Electronic – Operational Amplifier diagram

What exactly is an op-amp?

An opamp is a very simple device that does this: Vout = Gain*(Vin+ - Vin-). The Gain is fixed for each opamp and is usually 100,000 or higher. To make matters more interesting, this gain value is not very "tight". The gain of a real opamp might change over some large range depending on temperature, voltage, batch, frequency, etc. I would not be surprised to see this gain change from 100,000 to 1,000,000 in the same opamp.

Most practical opamp circuit require negative feedback to produce other gain values and to make the opamp behave more predictably.

What was the idea behind its creation?

I can't ask the guy who designed it, but simply put an opamp is one of the most flexible semiconductor devices available today. I am guessing that the original inventor knew this and that's why it was invented. So in a sense, this question is a bit like asking "why was the wheel invented". But this is a super generalization.

The op-amp itself is an integrated circuit, a chip. Before chips there were discrete transistors, and before that there were vacuum tubes. There are circuits that are functionally equivalent to opamps that were used back in the tube days. So the chips that we call "opamps" are "just" modern manifestations of those tube circuits.

What is an opamp composed of?

Modern opamps consist of transistors, diodes, resistors, and some other minor stuff. Of course they used to be made of tubes. The following image shows an internal schematic of one opamp:

Of course there are hundreds (or thousands?) of different opamps so there are many different internal schematics. Some of them are much simpler than this one.

I glossed over this before, so I'll restate it here: Most practical opamp circuits require negative feedback to make things work well. Without feedback, an opamp works nicely as a voltage comparator. For almost everything else, negative feedback is required. This feedback is the source of much magic, and the reason why opamps are so useful. Feedback is beyond the scope of this question/answer, but suffice it to say that this is where the important stuff happens.

Nothing wrong about the spec as far as I can see - it sounds like you want it couched in terms that are meaningful to your limited knowledge and application requirements.

• 1 Mohm minimum input impedance is fine - it's 1 Mohm guaranteed or greater - most folk want high input impedances so a guarantee is a good thing.
• Common mode input voltage range applies to differential amplifiers and it basically means, if you want to measure a small AC signal (say 100 mVp-p) it will do so even if both input wires are raised by a potentially interfering voltage of +/- 10V. Good spec I reckon.
• Common mode rejection - this tells you how good the input amplifier is when it is connected to a source voltage with different resistance connections. 70 dB at 60 kHz is pretty good for this sort of thing - you have got to envisage that some folk may not be able to adequately control the impedances of their signals that are to be measured by this device.
• Input imbalance adjustment - no amplifier is perfect and most will produce a small dc error at their output - this can be nulled-out by the adjuster.
• Linear output is telling you that it won't produce significant distortion (maybe due to clipping) for output peaks up to 10V. I also read this as meaning that the nominal rated full scale output level is +/-10V.
• Corner frequency is the point where the output starts to naturally fall when the input frequency gets above a certain value and for the default amp you are using this is 10kHz. It's adjustable.
• Cross talk of 80 dB RTI - referred to input is what RTI means. Cross talk applies to amplifiers with multiple channels and interference between one channel's signal and another channel.
• Gain resolution - if you want to adjust the gain there is a finite limit to how exact you may want it and this is the resolution.