The 4-20 mA current loop is the standard way industrial transmitters send a measurement, like pressure, level, or flow, to a control system as a proportional electrical current, where 4 mA always means the low end of the range and 20 mA always means the high end. It has survived decades of newer technology because it is simple, cheap to wire, and hard to fool.
In a 4-20 mA loop, the transmitter output is scaled so 4 mA represents 0% of the calibrated range and 20 mA represents 100%. A pressure transmitter ranged 0-100 psi will output 12 mA at 50 psi, a straight linear relationship. Any value in between is read the same way: percent of range, converted to current, converted back to engineering units by the receiving PLC or DCS analog input card.
This convention traces back to earlier pneumatic control signals (3-15 psi) and an intermediate 10-50 mA current standard used with magnetic amplifier equipment. As transistor electronics matured, 4-20 mA became the dominant signal because it needed less power and tolerated longer wire runs. The exact reasoning the standards committees used to land on these specific numbers is not well documented in surviving public records, though the underlying signal-compatibility rules are captured today in ANSI/ISA-50.00.01, originally issued under the ISA S50.1 designation.
The most important design decision in the whole standard is starting the scale at 4 mA instead of 0 mA. This is called a live zero. Because a healthy, powered loop always carries at least 4 mA, a reading of 0 mA is not a valid process value, it is a fault. That single property lets a control system tell the difference between "the tank is empty" (4 mA) and "the wire is broken or the transmitter lost power" (0 mA) without any extra diagnostic wiring.
Some transmitters and control systems apply a downscale or upscale burnout convention on top of this, driving the signal below 3.6 mA or above 21 mA on an internal fault so the failure is unambiguous. In Europe and Japan, this is commonly formalized through NAMUR NE43, which assigns specific bands: roughly 3.8-20.5 mA as the normal measurement range, with lower and upper bands reserved for alarm and fault states.
A current loop is a series circuit, so by basic circuit law the same current flows at every point in the loop, the transmitter, the wire, and the receiver all see the identical value regardless of wire resistance or small voltage drops along the cable. A voltage signal does not have that property: resistance in long cable runs and connections causes the voltage actually seen at the receiver to sag below what the transmitter sent, corrupting the reading over distance.
Current signaling is also inherently more resistant to electrical noise in a typical plant. Current-input receivers present a low impedance (on the order of a few hundred ohms), so induced noise voltage from nearby motors, drives, or switching gear drives very little extra current into the loop. Voltage-input receivers present a much higher impedance, so the same induced noise produces a proportionally larger, more disruptive error. This is a major reason plants run 4-20 mA out to field transmitters and reserve voltage signals like 0-10V for short, low-noise runs such as panel-to-panel wiring.
Most modern process transmitters are 2-wire, loop-powered devices: the same two wires that carry the 4-20 mA signal also deliver the DC power the transmitter needs to operate. A power supply, commonly 24V DC, sits in the loop along with the transmitter and a receiving resistor or analog input card, all wired in series.
The transmitter acts like a variable current regulator, drawing exactly the current that corresponds to its measurement regardless of the loop voltage, as long as enough voltage is available at its terminals. The supply voltage must be high enough to cover the transmitter's own minimum operating voltage, the voltage dropped across the receiver's input resistor, and any voltage lost in the field wiring, at the maximum loop current of 20 mA. This margin is called the loop's compliance voltage, and running short of it is a common cause of a signal that reads correctly at low current but flatlines or clips as the process approaches 100%.
HART (Highway Addressable Remote Transducer) protocol adds a small, symmetrical FSK digital signal on top of the 4-20 mA current, using two audio frequencies (about 1200 Hz and 2200 Hz, per the Bell 202 standard, at 1200 baud) to represent digital ones and zeros. Because that digital signal averages to zero DC offset over each cycle, it rides on the wire without changing the analog current value the control system is reading. This lets a technician pull configuration data, diagnostics, and even a secondary measurement from the same wiring used for the primary 4-20 mA signal, without adding conductors.
Most current loop problems show up as one of a handful of characteristic symptoms.
| Symptom | Likely cause |
|---|---|
| Reads exactly 0 mA | Open circuit: broken wire, loose terminal, blown fuse, or unpowered transmitter |
| Stuck at a fixed value, unresponsive to process changes | Failed transmitter, sensor at its physical limit, or a short somewhere in the loop |
| Reads above roughly 21-22 mA | Loop fault, transmitter driven into saturation, or a wiring short forcing excess current |
| Noisy, erratic reading | Poor grounding, damaged cable shield, or routing too close to VFD or motor wiring |
The standard field check is to break the loop and insert a milliamp meter in series, or, if you cannot break the loop, measure the voltage drop across a known precision resistor, commonly 250 ohms, which converts 4-20 mA into a convenient 1-5V range for a standard multimeter. Isolate the loop into segments (transmitter, wiring, power supply, receiver) and check each one in turn rather than guessing at the whole loop at once. Loose or corroded terminal connections and damaged insulation account for a large share of real-world field failures, so a visual and continuity check of the wiring is worth doing before assuming the transmitter itself is bad.
A 4-20 mA loop tells you the instrument failed. It does not tell you why the underlying asset drifted out of spec in the first place, and a live-zero fault often only gets caught after the process has already been running blind for a shift or more. Fabrico closes that gap by reading machine condition and OEE directly from the line with computer vision, catching degradation that a single point sensor and its wiring can miss, and auto-routing a work order the moment a real loss is detected. It is EU-built with EU data residency and holds ISO 27001, ISO 20000-1, and ISO 9001 certification. Book a Fabrico demo.
Starting the scale at 4 mA creates a live zero. Since a healthy, powered loop always draws at least 4 mA, a reading of 0 mA can only mean a fault, such as a broken wire or a dead transmitter, rather than a legitimate low process reading. A 0-20 mA signal would not let you tell those two situations apart.
Current loops are far less sensitive to cable resistance and length than voltage signals because the current is the same at every point in a series circuit. Practical distance is limited mainly by total loop resistance and the available compliance voltage from the power supply, not by signal degradation the way voltage signals are, which is why 4-20 mA is preferred for long field runs.
Yes. HART superimposes a low-level FSK digital signal on top of the 4-20 mA current using two audio frequencies. Because this digital signal has zero net DC effect, it does not alter the analog measurement, so both the analog value and digital diagnostics travel on the same two-wire loop.
Treat it as an open circuit until proven otherwise: check for a blown fuse, a loose or disconnected terminal, a broken wire, or a transmitter that has lost power. A true process reading can never produce 0 mA in a correctly functioning live-zero loop.
Login
Blog