This page is intended to aid proposers and observers as they select instrumental parameters for LETG/ACIS-S observations, specifically with regard to pointing offsets and subarray configurations. Most of the information presented here is also in the Proposer's Observatory Guide (see especially Table 9.3 and section 9.4.2/Offset Pointing). The additional contributions of this page are specific subarray configurations and links to the Spectrum Visualization Tool.
The default for Z-SIM is -8 mm. This puts the spectrum close to the ACIS readout, which reduces the effects of CTI-induced energy resolution degradation. With the current (late 2014) ACIS-S aimpoint (Y-offset=0, Z-offset=0) of (chipx,chipy)=(201,477) and a pixel size of 23.985 microns, an 8 mm shift moves the dispersed spectrum down to row 477-(8000/23.985) = 143. The default Z-Offset of -18" (-0.30') moves the spectrum up by 37 rows to row 180, and the subarray configurations listed below are appropriate for that row.
Originally, -8 mm was the largest Z-SIM value allowed with ACIS-S (because of fiducial light selection procedures), but that restriction was lifted in late 2005. In principle, a larger Z-SIM value can be used to put the spectrum as close to the readout as desired, yielding a very slight improvement in CCD energy resolution. This, however, also pushes the dispersed spectrum closer to the edge of the ACIS optical blocking filter; contamination increases rapidly near the filter edges and is less well calibrated than at the default location. Furthermore, dithering samples varying thicknesses of contaminant and therefore leads to periodicity in the source lightcurve. A few calibration measurements, usually to study contamination absorption, have been conducted at non-default Z-SIM values, including -11.5 mm.
Because of the changing thermal environment of the telescope, the aimpoint has drifted over time. The most recent (Oct 2014) analysis indicates that Y-offset=0 provides the best spatial resolution, without any concerns about the source dithering across the node 0/1 boundary on the S3 chip (which would slightly degrade energy resolution). Most LETG/ACIS-S observations have used Y-offset of 1.50' in order to place the important He-like O lines and the O-K absorption edge on the S3 chip; as seen in the figure below, the Backside-Illuminated (BI) chips, S1 and S3, have much higher QE at low energies than the Front-Illuminated (FI) chips.
Because of aimpoint drift, as well as increasing scatter in where a source falls on the detector (i.e., pointing errors, or aimpoint errors) we now recommend using a Y-offset of 1.20' for most observations. The 1σ errors on Y are now ±4" and errors of 10" or more have been seen. Keep this scatter in mind when choosing a value for Y-offset.
The Spectrum Visualization Tool displays where spectral features fall on the ACIS-S detector as a function of Y-offset and source redshift. One arcminute of Y-offset corresponds to a shift of 3.36 Å. The following offsets are of particular interest:
Very few LETG/ACIS observations use the full ACIS-S array. One advantage of using a subarray and/or fewer than 6 chips is a faster readout (shorter frametime) and thus less pileup in 0th order and the dispersed spectrum.
From the figure above one can see that S0 and S5 are useless for detecting 1st-order photons, although they may be useful for collecting higher order spectra in some cases. The S4 chip is also unlikely to be useful. Because of increasingly stringent ACIS thermal requirements, the S0 and S5 chips should be turned off unless there is a good reason for their use. Note that chips can be marked as "Optional"; these will be used unless Mission Planning schedulers determine that thermal constraints would otherwise be violated, in which case they will be turned off.
LETG/ACIS background is weak enough that users often don't need to worry about background subtraction, and minimal subarrays of 128 rows are commonly used. If background is a concern, however, an array of at least 200 pixels is needed to fill the default background regions (extending to 2.0 mm on either side of the spectrum) without dithered edges. Adding another 40 pixels allows for worst-case aimpoint errors. A larger subarray may also be needed if the source position is not well known or the source is extended. Of course, users may also reprocess their data using custom background regions in tg_extract.
To summarize, 3 or 4 chips will suffice for most observers. A 128-row subarray is probably fine but if you care about the background spectrum you should use at least 200 rows. Something like 4 chips with 230 rows is a good compromise between providing plenty of space on the detector and minimizing 0th-order pileup. Fewer rows or using 3 chips reduces pileup even more.
The tables below list subarray parameters for the traditional 1/2, 1/4, and 1/8 subarrays, and for optimized subarrays that maximize the number of rows for a given frametime. The listed Start row values will center the image in the cross-dispersion direction as closely as possible, and are based on the latest calibration of aimpoint drift.
The default Z-SIM value of -8 mm,
which puts the spectrum close to
the ACIS readout and minimizes the effects of CTI-induced energy
resolution degradation, is assumed, as is the
default Z-offset value of -0.30'.
Subarray Type = "Custom" in all cases.
|CCDs||4 chips||3 chips|
|CCDs||6 chips||5 chips|
|CCDs||6 chips||5 chips||4 chips||3 chips|
Frametime for m active CCDs, using n rows starting with row q, is given by the equation
T(msec) = 41.12*m + 2.85*n + 0.040*m*q - 32.99
and rounding up to the nearest 0.1 sec.
Last modified: 08/18/15
The Chandra X-Ray
Center (CXC) is operated for NASA by the Smithsonian Astrophysical Observatory.
60 Garden Street, Cambridge, MA 02138 USA. Email: firstname.lastname@example.org
Smithsonian Institution, Copyright © 1998-2004. All rights reserved.