What Are Analog Computers?

Analog computers are computing machines that solve mathematical problems by representing variables as continuously varying physical quantities, typically voltages or shaft rotations, rather than as discrete binary digits. Each variable in a problem corresponds to a measurable physical quantity in the machine, and the machine's internal dynamics evolve to produce a solution by direct physical analogy. This approach gave analog computers a natural advantage for problems governed by differential equations, where continuous processes in nature map directly onto continuous processes in the machine.

The earliest analog computers were mechanical, relying on rotating disks, gears, and linkages to perform arithmetic and integration. James Thomson described a machine for solving differential equations by integration in 1876, and Vannevar Bush at MIT produced the large-scale electronic Differential Analyzer in the 1930s. Electronic analog computers, using vacuum tubes and later transistors, displaced mechanical designs by the 1950s. As the Computer History Museum's account of analog computing notes, these machines excelled at modeling physical systems because they measured continuous values and could respond to parameter changes interactively, without restarting a computation from scratch.

Electronic Analog Computation

Electronic analog computers perform arithmetic using operational amplifier circuits. Summing amplifiers add weighted inputs, integrator circuits accumulate voltages over time proportional to an integral, and sign inverters change polarity. Multipliers, function generators, and coefficient potentiometers extend the repertoire to nonlinear operations. The output of each circuit block is a voltage that represents one variable in the problem being solved, and the circuit topology encodes the structure of the differential equation. Research by Yannis Tsividis and colleagues at Columbia University, described in IEEE Spectrum, demonstrated that modern CMOS fabrication technology can realize single-chip analog computers capable of solving equations from fluid dynamics and quantum chemistry with millisecond solution times and very low power consumption.

Programming and Patching

Configuring an analog computer required translating a mathematical problem into a wiring diagram called a patch program, then connecting circuit modules with cables on a patch panel. Setting initial conditions involved adjusting potentiometers to represent starting values, and scaling required normalizing all variables to fit within the machine's voltage range, typically plus or minus 10 or 100 volts. This programming process demanded careful attention to dynamic range and the stability of feedback loops. General-purpose analog computers from manufacturers such as Electronic Associates and Applied Dynamics International sold in large numbers to aerospace, nuclear, and biomedical research organizations during the 1950s and 1960s.

Summing Circuits and Coefficient Setting

Summing circuits in analog computers combine multiple voltage signals through precision resistors to produce a weighted linear combination at the output. The ratio of the feedback resistor to each input resistor sets the gain for that term. By cascading summing and integrating stages, an operator could implement the right-hand side of any linear ordinary differential equation as a single signal path. Nonlinear equations required additional hardware: diode function generators approximated arbitrary nonlinear relationships, and electronic multiplier modules handled products of two variables. These circuit-level operations are described in detail in IEEE textbook coverage of op-amp integrator and differentiator circuits.

Applications

Analog computers have applications in a range of fields, including:

• Aerospace engineering, for real-time simulation of missile trajectories and aircraft flight dynamics
• Nuclear reactor control, where differential equations govern neutron flux and heat transfer
• Biomedical research, modeling the electrical behavior of neurons and cardiac muscle
• Control systems development, representing plant dynamics during early-stage prototyping
• Educational demonstration of differential equations and feedback principles