Trouble balancing battery

Hi all.

I've built a 10s 6p battery pack, and I can't get the balancing BMS to balance. Looking at the specs for the BMS, it says that when a pack gets to 4.18V, a balancing charge of 35 mA is applied. But of course, even my smallest charger applies way more current than that, so the whole thing gets to the shut-off voltage before any real balancing can occur. When any one pack reaches 4.25V, the BMS shuts all charging down until that pack gets below 4.15V.

Neither of my chargers appears to be able to come back on after reaching max voltage. I've been trying to go through multiple cycles of letting it top out, then rest for awhile to drain the high packs, then unplug to reset and plug it back in again. But it really isn't working very well. After replugging, it only charges for a very short time and then shuts off again. And I'm afraid all the topping up is bad for the high cells. So I did some manual balancing, but I'd like to have a better solution for the long term.

I'm wondering if it would make sense to get a high-ohm, high-wattage resistor and put it in line with the charger output. Theoretically that would reduce the charge current so that I'm not overwhelming the balancing function. I'd just use it right at the end of charging every once in awhile to rebalance. Is this a good idea or no?

Thanks in advance.

ETA: It looks like the failure to re-start charging is with the BMS. The balancing function both turns on and releases at 4.18V, but the over-voltage shutoff does not release until 4.15. So if any one pack gets up to 4.25V, the BMS shuts off charging, and discharges that pack to 4.18V. But because it never gets down to 4.15V with the balancing current, the BMS never releases the over-voltage protection.

Stupidly designed BMS.

Use the right tool for the job, a dedicated balancer, or replace the BMS with an adjustable one,

higher balance current is what you need, and ideally starting at a much lower voltage.

Otherwise, get a power source that you can turn the current way down and sit at just over the (stupidly high) start-balance voltage, feeding just over the (stupidly low) balance current.

Or use a balancing charger.

Or break the pack and rebalance manually all cells in parallel - but I'm guessing the pack design makes that difficult?

J A G said:

Looking at the specs for the BMS, it says that when a pack gets to 4.18V, a balancing charge of 35 mA is applied.

Click to expand...

Is it a balancing 'charge' or resistance 'discharge' balancing?

J A G said:

But of course, even my smallest charger applies way more current than that, so the whole thing gets to the shut-off voltage before any real balancing can occur.

Click to expand...

Sounds like 'discharge' balancing of higher cells to match lower cell voltages which takes place when the charger is left plugged in after the green light comes one. Can take hours and hours (at 35mv )or even days with used salvaged cells and cheap second grade Chinese cells.

J A G said:

When any one pack reaches 4.25V, the BMS shuts all charging down until that pack gets below 4.15V.

Click to expand...

When you say "any one pack" do you mean any one parallel group? Sounds like typical BMS resistance 'discharge' balancing.

J A G said:

Neither of my chargers appears to be able to come back on after reaching max voltage.

Click to expand...

That's the way it's suppose to work. Overcharging the cells is a NO NO!

J A G said:

I've been trying to go through multiple cycles of letting it top out, then rest for awhile to drain the high packs, then unplug to reset and plug it back in again. But it really isn't working very well.

Click to expand...

Is sounds like your DIY 10S6P is madeup of used salvaged cells or cheap Chinese cells of questionable quality.

J A G said:

After replugging, it only charges for a very short time and then shuts off again.

Click to expand...

That's the way it should work. With new brand name cells the charger shouldn't come on again, at least not for a several hours.

J A G said:

And I'm afraid all the topping up is bad for the high cells.

Click to expand...

No kidding!

J A G said:

So I did some manual balancing, but I'd like to have a better solution for the long term.

Click to expand...

Split the 10S6P into two 5S6P packs so you can balance charge each using two RC Lipo/Lilo balance chargers (e.g. 2S-6S or a duo balance charger) so you have complete control over your balance charging ... bottom (e.g. 3.3v - 3.4v), middle (or bulk charge) and/or just top balancing (e.g. 3.9v - 4.05v). Depending on the quality of the cells and the age of the battery what technique one uses will vary over time, but you will extend the life of the battery by only balance charging to 4.05v.

Either eliminate the cheap BMS or wire in a switch so you can disconnect the BMS during TRUE balance charging. Resistance balance discharging of higher voltage cells until they match the lower voltage cells at or near the end of bulk charging by leaving the charger plugged in (with green light on) is a poor choice, but better than nothing. Such discharge balancing (leaving charger plugged in with green light on) can take hours or even days with a battery of used salvaged cells and cheap second grade cells of questionable quality.

eMark said:

Is it a balancing 'charge' or resistance 'discharge' balancing?

Click to expand...

It is resistance discharge balancing.

By one pack, I mean one parallel group.

The cells were purchased new, and were all tested for capacity (which averaged the nominal 3400mAh almost exactly), then sorted so that each parallel pack had approximately the same capacity. One of the balancing wires came loose without my realizing it, and one individual cell was not properly connected, so it was more a problem with my build than with the cells, I think. They're not off by much -- about 0.15V -- but I'd like to have the option to safely balance them. I'm going to try my resistor idea and see if that will work first. It theoretically should get all of them to 4.18V without having to tear the pack apart.

Once you get it balanced it should stay balanced for a long time. You could try putting a resistor in series with the charger to limit the current to under 30mA but depending on the charger it might not work because it shuts off when the current drops below a certain point.

If you have a cell that leaks voltage, there is nothing you can do to balance the pack. Even a balance charger will be of no help. You can troubleshoot it by finding the group of cells that has a lower voltage then the rest.

When you take those cells apart, right away they will read the same voltage, but after 8 hours, one cell will be almost dead while the rest read the same.

If that is your packs problem, you have to fix it right away. If you leave the pack unattended to fix later, a bad cell will completely drain a pack in a matter of weeks, destroying the good cells too.

If the charger you are using is a cc/cv charger, and the pack is not too far off in balance, it should lower the amps near the topend and allow the 35ma bms to do its job.

no that lowering of amps at the end is not due to the charge regulator, but battery resistance as you get toward the end of the CV stage.

Like at 99% SoC.

Ideally you are doing that at 4.05 or 4.1V not 4.2

The balancing should have started back at 3.9V in that case, and at a 1A rate.

A good BMS lets you change the settings.

Or a good charger lets you manually adjust the current rate down to 1A so if the stupid BMS forces them to spend 15 hours balancing at least you aren't murdering the cells (as much).

Breaking the pack apart and re-testing / winnowing / re-balancing all in parallel would be ideal.

fechter said:

Once you get it balanced it should stay balanced for a long time. You could try putting a resistor in series with the charger to limit the current to under 30mA but depending on the charger it might not work because it shuts off when the current drops below a certain point.

Click to expand...

I guess I'll find out. I have two chargers, so maybe that gives me a better shot at having one work.

jonyjoe303 said:

If you have a cell that leaks voltage, there is nothing you can do to balance the pack. Even a balance charger will be of no help. You can troubleshoot it by finding the group of cells that has a lower voltage then the rest.

...

If the charger you are using is a cc/cv charger, and the pack is not too far off in balance, it should lower the amps near the topend and allow the 35ma bms to do its job.

Click to expand...

I don't think leakage is my problem. I haven't charged it in months and it is not any more unbalanced than it was.

My charger does decrease to about 0.7 A at the end. But that's still 700 mA, which overwhelms the 35 mA balancing current.

john61ct said:

no that lowering of amps at the end is not due to the charge regulator, but battery resistance as you get toward the end of the CV stage.

Like at 99% SoC.

Ideally you are doing that at 4.05 or 4.1V not 4.2

The balancing should have started back at 3.9V in that case, and at a 1A rate.

A good BMS lets you change the settings.

Or a good charger lets you manually adjust the current rate down to 1A so if the stupid BMS forces them to spend 15 hours balancing at least you aren't murdering the cells (as much).

Breaking the pack apart and re-testing / winnowing / re-balancing all in parallel would be ideal.

Click to expand...

Can you recommend a good BMS that isn't too spendy?

First balance your pack before you install BMS.

Do you mean under $400?

Maybe those reco'd in the bluetooth BMS thread?

Chargery's not too much

The red Daly ones?

Personally I would not use a BMS, get a dedicated balancer kit, or maybe use a balance charger.

There aren't many at 10S though, splitting the pack 2x 5S gets you more and cheaper choices.

Can you post a picture of the BMS?

I've used many of the "cheap" style with good results. For badly out of balance packs I use a single cell charger to bring up the low cells. I found some solid strand copper wire that fits nicely into the BMS plug. One the cells were balanced, they have done nicely for years. The balancing function is pretty useless but having the protection against over and under voltage is important.

I wouldn't worry about bringing the cells up to 4.2v as long as you don't leave them there for weeks at a time in hot weather.

Yes that"s good but you must sit and watch and sit, Because as soon as you get and get a cookie it"s over 4.2 volt very bad. Sit.

I recently bulk charged my pack to 4.1 , but got distracted and let it get down to 0 amps for about 2 hours. Is it likely that damage was done?

What I don't understand 0 amps if you're doing constant voltage CV at 0 amps as long as you haven't got older your voltage is fine.

Define "damage".

Immediately noticeable reduced performance?

likely not

Reduced lifecycles off the back end?

maybe, some

in a gentler use case, for sure, but high C-rate usage, less likely to be significant

Danger burning down the house? 100% sure higher risk

eMark said:

Sounds like 'discharge' balancing of higher cells to match lower cell voltages which takes place when the charger is left plugged in after the green light comes one.

Click to expand...

Is it normal for a 'discharge balancing' type BMS to only balance when the charger is done charging 'when the green light comes on' ? Could you explain why the BMS would not start 'balance discharging' earlier (before the charger's green light comes on)?

Thank you!

it does. but the current the balancers can shunt around the cell is much lower than the current the charger is outputting, so they cant actually drain the high cells, or even prevent their overcharge.

so only after the charger shuts off (after the bms cuts off the input current, causing teh charger to shut off when it detects the current drop), do the balancers have a chance to begin draining the high cells.

999zip999 said:

What I don't understand 0 amps if you're doing constant voltage CV at 0 amps as long as you haven't got older your voltage is fine.

Click to expand...

I was using an LED driver to charge it . It was set at 57.6v (14s) @570 mA per cell(total 4A) , where I would normally terminate when it reached 50mA , but I got busy and came back like OH crap ! and it read 0 A . Meaning it charged until it wouldn't absorb anymore (I assume ) .

Some people would use such a profile

hold 4.1Vpc until trailing amps drop to zero

as a benchmark standard prior to running a timed CC Load cap test.

But only once a year, or every 100 cycles, not frequently.

I would caution to just try not do make it a habit.

Fact is you really won't be giving up much **actual usable** capacity / range

by just doing CC-only, don't hold that 57.6V for any Absorb/CV time, just "charge to and stop" for normal usage charge cycles.

Depending on the C-rate, likely getting to 98-99% SoC with that, and can just use a reliable HVC to terminate charging automatically.

Such a profile will be inherently less stressful, so you could also bump the voltage up closer to 4.2Vpc and shorten charge time a bit.

With good cells, a battery does not "need" a BMS.
Find out if you have good cells!
Manually balance your battery.
Apparently you can access your cells contacts to test voltage?
At "full charge" note the precise voltage of every bank of cells. "Note" means write down.
One, or a few, higher? Bleed down high V banks(s), I use a 12V 3A 3D printer heating element (in can of water), slow and predictable at 4.2V. 3D printer heat element
One, or a few, lower voltage? Charge up individual bank(s), I use the 1s setting of iMax B6 or run leads from single cell charger.
Run 1, or more discharge\charge cycles and note banks charged voltage. Fixed?

2nd level
After all banks of identical voltage, let set several days and note any voltage loss, if bank 1&2 then BMS is drawing power unevenly to function (or malfunction as I view BMSs). If any other bank has dropped in voltage, there is a bad (self-discharging) cell, and that bank requires disassembly and a cell replacement!
If all banks still equal, do a typical discharge, then precisely note bank voltages. If all banks equal then cells are good.
1 or more banks low? Then bad cell(s) or bad-poor connection at some cell. Requires connection repairs or cell replacement.

A personal note on BMSs
I variously describe a BMS as a security blanket - make you feel safe abusing battery?
or a
Band-Aid for a defective battery
It will not "fix" a defective battery, but it might give you some time to notice and, hopefully, repair the problem, before catastrophic failure.

I, typically, use recycled cells, with tried and true testing and build procedures, producing larger batteries, (to limit "strain"), of proper balance at "full" and "empty" and thousands and thousands of miles on each ... with no BMS, or balance charging (Annual ? tune-up).
Cheap BMSs might do my builds more harm than good, with parasitic draw, while a "good" BMS is liable to cost as much as my recycled cells! I'd rather monitor battery voltage and condition manually!

DrkAngel said:

With good cells, a battery does not "need" a BMS.
Find out if you have good cells!
Manually balance your battery.
Apparently you can access your cells contacts to test voltage? ... I'd rather monitor battery voltage and condition manually!

Click to expand...

AGREE!

fechter said:

The [35mv discharge] balancing function is pretty useless, but having the [BMS] protection against over and under voltage is important. For badly out of balance packs I use a single cell charger to bring up the low cells. Single Cell Charging Hookup.jpg ...

Click to expand...

Isn't this Single Cell Charging Hookup checking the average voltage of the six cells in each of his 10S6P parallel groups?
Would J A G have to partially disassemble the pack to test each single cell in a P group?



Have you (J A G) yet determined which parallel group(s) are most unbalanced? Have you yet tried using a "Single Cell Charging Hookup" (excellent suggestion with your problematic pack)? One or more of your parallel groups is most likely the cause of the problem--NOT the charger or BMS. First you need to perform the above to determine that in fact the problem is with one or more of the parallel groups and which one(s) and NOT the Charger or BMS.

Even though you say the cells are new are they name brand quality cells or cheaper Chinese cells (seconds) of questionable quality? Write down the cell voltage of each of the suspect parallel group(s) you test and post here. Then test the voltage of each cell in that suspect group and also post its voltage.

You may need to use true balance charging or possibly replace a cell or two depending on their variance. Consider adopting the following technique ... which is better than the best BMS ($400). You may need to partially disassemble and retest the voltages of each cell in the suspect parallel group(s).

john61ct said:

Personally I would not use a BMS, get a dedicated balancer kit, or maybe use a balance charger.
There aren't many at 10S though, splitting the pack 2x 5S gets you more and cheaper choices.

Click to expand...

Opting for TRUE balance charging is a great learning/teaching experience that you won't regret; even moreso with a DIY battery pack having new cells of questionable quality and for an aging pack to extend cycle life ... ... https://endless-sphere.com/forums/viewtopic.php?t=81203 ... using this procedure and only charging your battery pack to 40.5-40.7v and never discharging the battery pack below 33.5v will increase its charge/dischaarge cycle life longevity.

What is it you intend to mean with that "TRUE" qualifier?

I have owned dozens of different balance chargers thanks very much, and none were "pseudo" balancing.

If you mean IP, some of the "fake" clones were better build quality than the original, sometimes even improving the firmware.

eMark said:

Isn't this Single Cell Charging Hookup checking the average voltage of the six cells in each of his 10S6P parallel groups? Would J A G have to partially disassemble the pack to test each single cell in a P group?

Click to expand...

Yes by definition, a group of cells in parallel cannot be checked in any way as individuals.

The group needs to be atomized into cells each isolated from the others first.

However cells in a string can each be checked / tested independently without breaking up the string, since current does not flow from one series-connected cell to another.

Somehow you 've managed to misinterpret the content of my previous post

If you will reread it again you should see that it's a comparison between the BMS used by J A G that balances via discharge balancing of any cells having a higher parallel group cell voltage than the other parallel group cell voltages of his 10S6P pack. The reason for TRUE(all caps) balance charging is because J A G initially referred to his BMS as balance charging(misnomer) when his BMS uses discharge balancing. There's probably more than a few that think a BMS is balance "charging" which is FALSE. Whereas, TRUE balance "charging" is "charging" a parallel group having a lower voltage than the average voltage of the pack's parallel groups. So TRUE balance charging is best done by balance CHARGERS, and not by the majority of BMSs that balance by discharging the voltage of a parallel group having a higher voltage than that the other parallel groups. In fact one parallel group could be over 4.3v with a fully charged pack and take hours before a typical BMS lowers that parallel group to 4.2v. A BMS is more for protection (BPS), and not TRUE balance charging.

Thus, it's no surprise that J A G is using a typical BMS that discharge balances at say 35mv and even then not until his 10S6P pack reaches 4.18v. We know that discharge balancing is NOT TRUE Balance Charging. With the BMS used by J A G some of the cells (e.g. less expensive Chinese cells); especially a pack with more and more charge/discharge cycles and more cell variance with some cells possibly as low as say 4.0v; while other cells may be as high as 4.4v when fully bulk charging a pack with cells of questionble quality. Then if that isn't enough include a mediocre BMS that doesn't discharge balance his 10S5P pack until charged to an average parallel group of 4.18v or pack voltage of 41.8v. And then doing any discharge balancing only at 35-40mv which can take hours, days, or never, especially with a pack of second-grade cells even after just 100 cycles, if not sooner.

Packs with an IR variance of more than about 2mΩ/cell are more likely to get out of balance ... sooner than later. This is more noticeable with less expensive Chinese packs of questionable cell quality and aging packs with many charge/discharge cycles.

Bottomline is that a BMS is still a good idea if only for protection(BPS), even with TRUE Balance Charging. Again, using all caps (TRUE) because there are more than a few ebikers that think all BMSs balance by charging voltage which is FALSE ... the vast majority of BMSs balance by discharging a parallel group voltage that's higher than the other parallel group voltages. Depending on the condition of the pack discharge balancing could take at least 16 hours once a month, if not days, or never depending on the pack.[size]

eMark said:

Somehow you 've managed to misinterpret the content of my previous post.

Click to expand...

Ah now I see what you intended to mean, but in fact it's based on a misconception.

99.99% of the hobby balance chargers out there use the exact same passive bleeding / resistive method as do most of the BMSs.

In fact the methodology of the balancing has nothing to do with the category of device doing the balancing.

Active, capacitor based "charge shuttling" balancing is performed by other / fewer BMSs

and there are also shunt-based BMS as well, for example the Zephyer.

There are chargers that charge each cell / group with a separate regulator, each controlled completely independently of the others, as if you had a whole bunch of 1S chargers each attached to one pair of the pack's balance leads.

Best referred to as "node" chargers IMO.

Your generally low opinion of BMS as balancing devices is not warranted, plenty of them are out there that perform perfectly well in that role.

They just aren't the ones most people buy when they don't know what to look for, or think that the price of the BMS should not be high compared to the cells.

But the good quality balancing chargers aren't cheap either.