diy solar

diy solar

Eve 280Ah battery, strange behavior from one cell

mikemarmar

New Member
Joined
Mar 24, 2021
Messages
10
Hey all. A couple months ago I put together a 560Ah 12v pack (2p4s) using the 280Ah Eve cells (It is possible that I actually have the 272ah Lishen cells....the vendor on alibaba was a bit unclear).

Today I was charging up the pack from ~60% to 100% using my victron compact multiplus 12/2000. The charge rate was ~70A, and at around 80% SOC I noticed something strange. The voltage on one of the cells (actually 2 parallel cells) in the pack was rising and quickly hit my BMS cutoff of 3.65v. The other 3 cells were all around 3.35v. I figured that maybe I have a weak cell or they somehow got out of balance.

In order to keep charging I reduced the charge rate to about 15A, which kept that first cell voltage from hitting the cutoff. It hovered around 3.55v for almost the rest of the charge cycle while the rest of the cells stayed at ~3.35v. However at about 98% SOC when the other 3 cell voltages started to rise, the voltage of the "weak" cell dropped down. All four cells are now at around 3.45v, with the weak cell just slightly higher at 3.48v.

Is this normal behavior for a weak cell? Why would the voltage spike high earlier but then drop back down as the pack nears 100% SOC?

These voltages are measured by my BMS (daly smartbms) in case that is a factor...
 
Verify the bms voltage readings with your dvom.
Another thing that causes strangeness with a pack is a high resistance connection in the pack.
If the dvom agrees with your bms then you might want to discharge the pack a bit to give you room to experiment.
Start charging again at 70 amps or so and see if one or more of the busbars get warmer than the others.
Do you have temperature sensors on the body of the cells and/or on the busbars?
If you do that should give you a baseline.
 
Welcome to the forum.

Confirm all connections are secure, properly torqued and of high quality.

Hey all. A couple months ago I put together a 560Ah 12v pack (2p4s) using the 280Ah Eve cells (It is possible that I actually have the 272ah Lishen cells....the vendor on alibaba was a bit unclear).

Did you top balance them prior to build?

Was any meaningful testing done?

What has happened over the last "couple months?"

A chronology of events since cell receipt might be helpful.

Today I was charging up the pack from ~60% to 100% using my victron compact multiplus 12/2000. The charge rate was ~70A, and at around 80% SOC I noticed something strange. The voltage on one of the cells (actually 2 parallel cells) in the pack was rising and quickly hit my BMS cutoff of 3.65v. The other 3 cells were all around 3.35v. I figured that maybe I have a weak cell or they somehow got out of balance.

Parallel cells are essentially a single cell. A bad one in parallel with a good one makes them both appear bad.

Did you confirm any voltage readings with a separate meter?

In order to keep charging I reduced the charge rate to about 15A, which kept that first cell voltage from hitting the cutoff. It hovered around 3.55v for almost the rest of the charge cycle while the rest of the cells stayed at ~3.35v. However at about 98% SOC when the other 3 cell voltages started to rise, the voltage of the "weak" cell dropped down. All four cells are now at around 3.45v, with the weak cell just slightly higher at 3.48v.

Did you confirm any of these voltage readings with a separate meter?

Is this normal behavior for a weak cell? Why would the voltage spike high earlier but then drop back down as the pack nears 100% SOC?

No. Did you take note if any of the cells were warmer than others? Heat typically causes a voltage decrease.

These voltages are measured by my BMS (daly smartbms) in case that is a factor...

Not sure if it's a factor, but when one is having issues that seem to be related to voltage, a second opinion is a good idea.

Speculation:

The cell that hit high voltage first has excessive internal resistance causing a higher voltage increase for a given current higher than the other 3. Lowering the current reduced the impact of the higher internal resistance allowing all cells to continue to absorb charge. They all hit peak voltage at about the same time under the lower current (98% SoC).

Now that current is removed, voltages have stabilized.

After checking all connections to confirm they are high quality, I recommend you conduct a high current discharge - about 80% the rating of the BMS or cells, whichever is lower. Note the behavior of the cells. If you see the same cell that went high first, also show a consistently lower voltage than the other three, this is confirmation of a cell with high internal resistance.

IMHO, I would break the pack down into 2X 4S batteries for testing as parallel cells will mask issues.
 
Last edited:
Thank you for the quick replies. I will have to do some further testing it seems. To answer some of your questions:
Confirm all connections are secure, properly torqued and of high quality.
Good idea, I will do this first. The battery has been on one long roadtrip and something may have come loose.

Did you top balance them prior to build?
The cells were top balanced at 3.6v before the pack was assembled.

Was any meaningful testing done?
My brother in law did a capacity test on another battery built from the same shipment of cells. He pulled 250ah before stopping the test and everything seemed fine.

What has happened over the last "couple months?"
The cells were balanced and installed in my van where they have been since. They have been subject to some pretty low temperatures (went below freezing once, but not charged below freezing of course) and some rough dirt roads. They have spent the whole time between 50-100% SOC.

Did you confirm any voltage readings with a separate meter?
Not yet, but I have access to a fluke meter so I will test that when I get the chance.

The cell that hit high voltage first has excessive internal resistance causing a higher voltage increase for a given current higher than the other 3. Lowering the current reduced the impact of the higher internal resistance allowing all cells to continue to absorb charge. They all hit peak voltage at about the same time under the lower current (98% SoC).

Now that current is removed, voltages have stabilized.
This doesn't quite match what I observed. The charging current was steady around 15A from 80% to 98%. Right when cells 2-4 started to creep up, cell 1 started to drop. The current was still 15A at this time. The pack is now at "100%" and the charger is pushing 9A and dropping. The voltages of the cells are now all in the 3.55v range and in fact cells 3 and 4 are now slightly higher voltage than cell 1.

After checking all connections to confirm they are high quality, I recommend you conduct a high current discharge
Good idea, I will try this.
 
The cells were balanced and installed in my van where they have been since. They have been subject to some pretty low temperatures (went below freezing once, but not charged below freezing of course) and some rough dirt roads. They have spent the whole time between 50-100% SOC.

Has it been in regular use, or mostly sitting?

This doesn't quite match what I observed. The charging current was steady around 15A from 80% to 98%. Right when cells 2-4 started to creep up, cell 1 started to drop. The current was still 15A at this time. The pack is now at "100%" and the charger is pushing 9A and dropping. The voltages of the cells are now all in the 3.55v range and in fact cells 3 and 4 are now slightly higher voltage than cell 1.

You said you started charging at 70A, then switched to 15A. The lower charging current (15A) allowed the runner to keep charging below cut-off.

Did you witness any cell warming, particularly on #1? Higher temps can results in slightly lower voltages.
 
Has it been in regular use, or mostly sitting?
I would say mostly sitting. It has been powering some small loads the whole time, but the victron has mostly been plugged in and floating the pack at 13.2v, which tends to settle around 70% SOC and stay there. The batteries were in full use for a 2 week trip, and otherwise the van has been in my driveway.

I did not see much warming, although my only temperature probe is on cell 1. It rose by maybe 2c during the 70A charging.

I'm now discharging the pack at about 150A and watching the voltages. Currently at 75% SOC and cells 1 and 3 are mostly hanging around 3.0-3.1v while cells 2 and 4 are around 3.1-3.2v but they are all jumping around a bit in the 3.0-3.2v range.
 
That's quite a bit of variance. Can you take a picture of your battery?
While under load, does any one terminal feel warmer than the rest? If there's a bus bar/terminal connection problem, You should be able to feel a difference within 15 minutes of initiating a high load.
It does smell like a bad connection between a bus bar and a terminal, and the ONLY way to correct that is to take it apart, shine it up (e.g. 1000 grit sandpaper), de-oxidize it, and reassemble.
 
Hm yeah I might have to go ahead and disassemble and make sure my connections are good. I made my own bus bars from copper flat stock so it's possible that there is a burr I missed. Also, the terminals on these cells are very prone to stripping so I might not have tightened one of the bolts enough....
 
So after drawing down 500Ah, cell 1 is at 2.8v while cells 2-4 are around 3.15v. Cell 1 dropped below 3v fairly early on (around 40% SOC) and has stayed low. I guess the next step is to take the pack apart and try to figure out if it is a bad connection or if one cell is bad.
 
Cell 1 is the one that went high and tripped the BMS at 70A?

IMHO, their resting voltage response might identify an outlier:

Disconnect all cells.
Let them sit for 1 hour.
Record voltages.
Let them sit for 12+ hours.
Record voltages.
 
Yeah cell 1 is the same that went high while charging. Running that test again right now to make sure it is reproducible. Next step is to take everything apart, and run some of the tests you all have recommended. I'll let you know how it goes, thanks again for the help and suggestions!
 
Yeah cell 1 is the same that went high while charging. Running that test again right now to make sure it is reproducible. Next step is to take everything apart, and run some of the tests you all have recommended. I'll let you know how it goes, thanks again for the help and suggestions!
You don't have to take things apart to validate the bms with the dvom.
Its still worth doing that first.
Suggest to verify at rest and under load.
 
For that, you'll want to make sure your meter probes are touching the cell terminal and not the bus bars. If that's not physically possible, then that test won't rule out dirty contacts.
 
So this is interesting. I just tested charging from 50% to 100% at 70A charge rate. The same problem presented (cell 1 voltage goes high around 80%, then drops back down when the other cell voltages start to rise). However, it was much less dramatic this time. Cell 1 never really went above ~3.55v, and the BMS never tripped.

My theory is that there is a bad connection on the serial bar between cells 1 and 2. Thus when charging at high current, cell 1 will receive more than the others and when discharging at high current it will give more. This allowed cell 1 to get way out of balance (higher SOC) over time since most of my charges have been at a fairly high current (70A from the victron or ~150A from the alternator) and discharges have been mostly at low current. Yesterday when I charged the pack at a low current and then discharged the pack at a high current, that allowed the cells to come somewhat back into balance as the effect on cell 1 vs cells 2-4 was the exact opposite of what I have been doing in regular use.
 
It's possible, but voltage drop only manifests itself when there is current passing. BMS leads pass very little current and effectively measure open circuit voltage. Still worth checking, and I hope I'm wrong.
 
I ran the same experiment again (high current (150A) discharge, followed by lower current charge (60A)) and the imbalance is now pretty much gone. So whatever is going on, the imbalance gets larger when I do low current discharge + high current charge (normal usage) and the imbalance gets smaller when I do high current discharge + low current charge.

To me this seems like a poor bus bar / terminal connection, not a bad cell. Unless it is possible for a bad cell to behave differently in discharge vs charge. I will have to partially disassemble my pack to check all the connections, but the ones I was able to access did not get hot. EDIT come to think of it, if cell 1 has lower internal resistance than the other cells, that could explain this behavior too. It would take more charge than the others during high current charging, but the effect would be diminished during low current discharge. Over time it would drift high relative to the other cells...

As an aside, in feeling around for hot connections I almost burned myself on the terminals of my battery disconnect switch (bluesea dual circuit switch). The terminals were >200F and I measured 0.15v drop across them. I guess I have to take that apart too and see why it has poor contact...
 
Last edited:
I ran the same experiment again (high current (150A) discharge, followed by lower current charge (60A)) and the imbalance is now pretty much gone. So whatever is going on, the imbalance gets larger when I do low current discharge + high current charge (normal usage) and the imbalance gets smaller when I do high current discharge + low current charge.

To me this seems like a poor bus bar / terminal connection, not a bad cell. Unless it is possible for a bad cell to behave differently in discharge vs charge. I will have to partially disassemble my pack to check all the connections, but the ones I was able to access did not get hot. EDIT come to think of it, if cell 1 has lower internal resistance than the other cells, that could explain this behavior too. It would take more charge than the others during high current charging, but the effect would be diminished during low current discharge. Over time it would drift high relative to the other cells...

As an aside, in feeling around for hot connections I almost burned myself on the terminals of my battery disconnect switch (bluesea dual circuit switch). The terminals were >200F and I measured 0.15v drop across them. I guess I have to take that apart too and see why it has poor contact...

WOW yeah they shouldn't get anywhere near that hot. At 220 amps my cells went up all of 10 degrees. You've absolutely got bad connections. Dangerously bad.
 
Yeah...the cells don't seem to be getting hot so that's good. I will check the terminals I can't access sometime in the next few days. The bluesea switch on the other hand seems to have a bad internal contact, but that's a separate issue at this point.
 
Back
Top