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?
  • \$\begingroup\$ 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). \$\endgroup\$ – user_1818839 Feb 28 '20 at 14:18

Should I short the scope inputs

Yes but make sure you are still grounded to where you want to measure the noise from your amplifier - there may be some common-mode artefacts that bump up the noise even when the scope inputs are shorted to each other and, if the scope inputs were not grounded (but floating) you might not capture them (if any).

Does such a method neglect common mode related noise

If you keep the ground, grounded it should be more accurate. The formula is correct by the way and take several measurements to ensure agreement.

  • \$\begingroup\$ "Yes but make sure you are still grounded to where you want to measure the noise from your amplifier" I couldn't picture what is meant here? \$\endgroup\$ – pnatk Feb 28 '20 at 14:25
  • \$\begingroup\$ Short the scope input to the scope ground but also keep that ground connected to where you will make a ground on your amplifier when taking measurements. \$\endgroup\$ – Andy aka Feb 28 '20 at 14:27
  • \$\begingroup\$ One last question when you say "If you keep the ground, grounded", do you mean scope ground to be grounded to the mains earth ground? \$\endgroup\$ – pnatk Feb 28 '20 at 14:43
  • \$\begingroup\$ No, I mean keep the scope input and scope Input earth wire grounded to where you would also have the ground connection when measuring your amplifier output. In other words, don’t just connect scope input and scope input earth wire together and leave those unconnected to anything. Both scope input connections should be all connected together at the ground point on the amplifier when making this reference measurement. \$\endgroup\$ – Andy aka Feb 28 '20 at 14:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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