Issue with Headway 40152 battery pack

[hsors]

I have just finished assembling a new battery pack for my e-bike. This is a pack of 16 Headway 40152 LiFePO4 cells of 15Ah with a 16S BMS from BMAbattery. Discharge Current : 30-60A.
This issue: the pack doesn't behave like a LiFePO4 battery pack. Voltage drops progressively during the discharge cycle and goes down to 36-37V and the battery is never switched off. Therefore instead of getting always the same power as I should with a LiFePO4 battery, I loose power as the battery charge decreases. I also believe that leaving an LiFePO4 battery cell voltage going too low can damage it.
Has anybody heard about similar issue ?
Is there any chance the BMS is faulty ?
Any idea about how to fix this if this is fixable ?
Thanks for any help.
-Henri

 

[dogman dan]

Doesn't sound like your bms is set right to me. Should pop closer to 40 or 42v.

Voltage will drop as you ride, but you should be seeing a big drop at the start, followed by a slower drop through the meat of the ride.
What to do meanwhile? watch your pack with a voltmeter, and when you see that sudden drop, STOP! Note when it happened, and stop sooner next ride.

You have to stop riding for a sec, or at least get off the throttle and coast a few seconds. You can't judge by the voltage under load, especially if you are seeing a ton of sag under load. If you are really pulling 60 amps,, that's hammering the shit out of your battery.

 

[hsors]

Thanks for the hint.
In the meantime I have disassembled the pack and measured the voltage of each cell. The first and the last cell of the pack are at 0V. And when I plug to the charger, there is only half a volt. This seems to confirm the BMS is out of work, right ? (it's brand new...). Or the cabling is not correct (there isn't much instruction with this BMS from BMSBattery: https://bmsbattery.com/bmspcm/323-16s-lifepo4-bms-battery-management-system-bms-pcm.html)
Also do you know if an LiFePO4 cell that has gone down to 0V is definitively out of work ?
Thanks much,
-Henri

 

[dmwahl]

There's likely some damage to the 0V cells, but not necessarily enough to be a problem. LFP is more tolerant of abuse than some of the other Li-ion varieties. The longer you wait to recharge them the worse the problem will be. I've done the same accidentally on my lawnmower pack (8S2P, headway LFP). That was a few years ago and it still has a capacity of 20.5Ah (20Ah nominal). Get a single cell charger or lab supply and charge up all the cells to 3.60-3.65V, at which point the charging current should drop below a few hundred mA. Then reassemble and watch the voltages. Ideally you would record all the voltages at full, ride a while, record voltages, ride, record, etc until cells start dropping below 3.0V resting. If they all get to the 3.0V point roughly together, I'd say you're fine.

 

[redilast]

I would only charge the low voltage 0.5v cells up again if you suspect they have been under that voltage for under a week. Or at least that is what I have read. Much longer than a week and there can be some serious long term damage to the cell and you are better off recycling it at that point. Definitely sounds like the BMS is out of wack or not connected correctly. You may want to consider some cheap RC style voltage monitors so you can see the voltages of the cells during normal use to see if there are any problems.

 

[dingoEsride]

I had a couple of low volt cells on my headway pack that was stored for a while and the bms did mess with the charging so a cell balancing type charger helps top them all up then the bms may make sense of it,
the wiring should be like this,

 

[hsors]

The BMS I am using (from BMSbattery) - photo here below - has only P- and B- outputs on the board and a connector (second photo here below) with 16 outputs to be connected to the 16 cells. There is no documentation with the BMS and BMSbattery technical support (tech@bmsbattery.com) doesn't answer.
I have connected B1+ to the + of cell n°1, B2+ to the + of cell n°2... up to cell n°16 which + is connected to B16+ on the board.
Then I connected the + of cell 16 to the + of discharge (the motor) and to the + of charge (charger).
B- of the board is connected to the - of cell n°1
P- of the board is connected to the - of charger
And last, of course the - of cell 1 is connected to the - of discharge (motor)
The question is whether the BMS is faulty or if cells 1 and 16 (the two which are at 0V) were dead on arrival. The first few km I rode were OK (meaning all the cell voltages were probably OK) and then when power was starting to miss after a few km (2 ou 3 Ah) I checked the voltage and it was at 44V, meaning the two suspected cells were already at 0. The cells were delivered almost fully charged and the pack had been connected for only a few minutes to the charger before the first ride. So either the BMS has killed cells 1 and 16 or there were DOA.
Any hint from anybody with similar experience well appreciated !
Thanks,
-Henri

 

[dingoEsride]

The question is whether the BMS is faulty or if cells 1 and 16 (the two which are at 0V) were dead on arrival.

you will need to sort out the weak cells either with whom you bought them off or bring them back up to voltage manually with a cell charger before the bms will work

 

[hsors]

I understood that charging a LiFePO4 battery is not simple (need first to charge with constant current up to a certain voltage, then switch to contant voltage). Not an easy process...
Isn't the BMS supposed to do that ? Will it manage each cell individually and treat each one according to its initial voltage and bring all of them to the same final state ? Or should the user make sure all the cells are initially at the same level before they are "transferred under the BMS responsibility" ?
Thanks,
-Henri

 

[dustNbone]

B- of the board is connected to the - of cell n°1
P- of the board is connected to the - of charger
And last, of course the - of cell 1 is connected to the - of discharge (motor)

It looks like you've bypassed the ability of the BMS to disconnect the battery from the motor when it hits LVC on a cell. You need to connect the motor (-) to P- instead of directly to the negative end of the pack.

 

[hsors]

That makes sense, thank you !
Now do you know if the BMS can manage the charge of a pack with two cells at 0V and the others at 3.x V ?
Thanks again,
-Henri

 

[amberwolf]

If the BMS is functioning / designed correctly, it will prevent pack charging if any cell is below it's worst-case cell level lvc, to prevent a potential catastrophic conditiion.

Additionally, even if it does allow it to charge (I wouldn't trust any BMS that did allow this), it will take weeks or possibly months for it to be able to balance the entire capacity of a cell like that. (because it has to keep stopping the charger while it drains down all of the other cells from the too-high voltage they wlll each be charged to, and adds just a tiny amount of charge to each of the 0v cells, and repeating that process over and over and over until eventually someday it fills up the 0v cells (if it can, assuming they don't "leak" the charge out as fast as it goes in)).

So you would need to individually charge each of those cells up to the same state the others are at before you can easily use the BMS to manage charging the pack as a whole.