We designed a microchip to sense high pressures under water. The chip senses pressure through a thin diaphragm that deflects and a corresponding piezoresistor circuit that measures the deflection. For every pressure, the chip spits out a corresponding voltage output.
The chip contacts are soldered directly to a FR-4 PCB and then the whole thing is potted in urethane.
The chip works great outside of water, and underwater, but the transition causes problems. Say the chip is placed in a water tank and pressurized to 500 psi for a few hours. It measures 500 psi no problem. But when we depressurize the tank and remove the chip, the chip outputs 20 psi, or some other random number, when it should measure 0 psi.
We think that the FR-4 material is swelling from the exposure to high pressure water. This swelling would stress the pressure sensing diaphragm and cause the erroneous reading.
What are good PCB materials that swell less when exposed to high pressure (~500 psi) water?
Or, any other hypotheses as to what is happening?