BMS

[Burgerman]

A while back I said that I was going to use a BMS instead of the usuel PL8charger for the 230Ah battery pack. Either externally or internally on the chair.
Because the PL8 is no longer available. (I have 4!) And because many of you want a BMS instead...

I bought a JK-BMS (8S 200A with serial data so I ca use the PC based software as I hate crappy phones and silly little screens).
I looked at many BMS over the years and most are crap... For various reasons. The JKBMS is the best of a bad bunch. And at least has some PC software/interface if you get the correct versions.

They are physically a bit too big to hide on a chair uness you use a smaller cell to make room. Usually...

And theres a lot of things a BMS does that we dont need or want at all. So about half of its function is just a nusance. But in order to use a decent power supply or charger with a simple 2 wire interface we are forced to use a BMS.

When it comes to settings, I recommend that you follow this guys settings below. Why? Unlike 99.9% of the garbage on youtube he actually gets it!
With 1 difference. His choosen charge CV voltage is 3.45V. Thats great for solar. Where we want a more sensible charge and balance time I suggest you choose 3.55V
And set balance start point to the same as the charge voltage minus 0.05V. So charge to 3.55, balance start to 3.50V.
Set charger (ZXD preferably) to the same 3.55V. And set it to give 30 mins of CV time. (you dont need his 1 hour as you are charging at a higher voltage and starting balance where the difference is greater. This will result in a much faster charge and still never go near the battery high limits. You dont need float. We are not solar!

Watch his video here carefully and replay any bits you dont understand. And ask me if you need anything at all clarifying.
https://www.youtube.com/watch?v=xTKiOvJ_ejg


youtu.be/xTKiOvJ_ejg

[sergey0]

Hello everyone,
I have assembled two battery packs in an 8S1P configuration using EVE C40 (LiFePO4, 20Ah) cells.

The cells are specified at <3 mΩ internal resistance each. According to the BMS readings, the total resistance of the packs is 14.46 mΩ and 15.03 mΩ respectively.

BMS: JK-B1A8S10P.
I have attached my current settings below. I would appreciate any expert feedback on which parameters should be adjusted and the technical reasoning behind those changes.
Thank you in advance for your assistance.

[Burgerman]

https://www.wheelchairdriver.com/board/ ... =2&t=13106

Ask about each setting one at a time.

But 3.4 is a little early for balance to begin. See link.

Remember that the settings in that vid are for solar setups. So he doesent want to see them sat at 3.6V or so for long periods. So prefers a lower top voltage, lower balance voltage.

But in our case we want a higher balance start, and 3.55 to 3.60V charge CV voltage. And we then want charge to end after all cells reach hat point after around half hour and charge to stop. As it allows better, faster balance, and so quicker charge and less time at CV.

[shirley_hkg]

[sergey0]

But 3.4 is a little early for balance to begin. See link.

I don't intend to charge above 3.47–3.49V per cell. I'll set my charger to 27.75–27.85V. My reasoning is based on EVE's official documentation and numerous discharge tests: there is virtually no usable capacity in the 3.5–3.6V range, yet staying at these higher voltages negatively impacts the cycle life.

[Burgerman]

Look at the chart. Why do you think it has a differentiation between 3.6V charge and resting voltage.

In order to charge at all he has to be a large enough difference between battery voltage and and charger supply voltage. Whats more is that balance state at these such low voltages cant really be seen. In solar usage we dont mind if it takes days to balance or that it takes a day of sun to fully charge. Because some days that happens. So overall balance is maintained its just very slow.

If you never charge above 3.4V and preferably 3.5V you will never actually see a battery with bad balance. The difference between a 95% charged cell and a 99% charged one is 4% by capacity and well below the noise at below 3.5v. So it wont balance properly. The VOLTAGES may balance given adequate time/charges but the actual state of charge will not be equal across the pack.

Look at the shape of that curve here to see whay its problematic.
In order to actually SEE a difference in cell volts that relates accurately to state of charge (Ah) we need to be above 3.4Vpc.

In cyclic use. I charge and set 3.55V CV with only a short CV time (made possible because the cells are actually at a different voltage) so it doesent spend long at an elevated voltage. And sees a big cell voltage difference. But the difference between relaxed batt voltage and charge voltage is great enough to ensure rapid charge and rapid balance to full. I set balance to begin ONLY above 3.450V. Because doing so earlier results in actual unbalance at times initially that then gets corrected as voltage climbs.

[sergey0]

to Burgerman
Based on my screenshot, please specify which lines should be adjusted and what values you would recommend from your perspective?

[Burgerman]

OK a PICTURE is worth a 1000 words.

Here is a battery that I assembled. Then charged.
Take a careful look. As it charges I watched and was amazed at how well matched all of its cells were! They SHOULDNT be as these were just a random bunch of cells. But you CANNOT SEE any imbalance at all below 3.4 volts. Even though they were wildly out of balance... They all look just fine! Because thats the FLAT part of the curve.

You MUST pull them all above 3.500V and preferably above 3.550 to see the full level of actual balance. Only then can they be balanced.

Look carefully. You cannot balance anything if you cannot see it!!!
Initially looked great. Then as they cross 3.41V the balance issue begins to be visible

Then as the voltage climbs higher it REALLY shows the problem!!!
And now the charger CAN balance the cells properly.

You cannot balance below 3.4V AT ALL. First 1.6 hours looked perfect! Because all are below 3.4V.
And you cannot balance properly even at 3.45V, as it ust takes forever. Not enough difference visible. You can balance properly at 3.550V and no need to go to 3.60 or 3.65 though.

[Burgerman]

That above was a 6S pack. My BM3 used 13S (a 6S and a 7S pack in series).

Heres an 8S one that IS properly in balance. Charged at 3.550V. Balance set to 3.45V start. And 1A. But thats not like a BMS that is 1A across any or all cells. So equivelent to 7A the way a BMS works. And no start/stop voltages it just balances all cells and slows the charger rate to whatever is required.
With a PL8v2 charger that is more accurate and does a better job that a BMS but is external.

[Burgerman]

to Burgerman
Based on my screenshot, please specify which lines should be adjusted and what values you would recommend from your perspective?

Set it much the same as Shirley screenshot. But charge at 3.55V. Because as you can see from my graphs of real cells, theres isnt ANYTHING at all to balance at 3.40V. Thats only where they begin to diverge! (and very little to balance below 3.50V...)

Alao I woldnt worry about cycle life if you stay at or below 3.550V and seldom actually run this massive 230Ah pack low. With 230Ah you can do 100 miles. So charge weekly. It will outlive you. Dont run it low and you will get 10,000 cycles or more. And still have twice the range of lead...

[sergey0]

to Burgerman & shirley_hkg Thanks for your help

[Burgerman]

Basically the point as you charge where the DIFFERENCE in cell voltage due to state of charge becomes apparent starts at 3.41V. Before this point as you charge the cells will be the same voltage. (Or in fact really will read with a small FALSE voltage difference due to BMS or charger calibration difference between inputs, meaning it will attempt to UNBALANCE the battery at any voltage below this point).

And no consider that this 3.41V point, is the point where the cell voltages BEGIN to spread apart due to state of charge. So its not possible to balance at 3.4V either. This is the START of the spreading. In order to see this spread happen, and to allow for some BMS or charger calibration error too, we need to see some actual cell voltage spread. That begins to look realistic at 3.450V per cell. At this new voltage its JUST possible to balance, but it will take a very long time as the difference is tiny. Balance can therefore BEGIN at 3.4V without it doing any harm (it wont unbalance it) but it doesent do a lot of good either - at least in a sensible time frame, and be particularly accurate as far as STATE OF CHARGE is concerned. This may not matter much in a solar setup.

In cyclic use it does. We dont want to end up with an unbalanced pack because inadequate V differentiation across cells, or it to take many hours to balance badly. That doesent matter in solar use. We have all day every day.

So in cyclic use, sensible charge speeds we need to set balance to BEGIN at 3.45V and charge charge to 3.500V CV per cell or better still 3.550V as it works much faster/better with greater voltage spread visible to work on. This is still a lot less than the 3.650V that the typical system is configured to do. And it only needs to stay at this voltage for a very short time (unlike solar all day) to allow a full complete balance to happen because the spread is now clearly visible.
The charger/BMS or whatever setup is used, should be configured so that it holds cells at 3.550V per cell for around 30 to 40 mins max at end of charge and then turn off. Not critical but aim for 35 mins. This isnt an all day solar charge. Its a short term peak at a higher voltage to allow an imbalance to be rapidly corrected. At the following logic settings:

a. WHEN PACK OVERALL VOLTAGE IS REACHED AND IS THEN HELD AT 28.4V CV (and now current is LOW.)
b. WHEN ALL CELLS ARE BALANCED at 3.550V.
c. WHEN THIS HAS BEEN THE CASE FOR 30 to 40 MINS, END THE CHARGE. (time not critical, AIM for this!)

Also.

You see this chart below?
Look at the first 2 lines. CHARGE and RESTING voltage. These 2 voltages are the exact same thing.
Why are they different? Because 2 things...
1. The pack has RESISTANCE and you are charging at some CURRENT. That means the voltage DURING charge is always a little higher than the resting voltage. Because if you charge at the resting voltage zero current will flow!
2. Because of the chemical changes happening during charge that are temporary.
The difference between charge and resting voltage depends on the charge current not state ofcharge. And this difference happens across the charge from start to end. The voltage/fuel tank capacity only works at a fixed temperature and an unloaded "resting" battery.
All of which means that you cant choose a charge voltage based on resting voltages.

This is why EVE suggest 3.65V at a typical .5 to 1C charge rate. We charge at a lower rate, but still need to see an imbalance that is big enough to balance! So while a resting battery may be 3.4V you cannot balance at this voltage at all.