|The APEX-SZ Project|
The APEX telescope is a prototype, designed by the VERTEX Corporation , for the ALMA array. The telescope design is a on-axis Cassegrain mounted on an elevation over azimuth mount. The primary mirror has a focal length of 4.8 meters and is 12 meters in diameter. The telescope is designed for submillimeter observations, with an antenna rms surface accuracy of 18 microns, so design requirements for millimeter CMB observations are easily met. A mirror in the optical path within the cabin is used to switch between receivers. The telescope weights roughly 77 metric tons.
APEX will be located at 5000 meters in the Atacama plateau in Chile. This region is one of the driest places on earth. This makes the site ideal for millimeter-wave observations, as the limiting factor for ground-based observation sensitivity is noise from atmospheric water vapor fluctuations. The Atacama site can view most of the celestial sphere with good overlap with many other ground based telescopes, which will be needed for later follow-up observations to determine the redshifts of galaxy clusters. The Atacama site allows for drift scanning a patch of sky at different azimuths, which can be used to test systematics, and has year-round access.
Design of the tertiary optics is being done by the Berkeley group in collaboration with MPIfR. The meter diameter primary mirror and 0.75 meter diameter secondary form a Cassegrain focus inside the cabin. The field of view of the system is limited by the 0.75 meter diameter hole in the primary to be < 0.5 deg. Initially, the focal plane array will subtend a 0.4 deg field of view. Tertiary optics reimage the focal plane with f/# ~ 1.75 to cover the bolometer array, which will be 15 cm in diameter.
The tertiary optics create an image of the primary at a cold (4 Kelvin) Lyot stop. This Lyot stop is designed to truncate the beams on the primary mirror to reduce spillover, and results in a 0.8 arcminute FWHM beam on the sky.
The receiver will consist of a 300 element hexagonal close-packed bolometer array, fed by an array of smooth-walled conical horns with 2 f λ aperture diameter. We will use a single bolometer array for both frequencies of interest, and will swap between two horn arrays to control frequency sensitivity. The bolometers will be Si3N4 "spiderweb" Transition Edge Sensor (TES) bolometers, with quarter-wavelength backshort reflectors to maximize absorption of incident radiation. These detector arrays are being fabricated by our group at the Berkeley Microlab on segments of silicon wafers. The TES devices are low impedance devices that are less susceptible to microphonics than doped semiconductor bolometers. Fabrication of the TES devices by conventional optical photolithography and the development of SQUID multiplexers (which the Berkeley group is currently working on) will allow the development of large arrays.
The horn and bolometer arrays must be cooled to 300 milliKelvin for optimum operation of the TES bolometers. A pulse-tube cooler will be used to reach the 4 Kelvin operating point of the three-stage sorption refrigerator. The pulse-tube cooler has the advantage of not requiring liquid cryogens, reducing logistical and personnel demands.
All back-end electronics and data acquisition systems for the SZ receiver are also being developed by the Berkeley group. The TES bolometers are operated with voltage-bias, so the signal will be read out as a current measured by a shunt-feedback SQUID amplifier, with one SQUID per sensor. SQUID's are ideally matched to our sensors, as their low input impedance helps to maintain the constant voltage bias of the bolometers. Shunt feedback to the SQUID input coil improves the dynamic range of the readout and facilitates ac-bias operation. The Berkeley group is also currently working on a frequency domain SQUID multiplexer to read out large arrays of detectors, and this technology may appear in future generations of the APEX SZ receiver.