AAOmega is a dual-beam spectrograph. The slit units (one for each field plate) which each contain 392 science fibres are fed into a single collimator and then separate into the Blue and Red arms of the system via a dichroic beam splitter. There are two dichroics, one operates at 570nm and one at 670nm. Each arm of AAOmega uses a separate Volume Phase Holographic grating (VPH).

Contents:

• Selecting an AAOmega setup
  
• The AAOmega grating set
  
• Considerations in setup selection
  
• Grating Efficiencies
  
• Blaze Angles for VPH gratings
  

Selecting an AAOmega setup

AAOmega can be configured to observe the entire optical spectrum over the wavelength range 370nm-900nm, with a small overlap between the red and blue arms around the dichroic wavelength (570 or 670nm). The grating set available allows a range of resolutions between R~1,000 and R~10,000. The fibre spectra are recorded onto the 2Kx4K E2V CCDs with light dispersed along the 2K axis NOT the the 4K axis. Hence, at low resolution the entire accessible spectral range is recorded at once, but at higher resolutions the user must tune the wavelength range to that which best suits their requirements.

The full list of gratings can be found in the table below. An on-line AAOmega Grating calculator is available to simplify the process.

The AAOmega grating set

Grating	Blaze	Useful wavelengths	Coverage (single shot)	Angle	Dispersion	MOS Resolution*
	nm	nm	nm	Degrees	nm/pix	R
580V	450	370 to 580	210	8	0.1	1300
385R	700	560 to 880	320	8	0.16	1300
1700B	400	370 to 450	65	18	0.033	3500
1500V	475	425 to 600	75	20 - 25	0.037	3700
1000R	675	550 to 800	110	18 - 22.5	0.057	3400
1000I	875	800 to 950	110	22.5 - 25	0.057	4400
3200B	400	360 to 450	25	37.5 - 45	0.014	8000
2500V	500	450 to 580	35	37.5 - 45	0.018	8000
2000R	650	580 to 725	45	37.5 - 45	0.023	8000
1700I*	750	725 to 850	50	37.5 - 45	0.028	8000
1700D*	860	845 to 900	40	47-48	0.024	10000

* Resolution with the SPIRAL IFU will be slightly higher due to the smaller fibres used in the IFU.
* For high resolutions observations in the CaIII region, 1700D is the appropriate grating. The table above shows information at the central super blaze grating wavelength. For CaIII observations, 1700I does not have an increased coverage over 1700D. Wavelength dependent resolution is a property of the VPH gratings.

Considerations in set-up selection

1. Grating changes will not be performed during the night, only during the afternoon. Wavelength changes can be performed during the night, but there is an overhead.

2. Where possible, the blue arm of the system should be set to allow the strong 557.7nm skyline to fall within the observed spectral range. This will allow sky subtraction and fibre throughput calibration without the need for twilight flat fields or dedicated offset sky frames.

3. It is not required to have the two arms of the system overlapping in wavelength. However, leaving some overlap allows spectra to be spliced together.

4. Spectral curvature. As with all spectrographs, the spectra follow cuved paths on the CCDs and wavelength is not a constant function of X-pixel position between fibres. At low resolution this is barely noticed. However, at higher spectral resolutions there is a small mismatch between the observed wavelength of the central and outer fibres. This range of wavelengths is given in the AAOmega Grating calculator.

5. Departures from the standard default values for each grating are acceptable.

6. The blue arm CCD has a number of regions with closely-packed bad columns. For some projects it may be possible to tune the central wavelength to reduce the effects of these. The user should contact their support astronomer to discuss options.

7. Blaze angle. The collimator-to-VPH and VPH-to-camera angles are typically set to be equal. This gives the peak system throughput at the central wavelength, with a slow role off to shorter and longer wavelengths. For certain applications, one may wish to operate with an asymmetry in these angles which will boost the system sensitivity at shorter/longer wavelengths, but at the expense of sensitivity at longer/shorter wavelengths. Read the notes below, or call your support astronomer for a discussion of this very important concept which is specific to VPH gratings.

8. Due to the long fibre run (38m prime focus to Coude West) and the optics of the 2dF prime focus corrector, the system throughput below 370nm is very poor and there is little point attempting to observe at shorter wavelengths.
   

9. High resolution CaIII observations. The 1700D grating is specifically designed for observation of the CaIII lines at ~860nm. It gives a better response at this wavelength than the 1700I grating. However, it cannot be used at any other central wavelength. When observing the CaIII there is no advantage to observing with the 1700I grating over 1700D. If one wishes to observe at high resolution at red wavelength, but away from CaIII, then 1700I must be used.

10. For Service Applications consider the use of standard grating configurations as the probability of service observations being successfully undertaken is significantly higher for these settings

11. Ghost reflections: Like all diffraction gratings, the VPH gratings do induce some artifacts in the observed spectra. The dominant artifact is a prominent ghost reflection (essentially an out of focus 0th order image of the slit). The gratings are designed to throw the ghost out of the field of view for the most commonly used wavelength setups. For more unusual settings the user MUST visually check an arc frame to ensure that there are no ghost images that would damage critical wavelength ranges.

Grating Efficiencies

Preliminary efficiency curves for the AAOmega gratings are available here. All measurements were taken prior to AR coating, so all quoted efficiencies should be increased by a factor 1.08. All curves are approximate, with a 25mm aperture used to test the efficiencies. The different curves for each grating correspond to different grating angles; users can select whatever grating angle is most suitable for their observations. Note that the test may have been done with the grating (and hence slant angle) reversed with respect to the grating calculator. Note that altering the grating angle also has a 2nd-order effect on resolution and wavelength coverage; this becomes significant at high dispersion.

Low resolution	Medium resolution	High resolution
580V 385R	1700B 1500V (note angles +90deg) 1000R 1000I	3200B 2500V 2000R 1700I 1700D

Blaze Angles for VPH gratings

AAOmega uses Volume-Phase Holographic (VPH) transmission gratings. These have flexible blaze angles. Each grating has a specific design blaze angle which will give the absolute maximum efficiency with that grating (the super blaze). This peak efficiency reduces smoothly with wavelength away from that. The usual setup for most programs is therefore to have the grating set at its super blaze angle and the camera at twice this angle to centre the maximum efficiency wavelength on the CCD. The complication comes when the observer wishes to observe at a central wavelength which is some distance away from the super blaze angle for the grating. This would mean observing with the grating and camera angles highly asymmetric, and therefore operating on the low efficiency (and rapidly falling) part of the blaze envelope for the grating. The solution is to tune the grating and camera angles to new values. This shift in the grating angle will shift the blaze profile away from the super blaze, flattening the steep wings of the super blaze envelope and boosting system performance at the desired wavelength(s), with the expense of a slight reduction in overall peak performance in comparison to the super blaze setting.


Sarah Brough (sb@aao.gov.au)