Sessions that end unexpectedly are among the most frustrating faults to diagnose, because the evidence is transient and the vehicle has usually gone by the time anyone investigates.
The causes fall into a small number of categories, and the meter and status history usually distinguish them if the data was retained.
Supply conditions
Voltage outside the acceptable window, whether a sag under load elsewhere on the site or a wider supply disturbance, will cause a compliant charger to stop. This protects both the vehicle and the charger and is correct behaviour.
The signature is a stop that correlates with other site activity or with a time of day, and frequently affects several bays at once.
Thermal protection
Continuous current for hours in a sealed enclosure raises internal temperature, and direct sun or high ambient adds to it. A charger reaching its thermal limit will reduce output or stop.
The signature is seasonal, time-of-day correlated, and worse on units with more sun exposure. It also appears more on units that have been in service longer, because loosening connections generate additional heat.
Contactor and connection degradation
A contactor that no longer closes reliably, or a termination that has loosened, produces intermittent interruption under load. This is the category that worsens over time rather than appearing suddenly.
A unit whose mid-session stop rate is rising is describing a wear condition, and it is worth attending before it becomes a failure.
Vehicle-side stops
A vehicle reaching its charge limit, a departure timer, a battery management decision or a fault on the vehicle will all end the session, and none is a charger problem.
The distinguishing evidence is the reported status: a vehicle that stopped drawing while the charger continued offering energy is a vehicle-side stop, and treating it as a charger fault generates unnecessary visits.
Connectivity is rarely the cause and often blamed
A charger that loses its backend connection should continue an active session. If sessions stop when connectivity drops, that is a configuration or design position worth examining, because it turns a network event into a service failure.
Confirm whether this is the behaviour before attributing stops to connectivity, since the correlation may be coincidental.
What to capture to make this diagnosable
That last one separates a site supply event from a unit fault in a single glance, and it is free if the data is retained.
- Meter values at a fine enough interval to see the ramp down.
- Status transitions with timestamps, including which suspended state was reported.
- Any fault code and vendor-specific detail raised at the moment of stopping.
- Ambient and internal temperature where the unit reports it.
- Whether other bays on the same site stopped at the same moment.