Application Notes

2-Wire, 3-Wire or 4-Wire RTD? What the Wiring Does to Your Accuracy

2026-10-17

Why a 2-wire Pt100 can read several degrees high, how 3-wire compensation works, and when 4-wire is worth it — with the simple math (about 2.6 °C error per ohm) and practical fixes including head-mounted transmitters.

Two identical Pt100 probes can disagree by several degrees in the same process — and the cause is often not the sensor at all, but how it is wired. An RTD is measured by its resistance, and the copper leads between the element and your instrument have resistance of their own. The wiring configuration decides how much of that lead resistance ends up inside your reading.

The one number to remember

A Pt100 changes resistance by about 0.385 Ω per °C. So any unwanted resistance in the measuring loop converts directly into temperature error:

1 Ω of uncompensated lead resistance ≈ 2.6 °C of error on a Pt100.

Copper cable adds up quickly. Depending on the conductor cross-section, a lead can contribute roughly 0.1–0.4 Ω per metre of loop. A 2-wire probe on a 20 m cable run can easily carry 2–4 Ω of lead resistance — a permanent offset of 5–10 °C that no calibration of the instrument alone will remove, because the lead resistance also changes as the cable warms up (copper's resistance rises about 0.4 %/°C).

2-wire: simple, and honest about its limits

Both leads sit in series with the element, and the instrument sees element + leads as one resistance. Nothing is compensated.

Acceptable when: the probe is close to the instrument (short, fixed cable), the required accuracy is loose (several °C), or the sensor feeds a nearby transmitter. Many compact and OEM probes are 2-wire for exactly these reasons — short internal wiring inside a machine, straight into a local input.

3-wire: the industrial standard

A third lead lets the instrument measure the resistance of one lead and subtract it, assuming both current-carrying leads have the same resistance (same cable, same length). In bridge and transmitter inputs this cancels the bulk of the lead error; what remains is only the difference between the leads, usually a small fraction of the total.

Use it for: most process installations, PLC/DCS inputs and head-mounted transmitters. It is the default we recommend whenever the instrument supports it — the accuracy cost versus 4-wire is small, and it tolerates realistic cable runs.

4-wire: lead resistance eliminated

Two leads carry the excitation current; two separate leads sense the voltage directly across the element (Kelvin measurement). Because virtually no current flows in the sense leads, their resistance drops out of the result entirely.

Use it for: laboratory and calibration work, custody/quality measurements, Class AA elements, and Pt1000 sensors in precision loops. Note the flip side: a Pt1000 changes 3.85 Ω/°C, so 1 Ω of lead resistance costs only ≈0.26 °C — one reason Pt1000 is popular in long-cable, 2-wire OEM designs.

Matching the sensor to the input

Never land a 3-wire sensor on a 2-wire input (or vice versa) and expect compensation — the configuration is a pair: sensor wiring and instrument input must be set to the same scheme. Many modern transmitters are selectable; check the setting before blaming the probe.
If the sensor already exists and the cable run is long, the practical fix is a head-mounted temperature transmitter at the probe: the RTD wiring becomes a few centimetres, and the long run carries 4–20 mA (or RS485/HART), which is immune to lead resistance and far more noise-resistant.
Keep lead lengths and conductor gauges identical on multi-lead RTDs; compensation in a 3-wire circuit assumes symmetry.

Quick selection guide

Configuration	Lead error	Typical use
2-wire	Full lead resistance in the reading (≈2.6 °C/Ω on Pt100)	Short runs, OEM internals, loose accuracy
3-wire	Compensated, residual = lead mismatch	Industrial standard: process, PLC, transmitters
4-wire	Eliminated	Laboratory, calibration, Class AA, precision loops

Universal Ruike builds Pt100/RTD probes in 2-, 3- and 4-wire versions with the cable length and gauge specified to your installation, and can fit a head-mounted transmitter (4–20 mA / RS485 / HART) so the probe arrives PLC-ready. If a measurement point reads consistently high, tell us the cable length and wiring — we can usually tell you within one email whether it is the sensor or the leads.