R. C. Bless, J. W. Percival, J. S. Gallagher, J. G. Hoessel
Astronomy Department, University of Wisconsin-Madison, Madison, WI 53706 USA
D. J. Schroeder
Beloit College, Beloit, WI
Keywords: high speed photometry, integral field spectrograph
After it became clear that the High Speed Photometer was to be replaced by COSTAR, the HST project asked us to consider what sort of instrument might be made with the returned hardware and a modest amount of new hardware. The ground rules were minimal; any kind of instrument(s) could be considered, but the cost must be no more than that associated with a ``University Class'' project. We considered several options; the following describes the possibility we have most fully developed.
Two instruments are mounted in the old HSP axial enclosure: a simplified High Speed Photometer and an area (``3-d'') spectrograph (shown schematically in Figure 1). These instruments extend the science domain of the HST to fast UV/optical photometry, increase the scientific efficiency of HST by area spectroscopy, complement STIS capabilities, and cost much less than the limit we were given.
Figure: WIPS Optics and Detectors
Table 1: Optical Specifications of HSP2
Table 1 lists the salient features of the proposed photometer.
The time resolution of this instrument is such that it can sample all objects that vary as rapidly as millisecond pulsars. The complications caused by atmospheric scintillation in the regime from 1 to 100 Hz are eliminated and the ultraviolet is, of course, made available.
This instrument has fewer detectors, filters, and apertures than did the first HSP. There are no filter and aperture wheels; the various filter/aperture combinations are selected by pointing the telescope so that light falls on the desired filter/aperture pair, at the same time collecting photoelectrons only from the corresponding point on the image dissector cathode. This worked very well on the original HSP.
We have heard a lot about AGNs at this meeting so it is perhaps appropriate to point out the capabilities of this photometer for such objects. The high angular resolution isolates the AGN to a degree impossible from the ground; observations in the UV minimize the central bulge background; AGNs generally are more variable in the UV than in the optical; the high time resolution gives a high Nyquist frequency; long term precision extends variability searches. Thus it complements the wonderful high-resolution images we have seen in the past few days with time-resolved, spatially localized UV data otherwise unattainable with HST.
Table 2 gives some of the characteristics of the spectrograph.
Table 2: Optical Specifications of Spectrograph
It is fed by either of two fiber cables, the entrance end of each being located at an IDT, and the output ends at the spectrograph. The present concept has 37 fibers arranged at the entrance in concentric hexagons, with an outermost diameter of about 1.8 arcseconds. Each fiber core has an angular diameter of 0.14 arcseconds and all together they cover about 25 percent of the hexagonal area. Thus, 37 spectra with spectral resolution of about 6000 (or about 50 ) simultaneously sample an area about two arcseconds in diameter. Such an instrument has high throughput and is very efficient for objects of appropriate shape. In this sense it complements the STIS long slit mode. The spectrograph covers 2000Å at the peak of the CCD response in a spectral region rich with astrophysically important absorption and emission lines.
The spectrograph has two finders, namely, each of the photometer's two detectors. The fiber cable associated with the bialkali detector would be the more useful of the two, of course, because the solar blind detector would be limited to hotter objects. The fiber cores are made of pure silica and are in diameter. Their attenuation of light is very small, about 10--20 db/km. Two shutters are required to keep the inactive cable from sending light to the spectrograph.
Figure: FFS layout
The spectrograph is shown in Figure 2 It is very compact; it can be packaged in a 15-centimeter diameter tube about one meter long. It is fed by either of the two fiber bundles in a slit-like arrangement and a single lens serves as both collimator and camera lens. After light passes through the lens it passes through a transmission grating, is reflected by a reflection grating and retraces its path through the grating and lens and onto a STIS-type CCD. The reflection grating provides cross dispersion to separate the spectrum into four orders of 500Å each.
HST high resolution images have revealed a wealth of small structures such as jets, bright rims, and disks for which such an instrument could quickly provide useful line ratios and kinematic data. To mention AGNs again, this spectrograph could measure circumnuclear velocity fields in these objects, the interaction of jets with the interstellar medium, etc.
Within the instrument enclosure, a two-mirror system for each image dissector corrects the primary mirror spherical aberration, and at one point in the 50 arcsecond field, completely corrects the OTA astigmatism as well as minimizes coma. As in COSTAR, the spherical mirror M1 images the OTA pupil on the aspheric mirror M2 (see Figure 1); the f/24 OTA beam is converted to f/30. Unlike COSTAR, M2 is axially symmetric (thus much less costly) because we do not have COSTAR's tight space constraints.
Table 3 gives the optical tolerances. Only that of pupil centering is tight, so an on-orbit tip/tilt mechanism is required for M2. Laboratory work indicates that this mechanism should not be difficult to make. The other optical tolerances is met even in the all-aluminum HSP enclosure, especially given the high degree of temperature control provided by the HSP on-orbit adjustable thermal control system.
Table 3: Optical Tolerances
Several features contribute to the low cost of this instrument. First of all, the HSP was returned to earth in excellent condition. No contaminants were found on the instrument; functional test results were indistinguishable from those run before launch as were the various parameters of all electronic boxes. For example, the pre- and post-launch high voltage power supplies' output voltages agreed to about 0.04 percent. All flight units could be flown again. The HST environment is indeed benign.
As mentioned earlier, the HSP enclosure requires only relatively small modifications and the corrective optics are simpler than those in COSTAR. The new HSP commands, telemetry, and science data formats are backward compatible with the old HSP. The spectrograph is simple and stable; its optics are all off-the-shelf and inexpensive. Maximum use is made of STIS CCD electronics design, command macros, ground system software and, where possible, firmware.
We feel that such an instrument meets the design goals mentioned at the beginning of this paper.