Q-Dog
¯\_(ツ)_/¯
Is this one battery with a single BMS? or 2 or more batteries with individual BMSs?
Are my batteries cooked?
- Thread starter Steve Dally
- Start date
Brett V
Solar Wizard
Think of the cells in your battery as a chain and that chain is only as strong as the weakest link. Your battery capacity is only as much as the weakest cell. When the cells go out of balance, you stop charging and one cell is at a significantly lower state of charge than the other three.
Last edited:
sunshine_eggo
Victron's little biatch
Agree that there's sufficient imbalance to impact capacity and charging. Pretty much 0% chance it's @ 20% SoC. If cells are holding a given voltage at 0A current, high 3.3X is way closer to 100% than 80%.
I suspect this is a ~2-3% imbalance issue and a 97-98% user perception/true degradation issue.
SomethingAppropriate
Cannibalism starts with a kiss.
Wouldn't the fact that absorption time would solely be dependent on the pack voltage mean the spread could be so vast that the one at 3.6 hit that a long time ago and the absorption started when the low cell was at 3.30 and ended the second it hit 3.4?
Am I wrong in thinking that this could be the scenario?
sunshine_eggo
Victron's little biatch
It could be, but LFP cell voltages are pretty stable and find their resting voltage very quickly once charge is removed with very slow subsequent movement over time, and a drop from 3.39 to 3.30 is pretty unlikely unless there's an IR issue. If the voltages were observed at the exact moment the charge was terminated and then one immediately looked away, maybe there's a notable drop. My expectation is that the voltage was read casually when the battery was being supplied absorption voltage, but charge had terminated at least a minute or three before the voltages were noted.
To throw poo into the punch bowl, since you mentioned 3.30V explicitly, my experience is that cells resting at 3.29V are about 30% based on my as-received EVE cells and my as-received capacity test (out of the box I discharge each to 2.5V after recording voltage and IR).
@Steve Dally please identify BMS brand and supply settings.
Brett V
Solar Wizard
He said he’s charging to 14.0 so if balanced, that’s 3.5 per cell. He’s definitely got a balance issue but I’d make a rough guess it’s only costing him no more than 10% of his capacity.
sunshine_eggo
Victron's little biatch
With the low holding 3.394 @ 0A (assumption), I would be shocked at 10%. 2-3% is more likely, and even that may be high.
It's pretty astounding how even a small imbalance can interfere with charging. I've seen 0.1-0.2% imbalance interfere with normal charging.
Brett V
Solar Wizard
I’ll buy in to that. So let’s say it’s even a worse case scenario and he’s down 5%…then he’s got another problem. So either he’s not charging as high as he thinks, his loads are higher than he’s measuring, or the actual capacity of the batteries has degraded. The only way to know for sure is a load test on the individual cells.
Steve Dally
New Member
- Joined
- May 10, 2021
- Messages
- 30
To elaborate on this, BALANCE is about STATE OF CHARGE, not voltage. Voltage only correlates meaningfully OUTSIDE the normal operating range of 3.1-3.4V. Even a notably imbalanced battery will have a very tight cell voltages.
wut?
Thought I'd try a little something on a cheap 12V I've been testing...
Charge to 20A @ 13.6V (3.4V/cell), 2A tail current.
When testing at cell level and dealing with testers that don't have separate sense leads, lead resistance plays a big role, and I would typically observe cells get charged to about 95% SoC @ 3.4V.
Here's a plot of the charge:
0.2C charge
2.5 hour bulk
5 hour absorption
0.02C tail current
104.6Ah input (tested at 105.5Ah)
Have about a 0.12V drop in the leads @ 30A, so it's pretty negligible at 2A.
With sense leads, it would have spent more time in...
summary: A 100Ah cell held at 3.40V until the current drops to 2A will yield 99%+ SoC.
No. If you have a benchtop power supply, simply set it to 3.65V prior to connecting to cell and charge the cells starting with the lowest voltage first, cycling up through the cells and then hitting them again AFTER all 4 cells have been driven to 3.65V (takes just a couple minutes per cell) to ensure all are at 100% and very nearly the same voltage.
You can do this with the battery in-place and without disassembling. It will be helpful to remove loads.
What BMS? Please include your BMS settings.
BMS is RadioB Tech smart bms 200 amp 4s, I have 8 cells arranged 2p4s (I think that's the correct, 4 sets of 2 parallel cells.To elaborate on this, BALANCE is about STATE OF CHARGE, not voltage. Voltage only correlates meaningfully OUTSIDE the normal operating range of 3.1-3.4V. Even a notably imbalanced battery will have a very tight cell voltages.
wut?
Thought I'd try a little something on a cheap 12V I've been testing...
Charge to 20A @ 13.6V (3.4V/cell), 2A tail current.
When testing at cell level and dealing with testers that don't have separate sense leads, lead resistance plays a big role, and I would typically observe cells get charged to about 95% SoC @ 3.4V.
Here's a plot of the charge:
0.2C charge
2.5 hour bulk
5 hour absorption
0.02C tail current
104.6Ah input (tested at 105.5Ah)
Have about a 0.12V drop in the leads @ 30A, so it's pretty negligible at 2A.
With sense leads, it would have spent more time in...
summary: A 100Ah cell held at 3.40V until the current drops to 2A will yield 99%+ SoC.
No. If you have a benchtop power supply, simply set it to 3.65V prior to connecting to cell and charge the cells starting with the lowest voltage first, cycling up through the cells and then hitting them again AFTER all 4 cells have been driven to 3.65V (takes just a couple minutes per cell) to ensure all are at 100% and very nearly the same voltage.
You can do this with the battery in-place and without disassembling. It will be helpful to remove loads.
What BMS? Please include your BMS settings.
The xiou xiang app that connects to my phone controls all the parameters. I'll next try to take screen shot and upload it
Steve Dally
New Member
- Joined
- May 10, 2021
- Messages
- 30
Attachments
SomethingAppropriate
Cannibalism starts with a kiss.
Balancing tun on at 3.2 is crazy.
0.03 delta trigger is very tight, especially at 3.2.
I hope you can change that to 3.45.
That could be part of your problem.
0.03 delta trigger is very tight, especially at 3.2.
I hope you can change that to 3.45.
That could be part of your problem.
Steve Dally
New Member
- Joined
- May 10, 2021
- Messages
- 30
I don't have to disassemble first? Like I can leave them all connected to their bus bars? And if so, for the parallel pairs, do I treat each pair as a single cell, or should I charge each one in the pair separately?To elaborate on this, BALANCE is about STATE OF CHARGE, not voltage. Voltage only correlates meaningfully OUTSIDE the normal operating range of 3.1-3.4V. Even a notably imbalanced battery will have a very tight cell voltages.
wut?
Thought I'd try a little something on a cheap 12V I've been testing...
Charge to 20A @ 13.6V (3.4V/cell), 2A tail current.
When testing at cell level and dealing with testers that don't have separate sense leads, lead resistance plays a big role, and I would typically observe cells get charged to about 95% SoC @ 3.4V.
Here's a plot of the charge:
0.2C charge
2.5 hour bulk
5 hour absorption
0.02C tail current
104.6Ah input (tested at 105.5Ah)
Have about a 0.12V drop in the leads @ 30A, so it's pretty negligible at 2A.
With sense leads, it would have spent more time in...
summary: A 100Ah cell held at 3.40V until the current drops to 2A will yield 99%+ SoC.
No. If you have a benchtop power supply, simply set it to 3.65V prior to connecting to cell and charge the cells starting with the lowest voltage first, cycling up through the cells and then hitting them again AFTER all 4 cells have been driven to 3.65V (takes just a couple minutes per cell) to ensure all are at 100% and very nearly the same voltage.
You can do this with the battery in-place and without disassembling. It will be helpful to remove loads.
What BMS? Please include your BMS settings.
Q-Dog
¯\_(ツ)_/¯
If you really want to know what is going on it woudn't be a bad idea to break it all down and test each cell. Or just keep guessing -- your choice.
And would also explain if there is a big imbalance.
If you have the time to wait for that passive balaner, no, but depending on your imbalance it might take a while to recover. Easiest would be to just charge each cell separately to 3.5V or thereabout.
Yes, you can leave them connected even when manually balancing the cells. Personally, I would get an active balancer since some people are impatient with a lab power supply and tend to make mistakes (such as adjusting the voltage while connected to a cell) which ends up destroying the cell.
If you keep them connected, they will be considered as one cell. If you take it all apart, consider testing each cell individually and treating them as such. I don't think there is anything wrong with the cells; having that balancer running at 3.2V means over time it created a (potentially large) imbalance.
Of course, there might be something wrong with a cell, or multiple, so if this would end up on my desk or if it were my battery I would take it apart completely and test each cell.
sunshine_eggo
Victron's little biatch
Or the BMS balancing is inop or janky.
1,000% agree.
Correct.
Treat as single cell. Power supply lead on the (+) of one cell and the (-) of the other.
Again, adjust power supply to 3.65 (or your desired charge voltage) BEFORE connecting the leads.
3.40 for balance voltage. 3.2 is counterproductive, and I consider it a contributing factor to this issue.
Zwy
Emperor Of Solar
Worthless numbers when it comes to knowing cell balance.
What are the cell voltages above 3.4V?
Cell voltage near 100% or it is a waste of time unless very severely imbalanced.
RCinFLA
Solar Wizard
- Joined
- Jun 21, 2020
- Messages
- 3,847
If you made it to 14.2v without BMS shutting down charging for a cell overvoltage along way you are likely okay on balance.
Make sure you see some charging current at 14.2v and not just an open circuited BMS charge shut down with 14.2v absorb voltage from charger.
Last edited:
Steve Dally
New Member
- Joined
- May 10, 2021
- Messages
- 30
It's a single battery made of 8 cells arranged in 4 pairs, 2p4s I think is the way it's written. So so 4 pairs of two. Each pair is treated as a cell, then the 4 sets are connected in series to make a single 12v battery. I then use a single 4s bms (200 amp) to manage the whole thing.
Steve Dally
New Member
- Joined
- May 10, 2021
- Messages
- 30
That's helpful. I was thinking of it more as an average of all the cells. Thank you!
Zwy
Emperor Of Solar
I know the 3.394V number has been mentioned and I just happened to be top balancing some cells to 3.60V. I set the JK BMS battery capacity to 560Ah when cell average voltage was at 3.393V so we could track the Ah to 3.600V. As we can see, it shows charged capacity at 424Ah. Remember this number.
Charging to 3.600V average per cell with a delta of .008V, we have the following charged capacity of 453.2Ah.
That is 29.2Ah charged into the 280Ah cells, granted the EVE cells are new and slightly over 280Ah capacity plus there could be some inaccuracy of the BMS but my charger tracks Ah and it was reporting over 35Ah had been added.
29.2/280= 10.42%
If the OP has low voltage disconnect at 20% and the low cell is at 3.394V (90%),with a 560Ah pack, only 392Ah is the available capacity according to this test.
Brett V
Solar Wizard
SomethingAppropriate
Cannibalism starts with a kiss.
Good ole Jk BMS showing 81% at fully charged voltage.
Makes me all warm inside.
Makes me all warm inside.
Zwy
Emperor Of Solar
You missed the part about setting the Ah capacity to 560 Ah so I could track the Ah from 3.393V to 3.600V.
It's OK, we understand as you're from the land of the Cheese Heads.
Zwy
Emperor Of Solar
I've actually found the 6 that have been operating in the shop for about 20 days now track the Victron 1000A Smart Shunt very closely on SOC.
One has to properly set the BMS to have it read accurately. I use 3.51V for 100% SOC as I charge to 56.4V and it actually will set 100% slightly sooner.
Similar threads
