[sdiy] fanning grounds vs. dirty/clean ground

[jhaible]

>It seems in general Opamps don't like capacitors on their outputs.
>On a resent project using a single supply voltage I had to put 100R on the
>output of the opamp buffering the reference voltage to prevent oscillation.
>I might miss something here, but it seems to me the same problem could
>happen with a opamp buffered "ground".


Actually, opamps don't like capacitance _inside a certain range_ on their
outputs: If the pole from the (internal, open loop) output resistance and
the load capacitor is in the same range as the pole from the internal
phase compensation, you have that self-oscillating filter again.

For fast applications (audio), you
* must make the capacitive load very small, or
* isolate it with a resistor (as you did), or
* use two separate feedback paths (needs some calculation).

For DC buffer applications, just make the cap *big*. So big, that
the resulting pole is outside the critical range _on the other side_.
Example: LM324 as unity gain buffer for DC - > add 10uF directly
from output to GND.

JH.