Optical waveguides – Polarization without modulation
Reexamination Certificate
2002-01-15
2004-07-06
Sanghavi, Hemang (Department: 2874)
Optical waveguides
Polarization without modulation
C385S032000, C385S043000, C385S097000
Reexamination Certificate
active
06760495
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to optical fibers and optical fiber components. It relates in particular to an all-fiber depolarizer enabling the change of an optical signal polarization state, from a strongly polarized state to an unpolarized or depolarized state.
BACKGROUND OF THE INVENTION
Laser light transmitted through optical fibers is usually a polarized light because lasers and laser diodes are strongly polarized light sources. There are however some applications where the polarization of the light is not desirable. In test systems, the results are not influenced by polarization dependent loss (PDL) if the light source is unpolarized. In Raman amplification systems, the Raman effect is polarization dependent, thus the need for an unpolarized source for higher uniformity performance.
There are only a few truly unpolarized light sources available. An incandescent lamp or an arc lamp will produce unpolarized light, i.e. light that contains all states of polarization at the same time. However, these sources are big, produce a lot of heat and thus have low energy efficiencies. It is very inefficient to couple them to single-mode optical fibers. To build a bulk laser that is unpolarized, one can build a laser cavity that has no polarization-dependent element. This is possible with a gas laser that has dielectric mirrors which are not placed at angles with the axis of the cavity. Though these lasers can be coupled relatively efficiently to optical fibers through lenses, they are also bulky and not very energy efficient. To obtain higher power efficiency, one must use a semiconductor diode laser, either as a pump for fiber lasers or as the laser source itself Because of their structure, laser diodes are highly polarized light sources. Only a vertical cavity diode can be unpolarized, but it can not have a high power emission because of the extreme size length of the laser cavity. Thus such laser diodes cannot be used as pump lasers in Raman amplifiers.
With a non-birefringent rare-earth doped fiber, such as erbium doped fiber, one can build unpolarized light sources, such as a broadband amplified spectral emission (ASE) source or even a fiber laser, because they are made with non-polarization selective elements such as broadband WDM fused fiber couplers. These sources are powerful and easy to couple with fibers, but are much bulkier than semiconductor diode lasers. There is thus an interest to depolarize sources such as semiconductor diode lasers.
The purpose of the depolarizer is to transform the output of a polarized source, which has a high degree of polarization (DOP), typically close to 99%, and to reduce that degree of polarization to a very small value, typically a few percent. One cannot truly depolarize a polarized light source, but one can simulate an unpolarized source by shifting the state-of-polarization so fast that the system or the detector averages all the states of polarization. This can be done in time and/or in wavelength. To realize a depolarizer strictly in time, one can actively modulate the polarization. This is usually done with piezoelectric elements that modulate lengths of a birefringent fiber, thus rotating the state of polarization as fast as the piezoelectric element can be modulated. This method works, but does not insure that all-states of polarization are always on average passed through. It will depend on the speed of the detector or the system. This is also an active device, which is useful in test systems, for example because it is more or less universal. In systems or sub-systems however, such as Raman amplifiers, one prefers a passive depolarizer rather than an active one.
Passive depolarizers have been produced using fiber delay lines. The basic principle is to mix different states of polarization, delayed in time, through a fiber loop or fiber lengths. The principle behind passive depolarizers makes them sensitive to the bandwidth of the laser. A very broad source (tens of nanometers) is easy to depolarize, whereas very narrow band lasers, such as a single-mode laser, are almost impossible to depolarize completely because their coherence lengths are very long.
U.S. Pat. Nos. 5,933,555 and 6,205,262 describe a depolarizer based on recirculating loops. To achieve a depolarized state, a polarized light source is split through a 2×2 (2 inputs, 2 outputs) splitter, with a given amount of power being split in the recirculating loop. The rest of the power goes to the transmission fiber. The splitter's optical function is to split light from either of the input ports of the splitter to the two output ports, at a given ratio, i.e., 50%/50%, 33%/67%, etc. This split light goes through the loop back to the second input port of the splitter and thus is partly recombined in the initial transmission fiber. Because there is no control over the polarization in the loop, the amount recombined is in a random orientation with the initial input light. The rest goes back into the loop and again is recombined at a random state, etc. until the amount recombined is negligible. Though this principle does depolarize the light, it only performs a partial depolarization. The depolarization efficiency will depend on the coupling ratio and the polarization changes in the loop. To obtain a relatively uniform depolarization as a function of polarization input states and wavelength, one must use a broadband splitter and cascade many such recirculating loop depolarizers so that if the first loop reduces the degree of polarization to 40%, the second will reduce it to (0.4)
2
=16%. To obtain a very low DOP up to 15 such stages are required. One can make the device a bit more efficient by placing a polarization controller between the loops and either use birefringent fiber or place a further polarization controller in the loop, but this still produces a very bulky arrangement to obtain desired depolarization effects. Fewer loops are required to depolarize a broadband source than a narrower band source. Furthermore, if one wishes to depolarize without controlling the input polarization, theoretically one can never obtain a completely depolarized light. One can only hope that one will tend towards complete depolarization if one adds a great number of stages. The only advantage of this device is that it can be made with a low cost fused non-polarization maintaining couplers, but it requires several couplers, splices and a lot of optical fibers for all the different loops and thus is very bulky and lossy.
A more efficient recirculating loop is presented in U.S. Pat. No. 5,218,652 where control of the states of polarization is provided. Because the states of polarization are controlled, a single recirculating loop is required, reducing the amount of fiber and components. The principle is, however, the same as the recirculating loops discussed in U.S. Pat. No. 5,933,555. Part of the light is recombined with a delay and in a different state of polarization than the input light. In U.S. Pat. No. 5,218,652, the idea is to use a polarization preserving coupler to split the light and a birefringent fiber to preserve the polarization in the recirculating loop. It is stated that the best configuration for this device is to split the light with one part in the transmission and two parts in the loop and to recombine in the coupler at a 90° angle, in essence rotating the polarization maintaining (PM) fiber by 90° before splicing it to the second input of the coupler. In principle, one can achieve complete depolarization in this manner, if there is no optical loss in the coupler and the fiber loop, but in practice because the coupling ratio of the coupler depends on wavelength, the loop is lossy due to the length of the birefringent fiber and the precise control of the splice angle required in the loop. Because of such losses, the optical signal cannot infinitely loop back. However, because this configuration requires a single loop of polarization maintaining PM fiber, it is ultimately less lossy and bulky than the previously discussed design, but it still h
Dion Bruno Y.
Godbout Nicolas
Gonthier François
Villeneuve Alain
ITF Technologies Optiques Inc.
Primak George J.
Rojas Omar
Sanghavi Hemang
LandOfFree
All-fiber depolarizer does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with All-fiber depolarizer, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and All-fiber depolarizer will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3216176