Home › Blog › Valve Positioner Calibration: A Step-by-Step Field Guide
A positioner that is out of calibration sends the actuator to the wrong place. This guide walks through what calibration actually does, what you need before you start, and the five steps that cover most field work.
The question we hear most often is simple: why does the valve not reach the position the controller asked for?
The answer is usually that the positioner’s internal map — the relationship between the control signal it receives and the actuator pressure it sends out — no longer matches the valve it is bolted to. Calibration rebuilds that map. It does three things and nothing else:
• It learns the zero point (valve fully closed) and the span point (valve fully open).
• It records the travel direction, so a rising signal moves the valve the way the process needs.
• It stores a characterisation curve so that 50% signal really is 50% travel on this specific valve, not on a nominal one.
That is the whole job. If a positioner is hunting, oscillating or sitting at the wrong position after calibration, the fault is usually mechanical — linkage wear, air supply, or a sticking stem — not the calibration itself.
Calibration on a valve that is not mechanically sound simply teaches the positioner to memorise a fault. Check these first:
Step 1 — Isolate. Put the process in a safe state and confirm the valve is not carrying live flow. On a fail-safe actuator, be clear about what happens when air or power is removed.
Step 2 — Put the loop in manual. Take the controller off automatic so it cannot fight you while the positioner is learning its endpoints.
Step 3 — Run the auto-calibration. Most digital positioners have a single auto-calibrate command. It drives the valve to both mechanical stops, measures travel, and stores the endpoints. Let it finish without interrupting it — an aborted auto-calibration can leave a partial curve in memory.
Step 4 — Verify at three points, not one. Command 0%, 50% and 100% and read back the actual travel. A positioner that is correct at the ends and wrong in the middle has a characterisation problem rather than an endpoint problem.
Step 5 — Check the fail action. Remove the signal and confirm the valve goes to the position your safety case requires. This is the step most often skipped in the field, and the one that matters most.
Hunting or oscillation. Almost always too much gain for the actuator volume, or a leaking actuator diaphragm. Reduce gain, then check for leaks.
Slow response or a dead band. Restricted air path — a clogged filter, a kinked tube, or an undersized supply line. On a large actuator, a small-bore supply tube will visibly limit stroke speed.
Position drifts over hours. A leaking spool valve or a worn stem packing that is changing friction as the process temperature moves. Watch the actuator pressure gauge with the signal held steady.
Consistent offset in one direction. Lost motion in the linkage, or a feedback arm fitted one spline out. Re-check the mechanical zero before recalibrating.
Calibrate on commissioning, after any actuator or linkage work, after a stem or packing replacement, and whenever the travel read-back drifts by more than about 2% of span. Recalibrating on a fixed calendar schedule with no symptom is usually wasted effort; recalibrating after mechanical work is not optional.
On commissioning, after any actuator, linkage, stem or packing work, and whenever the reported travel drifts by more than roughly 2% of span. A fixed annual interval with no symptom is less useful than event-driven calibration.
Hunting is a control-loop problem, not a calibration problem. The usual causes are excessive gain for the actuator volume, a leaking diaphragm or tubing, or a sticking stem. Reduce the gain first, then leak-test the actuator.
Usually yes, if the process can tolerate the valve moving through its full stroke. You must still put the controller in manual and be certain of the fail-safe action before you start. If the valve is in a safety loop, follow your plant’s permit procedure.
Auto-calibration learns the endpoints and travel direction — the geometry of the valve. Auto-tuning adjusts the control dynamics — how aggressively the positioner responds. They are separate operations and are often confused. Calibrate first, then tune.