Overview
Connecting a 4-wire PT100 RTD to an analog input module designed for 3-wire termination is a routine field issue. The Omron CPM1A-TS101-DA provides one Pt100/JPt100 input with three terminals labelled PT1A, PT1B, and a second PT1B. When the field cable is a 4-wire PT100 (Kelvin construction), one lead becomes redundant. The correct installation is to leave that fourth conductor physically disconnected and insulated. This article documents the underlying lead-resistance compensation math, the step-by-step wiring procedure, CX-Programmer configuration, a verification method, and the error budget produced when the fourth wire is mis-terminated.
PT100 RTD Fundamentals
A PT100 is a platinum resistance temperature detector with a nominal resistance of 100.00 Ω at 0 °C. The resistance/temperature characteristic is defined by IEC 60751 as a polynomial with a nominal slope of 0.385 Ω/°C in the 0–200 °C industrial range. Selected reference points:
| Temperature (°C) | Nominal Resistance (Ω) |
|---|---|
| −50 | 80.31 |
| 0 | 100.00 |
| 50 | 119.40 |
| 100 | 138.51 |
| 150 | 157.33 |
| 200 | 175.86 |
Tolerance classes per IEC 60751:
- Class A: ±(0.15 + 0.002 |t|) °C — at 0 °C this is ±0.15 °C; at 100 °C it is ±0.35 °C.
- Class B: ±(0.30 + 0.005 |t|) °C — at 0 °C this is ±0.30 °C; at 100 °C it is ±0.80 °C.
- 1/3 DIN (Class AA): ±(0.10 + 0.0017 |t|) °C — at 0 °C this is ±0.10 °C.
Wiring Configurations
Three wiring topologies exist for an RTD probe:
- 2-wire: Excitation and sense share the same conductor. Lead resistance adds directly to the element reading. Acceptable only for short runs (≤ 2 m) where lead resistance is small compared to the element change.
- 3-wire: Two leads share the same resistance (assumed matched) and a third lead provides the compensation reference. The lead-resistance contribution is cancelled as long as leads are equal length, equal gauge, and same conductor material.
- 4-wire (Kelvin): Separate excitation and sense pairs. Sense leads carry only the high-impedance voltmeter input current (nA range), so their IR drop is negligible and lead resistance is mathematically eliminated.
| Configuration | Lead Resistance Error | Required Module Support | Typical Cable | Typical Application |
|---|---|---|---|---|
| 2-wire | Full lead resistance on both leads | None — direct ohms input | Twisted pair | Short probes, lab benches |
| 3-wire | Compensated assuming matched pair | Two sense terminals + reference | 3-conductor shielded | Industrial panels, ≤ 25 m |
| 4-wire | None (Kelvin separation) | Dedicated EXC+/EXC− and SENSE+/SENSE− | 4-conductor shielded | High accuracy, long runs, calibration |
Omron CPM1A-TS101-DA Hardware Reference
The CPM1A-TS101-DA is a CPM1A-series expansion module combining one Pt100/JPt100 input channel with one analog output (voltage or current). Refer to the Omron Industrial Automation portal for the operation manual revision shipped with your module. Pt100 input section terminals:
| Terminal Marking | Function | Internal Circuit |
|---|---|---|
| PT1A | Pt100 terminal A — excitation source and sense high | Constant-current source (typically 1 mA) and ADC positive |
| PT1B | Pt100 terminal B — sense return 1 | ADC negative input 1 |
| PT1B (2nd) | Pt100 terminal B — sense return 2 / lead compensation | Lead-resistance measurement node |
| AG (×2) | Analog ground | Common for input and output |
Key module specifications (from the CPM1A series datasheet):
- Input range: Pt100 0–200 °C; JPt100 0–200 °C.
- Resolution: 1/256 of full scale (≈ 0.78 °C/bit) for the RTD input channel.
- Conversion time: ≤ 250 ms per channel.
- Accuracy: ±1.0 % of full scale (subject to lead-resistance compensation).
- Isolation: photocoupler between analog section and PLC bus.
- Terminal screw torque: 0.5 N·m.
The two PT1B terminals are electrically distinct. The module drives a constant current through PT1A and the first PT1B; it then measures the voltage drop across the lead on the second PT1B and doubles it (since both B-leads are assumed to have the same resistance) to recover the true element resistance. This topology is fundamentally 3-wire and cannot accept a 4-wire Kelvin signal.
Lead-Resistance Compensation Theory
For a 3-wire PT100 with lead resistances R1 (on A-lead), R2 (on B-lead 1), R3 (on B-lead 2), the module performs two measurements:
- A-to-B1: V_AB1 = I × (R_element + R1 + R2).
- A-to-B2: V_AB2 = I × (R_element + R1 + R3).
With R2 = R3 (the matched-pair assumption) the element resistance is recovered as:
R_element = (V_AB1 / I) − (V_AB2 − V_AB1) / I = V_AB1 / I − R2
Any deviation from R2 = R3 produces a residual error equal to (R2 − R3)/2 at the element-resistance reading. This is the source of the +2 °C error example calculated later for a 10 m run when the 4th wire is tied to PT1B.
A true 4-wire Kelvin measurement would require separate excitation (EXC+, EXC−) and sense (SENSE+, SENSE−) terminals with independent current sources. The CPM1A-TS101-DA does not provide this. For a deeper treatment, see the Texas Instruments application note on 4-wire RTD measurement circuits with low-side reference.
Root Cause: Why the 4th Wire Must Stay Disconnected
The module's compensation algorithm assumes exactly three conductors reach the element: PT1A, PT1B, and the second PT1B. The two B-leads must share the same physical node at the probe end. Three improper terminations of the fourth wire produce errors:
Termination A — Tie to PT1A
Connecting the redundant A-lead to PT1A creates a low-impedance parallel path from the constant-current source back to the element's A-end. The element current splits between the two leads in inverse proportion to their resistances. The module's compensation reference, however, is built on the PT1B-side sensing and still subtracts (R2 + R3)/2 from V_AB1. Because R1 has effectively become R1 || R1' = R1/2, the measured element voltage drops by (I × R1/2) — a fixed offset — rather than scaling with lead length. The reading shifts positive by ≈ R1/2 / 0.385 Ω/°C. For a 10 m run this is ≈ 1.0 °C.
Termination B — Tie to PT1B
Connecting the redundant B-lead to either PT1B places two conductors in parallel between the terminal and the probe end. The effective B-lead resistance drops from R2 to R2 || R2' = R2/2. The compensation algorithm, which doubles the measured lead resistance, now over-corrects by R2/2. Residual error is +R2/2 / 0.385 Ω/°C, ≈ 2.0 °C for 10 m of 24 AWG copper.
Termination C — Tie to AG or Earth
Connecting the redundant wire to analog ground or chassis earth shorts one end of the element to ground through a low impedance. The constant-current source will drive into the short, the ADC saturates, and the module reports out-of-range. This is the worst-case failure mode and must be avoided during commissioning.
Correct Termination — Leave Disconnected and Insulated
The fourth conductor is allowed to float. To prevent antenna pickup and 50/60 Hz induced voltage, terminate it with a heat-shrink sleeve or terminal-block end-sleeve and dress it back into the cable jacket.
Prerequisites
- Omron CPM1A or CPM2A CPU with the TS101-DA module seated on the expansion bus.
- CX-Programmer 9.x or later (CX-One 4.x recommended).
- PT100 Class A or Class B probe with 4-wire tails; ferruled leads.
- Calibrated decade resistance box or PT100 simulator.
- True-RMS digital multimeter for lead-resistance verification.
Step-by-Step Wiring Procedure
- De-energise the CPM1A system. Lock out the 24 VDC supply feeding the I/O bus.
- Strip 9 mm of insulation from each PT100 lead. Crimp insulated ferrules.
- Identify the two leads bonded to each end of the element. With an ohmmeter, probe between every lead pair. A near-zero reading (≤ 0.1 Ω) identifies the same-end pair.
- Land one lead of the A-pair on terminal
PT1A. - Land one lead of the B-pair on the first
PT1Bterminal. - Land the other lead of the B-pair on the second
PT1Bterminal. - Insulate the remaining A-lead (the 4th wire) with heat-shrink. Tuck it back into the cable jacket. Do not land it on any terminal.
- Torque all terminal screws to 0.5 N·m. Confirm no stray strands bridge terminals.
- Restore 24 VDC and observe the input word in CX-Programmer online mode.
Schematic representation of the wiring:
PT100 Element (4-wire Kelvin construction)
+-------- A-end
| |
| (red lead 1) ---- PT1A (excitation + sense high)
|
| (red lead 2) ---- [HEAT-SHRINK INSULATED, NOT CONNECTED]
|
+-------- B-end
|
(white lead 1) -- PT1B (sense return 1)
(white lead 2) -- PT1B (sense return 2, lead compensation)
CX-Programmer Configuration
The CPM1A-TS101-DA reports the Pt100 input as a 12-bit value (0x0000–0x07FF) in the assigned IR input word (typically IR n). Scaling: 0x0000 = 0 °C, 0x07FF = 200 °C, so each bit ≈ 0.156 °C. CX-Programmer setup:
- In the PLC I/O Table, verify the TS101-DA is recognised and the word allocation is correct (input word = n, output word = n+1).
- Open the DM area. Locate DM
n(analog I/O range setting). Set bits 8–15 to0000for Pt100 0–200 °C, or0001for JPt100. - DM
n+1selects averaging count. Recommended value:0003(8-sample average) for mains-frequency rejection in 50/60 Hz panels. - DM
n+2enables the input. Set bit 0 to1to enable the Pt100 channel. - Read IR
nin a contact/PLC monitor window. Convert to engineering units: T(°C) = IR n × 200 / 2047.
Example conversion: IR n = 0x04B0 = 1200 decimal. T = 1200 × 200 / 2047 = 117.3 °C.
| Word | Function | Recommended Value |
|---|---|---|
| DM n (input range) | Sensor type select |
0000 = Pt100, 0001 = JPt100 |
| DM n+1 (averaging) | Averaging count |
0003 = 8 samples (≈ 2 s) |
| DM n+2 (enable) | Channel enable | Bit 0 = 1 (input enabled) |
| IR n | Pt100 reading (12-bit) | Read-only, scaling 0–200 °C = 0–0x07FF |
| IR n+1 | Output value | Write for analog output control |
Ladder Logic Example: Scaling the PT100 Reading
The following CX-Programmer ladder illustrates the conversion from the raw 12-bit input to a floating-point temperature in 0.1 °C engineering units:
|--[ IR 200 ]--[ MOV(21) #0 D200 ]--| ; Clear working register |--[ IR 200 ]--[ *F(456) #0.0977 D202 ]--| ; Scale: 200/2047 = 0.0977 °C/bit |--[ D202 ]--[ROUND(45) D202 ]--| ; Integer result in 0.1 °C |--[ D202 > 0 ]--[ MOV(21) D202 D204 ]--| ; Sanity check lower bound |--[ D202 < 2000 ]--[ MOV(21) D202 D204 ]--| ; Sanity check upper bound
For applications requiring 0.01 °C resolution, use double-precision arithmetic and a moving-average filter over 16 samples. Always check the result against the under-range (≈ −20 °C) and over-range (≈ +250 °C) module outputs before scaling.
Verification Procedure
- Substitute resistance box: Disconnect the PT100 probe and connect a decade box in 4-wire mode across PT1A, PT1B, PT1B, with the fourth box lead left floating.
- Set 100.00 Ω. Read IR n. Expected raw value = 0x0407 (decimal 1031) ± 1 LSB (≈ ±0.16 °C). Module tolerance plus box tolerance yields expected reading 0.0 ± 0.5 °C for Class B.
- Set 138.50 Ω. Expected raw value = 0x0589 (decimal 1417). Module reading 100.0 ± 0.8 °C.
- Set 175.84 Ω. Expected raw value = 0x0703 (decimal 1795). Module reading 200.0 ± 1.2 °C.
- Reconnect the actual probe. Place probe and a calibrated reference RTD in an isothermal block. Compare readings; deviation should be within sum of Class tolerances.
- Measure the floating 4th wire with a true-RMS DMM referenced to earth. Acceptable induced voltage: < 50 mV RMS. Above 200 mV indicates routing too close to VFD cables.
Lead Resistance Error Budget
| Run Length | Lead Resistance (R2) | Error if 4th Wire Tied to PT1B | Error if 4th Wire Tied to PT1A | Error if Floating |
|---|---|---|---|---|
| 2 m | 0.31 Ω | +0.40 °C | +0.20 °C | 0 (within calibration) |
| 5 m | 0.78 Ω | +1.01 °C | +0.51 °C | 0 |
| 10 m | 1.56 Ω | +2.03 °C | +1.01 °C | 0 |
| 25 m | 3.90 Ω | +5.06 °C | +2.53 °C | 0 |
| 50 m | 7.80 Ω | +10.13 °C | +5.06 °C | 0 |
| 100 m | 15.6 Ω | +20.26 °C | +10.13 °C | 0 |
The error scales linearly with run length. For runs longer than 25 m or applications requiring Class A tolerance, upgrade to a 4-wire-compatible module or use a remote 4-wire-to-4–20 mA transmitter at the probe head.
Alternative 4-Wire Compatible Modules
If the application demands the full 4-wire accuracy, consider these Omron modules from the CP1W series:
| Module | Channels | Sensor Support | Wiring | Resolution |
|---|---|---|---|---|
| CP1W-TS001 | 2 | Pt100/JPt100 | 3-wire only | 0.1 °C |
| CP1W-TS002 | 4 | Pt100/JPt100 | 3-wire only | 0.1 °C |
| CP1W-TS003 | 2 | Pt100/JPt100 | 4-wire compatible | 0.1 °C |
| CP1W-TS004 | 4 | Pt100/JPt100 | 4-wire compatible | 0.1 °C |
| CP1W-TS101 | 2 | Thermocouple + Pt100 | 3-wire | 0.1 °C |
| CP1W-TS102 | 4 | Thermocouple + Pt100 | 3-wire | 0.1 °C |
The CP1W-TS003 and TS004 accept the full 4-wire Kelvin connection. Refer to the Omron Industrial Automation portal for the specific CP1W-TS003/TS004 operation manual before retrofitting.
Cross-vendor equivalents with native 4-wire Pt100 inputs:
- Siemens SM 1231 RTD (6ES7231-5PD32-0XB0) — supports 2-/3-/4-wire RTD.
- Allen-Bradley 5069-IY4 (Logix 5000) — 4-channel universal RTD/thermocouple, 4-wire capable.
- Schneider Electric BMX ART0414 (Modicon M340) — 4-channel RTD, 2-/3-/4-wire.
Edge Cases and Field Diagnostics
Case 1 — Shielded cable with drain wire
If the PT100 cable includes a shield/drain, terminate the drain at the panel end only to the analog ground (AG) terminal. Do not ground at the probe end — this creates a ground loop that injects 50/60 Hz noise into the high-impedance sense lines.
Case 2 — Mixed-gauge leads
If the field installer has used different gauges for the A and B leads, the matched-pair compensation assumption is violated. For example, 18 AWG on PT1A and 24 AWG on PT1B produces a residual error proportional to (R_A − R_B). Replace the cable with matched conductors.
Case 3 — Probe replacement without rewiring
If a 4-wire probe replaces a 3-wire probe, the original cable has only three conductors. Connect both ends of the probe element to two of the three cable conductors as before; the third cable conductor provides lead compensation. The probe's internal 4th lead is bonded internally to one end — confirm by measuring element resistance.
Case 4 — Calibration drift over time
Pt100 Class A tolerance drifts at < 0.05 °C/year if the probe is operated below 400 °C. Above 500 °C, expect drift up to 0.5 °C/year due to platinum oxidation. Recalibrate annually for Class A work; biennially for Class B.
Case 5 — Common-mode voltage on long runs
On runs longer than 50 m, the cable capacitance to ground can store enough charge to develop a common-mode voltage relative to the AG terminal. Add a 100 kΩ resistor from each PT1B terminal to AG at the terminal block to bleed off static. This does not affect the DC measurement because the ADC input impedance is > 10 MΩ.
Safety and Standards Notes
Installation must follow:
- IEC 60751 — Industrial platinum resistance thermometers and platinum temperature sensors.
- IEC 60381-1 — Analogue signals for process control systems.
- NFPA 79 (Electrical Standard for Industrial Machinery) where the panel is part of a machine.
- Local electrical code for conductor sizing, segregation, and overcurrent protection.
Always de-energise the panel before changing sensor wiring. PT100 sensors in intrinsically safe areas require a barrier; consult the relevant ATEX or IECEx certificate before installing in hazardous locations.
Troubleshooting Matrix
| Symptom | Likely Cause | Diagnostic | Corrective Action |
|---|---|---|---|
| Reading ≈ −20 °C (under-range) | Open element or reversed wiring | DMM across PT1A to PT1B with probe connected; expect 100–175 Ω | Re-identify A/B pairs; re-land on correct terminals |
| Reading ≈ +250 °C (over-range) | Short to AG or 4th wire tied to AG | Insulation test 500 V from element to earth | Remove the bond; insulate the 4th wire |
| Reading 2–10 °C high | 4th wire tied to PT1B | Inspect terminal block for stray conductor | Disconnect and insulate the 4th wire |
| Reading 1–5 °C high | 4th wire tied to PT1A | Inspect terminal block | Disconnect and insulate the 4th wire |
| Reading fluctuates ±3 °C | Induced noise on floating 4th wire | DMM between floating wire and earth, AC coupled | Re-route cable; install ferrite; ground drain wire at panel only |
| Reading low at high temperature | Compensation lead on wrong PT1B terminal | Swap the two B-leads | Re-land on correct terminals |
| Module ERR LED ON | Sensor burnout or DM misconfiguration | Check DM n+2 enable bit; DMM resistance | Replace probe; correct DM settings |
| No reading (constant 0) | Module not seated or 24 VDC missing | Check expansion connector, supply voltage | Reseat module; restore power |
FAQ
Can I short the unused PT100 wire to PT1A?
No. Connecting the fourth lead to PT1A lowers the excitation loop impedance and adds a positive offset equal to roughly half the lead resistance divided by 0.385 Ω/°C. For a 10 m run this is about 1 °C of error. Always leave the fourth wire physically disconnected and insulated.
Why does the CPM1A-TS101-DA have two PT1B terminals?
The two PT1B terminals implement the module's 3-wire lead-resistance compensation. The internal ADC measures the voltage drop on the second PT1B lead, doubles it (assuming both B-leads share the same resistance), and subtracts the result from the PT1A-to-PT1B measurement to recover the true element resistance.
How much temperature error do I introduce if the 4th wire is accidentally tied to PT1B?
Approximately R_lead / (2 × 0.385) °C. For 10 m of 24 AWG copper (R_lead ≈ 1.56 Ω) the offset is about 2.0 °C; for 50 m it rises to 10 °C.
Will a 4-wire PT100 give better accuracy on this 3-wire module?
No. The CPM1A-TS101-DA is a 3-wire input only — the fourth lead cannot be used and provides no benefit. Upgrade to the CP1W-TS003 or CP1W-TS004 module for true 4-wire Kelvin measurement.
What should I do with the disconnected 4th wire at the panel end?
Insulate it with heat-shrink or a terminal-block end-sleeve. If the cable is in a high-noise environment, optionally ground the panel-end of the 4th wire to chassis ground (not AG) to act as a shield drain, keeping the probe end floating.