Four-wire Kelvin measurement is the preferred method for checking a Pt100 coil because current passes through one pair of leads while voltage is sensed across a separate pair, so lead resistance does not enter the resistance result [S1][S7].
A Pt100 element nominally measures 100.00 Ω at 0 °C, rises to about 138.4–138.5 Ω at 100 °C, and requires a meter capable of measuring approximately 20–400 Ω across a broad PT100 temperature range [S1][S3][S5].
What Four-Wire Measurement Actually Removes
A resistance meter normally forces a small current through the unknown resistance and measures the resulting voltage, using R = U/I. At 1 mA through 100 Ω, the expected voltage drop is 0.100 V [S4]. The four-wire arrangement separates that current path from the voltage-sense path, which is why the measurement is also called the Kelvin method [S7].
With a two-wire connection, the meter sees the sum of sensor resistance and both conductor resistances. The usual installation restriction is about 1–2 Ω per conductor, typically corresponding to roughly 300 ft of cable [S2]. A three-wire arrangement can cancel lead resistance when the two current-carrying branch resistances are equal, but that assumption is less robust than physically separating current and voltage connections [S2][S4].
The practical distinction is not the number of conductors alone. It is whether the voltage measurement has negligible current flow through its leads. A dedicated four-wire micro-ohmmeter can force current through the source pair and detect voltage with a high-impedance sense pair, producing a resistance reading without the first-order cable-drop contribution.
Pt100 Values and Expected Readings
PT100 identifies a platinum resistance thermometer with a nominal resistance of 100 Ω at 0 °C. Its resistance is approximately 138.50 Ω at 100 °C under the IEC 60751 relationship, giving a temperature coefficient of 0.00385 Ω/Ω/°C [S5]. The commonly cited local sensitivity near the upper part of that interval is approximately 0.384 Ω per 1 °C [S3].
The PT100 element is only one part of the check. A four-wire meter should first verify that the measurement terminals are open, then connect all four leads and confirm that the cable and terminal resistance has not displaced the result. IEC 60751 defines the Pt100 resistance-temperature relationship and probe tolerance classes, while IEC 60751-based calculation uses the ITS-90 polynomial equations [S5]. The stricter tolerance question is covered in IEC 60751 vs ASTM E1137.
For troubleshooting, compare the measured coil resistance with the expected value at the known reference temperature. At 0 °C, a nominal element should indicate 100.00 Ω; at 100 °C, the expected nominal value is about 138.5 Ω [S5][S8]. A stable offset across multiple readings points toward the meter, connection, or calibration, while resistance that changes as the probe warms can indicate self-heating or an unstable junction.
Comparing Two-, Three-, and Four-Wire Methods
Two-wire sensing is simple and inexpensive, but it requires a separate determination of lead resistance and cannot continuously correct the measured value [S2]. Three-wire sensing is common in industrial installations and can cancel equal lead resistances; stated guidance allows up to about 10 Ω for this configuration, with cable runs around 1,500 ft or more where necessary [S2].
Four-wire sensing provides the strongest lead-resistance rejection and can accommodate up to about 15 Ω of cable resistance in the cited arrangement, supporting runs near 3,000 ft [S2]. Its practical limit is the instrument and connection quality, not an intention to measure cable as part of the sensor. For bench verification of an RTD Pt100, a short, intact extension cable and stable terminal connections preserve the benefit of the topology.
The selection comparison is straightforward. Two-wire wiring wins on simplicity but has no active lead compensation. Three-wire wiring balances cost with partial compensation and assumes matched branch resistances. Four-wire wiring adds two conductors but gives the best accuracy for resistance measurement, especially where cable resistance is significant relative to the 100 Ω nominal element [S1][S2].
Instrument Setup and Self-Heating Control
The meter must support a four-terminal resistance measurement and match the intended resistance range. Typical resistance meters apply approximately 1 mA, while higher resistance ranges often use smaller currents; temperature transmitters may operate around 0.2 mA, with practical implementations spanning roughly 0.1 mA to several mA [S4].
Excitation creates self-heating in the platinum element, especially when the element is small or poorly coupled to its surroundings. The result is a resistance value higher than the undisturbed sensor temperature. To control this error, use the lowest test current that provides stable resolution, allow the reading to settle, and compare the measurement under the same current and connection conditions as the process instrument [S4].
For an embedded transmitter or data-acquisition design, ratiometric excitation can simplify conversion because the reference and sensor are measured relative to the same excitation. Texas Instruments describes a 1 mA ratiometric circuit with a 1.62 V reference and an RTD signal of approximately 20–400 mV for its -200 °C to 850 °C example circuit [S1]. The precision reference resistor matters in that design because a 0.01% reference contributes a gain error comparable to the ADC error [S1].
Where Four-Wire Measurement Fits
Four-wire measurement is appropriate for laboratory checks, calibration benches, commissioning, and low-error process instrumentation. It is not the preferred field arrangement when only two sensor wires can be installed or when transmitter terminals provide a two-wire connection. In those cases, a two-wire transmitter or a three-wire RTD input may be the only practical interface, although the insulation tester is a separate instrument from the RTD input and should not be used as the resistance measurement substitute. [S2]
Three-wire systems remain widespread in terminal heads with 4–20 mA transmitters and in duplex sensor arrangements [S2]. Four-wire circuits are more natural for high-accuracy measurement, analog input modules, laboratory instrumentation, and factory automation [S1]. For installations needing periodic verification without removing the sensor, an RTD transmitter can convert the element's resistance to a standard process signal, but loop current does not replace a true four-wire coil-resistance check.
Do not use an insulation resistance tester as a substitute for the micro-ohmmeter test voltage can stress or damage electronic RTD circuitry, and a high-voltage insulation result does not establish the 100 Ω coil value. The four-wire ohmmeter should be connected at the intended measurement plane, with the sensor de-energized or isolated from any transmitter loop.
Checks That Make the Reading Trustworthy
Begin with a lead check on the meter: short the Kelvin terminals and confirm a near-zero reading, then open the terminals and confirm the display does not retain a false low value. Connect the two excitation leads to one side of the PT100 element and the two sense leads to the other side. Keep terminal torque and contact resistance consistent, because poor contacts can masquerade as sensor instability [S4][S7].
For a standalone PT100 element, measure the coil and compare it with the IEC 60751 resistance-temperature relationship. At 0 °C, the nominal resistance is 100.00 Ω, and at 100 °C it is approximately 138.50 Ω [S5][S8]. If the coil is installed in a sheath, verify that the process temperature has stabilized before using the displayed value as a calibration check.
Record the meter model, test current, resistance, ambient or process temperature, and cable configuration. Repeat the measurement at least twice without changing the excitation setup; a changing result indicates thermal drift, contact variation, or self-heating. A stable value that disagrees with the expected Pt100 curve should then be checked against the sensor tolerance class and the reference temperature used for comparison.
The most useful next record is a four-wire coil value paired with a verified temperature: example, 100.00 Ω measured at a stabilized 0 °C reference. If the meter displays the nominal curve value with negligible effect from changing the four-terminal cable length, the circuit has demonstrated that lead resistance is excluded from the measurement [S1][S7].