diy solar

diy solar

Problem Cell?

DON'T PANiC

New Member
Joined
Dec 26, 2021
Messages
4
Cells: 16x EVE LF280k (new, just received from Jenny Wu @ Docan Tech)
BMS: Overkill 12v 4s
Configuration: 12v 4p4s each 4s with its own BMS
Top Balanced cells in parallel to 3.6-3.65vpc
Final installation: Sailboat. BMS will be set to 3.0vpc for low cutoff, 3.55vpc for top cutoff and chargers will be set to 14v (3.5vpc). Ideally I'll cycle the charge from 10-15% to 85-90% avoiding the knees.

I just completed top balancing the cells and running them through their first cycle.
Three of the batteries balanced fine without any issues noticeable to this noob.
One had some noticeable issues when discharging. I cycled this battery twice, including a second top balance and the same thing occurred.
Looking to see if i have a problem cell that I should return and get replaced.

When discharging one cell got way out of balance: Voltage diff 0.320
Cell 1: 2.781vpc
Cell 2: 3.101vpc
Cell 3: 2.856vpc
Cell 4: 2.831vpc
This was from the second cycle. I didn't record the first cycle, but it was worse since I let it trigger the 2.5vpc cut off. I recall it getting to V Diff 0.500+ Cell 2 was still over 3.0vpc while cell 1 (if I remember right) triggered the cutoff at 2.5vpc

As I recharged the cells back after the second balance & discharge to 90%ish I tracked the cell Voltage:
Time____________Cell1_____Cell2_____Cell3_____Cell4____V Diff_____ AH
21:40:00_________2.781____3.101_____2.856____2.831____0.320______0.00
22:39:07_________3.247____3.266_____3.249____3.250____0.019_____24.22
23:57:10_________3.295____3.323_____3.300____3.301____0.028_____55.22
00:51:40_________3.325____3.340_____3.328____3.330____0.015_____76.91
01:14:42_________3.330____3.341_____3.334____3.335____0.011_____86.02
01:15:35_________3.312____3.318_____3.314____3.316____0.006_____86.10___(Taken when battery was at rest, no charging)
Stopped the Charging at this point and went to sleep. I resumed charging the next morning. The AH in the battery was kind of shocking because nothing was attached to the battery to charge, everything was off. The shunt I had attached confirmed that nothing was charging, it showed 30% for both above/below. Bad BMS? After this point I didn't trust the AH capacity numbers.)
09:26:35_________3.296____3.301_____3.298____3.300____0.005____142.41___(Taken when battery was at rest, no charging)
09:29:48_________3.332____3.344_____3.336____3.338____0.012____143.50
10.43.05_________3.338____3.349_____3.342____3.342____0.011____172.42
11.43.41_________3.339____3.351_____3.343____3.343____0.012____196.30
12:44:09_________3.345____3.360_____3.347____3.350____0.015____220.15
13:44:30_________3.365____3.374_____3.367____3.369____0.009____243.90
14:45:05_________3.374____3.380_____3.375____3.375____0.006____267.64
15:45:37_________3.377____3.384_____3.379____3.380____0.007____289.31
16:45:53_________3.381____3.388_____3.382____3.384____0.007____289.31___(AH is not a typo, it didn't move since the previous recording. This capacity is close to what the shunt reported when I did the discharges on this battery)
17:01:29_________3.345____3.347_____3.346____3.347____0.002____288.83___(Taken at rest, stopped charging, Shunt reported 90%.)


1) Do I have a bad cell that should be returned/replaced?
2) If this cell is kept, will if degrade the other cells? The other batteries?
3) Do I have a bad BMS?
 
Without knowing the current put through the cells not much can be determined.

You need to check single series string of cells at a time. For 280AH cells you should check with a cell current of between 56 amps and 100 amps. I usually charge each cell to a little over 50% state of charge at 3.30v to 3.31 vdc after rested for 3 mins after removing from charging to check open circuit cell voltage, before making the load test. The cell voltage will drop a bit after taking off charger and it will decay exponentially for about 3 mins with no load until it reaches equilibrium open circuit voltage.

If cells are well matched the voltage slump with current load will be similar, even if the cell balancing spread is between 45% and 65% state of charge Do load test with load for 3 minutes and the cells will slump in voltage with load with a similar exponential decay for 3 minutes of load after which it will level out. Measure the cell voltage with load after it levels out, after 3 minutes of load current.

Used cells will have more voltage slump with current. Matched cells will have similar voltage slump for same cell current, even if they are used cells.

You should try to match series connected cell groups that have the same voltage slump with same load current. Make sure you check cell voltage on the battery terminals and not on the bus bar straps.

Measuring cell voltage slump with 0.2 to 0.4 CA load current is more valuable info than doing a time consuming capacity test at low discharge current rates. A new cell will slump 0.0.05v to 0.06v at 0.2 CA load current (56 amps for a 280 AH cell) after 3 minutes of load. A used cell can be 3 to 5 times this number. Anything greater than 5x this number is a very used garbage cell with high internal impedance.

Temp of cell effects the slump so test should be run near room temp 25 degs C and not lower than 15 degs C.

The greater the variance in cell voltage slump with same load current for series connected cells the tougher it will be to keep them in balance.
 
Last edited:
Charge Rate: 20-25 Amps
Discharge Rate: ~34amps
Room Temp: 22.5C (73F)

I do not have the ability to do higher current rates at home, I would need to move to the batteries to the boat with the bigger charger/inverter.
The Charge/Discharge tests were performed with one battery at a time. I did not have the batteries in parallel.
I did confirm with a multi-meter at the terminals that the cell voltages were correctly reported by the BMS

The imbalance I showed above in the vpc was there even at rest, it changed a little but not much, showing that V Diff wasn't a voltage slump caused the the load on the battery.

Voltage slump under load is not the problem I am seeing, from what I can recall, it is similar across the cells. The problem I am seeing is that at the battery discharges, one cell (#2) is staying at a high vpc compared to the other cells in the battery. This is something that did not occur with the other three batteries (12 cells) that I top balanced and capacity tested.

I just completed discharging the battery to about 40% for storage until installed on the boat. The numbers (at rest) from the battery are:
Cell1_____Cell2_____Cell3_____Cell4____V Diff
3.285_____3.291____3.289_____3.290____0.005
 
Since you did not say the specifics of how many AH you discharged the cells, you may have just depleted the battery.

When you get down near full discharge the cell voltage drops off quickly. Just slight imbalance cause some cells to drop off the cliff in voltage first.

This is difference between top and bottom balancing. Usually top balance is better since you want cells to be close near full charge to avoid one cell reaching overvoltage first and shutting down BMS for overvoltage cell.
 
Check all your connections. 50% of voltage problems posted on the forum are due to bad connections or bad crimps.

However, what is the chance that you created a bad connection twice on the same cell? Seems like a longshot to me, but check the terminal surfaces to make sure they are all parallel. One terminal surface that is off a few degrees could make for a lousy connection.
 
While it could be bad connections, it is slightly unlikely, the batteries have been reassembled multiple times between put in parallel and in series. I guess it could be a bad post/terminal. I will check it to make sure.

Let me try to rephrase my original question.
Is it normal for the V Diff of cells to vary over 0.5vpc at the bottom of a discharge?
None of the other cells-batteries varied by such an amount at the bottom of the discharge. All the batteries hit expected capacity.
 
Let me try to rephrase my original question.
Is it normal for the V Diff of cells to vary over 0.5vpc at the bottom of a discharge?
None of the other cells-batteries varied by such an amount at the bottom of the discharge. All the batteries hit expected capacity.

Top balanced cells are more likely to deviate further apart at low voltages. The thought behind that is that most batteries don't spend much time below a certain state of charge (say 50%), so the top balance is ideal in that case. Once the cells are charged back up, they tend to come back together. That won't happen if you have a cell that is simply bad.
 
I charged the same battery in a 12v config (4cells) last night. I used a 12v 15a victron smart charger. When I went out this morning they were in the bulk absorption mode and when I tested the votage on cell was 3.83v, another was 3.36.

I immediately turned off the charger which was running 0.1 amp at the time.

My question is do I have a bad cell(s)? Thoughts?
 
3.83V? Which BMS are you using? The BMS should have stopped charging.

Unless the BMS has some weird OVP settings configured I would replace it.
 
The batteries are brand new. It was the initial charge to top off before I top balance. Not sure a BMS is required.
 
Initital charge without a BMS is something you shouldn't do. Since the cells aren't balanced, 1 cell will overcharge without anything to prevent

ALWAYS use a BMS, or sit at least with a hawk eye next to it and check every minute
 
While it could be bad connections, it is slightly unlikely, the batteries have been reassembled multiple times between put in parallel and in series. I guess it could be a bad post/terminal. I will check it to make sure.

Let me try to rephrase my original question.
Is it normal for the V Diff of cells to vary over 0.5vpc at the bottom of a discharge?
None of the other cells-batteries varied by such an amount at the bottom of the discharge. All the batteries hit expected capacity.
It is normal if the capacity is not well matched. Less than 5 amp hours capacity difference will cause a very significant deviation in the knees.
 
I bought a cheap Daly 4s just for this purpose. Even though I was building a 48 volt battery, i first built 12v batteries WITH THE BMS to initially charge to 85- 90%.
Times 4. THEN parallel top balance. Then assemble in 16s with JBD bms.
 
If you have capacity values for each cell you can use


To tell you the best configuration of your cells in your 4s4p config...

I have a 88 cells at 55ah and it was to much to capacity test them all so I topped them all and then ran them down till they reached rated capacity at 55ah of discharge and immediately recorded the voltage of each cell and then put the voltage data into the program above (i entered the voltage instead of the capacity) and it worked great to match the capacity in the group.
 
BMS will be set to 3.0vpc for low cutoff, 3.55vpc for top cutoff and chargers will be set to 14v (3.5vpc). Ideally I'll cycle the charge from 10-15% to 85-90% avoiding the knees.
I am late to the party but if you're configuring your BMS as described above, then you're using your BMS wrong. The BMS is a safety cutoff that should be set near the limits of the cells/battery (2.5V - 3.65V). It is NOT made to control the charging to keep it at some SoC range.

Its the job of the charge controller (get it? it controls the charging?) to determine how high to charge your cells/battery. There are many thoughts on what is the best voltage to charge to. Ponder this widely accepted voltage to SoC chart:
LiFePO4 SoC chart.png

To meet your goal of charging to 85-90%, that would be about 13.35V (3.375Vpc) for 4S and 53.4V for 16S. These fall short of most minimum, conservative charging recommendations. I am a very conservative charger and charge to 13.8V (3.45Vpc) while many others recommend 14V and others even higher. Some of this depends on manufacturers specific recommendations and some also depends on your use case for your battery(s).

As for low voltage cutoff, 3.0V is also on the extreme conservative side but that's not as big of an issue. The issue that I see is that if you continually disconnect with your BMS, it will be short lived and components in your system will suffer from having the battery power cut, especially if part of your daily charge cycle. Your system should be setup so that you never hit the BMS. The BMS is there to save the cells/battery when there is a catastrophic failure somewhere.
Solutions for LVD include battery protect modules and inverters with programmable (or desirable) LVD settings that shut off before the BMS needs to interject.
 
Back
Top