AC Coupled In-Amp Gain vs DC Correction

[Riley-Pantheon]

I've been having poor CMR performance with an instrumentation amplifier circuit I have (mostly due to component mismatch of a high-pass filter on the input side), and thus have been exploring alternatives. The circuit I have amplifies a small voltage change from a Wheatstone bridge, where I need to eliminate the DC offset.

I'm considering using an integrator circuit to AC couple the instrumentation amplifier's reference voltage but all the documentation I have found (mainly from TI https://www.ti.com/lit/an/sboa242/sboa242.pdf?ts=1719948276394&ref_url=https%253A%252F%252Fwww.google.com%252F) shows an effective DC Correction Range versus Gain that I do not fully understand.



My amplifier currently has a Gain of around 750, but could be less. It will likely need to remain in the 100's though. Based on the TI docs though, it looks as though the "DC Correction Range" is only ±0.1 V to ±0.01 V for Gain in the 100's.

I need to remove a DC offset of around 1V-3V.

Can someone help explain to me what this "DC Correction Range" actually means?

[xvr]

Can someone help explain to me what this "DC Correction Range" actually means?

It mean that you can't compensate for DC of 1-3V with Gain around 750.

Input Instrumentation Amp should be capable handle DC offset with its desired Gain. With Gane of 750 and DC of 1V you should provide 750V on Ref input to compensate for DC 

[Riley-Pantheon]

It mean that you can't compensate for DC of 1-3V with Gain around 750.

Well that's a hinderance. Not sure what I was expecting. Seems pretty obvious now especially after looking at TI's reference design (https://www.ti.com/tool/TIPD191).

Thanks though.

[magic]

I'm not sure what the problem is supposed to be?

In chips like INA828, variable gain occurs in the first stage, which is a pair of opamps that apply differential gain to the signal while roughly preserving its common mode level (it may be slightly shifted due to matched but non-zero offset voltages of these opamps). Then there is a second stage which is a difference amplifier that works at a fixed closed loop gain, usually unity.

The external integrator applies feedback to the second stage, and the second stage shouldn't care whether its 1V differential input signal comes from a 1V input passed at unity gain by the frontend or from a 1mV input amplified 60dB

[xvr]

> I'm not sure what the problem is supposed to be?

Problem in DC application point. It applied as differential, not as common mode voltage. So first stage of INA128 have to amplify it up to 750V

[magic]

Right, the in-amp will clip with large differential DC at high gain

In audio preamps built with similar chips a capacitor is sometimes placed in series with the external gain resistor to block amplification of DC by the 1st stage. This capacitor needs to be fairly large, but its tolerance only affects filter cut-off frequency and not CMRR. Its ESR may affect gain, however.