JPH0480629A - Measuring device for frequency noise of laser - Google Patents

Abstract

PURPOSE:To enable accurate and simple measurement of frequency noise in a low-frequency range of a laser, such as flicker noise, by a method wherein the light frequency and propagation time of one branch light of a light to be measured are changed relatively in relation to the other branch light. CONSTITUTION:The light frequency and propagation time of one branch light of a light to be measured, which is made to branch by a light branch means A, are changed relatively in relation to the other branch light by a light frequency shifting means B and a delay means C. These branch lights are synthesized by a light synthesizing means D and a beat signal between the two synthesized branch lights is subjected to light heterodyne detection by a photoelectric conversion means E. Thereby frequency noise in a laser is down-converted into the frequency noise of the beat signal. A dispersion value, i.e., a frequency fluctuation dispersion value, obtained when the frequency of this electric signal is measured at a certain time interval, continuously and in a number of times is computed by a fluctuation dispersion measuring means F. Thereby the magnitude of a frequency noise spectrum of the original laser can be estimated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a device for measuring frequency noise of a laser oscillating with a single longitudinal mote, and particularly to a device for measuring frequency noise of a laser that oscillates with a single longitudinal mote, and in particular, it is capable of reducing flicker noise and random walking noise by 11 times. This invention relates to a laser frequency noise measurement device that can easily and accurately measure the magnitude of the dominant frequency ¥1 noise component in the frequency domain.

[Conventional technology]

Lasers, especially semiconductor lasers that oscillate in a single longitudinal mote, are used in optical frequency division multiplexing (FDM) communication systems that utilize the wide transmission band of several T (tera) llz of optical fibers, and optical heterodyne systems that have high reception sensitivity. / Homodyne detection type optical communication nostem, high) Recently, its importance has been increasing as a light source for various precenosanostem etc. with detection sensitivity. However, since the above-mentioned Nostem controls the previous frequency and phase, or adds information to them, it is easily affected by laser frequency noise (also called phase noise). Therefore, evaluating the magnitude of this frequency vl noise is extremely important for engineering applications.

The frequency noise of an oscillator is best evaluated by the frequency noise power spectrum. The frequency noise electric cassette vector of various existing oscillators can be calculated using a simple ζ series model, as described in Time and Frequency (by P. Kartanoyov, published by Enotifinok, 1980). It is known empirically that L is well approximated by .

For example, in a semiconductor laser, except for the frequency range above IGHz where the noise peak due to relaxation oscillation is observed,
The (one-sided) frequency noise large cass vector density S (f) below GHz is 5(f)-, /f' = k-3, # -k-,
It is expressed as (1). FIG. 4 schematically shows the frequency noise electric cassette vector given by equation (1). In this way, laser frequency noise has at least three components: white noise, flicker noise, and random walking noise, and the magnitude of each component is determined by the coefficient k.
It is evaluated by a (-4≦a≦-2).

FSK of laser light (frequency 7 soft key ink) and p
Using SK (phase knot key innoh), n several hundred to 4
The performance of high-speed coherent optical communication systems faster than b/s is determined by the white noise component of these three noise components.
Mainly limited. For this reason, energetic research has been carried out on the white frequency noise of lasers. And its magnitude 2 evaluation is the spectral linewidth Δ of the laser
This is done using ν. This is because, assuming that the frequency noise of the laser is only a white noise component, the field spectrum of the laser output light is of the Lorenots type, and its full width at half maximum Δν is -2 to the coefficient of the white noise component, and Δν ; π This is because it shows a simple relationship (for example, Coherent Optical Communication Engineering 1 Ohmsha, 19
1989). In fact, in the case of a semiconductor laser, the field spectrum observed θ is often of the Lorenot type, and it is possible to easily estimate it as a coefficient by measuring Δν.

However, slow coherent optical communication system of several Mb/s or less or feedback band of several 5
In an optical frequency stabilization nostem of IHz or less, the flicker noise component or ranogum walk noise component of the laser frequency noise can be a factor that greatly limits the performance of the nostem. Therefore, it is also extremely important for engineering applications to evaluate the magnitude of the frequency noise in the low frequency region of the laser, that is, to evaluate the magnitude of the frequency noise in the low frequency region, that is, to evaluate the coefficient in equation (1). Figure 9 shows the configuration of a conventional device for measuring the frequency noise of a semiconductor laser in such a low frequency range (see, for example, Kikuchi et al., Technical Report 0QE83-23 of the Institute of Electrical Communication Research Group). .

Here, the measurement principle using the f2 conventional device shown in Fig. 9 is as follows:
Explain briefly. The light from the semiconductor laser 21 is transmitted through an air filter 22 for preventing the oscillation state from changing due to return light from measurement devices, and then enters the Fabry-Perot etalon 231. The Fabry-Perot etalon 23 exhibits periodic transmission characteristics as shown in FIG. 10, and is used as an optical frequency discriminator by using the shoulder portion of the transmittance peak. That is, it functions to convert optical frequency fluctuations due to frequency noise into optical intensity fluctuations, that is, optical intensity noise. This situation is shown in FIG. 1I. The intensity of the transmitted light from the Fabry-Perot etalon 23 is converted into an electrical signal by the photoguioto 25. Therefore, the laser frequency noise electrical cassette vector is the intensity noise electrical cassette vector of this electrical signal as a spectrum
It can be measured by observing with the analyzer 26. On the other hand, a portion of the transmitted light from the Fabry-Perot etalon 23 is split by a beam splitter 24, and its intensity is monitored by a photodiode 27. Then, the oscillation frequency of the semiconductor laser 21 is controlled by the laser oscillation frequency control device 28 so that this intensity is constant.

[Invention or problem to be solved]

However, when measuring the frequency noise type cass vector of a laser in a low frequency region using the L conventional apparatus shown in FIG. 9, the following problems occur.

What is measured by the spectrum analyzer 26 is the intensity noise type cass vector of the light incident on the photodiode 25. However, the measured values include Fabry
The electrical flux vector is not due to the frequency noise of the laser converted by the Perot etalon 23, but is due to the laser intensity noise contained in the output light itself of the semiconductor laser 21 and mechanical fluctuations of other measurement equipment components. The electric scum vector due to the added light intensity fluctuation increases the weight.

However, unless the optical frequency discrimination sensitivity of the Fabry-Perot etalon 23 is sufficiently high, it is difficult to accurately measure frequency noise. Mako, from the measured value -3
Second, to estimate the frequency noise spectral density coefficient, it is necessary to determine the conversion coefficient from frequency noise to intensity noise by random measurement, and the measurement error at that time is directly reflected in the measurement accuracy. It will be done. Furthermore, this conversion coefficient depends not only on the transmittance characteristics of the Fabry-Perot etalon 23 but also on the intensity of the laser beam itself. Second...
The situation is schematically shown in Figure 12(a).

In addition, the oscillation optical frequency of the semiconductor laser 212 to be measured is controlled by the photodiode 27 and the laser oscillation frequency control device 28.
It is also a major drawback that there is a need for feedback control using . The control loop is a semiconductor laser 21
The transmittance characteristic of the Fabry-Perot etalon 23 fluctuates due to the oscillation frequency of , which serves to prevent the desired optical frequency discrimination characteristic from being obtained as shown in FIG. 12(b). Generally, spectrum measurements in the low frequency range require long measurement times. In the conventional device shown in FIG.
The oscillation frequency of No. 1 is extremely unstable and can easily fluctuate by several GHz. Therefore, the above-mentioned t=feedback control is indispensable in actual measurements.

An object of the present invention is to solve the above-mentioned problems, and to provide a foot hack that is less susceptible to the effects of intensity noise and intensity fluctuations, so that frequency noise in the low frequency range of a laser, such as flicker noise, can be accurately and easily measured. An object of the present invention is to provide a laser frequency noise measuring device that does not require any control.

[Means to solve the problem]

As shown in FIG. 1, the present invention includes a light branching means A that branches a laser beam to be measured into two, and a frequency ratio of one branched light branched by the light branching means A to a frequency of the other branched light. an optical frequency shifting means B, a delaying means C for delaying the other branched light relative to the one branched light, and the optical frequency shifting means B and the delaying means C shift the frequency and an optical multiplexing means for combining both branched lights whose propagation times have been relatively changed; and a photoelectric conversion means E for detecting an optical hetero gain of a beat signal between the two branched lights combined by the optical multiplexing means. , and a frequency fluctuation dispersion measuring means F for measuring a frequency fluctuation dispersion value of the heat signal.

[Effect]

According to the above configuration, the optical frequency and propagation time of one branched light of the light to be measured are changed relative to those of the other branched light, and the heat signal is converted into light between the other branched light and the other branched light. The frequency noise of the laser is down-converted to the frequency noise of the heat signal by 1,000-rodyne detection. Here, the center frequency of heat IN is the amount of frequency deviation in optical frequency deviation means B, so this value is, for example, several hundred MHz.
It can be set to several GHz. In other words, the frequency noise of a laser ranges from a few hundred MHz to around a scattered GHz.
(2) The electrical signal is converted into frequency noise. Then, by calculating the dispersion value when the frequency of this electrical signal is continuously measured many times at a certain time interval, that is, the frequency fluctuation dispersion value, the size of the original frequency noise spectrum of the laser can be estimated. becomes possible.

Note that the optical frequency and propagation time of one branched light of the light to be measured are changed relative to those of the other branched light, and the heat signal is detected by optical l\tero gain, and the frequency noise of the laser is detected by comparing the frequency noise of the laser with the heat signal.・The part that down-converts to frequency noise and measures it is a conventional technique that has already been proposed by Kikuchi et al. under the name of the delayed self-telodyne method (for example, the previously mentioned Copy Throat Communication Engineering, Chapter 3, 90-94). (see Peno). This delayed self-heterogy7 method observes the filter spectrum of the heat signal with a spectrum analyzer,
The line width Δ and the IF are used to evaluate the size of the white frequency noise component of the laser. The present invention combines the delayed self-hetero gain method with the measurement of the frequency fluctuation dispersion value of the heat signal. It can also be said to be a single device.

Two Real Zells 7 Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 shows the frequency of the laser according to the first embodiment of the present invention.
1 is a block diagram showing the configuration of a noise measuring device. In this figure, the laser beam to be measured is incident on an optical coupler 31 and is split into two. On the other hand, the 9-branched light is guided to the optical fiber delay line 32 and sent to the optical coupler 3.1), which takes a longer time τ than the other branched lights.
Give R Koke 9 delay and U5. The other branched light is knotted by an acoustic pre-modulator (0\1) 33b driven by an oscillator 33a with a frequency fl+', which is the optical frequency flF:+. Here, flF is set to about 250 Mtiz for Vj'i. The other branched light, which is frequency-knotted by the acousto-optic modulator 33b, is sent to the coupler 3 after passing through the polarization state controller 38.
4 and is combined with one of the branched lights. Both ends of the combined light are photoelectrically converted by the photodiode 35. Then, the center frequency flF C'') heat signal is detected by the optical heterodyno detection at 2'''1s''.In addition, in the two lights combined by the optical coupler 34, their polarized waves are orthogonal to each other. If the heat signal is not detected, the polarization state controller 38 adjusts the polarity so that they match.

The single heat signal detected by the photodiode 35 is amplified to an appropriate level by an amplifier 39 and applied to an electric signal frequency fluctuation dispersion measuring device 36. As the frequency fluctuation dispersion measuring device 36, for example, a frequency time interval analyzer HP5371A manufactured by Yokogawa Hulett Bakker 1~ is used.

Next, let me explain the knowledge of the frequency noise measuring device with the above configuration.

First, the general ζ relationship between the frequency fluctuation dispersion value, the frequency e, and the electronic cassette vector will be explained. The frequency fluctuation branch value σ2(τ) is obtained by measuring the average frequency at a time interval τ consecutively ＼ times with no idle time, and is calculated as σ′(r)−Σ(J++, −rt)′/2 I-D
−(3) is calculated. Here, fl is the first measured average frequency. If N is infinite, σ2(τ
) is a well-known measure of frequency fluctuation, the Allan variance σ
′(τ) and . is the average frequency, σ 2(τ
)-shi. ′ σ y'(τ )
(4) and L.5, which shows a simple relationship, that is, the frequency fluctuation variance σ2(τ) corresponds to the non-normalized Allan variance. Therefore, N
σ2(τ) at infinity is calculated from the frequency noise amount cass vector density S (f) by the following formula (for example, see the previous page 7 and Frequency-22).

For example, in the case of a semiconductor laser, by substituting equation (1) into the above equation, the following is obtained. This is schematically shown in Figure 3. As can be seen by comparing with Fig. 4, which schematically shows the frequency noise electric cassette vector, the frequency noise component in the low frequency range can be measured by setting τ large and measuring the frequency fluctuation dispersion value σ2 (τ). It is possible to evaluate according to the following. In particular, part of the magnitude of the flicker frequency noise component can be determined by measuring the frequency fluctuation variance value σ2 that does not depend on τ, so that −*= a′/21n2
-- It is possible to easily evaluate it as (7).

In addition, in the actual measurement of frequency fluctuation dispersion, if 5 is the evening limit, or \>10, the standard deviation of the measured value for σ at N infinity is \-1'' (however, K is a single digit). It is known that the constant 9 becomes (for example, see the time and frequency page 24 mentioned above).For example, \=10
00, the measurement error of σ caused by the finite number 5 can be kept to 10% or less.

Next, the relationship between the frequency fluctuation of the beat signal measured in the first embodiment of the present invention shown in FIG. 2 and the frequency noise of the laser to be measured will be derived. Let the propagation time difference between the two branched lights be τd.1 In this case, the density S of the frequency noise electric dust vector of the heat signal detected by the photodiode 35 IF(r) is the density S of the frequency noise electric dust vector of the laser. Using (f), 5IF(f)=23(f)・(1-cos2rfrd
) is given. Once, the heat signal 7 round cycle variance value σIF'(τ) can be calculated by substituting equation (8) into JS(r') 1'' in equation 5.
J Go 7-7,. White, Futonoka, Langeni, Walk [
The results of the frequency fluctuation dispersion value Ωnp due to the noise component are shown in Fig. 5 (a), (b: +, and (c) by the curved lines, respectively.

Way 1. The tonne line is 8 f(
,',fat"・92 times is plotted and returns 0.

If τ (τd) is 771, then the frequency fluctuation is measured by the uninvented first embodiment 9 device in Figure 2. Dispersion value: y, the frequency fluctuation of the laser corresponds to twice the dispersion l'1. R for chords added together
,-do.

Therefore, by setting the delay τd caused by the optical fiber delay line 32 and the fixed time interval τ of the frequency fluctuation θ to appropriate values, the device of the present invention can be used to detect the dominant t laser with the time interval τ. It is possible to measure the magnitude of 9-frequency noise components. For example, if a fiber with a length of 20 km is used as a delay line, τd ~ 100 μs, so τ
Frequency fluctuation dispersion value at ≦50μs or the measurement result is 05
A simple correction of multiplying can be used to obtain sufficiently accurate measurements. Further, the magnitude ka of the frequency noise can be evaluated using the relationship given by equation (6).

τ can be arbitrarily set between τ-600 ns and 8 s, for example, by using the previously described frequency/time interval analyzer HP 5371 A as the frequency fluctuation dispersion measuring device 36. Since this analyzer has a processing function to measure the frequency many times and obtain Allano dispersion, etc., it can be used as the frequency fluctuation dispersion measuring device 36.

On the other hand, when measuring the frequency noise of a semiconductor laser, for example, the magnitudes of typical white and flicker frequency noise components are determined by the linewidth - 30 MHz and the frequency fluctuation dispersion value σ' - (IM Hz) which does not depend on τ, respectively. 'That's about it.

In this case, using equations (2) and (7), the intersection point of both the white and flicker components is calculated to be approximately 5 μs in the model diagram shown in FIG. Therefore, τd=l
By setting OOμs1τ−50μs, the apparatus of this embodiment can easily evaluate the magnitude of the flicker frequency noise component of the semiconductor laser. Mani, τ〈τd
By measuring the dispersion value of the frequency fluctuation of multiple τ where a is a, the type of frequency noise component is identified using the τ dependence.
It is possible to evaluate 2 and O.

The above is explained with reference to FIG. 2. According to the first embodiment of the present invention, the frequency v! J. The fluctuation dispersion measuring device 36 measures the frequency fluctuation of the electrical signal.Secondly, by using dinotal circuits such as counters, it is possible to measure frequency noise essentially unaffected by intensity noise. . In addition, since optical frequency noise is directly converted into electrical frequency noise using the delayed self-telodyno method, there is no need to use parameters such as optical frequency discrimination sensitivity, which is seen in conventional equipment, and the accuracy is high. It is possible to easily realize the measurement.

Furthermore, even when measuring frequency noise in a low frequency range that requires a long measurement time, feedback control to the θ11 constant laser is not required, making it possible to easily measure the noise.

Note that in the f2 device shown in FIG. 2, the optical fiber delay line 32, the acoustic forward modulator 33b, and the polarization state controller 38 are installed in different branch optical paths, or are not limited to this. Shinodana(, As is clear from the above explanation, they may be placed in the order of Tora's path, Maf2, etc.).

Alternatively, the degree of freedom in setting the frequency r may be increased by using a plurality of acousto-optic modulators. Further, in place of the optical couplers 31 and 34, it is of course possible to use optical branching/combining means such as a half mirror. Additionally, the polarization state controller 38 can be omitted by using an optical wiring system based on polarization-maintaining fibers to ensure that the polarization states of both lights are not orthogonal at all times when combining at the optical coupler 34. Shiyoi.

In addition, as the frequency fluctuation dispersion measuring device 36, the HP
Instead of a counter-based device such as the 5371A, an electric frequency discriminator and an intensity fluctuation dispersion measuring device may be combined.

In this case, whether we lose the advantage of being essentially unaffected by intensity noise or lose it, we can easily achieve a steep and stable frequency discrimination characteristic in the electrical domain. It can be configured to In this case, unlike the prior art, for example, the intensity fluctuation caused by the fluctuation rJ of the polarization state between the two branched lights will change and affect the measurement. A block diagram showing the configuration of such an I?lI wavenumber fluctuation dispersion measuring device 36 is shown in Section 6V.AI and C control are based on feedback in the conventional device explained using FIG. In the conventional technology, light is interposed in the foot bag loop, but in the f2 circuit shown in Figure 6, it is closed in the electrical region, and high stability is easily achieved. I have something to do.

Next, FIG. 7 is a block diagram showing the configuration of a second embodiment of the present invention. The feature of this second embodiment is that in the first embodiment described above, a single photodiode 35 is used when performing optical heterodyne detection of the output of the optical coupler 34, which is an optical multiplexing means. Instead, a balanced photodiode 45 is used. ″
5. There are l. In this case, the branching ratio of the optical coupler 34 is set to 2.
The output light power from the two output photons is set to be equal. The configuration other than the above is shown in FIG. 2 and is completely the same as that of the first embodiment, so corresponding parts are given the same reference numerals and explanations thereof will be omitted.

By performing optical heterogain detection using the balanced photodiode 45 in this way, all of the two branched lights combined by the front coupler 34 are used for photoelectric conversion, reducing the power of the laser beam to be measured required for measurement. It becomes possible to do so. In addition, since it is possible to cancel the influence of additional light intensity fluctuation caused by the acousto-optic modulator 33b, the fiber delay line 32, and the circuit installed in the branched destination path, When using the frequency fluctuation dispersion measuring device 36 as explained using FIG. 6, it has the advantage that the influence of intensity noise can be reduced. In the measuring device according to such embodiment □, the first
All the features of the invention described in the embodiments apply.

Next, FIG. 8 is a block diagram showing the configuration of a third embodiment of the present invention. The feature of this third embodiment is that an optical amplifier 49 is used to compensate for optical loss in the optical fiber delay line 32. The configuration other than the above is exactly the same as that of the first embodiment shown in FIG. 2, so are there corresponding parts? The same reference numerals are used to omit the explanation.

Generally, the loss of optical fiber is, for example, 0.2 dB/km.
In order to set the delay time as large as τd-1 ms, a delay fiber of 200 km is required, and the loss in the fiber delay line 32 becomes as large as 40 dB. When measuring the frequency fluctuation dispersion value with a large value of τ, τdl must be set large, and in this case, a large laser light power is required. However, as shown in FIG.
By using 9, this requirement is relaxed and it becomes possible to measure large τ.

Also, in this third embodiment, it is possible to achieve higher detection sensitivity with a configuration similar to that described in the second example.

:Effects of the Invention] As detailed above, according to the present invention, it is possible to evaluate the magnitude of frequency noise in the low frequency region of a laser, that is, the flicker frequency noise component and the random walk frequency noise component.
This can be done extremely easily without stabilizing the laser under test through feedback control, and furthermore,
By measuring the frequency of electrical signals, it does not rely on optical frequency discrimination sensitivity like in the past, and the effects of intensity fluctuations caused by the intensity noise of the laser to be measured and other optical components are significantly reduced, making it more accurate. This has the effect of making it possible to measure frequency noise.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the principle configuration of the present invention, Fig. 2 is a block diagram showing the configuration of the first embodiment of the invention, and Fig. 3 is a graph for explaining a model of laser frequency fluctuation dispersion. , FIG. 4 is a graph for explaining the frequency noise power Svectorini model of the laser, and FIGS. 5(a) to (c) show the first embodiment of the present invention. L
Figure 6 shows an example of the configuration of the frequency fluctuation dispersion measuring device 36 of the first embodiment of the present invention. 7 is a block diagram showing the configuration of the second embodiment of the present invention, FIG. 8 is a block diagram showing the configuration of the third embodiment of the present invention, and FIG. 9 is a block diagram showing the configuration of the third embodiment of the present invention. A block diagram showing an example of the configuration of a conventional laser frequency noise measurement device. Figure 1O is a graph showing the optical frequency discrimination characteristic of the Fabry-Herot etalon 23 of the conventional laser frequency noise measurement device. Figure 1+ shows the frequency according to the discrimination characteristic. Graphs for explaining the conversion of noise into intensity noise. FIGS. 12(a) and 12(b) are graphs for explaining changes in optical frequency discrimination sensitivity and changes in optical frequency discrimination characteristics due to variations in optical intensity. . 32 ・ 3a 3b 35 ・・ 36 ・・ 38 ・ 39 ・ Optical coupler, optical fiber delay line, oscillator, acousto-optic modulator, tip coupler, ・Photodiode, ...frequency fluctuation dispersion measuring device, polarization state controller , ・Amplifier. S (f) Figure 4 d2 (τ) 09τ Figure 3 Temperature 8 Yuqiang Transmission S Light Qiang Figure 12

Claims (1)

[Scope of Claims] Optical branching means for splitting a laser beam to be measured into two, and an optical frequency shift for shifting the frequency of one branched light branched by the optical branching means compared to the frequency of the other branched light. means, delay means for delaying the other branch light relative to the one branch light, and both branches whose frequencies and propagation times are relatively changed by the optical frequency shifting means and the delay means. an optical multiplexing means for combining light; a photoelectric conversion means for optically heterodyne detecting a beat signal between both branched lights combined by the optical multiplexing means; and a frequency fluctuation dispersion for measuring a frequency fluctuation dispersion value of the beat signal. A laser frequency noise measuring device comprising: a measuring means;