JP2715224B2 - Resonator dispersion measurement method and apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for developing and adjusting an ultrashort optical pulse laser which generates an optical pulse having a time width of picosecond or less. In particular, it relates to a technique for measuring the wavelength dispersion characteristics of such a laser resonator.

[0002]

2. Description of the Related Art In recent years, techniques for generating optical pulses having a time width of picoseconds or less have been actively developed. As a result, when generating or transmitting a pulse with a short time width,
It has become clear that the wavelength dispersion characteristics of an optical component used for generation or transmission thereof or an optical path as an aggregate of the optical components greatly affects the shape of a pulse. For example, when a short optical pulse passes through an optical path where the wavelength dispersion characteristic changes rapidly, a waveform deformation phenomenon occurs. In addition, if an optical component whose wavelength dispersion characteristic changes rapidly is used, it is difficult to generate an optical pulse having a short time width. Especially for optical components in the cavity of ultrashort optical pulse lasers, it is necessary to control the chromatic dispersion characteristics with high precision because the waveform deformation is accumulated many times as light circulates through the cavity. It is. Under such circumstances, it is necessary to measure the wavelength dispersion characteristics of the laser resonator with high accuracy.

One method for measuring the wavelength dispersion characteristics of a laser resonator is to measure the wavelength dispersion of each optical component in the resonator with high accuracy, and estimate the wavelength dispersion of the resonator as the sum of these. Can be As a dispersion measuring method for this individual element, for example, Japanese Patent Application Laid-Open No. 2-134543 (Japanese Patent Application No. 63-287566), "Method and Apparatus for Dispersion Measurement", describes a method of measuring a dispersion on one arm of an interferometer using white light. A technique is disclosed in which an element is inserted, an interference waveform generated by changing a delay time difference is recorded, and chromatic dispersion of the element to be measured is obtained from phase information in a frequency domain obtained by performing a Fourier transform on the recorded waveform. Further, a waveguide type device is disclosed in JP-A-3-216530 (Japanese Patent Application No. Hei 2-216530).
11813), “Waveguide dispersion measurement method and device”
An optical coupling optical system is installed in both arms of the white light interferometer disclosed in Japanese Patent Application Laid-Open No. 2-134543 to cancel the wavelength dispersion of the optical coupling optical system for inputting and outputting light to and from the waveguide. A measuring instrument is disclosed.

However, some components in the laser resonator have wavelength dispersion characteristics that greatly depend on the incident direction or position of light on the element. For example, Optics Letter (Opti
csLetter), Vol. 9, pp. 150-152, 1984, describes a prism pair in which two isosceles prisms having apex angles formed so that the light incident and exit angles are Brewster angles. And anomalous dispersion characteristics in which the group delay time increases with respect to the wavelength are generated. At this time, the amount of dispersion generated depends very much on the optical path length passing through the glass in the prism, and thus on the incident position of light on the prism. In such a case, it is practically expected that the incident condition when measuring the chromatic dispersion of the individual element and the incident condition when the element is actually used in the laser resonator are completely the same. This cannot be done, leaving a large ambiguity in the estimation of the chromatic dispersion of the resonator. Further, performing measurements on each of the in-vibrator components that require a minimum of three or more is problematic both in terms of labor and time.

On the other hand, a technique for measuring the wavelength dispersion of the laser resonator itself (in this specification, the wavelength dispersion of the laser resonator itself is particularly referred to as “resonator dispersion”) has also been proposed. FIG. 5 shows such an example.

The conventional resonator dispersion measuring method shown in FIG.
Optics Letter, Vol. 17, pp. 514-516 1
It was used in 992 to measure the chromatic dispersion of a titanium sapphire laser resonator.

In this method, the measured laser resonator 5
Reference numeral 01 denotes a laser medium 502, a wavelength selector 503, an end face total reflection mirror 504, and an output coupling mirror 505.
When the laser medium 502 is excited by the continuous excitation device 506, the laser resonator 501 to be measured performs pulse oscillation. As the continuous excitation device 506, a continuous oscillation laser light source, a continuous lighting lamp, or a continuous current injection source is used. The oscillation wavelength λ of the laser resonator 501 to be measured is controlled by a wavelength selection device 503 therein.

The output of the laser resonator 501 to be measured enters the photodetector 508 from the output coupling mirror 505 as an output pulse train 507, and is converted into an electric signal pulse train. This electric signal pulse train is supplied to the frequency counter 509, and the pulse repetition frequency f (λ) of the pulse train is measured.
By repeating this measurement while sequentially changing the oscillation wavelength λ by the wavelength selection device 503, the pulse repetition frequency f (λ) for each wavelength is obtained.

Here, the repetition frequency f (λ) is expressed as f (λ) = c / L (1) using the resonator optical length L of the laser resonator 501 to be measured. c is the speed of light in vacuum. The resonator group delay time τ _d is expressed by an equation τ _d = (1 / c) [L−λ (dL / dλ)] (2) including the wavelength derivative of the optical length of the resonator. When this is rewritten into an expression including a repetition frequency, τ _d = (1 / f) [1+ (λ / f) (df / dλ)] (3) is obtained. The dispersion characteristic of the resonator is the resonator group delay time τ
_d is the change due to the wavelength λ of the third equation has become a basic equation representing the principle of measuring the cavity dispersion characteristic in the conventional method.

[0010]

However, the above-mentioned conventional resonator dispersion measuring method has the following problems.

The first problem is that the laser resonator to be measured must be excited by the continuous excitation device and perform pulse oscillation in this state. This requirement is not always met in view of the various laser devices.

As a method of realizing pulse oscillation in a laser, at least three types of forced mode locking, composite mode locking and passive mode locking can be considered. In the forced mode locking and the composite mode locking, a modulation signal or an excitation pulse that is repeated at a time interval equal to the laser circling time is externally applied. In this case, the obtained repetition frequency of pulse oscillation always exactly matches the repetition signal applied from the outside, and does not depend on the oscillation wavelength.
Therefore, a laser resonator employing forced mode locking or composite mode locking cannot be measured by the conventional resonator dispersion measurement method.

The second problem is that it is necessary that the laser resonator to be measured has a wavelength selecting device for controlling the oscillation wavelength therein.

In general, a laser device oscillates at a constant wavelength determined by a balance between the wavelength dependence of the gain of the laser medium and the wavelength dependence of the loss in the resonator. Since the dispersion characteristic of the resonator to be measured is a quantity including the wavelength derivative of the repetition frequency as shown in Equation 3,
It is essential to measure the repetition frequency at least at two different oscillation wavelengths. Therefore, it is necessary to forcibly change the oscillation wavelength using a wavelength selection device in the resonator.

It is known that in a certain kind of laser medium, for example, a semiconductor, the wavelength dependence of the gain changes depending on operating conditions such as pumping intensity or temperature, and changing these conditions changes the oscillation wavelength. Might be considered. However, such a change in the operating condition of the laser medium inevitably accompanies a change in the dispersion characteristics of the laser medium, so that the operating condition cannot be changed. This is because the resonator dispersion characteristics under certain operating conditions are measured here. Therefore, it is essential that the wavelength selection device is installed in the laser cavity to be measured, and for the purpose of the measurement here, the change in the dispersion characteristics of the element itself due to the selection of the wavelength is ignored. A wavelength selector as small as possible is needed.

However, in an ultrashort optical pulse laser, such a device is often not used because the wavelength selecting device in the resonator involves more or less expansion of the generated pulse width. In that case, temporarily installing the wavelength selection device only for the purpose of measuring the resonator dispersion is inconvenient, and furthermore, in order to obtain the dispersion of the resonator in a normal use state where the wavelength selection device is not installed. In addition, it is necessary to measure the dispersion measurement of the wavelength selection device using the dispersion measurement method for the individual element. Further, there is a case where it is impossible to additionally install a wavelength selecting device in a resonator after manufacturing a laser, for example, as in a monolithic mode-locked semiconductor laser.

These first and second problems limit the measurement object. In addition to this limitation, the conventional resonator dispersion measurement method requires that the photodetector used sufficiently respond to the resonator rotation time. For a laser having a long resonator length of, for example, 1.5 m, that is, a resonator circulating time of about 10 nanoseconds, this requirement can be easily satisfied with a commercially available PIN photodetector. In fact, measurements on such long-cavity lasers are made in the above-mentioned documents. However, a short cavity laser, for example, having an element length of 30
For a semiconductor laser of about 0 μm, the resonator circling time is about 7
As short as picoseconds, there is currently no photodetector that responds. Therefore, such a short cavity laser cannot be measured by the conventional technique. A similar high-speed response is required for a frequency counter disposed after the photodetector, which also makes measurement of a short-cavity laser very difficult.

As described above, according to the above-described conventional resonator dispersion measuring method, the laser resonator to be measured is provided with (1) pulse oscillation in a continuously excited state and (2) a wavelength selecting apparatus with a small change in dispersion characteristics. An extremely strong condition of controlling the oscillation wavelength by the laser is imposed, and the laser resonator to which it can be applied is extremely limited. In addition, it is difficult to measure short-resonance lasers because the photodetectors and frequency counters used must respond sufficiently to the cavity round-trip time.

An object of the present invention is to solve these problems and to provide a general-purpose resonator dispersion measuring method. In particular, irrespective of the type of means for causing the laser cavity to be measured to perform pulse oscillation such as continuous excitation, modulation signal application or pulse excitation, and irrespective of the presence or absence of a wavelength selection device in the laser cavity to be measured, It is an object of the present invention to provide a resonator dispersion measuring method in which the length of a measurable laser resonator is not limited by the response time of a photodetector or an electronic circuit.

[0020]

According to the resonator dispersion measuring method of the present invention, a parallel light beam is split into two light beams, and one of the two light beams is propagated to a first optical path having a fixed optical path length. The other of the two light beams is propagated to a second optical path whose optical path length is variable, and the light that has propagated through the first optical path and the light that has propagated through the second optical path are multiplexed and interfered. In the resonator dispersion measuring method for measuring the intensity while changing the optical path length of the second optical path, and obtaining the chromatic dispersion characteristic from the phase information in the frequency domain obtained by Fourier transforming the measured light intensity,
The parallel light flux is a light flux emitted from a laser resonator to be measured (hereinafter, referred to as a “laser to be measured”), and a relative optical path length difference between a first optical path and a second optical path is determined by the laser resonator. The optical path length of the second optical path is changed near an integral multiple of the resonator length of the second optical path.

When the product of the spread angle of the light beam generated by the laser resonator to be measured and the resonator length of the resonator is equal to or larger than the diameter of the light beam immediately after the resonator, that is, in the case of a diffused light beam, It is desirable to provide an optical device for increasing the parallelism of the light beam at the output of the vessel.

This method can be implemented using a conventional Michelson interferometer when the length of the laser resonator is short. However, when the cavity length of a dye laser or the like is long, for example, a cavity length of 1 m or more cannot be measured. In such a case, an apparatus for carrying out the above-described method, that is, the Michelson interferometer is provided with an incident end on which a light beam from the laser resonator to be measured enters, and the optical path length sweeping means includes two optical path length sweeping means. A resonator dispersion measuring device is required that is set near the range where the relative optical path length difference of the optical path is changed to an integral multiple of the resonator length of the laser resonator other than zero.

[0023]

After increasing the degree of parallelism of the light beam generated by the laser cavity to be measured, the relative optical path length difference between the two arms becomes an integral multiple of the cavity length of the laser cavity to be measured. The light is incident on a Michelson interferometer set in the vicinity of and the intensity of the generated interference light is converted into a voltage value by a photodetector and measured. At this time, every time the relative optical path length difference of the Michelson interferometer changes by a fixed amount, the output voltage value of the photodetector is sequentially and sequentially recorded, and the recorded data is obtained in the frequency domain obtained by Fourier transform. From the phase information, the wavelength dispersion characteristics of the laser resonator to be measured are measured. Thereby, regardless of the type of the means for causing the laser resonator under test to perform pulse oscillation, and regardless of the presence or absence of the wavelength selection device in the laser resonator under test, the measurable laser resonator can be used. The resonator dispersion characteristics can be measured without the length being limited by the response time of the photodetector or the electronic circuit.

This method uses a Michelson interferometer,
The point that the chromatic dispersion of the device under test is determined from the Fourier transform of the interference waveform is close to the technique disclosed in JP-A-2-134543. However, the device to be measured, in this case, the laser resonator is disposed outside the Michelson interferometer, not inside, and the light incident on the Michelson interferometer is not white light but a light beam from the laser resonator to be measured. One point is very different. Therefore, the measurement principle is also completely different from the technique disclosed in the above publication. The measurement principle will be described below.

The inventor of the present application has studied what caused the problems in the conventional measuring method described above. As a result, the conventional measurement method uses the phenomenon of pulse repetition frequency in pulse oscillation to find the cavity optical length for each wavelength of the laser cavity to be measured as the primary measurement quantity. It turned out not to be universally occurring.

As a method of avoiding such a problem and measuring the optical length of the cavity for each wavelength of the laser cavity to be measured through another phenomenon, the following method is considered. That is,
First, a wavelength selection device, specifically, a spectroscope, is installed between a laser resonator to be measured and a photodetector. This allows
A wavelength selecting device is not required in the laser resonator to be measured.
Next, the output electric signal of the photodetector is measured not by a frequency counter but by a spectrum analyzer. This eliminates the need for the laser resonator to be measured to perform pulse oscillation. According to such a configuration, even if the laser cavity to be measured oscillates continuously or even does not oscillate in the first place, on the spectrum analyzer, the cavity for the light of the wavelength selected by the spectroscope is used. A resonance peak is observed at a frequency position that is an integral multiple of the longitudinal mode frequency interval.
Since the resonator longitudinal mode frequency interval here is exactly equal to the pulse repetition frequency in the above-mentioned conventional measurement method, if this is applied to the third equation, the dispersion characteristics of the laser resonator to be measured can be obtained. .

However, this method also has the following difficulties.

First, in this method, similarly to the conventional measuring method, it is necessary that the photodetector to be used can sufficiently respond to the resonator rotation time. A similar high-speed response is, of course, required for the spectrum analyzer connected downstream of the photodetector, and this situation is the same as the requirement for the frequency counter in the conventional measurement method.

Second, in this case, the resonance peak observed on the spectrum analyzer at a frequency position that is an integer N times the resonator longitudinal mode frequency interval is the complex electric field amplitude at time t of the k-th longitudinal mode of the resonator. Is written as A _k (t), which is proportional to Ｗ W (λ _k ) ∫A _k (t) A ^* _{k + N} (t) dt (4) Where Σ represents the sum of k and W
(λ _k ) represents the power transmittance of the spectroscope for the wavelength λ _k of the k-th mode. This peak is generally weaker than the peak appearing at the zero frequency position on the spectrum analyzer. With respect to the peak appearing at the zero frequency position, the integral term of the fourth equation takes the form of ∫ | A _k (t) | ² dt (5), and the integrand of the integral term of the fourth equation is represented by an absolute value. The added integral ∫ | A _k (t) || A _{k + N} (t) | dt (6) has substantially the same magnitude as in the fifth equation. However, the value of the integral ＡA _k (t) A ^* _{k + N} (t) dt (7) included in the fourth equation is generally smaller than the value of the integral of the sixth equation. This is because the relative phases between the electric fields A _k (t) and A _{k + N} (t) of different longitudinal modes change every moment. However, when the laser cavity to be measured is performing mode-locked operation, the relative phase between the mode electric fields is fixed, so that the integral value of the seventh equation becomes equal to the integral value of the sixth equation. , A large resonance peak can be observed, and the resonator longitudinal mode frequency interval can be easily measured. On the other hand, when the laser resonator to be measured is oscillating continuously or has not yet oscillated, there is almost no relative phase relationship between the mode electric fields. The peak that appears is very small, on the order of one hundredth to one millionth, and it is practically very difficult to determine the cavity longitudinal mode frequency spacing therefrom.

The inventor of the present application has identified the principle advantages of this method,
That is, the characteristics that a wavelength selecting device is unnecessary and that the laser resonator to be measured does not need to perform pulse oscillation are maintained, and as a result of intensive studies to solve the above-mentioned difficulties, the present invention using an interferometer has been achieved. Was.

In the interferometer, a highly accurate time axis can be generated through a relative optical path length difference between two arms, that is, a delay time difference. The accuracy of this optical path length difference is 1μ even in a simple system.
m, several nm can be obtained by using the interferometric ranging method, which is equivalent to 3 to 0.02 femtoseconds in terms of delay time difference accuracy. The time response of the measurement is determined by the accuracy of the delay time, and the time response of the photodetector that receives the light emitted from the interference system is completely independent of the time response of the measurement. Thereby, it is possible to measure even a short resonator.

[0032] When the delay time difference of the interferometer tau, resulting interference signal S (tau), using the aforementioned _{symbols, Σ exp (jω k τ)} ∫ A k (t) A * k (t-τ ) dt... (8) Where Σ represents the sum of k and ω
_k is the angular frequency of the k-th longitudinal mode of the resonator. 8th
The meaning of the equation is that interference waveforms derived from each longitudinal mode oscillating at ω _k with respect to the delay time difference are superimposed with amplitude ∫ A _k (t) A ^* _k (t−τ) dt (9). That's what it means. Since the amplitude of the ninth equation is a correlation between the same longitudinal mode electric fields, it is similar to the fifth equation described above. Therefore, there is no disadvantageous property such as the equation (7), that is, the property that the property is large only when the laser resonator to be measured performs the mode-locking operation. Therefore, by using this interference signal for chromatic dispersion measurement, the measurement can be performed without any difficulty even when the laser resonator to be measured is continuously oscillating or has not yet oscillated. Ninth equation, near delay time difference τ is an integer N multiple of the cavity roundtrip _{^{time, t N (ω k) ∫}} | A k (t) | 2 dt ... (10) and can be considered. Here, t (ω _k ) is a complex number representing the change that the electric field undergoes when making a round in the resonator,
Its absolute value corresponds to a change in electric field strength, and its phase corresponds to a change in phase. The second half integral of the tenth equation is the optical power of the k-th longitudinal mode, that is, the optical spectrum U
(ω _k ). Therefore, the tenth equation is changed to t ^N (ω _k ) U (ω
_k ), and using this in equation (8), the delay time difference τ is represented by an interference signal S appearing near an integer N times the resonator rotation time.
_{N (τ)} is expressed as _{S N (τ) = Σ exp} (jω k τ) · t N (ω k) U (ω k) ... (11). By subjecting this signal to Fourier transform and extracting only a component that oscillates at ω _k , t ^N (ω _k ) U (ω _k ) is obtained. The light spectrum U (ω _k ) is always a positive real number. Therefore, the phase as a complex number of the Fourier transform result of the interference signal S _N (τ) is always the phase of t (ω _k ), that is, the phase change φ that the electric field receives when making a round in the resonator.
(ω _k ) only. That is, if the Fourier transform is represented by F, arg (F [S _N (τ)]) = Nφ (ω) (12) From the phase change φ (ω) obtained here, the resonator group delay time τ _d is obtained by τ _d (ω) = dφ (ω) / dω (13) Therefore, Equation 12 is a basic equation representing the principle of the resonator dispersion measurement according to the present invention.

[0033]

FIG. 1 is a diagram showing a basic configuration for implementing the present invention.

The parallel light beam generated by the laser cavity 101 to be measured enters a Michelson interferometer constituted by a semi-transparent mirror 102, a fixed mirror 104, a correction plate 103 and a movable mirror 105. Light that has passed through the Michelson interferometer enters the photodetector 108, and its intensity is measured. Movement of the movable mirror 105 changes the relative optical path length difference near the position where the relative optical path length difference 107 between both arms of the Michelson interferometer is equal to an integer N times the length of the laser cavity to be measured. Then, in the output signal of the photodetector 108, vibration caused by the light interference phenomenon appears. This output signal is stored in the waveform storage device 10.
9 is stored one by one.

Thereafter, the waveform of the signal recorded in the waveform storage device 109 is subjected to Fourier analysis by the computer 110. The phase of each angular frequency of light obtained as a result of this analysis,
That is, the phase of the Fourier component is
A phase characteristic for an integer N rounds of 1 is given.

Next, the principle of the present invention will be described in more detail.

Now, the delay time difference of the Michelson interferometer is swept from -T to T around a point where the relative optical path length difference is equal to an integer N times the length of the laser resonator 101 to be measured (FIG. 1). The movable mirror 105 is retracted in the arrangement (2), and the time interval ΔT during the sweep (when expressed by a relative optical path length difference, this ΔT may be multiplied by the light speed c).
When the interference signal is collected one by one every time, the number of data points of the collected interference signal is 2T / ΔT. When this signal data is Fourier-transformed by the computer 110, the step of angular frequency Δω (= π / 2) is obtained by the well-known sampling theorem.
T), it is decomposed into N / 2 Fourier components from ω = 0 to ω = π / ΔT. The upper end ω _NYQ of the angular frequency that appears here _expresses the well-known Nyquist frequency in the form of an angular frequency. The upper end ω _{NYQ needs} to be larger than the angular frequency ω _L = 2πc / λ _{L of} light corresponding to the short wavelength end λ _L of light generated from the laser resonator 101 to be measured, that is, ω _NYQ > ω _L. is there. This prevents aliasing caused by Fourier transform of discrete data, that is, if a frequency component exceeding the Nyquist frequency is present in the source signal, the frequency component is aliased and overlaps with a portion below the Nyquist frequency. That's why. Time step Δ during sweeping this condition
Rewriting to the condition of T, ΔT <λ _L / 2c (14) Further, when converted into the optical path length difference ΔL, the following condition is derived: ΔL <λ _L / 2 (15) Equation 15 indicates that, for example, when the short wavelength end of the fluorescence from the laser resonator 101 to be measured is 1200 nm, it is necessary to perform the measurement at intervals of the optical path length difference smaller than at least 600 nm. For this purpose, it is essential to measure the optical path length difference of the interferometer with an accuracy of at least several tens of nm.

The following methods are conceivable for realizing highly accurate optical path length difference measurement.

The first method is to use a two-frequency He-Ne stabilized laser which is already widely used.
In this method, since a resolution of 5 to 10 nm is achieved, the use of this technique makes it possible to measure the optical path length difference with the accuracy required by the present invention. However, in this length measuring method, a special polarization characteristic is required for the semi-transmissive mirror 102 in the interferometer.
Rather, the method described below is advantageous.

The second method uses a linearly-polarized monochromatic laser light source, for example, a He-Ne laser, as a reference light source having a length, and converts the linearly-polarized light into circularly-polarized light by one arm of an interferometer. This is a method of measuring light by separating polarized light. In this method, two interference signals having a phase difference of 90 degrees from each other are obtained. By using these two interference signals, a length measurement resolution of 1/50 or more of the wavelength of the reference light source can be easily achieved.

In addition, high resolution can be obtained even if a single interference signal is measured by a phase locked loop (PLL) without converting to a circularly polarized light. However, this method requires a higher uniformity of the sweep speed of the interferometer than other methods.

The present invention can be similarly implemented by using a method other than these three methods as a method for measuring the optical path length difference of the interferometer.

FIG. 2 is a diagram showing a specific embodiment of the present invention.

In this specific embodiment, the Michelson interferometer includes the cube beam splitter 205 and the movable mirror 20.
7. It is composed of a fixed mirror 206. In the cube beam splitter 205, the thicknesses of the transmitted light and the reflected light passing through the glass are always equal to each other, and the correction plate 103 as shown in FIG. 1 is not required. Between the Michelson interferometer and the laser resonator 201 to be measured, a single-mode optical fiber 202 as an optical device for increasing the parallelism of the light beam from the laser resonator 201 to be measured, and a coupling between both ends thereof. The lenses 203 and 204 are inserted.

As the linearly polarized monochromatic laser light source 211,
Here, a He-Ne laser is used. The reflecting mirror 212
The linearly polarized light emitted from the monochromatic laser light source 211 is reflected in the direction of the cube beam splitter 205. 1/8
The wave plate 213 converts linearly polarized light reflected by the cube beam splitter 205 via the reflecting mirror 212 into circularly polarized light. The reflecting mirror 214 reflects at right angles the interference light obtained by combining the linearly polarized light emitted from the Michelson interferometer through the fixed mirror and the circularly polarized light from the ８ wavelength plate 213. The polarization beam splitter 215 separates the interference light reflected by the reflection mirror 214 into polarization components. Photodetectors 216, 217
Detects the light intensity of the interference light of each polarization component separated by the polarization beam sputter 215. The trigger generator 218 generates a signal every time the interferometer optical path length difference changes by a predetermined step length based on two interference signals having a phase difference of 90 degrees output from the photodetectors 216 and 217. Generate a trigger signal. Waveform storage device 219
Is synchronized with the trigger signal from the trigger generator 218,
The level value (output voltage value) of the output signal of the photodetector 210 is captured and stored.

In this configuration, I in the 1.3 μm wavelength band
When measuring the wavelength dispersion of the nGaAsP semiconductor laser resonator, a germanium photodetector is preferably used as the photodetector 210. Further, when such a semiconductor laser is driven by a direct current, the fluorescence generated from the resonator is located at about 1.25 to 1.35 μm. In order to prevent the aliasing phenomenon, the interference signal may be measured at intervals of the interferometer optical path length difference of less than 625 nm.

In this example, the second method described above is employed as a length measuring method for achieving the optical path length difference increment with high accuracy. The monochromatic laser light source 21 is used as the length reference light source.
1. Specifically, a linearly polarized He-Ne laser oscillating at a wavelength of 632.8 nm is used. This monochromatic laser light source 21
The laser light from 1 is linearly polarized in the direction of 45 degrees on the plane of the paper, and is reflected by the reflection mirror 212 in parallel with the light from the laser resonator 201 through an optical device for increasing the parallelism of the light beam. Enter the interferometer. In this Michelson interferometer, the laser light from the monochromatic laser light source 211 is applied to the 腕 wavelength plate 21 by one arm of the interferometer.
3 and is reflected by the movable mirror 207 and again passes through the ８ wavelength plate 213 in the opposite direction. Therefore, an effect equivalent to passing through a quarter-wave plate is produced, and linearly polarized light is converted into circularly polarized light.

The laser light having a wavelength of 632.8 nm emitted from the interferometer passes through a reflecting mirror 214 and is polarized by a polarization beam splitter 21.
5 is separated into a polarized light component perpendicular to the paper surface and a light component parallel to the paper surface, and the light intensity of each component is converted into a voltage value by the individual photodetectors 216 and 217. These two voltage signals are input to the trigger generator 218. Trigger generator 2
At 18, a voltage pulse is generated from these two voltage signals as a trigger signal every time the interferometer optical path length difference changes by half of 632.8 nm, that is, 316.4 nm. With this trigger signal, the waveform storage device 219 sequentially stores the output voltage value of the photodetector 210 at the time when the voltage pulse is generated. The voltage signal value stored in the waveform storage device 219 is read by the computer 220 and subjected to Fourier transform.

FIGS. 3 and 4 show examples of the measurement results.
FIG. 3 shows an interference waveform obtained from the output of the photodetector 210, and FIG. 4 shows a chromatic dispersion characteristic obtained by performing an operation on the interference waveform. The interference waveform shown in FIG. 3 is obtained by sweeping the movable mirror 207 for a long distance before and after a fixed mirror contrast position 208 having a distance from the cube beam splitter 205 equal to that of the fixed mirror 206, and outputting the interference waveform output from the photodetector 210. This is a record of the whole situation.

The interference waveform shown in FIG. 3 has a large interference signal S ₀ near the interferometer optical path length difference (delay time difference) near zero, ie, near the position where the movable mirror 207 matches the fixed mirror reference position 208. (τ) has appeared. Also, on both sides,
The interfering signal S ₁ (τ) is separated by a circling time (7.46 picoseconds in this case) of the laser resonator 201 to be measured.
S _-1 (τ) appears. Here, a mutual interference signal generated at a position where the position of the movable mirror 207 is retracted from the fixed mirror reference position 208 is defined as S ₁ (τ). This mutual interference signal S
₁ (τ) is the measurement target here, and the delay time difference corresponding to the center is the resonator rotation time.

In order to find the position of the movable mirror where the interference signal waveform is to be recorded in the operation of the embodiment shown in FIG. 2, in the case of a short resonator as in this example, first, a large interference signal S
_It is effective to find a movable mirror position that coincides with the fixed mirror contrast position 208 as a position where ₀ (τ) appears, and then retract the movable mirror position from that position to find a mutual interference signal S ₁ (τ).

In the case of a long resonator length exceeding several cm, this method is not practical because the range of movement of the movable mirror moving mechanism is limited. Therefore, for example, the following method is used.
That is, the length of the resonator to be measured is calculated in advance from the thickness and the refractive index data of the optical components in the resonator and the interval between them, and the movable mirror 207 and the fixed mirror 20 are calculated.
The position of the movable mirror is temporarily set so that the difference in distance from the respective opposing surfaces of the sixth cube beam splitter 205 is equal to the calculated resonator length to be measured. Next, while moving the movable mirror back and forth around the position,
By observing the output voltage value of _No. 10, the position where the mutual interference signal S ₁ (τ) appears is found.

The overall operation of the embodiment shown in FIG. 2 will be described in more detail.

First, the movable mirror 207 transmits the mutual interference signal S
Move forward from the observed position of ₁ (τ) to the position where the signal disappears sufficiently. Here, the storage of the waveform storage device 219 is erased, and the data writing position is changed to the waveform storage device 21.
Reset to the starting address of 9. Next, when the movable mirror 207 is slowly retracted, a trigger signal is supplied to the waveform storage device 219 every time the interferometer optical path length difference changes by 316.4 nm, and the output voltage signal value of the photodetector 210 is stored in the waveform storage device 2.
19 is stored.

Here, "slowly" means that the trigger generator 2
The sweep speed is within a range in which the analog-to-digital conversion and writing operations of the waveform storage device 219 can follow the repetition of the trigger signal generated from. For example, when the analog-to-digital conversion and writing speed of the waveform storage device 219 is 20 kHz,
The maximum possible interferometer optical path length change speed is 20,000 [/ sec]
× 316.4 [nm] = 6.1328 mm / sec,
The maximum possible moving speed of the movable mirror 207 is half this, ie, 3.1.
64 mm / sec. The half here is the movable mirror 2
This is because, since the light is turned back on the surface of 07, twice the moving amount of the movable mirror 207 changes the optical path length.

The change range of the delay time difference required for the measurement is about twice the change amount of the group delay time of the laser resonator 201 to be measured in the wavelength range of the generated fluorescence. For example, in the 1.3 μm wavelength band InGaAsP semiconductor laser resonator illustrated in FIG. 2, the total change in the group delay time is at most about 0.5 picoseconds in the fluorescence wavelength range of 1.25 to 1.35 μm. The required delay time difference change range is about one picosecond. This is also confirmed from the fact that the observed delay time range of the mutual interference signal S ₁ (τ) does not exceed 1 picosecond in FIG.

In the actual measurement, the sweep was performed over a delay time difference change range of 4.4 picoseconds with a margin. When this delay time difference change range is converted into an optical path length difference change range, 0.66 m
m, and sweeping this range with the maximum possible moving speed of the movable mirror 207, the time required for signal measurement is only about 0.1 second. Even if the sweep is delayed with a margin, the signal measurement is completed within one second. It should be noted that a known moving mechanism was used for this sweep. The number of data points collected at this time is 4016 points, and the calculation of the Fourier transform of this data is performed by the computer 220 by 2 to 3 times.
Could run in seconds. FIG. 4 shows an example of the obtained wavelength dispersion characteristics. The measurement of the wavelength dispersion characteristic, that is, the wavelength dependence of the group delay time was completed within 5 seconds in total, and an extremely quick measurement of the wavelength dispersion characteristic was realized.

[0058]

As described above, the cavity dispersion measuring method of the present invention does not depend on the type of means for causing the laser cavity to be measured to perform pulse oscillation, and the wavelength selecting device in the laser cavity to be measured. Irrespective of the presence or absence of the laser cavity, the measurable length of the laser resonator to be measured is not limited by the response time of the photodetector and the electronic circuit, and the dispersion characteristics of the laser resonator can be measured for general purposes. INDUSTRIAL APPLICABILITY The present invention can be used not only for a test at the time of development of an ultrashort optical pulse laser, but also for a test after manufacture, and has a great industrial effect.

[Brief description of the drawings]

FIG. 1 is a diagram showing a basic configuration for implementing the present invention.

FIG. 2 is a diagram showing a specific example.

FIG. 3 is a diagram showing a result of measuring the cavity dispersion of an InGaAsP semiconductor laser resonator, and a diagram showing a measurement result of an interference waveform.

FIG. 4 is a diagram showing a result of measuring the cavity dispersion of the InGaAsP semiconductor laser resonator, and showing a wavelength dispersion characteristic obtained by performing an operation on the interference waveform of FIG. 3;

FIG. 5 is a diagram showing a conventional resonator dispersion measuring method.

[Explanation of symbols]

 101, 201, 501 Laser cavity to be measured 102 Semi-transparent mirror 103 Correction plate 104, 206 Fixed mirror 105, 207 Movable mirror 106, 208 Fixed mirror contrast position 107, 209 Relative optical path length difference 108, 210, 216, 217, 508 Photodetector 109, 219 Waveform storage device 110, 220 Computer 202 Single-mode optical fiber 203, 204 Coupling lens 205 Cube beam splitter 211 Monochromatic laser light source 212, 214 Reflector 213 ８ Wave plate 215 Polarization beam splitter 218 Trigger generation Device 502 laser medium 503 wavelength selecting device 504 end face total reflection mirror 505 output coupling mirror 506 continuous excitation device 507 output pulse train 509 frequency counter

Claims (3)

(57) [Claims] 1. A parallel light beam is split into two light beams, one of the two light beams is propagated to a first light path having a fixed optical path length, and the other of the two light beams is converted into a second light beam having a variable optical path length. Light propagating in the optical path, multiplexing the light propagating in the first optical path and the light propagating in the second optical path and causing them to interfere with each other, and changing the light intensity of the interference light to change the optical path length of the second optical path. In the resonator dispersion measuring method for obtaining chromatic dispersion characteristics from phase information in a frequency domain obtained by Fourier transforming the measured light intensity, the parallel light beam is emitted from a laser cavity to be measured. A light flux, the optical path length of the second optical path in the vicinity where the relative optical path length difference between the first optical path and the second optical path becomes an integral multiple of the resonator length of the laser resonator other than zero. A resonator dispersion measuring method, characterized by changing. 2. A Michelson interferometer for splitting a parallel light beam into two light beams, propagating the light beams to different light paths, and then combining and interfering with each other, and a relative light path between two light paths of the Michelson interferometer. Optical path length sweeping means for sequentially changing the length difference; measuring means for measuring the intensity of the interference light emitted from the Michelson interferometer in accordance with the change in the relative optical path length difference between the two optical paths; Calculating means for obtaining chromatic dispersion characteristics from phase information in the frequency domain obtained by performing Fourier transform on the obtained light intensity waveform, wherein the Michelson interferometer includes a laser resonator to be measured. The optical path length sweeping means has a range in which the relative optical path difference between the two optical paths is changed is an integral multiple of the resonator length of the laser resonator other than zero. Nearby Cavity dispersion measuring apparatus, characterized in that it is set to. 3. The resonator dispersion measuring apparatus according to claim 2, wherein optical means for increasing the parallelism of a light beam emitted from the laser resonator to be measured is provided at the incident end.