Calculate series and parallel wiring

The stored energy is the same either way — four 12.8 V / 100 Ah blocks always hold 5.12 kWh. What changes is the current: 4S gives 51.2 V at 100 Ah while 4P gives 12.8 V at 400 Ah, so 4P draws four times the current for the same load. Cable losses scale with current squared, which makes 4S one sixteenth of the cable losses of 4P.

The calculation assumes identical blocks, the same chemistry and a matched state of charge at the moment of connection, and it warns when you flag mixed blocks. Equalising currents between blocks at different states of charge, dynamic behaviour and resistance drift from ageing are not covered.

Current sharing is decided by the wiring method — taking both leads from the same end is the most common arrangement and the worst one. If you supply block and interconnect resistance from the datasheet, the calculator quantifies the imbalance; without those values it deliberately stays silent rather than inventing a number.

Four 12.8 V/100 Ah blocks always give 5.12 kWh — as 4S, 2S2P or 4P. But load current is P/U: at 1,000 W, 4P (12.8 V) draws about 78 A, 4S (51.2 V) only 20 A. Cable losses scale with the square of the current — 4P loses 16× what 4S loses in the same cable. Add thicker cables, larger fuses and current imbalance between strands. Higher system voltage is almost always the better choice if the loads allow it.

Capacity, chemistry, age and state of charge. Blocks of different capacity or chemistry in parallel are a critical error (hence the red MIXED_CAPACITY/MIXED_CHEMISTRY warnings); different ages let the strands drift apart. And when first connecting, the blocks must be at the same state of charge, otherwise high equalising currents flow. The load connection point additionally decides current sharing — the parallel-wiring calculator computes that.

In series the same current flows through all blocks, but small capacity and self-discharge differences accumulate: one block slowly runs fuller or emptier than the others until it hits the limit first during charge or discharge and caps the whole string — or gets damaged. A supervising BMS or balancer evens this out. Putting blocks with internal BMS in series is only permissible if the manufacturer explicitly approves it.

With typical values (block 5 mΩ, interconnect 1 mΩ, four blocks in parallel) the first block carries 80 % more current than its share — taken off diagonally it is zero. This figure appears since 5 August 2026 once you enter block and interconnect resistance from the datasheet; without them the calculator deliberately stays silent instead of naming a number it does not know. The detailed run with cable lengths lives in the parallel-connection calculator.

Because end-of-charge voltages and internal resistances do not match: a LiFePO4 beside an AGM forces the same terminal voltage on both — one is chronically over-, the other undercharged, and the equalising current flows permanently. The switch for this was previously unreachable through the interface; now the warning appears as soon as you set it.

Possible, but with an open warning: the older block has higher internal resistance and lower capacity — the new one takes more current and ages faster in turn. Rule of thumb: same chemistry, same type, age difference under a year. In series, mixing is stricter still, because the weakest block dictates the end of charge.

From the chosen chemistry's nominal: 12.8 V for LiFePO4, 12.0 V for lead types. If you wire 24 V blocks, enter the voltage via the switch — previously the field was unreachable and 24 V blocks could not be computed correctly at all.

Same energy (5,120 Wh with four 100 Ah blocks), but 19.5 A instead of 78.1 A at 1,000 W load — a quarter of the current, a sixteenth of the cable losses. In return 4S needs balancing across the string and a 48 V capable inverter. The system-voltage calculator runs the same trade-off with cable and converter costs.