I want to make a capacitance measurement on a sensor pad connected to the AD7746 over a long trace (24"+). Essentially, the sensor plates are far from the chip, and cannot be moved closer due to novel geometry. This introduces significant parasitic capacitance due to proximity with other nearby traces running in parallel (can't be avoided, as they all go to a similarly located sensor bank). I had first tried to accomplish this measurement using the AD7747, as having one of my capacitor plates connected to ground was desirable because it was more mechanically simple. The AD7747 and the AD7147 accomplishes a nifty shielding trick called Active Shielding, in which the chip outputs an identical excitation signal voltage to that used for sensing. This active shield excitation is applied to traces in close proximity to the sensing trace, creating a zero potential difference between the two and thereby cancelling most of the parasitic capacitance effects for the length of the trace that is shielded.
My question is this:
Can I use one of the excitation lines on the AD7745/AD7746 to create an active shield, by applying EXCB to the nearby parallel traces? And does this actually buy me anything in terms of performance, noise reduction, etc.? Dedicated shielding-traces can easily be applied in parallel with the capacitance sensing trace, but I am unsure if it is even worth it, and if I need to be careful which side of the DUT I shield in this way. Experiments to follow, but looking for wisdom before I dive into this circuit.