MILLIMETER ARRAY
PROGRAM PLAN
DESIGN AND DEVELOPMENT VOLUME II
JANUARY 1998
ASSOCIATED UNIVERSITIES, INC.
NATIONAL RADIO ASTRONOMY OBSERVATORY
TABLE OF CONTENTS
I. INTRODUCTION
II. OVERVIEW OF THE MILLIMETER ARRAY DESIGN AND DEVELOPMENT PLAN
2.1 Project Requirements 2.2 Science Requirements for the MMA Design and Development Phase 2.3 NSF Expectations for the MMA Design and Development Phase 2.4 NRAO Principles for the MMA Design and Development Phase 2.5 Project Goals
III. ORGANIZATION OF THE MMA PROJECT
IV. INSTRUMENTATION DEVELOPMENT FOR THE MMA
Table 4.1 Overview and Comparison of MMA Instrumentation Goals 4.1 Antenna Table 4.2 Antenna Specifications 4.2 System 4.3 Receivers 4.4 Local Oscillator 4.5 Correlator 4.6 Computing
V. IMPLICATIONS OF A REMOTE SITE FOR THE MMA: ROLE OF THE TEST INTERFEROMETER
Table 6.1 D&D Gantt Chart Table 6.2 Milestones and Deliverables by Task Table 6.3 Milestones and Deliverables: Chronological 6.4 Personnel Projections 6.5 Budget Table 6.4 Personnel Projections by Calendar Year Table 6.5 Personnel Projections by Task: Calendar Year 1999 Table 6.6 MMA Design and Development: Total Cost (Restricted Use Only - MMA Personnel)
VII. DESIGN AND DEVELOPMENT IN THE MMA PROJECT CONTEXT
7.1 Operations 7.2 Construction: Role of MMA Partnerships Table 7.1 Project OverviewGantt Chart
APPENDIX A T. J. Cornwell, M. A. Holdaway, and J. M. Uson "Radio Interferometric Imaging of Very Large Objects: Implications for Array Design"
APPENDIX B AUI/NRAO Organization Chart (not yet available on www)
APPENDIX C MMA Organization Chart (in PDF Format)
APPENDIX D Millimeter Array Memos
APPENDIX E MMA Memo 145: Antennas for the Millimeter Wave Array
APPENDIX F MMA Memo 190: A System Design for the MMA
APPENDIX G MMA Memo 151: Design of Planar Image Separating and Balanced Mixers
APPENDIX H MMA Memo 166: The MMA Correlator
APPENDIX I MMA Memo 164: MMA Computing Working Group Report
I. INTRODUCTION
The Millimeter Array is a revolutionary instrument, a direct result of
the revolution in astronomy
achieved in the twentieth century. While the present century will surely
be remembered for the
discovery that the universe was evolving from a discrete beginning, we
have been slow to
appreciate the corollary requirement that everything in the
universe also must be evolving from its
own beginning. This simple conclusion is now leading to a profound
reassessment of the
priorities for instruments necessary for astronomical research. Since
cosmic sources evolve from
matter that is cold to matter that is hot, if we hope to observe cosmic
objects as they form, tools
are needed to permit astronomers to see cold matter in the universe with
the same clarity of detail
that the Hubble Space Telescope, for example, allows us to see warm
matter. The Millimeter
Array is that tool.
The purpose of the Millimeter Array Program Plan, Design and
Development Volume II, is to
outline the steps to be taken to make the Millimeter Array a reality. The
overall project is divided
into three overlapping and interdependent stages: Design and Development,
Construction, and
Operations. This volume addresses the initial phase, a three-year design
and development phase
that leads to prototype hardware from which performance and cost
parameters for the array may
be assessed. It is the second synopsis of the Millimeter Array Design and
Development Plan; the
first was presented by Associated Universities, Inc., to the National
Science Foundation at their
request in September 1992. The present volume supplements and supersedes
the former in all
respects.
The Millimeter Array Design and Development Plan, outlined in the
sections below, includes a
description of the requirements for the design and prototyping activities
and the manner in which
the tasks defined to meet these requirements will be carried out. A
complete work breakdown
structure along with staffing and budgetary tables is included. Finally,
a skeleton outline is given
so as to place the design and development activities in the larger context
of the entire MMA
project.
II. OVERVIEW OF THE MILLIMETER ARRAY
DESIGN AND DEVELOPMENT PLAN
2.1 Project Requirements
The requirements for the Millimeter Array (MMA) Design and Development
(D&D) phase have
come from the astronomers who wish to optimize the MMA for their
scientific research, within
the stewardship of the National Science Foundation (NSF) which has the
ultimate responsibility
for the project, as organized by the National Radio Astronomy Observatory
(NRAO) in its role as
a National Observatory charged with fostering radio astronomy in the
United States. There is
universal agreement on the central requirement for the MMA, namely:
The Millimeter Array will provide sensitive, high-precision
astronomical imaging at
sub-arcsecond, 01, resolution at millimeter and submillimeter
wavelengths.
2.2 Science Requirements for the MMA Design and Development
Phase
The scientific requirements for the MMA have been established and
revised through a series of
MMA science workshops held from 1985 to 1995. The proceedings of these
workshops are
published and have been made available in electronic form on the World
Wide Web. In the MMA
astronomers are seeking an instrument that will allow them to make
spectroscopic images of the
gas, and continuum images of the dust, in normal galaxies such as the
Milky Way very early in the
history of the universe. They want to observe the details of the
formation of individual stars
throughout the Galaxy. They want to measure the isotopic composition of
the ejecta from giant
stars that are progressively shedding nearly all their mass in the final
throes of stellar evolution.
And they want to observe energetic events on the Sun, to establish the
chemical composition of
comets, monitor volcanic outbursts on Io, and assess the abundance of icy
asteroids beyond the
orbits of Jupiter and Neptune. These representative observational
programs, and many others that
could be mentioned, lead to three principal requirements for the MMA:
These requirements lead to some clear specifications for the MMA. In
particular, a large number
of antennas is needed to give the good Fourier (uv-plane) coverage that
produces precision
synthesis imaging; the antenna array must be have a physical extent of
approximately 3 km to
achieve 01 at one its prime millimeter frequencies, 230 GHz; and the
antennas must be
transportable to achieve precision imaging on all spatial scales up to 01.
However, these same
considerations lead to some practical consequences that define the work to
be done in the MMA
D&D phase. For example, the angular size of objects astronomers wish
to image--forming groups
of galaxies, interacting galaxies, regions of star formation--may be
smaller than approximately 3
arcseconds or they may be as large as 3 arcminutes. If astronomers want
to make an image of an
object as large as 3', and they want that object to fit in the primary
beam of the individual
antennas, then the diameter of the antenna can be no more than about 1000
times the wavelength
at which the observations are made. For observations done at 1 mm
wavelength this means the
individual antennas making up the array should have a diameter of less
than one meter. An
enormous number of such small antennas are needed to get enough collecting
area to realize much
sensitivity: 150 such antennas is needed just to achieve the same
collecting area as the present
NRAO 12 Meter Telescope. The cost of equipping such an array with
cryogenic receivers, and of
cross-correlating the data, makes such a solution impractical.
In order to preserve the scientific capability to image large objects,
and satisfy the practical
requirement that the antennas be of diameter large enough that a
reasonable number of them will
provide the needed sensitivity, means that the array must be capable of
producing not just
precision images, but precision mosaic images. Making such
images requires observing with
multiple, precision antenna pointings. In a seminal work Cornwell,
Holdaway, and Uson
(included here as Appendix A) showed that this can indeed be done if the
deviation from
perfection of the antenna figure is no more than three percent of the
observing wavelength and if
the pointing error of the antennas does not exceed five percent of the
primary beamwidth. These
severe technical requirements, deriving from a science requirement, are
the engineering
specifications for the antennas. Meeting such specifications in the
antenna design is a major task
to be accomplished in the MMA D&D program.
The need for sensitivity sufficient for the study of faint objects with
the MMA implies a
requirement for (1) development of broadband, quantum-limited receivers;
(2) design of antennas
of very low blockage so that the warm spillover is minimized; and
(3) choice of a site for the array
where the background emission and absorption from atmospheric water is
minimal. The first two
points are the focus of the D&D instrumentation development described
below. The site issue is
resolved by noting that since atmospheric water vapor is concentrated low
in the Earth's
atmosphere, the necessary site for the MMA is at high elevation. Two site
options have been
studied: Mauna Kea at 12,500 feet above sea level and Llano de Chajnantor
in the Chilean
Altiplano at 16,500 feet elevation.
The combination of requirements for a high-tech instrument to be
located in a remote site means
that great care must be taken in the design of the array. The MMA
instrumentation will need to
be reliable to minimize the failure rate and modular so that it can easily
be removed when
necessary for repair at a laboratory located at a lower altitude.
Together, these considerations
mean that the MMA design requires attention to maintenance issues. The
three-year MMA
Design and Development program is structured to make this possible.
The final astronomical requirement for the MMA to be addressed in the
MMA D&D phase
concerns ease of scientific access. Recognizing that the MMA will be
extremely fast--images of
small fields can be done in minutes--and suitable for a wide range of
scientific investigation,
astronomers seek to receive images as one of the data products
from the array. This goal should
not remove the ability of the sophisticated synthesis astronomer to refine
his or her image through
subsequent processing, but it should allow non-expert astronomers to use
the instrument easily
and effectively. It involves development of instrumentation and software
not presently part of
operating radio synthesis instruments. Fortunately, some of the ideas
that will go into this task for
the MMA can be tried and refined on existing instruments; that is the
thrust of the MMA D&D
effort in the area of data processing.
2.3 NSF Expectations for the MMA Design and Development Phase
Structure of the MMA project as three overlapping phases--Design and
Development,
Construction, and Operations--is a construct suggested by the NSF
Astronomy Advisory
Committee in 1992. It recognized that paper studies alone are not
sufficient and prototyping must
accompany design. With built prototypes one can realistically assess the
cost and performance of
the array to be constructed. The fact that a synthesis array such as the
MMA is composed of
multiple copies and variations of fundamental pieces means that with the
construction of a few
prototypes one can forecast the as-built cost and performance of the
complete instrument with
considerable confidence. This is a prudent approach for a synthesis
telescope project.
The deliverables to the NSF of the MMA D&D program are
these:
a firm, auditable, cost estimate for construction of the MMA
based on instrumentation and
software prototypes;
The requirement for partnership in the MMA speaks to issues that
transcend the task of designing
and prototyping a forefront scientific instrument. Nevertheless, the
interests shown in the MMA
by potential partners have the potential to enhance the scientific
capabilities of the instrument
significantly and, as such, this requirement may be both beneficial and
achievable.
2.4 NRAO Principles for the MMA Design and Development Phase
An instrument as capable as the Millimeter Array will have a profound
effect on astronomical
research in the U.S. and it will have a profound effect on the NRAO. A
major goal of the MMA
D&D phase is to develop an organization for the project such that the
MMA enhances all of the
research infrastructure of radio astronomy in the U.S. To this end two
specific principles have
been established:
2.5 Project Goals
The MMA D&D project requirements outlined above result in the
following goals for this initial
three-year phase of the project:
III. ORGANIZATION OF THE MMA PROJECT
The Millimeter Array project is an integral part of the NRAO, organized
as a construction project
of the Observatory under the supervision of the NRAO Director and the
management oversight of
Associated Universities, Inc. The relationships are shown on the
organization chart, Appendix B.
An unfortunate effect of separating the MMA project into an initial
three-year Design and
Development Phase to be followed, pending approval, by a six-year
construction phase, is to limit
the opportunity to gather the entire MMA team at one place early in the
project. The uncertainty
associated with the long-term prospect for the project beyond the first
three years implies that
staff may need to relocate for a period of less than three years only to
be faced with the prospect
of again moving should the construction phase of the MMA be delayed or
canceled. The only
practical alternative, adopted for the MMA D&D phase, is to make use
of the people involved in
the development and prototyping activities at the locations where they are
currently employed.
Development of the SIS mixer devices, the transistor (HFET) amplifiers and
design of the
correlator will be done at the NRAO Central Development Laboratory in
Charlottesville, Virginia;
the prototype receiver system (cryogenic dewar, refrigerator and control
instrumentation) and the
antenna design will be done at the NRAO facilities in Tucson, Arizona; the
array software
development, IF transmission system, operational planning, and system
integration on the test
interferometer will be done at the NRAO in Socorro, New Mexico. The
organizational challenge
to the management of the D&D program is coordination of the efforts of
these geographically
separate groups. Management of the NRAO as a whole involves these same
challenges, the issue
for the MMA is not unique.
The MMA D&D tasks will be conducted by a full-time staff assigned
to the project. These people
will be NRAO employees. The major D&D tasks will be managed by
Project Division Heads
whose responsibility it is to organize the efforts of the staff assigned
to the task. Each major task
will have a Working Group, a committee of experts made up of
individuals at the NRAO,
assigned to the MMA project, and individuals among the university groups
who can advise and
guide MMA work being done at the NRAO. There are four such joint
NRAO-university working
groups that meet at regular intervals:
In addition, there are two others: a site testing group made up wholly
of NRAO/MMA staff, and a
science working group, comprised wholly of university-based astronomers
whose purpose it is to
advise the MMA Project Scientist. Written reports are kept for all six
Working Group meetings
and these reports, together with the relevant ancillary information, are
posted to the WWW so as
to be available to all those interested in progress of the MMA project.
The organization chart for the MMA project, illustrating the activities
to which the working
groups contribute, is given in Appendix C. An important part of that
organization is the
Millimeter Array Development Consortium (MDC). The MDC is a collaboration
between the
NRAO and the university groups that operate millimeter arrays in the U.S.,
namely, the Caltech
Owens Valley Radio Observatory (OVRO) and the Berkeley-Illinois-Maryland
Association
(BIMA). By means of participation in the MDC Executive Steering
Committee, OVRO and
BIMA are fully involved in the decision making process for the MMA
development.
From its very inception, the Millimeter Array has been a collaboration
between the NRAO and the
U.S. astronomical community. The ideas that form the backbone of the
instrument definition are
contributed by interested individuals in the form of MMA Memos. The memo
series provides a
forum for considered analysis of the issues facing a project of the
magnitude and importance of
the MMA; it provides a permanent record of the views and analysis that
have gone into the
definition of the MMA. More than 100 people have participated as authors
of the series of MMA
memos that now spans the past sixteen years of MMA development. The MMA
memo series is
accessible via the WWW (http://www.mma.nrao.edu
/memos/memolist). Appendix D is a summary of the
titles and authors of the MMA Memos. It is an important and effective
means of maintaining
communication about MMA planning with the community of interested U.S.
astronomers.
IV. INSTRUMENTATION DEVELOPMENT FOR THE MMA
The instrumentation sought by astronomers for the MMA extends
significantly the capabilities
available at present from instrumentation on existing millimeter-wave
synthesis arrays. To achieve
the MMA specifications in some instances will require an escalation of
design techniques presently
in use; in others it will require a wholly new design approach. In either
case, the opportunity
provided by the MMA D&D phase will permit the design approach adopted
to be verified for
each of the major MMA instrumentation tasks.
A condensed overview of the MMA instrumentation design goals as
presently planned compared
to the current state-of-the-art on operational arrays (OVRO, BIMA, the
Nobeyama Radio
Observatory (NRO) array, and the Institut Radio Astronomie Millimetrique
(IRAM) array) is
given in Table 4.1 below.
TABLE 4.1 OVERVIEW AND COMPARISON OF MMA
INSTRUMENTATION GOALS
Spec Operating Arrays A brief description of the principal challenges for development of the
MMA instrumentation in the
areas of antennas, system, receivers, correlator, and computing is given
below; these comments
are amplified by material referenced in the Appendices to this volume.
4.1 Antenna
The antennas are the single most costly part of the MMA, the most
visible, and the most likely to
have the longest life in service. The scientific requirement that the MMA
have good mosaicking
capability has a strong effect on the antenna design: it means that the
antennas have to point
exceptionally well and that the sidelobe response cannot vary appreciably
with time or antenna
orientation. Because the array needs to be reconfigurable, the antennas
must be transportable and
this in turn means that they cannot be secured in an enclosure; they must
be in the open air and
meet their performance specifications fully exposed to the environment
(e.g. sun and wind).
Moreover, the MMA will be built at a high altitude, remote site. This
implies that the antennas
should be designed for low maintenance and long component life.
The MMA antenna specifications are described in detail in MMA Memo 145,
included here as
Appendix E. Table 4.2 shows a concise summary of the specifications.
TABLE 4.2 ANTENNA SPECIFICATIONS
< 25 RMS 75% of the time < 22 micrometers RMS 50% of the time
< 56 micrometers RMS 75% of the time The antennas proposed for the MMA in 1990 were conceived of as being 8
m in diameter. The
fiducial design was for a passive antenna, one with no active elements
working to adjust the
antenna shape or pointing. The possibility of securing partnership in the
MMA with the
Europeans or the Japanese, as described in Section VII, has served to
focus MMA antenna design
studies on a larger, 10 m diameter, design that would achieve the
scientific goals of the MMA and
the complementary goals of the Europeans and/or Japanese. Such a change
provides a foundation
for a partnership and yet allows a stand-alone MMA to be built of 36 such
antennas, should these
particular partnership initiatives fail.
The MMA antennas will be built under contract. In the MMA Design and
Development plan, a
contract will be let for an initial prototype antenna, with an option for
a second antenna. This will
be a design/build to performance contract. Although the ability of the
design to meet the
specifications will be the responsibility of the contractor, the MMA
antenna group will engineer a
concept design that they believe meets the MMA specs; that design will be
given to all contractors
interested in bidding on the antenna contract. At their discretion, they
may use and modify that
design or not. In either case, having the in-house design will give the
MMA antenna engineers a
tool with which to compare and assess the contractor's design. After the
design is accepted, the
MMA engineering team will monitor the progress of the contractor's
fabrication efforts and they
will be in a position to evaluate the desirability of making specific
engineering refinements prior to
contracting for the production suite of MMA antennas. The antenna design
and all the drawings
done by the contractor will become the property of AUI.
The production quantity of MMA antennas will be bid separately from the
prototyping work on
the initial one or two antennas. Quantity antenna procurement will be
done in the construction
phase of the MMA. The production procurement will be a build-to-print
contract, not a
build-to-spec contract. The purpose of the antenna prototyping in the
D&D phase is precisely to
allow us to assess the as-built design in sufficient detail that we can be
confident that there is little
risk associated with a build-to-print quantity antenna procurement. Such
an approach will enlarge
the pool of contractors interested in bidding on the MMA antenna contract
and capable of
performing the work satisfactorily. We anticipate a substantial cost
saving will be realized by this
approach and the competition it will foster.
4.2 System
The electronics system for a large synthesis array such as the MMA is
complex, with the signals
received by the antennas undergoing numerous frequency conversions using
local oscillators with
precisely controlled phases. The current concept for the MMA system
design is given in MMA
Memo 190, included here as Appendix F. The principal parts of the
electronics system are
receivers, local oscillator, wide bandwidth transmission system, and
correlator. Some of these
sub-systems are discussed in more detail below. The detailed design of
this system is an important
task for the design and development phase of the project.
Some of the major technical challenges for the overall system design
are the maintenance of phase
stability adequate for the highest observing frequency in the various
signal paths and provision of
an accurate total power measurement capability. The remote location of
the MMA requires that
the system be designed for easy operation and maintenance, implying a
monitoring system
adequate for off-site fault diagnosis and the packaging of all electronics
in easily replaceable
modules.
4.3 Receivers
The receiver plan for the MMA envisions use of transistor amplifiers,
HFETs (heterostructure
field effect transistor) for the frequency bands near 30 and 90 GHz, and
use of SIS mixers at
higher frequencies. For the 2.6 mm band that includes the CO(J=1-0)
transition at 115 GHz a
choice between HFET and SIS will be made based on the performance figures
demonstrated by
the prototype HFET amplifier in this band.
HFET amplifiers at 30 and 90 GHz with performance specifications
similar to those of the MMA
are being fabricated now at the NRAO Central Development Laboratory (CDL)
for use on the
Very Large Array (VLA), the Very Long Baseline Array (VLBA), and for the
NASA Microwave
Anisotropy Probe (MAP) spacecraft. Little work is necessary to refine
these designs for the
specific needs of the MMA.
SIS Mixers for use on the NRAO 12 Meter Telescope at frequencies from
70 to 300 GHz are also
produced as needed at the CDL. However, because the sites under
consideration for the MMA
are so dry with such little emission from atmospheric water vapor, there
are significant gains in
sensitivity to be realized if it is possible to provide the MMA with truly
quantum-limited SIS
mixer receivers. Presently the best SIS receivers have noise temperatures
in the range two to four
times the photon temperature, hf/k. This receiver noise contribution can
be exceeded by emission
from atmospheric water vapor in the unwanted (image) sideband and it can
be degraded by noise
from the local oscillator. The MMA goal is to minimize both these effects
through the use of
balanced, image-separating SIS mixers.
While most SIS mixer receivers respond to both upper and lower
sidebands, few astronomical
observations require this capability; most observations seek to employ one
sideband or the other.
Nevertheless, for a double sideband system the unwanted response of the
image sideband adds
atmospheric emission to the system temperature increasing the observing
time required to reach a
given sensitivity. The approach to be taken in design of the MMA SIS
mixers is to use
microfabricated LO or IF quadrature hybrids to combine the signal from a
pair of mixers and in
and out of phase so as to separate the sidebands. The approach to be
taken is outlined in
Appendix G.
Local oscillator power is usually coupled into a SIS mixer using a
directional coupler or beam
splitter. If the signal path loss through the LO coupler is to be kept
small, the LO loss will be
large, typically 15-20 dB. In addition to wasting LO power, noise
from the LO source in the
signal and image bands is coupled into the mixer. A balanced mixer
minimizes both these effects.
It has a separate LO port for efficient coupling to a pair of mixers so
that the LO power is
reduced relative to the single-ended mixer. Sideband noise is also
reduced by phase and
amplitude balance through the mixer. See Appendix G.
In the MMA Design and Development program a balanced, image-separating
SIS mixer will be
developed at 230 GHz. The device will be integrated with an HFET IF
amplifier for broadband
performance. The goal of the work is to demonstrate both that the design
approach is sound and
to produce an SIS design that can be scaled to all the MMA frequencies at
which SIS mixer
receivers will be used. The 230 GHz SIS mixer will be
incorporated in the prototype receiver that
will go on Antenna #1 in June 2001.
4.4 Local Oscillator
The MMA Design and Development plan provides support to parallel
efforts for development of
the local oscillator system: a conventional microwave source multiplied by
varactor diodes will be
designed and built for the 230 GHz band of the prototype receiver and,
simultaneously, a
photonic system will be built. The photonic approach offers, potentially,
greater simplicity and
reliability at lower cost but it will require substantial development
effort if it is to be adopted for
the MMA.
The conventional LO development planned in the MMA D&D phase will
be done in three phases.
First, several 100 GHz phase-locked LO chains will be built and evaluated
on the basis of
available power, as well as on phase and amplitude noise. Second, the
optimum design will be
adapted for the specific MMA needs (capable of appropriate fringe
rotation, tuning range). The
third phase will involve extending the 100 GHz system to 230 GHz
through the use of a
fixed-tuned, planar varactor frequency multiplier. Fiducial designs for
higher frequency bands will
follow.
The photonic LO will involve phase-locking the difference frequency of
two solid-state lasers
operating near 1550 nanometers. As applied to the MMA, the pair of
laser signals would be sent
along a single fiber (for each antenna) from a central building to the
antennas. There the signals
would be put into a photomixer with the difference frequency becoming the
receiver local
oscillator. A contract is in place with UCLA for development of a
velocity-matched
traveling-wave photodetector for the WR-10 waveguide band
(75-110 GHz). When this is
delivered the complete photonic local oscillator will be assembled and
compared with the
conventional LO for noise and stability. One of the two approaches will
be adopted for the MMA
and developed further.
4.5 Correlator
The plans for the MMA correlator development begin with the design and
fabrication of an
early-generation correlator that can be used with the test interferometer
to evaluate the first
antenna prototypes and to assess the performance of the initial prototype
instrumentation. This is
a single baseline cross-correlator, with spectroscopic capability, built
around the chip developed
for the spectrometer on the Green Bank Telescope.
Design of the MMA correlator itself will begin immediately but it is a
much longer term effort.
The plan calls for it to be built in a modular form such that it can be
delivered one-quarter at a
time. This staged delivery not only permits early analysis and debugging
of the correlator in an
operational setting but it also provides for a realistic appraisal of the
controlling software and for
an opportunity to use early subsets of the correlator to support interim
operations of the array as
it is assembled. The basic specifications for the correlator are that it
will support:
The correlator planning is outlined in MMA Memo 166, included here as
Appendix H.
4.6 Computing
Specification of the appropriate computing environment for the MMA
needs to combine the needs
of controlling the instrumentation in real time with the needs of
people and hardware to monitor
the performance of those instruments and with the needs of the astronomer
to interpret quickly
the scientific product of the observations. Fortunately, there is an
enormous amount of
experience at the NRAO and in the community that may be brought to bear on
the MMA
computing task. The MMA D&D planning emphasizes the need to recruit
that expertise.
Appendix I, MMA Memo 164, is a report of the MMA Computing Working
Group. It lays out
the high level requirements for the computing task. Especially important
among the conclusions
in this report are these:
Both of these requirements demand that the software supporting the MMA
have more
information available to it than is presently the case with operating
radio synthesis arrays. This
imposes a burden on the MMA hardware designs in many areas; it also means
that the computing
system must be capable of evolving as techniques that are useful to the
astronomer/users are
developed.
In the D&D program the opportunity will be taken to experiment with
software tools, techniques,
and interfaces on existing arrays through the Millimeter Array Development
Consortium (MDC)
collaboration, while at the same time sticking to the delivery schedule
needed for software for
support of early testing at the test interferometer.
V. IMPLICATIONS OF A REMOTE SITE FOR THE MMA: ROLE OF THE TEST INTERFEROMETER
Regardless of whether the MMA is located on Mauna Kea or on the Chilean
Altiplano,
construction of the array will involve trans-oceanic shipment of
materials. As long as such
materials can be packaged in standard ocean shipping containers the
shipping cost is determined
by the number of containers shipped: the cost is all in the loading and
unloading. Given this, one
may consider either to accumulate and ship construction materials for
assembly of the instrument
on-site, or alternatively, to fabricate and test large sub-assemblies in
the continental U.S. and ship
them as modular units. The latter approach is preferable because it
permits high-level MMA staff
to assemble and test MMA instrumentation in existing NRAO laboratories
where the staff are
currently located; there is no expense associated with relocating staff
with sophisticated technical
skills to a remote location for instrument assembly. Ideally, the work
on-site can be reduced to
that of connecting major sub-systems and testing.
Taking one step back from construction and considering this same issue
for the development
phase of the project one reaches the identical conclusion. Namely, the
process of verifying the
performance of prototype instrumentation for the MMA is best done by the
designers of that
instrumentation in a controlled, but realistic, environment whenever
possible. The MMA D&D
plan envisions the construction of a test interferometer comprised of two
prototype MMA
antennas, located at the VLA site and used to mount and test all the
prototype instrumentation
and software built during the D&D phase. This provides both a
comprehensive system test for
the MMA prototype assembly and it provides a test facility for the
evaluation of successive
iterations of hardware and software developments. Once performance and
system compatibility is
established on the test interferometer for each representative piece of
MMA instrumentation one
can build production quantities of that device and ship it directly to the
array site confident that it
will integrate smoothly into the overall array assembly on-site. The test
interferometer will also
be used for the initial training of the MMA operational staff.
At the conclusion of the three-year MMA D&D phase the prototype
antenna and representative
prototypes of all the MMA radiometric instrumentation will be complete and
delivered to the
VLA site. Specifically, the following deliverables, that are the product
of the D&D effort, will be
present on 1 June 2001 at the test interferometer:
Baseband converters are MMA design.
Fiber optic transmitters/receivers are MMA design. Spectroscopic capability.
Built around existing GBT chip, not final MMA.
Holography back-end. Single antenna and interferometer control.
Rudimentary operator interface.
Rudimentary astronomer interface.
Mapping program for testing purposes.
Fast-switch positioning capability. The next phases of the project, construction and operations, will begin
with the detailed
evaluation of the prototype hardware using the test interferometer (the
first two prototype
antennas) to be followed by design refinements and ultimately production
fabrication of
assemblies to be shipped directly to the array site.
The specific tests to be done with the prototype interferometer in the
first 6-12 months of its
operation in the construction phase of the MMA project include:
VI. WORK BREAKDOWN STRUCTURE
The tasks to be accomplished in the MMA Design and Development phase
are enumerated on the
D&D Gantt Chart, Table 6.1. This outlines the general steps to be
taken in each of the MMA
development areas, a time estimate for each and the personnel resources
needed to address the
tasks. The MMA D&D program is done to prepare for MMA construction
and as such the
deliverables are designs, decisions, and prototypes, not
production quantities of any of the array
hardware. In the final year of the program, approximately June 2000
to June 2001, many of the
D&D tasks will have been completed and their outputs delivered to the
appropriate site for testing
or incorporation in larger parts of the prototype hardware. The staff
involved with such
completed tasks will, at that time, be transferred either to the
construction phase of the MMA
which is anticipated to begin in FY2001 (October of 2000) or to the
operations phase of the
project which should begin in 2001.
Tables 6.2 and 6.3 present a summary of the Milestones and Deliverables
of the project as
abstracted from the Gantt Chart so that they may be easily reviewed either
by task (Table 6.2) or
by date (Table 6.3). Tables 6.4 and 6.5 summarize the personnel
assignments by skill and task
respectively.
Table 6.6 is an illustration of the breakdown of expenditures planned
in support of the Design and
Development work. The total cost for the three-year Design and Development
program given
here, $26.0M in current dollars, is only incrementally larger than the
$22.3M cost estimated for
the 1992 D&D plan, also in current dollars (see Volume I). An annual
inflation of 2.6 percent will
wholly account for this increase over the intervening six years.
TABLE 6.1 D&D GANTT CHART
This file in PDF Format Only - Design and
Development Gantt Chart (January 28, 1998).
TABLE 6.2 MILESTONES AND DELIVERABLES BY
TASK
TABLE 6.3 MILESTONES AND DELIVERABLES: CHRONOLOGICAL
6.4 Personnel Projections
The personnel required to carry out the tasks of the MMA Design and
Development program and
those needed to prepare for the construction phase of the project are
summarized in the tables
below. Table 6.4 presents the distribution of staff by calendar year.
These are numbers of
full-time employees, to be distinguished from years-worked for 1998 where
the project begins in
June, and for 2001 where the D&D phase is completed in June. Table
6.5 shows the distribution
of staff by task for 1999 where the project is fully staffed and the
project extends for the entire
calendar year. The personnel allocation for other years is essentially
identical to that in 1999; in
the final year personnel will begin to be transferred to construction or
operations.
The MDC university-based people working on the project are among those
included in this table.
The MMA D&D personnel projection given here, approximately fifty
full-time employees, is
identical to the number of employees projected as being needed in the MMA
D&D plan, Volume
I, as presented to the NSF in 1992. The distribution of skills required,
and the task assignments,
is also little changed from the 1992 plan. The goals of the plan as
presented here are, however,
much refined over the plan as given in Volume I.
6.5 Budget
The budget plan for the MMA Design and Development project is shown in
Table 6.6. Entries
are in dollars of the year of expenditure (current dollars). Personnel
cost estimates are consistent
with the personnel plan presented in Section 6.4 above. Cost estimates
for all of the electronic
sub-systems have been made by NRAO engineers experienced in building
equipment similar to
that needed for the MMA These tasks together account for approximately 75
percent of the
expense of the D&D program. The current allocation for contingency
is low but is considered acceptable because of the small fraction (approximately 25
percent of the total budget)
of the development work being done as contracts to commercial companies.
The cost of the
major contract, the antenna design and first element construction, has
been estimated from
budgetary figures provided at our request by three companies experienced
in millimeter
wavelength design and construction.
TABLE 6.4 PERSONNEL PROJECTIONS BY CALENDAR
YEAR
TABLE 6.5 PERSONNEL PROJECTIONS BY TASK:
CALENDAR YEAR 1999
TABLE 6.6 MMA DESIGN AND DEVELOPMENT: TOTAL
COST
VII. DESIGN AND DEVELOPMENT IN THE MMA PROJECT CONTEXT
7.1 Operations
At the conclusion of the Design and Development phase the prototype
antenna will be delivered
and accepted at the VLA where MMA system testing will be done. That
antenna will be outfitted
with the prototype receiver and controlled by prototype software written
for the purpose. There
will be a correlator to use as an autocorrelator for spectroscopic tests
and as a cross correlator
when the second prototype antenna arrives at the test site. As described
above, the test
instrument will be used to evaluate the antenna performance and to verify
the performance and
compatibility of successive iterations of all the MMA instrumentation.
The test interferometer
also plays a pivotal role in software development. All this can be done
in a facility convenient for
the MMA development staff.
The operating staff for the test interferometer will be the initial
members of the MMA operations
staff. These people will be among the first assigned to the MMA site
itself and they will be used
both to establish and document operational procedures, and to train
additional hires to operations.
Once the MMA site construction is in progress and some of these
individuals are assigned to that
location to initiate on-site operations, the MMA operations will be
expanded to include operation
of both the test interferometer on the VLA site and to interim MMA
operations at the MMA site.
A timely start to the gradual build-up of MMA operations beginning in
June of 2001 with the
access to a prototype antenna provided by the MMA D&D program is
important to the orderly
development of MMA operations.
7.2 Construction: Role of MMA Partnerships
Among the deliverables of the MMA Design and Development
initiative is an agreed partnership
in the array by foreign countries or by U.S. agencies other than the NSF.
Partners in the project
will have their own ideas as to the structure of the construction
project--they will need to be
included in the intellectual and financial description of MMA
construction--and, owing to this, it is
not possible to lay out the entire construction project in detail.
Moreover, some crucial aspects of
the instrument construction await the design decisions to be made in the
next three years of the
D&D effort. All this means that one can present an overview of how
the construction project
could go, understanding that this will need modification as the
partnerships and the D&D
progresses. Such an overview is given in the Gantt Chart, Table 7.1. Its
value is in the
illustration it provides for the interrelation between Design and
Development, Construction and
Operations. Note that the construction work begins promptly at the start
of FY2001 (October
2000) and smooths the transition between the D&D activities and the
construction efforts task by
task.
Two partnership possibilities may have a significant effect on the
construction phase planning.
These are the possibility of joining the MMA with the Japanese project,
the Large Millimeter and
Submillimeter Array (LMSA), or with the European Large Southern Array
(LSA). Both of these
initiatives have considerable support among their respective scientific
communities and the leaders
of both have expressed interest in discussing how their projects could be
joined with the MMA to
the benefit of all. Either combination with the MMA, or better, a
combination of all three, would
provide such a truly powerful imaging instrument. The U.S. community has
been supportive of
efforts by the MMA staff to secure such a partnership. One barrier to a
joint project is the
dissimilar antenna diameters considered by the three; the MMA has planned
8 m antennas, the
LMSA 10 m antennas, and the LSA 15 m antennas. Recent
discussions among the three groups
have led to successive compromises on the diameter to the range
1012 m. The MMA Design
and Development antenna design efforts will therefore focus on antennas of
10 m diameter to
facilitate a partnership with one or both of these groups. Should the
partnership initiatives fail,
the MMA construction budget estimates would allow an array of
approximately 36 antennas to be
built. Such an array could accomplish all the scientific goals projected
for an array of forty 8-m
antennas, but would do so with nearly fifty percent more collecting area.
This subtle change in
the baseline MMA planning is therefore an asset to be used to court
partnership with the LMSA
or LSA and an asset to the sensitivity of the MMA as a stand alone
instrument. Progress in
securing partnerships for the MMA, as indeed progress in realizing the
technology to achieve the
capabilities desired of the MMA itself, begins with the efforts outlined
in the MMA Design and
Development Plan.
TABLE 7.1 PROJECT OVERVIEW GANTT CHART
PDF Format Only - Project Overview
Gantt Chart
MMA
Capabilities of
Currently
ANTENNAS RSS Surface Accuracy
< 25 microns
30-80 microns Pointing Precision
08
> 3" Fast Switching
Cycle < 10s
No Capability Total Power Observing
Yes
No Capability RECEIVERS 28-45 GHz HFET
Yes
Special Purpose only 67-95 GHz HFET
Yes
No Capability 91-119 GHz SIS or HFET
Yes
Yes 125-163 GHz SIS
Yes
NRO only 163-211 GHz SIS
Yes
No capability 211-275 GHz SIS
Yes
Yes 275-370 GHz SIS
Yes
No capability 385-500 GHz SIS
Yes
No capability 602-720 GHz SIS
Yes
No capability 787-950 GHz SIS
Planned
No capability SIS Balanced Mixers
Yes
No SIS Image Separating
Yes
No SIS Integrated with IF
Yes
No Dual Polarization
Yes
No IF Bandwidth
2 x 8 GHz
2 x 1 GHz
Frequency Range
30 to 950 GHz Surface Accuracy
< 25 micrometers RMS Pointing Accuracy
< 08 RMS 50% of the time
Phase Stability
< 10 micrometers RMS 25% of the time
Dynamical
Performance
Switch 1.5 degrees within 1 second of
timeSubreflector Nutation3 beamwidths at 86
GHzClose Packing< 1.3 times the antenna
diameterPhysical DesignSimple and durable.
Antennas
The first prototype, designed and built under contract by
an antenna
fabricator. An option for the second prototype antenna will have been
exercised and funded from the MMA construction phase of the
project. Receivers
Prototype MMA cryogenic dewar and compressor. Three frequency
inserts,
at 30, 90, and 230 GHz. The initial 230 GHz SIS receiver is the
prototype,
balanced, image-separating mixer integrated with the broadband HFET IF
amplifier. The design is scalable to the other MMA frequencies. LO
Low frequency Gunn; multiplied by prototype broadband planar varactor
diodes for 230 GHz. IF
Lowest 1 GHz of the MMA 4-12 GHz system.
Correlator
Single baseline with 1 GHz bandwidth in each of two polarizations.
M/C
Software
Monitor and Control bus is MMA prototype.
ADMINISTRATION Deliver Initial MMA Cost Estimate
June 1998 Deliver Midterm MMA Cost Estimate
June 1999 Deliver Final MMA Cost Estimate
June 2001 SITE Site Recommendation
June 1998 Signed Array Site Use Permission
Dec 1999 Configuration Review
Mar 2000 Signed Support Facility Use Permission
Dec 1999 ANTENNA Preliminary Design Review
July 1998 Critical Design Review
Jan 1999 Bid Prototype Antenna (#1+option #2)
June 1999 Receive Bid Response
Sept 1999 Sign Antenna Contract
Jan 2000 Issue Transporter RFF
Sept 2000 Exercise Option Ant #2
Oct 2000 Receive Transporter Bids
Nov 2000 Sign Transporter Contract
Jan 2001 Deliver/Accept Ant #1
June 2001 Receive/Accept Transporter
June 2001 SIS MIXER Preliminary Design Review
Jan 2000 Deliver Proto 230 GHz
Apr 2000 Critical Design Review
Jan 2001 Deliver MMA 230 GHz
Apr 2001 HFET AMPLIFIER Procure HFET Wafer
Jan 1999 Deliver 30 GHz
Jan 1999 Deliver 90 GHz
Sept 1999 LOCAL OSCILLATOR Preliminary Design Review: Conventional
Sept 1998 Demo 230 GHz Doubler
Dec 1998 Preliminary Design Review: Photonic
Jan 1999 Critical Design Review: Conventional
June 1999 Breadboard PLL Demonstration
Sept 1999 Receive 3 mm Photodetector
Sept 1999 230 GHz LO Demonstration
Dec 1999 Deliver LO to Tucson
Mar 2000 Critical Design Review Photonic: Decision
June 2000 RECEIVER SYSTEM Receiver Package Preliminary Review
Apr 1999 Receiver Package Critical Review
Jan 2001 Deliver Prototype Receiver
Jun 2001 Deliver Holography System
June 2001 CORRELATOR PDR Correlator Design
Jan 1999 Deliver Test Correlator
Jan 2000 CDR Correlator Design
July 2000 SIGNAL TRANSMISSION Preliminary Design Review
Oct 1998 Critical Design Review
June 1999 COMPUTING PDR: Test Correlator Interface
Mar 1999 PDR: Single Dish System
Mar 1999 CDR: Test Correlator Interface
June 1999 Deliver Test Correlator Interface
Jan 2000 PDR: Computing System
June 2000 CDR: Single Dish System
June 2000 CDR: Computing System
Jan 2001 Deliver Single Dish System
Jan 2001 SYSTEM INTEGRATION AND TEST Project Book Version 1
June 1998 Project Preliminary Review
July 1998 Project Book Version 2
June 2000 Project Critical Review
Jan 2001 Test Interferometer Site Complete
June 2001
CALENDAR YEAR
1998 June
Deliver Initial MMA Cost
Estimate
Site Recommendation
Project Book Version 1 July
PDR: Antenna Design
Project Preliminary Review September
PDR: Conventional LO October
PDR: Signal Transmission System December
230 GHz LO Doubler Demonstration CALENDAR YEAR 1999 January
CDR: Antenna Design
Procure HFET Wafer
Deliver 30 GHz Amplifier
PDR: Photonic LO Design
PDR: Correlator Design April
Receiver Package Preliminary Review March
PDR: Test Correlator Interface
PDR: Single Dish Software System June
Deliver Midterm MMA Cost Estimate
Bid Prototype Antenna (#1+option #2)
CDR: Conventional LO Design
CDR: Signal Transmission System Design
CDR: Test Correlator Interface September
Receive Antenna Bid Response
Deliver 90 GHz HFET Amplifier
Demonstrate Breadboard PLL
Receive 3mm Photodetector December
Signed Array Site Use Permission
Signed Ops Facility Site Use Permission
230 GHz Conventional LO Demonstration CALENDAR YEAR 2000 January
Sign Antenna Contract Ant #1
PDR: SIS Mixer Design
Deliver Test Correlator
Deliver Test Correlator Interface March
Deliver Completed LO to Tucson
Configuration Review April
Deliver Proto 230 GHz June
CDR: Photonic LO, Decision to Continue
PDR: Computing System Design
CDR: Single Dish Computing System
Project Book Version 2 July
CDR: Correlator Design September
Issue Transporter RFP October
Exercise Option on Antenna #2 November
Receive Transporter Bids CALENDAR YEAR 2001 January
CDR: SIS Mixer Design
CDR: Receiver Package
CDR: Computing System
Deliver Single Dish Computing System
Project Critical Review
Sign Transporter Contract April
Deliver MMA 230 GHz SIS Mixer June
Deliver Final MMA Cost Estimate
Receive/Accept Antenna #1
Deliver Prototype Receiver to VLA Site
Deliver Holography System to VLA Site
Test Interferometer Site Complete
Receive/Accept Transporter
1998
1999
2000
2001 Scientist (S)
5.75
6.00
6.75
5.75 Engineer (E)
23.50
24.75
23.25
23.25 Programmer (P)
5.50
5.75
5.75
7.25 Technician (T)
11.00
12.00
12.5
12.00 Machinist (M)
1.00
2.00
2.00
2.00 Administrative/Other (A)
3.50
3.75
3.75
3.75
S
E
P
T
M
A
Total Administration
3.75
.50
3.50
7.75 Array Site
1.75
0.75
1.25
0.25
4.00 Antenna
Development
3.00
1.00
4.00 SIS Mixer
5.00
3.00
1.00
9.00 HFET Amplifier
1.00
1.00
2.00 LO: Conventional
1.00
1.00
2.00 LO: Photonic
1.50
1.00
2.50 Receiver Systems
3.0
2.00
1.00
6.00 Correlator
4.0
1.00
5.00 Signal Transmission
4.0
0.50
4.50 Computing
5.00
5.00 System Integration
0.50
1.50
0.25
0.25
2.50 TOTAL
6.00
24.75
5.75
12.00
2.00
3.75
54.25