One EVE 280Ah cell tripping BMS high voltage at top end

[LifeWAT]

I have a self-built 280Ah 48v battery pack, that was working great. I initially did a top balance, parallel to 3.65v



For a couple weeks now I've been having issues with one cell, #16 which happens to be my main positive for the entire battery. At the top end when the inverter is pushing the last bit of juice to fully charge the battery, cell No. 16 is running away from the rest and the BMS is tripped at 4.0v

Is there any way to fix this other than try:
1. Parallel all the cells again and top-balance again to 3.65v
2. Swap cell No. 16 with anther cell in the pack and see if that helps any
3. Install an active balancer on the pack if options 1 and 2 do not resolve the issue

Any other ideas?

 

[SeaGal]

Manually discharge cell #16 using a load such as a 12v car bulb when the pack is reaching the top. Keep monitoring the cell with a multi-meter as you do so, until it evens out with the other 15 cells.

Edited to add... you can do this as the pack it being charged - just keep track of voltages as you do so and be careful to not short anything out as you do so / take off jewellery / insulate-tape any rings etc.

 

[SeaGal]

Yikes.. just noticed how high cell #16 is... Would strongly recommend you set your BMS to trip at 3.65V, not 4V !

3.65V should be considered absolute max for top balancing... Many of us will only charge to 3.45V / 3.50V per cell in regular usage. You won't gain much kWh by going right up to 3.65V.

 

[LifeWAT]

Manually discharge cell #16 using a load such as a 12v car bulb when the pack is reaching the top. Keep monitoring the cell with a multi-meter as you do so, until it evens out with the other 15 cells.

Thank you! Just rephrasing to ensure I have it correct.

1. Connect PV input (or otherwise) and be charging all 16 cells still connected in series (48v)
2. As cell #16 starts to "run ahead" from the rest, connect a load such as a 12v car bulb to the positive and negative of cell #16.
3. Monitor all cell voltages and remove the "load" when they are all the same, or very close to the same within acceptable voltage variance with each other.
4. Theoretically they will all be "top balanced" now, and should move up and down more closely from this point forward.

I will consider lowering the trip point voltage after getting them balanced. I'm pretty sure 4.0v was the default high voltage cutoff in the JBD BMS.

 

[SeaGal]

Thank you! Just rephrasing to ensure I have it correct.

1. Connect PV input (or otherwise) and be charging all 16 cells still connected in series (48v)
2. As cell #16 starts to "run ahead" from the rest, connect a load such as a 12v car bulb to the positive and negative of cell #16.
3. Monitor all cell voltages and remove the "load" when they are all the same, or very close to the same within acceptable voltage variance with each other.
4. Theoretically they will all be "top balanced" now, and should move up and down more closely from this point forward.

Yes.. exactly. Start discharging #16 when the voltage goes above about 3.4V.

Because the load on cell #16 will reduce the cell's voltage due to the load itself, personally I'd take #16 a tiny bit lower than the average. This is because when you remove the load (and the charge) the voltage of cell #16 will raise up a bit again. You may need to repeat this process a few times to get them balanced close.

I will consider lowering the trip point voltage after getting them balanced. I'm pretty sure 4.0v was the default high voltage cutoff in the JBD BMS.

4V is incorrect for LiFePO4 chemistry cells. 3.65V is absolute max. You will risk damaging the cells with bloating by charging higher than 3.65V.

 

[LifeWAT]

@SeaGal Super helpful! Thank you.

 

[SeaGal]

Once you've fixed cell #16, you will find yourself wanting to do the same to cells #3, #4 and #6... It can get addictive ???

 

[LifeWAT]

Once you've fixed cell #16, you will find yourself wanting to do the same to cells #3, #4 and #6... It can get addictive ???

Yes, I was already thinking that

 

[Bossrox]

I've had an issue like that before that turned out to be a bad connection somewhere & that'll cause a false reading. The #1 & 16 differential is the clue that tells me that. Horrible balance, I have a bank that's like that, had to stack 5 balancers on it to bring 'em closer to together & still not that great.

 

[SeaGal]

I've had an issue like that before that turned out to be a bad connection somewhere & that'll cause a false reading.

Cause a false reading or cause an actual cell imbalance?

 

[TomC4306]

You can also lower your max charge voltage.
There is very little capacity above 56 volts (3.5).

 

[Bossrox]

Cause a false reading or cause an actual cell imbalance?

More likely the later but a bad bms cell wire can do it too.

 

[LifeWAT]

@SeaGal Thank you for your advice! I used an old 12v accessory truck light I had laying around with wires already connected, and worked on the offending high voltage cells while charging them up to the cutoff point. It's been a few days now and my issues have been resolved! Cell #13 runs just a little high at times, so I may bleed a little off when I get a chance.

I just wanted to come back and express my thanks, and helping me avoid the tedious job of re-wiring everything in parallel to fix a small but important issue.



I assume the inverse of this method is also true: If I have a cell running low, I could attach a bench charger putting out 3.65 volts and slightly increase the voltage of just that one cell to match the others?

 

[sunsurfer]

LifeWat, what BMS is that you are using? I really like that app.

 

[LifeWAT]

LifeWat, what BMS is that you are using? I really like that app.

It is the "JBD Smart 7S-20S 200A with Bluetooth+UART BMS". This is the link, but I think they updated the design a bit: https://www.alibaba.com/product-detail/Smart-BMS-8S-12S-16S-20S_1600296495265.html

I paid $87.90 in September, 2021

I believe this app can be used with all JBD BMS devices. My understanding is that there are two versions of the app though, and you probably want the pro version. Learn more here:

Currently I have mine set up with a security camera, and can view it from anywhere I have internet. I also just got solar-assistant.io so I will be connecting the BMS to Solar Assistant. https://solar-assistant.io/

I like the JBD BMS, but if I were purchasing today, I would purchase a BMS with active balancing included, like the JK BMS. From what I've seen those look nice.

 

[SeaGal]

I just wanted to come back and express my thanks, and helping me avoid the tedious job of re-wiring everything in parallel to fix a small but important issue.

Appreciated, thanks!

I assume the inverse of this method is also true: If I have a cell running low, I could attach a bench charger putting out 3.65 volts and slightly increase the voltage of just that one cell to match the others?

Yes, that's correct. Been there, done that

The only difference is I disconnected the charge leads first.

So, the way I did it was to charge the cells up to max via the inverter first (so there is bigger difference in voltage between cells). Then disconnect the battery pack from the inverter, attached bench power supply (with floating output, not grounded), to the lowest cell and charged it up just a little higher voltage than the highest cell so that when the PSU is disconnected it returned to similar voltage as highest cell. Then repeated for my other 'low' cell.

As before, monitor closely with multi-meter, remove jewellery / rings etc. first.

And remember that, unlike a 12V bulb, the power supply has a +ve and a -ve lead

 

[DIYrich]

Is that a normal range for balanced cells? The recommended range for 280ah for my JK BMS is 0.005v, much less than the 0.074 in the picture.

 

[SeaGal]

Is that a normal range for balanced cells? The recommended range for 280ah for my JK BMS is 0.005v, much less than the 0.074 in the picture.

Depends how far you are up the knee. I'd be surprised to see only a 5mV difference at higher voltages. IIRC my JBD doesn't start balancing unless there is a 15mV difference.