Optical Component Assembly: Is History Repeating Itself?

	Steve Adamson, Asymtek, Carlsbad, Calif.

Parallels exist between electronics packaging assembly and the new and growing field of optical component assembly. Those who recognize the similarities and can take advantage of the lessons already learned will benefit in the long run.

The IPC's roadmap forecasts optics to have a 65 percent growth rate over the next few years. To get to these high growth rates, it will be necessary to lower the cost of optical components. Each year we will have to double the function of the components and halve the price of them or better, similar to Moore's Law in the field of semiconductors.

In the assembly industry today, in particular dispensing, we see optics companies using manual, hand-held dispensing tools. We know these methods are slow and will not lead to rapid reduction in component costs. Some venture capitalist and industrial engineers compare optical component assembly to electronic component assembly 20 to 30 years ago.

For example, to dispense precise, small shots of adhesive and do it thousands of times a day, accurate pumps are required. In the 1970s the electronics industry was using air-pressurized dispensers. It was found that the volume of fluid from these dispensers was inconsistent over a shift. Operators were continually changing pressure or time because the walls of a syringe can flex and, as the syringe is emptied, more wall is exposed. This leads to a greater amount of flexing and a change in shot size.

As a result, in the 1980s, most dispensing systems used auger pumps with a screw-dispensing action to overcome the changes due to the syringe wall flexing. Other factors were still a problem, since the viscosity of some fluids fluctuated over time, which changed the volume of fluid that was dispensed. Mass flow control was introduced to monitor changes in fluid viscosity, and adjust the dispensing parameters to provide consistently dispensed volumes.

Then in the 1990s, positive displacement dispensing pumps were developed. These pumps were independent of viscosity and capable of dispensing accuracies of better than one percent with small (a few milligrams) shot sizes.

Assembly operations in the early 1980s used two-component materials. It was soon realized that a better quality product could be achieved by having adhesive formulators mix, package and freeze two-component materials. This practice eliminated the line operators from having to mix fluids, measure resin-to-catalyst ratios, and make sure that the filler in the adhesive did not fall to the bottom of its container. It also was found that when operators mixed ad-hesives, air was blended into the material, causing problems later on with "blow holes" in the encapsulation. Premixed, two-part materials can be put into automated machinery on high-speed assembly lines. Two-part mixing machines are available, but these have not found favor for applications of precise, small-shot, high-throughput dispensing.

Another lesson learned in dispensing is that a fluid's pot life (the time the fluid remains dispensable before it cures or doubles in viscosity) is very important. Short pot-life materials are not practical in high-accuracy pumps. The accuracy requirement dictates a pump with close tolerances, and this is reflected in the price of the pump. Line engineers have learned that it is unwise to use a short-pot-life material in expensive, precision pumps.

Curing the adhesive inside the dispenser is another innovation that has been developed for advanced packaging assembly lines. This technique could be applied to optical assembly since several optical component assembly companies use UV-activated adhesives. Curing the adhesive immediately following dispensing is possible with dual-action dispense heads, where one of the heads is equipped with a fiber-optic light to deliver the UV light.

Today's optical component assembly engineers have the task of bringing optics from the lab and into high-volume manufacturing, reducing production costs and enlarging the market. By leveraging the lessons learned in the last 30 years of electronics assembly, the smart ones will be able to reduce their learning cycle, avoid repeating mistakes of the past and bring low-cost components to market faster.