Sensor Technologies: Physics, Fabrication and Circuit

A: Option A looks attractive because MEMS parts are everywhere and cheap, but long cable runs plus ratiometric voltage output invite noise and offset drift unless you add local conditioning. Option C feels conservative and familiar, yet foil gauges creep over years and the bridge still needs stable excitation and trimming. Option D promises stellar stability, but now you've added a frequency‑to‑voltage stage in the control room that isn't on the GA and won't survive the schedule. The ceramic capacitive unit with current loop output shrugs off cable length, bakes in temperature correction, and stays within the drift budget without extra boxes.

A: A sounds logical if you've only seen self‑heating errors, but more airflow should pull the reading down, not push it up. B mixes up lead compensation basics—2‑wire makes resistance error worse, not better. D is tempting in an RF facility, yet EMI shows up as noise or jumps, not a steady offset tied to airflow. The fan upgrade changed how heat is stripped from the chassis versus the sensor pocket, setting up a gradient that biases the RTD unless the tip actually sees the same air.

A: A does happen on enclosures, but it doesn't explain the sensing element drifting while the box looks fine. C shows up as intermittent opens, not a slow RH bias. D needs strong UV exposure and years, not months. Salt ions and moisture soak into the polymer layer, altering its dielectric constant and hysteresis until the calibration walks off.

A: A shifts paperwork, not failure detection. B helps availability but doesn't reveal a stuck or biased sensor. D is seductive because MTBF numbers look scientific, yet they don't create diagnostics. Self‑test that injects a known stimulus or checks internal paths actually uncovers dangerous failures, which is what the safety case is after.