Level Loop Current and NAMUR Clamp: 3.5 to 20.5 Reality

A level transmitter does not necessarily stop at 20 mA when the level exceeds its configured range. With a 20.5 mA high-saturation setting, the output can rise above 20 mA before it reaches that limit and stops tracking further increases. The corresponding low-saturation setting is commonly 3.8 mA in manufacturer-documented NAMUR NE43 configurations—not a universal 3.5 mA floor. Confirm the actual transmitter settings before interpreting either value.

The important distinction is saturation versus a fault signal: a saturated current no longer uniquely represents the process level, while a configured fault current indicates that the measurement should not be treated as valid.

Distinguish Measurement Range, Saturation, and Fault Current

The normal 4–20 mA span maps the configured process endpoints to 4 mA and 20 mA. Saturation limits define how far the analog output may extend beyond those endpoints during otherwise valid measurement. Fault currents serve a different purpose: they signal a detected failure or invalid measurement.

For example, Emerson’s Rosemount 1408A Reference Manual, document 00809-0500-4480, Rev AB, Appendix A.2.2, lists the following settings for that model:

These values describe the 1408A’s documented behavior, not every level transmitter. The manual states that the current follows the measurement until it reaches the associated saturation limit, then freezes at that limit.

Receiver fault thresholds also need checking. The Endress+Hauser RMS621 Technical Information, TI092R/09/en/13.10, page 4, documents 3.8 mA and 20.5 mA under/overrange boundaries and fault recognition at ≤ 3.6 mA or ≥ 21.0 mA. A receiver’s fault threshold and a transmitter’s specified alarm-output value are related settings, but they are not necessarily the same number.

Consequently, 3.5 mA must not automatically be interpreted as “a low but valid level.” It falls below the low-fault threshold in the RMS621 example. Check the installed device’s diagnostics, alarm configuration, and loop condition rather than converting that current directly into level.

Calculate the Linear Current Before Applying Saturation

For a linear level output:

[
I_{\text{ideal}}
=
4\ \text{mA}
+
16\ \text{mA}
\left(\frac{h-h_4}{h_{20}-h_4}\right)
]

where:

• (I_{\text{ideal}}) is the calculated current before output limiting, in mA.
• (h) is the level relative to the selected datum, in m.
• (h_4) is the level assigned to 4 mA, in m.
• (h_{20}) is the level assigned to 20 mA, in m.

All level values must use the same unit and datum. This equation assumes that the output is linear with level. If a pressure transmitter supplies the signal, confirm the pressure-to-level relationship, density assumptions, and installation offsets first. Do not apply a square-root transfer function simply because the instrument measures differential pressure; the Rosemount 3051S manual identifies square-root output as a flow-proportional setting.

For an illustrative healthy-output configuration with saturation at 3.8 mA and 20.5 mA:

[
I_{\text{limited}}
=
\begin{cases}
3.8\ \text{mA}, & I_{\text{ideal}}<3.8\ \text{mA}\
I_{\text{ideal}}, & 3.8\ \text{mA}\le I_{\text{ideal}}\le20.5\ \text{mA}\
20.5\ \text{mA}, & I_{\text{ideal}}>20.5\ \text{mA}
\end{cases}
]

This limiting model applies only while the transmitter is producing a measurement-related output. An active fault, fixed-current test, or other operating mode must be interpreted separately.

Example: A Nonzero Level Assigned to 4 mA

Assume:

• 1.000 m corresponds to 4.000 mA.
• 5.000 m corresponds to 20.000 mA.
• The output is linear with level.
• Low and high saturation are configured at 3.800 mA and 20.500 mA.
• The sensor can measure the illustrated levels without a fault or another limit taking precedence.

The span is 4.000 m. At 6.125 m:

[
I_{\text{ideal}}
=
4+16\left(\frac{6.125-1.000}{5.000-1.000}\right)
=
24.500\ \text{mA}
]

The ideal extrapolation is 24.500 mA, but the configured output is limited to 20.500 mA.

The same calculation gives the following checkpoints. Values are displayed to three decimal places; the extra digits describe the arithmetic, not instrument accuracy.

Two details matter:

• Passing the 5.000 m upper endpoint does not immediately force the output to 20.500 mA. In this example, current continues increasing until the level reaches 5.125 m.
• A positive level assigned to 4 mA changes the mapping offset, not the configured low-saturation current. A suppressed-zero setting does not establish a universal 3.5 mA floor.

Why a Saturated Current Cannot Give an Exact Level

Before saturation, the inverse linear conversion is:

[
h
=
h_4
+
\frac{I-4\ \text{mA}}{16\ \text{mA}}
(h_{20}-h_4)
]

Here, (I) is the measured loop current in mA, and the level variables are in m.

For the example above, converting 20.500 mA gives 5.125 m. That is the level at which high saturation begins—not proof that the actual level is exactly 5.125 m. Both 5.125 m and 6.125 m produce the same limited current under the stated assumptions.

Use transmitter diagnostics and an appropriate independent process indication to investigate the actual condition. A digital process value may remain available beyond the analog range on some devices: the Rosemount 3051S manual, section 2.5.1, explicitly describes digital readings continuing within the sensor’s digital limits even when the configured 4–20 mA range is exceeded. That capability must be confirmed for the installed instrument and its current status.

Check Range Direction Without Silently Changing It

The calculation requires two distinct endpoints. If (h_{20}=h_4), the span is zero and the conversion is undefined.

If (h_{20}<h_4), the equation describes a reverse-acting mapping: increasing level produces decreasing current. Do not silently swap the endpoints, because doing so changes the intended signal direction. Confirm that reverse action is intentional and supported by the transmitter configuration.

For a direct-acting calculation, require (h_{20}>h_4) and flag a reversed entry for review. Rejecting a reversed range is an input-validation choice for that calculation—not a universal rule that all instruments refuse reverse action.

Field Checks for an Unexpected Loop Current

Before forcing an output, injecting current, or changing settings, follow the approved site procedure and coordinate any necessary control-mode changes or alarm handling. The Rosemount 3051S manual, section 2.2.1, specifically requires placing the process loop in manual before operations that disrupt the loop or change transmitter output; acknowledging a communicator prompt does not perform that change.

1. Confirm the signal assignment. Record the process variable, units, datum, and values assigned to 4 mA and 20 mA. Verify linear versus square-root configuration where relevant.
2. Read the actual limits. Check low/high saturation, alarm direction, and alarm-output settings. Do not substitute 3.5 mA or 20.5 mA merely because those values appear in a calculation.
3. Inspect operating status. Check for active diagnostics, fixed-current output, or a test mode. For the Rosemount 1408A, HART multidrop operation fixes current at 4 mA, so that current no longer represents changing level.
4. Compare the transmitter and receiver. Check measured loop current against transmitter-reported output, the analog input configuration, and the PLC/DCS engineering value. Preserve diagnostic status rather than allowing scaling alone to turn a fault current into a plausible level.
5. Separate scaling checks from sensor checks. An authorized current-injection test can check the receiving input and its scaling. It does not test the level sensor, hydrostatic installation, or complete measurement chain.
6. Use documented acceptance criteria. Set field tolerances from the applicable equipment specifications, test-equipment uncertainty, and site procedure—not from the decimal places shown in an ideal calculation.

If the calculation and field reading disagree, establish whether the cause is range configuration, output limiting, a diagnostic condition, or receiving-system behavior before considering adjustment. Correct arithmetic alone does not establish calibration acceptance or actual measurement accuracy.