Previous Article Next Article Table of Contents
IR News (1Mar95) (from CTIO, NOAO Newsletter No. 41, 1 March 1995) CIRIM Update Since the December 1994 Newsletter No. 40 we have conducted additional engineering tests with the CTIO IR Imager on the 4-m, 1.5-m, and 1-m telescopes. A table summarizing the performance on the different telescopes is given below. It should be noted, though, that the values presented are representative, particularly the background values, which can vary by factors of 2 depending on airglow (below 2.2 um) and temperature (2 um and above). CIRIM Performance ADU/sec on 15th mag star Background (mag/sq arcsec) Filter 1.5-m f/7.5 4-m f/7.75 1.5-m f/7.5 4-m f/7.75 J 871 135 16.5 16.1 H 887 136 14.5 14.2 K 608 94 12.8 12.4 Notes: The 1.5-m measurements were made on a colder night (8 deg C) than the 4-m measurements (15 deg C). Signals in the KS filter are slightly less than those in K, while backgrounds are approximately a factor of 2 lower. The gain of the instrument is 9 e-/ADU. Which telescope/focus should I use? Users who want to propose to use CIRIM obviously need to decide which telescope to use; it is also useful (and, in some cases, necessary) to decide which focal ratio will be used as well. The following rules should be helpful. They are intended for broadband (JHK or JHKS) work; prospective users of narrowband filters can use the rule of thumb that both signal and background in a 1% filter are about 1/20 those in the corresponding broadband filter. On the 1.5-m telescope, a practical limit for stellar photometry corresponds to a 1-sigma limit somewhat fainter than K=20, for about an hour elapsed time on the telescope. That is, if you just want to detect something (20% photometry), you can work down to at least K=18, but if you want reasonably precise photometry, you are limited to about K=16. This rule applies to aperture photometry in the case where the image is not badly under-sampled - i.e. the limits at f/7.5 on the 1.5-m will be several tenths of a magnitude brighter. If you have a field where you can do point-spread fitting and then calibrate the data using a brighter star, you can probably gain as much as a magnitude, although this has not been demonstrated as yet. On the 1.5-m, the choice of focal ratio depends on the program. For observations where field is important, or where the objects are reasonably resolved (galaxies, for example), the f/7.5 focus is preferable. The f/13.5 focus provides a better scale for stellar work, and may be better for extended objects which are significantly smaller than the field of view, because of the better scale. The pixel scale at this focal ratio still under-samples the telescope images - typical measured image FWHM is < 2 pixels - but for most aperture photometry this is not critical. For very precise photometry, it is probably desirable to illuminate individual pixels more uniformly, and for PSF fitting, it is necessary to have better sampling. In both of these cases, use of the f/30 focus is preferred in principle. The field of view is, of course, quite small (75") and the secondary is significantly undersized (it was originally intended to chop 20'), with consequently lower signal levels, and the background is presently somewhat higher because we don't have a precision Lyot stop with central obstruction for f/30. We therefore recommend use of the f/30 only for precise photometry (better than 2%) and for work in crowded fields. For other stellar photometry, we would recommend f/13.5, except for survey work, where f/7.5 is usually preferable. Note that it is possible to switch between f/13.5 and f/7.5 on the 1.5-m in ~20 minutes, while the f/30 top end must be scheduled in advance. Therefore, if you think you may need to use the f/30 focus but are not certain, it is desirable to consult with one of the local staff before submitting your proposal. On the 4-m telescope, there is not at present much advantage to use of the f/30 secondary. Our tests show that the image improvements made on the telescope have resulted in f/8 images that are very nearly as good - if not just as good - as those produced at f/30, which is to say < 0.8" FWHM. While this means that the images are somewhat under-sampled (usually 1.5-2 pixels), this is adequate for aperture photometry. In addition - and this is important - the automatic guider on the 4-m only works well at f/8; the available field of view of the guider and the field of view and scale of the guide probe are such that finding guide stars at f/30 is problematical, and guiding on them is difficult. Finally, relative to the f/8 focus, the f/30 focus shows higher background and lower signal levels (due to the small Lyot stop and undersized secondary respectively), just as on the 1.5-m. As a result, we recommend use of the f/8 focus for all 4-m programs; anyone who thinks his or her program may be one of the rare exceptions is advised to consult with us before submitting a proposal. IRS News New Gratings We have received and tested the new cross dispersed grating for the IRS. The grating has survived several thermal cycles in the lab and appears to be working as expected. The cross dispersed grating will offer full spectral coverage from 0.9 to 2.4 um in 5 orders in a single setting. The slit length available with the cross dispersed grating will be slightly reduced to avoid overlap between orders. The resolution is similar to that of the old 12 l/mm grating and will presumably replace that grating for all applications. Since it does not have to use a blocking filter and you get all the light from both orders it should be more efficient than using a normal grating. The cross dispersed grating will be tested on the telescope in early February so look to this space in the future for actual performance measurements. We have been able to test the 1.9-um blaze 75 l/mm grating and find it to work well. The true blaze wavelength appears to be somewhat shorter than the nominal value, with the result that it is recommended for use in first order rather than second order in the J band. We expect that for most programs at "low" spectral resolution, the grating of choice will be the cross-dispersed grating and not the 1.9-um 75 l/mm grating, but the latter does offer slightly higher resolution. Given the performance of both gratings, and the absence of a long slit in the IRS, we will no longer offer the 12 l/mm grating; its performance beyond 3 um is limited by high count rates and a higher resolution grating will, in general, give higher overall efficiency. Popular Gratings Since the IRS can have only two gratings installed at once, and since grating changes require warming up the instrument, opening it, and then cooling it down again, we try to block-schedule particular grating combinations, and in some cases we have to modify individuals' requests in order to schedule time for them. Based on our experience to date, the most popular gratings are the cross-dispersed grating, the 4.5-um 75 l/mm grating, and the 210 l/mm grating. The first offers a complete 1-2.5 um spectrum at low resolution; the second provides a resolution that is suitable for classification work without requiring multiple grating settings; the third provides higher resolution and is a good choice for the long wavelength windows. In particular, users who request the 1.9-um 75 l/mm grating should clearly state why the cross-dispersed grating is not useful to them; if the reason is inadequate spectral resolution they should also indicate why the higher resolution 75 l/mm grating (4.5-um blaze) is not preferable. Similarly, although the 632 l/mm grating does offer the highest possible spectral resolution, it is incompatible with 3 and 5 æm programs and thus harder to schedule; users who request it should indicate whether the 210 l/mm grating is acceptable or, if not, why. We will, of course, try to schedule whatever is requested and justified, but CTIO's block-scheduling policy does imply that an unusual IRS configuration with a TAC grade near cut-off will lose out in the scheduling process to a proposal that does not require an extra instrument set-up. J. Elias, R. Elston, B. Gregory
Previous Article Next Article Table of Contents