How does an operational amplifier (op-amp) work?

222 views

I have a pretty solid grasp of electronic theory but for some reason op amps are confusing the ever living *frick* out of me.

Edit:. Thanks to those of you trying to dumb it down for me! I have an electronics tech assessment with Garmin on Thursday, and though I’ve worked on electronics for 14 years, including 9 in the military, those 9 years were in the Marines. I understand things better when they’re sketched with crayons 🙂

In: Engineering

do you mean internally or externally ?

Externally: the simest model is, the output voltage Uout = G* (Uplus – Uminus), where G is really large.

Since G is very large, may even depend on temperature and such, it is not so useful on its own. The Opamp becomes very useful if you clamp the output back into the input, forcing additional relations between Uout and one (or both) of the inputs.

As used in a typical (negative feedback, amplifying) configuration: Imagine you are driving an incredibly powerful car. And when you step on the gas, the car throws you back so hard your foot withdraws from the gas pedal. And when it decelerates, your foot goes back on the gas. Now, by careful adjustment of the seat and seat belt, like careful selection of the components surrounding the op-amp in the circuit, you obtain a stable situation where your tiny pressure on the pedal translates to a tremendous pressure from the engine on the car.

An op-amp has two inputs (let’s call them A and B) and one output.

Very simply put, it compares the two inputs. If the voltage of input A is higher than B, the output is going to be high (close to its positive supply voltage), and if the voltage of B is higher than A, the output is going to be low (close to its negative supply voltage).

In practice the output voltage is going to depend on how large the difference between A and B is, since the gain of the op-amp isn’t infinite. So you may need a difference of say 1mV to get an output swing of 1V.

That’s the most basic function of an op-amp, but you can use it to build more complex circuits that do more than just provide a high or low signal at the output. For example, you can couple the output back to one of the inputs using resistors to achieve a specific gain, say if you want to amplify to low voltage signal by a factor of 100. In reality there’s probably a thousand different things you can do with an op-amp.

You have an inverting input (-) and a non inverting input (+). If the inputs are equal, the output stays the same. If + is higher than -, the output goes up. If – is higher than +, the output goes down.

A big thing to understand is that the output depends on the internal state of the opamp, not directly on the inputs. If the inputs are the same, the output can be high, low, anywhere. What you usually do is use some negative feedback so that you get the desired output for a given input.

For example, if you want to multiply the input by 5, you connect the input to +, and divide the output by 5 before feeding it back into the – input.

Internally, there’s a bunch of transistors that are arranged in a complicated way to give it properties that approximate the ideal op-amp. I don’t know about the internals.

I think the main point of the op-amp is the extremely high gain (amplifies the difference in voltage at input), which I think is useless without also using negative feedback.

The feedback forces the voltage at one of the inputs to be the same as the other, otherwise, the outputs are one of the rails (even an input difference of 0.01v will cause it to reach a 12v rail with typical op-amps)

Using the fact that the op-amp forces the input voltages to match allows us to pick clever circuits to get cool behaviour like:

– amplifying a signal
– signal buffering
– filtering
– oscillation
– transfer functions synthesis
—> brings control theory into reality
– more stuff

I think of it as a general purpose component. If you need a transfer function / amplifier of a particular gain, you can build it with whatever op-amp and some resistors and capacitors

Internally, there’s a bunch of transistors that are arranged in a complicated way to give it properties that approximate the ideal op-amp. I don’t know about the internals.

I think the main point of the op-amp is the extremely high gain (amplifies the difference in voltage at input), which I think is useless without also using negative feedback.

The feedback forces the voltage at one of the inputs to be the same as the other, otherwise, the outputs are one of the rails (even an input difference of 0.01v will cause it to reach a 12v rail with typical op-amps)

Using the fact that the op-amp forces the input voltages to match allows us to pick clever circuits to get cool behaviour like:

– amplifying a signal
– signal buffering
– filtering
– oscillation
– transfer functions synthesis
—> brings control theory into reality
– more stuff

I think of it as a general purpose component. If you need a transfer function / amplifier of a particular gain, you can build it with whatever op-amp and some resistors and capacitors.