When connecting analog transmitters to PLCs or data collection systems, choosing between 4-20mA current loop wiring and voltage-based signals like 0-10V is crucial. This article helps you calculate the required loop voltage, specify appropriate transmitters, and select wiring configurations for reliable signal transmission. We will also cover troubleshooting common issues to ensure system integrity.
Why the Loop Survived
A 4-20mA loop survived because three properties are useful at once. The current is the same at every point of a series circuit, so a long cable and a dirty terminal do not change the reading. The signal is current-driven, so series resistance and induced noise are rejected rather than integrated, unlike a voltage signal that is loaded by whatever impedance the receiver presents. And the range does not start at zero: 4mA means span low, so a broken wire and a genuinely low measurement are distinguishable.
That last point is a standard in its own right. NAMUR recommendation NE43 reserves the bands below 3.6mA and above 21mA for fault signalling, so a transmitter can report a failed sensor while producing a value that no controller should mistake for process. The IEC 60381 series fixed the 4-20mA analogue signal itself. Check that the receiver actually supports the NE43 bands before relying on them, because a plain resistor across the input cannot tell you which side of the range you left.
Against 0-10V the trade-off is not only noise. A voltage signal depends on the source impedance and the load divider, so a transmitter is fine across the room and wrong across the site as soon as the cable length changes. A current loop has no such dependency. Its limit is the voltage budget, and that is where almost every real field problem sits.
Two-Wire, Three-Wire and Four-Wire Transmitters
A two-wire transmitter is a current regulator in series with the loop. It has no separate supply and steals its operating power from the same 4-20mA it is reporting, which caps what its electronics may consume. That single constraint is why the loop voltage budget is a design task rather than a formality: at full scale the transmitter must still hold enough volts across its own terminals to stay regulated while everything else in series takes the rest.
V_supply >= I_loop x (R_load + R_cable + R_isolator) + V_transmitter
For a 20mA loop with a 250 ohm receiver resistor and a transmitter that needs 12V to regulate, the supply must clear 0.02 x 250 + 12 = 17V. A 24V supply leaves 7V of headroom, which reads as generous until the cable, the isolator and a termination at each end are added to it.
Three-wire devices take their supply on a separate conductor, so the signal current no longer carries the electronics’ appetite. The usual arrangement is positive supply, positive signal and a shared negative, or a negative supply with a sourcing output. Four-wire transmitters carry supply and signal on separate pairs entirely and can source as well as sink, which is what a mains-powered meter with a display or a heater needs.
The decision most installers never write down is whether the receiver input sinks or sources current. An input that supplies loop power is a transmitter-supply input and pairs only with a passive two-wire device. An input that expects to sink current from an externally powered transmitter reads nothing at all if you connect the wrong type. Several cards support both with a link or a software selection; confirm it before the cable is dressed.
Trade-offs Between the Three Configurations
The table below is the one to fill in per loop, because the columns that matter are power available, wiring count and what the receiver expects.
| Configuration | Power source | Conductors | What it suits | What it costs you | The trap |
|---|---|---|---|---|---|
| Two-wire passive | The loop itself | 2 | Remote sensors where running a supply is expensive | Lowest quiescent current, so the budget is tight | Forgetting the transmitter’s minimum terminal voltage |
| Three-wire | Separate supply | 3 | Loop-powered instruments needing more appetite | One extra pair to the field | Common-negative versus common-positive confusion |
| Four-wire | Separate, often mains | 4 | Meters with displays, heaters or relays | Supply and signal both to the field | Sourcing output into a sinking input |
| Two-wire with isolator | The loop, galvanically split | 2 plus isolator | Breaking a ground loop between two cabinets | Burden voltage and panel space | Budgeting the isolator at zero volts |
| 0-10V voltage signal | Separate supply | 2 or 3 | Short, quiet runs into a high-impedance input | Simplest possible receiver | Loading and cable-resistance error grows with distance |
Wiring and Commissioning Procedure
- Draw the loop as one series circuit: supply positive, transmitter, receiver input, conductor back to supply negative. Everything in series sees 20mA at full scale, so total the voltages before ordering anything.
- Add up the resistance budget. Load resistor, round-trip cable resistance (there and back, so double the one-way figure), any isolator, surge protection and terminal resistance.
- Check the headroom. Multiply the total resistance by 0.02A, add the transmitter’s minimum terminal voltage from its datasheet, and confirm the result sits comfortably under the nominal supply, not just under it on paper.
- Bond the screen at one end only, normally at the receiving cabinet. A screen bonded at both ends is a second path for ground current and an efficient antenna for the interference you fitted it to reject.
- Put the isolator where the ground path must break, not where the cable happens to be convenient, and keep any stub between isolator and transmitter short.
- Energise with the transmitter disconnected. Confirm the open-circuit voltage at the field terminals with the receiver in place, then connect and read the current in series rather than trusting a display.
- Ring out both ends. Force 4mA and 20mA from a calibrator and confirm the control system shows the correct lower and upper values, not transposed. Transposed scaling is common and reads like a sick transmitter.
- Prove the fault bands. Note what the system does at 3.2mA and at 21.5mA, so a cut conductor annunciates instead of holding the last good value.
Four Classic Loop Faults
The four faults below account for most callouts, and each has a distinct signature with a meter in series.
Symptom: the Loop Reads Nothing
Zero current means the series circuit is open, or the transmitter is not getting the voltage it needs to regulate. Measure the supply at the transmitter terminals with the loop connected. If it is correct and the current is still zero, the open is in the transmitter, its terminals or a cut conductor. If it is low, the budget has been exceeded and the fix is a stronger supply or a smaller load resistor, not a new transmitter.
Symptom: the Loop Sits at 20mA
A saturated reading is either a genuine over-range process or a transmitter in a fail-high state. Separate them at the sensor: if the process really is high, the loop is telling the truth and the plant has a different problem. If it is not, check scaling direction and supply polarity first, then the input itself, because a drifting measuring cell commonly ends up pinned at the top of the range.
Symptom: a Supply-Frequency Ripple
A 50 or 60Hz component on the signal is almost always a ground loop: two earth points at different potentials pushing current through the screen or the signal common. Break the path rather than filtering the symptom. Bond the screen once, insert a signal isolator, and confirm the ripple does not return when another device on the same circuit is switched.
Symptom: the Reading Moves on Its Own
Drift without a process change is usually thermal or mechanical. A transmitter on a pump discharge sees process temperature swing; a terminal that was never torqued sees vibration. Compare a calibrator reading at the transmitter with the control system value. Agreement at the source and disagreement at the display points at scaling, or at an input being loaded by a second device on a shared terminal.
Copper Resistance, Voltage Drop and the Real Distance Limit
Datasheets quote a maximum distance for a 4-20mA loop, and the number is less meaningful than the calculation behind it. There is no distance limit in the signal itself; there is only the voltage budget of the previous section, and cable resistance is one line in it. Distance becomes a problem when the copper in the budget eats the headroom.
Work with loop resistance, not one-way resistance. The current leaves on one conductor and returns on the other, so a run of L metres counts the resistance of 2L. Annealed copper at 20 C gives roughly these one-way figures: 0.5 mm square about 37 ohm per km, 0.75 mm square about 24 ohm per km, 1.5 mm square about 12 ohm per km, 2.5 mm square about 7.4 ohm per km. Twenty-two AWG is close to 0.35 mm square and sits near 54 ohm per km.
So a 500 m run of 1.5 mm square pair is 1 km of copper, about 12 ohm for the loop, and 0.24V at 20mA. The same run in 22 AWG is around 54 ohm and 1.08V. Neither number ends a 24V loop, which is why the distance limits that actually bite are set by the transmitter’s minimum terminal voltage and the receiver burden, not by the metres.
Then correct for the real site. Copper resistance rises about 0.39 % per kelvin, so a pair laid in a hot pipe trench carries more than the table value. Add every series element the loop actually contains: the surge protection resistors in the isolator, the terminals, the sense resistor inside a drive’s analog input, a display in series. Multiply the total by the maximum current the loop can reach, which is the NE43 upper fault band rather than 20mA, because a fault current must still be drivable.
Keep at least a volt or two between that result and the nominal supply, and re-run the arithmetic whenever anything is added in series. The rule that survives contact with the field is unglamorous: the loop is a series circuit, the supply is fixed, and every component you add without recalculating takes volts somebody else needs.
Turning the Fault Bands Into Alarms
A transmitter can signal a fault all day long and the control system can ignore it. The bands only work if the input is configured to react, and on a plain resistor termination that reaction is entirely software.
Set three limits per analog input rather than one. Below the lower fault band is a break, a failed sensor or an unwired input. Inside the normal range near the bottom and the top is a warning band, which is where a drifting instrument spends its last few weeks before it crosses into fault. Above the upper fault band is a shorted element or a saturated output. The transition points come from the instrument datasheet and, where NE43 compliance is claimed, sit at 3.6mA and 21mA. Confirm what the specific device does before copying those numbers, because not every transmitter is set to the recommendation.
Annunciation is not the only decision. A controller that holds the last good value during a fault quietly keeps a regulator running on a dead signal; one that drives to a predetermined output may trip a plant on a broken wire. Both are legitimate, and both belong in the loop document with the reason. On a safety-related function the answer is set by the safety requirement, not by the analog input manual.
Test the whole chain the way the fault will arrive. Inject 3.2mA at the cabinet terminals with a calibrator and confirm three things in order: the transmitter is bypassed, the input annunciates as a fault and not as a low reading, and the logic downstream of it does what the document says. Then do the same at 21.5mA. A loop that has never had its fault band exercised will have a configuration error in it somewhere; the calibrator finds it in five minutes, the plant finds it during a shutdown.
Finally, put the thresholds on the loop sheet next to the scaling and the burden resistor value. The next technician to touch that card did not read this article, and the terminal block is not self-documenting.
Scaling, and the Mistakes That Follow From It
The conversion is one line, and the endpoints are the whole skill:
Value = Value_low + (I_mA – 4) x (Value_high – Value_low) / 16
The divisor is 16 because the working span is 4mA to 20mA, not 0 to 20. A 0-16 bar transmitter reading 11.2mA reports 0 + (11.2 – 4) x 16 / 16 = 7.2 bar. At 4mA it reports 0 bar and at 20mA it reports 16 bar, while a technician who assumes 0-20mA scaling reads 8.96 bar and spends the afternoon blaming the instrument.
Run the calculation backwards at commissioning and put it on the loop sheet. On that transmitter a 3 bar alarm sits at 4 + 3 x 16 / 16 = 7mA. Inject 7mA at the terminals, watch the control system annunciate, and you have tested the whole chain including the scaling rather than only the sensor.
The same line exposes an over-ranged installation. If a vessel never sees more than 6 bar on a 0-16 bar transmitter, the working signal lives between 4mA and 9.5mA, and every milliamp of error is 0.63 bar of measurement. Specify the span so the process sits near the top of it.
Then the recurring errors:
- Measuring a loop with a voltmeter across it. A meter in series tells you the truth; a high-impedance voltmeter tells you almost nothing.
- Adding a second device into an existing loop without recalculating the budget. Every element in series takes volts the supply may not have.
- Bonding the screen at both ends for safety, which builds a ground loop with a very low impedance path.
- Assuming an input can power a transmitter, when sourcing and sinking analog inputs are not interchangeable.
- Calibrating the transmitter and never the controller scaling, which has not been checked since commissioning day.
Quick Selection Checklist
- Verify the required loop voltage is within the transmitter’s specifications.
- Ensure the loop resistance does not exceed the maximum allowed.
- Choose the appropriate transmitter configuration based on power and distance requirements.
- Check for compatibility with the PLC or data collection system.
- Confirm the transmitter is rated for the environmental conditions.
- Ensure the wiring is properly shielded and grounded to prevent EMI.
- Consider using a signal isolator if ground loops are a concern.
FAQ
What is the maximum distance for a 4-20mA current loop?
The maximum distance depends on the loop voltage and the resistance of the wiring. Generally, loops can extend up to several kilometers if the voltage drop is managed properly.
Can I use a 4-20mA transmitter with a 0-10V input?
Yes, but you will need a current-to-voltage converter. This device converts the 4-20mA signal to a 0-10V signal.
How do I detect a wire break in a 4-20mA loop?
A wire break will cause the current to drop below 4mA, which can be detected by the receiver as a fault condition.
What is the advantage of using a three-wire transmitter over a two-wire transmitter?
A three-wire transmitter provides more power to the transmitter, allowing for more complex circuitry and higher power consumption.
Sourcing the Loop Hardware
Xiamen Lisen Trading Co., Ltd supplies pressure and temperature transmitters, signal isolators, analog input modules and terminal hardware for 4-20mA field loops. Tell us the measured variable, the supply you have at the cabinet and the receiver card, and we will confirm sourcing or sinking and the burden resistor before quoting.
Related Transmitters and Input Modules
- Danfoss MBS3000 pressure transmitter
- Danfoss MBS5100 pressure transmitter
- Allen-Bradley 80190 analog variable frequency board

