Input Impedance OpAmp Circuit

message

Absolutely! Work experience is worth much more than book "larning" ;-) ...Jim Thompson

True but you can say the same thing about a "virtual ground". I think panfilero is actually right in that "virtual short" better describes the situation - it's just that nobody ever calls it that. The "virtual ground" term comes from the common case where the + input is in fact at "ground". The same term is then reused by analogy when it is at some other fixed potential.

Either term is misleading in it's own way. Since I'm not a big fan of the "virtual ground" model of op-amp operation* I don't have a great interest in figuring out if the confusion of changing the terminology on everyone is worth the change in perspective.

• It's much better to remember that the op-amp is in there, and that it's doing everything it can to keep V- equal to V+, as best as it can acting through the feedback impedance. Call it the "op-amp trying real hard" model. If you think of it that way, then not only does your mental model encompass the normal "virtual ground", it also encompasses the "virtual short". Further, it's easier to incorporate exceptions: as soon as you say "but the op-amp doesn't have high gain at high frequencies" it at least qualitatively encompasses the impedance and gain changes in the circuit as the frequency goes up; as soon as you say "the op-amp output can't exceed the rail" you can cope with that; and as soon as you say "the op-amp has delay (or phase shift)" you can understand the need for stability analysis.

My method for "solving" OpAmp circuits is assume V+ = V-, then the rest is Algebra. ...Jim Thompson

That's what I do if I know the circuit isn't hitting the rails, is well below the GBW product of the op-amp, etc.

If it's higher frequency I use the good ol' feedback equations -- usually I end up with a block diagram, because I'm more used to it.

If I know it'll be hitting rails or if there's diodes in it I'll either do it as piecewise linear or I'll use Spice -- then I'll back it up with some lab bench tests and maybe even some handcuff-and-lit-cigarette sessions with an apps engineer, because even when an op-amp hits a rail cleanly it doesn't necessarily come off it in a predictable way.