Solar Inverter Troubleshooting: 15 Common Problems

Most rooftop solar troubleshooting goes wrong in the first five minutes. The owner sees a blank app, assumes the inverter has failed, and calls the installer. Two days later an engineer arrives and finds a router that changed its Wi-Fi password. Meanwhile a genuine insulation fault on the next roof gets reset four times because the beeping was annoying.

The fix is not more electrical knowledge. It is order. Troubleshooting works when you establish safety first, then read the status indicators, then place the symptom on a triage ladder. Only then do you decide whether the inverter, the array, the grid, or the monitoring link is at fault.

This guide is that triage layer. It covers the safety rules that come first, how to read status lights and codes, and a ladder that sorts symptoms into four states. A symptom table then routes each case to the right next step. It also teaches the most useful skill here: separating an inverter fault from an array, grid, or monitoring fault. Finally, the evidence to gather before you call support.

Short version. Work in this order. Confirm nothing is smoking, arcing, or wet, and leave every cover closed. Note the time, the weather, the exact model, and the displayed code. Decide which of four states applies: normal, low, intermittent, or dead. Use the MPPT readings to separate an inverter fault from an array, grid, or monitoring fault. Then collect evidence and escalate.

Safety comes before diagnosis, and DC is the reason

A solar array is a generator you cannot switch off. Sunlight on the modules produces DC voltage at the inverter’s input terminals. That voltage is present whether the grid is up, the AC breaker is off, or the inverter is dead. Switching off the AC side does not de-energise the DC side. That single fact rules out most of what people want to try.

Three rules follow, and they are not negotiable.

Correct isolation, performed by a qualified person, runs from the grid inwards. Open the AC isolator, then the DC isolator. Then prove the circuit dead at the point of work. Doing DC first leaves the inverter feeding the AC side. Earthing and isolation practice sits in IS 732:2019 and IS 3043:2018 (Bureau of Indian Standards). Supply safety obligations sit in the Central Electricity Authority (Measures relating to Safety and Electric Supply) Regulations, 2023.

How to read the status indicators without guessing

Every string inverter talks in three layers: indicator lights, the display or app code, and the event log. Lights give the broad state. Codes name the event. The log says whether the event is new, recurring, or constant. Reading them in that order saves a wasted call.

Broad states are consistent across brands even when the exact colours are not.

Two cautions. An identical phrase can mean different things across brands and firmware revisions, so the current manufacturer manual defines your code, not a generic list. Our error code reference explains how code families are structured. And a warning state is the most under-used signal in Indian rooftop solar, because derating warnings accumulate for months before a summer shutdown.

The triage ladder: four states, four different first moves

Fifteen symptoms is too many to hold in your head. Four states is not. Place the system on this ladder first, because the state decides whether you are chasing performance or chasing a defect.

Step 1. Is it generating at all right now? Check in daylight, ideally between 11am and 2pm on a clear day. If AC power is above zero, you have a performance question, not a failure. Go to step 2. If it is zero in good sun, jump to step 4.

Step 2. Is it generating less than expected? Compare against a comparable day, never against the nameplate rating. A 5 kW inverter almost never produces 5 kW, and that is design, not failure. Soiling, shade, cable losses, temperature, and deliberate DC oversizing all reduce output legitimately. The low output diagnosis guide ranks the causes by likelihood.

Step 3. Is it faulting intermittently? Intermittent is the most informative state, because the timing carries the diagnosis. Midday clusters suggest heat or overvoltage. Dawn faults suggest low DC voltage start-up behaviour. Faults after rain suggest moisture ingress. Faults that track a neighbour’s motor load suggest the feeder.

Step 4. Is it dead? No display, no AC output, no app data, in full sun. This is the only state that usually needs a site visit. Before booking one, confirm the AC breaker position, confirm DISCOM supply is present, and confirm no firmware update is in progress.

The ladder matters because steps 2 and 3 get misdiagnosed as step 4. An owner who reports “inverter dead” when the real state is “derating at 2pm in May” gets the wrong engineer with the wrong parts.

The 15 common problems: cause, safe check and when to call

This table is a routing device, not a diagnosis. One displayed phrase can have several causes, and the manufacturer manual for your exact model and firmware supplies the definition. Every “safe check” is something you can do from outside the enclosure. The last column says when the problem stops being an owner task.

For code-level help on items 4 to 9 and 11, the solar inverter error codes guide explains how to group a fault code when you do not have the manual in front of you. When a problem reaches the last column, raise it on the Qbits support helpdesk{target=“_blank” rel=“noopener”} or find a partner through the authorized service partners directory.

Read across the whole row. The value of the table is that it pairs a plausible cause with an action that does not make things worse.

Inverter fault, array fault, grid fault, or monitoring fault

This is the most valuable distinction on the page. Four systems produce overlapping symptoms and route to four different people: the inverter brand, the installer, the DISCOM, and the monitoring platform. Getting it wrong costs weeks.

Three quick tests separate them without opening anything.

1. The clock test. Does the symptom correlate with time of day or weather? Time-correlated symptoms point at grid or thermal causes, not a failed component.
2. The local display test. If the unit’s own display shows live AC power while the app shows nothing, generation is fine and the monitoring link is broken.
3. The symmetry test. On a multi-MPPT inverter, compare the inputs against each other. Asymmetry is the array’s signature.

Worked example: separating an array fault from an inverter fault

This is arithmetic on assumed readings, shown to demonstrate the method. It is not field data from an installation.

Expected string voltage:

• Temperature correction = -0.28% x (55 - 25) = -8.4%
• Vmp per module = 42.0 x 0.916 = 38.5 V
• Vmp per string of 10 = 385 V

Expected string current:

• Imp scales roughly with irradiance = 13.0 x (700 / 1000) = 9.1 A

Now the readings an electrician takes at the inverter:

Read the result. Voltage on both inputs sits within 0.5 percent of each other, and within 1 percent of the calculated 385 V. Module count, wiring polarity, and the temperature assumption are therefore sound on both strings. Current on MPPT 2 is almost exactly half of MPPT 1. That is an array-side signature. A sub-string in a parallel pair has dropped out, a string fuse has cleared, a connector has degraded, or half the string is shaded.

An inverter-side MPPT failure looks different. The tracker cannot hold the maximum power point, so voltage wanders or collapses towards open-circuit while current stays low. Voltage that is correct and stable exonerates the tracker.

What a homeowner may safely do, and what needs an electrician

The boundary is simple. If it needs a tool, a cover, or a meter, it is not yours. Observation, recording, and the AC breaker you already use are yours.

Safe for a homeowner, from outside every enclosure:

1. Read and photograph the display, LEDs, and rating label.
2. Note the date, clock time, weather, and what changed recently.
3. Check whether grid supply is present elsewhere in the building.
4. Look, from a safe distance, for blocked ventilation, nesting, or vegetation against the heat sink.
5. Check whether the monitoring router or SIM is working.
6. Perform one, and only one, documented restart using the AC isolator if the manual permits it.

Requires a licensed electrician, without exception:

1. Any DC measurement, including string voltage, string current, and insulation resistance.
2. Any work behind a cover, including internal heat sink cleaning and fan replacement.
3. Any connector, gland, fuse, or terminal work.
4. Any ground fault, arc fault, or insulation resistance investigation.
5. Any change to grid protection settings, which follow the DISCOM connection standard rather than a preference.
6. Any repeated breaker trip.

Rule 1 deserves a plain explanation. String voltages on a residential array can exceed 400 V DC and climb further in cool weather. That is not a household measurement.

The contrarian bit: resetting is diagnosis, not repair

Resetting an inverter is the most common troubleshooting action in Indian rooftop solar, and usually the least useful. Three myths keep it alive.

Myth 1: if the reset clears it, it was a glitch. Usually it only means the trigger condition has passed. A midday overvoltage trip clears itself by evening. Nothing was fixed, and it returns tomorrow at the same hour. The useful data point is the pattern, not the clearance.

Myth 3: repeated beeping is only a nuisance alarm. Sometimes it is standby behaviour. Sometimes it is a battery or auxiliary supply condition that worsens quietly. Pattern and timing decide, which is why you log them.

One documented restart is a legitimate test. The second identical restart overwrites evidence the engineer needs.

Evidence to collect before you call support

Support quality tracks evidence quality. A ticket with a code, a timestamp, and a photograph of the rating label often resolves without a visit. A ticket that says “not working” guarantees one.

Collect all of the following:

Two things to keep out of the ticket. Never send passwords, OTPs, or portal credentials through a public support channel or a WhatsApp group. And do not photograph a code you have already cleared, because the timestamp is the part that matters.

For Qbits units, the datasheet library confirms the model and its rated specifications, and the authorised service partner network is the route for on-site attention. Check the partner locator for your area and confirm in writing which partner covers your site.

Warranty and RMA, at a high level

Know what a product warranty covers before you file. It covers the inverter. It does not cover the installation around it. Array wiring, earthing defects, mounting, DISCOM voltage conditions, and monitoring routers go to the installer or the utility instead. That division is the commonest reason a claim stalls.

Qbits publishes an expandable warranty, and its public datasheets do not define the base term, remedy, registration deadline or exclusions, so obtain the current written warranty for the exact quoted model. Ask what happens once a claim is approved, and get the committed action in writing, because Qbits does not publish a service response time. Your written warranty document and purchase contract govern the actual entitlement, including proof of purchase and commissioning records.

The process runs in a predictable order.

1. Identify the unit and confirm it is within the warranty period, using invoice and commissioning records.
2. Raise the claim with model, serial number, code, timestamp, and photographs.
3. Cooperate with remote diagnosis, which often resolves configuration and monitoring cases.
4. Accept a site inspection if requested, since it establishes whether the cause is the product or the installation.
5. Receive the approval decision, after which replacement logistics begin.
6. Keep the faulty unit available, because returning the defective item is normally a condition of replacement.

Our warranty claim walkthrough covers the documentation in detail, including what a rejected claim usually lacks. The hybrid range carries extra battery-side considerations, so check your hybrid inverter specification before assuming a battery symptom is an inverter fault.

Preventive maintenance that actually reduces faults

Most rooftop inverter faults in India are environmental, not electronic. Heat, dust, monsoon moisture, and rodents produce more tickets than component failure. That makes prevention unusually effective here, and it is mostly inspection rather than intervention.

Two items beat the rest of the list. Keep the heat sink clear and unshaded, because derating is the most common avoidable loss. And treat the monitoring history as a maintenance record, not a curiosity. Our maintenance guide for Indian conditions sets out the full schedule.

The Bottom Line

Troubleshooting a solar inverter is a routing problem, not an electrical one. Establish safety, read the state, place it on the ladder, then decide which of four systems owns the fault. That decision is usually the whole job. Getting it right is the difference between a five-minute app fix and a two-week misdirected ticket.

• Run the ladder before you call anyone. Generating, low, intermittent, or dead. Each state has a different first move, and reporting the wrong one sends the wrong engineer.
• Use the symmetry test to place the fault. Similar DC voltage with very different DC current across MPPT inputs means the array, not the inverter. Only a licensed electrician takes those readings.
• Bring evidence to the ticket. Model, serial, exact code, timestamp, screenshots, photographs. For Qbits equipment, confirm the model in the datasheet library, then contact the Qbits team with the record in hand.