# opamp Vos bias resistor

In a simple inverting amplifier I've seen the below topology in my text book, where R3 is said to be R3=R1||R2.

However, every time I see this, all three resistors are the same values! From Forest Mims, to op-amp app-notes. When I run the simulations in LT-Spice the equal value version has the lowest Vos, followed by R3 = R1||R2 followed by direct connection to ground.

So my question(s) are:

1. Why use this topology instead of a simple follower (no resistors), when the op-amp is unity gain stable.
2. All the academic texts say to use the parallel equation for the value of R3, but I never see it in the wild. Why?

A simple inverting follower (Gain of "-1") needs two equal resistors in the feedback path.

It is the non-inverting follower (gain "+1") which needs no resistors for 100% feedback.

The resistor R3 is - in most cases - not needed. The only purpose of this resistor (R3) is "offset compensation". Both input nodes draw a small DC current and this small current will produce a small DC voltage (which then will act as an input offset voltage Vo.)

This unwanted input voltage Vo will cause an unwanted DC output. But note that the input offset is multiplied with the noise gain (1+R2/R1) only - and not with the much larger open-loop gain.

This effect can be - up to a certain degree - compensated if app. the same DC input voltage is created at the other (pos.) input. For this purpose, R3 is added with a value of app. R1||R2.

If this compensation is necessary or not depends on the specific application.

• Yep, I got the non-inverting one mixed up. Thanks! – Aaron Dec 19 '18 at 21:10
1. The inverting configuration can lead to lower distortion with many op-amps since the common mode voltage remains at 0V.

2. I suspect that you're not seeing the correct value for the bias current compensation resistor because you're looking at some bad schematics.