George Davis, SensorPhysics, Inc.,
8425 S Timberline Road, Fort Collins, CO 80525
970-593-0383, fax 970-593-6999    www.sensorphysics.com

Published In Lasers & Optronics™ July 1998 page 40

Introduction

The measurement of the spatial profile of light beams of fiber optic components requires two special approaches. First at 1300 and 1550 nm wavelengths standard CCD cameras do not have sufficient sensitivity. This can be addressed by employing infrared sensitive cameras. In this report we use the Electrophysics 7290A Micro viewer camera that is sensitive from 700 to 1800 nm.

The second situation involves the magnification of the beam, since the output from a laser diode, LED, waveguide or fiber optic is typically on the order of a few tens of microns to a micron in size. We have worked with Olympus to specify long working distance objectives that make it practical to measure such fiber optical components.

Results

Figure 1
The test equipment employed is shown in Figure 1. The Olypmus "infinity corrected" long working distance M Plan Fluorite 50X and 100X objectives had working distances of 10.6 and 3.3 mm, respectively. Long working distances are essential to allow space to position the fiber of laser device under the microscope easily. The output of the MicronViewer camera is connected to the LaserTest beam profiler. The LaserTest is calibrated in microns by placing a grid of known dimensions under the microscope. A mounting fixture for the optical fibers was positioned under the objectives. A similar fixture can be fabricated for wafer level device tests, VCSEL devices, waveguides and laser diodes. The wavelengths were measured with the SensorPhysics EPP-2000 Optical Spectrum Analyzer.

Figure 2 shows the results of viewing a 62.5 um multimode fiber with a 50 X objective. The Fabry Perot laser diode (Mitsubishi ML 701B8R-01A packaged with a single mode ST connector by Oz Optics, Carp, Ontario, Canada) had a wavelength of 1304 nm. We also connected the same laser package to a 9 um single mode fiber. The resulting near Gaussian output is shown in Figure 3. These near field images of the fiber indicate the multimode near field image is considerably less uniform than that of the single mode fiber for the same laser diode source. The evaluation of multimode mode profiles (especially directly from the fiber port of a transmitter) is one of the design parameters that this method can be used to evaluate.

Figure 4 shows similar results for a single longitudinal mode DFB laser diode (Mitsubishi FU-627SDF) operating at 1552.8 nm connected to a 9 um single mode fiber. As in the single mode 1304 nm fiber case the output is nearly Gaussian.

With the Micron Viewer camera the minimum detectable signal is about 1 uW/cm2 at 1300 nm and 3 uW/cm2 at 1550 nm. In the case of too much output power from the device to saturate the camera, the Micron Viewer has a built-in filter slide that can accommodate 22mm diameter neutral density filter up to 2 mm in thickness. We have made similar beam measurements to 2.2 um with the Micron Viewer and to 12 um with the Electrophysics PV-320 camera.

Although the spatial resolution of the Micron Viewer camera is not as high as newer solid state cameras, it is clear from these tests that combined with the Olympus microscope system resolution of better than 1 um is easily achieved. This makes the overall system suitable for near field measurements from laser diode, VCSEL and waveguide devices. We have employed a similar approach to measure the far field pattern of such devices for laser safety class measurements as well.

Figure 2. Output at the end of a 62.4 um multimode fiber connected to a 1304 nm Fabry Perot laser diode. Viewed with 50 X objective.

Figure 4. Output from the end of a 9 um single mode fiber connected to a 1552.8 nm DBF laser diode. Viewed with 100X objective.