Sensitivity

Limiting Magnitudes

The limiting magnitude of bHROS is currently set by the relatively high read noise of the CCD and noise due to excess background within the instrument enclosure (see the scattered light page for more information). The table below lists the magnitude, m(nu) that will give a signal-to-noise of 10 in 3600s in 0.7 arcsec seeing for various grating and detector settings. This estimate includes read noise and noise from the excess light.

Exposure Times

The bHROS Integration Time Calculator is undergoing development and is not yet available. Until that time, the information below, summarizing the sensitivity of bHROS as a function of wavelength¹ can be used to estimate the exposure times needed to reach the desired signal-to-noise (Derived from observations of HR 9087 (29 Psc) on July 23/24 2005). The data in the plots and table below give the number of electrons per spectral pixel from the source target.

Important notes in the calculation of the S/N:

• In addition to Poisson noise, the read noise in the detector and the enhanced dark current can be important sources of noise, especially for faint targets near the limiting magnitude.

• In 3600s, the enhanced dark current will contribute ~1320 e^- per spectral pixel for the large Object-only fiber and ~600 e^- per spectral pixel for the small Object/Sky fibers.

• Due to the large spatial width of the orders, the read noise of 5.3 e^- can also be a significant source of noise. The unbinned orders produced by large Object-only fiber are 220 columns wide. The unbinned orders from the small Object/Sky fiber are 100 columns wide.

Plots give source e- per spectral pixel in 10,000s for m(nu) = 15.0 source (3.66 mJy). See below for details.
Object-Only Fiber (Fiber L5) Click image for larger version (opens in a new window)
Object/Sky Fiber (Fiber S8) Click image for larger version (Opens in a new window)
Object/Sky Fiber (Fiber S1) Click image for larger version (Opens in a new window)

bHROS observations of HR 9087 were obtained on July 23/24 2005, consisting of 100s exposures centered at 4300A, 4900A, 5500A, 6500A, 7500A, 8500A and 9500A (300s), at airmasses ranging from 1.124 to 1.196. Some settings were centered near the peak of the echelle blaze response, while others were centered nearer to the edges of the echellogram which, together with small variations in the seeing between successive exposures, leads to a small degree of scatter in resulting plots of instrumental response versus wavelength.

The measured seeing at the time of the observations was 0.69 arcsec FWHM. For these conditions 0.724 of the total flux from HR 9087 is predicted to have been encompassed by the 0.88x0.88 arcsec2 effective area of the large slicer aperture while 0.443 of the stellar flux is predicted to have been encompassed by the 0.57x0.57 arcsec2 effective area of the small (dual) bHROS apertures (S1 and S8).

Monochromatic m(nu) magnitudes for HR 9087 were taken from Hamuy et al. (1994, 106, 566), who tabulate them for 16A bandpasses at 16A steps, from 3300A to 10404A. The zero-magnitude flux calibration for m(nu) is 3.664x10-20 ergs cm-2 s-1 Hz-1 = 3664 Jy. Magnitudes in the broad-band V (5500A) system have the same zero-magnitude flux calibration in Jy as the m(nu) system but broad-band magnitudes at other wavelengths (e.g. B, R, I) do not. A source with an m(nu) magnitude of 15.00 at 4500A, 6800A and 8980A would correspond to B=15.18; R=14.76 and I=14.60, respectively.

The measured counts per spectral pixel per second from HR 9087 were extrapolated to predict the expected counts per spectral pixel in 10,000s for an m(nu) = 15.0 target (3.66 mJy) in 0.7 arcsec seeing.

Last update March 10, 2006; Steven Margheim
¹S/N plots and tables prepared by Mike Barlow (UCL), August 2005