It is so boring when things work.
Context:
It is quite a train smash when things go wrong. But once you “get it”, and understand what needs to be done to fix things quickly, the adrenaline has left one’s system and the fixing becomes mundane, it gets “boring” really fast.
Contradiction much, some will say.
Joke in the Co, I’m the “poop-scooper”… when things really go wrong, something “new” … wake me.
I always look for easier, simpler ways, to solve mundane repeat issues. THAT is fun.
15 posts were split to a new topic: Venus OS3.10: New ESS “Self consumption from battery” setting
Saw this accidentally.
Now 100% SOC is 3.45v.
I happen to check the BMS App, and the balancing working nicely … but the volts are hitting 3.55v!!!
We are talking like 63.9v … WHAT!!!
Check the Remote console, lekker charging going into the batt … the next second … both MPPTs drop to zero … and the entire load comes from the batts.
Goes on like that for a few seconds, and the MPPTs ramp up again.
All is calm now again on the batt front, balanced at 3.408v, 100% SOC.
I’m impressed … sure there is some code in there somewhere that did this on purpose.
There was a mention of a balancer on another thread 150mA is not bad …
Then I saw this NEEY … now we are talking balancers man …
Must say, if ever I have to get a balancer, I will look at NEEY range only.
This thing, maybe an idea to buy one to share … as it is WAY too much for my puny bank. 
Yes, I said that, but it is specifically within the context of other popular batteries that are even lower than that, and do just fine in the long run.
The most I’ve seen from a balancer, is 2A. I spoke to the designer of the BMS, and he said it was a bit of a waste because it’s only really used at the beginning of the live of the battery, and maybe after a deep discharge, but that they are doing it because customers are asking for it and willing to pay for it 
Yes, I’ve read and heard that all the last few years … since then, new info has come to light.
Info like the cells getting out of balance after 12/24/36 months, faster or slower, depending on use.
It is not easy to keep 15/16 individual cells perfectly in line for 3000/6000 cycles, as claimed, never actually tested in real-life conditions, and never with SA LS conditions either … and that is with BMSes that were just developed.
Andy saw his 3 x banks with good BMS’es going out of whack, watch the video, what he did.
Batteries, as we always tell each other, are not an exact science as it is a chemical reaction … it is complicated. We cannot even get the SOC exact, and precise. It is a guess.
BMSes are still evolving too. As are the balancers in them.
mA balancers are not going to cut it in 3/5/10 years …
If it gets out of balance faster than 50mA can correct… you have bigger problems
I know.
The problem is coming … is not about the first few weeks … it is the 3000/6000 claimed cycles to keep 15/16 cells nicely in line all the time, every time…
… just wait.
In other words, you think a bigger balancer will help at the EOL side of things?
I’m pessimistic about that (yeah, you can quote me ). Take Pylontech. EOL means it has 60% of the original capacity left. This happens after 6000 cycles. That doesn’t mean one cell is now dead and the other 14 are fine. It means all 15 have more or less 60% left, and if you top balance them, they continue to work.
Once one cell starts moving so fast that the balancer doesn’t keep up… that’s not EOL. That’s way beyond that, and something where a larger balancer will maybe buy you a few weeks at most.
Nope, EOL is EOL. If a cell goes, it goes, manufacturing defect that.
I think all these lithium cells in the market today, with the conditions in SA, will not last without bigger balancers, for various reasons.
Two reasons off the top of my head:
I compare it to Victron’s development over 5/10/20/30 years … how they learn in real market conditions what needs fine-tuning. The software updates, firmware updates … tuning.
Take my relay for example … they learned.
Lithium batts are on the “cutting edge” at the moment … learning as we go.
With LS and bad weather for weeks … you need a bigger balancer than mA balancing, generally speaking, or a problem starts developing ever so slowly.
Dude, I see your analogies, but they strike me as “They told Galileo I he was wrong, they are telling me I am wrong, Galileo was right, therefore I am right” reasoning.
If it is true that batteries will suddenly develop age-related imbalances after some time, imbalances so frequent and large that only a larger balancer can fix it… that would also be the point where you declare the battery dead. Constantly having to balance a battery is hugely inefficient too.
Even if it works, it will buy a few weeks at most.
There we go again.
Missing the point I’m making.
Next response, if you actually read what I’m saying … but it is the way you said it …
You have way too much faith in the software, BMS, balancers, and the individual cell construction of a new line of batteries this early in the game.
mA balancing is not enough. Finish en klaar.
I’m saying that “take my relay for example… they learned” doesn’t work. The authority of one field does not necessarily transfer to another. I would gladly discuss the other field on its own merits
So let’s break this down. The cell has a certian Ah capacity. It degrades over time, so there is a certain dc/dt (the change of capacity over time) first derivative, and this value is negative.
As long as the dc/dt for the cells are the same or close to each other, you need only a small bit of current to keep them top-balanced. This essentially is the same as saying “as long as they degrade at the same rate”. The unit for this will probably be something like mAh/cycle, since we know that the life of the cell depends more on the number of cycles than the age.
The second derivative is d^2c/dt^2, that is the rate at which the first derivative changes, or in layman’s terms, the tempo at which the wear accelerates. We expect this to be positive: Towards the end of the life of the cell, this will go faster.
Again, as long as the second derivative of the cells are the same, or close to each other, there is no problem.
The issue isn’t the factors themselves. It is the delta between them.
The crux of my argument, is that once that second derivative gets out of bounds so far that the balancer can no longer restore the balance on the next recharge, that’s when one of two things happens: Either you need a bigger balancer… or you throw away the battery/cell.
I’m not saying a bigger balancer won’t buy you time. I’m saying that if you need a 2-ampere balancer, you’re probably not far away from option 2.
A bigger balancer will buy you time. 50mA really is very low. I cannot prove this, as I don’t know the expected variance between the cells (maybe the people who make the batteries actually know this? ), but you may well be right that such a small balancer will prematurely junk a battery that had some life left in it.
But 2 ampere? Forget it… if you need that much, at the end of life… you’re just treading water, trying to avoid the inevitable.
Edit: Basically, if you lost 2Ah more on one cell than on the others, since yesterday, then the balancer will be able to balance that in about an hour. Having one cell degrade at 2Ah/cycle faster than the rest… is not a saveable situation.
Is 50mAh/cycle a saveable situation? Maybe… I have my doubts.
NO!!! Not since yesterday!!! … that is a dud cell.
After ±1.5 years Andy did the test on his bank. And he did lose some Ah, not a lot, that is inevitable, we agree. What he saw, between the 3 x BMSes, was how they handled the 3 banks differently.
This brings me to my 2nd point … you need to do maintenance now and then. Cannot just install a bank and walk away expecting it to work 6000 times perfectly. The hardware/software is not that intelligent … YET!.
We all are learning … with relays on inverters and BMS balancers on another “track” … the two have nothing to do with each other bar we are learning to refine their respective designs, based on actual experience.
EXACTLY my point.
The battery manufacturers, today, yesterday, last year, 5 years back, are doing the math the best they can in their lab conditions. You can see that at times. The specs on cells are changing, from the same manufacturer, within months for the same AH cells. They say it is a new cell design … maybe it is, maybe it is not … but that the cells are improving, that is a fact, or they last better with the new specs published … the fact is, they are learning based on actual experience, real-life applications across an immense field/range of use.
You can balance a bank over a week… or if you use a really big balancer, within minutes.
The question is … what is the right size balancer to use to ensure the life expectancy of the bank, and not damage it bit by teeny bit over 6000 cycles?
We don’t know that … yet!
We’re arguing about how big the balancer has to be. We’re not in disagreement about anything else.
My main point, which I am probably communicating poorly, is 2A is overkill. If the thing makes the wires hot… it is overkill. That thing is for getting things in line in the beginning.
Personally I think that if one cell loses 50mAh more than the rest in any one cycle, heck if it loses that on its own without comparison to the rest… the cell is toast already.
If you consider a battery that is EOL at 60% of original capacity, and let’s say you’re using really big 200Ah cells, then it degrades 80Ah over 6000 cycles, or less than 2mAh per cycle. If one of your cells is degrading twice as fast, it’s doing 4mAh/cycle. The balancer is 50mA. That’s an order more.
The thing about small balance currents… they require more smarts in other departments. As I said earlier, adjusting the charge voltage is necessary, since no commercially available inverter can accurately push less than 1A. This, perhaps, is where larger balancers makes sense: They work better with large inverter/chargers that cannot limit to less than a few amps. But again, you pay for it one way or another, either in software smarts, or components.
And I see that… but I’m not sure that’s just the amount of balancing current. That’s just that some BMSes are cheap junk. I don’t know which ones he was running against each other… I do know some of my colleagues are upset about how some brands get “official support” and then swap out the BMS for something cheaper later…
We are getting closer …
Exactly! The mA balancing falls off the cliff … causing teeny bits of problems compounded over weeks, especially if you have high LS cycles AND bad weather.
Further, as Andy also saw, using Eskom, us now, to charge a bank is more stable/controlled than using MPPTs as MPPTs are affected by various factors during the day. High amps, low amps, the balancers doing their best to keep up … over time.
So how do we “fix it”? We get into a messy “adjusting of volts” to cater for that “weakness” of balancers, where a bigger balancer that matches the inverter capability, would have been better suited than …
BTW, I’m not arguing, we are debating the merits of our views.
You just coined it in a way … to ensure the lifespan of the battery, it is a good idea to match the balancer to the inverter’s capability.
No adjusting (of volts/amps), let the balancer do the work.
He got rid of those ones fast …
I’m not going to even touch on that.
Again, I see this as more of a problem in the beginning. Once the cells are balanced, they should stay that way right until the end. Once they start to diverge so fast that larger balancers and more finely tuned charge current control is necessary… is it really worth the trouble?
In the end, that is the only question the manufacturer cares about. Should I spend money on this? Will customers pay for this?
As I noted earlier… some of them will.
And therein the challenge … the word should.
They don’t. Manufacturing is nowhere near the level of getting cells to behave 6000 times flawlessly.
Adding to that, many other factors are involved like temp, usage, how long the mA balancer actually balances… you get the idea.
You will see on your new bank, if you have the data on a screen in front of you, over months, of how the cells tend to work in “groups”, the effect when you have multiple LS and not using Eskom to recharge to 100% every day to get the balancing done … the small mA compounding over time to amps “problem” needing intervention like dropping volts, as you cannot drop the amps low enough, the problem compounding
Till you have to top-balance the bank again over days … cause the mA balacning takes so long …
Watching Andy, my own journey, seeing how my system is controlling the charging with NO link to the BMS like a pro, that the idea I have of checking the Delta, dropping the volts/amps - and the resultant impact that has on the system overall … maybe a bandaid to help the BMS balancer.
The better idea … find the right balancer for the inverter/bank I have, to handle it on BMS level with panache… year after year, irrespective of external factors…
The cost of that balancer compared to the cost of the earlier replacement of a bank, is that really a factor? It is a small % of the total cost of the bank, the right balancer.
NEEY seems to fit the “profile”, as they upgraded it, and improved the software, based on user feedback.