0
\$\begingroup\$

Alright, say I have this Inverting Opamp, but it so happens the positive input to ground is also connected with another sub-circuit that happens to produce a lot of noise (no pun intended):

schematic

simulate this circuit – Schematic created using CircuitLab

Now, say there's nothing I can do about the noise source, would it help if I were to put an LC filter on the positive input to ground. I would also like to use a common choke for the inductor, with the other side going at V_DD:

schematic

simulate this circuit

Or would this just be detrimental to the performance of the opamp?

Would there be a better way to filter/stabilize the ground leading to the opamp? There's also those Crystal, Ceramic, Saw filters that I know nothing about... Can you guys suggest the cheapest solution?

EDIT:

In response to Andy's comment, the V1 above has been added to denote the voltage source of the black box (It was not there in the first diagram). I'm not a 100% sure, but the the black box has a Boost Converter and the inductor might be the one causing the noise.

The Opamp by itself works well when powered by V_DD. The black box does so when powered with V1. The output of the opamp only appears after I ground it with the black box.

EDIT: It has been misunderstood by some that whole LC-common-choked filter are all at ground, so I'll clarify it:

schematic

simulate this circuit

schematic

simulate this circuit

but to think of it, I should've provided this diagram and not let the ground of the opamp "dangle" around.

\$\endgroup\$
1
  • \$\begingroup\$ You can get ground shift from your noise source, radiated noise and conducted noise needs layout details and CMRR of the OA can easily be degraded by these factors and unbalanced source impedance, wire impedance and R ratios. Your design goals and work is missing for CM gain, SNR etc. as well as neglecting the CM range of the OA. electronics.stackexchange.com/questions/79752/… \$\endgroup\$
    – Tony Stewart EE75
    Commented Feb 21, 2017 at 12:53

1 Answer 1

Reset to default
2
\$\begingroup\$

This is one of those cases where an exact circuit is needed. Showing the circuits like you have make no sense because the input source is not specified. It is also unclear what the power rails are. Attaching a block called "noise source" to 0V is also unhelpful and meaningless.

Also, 0V is 0V - that is the baseline - you cannot say 0V is noisy. However you can say the power rails are noisy relative to 0V.

If on the other hand your input comes from something that has a 0V rail that is wiggling around relative to your amplifier's 0V then a different solution is called for. Either transformer coupling or a differential amplifier is recommended in these cases.

The common-mode choke you have in your 2nd circuit does not achieve noise rejection as far as I can see.

\$\endgroup\$
7
  • \$\begingroup\$ Thanks for your help... Although, there's something that I don't understand... When you say 0V cannot be noisy, isn't that using the conventional current abstraction too much? In terms of actual electron flow, that's where the electron comes from. There should just be a periodic more positive voltage of a vast amount of charge to be filled to seriously divert the current away from the opamp. \$\endgroup\$
    – Dehbop
    Commented May 15, 2015 at 10:32
  • \$\begingroup\$ 0V is by definition 0V and is the reference point for all other nodes on the circuit. It is quiet by definition. However, if there are currents flowing through this node and there are small resistances and inductances associated with this "net" then clearly it can only be 0V at one single point - all other parts of that net will have a small noise voltage relative to it. \$\endgroup\$
    – Andy aka
    Commented May 15, 2015 at 11:43
  • \$\begingroup\$ Well the ground is designed and inherently DC. It's the boost that is inherently AC. So it holds, common ground can be noisy. Also, there are situation where other components are producing negative voltage that are then directly connected to common ground. No use defining that "there are no common grounds", just because it can be distorted. \$\endgroup\$
    – Dehbop
    Commented May 16, 2015 at 2:39
  • \$\begingroup\$ What you have drawn makes no difference because it cannot inject current. It has no return path to ground. It is current flowing from a noise source AND a non-zero ohm impedance ground rail that produces noise. \$\endgroup\$
    – Andy aka
    Commented May 16, 2015 at 8:54
  • \$\begingroup\$ You seem to misunderstand the diagram... The opamp is powered by V_DD. I had the power rails upside down than usually depicted, can't you see that?? Look at it again... (I'll also be redrawing it to make more obvious). The noisy ground rail of the opamp is top side and V_DD on bottom. The V_DD also is common-choked with said noisy ground as the steady DC V_DD will create a field, attenuating the noise at ground. \$\endgroup\$
    – Dehbop
    Commented May 16, 2015 at 13:52

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