how REALLY a comparator work?

i know comparator is a open loop opamp crkt that is used to compare a time varying signal to a ref level(given at the other input).but how that really happen ?then what is the difference between comparator and difference amplifier. Pls help me in this.

A comparator is effectively a difference amplifier with infinite (device maximum) gain. When the + and - input vary by a small amount, the output signal swings to a large voltage in an effort to to correct for the difference in the input signals. (Op amps try to make the + and - input be at the same voltage by adjusting the output voltage). Since the gain of a comparator is huge, a small difference of the input voltages produces an output voltage that swings to the supply voltage rails. Since there is no feedback, the changing output voltage never affects the input voltages, so the output voltage just stays maxed out.

A difference amplifier is generally constructed such that you define and control the gain of the stage with feedback (usually a combination of the input and feedback resistor values).

A standard op amp can be wired up for both of the circuit types. By giving it no feedback path, it acts like a comparator... output swings to the rails with only a small delta in the input voltages. Using the correct feedback technique and some algebra to pick the resistor values, an op amp will function as a fixed-gain difference amplifier.

Check out TI App Note 31 for a bunch of op amp circuit implementations. Sheet 1 shows a diff-amp, and sheet 4 show a comparator circuit. Search for "op amp comparator" in your favorite search engine for other examples.

The comparators I know about are different from an op amp in that the comparator output can go all the way to zero V. This is important for digital operations.

An op amp output does not go below 2V if you are using a single supply. (This rule of thumb does not apply to certain op amps but it was a chief distinguishing characteristic for years.) Hence you could not use one in digital circuitry, or if you did you had to go to the trouble of incorporating a negative supply.

The comparator output is internally connected to the collector of an NPN transistor. When the transistor is on, it pulls output low (zero).

When the transistor is off, the output shows high impedance. It does not go high. This is why we typically install a pullup resistor, connected to V+.

Op amps do not need the pullup resistor.

------------

Edited to add:

A comparator can have a feedback loop same as an op amp, if you devise a way to accommodate the idiosyncracies of the comparator, to get similar performance as with the op amp.

While op-amps and comparators are "theoretically" the same (differential high impedance input, infinite gain, low impedance output), in practice they are very different. In fact, in practice different types of op-amps are also very different (depending on what they attempt to optimize), and different types of comparators are also very different.

To build a simple dichotomy, an op-amp is built for operation in feedback, while a comparator is built for open-loop comparison. If you tried to use an op-amp as a comparator, you would find that it works, although its comparison performance sucks when compared to a normal comparator. If you tried to use a comparator as an op-amp, chances are your circuit wouldn't work at all. For purposes of stability during feedback, most op-amps are intentionally made slow using an internal "compensation" capacitor. (If they were too fast, then they wouldn't be stable). Comparators don't have this concern, so they can be made faster and have higher gain. (Sidenote: The more gain you have, the more difficult it is to stabilize a feedback loop; in some cases it's impossible).

It's your safest bet to use the right tool for the job: if you need comparison, use a comparator, and if you need analog amplification, use an op-amp.