Elliott Analogue Computer

From Object Wiki

Editing Elliott Analogue Computer

Manufacturer	Marconi-Elliott Avionic Systems Ltd
Production years	1957
Production location	(unknown)

Elliott Analogue Computer, the General-Purpose Analogue Computer (G-PAC), 1957 The G-PAC was first exhibited at the Farnborough Air Show in 1955. It was commissioned by the Ministry of Supply who funded Elliott Brothers to make a small and inexpensive analogue computer for the commercial market. These computers were used in the design of military aircraft and guided weapons.

1 How it works
2 Images
3 Memories
4 In the Science Museum
5 Related Objects

[edit] How it works

The purpose of analogue computers was to simulate (or model) real life dynamic systems such as aircrafts in flight, or missile systems. Such dynamic systems are vibrational in nature and can be represented mathematically by a several (simultaneous) differential equations. A simple example of such a dynamic system would be mass, suspended at the end of a spring. This would vibrate (or oscillate) at the rate dependent on the size of the mass and the strength of the spring. This system can easily be described by a single, second order, differential equation. To solve simultaneous differential equations (with respect to time) requires a process of integration and summation (subtracting as well as adding). In the case of the mass spring system, it would require two ‘integrators’ (because it is second order) and a ‘summer’.

The elements in an analogue computer that carry out integration and summation are called ‘operational amplifiers’. The Elliott GPAC (General Purpose Analogue Computer) featured in the photograph has 20 such operational amplifiers, all of which can be used as a ‘summer’ or an ‘integrator’.

The physical units of a system being modelled are represented by voltages and in the case of the Elliott analogue computer the voltage range is -100 volts to +100 volts and so for example, the extension or contraction of the spring could be represented by a voltage, i.e. the voltage would be an analogue (be analogous to) of the spring extension or contraction. E.g. a spring extension of 10 centimetres could be represented by +100 volts and a contraction of 5.5 centimetres could be represented by -55 volts. The results, or outputs, of an analogue problem set up (or ‘patched up’) on a computer would typically be displayed on a pen recorder. In the case of the mass spring system, this would appear as a sine-wave, gradually diminishing in amplitude as the oscillations are damped down. With a more complex problem such as an aircraft in flight, one would expect to see several waveforms representing pitch rate, pitch angle, roll rate, roll angle, speed etc.

The operational amplifiers (op-amps) in the analogue computer shown, are valve driven, each op-amp having about 10 valves and being slightly bigger than the size and shape of a carton of 200 cigarettes. In the present day, several operational amplifiers can be fitted into the smallest of integrated circuits! Although transistors were around at the same time as valve driven analogue computers, it was impossible to build a transistor op-amp of sufficient performance because of poor gain characteristics of the transistor compared to that of the valve. It wasn't until the early 1960's that first discrete (eg GAP/R's model P45) and then monolithic (eg 702/709) transistor op-amps became competitive.

The coloured panel in the picture, about half way down on the left-hand side is called the ‘patch panel’. This is where wires would link (‘patch') one op-amp to the next. The ‘patch panel’ can be removed by the operation of a lever (seen at the top of the ‘patch panel’) so that another simulation problem, which has been ‘patched up’ somewhere away from the machine, can be applied to the machine, without causing hold-ups. The rest of the computer’s front panel consists mainly of knobs that are connected to potentiometers that enables ‘coefficients’ to be set, for the particular problem being worked on.

This particular analogue computer was dedicated to the TSR2 (Tactical Strike and Reconnaissance aircraft) avionics project, in the mid 1960s, at Elliott Brothers, Rochester, Kent. After this project was cancelled, the computer was mainly idle, fulfilling odd jobs that may have happened to come along. Its last job, in 1971, was on modelling the hydraulic servo systems that drove the radar reflector on the Rapier missile system.

Within the defence, avionics and aviation industry during the 1960s to early 1970s, analogue computers were a more common sight than digital computers as transistor driven digital computers were just coming in, but were extremely expensive: a million pounds in the 1960s would have been a typical price! But, as digital computers became more plentiful and faster and since it was a relatively easy task to do ‘integration’ on a digital computer using a ‘step-by-step’ technique, more simulation problems were applied to digital computers. Eventually, analogue computers in around the 1980s, became confined to history.

John LePine