Patent Application
20210408755
The present invention relates to a swept light source, and more specifically, a swept light source usable in an optical coherence tomography device.
A swept light source is widely used for optical devices and electronic devices utilizing an optical imaging technique. For example, it is most frequently used for electronic devices for consumers such as cameras, printers, and facsimiles. Its use is also spreading to the medical field. For noninvasive imaging of sections in the body, attention has been focused on optical coherence tomography (OCT), which enables acquisition of information in the depth direction with high resolution by using low-coherence light. Recently, OCT is used also in the field of molecular imaging for interpreting biological phenomena at the molecular level and cell level in the body. The swept light source is an important and basic element for supporting the OCT technique as will be described later and has a great influence over the performance of an OCT device.
The molecular imaging described above mainly uses a method of using optical information to detect target molecules with high sensitivity, in which the OCT device is used. The types of OCT devices for acquiring section images by using coherent interference include time domain optical coherence tomography (TD-OCT) and Fourier domain optical coherence tomography (FD-OCT). FD-OCT is further classified into spectral domain optical coherence tomography (SD-OCT) and swept-source optical coherence tomography (SS-OCT). SS-OCT using the swept light source is particularly excellent in terms of high-speed responsivity.
Since the speed of response of SS-OCT is largely limited according to the performance of a light source, various types of high-speed and wide-range light sources are increasingly developed. In addition, in terms of an aspect of a sweep, an operation according to an application field is required for the swept light source like a wavenumber-linear light source in which a swept wavenumber linearly varies with time.
PTL 1 discloses a configuration example of a swept light source using an optical fiber loop. With reference to
PTL 1: Japanese Patent Laid-Open No. 2005-347668
PTL 2: Japanese Patent Laid-Open No. 2011-186218
PTL 3: Japanese Patent Laid-Open No. 2015-142111
NPL 1: Yoshiaki Yasuno et al., “Three-dimensional and high-speed swept-source optical coherence tomography for in vivo investigation of human anterior eye segments,” OPTICS EXPRESS, Vol. 13, No. 26, pp. 10652-10664, 2005
In the swept light source as disclosed in PTL 1, a loop gain and a lasing threshold generally differ according to wavelength. In the case of a laser, lasing occurs within such a wavelength range that a loop gain spectrum is larger than a lasing threshold (loop gain spectrum>lasing threshold). This wavelength range means a range of a loop gain larger than the lasing threshold (=loss). The wavelength range is hereinafter referred to as an unsaturated gain range.
In the swept light source for SS-OCT, it is preferable to keep a coherence length long at all swept wavelengths. To realize this, it is important that the unsaturated gain range (lasing possible wavelength width) is narrow at all swept wavelengths.
Since the unsaturated gain range is a wavelength width in which the loop gain spectrum exceeds the lasing threshold, it is only necessary to create such an operating state that the peak of the loop gain spectrum corresponds to the lasing spectrum for all swept wavelengths. However, it is difficult for a conventional swept light source to realize such an operating state because an injected current into a gain medium is temporally constant. In a laser oscillator, an injected current into a gain medium determines the shape and peak value of a loop gain spectrum. Accordingly, in the conventional swept light source in which the injected current is temporally constant, the loop gain peak value does not necessarily correspond to the lasing threshold at all swept wavelengths. This reason can be further explained as follows.
The loop gain and the lasing threshold are determined by a gain of a gain medium, various optical losses in a laser resonator, filtering characteristics of a band pass filter, a time spendable for lasing of each wavelength and the like. For example, in the swept light source disclosed in PTL 1, a time spendable for lasing of each wavelength and filtering characteristics of a band pass filter are constant irrespective of wavelength, whereas a gain of a gain medium decreases with distance from the center wavelength of the swept wavelengths. As a result, the lasing threshold is small near the center wavelength and increases with distance from the center wavelength.
In the conventional swept light source, since the injected current into the gain medium is constant for any swept wavelength, it is necessary to determine the injected current into the gain medium according to a wavelength with a large lasing threshold in order to secure a swept wavelength width necessary for a wide range (for example, about 100 nm in SS-OCT). When a constant current is thus injected into the gain medium, the wavelength width of the loop gain above the threshold (unsaturated gain range) is large at a wavelength with a small lasing threshold. In the swept light source disclosed in PTL 1, because of the injected current into the gain medium capable of lasing at the shortest swept wavelength and the longest swept wavelength, a coherence length is long near the shortest wavelength and the longest wavelength. However, the injected current becomes excessive with increasing proximity to the center wavelength and the peak value of the loop gain exceeds the lasing threshold to a considerable degree. Accordingly, roughly at the center of the sweep wavelength range, the unsaturated gain range is wide, the linewidth is wide, and the coherence length is short.
It is know that in the SS-OCT device, the strength of an interference signal increases as the coherence length of a beam from the swept light source becomes longer. The strength of one pixel of an OCT image obtained by frequency conversion of an interference signal, that is, a point spread function (PSF) is proportional to the energy of the interference signal (time integration value). Accordingly, in order to increase the strength of PSF, it is necessary to strengthen the entire interference signal and keep the coherence length long in the entire sweep wavelength range. In the conventional swept light source in which a temporally-constant current is injected into the gain medium, the coherence length cannot be kept long in the entire sweep wavelength range. This causes a problem that PSF strength that directly affects the quality of output images of the SS-OCT device declines.
If the PSF strength declines in the SS-OCT device, a difference between the PSF peak of a detected image and system noise becomes small, and accordingly, a weak detection signal from a target is buried into the system noise. This causes a problem that the image quality of the SS-OCT device deteriorates and the image detection sensitivity decreases.
The present invention has been accomplished in consideration of the above problems. An object of the present invention is to provide a light source capable of keeping a coherence length long over an entire wide sweep wavelength range. Description will also be provided of a method of generating semiconductor optical amplifier (SOA) current waveform data (optical gain control data) for optimally operating the swept light source of the present invention, a method of calculating the coherence length, and an optical deflector suitable for implementing the swept light source of the present invention.
An aspect of the present invention is a swept light source configured to output a beam having a continuously changing lasing wavelength, the swept light source comprising: a laser resonator including a gain medium configured to amplify a beam, means for changing a lasing wavelength, and means for changing an optical gain of the gain medium; wavelength control means configured to control the means for changing a lasing wavelength by using a wavelength control signal and sweep the lasing wavelength; and optical gain control means configured to control the means for changing an optical gain so that an optical gain in the laser resonator has an optical gain value slightly higher than an optical gain in a state of a lasing threshold over an entire wavelength range of a wavelength sweep.
According to another aspect, the optical gain control means can be configured to control the means for changing an optical gain by using an optical gain control signal based on optical gain control data obtained by combining a plurality of times from a start of a sweep obtained from the wavelength control means with an optical gain at which a coherence length of a wavelength swept beam becomes the longest at each of the times.
The means for changing a lasing wavelength can preferably include: a diffraction grating; a mirror on which a laser beam from the diffraction grating is incident and the incident laser beam is perpendicularly reflected; and an optical deflector configured to change an angle of incidence of the laser beam on the diffraction grating.
According to yet another aspect, the optical gain control means can be configured to detect an output light level from the laser resonator, generate an optical gain control signal based on the detected output light level so that the output light level has a predetermined output light level value, and control the means for changing an optical gain by using the optical gain control signal.
The means for changing a lasing wavelength can preferably include: a diffraction grating; and an optical deflector configured to change an angle of incident of a laser beam on the diffraction grating.
Another aspect of the present invention is a swept light source configured to output a beam having a continuously changing lasing wavelength, the swept light source comprising: a laser resonator including a gain medium configured to amplify a beam, means for changing a lasing wavelength, and means for changing an optical gain of the gain medium; wavelength control means configured to control the means for changing a lasing wavelength by using a wavelength control signal and sweep the lasing wavelength; and means for changing the optical gain by using an optical gain control signal based on optical gain control data obtained by combining a plurality of times from a start of a sweep obtained from the wavelength control means with an optical gain at which a coherence length of the wavelength swept beam becomes the longest at each of the times.
According to yet another aspect of the present invention, the optical deflector can comprise: an optical deflector with at least two electrodes formed on opposing surfaces of an electro-optical crystal, wherein in a case where a voltage is applied to the at least two electrodes, a non-uniform refractive index distribution occurs in the electro-optical crystal due to an electro-optic effect, thereby bending an optical path perpendicular to an electric field formed by the voltage; and an optical element arranged on at least one of an incident side and an outgoing side of the electro-optical crystal, the optical element having light condensing characteristics spatially distributed over a range of deflection of the beam.
Light condensing characteristics of the optical deflector can preferably change according to an angle of deflection and optical characteristics of the optical element can preferably be spatially distributed so as to compensate for a change in at least one of a beam waist position and a beam waist diameter of the outgoing side over the range of deflection of the beam.
As described above, the swept light source, the method of generating drive data for the swept light source, and the optical deflector of the present invention can constantly keep a coherence length of an output beam long during a wavelength sweep period.
An aspect of the present invention is a swept light source configured to output a beam having a continuously changing lasing wavelength, the swept light source comprising: a laser resonator including a gain medium configured to amplify a beam, means for changing a lasing wavelength, and means for changing an optical gain of the gain medium; wavelength control means configured to control the means for changing a lasing wavelength by using a wavelength control signal and sweep the lasing wavelength; and optical gain control means configured to control the means for changing an optical gain so that an optical gain in the laser resonator has an optical gain value slightly higher than an optical gain in a state of a lasing threshold over an entire wavelength range of a wavelength sweep.
For the optical gain control means, configuration examples according to different methods (first and second embodiments) are described. One is an example of calculating an SOA current for keeping a coherence length maximum in advance and operating the light source based on SOA current waveform data. The other is an example of detecting an output light level from the laser resonator and controlling an optical gain (SOA current) based on the detected output light level so that the output light level is a predetermined output light level.
In addition, various configuration examples of an optical deflector suitable for implementing the swept light source of the invention are described.
A swept light source of a first embodiment of the present invention can keep a coherence length of an output beam long over an entire sweep wavelength range. A gain of a gain medium is changed with time in response to a wavelength sweep, whereby the coherence length is kept maximum. That is, it operates so that the gain of the gain medium is kept close to an lasing threshold at each wavelength over the entire sweep wavelength range and an unsaturated gain range of the gain medium is narrowed. In other words, the optical gain is controlled so that a gain (optical gain) in the laser resonator has a gain value slightly higher than a gain in a state of the lasing threshold (lasing threshold gain or lasing gain). A linewidth of an output laser beam is constantly kept narrow by controlling the laser gain control means so as to further reduce a loop gain roughly at the center of the sweep wavelength range, which has an excessive gain in a conventional swept light source.
Although not limited to this, for example, the loop gain of the gain medium can be changed by an SOA current injected into the gain medium in the laser resonator to flow through a semiconductor optical amplifier (SOA). An SOA current for keeping a coherence length maximum in the swept light source of the present invention is generated based on SOA current waveform data. The SOA current waveform data is obtained by calculating a coherence length corresponding to each wavelength from a coherent beam of a light source output beam while associating different SOA currents with sweep time information in a state of sweeping the swept light source. At each wavelength, an SOA current for providing the maximum coherence length is selected from different SOA currents. A coherence length is calculated based on a plurality of coherent beams obtained from a wavelength swept beam while the swept light source is kept in a normal sweep operating state, and then an SOA current for keeping the coherence length maximum is calculated. The SOA current for keeping the coherence length maximum is supplied from a current source capable of outputting an arbitrary current waveform. The swept light source operates based on SOA current waveform data for implementing the SOA current waveform stored in storage means (memory). The gain of the gain medium of the laser resonator can be changed also by gain changeable means other than the SOA drive current.
The present invention can also be implemented as a method of obtaining SOA current waveform data for driving the swept light source while keeping a coherence length of an output laser beam long over an entire sweep wavelength range. Description is also provided of a novel method of measuring a coherence length for determining the SOA current waveform data suitable for the swept light source of the present invention and unique to the present invention. The swept light source of the present invention, the method of generating drive data for the swept light source, and the method of measuring the coherence length will be described below along with each example. First, the configuration and operation of the swept light source of the present invention are described as Example 1 of the swept light source of the first embodiment.
The beam deflected by the optical deflector 106 is diffracted by a diffraction grating 107 and reaches a mirror 108. Only a beam of a wavelength perpendicularly incident on the mirror 108 follows an optical path opposite to the optical path described above and returns to the SOA 102. The beam returning from the mirror 108 to the SOA 102 is reflected on a half mirror 104 between the SOA 102 and a coupler 103. A round-trip optical path is formed between the mirror 108 and the half mirror 104 and lasing occurs. Part of the laser beam passes from the SOA 102 through the half mirror 104 and is output as an output beam to the outside of the laser resonator 101 through the coupler 103 and an optical fiber 120.
As described above, since the round-trip optical path is formed only from wavelength components perpendicularly incident on and reflected on the mirror 108 and the lasing state arises, a lasing wavelength can be changed by an angle of incidence of a beam on the diffraction grating 107. For example, a configuration between the SOA 102 and the mirror 108 forms a band pass filter substantially in the form of a quadratic function and determines an unsaturated gain range of a loop gain.
The optical deflector driver 113 controls a direction of deflection (angle of deflection) of an outgoing beam from the optical deflector 106 by using the drive signal 115. As the drive signal 115 changes with time, the direction of deflection also gradually changes. In the swept light source, the drive signal changes periodically and repeatedly so as to smoothly increase or decrease a wavelength of a laser beam in only one direction and sweep the wavelength. In general, a predetermined interval is placed between a wavelength sweep and the next wavelength sweep. That is, part of a repeating period is used for an actual wavelength sweep. As described above, in the laser resonator 101, the lasing wavelength is changed temporally continuously by the drive signal 115 to implement a wavelength sweep operation.
The optical deflector driver 113 outputs a synchronizing signal 114 synchronized with a wavelength sweep together with the drive signal. For example, in the case of a wavelength sweep from a short wavelength to a long wavelength, the synchronizing signal 114 is output to an arbitrary waveform current source 111 at a time when an angle of deflection corresponds to the shortest wavelength.
The arbitrary waveform current source 111 has SOA current waveform data for generating an SOA current 116 to be used during a wavelength sweep operation and applied to the SOA 102 in an associated memory (storage means) not shown in
The memory may be provided in the arbitrary waveform current source 111, or in any place in the swept light source 100 outside the arbitrary waveform current source 111. The arbitrary waveform current source 111 generates an SOA current waveform from the current waveform data or information stored in the memory in synchronization with the synchronizing signal 114 from the optical deflector driver 113, and allows an SOA current according to the SOA current waveform to flow through the SOA 102. Incidentally, the SOA 102 is generally connected to a temperature controller configured to control the temperature of the SOA, but it is omitted here. The same applies to SOAs described hereinafter. In general, the swept light source further comprises a control unit including a CPU for control of the entire light source including control of a wavelength sweep operation. The control unit generally includes a memory, and an electronically controlled device that operates by using data stored in or read from a memory generally comprises a control unit. Accordingly, it should be noted that the control unit is not shown in
As the information stored in the memory associated with the arbitrary waveform current source 111, time-waveform per se may be stored, or data in an arbitrary form for generating time-wavelength may be stored. That is, time-waveform data per se, time-waveform data and a coefficient combined therewith, time-waveform data and a bias combined therewith, or a combination of time-waveform data, a coefficient, and a bias may be output.
In addition, as the information stored in the memory associated with the arbitrary waveform current source 111, data obtained by the Fourier transform of a current waveform may be stored. Also in this case, data obtained by the inverse Fourier transform of the data obtained by the Fourier transform per se, data obtained by the inverse Fourier transform and a coefficient combined therewith, data obtained by the inverse Fourier transform and a bias combined therewith, or a combination of data obtained by the inverse Fourier transform, a coefficient, and a bias may be output.
When the swept light source of the present invention is operated, the optical deflector driver 113 first causes the optical deflector 106 to perform a deflection operation. Upon the start of the deflection operation, the arbitrary waveform current source 111 loads current waveform data to be applied to the SOA 102 into the memory associated with the arbitrary waveform current source 111. The optical deflector driver 113 causes the optical deflector 106 to perform the deflection operation. The arbitrary waveform current source 111 applies an SOA current 116 corresponding to the current waveform data to the SOA 102 in synchronization with the synchronizing signal 114 from the optical deflector driver 111 based on the current waveform data stored in the memory. The operation principle in the present invention will be described in more detail.
As described above, in the swept light source used for the SS-OCT device, it is necessary that a coherence length of a laser beam from the laser resonator be kept long and a linewidth be as narrow as possible in the entire sweep wavelength range. In the configuration of the swept light source of the present invention shown in
The swept light source 100 of this example shown in
In the swept light source 100 of
For example, the means for changing a wavelength receives a wavelength control signal (deflector drive signal 115) output from wavelength control means corresponding to the optical deflector driver 113 in
As the optical deflector, for example, a KTN deflector using a potassium tantalate niobate (KTN) crystal as an electrooptic material can be used. Besides KTN, the following electrooptic materials can also be used: KLTN (K1-yLiyTa1-xNbxO3 [0<x<1, 0<y<1]), LiNbO3 (hereinafter referred to as LN), LiTaO3, LiIO3, KNbO3, KTiOPO4, BaTiO3, SrTiO3, Ba1-xSrxTiO3 (0<x<1), Ba1-xSrxNb2O6 (0<x<1), Sr0.75Ba0.25Nb2O6, Pb1-yLayTi1-xZrxO3 (0<x<1, 0<y<1), Pb(Mg1-2Nb2/3)O3-PbTiO3, KH2PO4, KD2PO4 (D is deuterium), (NH4)H2PO4, BaB2O4, LiB3O5, CsLiB6O10, GaAs, CdTe, GaP, ZnS, ZnSe, ZnTe, CdS, CdSe, and ZnO.
The means for changing an optical gain is controlled in response to an optical gain control signal (SOA current 116) output from the optical gain control means corresponding to the arbitrary waveform current source 111 in
The SOA current waveform data in the swept light source 100 shown in
Accordingly, the swept light source of the present invention can be implemented as a swept light source configured to output a wavelength swept beam having a continuously changing lasing wavelength, the swept light source comprising: a laser resonator comprising a gain medium configured to amplify a beam, means for changing a lasing wavelength (106, 107, and 108), and means for changing an optical gain of the gain medium (102); wavelength control means (113) configured to control the means for changing a lasing wavelength and sweep the lasing wavelength by using a wavelength control signal (115); and optical gain control means (111) configured to control the means for changing an optical gain by using an optical gain control signal (116) based on optical gain control data obtained by combining a plurality of times from the start of a sweep obtained from the wavelength control means with an optical gain in which a coherence length of the wavelength swept beam becomes the longest at each of the times.
The means for changing a lasing wavelength can preferably include a diffraction grating (107), a mirror (108) on which a laser beam from the diffraction grating is incident and the incident laser beam is vertically reflected, and an optical deflector (106) configured to change the angle of incidence of the laser beam on the diffraction grating. In addition, the means for changing an optical gain of the gain medium can be a semiconductor optical amplifier (SOA), the optical gain control signal can be an SOA current for driving the SOA, and the optical gain control data can be SOA current waveform data indicating an SOA current selected for each of the times.
As described above, the present invention is applicable to a swept light source configured to operate based on optical gain control data as long as the laser resonator 101 comprises means for changing a wavelength and means for changing an optical gain and there are wavelength control means and optical gain control means for controlling them. The same also applies to the examples of the present invention described below. The above example has been described based on the premise that there is already SOA current waveform data for operating the swept light source 100 of the present invention. In the next example, an apparatus and procedure for generating the SOA current waveform data will be described.
Time-coherence length correspondence data acquisition means 312 included in the current-time-coherence length correspondence data acquisition means 313 shows a step of acquiring a coherence length of a laser output beam particularly as an explicit functional block. In the swept light source of the present invention, it is necessary to acquire a coherence length of a laser output beam in order to determine an SOA current value at each wavelength when dynamically changing an SOA current. In the swept light source of the present invention, a unique and novel aspect is included also in a method and configuration for calculating the coherence length as a precondition to SOA current waveform data acquisition, which will be described as Example 3. Accordingly, the configuration of the time-coherence length correspondence data acquisition means 312 and the operation/procedure of calculating the coherence length will be described in detail in Example 3.
This example provides an overview of the procedure of obtaining the SOA current waveform data in the apparatus for generating the SOA current waveform data shown in
In Example 1, when the swept light source of the present invention is actually used as a light source, the arbitrary waveform current source 111 operates as one generating the SOA current dynamically changing with time in synchronization with a wavelength sweep based on already-acquired SOA current waveform data. In this example, the arbitrary waveform current source 111 operates as a current source that causes a current of a constant value to flow through the SOA 102 without any temporal change. It is hereinafter referred to as a current source 111 for simplification. The SOA current value in the current source 111 can be externally set based on SOA current setting value data.
The current-time-coherence length correspondence data acquisition means 313 inputs SOA current value data to the current source 111, sets a predetermined SOA current value to the SOA 102, and acquires time-coherence length correspondence data for each set SOA current value. That is, a data set of “time” and “coherence length” (time and coherence length) is obtained for each set SOA current value. The term “time” used here indicates a time from the start of a sweep in one wavelength sweep period. A coherence length is acquired under a condition of a set particular SOA current value and particular time by the time-coherence length correspondence data acquisition means 312. In short, the time-coherence length correspondence data acquisition means 312 shows a functional block that measures the coherence length under a particular condition. The time-coherence length correspondence data acquisition means 312 outputs time-coherence length correspondence data 316, that is, a data set (time and coherence length) 316. The current-time-coherence length correspondence data acquisition means 313 acquires the data set (time and coherence length) 316 for each set SOA current value and finally calculates a data set (SOA current value, time, and coherence length) including three types of data, “SOA current value,” “time,” and “coherence length.” The details of the data set will be easily understood from a flowchart of
As described above, the time-coherence length correspondence data acquisition means 312 obtains a wavelength swept beam from the laser resonator 101 in synchronization with the synchronizing signal 318 from the optical deflector driver 113 and measures a coherence length for each “time” with respect to the wavelength swept beam from the laser resonator 101. At this time, the time-coherence length correspondence data acquisition means 312 generates different coherent beams on different interference conditions from the wavelength swept beam while the SOA 102 is set at one SOA current value, and calculates a coherence length for each “time” from a plurality of items of coherent beam data corresponding to one SOA set current value. It should be noted that the SOA current waveform data for the wavelength swept light source of the present invention is obtained from a “wavelength swept beam” subjected to a wavelength sweep operation. That is, the coherence length is not calculated by setting the wavelength swept light source at a particular lasing wavelength and statically operating it, but the coherence length corresponding to each “time” of the wavelength sweep period is calculated based on the wavelength swept beam obtained under a condition of an actual wavelength sweep operation. Each “time” in the wavelength sweep period corresponds to each wavelength in the sweep wavelength range one by one as shown in the upper graph in
The maximum coherence length corresponding current selection means 314 acquires current-time-coherence length correspondence data 316 obtained by the current-time-coherence length correspondence data acquisition means 313 and selects/determines an SOA current value at which the coherence length becomes maximum for each “time” in the wavelength sweep period. In other words, the maximum coherence length corresponding current selection means 314 selects an SOA current value with the maximum coherence length from data sets having the same “time” among data sets (SOA current value, time, and coherence length) including three types, namely “SOA current value,” “time,” and “coherence length.” Finally, data on an SOA current value selected for each “time” (time-maximum coherence length current correspondence data), that is, current waveform data 317 is output. Next, the procedure of obtaining the SOA current waveform data will be described in detail with reference to a flowchart.
Procedure 400 of generating SOA current waveform data first sets various parameters in step 401. An initial value, final value, and current varying step of SOA current value data to be set to the current source 111 are set from the control unit 315 to the current-time-coherence length correspondence data acquisition means 313. Although what is set to the current source 111 is not a current but setting data corresponding to a set SOA current value, it is referred to as an SOA current value or the like for simplification in the description below.
Procedure 400 sets an SOA current initial value in step 402. The current-time-coherence length correspondence data acquisition means 313 sets an initial value of an SOA current value to the current source 111. The SOA current value is varied from the initial value by the current varying step and a coherence length at each time in one wavelength sweep is measured for each SOA current value in step 403 to be described later. The SOA current value may be first set to a small value and then gradually increased, or vice versa.
Procedure 400 acquires time-coherence length correspondence data in step 403. The procedure of calculating the coherence length will be described later in detail as Example 3, and accordingly a summary of the procedure is provided here. The time-coherence length correspondence data acquisition means 312 acquires a wavelength swept beam corresponding to one wavelength sweep from the laser resonator 101 in synchronization with the synchronizing signal 318 from the optical deflector driver 113. After the acquisition of the wavelength swept beam corresponding to one wavelength sweep, an interference condition is changed by varying a difference between optical path lengths of two arms (a sample arm and a reference arm) of the interferometer to be described later in Example 3 and a plurality of interference signals corresponding to different interference conditions are acquired. After each of the acquired coherent beams is rescaled, PSF for each optical path length difference is calculated for each time during the wavelength sweep. After that, a calculation is performed to obtain a differential value between two optical path length differences 2zPSFHALF1 and 2zPSFHALF2 having such a PSF peak value that the PSF peak value (strength) is ½ of the maximum value PSFMAX for each time. The absolute value of the differential value 2|zPSFHALF1−ZPSFHALF2| is defined as a coherence length Lc. Here, the positive and negative signs of the optical path length differences 2zPSFHALF1 and 2zPSFHALF2 are opposite to each other. The positive and negative of the optical path length differences and a mechanism of measurement will be described below.
On the assumption that a difference between an optical path length of the reference arm and an optical path length of the sample arm is an optical path length difference, the optical path length difference becomes negative when the optical path length of the sample arm is short and becomes positive in the opposite case. PSF is calculated for each optical path length difference. As the optical path length difference becomes closer to zero, interference becomes stronger and the PSF peak value increases. As the optical path length difference becomes longer, interference becomes weaker and the PSF peak value decreases. In the above measurement, both of the positive and negative optical path length differences are measured. Accordingly, the PSF peak value becomes maximum when the optical path length difference is close to zero and decreases as the optical path length difference becomes distant from zero. Since the coherence length is defined as a coherent distance until the interference strength is reduced by half from the maximum value, it is calculated as such an optical path length difference that the PSF peak value is reduced by half from its maximum value as described above.
If the PSF peak value decreases symmetrically with respect to the optical path length difference 0 in both the positive and negative optical path length differences, measurement may be taken only for either of the positive and negative optical path length differences. In this case, a PSF peak value (strength) when the optical path length difference 2z is close to zero may be defined as PSFMAX, a maximum value of PSF, an optical path length difference 2zPSFHALF when the PSF peak value is PSFMAX/2 may be doubled, and the length 4zPSFHALF may be defined as a coherence length at each time.
The PSF peak value when the optical path length difference 2z is close to zero is defined as PSFMAX, the maximum value of PSF, because it is impossible to calculate PSF when the optical path length difference 2z is zero. PSF is calculated from the interference signal. A result of the Fourier transform of an envelope of the interference signal after rescaling is substantially in the form of PSF. When the optical path length difference 2z is zero, a frequency corresponding to a carrier wave of the interference signal is zero and it is difficult to stably output the envelope of the interference signal. In coherence length measurement, PSF is generally acquired when the optical path length difference 2z is close to zero instead of acquiring PSF when the optical path length difference 2z is zero. Finally, correspondence data between each time and coherence length during a wavelength sweep (time-coherence length correspondence data 316) is output.
Procedure 400 acquires current-time-coherence length correspondence data in step 404. The current-time-coherence length correspondence data acquisition means 313 associates the time-coherence length correspondence data 316 output in step 402 with the SOA current value presently set to the current source 111. If there is already current-time-coherence length correspondence data, time-coherence length correspondence data newly associated with the SOA current in this step is added to the current-time-coherence length correspondence data. If there is no current-time-coherence length correspondence data, time-coherence length correspondence data newly associated with the SOA current in this step is used as current-time-coherence length correspondence data. As described above, the current-time-coherence length correspondence data is data sets (SOA current value, time, and coherence length) each including three types of data, “SOA current value,” “time,” and “coherence length.”
Procedure 400 determines whether the SOA current has the final value in step 405. The current-time-coherence length correspondence data acquisition means 313 determines whether the SOA current value presently set to the current source 111 is the final value. The procedure advances to step 407 if Yes and to step 406 if No.
Procedure 400 changes the SOA current value in step 406. The current-time-coherence length correspondence data acquisition means 313 changes the SOA current value set to the current source 111 from the present SOA current value to a new SOA current value according to the current varying step. Then, the procedure returns to step 403 and acquires the time-coherence length correspondence data 316 at the new SOA current value. That is, the data set (time and coherence length) 316 is obtained.
Procedure 400 acquires time-maximum coherence length correspondence current data (current waveform data) in step 407. The maximum coherence length corresponding current selection means 314 selects an SOA current value at which the coherence length becomes maximum for each time of the wavelength sweep for the current-time-coherence length correspondence data acquired in step 404. Then, data on the SOA current value (time-maximum coherence length current correspondence data 317), that is, SOA current waveform data is output at each time.
The coherence length is shown in each cell of Table 5 for each combination of the SOA current value (mA) and wavenumber k in Table 5. The coherence length is calculated for one SOA current value for each wavenumber of the center wavenumbers k(0·δt) to k(7·δt). For example, in the data in the uppermost row of the table where the SOA current is 100 mA, eight items of coherence length data 9.32 to 10.65 are calculated from left to right for center wavenumbers k(0·δt) to k(7·δt). These eight items of coherence length data correspond to the time-coherence length correspondence data 316 obtained in step 404 in the flowchart of
The coherence lengths in all the cells in Table 5 can be calculated by sequentially calculating the above eight items of coherence length data for different SOA current values. This corresponds to repeated execution of determining whether the SOA current value is the final value in step 405, changing the SOA current value to a new value in step 406, and sequentially acquiring the coherence lengths for different SOA current values in step 404 in the flowchart of
Further, in relation to step 407 in the flowchart of
Accordingly, the swept light source of the present invention may be implemented as one further comprising a memory associated with the optical gain control means (111) and storing the optical gain control data, wherein the optical gain control data is generated by controlling the means for changing an optical gain at a set optical gain value by using the optical gain control signal of a constant value over the wavelength sweep period, measuring the coherence length at each of the times from the start of a sweep for each of the different optical gain values, generates a data set including the set optical gain value, one of the times, and the corresponding measured coherence length, selecting the set optical gain value providing the maximum coherence length among the measured coherence lengths in one of the times, and generating the optical gain control data from the one of the times and the selected set optical gain value.
The procedure of obtaining the SOA current waveform data has been described with reference to
The current source 111 in
The current-time-coherence length correspondence data acquisition means 313 in
Similarly, the maximum coherence length corresponding current selection means 314 obtains optical gain-time-coherence length correspondence data from the optical gain-time-coherence length correspondence data acquisition means 313 and selects an optical gain with the maximum coherence length for each time in one wavelength sweep. Then, data on the optical gain selected for each time (time-maximum coherence length gain correspondence data), that is, optical gain control data is output.
When a time interval δt of data acquisition of current-time-coherence length correspondence data acquired by the maximum coherence length corresponding current selection means 314 is different from a time interval δt′ of a current controlled by the current source 111, these time intervals are made temporally consistent with each other. For example, although a time interval of time-maximum coherence length current correspondence data generated in the maximum coherence length corresponding current selection means 314 is δt, the time interval can be changed to δt′ by linear or nonlinear interpolation.
For example, if δt>δt′, current-coherence length correspondence data is acquired with temporal roughness and at low frequency. As a result, the current-time-coherence length correspondence data is also acquired with temporal roughness, which has the effect of reducing a processing time. When the highest frequency of the waveform of the time-maximum coherence length current correspondence data is f, all frequency components can be normally acquired without any aliasing distortion or the like if δt≤1/(2f).
Next, the procedure of acquiring the coherence length Lc in the time-coherence length correspondence data acquisition means 312 of
As described first, in the configuration shown in
The time-coherence length correspondence data acquisition means 312 acquires a wavelength swept beam 123 corresponding to one wavelength sweep from the laser resonator 101 in synchronization with the synchronizing signal 318 from the optical deflector driver 113. After the acquisition of the wavelength swept beam corresponding to one wavelength sweep, an interference condition is changed by varying a difference between optical path lengths of two arms (a sample arm and a reference arm) of an interferometer 501 and a plurality of corresponding interference signals are acquired under different interference conditions. After rescaling of each of the acquired coherent beams 521, time-optical path length difference-PSF correspondence data acquisition means 515 calculates PSF for each optical path length difference for each time during the wavelength sweep. After that, a calculation is performed to obtain a differential value between two optical path length differences 2zPSFHALF1=2qPSFHALF1δz and 2zPSFHALF2=2qPSFHALF2δz having such a PSF peak value that the PSF peak value (strength) is ½ of the maximum value PSFMAX for each time. The absolute value of the differential value 2|zPSFHALF1−zPSFHALF2|=2δz|qPSFHALF1−qPSFHALF2| is defined as a coherence length Lc. Here, the positive and negative signs of the optical path length differences 2zPSFHALF1 and 2zPSFHALF2 are opposite to each other. The positive and negative of the optical path length differences and a mechanism of measurement will be described below.
On the assumption that a difference between an optical path length of the reference arm and an optical path length of the sample arm is an optical path length difference, the optical path length difference becomes negative when the optical path length of the sample arm is short and becomes positive in the opposite case. PSF is calculated for each optical path length difference. As the optical path length difference becomes closer to zero, interference becomes stronger and the PSF peak value increases. As the optical path length difference becomes longer, interference becomes weaker and the PSF peak value decreases. In the above measurement, both of the positive and negative optical path length differences are measured. Accordingly, the PSF peak value becomes maximum when the optical path length difference is close to zero and decreases as the optical path length difference becomes distant from zero. Since the coherence length is defined as a coherent distance until the interference strength is reduced by half from the maximum value, it is calculated as such an optical path length difference that the PSF peak value is reduced by half from its maximum value as described above.
If the PSF peak value decreases symmetrically with respect to the optical path length difference 2z=0 in both the positive and negative optical path length differences, measurement may be taken only for either of the positive and negative optical path length differences. In this case, a PSF peak value (strength) when the optical path length difference 2z is close to zero may be defined as PSFMAX, a maximum value of PSF, an optical path length difference 2zPSFHALF having a PSF peak value of PSFMAX/2 may be doubled, and the length 4zPSFHALF may be defined as a coherence length Lc at each time.
The PSF peak value when the optical path length difference 2z is close to zero is defined as PSFMAX, the maximum value of PSF, because it is impossible to calculate PSF when the optical path length difference 2z is zero. PSF is calculated from the interference signal. A result of the Fourier transform of an envelope of the interference signal after rescaling is substantially in the form of PSF. When the optical path length difference 2z is zero, a frequency corresponding to a carrier wave of the interference signal is zero and it is difficult to stably output the envelope of the interference signal. In coherence length measurement, PSF is generally acquired when the optical path length difference 2z is close to zero instead of acquiring PSF when the optical path length difference 2z is zero. It is also possible to perform function f(2z) fitting of PSF in a position where the optical path length difference 2z is zero by using a plurality of PSF peak points (2zq, PSFpeak_q, where q is an integer for distinguishing the PSF peak points) of different optical path length differences instead of using PSFpeak_MAX, the maximum value of PSF, and define a value f(0) of the function at which the optical path length difference 2z is zero as PSFpeak_MAX, the maximum value of PSF. As the function f(x), for example, it is preferable to use a Gaussian function, a Lorentzian function, a polynomial of a function using x as an argument and the like.
Finally, correspondence data between each time and coherence length Lc during a wavelength sweep (time-coherence length correspondence data 316) is output.
Elements in the time-coherence length correspondence data acquisition means 312 in
The wavelength swept beam 123 from the laser resonator 101 is supplied to the interferometer 501. The interferometer 501 makes the wavelength swept beam 123 outgoing from the laser resonator 101 to a space by using a fiber collimator (FC) 502. The outgoing beam is split by a beam splitter (BS) 503 into two paths on the reference arm side and the sample arm side. The reference arm and the sample arm are equipped with a mirror 504 and a mirror 505, respectively. The beams incident on the respective arms are reflected, returned to the BS 503, and combined in the BS 503, and interferes. The coherent beam is incident on the photodetector (PD) 507 and is output as an electric signal (interference signal) 521 from the PD 507 to the outside of the interferometer 501.
The mirror 505 on the sample arm side is disposed on mirror position movable means 506 so as to externally control a distance between the BS 503 and the sample arm side mirror 505. That is, a difference z between the reference arm length and the sample arm length can be controlled from the outside of the interferometer 501.
After receiving a mirror movement completion signal 532 from the time-optical path length difference-PSF correspondence data acquisition means 515 to be described later, the interference signal acquisition means 511 acquires and outputs an interference signal 123 corresponding to one wavelength sweep in synchronization with the synchronizing signal obtained from the optical deflector driver 113. The mirror movement completion signal 532 is a signal indicating that the mirror position movable means 506 inside the interferometer 501 finishes one movement of the sample arm side mirror for changing the optical path length difference.
Rescaling means 512 rescales the interference signal 522 obtained from the interference signal acquisition means by using time-wavenumber correspondence data and outputs the rescaled interference signal 523, that is, i(k) (k is a wavenumber). Rescaling is described in detail in NPL 1. Operation is performed so as to resample the interference signal 522 and expand and contract the time axis. This converts the interference signal 522 into a signal having a temporally constant frequency.
Processing time generation means 518 generates time information 531 for performing processing in subsequent processes from a time T required for one wavelength sweep and a time step δt. More specifically, the processing time generation means 518 sequentially generates 0, δt, 2δt, (n−1) δt, where n=T/δt.
Extraction position calculation means 517 calculates a wavenumber corresponding to a time 531 obtained from the processing time generation means 518, namely pδt (p is from 0 to n−1), that is, an extraction position 536 from the time-wavenumber correspondence data, and outputs the extraction position 536. More specifically, the extraction position calculation means 517 outputs an extraction position (wavenumber) k(pδt).
Processing range extraction means 513 extracts the rescaled interference signal 523 obtained from a rescaling means 512, that is, i(k), in a range of ±Δk/2 having its center at k(pδt) by using an extraction width Δk and the extraction position k(pδt) obtained from the extraction position calculation means 518, and outputs the interference signal 523. At this time, for example, a window function having a center k(pδt) and a half width ±Δk/2 (such as a rectangular window, Gaussian window, Hanning window, Hamming window, or Blackman window) may be used for extraction.
Window function/frequency conversion means 514 applies a window function to the interference signal 524 obtained from the processing range extraction means 513, calculates a power spectrum (or a frequency distribution excluding phase, that is, the square root of a power spectrum) by using the Fourier transform, and outputs only a positive side. The output signal 525 is referred to as a point spread function (PSF).
It is not necessary to use the window function when a window in which the strength decreases with distance from the center such as a Gaussian window, Hanning window, Hamming window, or Blackman window is used as the processing range extraction means 513. Whether to use the window and the specification and type of window may be determined according to discontinuous points of the interference signal.
The time-optical path length difference-PSF correspondence data acquisition means 515 obtains PSF data from the window function/frequency conversion means 514 and stores the PSF data in association with a time pδt (p is from 0 to n−1, where n=T/δt) output from the processing time generation means 518 and a double of the arm length difference z of the interferometer 150 (2z=2qδz; optical path length difference). An optical path length difference between the reference arm and the sample arm from the FC 502 to the PD 507 is 2z.
Following that, the mirror position signal 533 is transmitted to the mirror position movable means 506 incorporated into the interferometer 501 to change the position of the sample arm side mirror 505. After the completion of the position movement, the mirror movement completion signal 532 is output to the interference signal acquisition means 511. The position of the sample arm side mirror changed by the mirror position signal 533 is determined by using a mirror position range ±Δz (Δz is a maximum distance from z=0) and a position step δz preset to the time-optical path length difference-PSF correspondence data acquisition means 515. For example, z=qδz, where m=Δz/δz and q is an integer from −m to m.
As the way the time-optical path length difference-PSF correspondence data acquisition means 515 detects whether the position movement of the sample arm side mirror 505 has been completed, it may be determined that the movement has been completed upon the acquisition of a movement completion signal 532 from the mirror position movable means 506. Alternatively, the position of the sample arm side mirror of the mirror position movable means 506 may be constantly monitored so that it is determined that the movement has been completed when the position corresponds to a mirror position signal or falls within a preset margin of error. Alternatively, it may be determined that the movement of the mirror has been completed after a lapse of time sufficient for the movement of the mirror.
The time-optical path length difference-PSF correspondence data acquisition means 515 outputs time-optical path length difference-PSF correspondence data 534 when p at the time pδt obtained from the processing time generation means 518 reaches n−1. In the output time-optical path length difference-PSF correspondence data 534, PSF data in a range of time from 0 to (n−1)δt and in a range of optical path length difference from −mδz to mδz is accumulated.
The time-coherence length correspondence data generation means 516 calculates a coherence length Lc of each time pδt (p is from 0 to n−1) from the time-optical path length difference-PSF correspondence data 534 obtained from the time-optical path length difference-PSF correspondence data acquisition means 515, associates pδt with Lc, and stores them. After all items of data corresponding to one wavelength sweep from time 0 to time (n−1)δt are acquired and stored as single time-coherence length correspondence data, the time-coherence length correspondence data 316 is output.
As an example of the way to acquire the coherence length Lc, a PSF peak value corresponding to each optical path length difference is calculated for each time pδt and a calculation is first performed to obtain a differential value between two optical path length differences 2zPSFHALF1=2qPSFHALF1δz and 2zPSFHALF2=2qPSFHALF2δz having a PSF peak value equal to ½ of PSFMAX, the maximum value of the PSF peak value. The absolute value of the differential value 2|zPSFHALF1−zPSFHALF2|=2δz|qPSFHALF1−qPSFHALF2| is defined as a coherence length Lc. If the PSF peak value is not just equal to PSFMAX/2 when 2z=2qδz, it is also possible to perform linear or nonlinear interpolation by using several PSF peak positions and peak values from both sides of a position of 2z where the PSF peak value is equal to PSFMAX/2 and calculate the position of 2z where the PSF peak value is equal to PSFMAX/2. Here, the positive and negative signs of the optical path length differences 2zPSFHALF1 and 2zPSFHALF2 are opposite to each other. The positive and negative of the optical path length differences and a mechanism of measurement of the coherence length Lc will be further described below.
On the assumption that a difference between an optical path length of the reference arm and an optical path length of the sample arm is an optical path length difference, the optical path length difference becomes negative when the optical path length of the sample arm is short and becomes positive in the opposite case. PSF is calculated for each optical path length difference. As the optical path length difference becomes closer to zero, interference becomes stronger and the PSF peak value increases. As the optical path length difference becomes longer, interference becomes weaker and the PSF peak value decreases. In the above measurement, both of the positive and negative optical path length differences are measured. Accordingly, the PSF peak value becomes maximum when the optical path length difference is close to zero and decreases as the optical path length difference becomes distant from zero. Since the coherence length is defined as a coherent distance until the interference strength is reduced by half from the maximum value, it is calculated as such an optical path length difference that the PSF peak value is reduced by half from its maximum value as described above.
If the PSF peak value decreases symmetrically with respect to the optical path length difference 2z=0 in both the positive and negative optical path length differences, measurement may be taken only for either of the positive and negative optical path length differences. In this case, a PSF peak value (strength) when the optical path length difference 2z is close to zero may be defined as PSFMAX, a maximum value of PSF, an optical path length difference 2zPSFHALF having a PSF peak value of PSFMAX/2 may be doubled, and the length 4zPSFHALF may be defined as a coherence length Lc. If the PSF peak value is not just equal to PSFMAX/2 when 2z=2qδz, it is also possible to perform linear or nonlinear interpolation by using several PSF peak positions and peak values from both sides of a position of 2z where the PSF peak value is equal to PSFMAX/2 and calculate the position of 2z where the PSF peak value is equal to PSFMAX/2. At this time, the coherence length Lc is 4z, which is a double of 2z.
As another way to measure the coherence length Lc, it is also possible to perform Gaussian fitting of PSF peak positions and peak values of different optical path length differences and define the full width at half maximum of the Gaussian function as the coherence length Lc. Besides the Gaussian function, a Lorentzian function, a polynomial of a function and the like are preferable.
If the PSF peak value decreases symmetrically with respect to the optical path length difference 2z=0 in both the positive and negative optical path length differences, measurement may be taken only for either of the positive and negative optical path length differences. In this case, a PSF peak value (strength) when the optical path length difference 2z is close to zero may be defined as PSFMAX, a maximum value of PSF, an optical path length difference 2zPSFHALF having a PSF peak value of PSFMAX/2 may be doubled, and the length 4zPSFHALF may be defined as a coherence length Lc at each time. The PSF peak value when the optical path length difference 2z is close to zero is defined as PSFpeak_MAX, the maximum value of PSF, because it is impossible to calculate PSF when the optical path length difference 2z is zero.
PSF is calculated from the interference signal. A result of the Fourier transform of an envelope of the interference signal after rescaling is substantially in the form of PSF. When the optical path length difference 2z is zero, a frequency corresponding to a carrier wave of the interference signal is zero and it is difficult to stably output the envelope of the interference signal. For this reason, in coherence length measurement, PSF is generally acquired when the optical path length difference is close to zero instead of acquiring PSF when the optical path length difference 2z is zero. It is also possible to perform function f(2z) fitting of PSF in a position where the optical path length difference 2z is zero by using a plurality of PSF peak points (2zq, PSFpeak_q, where q is an integer for distinguishing the PSF peak points) of different optical path length differences instead of using PSFpeak_MAX, the maximum value of PSF, and define a value f(0) of the function at which the optical path length difference 2z is zero as PSFpeak_MAX, the maximum value of PSF. As the function f(x), for example, it is preferable to use a Gaussian function, a Lorentzian function, a polynomial of a function using x as an argument and the like.
Next, the procedure of calculating the coherence length Lc by using the time-coherence length correspondence data acquisition means 312 will be described with reference to a flowchart of
Procedure 600 of calculating the coherence length Lc is started by setting various parameters in step 601. In step 601, the parameters are set as follows.
Procedure 600 sets an initial optical path length difference in step 602. The time-optical path length difference-PSF correspondence data acquisition means 515 outputs an initial mirror position signal 533 to the mirror movable means 506. The mirror position movable means 506 moves the sample arm side mirror 505 to a position (initial position) according to the mirror position signal 533.
Here, the mirror position signal 533 is a signal of some kind representing a position z of the mirror. The mirror position signal can use an electric signal of some kind. z may be indicated as a voltage value, and positional information about z may be transmitted by a command in an RS-232C or USB interface. That is, an electric signal of some kind corresponding to z=qδz is referred to as a mirror position signal. As described above, the mirror position signal 533 indicates z=qδz, where m=Δz/δz and q is an integer from −m to m. The sample arm is moved to a position of −mδz in the case of ascending order and moved to a position of mδz in the case of descending order. After moving the mirror, the time-optical path length difference-PSF correspondence data acquisition means 515 outputs the mirror movement completion signal 532 to the interference signal acquisition means 511.
Procedure 600 acquires the interference signal in step 603. Upon the receipt of the mirror movement completion signal 532 from the time-optical path length difference-PSF correspondence data acquisition means 515, the interference signal acquisition means 511 acquires an interference signal 521 corresponding to one wavelength sweep from the interferometer 501 in synchronization with the synchronizing signal 318 obtained from the optical deflector driver 113 and outputs the acquired interference signal 521.
Procedure 600 carries out rescaling in step 604. The rescaling means 512 rescales the interference signal 521 corresponding to one wavelength sweep obtained from the interference signal acquisition means 511 by using time-wavenumber correspondence data.
Procedure 600 generates an initial processing time in step 605. The processing time generation means 518 generates an initial time 531 from a time T required for one wavelength sweep and a time step δt and outputs the initial time 531. More specifically, the processing time generation means 518 outputs zero as an initial time when time information is generated in ascending order such as 0, δt, 2δt, (n−1)δt (where n=T/δt) as time information.
Procedure 600 calculates an extraction position in step 606. The extraction position calculation means 517 calculates a wavenumber 536 corresponding to the time pδt (p is from 0 to n−1) obtained from the processing time generation means 518, that is, k(pδt), based on the time-wavenumber correspondence data, and outputs the wavenumber 536. Here, it should be noted that the extraction position indicates a time for extracting the interference signal in the subsequent steps. It should be remembered that in a wavelength sweep period of the swept light source, a time from the start of a wavelength sweep corresponds to a wavelength or wavenumber of a laser output beam one by one.
Procedure 600 extracts a processing range in step 607. The processing range extraction means 513 extracts the rescaled interference signal 523 obtained from the rescaling means 512, that is, i(k), in a range of ±Δk/2 having its center at k(pδt) by using an extraction width Δk and the extraction position 536 obtained from the extraction position calculation means 517, that is, k(pδt), and outputs the extracted interference signal 524. At this time, for example, a window function having a center k(pδt) and a half width ±Δk/2 such as a rectangular window, Gaussian window, Hanning window, Hamming window, or Blackman window may be used.
Procedure 600 applies a window function to the extracted interference beam and carries out frequency conversion in step 608. The window function/frequency conversion means 514 applies a window function to the interference signal 524 obtained from the processing range extraction means 513, calculates a power spectrum (or a frequency distribution excluding phase, that is, the square root of a power spectrum) by using the Fourier transform, and outputs only a positive side of frequency. The output signal 525 is a point spread function (PSF). The use of the window function in this step is not essential and may be omitted. In particular, in the case of using a window function other than a rectangular window in step 607, it is not necessary to use the window function in this step.
Procedure 600 acquires and accumulates time-optical path length difference-PSF correspondence data in step 609. The time-optical path length difference-PSF correspondence data acquisition means 515 accumulates PSF data 534 obtained from the window function/frequency conversion means 514 in storage means in association with a time pδt and an optical path length difference 2z=2qδz. As the storage means, a memory associated with the swept light source can be used as described in Example 1.
Procedure 600 performs time update in step 610. The processing time generation means 518 updates a variable p at the time pδt. More specifically, p+1 is set as updated p. That is, (p+1)δt is set as an updated time.
Procedure 600 determines an ending time in step 611. The time-optical path length difference-PSF correspondence data acquisition means 515 determines whether the time pδt updated by the processing time generation means 518 is an ending time. If p is not n−1, the procedure returns to step 606. If p=n−1, the procedure advances to step 612.
Procedure 600 updates the optical path length difference in step 612. The time-optical path length difference-PSF correspondence data acquisition means 515 updates a variable q of z=qδz to update the optical path length difference 2z. The time-optical path length difference-PSF correspondence data acquisition means 515 performs update so that a value obtained by adding one to q is new q in the case of ascending order and performs update so that a value obtained by subtracting one from q is new q in the case of descending order.
Procedure 600 determines whether the optical path length difference update has been finished in step 613. The time-optical path length difference-PSF correspondence data acquisition means 515 determines whether z=qδz is a final value. If z=qδz is not the final value, a mirror position signal 533 indicating z=qδz is transmitted to the mirror position movable means 506. At this time, the mirror position movable means 506 moves the sample arm side mirror 505 to a position according to the mirror position signal 533. After the movement of the mirror, the time-optical path length difference-PSF correspondence data acquisition means 515 outputs the mirror movement completion signal 532 and returns to step 603. If z=qδz is the final value, the accumulated time-optical path length difference-PSF correspondence data is output. It should be noted that the final value of z=qδz is mδz in the case of updating z in ascending order and is −mδz in the case of updating z in descending order, where m=Δz/δz.
Procedure 600 generates time-coherence length correspondence data in step 614. The time-coherence length correspondence data generation means 516 calculates a coherence length Lc at each time pδt (p is from 0 to n−1) from the time-optical path length difference-PSF correspondence data acquired from the time-optical path length difference-PSF correspondence data acquisition means 515, associates pδt with Lc, and stores them. After all items of data corresponding to one wavelength sweep from time 0 to time (n−1)δt are acquired and stored as single time-coherence length correspondence data 316, the time-coherence length correspondence data 316 is output. Upon the completion of step 614, a coherence length corresponding to each wavenumber is calculated for one SOA current. That is, referring again to
Accordingly, in the swept light source of the present invention, the optical gain control data can also be implemented as one including a coherence length (316) calculated from a plurality of interference signals (521) based on the interference signals corresponding to respective different interference conditions generated by the interferometer configured to cause the wavelength swept beam (123) to interfere under different interference conditions, the swept beam (123) being output from the laser resonator set at the set optical gain value and obtained from one wavelength sweep of the wavelength sweep period.
In contrast to this, in the SOA current waveform of
Although
As an example of improvement in the coherence length Lc in the swept light source of the present invention, in a laser resonator using a galvano scanner and capable of varying the SOA current in a range between 0 to 600 mA, a coherence length Lc was able to increase 1.7 times over an entire sweep wavelength range when a wavelength is swept repeatedly at 20 kHz in a wavelength sweep range with a center wavelength of 1310 nm±40 nm and a wavelength sweep period is 25 μs.
As described above in detail, the coherence length of the wavelength swept output beam in the wavelength sweep period can be constantly kept long by the swept light source of the present invention and the method of generating drive data for the swept light source. This can increase the PSF strength in the SS-OCT device and improve the image detection sensitivity of the SS-OCT device.
In the swept light source of the first embodiment described above, the means for changing an optical gain (SOA) is controlled so that the optical gain in the laser resonator has an optical gain value slightly higher than the optical gain in a state of the lasing threshold over the entire wavelength range of a wavelength sweep. However, depending on the configuration and type of laser resonator, the coherence length may become maximum at an optical gain value far higher than that close to the lasing threshold. Even in this case, the coherence length of an output beam can be kept long over the entire sweep wavelength range by calculating the SOA current waveform providing the maximum coherence length like the swept light source of the present embodiment.
In the first embodiment described above, SOA current waveform data on an SOA used for the swept light source is acquired in advance and the swept light source is operated based on the acquired data. Accordingly, the step of acquiring the SOA current waveform data is necessary before the normal operation of the swept light source. In addition, the procedure of calculating coherence lengths for acquiring the SOA current waveform data described in Example 3 requires the interferometer 501 in the processing block diagrams of
The present embodiment shows an example of keeping the coherence length of a wavelength swept output beam in a wavelength sweep period constantly long like the first embodiment by using simpler constituent elements and processing based on output levels of the light source without calculating a coherence length at each swept wavelength.
As described above as to the conventional technique, in an SS-OCT system, an image becomes unclear in the case of observing a layer deeper than ¼ of the coherence length. Accordingly, a swept light source having a long coherence length is required. In other words, in the swept light source, an instantaneous laser linewidth should be kept narrow. The inverse of the instantaneous linewidth is a coherence length.
Here, the operation of the SOA is considered again. The SOA is characterized by its wide range and short carrier life and is excellent also in practicality because it can be driven directly by current injection by using an electric signal without the need for optical pumping. On the other hand, it is known that a laser linewidth increases with an increase in injected current not only in the swept light source but also in a general continuous-wave semiconductor laser.
The swept light source of the present embodiment operates so as to keep a long coherence length over the entire sweep wavelength range by a simpler control method by monitoring the output light level in real time during the normal operation of the swept light source. The injected current into the SOA is controlled based on the output light level in conjunction with the operation of the means (wavelength selection element) for changing a lasing wavelength in the swept light source. In this manner, the instantaneous linewidth of the light source is narrowed and the coherence length is maximized by driving the SOA near the lasing threshold at each wavelength in the sweep wavelength range. In the swept light source of the present embodiment, only repeating a control step in conjunction with a wavelength sweep to be described later several times or more enables the coherence length to immediately approach a state of extending to at least a certain extent and saves the need to measure the coherence length during the wavelength sweep operation, which is the normal operation of the light source. A target value of light output power to be described later can be determined only by a manufacture adjustment stage of the swept light source apparatus without the need for the interferometer, which is necessary for calculating the coherence length in the swept light source of the first embodiment.
One beam output from the optical amplifier (hereinafter referred to as an SOA) 701 is incident on the means for changing a lasing wavelength (hereinafter referred to as a wavelength selection element) 702. The wavelength selection element 702 here includes an optical deflector 703 and a diffraction grating 704. In the first embodiment, the diffraction grating is in Littman configuration. In contrast, in the swept light source of the present embodiment, the diffraction grating 704 is configured under a Littrow condition in which a diffraction beam is diffracted in the opposite direction of an incident beam in the same axis. As the wavelength selection element 702, it is also possible to use an element based on another wavelength selection principle such as a Fabry-Pérot filter.
As the optical deflector 703, it is possible to use a galvano mirror, MEMS mirror, KTN optical deflector or the like. The other output beam from the SOA 701 is partly reflected on the optical feedback element 105 and returned to a resonator configured between the diffraction grating 704 and the optical feedback element 705. The beam that has passed through the optical feedback element 705 is directly concentrated as an output beam by a lens 706, is combined with an optical fiber 707, and becomes an output beam 712 of the light source. The resonator in the swept light source of
The operation of narrowing a linewidth of an output beam in the resonator of the swept light source is described as follows. A beam output from the SOA 701 has a wide optical spectrum. The spectrum filtered in the wavelength selection element 702 and cut by the shape of the filter of the wavelength selection element 702 is incident on the SOA 701 again. At this time, the beam is amplified by the SOA 710 and output from the opposite end face of the SOA 701. Part of the beam is returned to the SOA 701 by the optical feedback element 705. In this manner, the beam loops in the optical resonator multiple times during a certain period. Since a selected wavelength of the wavelength selection element 702 is changed in the swept light source, the certain period is determined by a time when a filter spectrum crosses over a particular wavelength. A beam passes through the wavelength selection element 102 multiple times during the crossing time, with the result that a linewidth of the output beam is narrowed by the cumulative effect of filtering.
As described above with reference to
The light output power from the swept light source depends on the number of amplifications in the SOA 701. On the other hand, as described above, the coherence length depends on the number of passes through the wavelength selection element 702. Thus, each of the light output power and the coherence length is associated with the number of loops of the laser beam in the resonator. On the assumption that the transmission characteristics of the wavelength selection element 702 have a spectrum in the Gaussian shape, the shape function F(λ) is expressed by the following formula.
Here, η represents a maximum transmittance of the wavelength selection element 702, λc represents a center wavelength, and ΔλF represents a full width at half maximum of the spectrum. A state where a net gain Gnet made in a portion other than the wavelength selection element 702 when the beam loops one time in the optical resonator balances with a loss made in the wavelength selection element 702 is a state of a lasing threshold. Since a loss in the optical resonator is determined by the characteristics of each constituent element, the net gain Gnet changes mainly according to a gain of the SOA 701.
A loss L in the wavelength selection element 702 is expressed by the following formula using dB.
L=10 log{F(λ)} Formula (2)
From the condition of the lasing threshold, a relationship of the following formula is calculated between the net gain Gnet and the loss L.
G
net
L=0 Formula (3)
In addition, the following relational formulas are acquired by substituting Formula (1) and Formula (2) into Formula (3).
Here, the left side (λ−λc) of Formula (4) represents a half width of the wavelength range above the lasing threshold when the net gain Gnet is larger than the loss L. Thus, a full width Δλeff of the wavelength region above the lasing threshold is calculated by doubling the half width of Formula (4) and combining Formula (5) as follows.
Δλeff in Formula (6) represents an effective wavelength selection width related to lasing and is also referred to as an effective wavelength selection width. It should be noted that when the spectrum width ΔλF of the wavelength selection element 702 can be regarded as being unchanged irrespective of a wavelength (condition 1), the effective wavelength selection width Δλeff related to lasing is determined by the net gain Gnet, that is, a drive current of the SOA 701. This means that the effective wavelength selection width Δλeff can be kept constant by keeping the net gain Gnet constant at each wavelength. In addition, on the assumption that the selected wavelength of the wavelength selection element 702 changes with a uniform speed in the sweep wavelength range (condition 2), the number of beam loops in the resonator becomes constant at each wavelength if the effective wavelength selection width Δλeff is constant.
If the above two conditions are satisfied, the number of beam loops in the resonator, net gain Gnet, and effective wavelength selection width Δλeff become constant at each wavelength. That is, the light output power and the coherence length of the swept light source become constant irrespective of a wavelength in the swept range. Accordingly, the coherence length can be constant at each wavelength in the sweep wavelength range by controlling the SOA drive current so that the net gain Gnet becomes constant while monitoring the light output power as a supervisory signal beam. For this reason, it is preferable that a control voltage of the optical deflector 703 has such a waveform that the selected wavelength changes with a uniform speed in the sweep wavelength range.
However, in actuality, the wavelength selection element 102 may be driven by a sinusoidal wave as a control signal (wavelength sweep signal). In this case, the sweep speed differs according to a wavelength. The number of beam loops in the resonator differs according to a wavelength even if the effective wavelength selection width Δλeff is constant. Accordingly, the coherence length is not necessarily constant even if the light output power becomes constant. However, a rate of change at the selected wavelength of the wavelength selection element 702 near the center of the sweep wavelength range is approximately constant. Even in the case of driving by a sinusoidal wave, the coherence length is close to constant to some extent near the center of the sweep wavelength range. Although the above discussions are expressed by wavelength space, the advantageous result of the present invention can be similarly produced even if they are replaced with frequency space.
As described above, the net gain Gnet is brought close to the lasing threshold and the light output power is controlled to be constant, thereby further narrowing the effective wavelength selection width Δλeff and maximizing the coherence length near the lasing threshold state as shown in
In order to make the light output power constant, it is necessary to control the SOA drive current value to temporally increase and decrease with time (in synchronization with a corresponding wavelength) in conjunction with a wavelength sweep, which is an actual operation of the light source. The swept light source of the present embodiment can be implemented as a swept light source apparatus including the control circuit.
An output beam from the swept light source 700 is partly extracted by the optical branching element 722. One branched beam is detected as a light source output beam 727 and the other branched beam 728 is detected as a supervisory signal beam by the photodetector 723. In the photodetector 723, the supervisory signal beam is converted into a voltage signal 729 and sampled by the A/D converter 724. The sampled signal is input to the operation unit 725. In the operation unit 725, the input signal is compared with a setting value 730 to be described later, an error signal is generated, and a control waveform 731 according to the error signal is generated. The setting value 730 includes a threshold current for each wavelength, an initial current, a tolerance D, a target light output power value Vt and the like to be described later, and may be read from an unshown memory or storage device. The control waveform 731 is input to the arbitrary waveform generator 726. The arbitrary waveform generator 726 outputs the data stream of the input waveform as a control signal 711 in the form of a voltage waveform in which data points are smoothly interpolated. A current driver 708 in the swept light source 700 has the function of converting the voltage waveform input from the arbitrary waveform generator 726 into an SOA drive current waveform and injects an SOA drive current corresponding to the input voltage waveform into the SOA 701. The driver device 709 configured to drive the means for changing a lasing wavelength (such as an optical deflector) 702 supplies a trigger signal 710 synchronized with a wavelength sweep to the arbitrary waveform generator 726.
In the above description, the operation unit 725 is described as one that has the D/A conversion function and outputs the control waveform 731. However, waveform data may be output from the operation unit 725 and the control waveform 731 may be acquired by a separately provided D/A converter. The configuration from the A/D converter 724 to the arbitrary waveform generator 726 is not limited to that shown in
Here, N items of light output power data V0, V1, . . . , VN−1 that can be acquired by sampling by the A/D converter 724 in one wavelength sweep operation are considered. The N data points correspond to the items of light output power data of wavelengths λ0, λ1, . . . , λN−1 included in the entire sweep wavelength range, respectively. The items of light output power data correspond to voltage signals 729 of the supervisory signal beam in
In
E
n
=V
t
−V
n Formula (7)
The target light output power value Vt can be arbitrarily set as a numerical value used for light output power control. An optimum value of the target light output power value Vt should exist for each light source because it is determined by gain characteristics and loss characteristics of each constituent element of the swept light source 700. In the case of calculating the optimum value of the target light output power value Vt, before an actual operation as the swept light source, Vt is preset to a certain value by using a control flow described below, control is performed to make the light output power constant, and this control is repeatedly performed while changing the value of Vt to search for the optimum value. At this time, the search can be made by measuring the coherence length by, for example, the method described in the first embodiment.
The optimum target light output power value Vt thus determined is considered to be unchanged unless the output characteristics of the swept light source 700 largely change. Accordingly, once the target light output power value Vt is determined before an actual operation as the swept light source, the same value can be used thereafter. Thus, for example, it is sufficient that an optical interferometer or the like is provided as external adjustment equipment for manufacture adjustment of a swept light source apparatus at the time of manufacturing the swept light source so as to measure coherence lengths for determining the Vt value. If the target light output power value Vt is determined in the step of manufacture/adjustment of the apparatus and the Vt value is stored in each apparatus, the swept light source apparatus itself does not need to comprise the optical interferometer unlike the first embodiment.
In theory, it is only necessary to acquire I-P characteristics shown in
Iini,n=αIth,n Formula (8)
In the above formula, a value of the coefficient α is set so that the error signal En has a negative value when an initial current is injected at each wavelength. That is, at all data points of the N items of light output power data V0, V1, . . . , VN−1, the first light output power data is larger than the target light output power value Vt, that is, Vn>Vt when the initial current is injected. Next, a control step ΔIn when the SOA drive current is subjected to feedback control is defined as follows.
A tolerance between a light output power value Vn monitored by the photodetector 723 and the target light output power value Vt is defined as D. When the first feedback control is performed in the optimization process of the SOA drive current, an SOA current value In at the wavelength λn is an initial current value Iini,n defined by Formula (8). When the SOA drive current In is set and an absolute val0ue of the error signal En calculated from the light output power Vn monitored at a certain time point in a wavelength sweep is larger than the determined tolerance D, the optimization process requires such control as to update the value of In for the next wavelength sweep. Since the detected output power is smaller than a target value, that is, Vn>Vt (including zero), when En<0 (including zero), the injected current In should be reduced and the value of In is updated as follows.
In the above formula, M indicates that feedback control is an M-th one. “:=” indicates that the numerical value on the right side is substituted in the variable on the left side. In the above description, the case where En<0 (including zero) is indicated. When En>0 (including zero), the sign of ΔIn/2M in Formula (10) is positive. In a control loop, the current value In is not updated when |En|<D. If the size of the error signal En is equal to or less than the tolerance D at all the wavelengths, the control of the optimization process of the SOA drive current is stopped.
Accordingly, in the swept light source of the present embodiment, the optical gain control means (723 to 726) detects an output light level from the laser resonator, generates such an optical gain control signal that the output light level has a predetermined output light level value based on the detected output light level, and controls the means for changing an optical gain 710 by using the optical gain control signal.
Although the process of controlling the SOA drive current based on the light output power Vn described above is described as a series of control at a certain wavelength λn, in the swept light source apparatus 720 of the present embodiment, the N items of light output power data V0, V1, . . . , VN−1 can be acquired in one wavelength sweep period. Until the condition that |En|<D is satisfied, the SOA drive current value In is updated at a wavelength not satisfying the condition. That is, each time the wavelength sweep operation is repeated, update control of the SOA current value In for acquiring a target light output power value Vt at each wavelength is repeated by the optimization process of the SOA drive current shown in
Among the temporal waveforms of the time-SOA drive current, an optimum SOA drive current waveform 815 of the target value corresponding to the light output level for maximizing the coherence length is shown by a solid line. After the setting of the initial current value Iini,n, if the first feedback control is performed at each wavelength, a new time-SOA drive current waveform 813 (first update) is obtained. That is, the SOA drive current value is updated by Formula (4) at all the wavelengths and the second wavelength sweep is performed while the SOA is driven by the temporal waveform of the time-SOA drive current 813. During the second wavelength sweep, light output power data on the second new wavelengths λ0, λ1, . . . , λN−1 can be acquired from the photodetector 723. After the light output power data on all the wavelengths is newly acquired and the second control is finished, a temporal waveform of the updated time-SOA drive current 814 (second update) is obtained. The updated time-SOA drive current 814 is used for the third wavelength sweep. The time-SOA drive current waveform 814 becomes closer to the time-optimum SOA drive current waveform 815 at each wavelength as compared with the time-SOA drive current waveform 813.
In the example of
In the flowchart of
In the swept light source of the second embodiment, a bisection method is used as a method of generating the time-SOA drive current waveform in the operation unit 725. In the bisection method, the optical gain control means (723 to 726) performs temporally discrete control in conjunction with a wavelength sweep and gradually reduces by half a control amount of changing the optical gain (SOA drive current control step) with respect to a difference between the output light level and the predetermined output light level. The bisection method is a convenient feedback control method suitable for a case where a control target amount (light output level) is relatively stable with respect to an environmental change of the swept light source.
On the other hand, when the light source characteristics are more sensitive to an environmental change such as a temperature change, there is a case where the light output level cannot converge into the target value. In this case, the operation unit 725 can perform independent PID control of each wavelength λ of the sweep wavelength range instead of the bisection method to solve the problem. According to the PID method, the optical gain control means (723 to 726) performs temporally discrete control in conjunction with a wavelength sweep and performs control for changing the optical gain (SOA drive current operation) with respect to a difference between the output light level and the predetermined output light level by using the difference, the integral of the difference, and the derivative of the difference.
As described above with reference to
Since a sweep cyclic frequency in the swept light source may be 200 kHz or more, about 10 to 100 updates of the SOA drive current In require very short time in consideration of an actual use of the swept light source. Accordingly, there is no need to perform feedback control continuously each time a wavelength sweep of the light source is started. For example, steps S906 to S919 in the flow of
Once the SOA drive current optimization process is completed, the series of flow of
In the swept light source of the present embodiment, the means for changing an optical gain (SOA) is controlled so that the optical gain value is slightly higher than an optical gain in a state of the lasing threshold on the premise that the relationship between the SOA current and the coherence length as shown in
The above two embodiments show the configurations and control methods of calculating the SOA current for keeping the coherence length maximum, generating the SOA current waveform data, and operating the swept light source. An important element of the swept light source that affects the coherence length of an output beam is the optical deflector 106 shown in
A KTN (KTa1-xNbxO3) crystal or KLTN crystal (K1-yLiyTa1-xNbxO3) (hereinafter collectively referred to as a KTN crystal) that is an electro-optical crystal is known as a substance having a great electro-optic effect with a refractive index largely changing at low voltage. In addition, the use of Ti and Cr materials as electrodes enables charge injection into the crystal and an internal electric field produced by the injected charge can be used to implement a high-speed and wide-angle optical deflector. Therefore, in the case of using a general optic such as a lens, prism, or mirror for a purpose that requires high-speed operation, the optic may be replaced with a KTN crystal.
In a high-frequency operation, since a distance traveled by electrons injected from the electrodes by the applied voltage is shorter than a distance between the electrodes, an ideal space-charge control state is not realized and the angle of deflection is reduced. To solve the problem, a control method of applying a voltage of a burst waveform to the KTN crystal 1101 before applying the applied voltage for causing deflection to inject electrons into the crystal and trapping the electrons in a trapping level in advance is proposed (for example, see PTL 2).
When the voltage is applied for causing deflection, filling a trap in the KTN crystal 1101 with electrons in advance makes it possible to generate distribution or inclination of an electric field in the KTN crystal 1101 and realize optical deflection. As a method of trapping electrons in a trapping level, a method of superimposing an alternating voltage for optical deflection and scanning on DC bias voltage and applying it to the KTN crystal 1101 is also proposed (for example, see PTL 3).
If the DC bias voltage is continuously applied, electrons are supplied from a negative electrode so as to make up for electrons ejected from the trapping level again by thermal excitation or the like and accordingly a certain amount of electrons are left inside the crystal even after a long time. As a result, a deflector that keeps a constant amount of trapped electrons necessary for a high-frequency operation so that the angle of deflection is not reduced even after a long time can be realized.
A charge density N in the KTN crystal becomes close to uniform spatially when the permittivity of the crystal is high, but is accumulated only near a cathode at the time of voltage application and distributed non-uniformly in the crystal when the permittivity is low. When the charge density N is spatially non-uniform and the distribution depends on the position in the crystal like N (x, y, z), the effect of the electro-optical crystal as a lens differs according to a position of transmission of a beam deflected in the crystal. That is, the lens effect differs according to an instantaneous value of drive voltage for optical deflection.
In contrast, a method of arranging a variable focus lens, the focal length of which varies in synchronization with the voltage for optical deflection, on the outgoing side of the KTN crystal to acquire a deflected beam having constant light condensing characteristics irrespective of the applied voltage to the KTN crystal is proposed. This method requires an additional variable focus lens that dynamically operates in addition to the primary optical deflector, which produces a problem that configurations of a power source for operating the variable focus lens, an electric circuit for synchronizing the optical deflector with the variable focus lens and the like become complicated.
To solve the problem, the optical deflector suitable for the swept light source of the present embodiment operates so as to compensate for a spatial change in lens effect in the optical deflector having the lens effect. The optical deflector suitable for the swept light source of the present embodiment is an optical deflector having an electro-optical crystal with opposing surfaces on which at least two electrodes are formed, the optical deflector comprising: an optical deflector configured to, when a voltage is applied to the at least two electrodes, bend a path of a beam orthogonal to an electric field formed by the voltage by generating a non-uniform refractive index distribution in the electro-optical crystal due to the electro-optic effect; and an optical element arranged on at least one of the incident side and the outgoing side of the electro-optical crystal, the optical element having light condensing characteristics spatially distributed over a range of deflection of the beam. According to the optical deflector of the present embodiment, it is possible to compensate for a change in lens effect in the deflection direction by arranging, at the latter part of the optical deflector, an optical element having spatially distributed light condensing characteristics and optical path length.
Various configurations of the optical deflector suitable for the swept light source of the third embodiment of the present invention will be described below in detail with reference to the drawings.
As the electro-optical crystal, for example, potassium tantalate niobate (KTa1-xNbxO3 [0<x<1]; hereinafter referred to as KTN) or a lithium-doped crystal (K1-yLiyTa1-xNbxO3 [0<x<1, 0<y<1]; hereinafter referred to as KLTN) can be used. A charge is injected into the electro-optical crystal along with an electric field by voltage application. As a result, a spatial charge distribution formed by the injected charge or a trap charge distribution formed by the injected charge being further trapped in the electro-optical crystal occurs in the crystal. This charge distribution forms a non-uniform electric field distribution and a refractive index distribution occurs according to a change in the electric field distribution. This non-uniform electric field distribution expresses a gradient of the refractive index and bents a path of a beam orthogonal to the gradient of the refractive index. At this time, because of a nonlinear refractive index distribution in which the refractive index is high near the center of the electro-optical crystal, an incident beam 1205 that has passed through the electro-optical crystal 1201 is deflected and concentrated at the same time to be an outgoing beam 1206 or an outgoing beam 1208, and is diverged after the focal point.
A trap density N is not always spatially uniform depending on the composition distribution of the electro-optical crystal 1201 used as the optical deflector and the applied voltage for charge filling. In this case, the lens effect of the electro-optical crystal differs according to the position of transmission of a beam deflected in the crystal. That is, a focal position formed by the electro-optical crystal differs according to an instantaneous value of the voltage applied to the electro-optical crystal which is the optical deflector. A deflection operation with a constant focal length can be achieved irrespective of the instantaneous value of the voltage applied to the optical deflector by an optical element 1209 having light condensing characteristics spatially distributed so as to compensate for the variation.
For example, if there is no optical element 1209 having spatially distributed light condensing characteristics, the beam shape of the outgoing beam at the time of deflection is the outgoing beam 1206 or outgoing beam 1208 having a focal position different according to an angle. At this time, the focal position of the outgoing beam 1208 is close to the outgoing surface of the electro-optical crystal 1201 as compared with the outgoing beam 1206. This is converted into an outgoing beam 1210 having the same focal position as the outgoing beam 1206 by using the optical element 1209 arranged at the latter part of the optical deflector 1201 to make such a wavefront change that the outgoing beam is diverged. Alternatively, the outgoing beam 1206 is converted into an outgoing beam 1211 having the same focal position as the outgoing beam 1208 by arranging the optical element 1209 so that the focal position of the outgoing beam 1206 becomes close to the outgoing surface of the electro-optical crystal 1201.
When the light condensing characteristics of the electro-optical crystal 1201 used as the optical deflector are changed by the angle of deflection, the optical characteristics of the optical element 1209 configured to correct the angle of divergence of the outgoing beam are spatially distributed over a range of deflection of the outgoing beam. Such beam correction allows the outgoing beam to have an equal focal position at any angle of deflection. For example, in a laser printer or the like, there arises an advantage that a beam waist can be formed in any position on the surface of a photoreceptor. Further, the outgoing beam can be a parallel beam at any angle of deflection. For example, in the case of application to a spectroscope or the like, spectral characteristics of uniform resolution can be acquired without any degradation in resolution over the entire spectral range.
Next, more specific examples of the optical deflector suitable for the swept light source of the present embodiment described above will be described.
A KTN crystal of a thickness of 1.2 mm is used as the electro-optical crystal 1301 (an interval between electrodes is also 1.2 mm) and a beam having a beam diameter of 1.0 mm is incident on the crystal in linear polarization parallel to an electric field. The electro-optical crystal 1301 has a length of 4 mm. The incident and outgoing end faces of the electro-optical crystal 1301 have both a reflective film and an antireflective film. A distance of deflection of an incident beam by a round-trip optical path is 12 mm.
The control voltage source 1304 superimposes a sinusoidal AC voltage having an amplitude of ±360 V and a frequency of 200 kHz on a DC bias voltage of −240 V and apply it to the electro-optical crystal 1301. The voltage is applied via electrodes 1302 and 1303 formed of metal evaporated on the crystal surface. At this time, an angle of deflection of 110 mrad (about 6.5°) is obtained. Applying the DC bias voltage makes a trap density spatially non-uniform in the crystal. Since the trap density is non-uniform in the crystal, the lens effect of the electro-optical crystal differs according to the position of transmission of a beam diffracted in the crystal. If a prism 1309 does not exist, the focal position of a beam when the outgoing beam is diffracted differs according to the angle of deflection and the shape of the beam is like the outgoing beam 1306 or 1308.
In Example 4, a prism 1309 made from SF 11 glass and forming an angle of 30° between the incident surface and the outgoing surface is arranged at the latter part of the electro-optical crystal 1301. Since the prism 1309 increases the beam diameter of a beam passing therethrough, expansion of the beam after passing through the prism 1309 by diffraction is suppressed as compared with the beam before incidence on the prism 1309. The power of magnifying the beam diameter by the prism 1309 depends on the angle of incidence.
This arrangement further magnifies the outgoing beam 1308 at the angle of deflection (voltage) with the close focal position as compared with the outgoing beam 1306 at the angle of deflection (voltage) with the long focal length, whereby the effect of suppressing expansion by diffraction is enhanced and the outgoing beam 1310 is acquired. On the other hand, the outgoing beam 1306 is made into an outgoing beam 1311 by the prism 1309. This prism 1309 having light condensing characteristics spatially distributed over the range of deflection of the outgoing beam makes it possible to suppress the dependence of the focal position on the angle and acquire a deflection operation having a constant focal position irrespective of an instantaneous value of the AC voltage applied from the control voltage source to the optical deflector.
In Example 5, the light condensing effect dependent on the angle of deflection expressed by the optical deflector is corrected by using a free surface mirror 1409 at the latter part of the electro-optical crystal 1401. The free surface mirror 1409 is a concave mirror having a radius of curvature in the direction of deflection varying in the direction of deflection, and the curvature of the mirror decreases from an electrode 1402 on the ground side of the optical deflector to an electrode 1403 on the control voltage source side. That is, regarding the curvature of the free surface mirror 1409, a radius of curvature of a surface irradiated with an outgoing beam in the case of the angle of deflection (voltage) with the focal position close to the outgoing surface of the electro-optical crystal 1401 among the deflected outgoing beams is small and a radius of curvature of a surface irradiated with an outgoing beam in the case of the angle of deflection (voltage) with the distant focal position is large.
When the focal length of the free surface mirror 1409 is expressed by f(x) as a positional function, a free surface is formed so that a distance from the focal position of a beam incident on the free surface mirror 1409 in each incident position to the free surface mirror 1409 is constantly f(x). Accordingly, a beam reflected on the free surface mirror 1409 is a parallel beam at any angle of deflection. Although the parallel beam is generated in Example 5, the parallel beam is not necessarily important. The essence of Example 5 is keeping a constant angle of diffusion of a beam at each moment within the range of deflection of the outgoing beam.
In Example 5, since a negative DC bias voltage is applied as the applied voltage from the control voltage source 1404 and a charge density in the electro-optical crystal 1401 increases near the electrode 403 on the control voltage source side, the relationship between the change in curvature of the free surface mirror 1409 and the electrodes is as described above. In the case of applying a positive DC bias voltage as the applied voltage, a charge density near the electrode 1402 on the ground side in the electro-optical crystal 1401 increases. Accordingly, the arrangement may be made so that the curvature of the free surface mirror 1409 decreases on the electrode 1402 side.
In Example 6, the light condensing effect dependent on the angle of deflection expressed by the optical deflector is corrected by using an optical transmission element 1509 at the latter part of the electro-optical crystal 1501. The optical element 1509 is made from SF 11 glass. A free surface is formed on an outgoing surface, the curvature of the outgoing surface changes according to the angle of deflection of an outgoing beam, and the light condensing characteristics are spatially distributed over the range of deflection of the outgoing beam.
Although the optical transmission element 1509 is used in Example 6, an optical element having a refractive index distribution may be used inside the element to produce the same light condensing effect. In this case, since the optical element has a shape with a planar surface, it is easy to shape the optical element. This optical element can provide light condensing characteristics with a spatially high resolution as compared with the optical element 1509 providing light condensing characteristics by the shape of the outgoing surface, by adjusting the internal refractive index distribution.
According to Example 5 and Example 6, the angle of diffusion of the beam can be kept constant within the range of deflection of an outgoing beam of the optical deflector by arranging an optical element such as a mirror with a free surface on the outgoing side of the electro-optical crystal. If this optical deflector is used as a spectroscope in combination with a dispersion element such as a diffraction grating, spectral characteristics of uniform resolution can be acquired without any degradation in resolution over the entire spectral range.
Therefore, a spatial phase modulator 1609 is further arranged on the outgoing side of the cylindrical concave lens 1607. The spatial phase modulator 1609 is an optical element configured to modulate a phase in each place independently in the cross section of a transmission beam or reflected beam. A beam deflected in the electro-optical crystal 1601 is incident on the spatial phase modulator 1609. In the spatial phase modulator 1609, a phase modulation amount is continuously controlled so as to largely delay a phase from the center to the periphery. The phase modulation amount is large on a surface irradiated with an outgoing beam in the case of an angle of deflection (voltage) with a focal position close to the outgoing surface of the electro-optical crystal 1601 among the deflected outgoing beams and is small on a surface irradiated with an outgoing beam in the case of an angle of deflection (voltage) with a distant focal position.
In this manner, the spatial phase modulator 1609 does not make a phase change in the outgoing beam 1610 at the angle of deflection made into a parallel beam by the cylindrical concave lens 1607 and provides a phase modulation of such a spatial distribution as to counteract the light condensing effect of the outgoing beam 1606 at the angle of deflection, the light condensing effect of which was not counteracted by the cylindrical concave lens 1607. The angle of diffusion of the beam can be kept constant over the entire range of deflection of the outgoing beam by extending the focal length of each outgoing beam according to the focal length before the spatial phase modulator 1609 is inserted.
As the spatial phase modulator, a holographic optical element or diffraction optical element having a spatial phase modulation distribution may be used.
When a Gaussian beam is incident on an optical element having a lens effect, on the assumption that a position from the optical element of the beam waist on the incident side is s and a position of the beam waist on the outgoing side is s″, the relationship of the following Self s formula is established.
zR is the Rayleigh length and f is the focal length of the optical element.
In Example 4 to Example 7, an incident beam to the optical deflector (electro-optical crystal) is assumed to be a substantially parallel beam. The position of the beam waist on the incident side is distant from the optical deflector. At this time, as shown by arrow A in
Further, not only the beam waist position but also the beam waist diameter changes depending on the angle of deflection and requires correction. According to the Self s formula, the ratio between the beam waist diameters of the incident beam and outgoing beam, that is, magnification m is expressed by the following formula.
A compensation method in the case where the beam waist position is on the outgoing side of the optical deflector will be described. Compensation can be provided in the same method even in the case where the beam waist position is on the incident side of the optical deflector.
For example, a case where the configuration is the same as that of the optical deflector in Example 4 (
When the beam waist position monotonously increases and decreases according to the angle of deflection, the vertex of the prism is arranged on the side of the angle of deflection with a close beam waist position so that the width of the prism is large in the case of the angle of deflection with a distant beam waist position and the width of the prism is small in the case of the angle of deflection with a close beam waist position. Since the magnification of the beam diameter changes according to the incident and outgoing angles with respect to the prism, it is possible to compensate for the beam waist diameter. It is also possible to compensate for the beam waist position and the beam waist diameter at the same time by suitably determining the vertex of the prism and the angle of incidence on the prism.
If the prism is further added to the configuration of Example 4 (
Next, a case where the configuration is the same as that of the optical deflector in Example 5 (
Similarly, in the case of using a spatial phase modulator or a medium having a refractive index distribution, the beam waist position and the beam waist diameter of the outgoing beam can be adjusted and a change according to the angle of deflection can be suppressed by changing the phase modulation amount and the refractive index distribution according to the angle of deflection.
If the free surface mirror is further added to the configuration of Example 5 (
Similarly, in the case where the beam waist position is on the incident side, the optical element having spatially distributed optical characteristics may be arranged on the incident side of the optical deflector and compensation may be provided according to a distance between the beam waist position and the optical deflector.
According to Example 8, variations in either or both of the beam waist position and the beam waist diameter of the outgoing beam can be suppressed within the range of deflection of the outgoing beam of the optical deflector. Further, a compensation optical system can be optimized according to the beam waist position and the beam waist diameter of the incident beam. If the optical deflector is used as a spectroscope in combination with a dispersion element such as a diffraction grating, spectral characteristics of uniform resolution can be acquired without any degradation in resolution over the entire spectral range.
Although the optical deflector using KTN or KLTN has been described in Example 4 to Example 8, the optical deflector may use other crystals having the electro-optic effect. The optical deflectors from Example 4 to Example 8 can be suitably used for each of the swept light sources of the first embodiment and the second embodiment.
As described above in detail with various embodiments and examples, the present invention can provide the swept light source capable of keeping the coherence length long over the entire wide sweep wavelength range, thereby improving the image quality of the SS-OCT device.
The present invention can be generally used for an imaging apparatus, and can be particularly used as a swept light source of an SS-OCT device.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2016-239354 | Dec 2016 | JP | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 16466608 | Jun 2019 | US |
| Child | 17380309 | US |
