Theodore H. Kehl

LM 2 — A Logic Machine Minicomputer

In 1963 Clark and Molnar1developed the LINC computer (Laboratory Instrument Computer), principally for use in biomedical research. Several of the features of this machine (autoindexing, LINC tapes, 12-bit word, data break) later appeared in the PDP-5 and, still later, in the PDP-8. It is probably fair to give Clark and Molnar credit for thus starting the minicomputer revolution.

Interactive real-time computation

Abstract A conversational computer system allowing direct communication between scientist and computer-driven experiment is described. Important features are (1) raw data is retained in digital form, (2) fairly extensive editing and computation capabilities are available on-line and (3) a file structure is built in to the system enabling a scientist to have results from previous experiments available. This system has been implemented on a modest-size medium-scale computer (Raytheon PB440) and, whereas not dedicated to this system, could justifiably be so limited.

A digital system for studying interstitial transport of dye molecules

Abstract A method is presented for quantitatively describing the transport of absorbing dyes through the interstitial space in transilluminated preparations. A capillary bed is perfused with a vital dye, such as Patent Blue V, and the diffusion of the dye across the capillary wall and through the surrounding tissue is monitored by a microscope-closed-circuit television system. A slow-scan television camera, which scans a single line of the field 30 times per second, detects changes in optical density at various distances from the capillary. Changes in optical density are compared to those predicted by the equation for diffusion from a cylindrical source in an isotropic homogenous medium, and a value for the diffusion coefficient of the dye in the interstitium is obtained.

The regulation of stroke volume in the resting, unanesthetized dog☆

Abstract In the resting dog who displays a marked sinus arrhythmia, a model of stroke volume control was tested by a procedure of vectorial reconstruction. In the model, stroke volume is determined by preceding cardiac interval (I), arterial pressure (PA), and filling pressure (PDV). The fit of the model is good; changes in stroke volume are accurately predicted by changes in the other variables. A unit-vector solution indicates the relative importance of each variable. A set of (regression) coefficients indicates how a change in each variable changes the stroke volume. Agreement from run to run and from animal to animal is good. All of the variables—I, PDV, and PA—are necessary to achieve the best prediction of the stroke volume. Each of the variables makes a unique contribution to the stroke-volume changes. If another variable acts in conjunction with those named, its effect is less than that of any of the others. Sympathetic discharge to the heart appears to cause no changes in stroke volume in these resting animals.

Systems programming on-line

Abstract Use of the usually present oscilloscope display device on a real-time computer enhances the ability to write and debug programs. A text editor (CLIB) allows a programmer to manipulate massive source-statement character streams easily (or, indeed, any character stream) in preparation for assembly or compilation. A second related, but independent dynamic debug system (MACT) provides the programmer with the ability to “watch his program happen”. In both systems the oscilloscope provides very rapid displays of text and/or status information and places the programmer on-line to the program he is developing.

Basil architecture - an hll minicomputer

During the past several years a computer-aided design system (the Logic Machine) has been under development in our laboratory. Briefly, the Logic Machine consists of a microprogrammable control processor, one or more functional units, one or more bidirectional buses, and a microprogram all arranged to perform a specific digital algorithm. Our major goal has been to be able to construct any digital device with this system. We have been able to build a graphics display terminal (1), a floating point processor (2), string/array auxiliary processor (3), and a minicomputer (4). It has amazed us to see how simple and fruitful it has been to construct these devices. In this paper we describe the use of the Logic Machine design system to build still another digital device, a high-level language minicomputer. The motivation for this effort is probably obvious to all hardware designers; software is the most expensive part of a computer system and a high-level language computer will significantly reduce software costs. Not so obvious is the task simplification at the systems programming level, enabling a programmer to quickly review his work, decide on additions or corrections, and expand a program or system. These benefits reduce the layers of logic and, we are convinced, will enable much more sophisticated software systems to be built.

Design Automation for Custom Hardware in Computer Research

The equipment and software which permit a university group to perform computer hardware research is described. The objectives were to obtain a low-cost, minimally specialized hardware system coupled to computer-aided design techniques. Two measures demonstrated that these objectives were met: (1) students have been able to build reasonably large devices as part of their graduate research and, (2) the system has paid for itself many times over in the construction of computer devices which are not available commercially.

Computer data collection and editing for hemodynamic studies.

This paper presents a description of a computer system used in our laboratory for data collection, editing and analysis in cardiovascular research. Analog data are converted on-line into digital form by an A-D converter, and the raw data are stored on digital tapes in a format which expedites editing by the computer. Editing algorithms are discussed, with comments on their usefulness, reliability and possible errors. Computing modules available to analyze data are described very briefly. We further discuss the situations and types of experiments which justify the use of such a computer-based system.

Interactive continuous system simulation

Abstract In this paper an overview of an interactive digital analog simulator is presented. The objectives of simplicity and availability have been met while significantly adding to the power and flexibility of simulation in the biomedical research environment.

Applying computers in the research and development laboratory

To bring about a better understanding of the needs of laboratory researchers employing small computers and the potential of these computers for meeting such needs, a panel of two computer users, a system designer and a computer manufacturer representative will address the problem. The panel will discuss: (1) electronic and electromechanical devices comprising a computer system (hardware), (2) computer programs (software), and (3) connecting circuitry linking laboratory instruments and the computer (interface).