Effective Utilization Of Optical Spectrum Analyzers For Enrichment Of Undergraduate Photonics Laboratory Courses

The optical spectrum anlilyzers which we recently added to our photonics laboratory, thanks to an NSF ILI grant, has enabled us to introduce four new experiments into our two undergraduate laborato~ courses: Fiber Optics Laborato~ and Photonics Engineering Laboratory. The new experiments are (1) Spectral Attenuation of Optical Fibers, (2) Optical Wavelength Spectral Analysis of Laser Sources and Light-emitting Diodes, (3) Dynamic Narrowing of Linewidth and Changes in Modal Structure of Laser Diodes in the Vicinity of the Threshold Current, and (4) Spectral Responsivity of PIN Photodiodes. We have also prepared a video to demonstrate the dynamic changes in laser diode spectra as the drive currents are changed. We have effectively utilized the LabVIEW graphical programming environment to implement computer control of the experiments over a GPIB interface. This enhances the speed of data collectio~ and the sophistication of data processing in these experiments. Such computer control of the experiments is very helpfil in the dynamic measurements entailed in the experiments. This paper discusses the new experiments, their enriching effect on the courses, and their stimulating and motivating effect on the students. INTRODUCTION Photonics is a well established component of our electrical engineering program. About 105 students take the two undergraduate laboratory courses offered in our photonics laboratory annually. Our 1992 NSF ILI grant offered us the great opportunity of enriching these laboratory courses by adding one optical spectrum analyzer (OSA) to each of three work stations in our photonics laboratory. Our experience points to the unique importance of the QSA in a photonics laboratory and its role in stimulating learning in electro-optics as the only real-time, graphic display instrument in the optical frequency/wavelength range. Our OSA employs a double-pass monochromator which provides a high dynamic range (-55 dBm at 0.5 nrn from the signal peak) and high sensitivity (better than -85 dBm). It can display the optical spectrum over the wavelength range of 350 nm to 2000 nm. This range includes visible light and the inh.red region relevant to fiber optic communications. However, for calibrated display, the range is 600 nm to 1700 nm. It is capable of sweeping the fill wavelength range in 500 ms, thus saving hours in measurement time, compared to experimental arrangements which do not employ OSAs. The OSA can save data in many ways. In addition to being displayed, the experimental results can be sent directly to a plotter or printer, stored in the OSA memory, or can otherwise be accessed via computer control. .$iiii’) 1996 ASEE Annual Conference Proceedings ‘O,$!p;: . . ‘9 P ge 174.1 A main feature of our laborato~ is the use of the graphical programming environment of LabVIEW. Like other instruments in each workstatio~ the OSAS were cotigured for LabVIEW programming and computer control of experiments over a GPm interface. This greatly incraes the speed, quantity and sophistication of data acquisition and data processing. Four new experiments, breed on the OSAS, have been developed and added to the two undergraduate laboratory coursesA video demonstration of the dyntic changes in laser diode spectra in response to variations in the drive current has also been prepared. The new experiments and the video are briefly described below. Spectral Attenuation of Optical Fibers k this experiment the students measure the total spectral attenuation of optical fibers. The experimental setup is shown in Figure 1. The OSA is switched to the stimulus-response mode, which allows the output light from the monochromator to be passed through the device under test, tier which it is reinserted into the OSA for spectral analysis. The output light from the built-in light source is coupled into the double-pass monochromator with a short length of 62.5/125 mm fiber. The output of the ,monochromator is coupled into the photodetector input with a short (dummy) optical fiber of length LX = 2 w which is identical to the fiber to be tested. A measurement is take% which means that a series of values of optical power is recorded for corresponding wavelengths as a vector, say X. The short length of fiber is replaced with the fiber being tested, of length LY, about 1 km long or more. Another measurement is taken and stored as a vector, ~. Setup for calibrating normalization of the fiber measurement Figure la: Measurement of fiber spectral attenuation. ;#~ J4 }; 1996 ASEE Annual Conference Proceedings ... +.,,, N.,. O., P ge 174.2 Let x(l) be the scalar value of the vector X and y(k) be the scalar value of the vector ~ at a given wavelength k. The spectral attenuation as a finction of wavelength 1 may be displayed in a linear scale or a log (dB) scale. The spectral attenuation is given for a linear scale and a log scale at a given wavelength by equation (1) and equation (2), respectively. a(l) =10 [log, Oy(L) -log, Ox(k)] dB Q Km-’ 1 a(k) = y(k) i x(k) Km-l 2 Because the effective length of the fiber under test used for the computation in equation (1) is LY Lx, this procedure approximates the cut-back technique [1, 2]. The same comectors are used for coupling both the fiber under test and the short length of dummy fiber between the monochromator output and the photodetector input. This arrangement has the advantage of not requiring a cut back which reduces the length of the fiber being tested, each time the test is performed. Figure 2 shows the spectral attenuation obtained with this procedure when a 1 km long 100/140 ~ GRIN fiber was tested. .,.~.., ~{ } 1996 ASEE Annual Conference Proceedings ‘o,+,plly;~ 1.0 ! P ge 174.3