DEIMOS
Quarterly Report
Number 16
April 1 to June 30, 1998
Lens Element 8 was fractured at the Coherent Inc. facility while it was being cleaned in preparation for coating. Element 8 is aspheric and is part of a fluid filled doublet that comprises Group 4 of the camera. We have two remaining blanks for this element and have started to re-fabricate it. David Hilyard is expected to complete the new Element 8 by the end of November. Because of the fracture, the existing Element 8 has a large scallop out of the rear surface and one edge. We are planning to fill the edge with RTV and use the element in the camera for testing purposes.
All the other camera elements that need coating are at Coherent.
We received the window for the spectrograph from Kodak and are in the process of testing it. Based on tests done by Kodak and preliminary interferometric tests done on an 8-inch aperture at Lick, it meets the required specifications.
We have selected an optical couplant fluid, Cargille Laser Liquid 1074 at an index of 1.5570.
The design of the flat mirror cell and carriage has been completed, and parts are being fabricated. Design of the cells and carriage for the 6x8 gratings is in progress and nearing completion. The slide drive was found to be too noisy and a new design using a quieter roller chain is nearly done.
The dewar assembly is nearing completion. A mosaic of dummy CCD packages has been assembled on the backplane and cooled in the detector housing of the dewar system to a temperature of -120°C. The electronics boxes have been completed and mounted on the detector housing. Work is continuing on the final mechanisms for the dewar system.
The Renishaw encoders have been mounted, and we plan to test the rotational drive system through the next quarter. (Tests were delayed).
The fabrication of the science CCD controller chassis is complete, but checkout of this chassis awaits completion of the external power supply, which is still being wired.
Work continues on reading two CCDs in the test dewar. Major progress was made in the lab, getting the Leach-2 systems running on three different computer systems. This will allow testing of more than one system concurrently.
Delivery of the MIT/LL devices from the first epi lot in which DEIMOS has a share is currently scheduled to begin in early October.
Most software effort continues to be directed towards ESI, but much of this effort also benefits DEIMOS. The DSP software for the Leach-2 VME interface board was completed and tested. The non-mosaic Leach-2 CCD readout software is now operating on the Force 5CE CPUs in the ESI, DEIMOS flexure compensating, and DEIMOS science CCD VME crates. Work continues on shared general-purpose tools for engineering data visualization, GUI development, and collection of telemetry. In June, a prototype system was tested which automatically monitors and archives complete instrument state information. The RAID disk controller subsystem for the DEIMOS instrument computer was specified and ordered in June, with delivery in July. Our recruitment for a Deputy Software Manager/Developer was successfully concluded in April, and our top candidate, Robert Mykland, will join our scientific programming group in early July. Mykland will assist in both the development and management of the ESI and DEIMOS software.
A schedule prepared in early March set July 1 as the date to assemble the first subset of DEIMOS assemblies and begin testing. This did not occur (and still has not, as of September 15). Systems that fall seriously short of their July 1 date include the slit mask system, flexure compensation system, PA rotation system and cable take-up. The grating system, dewar, detector, and electronics stayed approximately on schedule. Analysis of the schedule in mid-July indicated that about half of the observed slippage was attributable to diverted manpower (largely to ESI) and half to budgeting too small a time to complete a task. During this tracked period, we accomplished roughly two-thirds of the intended work per available unit of resources. (Tracking ceased in mid-July, when mechanical work on DEIMOS largely halted due to work on ESI.)
2.1 Optics
The calcium fluoride lens Element 3 was finished on both sides in early May. It was transported to Coherent for coating on both sides and has been returned to the Optical Lab. The second sides of both Elements 4 and 6, reserved as the "pick-up" surfaces in the camera, were finished to the original specifications and transported to Coherent for coating. One side of Element 9 was repolished to remove the coating, which had a higher reflectivity than desired and then returned to Coherent for recoating on that side. The first surface of Element 1 was also coated in the same, flawed coating run. However, we elected to accept it since removing it would require re-polishing the aspheric surface. The average reflectivity is 3% rather than the hoped-for 1%. The reflected light will go back into the body of the spectrograph, where it will do little harm apart from the photons lost.
In early June, Coherent reported that they had fractured Element 8 while loading it into a coating fixture. We subsequently decided to use the fractured piece only as a test piece in the camera assembly and we have started fabricating a replacement (the extra blank is in stock). At the end of June, the new Element 8 had been generated, ground and finished polished on the spherical side and was ready for plunge grinding of the asphere on the second side.
The optics that are coated and returned to Lick include Elements 1, 2, 3 and the dewar window. Optics at Coherent for coating include Elements 4, 6, 7, and 9 (these have since been successfully finished). Element 5 is uncoated, and Element 8 is being re-made.
2.2 Mechanical Design
Dewar System
The Dewar system has been cold tested, and the mosaic of packages, some with chips attached, have been cooled to -120°C. Considerable work remains to fabricate and test the focus mechanism and the X-motion drive for the flexure compensation system.
Nasmyth mountingKeck has modified the design for the rail system on the Nasmyth platform. They are now planning to use the Nasmyth deck to switch the instrument positions, eliminating the turntable. Details remain to be worked out, but the concept is complete. Keck personnel are also working on the umbilical handling system. The design for reinforcing the Nasmyth platform for the kinematic mounts is also complete.
Cable Wrap
The original cable wrap design proved too short. A new design has been made, and a new cable carrier has been purchased. The fabrication drawings are being prepared. Most of the components built for the earlier design can be reused.
Grating System
Design of slider 2 (the flat mirror) is finished, and fabrication is approximately 75% done. Design of sliders 3 and 4 (the 6 x 8 gratings) is almost finished, but fabrication has not yet started. The slide drive was tested with ANSI #35 chain and sprockets and was found to be too noisy; a re-design with British Standard sprockets and Tsubaki roller chain was done. Fabrication has not started, but the new chain has arrived. This chain has rollers whereas the #35 chain had none. Further testing will await NC cutting of the new sprocket tooth form on the six sprockets and fabrication of the two chain-guard assemblies. Initial tests will be with oil, but we plan to ultimately use a dry moly-lube.
Slit Mask System
The
cassette actuator has excessive flexure about the Y-axis, and a fairly simple
redesign is in progress. A final check was made on vignetting by the slit mask
form in the critical form/grating collision zone, necessitating removal of some
material from the form. This will be incorporated into the final form when
it is fabricated.
Rotation Drive
The Renishaw encoders have been mounted, and preliminary tests have been conducted with encouraging results. The rotation drive was re-installed with no special alignment procedures and found to surge. It has been modified to facilitate alignment during assembly, re-assembled, and now operates within specification. New grease and a seal have been ordered. Further tests of the assembly procedure are planned when these components are ready for installation.
2.3 Detectors
Lincoln CCDs
We have received and tested several more CCID20s from the Consortium Phase-2 effort. The results are posted on the CCD lab web page:
http://gardiner2.ucolick.org/~ccdev/lincoln
Some of the Phase-2 CCDs were completed using the improved backside treatment process. This process reduces the "brick wall" QE variations by about a factor of two at 4000 A, compared to the old process used on all Phase-1 CCDs.
We expect to receive and test as many as six more Phase-2 CCID20 CCDs before the end of August. The Keck share of the Phase-2 effort was funded from the DEIMOS budget, so the devices we get from this work will be potential DEIMOS final mosaic CCDs. Based on present yields, we will get at least one and probably two high-grade devices, and at least one lower-grade device (this did occur).
The first lot of 150-mm epi wafers for Phase-3 (DEIMOS final devices) is out of fabrication and is undergoing backside processing. The latest schedule from Lincoln predicts that we will begin to see some of these CCDs by early September (now expected early October). The other two partners in this run (Subaru and the University of Washington) have agreed that Lick will test all of the CCDs from this run in order to have a uniform basis for comparison.
Lick/Orbit Interim Mosaic CCDs
Eight interim-mosaic Orbit CCDs have been packaged and tested. Two of the CCDs will be used in the test dewar to verify plans for locating the temperature-sensor diodes and the temperature control heater resistor and to test for cross-talk and other noise sources in a multi-CCD dewar. These two CCDs, along with the other six interim-mosaic CCDs, will then be assembled into the DEIMOS mosaic (to start October 15).
Lick/Orbit Flexure Control CCDs
The candidate FC CCDs were re-tested on the probe station to select the best devices for the DEIMOS detector arrays. Five devices were selected, two for each camera and a spare. Several test-grade devices were also identified. The aluminum nitride packages have been assembled. Attachment of the printed circuit board and connector will be done soon. For the first completed assembly we will use a test-grade FC CCD. If imaging tests on this device reveal no problems, the remaining CCDs can be finished within a few weeks.
2.4 Software
Hardware upgrades to the DEIMOS instrument computer
DEIMOS received a donation of 24 9GB disk drives from the Quantum Corporation, valued at $24,000. These drives will be used in a RAID disk controller to provide reliable and high-bandwidth data storage for DEIMOS images. The configuration for this controller was defined and an order placed with ISS Corporation, with delivery scheduled for July. Five 4GB internal hot-swap disk drives have also been ordered, along with 2 GB of RAM.
2nd-generation CCD VME crates now operational
In Quarter 15, we conducted our first tests of the second-generation SDSU (SDSU-2) CCD controller and successfully read out both Lick-Orbit and MIT-LL 2K by 4K CCDs. A spare first-generation CCD VME crate (identical to the crate currently used in LRIS) was used for those tests since the second-generation VME interface was not yet operational. During the present quarter, software for the second-generation VME interface board was completed, the VME crate CPU upgraded from a Sparc-1E CPU to a Force-5CE CPU, and the OS in the VME crate upgraded from VxWorks version 5.0.2 to version 5.2. This is the hardware and software configuration that will be shipped with ESI and DEIMOS.
Instrument telemetry and monitoring
A system was designed and implemented which automatically collects and archives (at regular intervals) all of the keywords from an instrument. This system, in conjunction with the engineering data visualization tools developed during Quarter 15, will be used for pre-ship testing, commissioning, and ongoing maintenance of ESI and DEIMOS. A prototype version of this system was tested in June using HIRES.
Impacts of ESI
ESI consumed a larger fraction of software resources during the quarter, and this trend will continue through Quarter 18. Our current plan is for the DEIMOS software effort to continue at a low level through the end of Quarter 18, and to ramp back up in Quarter 19 (January 1999).
Major expenditures for the quarter were for upgrades to the DEIMOS instrument computer and the RAID array.
2.5 Electronics
Considerable progress was made on the Science CCD Controller and power supply boxes. The power supply box has been finished and awaits completion of the CCD controller for testing. The CCD controller has had its front-to-rear-panel ribbon cables installed but is waiting on the power wiring. Some progress was made on the Galil motor programming. The SCSI-to-fiber boxes have been received and tested.
The four revised boards that comprise the SDSU-2 CCD controller system were received. All of the SDSU-2 support boards have been assembled and individually tested. The CCD preamplifiers were tested with their gain set for the Lick/Orbit CCD used in the interim mosaic. The three electronics boxes for both the science CCD controller and the flexure compensation CCD controller have been wired and tested with their accompanying boards for both continuity and functionality.
2.6 Flexure Compensation
The decision as of early March was to fabricate the bare minimum components for the FC system and leave the details and software until later if the system proved needed. The bare minimum system includes the X and Y actuators, the FC CCDs and signal chains, and the fiber feeds into the focal plane. Also needed for testing is an interim emission-line lamp, which need not cover the entire spectral range from 4500 A to 9500 A, as required for the final lamps. Based on successful tests by ESI, we adopted a Cu-Fe hollow cathode lamp for the broad-spectral FC source (we will also use the same type of lamp as the main wavelength calibration lamp for DEIMOS). We conducted brief laboratory tests to show that a simple Penray Ne lamp will serve as an interim light source; simply aiming one end of the fiber at the discharge tube at a distance of about 1-cm provides plenty of light. Design started but was not completed on the optical system for feeding the f/15 fiber beams into the focal plane.
The FC signal electronics are complete, and the FC CCDs are being packaged. The X and Y actuators have been largely fabricated but remain to be tested,
2.7 Alignment
We have located and made initial tests of a microscope for measuring the planarity of the assembled CCD mosaic. Use of a microscope rather than an interferometric or laser system will save about $4K in capital costs. The rms height error of the microscope system is 0.6 m, and the distance between the objective and the CCD is a safe 20mm. The surface of a front-side-illuminated Orbit CCD was readily focused. The current x-y stage has a z-height variation of 34 m over the x-y range of interest, which is too large for our purposes. We are designing a mosaic enclosure and support that will also serve as a translation stage.
We have followed in detail the initial optical measurements of the assembled ESI camera. We continue to follow their tests and document a procedure to be used for the assembled DEIMOS camera. This document will include the additional fixturing needed for the DEIMOS tests.
We have completed a detailed procedure for the assembly of Group 1 of the camera. The procedure has been reviewed by all parties involved (some through several iterations). The fixturing required for the assembly has been outlined.
A sample testing protocol was developed using the grating system as a prototype. Twenty-four performance tests were described, the great majority of which lend themselves to automated scripts and data recovery. Using this input, the software team started work on a web-based test interface tool that will allow testers to write scripts quickly, run them, and plot the output. A small number of these scripts will be refined eventually and delivered to CARA for regular monitoring and maintenance of the instrument.
[Tables and figures are not available via the web. Please contact Heather (heather@ucolick.org) for more information]
The project budget and spending are summarized in Table 1. Details are shown in Tables 2, 3 and 4. At the end of the quarter we have spent approximately $4.65 million on the project, or about 93% of the project budget.
Table 3 summarizes the expenditures of manpower. Approximately 16,400 man-hours of effort remain in the budget. These expenditures are graphed in Figure 1.
Expenditures on materials and supplies are summarized in Table 4 and are shown graphically in Figure 2. Approximately $240,000 for materials and supplies remains in the budget.
Major expenditures for the quarter were:
· Materials and rough machining of parts for the camera barrel.
· Parts for the CCD controllers.
· The remainder of the cost of optical coating of the camera elements.
· Raid Disk Controller.
· Ram for the instrument computer.
· Hard disks.
The
summary schedule as it stood on June 30 is shown in Figure 3. The critical
path of the project, shown in Figure 4, runs through the mechanical fabrication
of the camera and allows for 22 weeks of testing of the instrument. Most of
the effort in the shops has been directed towards ESI in the last months and
has resulted in a delay to DEIMOS. The current schedule shows a pre-ship review
in May 1999, which is approximately a two-month slip from the last report.
(Later estimates show a slip of six months and significant additional cost as
well.)
The following is a list of milestones for the quarter from the last Quarterly Report, together with the progress made on them:
1. Complete fabrication of the camera optics. Optics fabrication was completed; however Element 8 was fractured in the coating proces and needs to be re-fabricated.
2. Coat elements 1, 2, 3, 4, 5, 7, 8, 9 and the dewar window. Elements 1, 2, and 9 were coated. The others (except for 8) will be coated next quarter (Done).
3. Complete all parts for the camera barrel. All the major parts have been rough machined.
4. Build a mosaic of Orbit chips. Partial progress: two mosaic packages were fabricated, and jigging exists for the mosaic assembly process. We are currently waiting for two devices to be read out in the test dewar to start mounting Orbit packages on the backplane.
5. Install the Orbit mosaic of chips into the science dewar. Not started.
6. Complete the science CCD controller. The science controller was not completed by the end of the quarter. (Fabrication is now complete, and testing has started).
7. Start electrical tests of the science dewar. The dewar electronics boxes are complete and have been mounted on the dewar system. (Testing will start in October).
8. Read out two Orbit chips in the test dewar. One chip has been mounted and tested; the second will be mounted in October.
9. Complete and test the slit mask system. Further modifications to the slit mask system were required and have been designed. We are waiting for the priority ESI tasks to be completed before the modifications can be fabricated.
10. Complete the grating system. We have completed fabrication of the 8x12 slider. Minor modifications are required for the drive system. The mirror and remaining slides are in various stages of fabrication.
11. Complete testing of the Renishaw encoder on the PA rotational stage. Installation is complete; testing is delayed by ESI and is scheduled for January 1999.
12. Balance DEIMOS and continue testing of rotational drive. DEIMOS is balanced and has been rotated under computer control. The rotational drive has been reassembled but needs to be regreased. Testing is now scheduled for January.
13. Glue mounting hardware on the tent mirror. Delayed.
14. Order the filters. Delayed.
15. Select the optical couplant based on reactivity tests. Done.
16. Select the calibration lamp sources. Done.
17. Select the light source for the FC system. Done.
18. Finish the ghost pupil study. Delayed.
19. Produce an assembly and alignment plan for the camera optics.
20. Complete the design of the filter wheel. Final design awaits completion of a prototype and is now scheduled for December.
21. Complete the cable wrap. Little progress. (Design now done and awaiting fabrication.).
22. Specify and order RAID disk controller for the DEIMOS instrument computer. Done.
23. Begin work on converting HIRES-specific Image Rotator software into a generic image/instrument rotator task. Test using the NIRSPEC rotator. Done.
24. Implement and test waveforms for MIT/LL 2Kx4K CCD, and read out a single MIT/LL CCD using ESI dewar. Done.
25. Implement and test mosaic descrambling routines and waveforms, using these to read out a 2x1 mosaic of front-side illuminated 2Kx4K Orbit CCDs mounted in the new test dewar. Pending two chips in the test dewar; expected October.
26. Use engineering data visualization tool to analyze servo performance of various stages. Done.
Milestones for the next quarter:
1. Receive all but lens Element 8 back from Coherent.
2. Complete design drawings for the grating system.
3. Complete the design of the camera support.
4. Read out two Orbit chips in the test dewar.
5. Complete a detailed camera assembly plan.
6. Observe ESI camera testing to learn how we will test ours.
7. Complete the science CCD controller and all cables.
8. Mount an FC CCD on a test package.
9. Develop a final plan and the fixturing for measuring the planarity of the CCD mosaic using a microscope.
10. Rationalize the ESI and DEIMOS TV systems and convey the similarities and differences to CARA.
11. Install 5 x 4GB internal hot-swap drives and configure with Veritas software as an internal RAID partition.