Transmitter linearity testing verifies that the output signal of a process transmitter remains proportional to the measured input across its entire operating span, ensuring measurement accuracy, process safety, and regulatory compliance. Performed in accordance with recognized standards such as IEC 61298, IEC 60770, and the GB/T 17614 series, linearity testing quantifies the deviation from an ideal straight-line response and is essential for industries ranging from oil and gas to power generation. This article explains the technical principles, sample types, test items, procedures, and acceptance criteria involved in transmitter linearity evaluation.
Technical Overview
A transmitter is an industrial-process instrument that converts a measured physical quantity—such as pressure, temperature, level, or flow—into a standardized output signal, most commonly a 4–20 mA current loop, a 1–5 V voltage signal, or a digital communication signal. In an ideal transmitter, the relationship between the input and the output is a straight line over the entire calibrated span.
Linearity (also referred to as nonlinearity or conformity error) quantifies how closely the actual input–output characteristic follows a specified reference straight line. Linearity error is defined as the maximum deviation between the measured calibration curve and that reference line, expressed as a percentage of the output span. Depending on how the reference line is established, three definitions are commonly applied:
- Terminal-based (end-point) linearity: the reference line connects the output values at 0% and 100% of the input span; this is the most widely used criterion for pass/fail verification in industrial calibration practice.
- Least-squares (best-fit) linearity: the reference line is fitted to all calibration points by minimizing the sum of squared deviations; this method yields the smallest numerical nonlinearity and is frequently used for factory characterization.
- Independent linearity: the straight line is positioned to minimize the maximum deviation; it produces the lowest possible linearity error but requires more complex computation.
Transmitter linearity testing is governed by internationally recognized standards, including IEC 61298-2 (test methods under reference conditions for process measurement and control devices), IEC 60770-1 (methods for performance inspection of transmitters), the corresponding EN 61298 series adopted in Europe, and the GB/T 17614 series covering transmitters for industrial-process control systems.
Testing Samples
Linearity testing applies to virtually every category of process transmitter. Typical sample types submitted to accredited third-party laboratories and in-house calibration facilities include:
Pressure Transmitters
Gauge, absolute, and differential pressure transmitters, including smart transmitters with digital communication capabilities. Differential pressure transmitters used for flow and level measurement are among the most frequently tested devices, because any linearity error propagates directly into the accuracy of the computed flow or level value.
Temperature Transmitters
Head-mounted and rail-mounted temperature transmitters accepting thermocouple or RTD inputs. Linearity is verified either by applying mV or resistance signals that simulate the sensor input, or by testing the complete measuring assembly in a precision temperature bath or dry-block calibrator against a reference thermometer.
Level, Flow, and Multivariable Transmitters
Hydrostatic level transmitters, electromagnetic and vortex flow transmitters, and multivariable transmitters that compute compensated flow or density values. For these devices, the linearity of both the primary sensing element and the internal signal-conditioning chain is evaluated over the full operating range.
Testing Items
A linearity test campaign typically covers the following performance characteristics, all of which can be derived from the same set of calibration data:
- Linearity error: the maximum deviation of the output from the reference straight line, expressed in % of span or % of reading.
- Hysteresis: the maximum difference between upscale and downscale outputs observed at the same input value.
- Repeatability: the scatter of repeated readings at the same input when approached from the same direction.
- Zero error and span error: the deviations recorded at 0% and 100% of the measurement range.
- Output conformity: the agreement of the 4–20 mA, 1–5 V, or digital output with the nominal signal scaling.
- Combined accuracy: the root-sum-square or worst-case combination of linearity, hysteresis, and repeatability, compared against the manufacturer's rated accuracy.
Testing Methods
Test Conditions and Equipment
Tests are performed under reference ambient conditions, typically 23 °C ± 2 °C and 45–75 % relative humidity, with the device under test (DUT) energized and allowed to thermally stabilize before data collection. Traceable reference standards are required, such as a piston gauge (deadweight tester) or precision pressure controller for pressure transmitters, a temperature bath or dry-block calibrator with a calibrated reference probe for temperature transmitters, and a calibrated digital multimeter for output current measurement. The uncertainty of the reference standards should maintain a test uncertainty ratio (TUR) of at least 4:1 relative to the DUT specification.
Test Procedure
A minimum of five test points—0 %, 25 %, 50 %, 75 %, and 100 % of span—is applied, with nine or eleven points preferred for high-resolution characterization. The recommended sequence is: exercise the transmitter over the full range, approach the first test point from below, record the output at each point while ascending to full scale, then descend through the same points in reverse order. At least three complete up/down cycles are performed so that repeatability and hysteresis can be evaluated from the same data set.
Data Reduction and Acceptance Criteria
For each test point, the mean output value is calculated. The reference straight line is established using the end-point or least-squares method, and the linearity error is computed as: Linearity error (%) = (maximum deviation from the reference line / output span) × 100 %. Hysteresis is calculated as the maximum upscale–downscale difference divided by the span. Results are compared with the manufacturer's specification—typically ±0.075 % to ±0.2 % of span for high-accuracy smart transmitters and ±0.5 % for general-purpose models. If the measured error exceeds the limit, zero and span adjustments are performed and the full test sequence is repeated. The final report documents the test conditions, the reference standards and their traceability, the raw data, the calculated errors, the measurement uncertainty, and a clear pass/fail conclusion.
Applications
Transmitter linearity testing supports quality assurance, regulatory compliance, and metrological control across many industrial sectors:
- Factory quality control and type testing of new transmitters for manufacturers, performed in accordance with IEC 60770-1.
- Periodic calibration programs in refineries, chemical plants, power stations, and water treatment facilities.
- Custody transfer and fiscal metering applications, where linearity directly affects billing accuracy.
- Verification of transmitters used in safety instrumented systems (SIS) before commissioning and during scheduled proof tests.
- Incoming inspection of newly purchased instruments and post-repair verification of reworked transmitters.
- Independent verification by accredited third-party laboratories in support of tenders, audits, and regulatory submissions.
FAQ
What is transmitter linearity?
Transmitter linearity is the degree to which the output signal follows a straight-line relationship with the applied input across the calibrated span. It is quantified as the maximum deviation from a specified reference line, expressed as a percentage of the output span.
How is linearity different from hysteresis and repeatability?
Linearity describes deviation from the ideal straight line at any point of the span; hysteresis describes the difference between readings approached from below and from above at the same input; repeatability describes the scatter of repeated measurements under identical conditions. All three characteristics are derived from the same multi-point, multi-cycle calibration data set.
What is an acceptable linearity error?
The acceptance limit is the manufacturer's rated accuracy. High-accuracy smart pressure transmitters commonly specify ±0.075 % of span, standard industrial models ±0.1 % to ±0.2 %, and general-purpose transmitters ±0.5 %. Any result exceeding the specification warrants adjustment, repair, or reclassification of the instrument.
How often should linearity be tested?
Most industrial calibration programs verify linearity annually, although critical applications such as custody transfer or safety-related service may require semi-annual or quarterly verification. New, repaired, or relocated instruments should always be tested before being returned to service.
Which standards apply to transmitter linearity testing?
The principal references are IEC 61298-2 for test methods under reference conditions, IEC 60770-1 for transmitter performance inspection, the corresponding EN 61298 series, and the GB/T 17614 series for transmitters used in industrial-process control systems.
Can a transmitter with poor linearity be corrected?
Smart transmitters apply digital linearization and sensor trimming that can substantially compensate for nonlinearity when the error is stable and repeatable. However, digital compensation cannot remove hysteresis or non-repeatable errors, and sensing elements that have drifted mechanically typically require repair or replacement.