``... to observe this bright continuum source 
with the HETGS, the ACIS-S was operated in continuous
clocking readout mode... the observed HETGS-HEG spectrum from 1 to 9
keV is presented in Figure 22... the XSPEC model is a
powerlaw (PhoIndex = 1.3) plus lines of Mg-K, Al-K, Si-K, W-M (alpha
and beta), Cl-K, K-K, Ca-K, and Fe-K... observed line widths are
between 2 and 20 eV FWHM... the existence of the Cl-K line strongly
suggests that chlorine-based cleaning products are being used in the
XRCF source building...''
Although not an astrophysical observation, the above description of XRCF data gives a flavor of what is in store when the HETG is combined with the HRMA and ACIS-S to produce the High Energy Transmission Grating Spectrometer, HETGS. XRCF testing was the first and only ground testing of the full HETGS. Very preliminary analysis of the XRCF data indicates that the HETGS is performing at or above specification; the data have also indicated some low-level effects in the HETG that will be included in its instrument model.
The HETG can be inserted in the optical path behind the HRMA to intercept and diffract the converging X-rays. The focal plane image that results from a monochromatic source is a set of diffracted orders, m = 0, ± 1, ± 2, ..., spaced by a distance roughly proportional to the X-ray's wavelength, i.e., as 1/E:
where Y 
is the dispersion distance from
zero-order, X 
is the
HETG-to-detector spacing (8782.8 mm at XRCF, 8633.9 mm in flight), p
the grating period (see Table), hc is 12.3985 keV-A, and C 
is a useful
combination of the parameters (X hc/p). The m-th order image is to
first approximation the same size and shape as the un-diffracted (m=0)
HRMA image. Thus, a HRMA FWHM of order 50 
m in an image dispersed to
50 mm represents a resolving power of 1000.
| Grating | Period | Angle |  , XRCF | , Flight |
| (type) | (Å) | (degree) | (mm-keV) | (mm-keV) |
| HEG | 2000.95 | -5.19 | 54.42 | 53.50 |
| MEG | 4000.77 | +4.74 | 27.22 | 26.76 |
Figure 22: HETGS ``continuum'' source spectrum. The XSPEC ``data
and folded model'', log(counts) vs log(energy), is plotted from 1 to 3
keV in the top panel and 3 to 9 keV in the bottom panel. Trace
contaminants on the source have added discrete lines.
Figure 23: DEFOCUSSED HETGS observation of Cu-L (0.923 keV)
line complex. This figure shows the
full ACIS-S image. The MEG m = -2,-1,0,+1,+2
images run from lower-left to upper-right; the HEG m = -1,0,+1 images
run from upper-left to lower-right.
Figure 24: DEFOCUSSED HETGS observation of Cu-L (0.923 keV)
line complex. This figure is a close-up view of HEG-1,
MEG-1, and zero orders in Fig. 23.
Note that the HRMA ring structure is clearly
visible and different for the HEG and MEG.
An example of the effect of the HETG is shown in Figures 23 and 24. Here the ACIS-S has been DEFOCUSSED by 40 mm to spread the events over a larger region of the detector to reduce pile up. The central region contains the zero-order image - all four HRMA shells are visible. The HETG is made up of two grating types (HEG and MEG) that are oriented roughly +/- 5 degrees from the facility Y axis. The more dispersive HEG gratings are on the inner HRMA shells, 4 and 6, and the MEG gratings are on the outer HRMA shells, 1 and 3. The diffracted non-zero-order images show only the shells corresponding to their grating.
Figure 25: Top: HSI image of the MEG 3rd order Al-K
line complex. Here the image is in focus. Bottom: A spectrum is
obtained by binning the dispersion-derived wavelengths; the K-alpha
and satellite line (``sat.'') are separated by about 9 eV.
Between the HETG first-light on the morning of December 23rd 1996 and its last-ground-light on April 26 1997, over 550 XRCF tests were performed with the HETG inserted in the optical path. These tests took place in two calibration phases: Phase I with the non-flight HXDS detectors (FPC,SSD, HSI), and Phase II with the flight(-like) detectors (ACIS-2C,HRC-I,ACIS-S). In each phase, tests were performed to measure and model the i) point spread function PSF of the diffracted image and the resulting line response function LRF and resolving power E/dE, and ii) the HETG diffraction efficiency and/or the HETGS effective area. Detailed analysis of these data is on-going; a preliminary report was issued Sept. 1, 1997, and the final report is due June 1, 1998.
As an example of PSF/LRF data, Figure 25 shows an HSI image of the MEG 3rd order Al-K line. Preliminary analysis of data such as these confirms that the HETG is meeting or exceeding its resolving power specifications. One surprise indicated in this image is the gross ``roll'' misalignment of one of the MEG gratings - giving rise to the weak image, above the main image, at Facility Z around 4900. The cause for this and a handful of other less misaligned MEG gratings was traced to fabrication problems in a single batch of gratings. The effect is not severe on either throughput or resolving power and will be included in the instrument PSF model. (Discovering an effect like this is, of course, why we went to XRCF!)
Figure 26: Measured and predicted ratio of HEG+MEG first orders
to MEG+HEG zero order. The solid curve shows the expected
efficiency ratio due to the gold grating structures.
As an example of effective area / efficiency data, Figure 26 shows the measured ratio of the counts in the HEG and MEG plus first orders to the counts in the (combined) zero order. The solid curve is the prediction based on laboratory test data. The diamonds represent XRCF measurements made using the double crystal monochromator DCM source tuned to several energies with the HRC-I flight detector in the focal plane. At a preliminary analysis level (10% to 30%) these and other data confirm the laboratory-predicted efficiency curves; more refined analysis should yield a definitive HETGS effective area calibration.
For further details, the preliminary report, etc. please see
http://space.mit.edu/HETG/xrcf.html
or contact dd@space.mit.edu.
Dan Dewey for the MIT/HETG science team
