The Southern African Large Telescope (SALT) is a 10-m class 91-segment fixed altitude telescope located at Sutherland, South Africa. The segment alignment is maintained by inductively coupled sensors mounted on Sitall brackets beneath the segments. An extensive period of testing in environmental chambers and on the telescope has been conducted to establish the stability of the sensors and their response to temperature and humidity variations in the telescope chamber. We present some of the test results, including a demonstration of the ability of the sensors to maintain the alignment of the primary mirror over a period of 6 days.
The Southern African Large Telescope has till recently operated without active closed loop control of its Primary Mirror. The reason for this was that there were no suitable edge sensor system available on the market. Recently a system became available and SALT form Fogale Nanotech. The system consist of a sensor, cables and control electronics. The system was still under development and SALT was responsible for the integration of the sensors before deployment on the Telescope. Several issues still had to be addressed. One of these issues was the integration of the sensors at an appropriate production rate. The sensors was supplied as flexible pc boards with different types making up the transmitters and receivers. These flexible boards were bonded to ClearCeram Z L-Brackets before the appropriate connectors were installed. This paper describes the process used to integrate and test the sensors.
The development of an inductive edge sensor is in process for the control of the Southern African Large Telescope’s (SALT)1 segmented mirror primary. The original capacitive edge sensing system was not capable of maintaining the figure of the primary mirror due to excessive noise and a severe sensitivity to humidity despite exhaustive attempts at characterisation1. The prototype of the inductive edge sensor has progressed to a mature industrialised version that is in the process of being installed and commissioned on SALT. The performance of the sensor in response to temperature and RH is very good with a maximum error of 10nm typical after temperature compensation. The noise and control characteristics of the array have been simulated in order to establish the maximum cumulative error and error rate tolerable for the SALT specific case. It has been established through simulation that over the expected 5 day alignment cycle, a maximum cumulative error of 30nm can be tolerated.
This paper describes the cleaning of M5, one of the four mirrors that make up the Southern African Large Telescope's
Spherical Aberration Corrector. As the top upward-facing mirror in a relatively exposed environment, M5 had
accumulated a considerable amount of dust and dirt during the six years it had been on the telescope. With the corrector
on the ground for re-alignment and testing, we had the opportunity to remove, wash and replace the mirror. Various
cleaning techniques were investigated, including an unsuccessful trial application of First Contact surface cleaning
polymer film - fortunately only to a small region outside the mirror's clear aperture. Ultimately, "drag-wiping" with
wads of cotton wool soaked in a 10g/l sodium lauryl sulphate solution proved highly effective in restoring the reflectivity
of M5's optical surface. Following this success, we repeated the procedure for M3, the other upward-facing mirror in the
corrector. The results for M3 were equally spectacular.
The Southern African Large Telescope (SALT) recently (2008) abandoned attempts at using capacitive mirror edge
sensors, mainly due to poor performance at a relative humidity above ~60%, a not infrequent occurrence. Different
technologies are now being explored for alternative sensors on SALT. In this paper we describe the design and
development of a novel prototype optical edge sensor, based on the application of the interferential scanning principle,
as used in optical encoders. These prototype sensors were subsequently tested at SAAO and ESO, for potential
application on SALT and E-ELT.
Environmental tests, conducted in climatic control chambers, looked at temperature and relative humidity sensitivity,
long term stability and sensor noise. The temperature sensitivity for height and gap were, respectively, 10nm/°C and
44nm/°C, while for relative humidity they were 4nm/10% and 50nm/10%, respectively. These either met, or were close
to, the SALT specification. While there were significant lags in response, this was due to the sensor's relatively large
mass (~200 gm per sensor half), which was not optimized. This is likely to improve, should a revised design be
developed in future. Impressively the sensor noise was <0.015 nm RMS, over three orders of magnitude better than the
specification. Our conclusions are that optical edge sensing is a viable technique for use on segmented mirror telescopes.
At the Southern African Large Telescope (SALT), in collaboration with FOGALE Nanotech, we have been testing the recently-developed new generation inductive edge sensors. The Fogale inductive sensor is one
technology being evaluated as a possible replacement for the now defunct capacitance-based edge sensing system.
We present the results of exhaustive environmental testing of two variants of the inductive sensor. In addition to the environmental testing including RH and temperature cycles, the sensor was tested for sensitivity to dust and metals. We also consider long-term sensor stability, as well as that of the electronics and of the glue used to bond the sensor to its supporting structure. A prototype design for an adjustable mount is presented which will allow for in-plane gap and shear variations present in the primary mirror configuration without adversely disturbing the figure of the individual mirror segments or the measurement accuracy.
Construction of the Southern African Large Telescope (SALT) was largely completed by the end of 2005 and since then
it has been in intensive commissioning. This has now almost been completed except for the telescope's image quality
which shows optical aberrations, chiefly a focus gradient across the focal plane, along with astigmatism and other less
significant aberrations. This paper describes the optical systems engineering investigation that has been conducted since
early 2006 to diagnose the problem. A rigorous approach has been followed which has entailed breaking down the
system into the major sub-systems and subjecting them to testing on an individual basis. Significant progress has been
achieved with many components of the optical system shown to be operating correctly. The fault has been isolated to a
major optical sub-system. We present the results obtained so far, and discuss what remains to be done.
The SAMS (Segment Alignment Measurement System) is a
capacitance-based edge sensing solution for the active
alignment of the 10m SALT segmented primary mirror. Commissioning and calibrating the system has been an ongoing
task in an attempt to counteract the unfavourable response of the sensors to high humidity conditions and high dust
levels. Several solutions were implemented and tested including
real-time feedback systems and the application of
In parallel with the continuing efforts to improve the performance of the capacitive sensors, we have also been testing a
prototype inductive sensor developed by Fogale Nanotech that is of a very similar flexible plate construction.
In this paper we present the results obtained and performance gains achieved thus far with the capacitive edge-sensing
system as well as a performance comparison of the Fogale inductive sensor to the capacitive edge sensor.
Segmented primary mirrors dominate the current generation of 10m class telescopes as well as the designs for the next
generation of Extremely Large Telescopes (ELT's). The complex nature of these telescopes is demonstrated by the long
time periods associated with their commissioning and the difficulty of performing high precision optical alignments.
However, additional tools to provide in situ measurements of their optical alignment can be provided by making use of
the individual mirrors of a segmented primary; with the ability to move in six degrees of freedom, the individual mirrors
can be deployed to trace multiple optical paths through the telescope. In this paper we describe how it is possible to use
the segments themselves to create a number of Hartmann masks that allow focus and other aberrations to be measured
using a standard imaging camera rather than a dedicated wavefront sensor. The Southern African Large Telescope
(SALT), with a primary mirror composed of 91 1m segments, is used as an example. The segments were arranged to
create eight Hartmann masks to measure the optical alignment. Through imaging data obtained at the telescope, the
sensitivity of this method to changes in focus along with aberrations inherent in the system is demonstrated through
Zernike polynomial fits to the observed patterns. Finally, we present simulations of possible patterns for use on future
The 10-m class Southern African Large Telescope (SALT) at Sutherland, South Africa, was inaugurated in November 2005, following completion of all its major sub-systems. It is the largest single optical telescope in the southern hemisphere. The SAMS (Segment Alignment Measurement System) is a unique capacitive edge sensing solution for the active alignment of the SALT primary mirror. Twelve thin film edge sensors are bonded directly onto the edges of each of the 91 segments, with heat-generating control electronics housed remotely in temperature-controlled enclosures. The SAMS is capable of measuring the tip/tilt and piston of each segment, as well as the change in global radius of curvature, a mode normally undetected by such a system. The primary objective was to build a system that offered an excellent cost-to-performance ratio without sacrificing measurement accuracy, a very necessary requirement because of the scale and number of sensors required for large segmented mirrors. This paper describes the results obtained during the commissioning and calibration of the completed system.
On completion by the end of 2004 the Southern African Large Telescope (SALT) being erected in Sutherland, South Africa, will be the largest single optical telescope in the southern hemisphere. This paper addresses the process of designing, building and demonstrating a high performance primary mirror system for SALT that meets the overall telescope requirements. Throughout the process consideration was given to the fact that SALT is budget sensitive, which required careful allocation of funds among the various subsystems, innovative designs, and using COTS components where possible. The process delivered subsystems with a very high cost-performance ratio.
The design of the Southern African Large Telescope (SALT), which is based closely on the Hobby-Eberly Telescope (HET) at the University of Texas but includes advances incorporating lessons learned from HET, is briefly reviewed. The flowdown of requirements from the optical error budget to the primary mirror control subsystems is presented. The techniques and algorithms used by the Center of Curvature Alignment Sensor (CCAS) to measure segment tilt and piston and estimate the global radius of curvature of the primary are discussed in detail. The steps in the process that allows CCAS to capture and identify segments misaligned by more than 70 arcsec and bring them into alignment with residual errors less than 50milli-arcsec is fully described. Next, the hardware and software designs of CCAS are presented, as well as the results of laboratory performance testing. CCAS has been installed and integrated with the primary mirror control system. Performance results of the integrated system over a range of environmental conditions will be shown. Finally, the overall results of this project are summarized and suggestions for future improvements presented.