High impedance divider measurement with instrumentation amp

Question

I'm having trouble with getting an accurate measurement with this circuit and I was hoping for some more help.

See the circuit, which I have breadboarded and am testing. I step down a voltage through a 1 to 100 voltage divider. The intent is that I can connect to different parts of the divider, and get a different step down ratio.

If I leave the divider and the unity gain buffer out of the circuit, I get excellent performance from the AD620 in-amp. I can input a voltage between -1 and 1V, and get effectively exactly the same voltage out, which is what I want. My input point would be right before the two 1k resistors.

When I introduce the divider, the voltage I get at the output is no longer accurate. I thought it was because the impedance of the divider was too high for the 1nA input bias current of the AD620, so I added the buffer that you see on the positive input. But this didn't help things! I still don't have an accurate voltage at the output. The buffer is an OPA376AID.

Is there something wrong with my approach to stepping down the input voltage with the divider? I know this worked on another design with a single ended input. Is it because the positive and negative side of the instrumentation amp are being loaded differently?

Thanks

Answer 1

You have three DC voltage points: plus 5V supply, its counterpart which is drawn as ground symbol and the signal GND in the middle of +5V and its ground.How many volts is the voltage between Vinminus and the signal GND. If you cannot guarantee it's less than 2,5V you are out of the proper operating range of your components.

BTW. Your divider seems to be 1000:1 and you wrote 100:1