I have a reservoir of conductive liquid (~2000 μS). Inside that liquid are two sensors: a pH sensor and a conductivity cell.
The conductivity cell is a two-terminal device and submerged. When excitation voltage is not applied the sensor electrodes are tied to PCB ground with N-channel FETs.
The pH meter is a two-terminal device which is also submerged. One terminal comprises of a reference cell which is referenced to a ground potential.
The impedance between the measurement cell and the conductive liquid is on the order of 10 kΩ (through a ceramic liquid junction). The other terminal, the measurement cell, has a high impedance (~250 MΩ) due to the high insulation resistance of the pH glass.
A small voltage between ±500 mV is developed between the reference and measurement cells which linearly depends on pH.
The pH meter wires are coax, with the shield of the coax connected to ground and the reference cell. The center conductor is connected to the measurement cell and in turn connected to an electrometer grade (high impedance, low bias current) buffer. The conductivity cell has regular unshielded conductors.
The water is slightly conductive and there are two ground connections in the water (reference cell with an impedance of ~10 kΩ and the conductivity cell itself).
Questions:
Should I be concerned about this ground loop being influenced by nearby magnetic fields?
Can electrons migrate through the ceramic liquid junction?
How does the (relatively high) impedance of the ground loop change things?
Instead of tying the reference electrode to ground, would driving the electrode with a virtual ground (op-amp buffer) break the loop?
