Back to EveryPatent.com
| United States Patent |
6,044,195
|
|
Matsumoto
, et al.
|
March 28, 2000
|
Wavelength dependence correction method in optical variable attenuator
Abstract
The present invention provides a wavelength dependent compensation method
in a variable optical attenuator which can precisely set the amount of
attenuation within a wide range of wavelengths, and to provide superior
operability. An optical signal 20a is emitted from a white light source 20
incident on a monochromator 21, and under control of a CPU 24, only an
optical signal 20b of a specific wavelength is extracted and is made
incident on a variable optical attenuator 22; in optical attenuator 22
under control of a CPU 24, optical signal 20b is attenuated; the power of
attenuated optical signal 20c is measured by power measurer 23 under
control of a CPU 24, producing compensating data; and using the pulse
number corresponding to the rotation angle of the motor which activates
the variable optical attenuator 22 according to this compensating data,
the position of variable optical attenuator 22 is compensated to attain
the desired amount of attenuation.
| Inventors:
|
Matsumoto; Yoshinori (Tokyo, JP);
Sugimura; Hiroyuki (Tokyo, JP)
|
| Assignee:
|
Ando Electric Co., Ltd. (Tokyo, JP)
|
| Appl. No.:
|
043687 |
| Filed:
|
March 26, 1998 |
| PCT Filed:
|
July 29, 1997
|
| PCT NO:
|
PCT/JP97/02616
|
| 371 Date:
|
March 26, 1998
|
| 102(e) Date:
|
March 26, 1998
|
| PCT PUB.NO.:
|
WO98/04896 |
| PCT PUB. Date:
|
February 5, 1998 |
Foreign Application Priority Data
| Intern'l Class: |
G02B 006/00; H04B 010/00 |
| Field of Search: |
385/140,147
359/153,161,187,194
|
References Cited
U.S. Patent Documents
| 4516827 | May., 1985 | Lance et al. | 385/140.
|
| 4591231 | May., 1986 | Kaiser et al. | 385/140.
|
| 5226104 | Jul., 1993 | Unterleitner et al. | 385/140.
|
| 5325459 | Jun., 1994 | Schmidt | 385/140.
|
| 5432875 | Jul., 1995 | Korkowski et al. | 385/27.
|
| 5642456 | Jun., 1997 | Baker et al. | 385/140.
|
| 5694512 | Dec., 1997 | Gonthier et al. | 385/140.
|
| 5742725 | Apr., 1998 | Longobardi et al. | 385/140.
|
| 5745271 | Apr., 1998 | Ford et al. | 359/130.
|
| 5805759 | Sep., 1998 | Fukushima | 385/140.
|
| 5900983 | May., 1999 | Ford et al. | 385/140.
|
| 5963291 | Oct., 1999 | Wu et al. | 385/140.
|
| Foreign Patent Documents |
| 60-237328 | Nov., 1985 | JP | 385/140.
|
| 9-145540 | Jun., 1997 | JP | 385/140.
|
Primary Examiner: Healy; Brian
Attorney, Agent or Firm: Fish & Richardson P.C.
Claims
What is claimed is:
1. A wavelength dependent compensation method in a variable optical
attenuator, comprising:
a step in which an optical signal is emitted from a variable power optical
emission means;
a step in which an optical signal of a specific wavelength in said optical
signal is attenuated by an optical signal attenuation means that sets the
value of a predetermined wavelength and the value of the amount of
attenuation of the attenuated wavelength; and
a step in which the optical signal attenuated by the said optical signal
attenuation means is measured by a measuring means, and a step of
outputting compensating data supplying wavelength dependent loss
compensation to said optical signal attenuation means is output.
2. A wavelength dependent compensation method in a variable optical
attenuator according to claim 1, further comprising:
a step in which the wavelength dependency of the amount of attenuation is
compensated based on said compensation data to control said amount of
attenuation and the amount of attenuation of selected wavelengths.
3. A wavelength dependent compensation method in a variable optical
attenuator according to claim 1, wherein:
said optical signal attenuation means including a monochromator on which
the optical signal emitted from said optical emission means, and the step
in which said optical signal is attenuated further comprising a step in
which said optical signal of specific wavelength is extracted by said
monochromator.
4. A wavelength dependent compensation method in a variable optical
attenuator according to claim 3 wherein:
said optical emission means is a white light source, and the step in which
said optical signal is attenuated extracting a preset specific wavelength
from a plurality of wavelength components emitted from said white light
signal.
5. A wavelength dependent compensation method in a variable optical
attenuator according to claim 3 wherein:
said optical emission means being a semiconductor laser, and the step in
which said optical signal is attenuated extracting a preset specific
wavelength from a plurality of wavelength components emitted from said
white light source according to a preset value of said wavelength.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a wavelength-dependent compensation method
in a variable optical attenuator, and in particular to a
wavelength-dependent compensation method in a variable optical attenuator
which attenuates the wavelength of an optical signal over a wide range of
wavelengths in the optical signal, measures the power of the optical
signal with attenuated wavelengths, and supplies compensation data which
compensates wavelength-dependent loss.
2. Description of Related Art
It is known that the amount of attenuation of a natural density filter
(herebelow, abbreviated "ND filter") generally used to suppress the amount
of light in a variable optical attenuator depends on wavelength
characteristics. In order to achieve an accurate setting of the amount of
attenuation, it is necessary to limit attenuation to a single wavelength,
to actuate an ND filter with a motor drive member, to produce rotation of
the motor by the input of the pulse of its motor drive, and either to
compensate only the number of pulses corresponding to the angle of
rotation of the motor with respect to the amount of attenuation, or to
compensate the entire range of wavelengths to be set by using compensating
data for the amount of attenuation.
The method of compensating wavelength-dependency of an ND filter used in
conventional variable optical attenuators is explained in FIG. 9.
FIG. 9 is a block diagram showing a structure of a variable optical
attenuator applying a conventional compensation method for
wavelength-dependency. In FIG. 9, an optical signal 6a that transits ND
filter 1 is attenuated so as to equal the preset value of the amount of
attenuation set by ND filter 1 and is transformed into attenuated optical
signal 6b. The motor 2 is actuated by motor drive member 3 under control
of a CPU 4.
If the wavelength of optical signal 6a transiting ND filter 1 is uniform,
the amount of attenuation of the ND filter 1 is proportional to the angle
of rotation, and the angle of rotation of motor 2 is proportional to the
pulse number input into the motor drive member 3.
However, as is clear from the wavelength-characteristics graph of an ND
filter 1 shown in FIG. 2, where the abscissa is the pulse number and the
ordinate is the amount of attenuation (dB), generally in the attenuation
of an ND filter 1, the correction of the amount of attenuation is
conventionally limited to a single wavelength (e.g., 1310 nm) because
there is wavelength dependency.
Therefore, an accurate amount of attenuation of an optical signal differing
from the wavelength used for correction (e.g., 1310 nm) could not be
attained.
Because of this, when precisely setting the amount of attenuation with a
conventional variable optical attenuator, it is necessary to measure for
each wavelength, the relationship between the amount of attenuation of an
ND filter 1 and the number of pulses corresponding to the angle of
rotation of motor 2, and store the compensating data for each wavelength
in the compensating ROM 5 (see FIG. 9) under the control of a CPU 4.
However, as can be understood from the linear graph of an ND filter 1 in
FIG. 2, used when we explain below the embodiments of the present
invention, at each wavelength the relationship between the amount of
attenuation of an ND filter 1 and the number of pulses corresponding to
the angle of rotation of motor 2 are wavelength dependent (i.e., the
linear slope for each wavelength is different).
In addition, from the wavelength characteristic graph of an ND filter in
FIG. 3, where the abscissa is the wavelength (nm) and the ordinate is the
amount of attenuation (dB), the problem arises that even when the preset
value of the amount of attenuation is near 0 dB (insertion loss), there is
wavelength-dependency.
Therefore, in the conventional variable optical attenuators, in accurately
presetting the amount of attenuation with respect to an ND filter 1, the
wavelength is limited to a single wavelength, and either only the pulse
number corresponding to the angle of rotation (see FIG. 9) of motor 2 is
compensated, or the entire range of wavelengths to be preset are
compensated using compensating data of the pulse number corresponding to
the angle of rotation of motor 2 with respect to the amount of attenuation
of an ND filter 1.
However, in this case, for an optical signal whose wavelength differs from
the wavelength (in FIG. 2, 1310 nm) used in correction, accurate amounts
of attenuation cannot be obtained, and the problems arise that the number
of observation operations for compensating data becomes huge, the amount
of memory allocation in the compensation ROM 5 for the variable optical
attenuator becomes very large, and the cost becomes very high.
SUMMARY OF THE INVENTION
In order to solve the above mentioned conventional problems, the method of
wavelength dependent compensation in the variable optical attenuator of
the present invention is characterized in comprising a step wherein an
optical signal is generated from an optical emission means so that the
power and wavelength are variable, a step wherein a specific wavelength in
this optical signal is attenuated by an optical signal attenuation means
which sets the value of a pre-determined wavelength and the preset value
of the amount of attenuation of the attenuated wavelength, and a step
inputting the compensation data provided by a measurement means which
measures the optical signal attenuated with this optical signal
attenuation means for the wavelength dependent loss compensation in this
optical attenuation means.
According to this invention, by entering an optical signal produced from
the optical emission means, which can vary the power and wavelength, into
the optical signal attenuation means, a pre-determined wavelength of the
optical signal is attenuated, the power of the attenuated optical signal
is measured by the measuring means, the wavelength dependent compensation
data of the optical signal attenuation means is input, and this
compensation data supplies the wavelength dependent loss compensation.
According to the present invention, it is possible to precisely set the
amount of attenuation over a wide range of wavelengths, and achieve the
effect of superior adjustability because 1) a specific wavelength of the
optical signal from those in the optical signal generated by the optical
emission means is attenuated by the optical signal attenuation means, 2)
the power of the attenuated optical signal is measured with a measuring
means so as to obtain compensation data of the system of measurement of
the wavelength dependency, and 3) this compensation data supplies the
wavelength dependent loss compensation.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram showing the structure of the system of
measurement of wavelength dependency applied to the first embodiment of
the wavelength dependent compensation method of the variable attenuator of
the present invention.
FIG. 2 is a graph of the linearity of the ND filter in the system of
measurement of wavelength dependency of FIG. 1 used in the variable
attenuator.
FIG. 3 is a graph showing the wavelength characteristics of the 1300 nm
wavelength benchmark time of an ND filter used in the variable attenuator
in the measurement system of the wavelength dependency of FIG. 1.
FIG. 4 is a graph of the regenerated wavelength characteristics taking the
preset value 0 dB of the amount of attenuation as the benchmark in order
to confirm the linearity of the amount of attenuation of the wavelength
characteristics of an ND filter used in the variable attenuator in the
system of measurement of the wavelength dependency of FIG. 1.
FIG. 5 is a block diagram showing the structure of the system of
measurement of wavelength dependency applied in the second embodiment of
the wavelength dependent compensation method in the variable optical
attenuator of the present invention.
FIG. 6 is a graph produced by calculating the wavelength dependency for
each wavelength up to the preset value of 60 dB based on the data of a
preset value of 10 dB for the amount of attenuation of the variable
optical attenuator in the system of measurement of wavelength dependency
of FIG. 5.
FIG. 7 is a graph of the total wavelength dependency including the
non-linear regions of the preset value 0 dB of the amount of attenuation
of the variable optical attenuator in the system of measurement of
wavelength dependency of FIG. 5.
FIG. 8 is a graph showing the wavelength characteristics of the
compensating pulse number with respect the wavelength of the variable
optical attenuator in the system of measurement of the wavelength
dependency of FIG. 5.
FIG. 9 is a block diagram showing the structure of a variable optical
attenuator used in the wavelength dependent compensation method in the
conventional variable optical attenuator.
DESCRIPTION OF PREFERRED EMBODIMENTS
Embodiment 1
Below, a first embodiment of a wavelength dependent compensation method in
a variable optical attenuator of the present invention is explained with
reference to the figures. FIG. 1 is a block diagram showing the structure
of the system of measurement of wavelength dependency applied in the first
embodiment.
In FIG. 1, a white light source which emits optical signal 20a having a
wide range of wavelengths is used as an optical emission means, and below,
in the first embodiment, our explanation will proceed with this white
light source 20 as the optical emission means.
Optical signal 20a emitted from white light source 20 is incident upon the
optical signal extraction means for extracting an optical signal 20b of a
specific wavelength. In this first embodiment, a monochromator 21 is used
as an optical signal extraction means, and in the following explanation,
we will proceed using this monochromator 21.
The optical signal 20b emitted from monochromator 21 is incident on a
variable optical attenuator used as an optical signal attenuation means,
such as an ND filter, (we will explain the first embodiment using variable
optical attenuator 22 below), and the attenuated optical signal from this
variable optical attenuator 22 is emitted as optical signal 20c.
Optical signal 20c is measured by optical power measurer 23 (this measuring
means is explained along with the optical power measurer 23 below) used as
a measurement means. The white light generator 20, the monochromator 21,
variable optical attenuator 22, and the optical power measurer 23 are each
connected by an optical fiber 26.
Additionally, monochromator 21, variable optical attenuator 22, and optical
power measurer 23 are connected by GP-IB25 (control bus) to a CPU, which
is a controlling means, (as a controlling means, we will use a CPU
designated 24 below), and the monochromator 21, variable optical
attenuator 22, and optical power measurer 23 are controlled by CPU 24.
Next, we will explain the operation of the first embodiment constructed as
described above. First, a preset value of the wavelength of the variable
optical attenuator 22 is set to 1310 nm, for example, and a preset value
of the amount of attenuation is set to 0 dB.
The optical signal 20a is emitted from the white light source incident on
the monochromator 21 under the above conditions. The optical signal 20b is
incident on the variable optical attenuator 22 (amount of attenuation
being 0 dB), the attenuated optical signal 20c is input into the optical
power measurer 23, and its power is measured. The repetition of the above
measurements is carried out at 1 dB steps of the preset values of the
amount of attenuation from 0 to 10 dB, with a benchmark wavelength preset
value 1310 dB of the variable optical attenuator 22.
In FIG. 3, the intervals (amount of attenuation) between b, c, d, e, f, g,
h, i, j, and k are equal, while the interval between a and b (amount of
attenuation) is different. Therefore, from b to k, the intervals are equal
and linear, while the interval between a and b, in contrast, is
non-linear. In order to confirm the linearity of the amount of attenuation
in the wavelength dependency, in FIG. 4, the benchmark of the preset value
of the amount of attenuation is set to 0 dB, the abscissa is wavelength
(nm) and the ordinate is the amount of attenuation 0 dB.
However, as is apparent from FIG. 3, in contrast to the optical signal with
wavelengths differing from the standard, even at the preset value of 0 dB
for the amount of attenuation, there is wavelength dependency.
In FIG. 4, because the intervals (amount of attenuation) between b, c, d,
e, f, g, h, and, j, and k are equal, the amount of attenuation as a
function of wavelength dependency is linear for preset values of the
amount of attenuation from 0 to 10 dB.
Next, we compensate the wavelength dependence of the preset value of the
amount of attenuation from 0.about.60 dB of the variable optical
attenuator 22.
The output level of the optical signal 20a emitted from the white light
source is a low -40 dB, and when the set value of the amount of
attenuation of the variable optical attenuator is made 60 dB, the optical
signal 20c attenuated by the variable optical attenuator 22 has an output
level of -100 dB, and could not be measured with the light power measuring
device.
Here, a semiconductor laser 60 (wavelength: 1310 nm or 1550 nm) with a
large output level compared to the white light source 20 was used. By
confirming the linearity of the amount of attenuation of the wavelength
dependency, the preset value of the amount of attenuation from 0 to 60 dB
of the variable optical attenuator 22 is compensated.
Embodiment 2
FIG. 5 is a block diagram showing the structure of a measurement system for
wavelength dependency applied to the second embodiment of the present
invention when using a semiconductor laser 60 which emits the above
optical signal 60 of the wavelength 1310 nm/1550 nm.
In FIG. 5, the large output level optical signal 60a emitted from
semiconductor laser 60 is incident on the variable optical attenuator 22
via optical fiber 26, is attenuated, and is emitted from variable optical
attenuator 22 as optical signal 60b.
The optical signal 60b attenuated by the above variable optical attenuator
22 is incident on the light power measurer 23 which is the measuring
means, and there the power of optical signal 60b is measured. The variable
optical attenuator 22 and the light power measurer 23 are connected
through the CPU 24 by GP-IB25, the controlling means, and is controlled.
In FIG. 5, the monochromator 21 of FIG. 1 is omitted.
As is clear from the above, by being constructed in this manner, the
optical signal 60a (wavelength 1310 nm/1550 nm both greater than 0 dB)
emitted from semiconductor 60, are attenuated (0.about.60 dB) by the
variable optical attenuator 22, and the power of the attenuated optical
signal 60b is measured by the light power measurer 23.
A pulse number corresponding to the angle of rotation of motor 2 for
actuating the ND filter in the variable optical attenuator shown in FIG. 9
is sent to the motor drive member 3, and the amount of attenuation of the
variable optical attenuator 22 is determined by controlling the rotation
angle of the motor. As a result of the above measurements, the filter
characteristics shown in FIG. 2 were obtained.
In FIG. 2, because the graphs of the above wavelengths 1310 nm and 1550 nm
are linear when the amount of attenuation is a function of the pulse
number corresponding to the rotation angle of the motor 2 shown in FIG. 9,
the two wavelengths 1310 nm and 1550 nm can be judged to be linear.
Therefore, the other wavelengths can also be assumed to be linear.
With respect to wavelengths of the above two 1310 nm and 1550 nm
wavelengths, because we judge that the amount of attenuation as a function
of the pulse number corresponding to the angle of rotation of motor 2
shown in FIG. 9 is linear, FIG. 6 showing the wavelength characteristics
of the amount of attenuation up to a preset value of 60 dB is produced on
the basis of the data of the amount of attenuation up to the preset value
10 dB in each wavelength of FIG. 4 (the data of the amount of attenuation
of the set value 10 dB is increased 6 times).
In fact, as is apparent from FIG. 3, the wavelength dependency of variable
optical attenuator 22 has not only regions where the amount of attenuation
as a function of the pulse number corresponding to the rotation angle of
the motor 2 is linear, but it also has non-linear regions in the amount of
attenuation of the set value 0 dB. Thus the general wavelength
characteristics are shown in the following equation (1):
A.lambda.=A.lambda..sub.0 +A.sub.TT .lambda., (1)
where,
A.lambda. is the general wavelength characteristic,
A.lambda..sub.0 is the wavelength dependency of variable optical attenuator
22 of the amount of attenuation of the time 0 dB of the preset value, and
A.sub.TT .lambda. is the wavelength dependency at the time that the amount
of attenuation with a preset value 0 dB was the benchmark.
Therefore, in consideration of the above content, taking the abscissa as
the wavelength, the ordinate as the amount of attenuation (dB), and a 1300
nm wavelength as the benchmark, the wavelength characteristics of the
filter with discrete values of the amount of attenuation from 20 to 60 dB
are obtained, as shown in FIG. 7.
Next, the formula for wavelength compensation of the pulse number will be
described.
(1) The equation of the pulse number as a function of the amount of
attenuation is
P.sub.131 =.alpha.A.sub.131 +b (2)
where
A.sub.131 is the amount of attenuation dB at the time wavelength .lambda.
is 1330 nm,
P.sub.131 is the pulse number in A.sub.131 at the time wavelength .lambda.
is 1310 nm, and
a, b are the constants found by the method of least squares at the time
wavelength .lambda. is 1310 nm.
(2) The formula for a continuous wavelength is
P.lambda.=.alpha.(c.lambda.+(d.lambda./10).multidot.A.lambda.+A.lambda.)+b
(3)
where
P.lambda. is the pulse number at time wavelength .lambda.,
c.lambda. is the loss in the light source at the time wavelength .lambda.,
and
d.lambda. is loss at the set value 10 dB of the amount of attenuation at
the time of wavelength .lambda..
As explained above, in the first and second embodiments, even when the
pulse number corresponding to the rotation angle of the motor as a
function of the amount of attenuation of a variable optical attenuator
such as an ND filter is measured for each wavelength, in any wavelength
the pulse number corresponding to the amount of attenuation of the
variable optical attenuator and the angle of rotation of the motor is
linear, and thus if this linearity is used, when the above equations (2)
and (3) are stored in the compensating ROM of the variable optical
attenuator, and only the pulse number corresponding to the angle of
rotation of the motor as shown in FIG. 8 is compensated, the amount of
attenuation can be precisely set over a wide range of wavelengths with
only a single wavelength, realizing a superior adjustment operability.
In the above manner, when the compensation data obtained by the present
invention is used on the variable optical attenuator shown in FIG. 9, the
compensating data is stored on compensating ROM 5 of FIG. 9 and this data
is calculated by the above equations (1) for the general wavelength
characteristics and (2) for wavelength compensation by pulse number are
calculated in CPU 4, and then the pulse number corresponding to the angle
of rotation of motor 2 sent to a motor drive 3, but by controlling the
angle of rotation of motor 2, an accurate amount of wavelength attenuation
can be obtained no matter what the wavelength.
To give a concrete example of obtaining an accurate amount of wavelength
attenuation, because in fact there is adjustment, the P.lambda. of the
above equation (3) for a continuous wavelength is
P.lambda.=a.multidot.(c.lambda..multidot.e+(d.lambda./10).multidot.A.lambda
..multidot.F+A.lambda.)+b=a.multidot.{(d.lambda./10.multidot..function.+1).
multidot.A.lambda.+c.lambda..multidot.e}+b
The equation for the pulse number with respect to the wavelength
.lambda.=1550 nm and the amount of attenuation set value 10 dB (f=3=1) is
P.sub.155
=a.multidot.{(0.38/10.multidot..function.+1).multidot.10+0.5.multidot.e}+b
=a.multidot.{(0.38/10+1).multidot.10+0.5+b=10.88.multidot.a+b
Therefore, calculating the wavelength as 1310 nm makes the wavelength
attenuation 10.88 dB.
In this manner, taking the abscissa as the wavelength and the pulse number
as the ordinate, the wavelength characteristics of the compensated pulse
as in FIG. 8 showing the wavelength characteristics of the compensated
pulse number, no matter what the wavelength, an accurate amount of
attenuation can be attained from the compensated pulse number as shown in
FIG. 8.
Top