The Binospec spectrograph has been commissioned at the MMT telescope in fall 2017. Binospec is a multislit imaging spectrograph with two 8' x 15' fields of view, wide wavelength coverage, and high efficiency. There are two identical all-refractive spectrographs, one for each field. Each channel has several gratings and filters.
This page provides preliminary information about Binospec for potential users and proposers. Binospec was successfully commissioned in fall 2017 and is scheduled for shared-risk science observing in spring 2018.
Each side has a 4K x 4K deep depleted E2V CCD detector with 15 micron pixels and a scale of 0.24 arcsec/pixel. The resolution per pixel is tabulated below. The wavelength coverage of a slitlet depends on the distance from slitlet to optical axis along the X direction (along the short side of the masks). From near (central) side of the mask to the far side, the change in coverage is approximately 900 A to the blue for the 600 gpm grating.
Slitmasks will be machined by a laser cutter at the MMT. Slitmasks will be configured by software written at SAO and running on a server hosted at the MMT. Users will log into a web interface to interactively design masks and submit them. Masks must be configured ahead of time to allow time for machining. Masks with long slits are also provided; the widths available are 0.75", 1.0", 1.25", 1.5", and 5". One long-slit mask will always be loaded, typically the 1.0". If different long-slit widths are needed, please request these widths in your target catalog submission so they can be loaded ahead of time. 10 slitmasks can be loaded simultaneously. Masks can be changed during the daytime (and potentially during the night pending approval of this mode and training of staff).
|Grating lines/mm||Order||Blaze angle||Angle of incidence||Anamorph Demag.||Coverage (A)||Dispersion (A/pixel)||Pixels per 1'' slit||Resolution in 1'' slit|
|Grating lines/mm||Allowable central wavelengths|
|270||5501 - 7838 A|
|600||5146 - 8783 A|
|1000||4108-4683, 5181-7273, 7363-7967, 8153-8772, 8897-9279|
|Ghosts may be troublesome with 1000 gpm grating at
central wavelengths 5600-8500 A, worst near 7100 A.|
Throughput of 1000 gpm grating will be low in the red.
Binospec is expected to be several times more efficient than Hectospec due to throughput and improved sky subtraction. Here are on-sky throughput measurements made with the three gratings during commissioning, which should be taken as lower limits as the sky was not completely photometric. Three central wavelength settings are included for the 600gpm grating.
SAO has built an exposure time / sensitivity calculator for Binospec and Hectospec. A preliminary version is available at this link: Binospec exposure time calculator , which also gives you an option to download the package.
A direct link to the spectroscopic exposure time calculator for spectra is: http://hopper.si.edu/etc-cgi/TEST/sao-etc . For imaging see this calculator (select MMTcam): http://www.cfa.harvard.edu/mmti/megacam/obs_manual/exptime.html
A model of the optical train of Binospec is shown below, illustrating the single slitmask at the top, and the paired spectrographs, filters, gratings, and detectors.
Optical train of Binospec.
The paired fields of view on the sky are shown below in green - the separation between them is 3.2 arcmin. The yellow boxes are the guider patrol regions, and the blue box is the wavefront sensor patrol region. The slitmasks cover the guider regions and holes are machined into into the masks at the location of guide stars, to allow for slitmask alignment and guiding.
Binospec on the sky. Green rectangles are the spectroscopic fields, yellow are the allowed positions for guide stars. The blue rectangle shows the wavefront-sensor patrol area. Red dots are targets; blue are guide stars; short cyan lines show the slits assigned to targets.
Binospec observations will be done in a queue, similar to Hectospec and MMIRS. The queue will be administered by MMT staff and the instrument will usually be operated by MMT queue observers. PIs/guest observers will typically not come to the telescope. PIs will submit observing catalogs and mask designs through the MMT queue scheduling interface. Prior to the run, PIs will get an email asking them to log in to the queue scheduler to upload catalogs and observation details.
Binospec can hold 10 slitmasks, and these can be changed out freely during the day (unlike MMIRS), and potentially at nght. However, there is a lead time for machining masks.
The grating tilt controls wavelength coverage. We anticipate that a limited number of standard grating tilts will usually/initially be used, to ease calibration requirements and configuration changes during the night. The instrument can be changed from spectroscopic to imaging mode quickly, and will have g, r, i, z filters for each side. We do not anticipate initially offering mixed observing between the sides (ie imaging + spectra, or two different gratings).
Binospec uses guiders offset from the main field of view, patrolling the yellow box regions in the figure above. Guide stars can be selected automatically from GAIA (preferred) or GSC2 catalogs. Thus, it is necessary to supply coordinates that are on the GAIA or GSC2 system. The correct coordinate system is critically important for both longslit and slitmask spectroscopy.
Targets for longslit spectroscopy will be offset onto the slit by acquiring stars onto the guiders, whose positions have been calibrated relative to the slit. Thus, your target position needs to be on the same system as the guide stars. The pointing can be tuned up by tilting the grating to zeroth order and taking a direct image of the target through the slit. This should work for objects brighter than some magnitude (R~21). For faint targets, PIs should provide a nearby offset star on the same coordinate system as the target. Offset star coordinates and any other useful information about the target can be entered in the "notes" section of your observing catalog.
Acquiring using a slit-viewing camera requires a "mirrored" slit mask whose use has not yet been commissioned, so good coordinates that allow the above procedure to work are a requirement.
For multi-slit masks, the PI will provide a list of targets on GAIA or GSC2 coordinate systems. The BinoMask slitmask design program will add suitable guide stars and cut alignment holes in the slitmask, through which the guiders will image the stars and tune up the alignment of the mask.
Slitmask design will be done by PIs using the BinoMask web based interface. Masks will be machined with a laser cutter on the mountain by SAO staff. Masks will need to be submitted well in advance of the start of queue observations for a semester.
Please see the BinoMask slitmask design interface tutorial webpage for a more detailed walkthrough of using the software.
You will be running a web-based mask making program called BinoMask which will take a text file as input, containing your targets to be observed. If your objects are on the GAIA coordinate system, binomask can add guide and wavefront-sensor stars for you. If not, you must create a second catalog with such stars on the same coordinate system as your targets, extending to at least 0.4 degree in radius from the center of your target field. For guide stars, please select a magnitude range of r=14-18; for WFS stars, magnitude r=11-14.
The format for the catalogs is an ASCII file with columns. You may separate entries either with a comma or a tab. The first line of the file is a header:
name, ra, dec, magnitude, priority, pm-ra, pm-dec, epoch, typeThe following lines contain entries for those columns.
Only ra and dec are required, but name, magnitude, & priority are also recommended. (1=highest priority, then 2, 3, etc). Coordinates may be in decimal degrees or sexagesimal. Type should be 1 (or “target”), 2 (“sky”) or 3 (“standard”).
A queue scheduling interface similar to the MMIRS user interface allows PIs to submit mask designs and track their program progress. The materials/machining cost for slitmasks will be determined at a later date.
Mask making is integrated with telescope scheduling software. Thus if you have been granted time, you will receive an email from firstname.lastname@example.org with instructions to log in, make masks, and submit catalogs. You may also use the mask making software outside of that framework. To run the BinoMask software visit scheduler.mmto.arizona.edu/BinoMask/. Slit widths are also set by the user. As an aid, the MMT web site states that the median seeing is in the range of 0.77" - 0.85" based on data from 2003 to the present. 1 arcsecond slits are a good starting point.
Note that the differential atmospheric refraction across the 15 arcminute field is significant, so that the location of slitlets depends on the hour angle at which the mask is observed. In practice this means that the mask machining would change slightly depending on whether the mask is observed when HA is East or West. Thus the date and time of observation specified during mask design are significant. Observable hour angles will be taken into account during queue scheduling.
You may get the error “slits can not be made” which often occurs because the rotator setting is not correct. Try adding or subtracting 180 from the present setting.
In order to obtain the maximum scientific impact from Binospec, we will to continue the queue scheduling that has proven successful with Hectospec, Hectochelle and MMIRS. Queue scheduling offers the considerable advantage that we can flexibly adapt the observing program to the current conditions. We will reserve conditions of perfect transparency and superb seeing for those projects that can take best advantage of excellent conditions. Queue scheduling also allows observations to be obtained at low air-mass and allows continuous observation of high priority objects. The instrumeht will be operated by MMT staff queue observers, and proposers are not generally expected to travel to the telescope.
Binospec will have a data pipeline, written in IDL by SAO staff and publicly available, that is similar to the MMIRS pipeline. The SAO TDC is anticipated to return reduced data to the PIs. The timescale for data reduction is TBD. Availability of quick look reductions or quickly returning raw and reduced data are recognized as desirable for some programs and will be addressed. A quicklook for queue observers to assess data quality is in progress.
The raw data are stored as multiextension fits files; each CCD has 4 amplifiers, each of which is stored as an extension. Data taken are archived at the CfA. A pipeline under development (in late 2017) will be used to process the data and provide wavelength calibrated, relative-flux corrected spectra. In progress: the archive data system will send you a notification the day after any of your data is taken, at which time you may download the raw data. A few days later, you will receive another message concerning the reduced data. We are also developing an addition to the MMT queue catalog management pages that will allow you to view available data files and download them from your catalog page.
These images were taken during Binospec commissioning in fall 2017.
A multislit single exposure of 600 sec. Blue wavelengths are toward the center; red is to the left in the left frame and to the right in the right frame. Note the cosmic ray hits.
Combination of 4 exposures. Ca H and K absorption lines are visible as a doublet near the blue end of several spectra.
A 600 sec single exposure on a mask with many short, 2 arcsec long slits, using the 600 gpm grating
A direct imaging exposure, 50 sec in g band.
We will update this page as information is finalized during Binospec commissioning and preparations for the first semester of operations in 2018A. Please contact Benjamin Weiner, email@example.com, for further information.