Automated Astronomical Site-Testing Observatory Working Group 5th Meeting Friday, 24 May 1996 at 10am at UNSW in Sydney. Attendees: Michael Ashley (UNSW), Max Boccas (UNSW), Michael Burton (UNSW), Graham Davies (Software Engineering Associates, for UNSW), Mike Dopita (MSSSO), Gary Hovey (MSSSO), Andre Phillips (UNSW), Antony Schinckel (UNSW). Ashley suggested that the major goal for today's meeting should be to identify all the ``show-stopper'' engineering problems that we must solve before the AASTO can be remotely deployed. Examples include the optics-icing problem, and techniques to keep internal optics free of ice. GMOUNT Hovey described progress with the GMOUNT. The electronic systems are being constructed to fit into a Eurocard cage with an AT96 backplane. MSSSO have designed an adaptor card which will take a PC-104 card and convert it to AT96 (there was some concern expressed that the additional bus length may cause a problem for the low current driving capability of PC-104). MSSSO have designed a prototyping board that decodes the AT bus addresses. The boards are being made by Precision Circuits in Melbourne, and appear to be of excellent quality. Design work has begun on the axis control cards and MSSSO have almost finished the temperature and voltage measurement card. Eight temperatures are being measured (air, two motors, two preamps, the battery, and the circuit card (Gary - that's only 7?)), using 100 ohm platinum RTD sensors from Heraeus (cost AUD$13; Minco also has platinum RTDs). The sensors are IEC 751 DIN grade A, accurate to 0.15C (after linearity correction) in the range -100 to +200C; they are 30mm long by 2.6mm in diameter. All eight are being placed in series and fed with a constant current of 1.25 mA (anything larger would result in unacceptable self-heating), resulting in roughly 100 mV voltage drop per sensor. A wire connected (using ordinary 60/40 solder) to the junction between each pair of sensors gives a good approximation to a 4-wire current measurement. The voltages are being read by a 12-bit ADC with 34-channels. The AD590 temperature sensor was discussed. This is a nice unit since current is proportional to the temperature, and so a 4-wire measurement is not necessary. The disadvantage is that AD590s do not work reliably below -55C. The failure is catastrophic rather than just degraded performance, and is presumably due to differential thermal contraction. The GMOUNT cabling was discussed. There will be an 80mm diameter hollow central section for the cables (GMOUNT internal electrics, the ADIMM, and the AFOS). Hovey is searching for a DC-DC convert with a programmable or presettable current limit, so that he can design a power system for the GMOUNT/ADIMM/AFOS which is guaranteed to stay within the AASTO specifications. Cables from the AASTO to the GMOUNT need to be 50m long. The inductosyns for the GMOUNT were manufactured by Inductosyn International (used to be Farrand) - now the only manufacturer of these type of sensors. The angular resolution is 0.08 arcseconds, equivalent to 24 bits. Generic Tower No updates have been received from Yerkes on the GTOWER. Dopita will be at Yerkes next week and will discuss the interface between the GMOUNT and the GTOWER. A suitable compromise for the tower would be to have a smaller (4-metre?) steel tower this year, with a full size (12-metre?) carbon-fibre tower the year after. NISM/MISM Boccas described the changes to the NISM/MISM from the last meeting. The NISM and MISM will now be mounted in roof ports rather than through the side of the AASTO, as per the suggestion at the last meeting. They will sit in insulated cubic boxes of side 16-inches. A black-body (BB) source will be placed between the two boxes in such a way that both instruments can view it. The BB will be made of copper, with two blackened cones (50mm entrance aperture) drilled in the sides to provide the cavities. The NISM and MISM are now using reflective choppers, and will have two beams on the sky separated by 45 degrees. A narrow-band filter and lock-in amplifier will be used to measure the difference in the sky emission at the two positions. A platinum RTD will measure the temperature of the BB. The NISM and MISM will be filled with dry nitrogen to stop internal icing - there was some concern expressed about how to ensure that the nitrogen is truly dry, given the problems that people have experienced at the Pole. MSSSO have found a Honeywell microswitch that is rated for operation at -90C. ADIMM Dopita reported that tests are being done with the prototype ADIMM using the 30-inch telescope at Mt Stromlo for tracking/pointing. The microlens idea appears to be very successful, and gives an f/57 final beam. The separation between the spots provides a monotonic indicator of focus. Tests done during the daytime show that bright stars are usable. MSSSO are attempting to accurately mask half the CCD frame, to prevent contamination of the output signal during readout. Sutherland is working on the ADIMM acquisition software, and is making good progress. UV/VIS (now called AFOS - the Antarctic Fibre Optic Spectrometer) Boccas described the recent changes to the AFOS concept. The maximum swing-radius of the GMOUNT (0.60 metre), combined with the need for a 300-mm aperture (to collect enough light) and a f/3.5 beam (for injection into the fibres) has led to a decision to use a Newtonian configuration for the telescope optics. Discussions are being held with the AAO on the design of the fibre injection module. Three red and three UV fibres are being used, with a dichrohic beamsplitter immediately before them to eliminate the need for a filter assembly in the spectrometer. The fibres will have a 100 micron core (20 arcsecs on the sky at the image scale of 5 microns per arcsecond), surrounded by a polyamide buffer. No re-imaging optics will be necessary. The overall efficiency of the system will be 3% at 300nm, and 10% at 500nm. The fibre connectors will have 94% efficiency. There was considerable discussion about how many fibre segments to use between the telescope focal plane and the spectrometer in the AASTO. One segment will go from the focal plane and wrap around the GMOUNT altitude axis before terminating on the GMOUNT. Another segment will go from inside the GMOUNT, down a leg of the GTOWER, in a buried conduit to a mounting plate on the AASTO. A third segment will go from inside the AASTO to the spectrometer. It may be possible to combine these latter two segments into one, with consequent efficiency improvements, but slightly greater risk of failure. Ice detection/removable Various techniques for ice detection and removal on optics were discussed. Keeping the external optics well exposed to the atmosphere, with no edges or lips for ice to form, was considered good practice. Peltier devices may be useful if we need to generate heat using electricity, since the peltier effect assures greater efficiency that a simple resistor. Depositing resistive tracks on the internal surfaces of the optics seems a promising solution. Phillips proposed using a heat exchanger in the AASTO and a small fan to blow air through a large diameter well-insulated pipe out to the AFOS and ADIMM. Ashley discussed calculations he had made of the possibility of deflecting ice crystals using electrostatics (based on Dopita's suggestion for solving the icing of the microthermal sensors). The driver is that most of the icing on optics in Antarctica is due to micron-sized particles of ice (``diamond dust'') sticking onto the surfaces. If a mechanism could be found to deflect the particles, the icing may be reduced. The ice crystals form excellent dielectric cylinders, and so will be charged by induction when placed in an electrostatic field. The two equal and opposite charges on the ends of the cylinder will then experience a torque that will align the cylinder with the external field. If the external field is non-uniform, there will also be a net force on the cylinder which will cause it to move (not simply rotate). It is possible that this force could be used to deflect the crystals. Ashley's calculations showed that on the assumption of a 5 metre/sec wind, and a 300 mm aperture with a 1000 volt electrode at the centre, the force would be roughly 10E-6 times two weak. It may, however, turn out that with a better field geometry, and less pessimistic velocity assumptions, the deflection may do some good. It was suggested that an experiment be designed to test the idea, possibly this year at the Pole, using the IRPS ion pump power supply as a high voltage source. Data Acquisition Unit and Data Control Unit The Lockheed redesign of the DAU to use modern disk drives (the original magneto-optical drives used in the AGOs are no longer available) was discussed. The possible cost implications were a major concern. Power supplies Hovey is designing the GMOUNT power supplies to provide power for the ADIMM and AFOS. The following will be available: +/-12V at 0.8A; nominally for analog circuitry +/-24V at 2.0A; nominally for motors, heaters +5V at 3.0A; nominally for logic, computers The above supplies will be under the overall control of the GMOUNT computer, although each instrument is able to request that the power to itself be shutdown. In addition there is another +5V supply that will be continually available (provided the AASTO provides power to the GMOUNT). Next meeting The next meeting was scheduled for 10am Thursday 25th July at MSSSO. (NOTE: this was later changed to 10am Friday 26th July at UNSW). AASTO / Michael Ashley / mcba@newt.phys.unsw.edu.au