The lead-resistance problem
A Pt100 element changes resistance by about 0.385 Ω per °C. A metre of 0.5 mm² copper lead has a resistance of about 35 mΩ per metre at room temperature, which doubles up in a two-lead loop. For every metre of cable, you have two metres of copper carrying the current to and from the element.
Translate that into temperature error: 70 mΩ per metre of cable / 0.385 Ω per °C = 0.18 °C of error per metre of round-trip cable, on top of the element’s own tolerance. A 10-metre run gives 1.8 °C of fixed offset before you have even started. That offset also drifts with the ambient temperature around the cable, which makes simple zero-point calibration ineffective.
2-wire — cheap, short-cable only
Two wires — one to each end of the element. Simplest possible RTD assembly, and the cheapest. The lead resistance lives in series with the element and is read by the transmitter as part of the measurement. Most modern transmitters allow you to enter a fixed line-resistance correction; that cancels the DC offset but leaves you exposed to ambient drift of the cable.
When 2-wire is actually fine:
- Cable run < 1 m
- Ambient temperature around the cable is stable
- The element is a Pt1000 rather than a Pt100 — ten times the resistance change per degree means lead error scales down by 10×
- Class B (±0.3 °C) accuracy is sufficient
HVAC duct sensors and small appliance probes routinely use 2-wire Pt1000 for exactly these reasons. Battery-management systems do too, because the harness is short and the BMS chip has built-in 4-wire excitation for the cells anyway.
3-wire — the industrial default
Three wires: two to one end of the element, one to the other. The classical implementation uses a Wheatstone bridge in the transmitter; one bridge arm sees the element-plus-one-lead, the opposite arm sees just-one-lead. Subtract one from the other and the lead resistance cancels out, provided the two leads have the same resistance — which they do, because they are run side by side in the same jacket.
3-wire is the dominant industrial topology because:
- It uses one extra wire compared to 2-wire (small cost increment)
- It cancels lead resistance to first order, including thermal drift
- It works with the standard 3-wire RTD transmitter ecosystem (every PLC analog input card supports it)
The residual error from mismatched lead resistance is typically < 0.1 °C, well below the Class B element tolerance. For 99 % of process-industry, HVAC and motor-monitoring applications, 3-wire is the right answer.
4-wire — the Kelvin connection
Four wires: two carry the excitation current, two read the voltage across the element. The voltage-sense pair carries no current (or microamps of input bias for a modern instrumentation amp), so its resistance does not contribute to the measurement. The lead resistance disappears completely.
4-wire is mandatory when:
- You need Class AA accuracy (±0.1 °C at 0 °C)
- The cable run is > 30 m and even matched-pair 3-wire residuals matter
- The cable connectors degrade over time (corrosion, vibration) and you cannot trust lead-pair matching
- Calibration laboratory traceability requires the element resistance to be measured directly
The trade-off is that 4-wire doubles the cable count and requires a 4-wire-capable transmitter or data acquisition channel. The cost increment is small in itself but real at scale — a 100-channel 4-wire system is meaningfully more expensive than a 100-channel 3-wire system.
What about Pt1000?
Pt1000 elements are interchangeable with Pt100 in every wiring topology. The big advantage is that the lead-resistance error scales by Relement/Rlead: a 10× larger element means 10× less impact from the same cable. A 2-wire Pt1000 over 5 m of standard cable has similar error to a 3-wire Pt100 over the same cable.
Pt1000 is therefore the natural choice when you want to keep wire count low without giving up accuracy. It has displaced Pt100 in many HVAC and BMS applications for exactly that reason. The catch is that the higher impedance is more susceptible to EMI pickup — in noisy industrial environments, shielded cable becomes more important.
Cable shielding and noise
RTDs are low-impedance sensors and inherently quieter than thermocouples. That said, in factory environments with VFDs, switching power supplies and motor cables in the same tray, shielded twisted-pair cable is still the right default. Ground the shield at the instrument end only to avoid ground loops.
For very long runs (> 100 m) the standard fix is to put a 4-20 mA transmitter at the sensor and run the current loop back to the control system, which is immune to lead resistance and EMI alike. The transmitter cost is justified by the cabling savings.
Quick selection table
| Application | Element | Wiring | Why |
|---|---|---|---|
| HVAC duct probe, < 5 m | Pt1000 Class B | 2-wire | Lead error negligible |
| HVAC duct probe, 10 - 30 m | Pt100 Class B | 3-wire | Standard combination |
| Motor winding embedded, 1 - 5 m | Pt100 Class A | 3-wire | Per-class spec inside motor |
| Bearing temperature, 1 - 5 m | Pt100 Class B | 3-wire | Standard motor convention |
| Calibration laboratory | Pt100 Class AA | 4-wire | Traceability requires direct measurement |
| Long process run > 100 m | Pt100 Class B | 4-20 mA transmitter at sensor | Eliminates lead and EMI |
What Jianlu supplies
All Jianlu Pt100 / Pt1000 probes are available in 2-wire, 3-wire and 4-wire configurations. The JSF-M222A fluoropolymer Pt100 series is built with shielded 4-wire option (designated “P3L”) for long cable runs and high-EMI motor environments. The WZP probe series is available with M8/M10 thread, ceramic or metal tip, and shielded cable.