College of Mathematical & Physical Sciences
Department of Astronomy
|
|
The Ohio State University College of Mathematical & Physical Sciences Department of Astronomy |
Attendees: Pat Osmer, Bruce Atwood, Dan Pappalardo, Tom O'Brien, Darren DePoy, Jerry Mason, Jen Marshall, & Rick Pogge,
This is the first MODS meeting since the spring break at OSU.
Brief Notes
MODS Flexure Control System
Bruce Atwood gave an extended presentation on a concept for the Flexure Control System (FCS) will work for MODS.
The basic idea is as discussed earlier, namely that we will use the spectrograph optics itself to keep the optics aligned under varying gravity loads and temperatures using a "bypass grating" embedded in the science grating (at the location where the secondary mirror effectively "shadows" the grating).
The basic design criterion is that we want to design only one system for all gratings and configurations (angle of use, red/blue channel, etc.).
Specifically, Bruce is exploring a system using a 1.55-micron diode laser commonly used for fiber optics communications as the monochromatic light source, and an array detector located alongside the science CCD arrays. The use of an array rather than a quad cell is that this gives us plenty of latitude for "DC" tilt and perpendicular motion due to thermal expansion of the structure, different gravity loads, and hysteresis in the structure.
The detector under consideration is a 1282 InGaAs array made by Finisar. The device has 60-micron pixels and is 7-mm square.
The properties of the bypass grating are that it should have zero-th order at the middle of the science grating tilt range (which will in all likelihood be at or close to the nominal tilt location for first order), giving us multiple orders either side of this "nominal" tilt from zero-th order on the science grating to twice the nominal tilt at the other extreme. Given the 7mm array size, an order spacing of the laser spots at the detector of ~5mm will make sure that there is always one spot on the sensor, and times when there are two adjacent orders on the sensor (no blank spots in our coverage). This permits continous tilt control of the gratings rather than a set number of fixed positions as with ta quad cell.
A design for the bypass grating that Bruce put up is for a dual-blazed grating with 6-10 facets/mm and a 15-degree blaze. Each facet is essentially symmetric. For a tilt range of +/-0.5 radians, we would need about 100 spots, or 50 orders either side of zero-th order on the bypass grating. This puts the constraint on the bypass grating design that it deliver reasonable power in the laser spots in the n=50 order (note this is very different than the design of a science grating where you try to contrive the blaze to put most of the power into one order on one side of zero-th order). In principle, such a grating could be made in-house on our CNC mill, the characteristics of this grating being not that extreme as far as ruling is concerned. We will explore this option further, although other possibilities (e.g., variations on a theme of Ronchi rulings) are up for consideration.
Given a centering precision of 1-micron for the laser spot on the detector, in 50-th order this requires that the laser have a wavelength stability of 1 part in 350,000. This sounds daunting at first sight, but there is a family of commercial wavelength-multiplexing fiber optics systems that have a single-mode laser in a case with integrated thermoelectric cooler and reference sensor that can provide a wavelength stability of 1 part in 106 per 1-degree C on the outer case. Such units are available for about US$2K. If we were to go this route, we would buy one of these units, feed the laser beam into a 9-micron diameter fiber and into a microscope objective to create the reference beam that enters the MODS optics train.
The InGaAs detector would be readout in fixed format and probably fed to a "simple" DSP program to pick out and measure the spot centroid, and provide information to the motor control computer to effectively "guide" the laser spot, removing any temperature or mechanical flexure. We can active control the intensity of the laser via pulse-width modulation.
What mechanisms get moved to do the flexure correction? The current concept is that the camera primary will move. It has a total travel of 1-2-mm, by comparison the collimator has ~10mm of travel. Both optics are mounted on 3 actuators for tip/tilt/focus motion.
There was a some discussion of various aspects of the concept, and it was agreed to proceed to a more detailed design, with the idea of purchasing components to test the suitability of the system, get real numbers on stability of the laser, experience reading out the array, etc. Jen Marshall has expressed interest in helping with that effort now that her slit testing is nearing completion.
The next MODS meeting will be Tuesday, April 10 at 3pm in the Astronomy Conference Room.
R. Pogge, 2001 April 2