Tutorial on Optomechanical Beam Steering Mechanisms

OPTI 521 Tutorial

Doug Essex

62 N. Worth Ave.

Elgin, IL 60123

(847) 697-3286

Introduction

Crucial to any optical system is the ability to adjust the beam pointing, whether incoming or outgoing. Beams need to be moved in position and angle, and adjustments are easier if movements for each degree of freedom of the beam are kept separate. In this tutorial some examples of common optomechanical beam steering elements will be shown and their advantages and disadvantages compared. Additionally a new type of adjustable prism will be described and compared. Spatial light modulators, acoustooptical and electrooptical deflectors and other diffraction based devices will not be considered in detail, because with the exception of micro opto electro mechanical (MOEM) chips they tend to be less optomechanically intensive.

The application of these devices can be roughly broken down into two main areas: beam or alignment within a system and tracking or pointing at the target. Pointing at a target is conceptually very simple – all that is required is a 2 axis mirror to adjust tip and tilt angles until the target is illuminated (for an active system) or brought into the field of view (for a passive system). The complications tend to involve the subsidiary systems required to keep the optics pointed at the target. These pointing systems are generally electrically driven, and have feedback systems to assess target pointing, so bandwidths and slew rates and the existence of good control laws or models are important. System alignment requires moving the beam so that it is on or closely parallels to a required optical path, rather than simply hitting a target. Achieving this requires mechanisms that offset the beam position in addition to angle adjusting mechanisms. These mechanisms tend to be manually driven or controlled with loosely coupled or ad hoc feedback. The adjustments tend to be made once and then locked down and staked.

The functional requirements[1] for a beam steering device are

Maximum steering angle or offset

Control/resolution of steering

Beam divergence or imaging

Aperture/vignetting

Spectral range

Throughput

We will provide tables with some semi-quantitative comparisons of these parameters at the end of this paper.

Mechanisms to adjust beam position

Tilt Blocks

These are simple plane parallel plates of transparent material that offset the beam position through refraction. The offset (for small ) is given by

y = t (n-1)/n

Figure 1 Schematic of a tilt block

In order to get large displacements the thickness and/or the refractive index must be as high as possible and the tilt must be as high as practical also. When dealing with polarized light if the tilt is out of the plane of incidence then the surfaces will have to have high transmission for both polarization states. This is difficult to achieve for angles ~>45 degrees. High angles also either require use of oversize pieces or acceptance of significant foreshortening of the clear aperture in the tilt axis direction. As with all refractive systems dispersion limits the useable spectral bandwidth.

Periscopes

These elements are excellent for providing comparatively large offsets. As they are reflective they can be very achromatic. However this also means that they are prone to the common problem of mirrors: 2x magnification of angle errors. Using mirror matrices it can be shown that in order for the output beam to still be parallel to the input the second mirror must match any angle error in the first mirror.

If we start with the matrix for a mirror with its normal in the z direction and use the rotation matrices to tilt it to an angle of /4+ where  is an angular error we have


The second mirror of the periscope starts with the same matrix for a mirror with the normal in the z direction, but now we rotate 5/4+




Now multiplying the matrices for mirror 1 and mirror 2 we get

Which results in


Showing that the output beam is travelling in the same direction as the input beam. A popular method of assuring that the two mirrors are parallel is to make the periscope a monolithic rhomboid prism. The parallelism of the mirrors can be checked interferometrically to a fraction of an arcsecond and is fixed in the glass of the finished prism. If the offset of the periscope is made adjustable then straightness of the linear motion must be very good to ensure that no angle is introduced between the two mirrors.

Mechanisms to adjust beam angle

Fast steering mirrors[2],[3],[4]

These are typically the final output mirrors in active (beam emitting) systems. They are typically lightweight mirrors steered by electromechanical actuators on the back of the mirror. The choice of actuation mechanism is driven by the required frequency response of the mirror – high frequency, high load with small tilts or translations indicate using a piezoelectric driver, while medium frequency, low load or large tilts or translations indicate use of a voice coil. While angular errors still have a 2x effect on the output beam, since these systems generally have a sensor and feedback system the beam can generally be held on the target as long as the frequency response and actuator resolution limits are not reached.

Risley prisms[5],[6],[7]

Low angle wedges of transparent material deviate a beam by an angle  = (n-1) where  is the prism apex angle. One prism can only move the beam in a circle. Addition of a second prism moves the beam in a superposition of circles.

Figure 2 Risley prisms

Generally restricted to low wedge angles to keep the angle of incidence relatively low and therefore transmission relatively constant for polarized beams and to avoid dispersion therefore they have a low field of regard. The beam path follows a Lissajou type pattern dependant on wedge angle and rotation rate. If the wedges are not matched exactly there can be dead zones that cannot be pointed to. It is difficult to formulate control laws to position wedges to give predictable result – generally used manually to set angle and then fixed in place or in constant motion to scan.

Metastable adjustable prism[8],[9]

Described on and in U.S. patent 6320705 this new type of beam adjuster uses a doublet with a lubricating fluid filled interface between matching spherical surfaces.

Figure 3 Page from the patent describing the Metastable beam steering mechanism

One of the elements is decentered thereby creating an effective block with a wedge angle tan  = T/R where T is the amount of translation of one of the elements and R is the radius of curvature. The diameter of the clear aperture is given by DCA = D-2T. Because of this reduction of clear aperture as the device is adjusted it is preferable to use an oversize lens for the moving element Dmoving lens = Dnominal + 2T to make up for the 2T reduction when the mechanism is adjusted. The decentering can be performed on either axis independently, making easy orthogonal adjustments of beam angle possible. The entire assembly can be tilted much like a tilt block to translate the beam, thus combining 2 functions into one optical element. The surface tension of the lubricating film holds the lenses together, eliminating any need for guides to control the moving element.

This mechanism can also be adapted to make an extremely finely adjustable mirror mount. In conventional mounts the actuator translation distance divided by the lever arm between the pivot point and the actuator determines the increment of adjustment. Since mirror mounts are seldom much bigger than the mirror itself this lever arm distance is limited and consequently if fine adjustments are required then more precise actuators are required. If the flat surface of the moving lens of the Metastable mechanism is made reflective the change in angle of this surface is T (the actuator translation)/R (radius of the curved surface). This radius can be extremely accurately made up to a few meters long, potentially giving a fine adjust mirror 2 centimeters in diameter with a coarse threaded actuator an angular adjustment resolution 50x better than a conventional mirror mount.

Figure 4 Page from the Metastable patent describing the use of the mechanism for a mirror mount

Disadvantages of this mechanism include the need for a lubricating index matching fluid. This restricts the use to the visible and NIR with glasses, where index matching fluids exist. No fluids that match ZnSe (n=2.4) or Ge (n=4) are known, meaning the technique is not applicable in the mid to far IR. Decentered lenses systems have been described for use in the infrared but they rely on mechanical methods to keep the surfaces separated and moving in the correct orientation[10]. A common index matching fluid used in the UV, 3M Fluorinert FC-40 has too low a viscosity to lubricate well. The UV transparency of other fluids would have to be investigated before committing to using this mechanism in the UV. Laser damage threshold of the fluid may restrict the range of possible applications. The patent uses microscope immersion oil for best results, but used a variety of different lubricants, including 5W-30 motor oil!

Summary

Beam offset mechanisms / Tilt Block / Solid Periscope / Separate mirror Periscope
Max Offset / t(n-1)/n / As large as practical length of material allows / As large as a straight path can be made
Control/Resolution / Actuator controls this/orthogonal / N/A / Fixed or governed by straightness of path. Mirror angles must be controlled.
Beam div/imaging / Image shifted/aberration / Increased path length / Increased path length
Aperture/Vignetting / Dcos / /
Spectral Width / Limited by dispersion/ AR coatings/material transparency / Limited by AR coatings/ material transparency / Limited by reflector coatings – for metals very wide.
Throughput / ~99.5%, possibly less if high angle AR’s required / ~99.5% (AR’s + TIR bounces) / R2 ~ 98%

Table 1 Comparision of beam offset mechanisms

Beam angle adjustment mechanisms / FSM / Risley / Metastable
Max deviation / Many degrees / Typically minutes to a few degrees / T(n-1)/R
Control/Resolution / Orthogonal/actuator controlled resolution / Not orthogonal/resolution controlled by actuator / Orthogonal control/resolution controlled by actuator
Beam div/imaging / Limited by mirror flatness / Limited by transmitted wavefront / Limited by transmitted wavefront
Aperture/Vignetting / ~ / D / D-2T
Spectral Width / Limited by reflector coatings – metallic is very wide / Limited by dispersion/ AR coatings material transparency / Limited by material disperision/AR coatings/material (lube fluid?) transparency
Throughput / ~99% / ~96% / ~98%

Table 2 Comparision of beam steering mechanisms

References

1

[1] Tholl, “Novel Laser Beam Steering Techniques” SPIE 6397 Technologies for Optical Countermeasures III (2006)

[2] Sweeney, et. al. “Design Considerations for Fast Steering Mirrors (FSM’s)”, SPIE Vol. 4773 Optical Scanning 2002 (2002).

[3] Berta, et. al. “Development of a Commercial Line of High Performance Fast Steering Mirrors” SPIE Vol. 3787 Optical Scanning: Design and Application (1999)

[4] Zook, “Light Beam Deflector Performance: A Comparative Analysis”, Applied Optics 13 (4) April 1974 p. 875

[5] Marshall, “Risley Prism Scan Patterns”, SPIE Vol. 3787 Optical Scanning: Design and Application (1999)

[6] Ostaszewski et. al. , “Risley Prism Beam Pointer”, SPIE Vol. 6304 Free-Space Laser Communications VI (2006)

[7] Sanchez, et. al. ,“Control Laws for a Three Element Risley Prism Optical Beam Pointer”, SPIE Vol. 6304 Free-Space Laser Communications VI (2006)

[8] U.S. Patent 6320705

[9]

[10] Gibson, et. al. “Wide Angle Decentered Lens Beam Steering for Infrared Countermeasures Applications”, Optical Engineering 43 (10) Oct. 2004 p. 2312.