JPH052928B2 - - Google Patents

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a short optical pulse evaluation method, and more specifically, to a short optical pulse evaluation method, which changes rapidly with a time width comparable to or less than the response time of existing photodetectors. With these photodetectors, it is possible to detect short optical pulses in which the pulse waveform and changes in the wavelength of light in each part of the pulse, or changes in the frequency of light equivalent to the pulse waveform, cannot be measured and evaluated in detail. The present invention relates to a method that allows detailed measurement and evaluation of both the pulse intensity waveform and the pulse phase waveform corresponding to the integral of the change in instantaneous frequency when the pulse train forms a pulse train that repeats at a constant period.

[Conventional technology and its problems]

FIG. 2 is a schematic diagram showing a first conventional example of a short optical pulse measurement and evaluation device. The measured optical pulse flux is
A part of the light beam reflected by the semi-transparent mirror 24 is distributed to the spectrum measurement system, and the remaining part is supplied to the intensity correlation waveform measurement system. The spectrum measurement system is a wavelength sending device (scanning mechanism) that changes the set wavelength.
The spectrometer 27 is equipped with a high-resolution spectrometer 26, a diaphragm lens 25 that focuses a light beam onto an input slit of the spectrometer 26, a photodetector 28, and an integrating amplifier 29 that amplifies the output voltage of the detector 28.

In the intensity correlation waveform measurement system, an incident light beam is branched into two light beams by a semi-transparent mirror 1, one of which is reflected by a fixed prism 2, and the other beam is reflected by a movable table 4 on the optical axis. The light beam is reflected by a prism 3 that is movable in the direction, and the above two reflected light beams are combined on a semi-transparent mirror 1 to become one light beam again. This optical system consisting of a semi-transparent mirror and two prisms 2 and 3 is called a Michelson interferometer, and when the optical electric field before incidence is written as E(t) as a function of time t, the optical electric field after recombination is expressed as E(t). The optical electric field is [E(t)+E(t-τ)]/
It is expressed as 2. Here, the delay time τ is the prism 2
It is the amount obtained by dividing the relative optical path length difference l between the optical path passing through and the optical path passing through the prism 3 by the speed of light C, and since the prism 3 is movable by the movable stage 4,
τ can be freely changed. After this recombination, the light beam is focused by the lens 5 into the crystal 7 which has the ability to generate second harmonic light, and becomes [E(t)+E(t-τ)] ² .
Look at the generation of second-order harmonic light (light with half the wavelength of the incident light) whose electric field is proportional. The light emitted from the crystal 7 is focused onto a photodetector 9 by a lens 6.
At this time, only the second harmonic light is passed through the optical filter 8 and reaches the photodetector 9. Photodetector 9
The response time of is extremely long compared to the duration of the optical pulse to be measured and the second harmonic pulse generated by the crystal 7, so the waveform of the second harmonic optical pulse cannot be faithfully detected, and the output of the photodetector 9 Although the waveform of the voltage pulse becomes blunt, the time integral value, that is, the area of the pulse waveform becomes proportional to the energy of the second harmonic optical pulse. The output voltage pulse of the photodetector 9 is passed through an integrating amplifier 1 having a large time constant T.
0 is converted into a constant voltage proportional to the area of the output voltage pulse.

The short optical pulse measurement procedure using this conventional device is shown below.

First, turn the signal changeover switch 31 to the right,
The A/D converter 32 is connected to the spectrum measurement system. In this state, while scanning the center wavelength of the spectrometer 26 at a constant speed using the wavelength sending device 27, the output voltage of the photodetector 28 is read and stored in the computer 18 in time series via the A/D converter 32. let In this way, a data string of the optical pulse spectrum D(λ) is obtained. Next, the signal changeover switch 31 is turned to the left to connect the A/D converter to the intensity correlation waveform measurement system. In this state, while moving the movable table 4 at a constant speed, the output voltage of the integrating amplifier 10 is read and stored in the computer 18 in time series via the A/D converter 32. At this time, the cutoff frequency (reciprocal of time constant T) of the integrating amplifier 10 is set sufficiently lower than the frequency f ₀ =2V/λ ₀ determined by the center wavelength λ ₀ of the optical pulse to be measured and the moving speed V of the moving table. Ru. Therefore, the data string thus obtained is proportional to 1+2G ₂ (τ).
Here, G ₂ (τ) is the intensity correlation waveform, and using the pulse intensity waveform I(t), G ₂ (τ)=∫ _∞ I(t) ^-∞ I(t-τ)dt/∫ _∞ It is expressed as I(t) ^-∞ I(t)dt.

This completes the measurement stage and begins the analysis stage. Spectrum D initially expressed as a function of wavelength λ
Spectrum D(ω) with (λ) as a function of frequency ω
Convert to . Letting the optical electric field E(t) of the pulse to be measured and its Fourier transform be E〓(ω), E〓 is obtained from D(ω) to the 1/2 power according to |E〓(ω)|=√() The absolute value of (ω) |E〓(ω)| is obtained. After finding the phase φ(ω) of the pulse component when written as E〓(ω)=｜E〓(ω)｜exp〔iφ(ω)〕 (i: imaginary unit), we can calculate the phase of the optical pulse to be measured. The electric field E(t) can be completely reconstructed. Therefore, φ(ω) is expanded around the center wavelength ω ₀ of the spectrum, and φ(ω) = φ ₀ + (ω-ω ₀ ) p + (ω-ω ₀ ) ² q + (ω-ω ₀ ) ³ r In this case, the expansion coefficients, q, r (φ ₀ are arbitrary constants that have no physical meaning, and p only provides correction for the spectral center frequency ω ₀ and has no effect on the pulse intensity waveform. , only q and r can be determined using this method), G _tr (τ) = ∫ ^∞ / _-∞ lEtr (t)
Gtr (τ) calculated according to the following two formulas: l ² lEtr (t - τ) | ² dt／∫ ^∞ / _-∞ lEtr(t)l ⁴ dt is the one that best matches the measured intensity correlation waveform. The least squares method is used. As a result, φ can be obtained with accuracy up to the third order.
(ω) can be found approximately, so according to the above formula,
When Etr(t) is calculated, the intensity waveform of the pulse to be measured is obtained as |Etr(t)| ² , and the phase waveform of the pulse to be measured is obtained as arg(Etr(t)).
By time-differentiating the phase waveform of this pulse, the instantaneous frequency or instantaneous wavelength at each point on the pulse can be obtained. As described above, since only the third-order coefficients of φ(ω) can be obtained at the analysis stage, this method cannot, in principle, faithfully reproduce arbitrary shapes of pulse intensity waveforms and phase waveforms in detail. This can be improved by finding expansion coefficients of φ(ω) up to higher orders, but this is difficult due to the increase in calculation time.

Furthermore, there are many problems involved in using a spectrometer that requires extremely high resolution to measure spectra. For example, the full width at half maximum of the spectrum of an optical pulse with a width of 1 ps with a center wavelength of 0.6 μm is approximately 5 Å. The resolution required for the instrument becomes 0.1 Å or less, and as the width of the optical pulse to be measured becomes longer,
Proportionally good resolution is required.

Spectrometers that satisfy the above specifications include a diffraction grating spectrometer with a focal length of 1 m or more or a swept Fabry-Perot etalon, but the former has a large volume and is difficult to adjust the optical axis, and the latter has a larger capacity than the former. Therefore, careful optical axis adjustment was required, making easy measurement impossible.

As described above, in this conventional short optical pulse measurement evaluation method, in addition to the Michelson interferometer used to measure the intensity correlation waveform of the optical pulse, a separate high-resolution spectrometer is required for spectrum measurement. However, most of the volume and cost of measurement and evaluation equipment is occupied by the spectrometer, which hinders miniaturization, lower prices, and easier adjustment. Furthermore, the accompanying measurement data analysis method remains at a rough approximation, making detailed evaluation of the optical pulse impossible.

On the other hand, there is an invention described in Japanese Patent Application No. 61-211100 as a short optical pulse measurement and evaluation apparatus that does not use a spectrometer, and a schematic diagram thereof is shown in FIG. 3 as a second conventional example. This method is a method for measuring and evaluating short optical pulses that repeat at a constant period.The light beam to be measured is branched into two light beams using a semi-transparent mirror and sent through different paths, thereby reducing the relative optical path length difference. After the addition, the two light beams are combined again, the combined light beam is focused into a crystal that has the ability to generate second harmonic light, and the generated second harmonic light intensity is converted into a voltage value by a photodetector. In the optical system, the relative optical path length difference is changed at a slow constant speed, the output voltage value is recorded in time series, and the frequency interval of the interference fringe of the fundamental wave is obtained by Fourier analysis of the collected data. After raising the spectrum near zero frequency to the 1/2 power among the three distributed spectra, the intensity waveform of the measured wavelength optical pulse is reconstructed by inverse Fourier transform, and then the fundamental wave is The phase of the optical pulse to be measured is reconstructed through repeated calculations using twice the frequency of the interference fringe, the spectrum near the frequency raised to the 1/2 power, and the already determined intensity waveform. Now, 2
Let us define the harmonic optical electric field as a function of time t, u(t)
Then, the 1/2 power of the spectrum near the frequency twice the frequency of the interference fringe of the fundamental wave above is u
The absolute value of the Fourier transform u〓(ω) of (t) is |u〓(ω)|, and the 1/2 power of the spectrum near zero frequency is
Fourier transform I(ω) of intensity waveform I(t)=|u(t)|
The absolute value of |I〓(ω)| is. If u(t) is a so-called symmetrical pulse that is symmetrical with respect to a certain time, the above |u〓(ω)| and |I〓(ω)| are the pulse intensity waveforms I( t) and phase waveform (1/2) arg (u
(t)). The pulse intensity waveform and phase waveform reconstructed through repeated calculations using this method are accurate down to extremely detailed points, unlike the approximation down to low orders as in the first conventional example. However, if the pulse is not symmetrical, this alone is insufficient and the pulse cannot be reconstructed correctly. Therefore, although this method can reconstruct the intensity waveform and phase waveform in detail and faithfully for symmetrical pulses, it has the disadvantage that it gives erroneous results if the pulse is asymmetrical.

As described above, although this second conventional short optical pulse measurement and evaluation method is significantly superior to the first conventional measurement and evaluation method in terms of accuracy, compactness, and simplicity, when the pulse is asymmetric, Due to the principle error that occurs, its application has been limited to cases where it can be determined in advance that the pulse to be measured is a symmetrical pulse.

The present invention has been made in view of the above points, and does not require a high-resolution spectrometer that makes measurement and evaluation equipment large and expensive, and can be used in any manner, including when pulses are temporally asymmetric. It is an object of the present invention to provide a short optical pulse measurement and evaluation method that allows detailed measurement and evaluation of optical pulses in the case of.

[Means for solving problems]

Among the above purposes, in order to make the measurement and evaluation device simple and inexpensive, the present inventor used the device of the second conventional example of the prior art and investigated various causes that made it impossible to measure asymmetric pulses. As a result, if the spectrum of the optical pulse to be measured is known in addition to the dependence of the second harmonic intensity measured by the device on the relative optical path length difference, a sufficient set of data can be obtained to reconstruct the asymmetric pulse. I found out that it can be done. However, if a high-resolution spectrometer is introduced to measure the spectrum, the apparatus will become huge and expensive, and the purpose of the invention will not be achieved. In order to solve this problem, as a method that does not use a spectrometer, the present inventor invented a configuration in which the apparatus shown in the conventional example is newly equipped with a photodetector for measuring the fundamental wave light intensity.
If the output of this photodetector is recorded in time series while changing the relative optical path length difference, it is possible to measure the inverse Fourier transform of the spectrum of the optical pulse to be measured (called an interferogram). In particular, the spectrum can be obtained by Fourier transforming the obtained data.

[Effect]

According to the present invention, after a relative optical path length difference (delay time) is imparted by a Michelson interferometer, the superimposed light beam is focused into a crystal having a second harmonic generation ability. A second harmonic light is generated, and the intensity of the second harmonic light and the fundamental wave light having the same wavelength as the original pulsed light flux to be measured is converted into a voltage value and measured using separate photodetectors.

The output voltage of the photodetector that observes the intensity of the fundamental wave light is expressed as 1+Re[G ₁ (τ) exp (−iω ₀ τ)].
Here, the delay time ω ₀ =2πc/λ ₀ is the center angular frequency of the optical pulse to be measured, and the symbol Re means taking the real part of a complex number. Pulse optical electric field E ₀ (t)
When written as E ₀ (t)=Re[E(t)exp(-iω ₀ t)]...(1), G ₁ (τ) is expressed by the following formula and is called an electric field correlation function.

G ₁ (τ) = _∞ ∫ ^-∞ E(t)E ^* (t-τ) dt ...(2) This data G ₁ (τ) is Fourier transformed to obtain the Fourier value of the electric field E(t) of the pulse to be measured. The square of the absolute value of the transformation E〓(ω) is obtained. That is, _∞ ∫ ^-∞ G ₁ (τ)exp(iωτ)dτ = |E〓(ω)| ² ...(3) The right side of the above equation represents the spectrum of the optical pulse. This method of determining the spectrum of light by using an interferometer rather than a spectrometer is called Fourier transform spectroscopy.

Next, a photodetector 9 is used to observe the intensity of the second harmonic.
The output voltage of is proportional to the following formula.

1 + 2G ₂ (τ) + 4Re [F ₁ (τ) exp (−iω ₀ τ)] + R
e
[F ₂ (τ) exp (−2iω ₀ τ)] ...(4) Intensity of optical pulse I(t) and second harmonic optical electric field u(t)
is defined and introduced by the following formula.

I(t)=E(t) ^* (t) ……(5) u(t)=E(t)E(t) ……(6) Using these I and u, G ₂ (τ)( intensity correlation function) and F ₂ (τ) (second harmonic electric field correlation function) are expressed as follows.

G ₂ (τ)= _∞ ∫ ^-∞ I(t) I(t-τ)dt ……(7) F ₂ (τ)= _∞ ∫ ^-∞ u(t)u ^* (t-τ)dt …… (8) If these are Fourier transformed, the square of the absolute value of the Fourier transform I〓(ω) of the intensity I(t) of the optical pulse,
Fourier transform u〓(ω) of second harmonic optical electric field u(t)
The square of the absolute value of is obtained. That is, _∞ ∫ ^-∞ G ₂ (τ)exp(iωτ)dτ = |I〓(ω)｜ ₂ ......(9) _∞ ∫ ^-∞ F ₂ (τ)exp(iωτ)dτ = |u〓(ω) | ₂ ...(10) Here, we considered performing Fourier analysis on the second harmonic light intensity expressed by equation (4) minus the flat DC component. The upper left diagram of FIG. 4 shows the signal of equation (4), and the lower left diagram shows the results of its Fourier analysis. The period of vibration that is obvious at first glance on the signal matches the center wavelength of the light to be measured. This can be calculated by dividing the relative optical path length difference by the speed of light and replacing it with the delay time.
This is the period of optical oscillation of the light to be measured.

When this signal is subjected to Fourier analysis, three peaks appear as shown in the lower left figure. The peak in the middle is due to the vibration component that appeared at first glance in the signal, and the frequency of the optical vibration of the light to be measured is ω ₀
It's nearby. This ω ₀ is called the interference fringe frequency of the fundamental wave. As a result of Fourier analysis, peaks also appear near zero frequency and near 2ω ₀ .
These three peaks originate from the three terms in formula (4), respectively. As mentioned above, in equation (4), G ₂ (τ), F ₁
(τ) and F ₂ (τ) appear with a difference in carrier frequency ω ₀ that changes sufficiently faster than them, so Equation (4)
By performing Fourier analysis on the signal as a whole, these three types can be easily separated. Above, the peak near zero frequency is |I〓(ω)| ² in equation (9)
If we also extract the peak near 2ω ₀ and translate it by −2ω ₀ , we can obtain |u〓(ω)| ² in equation (10).

On the other hand, if the fundamental wave light intensity signal minus the flat DC component is subjected to Fourier analysis, one peak centered at ω ₀ will be obtained, as seen on the right side of FIG. 4. If we translate this by −ω ₀ , we get the equation
In (3), |E〓(ω)| ² is obtained.

Thus, by Fourier analysis of the data on the relative optical path length difference (delay time) dependence of the fundamental wave light intensity and the second harmonic light intensity, |I〓| ² , |u〓| ² , and |E〓| ² are obtained.
This situation is shown in Figure 4.

Next, in the present invention, the thus obtained |I〓|, |
An iterative calculation method is performed to reconstruct the optical pulse to be measured from u〓|, |E〓|. This calculation method is shown in FIG. In the figure, the symbol FT represents the Fourier transform, and the symbol I.
FT represents the inverse Fourier transform, and csqrt represents the square root of a complex number. Reconstruction of the pulse basically boils down to finding the phase of the Fourier transform E〓(ω) of the electric field E(t). The reason is that the absolute value of the Fourier transform |E〓(ω)| has already been obtained from the measured data, so if we know its phase, we can obtain E〓, and E can be found by inverse Fourier transform. . Therefore, from |E| ² , the intensity waveform of the optical pulse to be measured is
The phase waveform of the pulse to be measured can be obtained from arg(E). The iterative calculation method of the present invention utilizes the relationship between E, I, and u in equations (5) and (6).
It constructs the topology of E〓|, |I〓|, and |u〓|. That is, first, a phase of a random number is given to |E〓(ω)| to create E〓(ω). Using E obtained by inverse Fourier transform, calculations begin repeatedly from the bottom left of Figure 5. From this E, I and u are calculated according to equations (5) and (6), and then each of them is Fourier transformed. In I〓, u〓 obtained in this way, their absolute values |I〓|, |u〓| are replaced with |I〓|, |u〓| obtained up to the previous stage through Fourier analysis of the measurement data.
Here, the difference between |I〓| or |u〓| before and after the replacement is used to determine the convergence of the calculation. When the difference between them disappears, it is considered that the pulse to be measured has been completely reconstructed. Subsequently, those I〓, u〓 are inversely Fourier transformed to return I, u, and the absolute value of u is replaced by I obtained. The square root of u obtained in this way as a complex number is taken and returned to E. Furthermore, the absolute value of E〓 obtained by Fourier transforming E is replaced with the known |E〓|.
The difference between |E〓| before and after replacement is used to determine convergence, as in the case of I〓 and u〓. If E〓 after substitution is subjected to inverse Fourier transform, E can be obtained by returning to the starting point.
The E obtained by going through this calculation loop gives an electric field closer to the electric field of the optical pulse to be measured than the original E.
The degree of approximation is reflected in the differences between the absolute values of the three Fourier transforms obtained during the calculation loop, so calculations are repeated until these differences become sufficiently small.

The above analysis method is different from the approximate solution up to finite dimensions, such as the analysis method in the first conventional example, in principle, it incorporates up to infinite orders, and the intensity and phase of the pulse to be measured can be determined with high accuracy. Moreover, unlike the second conventional example, its effective range is not limited to symmetrical pulses, and is effective for pulses of completely arbitrary intensity and phase waveform.

〔Example〕

FIG. 1 shows the configuration of an apparatus for implementing the short optical pulse evaluation method according to the present invention. A Michelson interferometer is constituted by a semi-transparent mirror 1, a fixed prism 2, and a movable prism 3 by means of a movable stage 4.
Light from a helium neon laser 11 is used to calibrate the relative optical path length difference of this interferometer. A pulsed beam to be measured enters the Michelson interferometer in parallel with this light. This helium neon laser beam has linear polarization at an angle of 45° to the horizontal plane.
The light passes through a quarter-wave plate 19 placed in the middle of one arm of the interferometer so that its axis faces vertically. The helium neon laser beam emitted from the interferometer enters a polarization beam splitter 20 that separates vertically polarized light and horizontally polarized light, and each polarized light is guided to photodetectors 21 and 22. Both detectors output electrical signals proportional to their respective light intensities. Both outputs are input to sample trigger generator 23. The sample trigger generator extracts only the AC component of each input and generates a trigger voltage pulse at the time when the AC component crosses a zero value. Thus, the relative optical path length difference in the interferometer is 0.25× while the oscillation wavelength of the helium-neon laser changes by 0.63 μm.
Observe the generation of trigger voltage pulses a total of 4 times every 0.63 μm. This trigger voltage pulse is supplied to the waveform storage device 17, and the waveform storage device samples and records the input voltage value to each channel, which will be described later, at the time when the trigger voltage pulse occurs.

The pulsed light flux to be measured is reduced to 1/4 of the above amount in the interferometer.
passes through the interferometer without passing through the wave plate 19,
A lens 5 focuses the light in a crystal 7 (for example, a lithium iodate crystal) capable of generating second harmonic light, and the generation of second harmonic light is observed. The light emitted from the crystal 7 is made into a focused beam by the lens 6 and divided into two by the semi-transparent mirror 24. One of them is 2 which has twice the frequency (half the wavelength) of the pulsed light flux to be measured.
Only the harmonic light is extracted by the optical filter 8 and focused on the photodetector 9. The other one is 2 in 7 crystals.
Only the fundamental wave light having the same wavelength as the original pulsed light flux to be measured, which is emitted without being converted into harmonic light, is extracted by the optical filter 30 and focused on the photodetector 28. The photodetectors 9 and 28 each apply a voltage pulse having an area (time integral value) proportional to the second harmonic optical energy and the fundamental optical energy, and these voltage pulses are applied to the integrating amplifier 1, respectively.
0 and 29, the voltage is converted into a constant voltage proportional to the area, and then input to channels 1 and 2 of the waveform storage device 17. A computer 18 is connected to the waveform storage device 17 for reading and analyzing the stored data after measurement. According to the above explanation, measurement is performed at one position of the movable prism 3, in other words, at one point at one relative optical path length difference. In reality, the moving table 4 is moving at a constant moving speed V, and the difference in optical path length of the interferometer is the wavelength λ _ref of the reference helium neon laser.
Sampling is performed every time the waveform changes by 1/4 and is stored in the waveform storage device 17.

At this time, the time constant T of the integrating amplifiers 10 and 29,
The repetition interval P of the pulse to be measured and the moving speed V of the moving table are set so as to satisfy the following relationship: P<<T<<λ _ref /(8V) (11). Further, the wavelength λ _ref of the reference laser must be shorter than the center waveform λ ₀ of the pulse to be measured. This means that one sample is taken every time the relative optical path length difference of the interferometer changes by λ _ref /4, while the second harmonic with a wavelength of half of the fundamental wave of wavelength λ ₀ , i.e. λ ₀ /2, is correctly sampled. This is due to the fact that in order to observe it, samples must be taken at least twice while the relative optical path length difference changes by λ ₀ /2. As a result, the oscillation wavelength is 0.63μm.
When using a helium-neon laser as a reference for the optical path length difference of an interferometer, it is possible to measure optical pulses with a center wavelength longer than 0.63 μm. To measure a light pulse with a shorter wavelength, a laser with a shorter oscillation wavelength may be used as a reference.

In the following analysis step, first, the output voltage changes of the integrating amplifiers 10 and 29 stored in the waveform storage device 17 are read into the computer. These two sets of data are the correlation data shown in the upper part of FIG.

The spectra obtained by Fourier analysis of this data are shown in the lower part of FIG. 4. By using the data of |I〓(ω)|, |u〓(ω)|, and |E〓(ω)| obtained in this way, by performing the iterative calculation shown in Fig. 5,
The intensity waveform and phase waveform of the pulse are determined.

As a verification of the effectiveness of the above analysis method, data |I〓(ω)|, |u〓(ω
)
|, |E〓(ω)| are generated and intensity waveforms and phase waveforms are reproduced from them using the above analysis method. The results are shown in FIGS. 6 to 8.

The right half of each figure shows data assumed for each original waveform.

Regarding |E〓(ω)| and |u〓(ω)|, the values obtained from the spectrum analysis of the measured data are shown after parallel translation by −ω ₀ and −2ω ₀ , respectively.

Figure 6 shows the case where the original waveform is symmetrical, Figure 7,
FIG. 8 shows the case of asymmetric original intensity waveforms and original phase waveforms. From these results, it has become clear that both the intensity waveform and the phase waveform can be reproduced with high precision, and that even if the pulses are asymmetrical, there is no decrease in precision.

FIG. 9 shows intensity waveforms and phase waveforms obtained by measuring and evaluating optical pulses generated by a semiconductor laser using the short optical pulse evaluation method according to this embodiment. The semiconductor laser used is InGaAsP with a wavelength of 1.3 μm.
Using a DFB laser, a sine wave with a frequency of 200 MHz was superimposed on the injected current and modulated to generate pulses with a pulse repetition interval P = 5 ns and a pulse width of 20 to 30 ps. The moving table speed V is 3.88 mm/min,
Measurements were carried out under conditions in which interference fringes repeated at the wavelength of the fundamental wave appeared in the signal at a frequency of 2V/λ ₀ =2×3.88/60/1.3×10 ^-3 =99.5Hz.

Here, the time constant T of the integrating amplifier is set to about 100 μm so as to satisfy equation (11). Further, λ _ref /8V is approximately 1 ms, and the repetition interval P of the pulse to be measured is 5 ns as described above.

From the analysis results in FIG. 9, it can be seen that the pulse to be measured has a steep rise and a gentle tail, and the full width at half maximum is 23 ps.

It can also be seen that the phase waveform changes in a manner similar to the integrated intensity waveform.

〔Effect of the invention〕

As is clear from the above explanation, according to the short optical pulse evaluation method according to the present invention, the intensity waveform and phase waveform of the optical pulse to be measured can be measured with higher precision than the conventional method using a small and inexpensive measuring device. It becomes possible to ask for it.

The lower limit of the time width of a light pulse that can be measured by the present invention is considered to be about 10 times the period of optical oscillation of the light pulse. This is because it is necessary for the three peaks to appear sufficiently separately and independently when performing spectrum analysis of second harmonic light intensity changes.

Therefore, for example, for an optical pulse in the 1.3 μm band, the lower limit of the measurable time width of the optical pulse is 40 fs (40 ×
10 ^-15 s), or about 20fs in the 0.6μm band.

On the other hand, there is no upper limit in principle to the time width of a light pulse that can be measured by the present invention, and any long light pulse can be measured. From a practical point of view, this method can be implemented using small and inexpensive equipment, so
It is thought that it can be used to measure optical pulses up to a certain degree.

Furthermore, the present invention can measure optical pulses of any wavelength as long as a crystal with an appropriate second harmonic generation ability exists. As an example, when a β-BaB ₂ O ₄ (β-barium borate) crystal is used, it is possible to measure at a wavelength of 0.4 to 2.5 μm, and with a KTP (potassium trihydrogen phosphate) crystal, it is possible to measure up to a longer wavelength of 4.2 μm.

As described above, the present invention is applicable to light pulses having a wide range of time widths and wavelengths.

According to the present invention, the intensity and phase waveform can be obtained in detail even if the pulse is asymmetric, so it is extremely effective for adjusting and evaluating short optical pulse light sources.

Furthermore, by measuring optical pulses before and after passing through an optical fiber transmission line using the present invention, changes in their intensity and phase waveform can be observed in detail, making it possible to directly and easily evaluate the characteristics of the optical fiber line. It is. The same thing can be done for general optical systems and optical materials.

Furthermore, detailed observation of the pulse phase in the present invention can be achieved by optimizing a pulse compressor using optical fibers,
Provide a powerful means for coordination.

In this way, the present invention provides short optical pulse generation, shaping,
This is a basic method that is widely used in all fields related to short optical pulses, such as propagation.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of the configuration of an embodiment of the short optical pulse measurement and evaluation method according to the present invention, FIG. 2 is a schematic diagram of the configuration of the first conventional example of the short optical pulse measurement and evaluation method, and FIG. FIG. 2 is a schematic configuration diagram of a conventional example. FIG. 4 is a diagram showing the results of |I〓| ² , |u〓| ² , and |E〓| ² from the data measured by the method of the present invention and its Fourier analysis. FIG. 5 shows an iterative calculation method for reconstructing the original optical pulse to be measured from the above Fourier analysis results. Figures 6 to 8 show the intensity waveforms and phase waveforms (solid lines) of various original pulses used to verify the validity of the analysis method of the embodiment of the present invention, the data sets used for analysis, and the data. FIG. 3 is a diagram showing the intensity waveform and phase waveform (dotted line) of a pulse reconstructed from FIG. 1 using the analysis method of the example. 9th
The figure is a diagram showing an intensity waveform and a phase waveform of a pulse to be measured obtained by analyzing measured data in an example of the present invention. DESCRIPTION OF SYMBOLS 1... Half-transparent mirror, 2, 3... Prism, 4... Moving table, 5, 6... Lens, 7... Crystal having ability to generate second harmonic light, 8... Optical filter, 9...
Photodetector, 10... Integrating amplifier, 11... Helium neon laser, 12, 13... Mirror, 14...
Lens, 15...Photodetector, 16...Amplifier, 1
7...Waveform storage device, 18...Computer 1
8, 19... Quarter wavelength plate, 20... Polarizing beam splitter, 21, 22... Photodetector, 23...
...Sample trigger generator, 24...Semi-transparent mirror, 25
... Lens, 26 ... Spectrometer, 27 ... Wavelength sending device, 28 ... Photodetector, 29 ... Integrating amplifier,
30...Optical filter, 31...Selector switch,
32...A/D converter.

Claims (1)

[Scope of Claims] 1. Branching the optical pulse beam to be measured into two beams, combining the two beams while imparting a change in relative optical path length difference to the branched optical path, and combining the combined beam into two beams. The second harmonic light is made incident on a crystal having the ability to generate second harmonic light, and is converted into electric signals proportional to the intensities of the fundamental wave light and the second harmonic light that have passed through the crystal, and the relative optical path is Fundamental wave light and 2 for length difference
A short optical pulse measurement and evaluation method characterized by recording each intensity change of harmonic light, performing Fourier analysis of the recorded data, and calculating the intensity waveform and phase waveform of the optical pulse to be measured. 2 From the Fourier analysis of the intensity change data of the fundamental wave light, we found the Fourier transform of the electric field of the measured optical pulse as a spectrum near the fundamental wave optical frequency ω ₀ , and from the Fourier analysis of the intensity change data of the second harmonic light, we found the spectrum near 2ω ₀ . A patent claim in which the Fourier transform of the electric field of the second harmonic light is determined as the spectrum of the second harmonic light, and the Fourier transform of the measured optical pulse intensity is determined as the spectrum near zero frequency, and the iterative calculation is performed based on the Fourier transforms of the three components determined above. The short optical pulse measurement and evaluation method according to item 1.