Ep. 87 - 2.0v; Did I Kill My Battery? 

An easier way to use the precharge resistor would be to touch it to both sides of the switch when it is in the off position, rather than disconnecting wires. I have a momentary switch installed in my system with an incandescent light bulb in a lamp holder as the resistor. When you press the pre-charge button the bulb will light up, then dim when it is charged. While holding the button, rotate the switch to the on position.

Great quick changes and the rest of the video was pretty good as well.😊

Just ran across your channel Nice video. “Sparky” 😅 I watch Andy he’s over my head most of the time 😢. Great guy he took time to answer my question from a comment I had asked. ❤

I have been enjoyed, so thank you for delivering.

Merry Christmas and Happy New Year. I truly enjoy your videos.

Great video again. Its a blessing in disguise that one cell showed himself of being less than the others. Could save you a lot of trouble on the boat. If i where you i replaced it. Good luck building!

You go girl! Your up to date on all your knowledge and speak clearly and concisely and I must say you really know what your doing! You should be very proud of your work! Its easy to get discouraged by mean trolls but I truly believe that as long as you keep up a reliable video publishing schedule then you have the potential to become as big as you want! I just wanted to give you some words of encouragement speak to you as a friend I don't know what possessed me to make this comment cuz I never comment on videos but the universe spoke to me the universe spoke to me and told me to tell you that you rock. Have a nice day.

One thing that made me wince was using the all metal socket wrench. The thought of it getting dropped and shorting across battery terminals frightens me. The short cct current will be huge. A layer of heatshrink tube should give a good degree of protection.

Very good video. I too use a JK BMS, similar 280AH cells. My settings - 3.615 Over Voltage Protection, 2.600v Under Voltage Protection. Active balancing start voltage - 3.45v Great results. The 280AH LiFePO4 cells are more forgiving and tough than some give them credit. Thanks for sharing.

Nice to see you back at the batteries! What an awesome learning opportunity. Hope your holidays went well - and pet the kitties from me :) 10K subs coming up! Probably sooner than 1K on my channel.

You are one smart lady!

Definitely with you on the diy bench power supply 😊

Hi Maddy . I have just started to follow you. You are amazing ,the way you work through things. Best of luck for 2024

Always something to do, happy new year.

Merry Christmas have a blessed one and a great new year

Andy (Bombenfest) sure told you how to safe the battery.

They should make a 3 position rotary switch where the first position is the pre-charge resistor and second position is fully engaged. Those sparks are scary.

Thanks!

I love that cute little charger, just perfect for occasional use. Did you make a video for it? I also love that, living on a boat largely, that you face many of the equipment and infrastructure limitations that I do

Great video! If it ever happens again, I'd like to see the number of amps flowing into the cell that bottomed out when it gets reconnected. Even though you matched voltage beforehand, the resistance might possibly be different, especially if the other cells were topped off. They were all close-to-depleted in your case, so it seems like it was nothing to worry about, but even so, any capacity difference might create a significant current with one cell versus a large number of cells.

Merry Christmas!

Kudos to you for trying to figure what was going on. As you say voltage is really a difficult way with any battery to truly work out how much charge it has left, especially when not under load. A 48V battery has a might powerful kick which I've learnt with the fuse pre-charge procedure. I think next time I'm going to make sure I have a good contact maybe by using leads that are soldered to the fuse and using a switch?

So uou the battery lady right? Subbed for your knowledge. Lets see lesh gooo

Some notes on cell undervoltage recovery 1) in terms of damage what matters is both the magnitude of the undervoltage and the duration of time at which it sat at that undervoltage. Multiply these two togther to get an idea of the total damage. Ie a day down at 1v is about the same as a week down at 2v! 2) If you can be sure that a cell has NOT every been reverse biased (ie driven to a negative voltage by current being forced through it by other cells in series with it) then it is mostly likely "safe" to attempt to recharge. If a cell has been reversed biased DO NOT RECHARGE IT. This is because reverse biasing leads to conductive dendrites forming that can internally short the battery. This can happen at a significant period of time after your recharge it, so you could have a full cell that then internally shorts, which basically means you are getting a large fire (esp with cells of the capacity you are using!) 3) Dendrification that has already occured and lead to a short on a discharged cell, or a cell that was at such a low SoC when the dendrite formed and hence did not cause a fire, can be determined by attempting to recharge the cell at a very low power level. Effectively a dendrified cell looks to have a huge self discharge current, so if you say pput just 50mA into the cell, the voltage might climb, but as soon as you remove that 50mA current the voltage will fall again. Here the cell is not actually increasing in SoC you are simply pushing energy into the Internal resistance of the cell. If you attempt a higher current recharge then you are risking a fire 4) typically you end up with low charge mobility in cells that have been undervolted and this leads to a loss of capacity. However you tend to find that a fair amount of that capacity can be recovered if the cell is carefully charge cycled whilst not at a low temperature. ie carry out say 5 or 10 full charge cycles at 0.1C at around 25degC and you may see most of the capacity return. 5) If your cells are clamped (ie have a mechanical pressure load applied to them, do not remove the clamp load whilst trying to recharge the cell! A cell that has been clamped and then the clamp load is removed is likely to suffer a large loss of capacity, and reclamping that cell is very likely to cause an internal short (fire, again!) or lead to a significant increase in internal resistance Hope that is helpful info 🙂

Interesting channel Miss. I`m not quite sure what you do, but i`m in. I like tech and I work as an electrical test technician in a circuit board assembly plant. Even as a kid I would tear apart every piece of electronics in the house, and later on in my teens I discovered ham radio and got my license. Even today, I have to pick up every flat screen tv, radio, or computer left at the dumpster. LOL. Good luck with with what you do.

I have watched you entire battery build as I was planning mine. I enjoyed every minute! I also watched all of Andy's videos. I have the same 280ah cells and 200 amp JK BMS's. After filtering Andy's results I went with starting balance @ 3.45v per cell. Andy found that the 3.4v was to low for effective balancing, to low on the curve. Then set the target and absorption voltage more like 56.2v (3.51v) for a couple hours. But it sounds like you have the cell OPV @ 3.5 v instead of more like 3.625v or 3.6v. My Victron charge controllers are set to charge to 56.2 then go into absorption. My BMS starts balancing @ 3.45v and they stay perfect.

Andy is Really good at the electronic and Battery interface. ❤

HOWdy T-D-M, ... Back to BATTERY content 🙂 ... Thanks for Posting ... Happy HOLIDAZE ... COOP ... the WiSeNhEiMeR from Richmond, INDIANA ...

Be sure you use isolation (Like thin EVA foam or epoxy board (0.5-1mm) between your series cells as the bare metal under that blue covering is positive, especially on a mobile install. I just picked up 4 LF304 EVE cells to do my first build w/JKBMS. Figured it was a good way to start, eventually want a 16 cell pack in the garage.

the jkbms has a shutdown voltage where the bms will switch off itself to prevent paracitic drain of the bms pulling the cells below a certain voltage (power off Vol)

I really like your channel . It looks like you are going to be off grid in your home. I remember some of the boat build but this is the first I've saw of this in your house

A hack I found for using a precharge resistor in an adhoc way like that, is to solder short low current wires (about 6 inches / 150mm) to each resistor terminal with croc-clips on the other ends (I had a couple of suitable wires lying around). Then you can clip the wires to the busbar/terminal etc, and body of the lug you are attaching, and leave the resistor connected until you are sure the lug and terminal are making contact directly, before removing the croc-clips and tightening the terminals. I haven't had a spark since (as long as I don't forget the resistor).

Nice spark lecture again 😂. Great video for all of us with some bad cells. Note for top balancing and equilizing the charge. The voltage curve is very flat in the middle charge region. That's the reason why your cells were not gain same charge in few days. For so small difference in voltage you need ages to balance energy as well. Voltage will be equal but charge doesn't. Best way is to charge over 3.45V and then top balance with 1S 16P configuration.

If you'd left the mains power on, the Quattro charges its own capacitors. I also connected an analogue volt meter across the cables , so I can see exactly what voltage the Quattro is at. If you have the Cerbo connected up you can go to the VRM, go to device list, and at the bottom of the list is Remote VE Configure, you can download the settings file, edit it on the PC using VE Configure then upload it again. Think I've used my Mk3 cable once.

Hi Madi, great video as usual and nice to see the battery video's back.. You have been mentioned on Gridbusters (Sarah) YT channel along side Andy at OGG.. Sarah it a Brit living in France building her own Off grid setup.

Great video Maddy. I have always wondered what would happen if we go below 2.0 and I can wait to see the results. I am aligned with your frustration with the MK3 adapter.. I found you can program without the MK3 through your cerbo gx and remote VE configure. I am not sure if it is set up for OS yet.

Keplar, demonstrating eliptical orbits with his belly :)

The little coloured bar graphs on the charge screen indicate the cell is being balanced against the rest of the cells. To do this the charger discharges the higher voltage cells through an internal resistance which is where the heat was coming from.

you can configure a Quatro using a pc and the Cerbo using VRM as a gateway. I never plug in my MK anymore. I only keep it to ensure I would have access if the internet is down.

When charging or discharging a paralell battery bank with lots of connections, it is best to put the positive on one end of the bank and negative on the other.

Always unplug the balance leads on the bms if you are leaving them for a long time. Cheers

I recently had the exact same thing happen with a 4S-2P, 240AH pack. Two of the 8 cells were reporting ~2.2 volts while the rest were still at 3.3 volts. I charged the two cells back up to 3.45, charged the other cells up to 3.45, put them back into a pack and they have been cycling normally. There was a loose connection which somehow led to the initial problem.

To make sure you can't connect the battery without doing a pre-charge you can make a 3 position battery isolation switch using a 2 pole 3 position rotary switch. Don't connect contacts in position 1 to serve as "off". Connect the precharge resistor between the position 2 and 3 contacts and then connect the load to the position 3 contacts. This way there is no way to connect the battery without going through the 2nd position and using the pre charge resistor that's built into the circuit.

I'm sure it's been said (my first time watching your videos, hello!)... but a good option for the precharge would be a 2-bank power switch, the "first" bank position would be wired via the resistor and a jumper to the main feed from the switch, the 2nd bank switch position would be the main feed wiring. This way you switch to position "1", the circuit precharges, then you switch to "2" and only lose the discharge in for the moment it takes to flip positions. Nothing to disconnect and reconnect and only a very short wire/jumper in parallel with the main power line just behind/next to the switch (you could double the ring terminals on position 2 for the resistor connection from the "near" side of the resistor, and only need a 3" or whatever jumper wire to extend/reach out to the "far" side of the resistor, keeping it all close and tidy and only one short chunk of cable (or use 2 jumpers to "square up" the resistor (a 3" or whatever section on each end of the resistor to each set of terminals). 2 Bank switches aren't expensive (and only a few bucks more than one-position) and same physical dimensions typically, so it's an easy upgrade.

"Sorry about the mad rush to the boat"... WERE THERE AND BACK IN LESS THAN A MINUTE!!! Don't apologize, explain your wizardry :0 :0 :0

Another good idea, would be to use Equalizers on your bank's. Then the situation of a single cell dropping to 2V would not have happened.

After you use the precharge resistor, turn OFF the POS disconnect switch then connect the cable and then turn the POS switch ON

VERY curious if the SUSPECT Battery comes totally / reliably BACK to FULL CAPACITY & RELIABILITY Thanks AGAIN COOP ...

I live in the semi truck 3 weeks at a time then home 3 or 4 days. I'm living in a holiday rambler camper on solar and 840 ah at 24 volts with 4000 watt split phase inverter. 3 victron charge controllers with 3560 watts of solar

I'd still not trust that cell to not be weak /degrade quicker over next few years. but shocked that 6 months of BMS parasitic drain was all it took to deplete the battery from full.

Battery will be fine, is my guess. I would put them back in series and increase the voltage slowly to give the BMS time to balance. Let them sit for a while once balanced and watch this low voltage cell closely over time if it shows any sign of a lower voltage again. Edit: you did most of that anyway...😄 so, wall done. A tip for pre-charging: hold the resistor across the terminals of your red main switch while it is off. 5s is more than enough and you just turn on the switch afterwards. Saves you taking off the cables at the bus bars. @chevrofreak said this already in the comments.

Just a comment on the MPPT converter shutdown part: I don't know exactly what microcontroller Victron is using in the SmartSolar and how the power supply works but I can make some well educated guesses. I expect that they use a microcontroller with Bluetooth capability builtin (Nording nRF, STM32WB, ESP32,....) basically all of them run on 3.3V (however it's a bit odd that their Ve.Direct output is running on 5V voltage levels). The internal voltage for those is probably coming from a "high voltage" capable buck/step down converter so it likely will require quite a large minimum voltage drop which explains why the MPPTs shut down about 1 volt above the minimum 3.3v required to keep the microcontroller with Bluetooth up. A interesting thing I found about the Quattro is that they still seem to use all trough hole parts. This does hint to the fact that they would likely use a rather old microcontroller, most 3.3v controllers especially the wireless capable ones mentioned above, don't even exist as trough hole versions. So unless they completely redesign the Quattro and change then assembly line for them they likely won't become wireless capable.

I'd be more concerned if the cell was 0-1v but anything above 1v is usually very recoverable and likely with very little to any capacity loss. I will say that 12 is probably your weakest cell in that pack and likely hit the knee an AH or 5 before the others as you discussed at the end. Either way, the cell is probably fine but if you have spares that test higher that one weaker cell would do well as a spare or in a less critical application.

so you've got your cat in the room where there is a chance a lithium ion might burst into flames ...

Saving money any time you can Is very smart. 😃

55.2 volts charge 3.45 v balance. This will keep your cells top balanced. I listen to Andy.

I've noticed that with the JK bms, balance occurs much quicker with a load than without. I'm not sure but this may explain the cell with 2 volts. For some reason, and I've noticed this too with my JK set up the bms will slip at balance of particular cells over time without a load. Perhaps this is a glich in the program on the bms. Not sure. Plus, when I used a JBD bms that went defective and created very unbalanced cells, and I connected the JK, it was really slow at balance at idle. However, when I added a load along with a charge source, bingo. It was very quick at bringing all cells up to balance level and shut off the balance feature. Daly connected to same set up was completely inadaquate. I've tested several other brands/models as well but the JK remains my favorite. I use the lower balance model (as opposed to the more rapid balance model). Also when connecting a load to the bank, where capacitors are present, be sure to disconnect all outputs and turn the device off (inverter). This reduces the arc-spark you experienced. Isolation of the capacitor(s) helps quite a bit. I couldn't see what your wiring scheme looked like due to the camera angle and resolution. As soon as you perfect your system it will be replaced by newer technology.

You have a cell with a high self discharge rate. That should be coverd by warranty. But you can still use this cell in a setup with active balancer...but only if you use it on a regular basis.

Test all the cells, any below a volt are probably done for. Next gang all remaining cells in parallel. Use a DC arc welder to jump start them till you get 3V. Once there disconnect the cells from each other and reinstall with the BMS and charge normally.

@11:00 agree

5:02 dejavu! :)

When one cell voltage drops considerably before the others do, it 95% means that the cell has a problem. This happened to me with the TAICO manufactured batteries. Countless hours to balance, top balance, playing with BMS values, charging, discharging would not help. In your case, you are lucky, the damage is just a few amp-hours. Replacing the bad apple is the best and quickest way to go. In my case so many bad cells in every battery, I will take the to road of sacrificing one to heal the other batteries.

I have a whole long thing to post, but I'll skip to the end... To replace a bad or out-of-balance cell one needs a replacement cell _at the same SoC._ As you've said several times, voltage does not equal state-of-charge - 3.2v could be almost full, or almost dead. Connecting a 3.2v 90% cell in parallel to a 3.2v 10% cell _WILL NOT BALANCE THEM._ When there is no difference in voltage, there will be no current flow. Thus slightly overcharging them to make sure they're full. The shortest path is to _fully charge_ the new cell, and independently _fully charge_ the rest of the pack. Let both settle before putting them together, otherwise the BMS will attempt to balance them and actually pull them out of balance. (in a perfect world, it should be a small imbalance.) For a 16S, I'd bridge out the bad cell rearranging the balance leads to make a 15S. (or drag out a bunch of chargers, but I hope people are less insane than I am. I have two iCharger DUO's, and various EV Peak's around. I prefer the less prone to explosion iCharger -- the ev peak's _do not_ have any pre-charge logic in them... if you forget about that, the output caps _cease to exist._ Thus the Duff Beer Racing Patented Precharge System(tm).) Unless you don't have a setup to load cycle a single cell, I wouldn't mess with the whole pack. I don't remember the discharge rating for that little iCharger. The DUO (4010) I use can pass through 30A. (that'll melt my banana jacks) I use an external automotive 12v fan + heater module as the load. That setup can handle 300W, but obviously not from a single cell. -- A few things.. (1) in the off season, *physically* disconnect the BMS - pull the balance connector. no BMS has a true "off". and as you've seen, they can go crazy. (2) consult the specs for those exact cells. most simple aluminium cells like those need ("require") external compression. (I've had many turn into balloons because they weren't properly contained. and yes, the data sheet - if you read chinese - says to contain them, and the exact pressure required. they were in an EV, so they were being used "at their ratings.") (3) again, look at the data, but I think 2.0 for a short period is "safe", but you'll want to get it charged soon. And I'd fully load test it again to make sure it's still good. If that was LCO/NMC/etc. (4v Li), then yes, 2.0 is toast, but LiFePO4 is a lower voltage, and much more tolerant of deep dips. (and LTO is even lower.) I've never been a fan of massive parallel charging. But if 15 out of 16 are nearly full, it's a reasonable compromise. If I ever have a reason/need to get every cell to the same SoC, I like to do it one cell at a time, logging the process so I can judge how closely every cell is to each other. One cell charging faster than the rest would mean the BMS would have to throttle the charging and bleed the excess off. *Most BMS are crap at balancing; you'd need a fairly expensive active balancer to get beyond about 100mA balance current.* An _active_ balancer pulls charge from one cell for another; passive balancing, which is what you have in the iCharger and I suspect the other one, too, is a resistor, so it's just gone as heat. The bar graph on the iCharger is the amount of balance current it's drawing. Junsi makes some quality kit, but I trust my external meter 1000x more than the chinesium in the iCharger. There's a lot effecting its calibration.

While missing a battery is frustrating, at least you know before you installed it on your boat. And you know that the other batteries are reliable 😊 ... Stress testing the suspicious battery is a very good idea, and it will be interesting to see the results in your next video 😊 Also, can you switch off your JK BMS completely? Mine doesn't have a switch and draws 10 mA all the time. I can't store my bicycle battery over a winter without recharging it 🙁 Also also, I think it is crazy that devices with big input capacitors don't have a circuit to charge their capacitors "slowly". I bought a "soft start" for my chargers, which works pretty well. Maybe a soft start for high power amplifiers would be good enough for a big charger like yours.

To everyone! Remember to clearly mark all cells that have been exposed to undervoltage and then "restored" as "potentially damaged". Otherwise you might forget which is which and a "restored" battery might end up at sea. Even if they charge up normally they can be damaged internally increasing the risk of thermal runaway. A battery fire out at sea would be total disaster.

I used a 4-Position Marine Battery Selector, and wired Position 1 to my pre-charge resister as I turn the switch it passes Position 1 first and charges the capacitors, and then to Position 1+2 connects the inverter. No Spark ever.

You can chage settings with dip switches in the quattro

please cycle the battery atleast twice if it reaches a very low voltage if u were to salvage any and have it rest at full charge for atleast 1-2weeks to see if the internal resistance and the charge value doesnt change

good info. minor edit error double take at 5'02"

Maybe my Multiplus II is different but I can set the input current limit via the CERBO, no need for the MK3 cable.

15:40 I agree with you and at the same time dont. If you charge over one battery the charge will be a little uneven - it will still flow over to the other batteries but the "correct" way is to attach the charging points across the entire battery, "all the cells". So in opposite corners really. I think it may be more of an issue with REALLY big battery banks where there is a lot of distance between where you charge and the last cell in the battery. My tiny bank only has 4 batteries in it so i charge it the same way you do, over one cell only and then letting the parallel connections even it out.

Short idea, isn`t there a 3 step switch which first step could be on resistor and second step on battery? Cheers Marc

Mady the first thing I would have done on the suspect cell is a full charge as discharge test (only on the one cell), at least a couple of times, and then, assuming the cell came up to spec, put in a battery pack config, and charge and balance. I really really don't mean any disrespect (cause I have huge respect for you Mady) I'm just not understanding all the faffing around with making up the battery packs and charging etc, and you still can't be sure of that single cell, and you won't really know fully until you do test it singularly IMHO (food for thought, I'm not and expert!).

Replace the bad cell and replace the BMS from that battery. The cell has probably been damaged and the BMS probably caused it.

If you parallel charged them with the charg😊er connected at one end only, the batteries at 😅thenother end will not be charged as much as the cells at the charger end. You should have made a diagonal connection to the cells ( Don't know if you did this however, but it would explain why they were not equally charged).

One thing that needs caution is to remember that when you connect that many large cells up in parallel, without a BMS there to protect you from shorts, the short-circuit current can be 1000-2000A PER CELL. In otherwords, 16000A to 32000A with 16 cells, dumping 48kW to 96kW into the short. It is a huge risk to do that. At a minimum, if you ever do that again, please keep one side taped off at all times and even consider breaking up the parallel set with some inline fuses, or top-balancing in smaller groups. Safer methods: Leaving the series pack intact, holding at 3.55V/cell x 16 = 56.8V, and just letting the BMS get everything balanced, even if it takes a long time. And/or connecting up an active balancer. And/or one can apply some love to individual cells to speed things along while the balancers also do their magic at the same time. Hopefully your pack can pull nearly full capacity in your next test. It is almost certain that the low-voltage cut-off will be on that particular cell. Once it does cut-off, and if it is that particular cell that hit the LV, try putting an amp-hour into the cell (which you can do without taking the pack apart, with some care to ensure the pack is isolated before connecting up the charger)... and then continue the whole-pack discharge test and see how much more you can pull out of the whole pack. Repeat until some other cell hits LV cut-off. Keep track of the amp-hours of difference between that cell and the rest of the pack. If the difference is minor then that cell was not seriously impacted and you can keep using it. If the difference is significant, replace the cell with a new one. Of course, if you replace it, now that cell will be the high cell during discharge, which should be expected and "normal". Oh, put a physical label on the bad cell or write on the side with a sharpie if you haven't already to identify it for all time. And if replacing with a new cell, put a physical label on the new cell as well. Andy's channel is awesome. He's one of the most knowledgeable individuals on LiFePO4 cells on social media now.

I may have asked before, are the furbabies joining you on the boat in the future?

Just charge it up right where it's at. I had had lithium cells drop almost to 0 volts with only little damage. Recharged it as it was. Still in the pack. .you should be ok.

2:46 No, the voltage has to go to 0 volts before polarity reversal becomes a possibility. Not something that would happen without putting a load on the battery or letting the battery sit for an extended pperiod of time, one or two + years. Your BMS will prevent you from loading the battery so polarity reversal is not a genuine concern

Could be the BMS channel for that battery that is defective.

Sparks are just a part of working with high amperage, i shorted my torque wrench on my 48 volt battery and it literally welded all the metal parts together . little sparks like you had are nothing to be afraid of .

it is still able to recover , BUT it might not good as before, battery life might be shorter. start charging it with lower ampere and monitor their temp then leave it 1 or 2 day and monitor the voltage

You had me worried that you'd destroyed your entire battery. But it's just one cell, maybe. Now to watch the rest of the video to find out what happened.

Lifepo4 batteries are not significantly damaged by dropping to 2v. It's possible the cell is bad and that's why it dropped to 2v though.

3.45 is the best u only miss out on less than 2 percent capacity thate from off grid garage

25:50 The iChargerX8 tells you what it is limited by in the top left corner. It says I.CC, which means input constant current. In the bottom row we can see the iCharger is drawing 35A from the server psu, so clearly you have set the input to be 35A maximum. This means you could set a higher charge current (like 30A) without worrying about overloading the server PSU. The iCharger limits itself automatically. The charts next to the cell voltages represent the balancing current, not the voltage. The higher the chart, the more current the iCharger is pulling out of that particular cell through the balance lead.

Old sparky sticks again...lol

a friend bought new cells an one was 1v he charge it up an it work great

The key is the cell didn't reverse polarity . If keeping the bms ticking over is all the bank was doing the cell needs testing ! 1) It could be a faulty cell , self discharging BUT IT'S UNLIKELY . 2) The bms could be at fault . 3) The cell was way out of balance . 4) Or you just found the smallest AH cell in the bank ( that's the reason you don't discharge below 10 % ) .

It’s dangerous to handle a wrench on top of bare contacts like that 🙀

Just trickle charge the low battery until the voltage gets close and then use constant voltage to top it off

3.5v is too low for a high voltage cut off it is in the absorption range. I would do a parallel top balance with a charger at 3.625 to get them all properly balanced. Using 3.5v is probably why once cell is not taking a full charge and discharging more than the others. Once you have them all matched at 3.625 then you can lower the cell overvoltage at 3.6/3.615v which will keep them in the range. I think balancing ends up doing more harm than good once you have a properly top balanced pack and a decent low voltage cutoff (2.75v/2.8v) as they will only ever remove the same amount of power from each cell and only put back the same amount of power once in series. 1 cell effectively becomes the safety system. With balancing in the absorption range it throws off the pack individual cell balance due to individual cell spikes (taking power from that cell and chucking it in another) meaning that it removes all the good work you did with the top balance.

The LiFePO4 cell has a flat-line discharge curve. So, paralleling them and let them sit does not do anything. They'll just sit at their usual 3.2V. One might be almost empty, the other might be almost full. The only way to balance them is to charge them all individually until 3.65V until the charge current drops under 5% of their rated capacity. If yours are 100Ah, stop when it drops under 5A. So if you do 16 in parallel, your charger has to put out quite a bit more than 80A, so it will be able to stop when it drops below 80A. With such currents, you will need to connect the leads at opposite ends of the bus bars, so the current will spread out better over all cells. Even then, some cells might pull more than others. No cells are exactly the same. So, I don't recommend paralleling LiFePO4 cells for the purpose of balance-charging them. Your 2.0V cell is unlikely to be damaged. I found that damage might occur when they sit at 1.0V or lower. After finding it at a low voltage, to test for damage, first, charge with a voltage of just 3.00V and modest current until the charge current gets close to 0.0A. Disconnect and let it sit for a few hours. Measure the voltage. Should still be very close to 3.0V. Let it sit for a day and measure again. Should still be close to 3.0V. To be sure, follow-up with full charge, let it sit for a week and then do a capacity test discharge. Damaged cells develop an internal leak current, which will come to light with those tests. I don't know if such leak can develop into a full short but I'd rather not find out. I keep on using small LiFePO4 cells as alkaline, NiMH AAA, AA, CR123 replacements. When an empty LiFePO4 cell comes out of a device with less than 1V on it, I'll test it.

Active balancing is probably the best real world way to balance, is the internal resistance on cell 12 higher than the others?

Yeah, everything named Victron is extremely expensive.

You know how you said it doesn't matter where you put your charging leads on when you had all the cells in parallel? Test that assumption. I've had 12 cells in parallel charging, I placed the leads on the + and - of the first cell with the same assumption. After a bit of charging I noticed that the first cell was ahead of the last cell. And the difference was gradual in between. Placing one of the leads on the first cell and the other lead on the last cell overcomes this phenomenon.

u might find that funny... but JKBMS helped me find a faulty cell... removed that cell from the 16xcell pack and now running a 15x cell pack WITHOUT BMS at a MasterPower (Voltronic) 5kW inverter and it works JUST FINE :D imho either Chinese are mixing good with bad batteries or they try to save on expensive lithium component... it is probably a faulty battery... bought from a GERMAN company EVE cells and they refuse to replace the cell :(

You are the cuteist mermaid i ever met !! ❤❤❤❤❤❤❤❤

If your one cell dropped that low there is a reason why it happened. Not that you just charge it up and think everything is fine, it isnt. That cell is prone to fail again. I would have scrapped it immediately. Cells don't just randomly do things they do because they are bad. Replacing one cell in an old pack is problematic. New cell will always go out of balance and if you have very low balance current it takes forever to keep them in balance during charging big cells.

Seems odd that a unit like the Victron doesn't have inrush protection on power on. I'd complain big time. Then make yourself an inrush protection circuit and put it in a box so you can leave it connected permanently.