An optical disk "jukebox" mass memory system is being developed which will provide automatic access to any data in a store of 1013 bits within five seconds. This system contains a library of 128 optical disks with mechanisms for retrieving any disk, loading it onto a turntable, and recording or playing digital data at a rate of 50 Mb/s. The optical disks are housed in protective cartridges to facilitate handling by both the operating personnel and automatic disk handling mechanisms. Cartridges are moved from the store to a load station by a precision, belt-driven X-Y transport mechanism. The load station then mounts the disks onto a precision turntable where they are spun up to speed while still housed within their protective cartridges. Next, data is recorded or played back on these disks using focused laser beams. Finally, the disk is stopped and returned to its storage location so that the system can handle a new request. The disk handling mechanisms were designed to minimize mechanical shock and vibration while providing a rapid, smooth operation. A special centering hub design for the turntable minimizes disk eccentricities during multiple load/unload cycles and allows easy inter-changeability from machine to machine. This paper will describe the system, subsystem, and mechanism designs that were used in developing an engineering model of this jukebox concept. Test results from this engineering model will also be presented in this paper.

Optical Disc Recorder parameters are examined with respect to data and control management. Data processing requirements discussed include sync timing recovery and error management. Control processing requirements discussed include disc drive interface and user computer interface. In addition, the approach to data handling using an intelligent control processor is presented. The modular hardware controller described is adaptable to disc drive configurations: 1) to support data rates exceeding 100 Mb/s 2) to support multiple turntables per drive, 3) to support a jukebox library and automatic disc loader mechanism. The controller design considerations will be given along with the selected architecture.

The NASA Data Systems Technology Program is using the recent advances in optical disk storage technology to develop a prototype centralized archive for data collected by scientific satellites. Currently under development for Marshall Space Flight Center, the NASA Data Base Management System (DBMS) will ultimately demonstrate the feasibility of ingesting and recording data at rates up to 50 megabits per second and of retrieving and transferring the recorded data to a diverse community of users. A major component in the DBMS is an optical disk system which provides on-line storage for 1013 bits of data.

The relationship between errors in digital data retrieved from a storage medium and the fundamental properties of the medium are developed.

Optical recording systems include a transparent layer between the diffraction-limited objective and the recording surface, which prevents dust and dirt from obscuring the optical signal. A laser beam focused on the recording surface may be aberrated by thickness variations or birefringence in this cover sheet. This paper presents calculations of these aberrations and a discussion of the corresponding cover-sheet tolerances.

The design and the development of optical storage creates a significant impact on system architecture. Some of the highlights of optical storage include the following: both direct and sequential access are supported, large units of data are divided into bands that act as the unit of physical and logical data addressahility, and the user deals with logical records that can be of variable length, up to two megabytes. Ease of use is enhanced by the simplicity of data manipulation and access. Future consideratons will allow rapid retrieval of a platter from a library for subsequent automatic insertion of the cartridge into an optical storage unit.

Optical Disc Mass Storage Systems provide for storage of large quantities of digital data on a relatively compact medium with negligible access delays. Until now, however, all Optical Disc Systems have been designed for laboratory or computer room environments. For many years, RCA has been a leader in production of compact, rugged magnetic recording equipment for use in tactical and other non-benevolent environments. Now, this technology is being applied to Optical Disc Systems and the result is a design of a compact, high-data-rate, high-capacity data recording system for field operation. Discussion topics include system architecture and physical configuration.

Recognizing the need for overall system architecture design for the emerging optical disk storage systems, RCA designed and built a demonstration optical disk recorder system for the purpose of experimentation and evaluation of data handling and system control techniques. The demonstration system design is based on a previously developed semiconductor optical disk drive unit and an intelligent hardware/ software controller, configured as a digital image data file mass storage system.

Due to the relatively high raw error rate on present day Optical Media, a high performance, real-time error detection and correction system was necessary to achieve the kind of data reliability required in computer storage systems. This paper describes such an error detection and correction system, based on an interleaved Reed-Solomon code, that is capable of correcting multiple burst errors. A pipelined architecture combines several special-purpose processors designed for specific decoding functions. This system, operating at speeds of up to 30 megabits per second, achieves an increase in reliability of more than ten orders of magnitude.

A practical optical digital memory using Tellurium sub-oxide [TeOx (x = 1)] thin film disc has been developed for document and digital image storage. In the recording mode, the reflectivity of TeOx thin film recording layer evaporated on a pregrooved PMMA substrate is changed by the absorption of the irradiated thermal energy of focused recording laser beam and the signal is recorded on the disc in the form of a chain of micron-sized flat and non-deformed dots. The optical memory using a laser diode as a light source is capable of storing up to 700 MBytes in user data on one side of a preformatted and pregrooved disc, 200 mm in diameter with an access time below 0.5 seconds. Input NRZ digital data are converted to 4/5 MNRZ code which has less DC component and recorded on the optical disc rotating at 900 rpm. Recording data speed on the disc is 6.8 Mbps and the length of the shortest bits is 0.8 μm The error rate of less than 10-8 has been achieved by the implementation of an error correction scheme with 5.4 % redundancy.

The video and the audio disc systems have been realized and the progress of optical disc memory systems allows the new products such as the video file system and the document file system. The optical disc memory system has a number of advantages compared with the current magnetic disc memory. These are the large recording capacity, the high recording density (10 to 100 times more than ordinary magnetic disc) and the contact free record and retrieval of the information, which eliminates the "Head Crush" problems and realizes the reliable information storage system. However, high recording density and small track pitches (usually 1.6--2.5 micron meters) increases the probability of error rate of recording infomation. Bit error rate (BER) of the recorded optical disc sometimes presents 10-4-10-5 Which is much higher than the magnetic disc. This high BER strongly depends on the production process. Optical disc is preformatted with the address information and with the tracks for infor-mation recording. The pregrooved disc used to be produced by the optical mastering process and the U-V replication process, which needs the very high-grade clean room and the expensive facilities. Process control of the laser cutting on the resist coated glass master is quite sensitive and must be treated with extreme care. When the large quantity of mass produced optical memory discs is supplied, a new mastering process would be needed. One of the solutions of this problem is the dry mastering process with a mechanical cutting system.

The Cataloging Distribution Service (CDS) of the Library of Congress working with Xerox Electro-Optical Systems, Inc. of Pasadena, California has succeeded in developing and implementing one of the first image processing document delivery systems in the world using optical disk storage technology. The system uses automated file management and demand printing capability to replace a very costly, labor intensive, manual operation. The system includes a high resolution 480 bpi laser scanner, high capacity (4.5 gigabytes/surface) optical disk storage and a high resolution (480 bpi) laser printer. Through the implementa-tion of this system the Library of Congress has shown initiative and leadership, and has demonstrated that the automated systems of the future will process not only character code data but imaginal data in an attempt to move away from paper as a means of transmitting information.

The Computer Optical Disk Archival Memory, CODAM, development is a Defense Department program to ensure the availability of an advanced optical disk peripheral to satisfy expanding, archival, high performance computer memory needs. Ultimately, to satisfy user operational requirements, a number of such peripherals will be married to an automated disk handler. The resultant equipment will form an Optical Disk Library ODL, capable of over 1014 bits of on-line storage. The characteristics of the CODAM hardware and the ODL are described. The trade-offs that lead to the choice of the CODAM specifications are presented.

A document storage application is described in the context of the insurance industry requirements for storage of massive paper files. Optical disc will provide the storage, which is the last major piece that had to be invented. This paper describes the characteristics required in a large digital image storage system.

The development of a data capture and retrieval facility as a technology demonstration for future NASA data systems is described. The archival storage subsystem is a multi-disk optical recorder/reproducer under joint development by the U.S. Air Force (Rome Air Development Center) and NASA. This paper describes the application of the optical disk in the context of a data base management system which automatically constructs catalogs of incoming sensor data sets, and supports on line browsing and queries for data sets fulfilling search criteria specified by users. Data sets which satisfy user queries can be requested for immediate delivery to the user via a network of computers linked by communications services. The paper describes a solution to problems associated with interfacing the high speed storage subsystem (50M bit/sec.) to computers with inherently lower I/O bandwidths.

In the next few years several companies will bring to the marketplace an optical memory drive and a controller. These drive and controller systems will be an attempt at achieving very high capacity. high transfer rate and low cost per megabyte.

A run-length-limited code, designated 3Φ(1,7), has been applied to optical disc recorders for the purpose of increasing the information capacity of a disc. The 3Φ(1,7) code was selected by an organized search of an extensive collection of NRZ-like, phase, and run-length-limited codes. Theoretical examination indicated that greater than a 20% linear packing density increase could be expected under the conditions that the recorder was forming the smallest pits possible. The factors which led to this conclusion and the properties of the 3Φ,7) code are discussed. Details of an experimental verification of the conclusions reached herein are presented in a companion paper, "Implementation of 3Φ(1,7) code for improving packing density on optical disc recorders," by P. Hess and J. Waring.

The implementation and testing of a 3Φ (1,7) code on optical disk recorders for improved packing density is described with details concerning data formatting and CODEC electronics. Comparative test results presented relate off-disk bit-error-rate (BER) to linear packing density (LPD) for 3Φ (1,7), delay modulation (DM), and straight NRZ-L codes. 3Φ (1,7) improves LPD by more than 25% over the previously used DM code.

The KAD Digital Information Systems Laboratory As developed diode laser record and play-back equipment for use with Kodak Laser Write/Read (LWR) optical disks which have been described in previous sessions. The Digital Disk Breadboard (DDB) in operation for several years and the Advanced Optical Disk Demonstration Unit (AODDU) completed this year both offer convincing demonstrations of the excellent performance achievable with small light-weight apparatus. All our diode laser equipment utilizes lenses and optical assemblies of our own design and fabrication.

This paper describes a tracking servo system, with a new optical head, capable of locking an objective lens in place in the head. Using the new optical head, the lens position deviation from the center position in the lens movable range can be detected. Head and lens positions are controlled co-operatively using a detected lens deviation signal. In the tracking servo system, no lens vibration ocuurs, disk eccentricity does not cause track offset and track jump can be repeated as many time as necessary. As a result, stable and high speed accessing and accurate track following are achieved.

A simple and reliable optical head is needed for practical application of optical mass data storage. The so-called push-pull tracking method is widely used in pregrooved disc systems. A disadvantage of this tracking method is that track following range is limited to a few tens of microns, which is too little for practical use. The limited range is due to disc eccentricity, which displaces the axis of the reflected beam. The authors have overcome this problem by a new servo method which compensates electrically the undesirable influence by detecting the displacement from the driving signal of the tracking actuator, and have improved the track following range to 350 μm.

A high performance focus motor is an essential component in the optical path of a laser memory device. The demand for high performance is due in part to the small depth of focus of the optical system resulting from the high numerical aperture of the lens, in conjuntion with disk wobble and media surface roughness. The need for small overall size and weight adds to the already tough requirements. To obtain the maximum linear acceleration from a given focus motor volume, it is necessary to optimize the magnetic circuit and coil design. This requires one mathematical equation which models the entire electromechanical system. This equation can then be analyzed to discover the optimum values and dimensions for the design. This paper details the creation of such a mathematical equation. It includes, for an example problem, an optimum magnetic circuit and coil design which is within realistic constraints and limitations.

RCA has been developing high data-rate optical disk systems capable of recording at rates of 100 Mbps and above. Descriptions of the electrical, mechanical and optical designs, as well as experimental results, will be given for argon-based systems. Preliminary design and details will also be given for AIGaAs laser diode array systems, including some array parameters.

The development of an error budget must start with the identification of the read/write system parameters and how they are influenced by each sub-system in the drive. The read/write characteritics of the media are crucial to the read/write system and represent the most significant portion of any budget. Thus, the performance of the media must be modeled and the read/write chain designed to track any related disturbances. Presented are analytical techniques that can be applied to evaluate any budget and determine performance requirements. An example of budgeting and evaluating performance of a typical conceptual implementation is given within this paper. The budgeting of errors in any system should start with a block diagram of the system of interest and qualitative identification of error sources. For this paper consider the block diagram of an optical read/write system given in figure Before identification of error sources can be done. the mechanisms of the read and write processes must be understood.

General considerations involved in designing wide bandwidth low-noise photodetector preamplifiers are discussed. A specific circuit configuration is presented which meets the general requirements for high-speed low-noise operation. Relevant noise equations are presented and theoretical results are compared to actual data for a working 50 MHz version of the preamplifier. Test data confirms the usefulness of this configuration.