PV fault codes

How to interpret an inverter alarm safely without assuming that the same code means the same fault across manufacturers.

A PV fault code is a manufacturer and model-specific diagnostic message. The same number can mean different things on two inverters, and one fault can produce several secondary alarms. Use the manual or service information for the exact model and firmware.

The safe first response is to preserve the evidence, check for immediate danger and follow the approved user procedure. It is not to open an inverter, disconnect live DC plugs or change grid-protection settings.

Capture the complete message

Record:

exact code and full text
inverter make, model and serial number
date, time and whether the alarm clears after dark
PV power, grid voltage and battery status shown
recent power cut, storm, roof work or settings change
photographs of the display and any visible damage
local display status as well as the cloud app

A cloud portal can lag, lose communications or translate a code differently. The local event log is often the better evidence for the service engineer.

Common fault families

Grid voltage or frequency

The inverter monitors the public supply and must disconnect if conditions fall outside its permitted settings. Causes can include a network event, loss of supply, a loose AC connection or voltage rise within the installation during export.

Record whether neighbours lost power and whether the fault occurs only at high PV output. Do not widen the voltage or frequency limits. Grid settings form part of the G98 or G99 connection compliance.

DC insulation or earth fault

An insulation alarm can be caused by damaged cable, moisture, a module or junction-box fault, incorrect wiring or another path from the array to earth. It requires PV-specific testing.

Do not repeatedly reset it or touch module frames and wiring to look for a tingle. DC voltage can remain present whenever modules are illuminated.

Arc-fault indication

Where arc detection is fitted, an alarm can indicate a potentially dangerous discontinuity or connection problem. Treat it as a fault to investigate, not a nuisance feature to disable.

Loose or mismatched connectors, damaged conductors and failed terminals can arc. A competent contractor should follow the product’s diagnostic and isolation procedure before restart.

Overtemperature or fan fault

Blocked airflow, high ambient temperature, failed cooling components or excessive electrical stress can make the inverter derate or stop. Clear only owner-accessible external obstructions while the equipment remains closed.

Required clearances and serviceable filters or fans come from the exact manual. Do not remove covers or direct compressed air into energised equipment.

String, optimiser or MPPT fault

Low or missing input can follow shade, an open circuit, connector fault, failed optimiser, polarity issue or internal inverter problem. Compare channels under similar conditions and use the array layout.

A low string reading is not enough to instruct an owner to unplug modules. Diagnosis can require live and dead tests by a competent PV person.

Residual-current or RCD event

The inverter’s internal monitoring and an external RCD are different protective layers. A trip can relate to leakage, insulation, equipment failure, surge or an unsuitable protective-device arrangement.

Record which device operated. Do not replace an RCD with another type or rating without a design assessment.

Communications fault

Wi-Fi or portal loss can stop monitoring while generation continues. Check the local inverter power and meter before assuming an electrical outage.

Communications faults still matter because they can hide later alarms. Restore access through the approved owner process without factory-resetting grid or installer parameters.

When a user reset is reasonable

Some manuals permit one controlled restart after a transient condition. Use only the sequence for the exact system, normally at accessible AC and DC isolators in a defined order and with stated waiting periods.

Do not improvise a sequence or operate damaged, hot or wet switchgear. Stop if the alarm returns. Repeated cycling erases evidence, stresses equipment and can re-energise a genuine fault.

An isolator may be defective even if its handle moves. Signs of heat, crackling or melting require distance and urgent competent attention.

Emergency signs

If there is smoke, flame, visible arcing, melting, a strong burning smell or dangerous heat:

keep people away
call the fire and rescue service where there is immediate danger
do not climb onto the roof or open equipment
tell responders that a PV system is present

Turning off the AC supply does not make the illuminated array and its DC cables dead.

Grid faults versus installation faults

Repeated high-voltage trips around peak export may involve voltage rise in the service or installation, a network voltage issue, or both. The installer can record voltage at relevant points and calculate cable voltage rise; the DNO investigates its network where evidence indicates it.

Repeated low-frequency or loss-of-mains alarms may be genuine network events. A neighbourhood supply that appears normal by the time an engineer arrives does not invalidate a time-stamped inverter log.

Grid-protection parameters should remain at the approved settings. Altering them to keep the inverter online can create a safety and compliance problem.

Information for the service call

Provide the installer with:

commissioning pack and single-line diagram
inverter and module details
string layout and optimiser map
code history and monitoring export
photographs and recent work history
DNO connection documents
warranty and installer details

This shortens diagnosis and helps distinguish a product warranty claim from an installation, network or monitoring issue.

If the original installer is unavailable, use a competent PV contractor. For an unresolved MCS installation complaint, MCS publishes a consumer route, but immediate electrical safety should not wait for a contractual dispute.