Dead Zero in 4-20 mA Loops: Diagnosing 0 mA Faults

Problem Details: 0 mA Is Not 0%

A 4-20 mA current loop encodes the process variable linearly between a live zero of 4 mA and full scale at 20 mA. When the loop current collapses to 0 mA, the measurement chain has failed — the transmitter is no longer driving the loop at all. This condition is commonly called the dead zero problem.

Root Causes

Group causes by where the loop current is interrupted. The loop is a series circuit: power supply → transmitter → field wiring → receiver input resistor → back to supply. Any single break anywhere produces 0 mA everywhere.

Diagnostic Procedure

Work from the control room outward; each step localizes the break to one side of a boundary.

1. Read the raw counts, not the engineering-unit tag. Scaling can mask the fault. Confirm the raw input word is at or below the value that corresponds to 0 mA rather than sitting at the 4 mA endpoint.
2. Check whether neighboring loops are also dead. Multiple simultaneous dead zeros on one card or one power rail point to a shared supply, fuse, or breaker — not to a field instrument.
3. Verify the loop supply voltage at the marshalling terminals with the loop connected. A 2-wire transmitter needs its rated minimum terminal voltage; measure and compare against the transmitter datasheet before condemning the device.
4. Measure loop current in series at the marshalling terminals using a milliamp meter or the test jacks. If the meter reads a valid 4-20 mA there, the break is downstream in the input module or its wiring.
5. Move to the field terminals at the transmitter and repeat the series measurement. Valid current at the transmitter but 0 mA at the panel isolates the fault to the field cable, junction box, barrier, or isolator.
6. Force the transmitter output using its loop-test / fixed-current function (via the local keypad or a HART communicator, if the device supports it) and command a known value such as 12 mA. If the forced current does not appear, the transmitter output stage or its power feed is the problem.
7. Ohm out the cable with the loop de-energized and both ends disconnected. Check conductor continuity, conductor-to-conductor short, and conductor-to-shield/ground leakage.
8. Substitute a known-good source. Inject 4, 12, and 20 mA from a calibrator directly into the input channel. If the channel tracks the injection, the receiver is healthy.

Distinguishing Dead Zero from a Genuine Zero Process Value

The confusion arises when the operator sees 0 in engineering units on the HMI. Two very different loop states can produce a displayed zero, and the control logic must separate them.

Detection Logic and Configuration

Do not rely on the operator noticing a flat trend. Configure the analog input path to declare the fault explicitly.

• Enable open-wire / under-range detection on the analog input module where the hardware provides it, and map the resulting channel status bit into the tag's quality flag.
• Add an under-range comparison in logic. A current below the low fault threshold (commonly treated as roughly 3.6 mA and below) is not a measurement — latch a fault, freeze the last good value or substitute a safe default, and inhibit any PID loop using that variable from integrating on garbage.
• Do not clamp silently. Modules or scaling blocks that clamp everything below 4 mA to 0% will hide the dead zero completely. Scale from raw counts so an under-range value remains visible, or read the module's status word.
• Set the transmitter's failure mode deliberately. Devices that support a namur-style failure output can drive high or low on internal diagnostics, but note that an internal failure drive-low is still a nonzero current — a true 0 mA reading indicates a loss of the loop itself, not a diagnostic alarm.

Sample under-range interlock, expressed generically:

// Raw scaling: AI_Counts -> AI_mA (module dependent)
LOOP_FAULT   := (AI_mA <= 3.6) OR (AI_mA >= 21.0) OR NOT CH_Status_OK;
PV_GOOD      := NOT LOOP_FAULT;
IF LOOP_FAULT THEN
    PV := PV_LastGood;      // hold, do not fall to 0%
    PID.Auto := FALSE;      // drop the controller to manual / safe output
ELSE
    PV_LastGood := PV;
END_IF;

Repair and Verification

1. Restore the failed element: reterminate the conductor, replace the fuse or reset the breaker, replace the transmitter, barrier, isolator, or input module.
2. Re-energize and confirm the loop supply voltage at the transmitter terminals under load.
3. Force the transmitter output to 4, 12, and 20 mA and confirm the controller raw value and engineering-unit value at each point. Record the deviation.
4. Return the transmitter to normal output and confirm the reading tracks a real process change or a known reference.
5. Deliberately re-test the detection path: disconnect one loop conductor briefly (where safe and permitted) and confirm the LOOP_FAULT bit sets and the HMI raises a bad-quality alarm instead of showing a plausible 0%.
6. Clear the latched fault, confirm normal quality returns, and log the root cause against the tag so repeat failures on the same fuse, cable run, or device family become visible.

FAQ