Analog Computers
And what is that?
Well, it is the old method of making mathematical calculations by using analog signals rather than the usual digital systems of today.
How?
This is the building blocks that makes up an analog computing device.
A computational variable can be represented by an analog voltage or an angle performed by a servo.
A logical variable can be represented with a relay: On or Off.
In order to perform addition/subtraction of variables we are using operational amplifiers (op-amp´s) according to the following:
An op-amp has both a positive and a negative input and very high gain. The inputs has "virtually" ground potential,
that means that the inputs tends to keep their potential at zero voltage.
Due to very high input impedance there are no current in nor out through the inputs
(at least not compared to the current going through the feeding network components!).
By using different input and feedback resistors the resulting current through the feedback resistor
can be determined and thereby the output voltage.
More than one input can be added and/or subtracted based on input used.
By use of a servo a turning angle can be mechanised. Most often this angle is proportional to the input signal, but can also have non-linear functions built-in in the signal path..
The driving circuitry of the servo is very often carried out by a
magnetic amplifier.
That means a small amount of current can
change the magnetic field in a transformer core and thereby vary its magnetic gain.
Thus output a high current suitable for driving a servo motor.
If a servo now drives a potentiometer, the input analog signal to the potentiometer will be multiplied with the function that the servo angle represents. If the potentiometer is non-linear the multiplication will be carried out with that non-linear function of the servo angle...
A servo can carry out a division when driven by the difference between the divisor and the product of the dividend and the quotient. The equation a/b = c also means that the relation a = b*c is valid. Thus the servo has to find the position making a equal to the product b*c.
A servo can also be used to express an integration. The example with division is basically an integration of the error signal until it become zero. An integration just means there are no feedback, thus the servo will drift in direct proportion to both sign and amplitude of the input. And the position of the servo represents the integral of the input signal.
That´s the basics. Next is to find out those relations that shall be realized. All variables must be "scaled". Since this is done by using op-amp´s there will be a lot of them. And every multiplication need its own servo angle that drives the potentiometer realizing the function of that servo angle you will multiply with.
A potentiometer has an output voltage as a function of the angle. By using "taps" with external resistors put on different angles of the potentiometer the voltage ratio curve can be changed from strict linear to a more fancy curve. Thats the method to change a potentiometer to represent a desired function.
One use of non-linear curves is to make it represent a sinus curve for one revolution. If the servo represents actual heading the potentiometer then will deliver the Sin function of actual heading. Easy done - that´s it!
A feedback over an op-amp can be non-linear through diodes and extra resistors.
Sure - once in the early days maybe this happened you say.
The Swedish Air Force bought a number of analog simulators for the fighter Draken J35 around 1960. One of these has been sold to Austria after significant modifications. And was in service until 2005! Still with complete analog computational part for both aerodynamic and propulsion! A basic design that thus has been in service for over 40 years!
Find that PC or other "modern" computer that can compete!