TOKYO—Heads for reading data from optical disks have evolved from assemblies that look like miniature optical benches—in which separate lenses, beamsplitters, mirrors, and detectors are mounted to a common plane—to simple devices containing tightly integrated, multipurpose refractive or diffractive optical elements. Because heads for magneto-optic (MO) disks contain many polarizing optical components, they are more complex and more difficult to integrate. Researchers at Sony Corp. have been developing a MO head that is based on a design for a commercially available optical head but uses a birefringent prism to provide the MO signals. Their theoretical model of the prism has allowed them to pick the orientation of the crystal that provides the best performance. Based on their calculations, they have built an experimental version of their device that performs well.
The signal from a MO disk must be separated into two orthogonally polarized components; the MO signal is the difference of the two light intensities. In the commercially available optical head, a specially shaped prism allows the incoming data beam to strike two photodetector arrays in succession, providing information on focus error of the optical head. In the MO version, the prism is made of lithium niobate, a uniaxial birefringent crystal that separates the two polarizations enough that a dual-element detector can receive them separately.
In a birefringent material, the ordinary ray obeys Snell's law and the extraordinary ray does not. Therefore, a beam containing both polarizations passing into the material is angularly split in two. In addition, at each reflective surface within the material, the beam splits in two again, potentially resulting in eight beams striking the detectors.
This behavior made the Sony researchers' theoretical model more difficult. In their model, a geometrical spot diagram calculates the position of all eight spots for prisms with differing crystal axes. A diffraction model then predicts the intensities of all the spots. An additional constraint is the requirement that the spot width at the first reflection equal the spot width at the detectors, in order that a focus-tracking detector arrangement would work properly. From the results of the model, the researchers were able to choose the optimum direction for the crystal axis.
An experimental MO head was built around a 780-nm diode laser and an objective lens with a numerical aperture of 0.45. The device read a modulation signal from a MO disk that had a track pitch of 1.6 µm and a prerecorded signal with 0.29-µm channel bits. The result was an output of good quality and depth of modulation.