[dimitri]

The battery has been reading the same as the rest when fully charged, but when under load this one battery goes to a lower voltage faster than the rest till today when it got just over a volt lower than the rest!

Are you sure this new cell is fully charged? Did you top balance it to the rest of the pack?

Sounds like it just needs to be brought up to full charge along with the rest of the pack, i.e. fully charge the pack, then charge this one cell with single cell charger until red LED goes on on the BMS.

 

[dimitri]

I only top balanced it once initially. Other than that, I've just had the MiniBMS running till last week, and this week I've had the WIcom BMS added so I can see each individual voltage.

Should I try re-balancing again? I did fully charge the pack last night, then turned it on twice more to top it off, not a balance charge in parallel though. This morning, it kept above 2.9V, but the others were .1V to .2V higher (all under load on a 13 mile highway drive taking it easy at 55). I'm trying to top charge it again for the ride home. Then fully charge to pack voltage, then remove and top charge by itself.

I still can't gather from this if you charged just this one new cell separately before putting it into the fully charged pack. There is no need for balance charge in parallel when adding just one new cell. You just have to make sure the new cell is fully charged ( i.e. reaching into the upper knee range, i.e. having red LED lit on it at the end of charge ).

If you have a single cell charger, then procedure is simple:
1. Charge entire pack.
2. Top off the new cell until it reaches the same level as the rest of the pack.
Since you have miniBMS modules, its easy to tell when the cell catches up to others, when its red LED starts to glow.

NOTE: Topping off the new cell must ONLY be done if the rest of the pack is full, otherwise you will imbalance it in the other direction.

 

[dimitri]

My drive to work this morning was the same, no alarm, the battery didnt go under 2.9, but I can see it along with the others with the WIcom and its lower than the rest. Last nights drive was 25 miles and the low battery went way down, MiniBMS alarm was constantly going off and I could see with the WIcom it was a good volt under the other batteries.

Does it stand out from the rest only under load? Does it read same as others without the load? Do the terminals/bolts on this cell get noticeably warmer than others? Sounds like extra voltage drop is developed under load at this cell, but it could be inside the cell ( weak/defective cell ) or outside the cell ( poor/dirty/lose connection).

If BMS alarm goes off on this cell even when you let go of the throttle, that means the cell does not recover above LVC even when load is removed, which certainly points to a weak/bad cell. Or maybe this cell is simply much lower capacity than expected.

One trick I use to find lose/dirty connections is to measure the voltage drop between the actual body of the terminal and the busbar attached to it, while charging at max CC current. You should see just a couple millivolts or less on a healthy connection. Establish a baseline by checking all of them , most should be in same small range. If you find some significantly higher, clean and redo that connection.

 

[dimitri]

Just tried it on my own cells while charging at 20Amp. I measure 0.2-0.4mV between the side of aluminum nut holding the terminal and the top of the busbar near the black shrink tube. Anything that stands out by 2x-3x from the baseline I would consider a poor connection. Actual baseline value might vary due to sensitivity of the meter, amount of current, etc. The idea is to compare the baseline with the cell in question to establish if external connection is clean.

BTW, such test done earlier would have definitely saved your original cell from melting. Live and learn....

I spotted a few poorly torqued connections in my own pack back when I built it 2 years ago, using this method. Never had a problem since then, knock on wood...