Sloan Digital Sky Survey Telescope Technical Note 19980421
The Sloan Digital Sky Survey (SDSS) 2.5-m telescope is based on a wide-field fast two-mirror optical design, i.e., a 3° diameter field of view at f/5. Such a design is difficult to baffle adequately against stray light sources without excessive central obstruction of the entrance pupil. The adopted solution adds a conical baffle in the converging beam between the primary and secondary mirrors ("2.5-m telescope light baffles" SDSS Telescope Technical Note 19980215). Together with the usual primary and secondary baffles, the design rejects stray light efficiently and blocks only 28.6% of the pupil area on-axis.
A unique challenge is posed by the design of the conical baffle. It is suspended between the primary and secondary baffles. However, its suspension system may only minimally obstruct the light entering the telescope and reflecting from the primary mirror. Its location must be controlled to an accuracy of ±3 mm.
The conical baffle has three translational and two rotational degrees of freedom that can be controlled by adjusting the length of 8 wires using turnbuckles. In addition, the natural frequency of the torsional vibration mode of the baffle about its axis is determined by the tension in the wires. The number of degrees of freedom and adjustments imply that a systematic approach to adjustment is needed.
Prior to adjusting the baffle, the following steps should be taken.
The adjustment process is divided into four steps.
Initially, the tension in the upper wires should be uniform. The translation and rotation adjustments described below should not change the tension in the wires. However, tensions should be checked periodically. Note that when making adjustments, it is easier to turn the turnbuckles if loosening adjustments followed by the tighting adjustments.
Because of the weight of the baffle, the tension in the lower wires will be less that that in the upper wires. During the first three steps, it is easier to turn the turnbuckles if the tension is the wires is low. The natural frequency of the fundamental string vibration mode in the lower wires should be kept below about 60 Hz. (The frequency can estimated by plucking the wires and listening to the tone. Alternatively, a light stroboscope can be used to measure the vibration frequency.)
Each set of four wires along one diagonal of the conical baffle mounting frame acts largely independently of the other set. The exception is if the front set along one diagonal is tightened and the front set of the opposite diagonal is loosened. This causes uneven tension in the wires and distortion of the baffle. Consequently, it should be avoided.
The turnbuckle threads have a pitch of 635 microns (5-40 machine screws). It is convenient to adjust the turnbuckles in quanta of one-sixth turn since the turnbuckles have hex wrench flats machined on their surfaces. Since the turnbuckle is threaded on both ends, the adjustment quanta is 211 microns, more than fine enough for our purposes. The opening half-angle of the wires is 15°.
An axial translation adjustment consists of loosening the four rear turnbuckles one quantum and tightening the four front turnbuckles one quantum. This adjustment translates the baffle 815 microns toward the secondary mirror. Set the spacing between the secondary vertex and front crosshairs to 994.8 mm (39.17 inches).
A rotation adjustment consists of loosening the rear turnbuckle one quantum and tightening the front turnbuckle one quantum in one corner of the frame. In the diagonally opposite corner, these adjustments are reversed. This adjustment rotates the baffle about the opposite diagonal of the frame in the direction of the tightened turnbuckle of each corner pair. The front and rear crosshairs move in opposite directions 449 microns and 596 microns respectively per quantum.
A centering adjustment consists of loosening the turnbuckle pair in one corner of the frame one quantum and tightening the pair in the opposite corner one quantum. This adjustment translates the baffle 204 microns along the diagonal of the frame.
To adjust tension, first verify that the tension in the upper wires is uniform by plucking them and listening to the tone. Then, attach a spring scale to each of the lower anchor levers in turn. Adjust the tension in each of the lower wires so that a spring scale attached to the ball plunger end of the lever reads 35 N with the ball plungers just out of contact with the ball buttons. This step should not translate or rotate the baffle a significant amount as long as the initial tension was approximately correct.
Figure 1: One of eight conical baffle support wire anchors.
Table 1 gives the spacings of the baffles with respect to the optical surfaces for the imaging mode ("2.5-m telescope imaging optical design", SDSST Technical Note 19960410). These numbers are from the drawing "baffle layout97".
Table 1: Spacings of baffle features and optical surfaces. Feature x (mm) Focal surface 0 Common corrector front surface 784 M1 baffle tip 2065 Conical baffle rear tip 2697 Conical baffle front tip 3423 M2 baffle tip 3945 M2 vertex 4406
Feature
x (mm)
Focal surface
0
Common corrector front surface
784
M1 baffle tip
2065
Conical baffle rear tip
2697
Conical baffle front tip
3423
M2 baffle tip
3945
M2 vertex
4406
Table 2 gives the error in position of the critical features of the light baffles on May 23, 1998. They were measured with the telescope pointed at the horizon. Standing behind the focal surface facing the secondary with the telescope pointed at the horizon, +x is right, +y is up and +z is toward the focal surface.
All measurements are within the tolerances that were used to design the baffle system ("2.5-m telescope light baffles", SDSST Technical Note 19960616) except the error in z of the conical baffle location. It was adjusted to the wrong value. The location in the z direction is less critical than in x and y and gravity loading at smaller zenith angles will reduce the error. Consequently, it should have no effect. It will be set to the correct value the next time the baffle is adjusted. Visual inspection did not indicate any discrepancies, e.g., direct views of the sky from the focal surface.
Table 2: One of eight conical baffle support wire anchors. Feature x (mm) y (mm) z (mm) tolerance (mm) M1 baffle tip 1.9 -0.8 * 2 Conical back -0.3 0.5 * 3 Conical front -0.8 0.8 -4.7 3 M2 baffle tip 1.8 -0.6 -0.6 2 * Not directly measured but not likely to be out of tolerance.
y (mm)
z (mm)
tolerance (mm)
1.9
-0.8
*
2
Conical back
-0.3
0.5
3
Conical front
0.8
-4.7
1.8
-0.6
Date created: 4/21/98 Last modified: 2/12/99 Copyright © 1998, 1999, Walter A. Siegmund Walter A. Siegmund