Dual element RTD wiring: six leads, two elements, and why one element can't feed two inputs

The call usually goes like this. The PLC already reads a temperature off a Pt100, and now a second system needs the same number: a standalone controller, a recorder, a DCS getting added next to the old rack. Can they just parallel the RTD leads over to the new input?

No. And quite often the probe in that well already has six leads coming out of the head, which means the answer was sitting in the junction box the whole time.

That's what dual element RTD wiring is for: two independent elements in one sheath, each wired to its own input. Simple on paper. In practice I keep seeing leads landed by color, both elements on one card called redundant, and one element shared between two receivers. Here's how to sort the leads, where they go on the common Allen-Bradley RTD cards, and what to do when one temperature really has to reach two places.

Count the leads before anything else

Before any dual element RTD wiring starts, count what comes out of the probe. Inside are two separate elements, usually Pt100, sharing one tip but not connected to each other. The lead count tells you how each one is built.

 

Leads out of the probe	What's inside	What each element lands on
4	Two 2-wire elements, or one single 4-wire element	2 x 2-wire: a 2-wire input each, no lead compensation. 1 x 4-wire: one input, 4-wire or 3-wire with a lead left open
6	Two 3-wire elements (2 x 3-wire), the common build	A 3-wire input each
8	Two 4-wire elements	A 4-wire input each, or a 3-wire input with one sense lead left open

Lead count and what it means

Four leads is the sneaky one: a single 4-wire probe looks identical from the outside. The meter check below sorts it out: two separate pairs with open between them is two elements, all four leads reading each other is one 4-wire element.

Six leads can't give you two 4-wire elements. I get asked that more than I'd expect. Dual 4-wire means eight leads, and eight conductors back to the panel.

For a PLC job I rarely see the point anyway. The 1746-NR8, 1762-IR4 and 1756-IR6I run a 4-wire element as 3-wire with a lead left open, and the 5069-IY4 takes 2-wire and 3-wire RTDs only. Eight leads buys you one spare conductor per element. That’s it.

Sorting six leads into two elements with a meter

Most dual element RTD wiring mistakes start with colors. One common six-core extension cable for 2 x 3-wire probes has four red cores and two white. Which reds go with which white? The insulation won't say, and a probe re-terminated once in the field can be anything. I stopped reading colors a long time ago. A meter doesn't guess.

Disconnect the probe from the card first, so no excitation current sits on the leads.

1.    Measure any lead against the other five. Two read a resistance, three read open. That lead plus those two are one element. Tag them; the other three are element 2.

2.    In each group, find the pair that reads a fraction of an ohm. That's the compensation pair, both on the same end of the element. The third lead is the lone one.

3.    Pair lead to lone lead reads the element itself: 100 Ω at 0 °C on a Pt100, about 108 Ω at 20 °C. Both elements should agree within a few tenths of an ohm.

4.    Any lead of one group to any lead of the other, and every lead to the sheath: all open.

Step 4 is the one people skip, and for redundancy it's the one that matters. Anything but open between the elements, and that probe stays off a redundant loop. Whatever links them inside the sheath can take both readings down together.

Sorting leads for dual element RTD wiring: continuity groups first, then the near-zero pair, then the open reading between the groups.

Dual element RTD wiring to a PLC: one element, one channel

Each element is a complete 3-wire sensor. Give each one its own channel; the only thing they share is the cable shield.

The two pair leads are interchangeable, since both land on one end of the element. The lone lead isn't, and that's where it goes wrong.

 

Card	Lone lead	Compensation pair	Worth knowing
1746-NR8 (SLC 500)	Return x	RTD x and Sense x	4-wire element: the second sense lead is left open
1756-IR6I (ControlLogix)	IN-x/A	IN-x/B and RTN-x/C	No more than two wires on any one terminal
5069-IY4 (Compact 5000)	IN x+	IN x- and RTD-/COM	One wire per terminal; a 2-wire RTD in 3-wire mode needs IN x- jumpered to RTD-/COM

Dual element RTD wiring on three Allen-Bradley cards: where each element's leads land, per channel

A crossed lone lead doesn't give you a reading that's a little off. Run the compensation math with it swapped and the channel lands near zero ohms or below, depending on which two you crossed. Ugly, but nobody mistakes it for a real temperature.

Dual element RTD wiring for redundancy on two 1756-IR6I cards: lone lead to IN-0/A, pair to IN-0/B and RTN-0/C, one element per card.

Same card or two cards?

Depends on why you bought the probe. For a spare element, two channels on one card are fine. For redundancy, use different cards: a module fault or a pulled terminal block takes both readings out in one shot.

Channel isolation varies a lot between them too:

 

Card	RTD wiring it takes	Between channels	Lifecycle status
5069-IY4	2-wire, 3-wire	No isolation between input ports; ±10 V channel-to-channel common mode in 3-wire RTD mode	Active
1762-IR4	2-wire, 3-wire (4-wire runs as 3-wire)	±10 V DC channel-to-channel isolation	Active Mature
1756-IR6I	2-wire, 3-wire (4-wire runs as 3-wire)	Individually isolated channels	Discontinued March 31, 2017
1746-NR8	2-wire, 3-wire (4-wire runs as 3-wire)	5 V DC continuous between channels; ±1 V maximum common mode	Discontinued September 30, 2024

What each card takes, how its channels are separated, and lifecycle status as of October 2026

One element, two inputs: why it never works

Sharing one element is the dual element RTD wiring question I get more than any other.

An RTD input doesn't just listen. It pushes a constant current through the element and measures the voltage: 600 µA on a 5069-IY4 in 3-wire mode, 594 µA on a 1756-IR6I. Parallel a second input on and two currents flow through the element. When both flow at once, each card sees the voltage from both, divides by its own current only, and gets roughly double the resistance.

A Pt100 at 25 °C is about 110 Ω. Double that and the card reports somewhere above 300 °C. Whether the error holds steady or jumps with scan timing depends on how each card drives excitation. Useless either way.

It gets worse. Self-heating goes with the square of the current, so twice the current means four times the heat in the element. Each card's lead compensation assumes it's alone on those wires. And two input circuits are now tied together through one sensor; a 1746-NR8 is rated for ±1 V common mode, not much room for two systems on different grounds.

Left: two receivers on one element push two excitation currents through it, and each reads roughly double. Right: an RTD isolator excites the element once and feeds both receivers.

So what do you do when two systems need the same temperature?

 

Situation	What I'd do	Watch out for
Probe can be replaced, thermowell in place	Dual element probe, one element per receiver	Meter check first, especially the open reading between elements
One element, probe can't change	RTD isolator with one input and two 4-20 mA outputs (a Pepperl+Fuchs KFD2-UT2-Ex2 can be set up as a signal splitter)	Both receivers depend on one powered device; both inputs become 4-20 mA
A transmitter or controller already reads the RTD	Use its retransmit or second analog output	Ground the shield of that extra output at one end only
Both systems are on a network	Pass the value controller to controller	Alarm on lost comms, or the second system holds a stale value
No spare element and the process can't come down	Clamp-on surface RTD on the pipe as a stopgap	It reads the pipe wall, not the fluid, and it responds slower

Dual element RTD wiring and the other ways to get one temperature into two systems

Redundancy: what the second element covers, and what it doesn't

A dual element probe protects you against one element failing. It doesn't give you two independent measurements: same sheath, same thermowell, usually the same cable. Whatever happens to that hardware happens to both.

 

Comparing the elements catches	Comparing the elements misses
One element open or shorted	Probe pulled back and no longer touching the bottom of the well
One element drifting away from the other	Both elements drifting together after the same heat history
A broken or loose lead on one element	Buildup on the thermowell slowing both readings down
A failed channel or card, if the elements sit on different cards	A cut or crushed cable that carries both elements

What comparing the two elements can and can't catch

Setting the deviation alarm

This is where good dual element RTD wiring gets undone in the program. Two healthy elements are allowed to disagree.

Say both are Class B at 100 °C. Class B tolerance is ±(0.3 + 0.005|t|) °C, so ±0.8 °C there. One element 0.8 high and the other 0.8 low are both in spec, 1.6 °C apart. Class A, ±(0.15 + 0.002|t|) °C, allows up to 0.7 °C apart.

Then the cards. A 5069-IY4 channel in RTD mode is rated at 0.1% of full scale, calibrated, at 25 °C. On the 1 to 500 Ω range that's half an ohm, roughly 1.3 °C, times two channels. Stack it all up: about 4 °C apart on paper at 25 °C with nothing wrong. Over its full temperature range the same card carries 0.3% of full scale total error, so the paper worst case is wider still.

A 0.5 °C deviation alarm there just makes nuisance trips. I log the real difference for a couple of weeks after startup, then set the alarm comfortably above that and below the paper worst case.

And don't blindly average the two. An open element sends its channel off scale, or to whatever open-circuit value the card is set for, and the average jumps with it. Check each channel's fault status, drop the bad element, compare or average only when both are healthy.

With a dual-input transmitter, that logic lives in the transmitter. On a Rosemount 644 with the dual-sensor options, Hot Backup moves the primary variable to sensor 2 when sensor 1 fails, needs the PV set to First Good or Average, and stays on sensor 2 until reset over HART, from the local display, or by a power cycle. Sensor Drift Alert only raises a warning flag by default unless you set it to drive the output to alarm. Endress+Hauser's TMT162 offers the same pair as sensor backup and drift detection.

Fitting a dual element probe into the existing thermowell

The last objection to dual element RTD wiring on a retrofit: the new probe won't fit the existing well. Usually it does. Dual element probes are made on 6 mm sheaths too, and 6 mm is 0.236 in. A well bored for 1/4-inch elements typically has a 0.260 in bore, one for 3/8-inch elements about 0.385 in.

Check sheath diameter against the bore, insertion length so the tip still bottoms out, and the process connection and head. Then the cable: a 2 x 3-wire probe needs six conductors, and the cable for a single 3-wire element often has just three. The swap itself doesn't open the process. That's the point of the well.

Dual element RTD wiring questions I get asked

How many wires does a dual element RTD have?

Four, six or eight. Six (two 3-wire elements) is the common one. Four can also be a single 4-wire element; six is never two 4-wire elements.

What do I do with the unused element?

If a channel is free, wire it: a spare that's already reading is a spare you know works. If not, insulate each lead separately, land them on spare terminals and label them, never loose in the head.

Can one RTD element go to a PLC and a controller at the same time?

No. Each input's excitation current flows through the same element, and both readings come out wrong. That's the reason dual element RTD wiring exists. With a single element, use an RTD isolator with two outputs or a retransmitted 4-20 mA signal.

Is there a color code for dual element RTD wiring?

Not one I'd wire by. Four red and two white is a common scheme, and it can't tell you which red belongs to which element. Sort with a meter, tag the leads, and colors stop mattering.

RTD cards for a dual element loop, in stock

Two elements on two cards is easy when the cards exist. On older racks that's the hard part. The 1756-IR6I has been discontinued since 2017 and the 1746-NR8 since September 2024, and IQElectro keeps both on the shelf, along with the 1746-NR4, the 1762-IR4 for MicroLogix expansion racks and the 1769-IR6 for CompactLogix. Every card ships tested with a 2-year warranty. Adding a second card for a dual element RTD wiring job? Send us the catalog number off the one already in the rack and we'll confirm a match before you order.