# Op-amp stable voltage gain out of different input value

Can anyone tell me how to make a circuit with unstable input voltage signal value to a stable output value For example:

• A sinewave in the input between 1 Vrms to 5 Vrms
• The output must be a stable 10 Vrms

• Just build a sine wave oscillator outputting 10 V rms!
– user173271
Sep 17, 2022 at 12:33
• Hi James. I don't need to build a sinewave, I want to deal with an existing one with unstable amplitude. Sep 17, 2022 at 13:05

What you're looking for is an Automatic Gain Control (AGC) system. And it is not so simple to design one with such little information. I'd not suggest any implementation with such little details.

You have to start by defining the range of gains your amplifier will be handling.

Based on your schematic, sounds like your max. gain will 20dB, and your minimum will be 6dB.

It's easier to work with dBs, so your range, apparently, will be from 0dBV till 13.98dBV (or 14dBV, for practical matters). That means your amplifier needs to handle a 14dB input dynamic range.

After that, you have to define your gain steps. How fine will your gain grid be? 1dB, 2dB? You have to know what's the largest step your system can allow.

E.g. if you defined your gain step as 2dB, then, for instance, your signal might be wiggling between 0 and 2dBV and suddenly goes >2dBV, then you might wait a little (hysteresis) before switching from 20dB to 18dB. How to detect those? You need amplitude detectors that need to compared against some level reference. This can then trigger a gain switch from 20 to 18dB.

At my job, we're currently working on one. We have been designing it and improving it for the last 5 years. But mostly due to requirements being super tough on noise, linearity, etc. which you haven't mentioned in your example.

Sounds like you want an AGC (automatic gain control). There are many possible circuits.

One approach (simple, but relatively expensive, and only applicable over some range of frequencies) would be to measure the incoming RMS voltage and 4-quadrant divide the AC voltage with that voltage, scaled appropriately. So if the input voltage is 1V RMS you divide by 0.1, if it is 5 you divide by 0.5. (technically you only need 2-quadrant division since RMS voltage will always be >= 0).

Of course any similar method requires some time for the measurement to settle so if the amplitude is changing you’ll get additional errors and distortion.

There are much simpler AGC circuits possible- the details depend on the exact requirements.