Pt100 2-, 3- and 4-wire connections: why longer cables can read high

A temperature channel starts reading high after a probe cable is extended. The process has not changed, the value is steady, and replacing the Pt100 produces almost the same result. Before applying a software offset, look at what the input is actually measuring: in a two-wire circuit, the cable is part of the measured resistance.

Two-wire inputs include both lead resistances. Three-wire inputs compensate for lead resistance using a measurement method that depends on matching leads. Four-wire inputs use separate current and voltage connections to sense the element itself. The choice is therefore about where the measurement begins and ends, and whether the installed cable can meet the required error budget.

Why a longer cable looks like a hotter Pt100

A Pt100 is nominally 100 Ω at 0°C. An RTD input excites the resistance and determines its value from the resulting electrical measurement, then converts resistance to temperature using the selected sensor characteristic. With two conductors, the input cannot distinguish the element from the resistances in the outgoing and returning leads.

For an uncompensated two-wire connection: Rmeasured = RPt100 + Rlead 1 + Rlead 2. Added cable, a resistive splice or a poor contact can therefore raise the apparent resistance. Because platinum resistance rises with temperature over this example’s range, that additional resistance appears as a positive temperature error.

The first diagram separates the element from the two lead resistances. Its values are an illustrative calculation, not a test of a particular sensor or PLC module.

This is why a higher-accuracy sensor element alone does not fix a cable-induced error. The element may meet its tolerance while the measurement still includes several ohms of wiring. WIKA’s explanation of two-, three- and four-wire Pt100 connections describes this distinction and the matching condition for three-wire compensation.

A worked example: 0°C becomes about 10.2°C

Assume a nominal IEC 60751 Pt100 element at exactly 0°C, two leads of 2.00 Ω each at the time of measurement, and a correctly selected Pt100 input characteristic with no lead compensation. Ignore element tolerance and input error so that the cable contribution can be seen on its own.

For temperatures at or above 0°C, use R(T) = 100 × [1 + A × T + B × T²], with A = 3.9083 × 10⁻³ °C⁻¹ and B = −5.775 × 10⁻⁷ °C⁻². Solving for the near-zero positive root at 104 Ω gives approximately 10.25°C. The equation and coefficients are given in section 1.1 of TI’s A Basic Guide to RTD Measurements.

The useful diagnosis is the size of the error: four extra ohms can create roughly ten degrees of false temperature near 0°C. The example does not assign a cable length to 2 Ω. That would require the actual conductor material, cross-section, route length and temperature, plus any contact resistance.

A fixed correction can remove a known lead contribution at one condition. It cannot automatically follow later changes in cable or contact resistance. If the cable warms while the sensing element stays at the same temperature, a one-time correction may no longer describe the installed circuit.

What the third wire compensates—and what it assumes

A three-wire sensor brings two conductors to one end of the element and one conductor to the other. The input uses its circuit topology and additional measurements or excitation paths to cancel the lead contribution. That cancellation depends on the relevant lead resistances matching; an extra conductor by itself does not guarantee an accurate result.

Use the same conductor cross-section and route for the measurement leads, and pay attention to extensions and terminations. Replacing only one conductor with a different cable, adding a resistive splice to one path or exposing one section to a different thermal environment can undermine the matching assumption. Sensor wiring colours help identification, but the sensor’s actual connection diagram establishes which leads share an element end.

TI shows several three-wire measurement circuits in its guide. Their excitation and measurement arrangements differ, so a single residual-error equation should not be assumed for every PLC input. For selection, the practical question is whether the module’s specified compensation and permitted lead resistance cover the installed cable and required uncertainty.

Why four-wire sensing is different

Four-wire measurement gives each end of the Pt100 two connections. One pair carries excitation current. The other pair senses the voltage directly at the element ends. Because the sensing input draws very little current, the sense leads contribute very little voltage drop; voltage lost in the current-carrying leads sits outside the sensed element voltage.

The connection point matters. Joining the sense leads to the current leads in the cabinet still leaves the field cable inside the measured voltage path. Separate terminals at the input cannot remove resistance that lies between their joining point and the element.

Four-wire sensing addresses lead-resistance error. Element tolerance, input accuracy, self-heating and the way the probe exchanges heat with the process still contribute to the result. Select four-wire when controlling the lead contribution is important enough to justify the extra conductor and compatible input, rather than treating it as a promise of zero system error.

Match the probe, cable and PLC input as one measurement

Start with the allowed temperature error and the actual wiring route. Then compare connection options against that requirement. There is no universal cable length at which every Pt100 must change from two to three or four wires.

A concrete example is Siemens ET 200SP AI 4xRTD/TC 2-/3-/4-wire HF, order number 6ES7134-6JD00-0CA1. Its April 2026 equipment manual lists Pt100 RTD modes for two-, three- and four-wire connections. Section 4.1 describes a fixed resistance correction for the entire two-wire measuring loop and automatic three-wire compensation, with the same routing and cross-section for all three measurement cables. This is a specific module capability, not a rule for every ET 200SP analog input.

That manual also separates measurement type, measuring range and temperature coefficient in the channel configuration. A physically connected Pt100 still needs the matching RTD mode and characteristic. See the module’s wiring and parameter sections when checking a replacement. For the physical terminal carrier and potential-group context, the ET 200SP BaseUnit selection guide explains why the electronics module and BaseUnit must be considered separately.

If a transmitter is mounted near the probe and sends 4–20 mA to the cabinet, distinguish the short Pt100-to-transmitter wiring from the current loop. Apply the RTD lead analysis to the transmitter’s sensor terminals; assess the cabinet current input and loop separately. A two-wire 4–20 mA transmitter describes its power-and-signal loop, not necessarily a two-wire Pt100 sensor connection.

Find the first point where the temperature becomes wrong

For a channel that changed after cable work, compare the measurement in stages before changing the PLC scaling. Plan checks in a safe maintenance state: sensor disconnection can trigger wire-break behaviour or affect control, and resistance checks require an isolated circuit. Use the input’s service instructions and the plant’s isolation procedure.

A stable high value after a cable extension is consistent with extra series resistance in an uncompensated two-wire circuit; it is not proof on its own. If the input behaves correctly with a local simulator and the error appears only through the field wiring, the next useful action is to examine that wiring and compensation boundary. Adding an offset before locating that boundary can hide the fault without correcting the measurement.