Inverter Turning off due to Fault for low voltage battery on my DIY battery.

[Nonlin]

So my custom build solar system + DIY Battery (16s EVE 304ah Batteries from 18650batterystore) finally started doing something concerning yesterday.

At around 14% reported SOC from JK BMS the inverter just turned off. At first I thought it was my Growatt LVM 3000 48p failing but I noticed the SOC LED Display for my JK BMS jumped to 100% before going back down to 14% before again shutting off my inverter on my Growatt hybrid (This of course when I manually restarted the inverter on my Growatt to check what was going on).

So looking at the BMS I see a Cell Volt. Diff of 0.089 so I figured my cells need balancing.



But now that I'm at near 60% SOC reported the difference is only 0.001 between cells.


The settings on my JK are to start balancing at 3.45 volts and no cell has hit that yet.

is this an indicators that my cells are bad, very different internal resistances?

Or is something else going on?

My current goal is to just fully charge the battery atm, and see if the active balancer starts balancing at all

 

[Partimewages]

Do you have a smart shunt in the system?
What you are describing is some cells out of balance. More of an issue would be if you're only using your BMS to measure SOC. They are generally not a good gauge.

 

[Nonlin]

Do you have a smart shunt in the system?
What you are describing is some cells out of balance. More of an issue would be if you're only using your BMS to measure SOC. They are generally not a good gauge.

Yes I'm only using the BMS for SOC, no smart shunt.

 

[hwy17]

Might just be mis calibrated SOC?

You'd have to see the cell voltages under load near shutdown to tell if it was one going low or they were all going low. It looks like all low to me.

 

[Mattb4]

Though you might have a balance issue it sounds more like you simply had enough load on a nearly depleted battery. As your battery gets depleted it gets harder for it to maintain voltage. Also at lower voltage the more amperage is needed for the same inverter load. Once your inverter trips on low voltage the battery voltage jumps back up some.

 

[Nonlin]

Though you might have a balance issue it sounds more like you simply had enough load on a nearly depleted battery. As your battery gets depleted it gets harder for it to maintain voltage. Also at lower voltage the more amperage is needed for the same inverter load. Once your inverter trips on low voltage the battery voltage jumps back up some.

That's interesting? I usually just keep my EV plugged in to charge, consumes the same app/load. And I've had it drain my DIY battery down to 3% before shutting off before. This time it stopped at 14% so I wasn't expecting that to happen.

 

[Nonlin]

Might just be mis calibrated SOC?

You'd have to see the cell voltages under load near shutdown to tell if it was one going low or they were all going low. It looks like all low to me.

My understanding is I'd have to fully charge the bank once to get a "calibrated" SOC which I technically haven't done yet. I'm doing this now though.

 

[Partimewages]

How do you know that the bank is being fully charged?

 

[Steve_S]

Go into your AIO and check the Low Volt Disconnect setting there.
Personally I set LVD on Inverter to cut off at 2.800VPC or 44.8V on 48V system.
I then set the BMS to cut off (UVP) to 2.700 and recovery (UVPR) to 2.800
Set the BMS to start Active Balancing @ 3.42 Volts.

NOTE !! IMPORTANT !!!
Use a Multimeter and check the voltage @ your Battery Terminals (2 decimal point) Then check the Voltage at the Growatt AIO's terminals an correct for the losses between the physical battery & AIO. There is ALWAYS some losses and you need to voltage correct for that. SO DO THIS WITHOUT LOAD OR CHARGE GOING IN - AT REST Essentially. Then Check again with Load going out & when Charge is coming in. You WILL see a difference, these small variances must be accounted for.

Clarity:
A Bank is a set of complete batteries (cells with BMS) set in Series Parallel acting as one storage system.
A Battery is a group of cells with BMS operating as a complete unit.

 

[Mattb4]

That's interesting? I usually just keep my EV plugged in to charge, consumes the same app/load. And I've had it drain my DIY battery down to 3% before shutting off before. This time it stopped at 14% so I wasn't expecting that to happen.

Loads are in watts. If your battery voltage drops amperage it supplies must rise W=VA.

Capacity readings for batteries are not a precise thing. Even with all the shunts or other types of monitoring. Batteries are chemistry based and not equivalent to a volume measurement. The difference between a 14% capacity and 3% capacity is basically nothing if a condition changes such a temperature or measurement inaccuracy happens.

 

[Brucey]

My understanding is I'd have to fully charge the bank once to get a "calibrated" SOC which I technically haven't done yet. I'm doing this now though.

The BMS may not register small loads like for example a smartshunt would do. So it may think it's soc is at 14% but it's basically empty. Definitely give it a full charge and let the jk active balancing do it's thing at the top end.

 

[Nonlin]

Go into your AIO and check the Low Volt Disconnect setting there.
Personally I set LVD on Inverter to cut off at 2.800VPC or 44.8V on 48V system.
I then set the BMS to cut off (UVP) to 2.700 and recovery (UVPR) to 2.800
Set the BMS to start Active Balancing @ 3.42 Volts.

NOTE !! IMPORTANT !!!
Use a Multimeter and check the voltage @ your Battery Terminals (2 decimal point) Then check the Voltage at the Growatt AIO's terminals an correct for the losses between the physical battery & AIO. There is ALWAYS some losses and you need to voltage correct for that. SO DO THIS WITHOUT LOAD OR CHARGE GOING IN - AT REST Essentially. Then Check again with Load going out & when Charge is coming in. You WILL see a difference, these small variances must be accounted for.

Clarity:
A Bank is a set of complete batteries (cells with BMS) set in Series Parallel acting as one storage system.
A Battery is a group of cells with BMS operating as a complete unit.

I did do something like this to "Calibrate" the difference since the JK had that options, but maybe I did it poorly, pretty sure I had no charge or discharge when I did it though. Do I use the rest difference or the load difference Or take some average?

 

[Nonlin]

How do you know that the bank is being fully charged?

Voltage of the cells?

 

[Checkthisout]

So my custom build solar system + DIY Battery (16s EVE 304ah Batteries from 18650batterystore) finally started doing something concerning yesterday.

At around 14% reported SOC from JK BMS the inverter just turned off. At first I thought it was my Growatt LVM 3000 48p failing but I noticed the SOC LED Display for my JK BMS jumped to 100% before going back down to 14% before again shutting off my inverter on my Growatt hybrid (This of course when I manually restarted the inverter on my Growatt to check what was going on).

So looking at the BMS I see a Cell Volt. Diff of 0.089 so I figured my cells need balancing.

View attachment 206230

But now that I'm at near 60% SOC reported the difference is only 0.001 between cells.
View attachment 206231

The settings on my JK are to start balancing at 3.45 volts and no cell has hit that yet.

is this an indicators that my cells are bad, very different internal resistances?

Or is something else going on?

My current goal is to just fully charge the battery atm, and see if the active balancer starts balancing at all

Your first picture shows your battery at 45 volts or like 2.8 volts per cell.

The battery is dead. That's why the inverter shut down.

SOC can be ignored as it's only a rough indicator most times of actual charge level.

 

[Steve_S]

I did do something like this to "Calibrate" the difference since the JK had that options, but maybe I did it poorly, pretty sure I had no charge or discharge when I did it though. Do I use the rest difference or the load difference Or take some average?

OK here we go....
Calibrating the BMS/Inverter etc, you need a Really good DMM/DVOM that can do 2 decimal accuracy for one.
1) Check the voltage of the pack @ the Pack Terminals (NO Load/Charge - IDLE) (main batt terminals) and use that to set the calibrated voltage on the BMS.
The BMS is Independent and each one must see what the voltage is at "their own" terminals. If using multiple packs in parallel & interconnected via CAN/RS etc this is even more important.,

2) Then with the AIO IDLE (No AC going out & no Sun coming in) check the Batt Terminal Voltage on the AIO and note the difference.
This is your Idle Differential. THIS IS NOT THE SAME when you are pulling from Batt or when Charging Batt... THIS IS THE PITA because they are always different. BUT you can manually adjust (compensate) with the settings to make up the difference.
!~ Growatt as far as I am aware does not have OFFSET settings within, this is typically found in Premium Tier-1 gear. So you have to do it Manually.

3) The Voltage drop should be reasonably minimal from Battery Terminal to AIO but even a 0.01 Difference can result in "quirks". Keep in mind that every Lug, Terminal, Fuse/Breaker etc ALL add resistance & contribute to voltage losses along the path. Some folks are shocked at the losses through "some" fuses !

Where tweaking is required:
Inverter - Low Volt Disconnect Value
Inverter-Charger - Low & High Volt disconnect Values.
Solar Controller Settings - ENDAMPS/TAILCURRENT if applicable. *
* This is what triggers transition from Bulk/Absorb to FLOAT.
EndAmps/TailCurrent is calculated as follows: (100AH X 0.05 = 5A), 304AH X 0.05 = 15.2A

 

[Partimewages]

Voltage of the cells?

Voltage is a very poor indicator of SOC for lifepo4 batteries. You have some very good tips in this thread.

 

[Checkthisout]

Voltage is a very poor indicator of SOV for lifepo4 batteries. You have some very good tips in this thread.

I'm curious why we think this is the case when literally every Coulomb counter gets out of sync and displays the incorrect state of charge.

BMS's get out of sync, Inverters get out of sync, Separate shunts get out of sync.

The way they sync back up is by having the fully charged voltage held at a certain point for a certain period of time.

 

[AntronX]

Calibrate your BMS SOC counter. First screen shows 1% actual SOC and second is 50% based on cell voltage.

 

[hwy17]

Voltage is a poor indicator in the flat but it's a good indicator in the knees. Like we can be confident that a cell resting at 2.8v is basically running on emergency reserve and should be considered dead.

 

[Nonlin]

Voltage is a poor indicator in the flat but it's a good indicator in the knees. Like we can be confident that a cell resting at 2.8v is basically running on emergency reserve and should be considered dead.

Dead meaning it will never charge again or dead meaning its safe 0% SOC?

I think the spec sheet says it can go to 2.5 but I obviously don't want to go that low ever.

 

[hwy17]

Dead meaning it will never charge again or dead meaning it's safe 0% SOC?

Like 5% yeah, not dead dead like it's hurt.

My BMS cutoff is 2.5. Hitting 2.5 under load won't kill them.

You might have hit 2.5 and they bounced back to 2.8 at rest.

 

[AntronX]

Safe 0% SOC. See this chart below showing SOC vs. rested open circuit voltage (no amp load). No point in going below 5% SOC or 3.15v per cell.

 

[740GLE]

How are you charging these cells? The BMS will need adequate time where cells are above 3.45 for balancing to help the battery bank maintain top notch capacity.

Take a screen shot of the cell delta when the first cell hits 3.45 and balancing starts.

 

[AshleyL]

You sure you got the REAL 304AH battery?
The result looks like you have Grade B 280Ah battery instead......