Optical filter

Information

  • Patent Grant
  • 6633698
  • Patent Number
    6,633,698
  • Date Filed
    Wednesday, April 25, 2001
    23 years ago
  • Date Issued
    Tuesday, October 14, 2003
    21 years ago
Abstract
A main optical path disposed between light input and output ends and an auxiliary optical path disposed so as to correspond to the main optical path are optically coupled to each other with first, second, and third optical couplers. The respective optical path lengths of main and auxiliary optical paths differ from each other between the first and second optical couplers and between the second and third optical couplers. Between the first and second optical couplers and between the second and third optical couplers, at least one of the main and auxiliary optical paths is provided with temperature adjusting means for adjusting the temperature of the corresponding optical path.
Description




BACKGROUND OF THE INVENTION




1. Field of the Invention




The present invention relates to an optical filter which is suitably employable as a gain equalizer for equalizing the gain of optical amplification of signal light, or the like in an optical amplifier.




2. Related Background Art




An optical amplifier includes an optical waveguide, doped with a fluorescent material excitable with pumping light, for optically amplifying signal light; and pumping means for supplying pumping light to the optical waveguide. Such an optical amplifier is provided in a repeating station in an optical transmission system, or the like. In particular, it is important for the optical amplifier employed in a wavelength division multiplexing transmission system for transmitting signal light of a plurality of channels to optically amplify the plurality of channels in a collective manner at respective gains equal to each other, and to output each channel of signal light with a power held at a constant target value. Hence, for equalizing the optical amplification of signal light in such an optical amplifier, an optical filter having a loss spectrum in a form similar to that of the gain spectrum in the amplifying optical waveguide is used as a gain equalizer.




For example, literature 1—K. Inoue, et al., “Tunable Gain Equalization Using a Mach-Zehnder Optical Filter in Multistage Fiber Amplifiers,” IEEE Photonics Technology Letters, Vol. 3, No. 8, pp. 718-720 (1991)—and literature 2—G. H. B. Thompson, et al., “Planar Waveguide Filters for Dynamic Equalization of EDFA Spectra,” ECOC'99 (1999)—disclose techniques for flattening the gain of an optical amplifier by means of an optical filter using a Mach-Zehnder interferometer. The techniques disclosed in these literatures aim at adjusting the respective temperatures of individual optical couplers and individual branching optical paths in a Mach-Zehnder interferometer according to the input signal light power, so as to regulate the slope of transmission loss to wavelengths in optical filter, thereby compensating for the fluctuation in the slope of gain to wavelengths accompanying the fluctuation in input signal light power.




SUMMARY OF THE INVENTION




In the above-mentioned conventional techniques, if the power of signal light outputted from the optical amplifier is to be kept constant when the loss in an optical transmission line in front of the optical amplifier fluctuates due to some reason and thereby the signal light fed into the optical amplifier alters its power, it will be necessary for the optical amplification of signal light in the optical amplifier to change its gain. When the gain is changed, the wavelength dependence of gain, i.e., the slope of gain to wavelengths (gain slope), may fluctuate, thereby deteriorating the gain flatness of optical amplifier, which causes the respective powers of a plurality of channels of signal light outputted from the optical amplifier to deviate from each other. Therefore, the respective temperatures of individual optical couplers and branching optical paths in each Mach-Zehnder interferometer constituting the optical filter are adjusted according to the input signal light power, so as to adjust the slope of loss to wavelengths (loss slope) in optical filter, thereby compensating for the fluctuation in gain slope accompanying the fluctuation in gain slope. When the loss slope in optical filter is changed according to the input signal light power, however, the loss level in the signal light wavelength band may fluctuate, whereby the signal light outputted from the optical amplifier after being optically amplified may fluctuate and deteriorate its S/N ratio. Also, the number of heaters provided for adjusting the loss slope in this optical filter is 3 or 6, which is so large that the control of loss slope is complicated.




In order to overcome the problems mentioned above, it is an object of the present invention to provide an optical filter which is suitably employable as a gain equalizer or the like in an optical amplifier, whereas its loss slope is easy to control.




For achieving the above-mentioned object, the optical filter in accordance with the present invention comprises: (1) a main optical path for guiding light from an input end to an output end; (2) an auxiliary optical path, optically coupled to the main optical path with first, second, and third optical couplers, wherein the optical path length between said first and second optical couplers and an optical path length between said second and third optical couplers are different from the main optical path; (3) first temperature adjusting means, disposed in at least one of the main and auxiliary optical paths between the first and second optical couplers, for adjusting the temperature of the main and/or auxiliary optical paths; and (4) second temperature adjusting means, disposed in at least one of the main and auxiliary optical paths between the second and third optical couplers, for adjusting the temperature of the main and/or auxiliary optical paths.




In the optical filter in accordance with the present invention, the main and auxiliary optical paths and the first and second optical couplers constitute a first Mach-Zehnder interferometer, the main and auxiliary optical paths have respective optical path lengths different from each other between the first and second optical couplers, and the temperature of one or both of the optical paths is regulated by the first temperature adjusting means. Further, the main and auxiliary optical paths and the second and third optical couplers constitute a second Mach-Zehnder interferometer, the main and auxiliary optical paths have respective optical path lengths different from each other between the second and third optical couplers, and the temperature of one or both of the optical paths is regulated by the second temperature adjusting means. The temperature adjustment effected by each of the first and second temperature adjusting means makes it possible to set the slope of loss to wavelengths in a given wavelength band. Thus, this optical filter has a simple configuration and is easy to control the loss slope, whereby it can suitably be used as a gain equalizer in an optical amplifier, or the like, for example.




When the first and second temperature adjusting means are disposed in one of the main and auxiliary optical paths alone, it will be sufficient if only two pieces of heaters, Peltier devices, and the like employable as temperature adjusting means are provided, whereby it becomes further easier to control the loss slope.




Preferably, the optical filter further comprises control means for controlling the first and second temperature adjusting means so as to set a slope of transmission loss across the main optical path to wavelengths in a given wavelength band.




Preferably, between the first and second optical couplers and between the second and third optical couplers, the optical path length differences between the main and auxiliary optical paths are not longer than 42 μm. This configuration is suitable for setting the loss slope deviation from a given straight line to 1 dB or less.




Preferably, the transmission loss characteristics across the main optical path may be substantially independent of wavelength in a given wavelength band by adjusting the first and second temperature adjusting means.




The optical filter in accordance with the present invention may be configured such that the bandwidth of the given wavelength band is 25 nm or wider, the absolute value of the slope of transmission loss across the main optical path to wavelengths in the given wavelength is variable at least within the range of 0 to 5 dB/125 nm, a deviation from a given straight line is 1 dB or less, and the minimum value of the transmission loss across the main optical path in the given wavelength band is 2 dB or less. Alternatively, it may be configured such that the bandwidth of the given wavelength band is 36 nm or wider, the absolute value of the slope of transmission loss across the main optical path to wavelengths in the given wavelength is variable at least within the range of 0 to 5 dB/36 nm, a deviation from a given straight line is 1 dB or less, and the minimum value of the transmission loss across the main optical path in the given wavelength band is 2 dB or less. In these cases, the optical filter maintains sufficient linearity and sufficiently low loss, while the loss slope is variable in a sufficient range, in the wavelength band from 1535 nm to 1565 nm, which is a signal light wavelength band, for example.











BRIEF DESCRIPTION OF THE DRAWINGS





FIG. 1

is a diagram of the optical filter in accordance with a basic embodiment;





FIG. 2

is a graph showing wavelength characteristics of transmission loss across the optical filter shown in

FIG. 1

;





FIG. 3

is a diagram of a first embodiment of the optical filter in accordance with the present invention;





FIG. 4

is a graph showing wavelength characteristics of transmission loss across the optical filter shown in

FIG. 3

;





FIG. 5

is a graph showing a wavelength characteristic of power coupling ratio C in the first Mach-Zehnder interferometer in the optical filter of

FIG. 3

;





FIG. 6

is a graph showing a wavelength characteristic of transmission loss across the optical filter of

FIG. 3

;





FIGS. 7A

to


7


C are charts showing each of electric field components E


1


and E


2


in the optical filter of

FIG. 3

in terms of vector;





FIG. 8

is a graph showing wavelength characteristics of transmission loss across the optical filter of

FIG. 3

;





FIG. 9

is a graph showing the relationship between the loss slope deviation from a given straight line and the optical path length difference ΔL in the optical filter of

FIG. 3

;





FIG. 10

is a diagram of an optical amplifier using the optical filter of

FIG. 3

as a gain equalizer;





FIG. 11

is a diagram of a second embodiment of the optical filter in accordance with the present invention;





FIG. 12

to

FIG. 14

are graphs showing wavelength characteristics of transmission loss across the optical filter of

FIG. 11

;





FIG. 15

is a graph showing the relationship between the minimum loss of the optical filter of

FIG. 11

within a band and its loss slope;





FIG. 16

is a graph showing the relationship between the loss slope deviation from a given straight line and the loss slope width in the optical filter of

FIG. 11

;





FIG. 17

is a graph showing the relationship between the loss slope deviation from a given straight line and the optical path length difference ΔL in the optical filter of

FIG. 11

;





FIG. 18

is a diagram of an optical amplifier using the optical filter of

FIG. 11

as a gain equalizer;











DESCRIPTION OF THE PREFERRED EMBODIMENTS




In the following, embodiments of the present invention will be explained in detail with reference to the accompanying drawings. To facilitate the comprehension of the explanation, the same reference numerals denote the same parts, where possible, throughout the drawings, and a repeated explanation will be omitted.




Basic Embodiment




First, before explaining embodiments of the optical filter in accordance with the present invention, a basic embodiment of an optical filter utilizing a Mach-Zehnder interferometer will be explained.

FIG. 1

is a diagram of an optical filter


1


constituting this basic embodiment. This optical filter


1


is an optical planar waveguide type circuit formed on a substrate


100


; and comprises a main optical path


110


, an auxiliary optical path


120


, and a heater (temperature adjusting means)


130


.




The main optical path


110


is an optical path disposed between a light input end


101


located at one end face of the substrate


100


and a light output end


102


located at the other end face thereof. A first optical coupler


141


and a second optical coupler


142


are successively disposed from the light input end


101


to the light output end


102


. The main optical path


110


and auxiliary optical path


120


are optically coupled to each other at the first optical coupler


141


and second optical coupler


142


. The main optical path


110


, auxiliary optical path


120


, first optical coupler


141


, and second optical coupler


142


constitute a Mach-Zehnder interferometer.




The heater


130


is disposed on the main optical path


110


between the first optical coupler


141


and second optical coupler


142


. The heater


130


adjusts the temperature of main optical path


110


, so as to regulate the phase shift amount Δφ of main optical path


110


, thereby controlling the loss in light between the light input end


101


and light output end


102


.




The transmission spectrum T(λ) with respect to the light inputted to the light input end


101


and then outputted from the light output end


102


by way of the main optical path


110


is represented by the expression:








T


(λ)=1


−C




1




−C




2


+2


C




1




C




2


−2{square root over (


C





1





C





2


(


C





1


−1)(


C





2


−1))}cos[(2


πn




c


/λ)Δ


L


−Δφ]  (1)






where λ is the wavelength of light, C


1


is the power coupling ratio at the optical coupler


141


, C


2


is the power coupling ratio at the optical coupler


142


, n


c


is the effective refractive index of each of the main optical path


110


and auxiliary optical path


120


, ΔL is the difference between the respective optical path lengths of the main optical path


110


and auxiliary optical path


120


between the optical couplers


141


and


142


, and Δφ is the phase shift amount according to the temperature adjustment of main optical path


110


effected by the heater


130


.




The following can be seen from the above-mentioned expression (1). Namely, the transmission spectrum T(λ) of optical filter


1


periodically changes depending on the wavelength λ, whereas the wavelength period is shorter as the optical path length difference ΔL is greater. The wavelength λ


0


at which the transmissivity T is maximized (loss is minimized) can be controlled by adjusting the phase shift amount Δφ. For example, if the optical path length difference ΔL is set appropriately, then the transmissivity T can be maximized (loss can be minimized) at a predetermined wavelength (1550 nm) within a predetermined wavelength band (wavelength of 1535 nm to 1565 nm) when the temperature adjustment effected by the heater 130 is at a predetermined level (i.e., when the phase shift amount Δφ is at a predetermined value). Also, the temperature adjustment effected by the heater


130


can set the slope of transmission loss to the wavelength λ in the above-mentioned wavelength band.





FIG. 2

is a graph showing wavelength characteristics of transmission loss in the optical filter


1


in accordance with the basic embodiment. Here, C


1


=C


2


=0.5, n


c


=1.45, and ΔL is 6.95 μm. Further, the phase shift amount Δφ is set to 0°, 30°, 60°, 90°, and 120°. When no temperature adjustment is effected by the heater


130


, i.e., the phase shift amount Δφ is set to 0°, as can be seen from the graph, the transmission loss across the optical filter


1


is minimized at a wavelength of 1550 nm and yields a low value within the wavelength range from 1530 nm to 1570 nm, whereby the slope of transmission loss to the wavelength λ in the above-mentioned wavelength band is low. As the temperature of main optical path


110


is raised by the heater


130


(thereby increasing the phase shift amount Δφ), the slope of transmission loss to the wavelength λ in the above-mentioned wavelength band becomes greater. Here, the loss is minimized at a wavelength of 1535 nm on the shorter wavelength side. The optical filter


1


has such characteristics that the loss slope is variable within the range from 0 to 9.37 dB/30 nm when the phase shift amount Δφ is adjusted within the range from 0° to 120° within the wavelength range from 1535 nm to 1565 nm (with a bandwidth of 30 nm), the loss slope deviation from a given straight line is 1.55 dB or less, and the minimum value of transmission loss in the above-mentioned wavelength band is 2.84 dB.




When a Peltier device is provided in place of the heater


130


in the basic embodiment so as to raise or lower the temperature of main optical path


110


, the value of phase shift amount Δφ can be set not only positive but also negative. If the value of phase shift amount Δφ is thus changed within the range from −120° to 0°, then the loss slope can be set within the range from −9.37 dB/30 nm to 0 dB/30 nm within the wavelength band from 1535 nm to 1565 nm, whereupon the transmission loss is minimized at a wavelength of 1565 nm on the longer wavelength side. Also, a heater or Peltier device may be disposed on the auxiliary optical path


120


instead of main optical path


110


, or on each of the main optical path


110


and auxiliary optical path


120


.




In the optical filter


1


in accordance with the basic embodiment, however, the loss slope deviation from a given straight line cannot always be considered sufficiently small, and the minimum value of transmission loss in the wavelength band in use cannot always be considered sufficiently low. This is because of the fact that the optical filter


1


having a structure such as that shown in

FIG. 1

has such a tradeoff relationship between the linearity and low transmission loss that increasing the optical path length difference ΔL lowers the loss but deteriorates the linearity, whereas lowering the optical path length difference ΔL improves the linearity but increases the transmission loss. Hence, the configuration of optical filter


1


in accordance with this basic embodiment is hard to attain linearity and low transmission loss at the same time. The optical filter in accordance with the present invention realizes a configuration achieving linearity and low transmission loss at the same time, whose configuration will now be explained specifically.




First Embodiment




To begin with, a first embodiment of the optical filter in accordance with the present invention will be explained.

FIG. 3

is a diagram of an optical filter


2


in accordance with the first embodiment. The optical filter


2


in accordance with this embodiment is an optical planar waveguide type circuit formed on a substrate


200


; and comprises a main optical path


210


, an auxiliary optical path


220


, a heater (first temperature adjusting means)


231


, and a heater (second temperature adjusting means)


232


, as well as a controller


280


for controlling operations of the heaters


231


and


232


.




The main optical path


210


is an optical path disposed between a light input end


201


located at one end face of the substrate


200


and a light output end


202


located at the other end face thereof. A first optical coupler


241


, a second optical coupler


242


, and a third optical coupler


243


are successively disposed from the light input end


201


to the light output end


202


. The main optical path


210


and auxiliary optical path


220


are optically coupled to each other at the first optical coupler


241


, second optical coupler


242


, and third optical coupler


243


. The main optical path


210


, auxiliary optical path


220


, first optical coupler


241


, and second optical coupler


242


constitute a first Mach-Zehnder interferometer


251


. The main optical path


210


, auxiliary optical path


220


, second optical coupler


242


, and third optical coupler


243


constitute a second Mach-Zehnder interferometer


252


.




The heater


231


is disposed on the auxiliary optical path


220


between the first optical coupler


241


and second optical coupler


242


; and, under instructions from the controller


280


, adjusts the temperature of auxiliary optical path


220


, thereby regulating the phase shift amount Δφ


1


of auxiliary optical path


220


. The heater


232


is disposed on the main optical path


210


between the second optical path


242


and third optical coupler


243


; and, under instructions from the controller


280


, adjusts the temperature of main optical path


210


, thereby regulating the phase shift amount Δφ


2


of main optical path


210


. Namely, the controller


280


functions to regulate the transmission loss across the main optical path


210


by controlling operations of the heaters


231


and


232


.




The transmission spectrum T(λ) with respect to the light inputted to the light input end


201


and then outputted from the light output end


202


by way of the main optical path


210


is represented by the expression:








T


(λ)=1


−C−C




3


+2


C




3


−2{square root over (


CC





3


(


C


−1)(


C





3


−1))}cos[(2


πn




c


/λ)Δ


L




2


−Δφ


2


]  (2)






where λ is the wavelength of light, C is the power coupling ratio of the Mach-Zehnder interferometer


251


when assumed to be a single optical coupler and is represented by an expression similar to the above-mentioned expression (1), C


3


is the power coupling ratio at the optical coupler


243


, n


c


is the effective refractive index of each of the main optical path


210


and auxiliary optical path


220


, ΔL


2


is the difference between the respective optical path lengths of the main optical path


210


and auxiliary optical path


220


between the optical couplers


242


and


243


, and Δφ


2


is the phase shift amount according to the temperature adjustment of main optical path


210


effected by the heater


232


.




Namely, the configuration of the optical filter


2


in accordance with this embodiment is equivalent to that of the optical filter


1


of the above-mentioned basic embodiment when one optical coupler


141


is replaced by the Mach-Zehnder interferometer


251


. In the optical filter


2


, the power coupling ratio C of light in the Mach-Zehnder interferometer


251


is dependent on wavelength, whereas the wavelength dependence of power coupling efficiency C can be adjusted by the difference ΔL


1


between the respective optical path lengths of the main optical path


210


and auxiliary optical path


220


between the optical couplers


241


and


242


, and by the phase shift amount Δφ


1


according to the temperature adjustment of auxiliary optical path


220


effected by the heater


231


.





FIG. 4

is a graph showing wavelength characteristics of transmission loss in the optical filter


2


in accordance with this embodiment. Here, the power coupling ratio C


3


of light in the optical coupler


243


is 0.5, n


c


=1.45, ΔL


2


=16.57 μm, and Δφ


1


=0°. Also, the power coupling ratio C in Mach-Zehnder interferometer


251


is set to 0.2 and 0.5. In the optical filter


2


, as can be seen from this graph, the difference in loss between the respective cases where C=0.2 and C=0.5 becomes smaller as the wavelength is farther from 1550 nm, though a lower loss is attained near a wavelength of 1550 nm when C=0.5.




Hence, the power coupling ratio C of light in the Mach-Zehnder interferometer


251


is set so as to exhibit a wavelength dependence such as that shown in FIG.


5


. Namely, the power coupling ratio C of light in the Mach-Zehnder interferometer


251


is minimized near a wavelength of 1550 nm and is made greater as the wavelength is farther therefrom. Since the respective optical path lengths of the main optical path


210


and auxiliary optical path


220


between the optical couplers


241


and


242


differ from each other, and the phase shift amount Δφ


1


of auxiliary optical path


220


can be regulated when the temperature of auxiliary optical path


220


is adjusted by the heater


231


, the power coupling ratio C of light in the Mach-Zehnder interferometer


251


can have such a wavelength dependence. Such setting makes it possible to attain the linearity in transmission loss and a low transmission loss across the optical filter


2


at the same time.





FIG. 6

is a graph showing a wavelength characteristic of transmission loss across the optical filter


2


in accordance with this embodiment. Here, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.95 μm, ΔL


2


=16.57 μm, Δφ


1


=0°, and Δφ


2


=0°. In this case, the optical filter


2


exhibits a lower wavelength dependence of transmission loss near a wavelength of 1550 nm, and a lower loss over a wide wavelength range. Also, the width of wavelength band yielding a flat loss is broader than the bandwidth expectable from FIG.


4


.




This can be explained as follows. As shown in

FIG. 3

, the electric field component E


1


of light outputted from the Mach-Zehnder interferometer


251


to the main optical path


210


is represented by the following expression:








E




1


={square root over (1


−C


(λ))}


E




0


  (3)






where E


0


is the electric field component inputted to the Mach-Zehnder interferometer


251


. On the other hand, the electric field component E


2


of light outputted from the Mach-Zehnder interferometer


251


to the auxiliary optical path


220


is represented by the following expression:








E




2


={square root over (


C


(λ))}


E




0




exp[j


Δθ(λ)]  (4)






where θ(λ) is the phase difference between the electric field components E


1


and E


2


, and depends on the wavelength λ. When the phase difference θ(λ) is designed appropriately, it can broaden the wavelength band yielding a flat transmission loss as shown in FIG.


6


.





FIGS. 7A

to


7


C are charts showing each of electric field components E


1


and E


2


in terms of vector in the optical filter


2


in accordance with this embodiment. Here, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.95 μm, ΔL


2


=16.57 μm, Δφ


1


=0°, and Δφ


2


=0.

FIG. 7A

illustrates the case where the wavelength is 1530 nm, which yields a phase difference θ(λ) of 58.6°.

FIG. 7B

illustrates the case where the wavelength is 1550 nm, which yields a phase difference θ(λ) of 90.0°.

FIG. 7C

illustrates the case where the wavelength is 1570 nm, which yields a phase difference θ(λ) of 120.7°.




Thus, the phase difference θ(λ) between the respective electric field components E


1


, E


2


outputted from the Mach-Zehnder interferometer


251


to the main optical path


210


and auxiliary optical path


220


has such a wavelength dependence that it monotonously increases as the wavelength λ is longer. On the other hand, the phase difference between the main optical path


210


and auxiliary optical path


220


from the Mach-Zehnder interferometer


251


to the optical coupler


243


is represented by the expression of (2πn


c


/λ)ΔL


2


, and monotonously decreases as the wavelength λ is longer. Therefore, the phase difference between the electric field components E


1


, E


2


at the time when they reach the optical coupler


243


has a lower wavelength dependence. As a consequence, the wavelength dependence of loss in the optical filter


2


decreases, which makes it possible to broaden the wavelength band yielding a flat transmission loss as shown in FIG.


6


.





FIG. 8

is a graph showing wavelength characteristics of transmission loss across the optical filter


2


in accordance with this embodiment. Here, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.95 μm, and ΔL


2


=16.57 μm. Also, Δφ


1


=Δφ


2


, and they are set to each of 0°, 10°, 20°, 30°, 40°, 50°, and 60°.




As can be seen from this graph, when no temperature adjustment is effected by the heaters


231


,


232


, i.e., the phase shift amounts Δφ


1


, Δφ


2


are 0°, the transmission loss across the optical filter


2


is minimized at a wavelength of 1550 nm and becomes a low value within the wavelength range from 1530 nm to 1570 nm, thus yielding a low slope of transmission loss with an excellent linearity with respect to the wavelength λ in the above-mentioned wavelength range. Also, the slope of transmission loss to the wavelength λ in the above-mentioned wavelength range increases as the phase shift amounts Δφ


1


, Δφ


2


are enhanced by the temperature adjustment effected by the heaters


231


,


232


. Here, the transmission loss is minimized at a wavelength of 1535 nm on the shorter wavelength side.




The optical filter


2


has such characteristics that the loss slope is variable within the range from 0 to 7.8 dB/30 nm when the phase shift amounts Δφ


1


, Δφ


2


are adjusted within the range from 0° to 60° within the wavelength range from 1535 nm to 1565 nm (with a bandwidth of 30 nm), the loss slope deviation from a given straight line is 0.63 dB or less, thereby yielding an excellent linearity, and the minimum value of transmission loss in the above-mentioned wavelength band is 1.71 dB.





FIG. 9

is a graph showing the relationship between the loss slope deviation from a given straight line and the optical path length difference ΔL in the optical filter


2


in accordance with this embodiment. Here, the optical path length difference ΔL represents the longer one of the optical path length difference ΔL


1


between the main optical path


210


and auxiliary optical path


220


between the first optical coupler


241


and second optical coupler


242


, and the optical path length difference ΔL


2


between the main optical path


210


and auxiliary optical path


220


between the second optical coupler


242


and third optical coupler


243


. As shown in this graph, the loss slope deviation from a given straight line is greater in the optical filter


2


as the optical path length difference ΔL is greater. When the value of optical path length difference ΔL is 42 μm or less, the loss slope deviation from a given straight line becomes 1 dB or less in the optical filter


2


, which is favorable.




When Peltier devices are provided in place of the heaters


231


,


232


in this embodiment so as to raise or lower the temperatures of main optical path


210


and auxiliary optical path


220


, the values of phase shift amounts Δφ


1


, Δφ


2


can be set not only positive but also negative. If the values of phase shift amounts Δφ


1


, Δφ


2


are thus changed within the range from −60° to 0°, then the loss slope can be set within the range from −7.8 dB/30 nm to 0 dB/30 nm within the wavelength band from 1535 nm to 1565 nm, whereupon the transmission loss is minimized at a wavelength of 1565 nm on the longer wavelength side.




A heater or Peltier device may be disposed on the main optical path


210


instead of auxiliary optical path


220


between the optical couplers


241


and


242


, or on the auxiliary optical path


220


instead of main optical path


210


between the optical couplers


242


and


243


. Also, a heater or Peltier device may be disposed on each of the main optical path


210


and auxiliary optical path


220


.




The optical filter


2


is suitably employable as a gain equalizer in an optical amplifier.

FIG. 10

is a diagram of an optical amplifier


20


using the optical filter


2


in accordance with this embodiment as a gain equalizer. This optical amplifier


20


comprises a controller


21


, an optical amplifier section


22


, a branching section


23


, and a light-receiving section


24


in addition to the optical filter


2


as a gain equalizer. The branching section


23


branches out a part of signal light fed into the optical amplifier


20


, so as to output it toward the light-receiving section


24


and the rest toward the optical amplifier section


22


. The light-receiving section


24


is a photodiode, for example, which receives the signal light reaching there from the branching section


23


and outputs an electric signal corresponding to the power of signal light fed into the optical amplifier


20


. The optical amplifier section


22


has an amplifying optical waveguide (e.g., an optical fiber whose optical waveguide region is doped with Er element) and a pumping light source (e.g., semiconductor laser light source for outputting laser light having a wavelength of 1.48 μm as the pumping light). The optical amplifier section


22


optically amplifies the inputted signal light and outputs thus amplified signal.




The optical filter


2


as the gain equalizer has a loss slope set by the controller


21


, and inputs the signal light outputted from the optical amplifier section


22


and carries out gain equalization. The controller


21


inputs the electric signal outputted from the light-receiving section


24


and monitors, according to the electric signal, the power of signal light fed into the optical amplifier


20


. According to the input signal light power, the controller


21


controls the phase shift amount Δφ


1


caused by the temperature adjustment of heater


231


and the phase shift amount Δφ


2


caused by the temperature adjustment of heater


232


in relation to each other, thereby regulating the loss slope of optical filter


2


. For example, when the optical filter


2


has the configuration explained in

FIG. 8

, the controller


21


controls the phase shift amounts Δφ


1


and Δφ


2


such that they equal each other.




Thus configured optical amplifier


20


operates as follows. Namely, the gain in optical amplification of signal light in the optical amplifier section


22


is changed in order to maintain a constant power of signal light outputted from the optical amplifier


20


when the loss in the optical transmission line fluctuates for some reason and thereby the power of signal light fed into the optical amplifier


20


fluctuates. When the gain is changed, the wavelength dependence of gain in the optical amplifier section


22


, i.e., gain slope, fluctuates, thereby deteriorating the gain flatness of optical amplifier section


22


. However, by using the optical filter


2


in accordance with this embodiment as a gain equalizer in the optical amplifier


20


and adjusting the values of phase shift amounts Δφ


1


, Δφ


2


of optical filter


2


, i.e., value of transmission loss slope, the fluctuation in gain slope of the optical amplifier section


22


accompanying the fluctuation in input signal light power can be compensated for with the loss slope of optical filter


2


.




Here, even when the loss slope of optical filter


2


is altered, the loss in the signal light wavelength band is low, whereby the signal light outputted from the optical amplifier


20


would not deteriorate its S/N ratio. When the signal light wavelength band and/or center wavelength are set as mentioned above, in particular, the optical filter


2


in accordance with this embodiment can favorably be used as a gain equalizer for equalizing the gain characteristic of the optical amplifier


20


disposed in a repeater station or the like in an optical transmission system for transmitting signal light of a plurality of channels in a wavelength band of 1.55 μm or 1.59 μm. As in the foregoing, it will be sufficient if only two pieces of heaters or Peltier devices are provided as temperature adjusting means, whereby the loss slope is easy to control, and the optical filter


2


in accordance with this embodiment is suitably employable as a gain equalizer or the like in the optical amplifier.




Second Embodiment




A second embodiment of the optical filter in accordance with the present invention will now be explained.

FIG. 11

is a diagram showing an optical filter


3


in accordance with the second embodiment. The optical filter


3


in accordance with this embodiment further comprises heaters


331


and


332


in addition to the configuration of the optical filter


2


in accordance with the first embodiment.




The heater


231


is disposed on the auxiliary optical path


220


between the first optical coupler


241


and the second optical coupler


242


, and adjusts the temperature of auxiliary optical path


220


, thereby regulating the phase shift amount Δφ


1


of auxiliary optical path


220


therebetween. The heater


232


is disposed on the main optical path


210


between the second optical coupler


242


and the third optical coupler


243


, and adjusts the temperature of main optical path


210


, thereby regulating the phase shift amount Δφ


2


of main optical path


210


therebetween. The heater


331


is disposed on the main optical path


210


between the first optical coupler


241


and the second optical coupler


242


, and adjusts the temperature of main optical path


210


, thereby regulating the phase shift amount Δφ


3


of main optical path


210


therebetween. The heater


332


is disposed on the auxiliary optical path


220


between the second optical coupler


242


and the third optical coupler


243


, and adjusts the temperature of auxiliary optical path


220


, thereby regulating the phase shift amount Δφ


4


of auxiliary optical path


220


therebetween. Thus, the heaters


231


,


232


,


331


, and


332


adjust the transmission loss between the light input end


201


and light output end


202


.





FIG. 12

is a graph showing wavelength characteristics of loss in the optical filter


3


in accordance with the second embodiment. Here, as in the case of

FIG. 8

in the first embodiment, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


6.95 μm, and ΔL


2


=16.57 μm. Also, Δφ


1


=Δφ


2


=21° or 42° and Δφ


3


=Δφ


4


=0°, or Δφ


1


=Δφ


2


=0° and Δφ


3


=Δφ


4


=21° or 42°.




As can be seen from this graph, when no temperature adjustment is effected by the heaters


231


,


232


,


331


,


332


, i.e., all the phase shift amounts Δφ


1


to Δφ


4


are 0°, the transmission loss across the optical filter


3


is minimized near a wavelength of 1545 nm and becomes a low value within the wavelength range from 1527 nm to 1563 nm, thus yielding a low slope of transmission loss with an excellent linearity with respect to the wavelength λ in the above-mentioned wavelength range.




The slope of transmission loss to the wavelength λ in the above-mentioned wavelength range increases as the phase shift amounts Δφ


1


, Δφ


2


are enhanced by the temperature adjustment effected by the heaters


231


,


232


while keeping the phase shift amount Δφ


3


=Δφ


4


=0° with no temperature adjustment effected by the heaters


331


,


332


. Here, the transmission loss is minimized at a wavelength of 1527 nm on the shorter wavelength side.




On the other hand, the slope of transmission loss to the wavelength λ in the above-mentioned wavelength range increases as the phase shift amounts Δφ


3


, Δφ


4


are enhanced by the temperature adjustment effected by the heaters


331


,


332


while keeping the phase shift amount Δφ


1


=Δφ


2


=0° with no temperature adjustment effected by the heaters


231


,


232


. Here, the loss is minimized at a wavelength of 1563 nm on the longer wavelength side.




The optical filter


3


has such characteristics that, within the wavelength range from 1527 nm to 1563 nm (with a bandwidth of 36 nm), the loss slope is variable within the range from 0 to 5.06 dB/36 nm when the phase shift amounts Δφ


1


, Δφ


2


are adjusted within the range from 0° to 42° and is variable within the range from 0 to −5.33 dB/36 nm when the phase shift amounts Δφ


3


, Δφ


4


are adjusted within the range from 0° to 42°, the loss slope deviation from a given straight line is 0.55 dB or less, thereby yielding an excellent linearity, and the minimum value of transmission loss in the above-mentioned wavelength band is 0.40 dB or less.





FIG. 13

is also a graph showing wavelength characteristics of transmission loss across the optical filter


3


in accordance with the second embodiment. Here, differing from the case with

FIG. 12

, ΔL


1


=6.82 μm, and ΔL


2


=16.70 μm. Also, while Δφ


3


=Δφ


4


=0°, Δφ


1


=Δφ


2


=0°, 21°, 43°, 63°, and 84°. The wavelength dependence of transmission loss shown in this drawing is similar to that shown in FIG.


12


. However, while the values of phase shift amounts Δφ


1


and Δφ


2


are adjusted so as to regulate the loss slope in a positive range whereas the values of phase shift amounts Δφ


3


and Δφ


4


are adjusted so as to regulate the loss slope in a negative range in the characteristics shown in

FIG. 12

, the loss slope can be regulated in both positive and negative ranges by adjusting only the two kinds of phase shift amounts Δφ


1


and Δφ


2


in those shown in FIG.


13


.




The difference between the respective operations in

FIGS. 12 and 13

mentioned above is caused by the following difference in design. Namely, in the relational expression of








n




c




·ΔL=m·λ




0


  (5)






among the effective refractive index n


c


, optical path length difference ΔL, and center wavelength λ


0


, the respective values of optical path length differences ΔL


1


and ΔL


2


and effective refractive index n


c


are set appropriately. Here, if the value of parameter m is designed so as to become an integer or an integer +½, then the absolute value of loss slope of optical filter


3


can be minimized when all the phase shift amounts Δφ


1


to Δφ


4


are 0° as shown in FIG.


12


. In

FIG. 12

, m1=6.50 in the Mach-Zehnder interferometer


251


, whereas m2=15.50 in the Mach-Zehnder interferometer


252


. If the value of parameter m is designed so as to attain a value other than those mentioned above, on the other hand, then the optical filter


3


can have an initial slope when all the phase shift amounts Δφ


1


to Δφ


4


are 0° as shown in FIG.


13


. In

FIG. 13

, m1=6.38 in the Mach-Zehnder interferometer


251


, whereas m2=15.62 in the Mach-Zehnder interferometer


252


.




When the respective values of optical path length differences ΔL


1


and ΔL


2


and effective refractive index n


c


are set appropriately as in the foregoing, the loss slope can be regulated within both positive and negative ranges by adjusting two kinds of values of phase shift amounts Δφ


1


and Δφ


2


alone as in the case of FIG.


13


. Here, control is easier than that in the case shown in

FIG. 12

in which four kinds of values of phase shift amounts Δφ


1


to Δφ


4


are regulated. Even when the respective values of optical path length differences ΔL


1


and ΔL


2


and effective refractive index n


c


are set appropriately, however, each value is not always at its designed value in the final product. Therefore, the optical filter


3


comprises the heaters


331


and


332


, whose temperature adjustment regulates the optical path length differences ΔL


1


and ΔL


2


such that they attain their designed values.





FIG. 14

is another graph showing wavelength characteristics of transmission loss across the optical filter


3


in accordance with the second embodiment. Here, differing from the cases with FIG.


12


and

FIG. 13

, C


1


=0.04, C


2


=0.22, C


3


=0.25, ΔL


1


=6.03 μm, and ΔL


2


=9.31 μm. Also, while Δφ


2


=Δφ


3


=Δφ


4


=0°, Δφ


1


=0°, 30°, 60°, 90°, and 120°.




As can be seen from this graph, the optical filter


3


has characteristics that, within the wavelength range from 1570 nm to 1610 nm (with a bandwidth of 40 nm), the loss slope is variable within the range from 0 to 5 dB/40 nm when the phase shift amounts Δ 


1


is adjusted within the range from 0° to 120°, and the minimum value of transmission loss in the above-mentioned wavelength band is 0.25 dB or less.





FIG. 15

is a graph showing the relationship between the minimum loss of the optical filter


3


in accordance with the second embodiment in a band and the loss slope thereof. Here, as in the case with

FIG. 13

, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.82 μm, and ΔL


2


=16.70 μm. As can be seen from this graph, the minimum transmission loss is 0.5 dB or less within the wavelength band from 1527 nm to 1563 nm even when the loss slope is changed within the range of ±5 dB/36 nm.





FIG. 16

is a graph showing the relationship between the loss slope deviation from a given straight line and the loss slope width in the optical filter


3


in accordance with the second embodiment. Here, as in the case with

FIG. 13

, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.82 μm, and ΔL


2


=16.70 μm. The loss slope width is the difference between the maximum and minimum losses within the wavelength band from 1527 nm to 1563 nm. As can be seen from this graph, the loss slope deviation from a given straight line is 0.6 dB or less even when the loss slope is changed, whereby its linearity is excellent.





FIG. 17

is a graph showing the relationship between the loss slope deviation from a given straight line and the optical path length difference ΔL in the optical filter


3


in accordance with the second embodiment. Here, as in the case with

FIG. 13

, C


1


=0.276, C


2


=0.724, C


3


=0.5, n


c


=1.45, ΔL


1


=6.82 μm, and ΔL


2


=16.70 μm. This graph also shows the optical filter


1


in accordance with the basic embodiment. In order for the loss slope deviation from a given straight line to become 1 dB or less, as can be seen from this graph, it is necessary for the optical filter


3


in accordance with the second embodiment to have an optical path length difference ΔL of 36 μm or less, and it is necessary for the optical filter


1


in accordance with the basic embodiment to have an optical path length difference ΔL of 15 μm or less. The deviation is 0.20 dB in the optical filter


3


in accordance with the second embodiment when the optical path length difference ΔL is 16.6 μm, whereas the deviation is 0.38 dB in the optical filter


1


in accordance with the basic embodiment when the optical path length difference ΔL is 0.53 μm.




The optical filter


3


in accordance with this embodiment is suitably employable as a gain equalizer in an optical amplifier.

FIG. 18

is a diagram of an optical amplifier


30


using the optical filter


3


in accordance with the second embodiment as a gain equalizer. This optical amplifier


30


comprises a controller


31


, an optical amplifier section


32


, a branching section


33


, and a light-receiving section


34


in addition to the optical filter


3


as a gain equalizer. The branching section


33


branches out a part of signal light fed into the optical amplifier


30


, so as to output it toward the light-receiving section


34


and the rest toward the optical amplifier section


32


. The light-receiving section


34


is a photodiode, for example, which receives the signal light reaching there from the branching section


33


and outputs an electric signal corresponding to the power of signal light fed into the optical amplifier


30


. The optical amplifier section


32


has an amplifying optical waveguide (e.g., an optical fiber whose optical waveguide region is doped with Er element) and a pumping light source (e.g., semiconductor laser light source for outputting laser light having a wavelength of 1.48 μm as the pumping light). The optical amplifier section


32


optically amplifies the inputted signal light and outputs thus amplified signal.




The optical filter


3


as a gain equalizer has a predetermined loss slope in which the phase shift amounts Δφ


3


and Δφ


4


are appropriately adjusted and initially set according to the temperature adjustment by each of the heaters


331


and


332


under the control of the controller


31


, and inputs the signal light outputted from the optical amplifier section


32


, thereby carrying out gain equalization. The controller


31


inputs the electric signal outputted from the light-receiving section


34


and monitors, according to the electric signal, the power of signal light fed into the optical amplifier


30


. Then, according to the input signal power, the controller


31


controls the phase shift amount Δφ


1


caused by the temperature adjustment by the heater


231


and the phase shift amount Δφ


2


caused by the temperature adjustment by the heater


232


in relation to each other, thereby regulating the loss slope of optical filter


3


. When the optical filter


3


has the configuration explained with

FIG. 11

, for example, the controller


31


controls the phase shift amounts Δφ


1


, Δφ


2


such that they equal each other.




Thus configured optical amplifier


30


operates as follows. Namely, the gain in optical amplification of signal light in the optical amplifier section


32


is changed in order to maintain a constant power of signal light outputted from the optical amplifier


30


when the transmission loss in the optical transmission line fluctuates for some reason and thereby the power of signal light fed into the optical amplifier


30


fluctuates. When the gain is changed, the wavelength dependence of gain in the optical amplifier section


32


, i.e., gain slope, fluctuates, thereby deteriorating the gain flatness of optical amplifier section


32


. However, by using the optical filter


3


in accordance with this embodiment as a gain equalizer in the optical amplifier


30


and adjusting the values of phase shift amounts Δφ


1


, Δφ


2


of optical filter


3


, i.e., value of transmission loss slope, the fluctuation in gain slope of the optical amplifier section


32


accompanying the fluctuation in input signal light power can be compensated for with the loss slope of optical filter


3


.




Here, even when the loss slope of optical filter


3


is altered, the loss in the signal light wavelength band is low, whereby the signal light outputted from the optical amplifier


30


would not deteriorate its S/N ratio. When the signal light wavelength band and center thickness are set as mentioned above, in particular, the optical filter


3


in accordance with this embodiment can favorably be used as a gain equalizer for equalizing the gain characteristic of the optical amplifier


30


disposed in a repeater station or the like in an optical transmission system for transmitting signal light of a plurality of channels in a wavelength band of 1.55 μm or 1.59 μm. As in the foregoing, it will be sufficient if only two pieces of heaters or Peltier devices are provided as temperature adjusting means, whereby the loss slope is easy to control, and the optical filter


3


in accordance with this embodiment is suitably employable as a gain equalizer or the like in the optical amplifier.




Without being restricted to the above-mentioned embodiments, the present invention can be modified in various manners. The optical filters in the above-mentioned embodiments, in which individual constituents are formed on a substrate so as to be integrated, are favorable in that their handling is easy. However, the optical filter in accordance with the present invention may be configured such that each of the main and auxiliary optical paths is made of an optical fiber, whereas each optical coupler is made of an optical coupler. This case is preferable in that the increment of the transmission loss at the time of disposing the optical filter on the optical fiber transmission line is low.




The operating wavelength band of optical filter can be set arbitrarily. The operating wavelength band can be set when the respective values of optical path length differences ΔL


1


and ΔL


2


are set appropriately according to the above-mentioned expression (5).



Claims
  • 1. An optical filter comprising:a main optical path for guiding light from an input end to an output end; an auxiliary optical path, optically coupled to said main optical path with three optical couplers composed of first, second, and third optical couplers, wherein the optical path length between said first and second optical couplers and an optical path length between said second and third optical couplers are different from said main optical path; two asymmetrical Mach-Zehnder interferometers combined with said three optical couplers and sharing said second optical coupler; first temperature adjusting means, disposed in at least one of said main and auxiliary optical paths between said first and second optical couplers, for adjusting the temperature of said main and/or auxiliary optical paths; second temperature adjusting means, disposed in at least one of said main and auxiliary optical paths between said second and third optical couplers, for adjusting the temperature of said main and/or auxiliary optical paths; and control means for controlling said first and second temperature adjusting means so as to set a slope of transmission loss across said main optical path to wavelength in a given wavelength band.
  • 2. An optical filter according to claim 1, wherein between said first and second optical couplers and between said second and third optical couplers, the optical path length differences between said main and auxiliary optical paths are not longer than 42 μm.
  • 3. An optical filter according to claim 1, wherein the transmission loss characteristics across said main optical path is substantially independent of wavelength in a given wavelength band by adjusting said first and second temperature adjusting means.
  • 4. An optical filter according to claim 1, wherein the bandwidth of said given wavelength band is 25 nm or wider, the absolute value of the slope of transmission loss across said main optical path to wavelength in said given wavelength is variable at least within the range of 0 to 5 dB/25 nm, a deviation from a given straight line is 1 dB or less, and the minimum value of the transmission loss across said main optical path in said given wavelength band is 2 dB or less.
  • 5. An optical filter according to claim 2, wherein the bandwidth of said given wavelength band is 36 nm or wider, the absolute value of the slope of transmission loss across said main optical path to wavelength in said given wavelength is variable at least within the range of 0 to 5 dB/36 nm, a deviation from a given straight line is 1 dB or less, and the minimum value of the transmission loss across said main optical path in said given wavelength band is 2 dB or less.
Priority Claims (2)
Number Date Country Kind
P2000-126130 Apr 2000 JP
P2000-237363 Aug 2000 JP
US Referenced Citations (4)
Number Name Date Kind
5596661 Henry et al. Jan 1997 A
6285810 Fincato et al. Sep 2001 B1
6333807 Hatayama et al. Dec 2001 B1
6560381 Hatayama et al. May 2003 B2
Foreign Referenced Citations (3)
Number Date Country
0 382 461 Aug 1990 EP
0 926 532 Jun 1999 EP
1 041 424 Oct 2000 EP
Non-Patent Literature Citations (5)
Entry
K. Jinguji et al. Optical Half-Band Filters. Journal of Lightwave Technology, vol. 18, No. 2, pp. 252-259. Feb. 2000.*
K. Jinguji et al. Two-Port Optical Wavelength Circuits Composed of Cascaded Mach-Zehnder Interferometers with Point-Symmetrical Configurations. Journal of Lightwave Technology, vol.14, No. 10, pp. 2301-2310. Oct. 1996.*
K. Inoue, et al., “Tunable Gain Equalization using a Mach-Zehnder Optical Filter in Multistage Fiber Amplifiers,” IEEE Photonics Technology Letters, vol. 3, No. 8, pp. 718-720 ( Aug. 1991).
G.H.B. Thompson, et al., “Planar Waveguide Filters for Dynamic Equalisation of EDFA Spectra,” ECOC'99, pp. 1-320-321 (Sep. 1999).
Higher Order Dispersion Equaliser of Dispersion Shifted Fibre Using A Lattice-Form Programmable Optical Fiber, Takiguchi, et al., Electronics Letters, IEE Stevenage, GB, vol. 32, No. 8, Apr. 11, 1996.