hilyard visited Optovac and United Lens to discuss the progress on the CaF2 ingot being grown for Element 3, and to review plans for shaping the ingot into a lens blank. United Lens is ready to generate the curves on the blank after some preliminary shaping by Optovac. Optovac will face flat and edge the raw ingot to the required diameter, the operation involving the most risk, and then deliver the blank to United Lens for curve generation. The ingot is expected out of the furnace on November 7th. The machining operations are likely to take 2-3 weeks. We expect the lens blank here at Lick by early to mid December. We will need 4 to 6 weeks to grind and polish the lens to specifications. Optovac has another ingot in a different furnace that was started a few weeks after the first, that will be a back-up piece should anything happen to the first piece.

 

We have decided to use a back-up blank to remake Element 4, which had a curve error that is too large to correct back to the full design clear aperture. We will review the set-up for the generation of the new lens and begin generation as time permits.

 

Currently 14 of the 20 optical surfaces in the camera are finished. This includes the dewar window. The six surfaces left to be done are the aspherical surface on Element 1, both sides of Element 3, both sides of Element 4, and one side of Element 6. Two of those surfaces are to be done last as pick-up surfaces after the design has been optimized using the as-built lens parameters.

 

    1. Mechanical Design

Structure

 

The electronics enclosure ring has been installed. The insulation materials and the final aluminum skin for the electronics ring as well as the rest of the DEIMOS cylinder have been received. The fasteners for the electronics enclosure have been detailed, ordered and received.

 

The permanent solution for the drive motor surging has been designed and fabricated. A Renishaw linear encoder has been chosen for the on-axis encoder. We are exploring adjusting the gap at the tape ends to complete the circle without the complexity of a second read head.

Slit mask system

 

The cassette actuator, cassette and insertion air cylinder have been installed, and initial manual insertion tests were performed. Modifications to correct the initial problems were made and tests were conducted with the cylinder air driven. Preliminary designs for slit mask alignment tooling have been done.

 

Collimator

 

The collimator cell has been fitted and installed temporarily into the slides on the cylinder.

 

Camera

 

The camera cell design is in progress with Allen Schier. A design review has been scheduled for November 14th. Selection of the couplant has not been made pending analysis of chemical reactivity with the RTV lens potting material. The shutter air cylinders will be replaced with a newer cushioned type that was successfully used on the Primary Focus Camera.

 

CCDs

 

A glue fixture for the AlN flexure compensation parts was designed and fabricated.

 

Gratings

 

The grating system is nearing completion. We are doing a trial fit of the main beam and slide elements. The drive system is being assembled as well. The design for the slide and drive system for the 200-pound lead counter-weight is almost finished, and a few parts have been fabricated. Mahr gage mounts have been included in this support system for further flexure and repeatability testing. These tests have not been done yet.

 

Front End

 

The TV support hardware has been modified to clear the Photometrics cooling lines. The front nose has been modified and new parts designed for the front hatch doors. These will be pneumatic-operated. All the parts are here for the front end.

 

2.3 Detectors

Lincoln CCDs

 

Lick has been testing most of the Lincoln CCDs, as they have become available.

Given the rate of production we expect to hold the first round of CCD distributions to consortium members, near the end of the year.

 

Lincoln CCDs have generally very good performance. They have very low readout noise, excellent charge transfer efficiency, and the standard epi devices have very few hot or blocked columns. The high resistivity devices have significantly more hot columns. One of the most significant problems for all of the Lincoln devices is the "brick wall" pattern, which is a QE variation over the device caused by incomplete backside treatment. Lincoln has some sort of process problem with the boron-implant and laser-anneal, which they use on the backside of the CCD. This treatment is intended to eliminate the need for doing an UV-flood to achieve good quantum efficiency. Their current process leaves large QE variations, which are both color and temperature sensitive. Lincoln is planning to study the problem in the hope of improving the process and reducing or eliminating the QE variations. Check the web pages for detailed results.

 

Detailed results of the Lincoln CCD testing can be found at the following web site: http://gardiner.ucolick.org/~ccdev/lincoln/lincoln.html.

 

Lick/Orbit CCDs

 

In case an interim mosaic is needed, we have moved forward with fabrication of AIN substrates for Orbit frontside 2Kx4K CCDs. This fabrication step should be completed before the end of 1997. Aluminum nitride pieces for the flexure control CCDs have been designed and are on order.

 

The Lick thinning effort on the Orbit CCDs has produced the first working thinned CCD, a 2Kx4K device. Considerable work on the backside surface remains to be done to turn this CCD into a science-grade detector. However, the prospects for supplying at least some thinned devices for the interim mosaic have improved. We will process several more wafers to assess yields before we decide whether or not to thin any of the reserve interim mosaic frontside 2Kx4K CCDs. An image from this first Lick-thinned CCD can be found on our web page: http://gardiner.ucolick.org:80/~ccdev/lick-thin/lick-thin.html.

 

2.4 Software

Note: Items, which occurred during the first week of October (i.e., shortly after the September 30 end date for Quarterly Report 13) are enclosed in square brackets.

 

  1. Completed the formal response to DEIMOS Software CDR review board report. Response submitted September 30th. As part of that response, we have scheduled a DEIMOS GUI review to be held in Santa Cruz on December 16. (See response to review board report for more details, attached as an addendum.)
  2. With the software CDR completed, our efforts have switched from design and documentation to code implementation and test.
  3. Significant progress has been made in the following areas:

  1. User interface development tool (Dashboard): Clarke has succeeded in implementing all of the core functions required for the DEIMOS GUI. [These are now operational and included in Version 2.0, which was released at the start of October.]
  2. Using the features now available in Dashboard version 2.0, Phillips has started work on the DEIMOS GUI prototype and has completed the layout for approximately 60% of the main control window. This prototype will be operational and will be demonstrated at the GUI review on December 16.
  3. Tucker and Burrous have finished implementing approximately 80% of the low-level motor control software needed for operating the Galil motor controllers for DEIMOS. This software (the Galil controller firmware, the dispatcher process, and the keyword library) has been successfully tested using the Lick PFCAM instrument, which uses identical Galil hardware.
  4. We have been using the PFCAM instrument at Lick as an operational prototype for simulating both the high-level (GUI) and low-level software for control of DEIMOS stages (e.g., filter wheels, linear slides, etc.). [We have now successfully operated PFCAM using the version 2.0 Dashboard, along with Galil firmware, dispatcher software, and keyword library.] Much of this software was built using code generated automatically from the keyword database. This code generation capability will significantly reduce the time required to build the corresponding modules for DEIMOS and ESI.
  5. Allen has made significant progress on software infrastructure and image display software and has released version 1.0 of the code generation software used in item "d" above. A portable version of Figdisp is now under CVS and will be used for our initial CCD readout tests, which will commence this fall. Allen is also scheduled to visit NOAO on October 21 to check on progress of the NOAO mosaic image display server.
  6. Kibrick has started work on extending and improving the existing HIRES/LRIS CCD image readout and display software for use with DEIMOS and ESI, including definition of extensions to the Figdisp image display protocol to improve performance and reliability. [Working prototypes to support simultaneous image display and recording of CCD images to multiple sites (local and remote) were successfully tested in early October.]

  1. Sun Microsystems donated the DEIMOS instrument computer, an Ultra Enterprise 450, saving $35,000 in the computer hardware budget.

 

2.5 Electronics

Nearly all the time this past quarter was spent designing and doing the circuit board layout for the various Lick printed circuit boards that support the SDSU-2 CCD Controller. There are a total of eight different boards, many with active circuitry that perform many different functions and allow monitoring of signals to the CCD. The architecture of the boards and their cabling allows for true modularity of the CCD controller system. The system can be used for controlling both single CCDs o, in DEIMOS case, CCD mosaics. The boards and their associated cabling for the SDSU-2 system are expected to be completed by mid-October.

 

The building of the Flexure Compensation CCD Controller box is well under way with most of the wiring done, but we are waiting for completion of its internal boards by Leach. The 24-input analog boards are being assembled and tested. The first units seem to work as expected. The shutter control boxes have been upgraded to allow better feedback to the software by passing back the limit switch info and buffering the outputs.

 

2.6 Flexure Compensation

Aside from the FC electronics described above, no progress was made on the FC system this quarter

 

2.7 Alignment

Work on the alignment plan was postponed during this quarter to allow work on more urgent activities, such as the camera CDR.

 

 

3. Report from the PI:

 

Activities by the PI are included under other sections of this report.

 

 

4. Budget:

 

To the end of the quarter (September 30, 1997) we expended $3,600,000 of the project funds or about 72% of the budget. Of this amount $1,798,400 was spent on labor and $1,796,300 on materials/supplies. During the quarter we spent $219,000 on labor and $433,845 on materials/supplies as is shown on Table 1. The details of the budget are shown on Table 2.

 

Major expenditures during the quarter were as follows:

The cumulative expenses, and manpower usage are graphed in Figures 1 & 2.

 

For the start of the 14th quarter we revised the budget for various activities. The detail is shown as Table 5. In general terms we are planning to remove approximately $250,000 from the contingency, lowering its balance to approximately $78,000, and reducing the estimated manpower for software by 4,300 hours, to 13,900 hours. We plan to request additional funding for software development in the post-commissioning phase, once the tasks and effort required are better defined. The focus of these tasks are expected to be on data handling/reduction features.

 

Sun Microsystems donated three computers to the DEEP/DEIMOS project through the UC Berkeley Campus, and one of those computers is planned to be used as the DEIMOS data computer. This has allowed us to remove $35,000 from the computer hardware budget, as reflected in Table 3.

 

The funds have been redirected to cover emerging shortfalls in Mechanical and Optical budgets. Mechanical overruns are due principally to outsourcing the camera mechanical design to Alan Schier, and to overruns in the dewar, frame and skin (the L&F contract cost more), gratings, and rotation drive (whose encoders had to be reworked). Optics changes are running higher owing to extra labor charges for lens fabrication, to additional work at ORA, and to the fact that we were unable to get a free front window but had to buy one from Kodak.

 

Tables 4-6 show details of the budget revisions broken down into manpower and expenses. The percentages shown are the amounts consumed in each category at the end of the present quarter and reflect Budget Revision 9. At $78,000, the contingency stands at about 5% of remaining project funds.

 

 

5. Schedule:

 

Figure 3 shows a summary of the current schedule. The projected date of November 1, 1998 remains unchanged from the last report.

 

The time critical path is shown as Figure 4. The minor delays that have developed in the growing and generation of the CaF2 boule for Element 3 of the camera, has resulted in camera fabrication's being back on the critical path. The development of the detector system remains very close to the critical path with only a couple of week's float.

 

Conflicts with the ESI development could cause delay in the DEIMOS schedule, particularly in the instrument fabrication and testing, and in software. We are working to mitigate any conflicts in their schedules to the extent possible.

 

 

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

 

Milestones for the previous quarter:

 

  1. Install Renishaw encoder and complete PA rotation. The second version of the design is done, fabrication and testing expected to be done in January.
  2. Complete the slitmask system and begin testing. The slit mask system is not complete. However a slit mask has been deployed into the focal plane with the system. We plan to have the cassette powered by mid January.
  3. Complete the design of the grating system with the goal of having a fabricated system by mid-October. We now expect to have a working system to start testing by March.
  4. Install the collimator in its cell. This has been delayed until February to allow development of the grating and slit mask systems to mature to the point they were expected to be at this time and to accommodate other priorities in the Instrument Lab.
  5. Install the electronics enclosure. The frame for the electronics enclosure is now attached.
  6. Complete the design of the dewar system and start fabrication. The design is complete and fabrication has begun.
  7. Continue work on the alignment plan for spectrograph and camera. Progress was made on the camera alignment plan in preparation for the camera CDR.
  8. Start the detailed design of the camera barrel in preparation for the camera review in mid-October. Detailed design has begun. The review is now planned for November 14th.
  9. Complete the analysis of Element 8 asphere decenter and compensating schemes. Done
  10. Begin running MITLL CCDs in the test dewar. We expect to start running devices in the test dewar in January.
  11. Complete laboratory testing of the RTV CTE and YoungÆs modulus. Done.
  12. Choose filter thickness and adjust optical design. Done.
  13. Decide on a final plan for coupling Element 6: cement or liquid. Done, liquid.
  14. Complete the FC controller. Done.
  15. Complete the CCD interconnect boards. Done.
  16. Install the attachment points on the tent mirror and aluminize. Delayed to February.
  17. Take a TV image with the Photometrics camera/computer system in the rotating DEIMOS structure. Delayed indefinitely.
  18. Select the camera optical coatings. Delayed pending results from the ESI coatings.
  19. Prepare the camera CDR report. Delayed until November 14th.
  20. Consider the impact of ESI on the DEIMOS schedule. Continuing.
  21. Analyze Keck I and Keck II thermal data. Done.
  22. Select the camera coupling fluid. Candidate fluids were selected that we now must test versus or RTV, O-rings, and vinyl bladder material.
  23. Complete automatic motor-code generation software (CODEGEN) and use to build PFCAM motor software. Done.
  24. Start recruitment for extra Software person. Started.
  25. Prepare response to Software CDR report. Done.
  26. Complete new Software scheduling tool and use to analyze ESI and DEIMOS software schedules. Done.
  27. Complete Dashboard GUI version 2.0 including move-complete notification and test with PFCAM. Done.
  28. Complete Dispatcher version 2.0 and test with PFCAM. Done.
  29. Test DEIMOS slitmask alignment scheme and slitmask astrometry model and verify with LRIS. Done.
  30. Commission new DEIMOS-style CNC mill at Keck. Continuing.
  31. Generate CNC mill instructions directly from mask blueprint (bypass SURFCAM). Prototype done and output verified by mill operator at Lick.
  32. Reconcile Caltech and Lick Figdisp variants and merge. Done.

 

 

Milestones for the next quarter:

 

  1. Continue work on the Renishaw encoder and PA drive.
  2. Complete modifications to slit mask cassette; power cassette; continue testing system.
  3. Continue work on the grating system.
  4. Complete the dewar system and begin testing.
  5. Complete an analysis of the thermal environment of the camera within DEIMOS.
  6. Update the optical quality error budget for the Camera CDR.
  7. Hold the CDR for the Camera Mechanical Design November 14th Prepare CDR documents.
  8. Complete fabrication drawings for the camera following the CDR.
  9. Begin running MITLL CCDs in the test dewar.
  10. Select the camera optical coatings.
  11. Initiate reactivity tests of camera coupling fluids with RTV, O-rings, and vinyl bladders.
  12. Continue recruitment for extra software person.
  13. Complete our involvement in CNC mill commissioning at Keck.
  14. Start insulating and cladding DEIMOS.
  15. Receive Element 3 of the camera.
  16. Complete instrument simulator software.
  17. Hold DEIMOS GUI day with capable GUI and working instrument simulator.
  18. Hold the Lincoln CCD draft.
  19. Complete the set of light utility boards and cabling that connect the Leach-2 mosaic controller with the CCDs.
  20. Complete the FC controller.
  21. Fabricate mounting plates for motor control electronics on DEIMOS.
  22. Fabricate the motor control modules that sit near the motor.