ECE/PH 483/583 Laboratory Exercise

Cleaving Fibers and Measuring Numerical Aperture (NA)

Cleaving Fibers:

An optical fiber consists of a core with refractive index nco with a circularly-symmetric cross section of radius a (core diameter = 2a) and a cladding with refractive index ncl (with ncl < nco ) which surrounds the core and has an outer diameter d (d = 125 mm for most telecom glass fibers). A protective layer of acrylic (or polyimide) called the buffer layer covers the bare silica fiber to protect it from surface abrasions and water vapor. In most fibers telecom glass fibers this layer has a diameter of 250 mm.

Before measuring the diameters of the core and cladding of a fiber, the fiber must be cleaved. This is done by applying a small force on the cladding with a very sharp carbide or diamond blade which is used to cause a small local strain in the cladding. Evenly applied stress, applied by pulling and bending the fiber, causes a small crack to propagate from this point through the fiber and cleave it across a flat cross section of the fiber perpendicular to the fiber axis. A good cleave leaves a smooth, flat fiber end with no cracks across the core or any protruding glass edges. “Bad” cleaves have cracks that propagate into the core or glass “cusps” on the end edge of the fiber.

Before cleaving the fiber, the buffer layer coating must be removed so that the blade can cleanly contact the glass cladding surface. An acrylic buffer layer can be removed either mechanically using a “Miller tool” (or other mechanical stripper) removing only a few millimeters of buffer at a time or chemically by soaking the fiber end in a solvent such as dichloromethane which will dissolve the acrylic coating. [A polyimide buffer layer (not very common) requires either a hot acid solvent [ugh!] or heat (flame or laser) to remove this tough material.]

Cleaving and observing core/cladding of various fibers:

1. Use short fiber pieces in the round plastic boxes in lab to practice stripping and cleaving fiber. Label a box for your lab group. Use it to store the fibers for your group during the term.

2. Strip the acrylic buffer layer either mechanically or chemically. [Soak both ends of the fiber in the dichloromethane stripper for about 1 minute until the buffer layer starts to curl up on the fiber. Then remove the fiber and wipe the ends with tissue removing the softened buffer coating.] Mechanically using a Miller tool is simpler for 125 micron diameter fiber. [Since the Miller tool is precision ground to 126 microns, it doesn’t work well on 100/140 micron fiber. This fiber requires chemical stripping or use of another tool.]

3. Cleave both fiber ends by carefully using one of the many different cleavers in the lab (EG&G, Sumitomo, Corning, or Newport). The $800 Sumitomo cleaver works best on 125 micron fibers. Be gentle with it. It may look simple, but it is a precision instrument!

4. Look at the photomicrographs of the cleaved ends of three different fibers which were taken earlier using a measuring microscope camera. Initially, the far ends of the fibers were not illuminated, so the cores are not easily visible. Observe the smooth, flat, cleaved ends of the fibers. [If there is dust on the fiber ends due to static electricity, carefully and gently hold a small piece of Scotch tape out and touch the adhesive to the fiber end using only the tape itself without your finger behind it.] To see the cores where the light travels by total internal reflection, we shined light into the far ends of all three fibers by holding the ends up pointing toward a small table lamp bulb. Sketch in your notebooks the core/cladding views you see along with dimensions. Which fiber would you think would be easiest to couple light into? Hardest? Why?

Numerical Aperture:

For step index (SI) fibers the numerical aperture in air is defined as, where a is the maximum acceptance angle of the fiber measured from the fiber axis. The NA is a measure of the light-collecting ability of the fiber. It can be measured either at the input end or the exit end of the fiber. You will measure both ways and compare your results. When the step-index (SI) fiber is illuminated in air, the NA is equal to sina. This observation allows NA to be measured in the following manner. One end of the fiber under test is mounted on a rotating stage directly over the axis of rotation and is illuminated by a He-Ne laser, the other fiber end is connected to a power meter. By rotating the stage, the coupled power versus the incidence angle can be measured and the maximum acceptance angle can be obtained when the power drops to a predetermined amount (e.g. 10 dB below the maximum coupled power).


Cleaving/NA Page 2

For GRIN fibers, NA is not constant across the core, thus the above method provides an estimate of the average (over the core section) NA.

A quick approximation of the NA can also be obtained at the exit end of a fiber by placing a piece of paper a distance, L, away from the end of the fiber in a darkened room. Couple a HeNe laser beam into the opposite end of the ~2 m length of fiber. [Focusing the laser into the fiber will give a brighter pattern but should not change the measured spot size.] Then measure the diameter d of the spot on the paper and the distance, L, from the fiber to the paper. The NA of the fiber is approximately NA ~ d/2L for small angles. What is the correct value of NA in terms of d and L? Use this method to determine the NA of your fiber and compare it with the NA measured above at the input end. How much do the “correct” value and the approximation vary?

PRIMARY EQUIPMENT LIST

HeNe laser

Power meter

1 piece of 100/140 fiber (~2 m long) [100/140 means 100 μm core diameter and

140 μm cladding diameter] [Actually, GRIN fiber and not SI, but measuring “NA” still works.]

1 piece of 50/125 fiber (~2 m long) [Ours is also GRIN fiber, I believe.]

Alignment:

The alignment for this experiment is fairly critical. The reason is that the pedestal on which the fiber mounts rotates. Unless the fiber end is positioned precisely at the center of rotation (COR) and also at the central peak of the HeNe laser beam, the measurement will have significant error. In this experiment, the fiber must be aligned in five directions: yaw, pitch, x, y, and z (assuming z is along the fiber axis). The following alignment procedure is recommended.

1. Consider the table top as your reference plane.

2. Make certain the laser beam is parallel to the table top. Do this by shining the laser on a meter stick and measuring the beam's height above the table near the laser and far from the laser.

3. Align the fiber end at the center of rotation of the pedestal. Do this by shining the laser directly over the pointed allen screw marking the COR (the axis of rotation is perpendicular to the table top). When the fiber is properly aligned, the laser beam will remain exactly on the fiber end regardless of the rotation.

4. Align the translational (x-y) axes of the fiber holding stand. Do this by illuminating the fiber with the laser and turning the knobs, which control the translational position until the measured power peaks.

5. Check the alignment by rotating the stage and seeing that the laser spot does not move on the fiber end. Then change the other axes slightly and observe a drop in power level.

PROCEDURE

1.   Referring to Figure 1, illuminate the 50/125 test fiber with the HeNe laser.

Lab #1 Page 3

2. Rotate the input end of the test fiber until the power peaks. Measure and record this reference power.

3. Rotate the test fiber end clockwise in steps of 2 degrees, measure and record the power at each step. Repeat until the power drops to 3% of the reference. [Note: The exact power level is somewhat “fuzzy.” Some say 5%, some 3%, some 10 dB.] The sine of this angle is the NA.

4. Rotate the stage back to the starting position and repeat step 3, but this time the stage is rotated counter-clockwise.

5. Use a lens to couple more HeNe laser power into the fiber (now oriented straight on to the beam). Use a mode mixer to remove cladding modes and fill all the core modes. Measure the NA at the exit using the beam diameter of the brightest spot.

6. Replace the 50/125 test fiber with the 100/140 test fiber and recheck the alignment.

7. Repeat steps 1-4.

Lab Report:

Include a data table for each fiber. The tables should show a column for angle, received power in mW or W, normalized received power in dB or dBm, and the average (over the positive and negative angles) normalized received power in dB. In each table's title include the reference power (maximum power on axis through the fibers.) [Make sure your detector is not saturated at this maximum power value! Use the “microscope slide” test – slide in front of detector should drop measured power by ~8%. If less, may be saturating.]

Plot the average normalized received power in dB versus the angle (only positive angles) for each fiber. [Using a semilog plot would be helpful. You might also want to average the readings from both directions if they are not symmetrical.]

Calculate the maximum acceptance angle (amax) in degrees and the NA of each of the two fibers. To estimate a, determine the angle at which the power drops to 3% of the maximum. [Note: In some cases, the e -2 or 13.5% of maximum intensity is used instead of the -10 dB point.]

Questions:

Include a separate QUESTIONS section and answer the following questions.

[You might need to read in your text for the answers.]

A. Theoretically, the curve of the power vs. angle for the SI fiber should be flat for incident angles greater than the critical one (all core light should go down the fiber core). Is this true in your measurements? If not, can you explain why?

B. If the 100/140 fiber you tested is uniformly illuminated at one end, what is the maximum size of the cone of light escaping from the other end?

C. If one immerses the fiber end in a liquid with index of refraction, n0=1.3, should the maximum acceptance angle change? If so, what would be the new a? Does the NA change? If so, what is the new NA?

D. The NA is a measure of the light-gathering efficiency of a fiber: larger NA means more light efficiency. Consider two fibers, an SI one and a GRIN one, with the same NA at the core axis, which one is more efficient in gathering light? Use simple geometrical considerations to back up your answer.

REFERENCES

Fiber Optic Communications, J. Palais, Prentice Hall, Fourth Edition, 1998.

Section 2.4 “Numerical Aperture” pages 44-46

Section 4.4 “Coupling of the Waveguide” pages 87-90

Section 5.1 “Step Index Fiber” pages 102-105

Section 5.2 “Graded-Index Fiber” pages 105-108