# How can I extract the rms noise from the measurement's over all noise?

To clarify my question, here is an example: Let's say I have an opamp and I want to measure and quantify its output noise when its input is grounded. So after shorting the opamp's input to its ground, I couple its output to an oscilloscope and read the rms voltage. Call it $$\V_\mathrm{RMS_{all}}\$$.

Now some part of the the noise I read at scope screen is related to inherent noise from the scope itself. Call that $$\V_\mathrm{RMS_{scope}}\$$.

1. What is the practical way to extract only the opmap output noise ($$\V_\mathrm{RMS_{signal}}\$$) in this case? Should I short the scope inputs(and call the measurement Vrms_scope) and subtract it from the $$\V_\mathrm{RMS_{signal}}\$$? Does the following formula make sense: $$V_\mathrm{RMS_{signal}}^2 = V_\mathrm{RMS_{all}}^2 - V_\mathrm{RMS_{scope}}^2$$
2. Does such a method neglect common mode related noise?
• Add a 20dB gain stage before the scope. Measure its own noise with shorted input before adding the DUT. If that's not clean enough, repeat (40 or 60dB). – Brian Drummond Feb 28 at 14:18