Battery shutting off at 43v and will not take charge.

[MangDav]

About 5 months ago while riding, I noticed that my 48v 20ah battery died after 15 minutes of riding when it normally lasts me 2 hours and it would stay off until I plugged in a charger, after some troubleshooting, I noticed that the battery would shut off when the battery voltage sag reached 43.3v. The battery would also not shut off instantly but gradually drop voltage until it wasn't enough for display/controller to stay on. I recorded a video of this issue: https://www.youtube.com/watch?v=_xP2UpDr-Yw

I didn't have time at the time to deal with this and I had spare batteries anyway so I put this battery on the side and fast forward 5 months, I'm trying to diagnose and repair it, keep in mind, I only have basic tools and knowledge regarding ebike batteries so I'm looking for some guidance.

From sitting for 5 months, the battery discharged itself to roughly ~44v so I tried charging it again but the charger would shut off after 2 seconds, the charger is STC-8108LD so I bought another one just incase this one was bad, same model and same issue, only charges for 2 seconds before shutting off so I tested both chargers on working batteries and they charge different batteries just fine so I know it's not the chargers. First thing I tried was changing the BMS, the BMS that the battery had was a dual port so I replaced it with a dual port and same charging issue, I tried a different model of a 54.6v charger and it does charge the battery, I left it for a minute without an issue and the battery pack voltage did go up by 0.8v, not sure if this charger doesn't have some protection inside that causes the STC-8108LD to detect a problem with the battery and stop charging. At this point I was confused and this was getting out of my knowledge area, I did read somewhere that when a battery gradually shuts off like mine did, it could mean bad cells or bad bank of cells so I tested the series voltages at the BMS connector and these are the resaults with differences worked out for convenience:
1: 2.97v
2: 2.66v
3: 3.49v
4: 3.76v
5: 3.66v
6: 2.84v
7: 4.00v
8: 4.07v
9: 3.96v
10: 4.04v
11: 4.01v
12: 4.04v
13: 3.90v

Lowest is 2.66v and highest is 4.07v which is a big difference. I'm not sure where to go from here, should I try charging the battery with the charger that doesn't shut off and see if these cell banks will level off or would be safer and smarter to charge up the lower voltage banks to roughly 4v so they match the highest bank and see if the battery will continue shutting off at 43.3v? Even if that fixes it, could a bad BMS that stopped balancing cells cause this? But then why the new BMS isn't balancing the cells? Not sure if it helps but the cells inside this battery are: SINC ISR18650 3.7v 2600mAh cells. Thanks for any input!

 

[Hwy89]

It is so far out of balance that the BMS shuts it down immediately. You could manually rebalance but it would take a very long time and it isn’t a task that you can start and walk away from. You would need a single cell 4.2volt charger and a connector that will plug into two pins at a time on the balance cable removed from the BMS. Bring each group of cells to within .1volts.

 

[MangDav]

It is so far out of balance that the BMS shuts it down immediately. You could manually rebalance but it would take a very long time and it isn’t a task that you can start and walk away from. You would need a single cell 4.2volt charger and a connector that will plug into two pins at a time on the balance cable removed from the BMS. Bring each group of cells to within .1volts.

Thanks for the reply. Is there any clear indication on why these groups of cells went out of balance in the first place? Could it be just a bad BMS? Since I made the post, I did some more troubleshooting and noticed that when I plug in the STC chargers, the battery pack voltage drops from ~44v to just under 40v, then the charger shuts off and the battery pack voltage goes back up to ~44v, my guess is that maybe the BMS is trying to balance the lower banks but this causes a drastic voltage drop which causes the STC charger to cut off.

I don't have single cell charger, however, I do have 5v 1A power supplies and dc-dc buck converters, I could use the converter to drop the voltage from 5v to 4.2v, this should also drop the current to about 850mA, I hope this is fine. I understand this might take days to balance all the banks and would require constant attention with a multimeter.

 

[j bjork]

As it isn't just one or two groups or so that is lower than the rest, I dont really think it is a bms issue.
It can be partly that your bms doesn't balance, but the root cause is probably bad cells.

What kind of battery is it, quality or cheap Chinese?
What kind of cells, round 18650, 21700 or similar?
Do you know brand and model of cells?
Pouch or prismatics?
How big?

I think your best chance are to manually charge or discharge every group to the same voltage and see how it behaves.
Was it gradually loosing capacity slowly, or did it happen relatively fast?

 

[MangDav]

As it isn't just one or two groups or so that is lower than the rest, I dont really think it is a bms issue.
It can be partly that your bms doesn't balance, but the root cause is probably bad cells.

What kind of battery is it, quality or cheap Chinese?
What kind of cells, round 18650, 21700 or similar?
Do you know brand and model of cells?
Pouch or prismatics?
How big?

I think your best chance are to manually charge or discharge every group to the same voltage and see how it behaves.
Was it gradually loosing capacity slowly, or did it happen relatively fast?

The battery is a Chinese battery but I have several batteries of the same model from the same manufacture that are now 10+ months old that are drained and recharged everyday and I had no issues with them. This battery suddenly died after 3 months of daily use. I drained the battery to 39v and charged it overnight, the next day the battery was at 54v but died after 15 minutes when the voltage sag reached 43.3v, because the battery just "died" overnight, it made me think that it was BMS issue, I didn't notice capacity drop beforehand or even the day before this happened, the battery behaved normally and just went bad overnight.

The cells are cylindrical 18650s, model SINC ISR18650 2600MAH cells with total of 104 cells in this battery pack.

 

[Hwy89]

It is more than likely that you have some weak cells and the way to find out is to balance the pack then monitor it for self discharge for a period of time. If it seems to be stable then discharge/recharge and check again. The time and effort involved may not be worth it. Even if it comes back to life you won’t want to trust it.

 

[docw009]

[
I don't have single cell charger, however, I do have 5v 1A power supplies and dc-dc buck converters, I could use the converter to drop the voltage from 5v to 4.2v, this should also drop the current to about 850mA, I hope this is fine. I understand this might take days to balance all the banks and would require constant attention with a multimeter.

With #6 at 2.66V, that's so far below LVC, the BMS won't allow charging, If you get ,75A out of it, 3-4 hours of charging with the little supply should bring #6 to 4V. Do #1 and #2 too, Then if your BMS is a balance BMS, it should equalize overnight, Best to pull the balance connector when you do this, I've also done it this way with an unbalanced battery. Another possibility is to charge the three low groups post 3.3V, which shouldn't take long, and if you have a balance BMS, see if that can take over..

You might review your BMS order to see if it was a balance BMS. THey would list a balance current if it were, but even that is no indicator, I almost bought a BMS on amazon, and the ad listed a balance current, but reviews said it didn't balance.

So get all the cells back up to 4.1V or higher. Give it it a good run on the ebike til the pack is almost discharged and then look at the group voltages again to see how even they are. Hopefully, they're all the same. Then the problem is fixed. If not, then you have to decide if it's worth pursuing.

If you still see #1,#2, or #6 way too low, there is something wrong with them. Maybe they are self discharging on their own.Your BMS may be able to bring them back with a balance charge, but they might just droop on their own just sitting there. ANyway, that's speculation. Good luck.

 

[amberwolf]

The battery is a Chinese battery but I have several batteries of the same model from the same manufacture that are now 10+ months old that are drained and recharged everyday and I had no issues with them. This battery suddenly died after 3 months of daily use. I drained the battery to 39v and charged it overnight, the next day the battery was at 54v but died after 15 minutes when the voltage sag reached 43.3v, because the battery just "died" overnight, it made me think that it was BMS issue, I didn't notice capacity drop beforehand or even the day before this happened, the battery behaved normally and just went bad overnight.

The cells are cylindrical 18650s, model SINC ISR18650 2600MAH cells with total of 104 cells in this battery pack.

Since two of the packs are fine, but this one has multiple groups (not all sequential at the most negative end of the pack) that are very low, there is something wrong with the pack itself.

The most common problem is that the cell groups themselves have a defect, which could be:
--bad cells, internally failed in a way that drains other paralleled cells.
--bad cells, internally failed in a way that disconnects those cells from the groups (or otherwise greatly reduces the ability of those cells to supply the group)
--broken connections between cells in the same parallel group, reducing capacity on that group and making it discahrge faster (and to a lower voltage) than the other groups.

The other common problem is the BMS is damaged, and has balancers stuck on which are drainign those specific groups.


Does the BMS have balancers on it? They generally look like a bunch of identical groups of tiny parts, and may be under a drizzle of "goo" (hard or soft) so you can't see the parts themselves, but the stuff leading into and out of the goo still indicates a series of groups of parts under it. If there are no balancers, then the BMS is not likely to be the problem.


One test for the BMS vs cells is:

Disconnect the BMS from the pack's balance leads, and recharge the cell groups slowly (I would go a lot lower than 850ma, perhaps a quarter of that, for cells as low as you have there, until they are at least 3.something volts), until they are similar voltages to the rest of the cells, then let the pack sit for a few days.

If the cells ahve not changed in voltage, they are not likely self-draining. If they have dropped again, one or more cells in each low group has a problem internally and would need to be replaced to fix this.

If the cells aren't draining, reconnect the BMS balance cable. Let the pack sit for a few days.

If the cells have not changed in voltage, the BMS is probably not draining the cells. If they have dropped again, then the balancers on those low groups are stuck on and are draining the cells, and it's time for a new BMS.

 

[MangDav]

The battery is a Chinese battery but I have several batteries of the same model from the same manufacture that are now 10+ months old that are drained and recharged everyday and I had no issues with them. This battery suddenly died after 3 months of daily use. I drained the battery to 39v and charged it overnight, the next day the battery was at 54v but died after 15 minutes when the voltage sag reached 43.3v, because the battery just "died" overnight, it made me think that it was BMS issue, I didn't notice capacity drop beforehand or even the day before this happened, the battery behaved normally and just went bad overnight.

The cells are cylindrical 18650s, model SINC ISR18650 2600MAH cells with total of 104 cells in this battery pack.

Since two of the packs are fine, but this one has multiple groups (not all sequential at the most negative end of the pack) that are very low, there is something wrong with the pack itself.

The most common problem is that the cell groups themselves have a defect, which could be:
--bad cells, internally failed in a way that drains other paralleled cells.
--bad cells, internally failed in a way that disconnects those cells from the groups (or otherwise greatly reduces the ability of those cells to supply the group)
--broken connections between cells in the same parallel group, reducing capacity on that group and making it discahrge faster (and to a lower voltage) than the other groups.

The other common problem is the BMS is damaged, and has balancers stuck on which are drainign those specific groups.


Does the BMS have balancers on it? They generally look like a bunch of identical groups of tiny parts, and may be under a drizzle of "goo" (hard or soft) so you can't see the parts themselves, but the stuff leading into and out of the goo still indicates a series of groups of parts under it. If there are no balancers, then the BMS is not likely to be the problem.


One test for the BMS vs cells is:

Disconnect the BMS from the pack's balance leads, and recharge the cell groups slowly (I would go a lot lower than 850ma, perhaps a quarter of that, for cells as low as you have there, until they are at least 3.something volts), until they are similar voltages to the rest of the cells, then let the pack sit for a few days.

If the cells ahve not changed in voltage, they are not likely self-draining. If they have dropped again, one or more cells in each low group has a problem internally and would need to be replaced to fix this.

If the cells aren't draining, reconnect the BMS balance cable. Let the pack sit for a few days.

If the cells have not changed in voltage, the BMS is probably not draining the cells. If they have dropped again, then the balancers on those low groups are stuck on and are draining the cells, and it's time for a new BMS.

I decided to buy a bench power supply which is supposed to be delivered tomorrow, I'm going to need it for a future project so I decided to buy it now to use it on this battery pack, since I will have fine control over voltage and current, I can drop the current all the way to 100ma, I don't mind if it takes day to balance out all the groups, this is just a side project and I don't need this battery asap or anything like that so I'm not in a rush to get this done.

I kept the old BMS that I replaced in this battery, I replaced it with the same model so I took apart the one that is believed to be faulty and took pictures of it. I did try finding out if this BMS has balancing about a week ago, tried googling the part number but was not able to find data sheets or infomation on this BMS, I did find the manufacture's website and I found very similar BMSs to mine and they did have balancing built into it so I assumed that this one is likely to have balancing as well. Here's the picture of this BMS:

There's few components that look burned but they aren't, that section of the board is covered in clear layer of protective coat which is very reflective and it was hard to get a sharp picture.

 

[amberwolf]

I kept the old BMS that I replaced in this battery, I replaced it with the same model so I took apart the one that is believed to be faulty and took pictures of it. I did try finding out if this BMS has balancing about a week ago, tried googling the part number but was not able to find data sheets or infomation on this BMS, I did find the manufacture's website and I found very similar BMSs to mine and they did have balancing built into it so I assumed that this one is likely to have balancing as well. Here's the picture of this BMS:

There's few components that look burned but they aren't, that section of the board is covered in clear layer of protective coat which is very reflective and it was hard to get a sharp picture.

There's no typical parts visible for the resistive shunt type balancers on this side of this BMS board.

What is on the back side of the BMS board?


It is possible that this is a capacitive charge-shuffling type of balancing board, but it is also possible that it is not, and only detects HVC/LVC and possibly overcurrent and shuts charge / discharge off as needed based on those. If it is the charge-shuffling type, it is unlikely to have the stuck-on balancer type of failure that drains cell groups.

 

[MangDav]

I kept the old BMS that I replaced in this battery, I replaced it with the same model so I took apart the one that is believed to be faulty and took pictures of it. I did try finding out if this BMS has balancing about a week ago, tried googling the part number but was not able to find data sheets or infomation on this BMS, I did find the manufacture's website and I found very similar BMSs to mine and they did have balancing built into it so I assumed that this one is likely to have balancing as well. Here's the picture of this BMS:

There's few components that look burned but they aren't, that section of the board is covered in clear layer of protective coat which is very reflective and it was hard to get a sharp picture.

There's no typical parts visible for the resistive shunt type balancers on this side of this BMS board.

What is on the back side of the BMS board?


It is possible that this is a capacitive charge-shuffling type of balancing board, but it is also possible that it is not, and only detects HVC/LVC and possibly overcurrent and shuts charge / discharge off as needed based on those. If it is the charge-shuffling type, it is unlikely to have the stuck-on balancer type of failure that drains cell groups.

I didn't take a picture of the back because there's isn't much going on there, it's mainly just traces. There's only seven 205 resistors by the balacing connector and five R003 resistors by battery negetive terminal, nothing that would indicate any sort of major circut.

 

[amberwolf]

Then it isn't likely a shunt-type balancing system, if it does balancing at all, so it is probably not a BMS failure causing the cell group voltage issue.

That leaves the cells themselves, or their interconnects.

If it's interconnects, you may see different voltages on cells within a parallel group; that's not possible if the interconnects are intact.

 

[MangDav]

Then it isn't likely a shunt-type balancing system, if it does balancing at all, so it is probably not a BMS failure causing the cell group voltage issue.

That leaves the cells themselves, or their interconnects.

If it's interconnects, you may see different voltages on cells within a parallel group; that's not possible if the interconnects are intact.

The interconnect seem fine visually, I did charge all series to similar voltages with a bench power supply, recorded the voltages and left the battery to sit for about a week and then I'll re-check the voltages and see if they discharged at similar rate or not.

Another theory that I thought of that makes sense to me but perhaps might be impossible due to how BMS charging circuitry is designed but after riding all day without any issues and discharging the battery to 39-40v, all 13 series were around 3v, either during that day or the night when I left the battery to charge, a failure on the BMS caused 3 series not to charge and they remained at 3v while the BMS charged the 10 other series to correct voltages perhaps even over charging them to compensate for the 3 series at 3v, the next day after 15 minutes of riding while the 10 of 13 series were at correct voltages, above 4v, the 3 uncharged series discharged below 3v, around 2.7v and when the battery pack voltage sag reached 43.3v, those 3 series voltages dropped closer to 2v and that was causing the battery pack to drop voltage gradually until 16v like it was shown on the video, that would explain why the battery was fine the day before and why the battery pack was acting strange after 15 minutes the next day and why right now 10 series were close to 4v but 3 of them were 2.6-2.8v. I'm not sure if BMS charging circut can fail on certain series or if it fails, it just fails.

 

[amberwolf]

The BMS doesn't actually charge anything. All it does is monitor the cell groups for HVC and LVC (and some of them monitor for overcurrent), and turn off the appropriate FET set (charge or discharge or both) when one of those limits is exceeded.

The charger itself does all the charging, of the entire series string of cell groups all at the same time. ("bulk charging")

Some BMS also balance the cells (not all of them do); usually by top-balancing (at the "full" end of charge). Balancing is done by the BMS in one of two ways:

resistive "shunting" of current around a cell group while charge current is flowing, once a cell reaches it's "full" trip point below HVC, then if the cell goes above the HVC it turns the charge input off and drains a little bit of the cell group charge until it drops below the trip point, and the charge input is turned back on. This goes on until all the cell groups are "full".

charge shuffling; with a similar set of trip points, a charge pump is used to take a little charge from the high cell groups and pass it to the low cell groups. It's more efficient than the resistive method, and doesnt' have the stuck-on-balancer failure mode, but it is more complex; I don't know if it balances any faster. There is a potentially worse failure mode than the stuck-balancer, though: If the charge pupm system is not programmed to stop at some point, and there is a damaged cell group that is "leaking" internally, it could drain all the other cell groups over time (equally), trying to keep that damaged group at the same level as the others.

It won't cause the failure you see in that pack, though--that's got to be interconnects or cells.



There are some RC chargers (and other types) for BMSless packs that do individual cell charging, but your system is not like that. Your charger would have to have individual leads direct to each cell group and the main pack +/- to work this way.

 

[amberwolf]

The interconnect seem fine visually,

Unfortunately a visual check won't tell you if they are good, though some problems may be visible, a lack of visible problems doesn't mean there aren't any--welds could be broken under the interconnects, etc.

Also, if any of the interconnects ahve been touched, even a basic test of them may not show a problem, because touching one might cause it to connect again (handling the pack without the case that allows pressing on the interconnects, or twisting it slightly that causes connection changes, etc).

But the basic test you can try that's easy is (while the pack is mostly drained empty) to measure the voltages of each individual cell in the group. Meter on 20DCVolts, black lead to cell negative end, red lead to cell positive end. DON'T put the lead on the interconnects, ONLY put them on the cell itself. If the cell cannot be touched without touching the interconnects, then there's no good way to do this test. If you put the lead on the interconnect you're only testing the interconnect voltage.

IF the interconnect is disconnected from a cell in the group, you're not reading that cell's voltage, and if you push on the IC hard enough to read it anyway, it would reconnect it, and not show you the problem (it may even "heal" it temporarily, long enough to make it seem like the pack is now working normally).

If a cell is disconnected, and the pack is mostly discharged, and the cell got disconnected when it was at a higher state of charge, it'll be higher voltage than teh others in the group. Same thing if the pack was mostly cahrged, but the cell at lower state of charge. Etc.

Basically, if any cell in the group is a different ovltage, when measured *at the cell itself and not the interconnects*, then it is not connected to the rest of the cells, because if it were it would be the same voltage since it would be in parallel with effectively zero resistance to the others.