Transmitter & Loop Calibration: How to Calibrate a 4-20mA Loop

Calibration is the one task that separates an instrument that reads correctly from one that merely reads. A transmitter left unchecked for years can drift silently — reporting 82 °C when the process is at 87 °C, or showing 3.4 bar when actual line pressure is 3.1 bar. In a well-run plant those errors accumulate into off-spec product, safety margin erosion, or a regulatory non-conformance. Calibrating a transmitter — and verifying the entire signal chain through to the PLC — catches those errors before they cost anything.

This guide walks through every step: the terminology, the 5-point procedure, the as-found/as-left discipline, the difference between bench and in-situ loop calibration, and the controls-engineer perspective of confirming the scaled value at the HMI matches what you applied at the transmitter.

---

Why Transmitters Drift and Why Calibration Matters

Every sensor has a physical measurement element — a Bourdon tube, a thermocouple junction, a differential-pressure capsule, a strain gauge bridge. Over time, mechanical stress, thermal cycling, process contamination, and electronic component ageing shift the relationship between the physical input and the electrical output. This is called drift.

Drift is not a failure; it is an expected, predictable property of every measurement device. Manufacturers publish a drift specification, typically expressed as a percentage of span per year. A transmitter rated at ±0.05% of span drift per year will wander up to 0.05% of its configured range over twelve months under normal conditions. In many applications that is acceptable. In custody transfer, pharmaceutical batch manufacturing, or safety-instrumented systems, it is not.

Three reasons demand a calibration programme:

1. Process accuracy — Incorrect readings produce incorrect control. A flow transmitter that reads 4% high causes a flow controller to under-deliver product.
2. Safety — In a safety-instrumented loop, a transmitter that reads low may prevent a shutdown that should fire.
3. Regulatory compliance — FDA 21 CFR Part 11, ISO 9001 quality systems, ATEX/IECEx requirements for hazardous areas, and many industry-specific standards require documented calibration records at defined intervals.

Understanding the 4-20mA current loop that carries the signal is a prerequisite for any calibration work — a miscalibrated transmitter and a wiring fault produce similar symptoms; you need to know which you are fixing.

---

Calibration Terms Every Technician Needs

Before touching a calibrator, every member of the team should share the same vocabulary. Ambiguous terminology causes errors.

Term	Definition
Zero (LRV)	The Lower Range Value — the process input that produces 4 mA. Example: 0 bar produces 4 mA.
Span (URV)	The Upper Range Value — the process input that produces 20 mA. Example: 10 bar produces 20 mA.
Range	The difference between LRV and URV. In the example above, range = 10 bar.
Turndown	The ratio of the maximum range to the configured range. A transmitter with a maximum range of 0–100 bar configured for 0–10 bar has a turndown of 10:1. High turndown amplifies percentage error.
Accuracy	The maximum allowable deviation from true value, stated as ±% of span, ±% of URL (upper range limit), or ±% of reading.
Repeatability	The ability of the transmitter to return the same output for the same input applied multiple times under identical conditions.
Hysteresis	The difference in output between an upscale reading and a downscale reading at the same input value.
As-found	The transmitter's measured performance before any adjustment is made during a calibration visit.
As-left	The transmitter's measured performance after any adjustment. As-left must meet the acceptance tolerance.

Zero and span are the two primary adjustments available on any transmitter. Zero shifts the entire output curve up or down. Span changes the slope of the curve — it changes the output per unit of input. On a conventional analog transmitter these are hardware potentiometers. On a smart transmitter they are digital parameters adjusted through a HART communicator or configuration software.

---

The 5-Point Calibration Procedure

The 5-point calibration is the industry standard for verifying transmitter linearity. It tests the device at 0%, 25%, 50%, 75%, and 100% of the configured span, both ascending and descending. Running both directions checks for hysteresis.

Step 1 — Gather the Calibration Record and Tolerances

Pull the instrument's calibration record, data sheet, and the applicable tolerance class. Write down:

• Tag number
• Transmitter make, model, and serial number
• Configured LRV and URV
• Acceptable tolerance in mA (e.g., ±0.16 mA = ±1% of 16 mA span)
• Date of last calibration and as-left values

Step 2 — Apply the Known Inputs

Using a reference calibrator, apply each test point to the transmitter input:

Test Point	% Span	Applied Input (example: 0–10 bar)	Expected Output
0%	0	0.000 bar	4.000 mA
25%	25	2.500 bar	8.000 mA
50%	50	5.000 bar	12.000 mA
75%	75	7.500 bar	16.000 mA
100%	100	10.000 bar	20.000 mA

Measure the actual output current at each point. Record these values — this is the as-found data.

Step 3 — Record As-Found Results

As-found data is non-negotiable. Even if the transmitter is perfectly within tolerance, record the measured values before touching any adjustment. This data tells you:

• Whether the transmitter passed on arrival (useful for audit trails and failure analysis)
• Whether drift is trending in one direction, suggesting a root cause investigation is needed
• Whether a previous as-left record is consistent with today's as-found (a large jump could indicate process contamination, mechanical damage, or tampering)

Step 4 — Adjust If Required

If any as-found reading exceeds the acceptance tolerance, adjust zero and span:

1. Apply 0% input (0 bar in the example). Adjust zero until the output reads 4.000 mA.
2. Apply 100% input (10 bar). Adjust span until the output reads 20.000 mA.
3. Recheck 0% — span adjustment usually perturbs zero slightly. Re-trim zero as needed.
4. Repeat until both endpoints are within tolerance.

For a linear transmitter, correcting the endpoints corrects the midpoints. If midpoint errors remain after endpoint adjustment, there may be a sensor linearisation fault that zero/span cannot fix.

Step 5 — Record As-Left Results

Repeat the full 5-point ascending and descending sweep. Record every reading. These are the as-left values. Sign and date the record. All as-left values must be within the acceptance tolerance before the instrument returns to service.

---

As-Found vs As-Left: The Calibration Discipline

The as-found / as-left discipline is not optional paperwork. It is the core of a defensible calibration programme.

As-found answers: "Was this instrument in tolerance when we arrived?" If yes, no adjustment is required — record that it passed and move on. Making unnecessary adjustments introduces risk: you might accidentally make a transmitter worse.

As-left answers: "Is this instrument in tolerance when we leave?" Every instrument returned to service must have a passing as-left record.

Together, the as-found and as-left records over multiple calibration cycles tell you:

• Drift rate — how fast is this instrument moving over time?
• Calibration interval fitness — is the current interval too long (instruments frequently out of tolerance on arrival) or too short (instruments always in tolerance, interval could be extended)?
• Root cause evidence — a sudden large as-found error that was not present at the last calibration points to an event (process overpressure, temperature excursion, electrical transient) rather than normal drift.

---

Tools for Transmitter Calibration

Loop Calibrator

A loop calibrator is the primary tool. A modern unit in one device can:

• Source a known current (4–20 mA) to simulate a transmitter output for testing downstream devices
• Measure the loop current while the transmitter is operating (series connection into the loop)
• Source a known process variable input (pressure, temperature, frequency) to the transmitter under test

Calibrators certified to a higher accuracy class than the transmitter under test (typically 4:1 test accuracy ratio or better) are required for traceable calibration.

HART Communicator

Most modern transmitters are HART-capable. A HART communicator (handheld or PC-based) connects across the loop wires and provides access to:

• Zero and span digital trim
• Sensor trim (offset and gain adjustments applied before the 4–20 mA output)
• Process variable readback in engineering units
• Diagnostic information (sensor temperature, loop integrity, self-test results)
• Configuration data (tag, LRV, URV, damping, units)

For a smart transmitter with digital trim, HART calibration is faster and more repeatable than analog potentiometer adjustment.

Decade Resistance Box

When bench-calibrating a temperature transmitter that accepts a resistance input (RTD), a decade resistance box substitutes known resistance values for a Pt100 or Pt1000 sensor. This is more accurate and safer than immersing the sensor in a reference bath for routine calibration.

Pressure Hand Pump / Dead Weight Tester

For pressure transmitters, the reference input is a known pressure. A hand pump with a calibrated reference gauge covers most routine work. A dead weight tester applies traceable weights to a piston area for the highest accuracy level required in primary standard work.

Reference Standards

All calibration standards must be traceable to a national metrology institute (NIST, NPL, PTB, etc.) through an unbroken chain of comparison. Calibration certificates must document this traceability.

---

Bench Calibration vs In-Situ Loop Calibration

These are two different tasks with different scopes.

Bench Calibration

The transmitter is removed from the process and taken to a calibration workshop. A reference input (pressure, temperature simulation, resistance) is applied directly to the transmitter's sensing element. The output is measured directly from the transmitter terminals.

What it checks: The transmitter itself — sensor, electronics, output circuit.

What it does not check: The field wiring, junction boxes, intrinsic safety barriers, isolators, or the PLC analog input card. A transmitter can bench-calibrate perfectly and still produce a wrong PLC reading because of a fault elsewhere in the loop.

Advantages: Controlled environment, best accuracy, no process disturbance.

Disadvantages: Requires pulling the instrument, process isolation, and re-installation; loop integrity after reinstallation is not verified.

Loop Calibration (In-Situ)

The transmitter stays in place. A precision current source is injected at a convenient loop break point (often at a junction box or marshalling panel) to simulate transmitter output at known values, while the downstream reading (at the PLC or HMI) is observed and recorded.

Some technicians prefer to apply the process variable input to the installed transmitter using a portable reference (a hand pump for pressure, a temperature bath or dry-block for temperature), then read the loop current at the panel and verify the PLC value simultaneously.

What it checks: The complete loop from the injection point to the PLC display.

What it does not check independently: The transmitter sensor itself (when injecting current at the loop, the sensor is bypassed).

Advantages: Verifies the full signal chain including wiring and PLC card; no process interruption if the injection is done at a loop test point; fast.

Disadvantages: Does not independently verify the transmitter's sensor-to-current conversion.

Best practice is a combination: bench-calibrate the transmitter on its calibration schedule, and perform an in-situ loop check after reinstallation — and as a standalone verification when troubleshooting suspect readings.

---

Verifying the Whole Loop to the PLC and HMI

This is the controls view — and it is frequently skipped. A transmitter that reads correctly at the terminal strip can still produce a wrong engineering-unit value at the HMI if the PLC's analog input scaling is configured incorrectly.

The Scaling Path

For a 4–20 mA analog input, the PLC performs two conversions:

1. Current to raw count — The analog input card converts 4–20 mA to a raw integer count (commonly 0–4095 for a 12-bit card, or 0–32767 for a 15-bit card). The exact mapping depends on the card manufacturer. See PLC analog input scaling for the full conversion mechanics.

2. Raw count to engineering units — A scaling block in the PLC program converts the raw count to the engineering-unit value (bar, °C, m³/h). This block has four configurable parameters: raw low, raw high, EU low, EU high. An error in any of these — a typo, a unit mismatch, a wrong range — produces a display value that is systematically wrong across the entire range.

The End-to-End Verification Procedure

1. Inject 4 mA at the loop test point (or apply 0% process input to the transmitter). Note the raw count value in the PLC and the engineering-unit value on the HMI.

   • Expected raw count: matches card's 4 mA count (e.g., 819 for a 0–4095 card using 1–5 V range, or 0 for a card ranged to 4–20 mA)
   • Expected EU value: equals the configured LRV (e.g., 0.0 bar)

2. Inject 12 mA (50% of span). Note raw count and EU value.

   • Expected EU value: midpoint of configured range (e.g., 5.0 bar on a 0–10 bar range)

3. Inject 20 mA (100%). Note raw count and EU value.

   • Expected EU value: equals the configured URV (e.g., 10.0 bar)

4. Compare all three against calculated expected values. Any systematic offset (EU reads 0.5 bar high at every point) indicates a scaling parameter error. A non-linear error (off at midpoint but correct at endpoints) may indicate a raw count mapping error.

Common Scaling Errors to Catch

• LRV/URV transposed — The EU value reads backwards (high when low, low when high)
• Wrong raw count range — Card outputs 0–4095 but scaling block configured for 0–32767, causing a 7× gain error
• Unit mismatch — Transmitter calibrated in bar, PLC scaling block configured for kPa, with no conversion factor
• Damping mismatch — Transmitter has high damping set; PLC reading lags a long time after the injected current changes (not a scaling error, but it confuses the test if not accounted for)

Confirming the scaled PLC value matches the applied input at 0%, 50%, and 100% of span is a two-minute check that has prevented countless process upsets caused by control systems acting on systematically wrong measurements.

---

Calibration Interval and Records

Choosing the Interval

There is no single correct calibration interval. The interval should be based on:

• Manufacturer's drift specification — Provides the theoretical maximum interval if drift alone limited accuracy
• Historical as-found data — The most reliable basis; analyse your as-found records to find the actual drift rate in your process conditions
• Criticality — Safety-critical instruments (SIS, custody transfer) warrant shorter intervals regardless of drift data
• Process conditions — High-vibration, high-temperature, or corrosive environments increase drift rates

Common intervals in practice range from three months for critical SIS instruments to three years for low-criticality utility measurements. Industry standards such as IEC 61511 for safety instrumented systems and OIML regulations for custody transfer specify minimum requirements that override engineering judgement.

What Calibration Records Must Contain

A defensible calibration record includes:

• Instrument tag number, make, model, serial number
• Calibration date and technician name
• Calibration standard identification and traceability certificate reference
• Calibration procedure reference
• Configured LRV and URV at time of calibration
• Acceptance tolerance
• As-found readings at each test point (ascending and descending)
• Pass/fail determination for as-found
• Any adjustments made, and the method
• As-left readings at each test point
• Pass/fail determination for as-left
• Next calibration due date

Records should be retained for at least the life of the instrument, or longer if regulatory requirements demand it.

---

Frequently Asked Questions

How do you calibrate a transmitter?

Apply a series of known, traceable inputs to the transmitter's sensing element (pressure, temperature, resistance) and measure the resulting 4–20 mA output at each point. Compare each measured output to the expected output. Record these as-found values. If any point exceeds the acceptance tolerance, adjust zero (for the 0% endpoint) and span (for the 100% endpoint), then re-run the full 5-point sweep and record as-left values. All as-left readings must fall within tolerance before the instrument returns to service.

What is zero and span in transmitter calibration?

Zero is the transmitter adjustment that sets the 4 mA output point — the process input value that corresponds to the bottom of the range (LRV). Span sets the 20 mA output point — the process input that corresponds to the top of the range (URV). Zero shifts the entire output curve up or down; span changes its slope. On analog transmitters these are physical potentiometers. On smart transmitters they are digital trim parameters accessible via HART.

What is as-found and as-left in calibration?

As-found is the transmitter's measured performance before any adjustment is made during a calibration event. As-left is the measured performance after any adjustment. Recording both is mandatory: as-found tells you whether the instrument was in tolerance when you arrived (and drives decisions about calibration interval), as-left confirms it is in tolerance when you leave. Never adjust a transmitter without recording as-found first.

What is loop calibration?

Loop calibration (also called in-situ calibration) verifies the entire 4–20 mA signal chain — transmitter, field wiring, junction boxes, barriers, isolators, and PLC analog input card — as a complete system. A precision current source is injected at a point in the loop (often at a junction box) to simulate transmitter output at known values, while the receiving device (PLC or HMI) is read and recorded. Loop calibration catches wiring and scaling errors that bench calibration of the transmitter alone cannot detect.

---

Summary

Transmitter and loop calibration is a disciplined process, not a one-step adjustment. The 5-point procedure — applied ascending and descending, with as-found recorded before any touch and as-left recorded after — produces a traceable, defensible record that supports process quality, safety, and regulatory compliance.

The controls engineer's job does not end at the transmitter terminals. Injecting a known current and confirming the PLC and HMI show the correct engineering-unit value at 0%, 50%, and 100% of span is the only way to verify the complete measurement chain. Scaling errors are silent, systematic, and entirely preventable with a two-minute end-to-end check.

For the electrical fundamentals behind the signal itself, start with the 4-20mA current loop guide. For the pressure sensing element inside the transmitter, see pressure transmitter explained. For scaling the raw ADC count into engineering units inside the PLC, see PLC analog input scaling.