A novel method based on distributed-feedback laser diode (DFB-LD) continuous wavelength-scanning for acquiring precise spectra of phase-shift fiber gratings is presented. Compared to the traditional method, the spectral resolution retrieved by this method is only limited by the optical line-width of the light source, which can reach up to the order of femtometer and is much higher than that of high-resolution optical spectrum analyzer (generally on the order of picometer). In addition, a Signal-to-Noise Ratio (SNR) advantage can be provided owing to a much higher spectral density of DFB-LD than amplified spontaneous emission (ASE) source. Precise spectra of three phase-shift fiber grating samples have been obtained at a resolution of 23.5 femtometer.
© 2013 Optical Society of America
The phase-shift fiber grating written by ultraviolet light into the core of a rare earth doped fiber has developed into a critical component for many useful applications in fiber optic communication and sensor systems [1–3]. The phase shift open up a narrow transmission band inside the stop band of fiber Bragg grating (FBG), and advantages of distributed-feedback fiber laser (DFB-FL) based on such phase shift fiber grating include single polarization mode operation, intrinsically narrow emission line-width, remote pumping and compatibility with the transmission medium [4–6]. In recent years, asymmetric structure is also designed and DFB-FL with improved efficiency has been obtained [7, 8]. In addition, some other structures for DFB-FLs based on a chirped FBG where multiple π phase-shifts are introduced are also proposed and stable multi-wavelength fiber lasers are achieved . The favorable attributes of the DFB-FL are its small size, simplicity in design, all-fiber geometry, potentially low cost, the ability to accurately set the wavelength during manufacture . For the fiber gratings, the most distinguishing feature is the flexibility they offer for achieving different spectral characteristics. In any case, when the fiber grating is pumped by a 980 nm or 1480 nm pump source, a new laser beam would be generated at the position of the narrow transmission band which appears inside the stop band of the π phase-shift of the grating for the design wavelength. And also, the output laser characteristics, such as intensity noise , polarization property , mode profile , and line-width  are determined by the corresponding phase-shift grating spectral characteristics.
The phase-shift grating spectral characteristics will change in response to the variation of structure parameters of the grating. For example, if the phase-shift is changed to non-π, the narrow transmission band wavelength would deviate from the Bragg wavelength and its FWHM increases. FWHM of the narrow transmission band would decrease with the increasing refractive-index modulation depth. Similarly, for the other parameters, when changes are made, the transmission spectrum would change correspondingly. Therefore, accurate measurement of transmission spectra of phase-shift grating plays an important role in optimizing the design of the grating structure. Two main approaches are used to acquire the transmission spectra of the phase-shift grating: scanning by a tunable laser or an amplified spontaneous emission (ASE) light source. Traditionally, the two methods are simple and direct. But their scanning mode is discontinuous and they only apply to rough measurements. Take the Agilent 81940A tunable laser for instance, its wavelength step is usually about 0.1 picometer (Hereinafter referred to as pm), and this resolution is not enough for the precise spectra measurement of the fiber grating. For the ASE light source, an optical spectrum analyzer (OSA) with an ultrahigh resolution is needed. However, the measurement resolution of the commercially available OSA can only reach pm level. Recently, there has been a fair amount of work on analysis of the narrow spectrum of phase-shift fiber gratings. An optical vector network analyzer (OVNA) based on optical single-sideband (OSSB) modulation is the most effective approach [14, 15], which can measure the magnitude and phase response of an optical component by sweeping the frequency of one optical sideband. The measurement resolution of OVNA is mainly determined by the line-width of the laser source, and the bandwidth of the OSSB modulator is determined mainly by the spectral characteristics of the optical filter. A high resolution of 78 kHz has been achieved by using a laser source with a line-width of less than 100 kHz . However, the system is very complex and hence difficult to construct. The parameters of the optical components, such as the passband width and the tuning range of the tunable optical bandpass filter, are maintained within strict scope. Not only that, an OVNA is usually very expensive, which largely increases the cost.
In this paper, a novel method based on a distributed feedback laser diode (DFB-LD) continuous wavelength-scanning spectrum (DCWS) for acquiring the precise spectra of phase-shift fiber gratings is presented. Its biggest advantage is that it can get nearly continuous transmission spectrum of the phase-shift fiber grating. We experimentally demonstrate measurements of phase-shift fiber grating spectra with a resolution of 23.5 femtometer. A Signal-to-Noise Ratio (SNR) advantage can also be provided owing to a much higher spectral density of DFB-LD than ASE source. Three π phase-shift fiber gratings with different structure parameters are measured and the spectra of them are accurately acquired. The FWHM of their central narrow transmission bands are also obtained. Plenty of advantages, such as continuity, high SNR, simple structure, low price and flexibility of the DCWS method are demonstrated, which has potential for optimizing the design of the grating structure.
2. Experiments and results
Figure 1 shows the schematic diagram of the DCWS experimental system used to acquire the precise spectra of phase-shift fiber gratings. A single mode DFB-LD with a spectral line-width of 3 MHz and the maximum optical power of 13 dBm is taken as the wavelength scanning source, which can cover the wavelength range of the whole C waveband and can be changed according to the phase-shift grating wavelength. The central wavelength of the DFB-LD is chosen as 1533.07 nm in order to match the gratings we fabricated. The DFB-LD output wavelength is modulated by adjusting the laser temperature and the injection current. The laser temperature can be maximally adjusted 6 °C by using a micro temperature controlling chip (LTC1923, Linear Technology, USA). When the laser temperature changes 1 °C, the DFB-LD wavelength drifts 0.3 nm correspondingly. And therefore, the wavelength adjustment range of the DFB-LD is 1532.17 nm-1533.97 nm. This is also the wavelength range of fiber gratings which could be measured by this DFB-LD. Laser temperature is fixed to a suitable working point at the beginning of the experiment. The wavelength of the reference working point can be corrected by the spectrometer. And then, wavelength scanning of the DFB-LD is implemented by the current tuning. Compared to the thermal tuning, current-tuning can greatly accelerate the speed of the laser wavelength scanning as well as the speed of the spectral measurement. The maximum scanning frequency can reach to 10 kHz, which is thousands of times faster than thermal tuning. A saw-tooth wave analog voltage signal generated by the saw-tooth wave generator is converted into a current signal by the current drive circuit. It should be noted that the analog voltage signal can avoid the impact of the scanning step appearing in the digital signal, which can ensure a continuous and smooth change in current. The drive current converted from the digital voltage signal can lead to jump changes of the light wavelength due to the scanning step, which will affect the measurement resolution. The wavelength continuously drifts about 3 pm when the injection current changes 1mA. The injection current signal will drive the DFB-LD continuous scanning within the certain wavelength range. The wavelengths of several reference points including the starting point and end point within the scanning range are also calibrated by the spectrometer and multiple measurement and average are taken in order to reduce the errors. The optical signal propagated through the π phase-shift grating would be then converted into an electrical signal by the photo-detector (PD) with a responsivity of 0.9 A/W. Finally, the electrical signal would be displayed on the oscilloscope. The displayed electrical response of the oscilloscope at different points in time within a scanning period represents the optical response of the phase-shift grating at different wavelengths. Details of the spectrum could be revealed by changing the scaling factor of time axis and amplitude axis of the oscilloscope. For example, in the latter test, by increasing the scaling factor of time axis of the oscilloscope, there are more than 1000 sample points at least in a wavelength interval of less than 0.7 pm. This indicates that the sampling resolution of the oscilloscope is high enough to ensure that the measurement resolution of the DCWS method is only limited by the optical line-width of the laser source.
Several π phase-shift gratings with different design parameters are fabricated by the phase mask moving method . In the fabrication process, the transmission spectrum is measured in time to verify whether or not the grating is fabricated successfully. Three of them are chosen and measured. The fabrication conditions are set as: standard π phase-shift, grating length L = 0.04 m and without apodization. The descriptions of the others are given in Table 1 (Zπ represents π phase-shift position).
2.1 Measured results and analysis of G1
In order to verify the feasibility of the DCWS method, in our experiments, firstly, a π phase-shift grating used un-doped photosensitive fiber is fabricated. Its setting wavelength is 1532.52 nm. So we set the reference central wavelength of the DFB-LD at about 1532.52 nm by adjusting the temperature controlling circuit. And then a saw-tooth wave analog signal is generated, which period is 682 Hz and amplitude range is 310 mV-880 mV. Through a voltage-to-current conversion, the current signal range is 31 mA-88 mA.
Figure 2(a) is the complete transmission spectrum measured by the DCWS method. After adjusting the time resolution and amplitude resolution of the oscilloscope, the central narrow transmission band as shown in Fig. 3(a) is clearly observed. The curve is smooth and clear, and there are more than 1000 dots at least. FWHM of the central narrow transmission band is precisely acquired as 0.51 pm. This FWHM calculation is performed to an accuracy of two decimal places because the spectral resolution retrieved by this method is limited by the optical line-width (23.5 femtometer in wavelength corresponds to 3 MHz at the 1532.5-nm window) of the light source.
For comparison, the π phase-shift grating is also scanned by an ASE light source (CCFxxM OPEAK China). An ultra-high resolution OSA (AP2040A APEX France) is combined to acquire the transmission spectrum, which spectral resolution of up to 0.16 pm and is 62 times higher than that of the normal spectrometer (10 pm). The transmission spectrum and the central narrow transmission band spectrum are shown in Figs. 2(b) and 3(b). Firstly, the two spectra are compared in Fig. 2, although they show good agreement, some differences can be found. As indicated by the arrows, the proportion of the narrow transmission band height in the whole spectrum in Fig. 2(a) is much larger than which in Fig. 2(b). It can prove that the precision of the transmission spectrum measured by the DCWS method is much higher. The detail of the central narrow transmission band in Fig. 3(b) demonstrates that the spectrum is discontinuous and there are only 15 dots. The curve is so rough that it is not precise enough for estimating the FWHM. The result comparisons can be fully proved that the DCWS method is effective and high-resolution.
2.2 Measured results and analysis of G2
The transmission spectrum of an asymmetric π phase-shift fiber grating is quite different with the symmetric one. The height of the narrow transmission band of the asymmetric π phase-shift fiber grating is much lower. It means that for acquiring the precise transmission spectrum of this type of grating, in addition to the high resolution, high SNR is also needed.
G2 is an asymmetric π phase shift fiber grating which fabricated by placing the π phase shift at the position Zπ = 0.4L. When the grating is pumped by a 1480 nm semiconductor laser source, stable unidirectional output is obtained from the shorter grating end. Figure 4(b) is the transmission spectrum of G2 which scanned by the ASE light source and measured by an ultra-high resolution OSA (AP2040A APEX France) with a resolution of 0.16 pm. The spectrum does not reveal the central narrow transmission band and not appear even after it is zoomed in, shown as the inset. Then we use the DCWS method and measure the transmission spectrum of G2 again. The reference central wavelength of the DFB-LD is set at about 1532.85 nm by adjusting the temperature controlling circuit, and then a saw-tooth wave analog signal is generated to drive the DFB-LD scanning. The current signal range is 4 mA-98 mA. The result is shown in Fig. 4(a). It is also difficult to observe the central narrow transmission band in the full spectrum. But after adjusting the time resolution and amplitude resolution of the oscilloscope, a clear transmission band appears in the middle of the spectrum bottom, shown as the inset in Fig. 4(a). The comparison and analysis find that the SNR of the traditional method is not high enough. That is because the ASE has a wider line-width than the DFB-LD, which results in a lower spectral density. The same power is distributed over a broader range. The narrow transmission band signal is too weak to be drowned out completely by the noise background. However, a SNR advantage can be provided by DCWS owing to a much higher spectral density of DFB-LD than ASE source. And then we continue to adjust the time resolution and amplitude resolution of the oscilloscope, the precision spectrum of the narrow transmission band is obtained, as shown in Fig. 5. Its FWHM is 0.41 pm. It means the asymmetric fabrication of G2 is successful and the grating can completely work well. The results of G2 confirm that a high SNR can be provided by the DCWS method. It has considerable potential to help to optimize the design of the phase-shift fiber grating structure.
2.3 Measured results and analysis of G3
G3 is another grating which fabrication condition is as same as the G1, except the fiber is Er3+ doped fiber, as revealed in Table 1. In our experiments, we find that the laser output becomes unstable when the 1480 nm pump power increases and G3 could not work properly. To analyze the reason, the transmission spectrum of G3 is measured by the DCWS method. The reference central wavelength of the DFB-LD is set at about 1532.96 nm by adjusting the temperature controlling circuit. And then a saw-tooth current signal which amplitude range is 24 mA-91 mA is generated. The measured results are shown in Fig. 6. There are two narrow transmission bands in the spectrum, after adjusting the oscilloscope, the precise spectrum of the two transmission bands is obtained. As shown in the inset, the two transmission bands have different heights and the distance between them is only 5.83 pm. The spectrum can illustrate that G3 is fabricated when the fabrication system is in the event of a failure. When the grating is pumped by the 1480 nm pump power, the two narrow transmission bands are likely to generate lasers, and compete with each other. So, the output laser becomes unstable.
Although the G3 is a failure, the measured results can be used to prove the reliability and efficiency for the DCWS method. Not only the narrow transmission bands can be described particularly, the other parts of the transmission spectrum also can be revealed. Furthermore, the FWHM of the narrow transmission band and the space of the transmission bands can be obtained accurately. In our future work, the multi-phase shift structure will be tried. By using the DCWS method and analyzing the influences of structure parameters on the transmission spectrum and laser output, the new multimode operation with stable, efficient output will be obtained.
In conclusion, we have introduced a novel method for acquiring precise spectra of phase-shift grating. It worked by adjusting the laser temperature and the injection current to modulate the DFB-LD output wavelength continuous scanning. Compared to the traditional method which uses ASE as the light source and collects the transmission spectrum of the phase-shift grating with OSA, its biggest advantage is that it can get nearly continuous transmission spectrum of the phase-shift fiber grating. The spectral resolution retrieved by this method can be attained to the order of femtometer and is much higher than that of high-resolution OSA (generally on the order of pm). A SNR advantage can also be provided owing to a much higher spectral density of DFB-LD than ASE source. Three gratings with different structure parameters are chosen as the test object. With this method, the FWHM of a narrow transmission band in the middle of the transmission spectrum of a standard π phase-shift fiber grating G1 is precisely acquired as 0.51 pm, which could not be accurately obtained by using an ultra-high resolution OSA (resolution is 0.16 pm) combined with an ASE light source scanning. Furthermore, the central narrow transmission band of an asymmetric π phase-shift fiber grating G2 can be clearly distinguished by this novel method and its FWHM is 0.41 pm, which is not observed by the ultra-high resolution OSA for the SNR limit. The example of G3 also confirms the superiority of the DCWS method. The time resolution and amplitude resolution can be adjusted by changing the scaling factor of time axis and amplitude axis of the oscilloscope, and local details of the spectrum could be revealed exactly. The measured results illustrate that the DCWS method has plenty of advantages, such as, continuity, high SNR, flexibility and low cost. The result analysis also confirms that the DCWS method has potential for optimizing the design of the fiber grating structure.
This work was supported by Natural Science Foundation of China (60977058), Independent Innovation Foundation of Shandong University (IIFSDU2010JC002 & 2012JC015), and the key technology projects of Shandong Province (2010GGX10137).
References and links
1. P. St. J. Russell, J. L. Archambault, and L. Reekie, “Fiber gratings,” Phys. World 41–46 (1993).
2. I. Bennion, J. A. R. Williams, L. Zhang, K. Sugden, and N. J. Doran, “UV-written in-fiber Bragg gratings,” Opt. Quantum Electron. 28(2), 93–135 (1996). [CrossRef]
3. T. Erdogan, “Fiber Grating Spectra,” J. Lightwave Technol. 15(8), 1277–1294 (1997). [CrossRef]
4. K. P. Koo and A. D. Kersey, “Bragg grating-based laser sensors systems with interferometric interrogation and wavelength division multiplexing,” J. Lightwave Technol. 13(7), 1243–1249 (1995). [CrossRef]
5. G. A. Cranch, G. M. H. Flochhart, and C. K. Kirkendall, “Distributed Feedback Fiber Laser Strain Sensors,” IEEE Sens. J. 8(7), 1161–1172 (2008).
6. P. Wang, J. Chang, C. Zhu, Y. Zhao, Z. Sun, X. Zhang, and G. Peng, “Theoretical and experimental investigation of the intensity response of DFB-FL to external acoustic excitation,” Opt. Laser Technol. 49, 227–230 (2013). [CrossRef]
7. H. Qi, Z. Song, S. Li, J. Guo, C. Wang, and G. D. Peng, “Apodized distributed feedback fiber laser with asymmetrical outputs for multiplexed sensing applications,” Opt. Express 21(9), 11309–11314 (2013). [CrossRef] [PubMed]
8. K. Yelen, L. M. B. Hickey, and M. N. Zervas, “A New Design Approach for Fiber Dfb Lasers With Improved Efficiency,” IEEE J. Quantum Electron. 40(6), 711–720 (2004). [CrossRef]
9. X. Liu, “A novel dual-wavelength DFB fiber laser based on symmetrical FBG structure,” IEEE Photon. Technol. Lett. 19(9), 632–634 (2007). [CrossRef]
10. G. A. Cranch, M. A. Englund, and C. K. Kirkendall, “Intensity Noise Characteristics of Erbium-Doped Distributed-Feedback Fiber Lasers,” IEEE J. Quantum Electron. 39(12), 1579–1586 (2003). [CrossRef]
11. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and M. A. Nikulin, “Single frequency single polarization DFB fiber laser,” Laser Phys. Lett. 4(6), 428–432 (2007). [CrossRef]
12. S. Foster and A. Tikhomirov, “Experimental and theoretical characterization of the mode profile of single-mode DFB fiber lasers,” IEEE J. Quantum Electron. 41(6), 762–766 (2005). [CrossRef]
13. J. Ni, Y. Zhao, C. Wang, G. Peng, T. Liu, J. Chang, and Z. Sun, “Research on linewidth characteristics and broadening mechanism of distributed feedback fiber laser,” Acta Phys. Sin. 61, 0842051 (2012).
14. M. Xue, S. Pan, C. He, R. Guo, and Y. Zhao, “Wideband optical vector network analyzer based on optical single-sideband modulation and optical frequency comb,” Opt. Lett. 38(22), 4900–4902 (2013). [CrossRef]
15. A. Loayssa, R. Hernández, D. Benito, and S. Galech, “Characterization of stimulated Brillouin scattering spectra by use of optical single-sideband modulation,” Opt. Lett. 29(6), 638–640 (2004). [CrossRef] [PubMed]
16. Z. Tang, S. Pan, and J. Yao, “A high resolution optical vector network analyzer based on a wideband and wavelength-tunable optical single-sideband modulator,” Opt. Express 20(6), 6555–6560 (2012). [CrossRef] [PubMed]