Version 2.2 (19 March 1999)
Functional Specification for 6dF
Fred Watson, Quentin Parker,
Ian Lewis, Lew Waller, Ken Russell, Stan Miziarski,
Brian Hingley, Gabriella Bogatu, Chris McCowage & Chris Evans
Anglo-Australian Observatory
Coonabarabran NSW 2357
Contents
1. Scope
This document contains the scientific and technical specification for the 6dF system for the UKST. It incorporates the earlier preliminary specification included in the Phase A study, together with developments and details that have arisen since. It also incorporates amendments to Versions 1.0, 2.0 and 2.1 of the Functional Specification (dated 25 Sep 1998, 14 Dec 1998 and 21 Jan 1999).
Where appropriate, performance tests are suggested. Software requirements (I.J.Lewis) are included, and the electronic control specification (L.G.Waller) is appended.
2. Requirements
The 6dF instrument consists of:
The minimum functionality of the instrument when commissioned will be such as is provided by 1, 2 (one field-plate unit), 3 and 4(i), (ii) and (iv). The second and third field-plate units and the field-plate lifter will be post-commissioning, but before the completion of the project.
At its meeting in July 1998, AICAAT highlighted the following requirements:
These are addressed in this specification document.
In general, the area of commonality with OzPoz is limited to the mechanical design of the positioner (item (1) above) and some parts of the software (particularly 4(i)). The field-plate units, control electronics and remaining software are not common to OzPoz. Although some of the functionality is similar, the realisation is different (e.g. in the use of different programming languages to code the software).
Likewise, because 6dF and OzPoz are destined for very different applications, there is little common ground in the detailed specifications that follow. Requirements that are reasonably similar are indicated with an asterisk.
3. Basic characteristics and performance
3.1 Minimum number of fibres
Each field plate should carry 150 spectroscopic fibres (in addition to guide fibres).
The 150 fibres are the maximum that can be imaged on the existing CCD with the current spectrograph, and assumes the CCD is rotated by 90 degrees so the longest CCD dimension (578 pixels) is perpendicular to the dispersion direction. There is a corresponding 27% loss in spectral coverage.
3.2 Fibre type and core diameter
The preferred fibre type is Polymicro fibre drawn from a broad-band Heraeus preform. Jacket will be polyimide.
The current FLAIR II fibres are standard high-OH fibres from Polymicro. The Heraeus-type fibres offer freedom from absorption bands in the red, however.
Core-diameter will be 100um.
Core diameter of the FLAIR II fibres is 100um. Observed objects are typically B<17.1 galaxies (which subtend on average 22 arcseconds or ~330um) and 100um is well-matched to the cores of these objects. Because of the small aperture of the UKST, sky-contamination of point-source targets in these large fibres is seldom a problem, with read-noise representing the more serious limit.
3.3 Positional accuracy of fibre placement*
This should be ±15um (1.0 arcsec at the plate scale of the UKST) over the full 6-degree field, in the coordinate system of the positioner.
The tolerance represents 15% of the core of a typical galaxy observed with 6dF, but only 4% of its isophotal diameter.
Tests: Positioning accuracy to be tested by centroiding with the gripper camera. The integrity of the (r,theta ) carriage's coordinate system is established by means of fiducial marks (see 4.3).
3.4 Fibre placement speed
Must be such as to allow all fibres to be reconfigured (with any necessary intermediate park moves) within one hour. For ~150 fibres, this implies 24 seconds individual placement time.
This is in order to keep up with the anticipated minimum turn-around duration between observed fields when undertaking the 6dF galaxy redshift survey to B=17-17.5.
Tests: Straightforward timing tests.
3.5 Operating temperature range
The positioner and field-plate units must be able to work within the temperature range 0 to 30 deg.C.
Tests: Environmental testing is required.
4. Three-dimensional registration of fibre
array with targets
4.1 Target object coordinates
The positioner will require object coordinates accurate to 0.5 arcsec RMS.
These will come from some suitable database (e.g. APM, SuperCOSMOS). They will then be mapped to the field plate system with direct allowance for likely temperature effects etc.
4.2 Transformation to positioner coordinates
Coordinate transformations accurate to 1 arcsec RMS over the full 6-degree field will be required to relate sky coordinates to positioner coordinates.
These should be established once and for all using the gripper camera to measure a UKST film of an astrometric field tensioned over a field plate (2dF Poscheck analogue). Together with focus-determination and squaring-on requirements, this demands a photographic capability for the field-plate units, with some way of supporting the film in the focal surface rather than on the field-plate surface (whose radius of curvature is smaller). This is expected to be by means of an interlayer of suitable thickness.
Tests: Since there is no equivalent of the 2dF focal-plane imager on 6dF, the tests will also need to be carried out using photographic film.
4.3 Integrity of positioner coordinate system*
A grid of approximately 25 illuminated fiducial marks with accurately-known positions will be inlaid into each field plate.
The fiducial marks are for 2dF-type field surveys, although it is expected that only offset and rotation will change in routine use. It is likely that the disposition of the fiducial marks will be radial in the field. They will be illuminated by fibres.
Accurate determination of the positions of the marks can be either directly with a two-coordinate measuring machine (with correction of the measurements to the spherical surface) or photographically with a film overlaying the field plate (although this might require the machining of a mandrel with opposite curvature). The film could then be measured with SuperCOSMOS.
4.4 Temperature compensation*
The positioner must include compensation for thermal expansion of the field-plate at ambient temperature.
Temperature will be sensed (1) at the field-plate, (2) within the telescope, and (3) within the positioner enclosure.
4.5 Curvature of field-plate
The radius of curvature of the field-plates must be such as to allow the array of fibre-end virtual images to conform to the 3070mm radius of curvature of the focal surface with an accuracy of ±50um.
The value of the radius of curvature of the field plate depends on the final geometry of the fibre buttons, but must be such as to give the spherical surface defined by the virtual images of the fibre ends in their microprisms a radius of curvature of 3070 mm. The required uniformity of the spherical surface to maintain adequate focus over the full field is ±50um.
4.6 Coincidence of field plate with telescope focal surface
The field plate must be located within the telescope so as to render the fibre array coincident with the telescope focal surface to an accuracy of ±50um.
The method of achieving this coincidence with photographic and FLAIR plateholders is to obtain focus plates, allowing the plateholder to be squared-on relative to the telescope axis, and the best-focus position to be determined. They also reveal the extent of the surface uniformity. Similar exposures (on film) will be required for 6dF, and will be made with an interlayer to elevate the film above the field-plate surface to mimic the focal curvature (see 4.2).
4.7 Field acquisition*
There will be four guide fibre bundles disposed evenly around the field plate, each consisting of seven hexagonally-packed fibres.
The diameter of the fibres used in the guide bundles is yet to be decided. It is possible that they could be of the 33um/125um depressed-index cladding type as used (singly) in the current FLAIR system. The fibres will feed the existing intensified CCD camera. Pointing tests will need to be done with a guide bundle placed at the centre of each field-plate unit. However, it is not necessary for all the guide bundles to reach the field centre.
Because of the greater thickness of the guide bundles compared with the single fibres, an alternative method of retracting them will need to be devised (see 6.5).
4.8 Field rotation
Each field-place unit must be provided with a total of 10 arcminutes of rotation. The rotation must be controllable to 1 percent, and must be provided with an encoder.
The rotation facility is required to compensate for the rotational offset introduced by the polar-axis elevator. It is not necessary for the field-plate rotator to be in its mid-position when fibres are being positioned, since a rotational offset can be introduced into the coordinate transformations. Therefore, the piezo actuators can be relaxed during positioning.
4.9 Telescope guidance
Guidance of the telescope after acquisition of the field will be by means of the CCD autoguider on the UKST's north guide telescope.
The offset between the main telescope and guide telescope axes will have to be determined.
5. Fibre positioner
5.1 Gripper lifting force*
The gripper should be able to lift a mass of no less than twice the magnetic adhesion force with a stroke of approximately 14mm.
Tests: Can be checked using a balance.
5.2 Positioner location
The positioner will be situated off the telescope in a dust-free enclosure in the dome area.
The enclosure will be pressurised by means of a fan blowing in filtered air. The location is likely to be on the SW wall of the dome between the emergency exit and the dome-power switch-box.
5.3 Field-plate registration and identification
A means of registering and automatically identifying each field-plate will be provided.
The field-plate units will be mounted in the positioner on the same mounting lugs as those providing registration within the telescope.
5.4 Fibre identification*
A means of automatically identifying each spectroscopic fibre in a given field-plate unit will be provided.
As well as a miniature CCD TV camera, the gripper carriage will be equipped with light sources that can be used to forward-illuminate any fibre within the field of the TV. (These will be offset radially from the gripper axis, one on either side.) The positioner's back-illumination unit (for target-object fibres) will also house a TV camera that can inspect the whole of the fibre slit so that an individual forward-illuminated fibre can be identified. The pattern-recognition required for this must be included in the positioner control software.
5.5 Provision for manual positioning and checking of fibres*
The engineering interface should allow individual fibres to be parked/placed/shifted etc.
5.6 Post-observation positional verification*
There must be a capability to provide post-observation checking of fibre positions to determine any shifts from their initial positions that may have occurred during the loading or unloading process (see 7.2 item 14).
The gripper TV will be used to obtain and centroid an image of each back-illuminated fibre in turn. (In fact, this process will normally be carried out during the reconfiguration to the next field.)
5.7 Safety interlock*
The positioner must be equipped with a guard rail and an emergency stop button.
6. Field-plate units
6.1 Button type and diameter*
The magnetic buttons are cylindrical to match the circular collet of the 6dF gripper. Their greatest diameter will be such that the smallest separation of any two fibres will be 5mm (~5 arcmin).
6.2 Button manufacturing tolerance
For each button, the height of the virtual image of the fibre in its microprism above the button base will be within ±25um of the specified value.This is in order to achieve the required consistency of focus over the virtual fibre array.
Tests: Checks should be carried out on assembled buttons using a longitudinal travelling microscope.
6.3 Button adhesion to focal surface*
Each button should require a force to lift it from the focal surface that is great enough to provide resistance to lateral shocks, but will allow the gripper collet to lift it.Experiments carried out with the prototype buttons indicate that the adhesion is sufficient to withstand lateral accelerations on the order of 100g. These are far greater than will be experienced in normal service.
Tests: The proposed 6dF button design can be tested with a balance to demonstrate that its adhesion is comparable with the FLAIR interim upgrade design (done).
6.4 Microprisms*
Microprisms will be of SF5 glass with dimensions such that no vignetting of the telescope beam will take place.
SF5 is used so that total internal reflection of the full f/2.5 beam takes place. The prisms have a long (~4mm) extension in the z-direction so that the button can be elongated in this direction as required by the gripper design. The prism cross-section should be cylindrical to minimise the diameter of the button extension. The diameter of the cylindrical prism will be such as to allow the full f/2.5 beam to enter the input face; it is determined by the overall height of the button.
6.5 Fibre retractors
Fibres will retract under tension of 5–10 g into the body of the field-plate unit. All fibres (except guide fibres) must be able to reach to the centre of the field.
The fibres must not be subjected to a bending radius of less than 10mm.
A successful, novel, coiled spring retractor prototype has already been built. It is capable of providing the required fibre tension, which is sufficient to ensure retraction given the limited velocities and accelerations that each fibre button will be subjected to. With the circular field, all fibres must be able to reach to the centre of the field with a small safety margin. Investigation is required to determine how best to retract the guide fibres.
6.6 Interchangeability of retractors*
It is highly desirable that individual retractors or small groups of retractors must be replaceable in the field-plate unit.
Whether it is possible for retractors to be replaced when fibres are damaged depends on the details of the slit-unit construction.
6.7 Area of accessibility for fibre placement
The radius of the area of field within which fibres can be positioned will be 3.0 degrees (161mm), less a small annulus for fibre parking.
The physical area over which placement will take place should maximise field coverage on the sky. Tiling arrangement of surveys will not be dictated by the UKST photographic field size.
6.8 Field-plate unit physical dimensions and fittings
It must be possible to load each field-plate unit by means of the telescope plateholder elevator, so the units must have the same external dimensions, mounting lugs and dummy vacuum and nitrogen connections as the existing FLAIR and photographic plateholders.
The existing plateholders have a dark-slide, which is essential for the elevator interlocks to function, and provides a convenient means of carrying them.
6.9 Field-plate unit weight
The weight of each field-plate unit must not exceed 14 kg. If that is not possible, an external lifting device to assist with loading in the telescope must be provided.
There is concern over the weight of the existing FLAIR II fibre plateholders, which is about 17kg. This is not only an OH&S issue, but it is more difficult to avoid mechanical shock when handling an excessively heavy field-plate unit, with consequent risk of disturbing magnetic buttons. It may also be the cause of flexure that is observed as the fibre field is tracked across the sky.
Work on a lifting device is, in any case, in hand.
6.10 Fibre database*
There must be a database to record the status and progress in manufacture/repair of each fibre on each field-plate unit (see 7.2.13).
7. Software requirements
This section is designed to set up a framework of requirements which will form the basis of a formal software design to be performed by the authors of the software. It is not expected to be complete, nor intended to impose any particular solutions. Software tests will need to be specified in due course.
Three subsections cover the following areas of software:
It is not expected that the data reduction software will be delivered at the time 6dF is commissioned, and is seen as beyond the scope of the core project. The most likely source of the 6dF data reduction software is the Galaxy Survey project team. However the software does have implications for the required functionality of the pre-observation and positioner software, and these are considered here.
There is an additional software requirement to modify the code written by Paddy Oates to run the FLAIR CCD camera. This is to allow object/fibre data provided by the positioner to be written to the headers of each data frame, as well as simplifying some of the current functionality of the software.
The details of this requirement are not considered here, but it should be a priority to determine the amount of work required, and what resources are needed.
7.1 Pre-observation software*
1. Should use the same or compatible format to the 2dF configuration software. (ie. ASCII and SDS formats)
2. Should remain as similar as possible to the 2dF configuration software in terms of its use.
3. Should be a standalone utility running under Solaris for distribution to other university/observatory sites.
4. Output file should be ASCII or SDS format. It needs to be decided whether positional information should be in x,y or RA,Dec at this point.
5. A second optional output file should be a list of the unallocated objects in a suitable format for re-entry into the configuration program.
6. A third optional output file should be a FITS file containing the fibre allocation information. This may then be combined with the CCD image FITS file before archiving or reducing the data. This function may be superseded by the functionality of the positioner software below but should be included for testing purposes and to allow recovery of information in case of problems.
7.2 Positioner software
1. Controls all access to fibres.
2. Is capable of taking an ACSII or SDS output file from the configuration software and decoding it (and if necessary converting to useful x,y or r,theta coordinates).
3. Can work out the sequence of fibre moves to move a set of fibres from their current configuration to the future configuration without necessarily parking all fibres.
4. Organises movement of all fibres in the previously determined sequence.
5. Provides a simple, easy to use user interface for normal operation. Will only allow high level safe actions, e.g. select a file for configuration, set up a field, abort setting up a field and pause/restart setting up a field.
6. Provides an engineering interface with direct access to individual actions in the positioner software (e.g. open jaws, park a fibre, move a motor, centroid an image). Engineering interface must be able to run on a remote X-terminal and connect to a previously running system.
7. Must provide a useful mimic running on remote X-terminal. (This requirement yet to be clarified with Ian Lewis.)
8. All robot and fibre parameters must be configurable via the engineering interface and if required saved to disk.
9. The fibre handling software must provide full error condition checking at all stages during the fibre movement process and must halt immediately on an abnormal condition with full logging of the conditions that caused the fault. Further actions should be prevented until the fault status has been cleared.
10. The positioner software should have the ability to set up all of the required error tables (i.e. autoload each fibre button in turn).
11. The fibre handling software should use the positioning information to learn error tables automatically during normal use.
12. After setting up a field the positioner software should generate an output file in FITS format containing the fibre information for that field. This is the file that would generally be combined with the CCD image FITS file for archiving and data reduction. (Details to be clarified with Ian Lewis.)
13. A data-base will be required during fibre manufacture and assembly and during operations to provide the fibre to number to slit position correspondence.
14. The positioner software should also have the functionality to verify the position of all the fibres. After loading the field plate into the positioner the gripper TV system should be driven to the location of each fibre in turn and the position of each fibre measured. This information should be compared with the original fibre location after the fibre was placed on the field plate. This should allow monitoring of any jolting of the fibres during the load/unload procedure in the telescope and possibly account for missed target objects. It should be possible to generate a further FITS header file at this point giving the astronomer the option of using either file for inclusion with the CCD data.
15. The positioner software should be able to run in full simulation mode to allow software testing and development in the absence of the hardware and to allow debugging of situations potentially dangerous to the hardware.
7.3 Data Reduction
As part of the 6dF project, it is necessary to specify everything required for a pipeline data reduction (like 2dFdr) to operate in the future.
1. Will rely on fibre information in FITS header
2. Will require additional FITS header items in CCD data, e.g. arc, flat, object, sky, grating angle, central wavelength, lamp names etc.
3. Possible requirement for a permanent 6df utility to add certain FITS keywords to the CCD image data and combine the FITS fibre information.
4. A PC CDROM/DVD drive at UKST for archiving data is strongly recommended.
7.4 Detailed functionality required from user interface*
It is envisaged that only limited functionality will be required from the user interface, i.e. to select a file to be set up on a field plate and proceed to move the fibres, abort the setup or pause and restart the setup.
In addition, a means of copying files from remote locations will be required.
The detailed steps hidden behind this user interface may be as follows:
1. User selects file to be configured on the field plate.
2. Read in and decode the input file.
3. If necessary convert to robot coordinate system (probably x,y).
4. Identification of field plate is read from hardware.
5. Legality of field configuration is checked.
6. Sequence of moves required to go from current to next configuration is generated (and stored in a file on disk).
7. Turn on the back illumination for the fiducial marks.
8. Survey the fiducial marks to determine the coordinate transformation between the encoders and the real world of the field plate.
9. Move each fibre in the sequence previously determined.
10. Keep mimic display updated after each fibre movement.
7.5 Detailed functionality required from engineering interface*
This is simply a list (probably not complete) of the actions one would wish to be able to perform from the engineering interface based on experience with 2dF.
Note that most of the functionality of the user interface in the previous section is achieved by packaging up the functionality required below.
1. Set up a target field from the current field including the abort, pause, restart options.
2. Move a single fibre
3. Park a single or range of fibres
4. Survey the fiducial marks and if required save the results on disk.
5. Identify the currently loaded field plate.
6. Move a specified axis to a specified position.
7. Turn on/off back illumination.
8. Move the gantry as a whole, (ie to a specified XY or R,theta)
9. Initialise the hardware and software.
10. Open/close gripper jaws
11. Centroid the tv image, determine presence or absence of an image and its location, saving results if necessary.
12. Calibrate tv image scale and rotation.
13. Calibrate rotation centre in tv coordinate system.
14. Display, manually edit and set parameters for any fibre in the currently-loaded field plate.
15. Display, manually edit and set operating parameters for the positioner control system.
16. Provide differing levels of debug information/logging.
17. Park/unpark the gripper gantry.
APPENDIX
6df Positioner electronic control
Lew Waller
August 10, 1998
This document outlines a design proposal for the electronics system needed to control the 6dF Fibre Positioner. The Positioner contains a number of electro-mechanical systems, all of which are intended to be computer controlled.
It should be noted that this is very much a first draft of the design and that it is likely that further changes will be made during the course of the design phases.
The following is a description of the electro-mechanical systems in the Positioner, with regard to the means by which computer control for these systems, can be implemented.
The Theta drive is a servo axis, driven with a brushless DC rotary motor and with an incremental encoder providing position and velocity feedback. As this axis carries a number of cables, it cannot rotate past 400 Degrees. It has two sets of over travel limit switches (motion controller and amplifier) at each end of its travel. The servo axis is controlled using a servo motion controller driving an AC amplifier.
The R drive is a servo axis, driven with a brushless DC linear motor and with an incremental encoder providing feedback. It has two sets of over travel limit switches (motion controller and amplifier) at each end of its travel. The servo axis is controlled using a servo motion controller and an AC amplifier.
The Z drive is implemented with a pneumatic system, enabled with a solenoid which is computer controlled by a relay. Its position is indicated by limit switches at each end of its travel which are computer readable.
The Button Clamp is implemented with a pneumatic system, enabled with a solenoid, which is computer controlled by a relay. Its position is indicated by two computer readable limit switches, which indicate whether it is open or closed.
The Button Camera is a TV frame rate video camera (neither cooled nor intensified). The video signal from the camera feeds a frame grabber module, which can digitise TV frames from the camera. The computer has full control over the frame grabber and access to the digitised data.
A number of computer controllable relay outputs will be available to switch back illumination lamp sources located externally to the Positioner.
Since there are a number of pneumatically driven systems in the Positioner, it will be necessary to ensure that the supply for the pneumatic system is fully functional. There will be at least two pneumatic system pressure sensors in the system, which will indicate the status of the pneumatic system. These status sensors will be computer readable, and will have the capability of interrupting the computer should the pneumatic system fail at any time.
Control of the 6dF Positioner system is a variation of the models used to implement the 2dF System on the AAT or the VLT OzPoz system. The significant difference is that instead of using a VMEbus system running VxWorks to implement the low level control, the 6dF system will use a Windows NT workstation computer. The reasoning behind this is that a PC based system will be very much less expensive than a VMEbus system, and provides an environment which is familiar to software development personnel at the Schmidt telescope.
A block diagram of the proposed 6dF Positioner control system is available.
The 6dF Positioner system will be controlled through a combination of high level software running on a Unix workstation and lower level software running on a Windows NT workstation. The Unix workstation and the NT workstation will be interfaced to the Schmidt Telescope Local Area Network. The NT workstation will be located close to the Positioner hardware. Unlike 2dF, the 6dF Positioner hardware will not be located on the telescope. Instead, it is proposed that the 6dF Positioner be located in a small room on the main floor of the Schmidt telescope building.
The Windows NT workstation will include a number of interface boards, including a servo motion controller, a frame grabber and a digital input/output interface. The Windows NT workstation will be interfaced to an Amplifier rack and to a signal conditioning electronics rack for relay output and limit switch inputs, to implement the Positioner control system. There is a power supply unit providing analog voltage levels for the electro-mechanical systems.
The Windows NT workstation will be suitably configured with CPU, memory and disk space suitable for the task. A PCI graphics adaptor, Ethernet interface, a 17-inch monitor and a CD-ROM reader will also be included.
In addition to the normal PC peripherals, the Windows NT workstation will include the following:
Frame Grabber - This is a PCI bus based Video Camera frame grabber board (for example a Data Translation DT3155), which can digitise video frames from the Button Camera.
Digital I/O - The Digital Input/Output board (for example, a Data Translation DT2820) will interface to a number of relay drivers and limit switch input conditioners, to allow the computer to control various solenoids, and to read the status of various limit switches.
Servo Motion Controller - The servo motion controller will be used to control the R and Theta servo axes. A suitable PC Bus motion controller would be a Tech80 5650A, which has been used previously at the AAT and Schmidt telescopes. This device can provide closed loop control for one to four axes.
The amplifier rack contains the motor amplifiers for the R and Theta servo systems in the positioner. The servo amplifiers generally accept a torque command input and an enable input from the motion controller. They also send back a fault output to the motion controller. It is intended to use Kollmorgen SE03 SERVOSTAR Amplifiers for the R and Theta motors (which are also manufactured by Kollmorgen). The AC motor servo amplifiers include their own power supply, which will be located in the AC amplifier rack. This will be a Kollmorgen PA08 SERVOSTAR power supply.
The relay output and limit input signal conditioning rack contains modules with control relays and conditioning electronics for limit switches. The relay/limit modules will be identical and will include a number of relays and a number of limit switch inputs. The relay output and limit input rack is connected to the Digital I/O interface board in the NT workstation.
Control relays will be solid state devices, thus providing galvanic isolation between the analog supplies and the digital supplies in the PC.
All electro-mechanical status sources (limit switches, pressure sensors) will be powered from the analog supply. Their status signals will be conditioned using opto-couplers and thus will be isolated from the digital supplies in the PC.
The analog power supply rack contains power supplies to generate all the analog voltages required by the system. This includes:
The motion controller requires +12V and -12V supplies for its analog circuitry.
It is possible that additional power supplies (e.g. +24V) will be required for relays and solenoids.
The following table summarises control of the mechanisms in the Positioner:
|
Mechanism |
Controller |
Amplifier |
Method |
Encoder |
Limits |
|
Positioner Theta Drive |
Tech80 |
AC |
Servo |
Incremental |
Yes - Amp and Controller |
|
Positioner R Drive |
Tech80 |
AC |
Servo |
Incremental |
Yes - Amp and Controller |
|
Positioner Z Drive |
NT CPU |
Relay |
Yes |
||
|
Positioner Button Clamp |
NT CPU |
Relay |
Yes |
As much use as possible will be made of Commercial Off The Shelf (COTS) components. The only items not likely to be procured from COTS sources will be interfaces specific to the Positioner and its associated mechanisms, and any other items not available or easily adaptable from off the shelf components procured through commercial sources. It is not envisaged that there will be a large number of these. In addition, as much use as possible will be made of the concepts developed and used in the 2dF and OzPoz positioner systems.
All electronics racks, including the Amplifier rack, the relay output and limit input racks and the analog power supply rack will be installed in a 19-inch wide equipment cabinet. The PC case may also be mounted in this rack.
Cables shall be run between the Electronics Cabinet and the instrument hardware in a metallised duct. This duct will provide mechanical protection for the cables and RFI screening. There will be a connector breakout panel located on the instrument to permit all cables to be disconnected from the instrument.
A servo axis mechanism consists of a motor, an encoder and one or two over travel limit switches.
The motors proposed for implementing the R and Theta servo axes are DC brushless (i.e. an AC motor). These motors require require external commutation which will be done by the amplifier (using signals derived from Hall effect sensors in the motor).
All encoders used for servo mechanisms shall be incremental A/B quadrature encoders, with an index or home pulse. Sending critical encoder signals back to the motion controller along long cables has the high probability of loss of signal integrity through induced noise and interference. Encoder signals should be transmitted using balanced differential signalling, if that is not their default interface. Consideration could also be given to using low cost fibre optic signalling for encoder signal transmission.
All servo axis mechanisms (i.e. R and Theta) will have two pairs of limit switches at each end of the mechanism physical travel. The limit switches should be Normally Closed switches with current flow broken when the switch is activated. Limit switches will generally be implemented using contactless proximity switches.
The first limit switch activated at each limit of travel are "soft" direction overtravel limits which cause the motion controller to kill motion on that particular axis in that direction of travel. The second limit switch at each end of travel are "hard" overtravel limits and are connected to the amplifier. If a "hard" overtravel limit is activated, the amplifier is disabled and the motor is braked. If the amplifier used does not have an over travel limit switch function, these limits will be connected in series with the amplifier enable signal from the motion controller to the amplifier.
If proximity switches are used to implement the limit switch functions, they will be supplied with power from the motion controller analog circuitry.