0
\$\begingroup\$

I'm designing an input buffer stage for a low voltage, noise sensitive application. These buffer stages consist of dual package OpAmps, with each individual OpAmp buffering the + or - input to a differential amplifier.

I'm comparing similar OpAmps for their noise specs, mainly their input noise voltage (spec given in nV / sqrt(Hz)), which I understand to be related to the white noise spectral density, and their 0.1 - 10Hz input voltage noise (spec given in nVp-p), which I understand is mainly related to 1/f noise. My application is low frequency, ~1-1kHz and will be effected by both of these figures.

1/f Noise Noise Spectrum

My question is, is the noise from these sources common and directly related for the two OpAmps within the single package? In this case, will the noise then be removed by the differential measurement by the following amplifier stage?

\$\endgroup\$
1
  • \$\begingroup\$ LF ("flicker") noise, Johnson noise and shot noise will all be de-correlated, so adding 2 equal noise sources adds 3dB to the single opamp noise. Any other noise sources e.g. EM itnterference will be correlated, and cancel as you suggest. \$\endgroup\$
    – user16324
    Mar 24, 2020 at 20:56

2 Answers 2

2
\$\begingroup\$

No, the noise is not correlated between the two amplifiers.

There may be some tiny amount of noise that is coupled between the two of those, but in general it's so small that it's not even specified.

\$\endgroup\$
0
\$\begingroup\$

If there is a shared bias_generator, that noise will be correlated.

And power supply noise will be correlated.

But the noise in the input differential-pairs will not be correlated.

Have you thought of driving down the random noise floor, by performing an auto-zero operation, maybe 100 or 1,000 times a second?

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.