paralleling 280Ah + 100Ah

After installing the 280Ah SFK bank I attempted to rehome the hand-me-down 50Ah Chins for free (successful) and the 100Ah Rebel Batteries of a bottle of Bulleit rye (unsuccessful).

I’ve been dragging the 100Ah around since then, using it for projects or to run the air compressor. Mainly it sat there taking up space. 14 months of mostly storage has caused the BMS SoC estimate to be wildly off¹ and (I suspect) the cells to have drifted out of balance.² Based on self-discharge and my light use I’d guess actual SoC to be 70-80% instead of the indicated 99%.

It will probably be a while before the ~80mA passive balancing circuit gets it back under control.

before combining

The 100Ah was sitting at 13.15v and the 280Ah a little higher. I used the induction hob to make coffee, which pulled the main bank down to 13.18. I disabled the mounted array’s solar charge controller in preparation for the change.

I modified the thrift store Styrofoam cooler to make an insulated box for the 100Ah, popped in a spare lizard warming mat (15w, IIRC) and squeezed in the battery. We’re ready to go.

The shunt and the 280Ah both reported ~41% SoC.

combining

The actual reconfiguration went well.

The space is tight so I spent more time trying to get the two batts into their new position than anything else. No spark when they were combined because the voltages were so close. The POS is taken from the smaller batt and the NEG (via the shunt) taken off the larger. The respective BMS and shunt are reporting voltages within 0.01v of each other.

after combining

I enabled solar charging and current started flowing. Based on both theory and previous experience each batt in the bank should charge/discharge in relation to its share of bank capacity. The 280Ah bank has testing data showing it’s actually 295Ah so I’ll use a 3:1 constant for simplicity of math. If the bank were charging/discharging at net 10A the bigger batt should see ~7.5A and the smaller one 2.5A. Note: I can only look at one battery at a time in the app so numbers may not add up perfectly.

At solar startup net charging was 17.68A. The 280Ah was taking 14.4A of that (~80%) and the 100Ah took 3.61A (~20%). Maybe the smaller batt being at ~twice the SoC was skewing things. Or maybe it will be a stable pattern (calendar age? Older chemistry design?). I’ll update later with more observations.

Discharge update: made breakfast on the induction hub. -21.8A (71%) and -9.8A (32%).

The smaller batt already has a 25mV cell voltage delta and we are nowhere near the knee. The balancer is steadily working on cell 4 but it’ll be slow going. I’m not worried. If I ever open the case I’ll add an active balancer with minimum voltage cut-in like the SFK has.

resetting the monitor

I accidentally set the battery monitor to 100% SoC when trying to change capacity from 280Ah to 380Ah. :-/ No big deal but annoying. My best guess (assuming ~75% SoC on the 100Ah) is that it’s indicating 304Ah (80% SoC when it’s actually 189Ah (50% SoC).

It will finally be correct once the bank fully charges. Getting there may be tough as I’m still banging away with 100% electric cooking and haven’t consistently reached Float in the last two weeks. Close, but no cigar.

The long drive to Arizona next month should do the trick if nothing else does.

1. The coulomb counter has a minimum detectable current. In the case of this JBD BMS it appears to be 0.5A. The BT takes way less than that and self-discharge wouldn’t be counted either. ↩

2. Can’t really see it now but will show up in the upper knee. I expect the batt will shut off well before (actual) 100% SoC because of one or more runaway cells. ↩