Space Telescope Science Institute
Table of Contents Previous Next Index Print

Near Infrared Camera and Multi-Object Spectrometer Instrument Handbook for Cycle 17 > Chapter 7: NICMOS Detectors > 7.3 Detector Artifacts

7.3 Detector Artifacts
The NICMOS arrays exhibit a noiseless signal gradient orthogonal to the direction of primary clocking, which is commonly referred to as shading. It is caused by changes of the pixel bias levels as a function of temperature and time since the last readout (“delta-time”). The amplitude of the shading can be as large as several hundred electrons for some pixels under some circumstances. The first pixels to be read show the largest bias changes, with the overall shading pattern decreasing roughly exponentially with row number. The shading is a noiseless contribution to the overall signal, therefore it can be completely removed during pipeline processing once it has been calibrated with delta-time and temperature.
For a given delta-time (and temperature), the bias level introduced by the shading remains constant. For MULTIACCUM readout sequences (see Chapter 8) where the time between readouts is increasing logarithmically, the bias level changes with each successive read, and thus the overall shading pattern evolves along the MULTIACCUM sequence. We have calibrated the dependence of shading as a function of delta-time for each of the three NICMOS detectors. This information is used by the calnica pipeline to construct synthetic dark current reference files for NICMOS observations. The accuracy of this calibration is good (a few percent for most readout times).
The 1999 warm-up monitoring program has shown that the shading signal is temperature dependent. Nevertheless, the good temperature-stability of the NICMOS/NCS system has enabled accurate shading correction of NICMOS data with a single set of dark current reference files. Figure 7.10 presents the shading profiles for each camera at the operating temperature of 77.1 K.
The NICMOS group at STScI will continuously monitor both shading behavior and NICMOS temperature stability, and will provide additional calibration files should this become necessary.
Figure 7.10: Shading profiles for all camera/delta-time combinations measured at 77.1 K (NCS era). The profiles were created by collapsing a dark exposure of the respective integration time along the fast readout direction (after correction for linear dark current and amplifier glow).
Each quadrant of a NICMOS detector has its own readout amplifier situated close to the corners of the detector. Each time the detector is read out, the amplifier warms up and emits infrared radiation that is detected by the chip. This signal, known as amplifier glow, is largest in the array corners with ~80 e-/read, and falls rapidly towards the center of the detector where it is about 10 e-/read. The signal is cumulative with each non-destructive readout of an exposure. It is highly repeatable, and is exactly linearly dependent on number of reads. It also is constant with temperature, as shown in Figure 7.11.
Figure 7.11: Amplifier glow signal as a function of detector temperature.
In contrast to the shading, the amplifier glow is a photon signal, and thus is subject to Poisson statistics. It therefore contributes to the total noise in NICMOS exposures. Amp-glow images for all three cameras are shown in Figure 7.12. In case of an ACCUM exposure with multiple initial and final reads (see Chapter 8), the photon noise produced by amplifier glow can outweigh the read noise reduction from the multiple reads, especially close to the array corners producing a total noise reduction never larger than ~40 –50%. Similarly, the trade-off between improved cosmic ray rejection, reduced read noise, and increased photon noise in a MULTIACCUM sequence is complicated.
Figure 7.12: Amplifier glow for Cameras 1 (left) through 3 (right), on a uniform grayscale, and below a plot of rows (near the bottom) of each camera.
Effects of photon and cosmic-ray persistence are described in Section 4.6.
Electronic ‘‘bars’’ are an anomaly in NICMOS data taken during Cycles 7 and 7N. They appear as narrow stripes that cross the quadrants of an array, and occur identically in all 4 quadrants at the same rows/columns in each. The bars are caused by pick-up of an amplifier signal on one of the row/column address lines, causing a momentary change in the bias for that pixel.
Similarly, electronic ‘‘bands’’ are caused when one of the NICMOS detectors is reset while another is being read out. The reset pulse causes a sudden jump in the bias of the detector which is being read. The bias jump then appears as an imprint on the image that looks like a band.
The bars typically run the length of a quadrant (128 pixels), and are 3 pixels wide—the first pixel is lower than the mean, the second is at the mean level and the third is higher than the mean, giving the impression of an undersampled sinusoidal spike with an amplitude of up to ~10 DN peak-to-peak. If a bar appears in the 0th readout, it will be subtracted from all the other readouts as part of the normal calibration process, and will appear to be a negative of the above description. The bars run parallel to the slow readout direction, which is vertical in NIC1, and horizontal in NIC2 and NIC3. They are almost always broken in at least one place, with a shift of 2–10 pixels in the narrow direction. A more detailed description of the electronic bars and bands is given on the NICMOS WWW site:
Since Cycle 11, a modified readout sequence has been implemented for the three NICMOS cameras which reduces the probability that a detector will be reset while another is being read. This procedure is completely transparent to users and has significantly reduced the electronic bands problem.
Each NICMOS detector has a number of pixels that show an anomalous responsivity. Such “bad pixels” come in various flavors. So-called “hot” pixels have a higher than average dark current, and thus show excessive charge compared to the surrounding pixels. On the other hand, “cold pixels” are less sensitive to incident photons than the typical pixel. The anomalously low responsivity of a “cold” pixel could be due to either a lower intrinsic DQE of the pixel, or due to grot (see below). Some pixels do not even respond at all (“dead pixels”) to incoming light. Quantitative statistics of the hot/cold and grot pixels in the three NICMOS cameras are given in Table 7.1. It is important to note that the impact of bad pixels on the quality of NICMOS images can be minimized by dithering the observations.
On-orbit flat field exposures taken after the NICMOS installation in 1997 revealed a population of pixels with very low count rates that had not previously been seen in ground testing. It is believed that these pixels are at least partly obscured by debris on the detector surface, most likely small paint flakes that were scraped off one of the optical baffles during the mechanical deformation of the NICMOS dewar (Sosey, NICMOS ISR-99-008 and Boeker et al. 2001 PASP, 113, 859). Additional grot has collected on the detectors since its revival. NIC1 appears to be the most affected with an additional chunk of grot in the lower right quadrant. This so-called “grot” affects approximately 100–200 pixels in each NICMOS camera. The largest pieces of grot in NIC1 are shown in Figure 7.13. Again, dithering is recommended to minimize the impact of grot.
Figure 7.13: A NIC1 flat field image shows the largest of the groups of pixels affected by debris (“grot”). These bits of “grot” are roughly 5 by 9 pixels (upper left) and 5 by 6 pixels (lower right).

Table of Contents Previous Next Index Print