# Fully differential op amp offset nulling

1. ## Fully differential op amp offset nulling

Hi, I am looking for a circuit that allows for offset nulling at the inputs to a fully differential op amp without affecting the gain. Anyone have any ideas?
Thanks!

•

2. ## Re: Fully differential op amp offset nulling

Means you have the amplifier as component with four external resistors defining the gain? You might use current sources to inject an offset current to the inputs, e.g. a differential pair.

•

3. ## Re: Fully differential op amp offset nulling

Originally Posted by FvM
Means you have the amplifier as component with four external resistors defining the gain? You might use current sources to inject an offset current to the inputs, e.g. a differential pair.
I have tried connecting a pot to the negative input and in simulation this affects the gain.

•

4. ## Re: Fully differential op amp offset nulling

To measure current through a wire I used a 324 quad op amp because spec said its output goes down to 0V. Input up to tens of mV. Gain=200.

The op amp had an unwanted offset. Zero input did not produce zero output. To reduce the offset, I had to contrive a way to nudge the inverting output downward a little.

So I made a resistor divider from 3 low ohm resistors. I arranged it to apply a tiny voltage between 0 and supply+. It was a big headache to find the right combination. I suspect that it hindered accuracy and/or linearity. I'm not sure the 324 was the best choice for the task. And I wanted a single-ended supply.

Most likely a proper solution is to install a negative supply. Raise or lower its voltage in a manner that affects the output up or down. The voltage can be adjusted directly, or with resistors, or a diode/zener/led. Or even by putting a pot across a diode, to dial a fixed voltage drop. Etc.

5. ## Re: Fully differential op amp offset nulling

I have tried connecting a pot to the negative input and in simulation this affects the gain.
Yes it does. Even more problematic might be reduced CMMR in case of asymmetrical resistor circuits. A possible option is fixed offset compensation resistor pair, considered in the gain calculation, drive by an adjustable voltage source (e.g. a pot with resistance low compared to compensation resistors).

A current source or current source pair is the other option.

--[[ ]]--