Issue



Optoelectronics Packaging


09/01/2005







Despite the crushing blow dealt to the optoelectronics industry following the dot-com boom, research in the area continues because of the continuing allure of the benefits that technology promises. Fortunately, the world of photonics is not small, so not every facet of the sector was hit badly. While much attention has been given to telecom and datacom applications, photonics actually embraces many technologies, including data storage, image sensors, displays, and lighting, and serves a wider range of products, from cell phones to automobiles.

Still, the extreme bandwidth potential of optoelectronics is undeniably attractive. As a solution for long-haul data transmission, it is the technology of choice; however, there has been considerable reconsideration in determining where the bright line is relative to where optoelectronics should start and stop. That line is made less bright and more diffused, as simpler and more pervasive copper-based interconnection technologies continue to make significant performance strides. There has already been a lab demonstration of OC-768 performance (40 Gbps) in a 2-connector, 30-in. copper backplane channel, meeting a benchmark that was presumed to require an optical solution. But it is clear that photonics in general, specifically optoelectronics, will continue to play an important role, and research should explore new solutions in the anticipation that there will likely one day be a true optical transistor.

That aside, optoelectronics packaging has progressed slowly in the last few years, following the reduced pace of the telecom industry and its latent recovery. Recently, the industry has begun to see signs of a solidifying turnaround. A situation analysis prepared by Prismark for NEMI last year noted that global production of communications equipment driven primarily by Internet use was expected to increase at an average rate of 6.5% per year and reach $226B in 2008. Indications are that research and development continues at a modest and measured pace. With the wounds still sore, it appears that caution and certainty have replaced the unbridled risk-taking of the late 1990s.

With respect to optoelectronics packaging, the challenges remain the same. With reliable mass assembly methods and standards still elusive, cost continues to be a drag on expanded use of the technology. As a result, packaging cost remains high, and there is still much hand-soldering and many competing designs. Moreover, there are vexing, but solvable, problems associated with thermal issues related to the vertical-cavity surface-emitting lasers (VCSELs) used to transmit and receive photons. The lasers must be kept at exactly the same wavelength to communicate properly, and the operation temperatures of these devices must be controlled, lest the wavelength vary. This requires active cooling, which can prove to be both costly and energy-intensive.

As for general package types for optoelectronic applications, they include transmitters, receivers, modulators, VCSELs, and diode arrays. The packages can be non-hermetic or near-hermetic. TO-cans and butterfly packages remain common choices alongside newer BGAs. However, more advanced structures are being proposed, such as optical-RF modules and flip-mounted optoelectronic chip structures having VCSEL arrays with an optical underfill. The latter structures are designed to couple to optical-waveguide layers at a 90° angle. There is also ongoing research in developing thin film lasers and photo detectors, which could cause a change in direction or a new set of requirements.

Economic and technological challenges for optoelectronics are similar to those facing electronics-only packages; however, optoelectronics have their own set of concerns, such as waveguide coupling and fiber alignment. Components often have pigtailed fiber, requiring a connector or fiber splicing, or a connector receptacle or port as part of the package. RF connections also figure prominently because of the bandwidth and high frequency associated with optoelectronics. Because they tend to be hybrid in structure, integration of somewhat disparate technologies creates assembly challenges. Finally, optoelectronics are not immune to matters of reliability and environmental condition concerns.

One can say that photonics and optoelectronics have a bright future, but there will continue to be challenges to meet and overcome for full benefit in a cost-effective manner. Having challenges is important because, paraphrasing the poet Robert Browning, “Our grasp should exceed our reach, what else is research for?”

References

  1. Merritt, R., Board Design Revamped, Electronic Engineering Times, Oct. 13, 2003.
  2. Turbini, L. and Stafford, J., Optoelectronic, The 10th Anniversary NEMI Roadmap Workshop, June, 2004.
  3. Shibata, A., Status of IEC, IPC/JPCA Optoelectronics Standards, JISSO International Council Meeting, Herndon, VA, May, 2005.

Click here to enlarge image

JOE FJELSTAD, founder and principal, may be contacted at SiliconPipe, Inc., 992 DeAnza Blvd., San Jose, CA 95129; 408/973-1744, e-mail: jfjelstad@sipipe.com.