[Top] [Prev] [Next] [Bottom]
Basic Operations
NICMOS employs three low-noise, high QE, 256x256 pixel HgCdTe arrays in a passive dewar using solid N2 as a coolant. The detector design is based on the NICMOS 3 design that many IR observers may have used; however, there are differences between the two (see Chapter 7). Here we summarize the basic properties of the NICMOS detectors most relevant to the planning of your observations.
The NICMOS detectors have dark current of less than 0.05 electrons per second and effective readout noise for a single exposure of approximately 35 electrons.
The NICMOS detectors are capable of very high dynamic range observations and have no count-rate limitations. The dynamic range, for a single exposure, is limited by the depth of the full well, or more correctly the onset of non-linearity, which limits the total number of electrons which can usefully be accumulated in any individual pixel during an exposure to ~160,000 electrons.
NICMOS has four detector read-out modes that may be used to take data (see Chapter 8) plus a target acquisition mode (ACCUM, MULTIACCUM, BRIGHTOBJ, RAMP, and ACQ). For Cycle 7-NICMOS, only ACCUM, MULTIACCUM, and ACQ are supported and ACCUM mode observations are strongly discouraged.
The simplest mode is ACCUM which provides a single integration on a source. A second mode, called MULTIACCUM, provides intermediate read-outs during an integration that subsequently can be analyzed on the ground. A third mode, BRIGHTOBJ, has been designed to observe very bright targets that would otherwise saturate the detector. BRIGHTOBJ mode reads-out a single pixel at a time. Due to the many resets and reads required to map the array there are substantial time penalties involved. BRIGHTOBJ mode may not be used in parallel with the other NICMOS detectors. BRIGHTOBJ mode appears to have significant linearity problems and has not been tested, characterized, or calibrated on-orbit. RAMP mode implements a subset of the MULTIACCUM mode with onboard processing to avoid the transfer of large volumes of data. With the successful installation of the Solid State Recorder during the Second Servicing Mission, this mode is not necessary, has not been tested, and is not supported.
Users who require time-resolved images will have to use either ACCUM where the minimum exposure time is expected to be 0.6 seconds, and the minimum time between successive exposures is 20 seconds, or MULTIACCUM where the shortest spacing between exposures can be reduced to 0.203 seconds.
It is expected that MULTIACCUM mode will be used for most observations. It provides the best dynamic range and correction for cosmic rays, since post-observation processing of the data can make full use of the multiple readouts of the accumulating image on the detector. However, a sequence of ACCUM mode observations potentially offers a significant advantage for read noise limited observations using the NREADS option to decrease the effective read noise by up to a factor of ~2 (see Chapter 8). To enhance the utility of MULTIACCUM mode and to simplify the implementation, execution, and calibration of MULTIACCUM observations, a set of predefined MULTIACCUM sequences has been defined. The observer needs only to specify the name of the sequence and the number of samples which should be obtained (with defines the total duration of the exposure). For a set of fairly long exposure times these include sequences which accomplish an equivalent readout noise reduction to the use of ACCUM with multiple reads.
Comparison to CCDs
These arrays, while they share some of the same properties as CCDs, are not CCDs and offer their own set of advantages and difficulties. Users unfamiliar with IR arrays should therefore not fall into the trap of treating them like CCDs. For convenience we summarize the main points of comparison:
- As with CCDs, there is noise (read-noise) and time (read-time) associated with the reading out. In addition, there is an effect called shading which is an extraneous bias generated by the readout amplifiers. Furthermore, the dark current associated with NICMOS arrays, while low for IR arrays, is quite substantial compared to that produced by the current generation of CCDs.
- Unlike a CCD, the individual pixels of the NICMOS arrays are strictly independent and can be read-out non-destructively. Read-out modes have been designed which take advantage of the non-destructive read capabilities of the detectors to yield the optimum signal to noise for your science observations (see Chapters 7 and 8). Because the array elements are independently addressed, the NICMOS arrays do not suffer from some of the artifacts which afflict CCDs, such as charge transfer smearing and bleeding due to filling the wells. If, however, they are illuminated to saturation for sustained periods they retain a memory of the object in the saturated pixels. This is only a concern for the photometric integrity of back to back exposures of very bright targets, as the ghost images take many minutes to be flushed from the detectors.
Target Acquisition Modes
Most target acquisitions can be accomplished by direct pointing of the telescope. The user should use target coordinates which have been measured with the Guide Star Astrometric Support Package (GASP) to ensure the best accuracy with respect to the HST Guide Star Catalog. Particular care must be exercised with targets in Camera 1 due to its small field of view.
However, direct pointing will not be sufficient for coronographic observations since the achieved precision (~1 arcsec rms) is much larger than the 0.3 arcsec radius coronographic spot. Note that this is a function of the total HST pointing error and not only the result of uncertainties in the target's coordinates.
There are three target acquisition options for coronographic observations:
[Top] [Prev] [Next] [Bottom]
stevens@stsci.edu
Copyright © 1997, Association of Universities for Research in Astronomy. All rights
reserved.
Last updated: 07/24/97 15:17:04