• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Batteries suddenly wildly out of balance

smokerx

New Member
Joined
Dec 31, 2022
Messages
23
I have 2 AOlithium 12 volt 100AH with bluetooth and 2Humsienk 12volt 100AH parallel and series. I monitor their balance with the AOlithium SOC display on the app, monitoring their individual SOC. It hasnt reported falling out of balance by even 1 percent. This morning I woke up and 1 battery is in over discharge protection at 0% SOC and the other is reporting 38%. I had to go to AOlithiums website to look up Over discharge Voltage protection and release numbers, as I realized its not published in their manual. The one at 0% is slowly rising in voltage from being paralleled and series. But its at 2.72V per cell and saying it wont wake up until 3.0 So Im going to be here a minute without buying a "12volt" "smart charger"...nice.

I only had a 100 watt fridge on it, but it looks like the 400 watt defrost kicked in on the fridge at the crack of dawn and ruined me. I'm starting to think bluetooth SOC is just complete nonsense. Im not liking a 400 watt load slam hammering a battery into BMS protection at 38% reported SOC either. If the Battery would come out of protection, I can solar charge it. Ive pulled all the load off and the sun is out to produce the AIO inverters load.
 
Last edited:
Well, it sounds like one of the batteries has some severely imbalanced cells which wasn't obvious until the entire pack it was part of was drawn down far enough in SOC.

If the battery had never been down into that depth of discharge before, its own BMS wouldn't know about this issue either and whatever SOC it reported would not be realistic anyway.
 
Its cycled like 10 times and I never let it sit in that 100% float charge window. I usually do 50% to 80% as I dont like the idea of sun coming in and not harvesting it. Also I have my + - cables running to the AOLITHIUMS, where Ive seen diagrams of the poles coming off either side of a 4 way parallel series, to better distribute current. Also I dont have that square of cable wired into the center of a 2X2 either
 
Its cycled like 10 times and I never let it sit in that 100% float charge window. I usually do 50% to 80% as I dont like the idea of sun coming in and not harvesting it. Also I have my + - cables running to the AOLITHIUMS, where Ive seen diagrams of the poles coming off either side of a 4 way parallel series, to better distribute current. Also I dont have that square of cable wired into the center of a 2X2 either
Pretty much all BMSes are pretty bad at SOC accuracy. Voltage won't work for LiFePO4 in the middle ranges so the people writing the code went to using 100% current measurement and that also does not work because really accurate current measurement is hard.

If you do not get a real 100% charge on the battery every so often (say every week or so) then the SOC can be significantly wrong. The BMS adds up the current in and out of the battery and most (all?) BMS'es seem to have inaccuracies in their current measurement across the range. So if you charge in a range were current is 2% high, and discharge in a range were current is 2% low most of the time then the battery SOC % slips a few percent per week.

Some of the newer BMS firmware updates (for some brands) stopped reporting 100% until the battery was really 100%(stays at 99% until the voltage spikes and confirms 100%), prior to that it counted charge and reports 100% when it THOUGHT the battery should be 100%(based on current, ignoring voltage obviously not being 100%) and with that even charging to 100% every day fails because reported 100% becomes real 98% after a day, and slips a bit more each day, until eventually you have a SOC of 100% with a real SOC of 40%, and when you discharge that real 40% your normal 40% daily discharge the voltage will hit the low cutoff and the BMS will reset to the voltage indicated 0% in an instant.
 
Its cycled like 10 times and I never let it sit in that 100% float charge window. I usually do 50% to 80% as I dont like the idea of sun coming in and not harvesting it. Also I have my + - cables running to the AOLITHIUMS, where Ive seen diagrams of the poles coming off either side of a 4 way parallel series, to better distribute current. Also I dont have that square of cable wired into the center of a 2X2 either
Ok I checked it with a multimeter. The bluetooth is a liar. The battery is reported at 10.91 but reads 6 volts on my meter. BMS didnt protect nuffing. Im contacting AOlithium and seeing if theyre going to RMA this. No point in trying to charge it. That makes sense that the bluetooth monitoring would show them balanced at 38% SOC then suddenly the system starts beeping no battery is connected and the batteries SOC suddenly crashed to 0%. What the BMS did do is slam the over discharge protection state valve open when the battery crashed, is my guess. I also wondered why the AOlithium's suddenly disappeared off the market to never be restocked too. Hence, the 2 humsienks.
 
Last edited:
battery is reported at 10.91 but reads 6 volts on my meter.
If the battery is in protection, any meter reading is incorrect. All you are seeing is leakage across the 'off' fets. The battery won't reach recovery voltage until over 11 volts. This will need a 12v dumb charger or a 'jump start' form another 12v battery connected in parallel for a few moments. Then you can charge normally with a 12v charger.
When you have batteries in series they gradually diverge in state of charge, eventually, when in use, the lower charged battery will reach low volt protection and shut down the series pack.
Before putting the 12v batteries in series each needs a full charge with a 12v charger. Either break down the pack and repeat every few months or fit a battery balancer to each series pair. For a 24v system the ideal is to use 24v batteries.
 
Last edited:
Ok, mike fits, it doesnt hurt anything to try. I hooked both AO lithiums up on the counter in parallel...and to my amazement something actually happened and the protection state closed....Weird since its reading 2.87 volts, and it says it shouldnt come out of discharge protection until 3.0. But hey, Im going to take it. Since theres no reason to try to charge a 6 volt discharge...Nobody said anything about reparallelizing it on the counter...

I drew a picture of my wiring diagram to show whats going on here. Green X is the battery that slammed out to 0
 

Attachments

  • mceclip1.jpg
    mceclip1.jpg
    32.1 KB · Views: 22
Last edited:
A MikeFitz said, putting BMS-based batteries in series is very likely to lead to the exact situation you find yourself in. I can explain this in more detail if needed.

I would charge each battery individually.

Long term, use 24 volt batteries so you don't have to deal with the series issue. If this isn't an option then either get a battery balancer or schedule a manually charge of the (disconnected) batteries one at a time.
 
Ok I got my batteries back online, attempted to balance them, by surface voltage, by parallelling them on the table, then switched the negative lead over so its staggered. Now Im trying to charge them up on solar to get them to 100%. I'll keep charging them up on solar. then probably order a small 12 volt charger if I cant get them balanced and manually balance them, and start thinking about a battery balancer.

When I priced out 12 volts vs. 24 volt batteries, it was way cheaper to build a bank of 12 volts for a 24 volt system. 24volt batteries would have at least doubled the price.

Also I have a 12 volt inverter I can rob charge off single batteries with but no lithium 12 volt charger to manually balance them.
 
Also I have a 12 volt inverter I can rob charge off single batteries with but no lithium 12 volt charger to manually balance them.

An AGM charge profile is good. It may not get you to 100% state of charge, but it's safe. So if you charge all of them as far as possible on an AGM charge profile, they'll at least be at the same state of charge.
 
Also I have a 12 volt inverter I can rob charge off single batteries with but no lithium 12 volt charger to manually balance them.


If I am understanding what you are saying here it is probably the worst thing you can do.... hooking an inverter to 1 battery in a series string will cause a serious imbalance and the trouble you described at the beginning
 
it was way cheaper to build a bank of 12 volts for a 24 volt system
The suppliers don't help the situation with statements like 'up to 4 batteries in series or parallel'. The 12v batteries will be slightly different from each other, different capacities and internal resistance, so when charging as a series string they may not reach the same SOC. The BMS in each battery will keep the cells balanced in its battery but has no influence over the BMS in the other battery.
The idea of pulling down the SOC with a load on the higher capacity battery may work, but the Bluetooth SOC may not be a reliable indicator until the battery has cycled a few times.
When you get everything sorted and the pack operational, it may be useful to review the charge voltages you have set up in the charger.
If it's too high, its probable one of the series batteries will reach cell overvolts and shutdown charge in that series pack. This is undesirable and could compromise the balance between the series batteries. Since you have Bluetooth state of each battery this will be easy to detect and rectify if needed.
 
An external balancer for 4 batteries is about $50 and imo should be considered a non-optional part of the cost of making a 48v series string with 12v batteries. I have 4x 12v280ah with a balancer from day one and have never had an issue with it.
 
If I am understanding what you are saying here it is probably the worst thing you can do.... hooking an inverter to 1 battery in a series string will cause a serious imbalance and the trouble you described at the beginning
it says in the manual not to run it to 0% and leave it there. So I put my batteries back is 2s2p to charge them off the solar. To balance them by pulling load out with a 12 volt inverter, I would take them back out of 2s2p and use them solo.
The suppliers don't help the situation with statements like 'up to 4 batteries in series or parallel'. The 12v batteries will be slightly different from each other, different capacities and internal resistance, so when charging as a series string they may not reach the same SOC. The BMS in each battery will keep the cells balanced in its battery but has no influence over the BMS in the other battery.
The idea of pulling down the SOC with a load on the higher capacity battery may work, but the Bluetooth SOC may not be a reliable indicator until the battery has cycled a few times.
When you get everything sorted and the pack operational, it may be useful to review the charge voltages you have set up in the charger.
If it's too high, its probable one of the series batteries will reach cell overvolts and shutdown charge in that series pack. This is undesirable and could compromise the balance between the series batteries. Since you have Bluetooth state of each battery this will be easy to detect and rectify if needed.
I think its actually cheaper to get raw cells in these 12 volt cases. If need be, I have enough cells to create 2 24v 100AH batteries, cut 2 battery case sides off, plastic weld them together and replace the BMS, Before I do that Im going to try to balance them out, test them again with the Negative now attached to the other series pair, check them for balance then try wiring in a balancer.

Its different now that I have diagnosed this situation vs, waking up to a sudden SOC drop, with 3 hrs sleep, wondering why my battery is reading 6 volts. Im not thinking I can rely on bluetooth for any accurate description of any perimeter, now.
 
it says in the manual not to run it to 0% and leave it there. So I put my batteries back is 2s2p to charge them off the solar. To balance them by pulling load out with a 12 volt inverter, I would take them back out of 2s2p and use them solo..

So, the problem with what you are describing is you need to have the batteries fully 100% charged before you put them in series or parallel.

If you break them apart and use them, one or more, as independent batteries and then put them back in series to charge with solar ther is no surprise at all that one hits full before the other and it cuts off the charge.
 
So, the problem with what you are describing is you need to have the batteries fully 100% charged before you put them in series or parallel.

If you break them apart and use them, one or more, as independent batteries and then put them back in series to charge with solar ther is no surprise at all that one hits full before the other and it cuts off the charge.
Argument ad hominem, When the higher charged batteries hit 100%, Im taking them offline, discharging them to match the batteries that havent hit 100%, then reinstalling them 2s2p onto the "charger". Right now Im only using the solar system to charge the batteries in the sun. After the sun goes low in the sky, I turn the battery disconnect switch to off, turn the solar panel disconnect off and turn the inverter power switch to off. right now theyre reporting a difference of .09 volts according to the app. But the SOC is reporting 40% charge state difference.
 

diy solar

diy solar
Back
Top