Growatt PV Isolation Low: error 203, 125 and relay fault 117

On a Growatt MIN TL-X or TL-XH, Error: 203 beside the words PV Isolation Low means the inverter measured the insulation resistance between your array and earth, decided it was too low, and refused to connect to the grid. The same fault is numbered 125 on the 30-50K TL3-S/NS and carries no number at all on SPH BH-UP hybrids, so the number on its own tells you very little until you know the model. Error 117 is a different animal entirely: a relay fault on the TL3-S family, not the battery fault widely repeated online. And a flashing red LED on a Growatt datalogger is, in Growatt’s own manual, the state that means it has successfully connected to the inverter.

Work through the model selector first, then the fault itself.

First: which Growatt have you got?

Growatt uses four numbering schemes at once, and the same number means different things in different ones. The model number is on the label on the side of the inverter.

So there are three questions to answer before you look anything up: is it an Error or a Warning, does it come with a text string next to it, and which family is the inverter from. A page that just tells you “Growatt error 203 means PV isolation low” without asking any of that is wrong for a large share of the people reading it.

What Growatt PV Isolation Low actually means

Your panels, their frames, the mounting rails and the roof are earthed. The live DC conductors running from the strings to the inverter are not, and they are separated from everything earthed by insulation: the cable jacket, the connector bodies, the sealing around each junction box, the panel backsheet.

Before it connects to the grid, and continuously while it runs, a Growatt inverter measures the resistance between the DC side and earth. Dry and healthy, that resistance is very high. When water, dirt or damage gives current an easier path to earth, the measured resistance falls. Below the threshold set for that model, the inverter stops. That is not the inverter being fussy: a low-insulation array is a shock and fire risk, and stopping is what the safety standard requires it to do.

Two things follow, and they are what make this fault so awkward to catch.

The first is that the inverter is behaving correctly, so nothing else looks wrong. There is no smoke, no smell, no tripped breaker in the house and no burnt-out box, just a display showing an error and an export figure of zero.

The second is that water evaporates. An array wet enough to trip the inverter at first light in October fog can be dry enough to run again once the sun has been on it, at which point the inverter starts and produces normally for the rest of the day. If you check at lunchtime you will find a perfectly healthy system and no error on the screen at all.

The damp-morning pattern versus a real cable fault

You are trying to tell apart two situations with the same code.

Moisture ingress that dries out. The fault appears in the morning on damp, foggy or frosty days and after heavy rain. The inverter restarts by itself. The daily total is a bit down, but not zero. As the year gets drier the pattern fades, and it comes back in autumn. This is water getting into a connector, a gland or a junction box that is no longer keeping water out.

A genuine insulation fault. The code is there in the middle of a dry, sunny afternoon. It survives a restart. It appears on dry mornings as well as wet ones. Production is zero, not reduced. This is a cable chafed on a sharp roof edge, a rodent-damaged string, a damaged panel backsheet, or water that has got somewhere and is not coming out again.

The way you separate them is the log, not the display. Look at ShinePhone or ShineServer at the daily production graph for the past few weeks, not at today’s number. The damp-morning fault draws an unmistakable shape: a flat line each morning, then a normal curve once the array has dried. A real fault draws a flat line all day. If the shape looks right but the totals still look thin, that is a different question, worked through in solar panels producing less than expected.

We had solar panels installed in 2015 and everything was fine until February 2018. Then we received an email saying that our system had generated no electricity in the last 24 hours. Since then we have been checking the meter daily and, indeed, there are days when it does not generate.

That is the shape of an intermittent fault seen from the owner’s side: some days nothing, most days fine, and no way to tell which without looking every day.

Growatt PV Isolation Low: how to fix it safely, and what needs an electrician

So, how do you fix Growatt PV Isolation Low? Mostly you do not: you narrow it down and hand a technician the half of the picture only you have. Safe for you, at ground level, with the inverter still closed:

• Read the exact string. Photograph the display or the fault entry in ShinePhone. Error: 203 and Warning 203 are different faults and you will be asked which one it was.
• Note the time of day and the weather. Morning, fog, after rain, or dry midday: this is the single most useful piece of information you can give a technician, and only you have it.
• Check the daily graphs for the past few weeks and see whether the mornings are missing.
• One clean restart. Turn the AC breaker for the inverter off, wait a minute, turn it back on. Leave the DC isolator alone unless your own inverter manual tells you otherwise and you are certain which switch it is.
• Look at the outside of things. Standing on the ground: is a cable hanging where it can rub, has a conduit come loose, is there an obvious pool of water around an outdoor enclosure.

Not safe, at any point, no exceptions:

• Opening MC4 connectors, junction boxes or the inverter’s DC compartment. A string in daylight is live whenever there is light on the panels, it cannot be switched off by unplugging it, and DC arcs do not self-extinguish the way AC arcs do.
• Going on the roof to look at panels.
• Measuring insulation resistance yourself. Diagnosis needs an insulation tester used string by string, by someone qualified to isolate the array first.

What the electrician will do is disconnect the strings, megger each one separately to find which is low, and work along that string to the connector, cable run or panel that is failing. Ask for the measured resistance per string, in writing, before and after the repair. It is the only evidence that the fault is really gone rather than temporarily dry.

Growatt error 117 is a relay fault, not a battery fault

You may read online that error 117 means the battery voltage is outside the allowed charging range. That comes from third-party sites, not from Growatt. Two Growatt-authored guides for the 30-50K TL3-S family both describe 117 as a relay fault.

The relay in question is the AC-side isolating relay that connects the inverter to the grid. It is a part inside the sealed half of the inverter, not anything you own or can reach. Practically, that means:

• It is not something you can fix, and it has nothing to do with your battery, if you even have one.
• One clean restart is worth trying, in case the fault was a one-off.
• If it comes back, it is a warranty or service job. Relay faults are internal to the inverter.
• Growatt uses a second number for a relay fault elsewhere in the range: on the MIN TL-XH and SPH, Error: 405 is the relay fault, while Warning405 on the MIN means something completely different: the firmware version is not consistent.

If you have an SPH hybrid with a battery problem, the umbrella code documented in Growatt’s own live FAQ is error 217, not 117.

Growatt datalogger red light flashing: usually the healthy state

The ShineWiFi-X has a single tri-colour LED. From the official ShineWiFi-X user manual (EN 202311):

Growatt’s separate combined ShineWiFi-X / ShineWiFi-S indicator guide adds three states the X manual does not spell out:

Red is the stick’s first successful step, blue is the finish line, and the state that means failure is solid green.

A red light on the inverter itself is a different thing from a red light on the datalogger, and that is where most of the confusion comes from: the table above is the datalogger’s, and it does not describe the inverter. So the first question with any red light on a Growatt system is: which box is it on, the inverter or the little stick plugged into it? For the stick, the table above. For the inverter, the indicator table in your own inverter’s manual, and nothing else. Which box the light is on, and what each state on the stick means, are covered in Growatt datalogger red light.

I’m new to this - bought a house with panels recently … Recently I noticed that the readings don’t change and the red light is static.

The poster does not say which box the light is on, or which brand the system is, so the light alone cannot be read. Flashing red on a datalogger, with data still arriving in the app, is not a fault. A light on the inverter itself means whatever that inverter’s own manual says it means, and readings that have stopped moving are worth chasing either way. If you are still unsure which of the two you are looking at, the offline-or-not-producing triage separates a communication fault from a production fault.

The rest of the Growatt error code list

Every meaning below is Growatt’s own. The same fault changes number across families.

On SPF off-grid units the numbers are bare two-digit codes in a scheme of their own: 04 with a fault icon is battery voltage too low while 04 with a warning icon is a low-battery warning, 08 is bus voltage too high, 19 is battery disconnect and 20 is a BMS communication error. Do not read SPF codes off a MIN table or the other way round.

Error 111 is described as “neutral line charged” in one Growatt-authored guide and as an IGBT drive fault in another, so it is not in the table. The “20 minutes before a plant shows offline” figure repeated across the web is not in Growatt’s documentation, so do not rely on it.

If the inverter in front of you is a SolarEdge rather than a Growatt, the equivalent list is the SolarEdge error-code guide.

What an intermittent fault costs while nobody is watching

Worked for a 4 kWp array; put in your own kWp and tariff.

The second scenario is how owners describe finding these things:

Gutted to get back from holiday and find that the inverter stopped with an error on the day we left. 2 weeeks of prime August generation lost :( All it took was a turn off for 10 mins, then back on to resolve it but as we were away we lost out.

Which? found that 15% of solar owners had experienced inverter problems.

Who tells you the next time it happens

Not you, in all likelihood.

Growatt’s homeowner app is a monitoring app in the literal sense: Growatt documents no alerting in it. The information is there if you go and look for it.

Growatt’s alerts are in the web portal: three tick-boxes in ShineServer for alerts, daily reports and monthly reports. They go to the mailbox entered when the plant was set up, normally the installer’s address rather than yours. And they cover faults and warnings, not underperformance. Which inbox those messages reach is set out in who gets your solar alerts.

For a fault that appears at 07:00 and clears at 11:00, that adds up to a system nobody is watching. The inverter, ShineServer and the graph in your own app all know, but nothing reaches you, and by the time you look, the array is dry and the display says everything is fine. The check that catches this shape of fault is a daily comparison of yesterday’s production against what the weather should have produced. That is what Solmiro does: every day it works out what the weather over your roof allowed, and it e-mails you when production falls well short of that or stops. Connecting is free: Growatt monitoring, how it works and the guarantees. If you are not yet sure your panels are producing at all, how to know if your solar panels are working is the check to run first.