An IQ-series Rotork electric actuator (380 V) served the scoop tube of a feed-water pump on a 200 MW unit for automatic water-level control. After some time in service, the automatic loop could no longer be kept in service: for every 1% the controller outputted, the actuator moved 2–3%, the same in manual remote. Raising the dead band from 12% to 10% held for only two days; replacing the proportional control board did not cure it; tightening the dead band to 8% only induced hunting. The actuator worked long-term in the 85–98% opening range, and the fault paralyzed level control.
The troubleshooting took some wrong turns worth learning from:
- Dead band 12% → 10%: temporary relief for two days, then relapse — meaning the root cause was not the dead-band value itself;
- Replacing the proportional control board: no fundamental improvement — board hardware was ruled out;
- Tightening the dead band to 8%: more hunting — the wrong direction entirely.
The manufacturer's engineers already knew this problem and issued a clear, official remedy:
Short the actuator terminals 4-24-40 to 5-X-39, and set the dead band to 01 (0.1%) — the problem is then resolved. In other words, this application requires a special terminal wiring arrangement combined with an extremely small dead band; ordinary tuning can never reach the correct state.
The discussion added several application-specific points for scoop-tube control:
1. Dead-band rule of thumb: such control actuators are normally set around 1%; 12% is clearly oversized — itself one source of the "1% in, 2–3% out" behavior;
2. Signal calibration: the 4–20 mA command-to-position calibration must be performed in the second-level menu (e.g., 4 mA corresponding to the low position); with wrong calibration the input-to-motion ratio is wrong by definition;
3. Application fit: with the scoop tube working long-term at 85–98% opening, verify the control characteristics of that range against the coupling's characteristics (e.g., whether above 80% the scoop is already effectively "full open"), together with the mechanical crew;
4. Model fit: peers also suggested that for scoop-tube duty the IQM model (switch-type, negative-pole control) may be a better fit, avoiding the adaptation problem at the selection stage.
For gain anomalies like "1% commanded, 2–3% moved," check three things first: whether the terminal wiring matches the manufacturer's application scheme, the dead-band setting, and the 4–20 mA calibration. The core lesson: special-application configurations of imported actuators (terminal shorting and similar "factory knowledge") never appear in the general manual — for hard problems, ask the manufacturer for an application-specific solution, which beats repeated trial and error, swapped boards, and dead-band guesswork by a wide margin.
