Hey all, a while ago I got a new 24v 10ah LifePo4 battery from Golden Motor. I was surprised to find when charging it, the charging cut off abruptly at 27.1v. Other LifePo4s I have usually charge to 3.65v per cell, which means this battery should charge to 29.2v (which is also the output of the charger I'm using which works as expected on other 24v packs). I also noted, after testing, that the capacity was quite a bit lower than advertised (~200wh vs expected 240, but my meter could be off). The pack wasn't getting hot - no excessive voltage drop- the problem to me seemed like cell imbalance. I recently had on an older LifePo4 pack which had a similar early cutoff issue. I ended up fixing it by hooking it up to a small solar panel + controller for a few days, thinking that the daily slow charge / nighttime settling cycle might allow the BMS to balance. It did. From that point on it charged up fully again. So for this one I tried to do something similar but without solar - I simply got a DC converter with CC/CV, dialed it to 29.2v, and dialed the current all the way down (15w is about as low as it would go). Unfortunately, I've been unable to get the pack to go even a smidgen beyond 27.1v. It just rises at a constant rate and then instantly cuts. So I'm just confused at this point- what is going on in there? My technical understanding is that, if this is a typical passive balancing BMS, if a cell group went above 3.6v ish (standard LifePo4 balance voltage - maybe they custom set this somehow?) the BMS should start bleeding that group. That bleeding is slow though, so in theory, if a group was too much higher than the others, then during charging, that group's voltage could still rise above a safe limit, and thus the BMS would force all charging to stop abruptly. The lost total voltage and capacity would be from the other groups which didn't get a chance to fully charge before cut off. That's the only explanation I see. But if so, slower charging a few times should at least begin to correct this, no? Charging closer to the bleed rate would allow time for those high groups to bleed down and the low ones to catch up, before one trips over the limit. But it's just not happening. So, anyone have any ideas on what's up with this battery? I should mention I've had it a few months now and it does seem like a decent battery- this just bothers me. And also, I tried taking this issue to Golden Motor and they just had no idea what I was talking about. So here I am. Thanks for reading.
New battery won't charge all the way
- Thread starter Graham123
- Start date
docw009
1 MW
It's unbalanced, but there are different extremes. Is one cell too high, or one cell too low? The only way to tell is to open the battery and measure the group voltages, At that point depending on what you find, you may want to try manually equalizing the voltages and hoping that the BMS will maintain them.
j bjork
1 MW
Are you sure the bms has a balancing function?
Maybe it just cuts when a cell group goes high.
If you have given up on warranty I would open it up and measure what is going on
Maybe it just cuts when a cell group goes high.
If you have given up on warranty I would open it up and measure what is going on
You're right, it could be a low cell group. But wouldn't slow charging help to balance it regardless? Unless the BMS doesn't do balancing - but there definitely is a BMS, I know that much, and it's rare to find one that doesn't at least to passive balancing.
docw009
1 MW
I haven't seen a balance BMS in the last six batteries I've bought, They only have overcharge and low voltage checking, I have replaced some of those with balance BMS. I took for granted that they worked/
Last year, I built a half doxen batteries and now I was interested in whether they actually balanced as advertised. Some do. Some don't.
Last year, I built a half doxen batteries and now I was interested in whether they actually balanced as advertised. Some do. Some don't.
Last edited:
Unfortunately, BMS with no balance function have become fairly common, so you'd have to check if yours has that or not.
Since a balancer really only lets you use a problematic pack longer before it becomes unusable due to differences in cell characteristics, it might actually be better that way--it will show a buyer that their new pack is bad right when they get it, because it won't charge to full either at all or few cycles in.
If they'd use good quality cells *and* do cell property matching before building packs then they wouldn't need balancing (until the pack starts getting old).
Even if it has a balancing function, if hte imbalance is severe, it could take a very long time to correct--most of them have at most a 50mA shunting ability. If the problem cell(group) is several Ah different, say 8Ah (random#), then 8000mAh / 50mA is at least 160 hours to rebalance that. The BMS would turn off the input, let the high group drain down, reallow charge current, and repeat until all the groups are equal.
Also, the balance function could be broken--a channel could be stuck on so it constantly drains the cell (group) attached to it, or it could be unable to turn on, and never do any balancing on that cell (group).
You can find out what the problem is by measuring the voltages of all cell groups, starting at hte most negative and working your way ot the most positive, when the charger stops the first time.
If there are groups significantly different from the others, then you have the common imbalance problem.
If they are all the same, then it's not imbalanced, so it's likely the BMS is setup for that voltage per cell.
Etc.
Thank you for that. It reminds me - one interesting thing I'm noticing is that there is no settling going on after charging. The pack remains exactly at where it cut off - 27.1v, for 6 hours now (ill check again tomorrow). If a high group went over the limit, I'd expect to see the overall voltage dip as that group settled. I know LifePo4 hates to stay at 3.65- so this is pretty mysterious to me.
Then the most likely thing is that the BMS is designed or setup with a cell HVC of just under 3.4v (27.1v/8s), and all the cells are at that HVC and equal.
But you'd need to open it up to measure the cell(group) voltages to be certain of that.
3.4v is actually what LFP cells tend to sit at full when aged (bigger "block" prismatics may settle there on their own if you charge them and then let them just sit there), and if you look at their discharge curve there is not much capacity between 3.6 and 3.4v. (a few percent at most). If they're never charged more than that to start with, then they may even last longer (more charge cycles before loss of capacity).
So, it's not really a bad thing, unless you need those extra Wh. (the extra voltage would go away pretty quick on a ride, so probably not a factor).
But you'd need to open it up to measure the cell(group) voltages to be certain of that.
3.4v is actually what LFP cells tend to sit at full when aged (bigger "block" prismatics may settle there on their own if you charge them and then let them just sit there), and if you look at their discharge curve there is not much capacity between 3.6 and 3.4v. (a few percent at most). If they're never charged more than that to start with, then they may even last longer (more charge cycles before loss of capacity).
So, it's not really a bad thing, unless you need those extra Wh. (the extra voltage would go away pretty quick on a ride, so probably not a factor).
Maybe it's a whole pack HVC, not per cell? Otherwise, how would balancing ever occur? (unless they set a shunting threshold below 3.4)
I don't typically see a pack-level HVC or LVC on BMSes, just per-cell. The only pack level limits are typically current and temperature. Newer smart BMS (like those with BT that use phone apps to see their data) might have other pack level stuff; not sure.
Interesting, thanks. Sorry to keep on going with this -feel free get on with your life- but I just don't completely get the intent of the per cell HVC. So if any cell reaches above that, the whole charging cuts? That doesn't seem to allow any out of balance high cells to surpass a threshold where they'd start being shunted. Is it possible that threshold was set lower? I never hear of it being below 3.5ish. Over time the pack would simply cut off charging at lower and lower overall voltage, being limited by these wayward cells. And no way to fix it except to open it up. Am I on base here?
Correct. Once the cells are no longer matched, even if you had a balancing BMS, they'd continue to go down hill. Balancing would extend the life or help restore/maximize the remaining capacity. HVC is there to keep you from destroying the pack by stopping the charging. If your BMS had balancing, the HVC would cut off charging the pack and allow the balancing to occur.
I think my confusion is starting to dissipate. For years I've been trying to figure out how ebike batteries typically balance, always running into 2 conflicting ideas - one states that balancing occurs only after charging is cut off by HVC- you guys seem to say this is the case. The other says that balancing occurs only during charging and begins around 95% charge - this was my previous assumption which I had thought was verified by my own experience. Obviously it depends on the bms, but it's good to know what the general possibilities are.
There are a few variations from what I've read about, and some can balance while still charging, some work passively bleeding off the charge from the fully charged cells, which allows the charger start charging again then repeating the cycle, while others are active and can charge the low groups through the balance leads. At one point I was thinking of removing the BMS and adding in some active balancing modules, that are always balancing the groups (if the voltage variation is greater than a set threshold). I got lazy and so the modules are sitting in my parts bin now.
Yes. The point is to prevent any cell from going above it's safe charging limit. Without that, you can have cells in an imbalanced pack that get so far above their safe charging limit that they could be damaged in a way that can lead to a fire (or if severe enough, could actually start a fire right then).
This is the typical way almost any common BMS works, whether they have a balance function or not.
If there is no balancing function, that is correct.
If there is a balancing function, then the balancer will (if passive) drain the high cell(s) down to the point the BMS can turn charging back on. (if active balancing, it shuffles charge from high to low cells). Repeat this cycle until all cells are equal.
Note that balancing cells only makes them all the same voltage, not the same capacity or characteristics...so it doesn't really fix any pack, just lets you consistently use what capacity/etc it does have.
The balancing trigger point should be lower than the HVC. (if it isn't, then the balancing isn't able to work). Most BMS have multiple trigger points, each for different parts of it's functions. Programmable ones can be set by the user. Otherwise, they're either factory set or built into the hardware itself.
Balancing can start at any point in the process, depending on the type of balancing and the design of the BMS. You'd have to check the full specs for any particular BMS, but these are not commonly available.
999zip999
100 TW
I had one pack of lifepo4 cells with 1,480 charge cycles and found 3.45v as fully charged. As above that it 3,7v is just fluff. Are very little capacity. It can take over charge to 3.7v 3.8v with little damage and not recommended. My signalab BMS from Ping could go that high and then the blinking lights would come on and the charger would come off and would bleed them all down and balance. But most of the bms's Target 3.65 volt or so.
By the way that battery of mine had no BMS just for sets of 6s balance plugs and I just bulk charged after three rc balance chargers broke and just bulk charge.
How old is the pack would you feel comfortable opening it up and finding out which one of the eight parallel group cells or cell is low. And take a reading of each cell and write it down then bring up the low ones. Still how old is that battery and what size is it. Ah ?
By the way that battery of mine had no BMS just for sets of 6s balance plugs and I just bulk charged after three rc balance chargers broke and just bulk charge.
How old is the pack would you feel comfortable opening it up and finding out which one of the eight parallel group cells or cell is low. And take a reading of each cell and write it down then bring up the low ones. Still how old is that battery and what size is it. Ah ?
Dude just open up your pack and disconnect your cell sensing wire connector from the BMS and it will reset.
Money says it will charge up to the correct value.
This is only if the pack was charging up to the correct value to start with.
If it never did that's another story. Good luck.
Money says it will charge up to the correct value.
This is only if the pack was charging up to the correct value to start with.
If it never did that's another story. Good luck.
