Previous Article Next Article Table of Contents
GoldCam Comes to the 0.9-m As WhiteCam (1Jun92) (from KPNO, NOAO Newsletter No. 30, 1 June 1992) With the introduction of a big, blue-sensitive chip into the GoldCam camera, we suddenly have a capability hitherto unrealized for spectroscopy. We were anxious to try this instrument at the 0.9-m, as a reasonable physical slit width would project to a sufficient number of arcsec on the sky to provide accurate flux determination. Furthermore, the use of the GoldCam detector at the 0.9-m would provide capabilities not currently offered by the aging Intensified Reticon Scanner (IRS). Accordingly, we mated the GoldCam camera/detector combination to the White Spectrograph. Voila: WhiteCam. (The physical details of the White Spectrograph, other than its color, are identical to that of the Gold Spectrograph, GoldCam's normal home.) The instrument was used on three nights in mid-April to see just how well we could do. The upper two plots to the right show the spectrum of the A0I star Eta Leo. The upper of these simply shows the raw counts in pixel space. No flat-fielding has been done, and we can see that vignetting and loss of focus makes the detector unusable below pixel 400 or above pixel 2600. Still, this 2200 pixel "good" region provides approximately 800 spectral resolution elements (best focus is about 2.7 pixels, as discussed in the article above.) This may be compared to the IRS, where there are at most 300 spectral resolution elements. We see for the wavelength axis in the second plot that the detector is indeed plenty hot down to the atmospheric cutoff. We were able to detect the He I lamdba3187 line in our comparison spectrum. The DQE measurements of the Ford chip used to construct the plot in the article above shows a 57% DQE of the detector at 3200 A. This may be compared to the 8% DQE cited in the IRS manual for the response of the first photocathode down in the blue. Although the shot-noise is a little bit higher with the CCD spectrometer (15 e- read noise vs. the equivalent of 3 e-), the anticipated introduction of a new Ford chip in GoldCam this summer, and the factor of 7 improvement in sensitivity of the CCD over the photo-cathode of the IRS, makes up for this in any but the SNR < < 1 regime. (If you are in that regime, we suggest trying a bigger telescope!) On another night we pressed the instrument to see how far we could usefully observe into the red. Although the DQE measurements of the Ford chip shown in the article above stop at 9500 A, we can see in the plot of the planetary nebula NGC 6543 that the detector still has some sensitivity at the He I lamdba10830 line! To the best of our knowledge, this is the first detection of this line with an "optical" CCD at Kitt Peak. Although it is not clear from the plot of the PN spectrum, there is very significant fringing (25% peak-to-trough) in the red, which we could reduce only by factors of 2 using simple techniques. We are still working on it though. Again, for comparison, the IRS is essentially dead longward of 8000 A. [figures not included] The largest slit width that will not degrade the spectral resolution is about 4 arcsec, so it will generally be necessary to observe at the parallactic angle to avoid compromising the accuracy of the flux calibration. We mounted the spectrograph so the slit was N-S and simply kept our observations to within an hour of the meridian. None of the three nights were photometric, but on the two "OK" nights we were able to achieve relative fluxes of 3-5%, and absolute fluxes at the 10% level. For successful proposals that requested the IRS, we will give the observers the option to switch to WhiteCam for the Fall, and we anticipate that WhiteCam will be the detector of choice for spectroscopy in the 3200 A to 1 m region at the 0.9-m during subsequent observing semesters. For questions and details please contact one of us. Phil Massey, George Jacoby, Dave Silva, Jim De Veny, Di Harmer
Previous Article Next Article Table of Contents