Overview and project team
	Scientific case
	Science objectives Mask - Mode MOS vs. Multislit Assembly
	Mask design
	Overview How to use the Interactive Mask Design Interface (New! V.2.0)   How to use lists of celestial coordinates (blind mode)
	Mask manufacturing
	The control computer and software The milling machine Mask cutting and processing The masks: quality checks
	MOS observations
	Operation flow Pre-commissioning tests at Cima Ekar  MOS commissioning at the TNG
	Documents

Overview

The Mask-Mode multiobject spectroscopy (MM-MOS) at the Galileo National Telescope is an extended MOS capability complementary to the real-time built-in multislit assembly of the Low Resolution Spectrograph DOLoReS. The Mask-MOS facility represents a highly flexible tool for multiobject spectroscopy, with quality standards comparable to that of MOS facilities presently available at 4m class telescopes, at a very affordable cost (£ 40000 Euro) and with reasonable equipment size.

• MM-MOS will allow spectroscopy for many tens of objects (usually about 30-40) simultaneously over a useful field of 6' ´ 9'.5.
• The slitlets will have fixed width 1.1 or 1.6 arcsec and user-defined length and tilt. Stacking of two rows of spectra on a single mask is also possible using an appropriate filter. Up to 100 spectra will be obtained with this technique.
• Mask design will be easy using IMDI, a new, user-friendly IDL-based Interactive Mask Design Interface. Aperture masks will be efficiently designed using this portable mask design software both at La Palma and at home institutions.
• Up to 5 slit masks, manufactured at the TNG Observatory during daytime and mounted on removable plate holders, are available for observation each night. The number of masks that can be manufactured for each program is limited: see the TNG Instrument Pages for the current TNG policy. The masks are remotely positioned using the standard Slit Exchange Unit of DOLoRes.

Scientific motivations for an extended MOS mode

Science objectives

In the VLT era, very deep cosmological surveys will be conducted at 10m class telescopes using large-field spectrographs (e.g., VIMOS). Still, multiobject spectroscopy at 4m class telescopes still offers a good alternative for less extreme targets in fields with size of the order 10', such as:

• redshift and stellar populations of galaxies in clusters at intermediate distance;
• galaxy kinematics in cluster and field galaxies at intermediate redshift;
• individual stars in star clusters and Local Group galaxies;
• globular cluster systems in nearby galaxies.

Mask - Mode MOS vs. Multislit Assembly.

The multislit assembly provided with the Low Resolution Spectrograph DOLoRes (19 slits with fixed length of 20 arcsec) will be the standard option for most programs. It has several practical advantages, first of all a remotely controlled slit setup, possibly during the observing night, and optimal quality of the slitlets.

However, there remain a number of scientific cases where the uniform and fixed (along the spatial axis) coverage by the multislit assembly seriously affects observation efficiency. On this basis, an extended MOS mode was proposed by members of the Instrument Definition Team of DOLoRes (Held, Ramella & Cristiani 1995). A similar concept was also adopted for MXU at ESO.

In the Mask-Mode MOS (MM-MOS), the targets will be observed through multislit masks prepared for each target field using a mask manufacturing facility locally available at the TNG. The main advantage of this mode is its flexibility: mask can include a large number of slits with arbitrary position, orientation and shape, with very little constraints aside from those needed to prevent spectrum overlap. Obviously, masks also have some drawbacks: they must be prepared in advance of the observing night, requiring specialized staff and careful processing.
 
 

Mask design

The following flow chart summarizes the mask design and manufacturing process outlined in the following sections.
 
 

• The first step for making the masks will usually consist in collecting short exposure images of the target fields with DOLoRes well in advance of the observing run. This pre-imaging should be routinely taken in service mode for proposals approved by the TAC. In some special cases masks can be produced from a target list with astrometric precision better than 0.2 arcsec.

• Measuring the target positions from the image, setting up the slits, and producing the final drawing necessary to operate the cutting machine are performed using a newly written Interactive Mask Design Interface (IMDI), a graphical interface based on IDL widget procedures.

• IMDI was conceived to be fully integrated into the DOLoRes User Interface in the telescope/instrument control environment. A standalone version is provided to project PI's to design the masks at their home institution (important: IDL 5.2 or later must be available), using the pre-imaging observations.

Using the Interactive Mask Design Interface. The rectangles represent the location of the spectra, calculated from the instrument database. Most operations are executed using cursors, buttons and sliders in a self-explaining sequence. Instructions here.

Some features of IMDI:

• The basic setup parameters (wavelength range, spatial scale, etc ...) are directly read from a setup database and the image FITS header. The user only needs to select the preferred grism, load the image and start picking up the target objects.
• The centering, as well as the length and position angle of the slits, are interactively set under control of an anti-overlap algorithm. Stacking of two rows of spectra is also possible.
• The save/restore feature allows users to suspend and resume mask editing.
• The end products of IMDI include, in addition to the mask-cutting drawing in the standard Drawing Interchange File Format (DXF), further documentation specifying the setup of DOLoRes and telescope associated with the mask.
• All the files shall be sent to the TNG team (e.g., via ftp upload into a dedicated area) for mask production under the responsibility of mechanics staff.

Mask manufacturing
 

The control computer and software
 
 

A dedicated PowerPC computer, equipped with a network interface card and TCP/IP communication software, controls the cutting machine via the Quick Plate (QP) Software. The drawing files produced by the astronomer are converted into low-level instructions used to drive the milling machine.

 
 

The milling machine
 

A general view of the mask manufacturing facility. All the electronic components and the control computer are conveniently arranged in a dedicated rack close to the CNC machine. Mask production is managed by TNG staff only.

 

• The milling machine, heart of the system, is a high-precision commercial CNC cutting plotter by Scripta s.r.l., compliant with the required positioning accuracy (0.01 mm).
• The blank mask, a thin 10´ 10 cm square plate of brass, is securely clamped to the bed of the machine by a custom-made mask holder.
• The cutting tools are special cylindrical micro-mills with diameter 0.2 and 0.3 mm, corresponding to about 1.1 and 1.6 arcsec slits at the TNG.
• The cutting machine is equipped with a custom high-speed electrospindle reaching 90000 r.p.m.

Mask cutting and processing

		A detail of the machine head cutting a mask during our tests of different materials and tools.
An example of a test multislit mask where several slits were cut using two different tools (200 and 300 m m in diameter). This plate was made out of a 0.2 mm thick bronze sheet. After cutting, the masks are effectively cleaned by ultrasonic washing. Two reference holes ensure accurate repositioning of the mask in the plate holder. A third pinhole made during mask cutting (not shown here) will ensure correct mounting.

 
 
 

The masks: quality checks
 

Checking the quality of the slitlets is an important step of mask production. After inspection at the microscope, the plates are painted with an antireflection coating, mounted on the holding frames, and kept in a dust-free area until the frames are mounted in the slide assembly of DOLoRes.

 
 

		This figure shows a microphotograph of a 200 m m slitlet after ultrasonic cleaning. The enhanced contrast image allows assessing the degree of micro-roughness of the slit edges. Both the r.m.s of the slit edges and their parallelism over a 20'' length are better than 3 m m.

Several tests were made to obtain the best possible quality for the slits. In particular, we measured the micro-roughness and parallelism of the slit edges and the relative positions of the slits. We ended up with the following characteristics of the masks:
 
 

These characteristics basically meet the specifications, except for the minimum slit width that is constrained by the size of micro-mills available on the market.
 
 

MOS observations.
 
 

Operation flow
 

This flow chart outlines the telescope pointing and target acquisition procedures, once the masks have been mounted in the DOLoRes slide assembly. The adopted procedure assumes that the masks can be repositioned on the optical beam with high accuracy (i.e. 5 m m or better). The pointing sequence is guided by a MOS Pointing User Interface developed during the pre-commissioning tests.
 
 

Pre-commissioning tests
 
 
 

The mask design software and the pointing procedures were pre-commissioned in advance of th efirst light of TNG using the 1.8m telescope at Asiago Cima Ekar. The imaging low-resolution spectrograph AFOSC was employed as a working model of DOLoRes. MOS spectra of galaxies in three nearby Abell clusters have been obtained on the night of Jan. 11, 2000 using a scaled version of the masks for TNG.

Instrument history
 

• Since June 27, 2000 the milling machine is back to activity in the machine-shop of the TNG. The software and hardware have been succesfully tested by two of us (EH,CC) together with the staff of the TNG. Two special masks have been realized and used for tests of DOLoRes during the commissioning phase.

• The MM-MOS has been successfully commissioned in Dec. 2000. Since then, several guest observers have made use of MOS.

Documents
 
 

The following documents are available for download as gzipped postscript files:

• Held E. V., Ramella M., Cristiani S. 1995, Proposal to the Consiglio delle Ricerche Astronomiche for an extended MOS at the TNG.
• Held E.V., Ciattaglia S.C., Giro E., Zitelli V. 2000. Mask Mode MOS at the TNG: a flexible approach to multi object spectroscopy, in Telescopes, instruments and data processing for astronomy in the year 2000, G. Sedmak ed., 12-1 May 1999 S.Agata (NA), Mem.S.A.It., in press
• Held E.V. 2000, Report to the C.N.A.A. on the status of the MM-MOS project.

Questions, comments and suggestions are welcome.