JP2013114065A - Short pulse light generating device and bias voltage evaluation method for optical modulator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a short pulse light generating device that can optimally control a bias voltage applied to an optical modulator without directly measuring an ON/OFF extinction ratio from output light of the optical modulator composing the short pulse light generating device.SOLUTION: The short pulse light generating device comprises: a pulse transmitter 1 that generates short pulse light A having a repetition frequency f; an optical modulator 2 that removes specified short pulse light out of the short pulse light having repetition frequency fon the basis of an input signal, and causes the repetition frequency to be f/n (n is a natural number of 2 or more); a branching unit 3 that branches a part of the output light from the optical modulator 2; a bias control unit 5 that controls a bias voltage applied to the optical modulator 2; and a spectrum analysis unit 4 that analyses spectrum of light B2 branched by the branching unit 3. The bias control unit 5 controls the bias voltage so as to minimize a difference between the maximum value and the minimum value of light intensity of light incident to the spectrum analysis unit 4 in a desired spectrum frequency range.

Description

The present invention relates to a short pulse light generator, and in particular, removes a specific short pulse light from a short pulse light generated at a repetition frequency f ₀ and sets the repetition frequency to f ₀ / n (n is a natural number of 2 or more). The present invention relates to a short-pulse light generator that generates short pulses. The present invention also relates to a bias voltage evaluation method for an optical modulator that can be used in the short pulse generator.

  Consists of a combination of an optical frequency comb generator composed of a continuous light laser light source, an external EO modulator ("electro-optical modulator" or simply "optical modulator"), and a chirp compensator The short pulse light source has characteristics such as a high repetition frequency, high pulse shape and power stability, and low jitter, and is suitable as a terahertz spectroscopic light source, a frequency reference light source, and various measurement light sources (Non-Patent Document 1 or 2). reference).

  In the case of this method, when the modulation frequency is increased, there is an advantage that the spectrum of the modulated light becomes wider and shorter pulses can be obtained. Therefore, it is desirable to drive at a modulation frequency greater than 10 GHz.

  In this case, since the modulation frequency becomes the repetition frequency as it is, the repetition frequency also increases. Further, when the repetition frequency is large, the response of the optical amplifier does not follow, and the signal processing speed cannot keep up. On the other hand, a high repetition frequency makes it difficult to make a mode-locked laser. Considering these, there is a need to make a short pulse light source with a repetition frequency of about 100 MHz to 1 GHz.

  In order to reduce the repetition frequency, a pulse is picked out using a pulse picker. As the pulse picker, a modulator using an acousto-optic (AO) effect as shown in Patent Document 1 is used. However, since the operating band is limited to several hundred MHz, it cannot be used for dividing a high repetition frequency.

  In addition, if unnecessary light remains after thinning, it causes noise in measurement applications and the like, so the pulse picker modulator is required to thin out pulses with a high extinction ratio. For communication applications, 20 dB is sufficient, but for a pulse picker, 25 dB or more is required. As a modulator capable of obtaining such a high extinction ratio, for example, a nested modulator as shown in Patent Document 2 is known.

  It is not easy to match the pulse picker modulator to the optimum operating point in high-speed repetition with an extinction ratio of 25 dB or more and 1 GHz or more. Although a high-speed waveform can be seen with a digital sampling oscilloscope, since the dynamic range is about 10 dB, it is not easy to adjust the bias based on the high-speed waveform so as to obtain the maximum extinction ratio.

  In the case of a short pulse light source using a pulse picker, since it is composed of a combination of several modulators, the ON / OFF extinction ratio is statically measured for each device, and the ON / OFF of the short pulse light is accumulated by their accumulation. The extinction ratio can be estimated.

  However, the optical modulator used for the pulse picker needs to adjust the modulation degree, the bias, the skew, and the like. Unless these are optimized, the maximum characteristics cannot be obtained. Thus, since the ON / OFF extinction ratio after the pulse picker cannot be directly measured, conventionally, the optical modulator constituting the pulse picker could not be set to the optimum driving condition.

  In particular, since a high-speed LN modulator used as a pulse picker has a drift, it is necessary to control the bias so that it is always at an optimum point, and there is a means for monitoring the extinction ratio at a high speed and a high extinction ratio level. is necessary.

  On the other hand, as a conventional technique related to the driving and control of an optical modulator, as shown in Patent Document 3, a low frequency signal is superimposed on a driving signal corresponding to input data and signal light output from the optical modulator is output. A technique for controlling a DC bias voltage for adjusting an operating point of an optical modulator based on a detection result of a low-frequency signal component included in (modulated light) is known. However, when operating at a high extinction ratio as in this case, it is not desirable to use this method because the extinction ratio itself deteriorates due to the superposition of low frequency signals.

Japanese Patent No. 3592475 JP 2006-242975 A Japanese Patent No. 2642499

T. Sakamoto, T Kawanishi, and M. Tsuchiya, “10 GHz, 2.4 ps pulse generation using a single-stage dual-drive Mach-Zehnder modulator”, Opt. Lett. 33, pp.890-892, 2008 Y. Takita, F. Futami, M. Doi, and S. Watanabe, “Highly stable ultra-short pulse generation by filtering out flatoptical frequency components”, CLEO 2004 CTuN1, 2004

  The problem to be solved by the present invention is to solve the above-described problems and monitor the ON / OFF extinction ratio at a high speed and a high extinction ratio level from the output light of the pulse picker light modulator constituting the short pulse light generator. Another object of the present invention is to provide a short-pulse light generator capable of optimally controlling the bias voltage applied to the optical modulator. Another object of the present invention is to provide a bias voltage evaluation method for an optical modulator that can be used in a short pulse light generator.

In order to solve the above-described problem, the invention according to claim 1 is directed to a pulse transmitter that generates short pulse light at a repetition frequency f ₀ , and a short pulse light having a repetition frequency f ₀ based on an input signal. An optical modulator that removes pulsed light and has a repetition frequency of f ₀ / n (n is a natural number of 2 or more), a branching unit that branches a part of output light from the optical modulator, and the optical modulator A short-pulse light generator including a bias control unit that controls a bias voltage applied to the light source and a spectrum analysis unit that analyzes a spectrum of light branched by the branch unit, wherein the bias control unit includes the spectrum analysis unit. The bias voltage is set to be controlled so as to minimize a difference between a maximum value and a minimum value of light intensity in a desired spectral frequency range of light incident on the unit.

  The invention according to claim 2 is the short-pulse light generating device according to claim 1, wherein the spectral intensity distribution of the light has peaks at both ends of the frequency axis, and a region sandwiched between two peaks is the A concave shape is formed with respect to the peak, and the desired spectral frequency range exists in the concave shape portion excluding each peak.

  According to a third aspect of the present invention, in the short pulse light generation device according to the first aspect, the desired spectral frequency range is centered on the center frequency of the short pulse light input to the optical modulator. 2/3 times the frequency width connecting points obtained by subtracting a predetermined intensity from the maximum value of the light intensity in the spectrum intensity distribution at both ends of the envelope of the spectrum intensity distribution of the light incident on It is set within the width.

The invention according to claim 4, characterized in that the short-pulse light generator according to any of claims 1 to 3, wherein a desired spectral frequency range, that is covering the wider range than the repetition frequency f ₀ And

  According to a fifth aspect of the present invention, in the short pulse light generation device according to any one of the first to fourth aspects, the difference between the maximum value and the minimum value of the light intensity is a dispersion spectroscopy system having a resolution of 0.1 nm or less. It analyzes using the spectrum-analysis apparatus of.

According to a sixth aspect of the present invention, a short pulse light having a repetition frequency f ₀ is incident on an optical modulator, and the optical modulator is caused to enter a specific short pulse among the short pulse lights having the repetition frequency f ₀ based on an input signal. The light is removed, the operation is performed so that the repetition frequency of the output light from the optical modulator is f ₀ / n (n is a natural number of 2 or more), and a bias voltage is applied to the optical modulator by the bias control means Then, spectrum analysis means performs spectrum analysis on at least a part of the output light from the optical modulator, operates the bias control means, and operates the light incident on the spectrum analysis means in a desired spectral frequency range. A bias voltage evaluation method for an optical modulator, characterized in that a difference between a maximum value and a minimum value of light intensity is minimized and an applied bias voltage at that time is measured.

The invention according to claim 1, and a pulse generator for generating short optical pulses at a repetition frequency f _0, based on an input signal of the short pulse light of the repetition frequency f _0, to remove certain short optical pulses, repetition An optical modulator having a frequency of f ₀ / n (n is a natural number of 2 or more), a branching unit for branching a part of output light from the optical modulator, and a bias voltage applied to the optical modulator are controlled. A short-pulse light generator having a bias control unit for analyzing the spectrum of the light branched by the branching unit, wherein the bias control unit is configured to transmit the light incident on the spectrum analysis unit. Since the bias voltage is set so as to minimize the difference between the maximum value and the minimum value of the light intensity in the desired spectral frequency range, the ON / OFF cancellation is performed from the output light of the optical modulator. It is possible to optimally control the bias voltage of the optical modulator without directly measuring the optical ratio.

  According to the invention of claim 2, the spectral intensity distribution of light has peaks at both ends of the frequency axis, and a region sandwiched between the two peaks forms a shape that is depressed with respect to the peak, Since the spectral frequency range exists in the depressed shape portion excluding each peak, an indirect physical quantity corresponding to the ON / OFF extinction ratio can be measured more accurately.

  According to the third aspect of the invention, the desired spectral frequency range is centered on the center frequency of the short pulse light input to the optical modulator, and both ends at the envelope of the spectral intensity distribution of the light incident on the spectrum analysis unit. In the spectral intensity distribution, the maximum value of the light intensity is set within a frequency width that is 2/3 times the frequency width connecting the points obtained by subtracting the predetermined intensity from the maximum value of the light intensity in the spectrum intensity distribution. The indirect physical quantity corresponding to the ON / OFF extinction ratio can be measured more accurately by eliminating the influence of the peak portion shown.

The invention according to claim 4, a desired spectral frequency range, because it covers a wider range than the repetition frequency f _0, to eliminate the influence of repetition frequency f _0, indirect corresponding to ON / OFF extinction ratio It becomes possible to measure the physical quantity more accurately.

  According to the fifth aspect of the present invention, the difference between the maximum value and the minimum value of the light intensity is analyzed using a dispersion analysis type spectrum analyzer having a resolution of 0.1 nm or less, and therefore corresponds to the ON / OFF extinction ratio. Indirect physical quantities can be measured more accurately.

The invention according to claim 6, repeated pulsed light frequency f ₀ is incident to the optical modulator, the optical modulator, based on an input signal of the short pulse light of the repetition frequency f _0, the particular short pulse The light is removed, the operation is performed so that the repetition frequency of the output light from the optical modulator is f ₀ / n (n is a natural number of 2 or more), and a bias voltage is applied to the optical modulator by the bias control means Then, spectrum analysis means performs spectrum analysis on at least a part of the output light from the optical modulator, operates the bias control means, and operates the light incident on the spectrum analysis means in a desired spectral frequency range. Since the bias voltage evaluation method of the optical modulator is characterized in that the difference between the maximum value and the minimum value of the light intensity is minimized and the bias voltage applied at that time is measured. It is possible to easily and accurately measure the bias characteristics of the optical modulator used in the installation.

It is a figure explaining the outline of the short pulse light generator of this invention. It is a figure explaining modeling of short pulse light. It is a figure explaining the concept of the measurement spectrum which is the output light of a short pulse light generator. It is a figure which shows the relationship (theoretical curve) between the wave | undulation of a spectrum and an extinction ratio. It is a figure explaining the measurement concept of the spectrum power ratio using an optical spectrum analyzer (OSA). It is a figure which shows the correction | amendment curve which considered the dullness of the peak value in an optical spectrum analyzer (OSA) measurement. It is a measurement example using an optical spectrum analyzer (OSA), and is a diagram showing a state in which a 10 GHz component remains. It is a measurement example using an optical spectrum analyzer (OSA), and shows a state in which the 10 GHz component has disappeared but some undulations on the spectrum remain. It is a measurement example using an optical spectrum analyzer (OSA), and is a diagram showing a state where waviness (difference between a maximum value and a minimum value) on a spectrum is minimized.

Hereinafter, the short pulse light generator of the present invention will be described in detail using preferred examples.
Short pulse light generator of the present invention, as shown in FIG. 1, a pulse generator 1 for generating short optical pulses A at a repetition frequency f _0, based on an input signal of the short pulse light of the repetition frequency f ₀ , An optical modulator 2 that removes specific short pulse light and has a repetition frequency of f ₀ / n (n is a natural number of 2 or more), and a branching unit that branches a part of the output light from the optical modulator 2 3, a bias control unit 5 that controls a bias voltage applied to the optical modulator, and a spectrum analysis unit 4 that analyzes a spectrum of the light B2 branched by the branching unit. The bias control unit 5 controls the bias voltage so as to minimize the difference between the maximum value and the minimum value of the light intensity in the desired spectral frequency range of the light incident on the spectrum analysis unit 4. Set It is characterized by being.
The other output light B1 emitted from the branching unit 3 is used as a short pulse light such as a terahertz spectroscopic light source, a frequency reference light source, or various measurement light sources.

  FIG. 2 shows a model of short pulse light output from the optical modulator 2 constituting the pulse picker. Here, the pulse peak intensity is normalized by 1, and the thinned pulse intensity is a (amplitude is √a). Let n be the number of pulses included in one period T.

The extinction ratio R _E of the optical modulator 2 is expressed by equation (1).
R _E = 10loga (1)

The amplitude of the short pulse train in FIG. 2 is expressed using equations (2) to (4).
h (t) = f (t) × g (t) (2)
f (t) = 1 (0 ≦ t ≦ T / n) or a ^1/2 (T / n ≦ t ≦ T) (3)
g (t) = Σ _{l = −∞} ^∞ δ {t−T / 2n · (2l−1)} (4)
However, δ is a Dirac delta function, and the expression of g (t) is correct when it is assumed that the pulse width is sufficiently small with respect to the period T.

When h (t) in Expression (2) is expressed by Fourier series expansion, it can be expressed as Expressions (5) and (6).
_{h (t) ~Σ m = -∞} ∞ c m exp (im · 2π / T · t) (5)
c _m = 1 / T {exp (−iπ × m / n) + a ^1/2 Σl _{= 2} ⁿ exp (−iπ × m / n (2l−1))} (6)
Here, m is the order of series expansion, and corresponds to each discrete spectrum. C _m is a Fourier coefficient.

Let P _m be the intensity of the discrete spectrum (proportional to the second power of the amplitude) measured by spectral analysis. The intensity P _m of each spectral line is expressed by Equation (7).
P _m ∝ | c _m | ² (7)

When m = 0, n, 2n,..., the value of Expression (8) is obtained.
| C _m | ² = 1 / T ² · (1 + a ^1/2 (n−1)) ² (8)

When m is other than the above, the value of Expression (9) is obtained.
| C _m | ² = 1 / T ² · (1-a ^1/2 ) ² (9)

As shown in FIG. 3, the power difference (spectral power ratio) ΔP between the f ₀ component and the f ₀ / n component on the spectrum can be expressed by equation (10). However, it is assumed that a is sufficiently small.
ΔP [dB] = − 10 log {(1-2 · a ^1/2 ) / (1 + 2 · a ^1/2 (n−1)} (10)

A conceptual diagram of this spectrum is drawn as shown in FIG. Here, when the parenthesis in the log of the above equation is set to p and a is taken, equation (11) is obtained.
a = {(1-p) / 2 (1 + p (n-1))} ² (11)

Using formula (11), the extinction ratio can be obtained from ΔP of the spectrum shown in FIG. Specifically, assuming that f ₀ is 10 GHz, the spectral power ratio ΔP (corresponding to the magnitude of the wave of the spectrum) and the extinction when n = 2, 4, 8 (5.0, 2.5, 1.25 GHz). FIG. 4 shows the result (theoretical curve) of calculating the ratio R _E.

  It can be easily understood from FIG. 4 that the extinction ratio (dB) increases as the magnitude of the spectral waviness (dB) corresponding to the spectral power ratio ΔP decreases.

  Next, spectral analysis means for measuring the spectral power ratio ΔP (difference between the maximum value and the minimum value of light intensity) will be described. In general, there are the following three methods.

(1) Dispersion spectroscopy method A spectrum is spatially divided using a diffraction grating or a prism spectroscopic element, unnecessary light is removed by a slit, and the light intensity of a wavelength to be measured that has passed through the slit is measured. In the past, resolution has been difficult, but with recent advances in spectroscopic techniques, it is possible to measure up to about 0.01 nm (with a frequency of about 1 GHz). Spectral analysis can be realized at high speed and low cost.

(2) Interferometry system This is a Michelson interferometer configuration. Spectral analysis is performed by dividing the measured light into two light beams by a beam splitter, Fourier transforming the interferogram generated when the two light beams are superimposed after being given an optical path difference. The wavelength resolution is better than the dispersion spectroscopy, but the speed is slower than the dispersion spectroscopy, and the power measurement accuracy is also worse than the dispersion spectroscopy.

(3) Heterodyne type This is a method of measuring the spectrum intensity from the intensity of the beat generated when the frequency is brought close to the spectrum to be measured, using a local light source capable of sweeping frequency (wavelength) and having a small line width and little frequency fluctuation. Although the measurement accuracy is the best, high-speed measurement is not possible because local light must be swept. A so-called coherent receiver performs signal detection in this manner.

  When these three methods are compared, the characteristics shown in Table 1 below are shown. The evaluation for each feature is evaluated in the order of ◎ → ◯ → Δ → × from the most excellent characteristic to the inferior characteristic.

  Basically, the method to be selected should be selected according to the application, but in the present invention, the (1) dispersion spectroscopy method is more preferable from the viewpoint of measurement speed and power measurement accuracy. In the case of the heterodyne type, detection measurement is performed by heterodyne detection using a laser light source capable of sweeping the frequency. As shown in Table 1, the wavelength resolution and power measurement accuracy are accurate. However, since it is necessary to sweep the laser beam, it takes a long time to measure, and a separate light source is required, and the measuring device itself is expensive.

On the other hand, the measurement by the dispersion spectrum type optical spectrum analyzer (OSA) having a high resolution of 0.1 nm or less is simple because the optical spectrum can be directly visually observed, but the resolution is limited. Although the measurement accuracy is slightly inferior, it operates at high speed and is suitable for bias control. Since what is currently marketed is 0.01 to 0.02 nm, f ₀ is 1 GHz (when the resolution is 0.01 nm), which is the measurement limit.

Further discussion will be made on the measurement of the extinction ratio using OSA, particularly the dispersion spectroscopy. As shown in FIG. 5, the measurement accuracy depends on the following three conditions for the spectral line width W and the OSA resolution s. However, it is assumed that w << f ₀ / n (1) When f ₀ −w ≦ s, measurement is impossible.
(2) Correction is necessary when f ₀ / nw ≦ s <f ₀ −w. In particular, since power components of peaks other than f ₀ are included, it is necessary to exclude them.

  Since the difference in spectral intensity cannot be measured directly, the extinction ratio cannot be calculated directly from the theoretical formula (the extinction ratio is underestimated in the theoretical formula). However, this method (dispersion spectroscopy) can be used because the intensity difference only needs to be reduced if only the optimum value is found. In order to obtain the extinction ratio, it is necessary to subtract a spectrum component having a small intensity included in the maximum value of the intensity measurement. The result calculated by this method is the correction curve shown in FIG.

(3) When s <f0 / n−w, it is possible to calculate the extinction ratio from the intensity difference of the spectrum using Equation (11). That is, in the case of (3), since the maximum measurement result includes only one spectral component, the spectral intensity of m = 0, n, 2n can be measured accurately.

In general, a commercially available optical spectrum analyzer has a finer sweep step than the resolution, so it is possible to obtain a power value that completely includes the peak of f ₀ , but the sweep step is the same as the resolution. When the boundary of the slit is just over the spectrum, the peak power may straddle two measured values even if the above condition (3) is satisfied. In that case, correction may be performed by adding two corresponding measurement values and setting the resolution to 2 s.

In the case of f ₀ = 10 GHz shown in FIG. 4, correction is necessary when n = 4, 8, and FIG. 6 shows the result calculated in consideration of the correction.

Next, examples in which a short pulse spectrum is measured using a high-resolution optical spectrum analyzer (dispersion spectroscopy) having a resolution of GHz order among commercially available optical spectrum analyzers (OSA) are shown in FIGS. This is a case where f ₀ = 10 GHz and n = 8.

In FIG. 7, a slight f ₀ (10 GHz) component remains when viewed with an oscilloscope (an extinction ratio of about 10 dB is assumed). Further, FIG. 8 shows that 10 GHz is completely invisible on the oscilloscope, but the spectrum waviness is not yet minimal on the spectrum.

  In the above model, since the pulse train is a δ function (equation (4)), the spectrum has an infinite spread, but since the actual short pulse train has a finite width, the spectrum is also finite. . Therefore, since the outer part of the spread away from the center frequency is not consistent with the model, these peripheral parts need to be excluded.

  As shown in FIG. 8, the measured spectral intensity distribution of light has peaks at both ends of the frequency axis, and a region sandwiched between two peaks forms a shape that is recessed with respect to the peak. . The center depressed area excluding each peak is an area where the above model is established, and the entire width (the entire frequency width of the shape indicated by the spectrum intensity distribution as shown in FIG. This corresponds to a region of about 2/3 of a portion lower than the peak intensity by a predetermined intensity, for example, the frequency width of a −40 dB part lower by about 10 dB than the peak value in FIG. Such a recessed region is a portion showing the undulation of the spectrum.

  Further, as a desired spectral frequency range, the spectral intensity is centered on the center frequency of the short pulse light input to the optical modulator and at both ends of the envelope of the spectral intensity distribution of the light incident on the spectral analysis unit. It is also possible to set the frequency width to be within 2/3 times the frequency width obtained by connecting the points obtained by subtracting the predetermined intensity from the maximum value of the light intensity in the distribution. In this case, the predetermined intensity to be subtracted from the maximum value of the light intensity in the spectral intensity distribution may be, for example, about 10 dB, but other intensity such as 3 dB or 5 dB may be used as necessary.

  FIG. 9 shows a state where waviness is minimized on the OSA (corresponding to the best extinction ratio point). In order to minimize the undulation, the portion showing the undulation of the OSA is set as a spectral frequency range to be measured for determining the difference between the maximum value and the minimum value of the light intensity. The bias voltage applied to the optical modulator of the pulse picker is controlled so as to minimize.

  Next, when f0 = 10 GHz and n = 2, 4, and 8, the results of calculating the extinction ratio from the waviness of each spectrum are shown in Table 2.

Further, a desired spectral frequency range, it is necessary to target a wider range than the repetition frequency f _0. This makes it possible to measure the indirect physical quantity corresponding to the ON / OFF extinction ratio more accurately.

  In the above description, the configuration for controlling the bias voltage of the optical modulator incorporated in the short pulse light generator has been described. However, the present invention further relates to the light used in the optical modulator, particularly the short pulse light generator. It is also possible to use the modulator alone for test evaluation (measurement of bias voltage to be applied or measurement of change in bias voltage in a driving environment).

That is, as shown in FIG. 1, a short pulse light A having a repetition frequency f ₀ is incident on the optical modulator 2 to be measured, and the optical modulator 2 is caused to enter the short pulse having the repetition frequency f ₀ based on an input signal. A specific short pulse light is removed from the light, and the repetition frequency of the output light (B1 or B2) from the optical modulator is operated to be f ₀ / n (n is a natural number of 2 or more) A bias voltage is applied to the optical modulator by the control means 5, and at least a part of the output light B2 from the optical modulator (of course, B1 may be directly measured) by the spectrum analysis means 4. Analyzing the spectrum, operating the bias control means, minimizing the difference between the maximum value and the minimum value of the light intensity in the desired spectral frequency range of the light incident on the spectrum analysis means, and applying the bias applied at that time Measure voltage .

  The bias voltage thus measured can be used as a bias voltage evaluation of the optical modulator. For example, by measuring the optimum bias voltage of the optical modulator, it becomes possible to evaluate the ON / OFF extinction ratio of the optical modulator as a pulse picker. Needless to say, for the above-described “desired spectral frequency range” and the like, various methods and concepts described in the short pulse light generator can be used in the same manner.

  As described above, according to the present invention, the ON / OFF extinction ratio is monitored at high speed and with a high extinction ratio level from the output light of the pulse picker optical modulator constituting the short pulse light generator, and the optical modulator is It is possible to provide a short pulse light generator capable of optimally controlling the applied bias voltage. It is also possible to provide a bias voltage evaluation method for an optical modulator that can be used in a short pulse light generator.

1 Pulse transmitter 2 Optical modulator (pulse picker)
3 Branching unit 4 Spectrum analysis unit 5 Bias control unit

Claims (6)

A pulse transmitter for generating short-pulse light at a repetition frequency f ₀ ;
An optical modulator that removes a specific short pulse light from the short pulse light of the repetition frequency f ₀ based on an input signal and sets the repetition frequency to f ₀ / n (n is a natural number of 2 or more);
A branching unit for branching a part of the output light from the optical modulator;
A bias controller for controlling a bias voltage applied to the optical modulator;
A short-pulse light generator having a spectrum analyzing unit for analyzing a spectrum of light branched by the branching unit,
The bias controller is set to control the bias voltage so as to minimize a difference between a maximum value and a minimum value of light intensity in a desired spectral frequency range of light incident on the spectrum analysis unit. A short-pulse light generator characterized by comprising: 2. The short pulse light generator according to claim 1, wherein the spectral intensity distribution of the light has a peak at each end of the frequency axis, and a region sandwiched between the two peaks is recessed with respect to the peak. Formed,
The desired spectral frequency range exists in the recessed shape portion excluding each peak, and the short pulse light generator.   2. The short pulse light generator according to claim 1, wherein the desired spectral frequency range is centered on the center frequency of the short pulse light input to the optical modulator, and the spectral intensity of the light incident on the spectrum analysis unit. At both ends of the distribution envelope, the frequency width is set within 2/3 times the frequency width connecting points obtained by subtracting the predetermined intensity from the maximum value of the light intensity in the spectrum intensity distribution. A short pulse light generator characterized by the above. In short pulse light generating device according to any one of claims 1 to 3, wherein a desired spectral frequency range, short-pulse light generating device characterized in that it is intended for a wider range than the repetition frequency f _0.   5. The short pulse light generator according to claim 1, wherein the difference between the maximum value and the minimum value of the light intensity is analyzed using a spectrum analyzer of a dispersion spectroscopy system having a resolution of 0.1 nm or less. A short-pulse light generator characterized by: A short pulse light having a repetition frequency f ₀ is incident on the optical modulator,
The optical modulator of the short pulse light of the repetition frequency f ₀ based on the input signal, and remove certain short optical pulses, the repetition frequency of the output light from the light modulator f _{0 /} n (n is (Natural number of 2 or more)
A bias voltage is applied to the optical modulator by a bias control means,
Spectral analysis of at least part of the output light from the optical modulator by spectral analysis means,
The bias control means is operated, the difference between the maximum value and the minimum value of the light intensity in the desired spectral frequency range of the light incident on the spectrum analysis means is minimized, and the applied bias voltage at that time is measured. A bias voltage evaluation method for an optical modulator.

Images