Ebike battery wont charge

ee:

1 mW
Joined
Aug 29, 2019
Messages
10
I have a 48v ebike battery that was stored for a long time and will not longer charge.

When I attempted to charge it the charger shuts off when the battery gets to around 24 volts. I tried to charge through the output port and disconnected/reconnected the BMS but it still does the same thing.

Before I teardown the battery I was thinking of trying to replace the BMS, however I am not sure what to buy. Can I use any 13 wire BMS or is there a specific amperage I should be looking for?
 

Attachments

  • IMG_20230825_092911.jpg
    IMG_20230825_092911.jpg
    2.6 MB · Views: 0
  • IMG_20230822_175902.jpg
    IMG_20230822_175902.jpg
    2.7 MB · Views: 0
Have you checked the individual cell group voltages? The BMS might be working perfectly to protect you from a pack half full of dead cells.
 
Any 48v battery that is below 24v actual cell voltage total is probably damaged and unsafe to recharge--the cells could be damaged in a way that could lead to a fire. That fire could happen at any time, during use or just sitting there, and you cannot know if or when it is going to happen until it's on fire.

The BMS probably turned off the ports to prevent you from recharging it specifically so this wont' happen--that is what it is designed to do.

So your BMS probably works fine, but your cells are probably damaged and need to be replaced.


You can test this by measuring the individual cell group voltages at the back of the connector that plugs into the BMS. If you can't touch the metal of the contacts from teh back with your meter's probes, you can wrap a pair of needles or straight pins in tape so they can't short out, excpet for the tips of each end. Then use the pointy ends to poke into the wires or cotnacts back there, and the other ends to touch the meter probes to.

Set the meter to 20VDC

put the black probe on one end wire of the connector. The red goes on the next wire over. Write this voltage down.

move the black probe to where the red probe is. mvoe the red probe to the next wire over. write this voltage down.

Repeat the last step until your red probe is on the other end wire of the connector; you've now gotten all 13 voltages. Post them here.
 
Yeah if the pack is actually at 24v it's toast.
 
Yes I suspected as such. Can I just replace the faulty cell groups?

From left to right they are: 3.9 4.1 4.0 4.1 4.1 0 4.1 4.1 4.1 4.1 0.4 4.1 4.1.
 
That shows the cells are mostly not too far discharged, except for the 0v and 0.4v groups.


The rest are too different from each other, and could also trigger BMS shutdown. More than 0.1v difference can trigger a bms to prevent charge or discharge, because the cells have become too different in characteristics and the pack wont' operate the way it should (will become imbalanced easily and continue to get worse over time). Replacing the failed groups with new cells will exacerbate this, because they will have different characteristics from the others, too.


So, in this case, what to do about it depends...

What votlage did the pack start out at when taken out of storage?

How long did it try to charge on the charge port before it shutdown?

How long did it try to charge on the discharge port before it shutdown?

Have you ever done any battery repair or pack building?

What tools do you have available, or budget to get them?

What kind of cells does this pack use, and how are they interconnected?
 
Last edited:
I'm not sure about the starting voltage, but the bike did power on and had some torque for about 10 mins as I did a test ride when I took it out of storage.

The pack will charge for about 3-5 mins before it shuts down if I reset the BMS. I didn't charge through the output port for more than 5 mins or so as I heard that can be bad.

I have never done any battery work but I do have a Malectrics spot welder that I have never used before.

This is an older Unit Pack Power battery so I would assume they are probably generic Chinese cells. They are interconnected via spot welded nickel strips from what I can see.
 
Very strange that the bike would work at all with those two dead cells--even if they aren't actually dead and it's a connection fault to the BMS, the BMS should have turned off the discharge port for low voltage on them.

If it has now turned off that port, then it's not broken, with the port "stuck on" (common failure), and that would be strange. What voltage do you read at the discharge port, with it plugged into the bike's controller and the bike turned on?

It also shouldn't allow any charge (thru the charge port) at all, for any amount of time, with the BMS detecting dead cells.

Also...sometimes what kills cells dead is the BMS itself--if it has a balancing function, and it's using passive shunt resistors (most common kind) then sometimes they fail stuck on, so they constantly drain the groups attached to them, no matter what. Left sitting for a while, it'll drain the cells to zero.


So...you might need new BMS, too, if this one isn't protecting against potentially destructive failure modes that can lead to fires :(, and/or it is actually what killed the cells. If you do need a new BMS, you should match the A rating of the BMS so it doesn't allow overload of the cells (or use one with even less A rating, but that may cause shutdowns while using the bike). That's assuming the A rating of the BMS you choose isn't just what it is capable of without overheating, but also is a limit so taht the BMS would protect the pack against that overcurrent.

If you can access the cell groups at the spots the balance wires are soldered, I'd recommend doing the same test as before to test all the group voltages, but do it at those points instead of at the connector (to eliminate wires/connections to BMS as a fault, and be sure the actual cells are dead.
 
The two groups under 1.0 volt are toast. It's going to depend on how the battery was built. If the cells are in a plastic matrix, it's possible to cut the nickel strips and pull them off the bad cell groups. Then if you carefully remove the molded retainers in the plastic, the bad cells can be pushed out. If the cells are glued together, it's very messy. I don't recommend replacing cells as a viable repair method, but it might be a learning experience for a hobbyist.

What I learned about it is that if you have one bad cell group, you'll probably have other cells fail. If you have two bad groups, the odds are double more will fail.

Anyway, if you want to try it, you will need a new BMS too. The new cells will have a different discharge curve and lead to continued balance issues. and UPP does not use balance BMS,.


.
 
So I was able to tear down the packaging a bit further and find the offending cell groups. I tested directly from the battery terminals and they were the same as the BMS wire readings.

If I desolder the 4 wires from the 2 cell groups would the BMS still operate or would I need a BMS with less terminals? I am ok with having reduced capacity on this battery as I don't use it much. Or would that lower the voltage too far? I think it would be pretty close to 48v with a fully charged pack.

The batteries are glued together which would make taking them apart a messy endeavor. This ebike has rear rack, and the battery is in a small box which has plenty of space so it would be no problem for me to buy some cells and glue them together/solder the leads without having to disassemble the entire battery. But would newer cells cause problems by mixing them with the aged ones?
 

Attachments

  • IMG_20230830_105704.jpg
    IMG_20230830_105704.jpg
    3 MB · Views: 3
  • IMG_20230830_105655.jpg
    IMG_20230830_105655.jpg
    3.2 MB · Views: 3
  • IMG_20230830_105650.jpg
    IMG_20230830_105650.jpg
    3 MB · Views: 3
So I was able to tear down the packaging a bit further and find the offending cell groups. I tested directly from the battery terminals and they were the same as the BMS wire readings.
That's bad news for those groups. :(

If I desolder the 4 wires from the 2 cell groups would the BMS still operate or would I need a BMS with less terminals? I am ok with having reduced capacity on this battery as I don't use it much. Or would that lower the voltage too far? I think it would be pretty close to 48v with a fully charged pack.

You will need to physically bypass (short around) those cell groups with sufficient current capability to carry the full expected current of the system. The easiest way to do that without disassembling the pack is to spot weld strips across the +/- of the dead groups. Since they are already 0V dead, then it isn't going to hurt anything, there is no energy in them to short out and cause a problem.

Then you will need to "shift" all of the balance wires starting from the first dead group so the BMS is still reading consecutive groups, for the typical BMS, assuming it is capable of operating without correct voltages present at all of it's inputs. (some are, most aren't) To do this you would need to first disconnect all of the balance wires from the cells except the ones that are on the cells that are closer to the negative side than the first dead group. Then reconnect them one at a time starting at the more negative end of the connector and working your way to the most positive cell. (if you just shift them one at a time, you may end up with enough voltage across a BMS channel to damage it).

Some BMS that can be run without all balancers connected must have the unused balancers connected to each other in specific patterns, and that you'd have to find out from the BMS maker (or seller) if they have that information available. There are a very few threads scattered around here that have pictures of certain models of certain brands that show how to do this for those specific ones, but it's unlikely yours is one of them; may take more time than it is worth to locate these images to find out, as I don't know a good set of search terms likely to turn them up. Most are in battery build or repair threads.


Personally, I would replace the BMS anyway, since it allowed you to use *and charge* the pack with totally dead groups in it, which should *never* happen. Any group that is outside the safe limits should cause the BMS to completely turn off and disallow any connection to the cells from outside, to prevent further damage that can lead to a fire. If that doesn't happen, the BMS is not working and shouldn't be used.

The batteries are glued together which would make taking them apart a messy endeavor. This ebike has rear rack, and the battery is in a small box which has plenty of space so it would be no problem for me to buy some cells and glue them together/solder the leads without having to disassemble the entire battery. But would newer cells cause problems by mixing them with the aged ones?
Not directly, as long as the new cells are at least as capable as the originals, and are the same chemistry (preferably the same model).

The main issue you'll run into is that the older cells will have lower capacity and higher internal resistance, and vary more amongst each other, than the new ones, so they will fill up first before the new ones, empty first, and sag more under load. But if it still performs well enough for your usage, that doesn't matter--you just might have to leave it on the charger longer to keep it balanced to get more of it's capacity to remain usable. (assuming your BMS has a balancing function, many cheap ones don't).
 
Back
Top