Instrumentation and Process Measurement
What the exam asks
Expect loop signal conventions, transmitter selection, sensor physics, calibration concepts and P&ID literacy. Many items are conversion questions - going from a milliamp reading back to an engineering value, including square-root extraction for differential-pressure flow.
Core rules
- Current loop. 4–20 mA with a live zero; 0 mA means a broken loop, not zero process. A 250 Ω resistor converts the loop to 1–5 V. HART overlays digital data on the analog signal without disturbing it.
- Wiring. Two-wire (loop-powered) instruments take power from the loop; four-wire instruments have a separate supply. Intrinsic-safety work requires checking entity parameters against the barrier or isolator.
- Temperature. Pt100 is 100 Ω at 0 °C with roughly 0.385 Ω/°C, accurate and stable to about 600 °C. Thermocouples produce millivolts and need correct extension wire and cold-junction compensation.
- Flow. Orifice and venturi DP meters give Q ∝ √ΔP, typically 3:1 to 5:1 turndown. Magnetic meters need conductive fluid; Coriolis measures mass directly; ultrasonic clamp-on needs clean, full pipe.
- Level. DP with a sealed leg needs wet/dry-leg compensation; radar is unaffected by most vapours; ultrasonic struggles with foam and stratification.
- Terminology. Accuracy is closeness to true value; repeatability is spread over repeated readings; a span error is a gain error, a zero error is an offset.
Worked example
A DP transmitter is ranged 0–100 inH2O for 0–500 gpm. The loop reads 8 mA. What is the flow?
Signal fraction = (8 − 4) ÷ 16 = 0.25, so the transmitter is at 25% of its DP span (25 inH2O). Flow fraction = √0.25 = 0.50. Flow = 0.50 × 500 = 250 gpm. Reasoning: the milliamp signal is linear with DP, not with flow. Always strip out the 4 mA live zero before square-rooting.
Traps that cost marks
- Square-rooting the raw milliamp value instead of the DP fraction.
- Grounding a cable shield at both ends, creating a ground loop.
- Using copper wire on thermocouple extension runs, or omitting cold-junction compensation.
- Calibrating at 0% and 100% only and missing mid-scale linearity errors.