Battery cutting out

[PhilExplore]

Hi

My battery cut out the other day, unplugged and would go for a while until not very much power would cut it out again. Battery was charged before the ride and showed full voltage on the display after cutting out multiple times.

One cell showed very little voltage, everything was connected, all the other cells showed normal.

Put about 20 hours on the 1P lithium pouch cells

I reset the one cell and it did not charge to the normal voltage compared to the other cells.

I removed the cell and replaced it with a spare . I only made a 16 cell battery so I had 4 spare.

What could be the reason for the one cell going bad after 4 months of mild use? Of course it was one right in the middle of the pack

3 other cells are showing 1 to 2 V difference to the rest on full charge.
What Can I do to balance them ? Or is that a sign that the cells are not of the best quality. - Yes Alibaba .

I am using a Daly BMS that should balance them automatically.

Would using normal 21700 cells in parallel provide much better reliability than 1 P since the multiple cells in parallel would auto balance themselves in each group?
( Using a good known name brand this time )

Thanks for any input

 

[amberwolf]

That kind of voltage variation is huge, and indicates the cells are either junk (very common) or incapable of handling the loads placed on them. It's likely that at least some are defective, or damaged.

A good place to get decent cells is used EV pack modules, from the various battery recycling websites like batteryhookup, etc. If they came from a car's main battery pack, they are probably well-tested during pack manufacture, as they don't want to have to do warranty repairs on these things. Often every cell in them will still work perfectly, or there may be one that's not quite as good as the rest, etc., but if the modules have not been messed with by the reseller then a simple voltage check of them that shows all exactly the same voltage (within a hundredth of a volt or less) indicates they're almost certainly all still ok.

The ancient >decade old EIG cells I have were from EV usage before I got them back then, and while they've degraded in capabilities since then, they all degraded *equally*, which is not something you're likely to see with any random cell vendor on the internet.


Another issue with cell vendors, even if they claim to be selling a "name brand", is many of them either outright lie about what their cells can do (or even what they *are*, re-marking them however they feel like), or have no idea what they are so they make up numbers for their ads, etc. Others sell recycled garbage cells as if they were new, or factory-rejects that can't do what the specs say they should (if they work at all). So you can't ever be guaranteed that what you get is real, or even if it's real that it can do what it says it can.

Pictures mean nothing, because you'll find the same exact pictures on many sites selling the same thing, but if you order one from each of those sites, I can virtually guarantee (regardless of product type) that not all of them will be identical to each other, even visually (much less functionally), and they won't all (or even possibly *any*) be the actual thing in the pictures.

It's also highly unlikely that (unlike an EV pack) you'll ever get matched cells from a cell vendor. That means that your pack will *never* work as well as it would if they were all matched, because each series cell or group will react differently to the charge and discharge, and so rise or sag in voltage differently, and have different capacities, with the lowest capacity and highest resistance limiting your pack to those capabilities.


As for cells "auto balancing" in parallel, tha'ts true of *any* cells you do this with, not just 21700. The ones you have now would do the same thing...but it also means that defective cells would pull down good ones, and the pack wouldn't work any better than it is now.

Another issue with paralleled cell groups is that any interconnect issues or failures can leave you with some groups that don't have all the parallel cells connected, or not sharing current like they should under load, and put you in the same situation as before.


So...the best thing you can do is get cells that can be used in 1P, so you don't have parallel interconnect issues, that are significantly more capable than you need them to be***, and that are matched, identical in resistance, capacity, etc. Not *similar*, but *identical*.


***meaning, they have at least 1.5x the capacity you need, preferably 2x, at least 1.5x - 2x the current-delivery capability *continuous* than you need *max*; this allows both for aging degradation, and extra capability for future needs, excessively bad riding conditions, detours, etc, and lets you not fully charge or discharge the cells and still get the range you need, which will lengthen the lifespan of the cells.

 

[PhilExplore]

Ok thanks, after replacing the one bad cell, the 16S 1P voltage variation is 0.2V on 3 of the cells, the others are the same voltage. I keep on monitoring individual voltage on a ride basis.
Do you know by any chance what voltage difference does a bms cuts out at, or does it measure individual cell current / voltage and decide to cut power if anything exceeds the internal programming of the particular bms.? ( Daly - I am restricting the amp load to 60 Amp through the controller ).
Thanks again

 

[amberwolf]

0.2v is a huge variation, indicating a great difference in capabiility of the cells involved, at least some are probably defective, or damaged.

Some of the "smart" bluetooth BMS people have posted screenshots for show a default cell-delta-voltage-limit of 0.1v, which is large; when it's been reached the BMS shuts off to protect against the possibility of fire from damaged or defective cells, preventing recharge or discharge of the pack for safety reasons.

You'd have to check your specific BMS data or settings (if it has a setup app or program), or manual for it from the manufacturer or the website it was purchased from to find out what it's specific operation and limits are. If you don't have access to any of that, you would have to experiment to determine exactly where the BMS turns on or off the charge or discharge ports for any particular cell or pack condition.

Every BMS that is connected to individual cell groups monitors each of those groups and has limits for at minimum: cell HVC, cell LVC, and it may also have many other possible limits, such as cell voltage difference, balancing on or off, soft LVC, whole-pack current limits, etc. It's behaviors for each of these limits would be to turn the charge or discharge port on or off.

Some of the limits would cause it to keep the port(s) off until it is reset, either via a reset button (most don't have this) or by connecting the charger to it and starting the charge process (or finishing it).

Some of the limits may permanently turn off the ports, because the cells have reached limits indicating (like yours) that they are damaged or defective and are not safe to recharge or discharge. Some BMS have either hardware or software to "brick" the BMS so that it cannot be reactivated, so that the user does not cause a fire by recharging defective or damaged cells. Most don't, which is unfortunate because I'm sure quite a few fires have been caused by packs that were recharged anyway, either immediately or at some later date due to the damage already done to the cells, or the defects already in them.