The Ohio State University College of Mathematical & Physical Sciences Department of Astronomy

Attendees: Pat Osmer, Darren DePoy, Bruce Atwood, Paul Byard, Tom O'Brien & Rick Pogge.

Pat has received a letter from James Breckenridge, Program Manager for Advanced Technologies and Instrumentation at NSF, notifying us that our most recent MODS proposal was successful and that we have been awarded a grant to develop MODS. More details will be forthcoming.

Yahoooo!

Back to business...

Optics

Paul Byard reports that the optics bid process is continuing and we are starting to receive responses or promises of responses. Stay tuned...

Mechanical Systems

Tom O'Brien reports that the design of the grating cells is about done, reviewing the mounting system, counter weights, and various features. Mike Lannon (undergrad) is assisting full-time this summer. Tom finished a full-life test of the tilt-axis conical bearings last week and reports that they work great, with good friction characteristics. Parts for the prototype grating drive are arriving, and some machining will be required. All is on schedule.

When a grating is removed from MODS, the whole cell is as a piece, grating and all. Once a grating is mounted and aligned (front surface aligned so that it rotates about the tilt axis and the grooves aligned parallel to the tilt axis), it stays in its cell pretty much for life. We may want to consider building a permanent test fixture if we have to do a lot of gratings often, but we can probably do the alignment on-instrument (although with the instrument itself off-telescope).

Tom also presented a preliminary analysis of motion times related to changing masks and switching between imaging and spectroscopic mode.

Grating select (rotate between grating and imaging flat positions), the motion is bi-directional so the longest move is 180-degrees either way. Typical time is 5-10 seconds.

Grating Tilt: home-and-back typically requires 10-degrees of motion, taking about 45-seconds. Gratings with larger nominal tilt from "home", or used at higher order, will of course would take longer.

Mask select (move mask cassette to select a mask) is ~25 seconds to move 10 positions.

Mask insertion/retraction (move storage to deploy or the reverse) takes about 10 seconds, once a mask is selected.

These numbers are clearly preliminary and subject to change, but they give us something to work while trying to understand our observation setup overheads.

Acquisition & Guiding

Darren considered the question of how bright a guide star we would need to guide at 20Hz. In new moon, 0.8 arcsec seeing, no filter, and stock 24-micron pixel SITe CCD specs, we can get SN=50 on a V=21 M2v star in 1 second of integration. On average, there is one V=21 mag star per square arcminute of the sky. This means we should not lack for guide stars, and in principle we could guide on any star visible on the POSS plates. Adding more moon doesn't change this conclusions much. Darren points out that these numbers also remind us that LBT is a big telescope.

Paul looked at simple re-imaging optics, and finds that he can get good images in a 36" radius using a simple achromat lens providing 2:1 focal reduction. "Good" in this case is D₈₀<0.2-arcsec. Thus, making simple re-imaging optics will presents no particular challenges.

Wavelength Stability

We need to determine how much on-chip wavelength shift we can tolerate and still preserve stable flat-field and wavelength calibration vis-a-vis the nominal science goals for MODS. This will then provide us with a specification for flexure, which feeds into all aspects of the detailed mechanical design, as well as the active flexure compensation system. For example, it will help us to determine if an FCS can run open-loop (lookup table) or whether we need a closed-loop system.

One way to approach this is to look at the effect of small (pixel and sub-pixel) wavelength shifts on the ability to remove flat-field fringing effects at far-red wavelengths. To get some real numbers, Rick is experimenting with CCDS spectral flats, lamps, and object spectra to try to get a better feel for what this gets into. In particular, he is trying to better understand the rationale discussed in the DEIMOS spectrometer design documents, where they have settled on a target flexure specification of 0.25 pixels rms (DEIMOS is the new Keck multi-object optical spectrometer). In DEIMOS they will meet this spec by using a closed-loop flexure compensation system, while in the ESI spectrometer they are using an open-loop system (ESI is in operation, DEIMOS will probably go online in early-2001). Rick will continue this work and make a full report in later this summer.

R. Pogge, 2000 July 8