Replacement cells for older LFP-4812S 48V pack

[Pavelstorek]

Hello,

I have obtained ebike with Magic Pie and older 48V12Ah pack, but the previous owner indicated reduced capacity of the pack, supposedly because of faulty cell/-s. When I measured voltage across the 13 individual groups of 3, it indeed seems, that one group has measurably lower voltage and so, even if the rest of the pack would be able to continue, the BMS switches off the pack way too soon.

The solution would be either to somehow reduce the pack to 12 series (not yet sure, if this would be possible with the current BMS and charger?) or to obtain 3 new cells to replace the faulty group.

Is it feasible? Are the cells available somewhere? It seems like the size should be 26650 and capacity 4000mAh, correct? Would different source/manufacturer be a problem to combine with the originals?

Any advice, other than to purchase whole new pack would be appreciated.

Thanks,
Pavel

[Bikemad]

Hi Pavel andto the forum.

I just had a quick look on eBay and found these:

Click image for more details

and these:



But they don't state what the actual current output is, and it would be better if you can find some tagged cells so that you don't have to solder direct to the end of the cells.

I wouldn't worry too much about trying to match the manufacturer, because if you managed to get the same ones as you already have, the chances are they will discharge at a slightly different rate anyway as they would be new in comparison to the other used cells.

As long as you can find some with the same voltage and capacity, it's probably the best you will be able to do.

Alan
 

[Pavelstorek]

Hi Alan and many thanks for the help.

I found quite a lot of cells, but all with quite low discharge current and the ones you linked are also questionable (seen some rule of thumb, that you want to look for orange wrap, rather than blue, if you need high current rating...).

I am afraid, that the type originally used  (MNKE ??) are not in production anymore and I found only two possible sources with good parameters:

http://www.batteryspace.com/limnni-26650-battery-3.7v-8ah-29.8wh-20a-rate-w/o-pcb-for-diy.aspx

and
http://www.vaporalley.com/IMR-26650_p_94.html?AffId=10

Both on the expensive side,though.

Before taking the plunge, I will try to measure the pack a bit more, under load, to make sure that the whole exercise would make a difference. The connections are not particularly easy to disconnect and reconnect.

I am also afraid, that if I now replace one suspect group, that very soon, there may be a problem with another.

So, one direction would be to go with brand new pack (any practical recommendations? definitely want to saty with 48V and also would prefer high current rating, as I want to play with the settings in the controller a bit) The new 48V10Ah from GM seems a bit on the expensive side and also the max current is a bit low, I think.


I was also thinking about reusing the BMS and casing from the old pack, but to find good value cells that would also be a good fit dimension wise into the aluminium housing seems a bit of a tall order for me. Unless someone already did the ground work and can recommend cells?

For my purpose the total capacity (Ah) itself is not so critical, even 8-10Ah would be probably ok, as my trips are rather short. But the current rating is important, as I live in hilly area and to climb steep hill, I need to be able to draw as much as possible. Unfortunately, due to basic physics, it seems then to work out that I need rather high capacity pack, as the discharging current is in direct proportion to capacity, right?

The older LiMN cells had advantage in high current tolerance, as opposed to LiFePO cells, but had much shorter lifespan, I think.

Any ideas what direction should I look?

Goal 48V 8-10Ah ideally in the same casing as the original GM battery, with high current rating and not exorbitant price....

Many thanks,
Pavel

Hi Pavel andto the forum.

I just had a quick look on eBay and found these:

Click image for more details

and these:



But they don't state what the actual current output is, and it would be better if you can find some tagged cells so that you don't have to solder direct to the end of the cells.

I wouldn't worry too much about trying to match the manufacturer, because if you managed to get the same ones as you already have, the chances are they will discharge at a slightly different rate anyway as they would be new in comparison to the other used cells.

As long as you can find some with the same voltage and capacity, it's probably the best you will be able to do.

Alan

[Artem]

Hi Pavel.

Here I describe how I bypassed my faulty cell (if you can read in Russian, or you may use Google translate): http://electrotransport.ru/ussr/index.php?topic=7476.msg184798#msg184798

I'm also interested in new cells, may be even 6x, because I'm planning to have 14th triplet there. Let me know if you find something.