TROUBLESHOOT BAD BATTERY PACK

Hallo, I have been using for sometime a LIPO battery pack composed by 12 bricks (6s x 5000ma each) in configuration 4Px3S with a Sabvoton controller (72v60A). I have balance connector of paralleled bricks always connected in parallel. I do not own a bms but I always parallel charge with a balance charger the pack at 24v (1Sx12P).
I noticed that lately the cells at the position 6 (between wire 6-7 of the balance lead) were out of balance quite a bit.
I took apart the pack and at least 3 of the cells at position 6 of the pack were bad. I would like to ask if anyone of you is using the same configuration and had similar problem. Can it be that the cells gone bad are a result of bad configuration or it can be just the fact of one bad cell ruining the others paralled with it?

Hope it's clear

Thanks x help

Yes, if one is self discharging, it could suck the others down with it, and ultimately would prob have gotten the forth one as well if you hadn't caught it.

I see Voltron. Would it be a way to prevent this? I mean a way to realize that one cell is going bad before it actually compromise also the others?The thing is that when the single cell goes bad in a paralleled wiring you might not notice that because voltage is kept high from the good cells on the same line.

Yes, that is a drawback of the parallel first setup.. hard to catch a bad cell in a group. In large car sized packs that's where cell level fusing gets important.

But doing series first has drawbacks too. If you do a bunch of series ones that are then paralleled up just at the ends, not at cell level, you get a big drop in capacity from one cell going bad. But it's easier to find the bad cell.

The proper way to assemble 12 lipo bricks 3s4p like yours, is to make 4 series of 3 then parallel them. Then you can monitor every cell individually, and a weak cell can not hide behind others.

However, you may not want to charge the strings together anymore, because if you connect the balance leads and cells in one string are very different from those in another, it may cuase high current thru the leads trying to suddenly equalize that.

Another problem, if you *don't* parallel the balance leads during charging, is that any cells taht are low mean the other cells will be high by that amount, spread across them all by ratios of their internal resistances and other properties.

It isn't quite this simple, but an extreme example of a bad cell in such a string-built pack: if you have a 6s pack, all strings still paralleled at main +/- for charging, but not balance connectors not connected, then say for the one with the bad cell, the lowest cell is down at 2.5v and the rest at 3.7v. That's up to 1.2v to be distributed among the cells. If every cell gets .2v of that, it means the 5 cells will end up charging to 4.4v each (instead of 4.2v). The lowest cell won't fully charge, only reaching 3.2v. The 4.4v might not cause a problem this time around. (but it could).

Next time you charge it, let's say the difference in voltage is now 2v (because the disparity will get worse each time). The high cells will overcharge even more, while the low cell will undercharge even worse, while still resulting in the same total voltage across teh whole string. But now the overcharge could be enough to set the RC LiPo off, and there could be a fire. At best, it might just puff up all the previously-good cells in that string.

Let's say taht somehow you never manage to overcharge the good cells, but the other cell gets discharged down to empty well before the others do, and you keep riding because you don't know that's happened. If current continues to flow thru that now-empty cell, then it will begin to reverse-charge the cell. At best, it will puff it up and permanently damage it. At worst, it could burst into flame, lighting off the rest of the pack.


So...if you wire up the pack so balance wires are never connected between packs, and only main + / - is connected, then you must charge every pack separately, using it's balance connector to the charger, to rebalance them every time, to ensure the above scenario can't happen.



At least if you have balance wires always connected, then while you may drain down other paralleled cells to the bad one, you will probably never drain them down so far that they reach a dangerous point, as long as you monitor them in some way, even if it's just watching how much voltage sag you have under load as you get closer to empty, and stopping if it's really bad, to check the individual balance voltages.


If you simply charge each pack individually (and discharge them all in complete parallel), then you can see when you have bad cells because some of them are going to start out a bit low, and shoot up in voltage during the high-current phase of charging, becuase they have higher internal resistance (and lower capacity) than the rest. (similarly, if you run a discharge test on the pack, those same cells will drop in voltage faster than others under load). Then you know that pack is going to cause a problem, if not now, eventually. You can either stop using it or keep an eye on it, which ever you like.

Yep. When you monitor at the individual cell level (let’s say bulk charging because that is the best practical way) you always know which are your weakest cells. They get lower in discharge, and charge faster than the others. They also get hotter, a sign that they had lost not only capacity but also some of their C rate. Then you are following the wear of your cells, charge after charge, and know when it is time to take them off.

I am using many packs, most of them are about the same age and wear. When some cells need to be repaced, I disassemble one pack to have a supply of good equal cells to fix the other ones. Alternatively, I bypass or take off some of the cells to make a 23s out of a 24s, so I can continue using this pack for a while.

My usage limit is defined by heat, because I charge fast. When a pack start to build heat charging, I prefer discarding it than charge slower. Of all the cells in the pack, many can usually be testing good enough to replace some bad cells in another pack.

Individual cell monitoring is the only way to know the actual state of all of them. Parallel 2 cells, and you don’t know them anymore because they hide each other. One of them can go bad enough to start a fire, without you noticing any major difference in monitoring.