Tagged with: Alignment
ABSTRACT: Bessel beams are useful for alignment because they create a small diameter, bright, straight line image in space perpendicular to the Axicon grating producing a beam that is an exact analog of a single ray in a ray tracing program. Bessel beams are produced by plane Axicon gratings whose pattern is chrome on glass, evenly spaced, concentric circles that are illuminated by a point source of light on the grating axis. The grating produces a more nearly ideal Bessel beam than a cone shaped type Axicon. The plane grating also serves as a plane mirror in an alignment setup to define four degrees of freedom in space rather than the usual two a plane mirror does.
Most discussions of Bessel beams assume illumination with collimated light. We have found it advantageous to use a point source for illumination because it is easy to implement and less expensive using a single mode fiber as a source than a precision collimating lens the diameter of the Axicon. Besides, collimated illumination produces a Bessel beam of finite length in transmission while, in theory, a beam of infinite length is created using a point source.
With these assumptions about how Bessel beams are produced with plane gratings and details about the grating diameter and line spacing it is easy to calculate the useful length of the Bessel beam in reflection from the grating. Other practical matters are also discussed such as 4 degree of freedom lens centering with a test apparatus with no moving parts.
ABSTRACT: Traditionally a rotary table is used for optical centering because the table creates an axis as a reference. Previously, we showed that a Bessel beam also creates an axis useful for centering. The Bessel beam axis and a center of curvature of a surface makes it possible to center an optic simultaneously in tilt and decenter. We also showed that simultaneously sampling two arbitrary points along the Bessel beam also permits full adjustment of tilt and decenter of a powered optic. This makes centering possible without either a rotary table or a precision linear stage. In most common instances, however, sampling the beam at two points is unnecessary because of the inability to correct for both tilt and decenter. We discuss an alternative, simpler method using a Bessel beam.