An Optical Phase Reconstructor Based On Using A Multiplier-Accumulator Approach

Phase sensors that are most commonly used in the adaptive-optics area typically measure the gradient of the phase. A phase reconstructor is necessary to obtain the phase at the actuator positions of the deformable mirror. In the past reconstructors to obtain the optical phase from gradient measurements have been built using resistive nets. These nets simulate a least-squares reconstruction algorithm. There are other algorithms which can be used to mate wavefront sensors and deformable mirrors with different geometries or which can improve the noise performance by using the spatial correlation of the phase. These types of algorithms are difficult to implement and change using analog techniques. In addition, since the movement of an actuator can influence the position of adjacent actuators it is desirable to include this effect in the reconstructor. One may also want to remove the piston and the tip and tilt from the signal applied to the deformable mirror, and determine the values of the focus and tip and tilt terms in order to provide signals to auxiliary mirrors. A digital reconstructor can provide this capability. An approach to a digital reconstructor which can calculate an optical phase which is any linear function of the gradient measurements is described. This reconstructor is based on using a multiplier-accumulator circuit in each channel. A single phase value is calculated in each channel by summing the result of multiplying each gradient measurement by a stored matrix coefficient. Several sets of matrix coefficients are stored in memory to allow one to change the reconstruction algorithm quickly. The circuitry used and the time taken to perform the reconstruction will be described.