An arc-fault indication is a service event, not a reset challenge
If a solar inverter reports an arc-fault alert, keep clear of the equipment and arrange prompt qualified investigation using the exact model’s instructions. If there is smoke, flame, unusual heat or another immediate danger, move away and call emergency services. Do not open enclosures, touch connectors, climb to inspect panels, cycle breakers or disable arc protection to restore production. A subsequent return to normal output does not establish that the indicated problem is harmless.
The useful owner task is preserving the complete event, time and equipment identity while coordinating service. Avoid diagnosing the cause from weather, a graph or another owner’s reset video. This article is specifically about a protective arc-fault service pathway. The sunny-afternoon shutdown guide covers a broader interruption question; its grid or thermal discussion should not be used to dismiss an explicitly identified arc-fault event.
Match the manufacturer document before interpreting the message
Collect the full inverter model, serial and firmware from existing commissioning records or an established monitoring view that can be read safely. Keep the alert text and number intact, including the time shown and whether data appears current. Do not shorten a long message to “inverter error.” Similar product names can have different protection and restart behavior. Give the provider the actual identity rather than selecting a manual solely because its cover shows a familiar-looking enclosure.
The manufacturer example reviewed here is SMA’s US-41 documentation for SB3.0, SB3.8, SB5.0, SB6.0, SB7.0 and SB7.7-1SP-US-41, from firmware 04.00.03.R. Its validity page establishes that scope. It does not supply a universal event-number dictionary for other Sunny Boy generations, storage inverters or brands. This guide cites the complete relevant HTML sections, not a claim that every page of a large operating manual was read. Have the provider resolve applicability before any technical action.
Understand why renewed production is not clearance
SMA’s matched interfaces and functions section says AFCI indications should be investigated promptly and thoroughly by qualified people. Its AFCI section identifies several operation-inhibition and restart arrangements and marks that work for a qualified person. Those distinctions explain why an inverter can resume operation in some configurations without an owner reset. They do not authorize changing the configuration, deliberately producing more detections or waiting for an arbitrary event count before asking for help.
Record whether the app shows a resumed state, a continuing alert or another event, but do not induce a recurrence to strengthen the evidence. Ask the provider whether operation needs to remain restricted while investigation proceeds and who implements the matched instructions. An app’s present state is an observation at a particular time. The service record should state what was investigated, what was found and what evidence supports any decision to resume operation.
Leave switching and electrical investigation with qualified people
The reviewed SMA safety section warns that under fault conditions direct actions on the product can be dangerous, and explicitly says not to operate the inverter’s DC load-break switch during an error. This is why a generic online instruction to “turn everything off and on” is unsuitable. The same safety document explains that illuminated PV modules produce hazardous DC voltage. A dark display or interrupted production does not prove the whole array, wiring or connected storage is de-energized.
This homeowner guide gives no isolation sequence, connector handling, voltage test, enclosure-opening instruction or restart parameter. Tell the electrician or solar service provider about the specific error and known system configuration, including batteries and other inverters. If safe access is uncertain, keep people away and let qualified responders or service personnel determine the next action. Reading a service manual does not provide the training, equipment or site assessment needed to perform its technical work.
Use observations to organize the response
A clear service request distinguishes observed conditions from proposed diagnoses. For instance, saying an alert appeared after rain is useful history; saying rain caused a connector fault is an unsupported conclusion without investigation. Retain the original screenshot and add context separately. If the timestamp is stale or the monitoring connection is uncertain, report that limitation instead of silently treating it as the present equipment state.
| Observation | Useful owner response | Conclusion to avoid |
|---|---|---|
| Smoke, flame or unusual heat | Move away and contact emergency services | A remote reset will make access safe |
| Arc-fault alert without visible emergency | Report complete message and arrange qualified investigation | No visible damage means no problem |
| Production resumes automatically | Preserve event and ask about continuing restrictions | Resumed output certifies correction |
| Repeated recorded alerts | Escalate recurrence without causing another event | Changing a threshold is maintenance |
| Service visit completed | Obtain findings, corrections and authorized handover | A cleared screen proves all work was done |
Prepare a targeted service checklist
Provide the installation location, complete alert, time and reporting freshness, model and firmware, commissioning documents, known modifications, previous faults and relevant weather or work history. Include existing safe photographs and the manufacturer case if one has been opened. Do not approach a suspect inverter or array to obtain a better photograph. State which information is missing so the provider can plan identification and safe access as part of the visit.
Ask the quote to identify the investigation scope, qualified personnel, manufacturer coordination, documentation and any separate corrective authorization. The owner need not prescribe a test method or a replacement component before findings exist. Require a distinction between an assessed cause, an unresolved lead and work outside the visit. If the original installer has closed, use the manufacturer and appropriately qualified local service pathway; preserve equipment ownership documents rather than relying on an inaccessible contractor portal.
Require a closeout that explains the authorized restart
The handover should identify the actual equipment and alert, relevant findings, corrective work, changed components where applicable, and who authorized return to operation under the matched instructions. Ask how any unresolved limitation affects use and what recurrence should prompt another call. A technician’s detailed record serves a different purpose from an owner screenshot showing energy production. Keep both, with their dates, instead of replacing the fault history with a reassuring final graph.
Do not accept a homeowner-facing instruction to defeat AFCI as a routine remedy or conceal recurring indications because a system still generates energy. Local licensing, electrical rules, utility requirements and the actual installation affect service responsibility. No general article can promise fire prevention, certify every connection or guarantee a repair outcome. DOE’s fire-safety guidance reinforces the need for qualified installation and response; it is not evidence that this particular alert has been resolved.
Illustrative scenario: the alert clears before the appointment
Suppose a homeowner sees an arc-fault notification in the morning and normal production later that day. This is an illustrative scenario. Rather than assuming a software glitch, the owner saves the event and exact model information, keeps clear of the equipment and contacts qualified service. The provider checks which manufacturer documentation and restart arrangement apply to the installed unit before determining the safe investigation plan.
The eventual report distinguishes the alert, examined components, findings, corrective work and any remaining uncertainty. If a restart is authorized, the file records that decision and the follow-up expectations. The owner does not experiment with switches while waiting or reproduce a reset procedure from a different inverter. This approach preserves the evidence needed to answer the fault question without turning the household into a test crew or claiming that one resumed afternoon proves future safety.
Frequently asked questions
Can an inverter restart itself after an arc indication? Some configurations do; that is not owner clearance. Can I reset the alert to see whether it returns? Leave that decision and action with qualified service. Is every alert false if the panels look intact? No, visual appearance does not resolve it. Can the SMA example be used for another model? Not without actual applicability. Does shutting off a screen remove all PV voltage? No. Should I change AFCI sensitivity or disable it? Not as homeowner troubleshooting. What should I send the provider? The complete message, time, equipment identity and existing history. Does this guide prescribe a switching sequence? No.