Abstract

A component vital to the successful implementation of single-mode fiber signal processing and communications systems is an integrated in line Bragg cell that can function as a modulator, SSB frequency shifter, and tunable WDM tap. A number of reported devices^1,2 go toward achieving this goal, the most efficient to date being the flexural-wave modulator in a dual-mode single-core (DMSC) fiber developed by Kim et al.² In their device the LP₁₁ and LP₀₁ normal modes (NMs) of the DMSC structure are coupled together by an acoustic flexural wave whose wavelength matches the intermodal beat period. Launching light into the even mode results in a frequency-shifted signal in the odd mode. In practice, however, it is difficult to excite just one NM, and a frequency-shifted signal in the LP₁₁ mode is not appropriate for efficient pigtailing to a single-mode fiber. The dual-core (DC) device we have developed avoids these problems by permitting low-loss fusion splicing of a single-mode fiber to either core and having negligible intrinsic coupling. This is achieved by designing the DC to be phase velocity mismatched to such a degree that the NMs of the DC coupler are almost identical to the modes of each core in isolation. Provided that there is sufficient overlap of the modal fields, distributed feed-forward NM (and hence intercore) coupling will be achieved³ in the presence of a flexural wave at the correct frequency. We fabricated two different DC fibers, DC1 and DC2, with experimentally determined beat lengths of 0.43 mm and 1.205 mm, respectively at λ = 1μm. Flexural waves were excited by using fused silica horns bonded to the DC fibers with glass solder. The intrinsic coupling in DC1 was too small to measure with our setup, and 8.5% acousto-optical coupling was obtained at 4.6 MHz and 962.5 nm. The instrinsic coupling of the DCS was approximately 4%, and 100% (and beyond) acousto-optical coupling was achieved at 560 kHz and 1.064 μm. The length of DC fiber in each case was 30 cm. Both devices were widely tunable. The results of a heterodyne experiment with DCS, illustrated in Fig. 1, are presented in Fig. 2. The high intrinsic coupling of DC2 meant that the carrier was not fully suppressed, and the 9 cm of fiber on the opposite side of the horn meant that sideband suppression was not perfect. We believe that with additional development the modulator has considerable potential as a practice device.

© 1990 Optical Society of America