How can I charge/test a 4-cell pack in a single slot of a charger?

[alexon]

Hello all,

I have a 36V eBike battery made up of forty 18650 cells (10series / 4parallel).

Sometimes, when charging the whole battery, the BMS doesn't balance the voltage correctly among the 10 packs. When that happens (which is often), I use an Imax B6 charger to charge each pack individually up to 4,2V:

https://secondlifestorage.com/imghost/glibor2.jpg

This can take up to 2 / 3 hours per pack, BTW :?

I just bought an Opus C3100 charger/tester and I want to use it to charge and test each 4p pack to determine their health (I'm pretty sure that 3 out of the 10 packs have defective cells).

I know I could tear apart the whole battery and separate each 18650 cell to be able to stick them in the Opus one at a time, but I REALLY don't want to do that! (it's a bottle battery and it would be HELL to reassemble).

The question, then, is: can I charge/test a whole 4p pack by rigging a cable from the BMS connector to an Opus' slot +/- pins?

I think some people are already doing it, but with fewer cells per pack:

https://secondlifestorage.com/imghost/ycsxfm2.jpg

Does anyone know if this would work safely for 4 cells welded in parallel (samsung ICR18650-22p / 2200 maH each), with a TOTAL OF 8800maH? Would the Opus give an accurate reading of the pack's capacity?

THANKS A LOT,

Alex

 

[fechter]

According the manual for the Opus C3100:

The maximum charging capacity can be up to 20,000mAH

So you should have no problem charging a 4p group.

 

[Punx0r]

It might also be worth investing some time to find out why your battery keeps going out of balance and also why the BMS is not bringing it back into balance.

 

[Matador]

You can. The opus will charge each slot with 1A current. This current is low enought to not overheat your small BMS wires... So yes you could technically do it. But this is temporary and wont fix your defective cells.

 

[alexon]

Thanks a lot for the quick replies!

I just want to identify which of the 10 packs contain damaged cells so I can try to replace those without tearing apart the whole blessed battery

Once I know which packs are affected, I'll have to rip out the corresponding cells, test them individually with the Opus and discard the dead ones.

 

[izeman]

It was already confirmed, and i will also confirm
This will work w/o issues and you really need to do it. If a BMS is not able to balance a pack then there is something wrong for sure. I would even say that you're playing with fire as it is right now. If one cell in each pack is more or less dead, bad things can happen ...
So proceed with your tesing ASAP.

edit: you can do the testing with the imax as well! Just charge the 4P pack to 4.2V and then discharge to 2.8-3.0V depending on cells used) and note capacity. Limit charge and discharge current to 1A to not overheat the thin balance wires. Each cycle will take 4-8h (depending on SOC), but it's necessary.

 

[alexon]

Thanks again!

That's exactly what I'm doing, discharging the suspect 4P pack with the IMAX. To be on the safe side, I started with 0,5A, but it takes forever. I'll switch to 1A. The pack is supposed to have a 8800 mah capacity, but it's looking like it will reach around 6000.

Will report back

 

[alexon]

Hi there,

It’s me, Alex, reporting back.

I went and tested 6 of the 10 packs of my 10S4P battery and got the following results (the official capacity of each pack is 8800 mAh):

(By "Cell 1" I meant "Pack 1", etc)

That surprised me, because I was pretty sure that packs 1, 3, 4 and 5 where the most likely to be damaged (on account of their usually being more discharged than the rest). It turns out that they’re mostly OK, except 5 which is clearly not.

After this testing, with some packs completely depleted and others close to full capacity, I hooked the battery to the 42V charger to see how the BMS (which may or may not be working correctly) would handle the balancing. I unplugged the charger every 15 minutes and measured each pack.

This is what I got:

The results left me puzzled, really.

At some point (around 60 minutes into the process I think), the red light in the charger turned green, which I took to mean that the process was done and the charger had stopped supplying current to the battery. This is clearly not the case; I think the green light comes on when some of the packs are fully charged, but the charger still feeds the battery so that the BMS can continue with the balancing.

Then I left the battery alone for 3 days and measured the voltages to see if any of the voltages had dropped significantly : only #1 had (3,26V to 3,18V).

Then I hooked up the charger, but the green light came on immediately and no charging seems to have taken place after that (even though the voltages are all over the place and the total voltage is barely 36,94V!).

Can we conclude anything about the BMS from this little experiment?

Thanks again,

A

 

[Punx0r]

Assuming each cell group started the discharge test fulluy charged, the widely differing capacities measured indicate the battery is rather ill, unfortunately. Group 5 in particular looks like it may have a completely dead cell in it.

The balancing function on most BMSs is very slow - 50mA is typical, some are less. Starting, as you did in your test, with some groups fully charged and others discharged you'd need to leave it charging for several days to let it balance.

I'm not completely sure how the lights on your charger work, but the green light probably just indicates it thinks the battery should be at full voltage - it has no real way of knowing if the pack is balanced.

 

[DrkAngel]

BMS stops the charging when 1 bank hits HVC.
BMS, typically, only discharges highest voltage cells, it will not charge up low voltage cells to balance.

Recommend you manually charge-balance all banks to identical voltage.
Leave set for several days, mark any banks with notable self-discharge - problem #1.
Recharge-balance.
Discharge entire battery till pack very near empty or faults.
Open and measure voltage of each bank and mark, this will tell you which banks have lower capacity - problem #2.

Either indicates a bank that requires separation, individual cell test and repair = cell replacement.

 

[DrkAngel]

BMS stands for Battery Management System, unfortunately there is a great variance in the function and quality of management.

1st level includes:
High pack voltage cut off
Low voltage pack cut off
High cell pack cut off
Low cell pack cut off
Amperage exceeded cut off

2nd level additionally includes:
basic balance function - bleeds down cells with excessive high voltage - does not charge low voltage cell after 1 cell or bank hits HVC
Might partially discharge excessive voltage cell(s) then begin charge cycle again - repeat-repeat-repeat over many hours!

3rd level:
a true balance function that bleeds down high voltage cells and individually charges up low voltage cells

Your BMS might be a 1st level and is certainly no better than a 2nd level.

 

[alexon]

Thanks a lot, Punx0r and DrkAngel, for your helpful feedback!

The purpose of this whole experiment was twofold: to identify which battery packs would need a cell replacement (which would entail individual cell testing with the Opus), and to assess how well the BMS is working; as there is often quite an imbalance between packs at the end of a typical battery use or after charging.

As to your points:

* Yes, everything indicates that the battery as a whole is in poor shape and may not be worth the hassle. I want to learn about this stuff, so I’ve decided to investigate further, but if I had a repair shop I would probably not dedicate a bunch of hours trying to fix it.

* It’s probably the case that my BMS is not malfunctioning; it’s just a crappy BMS. The balancing function appears to be slow and crude, and seems to assume that the packs more or less discharge and charge at the same rate, which is clearly not the case in my battery. For example, the first few times I measured each pack, some of them were charged up to 4,27V, which worried me (that’s why I don’t leave the battery charging for hours on end anymore). But if I don’t trust the 42V charger + BMS combo to do the job, that means I would have to open the battery every day and charge the packs individually using the Imax (or the Opus). Clearly not worth it.


In any case, I’ve purchased all the components to make a new battery, so I will concentrate on that for a while. The new BMS I got is one of those with Bluetooth and a Android app showing the health of each pack in real time. Sounds like fun.

Once I get that working, I’ll tear my current battery apart and try to salvage as many cells as possible, to then rebuild it as a backup battery. In the process I’ll try to think of ways to improve the design of these batteries, which are CLEARLY not meant to be easily diagnosed and repaired. I find that economically and environmentally abominable, BTW.

Will report back…

 

[izeman]

Imho it doesn't make sense to take batteries from an old pack an mix them with new ones. Your pack will ALWAYS be unbalanced. Those old cells always sag a lot more, even if they have similiar capacity.
And your BMS will never be able to adjust for very different powerful batteries. As already mentioned, the balancing current is quite low, and it can only help to improve slight imbalance. The manual of those BMS states that as well: Take care that all cells are equal voltage when you connect the BMS - it will not be able to balance a pack of empty and full (old and new) cells in an adequate amount of time.
There are quite long thread about those BT enabled BMS. I like them very much. Very cabable, and being able to see how your pack's state is with a phone is very comfortable.

 

[DrkAngel]

Good cells, when well treated, don't even "need" any "balance", or even any BMS.
Matched capacity, quality cells will be equal voltage at near full discharge and will naturally recharge to same full voltage they started at. I am not advocating against using a BMS ... I have faith in my builds and do monitor my batteries.
Keep in mind that your Samsung cells are effectively empty at a static 3.40V, discharging deeper is abusive and will damage cells.
I built 25.9V 26.4Ah last year and still maintains equal charged-discharged voltages within 2/100th V.




Most important is to remove any cells exhibiting self-discharge, test all ... even new cells!
next, build banks of equal capacity
then, moderate C-Rate to prolong performance and longevity
either moderate throttle or build battery of larger capacity
Discharging cells at "rated" C-rate will cause rapid deterioration, short lifespan and premature failure!