MadSci Network: Physics |
Hi John, The diffraction does not create a loss inside fiberoptic cable. The diffraction occurs at exit and entrance of fiberoptic cable. Losses can occur when the light is "launched" into the optical fiber. There are two methods whereby light is coupled into a fiber. One is by pigtailing. The other is by placing the fiber's tip in very close proximity to an LED or LD. When the proximity type of coupling is employed, the amount of light that will enter the fiber is a function of one of four factors: the intensity of the LED or LD, the area of the light emitting surface, the acceptance angle of the fiber, and the losses due to reflections and scattering. Following is a short discussion on each: Intensity: The intensity of an LED or LD is a function of its design and is usually specified in terms of total power output at a particular drive current. Sometimes, this figure is given as actual power that is delivered into a particular type of fiber. All other factors being equal, more power provided by an LED or LD translates to more power "launched" into the fiber. Area: The amount of light "launched" into a fiber is a function of the area of the light emitting surface compared to the area of the light accepting core of the fiber. The smaller this ratio is, the more light that is "launched" into the fiber. Acceptance Angle: The acceptance angle of a fiber is expressed in terms of numeric aperture. The numerical aperture (NA) is defined as the sine of one half of the acceptance angle of the fiber. Typical NA values are 0.1 to 0.4 which correspond to acceptance angles of 11 degrees to 46 degrees. Optical fibers will only transmit light that enters at an angle that is equal to or less than the acceptance angle for the particular fiber. Other Losses: Other than opaque obstructions on the surface of a fiber, there is always a loss due to reflection from the entrance and exit surface of any fiber. This loss is called the Fresnell Loss and is equal to about 4% for each transition between air and glass. There are special coupling gels that can be applied between glass surfaces to reduce this loss when necessary. The Optical Fiber-Losses in Optical Fiber Other than the losses exhibited when coupling LEDs or LDs into a fiber, there are losses that occur as the light travels through the actual fiber. The core of an optical fiber is made of ultra-pure low-loss glass. Considering that light has to pass through thousands of feet or more of fiber core, the purity of the glass must be extremely high. To appreciate the purity of this glass, consider the glass in common windowpanes. We think of windowpanes as "clear," allowing light to pass freely through, but this is because they are only 1/16 to 1/4 inch thick. In contrast to this clear appearance, the edges of a broken windowpane look green and almost opaque. In this case, the light is passing edgewise into the glass, through several inches. Just imagine how little light would be able to pass through a thousand feet of window glass! Most general purpose optical fiber exhibits losses of 4 to 6 dB per km (a 60% to 75% loss per km) at a wavelength of 850nm. When the wavelength is changed to 1300nm, the loss drops to about 3 to 4 dB (50% to 60%) per km. At 1550nm, it is even lower. Premium fibers are available with loss figures of 3 dB (50%) per km at 850nm and 1 dB (20%) per km at 1300nm. Losses of 0.5 dB (10%) per km at 1550 nm are not uncommon. These losses are primarily the result of random scattering of light and absorption by actual impurities within the glass. Another source of loss within the fiber is due to excessive bending, which causes some of the light to leave the core area of the fiber. The smaller the bend radius, the greater the loss. Because of this, bends along a fiber optic cable should have a turning radius of at least an inch. 1. I do not lose more and more light the longer the pipe is due to diffraction but it affects at the entrance and exit of non TIR pipes 2. Diffraction does not make any loss in the middle of the fiberoptic. 3. The light does not diffract as light exits the fiberoptic. You are right "I know it diverges (since light exiting isn't collimated)" but some of that divergence is not due to diffraction. (i.e. if the light were collimated, it only would still diffract at the fiberoptic exit.) So, it might diffract at the exit. Best Wishes,
Try the links in the MadSci Library for more information on Physics.