Skip to content
Field note · Case study · Solis · Pylontech · Aug 2026

The MCB that looked like it was on

A battery that had stopped working, a BMS still cheerfully reporting 78%, and almost no voltage at the inverter. The DC breaker had tripped, but the enclosure lid stopped the toggle dropping, so it read as on from the outside.
  • Real customer fault
  • Diagnosed from the data
  • No parts, no return visit
Battery stopped working, but the app still shows a charge?

That combination is a strong clue, not a mystery. A BMS reporting a healthy pack while the inverter sees no voltage points at the DC path between them. Send us your monitoring and we will tell you where to look.

Ask about a dead batterySolis fault codes

Brilliant support to get my solar battery working again. I didn’t expect help on a Saturday but Ron answered the phone, listened and sent me the information I needed to get it going, answered questions etc. A brilliant service I’d happily recommend.

Alison F. Cockerill · Jun 2026 Google

Genius when it comes to solar and off grid installations we spent days trying to work out what was causing our problems and Ron resolved it in less than an hour! Cant reccomend enough

Ben Carroll · Aug 2026 Trustpilot

Fantastic support. We have a Givenergy system where one of our batteries was not working. Ronald successfully helped us get it back up. His perseverance was amazing….and it shows what REAL customer service means. Can’t thank you enough Ron!

Martin · Aug 2026 Google

I work for a Solar Company and had a customer with a GivEnergy system that we are not versed in. Ron took the time to explain the issues my customer was having and between us managed to rectify the issues. 10 Stars ... Thanks again STS

Craig Daly · Jun 2026 Google

Ron is a life saver sorted out a connectivity problem and had to replace one of my inverters all good now so I can stop stressing. Thank you again

Glenn Hempstead · Oct 2026 Google

Very happy with the support I got from Ron at Solar Tech Support with my GivEnergy battery problem. The battery having operated perfectly for 3 years suddenly stopped discharging. He sorted it out in minutes.

Paul Elliott · Sep 2026 Google

Some faults announce themselves. A burnt connector, a cracked case, a smell. You find those in the first five minutes. This one was a switch that was already off and looked like it was on.

The call

Customer with a Solis inverter and a Pylontech battery stack. The battery had simply stopped working, with no charge, no discharge, and the house back on full grid import.

What the portal said

Before going anywhere, we looked at the data. Two things stood out.

On the evening it failed, the system was entirely normal right up to 18:00: battery voltage 48.6 V, battery current 63 A, power flowing, status online. A working battery doing exactly what it should on a summer evening. Then it stopped.

SolisCloud chart for the day the system failed. Battery voltage holds near 48.6 V all day and battery current is active, then both collapse to zero just after 18:00. The tooltip reads 18:00:05, total power 0.181 kW, battery voltage 48.6 V, battery current 63 A, status Online.
The day it failed. Battery voltage sat near 48.6 V and the battery worked right up to 18:00, the tooltip catching one of the last healthy readings, and then everything drops to zero at once.

With the system down, the same portal was reporting battery voltage 0.5 V. And here is the part that matters: at that same moment it was also reporting battery state of charge 78% and a maximum discharge current from the BMS of 80 A. An alarm had been raised against the system.

SolisCloud chart from a day with the system down. Solar is still generating through the day, but the alarm trace runs red along the bottom from first light and the battery state of charge sits flat. The tooltip reads 18:58:46, total power 0.007 kW, battery voltage 0.5 V, battery state of charge 78 per cent, maximum discharge current from the BMS 80 A, with alarm 1055 flagged.
With the system down. Solar still generates, but the alarm trace runs along the bottom all day and the battery does nothing. The tooltip holds the combination that gives the fault away: 0.5 V at the inverter, and a pack simultaneously reporting 78% and an 80 A discharge limit. Note this is a different day from the chart above; the generation curve and the power axis are not the same.

Stop and think about what that combination actually means.

  • The BMS is alive. It is reporting a state of charge, and it is reporting its own discharge limit. Those numbers come from inside the battery. If the battery were dead, or disconnected from the inverter entirely, they would not be there.
  • The pack is 78% full. Whatever has happened, it has not been drained.
  • But the inverter is seeing effectively no voltage at its battery terminals.

A charged battery that is still talking to you, and no voltage where the voltage should be. That is not a flat battery and it is not a dead BMS. That is an open circuit somewhere in the DC power path.

Why comms and power tell you different things

This trips people up constantly, so it is worth being explicit.

The battery talks to the inverter over a separate communications cable, CAN or RS485, a thin data lead. The actual energy travels over big DC cables, through an isolator or a DC-rated MCB, into the inverter's battery terminals.

Two different paths. Break the power path and the comms path carries on regardless. The BMS keeps chatting away, cheerfully reporting a healthy 78% pack, while the inverter sits there seeing nothing at all.

So when you get a healthy battery and zero volts, the fault is almost always somewhere between the two: the DC cabling, an isolator, or a breaker. That is the same signature behind a lot of batteries that will not discharge, and it is worth separating from a genuine battery communication fault, where the data stops arriving too.

The breaker

Which brings us to a small grey enclosure on the wall, marked in marker pen: BATTERY BREAKER.

The battery breaker enclosure on a white-painted brick wall, mounted below the inverter and above a Pylontech US5000 battery module. Red and black DC cables run down from the enclosure to the battery.
The enclosure in context, between the inverter above and the Pylontech US5000 below. The red and black DC cables leaving the bottom are the entire power path between the two.

We looked at it. The toggle was up. Up means on.

Close-up of the enclosure, labelled BATTERY BREAKER in marker pen. Through the clear hinged lid the MCB is visible, a GivEnergy PEBS-L 100 A device, with its toggle sitting in the upper position.
Through the lid, the toggle reads as up. It was not: the lid was fouling it and holding it there. The breaker is a GivEnergy PEBS-L, which is worth a word on a Solis and Pylontech job. Installers reuse DC breakers across brands, and nothing about this fault is brand-specific.

Except it was not.

The enclosure lid was fouling the breaker's toggle and physically preventing it from dropping. The MCB had tripped internally, the mechanism had done its job and broken the circuit, but the arm could not travel far enough down to show it. From the outside it sat in the on position, looking for all the world like a healthy breaker.

We opened the enclosure and the toggle dropped down on its own the moment it was free.

That is the whole fault. A tripped breaker that had no way of telling anyone it had tripped.

The fix

Reset the MCB, power cycle the inverter, and the system came straight back up. Battery charging and discharging normally, customer back to saving money the same afternoon.

No parts. No return visit. A fault that would have looked, to almost anybody, like a dead battery.

What we told the customer

A tripped breaker is a symptom, not a cause. Something made it trip, and we do not yet know what. So the advice was:

  • Check the breaker periodically. Now you know the toggle cannot be trusted from the outside, open the enclosure to look properly.
  • If it trips again, tell us. Once is an event. Twice is a fault, and the next step is to find out what is drawing enough current, or what is degraded enough, to keep operating it.
  • Do not just keep resetting it. A breaker that trips repeatedly is doing its job. The answer is to find out why, not to keep putting it back up.

The lesson

If your battery has stopped working, check the DC breaker before you conclude anything expensive, and open the enclosure to check it. Do not trust the position of the toggle through a lid that might be holding it. If you need to isolate the system properly first, our shutdown guide covers the order to do it in.

More generally: when the monitoring shows you a battery that is clearly alive and clearly charged, but the inverter cannot see any voltage, resist the urge to condemn the battery. The energy and the information travel by different roads. Follow the one that is broken.

Solar Tech Support is an independent solar and battery repair company. We are not affiliated with Solis, Pylontech, or any manufacturer. Most faults we diagnose remotely, free, no fix no fee.

FAQ

Frequently asked questions

It usually means the power path is broken while the data path is still working. The battery talks to the inverter over a thin communications lead, and the energy travels separately over large DC cables through an isolator or a DC-rated breaker. Break the power path and the BMS carries on reporting a healthy pack over the comms lead, because nothing is wrong with the battery. So a pack that reports a sensible state of charge while the inverter sees little or no voltage points at the DC cabling, an isolator or a breaker between the two, not at the cells.
Yes, and it is easy to miss. A breaker trips by releasing its mechanism internally, which normally lets the toggle fall to the off or mid position so you can see it has gone. If anything obstructs that travel, the mechanism can open the circuit while the toggle stays up. On this job the enclosure lid was fouling the toggle, and it dropped on its own the moment the lid came off. If a battery has stopped working, open the enclosure and look at the breaker properly rather than judging it through a lid.
No. It only proves the BMS is powered and talking. State of charge, cell voltages and the maximum charge and discharge currents a BMS advertises all come from inside the battery and travel over the communications cable. They will keep arriving even when the DC power connection is open. Treat them as evidence the battery is alive, which is genuinely useful, but not as evidence that the energy has anywhere to go.
Once, having looked at it, is reasonable. Repeatedly is not. A breaker that trips again and again is doing its job and reacting to something real, and resetting it over and over hides the cause while any damage gets worse. If it trips a second time, stop resetting it and get the DC side inspected by someone qualified. If anything is warm, discoloured or smells hot, leave it off and get it looked at urgently.
No. Nothing about this fault was brand-specific. The same combination of a live BMS, a healthy state of charge and no voltage at the inverter turns up on GivEnergy, Fox ESS, SolaX, Sunsynk and the rest, because they all separate the comms path from the power path in the same way. What matters is that whoever looks at it works the DC path rather than condemning the battery on the app reading.
The diagnosis was made before anyone travelled. The portal data on its own said the BMS was alive, the pack was 78% full and the inverter was seeing effectively nothing, which narrows the fault to an open circuit in the DC path and rules out the expensive answers. What could not be done remotely was opening the enclosure and looking at the breaker. That is the usual split on this kind of job: the data tells you where to go, and someone still has to put their eyes on it.

Battery stopped working but still showing a charge? Send us the data.

Tell us what the app is showing — state of charge, any voltage or current readings, and any alarm on the portal. That combination usually narrows the fault before anyone travels, and if it is something you can safely check yourself, we will say so.

By submitting you agree to be contacted about your diagnostic request. We don't share your data with 3rd parties.

Call 07944 877 329 Book free diagnostic