# What happens to second stage offset in a 2-stage amplifier

#### jaydnul

Joined Apr 2, 2015
175
If you had two separate 10x 1-stage amplifiers cascaded you would get 100x gain. The first amplifier offset would dominate because the second amplifier's offset would be divided by 10x.

However if I have a 100x 2-stage amplifier (both stages in the same feedback loop), what happens to the second stage offset? What is the advantage of having both stages inside the same feedback loop rather than having two separate 10x amplifiers?

Thanks

#### Ian0

Joined Aug 7, 2020
9,522
That’s why they are usually AC coupled.
Two amplifiers inside the same feedback loop is likely to cause instability.
The best solution is a single amplifier with a gain of 100.
If you put two amplifying elements inside the same feedback loop, then there is only one offset.

#### jaydnul

Joined Apr 2, 2015
175
Hmm, I mean what's the advantage of a standard 2-stage miller compensated amplifier with 100x gain vs. having 2 separate single stage 10x amplifiers with their own independent feedback loops cascaded one after another? What happens to the second stage offset when its inside the 2-stage amplifier?

#### Ian0

Joined Aug 7, 2020
9,522
Each Miller-compensated amplifier will have a phase-shift of 180°. Put two inside a feedback loop and you have an oscillator. If you can manage to keep it stable, the outer loop only will determine the offset.
This assumes that the inner amplifier has gain that is constant down to DC. If it has no DC gain, then the output of the first stage will be undefined, and may well be stuck at one of the supply rails.

With Two cascaded amplifiers, if they have DC gain, then the offset will be the offset of the first multiplied by the gain of the second, plus the offset of the second.
If both have AC gain only, the it is a perfectly acceptable solution, if a little noisier than a single x100 amplifier.

#### jaydnul

Joined Apr 2, 2015
175
Oh ok, thanks.

Think i came to my answer: the errors of the second stage are divided by the OPEN LOOP gain of the first stage, which is way higher than two separate amplifiers where the errors of the second amplifier are only divided by the CLOSED LOOP gain of the first amp

#### BobTPH

Joined Jun 5, 2013
8,677
Okay, you just changed the problem from offsets to errors signals. These are not the same thing.