Automating fiber optic assembly through inspection ensures quality cables, components, and connectors in addition to increasing yield and throughput. In fact, advances in fiber optics techniques are increasing the need for better fiber inspection and submicron measurements.
Manufacturers are seeking to decrease fiber inspection times from tens of seconds to milliseconds while eliminating subjective human factors such as different operators, operator fatigue, etc. New automation techniques include using a camera to acquire information from the fiber ends, measure the important parameters, and analyze the results in a computer-based system. Integrating machine vision, motion control, and data acquisition enables engineers to make multiple measurements resulting in greater flexibility and productivity.
Fiber Basics
Optical networks require the highest integrity of fiber and fiber ends for optimal data transmission. The three main components of a fiber are: first, the core; second, the actual conduit for the light, or cladding, a glass material which by doping achieves a certain index of refraction to cause total internal reflection; and third, layers that provide strength and insulation from harsh conditions. Light travels down an optical fiber because of the indices of refraction and angles of reflection. There are two generic types of fibers: multimode fiber (core diameter: 50 µm to a few mm) and single mode fiber (core diameter: 4 to 9 µm). These measurements must be accurate because the diameter and concentricity of the core and cladding might affect such things as the modes of light that can travel in the optical fiber, the numerical aperture of the fiber, or the coupling of light into and out of the fiber.
Automation
There are two areas where automation adds value. One is how the inspection of newly cleaved or manufactured fiber is important because imperfections can cause severe power attenuation when employed in systems. The other involves how the alignment and bonding of small optoelectronic devices together, including fiber optics, is time consuming and requires nanometer precision.
Fiber Inspection
Typical fiber inspection measurements include core-cladding diameter, core-cladding concentricity, and non-circularities. Typical identification measurements include: edge defects, blemishes, dust and scratches. These measurements can be obtained using optical inspection techniques in tandem with image processing and analysis algorithms.
Using a CCD camera and NI-IMAQ image acquisition software and hardware, the optical inspection system can measure the core and cladding diameter using edge detection functions crossing through the core and cladding boundaries of the fiber; functions within National Instruments Vision Development Module. An annulus or circular region of interest tool creates an inner and outer circle around a boundary. Line profiles are automatically placed perpendicular to the cladding boundary between the regions of interest. NI Vision edge-detection functions precisely locate the boundary of the cladding and core. Once you have located the cladding and core boundary, the center of the fiber, circularity, and diameter are all calculated using built-in NI Vision functions. To automate the process completely, engineers can set up a coordinate system to account for any variance in fiber location within the field of view of the camera, using pattern matching or morphology tools to locate the core. This location will serve as a reference point and automatically position and adjust all subsequent regions of interest and measurements.
Scratches, blobs and defects, which may appear on the fiber end face, are detected using various methods. The most popular includes morphological techniques. By performing an image threshold and then manipulating the resultant image, the software can highlight surface abnormalities.
Fiber Alignment
A perfect example of automation in the optoelectronics space is aligning devices with another device such as an optical fiber to a transmitter. Nanometer accuracy is required in order to maximize the amount of light and energy passed from the optical fiber to the component. A machine vision system in conjunction with motion control is typically used to perform a coarse alignment, a general component alignment that has accuracy in the micron range. A clever technique using a mirror can allow one CCD sensor to simultaneously align two axes. Precise alignment in the nanometer range is achieved using data acquisition together with motion control. A common technique uses the scan clock from the data acquisition device to perform a high-speed capture of the position on the motion controller. Both of these signals are available through a RTSI (Real-time System Integration) connection, a technology specially designed for device synchronization. This technique acquires optical-power measurements at a rate determined by the scan clock on the data acquisition device and simultaneously collects position data from the encoders on the motion stage. This process is called time-based position measurements and is much more accurate and faster than using a traditional move, stop, and measure approach.
Dave Barker, director of engineering at G-Systems Inc, says, "National Instruments vision products enable us to consistently provide cost-effective and creative machine vision solutions to our customers. We prototype complex vision algorithms using Vision Builder then quickly transition scripts into LabVIEW source code. The Vision software provides analysis routines for creating sophisticated solutions without reinventing the wheel. Furthermore, the breadth of NI IMAQ hardware offerings helps us specify the best components for each project that we can be confident will seamlessly integrate with our custom software solutions and other hardware components such as motion control. NI vision is our tool of choice for all vision system needs."
Conclusion
By tightly integrating different test and manufacturing functions and processes, optical component manufacturers can build highly automated systems that result in cost savings, shorter lead times, higher yield, and better quality. The main advantage of an automated approach lies in removing time pressure and reliability concerns of labor-intensive procedures. Productivity increases as a result of higher quality. An open-ended system using NI LabVIEW, NI Motion, and NI vision software can easily be augmented with National Instruments image acquisition and motion control devices. In addition, you can also add data acquisition hardware coupled with a light-emitting source to perform beam profile analysis and quantify power dissipation. As new test or inspection requirements arise, you can easily modify the system because you have used a flexible open platform based on off-the-shelf components such as PXI and LabVIEW.