Challenge
Northline Energy runs pumping stations across remote sites where the nearest technician can be four hours away. Their off-the-shelf motor controllers failed every few months: heat in summer, condensation in winter and constant vibration from the pumps they drive.
Each failure meant a truck, a day of lost pumping and a controller that nobody could diagnose afterwards. Northline needed hardware that simply keeps running, and that tells them when it is getting close to its limits.
Approach
We designed the power stage, the enclosure and the firmware as one system, starting from the heat. A controller that cannot get rid of its heat will fail no matter how good its components are.
Thermal path first
The power stage is bonded directly to a finned aluminium heatsink that forms the lid of the enclosure. Thermal paths were modelled before the board layout started, and the hottest components sit closest to the fins. At full load the internal temperature stays 18 °C below the rating of the weakest part.
Sealed, not ventilated
The enclosure is sealed to IP67 with no fans and no vents. Cable entries use moulded glands, and the internal wiring is routed and strapped so that nothing can rub through under years of vibration.
Firmware that reports its own health
The controller logs temperature, current ripple and supply quality, and sends a warning when trends drift, long before a fault. Northline's operators see a health score for every station instead of an alarm after the fact.
Validation
Prototypes ran through combined thermal, vibration and humidity cycling equivalent to five years in the field, then a further 3,000 hours on a load bench at full rated current. Every failure in testing led to a design change before the first unit was installed.
Outcome
The controller runs from −40 to +85 °C and is specified for a five-year service interval. Across the fleet, field failures have dropped to zero since rollout, and the health data has already caught two worn pump bearings before they failed.
