ECEN 4606, UNDERGRADUATE OPTICS LAB Lab 12: Fiber optics SUMMARY: In this lab you will become familiar with procedures for working with fiber optics including cleaving, splicing and coupling. PRELAB: HOMEWORK PROBLEM 1: Pedrotti 3 10-4. Derive the equations for parts a and b starting with the diagram below. The critical angle at the core/cladding boundary occurs when the ray direction in the cladding is parallel to the boundary. An equivalent way to think about this is that the spatial period in the cladding is the smallest it can possibly be (the vacuum wavelength over the cladding index). n=1 NA = sinθ x a z n cl n co n cl HOMEWORK PROBLEM 2: Pedrotti 3 10-11. Basic loss calculations. HOMEWORK PROBLEM 3: Pedrotti 3 10-15. There are multiple sources of dispersion (pulse spreading) the modal dispersion in this example is one of the easiest to understand. Express the results of part b in units of ps (of differential delay) / km (of fiber length). Assuming that the modal dispersion should broaden pulses by at most by 1/10 the pulse width, derive a limit on available information bandwidth (Hz) as a function of fiber length (km). TECHNICAL RESOURCES: TEXTBOOK: Pedrotti 3 chapter 10 LECTURE NOTES: Lecture 13, Fiber optics. EQUIPMENT AVAILABLE: 125 micron cladding diameter, 62.5 micron core diameter multi-mode fiber. 125 micron cladding diameter, 8.2 micron core diameter SMF-28 fiber. This is single mode at 1550 nm and thus will carry a small number of modes at 632 nm. Coating stripper tools of several varieties. Version 1.0, 12/03/09 Robert McLeod 1
A Fujikura high precision fiber cleaver and a Furukawa fiber cutter. A fiber microscope for inspecting cleaves. Fitel S182K fusion splicer. A spatially-filtered, JDS Uniphase 1103P-3020 Helium Neon laser. The laser wavelength is 633 nm. Collimation and focusing lenses. A sheet polarizer. A Newport fiber coupling mount and bare-fiber adaptor. An optical power meter. LAB PROCEDURE: Note: We do not have multiple splicers or fiber microscopes, so you will have to use this equipment in series. However, the steps below do not have to be performed in order, so you can start on later steps while waiting for access to the cleaver. Single-mode coupling is difficult, so you should probably leave that step for last. STEP 1: FIBER STRIPPING AND CLEAVING Optical fiber consists of a high index glass core where the light is guided, surrounded by a lower-index glass cladding. The typical outer diameter for that cladding is 125 microns. To protect the delicate glass, this cladding is typically surrounded by a 245 micron diameter polymer coating. Often, further protection is provided by an additional tight buffer of 900 micron diameter. Fiber optic cables and jumpers build out from this structure. Figure 1. Fiber optic cable structure and terminology. Version 1.0, 12/03/09 Robert McLeod 2
Prepare your work-space. You will be accidentally creating a number of ~1 long, 125 micron diameter slivers of glass. Another name for these is the world s most painful splinter. They particularly like to spring away from your tool and land, point up, in the seats of cloth chairs. This adds the adjective embarrassing to the description. So, work on a well-lit work-bench with no clutter (backpacks, coats, general junk) this will give you the best chance of finding errant shards. Clear all fabric chairs from the area. Before working, obtain several lengths of cellophane tape. The moment you create a fiber fragment, capture it on the sticky side of the tape. That means right on the tool if it is still laying there or on the bench or floor if it gets away. This is a real safety issue find and capture them all. Replace the tape as needed. Dispose of the full tape in the garbage. Step one in this process is to get access to the glass by removing the coating from a section of SMF-28 fiber. The tool for this looks and works much like a wire stripper used to remove the insulator from wire, however it must be much more precise since the coating is quite thin and the glass is quite brittle. Even the smallest scratch of the glass cladding will cause the fiber to break you will do this multiple times before you get the hang of it. Try both kinds of strippers including the no-nick and the metal set that look just like wire strippers. Even though they look similar, don t ever use fiber strippers as wire strippers this immediately destroys the knives. Strip back about an inch if coating you may find that several small sections are easier to remove than one large one. Important: Use lens tissue and methanol to clean all residue of the coating from the exposed glass any remaining coating will be vaporized in the fusion splicer and ruin the electrodes. Unlike copper wire, the glass that makes up the functional portion of the fiber cannot be just cut with a sharp tool it will shatter. Instead, the glass is forced cleave in a flat plane perpendicular to the fiber axis. This process starts with a microscopic fracture or scratch, much that used to cut plate glass. The fiber is then bent, causing it to break. You want about 4 mm of exposed glass fiber beyond the end of the remaining coating. Remember to immediately capture the excess fragment on your tape. Inspect your cleave using the fiber microscope. You should see a perfectly flat cylinder of glass. If you observe scratches or the plane of the cleave is not flat, strip a new section of fiber and try again. Do this until you have two fiber ends that are both flat. Carefully place both fibers into the fusion splicer and lock them in position with the clamps, then start the machine. A video microscope in both the x and y planes will attempt to locate the fiber cores and automatically align them, then bring the two fibers together in z. Finally, an electrical arc is used to melt the glass, effectively making one fiber. The fuse region is extremely delicate and so is usually re-coated or protected in a special splint. The splicer will report its estimate of the coupling loss record this value. A good splice should have under 0.1 db loss. The loss relates to the xy offset by the formula 2 2 ( ) exp x x η x = 4.34 [ db] w0 w0 Version 1.0, 12/03/09 Robert McLeod 3
where w0 is the radius of the fiber mode, about 4 microns. Calculate % loss and the offset for your splice. STEP 2: COUPLING INTO MULTI-MODE FIBER Collimate and focus the HeNe laser at ~0.1 NA. Following the procedures above, obtain 1 to 2 meters of multi-mode fiber from one of the reels and put good cleaves on both ends. Mount the fiber in the chuck on the fiber coupler and place the chuck in the mount at the focus of the HeNe. Lay the fiber on the table making a large, smooth loop and gently tape it down in several locations to keep it motionless and also to help you avoid accidentally snagging it. Place several tape hold-downs near the fiber chuck to serve as strain-relief for later steps. Tape the other end of the fiber to a post top and point it at a white screen. Hint: Fold over the end of each piece of tape as a handle so you can just grab and peel. At the end of the lab, there should be zero tape left on the benches or tables. Couple the HeNe into the fiber by observing the output on the screen. This should not be too difficult due to the large core size. Measure your efficiency with the power meter. Observe the pattern on the screen does it appear to be single mode? Estimate the NA of the fiber. Carefully untape the fiber loop, leaving the tape near the fiber chuck to minimize any displacement of the fiber in the chuck. Form the fiber into a loop and gradually decrease the radius. Observe the pattern on the screen and (in the dark) the fiber in the loop. Explain what is going on. Comment on how your observations would impact a) the minimum size of fibercoupled optical systems and b) the use of different modes of the fiber as independent communication channels. STEP 3: COUPLING INTO SINGLE-MODE FIBER Repeat step 2 using the SMF-28. This fiber is single mode in the 1550 nm telecommunications band but will carry several modes in the red. The core size is 8.2 microns, almost a factor of ten smaller than the multi-mode fiber. This alignment will thus be much more challenging. Place an iris in the collimated beam and observe the reflection of the focus from the glass face of the fiber. If the focus is exactly on the glass and normal to it, the Fresnel reflection will retrace the beam path and return through the aperture. Use this to get three degrees of freedom (z, and the two angles) correct. Then scan in x,y until you detect a gleam from the far end of the fiber Then optimize the coupling, much as you did for lab 1. You may have to remove your fingers from the micrometer at each step to avoid perturbing the mount by a few microns. Once you have become sufficiently frustrated, repeat the steps above. Compare the bend radius and observed light leakage to the multi-mode fiber and comment on what this implies about this fiber relative to the multi-mode one. Does this agree with the NA estimates of the two fibers? Additionally, with the fiber in a minimally-stressed state, use a polarizer to examine the polarization of the output. Then pull the fiber into a small loop and Version 1.0, 12/03/09 Robert McLeod 4
repeat. The bend stretches and compresses the glass, causing a small degree of orientation of the molecules and thus making the glass birefringent. This retardance modifies the polarization. This is how you make wave-plates in fiber. Examine the paddle polarization controller to see how this is implemented. Version 1.0, 12/03/09 Robert McLeod 5