Site: Hitachi-Maxell, Ltd.
6-20-1 Kinunodai, Yawara-mura
Tsukuba-gun, Ibaraki, 300-24 Japan
Date Visited: November 15, 1994
Report Author: D. Keck
What is today Hitachi-Maxell was established in 1961. The sales of the company and consolidated subsidiaries are about $1.5 billion, with 4,000 employees and $79 million in capital investment. The company started as a battery producer, and the name Maxell derives from Maximum Capacity Battery Cell. The company's layered media technology led it to produce magnetic recording media, and in 1960, it produced its first audio magnetic tape. In 1961, Maxell joined the Hitachi Group. Hitachi controls 52% of Maxell, and the rest is publicly owned. In 1980, Hitachi-Maxell began producing both video magnetic tapes and floppy discs for computers.
The Hitachi Central Laboratory started research on optical recording 20 years ago. Maxell began picking up the technology in 1984; the company entered the optoelectronics manufacturing arena in 1994 with a 2.6 Gbit, 12 in. write-once disk. Today the company believes it is Japan's largest video tape producer and second largest audio tape producer.
Twenty years ago, the current Hitachi-Maxell president, Hiroshi Watanabe, was the general manager of the Hitachi Central Research Laboratory (CRL). Like many Japanese companies, Hitachi-Maxell's management is heavily technology-trained. The Hitachi-Maxell research has been little affected by the current recession.
Figure Maxell.1 shows the Hitachi-Maxell organization. Two manufacturing and one engineering research divisions are concerned with optical disk activities. Dr. Toshikatsu Manabe heads the Engineering Research Lab (ERL). He came from Hitachi's CRL 12 years ago and has a seat on the Maxell Board. Dr. Ohta, the JTEC team's host, also came from CRL 4 years prior to the JTEC visit. He reports both to Manabe as Manager of ERL's Second Department and to the head of the Digital Media Division (DMD-manufacturing) as Manager of the Optical Disk (OD) Development Department.
The company's optical storage activities involve about 100 people: 20 in the ERL OD, 30 in the DMD OD department, and the remaining in the Hitachi CRL. Dr. Ohta is responsible for both engineering and production design. The JTEC panel visited both the ERL and the Digital Media Division Factory.
Figure Maxell.2 shows the product evolution scenario that Hitachi-Maxell envisions. The company's first optical product in 1984 was a 12 in. write-once disk used primarily by banks and government for archival storage. Maxell's 650 MB CD-ROM was produced in 1987. The first rewritable was a 640 MB 5 in. magneto-optic (MO) unit in 1989. Last year the company achieved 2 GB in a magneto-optic rewritable disk, with an access time of 40 ms. This is being used in a jukebox connected to computer hard disks for central storage, replacing magnetic tape units. The price is $200/disk or about $0.10/MB. A library system for the United States is being made by Hewlett-Packard. A half-height 5 in. unit has been introduced, but is expensive at $4,000; the selling feature is disk removability.
Dr. Ohta does not think that OD can compete with magnetic hard disks, even with the removable feature, but he does see being able to compete with floppy disks and tapes. He also does not think the 2.5 in. format will replace the 3.5 in. format.
Maxell planned to ship a 3.5 in. 650 MB rewritable MO disk in 1995 (same capacity as the current CD-ROMs). The company also has a 5 in. CD-R write-once at 650 MB.
Maxell will have a 2.6 GB 5 in. MO in production in 1996. With MPEG-2 standards currently being adopted (8X compression) this will accommodate one-hour video storage. The company expects to move this 2.6 GB capacity to the 3.5 in format by about 1999.
The company's CD-R product will require a new organic material to achieve 2.6 GB capacity. Management thinks MO will win over phase change except for Maxell's circa-1998 CD-E erasable replacement for the CD-ROM product.
The 12 in. write-once market is steady to slightly increasing. Maxell has no plans to move to a 14 in. product, because it is not needed for capacity and it doesn't fit the company's equipment format.
Fig. Maxell.1. Hitachi group.
Fig. Maxell.2. Engineering Research Laboratory.
To achieve the product evolution described above, Hitachi-Maxell will change its technology. Generally, for its 0.128 to 0.230 GB products it has used 780 nm lasers and constant velocity. To move from 0.230 to 0.650 GB capacity, the company will move to 680 nm lasers. The move from 0.650 to 1.3 GB capacity will rely on either the company's Multi-Valued MO or Sony's superresolution technology. To increase from 1.3 to 2.6 GB will require shorter-wavelength technology.
Mr. Katsusuke Shimazaki described the Hitachi-Maxell Multi-Valued technology, which was presented at the summer 1994 Magneto-Optic Recording Symposium, a conference for which Dr. Ohta served as program chair. (Sony and other companies presented ten papers on superresolution, but Dr. Ohta thinks next year there will be that many on multi-valued recording.) Multi-valued recording involves depositing multiple layers of MO material. Currently Hitachi-Maxell's engineers will deposit only two. Each layer can independently have a different magneto-optic rotation, thereby giving 2n (n is number of layers) values. Therefore for the 2 layers, there will be a total of 4 bits/storage location. The layers are of slightly different thicknesses, so the light can pass through. The readout requires no new hardware other than 2 coils and the chip for readout. Dr. Ohta said both of these are very inexpensive. The company expects multi-valued technology to be in production in 2-3 years. Shimazaki will transfer to the factory to move the technology and return when it is complete ÷ he has done this on two earlier occasions.
Speed of data access in optical recording is governed by the head weight. A magnetic head weighs 2 grams, while an optical head weighs about 20 grams. Hitachi-Maxell is working on concepts of multiple heads to overcome this limitation.
The JTEC team visited the Engineering Research Laboratory, located in a facility about 3 years old. There are currently about 120 people on the staff. The team saw state-of-the-art equipment in the analytic laboratory, including Auger spectroscopy (Quantum) and ESCA (Perkin-Elmer) machines. Each machine is staffed, but people are cross-trained.
The team also toured one operatorless floppy disk production line at the Digital Media Division. The machines were developed at Hitachi-Maxell itself, and it does not sell to others. Over 15 million disks/month are produced. Dr. Ohta declined to say how many lines of these machines are required for this level of production.
The optical disk line was in class-100 to -1,000 clean rooms. Disk recording layers are deposited by RF sputtering in large rather than small vacuum chambers. Line workers are mostly high school graduates. Testing is by sampling at early stages of the process but 100% at output. Disk tracks are stamped by molding machines costing about $0.3 million each. Several ten thousands of disks are produced by each mold.
The JTEC team had a discussion of technology transfer with Drs. Ohta and Yorozu and with K. Shimazaki and R. Tamura.
Dr. Yorozu joined Maxell from another small company, Optical Storage Corp., a subsidiary of Sumitomo Metal Mining and Sumitomo Chemical. He is Maxell's manager of disk design and development in the Digital Media Division. He described it as difficult to go from research into manufacturing -- schedules are very rigid, difficult criteria must be met, and initial yields are not high in the factory but must be increased.
Hitachi-Maxell transfers products to a pilot line using a joint research and production team. From the pilot line, a product is moved to a main production line and the pilot line begins a new product. In a fairly typical fashion, defect testing of materials is done in the ERL and product testing is done in the factory.
Shimazaki will come and work for Yorozu to transfer his multi-valued concept. He is typical of Japanese researchers who see creative ideas as coming not only from research but also from manufacturing operations. They see Japanese companies as telling researchers that they must be flexible and do more than just research. Scientists are attracted to industry rather than university research in Japan because of the "better" equipment in industry. Once in the company they must become flexible.
Tamura is a researcher doing phase-change media work. He did a two-year sabbatical in the United States, returning one year prior to the team's visit. He went from Hitachi CRL to the Hitachi San Francisco office to learn the U.S. market.