 |
Claims  |
|
|
What is claimed as new and desired to be secured by Letters Patent of the
United States:
1. A physical sensor comprising:
input optical signal means for generating an input optical signal, said
input signal including a first coherent light component at a first
wavelength and a second coherent light component at a second wavelength,
said first wavelength and said second wavelength differing by a wavelength
difference;
input coupler means for receiving said input optical signal and for
splitting said input optical signal into first and second beams;
sensing light path means for receiving said first beam from said input
coupler means and for transmitting said first beam, said sensing path
means having a first path length;
transducer means coupled to said sensing path means for modulating the
phase of said first beam being transmitted by said sensing path means in
response to a physical quantity;
reference light path means for receiving said second beam from said input
coupler means and for transmitting said second beam, said reference path
means having a second path length, said first and second path lengths
differing by a path length difference;
output coupler means for receiving said first beam transmitted by said
sensing path means, for receiving said second beam transmitted by said
reference path means, for combining said first and second beams, and for
splitting said combined first and second beams into third and fourth
beams;
first detector means coupled to receive said third beam for producing a
first electrical output signal proportional to the intensity of said third
beam, said first output signal including a first signal component due to
said first light component at said first wavelength and a second signal
component due to said second light component at said second wavelength,
the amplitudes of said first and second signal components being
proportional to said physical quantity, the phase of said first signal
component differing from the phase of said second signal component by a
phase difference, said phase difference being proportional to said
wavelength difference and said path length difference;
phase adjustment means comprising wavelength control of said first and
second wavelengths for adjustment of said phase difference; and
comparator means for distinguishing and comparing said first and second
signal components;
whereby said phase difference may be adjusted such that the sensitivities
of said first and second signal components to said physical quantity are
never simultaneously at a minimum.
2. The physical sensor as recited in claim 1, which further comprises:
second detector means coupled to receive said fourth beam for producing a
second electrical output signal proportional to the intensity of said
fourth beam, said second output signal being out of phase with said first
output signal, said second output signal including a third signal
component due to said first light component at said first wavelength and a
fourth signal component due to said second light component at said second
wavelength, the amplitudes of said third and fourth signal components
being proportional to said physical quantity, the phase of said third
signal component differing from the phase of said fourth signal component
by said phase difference.
3. The physical sensor as recited in claim 1, wherein:
said sensing path means comprises a first optical fiber coupled between
said input coupler means and said output coupler means; and
said reference path means comprises a second optical fiber coupled between
said input coupler means and said output coupler means.
4. The physical sensor as recited in claim 3, wherein said phase difference
(.DELTA..phi.) is given by:
.DELTA..phi..apprxeq.(-2.pi.N.sub.eff /.lambda..sup.2)[S(.DELTA..lambda.)]
where:
N.sub.eff is the effective refractive index of said first and second
optical fibers;
.lambda. is said first wavelength;
.DELTA..lambda. is said wavelength difference; and
S is said path length difference.
5. The physical sensor as recited in claim 2, wherein:
said sensing path means comprises a first optical fiber coupled between
said input coupler means and said output coupler means; and
said reference path means comprises a second optical fiber coupled between
said input coupler means and said output coupler means.
6. The physical sensor are recited in claim 5, wherein said phase
difference (.DELTA..phi.) is given by:
.DELTA..phi..apprxeq.(-2.lambda.N.sub.eff
/.lambda..sup.2)[S(.DELTA..lambda.)]
where:
N.sub.eff is the effective refractive index of said first and second
optical fibers;
.lambda. is said first wavelength;
.DELTA..lambda. is said wavelength difference; and
S is said path length difference.
7. The physical sensor as recited in claim 1, wherein said comparator means
further comprises:
means for separating said first and second signal components from said
first output signal.
8. The physical sensor as recited in claim 7, wherein said separating means
comprises:
first modulator means for modulating said first light component of said
input optical signal with a signal having a third wavelength;
second modulator means for modulating said second light component of said
input optical signal with a signal having a fourth wavelength;
first band pass filter means centered at said third wavelength for
receiving said first output signal from said first detector means for
separating said first signal component from said first output signal; and
second band pass filter means centered at said fourth wavelength for
receiving said first output signal from said first detector means for
separating said second signal component from said first output signal.
9. The physical sensor as recited in claim 8, wherein:
said sensing path means comprises a first optical fiber coupled between
said input coupler means and said output coupler means; and
said reference path means comprises a second optical fiber coupled between
said input coupler means and said output coupler means.
10. The physical sensor as recited in claim 9, wherein said phase
difference (.DELTA..phi.) is given by:
.DELTA..phi..apprxeq.(-2.pi.N.sub.eff /.lambda..sup.2)[S(.DELTA..phi.)]
where:
N.sub.eff is the effective refractive index of said first and second
optical fibers;
.lambda. is said first wavelength;
.DELTA..lambda. is said wavelength difference; and
S is said path length difference.
11. The physical sensor as recited in claim 2, which further comprises:
means for separating said first and second signal components from said
first output signal and for separating said third and fourth signal
components from said second output signal.
12. The physical sensor as recited in claim 11, wherein said separating
means comprises:
first modulator means for modulating said first light component of said
input optical signal with a signal having a third wavelength;
second modulator means for modulating said second light component of said
input optical signal having a fourth wavelength;
first band pass filter means centered at said third wavelength for
receiving said first output signal from said first detector means for
separating said first signal component from said first output signal;
second band pass filter means centered at said fourth wavelength for
receiving said first output signal from said first detector means for
separating said second signal component from said first output signal;
third band pass filter means centered at said third wavelength for
receiving said second output signal from said second detector means for
separating said third signal component from said second output signal; and
fourth band pass filter means centered at said fourth wavelength for
receiving said second output signal from said second detector means for
separating said fourth signal component from said second output signal.
13. The physical sensor as recited in claim 12, wherein:
said sensing path means comprises a first optical fiber coupled between
said input coupler means and said output coupler means; and
said reference path means comprises a second optical fiber coupled between
said input coupler means and said output coupler means.
14. The physical sensor as recited in claim 13, wherein said phase
difference (.DELTA..phi.) is given by:
.DELTA..phi..apprxeq.(-2.pi.N.sub.eff /.lambda..sup.2)[S(.DELTA..lambda.)]
where:
N.sub.eff is the effective refractive index of said first and second
optical fibers;
.lambda. is said first wavelength;
.DELTA..lambda. is said wavelength difference; and
S is said path length difference.
15. The physical sensor as recited in claim 7, wherein said separating
means comprises:
means for alternatively applying said first light component at said first
wavelength and said second light component at said second wavelength to
said input coupler means;
said first output signal including said first signal component when said
first light component is applied to said input coupler means and including
said second signal component when said second light component is applied
to said input coupler means.
16. The physical sensor as recited in claim 15, wherein said means for
alternatively applying said first and second light components comprises:
first coherent source means for generating said first light component at
said first wavelength and for supplying said first light component to an
input of said input coupler means;
second coherent source means for generating said second light component at
said second wavelength and for supplying said second light component to
said input of said input coupler means;
signal level comparator means coupled to receive said first output signal
from said first detector means for determining when the sensitivity of
said first and second signal components to said physical quantity is
maximized and for producing a control signal in response to this
determination; and
logic means for alternatively energizing said first and second source means
in response to said control signal;
whereby the sensitivity of said output signal to said physical quantity is
maximized.
17. The physical sensor as recited in claim 16, wherein:
said sensing path means comprises a first optical fiber coupled between
said input coupler means and said output coupler means; and
said reference path means comprises a second optical fiber coupled between
said input coupler means and said output coupler means.
18. The physical sensor as recited in claim 17, wherein said phase
difference (.DELTA..phi.) is given by:
.DELTA..phi..apprxeq.(-2.pi.N.sub.eff /.lambda..sup.2)[S(.DELTA..lambda.)]
where:
N.sub.eff is the effective refractive index of said first and second
optical fibers;
.lambda. is said first wavelength;
.DELTA..lambda. is said wavelength difference; and
S is said path length difference.
19. A physical sensor comprising:
input optical signal means for generating an input optical signal including
a first coherent light component at a first wavelength and a second
coherent light component at a second wavelength, said first and second
wavelengths differing by a wavelength difference;
a Mach-Zehnder interferometer including an input port, an output port, a
sensing light path, transducer means coupled to said sensing path for
modulating the phase of light signals passing through said sensing path in
response to a physical quantity, and a reference light path, the path
length of said reference path differing from the path length of said
sensing path by a path length difference, said input port receiving said
input optical signal;
detector means coupled to receive an output optical signal emanating from
said output port for producing an output electrical signal proportional to
the intensity of said output optical signal, said output electrical signal
including a first signal component due to said first light component at
said first wavelength and a second signal component due to said second
light component at said second wavelength, the amplitudes of said first
and second signal components being proportional to said physical quantity,
the phase of said first signal component differing from the phase of said
second signal component by a phase difference, said phase difference being
proportional to said wavelength difference and said path length
difference;
phase difference adjustment means comprising wavelength control of said
first and second wavelengths for adjustment of said wavelength difference;
and
comparator means for distinguishing and comparing said first and second
signal components;
whereby said phase difference may be adjusted such that the sensitivities
of said first and second signal components to said physical quantity are
never simultaneously at a minimum. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
BACKGROUND OF THE INVENTION
The present Invention relates in general to novel fiber optic
interferometric physical sensors and, more particularly, to novel fiber
optic interferometric sensors which utilize light energy inputs at more
than one wavelength.
Guided wave interferometers using single-mode fibers have been proven to be
useful tools in detecting various physical quantities such as acoustic
waves, magnetic fields, and electric fields. In order to achieve high
sensitivity, the arm lengths of the interferometers are made long (>10
meters). The long arm lengths, however, cause large phase drifts in the
output signals, mainly due to changes in ambient temperature and pressure.
The optical output (P.sub.1) from an interferometer is represented by
P.sub.1 =C.sub.1 +C.sub.2 cos .phi.(t), where C.sub.1 and C.sub.2 are
constants and .phi. is the phase difference between two waves propagating
along the two interferometer arms, respectively. The differential phase
.phi. is composed of two parts: one part due to signal (.phi..sub.S) and
the other part due to noise (.phi..sub.N). When the ambient temperature
changes, for example, .phi..sub.N drifts at a rate of about 18000 degrees
per meter per degree (.degree.C.). Since the sensitivity of the
interferometer is proportional to .vertline.dP.sub.1
/d.phi..vertline.=.vertline.C.sub.2 sin .phi.(t).vertline., the
sensitivity fluctuates wildly with temperature change. This causes signal
fading problems and results in system down-time due to lack of sensor
sensitivity which occurs at the condition .vertline.dP.sub.1
/d.phi..vertline.=0 when .phi.=m.pi., where m is an integer.
This problem has been approached in various ways. One solution utilizes a
feedback scheme in which a piezoelectric cylinder is wound with a
plurality of turns of optical fiber. A feedback signal is applied to the
cylinder to cause stresses in the optical fiber resulting in optical path
length changes in the fiber proportional to the feedback signal. Such a
system is described by D. A. Jackson et.al. "Elimination of Drift in a
Single Mode Optical Fiber Interferometer Using a Piezoelectrically
Stretched Coiled Fiber", Applied Optics, Vol. 19, No. 17, pp. 2926-2929.
The disadvantage of this approach is that it requires bulky piezoelectric
cylinders which add volume and weight to the device. Other prior art
solutions have similar disadvantages.
SUMMARY OF THE INVENTION
Accordingly, one object of the present Invention is to provide a novel
physical sensor.
Another object is to provide a novel physical sensor utilizing a fiber
optic interferometer.
Still another object is to provide a novel physical sensor utilizing a
fiber optic interferometer which exhibits high sensitivity while avoiding
sensitivity variations due to phase drift.
These and other objects and advantages are provided by a novel physical
sensor according to the present Invention which includes a Mach-Zehnder
interferometer having an input port, an output port, a sensing light path,
and a reference light path. A transducer is coupled to the sensing path
for modulating the phase of light signals passing through the sensing path
in response to a physical quantity. The path lengths of the sensing and
reference paths differ by a path length diffference. The input port is
supplied with an input optical signal including a first light component at
a first wavelength and a second light component at a second wavelength.
The first and second wavelengths differ by a wavelength difference. A
detector is coupled to receive an output optical signal emanating from the
output port for producing an output electrical signal proportional to the
intensity of the output optical signal. The output electrical signal
includes a first signal component due to the first light component at the
first wavelength and a second signal component due to the second light
component at the second wavelength. The amplitudes of the first and second
signal components are related to the physical quantity. The phases of the
first and second signal components differ by an amount proportional to the
wavelength difference between the first and second light components of the
input optical signal and the path length difference between the sensing
and reference paths. The phase difference may be adjusted such that the
sensitivities of the first and second signal components to the physical
quantity are never simultaneously at a minimum.
BRIEF DESCRIPTION OF THE DRAWINGS
A more complete appreciation of the present Invention and many of the
attendant advantages thereof will be readily obtained as the same becomes
better understood by reference to the following detailed description when
considered in connection with the accompanying drawings, wherein:
FIG. 1 illustrates an interferometer according to a preferred embodiment of
the present Invention;
FIG. 2 illustrates output curves for the interferometer shown in FIG. 1
plotted as a function of phase;
FIG. 3 illustrates output curves for the interferometer shown in FIG. 1
plotted as a function of phase for a particular phase relationship between
the outputs;
FIG. 4 illustrates an interferometer according to another preferred
embodiment of the present invention which makes use of the phase relation
shown in FIG. 3;
FIG. 5 illustrates an output signal normalization system for use with the
preferred embodiment of the interferometer according to the present
Invention shown in FIG. 4;
FIG. 6 illustrates an interferometer according to still another preferred
embodiment of the present Invention;
FIG. 7 illustrates a signal processing system for use with the
interferometer shown in FIG. 6;
FIG. 8 illustrates output curves plotted as a function of phase for various
input wavelengths for an interferometer according to the present
Invention; and
FIG. 9 illustrates an interferometer according to yet another preferred
embodiment of the present Invention which makes use of the output
relationships shown in FIG. 8.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to the drawings, wherein like reference numerals designate
identical or corresponding parts throughout the several views, and more
particularly to FIG. 1 thereof, an all fiber optic MACH-ZENDER-type
interferometer 10 is illustrated as including a single mode optical fiber
sensing arm 12 and a single mode optical fiber reference arm 14 coupled
between a first optical coupler 16 and a second optical coupler 18. The
sensing arm 12 includes a sensing portion 20 which may take the form of a
coil or other appropriate configuration, as is known in the art. The
sensing portion 20 is subjected to a physical quantity F which acts to
produce proportional changes in the effective optical length of the
sensing position 20 and thus in the effective optical length of the
sensing arm 12 in response thereto.
The optical couplers 16 and 18 are preferably all fiber couplers of a type
known as "bottle" couplers which couple optical signals traveling in
optical fibers by means of evanescent wave coupling. The coupling
coefficient of these devices can be adjusted by adjusting various device
parameters. In the present Invention 3 db coupling is preferably used.
Such "bottle" couplers are the subject of U.S. Pat. No. 4,264,126 and U.S.
Application Ser. No. 217,338 filed by S. K. SHEEM on Dec. 17, 1980, now
abandoned. Additionally these devices are described in S. K. SHEEM and T.
G. GIALLORENZI, "Single-Mode Fiber-Optical Power Divider: Escapsulated
Etching Technique", Optics Letters, Vol. 4, No. 1, January 1979, pages
29-31. The teachings of these documents are specificantly incorporated
herein by reference. Other optical coupling means can also be utilized as
should be apparant to the skilled reader.
An input laser light signal P.sub.IN (.lambda.) is coupled to a first input
of the first optical coupler 16 by an optical fiber 22 while a second
laser light input signal of a slightly different wavelength P'.sub.IN
(.lambda.') is coupled to a second input of the coupler 16 by an optical
fiber 24. The expression P.sub.IN (.lambda.) refers to the optical power
coupled into the first input of the first coupler at a wavelength
.lambda.. Similarly, P'.sub.IN (.lambda.') refers to the optical power
coupled into the second input of the first coupler at a second wavelength
.lambda.'. Alternatively, the two input light signals could be jointly
coupled to either the first input or the second input of the first optical
coupler 16. These light signals could be supplied by separate sources or
by a common source. The optical coupler 16 forms a beam splitter which
divides each of the inputs P.sub.IN (.lambda.) and P'.sub.IN (.lambda.')
into two, respectively, and outputs the divided signals into the sensing
arm fiber 12 and into the reference arm fiber 14, respectively. Thus each
of the split beams travels along the two paths formed by the sensor and
reference arms.
While traveling, each beam experiences a phase shift .phi.(.lambda.) where:
.phi.(.lambda.)=-2.pi.SN.sub.eff (.lambda.)/.lambda. (1)
The derivation of equation (1) is presented by S. K. SHEEM and R. P.
MOELLER in "Single Mode Fiber Wavelength Multiplexer", Journal of Applied
Physics, Vol. 51, No. 7, August 1980, pp. 4050-4052. The teachings of
this document are specifically incorporated herein by reference.
In equation (1), S is the difference between the lengths of the
interferometer sensor arm 12 and the reference arm 14, and N.sub.eff is
the effective index of refraction for the particular guided mode carrying
the optical beam. Here we assume that the fibers 12 and 14 are identical.
This does not cause any loss in the generality of this discussion because
when different fibers having different lengths are used, the quantity
SN.sub.eff in equation (1) can be replaced by the quantity L(14)N.sub.eff
(14)=L(12)N.sub.eff (12), where L(14) and L(12) are the optical path
lengths of the reference arm 14 and the signal arm 12, respectively.
Therefore, it is clear that the phase of the light signals traveling in
the interferometer is a function of wavelength. Thus:
.phi.(.lambda.)=-2.pi.SN.sub.eff (.lambda.)/.lambda. (2)
.phi.(.lambda.')=-2.pi.SN.sub.eff (.lambda.')/.lambda.' (3)
When the two optical signals at wavelengths .lambda. and .lambda.' arrive
at the second optical coupler beam splitter 18 after traveling over the
reference arm 14 and the sensing arm 12, they interfere together to
produce two output signals P.sub.1 (.lambda.) and P.sub.1 '(.lambda.'),
which are coupled to a first intensity detector 26 by means of a fiber 28,
and to produce another two output signals P.sub.2 (.lambda.) and P.sub.2
'(.lambda.') which are coupled to a second intensity detector 30 by means
of a fiber 32. Thus:
P.sub.1 (.lambda.)=(P/2)[1+cos .phi.(.lambda.)] (4)
P.sub.1 '(.lambda.')=(P'/2)[1+cos .phi.(.lambda.')] (5)
P.sub.2 (.lambda.)=P-P.sub.1 (.lambda.)=(P/2)[1-cos .phi.(.lambda.)](6)
P.sub.2 '(.lambda.')=P'-P.sub.1 '(.lambda.')=(P'/2)[1-cos
.phi.(.lambda.')](7)
where:
P=P.sub.IN (.lambda.)-P.sub..alpha. (8)
P'=P'.sub.IN (.lambda.')-P'.sub..alpha. (9)
with P.sub..alpha. and P'.sub..alpha. being the optical losses for each
wavelength signal in the interferometer 10.
We may define .phi.(.lambda.') as follows:
.phi.(.lambda.')=.phi.(.lambda.)+.DELTA..phi. (10)
Combining equation (10) and equations (2) and (3), we obtain:
##EQU1##
We note that N.sub.eff is practically constant with respect to .lambda. if
the difference in wavelengths between the two signals P.sub.IN (.lambda.)
and P'.sub.IN (.lambda.') is small, eg. (.lambda.-.lambda.')<100 .ANG..
Thus, for this condition, equation (11) may be approximated by:
##EQU2##
and,
.DELTA..phi..apprxeq.(-2.pi.N.sub.eff
/.lambda..sup.2)[5(.DELTA..lambda.)](13)
where:
.DELTA..lambda.=(.lambda.-.lambda.').
Thus, for given values of N.sub.eff and .lambda., we can determine the
product S(.DELTA..lambda.) which gives a desired value of .DELTA..phi..
For example, if .DELTA..phi. is conveniently selected to be .pi./2,
equations (4) through (7) become:
P.sub.1 (.lambda.)=(P/2)[1+cos .phi.] (14)
P.sub.1 '(.lambda.')=(P'/2)[1-sin .phi.] (15)
P.sub.2 (.lambda.)=(P/2)[1-cos .phi.] (16)
P.sub.2 '(.lambda.')=(P'/2)[1+sin .phi.] (17)
Equations (14) and (15) are drawn in FIG. 2 as a function of .phi..
Equations (16) and (17) are identical to equations (14) and (15) except
for a phase shift of .pi. (due to the optical coupler (18), and thus have
not been illustrated.
The phase .phi. is composed of two parts: .phi..sub.F due to the physical
quantity F and .phi..sub.N due to noise. Thus:
.phi.=.phi..sub.F +.phi..sub.N (18)
Usually, .phi..sub.N is very large while .phi..sub.F is very small.
Actually, .phi..sub.N can be much larger than .pi., in which case the
outputs P.sub.1 (.lambda.) and P.sub.1 '(.lambda.') swing along the curves
shown in FIG. 2. .phi..sub.F, being small, produces small differential
changes along the curves shown in FIG. 2.
The sensitivity of the interferometer 10 is proportional to dP/d.phi.. This
is illustrated graphically in FIG. 2. When the operating point is sitting
at the quadrature point A along curve P.sub.1 '(.lambda.') (as determined
by the noise .phi..sub.N), the sensitivity dP.sub.1 '(.lambda.')/d.phi. is
at a maximum (F.sub.OUT /F.sub.IN is maximized), while at the same value
of .phi..sub.N the operating point is at point B along curve P.sub.1
(.lambda.) and the sensitivity dP.sub.1 (.lambda.)/d.phi. is at its
minimum (F.sub.OUT /F.sub.IN is minimized). However, since P.sub.1
(.lambda.) and P.sub.1 '(.lambda.') have the relative phase relationship
as represented by equations (14) and (15) and as shown in FIG. 2, the
sensitivities of the interferometer for signals at wavelengths .lambda.
and .lambda.' are mutually complimentary. Thus, an output signal is
available from the interferometer 10 at all times, regardless of the noise
.phi..sub.N.
FIG. 3 illustrates a method for taking advantage of the relationship
between P.sub.1 (.lambda.) and P.sub.1 '(.lambda.') when
.DELTA..phi.=.pi./2. From the curves of P.sub.1 (.lambda.) and P.sub.1
'(.lambda.') it can be seen that at any given phase value .phi., the
operating point lies along the "linear" portion of one of the curves
P.sub.1 (.lambda.) or P.sub.1 '(.lambda.'). Thus, in region I, P.sub.1
(.lambda.) is essentially linear; in region II, P.sub.1 '(.lambda.') is
essentially linear; in region III, P.sub.1 (.lambda.) is essentially
linear; and so on. This may be accomplished by alternately switching laser
inputs P.sub.IN (.lambda.) and P.sub.IN '(.lambda.') on and off at
appropriate times so as to select the appropriate curve P.sub.1 (.lambda.)
or P.sub.1 '(.lambda.') for any given phase angle. In this way there is
always an output signal from the system although the sensitivity will vary
even along the "linear" portions of the curves. Although the above
discussion has been with respect to the curves P.sub.1 (.lambda.) and
P.sub.1 '(.lambda.'), it should be understood that an exactly analogous
situation exists with respect to the curves P.sub.2 (.lambda.) and P.sub.2
'(.lambda.').
FIG. 4 illustrates a block diagram of a system for implementing the method
shown in FIG. 3. In FIG. 4, a first laser 50 operating at a wavelength
.lambda. and a second laser 52 operating at a wavelength .lambda.' supply
inputs P.sub.IN (.lambda.) and P'.sub.IN (.lambda.'), respectively, to the
input fibers 22 and 24, respectively, of the interferometer 10. The first
laser 50 and the second laser 52 are turned on and off by signals
appearing on control lines 54 and 56, respectively. The interferometer 10
and other correspondingly numbered parts are as shown in FIG. 1 and as
described above. As previously described, both laser inputs could be
supplied to the same input of the interferometer 10.
The outputs of the interferometer 10, appearing at the outputs of fibers 28
and 32, are detected by photodetectors 26 and 30, respectively. Thus, when
the first laser 50 (.lambda.) is turned on, photodetector 26 detects
P.sub.1 (.lambda.) and photodetector 30 detects P.sub.2 (.lambda.).
Similarly, when the second laser 52 (.lambda.') is turned on, P.sub.1
'(.lambda.') and P.sub.2 '(.lambda.') are detected by the photodetectors
26 and 30, respectively.
The output of the photodetectors 26, P.sub.1 (.lambda.) or P.sub.1
'(.lambda.'), is filtered by a low pass filter 58 to eliminate any signal
components due to the detected physical quantity F. The output of the low
pass filter 58 is fed to two comparators 60 and 62 which compare the
filtered signal to a voltage representative of the upper and lower limits,
respectively, as illustrated in FIG. 3. The outputs of the comparators 60
and 62 are fed to a logic circuit 64 which analyses the comparator signals
and determines which laser, 50 or 52, to turn on via the control lines 54
and 56. Thus laser 50 is energized during periods I and III, while laser
52 is energized during period II, as illustrated in FIG. 3. Therefore,
output signals representative of the detected physical quantity F are
present at the outputs of the photodetectors 26 and 30 at all times.
Although the system of FIG. 4 illustrates the use of the output signal of
photodetector 26 to develop control signals for the lasers 50 and 52, the
output of the photodetector 30 could also be used, as should be apparent
to the skilled reader.
The logic circuit 64 is a simple logic implementation of a simple truth
table based on the four possible output combinations of the comparators 60
and 62. An initialization function should also be included. The design of
this unit is fully within the capabilities of any person of skill in the
art. The remaining portions of the system shown in FIG. 4 are readily
available standard hardware items.
The outputs of the photodetectors 26 and 30 are fed to a differential
amplifier 66 which outputs a signal proportional to P.sub.1
(.lambda.)-P.sub.2 (.lambda.) or P.sub.1 '(.lambda.')-P.sub.2
'(.lambda.'), depending upon which the particular laser 50 or 52 which is
energized. The output signal is passed through a high pass filter 68 to
remove the spectral components due to environmental perturbations (noise).
The remaining signal, representative of the physical quantity F, can be
fed to a spectrum analyzer 70 for analysis or fed to a computer 74 via an
analog-to-digital converter 72. Other means for analysing the output
signal should be obvious to the skilled reader.
The system of FIG. 4 provides an output at all times; however, as
previously described, the sensitivity of the interferometer varies
depending upon the actual operating point of the interferometer along the
linear portions of the curves P.sub.1 (.lambda.) or P.sub.1 '(.lambda.')
shown in FIG. 3. Thus, the amplitude of the output of the interferometer
varies as a function of the phase .phi., and in particular as a function
of the noise component of the phase .phi..sub.N. FIG. 5 illustrates a
normalization system for normalizing the sensitivity of interferometer
with respect to the operating point such that the sensitivity effective
remains constant with respect to the phase noise .phi..sub.N.
In FIG. 5, the normalization system receives, as its input, the output of
the differential amplifier 66 shown in FIG. 4. Assuming that the
interferometer is operating at the wavelength .lambda., from equations
(14), (16), and (18) the output of the differential amplifier 66 is as
follows:
P.sub.1 (.lambda.)-P.sub.2 (.lambda.)=P cos (.phi..sub.N +.phi..sub.F) (19)
If we assume the physical quantity F to be a sinusoidal function, equation
(19) can be rewritten as:
P.sub.1 (.lambda.)-P.sub.2 (.lambda.)=P cos [.phi..sub.N +C(.omega.) sin
.omega.t] (20)
where C(.omega.) is a function of the frequency .omega. of the physical
quantity F. Expanding equation (20), we obtain:
P.sub.1 (.lambda.)-P.sub.2 (.lambda.)=P cos .phi..sub.N cos [C(.omega.) sin
.omega.t]-P sin .phi..sub.N sin [C(.omega.) sin .omega.t] (21)
If we further assume F to be a small signal, the quantity cos [C(.omega.)
sin .omega.t] will be approximately equal to 1 and the quantity sin
[C(.omega.) sin .omega.t] will be approximately equal to C(.omega.) sin
.omega.t. Therefore, equation (21) becomes:
P.sub.1 (.lambda.)-P.sub.2 (.lambda.).apprxeq.P cos .phi..sub.N
-PC(.omega.) sin .phi..sub.N sin .omega.t (22)
Returning to FIG. 5, the input signal as represented by equation (22) is
separated into AC and DC components by means of a high pass filter 76 and
a low pass filter 78, respectively. Thus, the output of the high pass
filter 76 is:
[P.sub.1 (.lambda.)-P.sub.2 (.lambda.)].sub.AC .apprxeq.-PC(.omega.) sin
.phi..sub.N sin .omega.t (23)
This signal is rectified to obtain its absolute value via a rectifier 80.
The output of the low pass filter 78 is:
[P.sub.1 (.lambda.)-P.sub.2 (.lambda.)].sub.DC .apprxeq.P cos .phi..sub.N (
24)
The phase angle due to noise .phi..sub.N is calculated in an angle computer
circuit 82 as follows:
##EQU3##
The quantity sin .phi..sub.N is then computed via a sinusoidal calculator
circuit 84. The output of the rectifier 80 is then divided by sin
.phi..sub.N in a divider circuit 86 as follows:
[P.sub.1 (.lambda.)-P.sub.2 (.lambda.)].sub.AC /[P.sub.1 (.lambda.)-P.sub.2
(.lambda.)].sub.DC .apprxeq.PC(.omega.) sin .omega.t (26)
Thus, the output of the normalizer system is independent of the phase noise
.phi..sub.N and thus is independent of the operating point of the
interferometer along the linear portion of the curve P.sub.1 (.lambda.)
(and the curve P.sub.2 (.lambda.)-not illustrated). The normalization
system operates identically when the interferometer is operating at the
wavelength .lambda.'. The angle computer circuit 82, the sinusoidal
calculator circuit 84, and the divider circuit 86 may be implemented with
well known analog circuitry as should be obvious to the skilled
practitioner. Alternatively, these functions may be performed digitally or
via a computer.
Another method for taking advantage of the relationship between the curves
P.sub.1 (.lambda.) and P.sub.1 '(.lambda.') when .DELTA..phi.=.pi./2
involves supplying inputs to the interferometer at two appropriate
wavelengths .lambda. and .lambda.' and then separating the interferometer
outputs to obtain intensity signals representing equations (14) through
(17). As previously mentioned, the sensitivity of the interferometer is
proportional to dp/d.phi.. Thus, the sensitivity at wavelength .lambda. is
from equation (14):
dP.sub.1 (.lambda.)/d.phi.=-(P/2) sin .phi. (27)
The sensitivity at wavelength .lambda.' is from equation (15):
dP.sub.1 '(.lambda.')/d.phi.=-(P'/2) cos .phi. (28)
If the coefficients P and P' are made equal, we can add the squares of the
absolute values of the sensitivities given in equations (27) and (28) as
follows:
##EQU4##
Thus, the sensitivity of the interferometer is effectively a constant.
FIGS. 6 and 7 illustrate a preferred embodiment according to the present
Invention which maintains constant sensitivity following the above
described method. In FIG. 6, the light output P.sub.IN (.lambda.) of the
first laser 50 is modulated by a signal C.sub.1 cos .omega..sub.1 t at a
frequency .omega..sub.1 supplied by a generator 100. Similarly, the light
output P'.sub.IN (.lambda.') of the second laser 52 is modulated by a
signal C.sub.2 cos .omega..sub.2 t at a frequency .omega..sub.2 supplied
by a generator 102. The lasers 50 and 52 may be modulated by modulating
their input curr | | |
