Help with audio test equipment design

[jackocleebrown]

Hello,

I'm trying to design a high impedance differential input buffer. The input signal will be up to around 40V pk-pk and may have a relatively high common mode offset. Because the signal voltage is a high I need to attenuate the signal to avoid clipping the opamps.

One option which works reasonably well is a classic instrument amp topology with an attenuator at the input. To keep the input impedance high I have to use very large values in the attenuator. Additionally, If I follow the application notes and add resistors to ground/ref from the two opamp inputs (to add an input current bias return path) then this significantly reduces the input impedance and the CMRR.

As a second option I came up with is to inverting amplifier buffers in the instrument amplifier with gain 0.1 (see attachment) which avoids the requirement for the bias current resistors. On paper this design is working better because it has higher overall input impedance.

In both topologies I need large resistors on the input and I'm concerned about noise. Does anyone have experience of using inverting topologies with high input impedance? How much of a noise issue can I expect?

Thanks for you help!

Jack.

[godfrey]

IMHO, At the output, about 12uV rms (in a 20KHz bandwidth) from the resistors, plus a bit extra from the opamps.