600 Mbyte

Testing parameters in a common test method for f 90 mm/600 Mbyte phase-change optical disks Seiya Yamada Takahiro Kubo Kouhei Yamauchi Junji Ohtsubo, MEMBER Shizuoka University 3-5-1 Johoku Hamamatsu, 432 Japan

SPIE

Masahisa Shinoda Takeshi Utakouji Mitsubishi Electric Corporation 1 Zusho-Baba Nagaokakyo, 617 Japan Toshiaki Iwai Hokkaido University 12-6 Kitaku Sapporo, 060 Japan

1

Abstract. Six samples of phase-change optical disks made by several different manufacturers are evaluated using the developed tester. It is observed that the dependence of recording and erase characteristics on the linear velocity of laser beam scanning on the track and frequency of recording signal were very much affected by laser power. The phase jitter is measured by a proposed test method and the effects of disk tilt and mark distortion on the measured values are indicated by a computer simulation. Common testing methods are discussed and some results of common tests are presented. © 1996 Society of Photo-Optical Instrumentation Engineers.

Subject terms: phase-change optical disk; common test method; linear velocity; disk tilt; mark distortion; phase jitter. Paper 20125 received Dec. 12, 1995; revised manuscript received July 19, 1996; accepted for publication July 23, 1996.

Introduction

Phase-change ~PC! optical disks have many technical merits, including overwritable operation, high density, and no need for a magnetic head. These lead to high data-transfer rates, motion picture recording, and a simple optical head. Magneto-optical ~MO! disks are already standardized, but PC optical disks are just beginning toward standardization. The method of the common tests for these disks has not yet been established. Disk manufacturers may have their own testing methods to measure physical values using their experimental machines. Therefore, it is essential for various disk manufacturers to investigate the interchangeability of PC optical disks. Common specifications for the PC disks were developed by PC optical disk workshop members including us.1 The disks are f 86 mm, double-sided disks, with two substrates 0.6 mm thick and have 600 Mbytes on both sides with a track pitch 1.0 m m and a bit pitch of 0.7m m. The first common test was performed2 in 1993. Common tests for PC disks are very effective in maintaining interchangeability. In 1995, the new common testing machine ~we call it ‘‘Tester-K’’! was developed. The performance of Tester-K was checked by the round-robin test, and six sample disks, which were developed by five manufacturers, were tested using this tester.3 All sample disks had similar recording sensitivities. It was observed, however, that the recording and erase characteristics decreased on the outer areas of the disks. Some differences in crosstalk from tracks (n61) were caused in the radial range of 24 to 37 mm on the disk. Thus, this cause was investigated using a computer simulation.4 In this paper, we present some common test results, which were evaluated with testing parameters such as laser power, linear velocity of laser beam scanning on the track, and frequency of the recording signal. With the test method proposed in this paper, phase jitter was also 3602 Opt. Eng. 35(12) 3602–3607 (December 1996)

measured. Effects of disk tilt and laser power on the observed jitter are theoretically discussed. These studies can enable disk manufactures to know the relative levels of their disk characteristics, and are very helpful for modifying and improving their disks. These are useful for discussion of a first International Standardization Organization/ International Electronic Commission ~ISO/IEC! standard for PC optical disk having 65032 Mbyte capacity. 2

Experiments and Results

2.1 Common Testing System The common testing system is shown in Fig. 1. Tester-K is controlled by a disk driver unit and laser power is modulated by the laser diode ~LD! controller. Read and write laser powers are controlled automatically. This tester was checked by the round-robin test, which was done by comparing it with the other testers used at three companies. Tester-K had the best record characteristics. The optical head of Tester-K includes a laser source with a wavelength of 685 nm and an objective lens with a numerical aperture of 0.6. These values are based on the common specifications. The mask unit has the role of removing signals from sector marks on sample disks to analyze only the signal from recording area using a spectrum analyzer. The time interval analyzer is used to measure time intervals of signals for phase jitter. In this system, the speed of disk rotation is constant at 3600 rpm. The pulse width and modulation code of the signal are 39 ns and ~2.7!RLL code. 2.2 Results of Common Tests The performance of PC sample disks deteriorated in the outer area of the disks. The format of these sample disks is CAV ~constant angular velocity!, so the linear velocity of the outer area is larger than that of the inner area. Thus, the predicted reason is that the short irradiating duration of the laser power prevents the recording layer from crystallizing and amorphizing. First, the effects of the linear velocity

0091-3286/96/$6.00

© 1996 Society of Photo-Optical Instrumentation Engineers

Yamada et al.: Testing parameters in common test method . . .

Fig. 3 Peak power dependence of D 1 NBSNR, and the erase power dependence of D 1 erasability.

Fig. 1 Common testing system using Tester-K.

were also evaluated. The dependence of the narrow-band signal-to-noise ratio ~NBSNR! and the erasability of the linear velocity are shown in Fig. 2. An NBSNR of more than 47 dB is requested for recording in the common specifications. The erasability here is defined as a ratio of the 3T signal amplitude before and after the overwriting by the 8T signal. An erasability of more than 25 dB is requested for overwriting. The recording frequencies of 3T and 8T signals were 8.58 and 3.22 MHz, respectively. Sample

Fig. 2 Dependence of the NBSNR and the erasability on linear velocity. Peak power was 11.5 mW, bias power was 6.0 mW, and rotational speed was 3600 rpm.

disks used in this test are in accordance with the common specifications. Five samples out of six disks were made in 1993 and the other was made in 1995. The recording ability and the erase characteristics of the sample disks both deteriorated with an increase of linear velocity. The same tendency is observed when the rotation speed of the disk drive is changed. This result was produced with the optimum laser power, which we defined. Subsequently, these effects of the linear velocity were similarly evaluated at an other laser power. Figure 3 shows plots of the dependence of the NBSNR and the erasability on linear velocity versus laser power. Symbols D 1 NBSNR and D 1 erasability were defined as the dependence on linear velocity or the deterioration at outer area. At the peak power more than 11 mW, D 1 NBSNR became very small and D 1 erasability indicated the minimum at the bias power of 6.0 mW. NBSNR had broad margins for recording power and linear velocity. But erasability of more than 25 dB was obtained within narrow erase power and linear velocity. The dependence on the linear velocity largely varies with laser power, so that this point should be taken into consideration to determine the laser power. The standardization of a PC optical disk that has a capacity of 650 Mbytes32 is now under discussion in ISO/ IEC.JTC1. The format of the disk is ZCAV ~zone constant angular velocity!, so the recording frequency and the data transfer rate increase at the outer area of the disk. Second, the effects of the recording frequency were also investigated. The frequencies of the recording signals range same 8.58 to 13.33 MHz ~3T signal! and 3.22 to 5.00 MHz ~8T signal!. They are increased to make the mark pitches short at a large track radius. Figure 4 shows the results for the dependence of the NBSNR and the erasability on recording frequency using the optimum laser power. Good recording and erase characteristics were obtained on all sample disks. However, the observed values for other laser powers varied greatly within the recording frequency range. Figure 5 shows plots of the dependence of the NBSNR and the erasability on recording frequency versus laser power. Here D f NBSNR and D f erasability are defined as the dependence on recording frequency or the deterioration at the Optical Engineering, Vol. 35 No. 12, December 1996 3603

Yamada et al.: Testing parameters in common test method . . .

Fig. 6 Proposed method for testing phase jitter, considering the effects of tangential disk tilt. Fig. 4 Dependence of the NBSNR and the erasability on recording frequency. Peak power was 11.5 mW, bias power was 6.0 mW, and rotational speed was 3600 rpm.

outer area. As can be seen from Fig. 5, the range of laser power where D f NBSNR and D f erasability became small was very narrow. As such results show, the peak power of 11.5 mW and the bias power of 6.0 mW correspond to the best choice of the parameters obtained in the common tests, and PC optical disks have sufficient recording ability and technology to be standardized as a high-density recording media. 3

Method of Phase Jitter Measurement and Calculation One of the methods of high density recording, pit-width modulation ~PWM! recording, has been recently used in PC optical disk drives. In the case of PWM recording, the edge of the recording marks should be detected, because with a time lag in the reproduced signal due to disk tilt and the mark distortion some bit errors occur. Therefore, it is very important in the evaluation of the reproduced signal to measure the jitter. The jitter has been measured using many

Fig. 5 Peak power dependence of D f NBSNR and the erase power dependence of D f erasability. 3604 Optical Engineering, Vol. 35 No. 12, December 1996

methods of disk manufacture.5,6,7 To know the relative levels of their disk characteristics, a common method of jitter measurement must be established. However, a common test method has not yet been defined, so we measured the phase jitter of the pit-position modulation ~PPM! using two of our proposing methods as is described. First, the effects of only disk tilt was investigated. As shown in Fig. 6, the negative time, which was just between 3T and 8T signals was chosen as the ideal negative time Dt 1 . In case some tilt angle exists on the sample disk, the negative time is reproduced as Dt 2 . Thus, phase jitter is defined as the difference between Dt 1 and Dt 2 . Figure 7 shows plots of the phase jitter and tangential tilt on disk A. The negative time was measured by averaging 104 samples using a time interval analyzer. The tangential disk tilt on the sample disk was measured using an optical fiber probe. The axial deflection on the radius of a rotating disk can be monitored. By differentiating the deflection, tangential disk tilt on the radius of a sample disk is shown in this figure. This indicates the absolute value, which includes turntable tilt. A tilt less than 5

Fig. 7 Relationship between phase jitter and disk tangential tilt on disk A.

Yamada et al.: Testing parameters in common test method . . .

mrad was obtained at all disk positions, which corresponded to the specification. Phase jitter was less than 7 ns, which corresponded to 20% of the recording laser pulse. The shapes of the two curves are very similar. This figure also indicates that the phase jitter greatly depends on tangential disk tilt. The effects of disk tilt on phase jitter were also investigated by carrying out a computer simulation. The calculated method is the same as the testing method. This was calculated according to scalar diffraction theory based on Hopkins’ model.8 Distributions of light g 0 (x 1 ,y 1 ) passing through the objective lens should form the image of diffraction g 1 (x f ,y f ) on the surface (z 5 f ) of the disk as follows: 1 g 1 ~ x f ,y f ! 5 jl

EE

exp~ jkr ! dx 1 dy 1 , g 0 ~ x 1 ,y 1 ! r 2` `

g G ~ x 1 ,y 1 ! 5A exp~ 2x 21 /w 2x 2y 21 /w 2y ! .

1 ~ x 21 1y 21 D 2 /4!

~4!

where D is the diameter of objective lens. The intensity distribution that received the phase shift term is given by p L ~ x,y ! 5exp$ jk @ f 2 ~ f 2 1x 21 1y 21 ! 1/2# % .

~5!

Therefore, the distributions of light is expressed as ~6!

From Eqs. ~1! to ~6!, g 1 (x f ,y f , f ) is obtained by simple Fourier transform as follows:

FS

3@ g 0 ~ x 1 ,y 1 !# .

DG

F ~ x f /l f , y f /l f

2`

r ~ x f ,y f ! g 1 ~ x f ,y f , f !

exp~ jkr ! dx f dy f , r

~9!

g 82 ~ x 1 ,y 1 , f ! 5

21 exp$ jk @ f 1 ~ f 2 1x 21 1y 21 ! 1/2# % ~ jl f !

H

3F x f /l f ,y f /l f r ~ x f ,y f ! g 1 ~ x f ,y f , f !

FS

3exp jk

2f

r ~ x f ,y f ! 5 AR d exp~ j2kG d !

3

H

Rm Rd

1/2

exp~ jkW C !

p 8

~ Groove!

~11!

,

where R m is the mark reflectance, and R d is the disk reflectance. When this light goes back to the object lens, it receives the aperture function @Eq. ~4!# and the phase function @Eq. ~5!#. Thus, distributions that are just passing through the object lens are given by g 2 ~ x 1 ,y 1 , f ! 5

2exp~ j2k f ! m A ~ x 1 ,y 1 ! F x f /l f ,y f /l f ~ l f !2

FS

H

E E `

0

2`

2`

~7!

Generally, wavefront aberration of the coma is caused when a laser beam passes a parallel plate at a tilt angle of u to the optical axis of the beam.9 The wavefront aberration coefficient is given as W c 51/2d ~ n 2 21 ! n 2 ~ n 2 2sin2 u ! 25/2 sin u cos u NA 3 , ~8! where d is the disk thickness, and n is the refractive index. The distributions of light g 28 (x 1, y 1 , f ), which were reflected at the disk surface in consideration of the tilt angle, recorded mark, and track form, are given by

x 2f 1y 2f 2f

D GJ

.

~12!

From Eq. ~12!, the reproduced amplitudes are calculated by integrating the intensity distribution on two divided detectors as follows: ~ I sum! p2p 5I 1 1I 2 5

!

~10!

,

~ Land!

0 G d5

D GJ S D

x 2f 1y 2f

3 r ~ x f ,y f ! g 1 ~ x f ,y f , f ! exp jk

g 0 ~ x 1 ,y 1 ! 5g G ~ x 1 ,y 1 ! m A ~ x 1 ,y 1 ! p L ~ x 1 ,y 1 ! .

x 2f 1y 2f 1 exp jk f 1 jl f 2f

`

i.e., the Fourier transform of g 1 (x f ,y f , f ). Subsequently, g 28 (x 1 ,y 1 , f ) is given as follows:

~3!

To take into consideration the lens aperture, this aperture function is given by

g 1 ~ x f ,y f , f ! 5

EE

~2!

where k is 2 p /l to assume a light laser as the beam of one wavelength. An intensity distribution g G (x 1 ,y 1 ) to the objective lens is assumed as the Gaussian beam given by

H

1 jl 3

~1!

r5 @ f 2 1 ~ x 1 2x f ! 2 1 ~ y 1 2y 1 ! 2 # 1/2,

m A ~ x 1 ,y 1 ! 5

g 82 ~ x 1 ,y 1 , f ! 5

1

E E `

2`

g 2 ~ x 1 ,y 1 ! 2 dx 1 dy 1 `

0

g 2 ~ x 1 ,y 1 ! 2 dx 1 dy 1 .

~13!

As shown in Fig. 7, the negative time of the 3T signal is calculated as follows: Dt 1 5 ~ I 1 ! 80 2 ~ I 1 ! 90 , Dt 2 5 ~ I 2 ! 80 2 ~ I 2 ! 90 .

~14!

Thus, phase jitter is given by phase jitter5Dt 1 2Dt 2 .

~15!

Optical Engineering, Vol. 35 No. 12, December 1996 3605

Yamada et al.: Testing parameters in common test method . . .

Fig. 8 Dependence of phase jitter on tangential tilt. Measured values are shown by dots and the calculated values by a solid line. Mark distortion of 0.05 m m was added as a condition.

From the calculated results, the laser beam with tilt has a larger sidelobe than that without tilt. The intensity of the laser beam was asymmetrical. Therefore, as shown in Fig. 6, a reproduced waveform with disk tilt has a time lag. In Fig. 8, the measured and calculated phase jitter are denoted by dots and squares and a solid line, respectively. Generally, phase jitter depends on not only disk tilt but also mark distortion, so that these calculated plots are results that took into consideration a mark distortion of 0.05m m in addition to tangential disk tilt. As can be seen from the calculated values, phase jitter became larger as the tangential tilt increased and phase jitter was less than 7 ns within a tangential tilt of 65 mrad. The measured and the calculated values were very similar. The effects of disk tilt on phase jitter were theoretically confirmed. Subsequently, the effects on the mark length distortion were investigated. As shown in Fig. 9, 3T and 8T random signals were recorded on an unrecorded track. The phase jitter for other random mark length was not measured in this common test, because the difference of the time inter-

Fig. 10 Dependence of phase jitter on mark shift. Measured values are shown by dots and the calculated values by a solid line.

val between 3T and 8T signals was larger than for other signals. The slice level was always set up as a 50% ratio of 3T signal. The slice level of 3T signal varies with the record laser power, the positive time (Dt 2 ) varies with the slice level and the recorded mark form. Therefore, the positive time of the 8T signal in the case that was recorded by the optimum laser power was defined as the ideal positive time (Dt 1 ). Thus, phase jitter was measured as the difference Dt 1 2Dt 2 . To minimize the effects of disk tilt, the relative position of a sample disk and a turntable was constantly fixed. A computer simulation was similarly carried out to confirm the effects of mark length distortion. The mark length distortion was caused by changing write laser power in the experiment, and, on the other hand, by changing the mark length in the calculation. Figure 10 shows the effects of mark shift and write laser power on phase jitter. Measured values are shown by dots, squares, and triangles and the calculated values by a solid line, where the measured values were related to the write power is shown on the upper axis. Phase jitter of less than 8 ns was obtained for three sample disks. As seen from the calculated results, the jitter was also influenced by mark shift and less 10 ns within the range of 60.1m m. Thus, the effects of mark distortion was theoretically confirmed and it was obvious that these sample disks had broad margins for their write laser power. In this paper, the effects on the mark edge distortions were investigated, but it is similarly assumed that these distortions of PC optical disks cause the increase of phase jitter. 4

Fig. 9 Method of testing phase jitter considering the effects of mark distortion. 3606 Optical Engineering, Vol. 35 No. 12, December 1996

Conclusions

Interchangeability tests for disks and drives are indispensable for standardization of phase-change optical disks. In these common tests, the linear velocity of the laser beam and the frequency of the recording signal should be taken into consideration to determine the laser power. Phase jitter, whose testing method has not been defined, was also measured by a proposed method. The effects of disk tilt and mark distortion on the observed values are theoretically confirmed using a computer simulation. From such results,

Yamada et al.: Testing parameters in common test method . . .

PC optical disks are shown to have sufficient interchangeability. The capacity of these sample disks in accordance with the common workshop specifications is 600 Mbytes. Being based on the fruits of these developed disks the first standard for PC optical disks of 1.3 Gbyte capacity is discussed in ISO/IEC. JTC1 SC23.

Kouhei Yamauchi received his BS degree in mechanics from Shizuoka University in 1995 and is a graduate student in Prof. Takahiro Kubo’s lab. His current research interest is computer graphics.

Acknowledgments The authors are very grateful to S. Tagaya and T. Kinoshita of PULSTEC Co. for helpful discussion on the disk drivers. The PC optical disk samples tested in this study were made by the following five manufacturers: Asahi Chemical Industry Co., Ltd., Toshiba Corp., Toray Industries, Inc., Hitachi Ltd., and Matsushita Electric Industrial Co., Ltd.

References 1. Phase-Change Optical Disk Workshop, Phase-Change Optical Disk Common Specifications, No. WS 13-07, Budapest, Hungary ~1993!. 2. T. Kubo, T. Sugaya, I. Sato, J. Ohtsubo, and T. Iwai, ‘‘Testing of phase-change ~PC! optical disk for common specifications,’’ Proc. SPIE 1983, 460–461 ~1993!. 3. S. Yamada, T. Kubo, T. Ito, and K. Yamauchi, ‘‘Exchangeability test on f 90 mm/600 MB phase-change ~PC! optical disks for common specifications,’’ in Int. Soc. Opt. Memory ’95 Technical Digest, pp. 97–98 ~1995!. 4. S. Yamada, T. Kubo, and K. Yamauchi, ‘‘Exchangeability test on f 90 mm/600 MB phase-change ~PC! optical disks for common specifications,’’ Jpn. J. Appl. Phys. 34, 481–485 ~1995!. 5. H. Kobori, H. Hasegawa, and T. Sugaya, ’’High density recording providing over 1 GB capacity for a 90 mm phase-change optical disk,’’ in Optical Data Storage ‘94 Technical Digest, pp. 67–68 ~1994!. 6. T. Nishida, H. Sugiyama, and S. Horigome, ’’Sn-Sb-Se phase-change media for high-density write-once optical disk,’’ in Optical Data Storage ‘94 Technical Digest, pp. 63–64 ~1994!. 7. N. Miyagawa, E. Ohno, K. Nishiuchi, and N. Akahira, ‘‘Phase change optical disk using land and groove method,’’ in Int. Soc. Opt. Memory ’95 Post-deadline papers Technical Digest, pp. 25–26 ~1995!. 8. H. H. Hopkins, ‘‘Diffraction theory of laser read-out systems for optical video discs,’’ J. Opt. Soc. Am. 69~1!, 4–24 ~1979!. 9. S. Kubota, ‘‘Aplanatic condition required to reproduce jitter-free signals in an optical digital disk system,’’ Appl. Opt. 26, 3961–3973 ~1987!. Seiya Yamada received his BS degree in mechanics from Shizuoka University in 1994 and is presently pursuing his MS in precision engineering. He is currently studying methods of testing optical disks. He is a member of the Japanese Society of Applied Physics.

Takahiro Kubo received the BEng and DEng degrees in applied physics from Osaka University in 1961 and 1974, respectively. From 1961 to 1990 he worked at Mitsubishi Electric Corporation, where he was engaged in research on optical pumping, holography, and optical disk memories. He is presently a professor of engineering at Shizuoka University. He is a leader of the Optical Disk Study Group (OITDA), co-leader of the Phase-Change Optical Disk Workshop, member of the Japanese National Committee for ISO/IEC SC23, and advisory committee member of the International Symposium on Optical memory.

Junji Ohtsubo received his BS degree in electronics from the Kyusyu Institute of Technology in 1973 and the MS and PhD degrees in electronics from Hokkaido University in 1975 and 1978, respectively. In 1978, he joined the Mechanical Engineering Laboratory. During 1981 and 1982, he was a Research Associate at the Institute of Optics, University of Rochester. He joined Shizuoka University as an associate professor in 1985 and is presently a professor. His current research interests are statistical optics, speckle, optical metrology, active interferometry, optical information processing and computing, and nonlinear dynamics in optics. He is a member of SPIE, the Optical Society of America, IEEE, the Japanese Society of Applied Physics, the Optical Society of Japan, and the Laser Society of Japan. Masahisa Shinoda received his BS and MS degrees in physics from Osaka University in 1979 and 1981, respectively. He then joined Mitsubishi Electric Corporation, where he is currently engaged in research on optical design for optical disk memories at the Video Disc Business Development Center.

Takeshi Utakouji received the BE and ME degrees in 1988 and 1990, respectively, from Shizuoka University. He joined the Consumer Electronics Development Laboratory at Mitsubishi Electric Corporation in 1990. He is now with the Optoelectronic and Microwave Devices Laboratory, Itami, Hyogo, Japan. He is engaged in the development of short wavelength laser diodes. He is a member of the Japan Society of Applied Physics. Toshiaki Iwai received BEng, MEng, and DEng degrees in electronics from Hokkaido University in 1979, 1981, and 1984, respectively. He was a postdoctoral fellow of the Japan Society for the Promotion of Science from 1984 to 1985. He joined the Shizuoka University engineering department in 1985 as an assistant professor and was an associate professor from 1986 to 1993 in the Department of Electronic Engineering. In 1993, he joined the Research Institute for Electronic Science, Hokkaido University, as an associate professor. His current research activities are in the spacetime statistical properties of laser light scattering from dense medium, analog optical computing and its application to metrology, and inverse optical scattering. He is a member of the Optical Society of America, the Optical Society of Japan, the Japan Society for Applied Physics, and the Institute of Electronics, Information, and Communication Engineers. He received the award from the Japan Society of Applied Physics in 1995. Optical Engineering, Vol. 35 No. 12, December 1996 3607