We are looking at a strain gauge. I assume it is a wheatstone bridge inside of the gauge but the electrical schematic is not available. We are using the LMP90099 and I have 1Meg pull ups on one input and a 1Meg pull down on the other input for open circuit diagnostics. Then we have 1K series resistance on each input to protect against high voltages.
If we put a 10uF cap common mode cap between the two inputs, effectively across the LMP90099’s input pins, we get the noise cancelation we want, but there seems to be a very slow drift added, which eventually stops after 1minute or so.
Can you explain this? Have we added some weird crosstalk between sampled and unsampled inputs?
Should we just put caps to ground and not couple inputs together?
Is there some setting in the chip that makes this work better?
Thanks for the feedback,
Jason