Strada Fabricilor 7, Rădăuți, Suceava county · Mon–Fri 08:00–17:00 · +40 748 238 642
B2B Platform - Online [email protected] RO EN
Frequently asked questions

What the inverter is telling you

The situations we meet most often in service, on the makes we work with. Many of them clear up in two minutes, without anyone coming out to you.

Codes and behaviour differ between models and between firmware versions. What follows is the situations we see most often, not the manufacturer's full list. If what you are seeing looks like nothing here, send us a photograph of the display — it is the quickest way to find out what it is.

General questions, whatever the make

The system shuts down by itself around midday, in the strongest sun. Why?

This is the most frequent situation we see, and it is almost never a faulty inverter. When many systems in one area export at once, the grid voltage rises. Above the legal limit (typically 253 V per phase) the inverter is required by standard to disconnect — otherwise it would damage appliances in the house. It shows as a grid overvoltage: code 2001 on Huawei, 002 or 003 on Sungrow, OV-G-V on Solis. It reconnects by itself once the voltage drops. If it happens daily, the cause is either an undersized or over-long AC cable (which can be fixed) or the grid in your area (in which case a report goes to the network operator, and we help you with the monitoring data).

It is producing less than last year. Has something broken?

First of all, compare equivalent months rather than this month against last: in December a system makes a fraction of what it makes in June, and that is normal. If a year-on-year comparison still shows a drop, the usual causes, in order of frequency, are: dirty or snow-covered panels, new shade (a tree that has grown, a building, an aerial), a string that has dropped out — in which case you see one MPPT at zero in the app — or a tripped breaker. You can check the first two yourself; for the rest, look at what each input is showing in the app and send us what you see.

The app shows nothing, but the system seems to be running.

Almost certainly the communication rather than the production. Check on the display whether the inverter shows it is producing: if it is, the system is doing its job and only the reporting is interrupted. The causes and what to do about them are in the Monitoring and WiFi connection section below.

What counts as "the system is completely down"?

That it produces nothing and powers nothing: a blank display or one showing an error, zero production in the app, and if you have backup, that is not holding the loads either. It is different from "producing less" or "it stopped for an hour at midday". When you tick that box on the support form, the request goes to the top of the queue.

Can I reset the inverter myself?

A restart from the inverter's own button or app, where the manufacturer puts it within the user's reach, is fine — many temporary alarms clear that way. What we never recommend: opening the case, or touching the DC wiring or the PV board. The DC side stays live for as long as there is daylight, whatever breakers you have switched off, and it does not forgive mistakes. If the alarm comes back after a restart, leave that part to us.

Do you take on systems installed by somebody else?

Yes. Just tell us who installed it and, if you know, whether the company held an ANRE licence — not so we can refuse the work, but so we know what we are starting from and what the warranty covers. An installation that did not follow the manufacturer's instructions changes what can be repaired under warranty and what is chargeable, and it is fairer that you know that beforehand rather than afterwards.

Monitoring and WiFi connection

The inverter reaches the internet through a communication module — a WiFi dongle or a logger. When that loses its connection, monitoring stops but the system goes on producing. Almost every situation below comes from the home network rather than from the inverter.

I have changed my internet router. What do I need to do?

Our recommendation: keep the same network name (SSID) and the same password on the new router, and leave the 2.4 GHz network enabled. That way the system reconnects by itself, with nobody touching it — the inverter has the credentials stored and does not know the router changed. If the name or the password changes, the communication module has to be reconfigured, because it goes on looking for a network that no longer exists. Tell the internet installer this before they leave: it is far simpler then than afterwards.

How do I reconnect a Huawei inverter to a new network, without an installer?

It can be done from the FusionSolar app with your own owner account — no installer password needed. On recent versions: Services → WLAN Configuration, scan the QR code on the inverter or the dongle, the phone joins its network, you log in as User and land straight in the router settings; pick the 2.4 GHz network, enter the password, Connect. On older versions the path is Me → Commissioning, log in as User with the default password 00000a, then "…" top right → Settings → Router connection settings. If whoever installed the system changed the User password this will not work and you need us — write to us.

The same on Sungrow.

From the iSolarCloud app: press the button on the WiNet module three times to start its hotspot, then in the app go to Support → Local Access and log in with the "user" account and the default password pw1111 ("admin" with pw8888 is the installer account, which you do not need). Then More, bottom right → WLan configuration → choose the 2.4 GHz network, enter the password, confirm. The app needs to be a recent version, or the Local Access menu looks different.

The same on Solis.

On Solis the reconfiguration happens on the monitoring stick rather than the inverter. The datalogger broadcasts its own network, usually named starting with "Solis" or "AP_" followed by the serial number; join it from your phone and open 10.10.100.254 in a browser. The user is admin, and the password differs between models — 123456789, Solis1 or admin. Then Quick Set → choose the 2.4 GHz network → password → Save. Recent SolisCloud versions also have a WiFi configuration tool in the app, but not all of them do. If none of the passwords gets you in, write to us.

Why does it have to be 2.4 GHz?

The inverters' communication modules — Huawei, Sungrow, Solis, Deye — work on 2.4 GHz only. They will not connect to 5 GHz, whatever the settings. This is not a quality limitation but a deliberate choice: 2.4 GHz travels better through walls and reaches the garage or the loft where the inverter is mounted, which 5 GHz does not. If the new router only has 5 GHz enabled, there is nothing for the system to connect to.

My router broadcasts 2.4 and 5 GHz under the same name. Is that a problem?

It can be, and few people suspect it. Modern routers often combine the two bands under a single network name and move devices between them by themselves — the feature is called "band steering" or "smart connect". The inverter connects, the router tries to move it to 5 GHz, and it drops. The typical sign: monitoring works for a few minutes and then cuts out, over and over. The fix is either to disable the feature or to create a separate 2.4 GHz network with its own name and put the inverter on that.

I only changed the WiFi password, not the router.

The same effect as changing the router: the module keeps trying the old password. It has to be reconfigured with the new one. Which is why, when you change the password, it is worth remembering there is another device in the house using it — one that does not complain, it just goes quiet.

The WiFi signal is weak where the inverter is mounted.

A common situation: the inverter is in the garage, the loft or on an outside wall, and the router is at the other end of the house. If monitoring drops often with nothing having changed on the network, the signal is the first suspicion. A repeater or an access point closer by solves it; on some models there is also a wired option, steadier than any WiFi.

The system is offline in the app. Am I losing anything?

Not production — the inverter carries on and delivers energy exactly as before. What is lost is the history for the period it was offline and, more importantly, our ability to see what happened if something goes wrong. That is why it is worth reconnecting even if you do not look at the app daily: when you call about a fault, the first thing we look for is the alarm history.

Backup, ATS and MTS

When a system has backup, the question comes up of how the house switches between the grid and the inverter. We have a clear position on this, and it is worth stating beforehand rather than after somebody has bought one online.

What is an ATS and what is an MTS?

Both are switches between two supplies. ATS stands for Automatic Transfer Switch — it changes over by itself, with nobody involved, when it detects that one of the supplies has gone. MTS stands for Manual Transfer Switch — it changes over only when a person turns it. That is exactly the difference: automatic versus manual.

Do you install an ATS?

No, and we do not recommend one. There are cheap ATS units on the market, and most of the ones we have come across are faulty. A switch that has to work on its own, with nobody watching it, is precisely the part not to economise on: when it fails, it fails at the exact moment you need it. We install MTS only.

How do you install the MTS?

As a bypass for the whole inverter assembly, where the system has backup. Its purpose is to let you take the inverter out of the circuit in the event of a fault without leaving the house without power. The wiring is simple: the house is connected to the MTS load port, and its two inputs are the grid and the inverter's backup output. Switch to the grid and the house carries on, whatever the inverter is doing.

Why does a bypass matter if the inverter already has backup?

Because backup covers the grid failing, not the inverter failing. If the inverter itself stops — an error, a part to be replaced, a planned visit — everything that passes through it stops with it, the backup loads included. With an MTS you switch the house straight onto the grid in a few seconds and wait calmly for the part or the technician.

Huawei SUN2000

Alarms appear in the FusionSolar app under Device → Alarm History, and as active alarms on the display. The code has four digits and starts with 2.

Code 2001 — grid overvoltage

The grid voltage has exceeded the permitted upper limit. Most of the time this is a passing disturbance and the inverter recovers by itself. If it appears often in the alarm history it is worth investigating: either the AC cable or the grid in your area. See also the first question in the general section.

Code 2032 — grid loss

The inverter can no longer see the grid. The first check, and one you can do yourself: whether the PV system's breaker in the consumer unit has tripped. If it has, and it trips again immediately after you reset it, do not persist — call us. If the breaker is fine and the house has power, the AC connection needs checking.

Code 2002 — DC arc fault

The inverter has detected an electrical arc, usually from a loose contact on a string cable. This is not an alarm to ignore, and not one to deal with yourself: it needs the DC connections checked with the system isolated in the correct order. Send us a request.

Code 2011 — PV string polarity reversed

The positive and negative of one string are reversed. In practice this only appears after work or an extension, never on its own. It is corrected by an electrician, in the evening, once the string current has dropped far enough.

Code 2012 — too few modules in a string

One string has fewer panels in series than the others. In practice: a panel or a connector on that run has become disconnected. Checked on site.

Code 2063 — over-temperature

The inverter is overheating. The usual cause is where it is mounted: not enough ventilation, direct sun on the case, too tight a space. Check whether something has been put next to or over the inverter since, and whether the vents are clogged with dust. If the location is unchanged and the alarm is new, it may be the fan.

Code 2064 — high residual current / earth leakage

The inverter has measured a leakage current above the permitted threshold. This is a safety alarm and is checked by an installer — not something to keep resetting until it goes away. Open a request.

Sungrow (SH and SG series)

Sungrow codes have three digits and appear in the iSolarCloud app and on the display. The 0xx range covers the grid and currents, the 7xx range the battery.

Codes 002, 003, 004 — grid voltage out of range

002 and 003 mean overvoltage (003 a brief one), 004 undervoltage. As with any make, the inverter disconnects because the standard requires it to, not because it has failed. If it repeats daily at the same hour, the AC run or the local grid is worth analysing.

Codes 008 and 009 — frequency out of range

The grid frequency has left the permitted band. The cause is in the grid, not in the system, and the inverter recovers by itself. If it happens very often it is worth reporting to the network operator.

Code 714 — battery communication lost

This is the most frequent battery alarm on Sungrow. The inverter can no longer talk to the battery's BMS, so it stops charging and discharging. Usually it is the RJ45 communication cable between inverter and battery: loose, in the wrong socket, or replaced with an ordinary network cable, which is not wired the same. It is not a cell fault.

Code 039 — insulation / Code 012 — leakage current

Both are safety alarms: 039 shows too low an insulation resistance on the DC side (often damp in a connector or a damaged cable), 012 a leakage current above the threshold. Both are checked by an installer, with instruments. Neither should be repeatedly reset.

Code 088 — arc fault

Arc detection on the DC side, most often from a loose contact. It needs the connections checked on site.

Codes 070, 082, 083 and 619 — fans and temperature

The three in the 0xx range report fan problems, 619 a heat-sink over-temperature. Check first that the inverter has air around it and that the vents are not clogged. If the location is unchanged, a fan is a part that gets replaced.

Solis (Ginlong)

Solis shows its alarms as text rather than numbers, which is why you recognise them by the abbreviation. They appear on the display and in SolisCloud.

OV-G-V — grid overvoltage

The inverter measures too high a grid voltage and disconnects. On Solis a frequent and fixable cause is an undersized or over-long AC cable: its resistance raises the voltage measured at the inverter's terminals even when the voltage at the board is fine. Moving to a larger cross-section often solves it. If the run is correct, the local grid is what remains.

NO-Grid — no grid detected

The inverter measures no grid voltage. Check first the system's breaker in the consumer unit, and whether the rest of the house has power. If both are fine, it may be the AC connection or the grid standard setting — the latter mostly after an inverter has been replaced.

ISO-PRO and ILeak-PRO — insulation and leakage current

Safety alarms on the DC side. ISO-PRO often appears after long spells of rain, from damp in a connector, and sometimes clears as things dry out — but if it comes back it has to be measured. Neither is solved by repeated resets; both need an installer with instruments.

ARC-01 — arc fault detection

The inverter has detected an arc on the DC side. As with the other makes, the typical cause is a loose contact, and the check is done on site with the system properly isolated.

Why does Solis show more grid alarms than other makes?

Not necessarily because it is more sensitive, but because Solis also reports the brief events that other inverters ride through without logging. A single OV-G-V that clears within seconds is not a fault. What matters is whether it repeats and how long the system stays off — and that is visible in the SolisCloud history.

Deye (hybrid)

Deye uses F codes, from F01 to F64. Several of them appear in the history even when nothing has failed — they are event records, not defects.

F13 — working mode or grid change

Appears after a configuration change: working mode, battery type, grid type or frequency. If F13 appeared exactly when settings were changed and the inverter went straight back to normal, it is an event record, not a fault. It is only worth looking into if it returns without anybody touching the settings.

F18 — AC over-current

The hardware over-current protection on the AC side. In practice it most often means too much load on the backup or the load output. Check what you have running on backup — a water heater, a pump and a washing machine started together can exceed the output rating on their own. Once the load is reduced, restart.

F56 — DC bus voltage too low

Almost always battery-related: low voltage, a very low state of charge, or limits imposed by the BMS. Check the battery level in the app. If the battery is deeply discharged it has to be charged from PV or the grid before the system returns to normal.

BMS_Err_Stop — the inverter has lost communication with the battery

When BMS communication is lost the inverter stops and reports the fault — that is protective behaviour. The usual causes: the communication cable is not in the right socket, or the battery type selected in the inverter does not match the battery fitted. The second happens mostly after a battery has been changed or extended.

The battery will not charge even though I have production. Why?

On a Deye hybrid this depends almost entirely on settings rather than hardware: the charging time windows, the state-of-charge limit, the priority between loads, battery and grid. If the system was working correctly and the behaviour changed with nobody touching it, tell us — but in most cases it is a setting we can correct remotely, if the inverter is on monitoring.

Dyness batteries

Batteries have no display of their own: what happens to them shows as an alarm on the inverter they are connected to. That is why the first useful piece of information is what the inverter says, not what lights up on the battery.

The inverter reports a battery communication error.

Most BMS alarms on Dyness are configuration problems, not dead batteries. The four causes we find most often: two modules on the same address (DIP switches not set after an extension), a loose cable between modules, a module count set in the inverter that does not match what is actually fitted, or a firmware mismatch between the BMS and the inverter. None of them needs new parts.

I added a module and now nothing works.

The classic case. An extension requires the addresses to be set on the new modules and the module count updated in the inverter. If two modules are left on the same address, the stack stops reporting correctly and the inverter halts charging and discharging. It is not a fault — it is an unfinished commissioning.

Which installation mistakes cause the most alarms?

In order: an ordinary network cable used instead of the battery's communication cable (they look identical but are not wired the same), the CAN cable put into the meter, RS485 or logger port instead of the battery port, the wrong battery make selected in the inverter, and DIP switches left unset. If the battery worked for a while and then stopped, the first suspicion is a loose cable, not the cells.

The battery shows a state of charge that makes no sense.

The state-of-charge estimate can drift over time, especially if the battery never reaches fully charged or fully discharged. It usually corrects itself after a few full charge cycles. If a large discrepancy remains after that, it gets checked.

How do I find out exactly what happened?

From the inverter's alarm history, not from the battery. When you write to us, send the exact code or text from the inverter app and, if you can, a photograph of the module stack — it shows how many modules there are and how they are connected.

Hailei batteries

51.2 V LiFePO4 batteries (16 cells in series), wall-mounted or rack, with CAN and RS485 communication. As with any 51.2 V battery, what matters at commissioning is the protocol selected in the inverter.

Which inverters does it work with?

With hybrid inverters that have a configurable lithium battery profile or direct BMS communication over CAN or RS485 — among them Deye, Growatt, Goodwe, Solax, SMA, Voltronic and Sorotec. Importantly, "works with" does not mean "is recognised automatically". The correct protocol has to be selected in the inverter, or the battery is seen as not there.

The inverter does not see the battery at all.

Three checks, in this order: the protocol or battery make selected in the inverter, the port the communication cable is in (CAN and RS485 look identical but are not interchangeable), and the cable itself — a battery communication cable is not an ordinary network cable. Only once all three are right does a fault become worth considering.

Can I put a Hailei next to a battery of another make?

No. Modules of different makes, or even from different series of the same make, do not go in the same stack: the BMSs do not coordinate, and the modules end up charging and discharging unevenly. If you want more capacity, the extension is made with modules identical to the ones fitted.

How many cycles does it last?

The manufacturer quotes up to 6000 cycles, which is typical for LiFePO4 chemistry. In real use, the life depends a great deal on how deeply it is discharged each day and on the temperature of the room it is in — a battery kept cold in an unheated garage ages differently from one in a space at a steady temperature.

The battery will not charge in winter.

LiFePO4 batteries have a temperature below which the BMS stops charging, to protect the cells — that is protection, not a fault. Discharging generally stays permitted at lower temperatures than charging. If the battery sits in an unheated space and the behaviour repeats every winter, the answer is where it is mounted, not the battery.

Did not find the answer?

Send us the details and a photograph of what the inverter is showing. With the alarm code and the make of the equipment, we often know the cause before anyone gets in the van.

Open a support request →