Trouble charging (one cell triggers OVP at 35% charge)

[jameshowison]

Could someone suggest troubleshooting steps?

Server rack LFP battery. Was stored (at 95%) for 2 months. Previously working fine.

The battery can’t effectively charge because one cell triggers OVP. It shows at 3.5v and rises to 3.7 if any substantial current is applied. Other cells are all around 3.3v. Capacity shows 25% but up to 35% after charging at 2A overnight.

Worryingly that same cell shows low voltage (drops to showing 3.1) when some decent load applied.

Any chance this is just balancing or similar? Should I reduce charging voltage?

It’s hard to extract from RV, but I guess check connections? Could a loose connection show those symptoms? (I would have thought high resistance would mean slow charge and discharge?)

Other suggestions?

Thanks,
James

 

[BentleyJ]

Based on your description, it seems like the cell in question is potentially damaged. When one cell gets out of balance (which can happen with storage and is corrected by balancing) either on the low or high side it will exhibit either over or under voltage excursions but not both conditions at the same time.

 

[740GLE]

What brand server rack battery?

 

[brum]

Does the cell still show higher voltage when there is no current going through the pack? If yes - most likely damaged cell. If no - most likely bad cell connection.

Either way - you'll have to take out the battery from the RV. Once out you can inspect the issue in details.

And don't store cells at 95% charge. Recommended storage charge is ~50%. But this is not the reason for the failure.

 

[Will Prowse]

That's going to be a warranty claim. What manufacturer and which distributor? Can you post your cell voltages as a picture? And the ovp message?

 

[Will Prowse]

Only other possibility is if the SOC meter is not calibrated properly and it's hitting 100%, not 35%. Can you post pictures of your entire system? How many batteries are in parallel? What is the SOC of the other batteries?

And if that one cell is dipping under load, it is probably defective. The distributor should start a warranty claim.

 

[RCinFLA]

SoC meter should not limit charging, only cell overvoltage, overcurrent, or over temp. Most likely a cell went overvoltage during charging

If pack was charged to 95% and not used, it should not have had enough internal leakage or BMS overhead consumption to drop to 50% SoC in a couple of months.

You can keep absorb level charging for a week to see if it rebalances cells or open the pack up and individually discharge the overvoltage cell(s).

You should set absorb voltage to about 3.5v times number of series cells to allow balancing. You may get several BMS overvoltage shutdowns during the balancing time. The pack should bleed down the overvoltage cell and start charging again in a few minutes. Eventually, it will rebalance but may take a long time.

If one cell is tripping overvoltage shutdown and another cell is at about 3.30v you have close to a 50% difference in state of charge. If 100 AH battery, that is 50 AH misbalance. With 100 mA resistor dump balancer that is 50 AH / 0.1 amp balancing = 500 hours of required balancing time !!!

It is likely you will be told by supplier you need to balance battery first before they consider a warranty claim. It may help if you tell them the voltage you see on lowest cell indicating a near 50% difference in cells' SoC. 500 hours is a pretty unreasonable time to expect someone to do rebalancing.

If what you stated is true, there is something wrong with cells or BMS. I am assuming when you say 'stored', it means the battery pack was not connected to anything that may have drained some extra current.

Some cheap BMS's have poor power management and can have fairly high average idle overhead current drain of 5 to 30 mA or so. This may happen if you leave a digital interface comm line plugged in to battery pack.

 

[jameshowison]

This is a Jakiper Gen 1 battery. SoC indicator on battery says 35%. Yes, under load that cell is dipping to 2.9v (while others stay at 3.3v)


then switching to charging at anything over 5Amp that cell spikes to 3.6 (pic shows 3.4 but that was ~3.5Amp charging).



Under no load it shows 3.19, while the others show 3.34.



I'm already talking with Jakiper support and they have been very responsive, they'll ship a replacement cell and then to insert it, it will have to be rebuilt (they have a video showing how to do it). It's a bit involved (remove all cell holders and bus bars) but looks straight forward enough, given time.

--J

 

[jameshowison]

SoC meter should not limit charging, only cell overvoltage, overcurrent, or over temp. Most likely a cell went overvoltage during charging

If pack was charged to 95% and not used, it should not have had enough internal leakage or BMS overhead consumption to drop to 50% SoC in a couple of months.

Apologies, the battery stayed at a reasonable level of charge after storage, I can't quite remember exactly but that wasn't a concern. Then I ran loads on it for a week or so (~8amp draw, sometimes with the charger providing power) and it ran down much faster than the other battery that is in parallel. I have not been able to bring the problematic one back up above 35% SoC (the other battery charges fine, done with the problematic battery switched off at its circuit breaker).

 

[Lt.Dan]

Atleast they are already opting to take care of you. Thank goodness its a user serviceable battery and you don't have to ship it back or anything.

 

[Will Prowse]

SoC meter should not limit charging, only cell overvoltage, overcurrent, or over temp. Most likely a cell went overvoltage during charging

If pack was charged to 95% and not used, it should not have had enough internal leakage or BMS overhead consumption to drop to 50% SoC in a couple of months.

You can keep absorb level charging for a week to see if it rebalances cells or open the pack up and individually discharge the overvoltage cell(s).

You should set absorb voltage to about 3.5v times number of series cells to allow balancing. You may get several BMS overvoltage shutdowns during the balancing time. The pack should bleed down the overvoltage cell and start charging again in a few minutes. Eventually, it will rebalance but may take a long time.

If one cell is tripping overvoltage shutdown and another cell is at about 3.30v you have close to a 50% difference in state of charge. If 100 AH battery, that is 50 AH misbalance. With 100 mA resistor dump balancer that is 50 AH / 0.1 amp balancing = 500 hours of required balancing time !!!

It is likely you will be told by supplier you need to balance battery first before they consider a warranty claim. It may help if you tell them the voltage you see on lowest cell indicating a near 50% difference in cells' SoC. 500 hours is a pretty unreasonable time to expect someone to do rebalancing.

If what you stated is true, there is something wrong with cells or BMS. I am assuming when you say 'stored', it means the battery pack was not connected to anything that may have drained some extra current.

Some cheap BMS's have poor power management and can have fairly high average idle overhead current drain of 5 to 30 mA or so. This may happen if you leave a digital interface comm line plugged in to battery pack.

I'm meaning it's showing 35% but it's not. We actually had this issue on packs in the past. Some bluettis had shunt calibration issues as well. what a good BMS will do us reset to 100% at ovp. But some do not. I don't know what battery they are running. Or the BMS.

 

[Will Prowse]

SoC meter should not limit charging, only cell overvoltage, overcurrent, or over temp. Most likely a cell went overvoltage during charging

If pack was charged to 95% and not used, it should not have had enough internal leakage or BMS overhead consumption to drop to 50% SoC in a couple of months.

You can keep absorb level charging for a week to see if it rebalances cells or open the pack up and individually discharge the overvoltage cell(s).

You should set absorb voltage to about 3.5v times number of series cells to allow balancing. You may get several BMS overvoltage shutdowns during the balancing time. The pack should bleed down the overvoltage cell and start charging again in a few minutes. Eventually, it will rebalance but may take a long time.

If one cell is tripping overvoltage shutdown and another cell is at about 3.30v you have close to a 50% difference in state of charge. If 100 AH battery, that is 50 AH misbalance. With 100 mA resistor dump balancer that is 50 AH / 0.1 amp balancing = 500 hours of required balancing time !!!

It is likely you will be told by supplier you need to balance battery first before they consider a warranty claim. It may help if you tell them the voltage you see on lowest cell indicating a near 50% difference in cells' SoC. 500 hours is a pretty unreasonable time to expect someone to do rebalancing.

If what you stated is true, there is something wrong with cells or BMS. I am assuming when you say 'stored', it means the battery pack was not connected to anything that may have drained some extra current.

Some cheap BMS's have poor power management and can have fairly high average idle overhead current drain of 5 to 30 mA or so. This may happen if you leave a digital interface comm line plugged in to battery pack.

Yes exactly, cells or bms at fault. They need to start a warranty claim.

 

[Will Prowse]

This is a Jakiper Gen 1 battery. SoC indicator on battery says 35%. Yes, under load that cell is dipping to 2.9v (while others stay at 3.3v)

View attachment 153157
then switching to charging at anything over 5Amp that cell spikes to 3.6 (pic shows 3.4 but that was ~3.5Amp charging).

View attachment 153156

Under no load it shows 3.19, while the others show 3.34.

View attachment 153158

I'm already talking with Jakiper support and they have been very responsive, they'll ship a replacement cell and then to insert it, it will have to be rebuilt (they have a video showing how to do it). It's a bit involved (remove all cell holders and bus bars) but looks straight forward enough, given time.

--J

Yeah that's a bad cell or bad voltage sensing. Probably a bad cell because that's a good BMS.

 

[timselectric]

In the future, don't store your battery with that high of a SOC. You should be storing them at around 60%.
Can't say one way or the other, that it had anything to do with the issue.

 

[jameshowison]

Interim follow-up here.

We took the camper on the road, with both batteries in place. I had the problematic one turned off. After about two weeks, including national forest camping down a quite bumpy unpaved road (4 low gear) the other battery started showing issues. It would shut off under anything like load (it was ok for led lights, but not for ~800w draws), with one cell showing very low voltage under load, but also quite higher voltage just after load turned off.

A few days later we limped into town. I bought a torque wrench (at great expense, mountain towns man, but at least I could get one) and I spent a day pulling the batteries out.

The good news was that I worked on the second battery to have issues first and I tightened all the battery connections to the recommended torque. Some were quite loose, perhaps a half-turn or more. Nothing was going to fall off and flop around inside the battery, though. This appeared to entirely fix the issue and that second battery is back to happy service. Unfortunately, I did not have time to adjust the torque on the first battery, but I'll do that when I get home in a few weeks and report back. I feel pretty confident that I will also find some below spec battery connections, which I think could explain the pattern of cell voltage/behavior that I saw earlier in this thread.

In conversation with the very responsive customer service I commented that the battery connections don't have lock-washers. They are serrated nuts. I asked if I could replace those serrated nuts with lock-washers and retain warranty and they said that I could. I think that will be more appropriate for a vibration environment like an RV. Perhaps this is also an argument for the welded cells connections in RVs (although the downside would be that if it was a bad cell, then very much out of luck).

Thoughts? This thread suggests nord-lock washers: https://diysolarforum.com/threads/v...ections-possible-solutions.50242/#post-638097

 

[timselectric]

Serrated nuts do the same thing as lock washers.
Some might say better, because they are fixed on one side.

 

[jameshowison]

Serrated nuts do the same thing as lock washers.
Some might say better, because they are fixed on one side.

Hmmm, learning over here. I thought lock-washers created tension, but now I read they just create additional friction with the cut edges.

I think I'll explore the nordlock washers.

 

[timselectric]

Another option is to add blue loctite to the threads.
Not the red, you'll never get them apart, again.

 

[wild01]

nordlocks work very well if you can find them for a decent price. I regularly use them in industry, never seen one fail.

 

[Checkthisout]

Lockwashers give "crush" as in pretension. The busbars themselves are usually pretty soft and give this same effect.