4-20 mA Signals from the Instrument Side

Why current and not voltage

A voltage signal loses some of itself in the wire. A 1 to 5 V signal sent down a long run arrives smaller than it left, by an amount that depends on the wire and the input, and the receiver cannot tell the difference between a lower measurement and a longer cable. Current does not have that problem. In a series loop the current at the transmitter is the current at the receiver, whatever the wire resistance, so the reading is the reading.

The range starts at 4 mA rather than zero for two reasons. A transmitter that has a few milliamps to run on can be powered by the signal itself, which is what makes two-wire instruments possible. And a loop that reads 0 mA is broken, not empty; a real minimum reads 4 mA. That live zero is the single most useful diagnostic property the standard has.

The three parts of the loop

The supply

A DC source, almost always 24 V, either a panel power supply or the internal supply of an input card. It provides the voltage that drives the current around the loop.

The transmitter

A variable resistance in effect. It throttles the current so that exactly 4 to 20 mA flows in proportion to the measurement, and it needs a minimum voltage across itself to do so, typically 10 to 12 V.

The receiver

The input resistance the current flows through, usually 250 ohms inside a PLC or SCADA input, sometimes a 250 ohm resistor added to a voltage input. The current produces a voltage across it, 1 V at 4 mA and 5 V at 20 mA, and that voltage is what the analog to digital converter actually measures.

Two-wire, three-wire, and four-wire

A two-wire transmitter is loop powered. The same pair carries the supply to the instrument and the signal back, and the instrument runs on the current it is regulating. It is the most common arrangement for pressure, level, and temperature transmitters because it is cheap to wire and intrinsically limits how much power can reach the field.

A four-wire transmitter has its own power connection, either 24 VDC or line voltage, and a separate signal pair on which it sources the 4 to 20 mA. Magnetic flowmeters, most analyzers, and anything with a heater or a display that needs real power are four-wire. The signal loop is then powered by the transmitter, not by the receiver, and connecting it to an input that also supplies loop power will damage one of them.

The voltage budget

The supply voltage has to cover everything in the loop at 20 mA: the drop across the receiver, the drop in the wire, and the minimum the transmitter needs to operate. Whatever is left is the margin. The arithmetic is short and it is the single check that prevents a loop from working on the bench and failing in the field at full scale.

With a 24 V supply and a transmitter that needs 12 V, the loop can carry 600 ohms. A 250 ohm input uses 250 of that. A loop-powered indicator in the field might use another 50 or 100. Long runs of small wire use the rest quickly, and the loop that read correctly at 4 mA will clamp somewhere below 20 mA at full scale because the transmitter can no longer push the current through the total resistance. The wire gauge table on this site gives resistance per foot for the arithmetic.

Wiring and shielding

Use twisted, shielded pair. Twisting rejects magnetic pickup; the shield rejects capacitive pickup. Ground the shield at one end only, normally the panel end, and leave it insulated and floating at the instrument. A shield grounded at both ends becomes a conductor for the difference in ground potential between the two locations, and that current induces noise into the pair it was meant to protect.

Keep signal pairs away from power conductors, and particularly away from variable frequency drive output cables, which radiate high-frequency noise that a 4 to 20 mA loop will pick up as a jittery reading. Separate conduit or separate tray sections with distance between them is the remedy; a shield alone is not enough beside a drive cable.

Fault signalling

The standard range leaves room above and below for the transmitter to say it has failed. NAMUR recommendation NE 43 defines the convention most transmitters follow: a failure is signalled by driving the loop below 3.6 mA or above 21 mA, and the valid measurement range, including a little over-range, stays inside 3.8 to 20.5 mA. A receiver that treats anything outside that window as a fault rather than a reading will not run a pump on a dead transmitter.

Checking a loop with a meter

1. 1

   Read the current in series.

   Break the loop at a convenient terminal, put a milliammeter in series, and read the current. It should be between 4 and 20 mA and should agree with the transmitter display and the control system reading. A clamp meter rated for DC milliamps reads it without breaking the loop.

2. 2

   Or read the voltage across the input resistor.

   Measure across the 250 ohm receiver input. 1.000 V is 4 mA, 3.000 V is 12 mA, 5.000 V is 20 mA. This is the check that does not interrupt the loop and it is the one to use on a running process.

3. 3

   Check the voltage at the transmitter.

   Measure across the transmitter terminals at full scale. If it is below the minimum on the datasheet, the loop is out of voltage budget and the reading will clamp.

4. 4

   Simulate.

   Most transmitters can be commanded to output a fixed current from their configuration menu or a HART communicator. Drive 4, 12, and 20 mA and confirm the control system reads 0, 50, and 100 percent of range. Disagreement here is a scaling problem, not a loop problem.

HART on the same pair

HART superimposes a small digital signal on the 4 to 20 mA current, which lets a communicator or an asset management system read diagnostics and change configuration without disturbing the analog value. It needs at least 230 ohms in the loop to work, which a 250 ohm input provides, and it raises the transmitter minimum voltage slightly. A loop that has no resistance, such as one wired straight to a supply with the input bypassed, will carry the current fine and will not talk HART at all.