Before its failure in January 2007, the High Resolution Channel of ACS was the
primary camera for near-UV imaging. HRC provided high throughput at blue wavelengths and better PSF sampling than either the WFC or other CCD cameras on HST
. HRC’s pixel size critically sampled the PSF at 6300 Å and undersampled the PSF at the blue end of its sensitivity range (2000 Å) by a factor of 3.0. Well-dithered observations with the HRC led to a reconstructed PSF FWHM of 0.03 arcsec at ~4000 Å, increasing towards longer wavelengths. For these reasons, HRC was used mostly for UV and blue imaging. It was also used for red imaging when PSF sampling was important. The photometric accuracy of the HRC was generally higher than that of the WFC. HRC also included a coronagraph that is discussed in Section 6.2
. HRC’s CCD scattered red light into a wide-angle halo (as does STIS’s CCD). Production constraints prevented the remediation of this halo by front-side metallization, which was done for WFC’s CCD. Although most of the HRC imaging occurred in the UV, users are cautioned to take into account the effect of the long wavelength halo when the HRC was used in combination with near-IR filters (see Section 5.6.5
The HRC-specific filters were mostly UV and blue. The set included UV and
visible polarizers (discussed in Section 6.1
), a prism (PR200L, discussed in Section 6.3
), three medium-broad UV filters (F330W, F250W, and F220W) and two narrow band filters (F344N and F892N). Use of the UV filters with the WFC is not supported because of the uncertainty of the WFC silver coating transmission below 4000 Å. All broad, medium and narrow band WFC filters could be used with the HRC whenever better PSF sampling was required. Generally, the throughput of WFC was higher than that of HRC where their sensitivities overlapped. Only some of the WFC ramp segments -- the FR459M and FR914M broad ramps, and the FR505N [OIII], FR388N [OII], and FR656N (H) narrow ramps -- could be used with the HRC because only the middle segment overlapped with the HRC FOV.
Unlike the WFPC2 CCD, the HRC CCD was directly sensitive to UV photons and
was much more effective in detecting them. When a detector has non-negligible sensitivity over more than a factor of two in wavelength, however, it is possible for a UV photon to generate more than one electron and thus be counted more than once. This effect was noted during ground testing of the HRC CCD and also has been noted for the STIS CCD. The effect is only important shortward of 3200 Å, reaching approximately 1.7 e-/photon at 2000 Å. Multiple counting of photons must be accounted for when estimating QE and the noise level of a UV observation because multiple photons distort the Poisson noise distribution of the electrons.
When designing a UV filter, high suppression of out-of-band transmission,
particularly at red wavelengths, must be balanced with overall in-band transmission. HRC’s very high blue quantum efficiency made it possible to obtain red-leak suppression comparable to that of WFPC2 while using much higher transmission filters. Red leak calibration data was obtained in Cycle 14 and are described in ACS ISR 2007-03
. Table 5.6
shows the ratio of in-band versus total flux for a few UV and blue filters in the now-defunct HRC, where the boundary between in-band and out-of-band flux is defined as the 1% transmission point. The same ratio is also listed for the equivalent WFPC2 filters. Correction factors for different stellar spectral types and non-stellar spectra are found in ACS ISR 2007-03
. Red leaks were not a problem for F330W, F435W, and F475W. Red leaks were more important for F250W and F220W. In particular, accurate UV photometry of M stars requires correction for the F250W red leak and is essentially impossible in F220W. For F220W, red-leak correction is also necessary for G and K stars.