A new kind of bolometric architecture has been successfully developed for the PACS photometer onboard the Herschel submillimeter observatory. These new generation CCD-like arrays are buttable and enable the conception of large fully sampled focal planes. We present a feasibility study of the adaptation of these bolometer arrays to ground-based submillimeter telescopes. We have developed an electro-thermal numerical model to simulate the performances of the bolometers under specific ground-based conditions (different wavelengths and background powers for example). This simulation permits to determine the optimal parameters for each condition and shows that the bolometers can be background limited in each transmission window between 200 and 450 microns. We also present a new optical system that enables to have a maximum absorption of the bolometer in each atmospheric windows. The description of this system and measurements are shown.

Astronomical observations at sub-millimetre wavelengths are limited either by the angular resolution of the telescope or by the sensitivity and field of view of the detector array. New generation of radio telescopes, such as the ALMA-type antennas on Chajnantor plateau in Chile, can overcome these limitations if they are equipped with large detector arrays made of thousands of sensitive bolometer pixels. Instrumentation developments undertaken at CEA and based on the all silicon technology of CEA/Leti are able to provide such large detector arrays. The ArTeMiS project consists in developing a camera for ground-based telescopes that operates in two sets of atmospheric windows at 200-450 μm (channel 1) and 800-1200 μm (channel 2). ArTeMiS-1 consists in grid bolometer arrays similar to those developed by CEA for the Herschel Space Observatory. A prototype camera operating in this first atmospheric window was installed and successfully tested in March 2006 on the KOSMA telescope at Gornergrat (Switzerland) in collaboration with the University of Cologne. ArTeMiS-2 will consist either in antenna-coupled bolometer arrays or specific mesh bolometer arrays. By the end of 2008, ArTeMiS cameras could be operated on 10m-class telescopes on the Chajnantor ALMA site, e.g., APEX, opening new scientific prospects in the study of the early phases of star formation and in cosmology, in the study of the formation of large structures in the universe. At longer term, installation of such instrumentation at Dome-C in Antarctica is also under investigation. The present status of the ArTeMiS project is detailed in this paper.

High sensitivity submillimeter-wave focal plane array using SIS photon detector is being developed, which we call SISCAM, the superconductive imaging submillimeter-wave camera. In the course of the detector evaluations, we have measured performance of the SIS photon detectors under various operating conditions. Advantages of the SIS photon detectors are explained by the nature of antenna coupled quantum detectors. Their input coupling can be designed to have band-pass characteristics owing to the distributed junction design. This reduces requirements for infrared blocking filters and enhances optical efficiency. The detector performance is evaluated under background loading and they show background limited performance. Measurement at 4 K shows the SIS photon detector operates under shot noise limit of thermal leakage current and its NEP is 1x10^-14 W/Hz^0.5, that is better than bolometers at 4.2 K, whereas the same detector has NEP of 10^-16 W/Hz^0.5 at 0.3 K. Dynamic range of SIS photon detectors is estimated to be higher than 10⁹, which surpass the dynamic range achievable with TES bolometers. Nine-element array of SIS photon detector, SISCAM-9, is developed and their performance is evaluated in a submillimeter-wave telescope. With a development of integrated electronics with GaAs-JFET charge integrating readout circuit, the SIS photon detector will be an ideal imaging array in submillimeter-wave region. Due to its large dynamic range and shot noise limited performance under various operating condition, SIS photon detectors can be used for various astronomical instrumentations as well as for other fields of terahertz technologies.

Photoconductivity spectra of unstressed and stressed Ge:Ga detectors were measured. The experiments were performed with a polarizing step scan Fourier transform spectrometer using the synchrotron facility BESSY, which was operated in a dedicated mode with a low momentum compaction factor. By this way powerful and coherent synchrotron radiation below 50 cm^-1 was generated. We observed a significant response of unstressed and stressed Ge:Ga detectors below 50 cm^-1 and 25 cm^-1, respectively. This response can be attributed to transitions between bound excited states or from bound excited states to the valence band. The results indicate that in germanium detectors a fraction of the recombining holes is captured into bound excited states.

We are developing a hot-electron superconducting transition-edge sensor (TES) that is capable of counting THz photons and operates at T = 0.3K. The main driver for this work is moderate resolution spectroscopy (R ~ 1000) on the future space telescopes with cryogenically cooled (~ 4 K) mirrors. The detectors for these telescopes must be background-limited with a noise equivalent power (NEP) ~ 10^-19-10^-20 W/Hz^1/2 over the range ν=0.3-10 THz. Above about 1 THz, the background photon arrival rate is expected to be ~ 10-100 s^-1, and photon counting detectors may be preferable to an integrating type. We fabricated superconducting Ti nanosensors with a volume of ~ 3×10^-3 μm³ on planar Si substrate and have measured the thermal conductance G to the thermal bath. A very low G=4×10^-14 W/K, measured at 0.3 K, is due to the weak electron-phonon coupling in the material and the thermal isolation provided by superconducting Nb contacts. This low G corresponds to NEP(0.3K) = 3×10^-19 W/Hz^1/2. This Hot-Electron Direct Detector (HEDD) is expected to have a sufficient energy resolution for detecting individual photons with ν > 0.3 THz at 0.3 K. With the sensor time constant of a few microseconds, the dynamic range is ~ 50 dB.

Microstrip-coupled Transition Edge Sensors (TESs) are important because they can be combined with waveguide-horn technology to produce sensitive bolometric detectors with well-defined, single-mode beam patterns and polarisation characteristics. They also allow superconducting RF filters to be included on the detector chips. Our own design of TES uses a finline taper to transform between waveguide and superconducting Nb microstrip. The microstrip transports the signal to a matched Au-Cu resistor, which is deposited on a thermally isolated SiN membrane. The dissipated RF power causes the resistance of a Mo-Cu TES bilayer to increase, and the resulting reduction in bias current is read out by a SQUID. We have fabricated TES bilayers with critical temperatures of 400 to 600mK, and deduced dark NEPs as low as 3x10^-17W/√Hz at 150GHz. In this paper we describe a number of experiments that were carried out in order to investigate the electrothermal behaviour of microstrip-coupled TESs. We show that the electrothermal behaviour of microstrip-coupled TESs can be as good as that of free-space TESs, and therefore that they are suitable for high-performance astronomical applications.

To determine the lowest attainable phonon noise equivalent power (NEP) for membrane-isolation bolometers, we fabricated and measured the thermal conductance of suspended Si₃N₄ beams with different geometries via a noise thermometry technique. We measured beam cross-sectional areas ranging from 0.35 × 0.5 μm² to 135 × 1.0 μm² and beam lengths ranging from 700 μm to 8300 μm. The measurements directly imply that membrane-isolation bolometers are capable of reaching a phonon noise equivalent power (NEP) of 4×10^-20 W/Hz^1/2. This NEP is adequate for the Background-Limited Infrared-Submillimeter Spectrograph (BLISS) proposed for the Japanese SPICA observatory, and adequate for NASA's SAFIR observatory, a 10-meter, 4 K telescope to be deployed at L2. Further, we measured the heat capacity of a suspended Si₃N₄ membrane and show how this result implies that one can make membrane-isolation bolometers with a response time which is fast enough for BLISS.

Excess phase noise has been observed in microwave kinetic inductance detectors (MKIDs) which prevents the noise-equivalent power (NEP) of current detectors from reaching theoretical limits. One characteristic of this excess noise is its dependence on the power of the readout signal: the phase noise decreases as the readout power increases. We investigated this power dependence in a variety of devices, varying the substrate (silicon and sapphire), superconductor (aluminum and niobium) and resonator parameters (resonant frequency, quality factor and resonator geometry). We find that the phase noise has a power law dependence on the readout power, and that the exponent is -1/2 in all our devices. We suggest that this phase noise is caused by coupling between the high-Q microwave resonator that forms the sensitive element of the MKID and two-level systems associated with disorder in the dielectric material of the resonator. The physical situation is analogous to the resonance fluorescence in quantum optics, and we are investigating the application of resonance fluorescence theory to MKID phase noise.

We are developing a large-format, versatile, bolometer array for a wide range of infrared through millimeter astronomical applications. The array design consists of three key components - superconducting transition edge sensor bolometer arrays, quarter-wave reflective backshort grids, and Superconducting Quantum Interference Device (SQUID) multiplexer readouts. The detector array is a filled, square grid of bolometers with superconducting sensors. The backshort arrays are fabricated separately and are positioned in the etch cavities behind the detector grid. The grids have unique three-dimensional interlocking features micromachined into the walls for positioning and mechanical stability. The ultimate goal of the program is to produce large-format arrays with background-limited sensitivity, suitable for a wide range of wavelengths and applications. Large-format (kilopixel) arrays will be directly indium bump bonded to a SQUID multiplexer circuit. We have produced and tested 8×8 arrays of 1 mm detectors to demonstrate proof of concept. 8×16 arrays of 2 mm detectors are being produced for a new Goddard Space Flight Center instrument. We have also produced models of a kilopixel detector grid and dummy multiplexer chip for bump bonding development. We present detector design overview, several unique fabrication highlights, and assembly technologies.

The Millimeter Bolometer Array Camera (MBAC) is sheduled for installation on the 6-meter Atacama Cosmology Telescope (ACT) in 2007. MBAC will eventually contain three diffraction-limited, 1024-pixel, focal plane arrays of Transition Edge Sensor (TES) bolometers developed at NASA Goddard Space Flight Center. We present instrument designs and results from optical and electrical measurements made with the MBAC prototype instrument, the Column Camera (CCam). Initial studies with CCam include measurements of the properties of TES detectors, the time-domain SQUID multiplexing readout system developed at NIST Boulder, the Cryoperm magnetic shielding, and preliminary optical measurements with the detectors. We also discuss the status of first light measurements with CCam mounted on a 1.5-meter telescope.

We describe a miniature dilution refrigerator (MDR), operated in continuous mode and suitable for many detector applications at temperatures down to 50 mK. It distinguishes itself from other refrigerators in that it is self-contained and benefits from an internal cycle of the ³He gas. As a result, no external gas handling system is required so size, weight and complexity of the system is dramatically decreased. The system has no fine capillaries, moving parts or cooled O-rings. It is therefore mechanically very reliable, has no risk of blockages and is unlikely to develop cryogenic leaks. One direct application is balloon-borne or ground-based observations of the CMB using large detector arrays. When these experiments are operated remotely on platforms or at sites with limited infrastructure and maintenance support, a compact and reliable dilution refrigerator becomes essential. We describe a complete system incorporating an MDR which we have built and integrated with a pulse-tube refrigerator to achieve a cooling power of several micro Watt at 100 mK. This system is being developed for a CMB polarization experiment (CLOVER) which requires three independent cryostats to cool large TES detector arrays.

The ESI instrument (European SPICA Instrument) is a proposed imaging spectrometer for the 30-210μm band for the JAXA SPICA mission. The instrument will have unprecedented spatial resolution and sensitivity due to the large 03.5m telescope aperture, cold fore-optics (~5K) and high sensitivity detectors (NEP~10^-19W/√Hz). One of the key technical challenges of the design of the instrument is the thermal architecture due to the mass and cryogenic heat load constraints and the need for very low temperatures. Two candidate detector technologies have been pre-selected for inclusion in the instrument Phase-A study; Photoconductors and TES Bolometers. An overview of thermal architecture of the SPICA spacecraft is presented in order to explain the thermal interface constraints imposed on the instrument. Proposed thermal architectures for the instrument for both the TES and the Photoconductor options will be outlined including a novel design for a lightweight hybrid cooler for achieving sub 100-mK detector temperatures. This novel cooler architecture utilizes a combination of ADR and sorption coolers. Several design solutions for achieving high thermal isolation generic to both detector options are presented.

Thermal performance testing of a large cryogenic system of an open geometry (not contained within a dewar) is problematic due to the proximity of room temperature and very low temperature components. These tests require careful attention to the closeouts between temperature zones to prevent unwanted radiation transfer and to avoid perturbing the system under test. We are demonstrating the possibility of performing these sensitive tests on a small scale which makes the tests faster and cheaper. It also offers the possibility of refining these close-outs, which may not be practical on a full size system. This paper describes a test item modeled on a larger passive and actively cooled future space telescope, and the apparatus used to test it down to 4 K.

APEX, the Atacama Pathfinder Experiment, is collaboration between Max Planck Institut fur Radioastronomie (MPIfR) with Astronomisches Institut Ruhr Universitat Bochum, Onsala Space Observatory and the European Southern Observatory (ESO). The telescope was supplied by VERTEX Antennentechnik in Duisburg, Germany, and is a 12 m antenna with 15 μm rms surface accuracy operating at the Atacama Desert Llano Chajnantor, in the Chilean Andes at 5100 m altitude. APEX heterodyne single pixel facility receiver are placed in the telescope Nasmyth cabin A. The receivers are coupled to the antenna via relay optics providing possibility to operate either one of the two different PI-type instruments or a multi-channel facility heterodyne receiver to cover 211 - 1500 GHz frequency range. In this report, we present the optical design for APEX single-pixel facility heterodyne receiver providing frequency independent illumination of the secondary for all the receiver channels. We present design of the two-channel facility receiver APEX A, installed and operating since June 2005, and of the coming 6-channel APEX facility receiver. The report includes a brief review of the mixer technology development status for APEX Band 1, 211 - 270 GHz, using sideband separation technology (2SB), Band 2, 270 - 370 GHz, 2SB, Band 3, 385 - 500 GHz, 2SB, and Band T2, 1250 - 1390 GHz, HEB waveguide balanced mixer, those on the development at Onsala Space Observatory. We present description of the receiver control system and example observation of APEX 2a receiver.

Quantum cascade lasers (QCLs) operating at 2.5 THz have been used for gas phase spectroscopy and as local oscillator in a heterodyne receiver. One QCL has a Fabry-Perot resonator while the other has a distributed feedback resonator. The linewidth and frequency tunability of both QCLs have been investigated by either mixing two modes of the QCL or by mixing the emission from the QCL with the emission from a 2.5 THz gas laser. The frequency tunability as well as the linewidth is sufficient for Doppler limited spectroscopy of methanol gas. The QCLs have been used successfully as local oscillators in a heterodyne receiver. Noise temperature measurements with a hot electron bolometer and a QCL yielded the same result as with a gas laser as local oscillator.

A 16 pixel heterodyne receiver for 2.5 THz has been developed based on NbN superconducting hot-electron bolometer (HEB) mixers. The receiver uses a quasioptical RF coupling approach where HEB mixers are integrated into double dipole antennas on 1.5μm thick Si₃N₄/SiO₂ membranes. Spherical mirrors (one per pixel) and backshort distance from the antenna have been used to design the output mixer beam profile. The camera design allows all 16 pixel IF readout in parallel. The gain bandwidth of the HEB mixers on Si₃N₄/SiO₂ membranes was found to be 0.7÷0.9 GHz, which is much smaller than for similar devices on silicon. Application of buffer layers and use of alternative types of membranes (e.g. silicon-on-insulator) is under investigation.

The Atacama Pathfinder EXperiment (APEX) is a 12 m antenna operating at the Atacama Desert on the Chilean Andes at about 5000 m altitude. APEX would be equipped with a suit of single-pixel heterodyne receivers covering 211 - 1500 GHz frequency range. We present here a design of a sideband-separating superconductor-insulator-superconductor (SIS) mixer for the APEX, receiver Band 3, operating in 385 - 500 GHz band. The receiver uses quadrature scheme with the RF signal passing via a 90-degree waveguide 3 dB hybrid and the LO is divided by a waveguide E-plane Y-junction. The outputs of the waveguide hybrid are coupled to the mixer SIS junctions through an E-probe with integrated bias-T. For the LO coupler, conventional branch waveguide couplers are difficult to manufacture at this high frequency with required accuracy as the branch waveguides become extremely narrow. In order to solve this problem, we propose an on-chip LO injection, where the LO coupler is integrated onto the mixer chip and fabricated together with the SIS junction and the tuning circuitry. The on-chip LO coupler is made of superconducting lines, which gives almost a lossless solution and provides fabrication accuracy better than 0.5 μm by using optical lithography only. Furthermore, the mixer design includes a novel component, an ellipse termination for the idle LO port, made of thin-film resistive material with sheet resistance equal to the transmission line characteristic impedance, which gives very broadband performance using extremely compact area.

GREAT, the German REceiver for Astronomy at Terahertz frequencies, is a first generation SOFIA dual channel heterodyne PI-instrument for high resolution spectroscopy. The system is developed by a consortium of German research institutes. The receiver will allow simultaneous observations in two out of the following three far-infrared frequency bands: a 1.4-1.9 THz channel for e.g. the fine-structure line of ionized carbon [CII] at 158μm; a 2.4-2.7 THz channel for e.g. the 112μm transition of HD; and a 4.7 THz channel for the 63 μm fine-structure line of neutral atomic oxygen. Hot electron bolometers (HEB) mixers provide state of the art sensitivity. A spectral resolving power of up to 10⁸ is achieved with chirp transform spectrometers, and a total bandwidth of 4 GHz at 1 MHz resolution is reached with wide band acousto-optical spectrometers. The modular concept of GREAT allows to observe with any combination of two out of the three channels aboard SOFIA. A more complete frequency coverage of the THz regime by adding additional GREAT channels is possible in the future. The adaptation of new LO-, mixer- or backend-techniques is easily possible. We describe details of the receiver and the results of first performance tests of the system at 1.9 THz.

CHAMP⁺, a dual-color 2 × 7 element heterodyne array for operation in the 450 μm and 350 μm atmospheric windows is under development. The instrument, which is currently undergoing final evaluation in the laboratories, will be deployed for commissioning at the APEX telescope in August this year. With its state-of-the-art SIS detectors and wide tunable local oscillators, its cold optics with SSB filters and with 2 GHz of usable IF bandwidth per pixel, CHAMP⁺ will provide unmatched observing capabilities for the APEX community. The optics allows for simultaneous observations in both colors. For both sub-arrays a hexagonal arrangement with closest feasible spacing of the pixels on sky (2×Θmb) was chosen, which, in scanning mode, will provide data sampled with half-beam spacing. The front-end is connected to a flexible autocorrelator array with a total bandwidth of 32 GHz and 32768 spectral channels, subdivided into 32 IF bands of 1 GHz and 1024 channels each.

We report on the development of SuperCam, a 64 pixel, superheterodyne camera designed for operation in the astrophysically important 870 μm atmospheric window. SuperCam will be used to answer fundamental questions about the physics and chemistry of molecular clouds in the Galaxy and their direct relation to star and planet formation. The advent of such a system will provide an order of magnitude increase in mapping speed over what is now available and revolutionize how observational astronomy is performed in this important wavelength regime. Unlike the situation with bolometric detectors, heterodyne receiver systems are coherent, retaining information about both the amplitude and phase of the incident photon stream. From this information a high resolution spectrum of the incident light can be obtained without multiplexing. SuperCam will be constructed by stacking eight, 1×8 rows of fixed tuned, SIS mixers. The IF output of each mixer will be connected to a low-noise, broadband MMIC amplifier integrated into the mixer block. The instantaneous IF bandwidth of each pixel will be ~2 GHz, with a center frequency of 5 GHz. A spectrum of the central 500 MHz of each IF band will be provided by the array spectrometer. Local oscillator power is provided by a frequency multiplier whose output is divided between the pixels by using a matrix of waveguide power dividers. The mixer array will be cooled to 4K by a closed-cycle refrigeration system. SuperCam will reside at the Cassegrain focus of the 10m Heinrich Hertz telescope (HHT). A prototype single row of the array will be tested on the HHT in 2006, with the first engineering run of the full array in late 2007. The array is designed and constructed so that it may be readily scaled to higher frequencies.

Active surface correction of the Caltech Submillimeter Observatory (CSO) primary mirror has been accomplished. The Dish Surface Optimization System (DSOS) has been designed and built to operate at the CSO, on Mauna Kea, Hawaii. The DSOS is the only active optics system of its kind in the world. There are 99 steel rod standoffs that interface the dish panels to its backing structure. Each standoff is now fitted with a heating/cooling assembly. Applying a controlled potential to each of the 99 assemblies adjusts the surface of the dish. Heating elongates and cooling shortens the standoffs, providing the push or pull on the primary's panel surface. The needed correction for each standoff, for a given elevation, is determined from prior holography measurements of the dish surface. Without the DSOS the optimum surface accuracy was 25-μm RMS, yielding a beam efficiency of 33% at the 350-μm-wavelength range. With the DSOS on, this has been improved to 10-μm RMS. The best beam efficiency obtained is 56%, with an average beam efficiency of 53%. The DSOS has been in operation on the CSO since February 2003. Observers using the SHARCII (a 384 pixel submillimeter high angular resolution camera) and the 850 GHz heterodyne receiver, have been able to detect new weak and/or distant objects including detection of an earth-massed planet in Fomalhaut with the help of this unique active optics system.

Microwave, submillimetre-wave, and far-infrared phased arrays are of considerable importance for astronomy. We consider the behaviour imaging phased arrays and interferometric phased arrays from a functional perspective. It is shown that the average powers, field correlations, power fluctuations, and correlations between power fluctuations at the output ports of an imaging or interferometric phased array can be found once the synthesised reception patterns are known. The reception patterns do not have to be orthogonal or even linearly independent. It is shown that the operation of phased arrays is intimately related to the mathematical theory of frames, and that the theory of frames can be used to determine the degree to which any class of intensity or field distribution can be reconstructed unambiguously from the complex amplitudes of the travelling waves at the output ports. The theory can be used to set up a likelihood function that can, through Fisher information, be used to determine the degree to which a phased array can be used to recover the parameters of a parameterised source. For example, it would be possible to explore the way in which a system, perhaps interferometric, might observe two widely separated regions of the sky simultaneously.

We describe a modal theory of interferometry, suitable for the modelling of multimode bolometric interferometers, and present the first simulations of such an interferometer. The motivation for the work is the wish to combine the low noise properties of bolometers with aperture synthesis techniques to design far-infrared and sub-mm interferometers, without the need for single-mode components that restrict the power throughput and therefore reduce the signal to noise ratio. The analysis of such interferometers has proved impossible in the past because optical systems at sub-mm wavelengths are partially coherent, and bolometers respond not only to the intensity but also to the correlations in the field. In order to assess the viability of bolometric interferometers, we therefore require a multimode theory of interferometry. We show that the appropriate modes for describing an interferometer are the Hilbert-Schmidt decompositions of the kernels of the integral operators describing the individual telescopes, and demonstrate that these modes provide both a clear conceptual understanding of the operation of an interferometer and an extremely fast method of computation. We present simulations of idealized Michelson and Fizeau interferometers, and show that the normal behaviour of such interferometers is recovered. We show simulated sources, and dirty maps obtained from uv data for a simulated Michelson interferometer. We discuss the application of the theory to real instrument design.

The characterization and calibration of far-infrared (FIR) detectors is a delicate task that requires good knowledge of the incident flux and its spectral composition. In many test setups the FIR flux to the detectors is provided by means of an external or internal black body and a set of cold attenuation, band pass, and blocking filters. For scientific instruments (e.g. PACS aboard ESA's Herschel satellite) band pass and blocking filters are used to achieve the desired spectral throughput either as order sorting filters in spectrometers or for selecting a wavelength range in imaging cameras. In all cases a detailed knowledge of the spectral transmittance of the used filters is mandatory for an accurate calibration of the system. We have build a test platform that allows to measure the transmission of cold (T ~ 4K) filters in the far-infrared. The setup uses a dual grating monochromator with excellent spectral purity and a resolution up to 800, which is operated under a dry nitrogen atmosphere to eliminate water vapor absorption bands. An Si-bolometer is used as detector and is read out by a cryogenic low noise trans-impedance amplifier circuit with common mode rejection and a warm electronics using a lock-in amplifier and a 22 bit analog-to-digital converter. A cryogenic filter slider in the setup allows for differential measurements between filters and the use of cold order sorting filters. We present initial results for FIR cut-on and attenuation filters, demonstrating that our setup is suited to measure transmissions as low as 10^-4 over the covered wavelength range.

Recent advances in the development of large format detector arrays for wide field astronomical applications have required the use of large aperture cryostat windows. An important consequence is the significant increase in thermal loading on the cryogens and in the stray thermal emission reaching the detectors. We have identified that a significant part of this unwanted radiation comes from re-emission of thermal power absorbed by the dielectric substrate of the metal mesh rejection filters. To overcome this we have developed a thermal filter, which preferentially reflects radiation near the 10μm thermal peak. These filters have essentially no absorption, and hence negligible emission, at these wavelengths whilst allowing high transmission of the wanted sub-millimetre bands. We report here on thermal problems with existing metal mesh filter technology and give performance data for the new filters. As a proof of concept we present data for a medium aperture instrument (QUaD¹), which utilises a 200mm diameter window, both with and without thermal filters.

The Astronomical Instrumentation Group at Cardiff University has been developing metal mesh optical filters for more than 30 years, which are currently in use in many ground-, balloon- and space-based instruments. Here we review the current state of the art with respect to these quasi-optical components (low-pass, high-pass and band-pass filters, dichroics and beam-dividers) as developed for the FIR and sub-millimetre wavelength region. We compare performance data with various modelling tools (HFSS, transmission line theory or Floquet mode analysis). These models assist with our understanding of the behaviour of these filters when used at non-normal incidence or in the diffraction region of the grid structures. Interesting artefacts, such as the Wood anomalies and behaviour with S and P polarisations, which dictate the usage of these components in polarisation sensitive instruments, will be discussed.

Filled arrays of planar bolometers are finding astronomical applications at wavelengths as long as several millimeters. In order to minimize detector size while maintaining element number and performance it is common to push these arrays to scales approaching the single-mode limit. Doing so introduces several new challenges that are not experienced with multi-mode far-infrared detectors having similar pixel size. First, diffractive effects by the pixels themselves are no longer insignificant and will ultimately limit the resolution and polarization response of the optical system. Second, it is necessary to re-examine the coupling between the incident radiation and the individual bolometer absorbing elements. The relatively low f-numbers often employed with millimeter wavelength focal planes can make ideal backshort performance and construction problematic. In addition, typical methods for stray light control that rely on multiple reflections in a lossy medium can fail due to physical size constraints. For these applications, we find resonant absorbers and anti-reflection coatings as effective strategies which can be implemented in the available focal plane area to control stray light.

We present a new theory for the description of detectors in terms of modes. Although the theory is very general, it is expected to be of particular use in the modelling of far-infrared and submillimeter instruments. Such a theory is needed because at far-infrared frequencies, both optical systems and detectors show partially coherent behaviour. That is to say, even when the instrument is illuminated by an incoherent source, the resultant field at the detector is partially coherent, and the detector itself is sensitive to the coherence properties of the field and not just the intensity. We have previously developed a modal description of optical systems at far-infrared wavelengths; here we describe a modal theory for the detectors themselves. The theories can be combined to provide a complete modal description of far-infrared instruments. The theory presented here applies equally well to pulsed or ergodic radiation, and incorporates polarisation effects. We also show how the statistics of the detector output can be determined from the theory. This is important for instruments such as bolometers, where the internal noise of the detector is very low, and either sky noise or radiation from the optical components dominate. We illustrate our work with a number of simulations, showing signal to noise ratios for different detector types, and we show how how the theory may be applied to the array packing density problem.

The Millimeter-Wave Bolometric Interferometer (MBI) is designed for sensitive measurements of the polarization of the cosmic microwave background (CMB). MBI combines the differencing capabilities of an interferometer with the high sensitivity of bolometers at millimeter wavelengths. It views the sky directly through corrugated horn antennas with low sidelobes and nearly symmetric beam patterns to avoid spurious instrumental polarization from reflective optics. The design of the first version of the instrument with four 7-degree-FOV corrugated horns (MBI-4) is discussed. The MBI-4 optical band is defined by filters with a central frequency of 90 GHz. The set of baselines determined by the antenna separation makes the instrument sensitive to CMB polarization fluctuations over the multipole range l=150-270. In MBI-4, the signals from antennas are combined with a Fizeau beam combiner and interference fringes are detected by an array of spider-web bolometers with NTD germanium thermistors. In order to separate the visibility signals from the total power detected by each bolometer, the phase of the signal from each antenna is modulated by a ferrite-based waveguide phase shifter. Observations are planned from the Pine Bluff Observatory outside Madison, WI.

We report on the current progress of the water vapor radiometer (WVR) phase correction project for the Combined Array for Research in Millimeter-wave Astronomy (CARMA). CARMA is a new millimeter array that merges the Owens Valley Radio Observatory (OVRO) array, the Berkeley-Illinois-Maryland Association (BIMA) array and eventually the Sunyaev-Zel'dovich Array (SZA). WVRs are designed for phase correction by monitoring the water vapor in the atmosphere along the line of sight toward astronomical sources. In addition, we discuss the stability of the current OVRO water vapor radiometers in preparation for testing at the CARMA site. We will systematically analyze the receivers with atmospheric correlations to decouple the effects of instrumentation and atmospheric noise. Finally, we report on the status of the correlation receivers in development.

With the advent of large-format submillimeter wavelength detector arrays, and a new 25 m diameter submillimeter telescope under consideration, the question of optimal wavefront sensing methods is timely. Indeed, not only should bolometric array detectors allow the use of a variety of wavefront sensing techniques already in use in the optical/infrared, but in some cases it should actually be easier to apply these techniques because of the more benign temporal properties of the atmosphere at long wavelengths. This paper thus addresses the fundamental limits to wavefront sensing at submillimeter wavelengths, in order to determine how well a telescope surface can be measured in the submillimeter band. First several potential measurement approaches are discussed and compared. Next the theoretical accuracy of a fringe phase measurement in the submillimeter is discussed. It is concluded that with Mars as the source, wavefront sensing at the micron level should be achievable at submillimeter wavelengths in quite reasonable integration times.

We report on the status of Z-Spec, including preliminary results of our first astronomical measurements. Z-Spec is a cryogenic, broadband, millimeter-wave grating spectrometer designed for molecular line surveys of galaxies, including carbon monoxide redshift measurements of high-redshift submillimeter sources. With an instantaneous bandwidth of 185-305 GHz, Z-Spec covers the entire 1 mm atmospheric transmission window with a resolving power of 200-400. The spectrometer employs the Waveguide Far-Infrared Spectrometer (WaFIRS) architecture, in which the light propagation is confined within a parallel-plate waveguide, resulting in a minimum mechanical envelope. Its array of 160 silicon-nitride micromesh bolometers is cooled to below 100 mK for background-limited performance. With its sensitivity, broad bandwidth, and compactness, Z-Spec serves as a prototype for a future far-IR spectrometer aboard a cold telescope in space. Z-Spec successfully demonstrated functionality with a partial array of detectors and warm electronics during a week-long engineering run at the Caltech Submillimeter Observatory in June, 2005. We describe the instrument performance evaluated at the telescope and in subsequent laboratory tests and compare these results with design specifications. Following several modifications we returned to the telescope in April, 2006. We present a preliminary astronomical spectrum and discuss our plans to improve sensitivity and throughput to achieve our ultimate science goals.

We present a new generation of very flexible and sensitive spectrometers for radio astronomical applications: Fast Fourier Transform Spectrometer (FFTS). The rapid increase in the sampling rate of commercially available analog-to-digital converters (ADCs) and the increasing power of field programmable gate array (FPGA) chips has led to the technical possibility to directly digitize the down-converted intermediate-frequency signal of coherent radio receivers and to Fourier transform the digital data stream into a power spectrum in continuous real-time with no gaps in the data. In the last years FPGAs have become very popular for building fast and reconfigurable hardware. State-of-the-art chips include several hundred dedicated 18 bit×18 bit multipliers, which allow up to 80 billion multiplication and nearly 500 billion 36-bit additions per second. This extremely high computing power makes it now possible to implement real-time FFTs to decompose a 1GHz frequency band into 16384 spectral channels. In this paper we present the technological concept and results of our novel broad-band 1GHz FFTS with 16k frequency channels, which is installed at the APEX telescope. This backend can be considered prototypical for spectrometer developments for future radio astronomical applications.

The ESA Planck mission is the third generation (after COBE and WMAP) space experiment dedicated to the measurement of the Cosmic Microwave Background (CMB) anisotropies. Planck will map the whole CMB sky using two instruments in the focal plane of a 1.5 m off-axis aplanatic telescope. The High Frequency Instrument (HFI) is an array of 52 bolometers in the frequency range 100-857 GHz, while the Low Frequency Instrument (LFI) is an array of 11 pseudo-correlation radiometric receivers which continuously compare the sky signal with the reference signal of a blackbody at ~ 4.5 K. The LFI has been tested and calibrated at different levels of integration, i.e. on the single units (feed-horns, OMTs, amplifiers, waveguides, etc.), on each integrated Radiometric Chain Assembly (RCA) and finally on the complete instrument, the Radiometric Array Assembly (RAA). In this paper we focus on some of the data analysis algorithms and methods that have been implemented to estimate the instrument performance and calibration parameters. The paper concludes with the discussion of a custom-designed software package (LIFE) that allows to access the complex data structure produced by the instrument and to estimate the instrument performance and calibration parameters via a fully graphical interface.

The design of the 300 mK system for Herschel-SPIRE is complex, with many difficult, sometimes conflicting, requirements and constraints placed upon it. Five detector arrays, mounted from a 2 K box, are linked to a single ³He sorption-cooler tip by a high-conductance copper strap network. This strap retains high thermal conductance, even though it incorporates an electrical break to comply with the SPIRE grounding scheme. It requires stiffness to withstand launch vibrations, but needs compliance to avoid transmission of loads to the detector arrays. The strap is stiffly supported by novel, compact cryogenic stand-offs which provide a high degree of thermal isolation from the 2 K stage. An additional complication is that the detectors reside in a 2 K environment, whilst the cooler tip is in a 4 K environment. Two of the cryogenic stand-offs also act as light-tight feed-throughs to pass the strap from the 4 K environment to the inside of the 2 K detector boxes. Active thermal control is provided on the 300 mK system to address the detector stability requirements. This paper describes the system, and gives results of the performance in SPIRE flight model ground tests.

The SPIRE instrument for the Herschel Space Observatory has two on-board calibration sources. The photometer calibrator, PCAL, is an electrically-heated thermal source which can be seen by all detectors, including the spectrometer. It is not an absolute calibrator, but a repeatable source of sub-millimetre radiation which may be used in the overall calibration scheme. The purpose of the spectrometer calibrator is to null the background emission from the Herschel telescope, thereby improving the dynamic range of the spectrometer detectors. This paper details the final flight design of the calibration units, and presents results from SPIRE flight model ground testing.

The ESA Herschel space observatory will be launched in 2008 into the Earth-Sun L2 orbit and the three instruments onboard will be exposed to cosmic radiation during the 4 years lifetime of the satellite. To study the impact of ionizing radiation on the Ge:Ga photoconductors of the PACS instrument (Photodetector Array Camera and Spectrometer), we performed a series of irradiation measurements at the cyclotron of the University of Louvain la Neuve, Belgium simulating the in-flight predicted proton fluxes including solar flare events. The PACS integral field spectrometer contains two 25×16 pixel arrays of Ge:Ga crystals: a low stressed configuration is used in the wavelength range from 55 to 105 μm, and a high stressed device covers the range 105 to 210 μm. Calibration of the detector modules under realistic IR background fluxes is done at MPE Garching and MPIA Heidelberg. 70 MeV protons were generated at the cyclotron test site. They were attenuated on their way to the detectors by beam conditioning elements and the metal shields of the cryostat before they reached the Ge:Ga crystals with a mean energy of 17 MeV and a standard deviation of 1.5 MeV. According to predictions the expected proton fluxes were set to nominally 10 ps^-1cm^-2 and to 400 ps^-1cm^-2 simulating solar flares. We observed radiation-induced glitches in the detector signal, changes in responsivity, increase in noise and transient behavior. The ongoing data evaluation indicates optimal operating parameters, the best curing method and frequency, calibration procedures and data processing algorithms aiming for a high photometric accuracy.

In the framework of the Photodetector Array Camera and Spectrometer (PACS) project IMEC designed the Cold Readout Electronics (CRE) for the Ge:Ga far-infrared detector array. Key specifications for this circuit were high linearity, low power consumption and low noise at an operating temperature of 4.2K. We have implemented this circuit in a standard CMOS technology which guarantees high yield and uniformity, and design portability. A drawback of this approach is the anomalous behavior of CMOS transistors at temperatures below 30-40K. These cryogenic phenomena disturb the normal functionality of commonly used circuits. We were able to overcome these problems and developed a library of digital and analog building blocks based on the modeling of cryogenic behavior, and on adapted design and layout techniques. We will present the design of the 18 channel CRE circuit, its interface with the Ge:Ga sensor, and its electrical performance. We will show how the library that was developed for PACS served as a baseline for the designs used in the Darwin-far-infrared detector array, where a cryogenic 180 channel, 30μm pitch, Readout Integrated Circuit (ROIC) for flip-chip integration was developed. Other designs and topologies for low noise and low power applications will be equally presented.

We describe the performance of the Band 3 and Band 4 Flight Model mixer units for Herschel/HIFI Instrument. These units are part of the Focal Plane Unit of HIFI. The band 3 and 4 mixer units cover the 800-960 GHz and 960-1120 GHz frequency range and have a 4-8 GHz IF frequency band. The sensitivities of the mixers within the HIFI setting are excellent and are the best reported to date. The DSB receiver noise performance in the HIFI FPU environment ranges from 150 K at 800 GHz to 350 K at 1120 GHz. This sensitivity and the absence of atmospheric attenuation will reduce the necessary observation time for astronomical observations in this frequency range by at least two orders of magnitude compared to ground based facilities.

There are five bolometric detector arrays for the SPIRE instrument on board of the Herschel Space Observatory. Our first report (Nguyen et al., 2004) presented the measurement of the two spectroscopic detector arrays. In this paper, we report the performance of the remaining three units for the Photometer, including the photometric long, medium and short wavelength (PLW, PMW and PSW). We note that all five SPIRE detector arrays meet the requirement.

We present the experimental results of voltage-biased superconducting bolometers (VSB) on silicon nitride (Si₃N₄) membranes with niobium wiring developed in collaboration between the Institut fur Physikalische Hochtechnologie (IPHT), Jena, Germany and the Max-Planck-Institut fur Radioastronomie (MPIfR), Bonn, Germany. The bolometer current is measured with the superconducting quantum interference device (SQUID), and as expected, the current responsivity is proportional to the inverse of the bias voltage. The experiments were performed with bilayer gold-palladium molybdenum thermistor at 300 mK ³He cooled cryostat and the desired transition temperature of T_c = 450 mK is achieved. The strong negative electro-thermal feedback of the VSB maintains the constant bolometer temperature and reduces the response time from 4 ms to 100 μs. We have tested thermistors of various size and shape on a continuous membrane and achieved a noise equivalent power (NEP) of 3.5 × 10^-16 W/√Hz. The measured NEP is relatively high due to the comparatively high background and high thermal conductance of the unstructured silicon nitride (Si₃N₄) membrane. We have fabricated 8-leg spider structured membranes in three different geometries and the relation between the geometry and the thermal conductance (G) is studied. Using the COSMOS finite element analysis tool, we have modeled the TES bolometers to determine the thermal conductance for different geometries and calculated the various parameters. Due to the demands of large number pixel bolometer camera we plan to implement multiplex readout with integrated SQUIDs in our design.

The Penn Array Receiver (PAR) is a camera designed for rapid, high angular resolution imaging at 90 GHz (3.3 mm). When installed on the 100 m Green Bank Telescope it will have a 32" × 32" field of view and 8" resolution. PAR has an eight by eight planar array of superconducting Transition Edge Sensor bolometers. Currently it is in the commissioning phase and after that it will become a user instrument capable of mapping a 5' × 5' area of sky to a noise level of 40 μJy in one hour.

We present images taken with the first deployed astronomical instrument to use multiplexed superconducting bolometers. The Fabry-Perot Interferometer Bolometer Research Experiment (FIBRE), a broadband submillimeter spectrometer, took these images as a detector investigation at the Caltech Submillimeter Observatory (CSO). FIBRE's detectors are superconducting bilayer transition edge sensor (TES) bolometers read out by a SQUID multiplexer. An order-sorted Fabry-Perot provides illumination of a 16-element linear bolometer array, resulting in five orders at a spectral resolution of around 1200 covering the 350 micron atmospheric band. We present multiwavelength images of Jupiter, Venus and the high-mass star-forming region G34.3+0.2 taken with this instrument at several wavelengths in the 350 micron band, separated by approximately 8 microns. These images have validated the use of multiplexed superconducting bolometers in an astronomical application and have helped inform the design of our future instruments.

We are building a bolometer camera (the Goddard-Iram Superconducting 2-Millimeter Observer, GISMO) for operation in the 2 mm atmospheric window to be used at the IRAM 30 m telescope. The instrument uses a 8x16 planar array of multiplexed TES bolometers which incorporates our newly designed Backshort Under Grid (BUG) architecture. Due to the size and sensitivity of the detector array (the NEP of the detectors is 4×10^-17 W/√Hz), this instrument will be unique in that it will be capable of providing significantly greater imaging sensitivity and mapping speed at this wavelength than has previously been possible. The major scientific driver for this instrument is to provide the IRAM 30 m telescope with the capability to rapidly observe galactic and extragalactic dust emission, in particular from high-z ULIRGs and quasars, even in the summer season. The 2 mm spectral range provides a unique window to observe the earliest active dusty galaxies in the universe and is well suited to better confine the star formation rate in these objects. The instrument will fill in the SEDs of high redshift galaxies at the Rayleigh-Jeans part of the dust emission spectrum, even at the highest redshifts. The observational efficiency of a 2 mm camera with respect to bolometer cameras operating at shorter wavelengths increases for objects at redshifts beyond z ~ 1 and is most efficient at the highest redshifts, at the time when the first stars were re-ionizing the universe. Our models predict that at this wavelength one out of four serendipitously detected galaxies will be at a redshift of z > 6.5.

SCUBA-2 is an innovative 10,000 pixel submillimeter camera due to be delivered to the James Clerk Maxwell Telescope in late 2006. The camera is expected to revolutionize submillimeter astronomy in terms of the ability to carry out wide-field surveys to unprecedented depths addressing key questions relating to the origins of galaxies, stars and planets. This paper presents an update on the project with particular emphasis on the laboratory commissioning of the instrument. The assembly and integration will be described as well as the measured thermal performance of the instrument. A summary of the performance results will be presented from the TES bolometer arrays, which come complete with in-focal plane SQUID amplifiers and multiplexed readouts, and are cooled to 100mK by a liquid cryogen-free dilution refrigerator. Considerable emphasis has also been placed on the operating modes of the instrument and the "common-user" aspect of the user interface and data reduction pipeline. These areas will also be described in the paper.

We present the results of characterization measurements on a 1280 pixel superconducting bolometer array designed for operation at wavelengths around 450 μm. The array is a prototype for the sub-arrays which will form the focal plane for the SCUBA-2 sub-mm camera, being built for the James Clerk Maxwell Telescope (JCMT) in Hawaii. With over 10 000 pixels in total, it will provide a huge improvement in both sensitivity and mapping speed over existing instruments. The array consists of molybdenum-copper bi-layer TES (transition edge sensor) pixels, bonded to a multiplexer. The detectors operate at a temperature of approximately 175 mK, and require a heat sink at a temperature of approximately 60 mK. In contrast to previous TES arrays, the multiplexing elements are located beneath each pixel (an "in-focal plane" configuration). We present the results of electrical and optical measurements, and show that the optical NEP (noise equivalent power) is less than 1.4 × 10^-16 W Hz^-0.5 and thus within the goal of 1.5 × 10^-16 W Hz^-0.5.

We present a first cut instrument design package for the proposed 25 meter Cornell-Caltech Atacama Telescope (CCAT). The primary science for CCAT can be achieved through wide field photometric imaging in the short submillimeter through millimeter (200 μm to 2 mm) telluric windows. We present strawman designs for two cameras: a 32,000 pixel short submillimeter (200 to 650 μm) camera using transition edge sensed bare bolometer arrays that Nyquist samples (@ 350 μm) a 5'×5' field of view (FoV), and a 45,000 pixel long wavelength camera (850 μm to 2 mm) that uses slot dipole antennae coupled bolometer arrays with wavelength dependent sampling that covers up to a 20' square FoV. These are our first light instruments. We also anticipate "borrowed" instruments such as direct detection and heterodyne detection spectrometers will be available at, or nearly at first light.

SHARC-II is a 32 × 12 pixel submillimeter camera that is used with the ten-meter diameter Caltech Submillimeter Observatory (CSO) on Mauna Kea. This camera can be operated at either 350 or 450 microns. We developed a module that is installed at the CSO Nasmyth focus in order to convert SHARC-II into a sensitive imaging polarimeter, which we refer to as "SHARP". SHARP splits the incident beam into two orthogonal polarized beams that are then re-imaged onto different halves of the SHARC-II bolometer array. When this removable polarimetry module is in use, SHARC-II becomes a dual-beam 12 × 12 pixel polarimeter. Sky noise is a significant source of error for submillimeter continuum observations. Because SHARP will simultaneously observe two orthogonal polarization components, we are able to eliminate or greatly reduce this source of error. Here we describe the design of SHARP and report preliminary results of tests and observations carried out during our first two runs at CSO in August 2005 and January 2006.

AMiBA, as a dual-polarization 86-102 GHz interferometer array, is designed to measure the power spectrum of fluctuations in the cosmic microwave background (CMB) radiation, and to detect the high-redshift clusters of galaxies via the Sunyaev-Zel'dovich Effect (SZE). The operation of AMiBA is about to begin after installation of the first two receivers and correlators onto the 6-meter diameter platform by the end of 2005. The initial setup of the array will consist of 7 antennas with 60 cm diameter reflectors in a hexagonal configuration, aiming at multipoles l ~ 3000. Signals from receivers are cross-correlated in analog lag correlators. The initial operation will focus on characterizing the systematics by observing various known objects on the sky. The expansion to 13 elements with larger dishes will commence once the 7-element array testing is completed.

PAPPA is a balloon-based experiment designed to measure the polarization of the Cosmic Microwave Background using candidate technology for an eventual Einstein Inflation Probe mission. It will survey a 20° × 20° patch of sky with 0.5° angular resolution covering 3 passbands centered at 89, 212 and 302 GHz. Detection will be accomplished via antenna-coupled transition edge sensors (TESs) with SQUID-based readouts. In the eventual flight package, band defining filters and MEMS-based polarization modulators will be incorporated into the superconducting microstrip transmission lines that terminate in resistors that are thermally coupled to the TESs. The MEMS switches will allow on-chip polarization modulation that is faster than significant detector gain variations. The initial configuration will incorporate a simplified focal plane augmented by quasioptical polarization modulation. We describe the overall instrument design and present a summary of the current progress.

The Robinson Telescope (BICEP) is a ground-based millimeter-wave bolometric array designed to study the polarization of the cosmic microwave background radiation (CMB) and galactic foreground emission. Such measurements probe the energy scale of the inflationary epoch, tighten constraints on cosmological parameters, and verify our current understanding of CMB physics. Robinson consists of a 250-mm aperture refractive telescope that provides an instantaneous field-of-view of 17° with angular resolution of 55' and 37' at 100 GHz and 150 GHz, respectively. Forty-nine pair of polarization-sensitive bolometers are cooled to 250 mK using a ⁴He/³He/³He sorption fridge system, and coupled to incoming radiation via corrugated feed horns. The all-refractive optics is cooled to 4 K to minimize polarization systematics and instrument loading. The fully steerable 3-axis mount is capable of continuous boresight rotation or azimuth scanning at speeds up to 5 deg/s. Robinson has begun its first season of observation at the South Pole. Given the measured performance of the instrument along with the excellent observing environment, Robinson will measure the E-mode polarization with high sensitivity, and probe for the B-modes to unprecedented depths. In this paper we discuss aspects of the instrument design and their scientific motivations, scanning and operational strategies, and the results of initial testing and observations.

We have developed a completely lithographic antenna-coupled bolometer for CMB polarimetry. The necessary components of a millimeter wave radiometer - a beam forming element, a band defining filter, and the TES detectors - are fabricated on a silicon chip with photolithography. The densely populated antennas allow a very efficient use of the focal plane area. We have fabricated and characterized a series of prototype devices. We find that their properties, including the frequency and angular responses, are in good agreement with the theoretical expectations. The devices are undergoing optimization for upcoming CMB experiments.

Precise astronomical polarization measurements generally require the use of polarization modulation. We have developed a new modulator, the Variable-delay Polarization Modulator (VPM) which uses two modified Martin- Puplett interferometers to induce a physical path length difference between polarization components. This highly durable and efficient device can easily be adapted to a wide range of wavelengths and temperatures, making it well suited for air- and space-borne facilities. This paper discusses the basic modulator design and a comparison to the half-wave plate, as well as details of VPM tests conducted at the Submillimeter Telescope Observatory (SMTO).

Kinetic inductance detectors (KIDs) provide an attractive solution to the production of large detector arrays for use in ground and space based 200μm astronomy. KIDs work by measuring the change in quasi-particle density upon photon absorption in a high Q superconducting resonator. A change in quasi-particle density is measured by a shift in phase of a microwave probe signal of frequency equal to that of the resonant frequency of the KID. Such detectors have a fundamental noise limit owing to the quasi-particle recombination rate, which, in a KID fabricated from a high quality Niobium film can give sensitivities of 10^-18W√Hz at 1K. Constructing KIDs of varying resonant frequencies coupled to a single transmission line provides a multiplexed detector array with simple low temperature electronics. Here we discuss the theoretical requirements for both ground and space based 200μm cameras with various radiation coupling schemes for this wavelength range using distributed and lumped element high Q resonators.

We describe a new type of terahertz (THz) detector for astronomical observation using a two-dimensional electron gas (2DEG) as the absorbing medium. The detection principle is based on the hot electron effect in 2DEGs. Electrons are heated by THz radiation and the electron temperature is read out by two symmetrical superconductor - 2DEG tunnelling junctions. Hot electrons are removed via tunnelling through a barrier into the superconducting contacts. The energy gap in the superconducting contacts prevents the escape of the colder, non-photoexcited electrons from the 2DEG. The high mobility 2DEG itself is created within AlGaAs/GaAs heterostructure with a single quantum well. In this paper we present low temperature DC measurements of 2DEG detectors, and measurements of the electron-phonon thermal conductivity of a 2DEG at 4.2 K and 300 mK as a function of electron temperature and magnetic field (in the 4.2 K case). From these measurements we estimate the noise equivalent power (NEP) of an element in a filled array of S-2DEG-S detectors at 4.2 K to be on the order of ≈ 10^-14W/√Hz with a response time of ≈ 1ns; at 300 mK, an NEP on the order of ≈ 10^-19W/√Hz and a response time of ≈ 0.1μs. Using measured parameters for the normal resistance of the S-2DEG-S contacts, we calculate the effect of using a voltage bias to cool the electrons in the absorber to significantly below a 300 mK base temperature. In this configuration, S-2DEG-S detectors can achieve sufficient sensitivity to detect individual THz photons.

Ever since the first proposal of the voltage-biased transition-edge bolometer the astrophysics community desired bolometer arrays with as many pixels as possible. With respect to the technical problem due to the need of lots of readout SQUID sensors only with multiplexing it is possible to go beyond a few hundred pixel. A technology which allows the manufacture of detector and readout on one chip would simplify this task substantially. Here we demonstrate the fabrication of a transition edge sensor based on a thermistor out of a molybdenum / gold-palladium bilayer. The alloy of gold-palladium (Au-Pd), which allows the tuning of molybdenum's critical temperature by one order of magnitude, is taken from our foundry process for SQUID manufacturing. Au-Pd can further be used for shunt resistances, absorber patterns and bond pads, and, therefore, it is a good choice for a combined technology. The thermistor is placed on a moderately patterned silicon nitride membrane in the shape of an 8-legged spider. The radiation band of interest is coupled via a conical feed horn to a simple grid of dipole-like antenna patterns. This removes the need for the poorly reproducible high-resistance absorption films for the matching of the free space impedance. The simple detector technology is compatible with the SQUID manufacturing. Hence, some of the SQUID layers can be merged with the corresponding detector layer, i.e. the thermistor wiring and the SQUID washer are made in a single niobium layer. The concept of feed horn coupling eases the design requirements, consequently the SQUID can be placed close to the detector, thereby allowing a simpler wiring to be used and in theory a better performance to be obtained.

We describe progress toward realizing a new architecture for focal plane arrays for the Submillimeter and Far-Infrared (FIR) bands. This architecture is based on a detector design utilizing distributed hot-electron transition edge sensors (TES) coupled to slot antenna elements. Arrays utilizing this type of detector can be considerably easier to manufacture than membrane-isolated TES arrays, because the need for micro-machining is eliminated. We present background and rationale for this new array architecture and details of a new antenna design for an imaging polarimeter, which yields greater bandwidth than past designs. In addition, we describe a cryogenic facility for testing these arrays.

Gallium arsenide is a promising material for large photoconductor arrays to be operated at submm wavelengths, where currently small stressed germanium arrays are used. The smaller binding energy of shallow donors in GaAs compared to Ge results in response at longer wavelengths without having to apply uniaxial stress. Use of n-type GaAs will greatly simplify the production of detector arrays and therefore allow much larger numbers of pixels. We have grown n-doped GaAs epitaxial films and demonstrated high absorption coefficients at wavelengths exceeding 300 μm. Combined with a high purity GaAs layer, a blocked impurity band (BIB) detector can be formed in order to simultaneously achieve efficient absorption and low dark currents. Recent progress in GaAs epitaxy technology allows production of such multilayer devices in wafer size. We are presenting the characterization results of our preliminary GaAs BIB structures.

The Atacama Pathfinder EXperiment (APEX) is a 12 m antenna now operating at the Llano Chajnantor on the Atacama Desert in Northern Chile at 5100 m altitude. APEX will be equipped with single-pixel heterodyne receivers covering 211 - 1500 GHz frequency range. We present a sideband separation (2SB) mixer using superconducting-insulator-superconductor (SIS) junction for the APEX band 2, 275-370 GHz. The 2SB mixer is based on a previous development of a double sideband (DSB) mixer, which is currently installed at the APEX telescope. This DSB receiver has a noise temperature of about 40-50 K across the band, and, as installed on one of the best site on the Earth, yields total DSB system noise temperatures of about 100 K for excellent weather. The 2SB mixer layout uses a modular approach with two identical DSB mixers, independently tested, having similar characteristics, and combined with an intermediate waveguide block, containing a 3 dB-90° branch-line coupler for the RF signal and a 3 dB-180° divider for the LO signal. The LO signals are injected into the mixer using a novel waveguide directional coupler based on 2 quartz chips containing E- probes, allowing to couple -15 dB of the LO to the RF path. At the conference, we will present the first measurements of this 2SB mixer, together with the current performance the DSB receiver at the APEX telescope.

We summarize the development and the delivery of two SIS mixers for the 1.1-1.25 THz band of the heterodyne spectrometer of Herschel Observatory (HSO). The quasi-optical SIS mixer has two Nb/AlN/NbTiN junctions with the area of 0.25 um². The Josephson critical current density in the junction is 30-50 kA/cm². The tuning circuit integrated with SIS junction has the base electrode of Nb and a gold wire layer. With the new SIS mixers the test receiver maximum Y factor is 1.41. The minimum receiver uncorrected DSB noise temperature is 450 K. The SIS receiver noise corrected for the loss in the optics is 350-450 K across the 1100-1250 GHz band. The receiver has a uniform sensitivity in the full IF range of 4-8 GHz. The sub-micron sized SIS junction shape is optimized to ease the suppression of the Josephson current, and the receiver operation is stable. The measured mixer beam pattern is symmetrical and, in a good agreement with the requirements, has the f/d =4.25 at the central frequency of the operation band. The minimum DSB SIS receiver noise is close to 6 hv/k, the lowest value achieved thus far in the THz frequencies range.

We report the measurement results and compensation of the antenna elevation angle dependences of the Submillimeter Array (SMA) antenna characteristics. Without optimizing the subreflector (focus) positions as a function of the antenna elevation angle, antenna beam patterns show lopsided sidelobes, and antenna efficiencies show degradations. The sidelobe level increases and the antenna efficiencies decrease about 1% and a few %, respectively, for every 10° change in the elevation angle at the measured frequency of 237 GHz. We therefore obtained the optimized subreflector positions for X (azimuth), Y (elevation), and Z (radio optics) focus axes at various elevation angles for all the eight SMA antennas. The X axis position does not depend on the elevation angle. The Y and Z axes positions depend on the elevation angles, and are well fitted with a simple function for each axis with including a gravity term (cosine and sine of elevation, respectively). In the optimized subreflector positions, the antenna beam patterns show low level symmetric sidelobe of at most a few%, and the antenna efficiencies stay constant at any antenna elevation angles. Using one set of fitted functions for all antennas, the SMA is now operating with real-time focusing, and showing constant antenna characteristics at any given elevation angle.

Many types displacement sensors used in active optics in many astronomical telescopes over the world ware described and the measurement theory of different sensors were explained in this paper. Based on the displacement sensor's specification of LAMOST, a test bed to check it which used dual-frequency laser interferometer was established. Some main parameters affect on the measurement accuracy of the test bed such as temperature, vibration as well as some mechanism characteristics of fixing devices were analyzed in detail, and correspond solutions were adopted. According to theoretic analysis, rectification method was brought forward to compensate the errors of sensor cause by tip-tilt, thermal shift, non-linearity, and combined with the test results of some types sensors, their advantages and disadvantages was concluded. At last it forecasted the most appropriate displacement will be used in future active optics based on the discussion above and the developing trend of large telescope.

Atmospheric water vapor causes significant undesired phase fluctuations for the SMA interferometer, particularly in its highest frequency observing band of 690 GHz. One proposed solution to this atmospheric effect is to observe simultaneously at two separate frequency bands of 230 and 690 GHz. Although the phase fluctuations have a smaller magnitude at the lower frequency, they can be measured more accurately and on shorter timescales due to the greater sensitivity of the array to celestial point source calibrators at this frequency. In theory, we can measure the atmospheric phase fluctuations in the 230 GHz band, scale them appropriately with frequency, and apply them to the data in 690 band during the post-observation calibration process. The ultimate limit to this atmospheric phase calibration scheme will be set by the instrumental phase stability of the IF and LO systems. We describe the methodology and initial results of the phase stability characterization of the IF and LO systems.

We present the preliminary design of FTS-2, an imaging Fourier transform spectrometer (IFTS) for use with SCUBA-2, the second generation, wide-field, submillimetre camera currently under development for the James Clerk Maxwell Telescope (JCMT). This system, which is planned for operation at the start of 2007, will provide simultaneous broadband spectral imaging across both the 850 and 450 μm bands with variable resolution ranging from resolving powers of R ~10 to 5000. The spectrometer uses a folded Mach-Zehnder configuration and novel intensity beam dividers. The mechanical and optical design of FTS-2 as of the Critical Design Review stage of the project are discussed, along with the interfaces with SCUBA-2 and the JCMT.

We propose an instrument by applying the aperture synthesis technique to the Martin & Puplett type Fourier Transform Spectrometer in millimeter and sub-millimeter waves. We call this equipment Multi-Fourier Transform interferometer (MuFT). MuFT realizes a wide band imaging, spectroscopy and polarimetry in millimeter and sub-millimeter waves. The direct detectors, eg. bolometer, SIS video detector, can be used as the focal plane detectors. These type of detectors have a great advantage in FIR band since they are free from the quantum limit of the noise which limits the sensitivity of the heterodyne detectors used in the usual interferometers. Further, the direct detectors are able to make a large format array contrary to the usual interferometers in which usage of array detector is practically difficult. Above three characteristics make one be possible to develop high sensitive super broad band FIR interferometer with wide field of view. Fundamentals of the MuFT with results of laboratory experiments and current status of astronomical observations with MuFT are summarized. We did test observation with this system in the winter 2005. We also report concerning the observational result in 2006.

We report the design of a sub-millimeter Fourier Transform Spectrometer of the Martin Puplett type (FTS-MP here after). The instrument will be installed on Sierra La Negra, and will operate from ~215 GHz to 1 THz approximately with a moderate resolution of 500 MHz. The main motivations of the work are the development of basic instrumentation for characterizing the LMT site (Large Millimeter Telescope) as well as optical components and to provide a portable broadband system for site testing. The collected data will be used for transmission model validation. The data also will influence the design of the new generation of sub-millimeter and millimeter cameras for the LMT. In the present work we emphasize cryogenic detector system design (bolometer detector and coupling optics) and the optical system layout for the FTS-MP. Test measurements in the laboratory are reported in this work.

SCUBA-2 is a new wide-field submillimeter camera under construction for the James Clerk Maxwell Telescope on Mauna Kea in Hawaii. SCUBA-2 images simultaneously at 450 and 850 μm using large-scale arrays of superconducting bolometers, with over five thousand pixels at each wavelength. Time division multiplexed readouts and cryogenic amplifiers, both based on superconducting quantum interference devices (SQUIDs), are also used in the design. The SCUBA-2 detector arrays must be well shielded against magnetic fields, since the performance of the bolometers can be seriously affected by the presence of a strong field, and the SQUIDs are themselves sensitive magnetometers. This shielding is to be provided by a combination of high-permeability and superconducting layers on both the ambient temperature and cryogenic stages of the instrument. To optimise and demonstrate the effectiveness of the shielding design, a finite-element modelling method was employed, using the Ansoft^(R) Maxwell 3D^TM package. Although a number of approximations had to be made in the modelling, the finite-element results allow a good estimation of the effectiveness of the shielding at attenuating external magnetic fields to be made. This paper describes the modelling process, outlines the key results and summarises the final shielding design.

We have developed the superconductive imaging submillimeter-wave camera with nine detector elements (SISCAM-9) for Atacama Submillimeter Telescope Experiment (ASTE). SISCAM-9 has nine SIS photon detectors as focal plane array detector at 680 GHz. To obtain background noise limited sensitivity, we need to operate detectors under the condition that noise is dominated by shot noise of background photo current. To realize this condition, we fabricated low noise readout circuits using Si-JFETs. In laboratory, we evaluated performance of nine SIS photon detectors. This is the first demonstration of 2D array of SIS photon detector for SISCAM-9. Measuring I-V characteristics, detector gap voltages were 4.9 mV and photo currents were 3 nA. Measuring spectral responses, they had almost same center frequency of 679 GHz and bandwidths of 77 GHz. They almost matched to 675 GHz atmospheric window from ASTE site. Detector noise under 300 K radiation were only a few times as large as the shot noise of photo current. Detector NEP was 1.7x10^-15 W/Hz^1/2,with the detector quantum efficiencies of 12%. For the first time, SIS photon detectors worked under shot noise limited condition under 300 K background radiation. We have developed observation system, SISCAM-9, to realize the first astronomical observations. Detectors were mounted in a cryostat that can be remotely operated to cool the detectors to 0.46 K. We installed SISCAM-9 in the ASTE telescope and measured system performance such as photo current and noise characteristics. For the first time, SIS photon detectors operated under the observing condition. We succeeded in making the first astronomical observation of the moon.

C_ℓOVER is an experiment which aims to detect the signature of gravitational waves from inflation by measuring the B-mode polarization of the cosmic microwave background. C_ℓOVER consists of three telescopes operating at 97, 150, and 220 GHz. The 97-GHz telescope has 160 horns in its focal plane while the 150 and 220-GHz telescopes have 256 horns each. The horns are arranged in a hexagonal array and feed a polarimeter which uses finline-coupled TES bolometers as detectors. To detect the two polarizations the 97-GHz telescope has 320 detectors while the 150 and 220-GHz telescopes have 512 detectors each. To achieve the required NEPs the detectors are cooled to 100 mK for the 97 and 150-GHz polarimeters and 230 mK for the 220-GHz polarimeter. Each detector is fabricated as a single chip to guarantee fully functioning focal planes. The detectors are contained in linear modules made of copper which form split-block waveguides. The detector modules contain 16 or 20 detectors each for compatibility with the hexagonal arrays of horns in the telescopes' focal planes. Each detector module contains a time-division SQUID multiplexer to read out the detectors. Further amplification of the multiplexed signals is provided by SQUID series arrays. The first prototype detectors for C_ℓOVER operate with a bath temperature of 230 mK and are used to validate the detector design as well as the polarimeter technology. We describe the design of the C_ℓOVER detectors, detector blocks, and readout, and give an update on the detector development.

Several technologies are now being considered for modulating the polarization in various B-mode instruments, including rotating quasioptical half-wave plates in front of the focal plane array, rotating waveguide half-wave plates and Faraday rotators. It is not at all clear that any of these techniques is feasible without heavy penalty in cost or performance. A potentially much more efficient method is to use a pseudo-correlation polarimeter in conjunction with a planar circuit phase switch. We investigate three different devices for use as mm-wave switches, SIS tunnel junctions, capacitively coupled superconducting nanostrips and RF MEMS. The SIS tunnel junction switches operate by switching between two different bias voltages, while the nanostrip switch operates by changing the impedance of a resonant circuit by driving the nanostrip from the superconducting to normal state. In each case the RF signal sees two substantially different complex impedance states, hence could be switched from one transmission line branch to another. In MEMS this is achieved by mechanical movement of one plate of a parallel plate capacitor system. Although RF MEMS have been reported at high microwave and low mm-wave frequencies, in this work we have investigated cryogenic MEMS for operation at high mm-wave frequencies (225 GHz) using superconducting transmission lines. We present and compare designs and simulations of the performance of phase switches based on all three switching technologies, as well as preliminary experimental results for each of the switches. Finally we also present designs of phase shift circuits that translates the on/off switching into phase modulation.

In this paper describes the system configuration and the some performance test results of the 15 pixels digital autocorrelation spectrometer to be used at the Taeduk Radio Astronomy Observatory (TRAO) of Korea. This autocorrelation spectrometer instrument enclosed in a 3-slot VXI module and controlled via a USB port by a backend PC. This spectrometer system consists of the 4 band-pass filters unit, the digitizer, the 512 lags correlator, the clock distribution unit, and USB controller. And here we describe the frequency accuracy and the root-mean-square noise characteristic of this spectrometer. After some calibration procedure, this spectrometer can be use as the back-end system at TRAO for the 3x5 focal plane array receivers.

We are developing cryogenic readout circuits for the array of superconducting tunneling junctions (STJs) at submillimeter wavelength SISCAM (Superconductive Imaging Submillimeter-wave CAMera). A current conceptual design of SISCAM will employ a direct hybrid array system just like CMOS image sensors widely used at optical and infrared wavelength. Because of relatively large impedance of the STJ fabricated by RIKEN (~10 MΩ in a dark condition), it requires readout preamplifier with low current noise. Therefore, it is not suitable for the STJ to use a readout system by Superconductive Quantum Interferences Devices as for Transition Edge Sensor. Instead, we selected capacitive transimpedance amplifier (CTIA) using a SONY n-type GaAs Junction Field Effect Transistor (JFET). However, the CTIA has not been used as the readout of the STJ. Therefore, we measured the photocurrent of the STJ by the CTIA with Silicon JFETs and by transimpedance amplifier (TIA), which is a conventional readout for the STJ, in the same bias condition, and confirmed both results are in good agreement. Additionally, we report development of readout integrated circuits with GaAs JFETs. In order to design the CTIA circuit with the GaAs JFETs, we fabricated the independent GaAs JFETs and matched pairs of them. We measured electrical characteristics of these GaAs JFETs at the cryogenic temperatures less than 4.2 K. We demonstrated performance of an operational amplifier fabricated with the GaAs JFETs measuring a differential amplifier with the dual GaAs JFET, and additionally estimate amplifier gain, offset voltage, and power consumption of the CTIA by the circuit simulation using the PSPICE. In consequence, the expected performance fulfills the requirements for the readout amplifier of the STJs except for the noise performance.