To illustrate the change in sensitivity, we compare the performance attained by WFC3 with FPA#64 vs. the new substrate removed FPAs in Fig. 4. We consider a few representative cases: 1) FPA#64 and WFC3 as-built; 2) FPA#64 with 1 e-/s/pixel of extra glow due to cosmic rays (worst case scenario); 3) FPA#64 with 0.25 e-/s/pixel extra glow (best guess scenario); 4) Second generation FPA with same RON of FPA#64 (24 e-/pixel DCS); 5) Second generation FPA with high RON (40 e-/pixel DCS); 6) Second generation FPA with low RON (15 e-/pixel DCS). We consider 4 filters, F110W and F160W (approximately the standard J and H passbands); F126N (a representative narrow-band filter) and F093M, roughly corresponding to the z-band often used with CCD detectors (e.g. ACS/HST).
Figure 4 shows the discovery efficiency (1/observing time needed to reach a given SNR on a faint point source x the detector area in square arcmin) normalized to the current NIC3 values (to ACS for F093M). The second generation detectors could nearly double the discovery efficiency at short wavelengths and also be competitive with ACS in the F850LP (note however the different pixel scale: 013" of WFC3-IR vs. 0.05 of ACS); for broad-band surveys the gain in blue QE easily offsets a potential increase of readout noise. On the other hand, a luminescence of 1 e-/s/pix would have a dramatic impact on the discovery efficiency with FPA#64.
Was this article helpful?