A design concept has been proposed for an optomechanical apparatus that would implement a variable optical delay line with a fixed angle between its input and output light beams. The apparatus would satisfy requirements that emphasize performance in interferometric applications: to contain a minimum number of optical surfaces, each used at low angle-of-incidence, and to be nominally free of shear (transverse motion of the beam) on any optical element. As an additional advantage, the apparatus would afford partial compensation of vibration disturbances associated with adjustment of the optical delay by both reducing the amount of motion required to achieve a desired optical delay and by splitting the total motion between two assemblies. As compared to prior art implementations of delay lines, the only disadvantage of the concept is that the motions of the optical elements must be well coordinated through mechanical linkages or electronic controls.
The figure depicts a typical configuration of the apparatus. The optical elements would be two flat mirrors - M1 and M2 - mounted on linear actuators. The actuation axes of M1 and M2 would be parallel to the incoming and outgoing light beams, respectively. M1 would be mounted on its actuator at a fixed angle required to aim the beam reflected from it to the center of M2. In turn, M2 would be mounted on its actuator at a fixed angle required to aim the outgoing beam in the desired direction. Moreover, the angles of M1 and M2 would be chosen so that the angle between M1 and the incoming beam equals the angle between M2 and the outgoing beam.
All of the properties of this apparatus that make it preferable to prior variable optical delay lines depend on making M1 and M2 move by equal and opposite amounts to vary the length of the optical path: In shortening (or lengthening) the optical path, one must move M1 a required distance along the input beam path toward (or away from) M2 while moving M2 along the same distance along the output-beam path toward (or away from) M1. It is noted that the path length change introduced by the linear motion of each mirror is greater than just the distance actually traversed by the mirror. In most configurations, the path length change effected by the delay line is more than 3 times the actual distance moved by either mirror.
As a result of this geometric arrangement and coordination of motions, the incoming beam would always strike M1 at the same point, the beam reflected from M1 would always strike M2 at the same point, and the outgoing beam would always strike the next optical element in the output path at the same point, giving zero beam shear at all times. Assuming that the mirrors and their associated mounts would have equal masses, the vector component of the motions of the mirrors along the line joining the centers of the mirrors would introduce no net momentum disturbance, and thereby no significant vibrational perturbations into the surrounding structure. There would remain a small, uncompensated vector component of momentum disturbance along the direction perpendicular to the line between the centers of the mirrors; optionally, one could compensate for this component of momentum disturbance by use of a relatively small auxiliary moving mass.
This work was done by Jeffrey Oseas of Caltech for NASA's Jet Propulsion Laboratory.