Inverting Amp with Gain <1 versus Buffered Voltage Divider?

Thread Starter


Joined Feb 28, 2018
Circuit A is a buffered voltage divider. Circuit B is an inverting amplifier. Both circuits have the same less than unity gain magnitude = 0.2.

I don't care about the polarity inversion. Is there any inherent advantage or disadvantage to either? Stability or noise perhaps?



Joined Jan 3, 2011
It depends what aspects of the signal are most important to you, and how it would be used. The non-inverting configuration has common-mode voltage equal to the voltage at the input pin. This can have higher distortion than the inverting amp, which has essentially 0V common mode. Although you’re attenuating it, so it may be less of an issue. Plus, “more distortion” is relative to what is probably a very low distortion figure anyway, but in case you’re worried about 0.001% distortion, there you have it.


Joined Feb 19, 2010
I think the buffer circuit will have less influence on the circuit down stream.

Both circuits use three parts.

I would go with buffer. But it might use more physical space.


Joined Mar 14, 2008
The inverting amp will have a 20% lower bandwidth, and 20% higher DC offset voltage and noise.

An inverting amp has a bandwidth equal to the gain-bandwidth divided by (1+gain) with the noise and offset multiplied by (1+gain).
Last edited:


Joined Mar 14, 2008
Its important to check the OPA d/s stability data, when setting Gains < 1.
The worst-case for closed-loop stability is a follower with G=1, since it has the maximum feedback.
An inverter with a gain of less than one, has a feedback factor of less than one (the same as a non-inverting amp with a gain of slightly over one) so will be stable if the G=1 configuration is stable.


Joined Mar 14, 2008
In my experience OPA's can be unstable with gains less the one, depending upon the type of load being driven
If you had instability problems it's likely from reasons other than an inverting gain of less than one.
Certainly a capacitive load can cause instability problems no matter what the closed-loop gain.