During my time at IPG I had the pleasure to work on the testing and development of the Fiber-to-Fiber Optic Switch and Coupling units. These devices are popular because they allow one single Laser unit to serve several work-cells through computer controlled mirrors and lenses. They also provide stray energy monitoring, ease of maintenance and easy work-cell reconfiguration. These devices were meant to compete directly with those developed by Optoskand and TeraDiode. I was involved mostly in the assembly and testing of the devices. The result of these activities provided valuable insight on how to combine Quartz, metal and high energy light to create a reliable Laser delivery system. You can read more about the workings of the switch in its Google patent page and IPG's website.

Feature goals:

• Operation Wavelength: 1060 nm • Time switching or Energy sharing • Fast Fiber Break Detection • Control and Safety Electronics • Air-cooled • Air-tight beam chamber • Number of Outputs Available: 2, 4, and 6 • Optical Material: Quartz • Optical Power Handling: 2 kW.

The Input and Output focus of the device was meant to be set at the lab. so it would be ready to use upon delivery. The mechanism consisted of two (2) 90° spring-loaded, wedge-joint, screw-driven rings which provided adjustment in the X,Y plane. These parts are marked by numbers 5, 6, 9, 10, 11, 14, and 17 in figure 3. The Z-axis adjustment consisted of a spring-loaded, pivot-joint, screw-driven lever that pushed the Collimating lens housing in and out along the focus axis. Part numbers 2, 3, 4, 7, 8, 12, 15 and 18 (Fig. 3) comprise most of the Z-axis focus mechanism. The X,Y displacement was 4 mm in each axis, the Z-axis had a 6 mm displacement.

The bayonet interlock mechanism, shown in figure 4, secures the fiber laser bayonet to the focusing mechanism. Three steel pins are rotated until the bayonet is lodged into a funnel-shaped seat that effectively locks it and prevent movement in all degrees of freedom (Fig. 4 parts 1-5).

Taking some elements from other optical machines I developed a slightly different housing assembly that minimize shifting by incorporating yet another spring element. Figure 5 shows the lens housing sitting on two pins that acted as rails along the Z-axis and a spring-loaded plunger. All three support points allow the housing to move in and out of the page by sliding along the low-friction line (Z-axis). The previous set-up relied on the lens housing (LH) sitting in a clearance fit within the X-axis ring. This clearance fit lead to the uncertainty of final position of the LH after shaking. The spring plunger in figure 5 guarantee the lens will return to the same position it was before the shaking test.

The switches and couplers are assembled in a class 100 clean room, where all employees are required to wear shoe covers, lab. coat, hair bonnets, gloves and even face masks. All parts coming into the lab. where thoroughly clean in an alcohol bath solution while subjected to ultrasonic vibrations. Cleanliness was very important because foreign substances inside the laser chamber (Fig. 1: part 14) could quickly evaporate or out-gas upon reaching a high temperature, these dirt or gas particles would then accumulate on the optics, blocking the light and thus reducing beam quality.

Keeping the beam chamber clean during assembly was only the first challenge in that regard. The next one was just as difficult, keeping outside particles from being drawn into the chamber. During normal operation hot, expanding air would keep the chamber at a higher pressure than the outside, however, upon switching off the device the cooling, contacting air would produce a vacuum that was certain to bring in particles from the factory floor; eventually reducing beam quality over several cycles. The way we tried to address this scenario was to seal off the assembly as best we could with features such as o-rings, Sealtight® fasteners, tight-tolerance flat matting surfaces, and hermetic electrical feedthroughs. The chamber was designed to hold a 2 atmospheres of pressure when the beam was off. This could be done by injecting nitrogen, argon or other inert gas through a valve. O-rings were place on every module that plugged directly into the switch block housing (Fig. 2: parts 12, 24, 34, and 36). Even the laser bayonet featured one O-ring in the stem.

In order for the switch to provide the Energy Sharing feature, the motors had to be capable of moving the mirrors very quickly in and out of the laser path (Fig. 2: parts 20, 34). This affected the longevity of its parts, particularly the ball bearings. We found that bearing pre-load was a major factor affecting motor life, so we carefully recorded pre-load settings and tested the motor by fast switching through thousands of cycles in order to determine optimum factory settings. All measurements and data recording was done through LabVIEW virtual instruments and Data Acquisition hardware PXI.

To simulate shipping conditions, the switching device was tested for vibration endurance. The switch was intended to be ready to use upon arrival to the customer's factory floor. The focus settings were supposed to remain unchanged throughout the trip. The test consisted in bolting the device to a vibrating bed table. The table vibrated from 0 to 29 Hz for one hour. Most of the switches did not maintain the focus setting because the lens housing shifted through the tolerance clearance stack and spring displacement.

I value the skills obtained during my time developing fiber laser devices; these include Design for manufacturing/assembly (DFMA), Tolerance Stacks analysis, Adhesive Bonding Assembly, and Geometric Dimensioning & Tolerances (ASME Y14.5).