Is using a buffer before a non-inverting input redundant?

Question

I came across a circuit in someone's paper and noticed he buffers the input signal and routes it to the non-inverting input of the opamp:

enter image description here

I thought this is redundant since non-inverting input has same input impedance as a buffer. Is this a good practice or such extra buffer is redundant?

• Jonathan S.: What's the input voltage range, and what's the part number of U2? Where does the signal come from?
• LvW: No - it is not the same because of different signal feedback (wich always enlarges the input impedance). However, under practical aspects this difference will be negligible and will be overshadowed by other (parasitic) effects .
• citizen: @floppy380 assuming A & B are equivalent parts in a dual part device, maybe the author wanted to make use of the spare device as an additional buffer to provide extra impedance buffering to U2B.
• floppy380: @JonathanS. Input voltage range is between 0 to 100mV. The part is OPA344.
• floppy380: @citizen "provide extra impedance buffering to U2B." But as LuW mentions isnt the effect of extra buffer negligible?

Answer 1

This is redundant, but there may be valid reasons to do so.

For example, best practice is to not leave unused op amps unwired, as if they start oscillating in open loop, that noise on the board might do nasty things. To avoid placing them in saturation, one might configure them as a buffer about halfway between the rails. Just adding a buffer here might save placing a few resistors.

Whether adding the buffer causes unacceptable distortion would depend on what this is for.

Answer 2

Given that the input doesn't have to withstand any overvoltage conditions with its 0..100mV voltage range, this is just a simple way to make use of an unused OpAmp section in a dual OpAmp IC (as you shouldn't leave unused OpAmp input pins floating).