Dispersion Compensating Fiber
Click to Enlarge
Single Mode Fiber
Thorlabs offers dispersion compensating optical fiber for custom solutions across a broad spectral range in the telecom region. DCF3 fiber is a non-zero dispersion-shifted fiber (NZ-DSF) with negative dispersion and low positive dispersion slope that is optimal for medium distances and wide band WDM systems. Optical dispersion across the entire C-band enables effective dispersion compensation and suppresses nonlinear impairments. DCF38 has dispersion designed specifically to match and compensate Corning SMF-28e+ or Vascade L1000 fiber. Please see the Dispersion Tutorial tab for more detailed information about dispersion compensating fibers.
Dispersion compensating fibers are fully compatible with typical connectors and termination tools. Loss is slightly higher than typical SM fibers at around 1 dB. Lower losses can be achieved by splicing. Various compatible connectors and tools are summarized in the table to the upper right.
Splicers and tooling designed for Ø125 µm fiber can be used with this fiber. To achieve an optimized fuse, the program should use a shorter fusion time than with conventional SM fibers. This is due to excess diffusion of the core dopants, which alters the guiding properties of the fiber in the splice region. Alteriatively, DCF3 can be used as a bridge between DCF38 and SMF-28 to reduce splice losses. For more information, see "New Technique for Reducing the Splice Loss to Dispersion Compensating Fiber", Edvold B., Gruner-Nielsen, L., Optical Communication, 2, 245-248 (September 19, 1996).
Waveguide dispersion offsets chromatic dispersion to produce zero dispersion at 1.31 µm in step-index SM fiber (Click to Enlarge).
Dispersion in Optical Fiber
Chromatic dispersion is a property of optical fiber where different wavelengths of light propogate at different velocities. Chromatic dispersion is a function of wavelength, and is the sum of two components: material and waveguide dispersion. Material dispersion arises from the change in a material's refractive index with wavelength, which changes the propogation velocity of light as a function of wavelength.
Waveguide dispersion is a separate effect, arising from the geometry of the fiber optic waveguide. Waveguide properties are a function of wavelength; consequently, changing the wavelength affects how light is guided in a single-mode fiber. For example, decreasing the wavelength will increase the relative waveguide dimensions, causing a change in the distribution of light in the cladding and core. In general:
Dispersionchromatic(λ) = Dispersionmaterial(λ) + Dispersionwaveguide(λ)
Since material and waveguide dispersion are wavelength dependent, the dispersion is a function of wavelength. The dispersion slope can be positive or negative.
A fiber designed with more waveguide dispersion shifts the zero-dispersion wavelength to 1.55 µm (Click to Enlarge).
In standard step-index single-mode fiber, the sum of the material and waveguide dispersion is zero near 1310 nm, which is called the zero-disperion wavelength. By varying the fiber's waveguide structure, the waveguide dispersion can be shifted up or down, thus changing the zero-dispersion point. Fiber in which the zero-dispersion wavelength has been changed is called zero dispersion-shifted fiber.
An initial strategy was to alter the waveguide structure to shift the zero-dispersion point to the signal wavelength of 1550 nm, creating zero-dispersion shifted fiber (see the diagram to the right). Unfortunately, fixing the dispersion problem is not so simple. When multiple optical channels pass through the same fiber at wavelengths where dispersion is very close to zero, they suffer from a type of crosstalk called four-wave mixing. The degradation is so severe that zero dispersion-shifted fiber cannot be used for dense-WDM systems. To avoid four-wave mixing, the zero-dispersion wavelength is moved outside the transmission band. So-called nonzero dispersion-shifted fibers have a dispersion that is low, but nonzero in the 1550 nm band (typically 0.1 to 6 ps/nm*km). Although dispersion is minimized, it is still present.
Only total dispersion is shown in this graph. (Click to Enlarge)
Since dispersion is inevitable in optical fibers, dispersion-compensating fibers, such as those sold on this page, can be incorporated into optical systems. The overall dispersion of these fibers is opposite in sign and much larger in magnitude than that of standard fiber, so they can be used to cancel out or compensate the dispersion of a standard single-mode fiber, such as a nonzero dispersion-shifted fiber. A negative dispersion slope enables effective cancellation of dispersion over a larger wavelength range, since the dispersion slope of standard fiber is usually positive. Generally, a short length of dispersion-compensating fiber is spliced into a longer length of standard fiber to compensate for dispersion, as in the example below.
Dispersion can cause various penalties in signal transmission in optical communications systems. Thus, dispersion management is a very important part of designing a fiber optic transmission system. The following table, provided by ITU* standards, which gives the maximum distances for different transmission bit rates and fiber types at around 1550 nm as limited by dispersion.
CDtotal = CDfi + CDDCM + CDother
A dispersion limit, CDlimit, is provided by ITU standards providing the maximum allowable accumulated chromatic dispersion. In general, the relation CDlimit ≥ CDtotal should be true. When CDlimit= CDtotal , a 1 dB decrease in signal strength as a function of bit rate will be present.
Dispersion Compensating Planning Example
Transmitted Power: 4 dBm
First, is dispersion compensation necessary? CDfi = Dispersion x Length = 18.00 ps/(nm x km) x 100 km = 1800 ps/nm. The dispersion limit for this system is CDlimit = ±1000 ps/nm, and so we need dispersion compensation. For this example, we need CDlimit - CDDCM ≥ CDfi.
To reach the positive limit:
Thus, we need -2800 ps/nm ≤ CDDCM ≤ -800 ps/nm. Our DCF38 fiber has dispersion -38.0 ps/(nm x km), so we can use two 13.2 km segments for a total CDDCM of: CDDCM = 2 x 13.2 km x -38.0 ps/ (nm x km) = -1003.2 ps/nm.
Our total dispersion is then CDtot = -1003.2 ps/nm + 1800 ps/nm = 796.8 ps/nm, which is below the dispersion compensation limit.
Click to Enlarge
Undamaged Fiber End
Click to Enlarge
Damaged Fiber End
Laser Induced Damage in Optical Fibers
The following tutorial details damage mechanisms in unterminated (bare) and terminated optical fibers, including damage mechanisms at both the air-to-glass interface and within the glass of the optical fiber. Please note that while general rules and scaling relations can be defined, absolute damage thresholds in optical fibers are extremely application dependent and user specific. This tutorial should only be used as a guide to estimate the damage threshold of an optical fiber in a given application. Additionally, all calculations below only apply if all cleaning and use recommendations listed in the last section of this tutorial have been followed. For further discussion about an optical fiber’s power handling abilities within a specific application, contact Thorlabs’ Tech Support.
Damage at the Free Space-to-Fiber Interface
There are several potential damage mechanisms that can occur at the free space-to-fiber interface when coupling light into a fiber. These come into play whether the fiber is used bare or terminated in a connector.
Unterminated (Bare) Fiber
The effective area for SM fiber is defined by the mode field diameter (MFD), which is the effective cross-sectional area through which light propagates in the fiber. A free-space beam of light must be focused down to a spot of roughly 80% of this diameter to be coupled into the fiber with good efficiency. MFD increases roughly linearly with wavelength, which yields a roughly quadratic increase in damage threshold with wavelength. Additionally, a beam coupled into SM fiber typically has a Gaussian-like profile, resulting in a higher power density at the center of the beam compared with the edges, so a safety margin must be built into the calculated damage threshold value if the calculations assume a uniform density.
Multimode (MM) fiber’s effective area is defined by the core diameter, which is typically far larger than the MFD in SM fiber. Kilowatts of power can be typically coupled into multimode fiber without damage, due to the larger core size and the resulting reduced power density.
It is typically uncommon to use single mode fibers for pulsed applications with high per-pulse powers because the beam needs to be focused down to a very small area for coupling, resulting in a very high power density. It is also uncommon to use SM fiber with ultraviolet light because the MFD becomes extremely small; thus, power handling becomes very low, and coupling becomes very difficult.
Area = πr2 = π(MFD/2)2 = π • 1.22 µm2 = 4.52 µm2
This can be extrapolated to a damage threshold of 11.3 mW. We recommend using the "practical value" maximum power density from the table above to account for a Gaussian power distribution, possible coupling misalignment, and contaminants or imperfections on the fiber end face:
250 kW/cm2 = 2.5 mW/µm2
4.25 µm2 • 2.5 mW/µm2 = 11.3 mW
The scattered light propagates into the epoxy that holds the fiber in the ferrule. If the light is intense enough, it can melt the epoxy, causing it to run onto the face of the connector and into the beam path. The epoxy can be burned off, leaving residue on the end of the fiber, which reduces coupling efficiency and increases scattering, causing further damage. The lack of epoxy between the fiber and ferrule can also cause the fiber to be decentered, which reduces the coupling efficiency and further increases scattering and damage.
The power handling of terminated optical fiber scales with wavelength for two reasons. First, the higher per photon energy of short-wavelength light leads to a greater likelihood of scattering, which increases the optical power incident on the epoxy near the end of the connector. Second, shorter-wavelength light is inherently more difficult to couple into SM fiber due to the smaller MFD, as discussed above. The greater likelihood of light not entering the fiber’s core again increases the chance of damaging scattering effects. This second effect is not as common with MM fibers because their larger core sizes allow easier coupling in general, including with short-wavelength light.
Fiber connectors can be constructed to have an epoxy-free air gap between the optical fiber and ferrule near the fiber end face. This design feature, commonly used with multimode fiber, allows some of the connector-related damage mechanisms to be avoided. Our high-power multimode fiber patch cables use connectors with this design feature.
Combined Damage Thresholds
The graph to the right shows the power handling limitations imposed by the fiber itself and a surrounding connector. The total power handling of a terminated fiber at a given wavelength is limited by the lower of the two limitations at that wavelength. The fiber-limited (blue) line is for SM fibers. An equivalent line for multimode fiber would be far above the SM line on the Y-axis. For terminated multimode fibers, the connector-limited (red) line always determines the damage threshold.
Please note that the values in this graph are rough guidelines detailing estimates of power levels where damage is very unlikely with proper handling and alignment procedures. It is worth noting that optical fibers are frequently used at power levels above those described here. However, damage is likely in these applications. The optical fiber should be considered a consumable lab supply if used at power levels above those recommended by Thorlabs.
Damage Within Optical Fibers
In addition to damage mechanisms at the air-to-glass interface, optical fibers also display power handling limitations due to damage mechanisms within the optical fiber itself. Two categories of damage within the fiber are damage from bend losses and damage from photodarkening.
A special category of optical fiber, called double-clad fiber, can reduce the risk of bend-loss damage by allowing the fiber’s cladding (2nd layer) to also function as a waveguide in addition to the core. By making the critical angle of the cladding/coating interface higher than the critical angle of the core/clad interface, light that escapes the core is loosely confined within the cladding. It will then leak out over a distance of centimeters or meters instead of at one localized spot within the fiber, minimizing damage. Thorlabs manufactures and sells 0.22 NA double-clad multimode fiber, which boasts very high, megawatt range power handling.
Germanium-doped silica, which is commonly used for the core of single mode fiber for red or IR wavelengths, can experience photodarkening with blue visible light. Thus, pure silica core single mode fibers are typically used with short wavelength visible light. Single mode fibers are typically not used with UV light due to the small MFD at these wavelengths, which makes coupling extremely difficult.
Even with the above strategies in place, all fibers eventually experience photodarkening when used with UV light, and thus, fibers used with these wavelengths should be considered consumables.
Tips for Maximizing an Optical Fiber's Power Handling Capability
With a clear understanding of the power-limiting mechanisms of an optical fiber, strategies can be implemented to increase a fiber’s power handling capability and reduce the risk of damage in a given application. All of the calculations above only apply if the following strategies are implemented.
One of the most important aspects of a fiber’s power-handling capability is the quality of the end face. The end face should be clean and clear of dirt and other contaminants that can cause scattering of coupled light. Additionally, if working with bare fiber, the end of the fiber should have a good quality cleave, and any splices should be of good quality to prevent scattering at interfaces.
The alignment process for coupling light into optical fiber is also important to avoid damage to the fiber. During alignment, before optimum coupling is achieved, light may be easily focused onto parts of the fiber other than the core. If a high power beam is focused on the cladding or other parts of the fiber, scattering can occur, causing damage.
Additionally, terminated fibers should not be plugged in or unplugged while the light source is on, again so that focused beams of light are not incident on fragile parts of the connector, possibly causing damage.
Bend losses, discussed above, can cause localized burning in an optical fiber when a large amount of light escapes the fiber in a small area. Fibers carrying large amounts of light should be secured to a steady surface along their entire length to avoid being disturbed or bent.
Additionally, choosing an appropriate optical fiber for a given application can help to avoid damage. Large-mode-area fibers are a good alternative to standard single mode fibers in high-power applications. They provide good beam quality with a larger MFD, thereby decreasing power densities. Standard single mode fibers are also not generally used for ultraviolet applications or high-peak-power pulsed applications due to the high spatial power densities these applications present.