If your robot faults on the first startup of the morning with communication or bus errors, then runs fine by mid-shift, don't chase a broken component: look for water. Condensation inside the controller cabinet is one of the most underrated causes of intermittent faults on ABB, KUKA and FANUC robots, and it rarely throws an alarm saying "I have moisture inside".
How condensation forms inside the cabinet
Plant air contains water vapour. When the temperature of an internal cabinet surface (a board, a drive heatsink, a metal terminal) drops below the dew point, that vapour condenses and liquid water appears on the electronics. This happens mainly in three situations:
- Overnight stop in a non-climate-controlled plant. The cabinet cools all night, ambient humidity rises and it condenses at dawn. That's why faults appear on the first shift and clear once the electronics warm up.
- Cold rooms, meat processing, food industry and washdown areas. Environments with permanently high relative humidity.
- Robots outdoors or plants with open doors. Sharp temperature swings and humid air ingress.
The key point: condensation doesn't need water to come in from outside. It forms from the air already inside the cabinet if the thermal swing is large enough.
Symptoms: telling moisture apart from other faults
The classic signature of a condensation problem is its correlation with the thermal cycle and the time of day:
- Faults concentrated on the first startup after a long stop (weekend, holiday, overnight).
- The robot "heals itself" after 30-90 minutes of running, with nobody touching anything.
- Intermittent fieldbus, encoder or axis-board communication errors that don't repeat consistently.
- RCD trips or earth leakage that come and go.
- Visible corrosion: green or white stains on terminals, rust on hardware, boards with dried water marks.
If the fault is constant and reproducible, it's probably a degraded component, not moisture. Humidity is treacherous precisely because it's intermittent and temperature-related. If you also see interference and odd signal behaviour, don't rule out combining it with an electrical noise and EMI problem.
What to check and in what order
Always work with the cabinet de-energised except for measurements that require it running, and follow lockout procedures. Recommended order:
- Visual inspection with a torch, cold and first thing. Open the cabinet right after the long stop, before startup. It's the best time to catch droplets or fresh condensation. Check the cabinet's inner roof, drive heatsinks, the underside of boards and the cable glands.
- Look for dried water traces. Whitish rings, incipient rust on hardware, verdigris on copper and brass. They prove repeated condensation even if there are no droplets today.
- Check the cabinet seal. Door gasket in good condition, full closure, glands tight and no cables coming in through open holes. A poorly sealed cabinet "breathes" and lets moisture in.
- Check filters and ventilation. Clogged filters stop the heat exchanger or fans from keeping a stable temperature, encouraging dew. Don't confuse this with a pure heat problem; for that, see our article on thermal management and ventilation.
- Measure temperature and relative humidity inside the cabinet. A cheap datalogger over 48-72 h tells you whether relative humidity climbs near 100 % overnight. That's the definitive proof.
- Verify the anti-condensation heater if the cabinet has one. Many controllers and cell cabinets include a heater with a thermostat/humidistat; check it exists, is powered and isn't faulty.
Common mistakes made on the shop floor
- Opening the cabinet warm in a humid plant. You let in humid air that will condense as it cools. If you must open it, keep it brief.
- Blowing with the plant air gun. Compressed air usually carries water and oil. You're injecting moisture, not drying it. Use dry air or vacuum.
- Blocking vents or filters "to keep dirt out". You kill the designed airflow and lose control of internal temperature.
- Disconnecting the anti-condensation heater because it "draws power" or "heats up". That's exactly what prevents dew during stops.
- Wiping with a rag and closing up without finding the cause. The droplet returns the next night. You have to attack the root: sealing, heater or ambient climate.
How to prevent condensation for good
- Keep the anti-condensation heater working with its humidistat/thermostat. It's the most effective measure in non-climate-controlled plants.
- Seal the cabinet: door gasket in good condition, correct glands, no cables through open holes. A well-sealed cabinet condenses far less.
- Don't let the cabinet go fully cold all night if the environment is critical: standby electronics or the heater keep the temperature above the dew point.
- Clean filters and checked ventilation within your preventive maintenance plan. Schedule the check on a rough quarterly or half-yearly basis depending on how dirty the environment is.
- Desiccant cartridges as a temporary measure in critical cabinets, knowing they saturate and need replacing.
- In very harsh environments (washdown, cold, food), consider cabinets with a higher protection rating or active climate control. Here an environment audit beats patching things up.
When to stop and call a specialist
Stop immediately and don't force the startup if you find liquid water on boards or drives, if there are repeated RCD trips, or if you see advanced corrosion on power connectors or terminals. Starting up with wet electronics can turn a cheap problem into replacing an axis board or a power module. Dry it out properly, identify the cause and, if there is corrosion or affected components, contact technical service to assess repair or a module exchange before resuming production.