Zero Impedance Output

[ioplex]

Can someone explain the "zero impedance output" circuit in Small Signal Audio Design by Douglas Self page 426?

More specifically, I would like to model this circuit to understand why it supposed to be "not affected by load capacitance" as described on page 382.

Is it not possible to model this performance improvement in LTSpice?

 

[Andy Peters]

Can someone explain the "zero impedance output" circuit in Small Signal Audio Design by Douglas Self page 426?

More specifically, I would like to model this circuit to understand why it supposed to be "not affected by load capacitance" as described on page 382.

Is it not possible to model this performance improvement in LTSpice?

The 68-ohm build-out resistor is inside the op-amp's feedback loop, so its effect on the output impedance is removed by the op-amp action.

 

[joaquins]

A capacitive load also will try to take down some highs (RC with 68) but as seen by feedback op amp take highs up again making it flat and no affecting op amp stability by direct capacitive loading.

JS

 

[JohnRoberts]

It IS affected by load capacitance, but the HF and LF feedback components are separated by a build out resistor so load capacitance doesn't phase shift the negative feedback enough at the opamp that it oscillates. (Opamps become unstable when negative feedback shifts 180' becoming positive feedback). 

A capacitive load will cause a one pole roll-off after the 68 ohm, but since the LF negative feedback is grabbed after that rolloff, the opamp output will actually boost to neutralize that loss. The 47 pf provides very HF feedback so the opamp remains stable. However this output boost to compensate for the roll-off comes out of signal headroom, so capacitive loading will reduce signal swing when the opamp clips sooner. 

This is generally a technique to prevent opamps from being unstable when exposed to capacitance on outputs. The frequency response is flatter, until it isn't. With some combinations of components and loading I've seen HF bumps, before final dip.

JR

 

[ioplex]

I was discussing this casually with an EE and we arrived at another somewhat unexpected conclusion which is that the 68 ohm resistor actually limits the output current so as not to "saturate" the op amp internally. Meaning if you're driving a circuit with a lot of capacitance like a really long cable or you're directly driving an inductor circuit, you don't want to draw so much current that the op amp cannot operate appropriately. So the "Zero Imedance Ouptut" is somewhat of a misnomer in that the 68 ohm actually limits the output impedance. Without it you might draw so much current that internally the op amp is in a state that it cannot recover from quickly. So depending on the load, this "current limiting" resistor could be somewhat important and I'm surprised we don't see this used more often. After thinking about it in this way, I found 3 places in my immediate project where I decided to add the current limiting resistor. One was driving a filter that had a very low imput impedance at certain frequencies. Another was driving an inductor. And the third was the line out.