Mystery battery build

[PhilExplore]

Hi

I built a 60V battery out of Samsung 50S cells, double checked spot welder - tested. 6 cells in parallel, amp draw could not be more that 60 amp as that was the controller limit, with 6 cells in parallel, Samsung 50S could handle 125amp easily.

Battery has JK bms with balancing and a lcd screen, these are new cells and my problem started on my 3rd ride.

I made sure to leave the battery to self balance before charging, and again after charging, to ensure all groups were equally charged. ( to 96%)

Basically on the ride I was checking the screen every 30 mins to change the battery with a smaller spare one to get home when it reached 3.15 V overall.

I should have stopped as one bank was showing 3.06V while the rest was at 3.12V, then the battery died, I have good gearing on my bike so I could pedal home - slowly.

After dissection, 5 cells out of 6 showed mV or reverse voltage. I am guessing these cannot be fixed?

This is the 2nd time this happened, I got new cells to replace those that went mad, again in a different group. I did notice on my 2nd ride that the voltage in another group , not the one that had the issue , got lower so I let it self balance before charging.

Has anyone have a clue why this is happening, I think I am just going to buy ready made batteries from now on.

I redid the spot welder on a few cells including the one in the picture to the right. Meter shows the lowest bank next to the cells that have failed in blue, marked a good cell by mistake.

Thanks

 

[amberwolf]

Cells that drop below 2v shouldn't be reused. If they actually reversed it's potentially much worse--damage can have occured that could cause a fire at any point in the future, and you won't know about the damage until the fire happens.

If this is the same battery as this one

Battery cutting out

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...

endless-sphere.com

where there were cells wiht 0.2v difference between them, it's likely that the cells you have are wildly mismatched in characteristics (capacity, internal resistance, etc), which means that they won't charge or discharge equally, and you may have to rebalance them on every charge cycle just to be able to fully charge the battery.

Balancing doesnt' fix anything, it just makes them all the same voltage at full charge, but it means that at least the cells all have the max charge that they are capable of holding, for their individual characteristics.

Note that buying premade packs is not likely to be much different, in that unless they are from a seller that spends the money to test and sort all the cells they use so that each pack is built of identical cells, they'll be built of mismatched ones to one degree or another. (the degree of mismatch will be completely unknown, as no testing to match them will have been done in these cases).

Almost no pack builder is going to do this matching; it's time consuming and expensive. Perhaps EM3EV does it; you'd have to ask them. If a pack builder says they do this but can't or wont' tell you the characteristics tested for and the limits used to determine a match, they don't actually do it, or they use limits so wide that they may as well not have done it.

Large-EV manufacturers (cars, etc) *do* do this so the used modules out of those large EVs are generally going to be built of cells that are close to identical (much closer than in ebike packs). Places like greentecauto and batteryhookup etc that sell used EV modules are a source of these.

Even better are the modules not made of tiny small capacity cells, but instead of larger high capacity cells, as those cells are often better manufacturered, designed to be capable of high currents for long periods in EV use, and don't have a bajillion interconnects as potential failures just wiating to happen in the packs made from them.

 

[PhilExplore]

Hi I got the all the first batch at 18650batterystore.com and checked voltage of all the cells, worked well for first few shorter rides, the issues only came up on longer rides, the only indication I had was that one bank wa lower than the rest, I guess if it gets more than 0.5 v difference then its time to stop. ( over 3.1V). I did put the bms cut off at 3.05V, but I guess that did not stop the damage.

 

[amberwolf]

0.1v difference between cells (that were charged to the same voltage initially is a massive difference.

0.5v difference means there's a bunch of garbage in there--massive differences in cell characteristics (resistance, capacity, etc).

Checking voltage on cells before using them to build a pack doesn't tell you anything other than what the cell votlage is, and doesn't help you build a useful pack.

I highly recommend reading the various cell-testing threads (pajda, etc) for better testing methods to help ensure you build packs out of matched cells.

Without buildng from matched cells, you're going to continue to have problems like this.

 

[ak136]

Your spot welds dont look consistant, its no wonder you get problems with disbalance

 

[docw009]

On this strip, the nickel was misaligned and one of your welds was over the edge of the cell. You have a PVC washer and shrink wrap there, but a 400A weld on a nickel strip with nothing below it could easily punch thru the insulating layers and cause a potential short circuit of the cell. That's why most batteries, even the commercially built models, use paper washers for added protection on the positive ends.

It's kind of unlikely, but maybe there was a carbonized pinhole under one of these misplaced welds, enough to slowly drain it and the rest of the group to zero volts.



With spotty spot welds, it's also possible one of them lifted on one cell within a group. That makes it a 5P or lower, and those groups will discharge faster and go unbalanced.

Otherwise, how do healthy cells discharge to 0V in a battery with a working BMS? They cannot. I don't have a JK BMS with self balancing, but maybe you set it up incorrectly. Is that an active balance where it's always working (seems silly in my opinion) or a top balance scheme where it works during charge?

If you got a couple of leakers in your battery, those will discharge a cell group, but you bought them new from 18650batterystore? They would have been low before you built the battery. Anyway, that's where I get cells when they go on sale, and I recently bought 50E's from them, It's a very good cell.

 

[PhilExplore]

Ok, thanks. Cells were the same voltage and I charged them to 98% and let them self balance. I think the first charge I had to let the battery sit for a few hours to get all the groups to the same voltage. Riding after several hours showed 0.3V difference between one group and the rest of the groups, so maybe you all are correct.

I will rip it all off, clean up and start again. Maybe solder as well. I made another battery the same way and did not have any issues. I think I will just get a professionally made battery with bluetooth so I can see the individual groups voltages and stop riding in any difference in the groups are detected. Still learning. Thanks for all your help

 

[fechter]

The BMS should have stopped things way before any cells reached zero. Not sure how that could happen.

 

[d8veh]

It's good practice to use insulating rings on the positive ends of the battery. I can see that you fitted one. Why not the others. It''s really dangerous to not use them because the nickel strips and their cut ends can wear through the plastic heat-shrink cell covering on the corner of the cell,at the top, which would cause an instant fire. At the very least it would short out a cell-bank and give the problem you've described.