EAI Electronic Associates Newsletter No. 005 (January–February 1966)

This document is an English translation of the original German-language newsletter published by EAI Electronic Associates GmbH, Aachen, January–February 1966, issue No. 005.

---

Brush Mark 280 Recorder

The company BRUSH Instruments is one of the leading American manufacturers of direct-writing measuring and recording systems. Instruments from this company find application in American defense systems, in the navigation system of Polaris submarines, as well as in satellite and space projects.

The 2-channel recorder Mark 280 is a portable, fully transistorized instrument suitable for both mobile and stationary use. It allows the evaluation of recorded processes with an accuracy previously unattainable in mid-frequency recorders. For the first time, a static and dynamic accuracy of 0.5% has been achieved with a channel width of 80 mm. The upper frequency limit at full utilization of the 80 mm wide channel is 35 Hz; however, frequencies up to 200 Hz can be processed when amplitudes are reduced.

The system owes its high accuracy to a feedback control loop. The position of the writing stylus is continuously monitored by a sensor; the difference between the commanded and actual position is detected and an appropriate signal is fed to the drive coil after sufficient amplification. The sensor operates without contacts, so accuracy degradation due to wear is excluded. The system has no restoring spring; the drive coil therefore responds only to the signal and remains unaffected by mechanical forces. Virtually no energy is required to hold the stylus at a given position; drive coil and amplifier therefore do not heat up, preserving their stability.

Geometrical errors inherent in conventional systems are eliminated through a linkage mechanism. The motion of the writing stylus is rectilinear, making it possible to use paper with a rectangular coordinate grid, which facilitates reading and evaluation.

The writing system of the Mark 280 operates with ink. This has the advantage that no expensive special paper need be used — costs amount to only one-third of the next cheapest method. The disadvantages normally associated with liquid-writing systems are largely eliminated here. The writing fluid is pressurized and is pressed into the paper surface. The pressure is generated by a motor that adjusts it to prevailing requirements and shuts it off when stationary. The result is a clean, sharp line of uniform width and high contrast at all writing speeds. There is no ink spatter or leaking.

On request, the instrument can be supplied with a calibrated zero-suppression feature. Adjustment is made on a micro-scale with an accuracy of 1/4000, independent of the selected sensitivity.

Thanks to its favorable characteristics, the Mark 280 can be used universally in research and industry wherever small, low-frequency voltage signals must be measured and recorded with high accuracy without loading the signal source. In addition, any variable that can be converted into a voltage change can be recorded. Examples of application areas include:

Physics: Gas chromatography, X-ray spectrography, nuclear measurements, temperature monitoring with thermocouples and resistance thermometers.

Mechanical Engineering: Recording of mechanical quantities such as pressure, acceleration, force, torque, strain, etc.

Medicine: Physiography, biotelemetry, monitoring of body temperature, blood pressure, pH values, cardiographic and encephalographic studies.

Specifications

Electrical:

• No. of channels: 2 analog, 2 event (3-position)
• Channel width: 80 mm, 50 divisions per channel
• Trace width: 0.01” nominal
• Presentation: true rectilinear
• Writing method: pressurized fluid (closed system)
• Chart capacity: High-contrast paper — 275 ft; Reproducible paper — 400 ft
• Writing fluid: approx. 2 oz, capacity for approx. 50 rolls, normal recording
• Major dimensions: 18-3/4” high, 10-1/2” wide, 7-1/2” deep
• Weight: 76 lbs
• Power requirement: 105–125 V, 60 cps, 200 W (50 cps and 400 cps and/or 220 V models also available)
• Sensitivity: 0.5 mV/chart div.
• Measurement range: 500 V full scale
• Frequency response: 80 mm peak-to-peak to 35 cps (flat within 0.4 div.); 40 mm peak-to-peak to 60 cps; 15 mm peak-to-peak to 100 cps; 4 mm peak-to-peak to 200 cps
• Signal input: 1 megohm impedance, constant, floating, guarded
• Common mode rejection: DC — 120 dB; Maximum voltage off ground: ±500 V
• Noise: 0.2 division peak-to-peak with 100 kohm source impedance, any attenuator position
• Linearity: ±0.25% full scale, AC & DC
• Hysteresis: less than 0.1% full scale
• Repeatability: better than 0.1% full scale
• Stability: 1/5 div./8 hrs
• Typical coefficients — Temperature: 0.06%/°C over 20–40°C range; Line voltage: 0.02% per volt (105 V to 125 V)
• Zero suppression: continuously variable, 500 division range readable to 1 part in 1000; Stability: 0.01%; Accuracy: within 1% of full scale suppression

---

Squaring and Square-Root Operations with EAI Multipliers of the TR-48 and TR-20

In addition to performing multiplications and divisions, the parabola multipliers can also be used as function generators, specifically to compute the square and the square root of a quantity. For squaring, a parabolic characteristic is placed in the input circuit; for square-root extraction, it is placed in the feedback of a computing amplifier.

When performing these operations, the input quantity X may assume positive, negative, or both signs, and a particular sign may be required for the output quantity as well. Based on these criteria, the programming engineer must select the most favorable circuit from a range of possible configurations, with respect to the number of amplifiers and multipliers consumed.

The various circuit possibilities arise from the number of parabola segments in a multiplier (previously 4 half-parabolas, and more recently 2 half-parabolas in the “bi-polar” multiplier 7.417 or 7.137), and especially through the possible direct access to the parabolas from the programming patch panel. Many such circuits are given in the data sheets and handbooks (Operator’s Manuals, Reference Handbooks).

The following table summarizes, for a range of computing operations, the circuits of multipliers 7.117 (TR-48) and 7.137 (TR-20), 7.096, 7.099 (both TR-48), and 7.045 (TR-20). It indicates whether the circuit can be found in one of the handbooks or in the supplement to this report. A further note states how many times the computing operation can be executed simultaneously by one multiplier, and finally how many computing amplifiers are required for that purpose.

Further circuit possibilities found by readers are welcomed.

Summary Table (Supplement)

The supplement tabulates squaring and square-root circuit configurations for multipliers 7.117, 7.137, 7.096, 7.099, and 7.045, covering cases such as:

• x² for various sign ranges of x (e.g., 0 ≤ x ≤ +1; −1 ≤ x ≤ 0; both signs)
• √x and √|x| for various sign ranges, with or without absolute-value preprocessing
• −x² and variations
• Implementations requiring 1 or 2 amplifiers; implementations feasible or not feasible with a single multiplier

Reference codes A through M indicate supplemental circuit diagrams. The schematic diagrams in the supplement show patch-panel wiring for each case, including connections to summer/inverter amplifiers and the parabola input/output terminals.

---

DES-30 — A Step Toward the Hybrid Computer

The Digital Extension System DES-30 substantially extends the range of applications of small to medium-sized analog computers (TR-20, TR-48). It can also be used as an independent digital experimentation device or for the logical control of arbitrary processes. This report briefly describes the technical characteristics and some application possibilities of the Digital Extension System DES-30.

Technical Characteristics of the DES-30

1. Construction

The instrument consists of a completely wired chassis with appropriate power supplies and control facilities, and a complement of various digital plug-in units whose inputs and outputs are accessible at an exchangeable patch panel. Programming is therefore carried out exactly as with an analog computer — by interconnecting the digital elements with patch cables. The complement of the system can be varied within wide limits and thus adapted to specific requirements. The front panel is divided into four horizontally arranged fields for receiving the plug-in units.

The topmost field (Field 0) contains decade counters, monostable flip-flops, digital differentiators, microswitches, and indicator lamps. The field below it, Field 1, accommodates a maximum of 8 plug-in units with 6 AND gates each (a total of 48 AND gates). Field 2 provides space for 8 4-bit registers and 2 plug-in units for the input and output of signals to and from the analog computer. The bottom field, Field 3, contains the indicator lamps for displaying counter and register states.

2. Synchronization

All storage elements of the DES-30 are clocked (synchronous logic): counters, registers, monostable flip-flops, and even manually operated microswitches. These units change their output state only in coincidence with a clock pulse — specifically on its falling edge.

Available clock frequencies are: 1 MHz, 1 kHz, 1 Hz, and manual (for test purposes).

The clock thus provides a precise time base for all DES-30 operations, and further enables operating-mode control of the computing sequences. As with the analog computer (which has three operating modes: initial conditions, compute, and hold), the DES-30 also has three modes:

• Run — the clock reaches the digital elements and activates them.
• Stop — the clock is inhibited; the process halts.
• Clear — the clock is inhibited and all storage elements are reset to zero.

Selection of operating mode can be performed manually or by external signals.

3. Communication Between the Analog Computer and the DES-30

The analog computer generates logical signals (zero or one) via comparators (analog-to-digital switches), which reach the DES-30 through so-called Analog-Digital Trunks (interconnections). Since the voltage levels of the logic signals are not the same for both instruments, the trunks provide the necessary adaptation.

In the reverse direction, the digital signals of the DES-30 influence the processes on the analog computer in a variety of ways:

• By controlling the operating modes (initial condition, compute, hold, repetition) of all integrators;
• By controlling the operating modes of individual integrators or groups of integrators;
• By controlling the integrator time constants;
• By controlling electronic switches (Digital-to-Analog switches);
• By controlling analog storage elements (“Track–Store”).

For analog computers with electronically controlled integrators, a level adaptation of the logic signals within the integrators is sufficient, while for relay-controlled integrators a series of Digital-to-Analog Trunks with relay amplifiers is provided.

4. Logic Elements

a) Logic inverters are unnecessary, since almost all elements have complementary outputs.

b) AND gates have two inputs and complementary outputs. To combine more than two signals, the negated outputs of two or more gates are simply connected together. This allows all operations of switching algebra to be realized — including disjunctions, exclusive-OR, NOR, etc.

c) 4-bit registers can initially be used in three different ways: as 4 individual flip-flops, as a binary 4-bit shift register, or as a binary counter. Appropriate programming yields BCD counters, left-shift and right-shift registers, up-counters and down-counters, cascade connections for more than four bits, etc.

d) Presettable decade counters can be preset to a manually adjustable count between 00 and 99. When the input is high, 1 is subtracted from the counter content with each clock pulse. When the counter reaches 00, it emits a pulse of one clock period in length. This pulse can, among other things, reset the counter to the manually set value.

e) Monostable flip-flops — A one at the input sets the monostable flip-flop for a manually adjustable duration that can be selected between 10 μs and 10 s. After this time elapses, the output returns to zero on the next clock edge.

f) Differentiators — A transition from zero to one at the input causes a pulse of one clock period duration at the differentiator output.

With all these elements, quite extensive programs for binary logic or arithmetic can be constructed. For many subtasks, established subroutines exist. However, for more complex control tasks, various approaches always lead to the goal, and the programmer’s creativity is given wide scope.

Applications of the DES-30 in Combination with an Analog Computer

The enormous number of individual application possibilities can here only be grouped into a few characteristic categories.

1. Computing Runs at Different Speeds

This is made possible by individual control of the operating modes and time constants of different integrators. Although the analog computer actually has only one independent variable — computing time — multi-dimensional problems can be represented in repetitive operation through various nested repetition times. In the same time that suffices to traverse one dimension’s range, another dimension is traversed n times (diagram t₀, tₑ, tₐ; classic example: computation of multiple integrals).

The method of alternating computing runs can also be applied successfully when the results of one process represent the initial conditions of another and vice versa (in the diagram, t₀ and t₁). Fast-repeating subroutines are also used conveniently for function generation.

2. Boundary-Value Problems and Optimization Tasks

For boundary-value problems, the computer compares computed results with specified boundary values and derives modifications to the initial values or system parameters from the deviations, until all boundary conditions are met. The various iterative methods used differ with respect to convergence and computational effort.

For optimization tasks, the permissible parameter space can be systematically scanned, with the computer storing the best intermediate value found along with its associated parameters until a better value is found. More effective but also more demanding are gradient methods (“steepest ascent”). Here, the direction of progress in parameter space is determined for each point by a vector whose components arise from the partial derivatives of the function to be optimized with respect to the respective parameters.

3. Function Storage

Through appropriately controlled analog storage elements (“Track–Store”), entire curves can be held point by point. On output, a reasonably smooth function is obtained from the resulting step function through DES-controlled interpolation of first or higher order. This allows dead times to be simulated, which can also be varied. With certain limitations, partial differential equations can also be solved iteratively by this means (serial solution).

4. Applications for Statistical Studies

Particularly in combination with a noise generator for excitation of simulated systems, effective studies (e.g., correlation analyses) can be conducted using comparators and digital counters. Smaller hybrid systems are also well suited for the statistical recording of arbitrarily occurring measured values.

5. Simulation of Discontinuous Processes

In part, such problems can be solved with conventional analog computing components such as comparators and diodes. However, with more complex logical structures, only the digital extension of the analog computer will enable a solution (simulations in aeronautics and astronautics, non-continuous control systems: stepping controllers, sampling systems).

In summary, it can be stated that complete control over the important independent variable of the analog computer — namely, the computing time — is achieved only with the aid of a digital extension device. The benefit is already evident from tasks as simple as the evaluation of definite integrals.

---

Announcement: Brochure on Electronic Evaluation in Measurement Technology

On request, a copy of the newly completed brochure “Electronic Evaluation in Measurement Technology” (Elektronisches Auswerten in der Messtechnik), written by Dipl.-Ing. G. Buschmann, Aachen, will be sent. The brochure describes some widely proven electronic evaluation methods.

---

MC Laboratory Connectors with Snap-On Ferrules

Previously unattainable secure contact transition resistance — less than 0.1 mOhm

Simplest connection of stranded wire and plug with or without soldering

The MC snap-on ferrule provides the ideal plug-in isolation. Thread through — push on — snap closed!

Maximum versatility, minimum dimensions.

The MC laboratory connector is a springy tube made of hard gold-plated, hardened beryllium bronze — it is a hybrid: both plug and socket. It therefore has the highly desirable properties and the most versatile applicability:

1. The MC laboratory connector is a large-area contact with extremely low contact transition resistance (less than 0.1 mOhm), highest constancy; it is completely vibration-proof.
2. Disturbances due to thermal voltages are vanishingly small.
3. The MC laboratory connector is very space-saving. The connectors can be manipulated without difficulty in patch panels with 8 mm grid spacing, and any socket can be branched almost arbitrarily often.
4. The patch pattern of socket panels is preserved even when the panel is fully populated with MC connectors.
5. Two MC laboratory connectors can be plugged together front-to-front to extend leads.
6. For creating multi-way junctions, they are very easily pluggable in series.
7. The MC laboratory connector can be used in rigid and spring-type sockets.
8. It can also be plugged from front and back onto rigid and spring-type pins.
9. Thanks to the serrated slot, the MC laboratory connector can also be used as a clip-on probe.
10. The MC laboratory connector makes contact in sockets of 4–4.3 mm diameter and therefore bridges the European (4 mm) and American (4.3 mm) standards. It can even be pushed onto a 5 mm pin (telephone plug, Schuko plug).

The MC snap-on ferrule is a completely novel type of connector insulation, molded in one piece from the high-quality insulating material RILSAN. Pushed onto the MC connector and snapped closed, the MC ferrule holds firmly in place and enables — even without soldering (for provisional connections) — a reliable connection of stranded wire and plug.

MC laboratory cables made from highly flexible copper stranded wire (1 mm², 512 individual wires of 0.05 mm diameter), fitted with MC laboratory connectors and snap-on ferrules, represent the modern laboratory measuring leads for the highest demands. Available in black, white, red, blue, yellow, green, violet, and gray, and in 6 standard lengths.

Prices (at EAI Aachen):

Length	Unit price (DM)
25 cm	2.35
50 cm	2.55
75 cm	2.75
100 cm	3.00
150 cm	3.60
200 cm	4.10

Quantity discounts available on request when quantity is specified. Delivery time: generally 8 to 14 days after receipt of order.

---

Heinzinger, Munich — High-Stability Transistor Power Supplies in All-Silicon Technology with Automatic Current-Voltage Regulation

A transistorized power supply with 0–25 V and 0–500 A was built for a major industrial facility. Voltage and current are continuously adjustable via 10-turn potentiometers.

The control unit with indicating instruments, voltage and current settings, and mains switch can be positioned at any distance from the main unit for remote control purposes; both parts are then connected via a multi-wire cable. The unit also has sensing leads to compensate for external voltage drops in the supply leads to the load. A battery can be connected in parallel to the power supply for drawing larger currents. The resulting total current is also displayed and regulated by the power supply.

The instrument is primarily used to drive electric motors. The resulting inductive spikes and back-voltages from a parallel battery cannot damage the instrument.

Voltage and current regulation against load and mains variations is in the range of 10⁻⁴. The ripple voltage at the output is less than 10 nV rms under all conditions.

Instruments of this type are supplied up to approximately 10 kV or up to 1000 A or up to 70 kW. All instruments operate with air cooling.

Doubly Stabilized High-Voltage Power Supplies in All-Silicon Transistor Technology with Current and Voltage Regulation

Model	Voltage Range	Max. Current	Price (DM)
HN 500-500	0–500 V	max. 1 A	4,240
HN 1000-500	0–1000 V	max. 0.5 A	4,310
HN 2000-500	0–2000 V	max. 0.25 A	4,390
HN 5000-500	0–5000 V	max. 0.1 A	5,870
HN 500-1000	0–500 V	max. 2 A	6,750
HN 1000-1000	0–1000 V	max. 1 A	5,970
HN 2000-1000	0–2000 V	max. 0.5 A	7,540
HN 5000-1000	0–5000 V	max. 0.2 A	7,890
HN 1000-5000	0–1000 V	max. 5 A	9,880

---

Analog Computing Courses 1966

In the course of 1965, EAI GmbH organized a total of 7 analog computing courses. Four of these courses took place in Aachen and the remaining three were held for specific EAI customers in Homburg/Saar, Stuttgart, and Darmstadt. A total of 195 persons participated in these computing courses, of whom 10 were from abroad.

The strong participation in such courses has prompted planning of several analog computing courses for the current year as well. The first course will take place from March 14 through 18 (inclusive). Depending on the number of advance registrations received, further course dates will be set and announced in due time.

The program to be covered is outlined by the following points:

• Construction and basic operating principles of the analog computer
• Linear computing elements
• Programming of simple linear problems
• Input and output possibilities
• Nonlinear computing elements and special components
• Special computing circuits
• Introduction to scaling techniques
• Application of nonlinear computing elements in various examples such as function generation, solving differential equations with variable coefficients and nonlinear differential equations
• Control engineering problems
• Fundamentals and various applications of iterative analog computing techniques
• Fundamentals and applications of hybrid computing using the digital extension system, type DES-30

For non-customers of EAI, a course fee of DM 150 is charged. For EAI customers, participation is free of charge, though the maximum number of non-paying participants is reserved.

Anyone interested in an analog computing course is requested to send an advance registration soon. The number of advance registrations will be decisive for planning additional courses, since the number of participants per course should not exceed approximately 30. For assignment to the first course from March 14–18, the chronological order of received advance registrations will be considered.

Hotel reservations will be made on request.

---

High-Speed Printer, Type 6.610

For connection to the EAI digital voltmeters, types 6.000, 6.001, and 6.101, the high-speed printer type 6.610 has been available for some time. This fully transistorized unit with a maximum speed of 20 lines per second is currently one of the fastest devices of this type.

Specifications

Print format: 12 columns

• Column 12: Range — 1 Volt (1), 10 Volt (2), 100 Volt (3), 1000 Volt (4)
• Columns 7–11: Measurement — 4 digits plus 1 overflow digit
• Column 6: Function symbol

Function codes:

• −VDC
• +VDC
• Ohm O (for Digital Volt-Ohmmeter 6.101 only)
• Overload (for 6.001 and 6.101 only)
• Columns 1–5: additionally available (individually), can be used e.g. for addressing a measuring-point selector

Input logic and characters: 8, 4, 2, 1 code with the following available characters: 0, 1, 2, 3, 4, 7, 8, 9, +, −, ·, /, !

Voltage levels: Standard “0” = 0 V; “1” = 6 V

Speed: Maximum 20 DC measurements per second

Operating modes:

• High automatic speed: 20 lines/sec
• Low automatic speed: 2–3 lines/sec
• Triggering: Manual — 1 line per button press
• External: One command from outside initiates one measurement cycle

---

EAI Electronic Associates GmbH — 51 Aachen, Bergdriesch 37 — Tel. (0241) 44041