Top balancing with BMS and invertor - any issues to expect?

[uapower]

I'd like to validate my idea of top balancing of battery pack with BMS 1A balancer only.

My setup is 16s EVE LF280K and B1A20S15P JK BMS which provides 1A balancing current. Iverter is Axioma Energy BFP5600 which is known as Axpert/Voltronix 5600 version. Initially all cells were at ±3.3v and, as expected, they are not balanced.

I read almost all forum topics on balancing and know the recommeded procedire. Still, I decided to skip top balancing wih power supply for two reasons. First, I do not have access to one and cannot wait for it to ship (I also build the battery for myself so probably this would be the only single use of that PSU). Second, I want to connect and use the batteries ASAP, even not finely balanced. As of now here in Ukraine we have grid electricity 4 hours on / 8 hours off or vice versa, because terrorist state of russia destroyed big part of or grid infrastructure with missiles...

Back to my idea.
1. Set BMS cell top at 3.6v (recoery at 3.55v), start balance at 3.45
2. Set invertor voltage to 54.2v (3.375*15 + 3.55) - when during charging the first cell goes to 3.5v, others are around 3.375
3. With time, more cells go to higher range, so I gradually increase the invertor voltage until it reaches 56v (3.5*16)
I expect it to finish balancing in day or two from now.

Initial state (less than 1d of runtime)



And now (almost 4d running) - 8 cells already over 3.5 and diff is 0.123v

 

[MisterSandals]

Second, I want to connect and use the batteries ASAP, even not finely balanced

To do the balancing quicker, you could manually bleed off the higher cells. You could do this with almost any load like an incandescent light bulb (may not light but should burn off ~1/4 rated wattage (230Vac vs 55Vdc).

A clothes iron or any simple heater would likely be more watts.

You would become a passive balancer many times faster than your BMS.

 

[TorC]

Seems odd to me to try to bleed from the full pack to help balance, unless you need to set the tail current higher than the balancer can make use of. I'd look toward a 12V incandescent bulb or three, and put one across a spiking cell. IIRC, car headlights are generally around 55W at 12V, so you'd get about 15W, 4.5A from the cell you put it on. Only reason for more than one bulb is to do multiple high cells at once.

ETA: It's hard to get any 230V load that can directly drain from a cell. Maybe a hot water heater element would do something, but still looks like a 4.5kW element would only bleed off ~290mA.

 

[MisterSandals]

Seems odd to me to try to bleed from the full pack to help balance

Ah, i see why you say that. I did say "bleed off higher cell" but using 55V in my math was wrong.

Yes, only about 1.5% of rated wattage for 230V appliance bleeding 3.6V cell. 12V much better as you suggest.

 

[time2roll]

As long as all cells are 3.400+ volts, I would put the battery in service. Check in a month if you want to increase voltage.

 

[TorC]

Looks like I took the wrong part of your post as the authoritative section. Cheers.

 

[uapower]

Looks better, 9 cells around 3.575, and the rest around 3.415.

 

[TorC]

Looks like they're all above the knee together. Should mean <5% capacity, and likely <2.5% lost in the lack of balance. Your balancer will probably eke out a little more evenness of charge over time, and thus usable capacity, but I don't see any reason left to bother with manually discharging "runner" cells. Using the pack will probably take care of that better now, anyway.

 

[uapower]

After 5 days it’s finally done.

 

[Raikoh]

I just got same setup and im trying to balance the battery the same way with same BMS.
Did you use power from the battery during those 5 days or you let them balance only?

 

[zorlig]

You did exactly what I would have done, except I would have just put the cells into service right away and increased the voltage only every few weeks or good sun days.

 

[Hans Kroeger]

1. Set BMS cell top at 3.6v (recoery at 3.55v), start balance at 3.45

I recommend to set the start balancing voltage to 3.37 Volts. Above this voltage level you can be sure, that the cells are charged, even with Low current values. You can charge with a very Low voltage down to 3.4 Volts per cell and still have your cells balanced.

 

[zorlig]

I recommend to set the start balancing voltage to 3.37 Volts. Above this voltage level you can be sure, that the cells are charged, even with Low current values. You can charge with a very Low voltage down to 3.4 Volts per cell and still have your cells balanced.

I can second that call out, my notes are long gone but 3.35-3.37 is about where the cells actually fill and then voltage spikes. The voltage varies slightly per cell so 3.4 is the ultimate safe voltage guaranteed to be past the charging phase.

 

[uapower]

I just got same setup and im trying to balance the battery the same way with same BMS.
Did you use power from the battery during those 5 days or you let them balance only?

Yes, battery was in service (discharge for up to 4 hours / 40AH max). The key to success there IMO is slow charging (10A or so), gradual increase of total voltage and some patience.

 

[uapower]

And now my typical charge curve is consistent. At the beginning of absorbtion (C.V.) phase, the cell voltage delta is around 10-15mv, and quickly reduced to ~5mv. There is how it looks on my grafana dashboard (work in progress using mpp-solar / telegraf / influx via ble):

 

[Hans Kroeger]

And now my typical charge curve is consistent. At the beginning of absorbtion (C.V.) phase, the cell voltage delta is around 10-15mv, and quickly reduced to ~5mv. There is how it looks on my grafana dashboard (work in progress using mpp-solar / telegraf / influx via ble):
View attachment 134776

Hi there,
May I ask which Equipment did you use to sample the cell voltages and the current?
Regards Hans

 

[kanelr]

Seems odd to me to try to bleed from the full pack to help balance, unless you need to set the tail current higher than the balancer can make use of. I'd look toward a 12V incandescent bulb or three, and put one across a spiking cell. IIRC, car headlights are generally around 55W at 12V, so you'd get about 15W, 4.5A from the cell you put it on. Only reason for more than one bulb is to do multiple high cells at once.

ETA: It's hard to get any 230V load that can directly drain from a cell. Maybe a hot water heater element would do something, but still looks like a 4.5kW element would only bleed off ~290mA.

I mentioned elsewhere but i use a cheap modified sine wave inverter with alligator clips and a small space heater plugged into that. It pulls the hot one down fairly quick, when its balance day (tomorrow- yuck)

 

[TorC]

@kanelr Huh? How did you find a 3V inverter to draw down a single cell? Or, if 12V, what would you do if the hot cell is right between a couple low cells, short of disassembling the pack?

 

[Beene Brothers]

I feel it's overkill in most cases to try and bleed off individual cells when there is an active balancer involved. I will assemble, verify BMS settings, attach my 5a active balancer and after a few hours of the battery getting "full" (first cell hits 3.65v) the whole pack is in line and ready to go. Just too much hassle IMO to assemble the battery for top balancing at 3.65v, wait a week, then reassemble into a real pack.

Well done OP! I see your pack is all balanced out. Stay safe over there! Hoping Russia backs off or is backed off soon, senseless killing going on over there IMO.

 

[kanelr]

@kanelr Huh? How did you find a 3V inverter to draw down a single cell? Or, if 12V, what would you do if the hot cell is right between a couple low cells, short of disassembling the pack?

Ha. great question. I am 5s2p so I don't use a dedicated bms, though I do constantly monitor top and bottom voltage differences with a good bluetooth battery monitor. I have to pull 10 acorn terminal bolts for my manual balance but ity's not too bad everything stays put an inch from where it was connected, etc.