Next: 2 Summary of Changes
Up: Part I: Overview and
Previous: Part I: Overview and
This section provides an overview of the CDR document which consists
of the following five parts:
I. Introduction and Overview
II. Subsystem Designs
III. DEIMOS Calibration
IV. Flexure Compensation
V. Budget and Schedule
Since the document is very long, we list here the sections to which we
would particularly like to direct the review board's attention. We hope
that all board members will have a chance to read these essential sections
prior to the CDR.
Part I, Appendix A: Contains an updated and detailed
response to the report of the DEIMOS Software PDR Review Board. It describes
the specific actions we have taken in response to the board's recommendations
and provides a comparison of projected versus actual progress over the
interval between the PDR and CDR. This appendix also highlights
several critical personnel and management issues which urgently need to be
resolved; see I.A.4.2. A summary of this appendix is provided in
Part I, Chapter 3: Describes those areas of the
DEIMOS system in which we have a solid proof of concept for the underlying
software and/or hardware components. In many cases, these proofs are in
the form of fully functional software tools (operating within a well-defined
software infrastructure) that have been successfully used to complete and
document the software designs presented in this volume.
Part I, Chapter 4: Describes areas of risk for which neither a
detailed design nor proof of concept yet exists. Contingency plans for
the Image Display subsystem are discussed, and an abbreviated set of
functional requirements for the Flexure Compensation and Calibration
subsystems is provided. (A more detailed discussion of the Calibration
and Flexure Compensation subsystems is provided in Parts III and IV,
Part II, Chapter 1: Describes the design of the software and
operating procedures for the slitmask design and fabrication subsystem.
This design has evolved significantly from what was presented at the PDR.
Since the inventory and production control aspects of this subsystem are
ones that we have not faced in previous instruments, these aspects should
be afforded careful review.
Part II, Chapter 7: Describes the information management (database)
subsystem, which is integral both to the operation of the overall instrument
and to the design and documentation of all of the DEIMOS subsystems (including
itself). This chapter addresses various concerns about this subsystem that
were raised at the PDR, including the concern that the database component
per se will delay the project schedule or consume more than
its share of resources. This chapter illustrates that the database subsystem
now has a more limited and well-defined scope and that so
far it appears to be ahead of the other major instrument subsystems:
the design of this subsystem is nearly complete, the implementation of most
of its components is quite advanced, and many of its components are already
operational. The various documentation products that are automatically
generated by this subsystem offer solid proof of the utility of its design
and its cost effectiveness.
Part V: Describes the DEIMOS software budget and schedule.
This part was not ready in time for the pre-CDR distribution
of documents, but will be presented at the CDR. A significant block
of time is being allocated for its presentation and subsequent discussion
at the CDR.
The remaining portions of this CDR document are less important. In particular,
parts III and IV consist of supplementary material:
Part III is intended to become a separate document
on DEIMOS Calibration,
but it is not yet complete. Its contents are summarized in section I.4.3
and it is included in draft form for reference only. It defines
the procedures and mathematical relationships required
for the calibration of DEIMOS data but does not specifically address
the detailed design of the calibration software.
Accordingly, Part III is a preliminary
step in defining the functional requirements for that subsystem.
(Since implementation of pipeline data reduction is a goal rather than
a commitment, it will likely not occur until much closer to or even after
instrument commissioning. Recall that the DEIMOS software effort is
budgeted to continue for 2 years beyond first delivery.)
A complete detailed software design of this subsystem will not commence
until later in the project. However, the material in Part III indicates
how well the requirements of this subsystem are understood, and the
magnitude of the design and implementation task which remains.
An understanding of these
points is needed to assess the budget and schedule impacts of implementing
pipeline data reduction.
Part IV is intended to become a separate document on the Flexure
Compensation (FC) Subsystem, but it also is not yet complete. Its contents
are summarized in section I.4.2 and it is included in draft form
as part of the CDR document to serve as a preliminary set of functional
requirements for this subsystem. However,
the FC subsystem is a required part of the instrument,
so the detailed design of its software must proceed now.
To date, the overall FC system (optical, mechanical, electronic, and software)
has not been subject to any detailed independent review. This software CDR
represented the first opportunity to review much of the material contained
in Part IV, and while the completed portions of this FC document do not
specifically focus on software, they do indicate the level to which the
requirements of this subsystem are understood, and the magnitude of the
design and implementation task which remains. We recommend that the
overall FC system be reviewed at a later date.
In addition to the five parts of this CDR document, we have also produced a
preliminary draft of Lick Observatory Technical Report ???: The DEIMOS
Keyword and Database Dictionary. This report (commonly referred to as
LOTR ??? or ``The Dictionary") provides a dictionary defining all of the
DEIMOS keywords and database tables. Example pages from this dictionary
are included in section I.3.4 of this document. However, due to its large
size, we are not including LOTR ??? as part of the pre-CDR distribution of
documents to the review board. Copies of LOTR ??? will be available for
inspection at the CDR. An on-line copy of LOTR ??? will be available on or
before June 13 via the WWW at the following URL:
For those reviewing this dictionary, remember that it is automatically
generated from the data contained within the various tables of the DEIMOS
database, using the document generation tools which are described
in section I.7.1.1. Not all data are complete, but the dictionary
currently represents approximately 80% of its finished equivalent.
This section provides a global overview of the DEIMOS software system.
It includes two diagrams showing all of the major subsystems and their
interconnectivity (see Figures 1.1 and 1.2).
The DEIMOS system consists of the following major software subsystems,
which are described in the indicated sections of this report.
Mask Design: Part II, Chapter 1
Mast Fabrication: Part II, Chapter 1
DataBase: Part II, Chapter 7
Mask Management: Part II, Chapter 1, Section 5
Spectrograph Control: Part II, Chapter 3
Header Collection: Part II, Chapter 4 (available at CDR)
Science Data Capture: Part II, Chapter 4 (available at CDR)
Image Display: Part II, Chapter 6
STB: Existing system at Keck
Pipeline Reduction: Part III
Mask Alignment: Part II, Chapter 2
Detector Control: Part II, Chapter 4 (available at CDR)
DCS and Guider: Existing system at Keck
Rotator Control: follows model of HIRES rotator (existing system)
Flexure Compensation: Part IV
Documentation Facility: Part II, Chapter 7
Calibrate Spectrograph: Part III
Logging Facility (KTLwatch): Part II, Chapter 8
Engineering Data Visualization: Part I, Chapter 3, Section 2
KTL User Tools: Part I, Chapter 3, Section 2, plus existing KTL UIs
Figure 1.1: Overall system data flows
Figure 1.2: KTL User Tools data flows
The basic optical properties and capabilities of DEIMOS were
described in Chapter 2 of the
Software PDR Report
presented on March 22, 1996.
We list here the significant changes that have occurred
since then that are relevant to software development.
The final choice of CCDs has not been made, but the
probable properties are better known, as discussed at the
recent Detector/Controller CDR held on May 20, 1997.
Copies of that report are available. A summary is given here
in Part I, Chapter 2.7.
are now flat sheets of aluminum as carried
in the slitmask cassette holder and are deformed to the curvature
of the focal plane when inserted. The curved rigid holders have
been eliminated. As a result we can now carry 12 slitmasks plus
one longslit, an increase of 3 slots. Mounting and handling will
be greatly simplified.
The TV CCD
is a Photometrics 1K 1K SITe chip with
24 pixels, a scale of 02 px , and a total FOV of
33. The Photometrics camera looks at the telescope focal plane
in staring mode. The visible focal plane
is divided into two parts. One area is covered by a curved
mirror permanently mounted in the telescope focal plane beside
the slitmask area. This mirror is always visible regardless of
whether the slitmask is in or out and provides offset guider
capability when objects are not visible on the slits (e.g., during
direct imaging). The dimensions of the offset guider area are
. The remaining area of the TV FOV images
either the cylindrical slitmask or a spherical mirror with a long
slit, depending on which is inserted.
The long slit is located 45 off axis from the center of the telescope
focal plane. The dimensions of the
imaged slit area are .
The entire long slit area is unvignetted in the TV by virtue of
its spherically curved mirror. Because the slitmask is cylindrical
rather than spherical, objects will be visible only in a strip approximately
30 wide surrounding the longslit. However, the top surface
of the masks is reflective,
and it should be possible to place objects directly into
the slitlets using the TV.
The grating slide
still has slots for the imaging flat mirror
plus three gratings. However, there is now room for only
one -in grating (in Position 1), as Position
3 has been trimmed to accept only -in gratings to
simplfy its encoder design.
The option of tilting all grating positions to the flat
mirror imaging angle has been retained, as it
appears likely that notch mirrors for narrow-band imaging can
be fabricated, giving DEIMOS a narrowband imaging capability
that is otherwise not possible. Extra interlock hardware and software
is needed to prevent collisions when grating slots are tilted
to the mirror positions. ??True?
The stated image stability goal of 0.25 px rms in
each coordinate may not be achieved despite the Flexure Compensation
system owing to thermal changes in the camera plate scale. These may cause
image motions (contraction, expansion) of a few
pixels at the edge of the detector
if not successfully compensated passively in the thermal
design of the camera lens barrel. Residual motions of
this magnitude would have adverse implications
for flat-field accuracy and wavelength stability, as described
in Part I, Chapter 4.3
and Part III of this report.
The slitmask milling machine has been received and
debugged. It operates successfully.
The SDSU-2 controller has been received and tested.
It meets our needs, as described in the recent Detector/Controller
CDR, held on May 20, 1997. Copies of that report
A barcode reader has been
added inside the spectrograph
to read the IDs of slitmasks after the masks are loaded.
These IDs are checked against plan and then loaded into the
instrument control program and database.
Remote observing from
Keck Headquarters at Waimea has
been demonstrated using LRIS in multislit mode and is highly efficient,
including slitmask alignment. Our default plan for DEIMOS after
instrument commissioning is to observe remotely from Headquarters.
Next: 2 Summary of Changes
Up: Part I: Overview and
Previous: Part I: Overview and
DEIMOS Software Team <firstname.lastname@example.org>