The real reason why your battery won't balance

[ChrisG]

Speaking of manual balancing, check your bench charger with a meter before starting each cell. My charger was set to 3.650 and the output was actually 3.673. I don’t think that much over would do much harm but better to be safe.

 

[Recovering EE]

That was never going to work, you can't fit 22 oz of beer in a 16 oz cup.

You can if you pour it over your open mouth!

 

[Alarchist]

I would also add to Will's list in post #1

6. Poor busbar connections.

I've seen a few others with this, which looks like a strange high overpotential voltage on a cell during bulk, but then that same cell will be a straggler at absorption voltage.

I had a pack that even though for all cells: I cleaned the contact points with a scouring pad, IPA and then torqued all connections to 4Nm, but one cell would always be first to hit absorption voltage early with some weird overpotential voltage. Then it would balance down until the other cells got up to abs voltage, and at that point it was now the lowest cell voltage, cycling them daily ended up pushing that cell further down from the others at abs voltage.

I finally bothered to open up the pack and cleaned up and re-torqued that cell to 5Nm and now it's no longer a problem. The JK BMS never showed any difference in balance lead resistance from the others, and the nut wasn't loose on the terminal post or any obvious sign of poor contact.. but re-torqueing worked.

In London, UK and off grid (Boat), like Upnorth we don't get enough Sun from the end of October until February to spend any decent time balancing, but I do try to run the genny on sunny days to try and get it there around once a month, but it's not the same as 100% every day.

My cells are the Gotion 340Ah and I charge with 0.25-0.3C to 3.5v and now I see it takes 15mins to balance <10mV and go to float once they hit my 28V pack voltage, which they can now do every day.

 

[BarracudaBob]

Amazon.com

 

[Skypower]

View attachment 295472

Amazon.com

That seller must have bought a ton of those

 

[fuzzmunky415]

I would also add to Will's list in post #1

6. Poor busbar connections.

Will needs to adjust title: "6 reasons why..."

 

[fnnwizard]

Can anyone guess about how much imbalance (% wise to stay applicable to all capacities) there is when the low cell is at 3.480V vs high cell at 3.650V = 170mV delta @ charge current .02C? I'm curious to see the answers.

Man, no one has data to share?

I guess I will start. at .02C (5.6A) charge rate.
the % delta is .......

less than .15% (.382/280). This is better than any BMS measurement btw.

In a new, good quality 280Ah cell, when the hi cell is 3.65V, the low one almost 3.480V the capacity difference is 331mAh -382mAh over 8 random cells tested in various 2s, 4s, 8s.

Here's a zoomed in graph of one of those test. the irregularities in the top (cell 1) final charging to 3.65V was me tightening and loosening the ring terminal to make sure torque was spot on.

Note that the lower V cell ( written as lower cell in graph) had termination current dropping due to the power supply setting at 3.65V vs when charging both cells and total V setting at 7.300V. But all that matters is the mAh reading when the cell reached 3.650V.

 

[42OhmsPA]

Man, no one has data to share?

Before I read the rest of your post I was thinking less than 1% because the tail current is so low.

 

[fnnwizard]

Before I read the rest of your post I was thinking less than 1% because the tail current is so low.

A lot of people are surprised how little imbalance there actually is at max delta (200mV) when the low cell has hit 3.45 at <.05 rates. It's less than 500mAh for 280Ahr cells. ~ .18%.

It wasn't always like that. With manufacturing processes much better, the cells are very consistent. And the costs are silly low if we think about it.

My first "large" outlay for LifePO4 are these dewalts compared to my last large outlay . Too embarrass to admit what it costs even at wholesale for what we were doing with them back then.

But man, they were good compared to everything else tolerance wise. When the counterfeits ( or rejects) started popping up we didn't know who to buy from as even some very reputable dealers got taken.

The good'ole dyas are actually now !

 

[robbob2112]

Speaking of manual balancing, check your bench charger with a meter before starting each cell. My charger was set to 3.650 and the output was actually 3.673. I don’t thank that much over would do much harm but better to be safe.

It hadn't even occurred to me to check the bench PSU -- will do that next time I am working on things

 

[740GLE]

A lot of people are surprised how little imbalance there actually is at max delta (200mV) when the low cell has hit 3.45 at <.05 rates. It's less than 500mAh for 280Ahr cells. ~ .18%.

It wasn't always like that. With manufacturing processes much better, the cells are very consistent. And the costs are silly low if we think about it.

My first "large" outlay for LifePO4 are these dewalts compared to my last large outlay . Too embarrass to admit what it costs even at wholesale for what we were doing with them back then.

But man, they were good compared to everything else tolerance wise. When the counterfeits ( or rejects) started popping up we didn't know who to buy from as even some very reputable dealers got taken.

The good'ole dyas are actually now !

But if cells climb to 300mv and over voltage protection starts kicking in before those bottom cells are able to charge.

3.3 low and 3.6 high is a wildly different state of charge.

 

[ringi]

It seems like you do initial reviews of build quality and maybe peak current capability. I’d love to see a 300 cycle review. Simulate a basic system with generic charging profiles let it charge and discharge say 2-3 times a day. See which units are able to keep balance, which units can’t

A battery charged to 100% on ofpeak electricity every night and only fully discharged on some days would perform better then such a test indicates. Personal I think it better to have more capacity of B cells, rather then fully cycle A cells every day.

 

[Zork Arg]

I don't think Joe Sixpack is mentally differentiating Lithium blah-blah-blah from Lithium scooby-dooby-doo. ;-) It seems to all mean the same or something and they hear "Don't charge Lithium batteries to 100%" and they see their EVs warning them "We REALLY recommend you stop charging at 80%". Not understanding Lithium NMC is different from Lithium LFP.
I was just mentioning to another utuber how his 3 year old video on EV batteries is still current 3 years later but people won't look at it that way and won't watch it thinking it won't apply. There were no comments since 3 years ago. Maybe utube would frown on it but it might be handy to re-edit an old video with a new intro stating the data is valid and doing an overview then go into the old video while ending it with any extra thoughts.
You did lots of work, have good data but why they're not looking at it seems a mystery.

Speaking as Joe Sixpack, the only knowledge about batteries I've gotten so far has been from this forum, YouTube videos from people like Andy OffGridGarage, and Will's text - my cronies and I would do well to find, and read, real battery textbooks, and attend some local battery university classes. (if we could only find them...)

 

[fnnwizard]

But if cells climb to 300mv and over voltage protection starts kicking in before those bottom cells are able to charge.

3.3 low and 3.6 high is a wildly different state of charge.

Yep, any cell not hitting at least 3.45V at low charge rates will not be charged enough so the imbalance is greater if other cells are >3.450V. So even 200mV at 3.400 and 3.600 starts to show greater rate delta of the V delta.

Higher charge rates skews the imbalance scales due to I^2 since as cells approach fully charged their impedance ramps up = V (over potential) ramps up (I know, stating the obvious here). resistance curve similar to uncambered snow skis.

Someone mentioned busbars as also a possible culprit, which is very true for the Hi-rate case = don't assemble batteries while eating a greasy burger. 99% C3H8O is a terminal connections' best friend and cheap insurance

BMS makers should know to have SOC only register 100% when the lowest cells has at least gone past 3.45V. We saw in another thread that some Eg4 batts register 100% at any cell = 3.55 with no other criteria, and if in coms soc mode, their inverters will stop charging so gives much less time for the cells to balance = cells become even more imbalance from the 3.55 trigger.

 

[rcguymike]

I've been recommending 58 volts for absorption for quite a few years now:

View attachment 294975

Works great to balance the packs. I've also been recommending people charge to 100%. If they don't, it will not be top balanced properly. And they will have the issue that you are speaking of.

Now if it comes imbalance brand new, I think the points I mentioned in the beginning are pretty valid. Does anyone think I missing you anything from that list? Besides software issues

The new ecoworthy batteries I just got are requesting 58.4 charge voltage from the BMS...the EG4 LL-S I have requests 56V...crazy difference. I currently have absorption at 56.8V per Victron. How long do you set absorption for? I think Victron has 2 hours for the default and 54V float? Maybe I'll bump those up. EG4 also has a flier now that says to charge some of their batteries to 57V until a cell overvoltage is triggered to reset the BMS. Why is everyone doing it so differently are their internal chemistries that different?

 

[Enketchum]

Not charging to 100% often, running high c rates, and using pack in extreme cold can create imbalance pretty easily.

That's why all my new recommendation videos say you should charge these things to 100%.

For some reason people charge to lower voltages and they think it will allow their batteries to last longer. With these solar systems, they are rarely held at 100%. Usually they are at 50-70% for most of the time if days of autonomy is designed properly.

How 'often' is 'often' ? I talk to a lot of people in the Facebook solar and RV groups about making the switch from lead to lifepo4 and keeping their stock converter. How often should they be charging to full with an external charger for top balancing when keeping the stock converter? Is it a number of times every month or a number of times every few cycles?

Everyone thank you.

 

[rcguymike]

How 'often' is 'often' ? I talk to a lot of people in the Facebook solar and RV groups about making the switch from lead to lifepo4 and keeping their stock converter. How often should they be charging to full with an external charger for top balancing when keeping the stock converter? Is it a number of times every month or a number of times every few cycles?

Everyone thank you.

I'm probably not an expert compared to most on here, but I think it's going to really depend on BMS programming. Something like an EcoFlow power station I've been doing ~3x per month.

Otherwise on something like a JK BMS I'm leaning towards twice a month for daily cycling more than 40%... that's going to be pretty hard to do during the dead of winter for me though, with my recent battery upgrade, unless I hold 90%+ through winter.

 

[westcoaststank]

Problem is that many people do top balancing wrong and destroy their cells in the process. An active balancer as part of the BMS eliminates this.

How do people top balance wrong and destroy their cells? Honest question as I'm in the process of trying to top balance my 4 "Dumfume" 12v lifepo4's with a dumb non-bluetooth BMS and glued case (so I cant measure the voltage of each cell)

 

[upnorthandpersonal]

How do people top balance wrong and destroy their cells?

- Change the lab power supply voltage while connected to cells
- Change the voltage so 'it goes faster' instead of using proper sized leads (alligator clamps, thin wires, leading to losses/low current to the cells)
- Charging to a too high voltage/keeping cells at too high voltage for too long/...

 

[westcoaststank]

- Change the lab power supply voltage while connected to cells
- Change the voltage so 'it goes faster' instead of using proper sized leads (alligator clamps, thin wires, leading to losses/low current to the cells)
- Charging to a too high voltage/keeping cells at too high voltage for too long/...

So if I fine tuned the voltage while it was connected but stayed well under 14.6v I’d be okay?

 

[upnorthandpersonal]

So if I fine tuned the voltage while it was connected but stayed well under 14.6v I’d be okay?

You can't fine tune the voltage while connected. You're (likely) in constant current mode, so the voltage at the power supply drops to whatever the battery voltage is at. Changing the voltage won't be visible at the supply.

I had to demonstrate this in the past because someone didn't believe this:

 

[Lt.Dan]

So if I fine tuned the voltage while it was connected but stayed well under 14.6v I’d be okay?

Sounds like you are also connecting to a battery through a BMS, which means the PSU could be set to 20v, but the BMS should shut down before damage occurs anyways.

Typical damage from top balancing comes from putting the cells in parallel with no BMS, and charging over 3.65v

 

[Skypower]

You can't fine tune the voltage while connected. You're (likely) in constant current mode, so the voltage at the power supply drops to whatever the battery voltage is at. Changing the voltage won't be visible at the supply.

I had to demonstrate this in the past because someone didn't believe this:

Bingo! Probably the number one cause of newbie cell death.