Cryocoolers have long been demonstrated to be a dominant source of vibration that have caused significant problems with AO systems on large telescopes. Existing large telescopes have already imposed strict vibration requirements on instruments in response to existing problems, and have often struggled to achieve them. As the field moves into the next generation telescopes with GMT, TMT and eELT, vibration requirements continue to get ever tighter. Instrument teams must respond to these more demanding requirements by careful selection of cryocoolers and thoughtful design of cryocooler mounts that are matched closely with the specific requirements of the telescope. As we will demonstrate in this paper there is not a one-size-fits-all solution for every instrument and every telescope. In this paper we demonstrate a general method of deriving the required performance for an anti-vibration mounts for cryocoolers. First we characterize a linear Stirling-type cryocooler as a source of vibration, and determine what compliant mounts would be required to make them acceptable for use on the VLT, GMT and TMT. Measurements are taken of vibration from a Cryotel GT linear Stirling cooler (with active vibration cancellation enabled). By comparing the measured vibration against the requirements of each telescope, we are able to determine the required transfer function and therefore the required spring rate for compliant mounts. The results indicate that while some simple rubber mounts may be sufficient for use with the VLT and TMT, but a compliant mount with natural frequency below 14 Hz must be used for GMT.
Ground-based infrared observations are often limited by atmospheric absorption and emission. Space-based instruments avoid this, but introduce unique technical challenges. We present the design of a flexible, compact, and cost-effective detector controller for space, based on commercial off-the-shelf components. Its architecture provides up to 50 configurable clock sequences and 16 biases, 32 16-bit video channels and several genera lpurpose ports. This allows for full control of CMOS detectors including Leonardo ‘SAPHIRA’ avalanche photodiode arrays (APD) which represent current state of the art in low-noise infrared imaging.
The Leonardo SAPHIRA is a HgCdTe linear avalanche photodiode array enabling high frame rate, high sensitivity, low noise, and low dark current imaging at near-infrared wavelengths. The ANU utilised the Leonardo SAPHIRA to develop a high cadence “Lucky Imager” which was successfully tested on sky at Siding Spring Observatory. The cryogenic electronics and cryostat were designed and built by the ANU. The cryostat was cooled with a compact Stirling cycle cryocooler with active vibration damping. Various detector control systems were tested, including an ESO 'NGC' system and also a 32 channel ARC SDSU Series III. Images were ultimately captured at a windowed frame rate of 2.2 kHz with the ESO NGC controller.
‘Emu’ is a compact wide-field photometer destined for a 6-month mission on the exterior of the International Space Station (ISS), commencing in 2021. Emu will undertake a sky survey in the 1.4 μm ‘water band’, as a method of estimating oxygen abundance in the atmospheres of cool stars down to a magnitude of mAB≈13 (H-band).
SAPHIRA detectors, which are HgCdTe linear avalanche photodiode arrays manufactured by Leonardo, enable high frame rate, high sensitivity, low noise, and low dark current imaging at near-infrared wavelengths. During all University of Hawaii Institute for Astronomy lab testing and observatory deployments of SAPHIRA detectors, there was approximately one meter of cables between the arrays and the readout controllers. The output drivers of the detectors struggled to stably send signals over this length to the readout controllers. As a result, voltage oscillations caused excess noise that prevented us from clocking much faster than 1 MHz. Additionally, during some deployments, such as at the SCExAO instrument at Subaru Telescope, radio-frequency interference from the telescope environment produced noise many times greater than what we experienced in the lab. In order to address these problems, collaborators at the Australia National University developed a cryogenic preamplifier system that holds the detector and buffers the signals from its outputs. During lab testing at 1 MHz clocking speeds, the preamplifiers reduced the read noise by 45% relative to data collected using the previous JK Henriksen detector mount. Additionally, the preamplifiers enabled us to increase the clocking frequency to 2 MHz, effectively doubling the frame rate to 760 Hz for a full (320x256 pixel) frame or 3.3 kHz for a 128x128 pixel subarray. Finally, the preamplifiers reduced the noise observed in the SCExAO environment by 65% (to essentially the same value observed in the lab) and eliminated the 32-pixel raised bars characteristic of radio-frequency interference that we previous observed there.
We present a summary of the cryogenic detector preamplifier development programme under way at the ANU. Cryogenic preamplifiers have been demonstrated for both near-infrared detectors (Teledyne H1RG and Leonardo SAPHIRA eAPD as part of development for the GMTIFS instrument) and optical CCDs (e2v CCD231-84 for use with the AAT/Veloce spectrograph). This approach to detector signal conditioning allows low-noise instrument amplifiers to be placed very close to an infra-red detector or optical CCD, isolating the readout path from external interference noise sources. Laboratory results demonstrate effective isolation of the readout path from external interference noise sources. Recent progress has focussed on the first on-sky deployment of four cryogenic preamp channels for the Veloce Rosso precision radial velocity spectrograph. We also outline future evolution of the current design, allowing higher speeds and further enhanced performance for the demanding applications required for the on instrument wavefront sensor on the Giant Magellan Integral Field Spectrograph (GMTIFS).
The Australian National University (ANU), we are undertaking to deploy a Lucky Imaging instrument on the 2.3 m telescope at Siding Springs using a Leonardo SAPHIRA near-infrared electron Avalanche Photo-Diode (eAPD) array, capable of high cadence imaging with frame rates of 10 - 5,000 Hz over the wavelength range of 0.8 μm to 2.5 μm. compact cryocooler capable of cooling the Leonardo SAPHRA APD and associated cryogenic electronics to temperatures below 100K with little to no vibration. An ideal candidate cryocooler is the Sunpower Cryotel GT with active vibration cancellation. The Cryotel GT is an orientation independent, Stirlng cycle cooler with water jacket heat rejection. This cooler will meet the system cooling requirements. The cryocooler has been integrated with the APD Lucky Imager cryostat through 3 rubber isolating mounts and bellows and tested while suspended from a stable frame. The tethers supporting the cryostat and cooler assembly are not attached to the cryostat and cooler. The exported vibration was measured simultaneously in all 3 axis on the external cryostat wall and internally on the cryostat getter attached directly to the cold tip of the cooler. The test results were collected while the cryocooler was cooling and at the stable set point, at various levels of cooling power and with thermal control enabled and disabled.
As space debris in lower Earth orbits are accumulating, techniques to lower the risk of space debris collisions must be developed. Within the context of the Space Environment Research Centre (SERC), the Australian National University (ANU) is developing an adaptive optics system for tracking and pushing space debris. The strategy is to pre-condition a laser launched from a 1.8 m telescope operated by Electro Optics Systems (EOS) on Mount Stromlo, Canberra and direct it at an object to perturb its orbit. Current progress towards implementing this experiment, which will ensure automated operation between the telescope and the adaptive optics system, will be presented.
We report the design evolution for the GMT Integral Field Spectrograph, (GMTIFS). To support the range of operating modes – a spectroscopic channel providing integral field spectroscopy with variable spaxel scales, and a parallel imaging channel Nyquist sampling the LTAO corrected field of view - the design process has focused on risk mitigation for the demanding operational tolerances. We summarise results from prototype components, confirming concepts are meeting the necessary specifications. Ongoing review and simulation of the scientific requirements also leads to new demonstrations of the science that will be made possible with this new generation of high performance AO assisted instrumentation.
Veloce is an ultra-stable fibre-fed R4 echelle spectrograph for the 3.9 m Anglo-Australian Telescope. The first channel to be commissioned, Veloce ‘Rosso’, utilises multiple low-cost design innovations to obtain Doppler velocities for sun-like and M-dwarf stars at <1 ms -1 precision. The spectrograph has an asymmetric white-pupil format with a 100-mm beam diameter, delivering R>75,000 spectra over a 580-930 nm range for the Rosso channel. Simultaneous calibration is provided by a single-mode pulsed laser frequency comb in tandem with a traditional arc lamp. A bundle of 19 object fibres ensures full sampling of stellar targets from the AAT site. Veloce is housed in dual environmental enclosures that maintain positive air pressure at a stability of ±0.3 mbar, with a thermal stability of ±0.01 K on the optical bench. We present a technical overview and early performance data from Australia's next major spectroscopic machine.
The recent availability of large format near-infrared detectors with sub-election readout noise is revolutionizing our approach to wavefront sensing for adaptive optics. However, as with all near-infrared detector technologies, challenges exist in moving from the comfort of the laboratory test-bench into the harsh reality of the observatory environment. As part of the broader adaptive optics program for the GMT, we are developing a near-infrared Lucky Imaging camera for operational deployment at the ANU 2.3 m telescope at Siding Spring Observatory. The system provides an ideal test-bed for the rapidly evolving Selex/SAPHIRA eAPD technology while providing scientific imaging at angular resolution rivalling the Hubble Space Telescope at wavelengths λ = 1.3-2.5 μm.
GMTIFS is the first-generation adaptive optics integral-field spectrograph for the GMT, having been selected through a competitive review process in 2011. The GMTIFS concept is for a workhorse single-object integral-field spectrograph, operating at intermediate resolution (R~5,000 and 10,000) with a parallel imaging channel. The IFS offers variable spaxel scales to Nyquist sample the diffraction limited GMT PSF from λ ~ 1-2.5 μm as well as a 50 mas scale to provide high sensitivity for low surface brightness objects. The GMTIFS will operate with all AO modes of the GMT (Natural guide star - NGSAO, Laser Tomography – LTAO, and, Ground Layer - GLAO) with an emphasis on achieving high sky coverage for LTAO observations. We summarize the principle science drivers for GMTIFS and the major design concepts that allow these goals to be achieved.
To achieve the high adaptive optics sky coverage necessary to allow the GMT Integral-Field Spectrograph (GMTIFS) to access key scientific targets, the on-instrument adaptive-optics wavefront-sensing (OIWFS) system must patrol the full 180 arcsecond diameter guide field passed to the instrument. The OIWFS uses a diffraction limited guide star as the fundamental pointing reference for the instrument. During an observation the offset between the science target and the guide star will change due to sources such as flexure, differential refraction and non-sidereal tracking rates. GMTIFS uses a beam steering mirror to set the initial offset between science target and guide star and also to correct for changes in offset. In order to reduce image motion from beam steering errors to those comparable to the AO system in the most stringent case, the beam steering mirror is set a requirement of less than 1 milliarcsecond RMS. This corresponds to a dynamic range for both actuators and sensors of better than 1/180,000.
The GMTIFS beam steering mirror uses piezo-walk actuators and a combination of eddy current sensors and interferometric sensors to achieve this dynamic range and control. While the sensors are rated for cryogenic operation, the actuators are not. We report on the results of prototype testing of single actuators, with the sensors, on the bench and in a cryogenic environment. Specific failures of the system are explained and suspected reasons for them. A modified test jig is used to investigate the option of heating the actuator and we report the improved results. In addition to individual component testing, we built and tested a complete beam steering mirror assembly. Testing was conducted with a point source microscope, however controlling environmental conditions to less than 1 micron was challenging. The assembly testing investigated acquisition accuracy and if there was any un-sensed hysteresis in the system. Finally we present the revised beam steering mirror design based on the outcomes and lessons learnt from this prototyping.
A representative range of the rotary mechanisms proposed for use in GMTIFS is described. All are driven by cryogenically rated stepper motors. For each mechanism, angular position is measured by means of eddy current sensors arranged to function as a resolver. These measure the linear displacement of a decentered aluminum alloy target in two orthogonal directions, from which angular position is determined as a function of the displacement ratio. Resolver function and performance is described. For each mechanism, the mechanical design is described and the adequacy of positioning repeatability assessed. Options for improvement are discussed.
This paper describes the software systems implemented for the wide-field, automated survey telescope, SkyMapper. The
telescope is expected to operate completely unmanned and in an environment where failures will remain unattended for
several days. Failure analysis was undertaken and the control system extended to cope with subsystem failures,
protecting vulnerable detectors and electronics from damage. The data acquisition and control software acquires and
stores 512 MB of image data every twenty seconds. As a consequence of the short duty cycle, the preparation of the
hardware subsystems for the successive images is undertaken in parallel with the imager readout. A science data pipeline
will catalogue objects in the images to produce the Southern Sky Survey.