Valve Positioners: Types, Calibration and Diagnostics

The positioner is the closed-loop brain between the control signal and the stem. Without it, packing friction and pressure forces decide where the valve sits; with it, the stem goes where the controller says. This article covers how positioners work, when you actually need one, calibration steps and the diagnostic patterns every technician should recognise.

Globe control valve with positioner
Photo: Bitjungle — Wikimedia Commons, CC BY-SA 4.0

1. What a positioner does

A positioner compares the commanded position (the 4–20 mA signal, converted internally) with the measured stem position (mechanical linkage, or an electronic sensor). It then servo-controls the actuator pressure until the error is zero. In effect it wraps a local feedback loop around the valve, shielding the control loop from:

2. When a positioner is needed — and when not

SituationPositioner?
Throttling control, any accuracy expectationYes
Split-ranging (one controller, two valves)Yes — this is inherently a positioner function
Actuator pressure required differs from available signal pressure (e.g. 3–15 psi signal vs 6 bar supply)Yes
Two-position (on-off) valve with fast strokingUsually no — a solenoid is faster and simpler
Very fast loops (compressor anti-surge)Debatable — positioner response can lag; consult dynamics
Vibration-heavy serviceYes, but mount remotely and use high-gain damping

3. The three families

4. Calibration in practice

  1. Zero. With the input at 4 mA (or 3 psi), the stem must sit exactly at the rated start of travel (0 % for air-to-open). Adjust the zero/spanspring until it does.
  2. Span. Apply 20 mA; the stem must reach 100 % of rated travel. Adjust span, then re-check zero — the two interact.
  3. Linearity points. Walk 0-25-50-75-100-75-50-25-0 %; error at each point should be within the positioner spec (typically <0.5–1 % of span on smart units).
  4. Smart units: run the automatic autocal (auto-travel finding, friction characterisation). The autotune builds the stiffness/derivative set that determines the hunting behaviour later — do not skip it after packing replacement.
  5. Re-torque the packing before calibrating, not after — calibration done against a loose gland is calibration against a fiction.

5. Field diagnostics — the patterns that matter

SymptomMost likely causesFirst checks
Hunting / oscillation at steady set pointPositioner gain too high; sticky packing; process interactionStep response: slow tuning down; grease/lubricate; check stiction with small steps
Valve reaches position but "creeps"Air leak in actuator or tubing; supply too lowSoap-test fittings; watch supply gauge under stroke
Position follows command with large offset under flowUnbalanced plug force exceeding positioner authority; undersized actuatorCompare shutdown vs flowing position; verify actuator sizing
Works in one direction onlyBroken stem-to-travel pin; slipped linkage; mis-set fail directionWatch stem vs feedback linkage during stroke
4–20 mA fine, valve does nothingPositioner in manual; analog input mis-wired; current loop open at the positionerCheck positioner front panel / loop current at the device
Related reading: dead band and hysteresis definitions · positioner types in the accessories section · bench set for the actuator side.
Editorial note. Calibration sequence follows common industry practice; exact menu paths and error limits are device-specific (e.g. DVC-family, SAMSON 3730-series). Follow the manufacturer instruction manual for your model. Engineering content is educational.