Author Topic: Battery balancing - big cells  (Read 7721 times)

0 Members and 2 Guests are viewing this topic.

Offline mtdoc

  • Super Contributor
  • ***
  • Posts: 3575
  • Country: us
Re: Battery balancing - big cells
« Reply #50 on: April 24, 2018, 05:25:47 am »
I'm sorry. I was wrong. I should have just walked away after my first response to this link. Please accept my apology. I was confusing balancing (initial balancing) and active balancing. And confusing posters.
Accepted and kudos for acknowledging the error.

Quote
As he says, you have to avoid over charging, as well as over discharging. And completely removing any cell monitoring from the bottom would just shift that burden to the top end, would it not?

Perhaps it's a simple matter of LiFePO4 are harder to damage by overcharging and easier to damage by overdischarging than Li ion. Or maybe it's harder to detect the lower threshold by voltage monitoring. (Pretty stupid easy with Li ion cells).  But this is just speculation. Perhaps there is someone out there that can explain why the hell they are doing it this way. If not, just say it works better and I don't know why.

Have a look at the second link I posted in my first post in this thread. If you read through the entire thread in that link I think it will make more sense. There are a couple of engineer posters there who do a better job explaining it than I likely could. Again, it’s no panacea but it has been shown to be a practical method that works for these large prismatic LiFePO4 cells.
« Last Edit: April 24, 2018, 05:29:15 am by mtdoc »
 

Offline KL27x

  • Super Contributor
  • ***
  • Posts: 4099
  • Country: us
Re: Battery balancing - big cells
« Reply #51 on: April 24, 2018, 07:04:17 am »
Thanks. I was hit in the head pretty hard as a kid. That's my latest excuse for being an occasional ass.

On the forum linked, it seems when referring to top balancing, they are actually talking about active BMS system (specifically a "Vampire" board) but bottom balancing they are referring to initial balancing just once with no active measures. Same thing on the first web page, which was part of my initial confusion where he was talking about active balancing on top, then going to claim bottom balancing was better. (I thought he was talking about active bottom balancing). I was assuming there was cell voltage monitoring in both cases.

The things that stand out to me which sound accurate and appropriate are:

Quote
Most who use LFP batteries in EV's and solar use either the Motor Controller LVD in EV's, and the LVD built into the Inverter Both those use the total pack voltage to protect the battery.
and
Quote
A voltage that allows you to go close to full discharge under heavy load will not trip until it is too late when discharging under very light load

These seem to be pretty obvious issues. Of course you can't rely on a "Vampire board" to save you from overdischarging your battery if it has nothing to do with this. And relying on the minimum voltage which your vehicle's motor controller will accept before it stops functioning? It seems perfectly reasonable that a manufacturer of an EEV will design the motor controller to accept a minimum voltage which MIGHT still be safe for the S's and chemistry of the expected battery. It's not up to the motor controller to limit your mileage, arbitrarily. Unless it talks to a battery management system on the battery, of course it doesn't do this job.

If bottom balancing and simply running until the motor controller shuts down work for you, then great. I don't understand why you want to run out of battery in an EV, though. Seems like you'd be stranded. I would rather know when the battery is dangerously low and have the option to kill a cell if I really wanted to. :)

Then there is a lot of explanation that is less realistic.

Take a complaint that the Vampire board allows overcharging of cells.
Quote
First and the lessor of the issue they do not prevent you from over charging the battery. On my 100 AH batteries in my EV, I have a 50 amp charger. When that first Vampire Board turns on it shunts only 1 amp around the battery, while the remaining 49 amps continues to flow through the fully charged cell. Second, third cell and so one turns on until the last battery finally reaches 100% and terminates the charge. So in my 16S pack I have 15 overcharged cells cells, and one happy cell.
This is specific to the Vampire board and the specific battery charger and battery being charged, firstly. If 1 amp shunt isn't enough to prevent overcharge, that is potentially bad design of the Vampire board or it is maybe user error, using the board for something out of its specification. But he is probably not aware that his battery/charger is going to hit the constant voltage phase well before the first cell reaches max voltage, and the current will have probably dropped well below 50A by that time. But maybe 1 A isn't enough. This is a fault of the overall system and/or user error.

There is also the repeated notion that a cell dies because of reverse polarity, when good cells push current through a dead cell. This is news to me. With regular li ion, the cell doesn't need any help to be killed by overdischarge. You have to actively stop using them before this happens whether they are all balanced or not. So I know LiFePO4 are more forgiving, but I would personally be doing something a little more, here, that just balancing them on the bottom and running them dry.

Quote
Only the weakest cell ever reaches 3.5 volts or roughly 90% SOC. No cell will ever see 100% SOC. All others will be slightly lower.
This explains what happens on the top end. So after bottom balancing, you have to cut charging when the weakest cell reaches full charge (or your target goal, 90% or whatnot). Perhaps this is easier to implement than monitoring voltage at the bottom for many applications. In charging a batt in your home under controlled conditions that can't leave you stranded. I can see the advantage, and you could even simply set a charger to a conservative voltage and then physically do a voltage check on your cells (assuming you have access to all the terminals).

Quote
But all cells will have the same capacity in a Bottom Balanced system
Same amp hours, slightly lower energy capacity, though.

But I really like the idea of bottom balancing, esp if this is enough to do the job on LiFePO4 batts. Esp because the weakest cell ironically will take the longest to charge when actively top balancing, at least with Li ion. IME, it's not ever a matter of one cell having slightly less capacity than the others because it didn't get as much of the good stuff in there... it has always been found, IME, that one cell is slightly defective when there's a significant difference in capacity. By bottom initial/pack-balancing, you will get the weakest cell to finish charging first or close to first, and this is very simple and elegant and the balance should remain stable for a very long time. But I think eventually, the stronger cells are going to generally charge up a little more and discharge ever so slightly less, and you end up out of balance. But starting from bottom balanced might essentially give you a head start and a longer time between manual pack balancing (compared to top pack-only balancing, of course). I might have to try it, but it is a lot of work to do this.... I have a circuit for draining cells, somewhere.










« Last Edit: April 24, 2018, 09:27:12 pm by KL27x »
 

Offline MiyukiTopic starter

  • Frequent Contributor
  • **
  • Posts: 896
  • Country: cz
    • Me on youtube
Re: Battery balancing - big cells
« Reply #52 on: April 24, 2018, 07:54:35 am »
And if you have a better concrete answer for the OP as to were to find a high quality, affordable BMS system for very large cells, please share.  The OP mentioned 58Ah cells which are almost certainly LiFePO4 prismatic cells.
I was wrong they are not single cell but two parallel rated at 33Ah flat LiMn2O4
But it is not that big difference
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26682
  • Country: nl
    • NCT Developments
Re: Battery balancing - big cells
« Reply #53 on: April 24, 2018, 12:03:10 pm »
You can't read the charge from the voltage of a Li-ion battery. There is a huge section of the charge/discharge curve which is flat. This is the reason why you need a coulomb counter calibrated for the specific chemistry to determine the charge in a battery pack. The only thing the voltage indicates is whether you can charge or discharge a Li-ion battery any further.
Yes, absolutely right. I believe this is part of the reason why traditional “ top balancing” of series strings of large LiFePO4 so often fails.
Without knowing the BMS architecture that is hard to determine. In the links you provide it seems people have destroyed their batteries due to a BMS which doesn't care about the individual cell voltages. That is a big red herring (ofcourse). I'm also not sure whether using very large cells is a good idea to begin with. The battery packs  I've been around (mostly for small motorcycles and small vehicles) all use 18650 cells connected in parallel/series. Tesla does the same and apparantly for a good reason.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline paulca

  • Super Contributor
  • ***
  • Posts: 3989
  • Country: gb
Re: Battery balancing - big cells
« Reply #54 on: April 24, 2018, 01:18:28 pm »
What some people may not be aware of here is that for systems that have high S do not usually charge cell individually, they charge the stack in series. When a cell in the stack reaches man V, charging MUST be stopped. So yes, high S packs must be balanced during charging. Or be shut down when a cell reaches max V and the stack remians unbalanced, and gets more unbalanced over time.
Agreed. And being involved in battery packs myself I think that balancing a fully charged cell is better than a nearly discharged cell. One of the problems is that it is much more unlikely a pack gets discharged enough to do the balancing at all. Also the voltage of a cell (usually a group of cells in parallel) may drop below the shutdown voltage which means the BMS has to shut down and cannot complete the balancing operation. On the other hand most packs will be fully charged when they are charged so basically the balancing will happen for each charge/discharge cycle.
Ofcourse the BMS has to shut down charging and discharging when the cell voltage limits are exceeded (usually these are not the absolute minimum and maximum voltage for the cells but have some margin to prolong battery life).

The chargers I have used balance charge at different stages within a set of conditions.  Balancing involves resistive loads and reducing charge current.

If the cells are unbalanced, it will balance them to bring them into a tolerance before it will open up the charge current.  As an example I put a LiPo on the other night and got 3.1V 2.9V 3.3V cells.  Even though I selected 2Amp charge rate it only charged at 0.2A will it balanced the cells to within about 100mV of each other, then it ramped the current up to 2Amp.

My current once supports unbalanced charging which omitts this step and leave it till the end.

If also has "Fast charge" which ignores cell balanacing completely and just charges at whatever current you ask for until a single cell hits 4.20V when it ends the charge.

Finally there is storage charge which balances the cells at 3.7V per cell, either charging or discharging to achieve that.

Lithium batteries are best stored at about 60% charge in the fridge.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf