Replace Cells?

pollywaffle

10 mW
Joined
Sep 25, 2018
Messages
23
Hey,

I have a 48v battery pack. Of the 13 cell banks 12 are measuring about 4.2v but one is down to 3.5v.

Should i replace these low voltage cells?

Will any damage be done by just leaving them for a few months?

Sorry, I don’t know much about this stuff.
 
If the cells are low due to physical damage from your previous crash or water immersion, they need to be replaced.

If they are low due to being drained by the damaged BMS, then you can simply leave the pack on the charger continuously for a few days (or more likely a few weeks), and the BMS will rebalance the pack.

If it does not, or the problem gets worse, then the BMS is probably defective and is killing that group of cells.

If you measure a voltage across the larger black shunt resistor(s) on the BMS, even when the pack is not on the charger or bike, then the BMS is defective.
 
ok. I think i will take it to a local dealer and get them replaced. Thanks for the scary info. :)
 
Better scary info than a scary incident. ;)

Before you replace them, you should test teh BMS as I noted above, or else it will just kill any new cells, too.
 
yes, you should replace those low voltage cells, the first step to do a battery pack is cell sorting, to make sure cell consistency.
voltage difference should be less than 5mv.
because faulty cell lowers overall voltage, causing early cut-off
Weakest cell is stressed most; stack deteriorates quickly
 
amberwolf said:
If they are low due to being drained by the damaged BMS, then you can simply leave the pack on the charger continuously for a few days (or more likely a few weeks), and the BMS will rebalance the pack.

In the past, I was led to believe that BMS often run their continuous processes from the power of a single cell/group, and that if the battery is left idle for a long time, the one cell or group could be drawn down to cutoff voltage while the rest of the pack is nearly full. Is this still true? If so, why would the BMS qualify as "damaged"?
 
Can i also ask how many cells to each group of cells? Like it may only be one or two cells in that first bank that are potentially damaged right?

Sorry if this is a dumb question.
 
The typical cheap BMS draws the power for the control circuit from cells 1-3. Over a really long time, those cells will get lower than the rest. If only one group is low, it could be a bad channel on the BMS. 3.5 is not dangerously low and when a BMS channel fails, it typically drains the cell to near zero.

I would try charging just the low group to bring it up to the same as the others and then watch the voltage for a few days resting. If the voltage starts dropping more than the others, something is bad. It could be the BMS or it could be a bad cell.

If the voltage still drops, then you want to recharge that cell and disconnect the BMS and watch again for voltage drop. This will tell you whether it's the BMS or the cells.
 
You can use an old 5v phone charge via the bms jst connector.
Cut the charge connector off the phone charger end, bare and solder a small metal pin (one's that snugly fits in to the jst connector) to each wire end and insulate as much as poss leaving 5 or 6mm of pin showing. Insert the V+ pin from the charger in to the corresponding position on the jst of the low cell group and the V- pin in to the pin before the low cell group. Turn power on and monitor voltage with a meter until it equals the other cell readings.
 
Ahhhh, clever. So the negative pin just goes into the pin with the black wire and the positive into the pin going to the low cell group?
 
Chalo said:
In the past, I was led to believe that BMS often run their continuous processes from the power of a single cell/group, and that if the battery is left idle for a long time, the one cell or group could be drawn down to cutoff voltage while the rest of the pack is nearly full. Is this still true? If so, why would the BMS qualify as "damaged"?
If that's what's causing it, it wouldn't be damaged, just not a good design for packs that aren't charged enough to rebalance frequently. ;)

At least some of the cheap BMSs do this, probably most of them do--run off either the most negative group of cells, or a few of them. There've been plenty of threads about batteries this has happened to, when stored over winter / etc.

But there seem to be more and more of them that run off teh whole pack voltage, based on reports here and there on ES.

The latter are probably the ones that use an MCU of some type, or a BMS chip of some type, vs the ones like the old Ping / Signalab BMS and similar that are all discrete components.

There's also more and more of them that have pads for an "off" switch on the BMS that disables all the power draw, though having that function actually wired out to a user-accessible switch anywhere isnt' a given. Seems like a lot of the tube-mounted battery cases do have an on/off switch (either discrete or in the keylock), and that's where it usually connects.
 
pollywaffle said:
Ahhhh, clever. So the negative pin just goes into the pin with the black wire and the positive into the pin going to the low cell group?
The wires have to go across the low cell. They will be adjacent holes in the BMS connector. Since the phone charger can over charge the cell, you need to babysit the voltage and disconnect it before it gets above 4.2v. A diode in series with the phone charger will drop the voltage a little to make it safer.

Like this:
 
Thanks, i got it. Lucky for me my BMS has blue tooth connectivity, otherwise i would have never known. Must be people riding around out there in ignorant bliss, until....
 
pollywaffle said:
Thanks, i got it. Lucky for me my BMS has blue tooth connectivity, otherwise i would have never known. Must be people riding around out there in ignorant bliss, until....

Well, until the pack dies way sooner than expected like the one in my picture. That one had a bad BMS channel that drained the cell to near zero. BMS was replaced but that cell group was toast and never worked properly afterward.
 
fechter said:
pollywaffle said:
Thanks, i got it. Lucky for me my BMS has blue tooth connectivity, otherwise i would have never known. Must be people riding around out there in ignorant bliss, until....

Well, until the pack dies way sooner than expected like the one in my picture. That one had a bad BMS channel that drained the cell to near zero. BMS was replaced but that cell group was toast and never worked properly afterward.

I might know exactly how you felt soon enough. That group charged up 7mv in an hour then only 9mv in the next 2.5 hours. We’ll see. Thank you for the info.
 
Not good.

Is that with the BMS attached or not? It sounds like one of the balancing shunts is stuck on.
 
That’s with bms connected. Charging that group up now to see what happens without bms connected.

It’s a brand new BMS as I destroyed the last one in a crash with total water submersion.
 
Good to repeat the test. A new BMS can blow up like that pretty easily when it's being connected. There is a sequence that will minimize the risk of blowing it up.

If the cell holds voltage, then you could reconnect the BMS and verify one of the balancing shunts is stuck on by measuring the voltage across each shunt resistor. If one is stuck on, you need a new BMS. Replacing the parts on the board is very challenging.
 
hemo said:
You can use an old 5v phone charge via the bms jst connector. Cut the charge connector off the phone charger end, bare and solder a small metal pin (one's that snugly fits in to the jst connector) to each wire end and insulate as much as poss leaving 5 or 6mm of pin showing. Insert the V+ pin from the charger in to the corresponding position on the jst of the low cell group and the V- pin in to the pin before the low cell group. Turn power on and monitor voltage with a meter until it equals the other cell readings.
I wouldn't do that. Too easy to look away as the cell gets near 4.2 volts and have a fire. And even if you get the cell to the same voltage, the instant you disconnect it's going to drop fairly significantly.

4.2 volt supplies intended for this application (charging single cell li ion batteries) are readily available on Amazon for under $10. Use it to charge all the cells in the pack. Voila, pack completely charged and balanced. If it quickly gets unbalanced again - you know you have a problem.
 
Okay, that cell group has dropped to 3.99v overnight without the BMS connected. The cells are damaged like I thought in the first place.

This has been a great learning experience but I truly have no idea how to replace a cell group.
 
Post a picture of your pack.

Replacing a cell group may be challenging to impossible depending on pack construction and your skill level. I have a couple packs with bad cell groups but the welding and gluing was very strong and hard to take apart without destroying the good parts. Never fixed them. One thought I had was to just cut the welds on the bad group and bypass it with a jumper wire so the pack now has one less cell group. Alternately I could use a big resistor to totally discharge the bad group, then solder a jumper across it and not cut the welds. It could be used for something but would need a different charger. My particular BMS would work fine with one cell jumpered. I'm thinking making it into a sort of powerwall device to run things during a power failure.

I've seen similar problems with several packs where one balancing shunt on the BMS gets stuck on and totally drains the cell group. This, by itself, wouldn't be destructive. The problem is when you next attach the charger, the low cell group is getting full charge current. This is what destroys the cells. If you know a group went dead and charge it very slowly (C/50) until it reaches around 3.0v, then it can most likely be recovered to full capacity.
 
Back
Top