European ALMA
Newsletter
January 2006 7th edition
Download pdf here
This is the seventh edition of the quarterly European ALMA newsletter. The editors are Tom Wilson, Carlos De Breuck and Martin Zwaan (ESO).
|
ESO, the European Organisation for Astronomical Research in the Southern Hemisphere, announced on December 6 that it has signed a contract with the consortium led by Alcatel Alenia Space, European Industrial Engineering (Italy) and MT Aerospace (Germany). The consortium will supply 25 antennas for the ALMA project, along with an option for another seven antennas. The contract, worth 147 million euros, covers the design, manufacture, transport and on-site integration of the antennas. It is the largest contract ever signed in ground-based astronomy in Europe. The ALMA antennas present difficult technical challenges, since the antenna surface accuracy must be within 25 microns, the pointing accuracy within 0.6 arc seconds, and the antennas must be able to be moved between various stations on the ALMA site. This is especially remarkable since the antennas will be located outdoors in all weather conditions, without any protection. Moreover, the ALMA antennas can be pointed directly at the Sun. ALMA will have a collecting area of more than 5,600 square meters, allowing for unprecedented measurements of extremely faint objects. Pascale Sourisse, Chairman and CEO of Alcatel Alenia Space, said: "We would like to thank ESO for trusting us to take on this new challenge. We are bringing to the table not only our recognized expertise in antenna development, but also our long-standing experience in coordinating consortiums in charge of complex, high-performance ground systems." The completion of the provisional acceptance of the first antenna is scheduled for September 21, 2008. The provisional acceptance of the 25th antenna shall be completed by December 6, 2011, i.e., exactly 72 months after the contract signature. Antenna transporters One of the unique features of ALMA is the possibility to move the radio telescopes to well defined positions around the high altitude plateau of Chajnantor and to transport antennas from the Operations Support Facility to the observing site. In order to do this, specially designed transporters, meeting all environmental conditions at an altitude of 5000 meters, need to be designed and manufactured, and delivered to the Atacama desert. ESO has signed a contract with Scheuerle Fahrzeugfabrik GmbH, a world-leader in the design and production of custom-built heavy-duty vehicles for two antenna transporters.
Given their important functions, the vehicles must satisfy very demanding operational requirements. Each transporter has a mass of 150 tonnes and is able to lift and transport antennas of 110 tonnes. They must be able to place the antennas on the docking pads with millimeter precision. At the same time, they must be powerful enough to climb 2000 m reliably and safely with their heavy and valuable load, putting extraordinary demands on the 500 kW diesel engines. This means negotiating a 28 km long high-altitude road with an average slope of 7 %. Finally, as they will be operated at an altitude with significantly reduced oxygen levels, a range of redundant safety devices protect both personnel and equipment from possible mishaps or accidents. The first transporter is scheduled to be delivered in the summer of 2007 to match the delivery of the first antennas to Chajnantor . 
The Atacama Large Millimetre Array is designed to produce excellent images in spectral lines and continuum, to detect galaxies like our Milky Way at a redshift z=3, and to image proto-stellar disks in the nearest molecular cloud. To accomplish these goals, signals from the ALMA antennas must be combined. For this, one must have a back-end that is loss-free, reliable and flexible. Back-end The cooled Front End (FE) in each antenna down-converts the received signal (up to 950 GHz in 10 different receiver bands) and outputs two intermediate frequency (IF) signals per selected receiver, one for each polarization, in the 4-12 GHz bandwidth range. The Back-End (BE) transfers the signals to the Technical Building (TB) where the correlator is located and converts the signals to a form suitable for the correlator. The BE includes equipment both in the antennae and in the TB. The equipment that carries out the digitization and data transmission is the Data Transmission System (DTS). In September 2005 the first prototype of the DTS successfully underwent the complete end-to-end test, from the Digitizer Input to the Correlator Output. 
 Data transmission system The Data Transmission System in the Antenna performs two main tasks: it digitizes the processed IF ‘low’ frequency signals and converts them in a format suitable for transmission along one single optical fibre. The two 4-12 GHz signals produced by the FE receiver in the antenna are further down-converted by the IF processor to four pairs of 2-4 GHz bandwidth signals. Those signals are sampled and quantized at a 4 Gbit/s sampling rate, with 3-bit resolution. Each 3-bit 4 Gbit/s data stream is converted into a 16-bit parallel word at a rate of 250 MHz. Each pair of these 3-bit 4 Gbit/s data stream is formatted and multiplexed up to three 10 Gbit/s (some overhead is added by the communication protocol) serial streams each one driving a laser emitter. The twelve emitters are tuned to produce a slightly different ‘colour’ so as to implement a Dense Wavelength Division Modulation (DWDM) which allows the twelve light beams to be mixed together and then injected into one single optical fibre. In the Technical Building the incoming light beam is optically de-multiplexed to separate the 12 original beams. These are converted to electrical signals and fed into the correlator. In order to reliably operate at the required light frequency and with the proper power and sensitivity throughout their whole lifetime both the laser emitters and receivers need a sophisticated control system for current and temperature. The Digitizer
The heart of the DTS transmitter is the Digitizer chip, “VEGA”, a band-pass 2-4 GHz flash Analog to Digital converter operating at the sampling rate of 4 Gbit/s. Although at present the market offers components that nominally meet some of the requirements, the suitable combination of sampling rate, resolution, maximum input frequency (4 GHz) and power consumption is not available, therefore an Application Specific custom device (ASIC) needed to be developed. Low power consumption is a key requirement with respect to the reliability VEGA Specifications In order to meet the high speed and low power requirements, the digitizer uses the Si-Ge technology and BiCMOS 0.25μm process. This technology allows the fabrication of high speed and low dissipation hybrid analog/digital devices. The nominal voltage supply is 2.5V and the average power consumption is 1.5W. The sampling rate exceeds 4 Gbit/s and an external Sample/Hold is not required thanks to the short aperture time. The output is Gray-coded and the output lines are differential (LVDS). The device shows a high temporal stability (the Allan Variance tests show stable operation for more than 100s), a key requirement for Radio Astronomy. A self-test feature is implemented to allow a simple test of the device enabling an internal low frequency free-run clock.
The De-Multiplexer PHOBOS A companion chip, the de-multiplexer “PHOBOS”, was also developed. It parallelises an incoming 4 Gbit/s serial stream in 16-bit words at a rate of 250 MHz. One of these devices is connected to each VEGA output line. Despite the higher complexity and similar power dissipation constraints, the development of PHOBOS was less critical being a fully digital device with no analogue parts. An option considered during the development was the integration in one single chip of both analogue to digital (AD) conversion and de-multiplexing functions. However, the concern of keeping the analogue part as much separate as possible from the digital part and the complexity of the packaging due to the high pin count (the chip would have 3×16×2 outputs) led to two separate devices. The Development Process VEGA and PHOBOS are the result of a combined effort between the Observatoire de Bordeaux, the IXL laboratories of the Université de Bordeaux and a commercial partner. All design, simulations and qualification tests have been performed by the two institutes while the commercial partner provided the software tools and the production facilities. The development of the devices took several years. A first simplified implementation of the digitizer was designed and produced in 2002 to test the suitability of the technology to meet the requirements. The first functional 3-bit version, ALTAIR, was developed and tested in 2002. An improved version, ANTARES, optimizing power consumption and pin-out allocation was produced in 2003. The final fully engineered version, VEGA, including ESD protection, was prototyped in 2004 and mass produced in 2005. Prototyping costs have been kept affordable by making use of the Multiple Project Wafer service provided by the industrial partner where limited amount of parts can be produced sharing the silicon foundry production costs among several projects/customers and thus reducing the overall Non-Recurring Expenses (NRE). (Contributed by Fabio Biancat Marchet)
General view of construction at the Array Operation Site (AOS) at 5000m altitude.
What is the Design Reference Science Plan? The ALMA Design Reference Science Plan (DRSP) grew out of the need to have a detailed view of what the first 3-4 years of full ALMA operations will look like. Based on the projects that astronomers will want to carry out with high priority, ALMA's development can be optimized. For example, ALMA's specifications can be tested for realistic scenarios, or plans can be made which configurations or frequency bands to commission with high priority. The DRSP can also be used to determine observing strategies, data rates, and use-cases. Finally, and most crucially, the impact on the science (and ALMA's primary Science Drivers) from any changes in specifications can be quantitatively assessed. What is the DRSP not? The DRSP is not a set of observing proposals. Although they look like proposals, they will not form the basis of any kind of ALMA program, and do not imply any claims on particular observations. The DRSP is also not set in stone. Science priorities will change over time, and the DRSP is only the current reflection of what the community wants to do with ALMA. The current DRSP In total, by December 2003 128 DRSP projects were submitted for a total of ~25,000 hours, distributed over four main science areas: Galaxies & Cosmology (41% of time), Star & Planet Formation (35%), Stars & Their Evolution (10%), and Solar System (14%). These projects were written by more than 75 astronomers, and 'peer reviewed'. The results are collated at a web site (see address below). From the DRSP, one can, for example, learn that the foreseen use of receiver bands (3/6/7/9 = 20%/30%/37%/13%) is roughly consistent with expected weather statistics. While band 6 is heavily used for spectral line work, bands 7 and 9 are the most requested for continuum observations, especially for extragalactic targets. Roughly 10% of the proposals employ the total-power capability of the array. How to use the DRSP The DRSP can be accessed at the web site given below. The individual projects can be downloaded together with their review reports. Spreadsheets are also available with overviews of all program statistics. These have been used, e.g., to get estimates of the calibration requirements, or to assess the impact of various re-baselining decisions. The DRSP is a valuable resource for anyone wishing to get a realistic and detailed view of ALMA's capabilities and foreseen use. The DRSP is a living document The DRSP can only be an accurate reflection of future ALMA use if it is continuously updated. New projects can be added at all times, and existing projects can be augmented as the science questions evolve or instrument specifications change. This evolving aspect of the DRSP is crucial, because planning decisions are based on the DRSP. The DRSP is being maintained for the ALMA Science IPT by Michiel Hogerheijde, and suggestions for additional DRSP projects can be e-mailed to him at any time ( michiel@strw.leidenuniv.nl ). For more information, go to http://www.strw.leidenuniv.nl/~alma/drsp.html (Contributed by Michiel Hogerheijde)
Complex Molecules in Space: present status and prospects with ALMA To be held from 8 to 11 May 2006 in Fuglsøcentret, near Aarhus, Denmark. The deadline for registration in 31 March. For more information see: http://www.isa.au.dk/meetings/alma06/ Astronomical Telescopes and Instrumentation 2006 24 to 31 May 2006 in Orlando, Florida USA. For more information see: http://spie.org/Conferences/Calls/06/as/ Science with ALMA: a new era for Astrophysics The second ‘Global ALMA Meeting’ will be held in Madrid on 13 to 16 November 2006. This will be the first world-wide ALMA science meeting since the Washington DC meeting in 1999. The deadline for contributions is 15 June and the final registration deadline is 15 September. For more information see: Water Vapor Radiometry There are plans for a workshop on water vapor radiometry in Cambridge, UK, toward the end of 2006. |
The signing ceremony took place on December 6, 2005 at ESO Headquarters in Garching, Germany. "This contract represents a major milestone. It allows us to move forward, together with our American and Japanese colleagues, in this very ambitious and unique project," said ESO's Director General, Dr. Catherine Cesarsky. "By building ALMA, we are giving European astronomers access to the world's leading submillimetre facility at the beginning of the next decade, thereby fulfilling Europe's desire to play a major role in this field of fundamental research."
issue: lower dissipation means lower operating temperature which leads to a longer lifetime and hence higher reliability and less maintenance. The power dissipation is a concern as the low air density at the operating altitude makes the ventilation rather inefficient, especially considering that the digitizer assembly, to reduce RFI, needs to be encapsulated in a sealed case. 
The chip is embedded in a 44-Pin VFQFPN (Very Fine pitch Quad Flat Pack No-lead) package (7x7x1mm) that provides a dissipation lead that reduces the thermal resistance between the die and the environment. A synthetic figure of merit combining resolution, speed and power consumption for flash converters is given by the formula: 
Preben Grosbøl 
Bogdan Jeram
Alessandro Caproni