2 battery setup - disconnected and working only on one battery

I posted this on the Facebook group but had no replies so thought I would try on here.

I have another "issue" with my 2 battery setup - which Martin has previously suggested a calibration - which sort of worked in that they did both come back online and work together.
But the other day, the 8.2kWh Gen1 decided to just sit there and not charge or discharge. So I waited until the evening and set a calibration off, which it started at the prescribed time and started to dump to grid.
Next morning, it was still in calibration mode and the master 9.5kWh Gen 3 battery had discharged, but the 8.2kWh battery was just the same.
Rebooted the inverter and the BMS voltage dropped to something like 8v - so I turned off the 8.2 and rebooted and all was well and the system worked fine all day (But reduced capacity)
I think the basic issue is a mismatch of battery voltages, so as calibration isn't sorting the 8.2 out, I thought I would force discharge the 9.5 and see if they can get close enough to be happy to re-join.

No such luck and the 8.2kWh is sitting at about 11% SOC and 3.2V per cell (51.268 total)
The 9.5kWh has recharged overnight and is sitting at 3.31V per cell and 53.051 overall.
Main overall BMS voltage is 51.31V

Anyone got any ideas what I can do to sort out my system? Its not particularly urgent at this time of year, but would be nice to be able to sort it out and whilst I will contact Giv Support officially if i can't sort it out, they probably have more urgent issues to deal with at the moment!

Update from this morning. The primary battery was discharged fully last night, with the two battery voltages at one point At leat, being aligned but it didn’t reconnect the second battery and it’s still hanging.

Welcome back!! I only have one battery, so I have no experience that can help with your issue.

I know you run HA, but are you running my Android app as well? If so, is there anything useful recorded in the fault log?

When doing a calibration, always record the "full" capacity for each battery before and after. What are the current capacities? What firmware version are you running on the batteries?

I too don't have 2 batteries, but if you turn off the one that currently works, can you force the other to 100% and see if it catches it at the top end? (You may need to do a bit of see-saw battery on/off to get them both to 100%)
I will admit I don't know how smart the whole Master-Slave thing is - the system may simply not do any charging if it can't see the master I guess.

There is a command Givenergy (& meddlers if you try hard enough) can use to force a pack online - mainly used to try and awaken ones that have gone too low voltage wise & hibernated - but might be of some use here?

edit - i guess the rs485 cabling may have come loose so the system can't "see" it? Does the portal still show 2 packs & the right AH your expecting?

Thanks for the hello on my return! Been busy and now semi retired (aiming for full retirement no later than Christmas!)

I can see the second battery via givtcp and the givenergy app also sees the second battery so I have no reason to suspect the modbus cabling.

I am pretty sure if I turn off the master battery it won’t work at all - I’ve not tried it and can’t until Wednesday at the earliest as we are away in deepest darkest wales at the moment !

When went the first time Martin suggested a calibration- which I did but it didn’t affect the slave battery and it got stuck and I had to turn off the slave to make it do anything. (I had turned off the whole system prior to this and the common battery voltage was around 8v - despite both batteries reporting around 50-53v

This time I tried again before turning anything off and was the same as before - got stuck and didn’t affect affect the slave. Though this time when I turned the slave back on the common voltage was ok and has been fine since. Just working on one battery.

The portal sees two packs but only the capacity of the master.

Rubikcube oh and I run iOS so don’t have your app.

Actually I do have a tablet on Android - I assume I need to be local for it to work?

anglefire Same as GivTCP, app just needs an IP Address (of your inverter). Local is simplest, but some people have it running over a VPN. The networking is up to you. A tablet is much better than a phone (less mobile and larger screen).

https://givenergymonitor.wixsite.com/tutorials/getting-started

anglefire deepest darkest wales

Can be found here https://www.go-below.co.uk/

I will look when I’m back home. Currently away in wales - in the rain now after a scorching weekend!

I have two 9.5kWh GivEnergy batteries. Occasionally they go out of sync – meaning there's a voltage difference of more than 0.5V between them. They often hover close to that limit, as shown here:

If the voltage difference grows too large and one battery exceeds around 55V, that battery stops responding. It locks at 100% and stops responding while the other battery does all the work and drains itself flat. The only fix I’ve found is to run an overnight charge, which brings the lower battery’s voltage back in line with the “frozen” one.

I posted this in the GivEnergy Owners Facebook group back in May 2024:
“I'm so annoyed with my GE system right now. I have a Gen 2 5kW inverter with 2x 9.5 GE batteries. The Secondary battery is always borking. When it hits 100%, it then overvoltages, goes out of sync with the Primary, and shuts down. This has happened twice this week and about 20 times in total. Today it was worse – the Secondary battery actually crashed the inverter at 4pm, right when it was meant to start exporting. Of course I was at work and couldn’t do a thing. I’m fed up with how unreliable it is – I can’t trust it to behave. Yes, it usually sorts itself during the 2–5am Flux charge, but once again I missed out on peak export from 4–7pm, and that’s money lost.”

Since moving to Home Assistant in Sept 2024, I’ve only had this happen once. I realised the problem tends to occur on very sunny days when both batteries hit 100% before 2pm. But if both batteries are full before 2pm, and it’s still sunny, the inverter keeps trying to charge them – which pushes the voltage too high and triggers a battery freeze.

To avoid that, I created Home Assistant rules:
From 8:00 to 14:00: Export to grid if SOC is above 80% (reduced if not summer, or poor weather forecast)
From 14:00 to 16:00: Allow normal charging
From 16:00 onwards: Begin exporting for the Octopus Flux window

This strategy delays full battery charge until later in the day, reducing the risk of overvoltage. Even if the batteries don’t quite reach 100% by 16:00, it’s worth it to avoid the freeze issue. Since setting this up, I haven’t had a single battery lock up.

miffy the secondary inverter isn’t charging even when the system is force charged when it’s off peak. I think what I shall do when I get back home is discharge the primary until it’s voltage is about the same and do a system shutdown and restart and hope that the voltages are close enough to then reconnect because they aren’t doing it automatically.

I don’t actually know why they disconnected originally because it’s been really stable since I had the second battery (and uprated inverter installed back in February) until a couple of weeks ago. The batteries never fully charge - get to about 98% ish - the secondary often a bit lower than that (maybe the issue!) as it’s all controlled by Predbat

Last night I managed to get the primary battery voltage to within 0.5v of the secondary, reset the batteries and inverters and they reconnected as one and the capacity returned to the full 17kWh

Re-enabled Predbat which started the overnight charge, woke this morning to find they had disconnected again and the secondary battery still at 9%

Looks like I’m going to have to contact support. Bugger.

Raised a support ticket on mine this morning after trying the live chat (Which didn't get answered)

Case number 102463 - obviously there is no one working until monday - but even then its upto 5 days for a first response.

After reading a post on FB from Paul about another system which said to try turning off the batteries for 5 minutes in case they had gone to sleep, I tried it last night - after turning them back on I found the 8.2 DC breaker off - switched it back on, but then it went into the common battery voltage flipping between 7 and 52v (And the reported capacity switching between 8.2 and 9.5kWh)

So I turned off the 8.2 battery again - or so I thought - as I woke this morning to find both batteries charged up..... Must have just done a reset not a power off. Very odd!

But happy its working again ok!

[unknown] that's over 17k support tickets since mine on 22nd May - ouch.

DC Breaker on the secondary battery tripped again last night - it coincided with the primary battery getting to 100% - which is a bit odd.
Not sure why a 100A breaker would trip though - should be in the order of 70A max.
No sign of overheating.
Reset it this morning and Predbat decided to charge it up a bit (Not really sure why, but not overly concerned) - which it did for several minutes at seemingly full chat. I do have a DC clamp meter - but getting it onto the tails safely is a challenge!

I guess if pack 1 is full and pack 2 isn't then all the current get sent to pack 2 rather than shared, which would lean me towards the MCB terminals being loose.
Other thoughts:
I'm unsure how the BMS manages current flow when in multiples as I've never experienced it BUT if it's not the usual (soggy terminals coming loose in the MCB) could it be related to the pack imbalance & high current flow from the full one to the empty(er) one?
I imagine that is why they have the 0.5V restriction between packs to stop a 100% full pack discharging 100's of amps into an empty one? I don't really know what comms get passed between masters & slaves to be honest and how this kind of thing is all managed beyond the only real ability it has which is to go open circuit at anything it doesn't like.

My thought was terminals being loose in the breaker too - but I've checked and they seem ok to me - not that I have a torque screwdriver to check actual tightness - but they are properly crimped - having said that, there is no requirement for a DC isolator between batteries according to the install instructions.

hoggy Oh and yes you are correct - but in a single battery the same current is pushed in as the limit is the inverter not the battery as such.

Currently the primary battery is drifting down slowly from 100% and is now 84% since this mornings reset and the secondary is now 48%, up from 30% after its initial grid charge - about 230w is going in from the solar at the moment.

Bit the bullet and did a calibration yesterday - started at 15:08, took until 19:02 to empty both batteries and then charged backup until 2:45 this morning.
Overnight Predbat did its thing and now they are within a nats cock of each others voltage and the primary is 90% and secondary 90% - which is just an estimate obviously. Will see how it goes over the next few days.

Gone again today - secondary battery disconnected from the primary, total capacity dropped back to 9.5kWh.

DC breaker is fine and hasn't tripped.

Bit of a loss as to why it is doing it - has worked fine for 6days.

I guess it may rejoin again when the battery re-charges tonight.