A charger that has stopped, trips intermittently, or has quietly become slower is usually not a dead charger. Diagnosis starts at the circuit, because that is where the fault normally is.
Heat cycling is the local failure mode
Every metro has a characteristic way that EV circuits fail. On the coast it is corrosion. In the north it is freeze-thaw and moisture ingress. Here it is heat.
A garage in Central Texas swings through a wide daily temperature range for months, and an attic-routed conductor spends August far above the outdoor reading. Metal expands and contracts. A termination that was torqued correctly on the day gradually relaxes. Resistance rises at the joint, which produces heat, which relaxes it further.
The symptom set is recognisable: a breaker that held fine through two winters and starts tripping in July, charging that becomes intermittent, or in the worse cases discolouration at a connection point. The cause is a joint, not a circuit board.
So we check and re-torque terminations as a first move, and thermal-image them where an intermittent fault points that way. It is unglamorous and it resolves a large share of the calls we get.
Throttling is not always a fault
A charger that has become slower on hot afternoons is frequently doing exactly what it was designed to do. Units protect themselves by reducing output as internal temperature rises, and one mounted on an unshaded west-facing wall in this climate will hit that point most summer days.
That is a siting problem with a siting fix — shade, a small relocation, occasionally a different mounting position on the same wall. It is worth distinguishing from a genuine fault before anyone spends money on replacement hardware, because the replacement will do the same thing in the same position.
Nuisance trips and ground faults
Where a receptacle-fed installation trips without an obvious cause, two suspects come first: the interaction between the receptacle’s required GFCI protection and the charger’s own internal ground-fault detection, and moisture ingress at an outdoor enclosure after a spring storm.
Both are diagnosable. Neither means the charger has failed.
Measuring rather than guessing
Voltage gets measured under load, not at rest — a circuit can read perfectly at no load and collapse under forty amps, and the no-load reading is what makes a marginal connection look healthy. Insulation resistance and ground continuity are verified. Conductor sizing is checked against the actual run length, because a fault that appears years later sometimes turns out to be an original undersizing that only shows up once the weather is against it.
Installations done by someone else
We assess them without comment about who did the work, including installs that were never permitted. You get a plain account of what is there, what is wrong, and what putting it right involves — including the route to getting previously unpermitted work inspected, which is a far better position than leaving it to surface during a sale or a claim.