Gemini Near-Infrared Integral-Field Spectrograph (NIFS)
|
|
|
Instrument Control System
The NIFS instrument is a complex arrangement of optical elements, mask wheels, filter wheels, gratings, and sensors which need to be thermally controlled as well as configured with high precision in order to assure maximum scientific potential and versatility. It is the function of the Instrument Control System (ICS) to control these operations, thus leaving the detector controllers unfettered. Broadly speaking, the ICS may be thought of as two separate control systems operating under the same VME single-board processor, interfaced to the VME backplane via a transition module which provides industry standard connectors to the computer. The detailed operation of the Optical Component Controller (OCC) and the Temperature Controller (TC) is described in this section.
NIFS will use the same motors and encoding philosophy as NIRI. This allows the NIFS ICS to be nearly a direct copy of the NIRI ICS. This section describes the operation of the NIRI ICS in the NIFS context, and also highlights areas where the NIFS ICS will differ from that used in NIRI. IfA will duplicate the NIRI ICS for NIFS. An overall schematic of the NIRI ICS is shown in Figure 1. Most of the information presented in the section has been derived from NIRI design review documentation. The explanations of operating principles reflect the interpretation by RSAA of available NIRI information.
Figure 1: NIRI Instrument Control System schematic.
7.1 Optical Component Controller (OCC) Electronics
The Optical Component Controller (OCC) is responsible for moving and positioning all of the optical elements associated with the NIFS cryostat. There are two major assemblies within the cryostat, the OIWFS and the NIFS spectrograph. Externally there is also a light tight window cover.
The NIFS OCC will control and position the following 7 active elements:
|
OIWFS X-Axis Gimbal |
linear positioned |
|
OIWFS Y-Axis Gimbal |
linear positioned |
|
OIWFS Filter Wheel |
rotary positioned |
|
Spectrograph Focal Plane Mask Wheel |
rotary positioned |
|
Spectrograph Order Blocking Filter Wheel |
rotary positioned |
|
Spectrograph Grating Wheel |
rotary positioned |
|
Environmental Cover |
open/closed |
7.1.1 Control Strategy
The NIFS active elements will be controlled using copies of NIRI mechanisms. An open loop scheme using stepper motors and dead reckoning is employed using Hall effect sensors to provide "home" or "datum" positions. Once a zero-point is established, the number of steps taken by the stepper motor determines position.
Hall effect sensors consist of a small block of semiconductor material, such as indium, sandwiched between two electrodes. They produce a weak DC voltage, typically 10 millivolts per kilogauss across the electrodes, in the presence of a magnetic field. The polarity of the potential difference is a function of the "pole orientation" of the magnet field (north-seeking or south-seeking).
The position monitoring and calibration system will use stationary Hall effect sensors and small position-indicating magnets mounted on specific moving parts of all stepper motor driven optical component mechanisms. Additional (redundant) Hall effect sensors, which are software selectable, will be incorporated to improve reliability.
All movable elements within the NIFS cryostat can be categorized as either rotary or linear positioned with the environmental window being unique in that it is an “open/closed” type of mechanism.
7.1.1.1 Rotary Positioned Elements
The OIWFS Filter Wheel, spectrograph Focal Plane Mask Wheel, spectrograph Order Blocking Filter Wheel, and spectrograph Grating Wheel are rotary positioned elements. All rotary parts are positioned absolutely by a detent system, and driven by a modified "Geneva Drive" (Figure 2). "Home" position is determined by a single magnet and Hall effect sensor. The magnet is mounted on the rim of the wheel and passes perpendicularly under the stationary Hall effect sensor.
Figure 2: Modified Geneva drive with locking detent pin.
The output from the Hall effect sensor, as the magnet passes under it, rises to a peak as more magnetic flux passes through the area of the Hall effect sensor and then declines again as the magnet moves away (Figure 3). Figure 4 shows how the Hall effect sensor output varies in relation to the steps of the stepper motor.
Figure 3: Hall effect sensor output voltage versus magnet angle.
Figure 4: Hall effect sensor output voltage versus number of stepper motor steps.
The "Home" position is found by finding the midpoint between two points of identical fluxes (output voltage from the Hall effect sensor). This should provide a great deal of repeatability and be tolerant to sensor sensitivity and DC drifts or offsets introduced by the Hall effect sensor amplifiers, as well as the non-linear wheel motion resulting from the use of Geneva mechanisms.
Redundancy is provided by placing another Hall effect sensor in parallel with the first. Rather than mounting it on top of the primary, it is placed to the side on a slightly smaller radius. By then adding additional magnets of variable field strengths, a two track system quasi-positioning system can be incorporated with little additional effort.
The position indicating magnets are Samarium Cobalt rod shaped magnets, which produce a magnetic field intensity of approximately one kilogauss. Intensity will be controlled by adjusting the depth of the mounting hole.
Being a rotary system there is no need for limit switches or hard stops to prevent excess travel.
7.1.1.2 Linear Positioned Elements
The OIWFS X-Axis Positioner and OIWFS Y-Axis Positioner are linear positioned elements. All linear elements move back and forth on lead screws over a finite travel distance. Position is determined by a combination of Hall effect sensors and magnets (Figure 5) defining coarse and fine positioning.
Figure 5: Encoding scheme for linear positioned elements.
Coarse position is determined by measuring the flux between a North-South pair of magnets. The opposite polarities of the coarse positioning, or "linear", magnets produce two adjacent peaks, one positive the other negative. Between the peaks there is a ramp, which is (almost) linear (Figure 6). The stage is designed so that it is always operated in the linear region, providing a useful although not extremely accurate encoding of the position stage. Note that any changes in the sensitivity of the sensor electronics or DC offset can affect this encoding scheme. This is adequate, because the encoding scheme needs only to position the stage datum to within one turn of the main gear. A zero-crossing comparator is provided to rapidly find the "Coarse Home" position, the point in the middle of the two magnets where the Hall effect output crosses from negative to positive or vice versa.
Figure 6: "Coarse Home" and "Datum" positions for linear positioned elements.
Fine positioning is determined by measuring the rotary position of the lead screw in a method similar to that detailed in §7.1.1.1. A single magnet and Hall effect sensor pair is used to seek the peak response while travelling outwards from the "Coarse Home" position. Backlash in the gears is compensated for by travelling in only one direction. This lead screw peak position is used as the "Home" or "Datum" position for all further screw displacements (Figure 6).
Multiple positions for a single magnet are milled into the lead screw spur gear so the choice of "Home" position is made to provide the most versatile "Home" position in relationship to the "Coarse Home" which may move slightly due to DC drift. This is determined empirically.
For redundancy, all Hall effect sensors operate in pairs, mounted in parallel with a software switch choosing between primary or backup sensor. Should one Hall effect sensor fail, the computer will switch to the backup sensor. There will be a slight difference in peak position because it is not possible to ensure that the sensors are at the same angle with respect to the axis. This must be compensated for in software. The magnitude of compensation is determined empirically.
No provision is made for backup limit switches or hard stops to limit linear travel. Limit switches have a high rate of failure in cryogenic conditions so were deemed unsuitable for NIRI and it was thought that the current system of using backup Hall effect sensors was reliable enough so that hard stops were unnecessary. The operation of this system in NIRI will be monitored.
7.1.1.3 Open/Closed Elements
The Environmental Cover is an open/closed element. The Environmental Cover is unique in operation in NIFS; it has only two relevant positions, "Open" or "Closed". The cover is a two vane sliding cover, driven by a flexible belt with limit switches signaling the end of travel (Figure 7).
Figure 7: Environmental cover assembly drawing.
In the "Closed" position, the cover should be light tight and protect the cryostat window from dirt, dust, and moisture likely to be encountered during storage or shipping. In the "Open" position during operation, the cover should be retracted and provide an unobstructed path for light to enter the cryostat.
An Ioniser Sprayer Head (item 14 in Figure 7) is separate to the mechanism for opening and closing the Environmental Cover but part of the same sub-assembly. This head is responsible for venting de-ionized, dry, non-turbulent air over the cryostat window in order to prevent moisture condensing during astronomical observations.
A Cleaning Spray Bar (item 8 in Figure 7), is also part of the Environmental Cover sub-assembly. It uses a high-pressure jet of dry air to clean the lower vane and cryostat window of large detritus. The air from this jet is highly turbulent and will not normally be vented during astronomical exposures.
7.2 Temperature Controller (TC)
The Temperature Controller is responsible for temperature regulation and temperature control during cool-down and during warm-up. The NIFS cryostat consists of three main thermal sub-assemblies; the OIWFS detector, the spectrograph detector; and the cold plate thermal mass.
In NIRI, each thermal sub-assembly is regulated by its own Omega CYC321 Temperature Controller. For NIFS the control of the spectrograph detector temperature has to be regulated to the milliKelvin level (§8.3.7.3). It is proposed to use a Lakeshore Model 340 Temperature Controller for the spectrograph detector. This is a dual PID controller with 100 W primary and 1 W secondary which when incorporating a Cernox™ temperature sensor gives temperature regulation to 1 mK. Temperature set-point, heater On/Off, curve data, and output data may be controlled over a RS232 link via a XYCOM XVME-400 Serial I/O board located in the OCC VME Crate.
Additional thermistors will be mounted at various locations within the cryostat for use during Accelerated Warm-up and for thermal diagnosis. Preamplifiers for these thermistors are located on the same board as the Hall effect sensor preamplifiers. The amplified thermistor signals are connected to inputs on the XYCOM Inc. XVME566, High Performance Analog Input Modules, which are located in the OCC VME Crate.
NIRI uses two Leybold Coolpower 130 Cryocoolers, but it is not clear whether these are still available. Even if available, different coolers may be chosen for NIFS to reduce vibration problems which are currently a concern for NIRI (§6.4). This possible change is unlikely to cause significant electrical control system design changes.
The two closed-cycle helium coolers are attached with cold straps to internal mechanisms. Varying the motor speed within each cryo-head regulates the rate of cooling. During normal cool-down operation, the speed of the cooler will be set at 140 rpm, which is the factory recommended nominal operating speed. During emergency cool-down operation only (minimum cool-down time), the speed will be set to 200 rpm. Once a nominal cryostat operating temperature has been achieved, the speed of the cooler will be reduced to a value consistent with normal operation heat-load requirements. It is estimated that reducing the cooler speed to ~ 30% of maximum would be a nominal operating speed. The actual value will be determined during NIFS commissioning.
Heater sub-systems will be used to control heat flow such that the OIWFS and science detectors are always maintained at a slightly positive temperature differential with respect to their surrounding environments in order to minimize sublimation of any residual matter onto the detector surfaces.
The first stage of the cryo-head, which has the greater refrigeration capacity, is cold-strapped to the Cold Work Surface thermal mass (i.e., the CWS plate) and the second stage, which can achieve a lower temperature, is cold-strapped to the spectrograph detector. The OIWFS detector is cooled by a cold strap to the Cold Work Surface thermal mass.
Controlled warming of the NIFS cryostat will be accomplished by turning off the coolers and raising the set point of the thermal regulation system. Accelerated warm-up is possible by diverting power normally dedicated to the stepper motors and the cooler motors to additional heater resistors located on the Cold Work Surface thermal mass. An additional 800 W will become available to provide the accelerated warm-up. Typical locations for these additional heater resistors are the CWS plate and the charcoal getters.
The OCC CPU controls the temperatures of the CWS plate and the detector thermal masses during the accelerated warm-up operation so that positive detector temperature differentials are maintained to prevent detector contamination. During accelerated warm-up operation, up to one kilowatt of power will be applied to the NIFS cold surfaces under the control of the OCC Heater System. The OCC utilizes the computer-independent Auto-shutdown and Interlock Safety System to prevent overheating of critical cryostat parts should any of the accelerated warm-up systems fail.
Initiation of the accelerated warm-up operation requires active human action. To start an accelerated warm-up operation, the following must be accomplished:
1. A panel circuit-breaker has been set.
2. A panel timer has been set.
3. All cryostat thermostat switches are below 40 C (100 F).
4. A panel spring-loaded normally-off key-switch has been momentarily switched on.
5. UPS and telescope mains power are present.
The safety interlock system will:
1. Prevent accidental turn-on of the high power heater system.
2. Require periodic "human verification" that cryostat temperatures are within normal limits.
3. Automatically shut-down the high power heater system if an UPS power failure should occur which could affect the accelerated warm-up operation.
4. Automatically shut-down the high power heater system if cryostat temperatures exceed 40 C (100 F).
5. Provide a computer-independent temperature display of cryostat temperatures so that a human being can verify nominal warm-up operation.
During accelerated warm-up, the Auto-shutdown System has the ability to disconnect the 110 V to the power supply powering the cryostat heaters. Thermal switches, located inside the NIFS cryostat, will trigger the shutdown. The Auto-shutdown System will also be equipped with a timer that must be periodically reset in order to continue the warm-up operation. This part of the Auto-shutdown System will terminate the warm-up operation should weather conditions prevent observatory access. Additionally, a reliable temperature monitoring meter will be provided so that crew personnel can visually check the cryostat temperatures.
7.3 Instrument Control System Risks
7.3.1 Mechanism Encoding Method
Problems with the magnetic position encoding methods were uncovered during NIRI testing. Any NIRI mechanisms that are to be duplicated for use in NIFS must incorporate the NIRI design changes needed to make the encoding methods reliable. The design of NIFS specific mechanisms must also incorporate these design changes. The mechanism encoding method represents a small risk until it is demonstrated to perform at a satisfactory level in NIRI.
7.3.2 Science Detector Temperature Control
Controlling the science detector temperature to 1 mK will be difficult. The commercial temperature controller that has been proposed is capable of controlling the detector thermal block to this level if sensors and heaters are mounted carefully. However, the detector will need to be closely coupled to this block to prevent self-heating effects causing temperature variation at the detector in excess of 1 mK.