The electric actuator on a power plant's deaerator developed an unusual but highly instructive problem: whenever someone talked on a walkie-talkie nearby, the actuator was disturbed — the feedback received by the DCS fluctuated, and the actuator on site even moved by itself. The unit was extremely simple, with only six terminals for power, command, and feedback. The maintenance crew tried shielding several times without any effect, and even the manufacturer failed to fix it.
The value of this case lies in a reproducible test method and a set of countermeasures with verified effectiveness:
First, run comparison tests to locate the interference path. One I&C engineer conducted dedicated experiments with a FLUKE source and a FLUKE multimeter: injecting a standard current signal while disturbing it with a handheld radio at close range, then comparing suppression methods one by one. The conclusions were very specific: a ferrite ring wrapped around the walkie-talkie antenna worked best; clamping ferrites onto the cable was somewhat less effective; twisted-pair wiring also helped. Ordinary TDK anti-interference ferrite rings cost only a few yuan — use enough of them to envelop the cable as completely as possible.
Administrative measures work immediately: during calibration, one person operates while the other communicates with the control room from at least one meter away; in daily operation, keep radios away from actuators.
Circuit-level measures:
- Check the shield grounding of command/feedback cables — ground at one end only (normally the control-room side); grounding both ends introduces a ground-loop interference;
- Install an active signal isolator on the signal loop — note "active": passive isolators are too weak to help much;
- Twisted pair, shielding, ferrites, and isolators usually need to be combined.
The discussion converged on one judgment: the actuator's internal circuitry simply had poor RF immunity — a design/selection defect. Both units from that purchase batch showed the same behavior, several layers of shielding did nothing, and even a replacement main board from the manufacturer did not cure it. Similar lessons are not isolated: one plant used a high-precision split-type positioner in a retrofit project — its internal circuitry was so sensitive that even welding nearby disturbed the feedback, and the problem was only solved by switching back to a proven common model. The peers' conclusion was blunt: power plant selection should favor mature, reliable, stable products — do not become a test bed for new products.
1. Keep walkie-talkies away from actuators (over one meter during calibration);
2. Ferrite rings: on the disturbance source's antenna first, and around the affected cables as well;
3. Verify single-end shield grounding; convert signal lines to twisted pair;
4. Add active signal isolators;
5. Ask the manufacturer for a main board with proper RF immunity;
6. If all fails and the problem is batch-wide, replace the model without hesitation.
RF interference (radios, walkie-talkies, welders) causing feedback fluctuation or misoperation is a combined problem of "sensitive device + disturbance source on site." The order of attack: administration (distance) first, suppression (ferrites/grounding/twisted pair) second, isolation (active isolator) third, hardware replacement (board/model) last. Shielding that does not work does not mean the interference cannot be beaten — with the right method, a ferrite ring worth a few yuan may outperform a major retrofit.
