1. Theory of Operation: The Seebeck Effect
A thermocouple is not a resistor. It is an active generator of voltage. The Seebeck Effect describes how a voltage (Electromotive Force or EMF) is generated when there is a temperature gradient along a conductive wire.
Crucially, the voltage is generated along the length of the wires where the temperature gradient exists, not just at the junction. The junction connects the two dissimilar metals to complete the circuit. The net voltage measured at the open ends is the difference between the Seebeck voltages of the two different alloys.
Where $S_A$ and $S_B$ are the Seebeck coefficients of the two wire materials.
2. Cold Junction Compensation (CJC)
A thermocouple measures the difference in temperature between the Hot Junction (Process) and the Cold Junction (Instrument Terminals). It does NOT measure absolute temperature directly.
The Law of Intermediate Temperatures: To find the process temperature, the instrument must know the temperature of the terminal block where the TC wires land. This is the "Reference Junction". The instrument measures this using a precision thermistor or RTD, calculates the equivalent millivoltage for that ambient temperature, and adds it to the measured thermocouple millivoltage.
Example: If a Type K TC is in 100°C water, and the meter is in a 25°C room, the TC generates approx 3.0 mV (equivalent to 75°C difference). The meter measures 25°C CJC (approx 1.0 mV). Total = 4.0 mV, which corresponds to 100°C in the NIST tables.
3. Wire Grades: Extension (X) vs. Compensating (C)
Running high-purity thermocouple wire thousands of feet to a control room is prohibitively expensive. Industry uses specific cable grades to save cost.
- Extension Grade (e.g., KX, JX, EX): Uses the exact same alloys as the thermocouple but with wider purity tolerances. It is usually limited to 200°C (392°F) due to insulation limits. The "X" denotes Extension.
- Compensating Cable (e.g., KCB, RCA, SCA): Uses completely different alloys (usually copper/copper-nickel) that mimic the EMF curve of the expensive thermocouple (like Platinum Type R/S) over a limited ambient range (0-100°C). This saves massive costs on noble metal installations.
Warning: Never use standard copper wire to extend a thermocouple signal. This creates new "parasitic junctions" at the connection point, causing massive errors equal to the temperature difference between the connection and the instrument.
4. Detailed Type Analysis
Type K (Chromel / Alumel)
- Range: -200°C to 1250°C.
- Pros: Most common, cheap, good linearity, decent oxidation resistance.
- Cons (Green Rot): Between 800°C and 1050°C in low-oxygen (reducing) atmospheres, the Chromium in the positive leg oxidizes preferentially, turning the wire green and causing a massive drift (reading low).
- Magnetism: The negative leg is magnetic. Handy for ID.
Type J (Iron / Constantan)
- Range: 0°C to 750°C.
- Pros: Very high sensitivity (~50 µV/°C). Good for reducing atmospheres.
- Cons: The iron leg rusts. Not recommended below 0°C due to embrittlement.
- Magnetism: The positive (Iron) leg is strongly magnetic.
Type N (Nicrosil / Nisil)
- Range: -270°C to 1300°C.
- Pros: Designed by NASA/compatriots to fix Type K's flaws. Silicon is added to create a protective oxide layer. Much more stable at high temps. Immune to Green Rot.
- Cons: Slightly more expensive and less available than K.
5. Troubleshooting: Ground Loops & Noise
Thermocouple signals are small DC voltages (often < 20mV). They are easily corrupted by AC noise from VFDs, motors, or ground loops.
- Ground Loops: Occur when the sensor is grounded at the pipe (Grounded Junction) AND the shield is grounded at the panel. Current flows through the shield. Solution: Ground the shield at ONE END ONLY (typically the instrument side). Use Ungrounded Junction probes if noise persists.
- Burnout: If a wire breaks, the circuit opens. Most transmitters inject a tiny "Burnout Current" (nano-amps) to drive the reading to full scale (Upscale Burnout) or zero (Downscale Burnout) to alert the operator.