How op-amp can be used in signal processing applications? Can anyone explain? I know that op-amp can be used for solving linear analog mathematical problems.
But how op-amp can be useful in signal processing application?
Op-amps are the standard building block of many analogue circuits. They can be used to make many circuits like:
- Inverting, non inverting, buffer and differential amplifier.
- Analogue adder and subtractor.
- Active filter: high-pass, low-pass, band pass.
- Differentiator and integrator.
- Current to voltage converter.
- Negative resistor
- Peak detector
- Analog multipliers (by using log and antilog amps)
- Gyrators (effectively an inductance made from opamps and capacitors and often a very large inductance value)
- Different types for different applications
(I'll leave this open for community so you can all add your favorite op-amp application)
Op Amps are used as amplifiers in signal processing applications. And as inductors, using a capacitor feedback circuit.
And particularly, as the output drivers of Switched Capacitor filter stages. https://en.wikipedia.org/wiki/Switched_capacitor
Analog electronics as an opposite to digital was originally signal processing concept. When digital computers became in 1950's popular their calculation power was insufficient for many urgent analysis problems. For ex. simulating power plant processes or dynamics of a ship etc...was simply out of the available capacity.
The development of the electronics had created device named operational amplifier (not as IC in 1950's, but discrete) which was useful for ANALOG COMPUTATION.
Analog means literal analogy. Engineers made a circuit which had formally the same differential equations than the actual interesting process. With such circuit it was well possible to test for ex how a ship behaves. Plotters plotted on paper curves how voltages evolved in the circuit and engineers saw as directly as possible how well a ship performed certain manoveur in given conditions. The only more exact way would have been to test with a ship, but it was nonexistent. The simulation was needed to design it properly.
Analog computation models used opamps in integrators, amplifiers, adders, multipliers, comparators, filters, signal generators and circuits with common non-linearities. Even today hobbyists can buy analog computer kits which have a bunch of these blocks in easy to use format "insert only the wires".
Analog computers were used widely in serious simulations as late as in 1970's.
How to use analog circuits to solve a problem?
Circuits do not invent anything. With a circuit one can simulate a phenomena if its math law is known. Many phenomenas are known as differential equations i.e. the phenomena is characterized by values of certain state variables and the math law states 1) the time derivatives of the state variables as functions of reached values and input variables 2) how the visible variables depend on state variables.
The state variables are stored to integrators and a network is built to feed the integrators with the calculated derivatives. The network gets integrator outputs and the process input signals as its inputs.
Then there's another network which calculates the interesting visible variables if the happen to be something which must be calculated from state variables, not the state variables itself.
The simulation needs initial starting voltages to the integrators and some signal generators to feed the process input variables. Plotters can record the outputs. Or they can be stored to some other storage media such as audio recorder.
Shouldn't differential equations need differentiators? (They were not mentioned)
In theory they can be used, but it's useless because the limited frequency and voltage ranges of the opamps would make the result unusable. Noise would easily use all of the available voltage range and it would still be hopelessly inaccurate. That's because the gain should grow linearly as the frequency increases. The math law must be transformed to such form that integrators can be used but no differentiators are needed.
So called RC differentiators are well usable, but they do not make derivatives, their gain do not increase over certain limit as the frequency grows.
If we consider creating interesting sounds for music and effect purposes a problem, that problem is still often solved by using analog signal processors and analog synthesizers. Modular analog synthesizers became popular in 1960's. They were analog computers for signal generation.
They are still available, but prices are high when compared to DSP based products. But many people still want them because the complex non-idealities of the analog processing make the last touch for the sound and that's non-existent in precise digital stuff they say.