Based on the technology of real-time laser noise measurement which had been developed for the red-light DVD optical storage system before, this study investigates the laser noises corresponding to different feedback configurations by using the optical system of a blue-light HD-DVD optical pickup head (OPH) with the numerical aperture (NA) of 0.65 and the laser wavelength of 405 nm. It combines with a digital servo controller for dynamically measuring the blue laser noise while operated in an OPH. Experimental results show the variations of relative intensity noise (RIN) values corresponding to focusing errors, lens off-axis errors, disk rotation speeds, and laser emitting powers, respectively.
©2009 Optical Society of America
The semiconductor laser with the wavelength of 405 nm had been proposed as the light source of the new generation optical storage system for increasing the disk capacity [1,2]. However, as the recording density of an optical memory increases, the signal-to-noise ratio (SNR) will consequently deteriorate . In order to achieve the large disk capacity and improve the SNR, the opto-electronic specifications of a blue-light optical pickup head (OPH) are tightly determined, while compared to those of the current red-light DVD systems. Thus, the laser noise consideration that rises from the optical feedback of the external cavity of the blue-light OPH is an important issue and worthy of being highly explored [4,5].
The conventional measuring system of the laser feedback noise without the servo close-loop control is a static system and difficult to realize the real laser noise distribution of an optical drive [6–12]. For the reason, we had developed a real-time laser feedback noise measuring system to investigate the genuine laser noise distribution of the red-light DVD system while operating in an optical drive by modifying the optical system of a commercial DVD pickup head . Based on the previous works, this study further advanced this technology to be applied for the blue-light optical storage system. The measuring system was implemented by a blue-light OPH with the numerical aperture (NA) of 0.65. It was integrated with precision mechanical design, optical design, servo control design, and opto-electronic signal inspection for providing the dynamic real-time feedback noise measurement of the 405 nm laser. The variations of relative intensity noise (RIN) corresponding to different feedback and servo configurations, including the focusing errors under various laser emitting powers, lens off-axis errors, disk rotation speeds, and with or without a quarter wave plate (QWP), were discussed in this paper. The experimental results are interesting and could provide the laser noise information for the development and manufacturing of the blue-light OPHs.
2. System design
2.1 Optical system design
We adopted a standard blue-light HD-DVD OPH that was developed by the Industrial Technology Research Institute (ITRI) in Taiwan as the main optics of the measuring system design. It is an infinite-conjugate optical system with the objective NA of 0.65 and the laser wavelength of 405 nm. Figure 1(a) shows the schematic diagram of the OPH and its optical parameters are listed in Table 1 . The laser diode was produced by SANYO corp. in Japan with the model number of DL-4146-301 and the brief specifications tabulated in Table 2 . The beam-splitter reflects the laser incident light and transmits the return beam from the disk to the photodetector. The astigmatic servo mechanism was used as the focusing detection method in this OPH. Its astigmatic aberration could be brought out from the thick beam-splitter plate. A photodetector with quadrant detection areas and sensitive to the 405 nm light was adopted for generating the servo and readout signals. The OPH was modified for the laser feedback noise measurement as shown in Fig. 1(b). In order to monitor the laser output, the fully reflective folding mirror was replaced with a partially reflective mirror (beam-splitter 2) and a converging lens with the NA of 0.1 was set on the forward path for collecting the sampling laser beam into the photodetector 2. This beam is not modulated by the optical disk and faithfully presents the laser output quality. The photodetector 2 is of high-speed PIN structure, which has a built-in internal amplifier of 1 GHz electronic bandwidth. The testing HD-DVD recordable disk was manufactured from the CMC corp. in Taiwan and recorded with random patterns in advance.
2.2 Measuring system design
The complete measuring system was set up as the configuration shown in Fig. 2 . There are two electric outputs from the photodetector 2. One was connected to a digital oscilloscope for measuring the laser DC intensity whereas the other was connected to a spectrum analyzer for detecting the laser AC fluctuation. The noise level could be acquired from the expected frequency domain among the AC fluctuation. A servo controller that was dedicatedly designed to receive the focusing error signal (FES) from the photodetector 1 and then to control the focusing actuator in close-loop. Besides, it also has the function of driving the spindle motor for obtaining the disk rotation in constant linear velocity. A constant current driver with automatic power control (APC) was adopted for lighting the laser diode. A personal computer was used to communicate with and collect data among the controller and equipments.
The sampling noise frequencies of interest are determined by the data rate for the specific drive application being investigated. For the 1X-speed HD-DVD data retrieving, the channel bit rate is 64.8 MHz. It corresponds to the central frequency of 10.8 MHz for the recording frequency of 3T data pits (T is the system clock length.) and is generally associated with a bandwidth of 10 kHz for investigating the laser noise. The noise level is usually expressed as the RIN defined as the following:
2.3 Servo controller design
We designed the servo controller based on the previous study , whereas the servo specifications for the blue-light HD-DVD system were tighter than those for the red-light DVD system. The controller was implemented by using a high speed digital signal processor (DSP) as the core component (model TMS 320F2812 developed by Spectrum Digital Inc.), and combined with other components, such as the preamplifier, analog-to-digital (A/D) converter, digital-to-analog (D/A) converter, actuator power driver, and spindle motor driver, etc. The phase-lead and -lag technology was adopted for the compensator design of the servo controller. It should be optimized to provide the compensation of the system for cross frequency greater than 4.5 kHz, gain margin greater than 72 dB, and phase margin greater than 35 degrees. Figure 3 shows the schematic diagram of the servo controller that could provide the close-loop control of the OPH and the dynamic measuring for noise while retrieving data from the disk. The block diagram of the digitized controller is sketched in Fig. 4 . It could generate a focus-on signal (FOS) to move the objective lens for compensating the focus position error. The reference voltage should be tuned for obtaining the maximum radio-frequency signal (RFS).
3. Experimental results
3.1 System verification
We first examined the proposed noise measuring system by evaluating the capability of the servo controller. The linear range of the FES (so-called S curve) was designed to be 4 μm and implemented to be from 2.035 V to 3.085 V with the center level (best focus position) at 2.5 V as shown in Fig. 5 . It corresponds to the translation of 3.81 μm/V. Figure 6 shows the verification results corresponding to the transient states while the controller was switched into the close-loop status. The frequency generator signal (FGS) was generated by the driving circuit of spindle motor for controlling the disk rotation speed. The frequency of the FGS was six times higher than that of the disk rotation speed. The FOS was sent to the focus coil after the FES was found from an open-loop search for focus. Once the focus was locked by the servo controller, a high-level RFS emerged. In the experiment, only the focusing was applied to the system whereas the tracking stayed open. Therefore, the eye-pattern presented blurring eyes in Fig. 6, but it did not affect the following measurements corresponding to respective configurations.
3.2 RIN corresponding to focus errors under different laser emitting powers
By adjusting the reference voltage of the controller, we can get the defocus configurations for measuring the laser noise distribution. The RIN values were calculated by the acquired data from the oscilloscope for the DC part and from the spectrum analyzer for the AC part according to the equations described in our previous work . Figure 7 shows the RIN data relating to various focus positions of the objective lens under 1X standard HD-DVD rotation speed (central noise frequency of 10.8 MHz). The testing blue laser diode was operated at room temperature (27 °C) with the threshold current of around 26 mA. There are three RIN distributions, noted as curve ‘a’ to ‘c’, measured corresponding to three laser emitting powers in the experiment. They are 4.24 mW operated at 30.5 mA (curve ‘a’), 6.67 mW at 32.1 mA (curve ‘b’), and 9.09 mW at 33.6 mA (curve ‘c’), respectively. The RIN distribution can be effectively suppressed from curve ‘a’ to ‘c’ by increasing the laser output.
The experimental results reveal some interesting phenomena that didn’t be found in our previous red-light study . At the curve ‘a’, the laser beam on the exact focus point is noisier than those on the defocus points. Once the laser was operated at a higher output power, the RIN distribution would be gradually reversed. For the curve ‘a’, it makes sense with our previous result that the RIN presents the maximum value while the objective lens is positioned at the best focus point because the laser diode is fed with the largest amount of feedback light at that point. However, the curve ‘b’ and ‘c’ contradict this deduction. By inspecting the experimental result presented in the Fig. 10 of reference , it had shown that the RIN value of a GaN blue laser is dependent on the laser output and has the tendency of lower RIN toward higher output. Combined with the general knowledge that the RIN is also dependent on the feedback light percentage, we guess that the RIN distribution versus feedback percentage might be shiftable along the feedback percentage axis by adjusting the laser output and, therefore, it results in the reversed distributions from curve ‘a’ to ‘c’. In order to investigate the longitudinal mode variation during operation, we also measured the laser output spectrum corresponding to different feedback configurations. Figure 8 shows the experimental results. Figure 8(a) presents that the central laser wavelength is not responsive to the laser emitting power, whereas Fig. 8(b) to 8(d) reveal that the central laser wavelength is slightly sensitive to the focus position. It comes from the reason of creating external cavities by different focus positions. These data also show that the higher the laser output is, the more apparently the central wavelength shift will be affected by the focusing.
3.3 RIN corresponding to focus errors under different disk rotation speeds
The above experiment of noise measurement was repeated by changing the disk rotation speed for investigating its effect on the laser noise. Figure 9 shows the RIN distributions while the disk rotation speeds was selected to be 0.25X and 0.5X. They present the same trend as that revealed in Fig. 7. Figure 10 summarizes the RIN distributions from Fig. 7 and Fig. 9 for three central noise frequencies of 2.7 MHz, 5.4 MHz, and 10.8 MHz corresponding to disk rotation speeds of 0.25X, 0.5X, and 1.0X, respectively, at the laser output of 6.67 mW. It indicates that the overall distribution of RIN will decrease while the rotation speed increases. This tendency is similar as what had been discussed in the previous red-light laser study . As we know, the data pit length formed on a HD-DVD disk randomly distributes from 3T to 14T. The 3T data is the highest frequency that modulates the external feedback and contributes the noise of this frequency to the corresponding data signals, but it is not the unique feedback noise frequency. Therefore, the feedback noise generally has a wide spectrum spreading from the frequency of 14T to that of 3T and this spectrum is shiftable by changing the disk rotation speed. For the reason of paper length consideration, we only discussed the effect of noise distribution at the 3T frequency.
3.4 RIN corresponding to the lens off-axis errors
In order to separate the effect of focusing (on-axis) offset from that of tracking (off-axis) offset, only the focusing servo was applied to the OPH and the tracking control was kept open in the above measurements. If the tracking servo was simultaneously added to the system, the track following would cause the objective lens continuously moving in the radial direction and the noise level could not be certainly measured. Therefore, we adopted the open-loop control by applying different DC voltages to the tracking coil for shifting the lens and observing the RIN variations corresponding to the lens off-axis errors. The laser output was operated at 6.67 mW and the objective lens was kept on the best focus position under 1.0X rotation speed. Figure 11 shows the experimental data and indicates that the RIN is not sensitive to the lens off-axis offset. It is reasonable that owing to the optical conjugation of the image and object, the returning light from the disk always focuses and illuminates back onto the original emitting point even with small amount of radial lens offset. The percentage of optical feedback is not remarkably changed and therefore does not affect the noise level.
3.5 RIN corresponding to with or without QWP
Generally, the optical feedback is immunized by modulating the laser in high frequency oscillation or adopting a QWP to change the feedback light polarization . In this experiment, we inserted a QWP, which was designed for the 405 nm wavelength, between the objective lens and beam-splitter 2. Figure 12 shows the experimental result under the 1.0X rotation speed. The RIN values are apparently suppressed by around 20 dB/Hz, while compared to those without a QWP. The overall distribution of RIN meets the specification of being lower than −125 dB/Hz for the blue-light system even though the laser stays at the noisier level of being operated at 4.24 mW output.
6. Discussion and conclusion
A real-time dynamic system for measuring the laser noise distribution of a blue-light OPH corresponding to different feedback configurations has been achieved. The experimental results show that the blue laser feedback noise is responsive to the defocus distance, whereas it is not sensitive to the lens off-axis offset. The distribution tendency of the RIN versus defocus distance is dependent on the laser operating power. Also, the central laser wavelength is slightly sensitive to the focus position, especially apparent at higher laser output. Besides, the experimental results also present that the noise level will decrease under either increasing the disk rotation speed or boosting the laser emitting power. Furthermore, this study verifies that the noise level could be effectively suppressed to meet the blue-light storage system specification by using a QWP in the optical path.
The authors acknowledge Mr. Kevin Cheng for his help with the experimental data collection and Dr. Donyau Chiang for his providing of the testing HD-DVD disk. The authors are also grateful for the support of the research in part by the ITRI, by the National Science Council (NSC) under grants NSC95-2221-E-005-148- and NSC 92-2218-E-005-012-, and by the Ministry of Economic Affairs (MOEA) under grant 96-EC-17-A-07-S1-011.
References and links
1. I. Ichimura, F. Maeda, K. Osato, K. Yamamoto, and Y. Kasami, “Optical disk recording using a GaN blue-violet laser diode,” Jpn. J. Appl. Phys. 39(Part 1, No. 2B), 937–942 (2000). [CrossRef]
2. T. Maeda, M. Terao, and T. Shimano, “A review of optical disk systems with blue-violet laser pickups,” Jpn. J. Appl. Phys. 42(Part 1, No. 2B), 1044–1051 (2003). [CrossRef]
3. S. Miyanabe, H. Kuribayashi, and K. Yamamoto, “New equalizer to improve signal-to-noise ratio,” J. Appl. Phys. 38(Part 1, No. 3B), 1715–1719 (1999). [CrossRef]
4. J. P. Wilde, A. A. Tselikov, G. R. Gray, Y. Zhang, and S. Gangopadhyay, “Magneto-optical disk drive technology using multiple fiber-coupled flying optical heads. Part II. Laser noise considerations,” Appl. Opt. 41(5), 884–894 (2002). [CrossRef] [PubMed]
5. M. Itonaga, S. Chaen, E. Nakano, H. Nakamura, F. Ito, K. Iwata, T. Kondo, E. Nakagawa, T. Kojima, A. Nishizawa, and K. Miyazaki, “Optical disk system using a high-numerical aperture single objective lens and a blue LD,” Jpn. J. Appl. Phys. 39(Part 1, No. 2B), 978–979 (2000). [CrossRef]
6. J. W. M. Biesterbos, A. J. D. Boef, W. Linders, and G. A. Acket, “Low-frequency mode hopping optical noise in AlGaAs channeled substrate lasers induced by optical feedback,” IEEE J. Quantum Electron. 19(6), 986–990 (1983). [CrossRef]
7. P. A. Andrekson, P. Andersson, and A. Alping, “Electrical noise measurements on laser diodes for monitoring of optical feedback and mode hopping,” Electron. Lett. 22(4), 195–196 (1986). [CrossRef]
8. A. G. Dewey, “Measurement and modeling of optical disk noise,” Proc. SPIE 695, 72–78 (1986).
9. A. Arimoto, M. Ojima, N. Chinone, A. Oishi, T. Gotoh, and N. Ohnuki, “Optimum conditions for the high frequency noise reduction method in optical videodisc players,” Appl. Opt. 25(9), 1398–1403 (1986). [CrossRef] [PubMed]
11. A. G. Dewey, “Optimizing the noise performance of a magneto-optic read channel,” Proc. SPIE 1078, 279–286 (1989).
12. M. O. Freeman, H. F. Shih, T. P. Yang, J. K. Wang, T. M. Yang, K. L. Lei, Y. H. Chuang, and M. W. Chang, “A laser noise reduction method for holographic pickup heads,” IEEE Trans. Magn. 34(2), 474–476 (1998). [CrossRef]
13. H. F. Shih and D. W. Hsu, “Dynamic laser feedback noise measurement for the optical information storage system,” Opt. Eng. 46(5), 055202 (2007). [CrossRef]